Problematic Anhyrdrite floor!

[Perfect Tiling]

Hi there. I have recently tiled a couple of floors in a new house and they have started to come loose. The floor is called hemihydrite which is a new type of anhydrite that does not produce a latent skin so does not need sanded prior to tiling etc (thats what the spec sheet says). The floor with UFH was laid in mid sep at 75mm depth and heating was on for 3 weeks prior to me tiling. Prepared as per the book, with 2 coats of APD primer and used SPF adhesive. A carpenter on site did a moisture check prior to him laying an oak floor and said it was fine. Tiles were laid 2 weeks ago and up until last fri were solid, however today they are loose in places with grout cracks. Other floors have been laid with travertine in the same way and are still solid. The only thing different is the customer has put the heating up to 20 degrees since then and the floor was warm. On lifting a tile it came up with some adhesive on it and the remaining adhesive came up with the primer on it. I think the floor still has moisture in it and this has been trapped under the tiles and caused them to lift due to the heating being put on too much too soon...any ideas. If this is the cause...who is to blame. Customer says he was told by screed installer that floor takes 75 days to dry...however I called the company today who said 120 days...depending on conditions...cold and wet up ere. Many thanks in advance.

 

[Colour Republic]

heating should not have been on after the tiles had laid for 28 days

 

[Ajax123]

Hemihydrate is not a new type off anhydrite. It is alpha hemihydrate ... It is a different chemical arrangement and is by no means new having been around for decades. From your point of view it should be treated exactly the same as anhydrite screed.

All gypsum screeds, and for that matter all screeds, should be sanded in order to prep them for tiles. This is to remove any loose debris or extraneous surface contamination.

Drying time is 1mm per day for the first 40mm depth and 2days per mm there over so approx 120 days. This depends on site conditions as it is based on 20degrees c and 60% rh. Underfloor heating should be comissioned and run up to temperature and then down again before moisture testing. Electronic moisture meters should not be used on these screeds except as an indicator. The correct method is the floor hygrometer or carbide bomb test.

My suspicion is moisture rising to the surface following increasing the temperature. Electronic meters tend oly to measure the top part of the screed so if there was moisture within the body of the screed it might not have been detected.

Underfloor heating should not be applied to a tiled floor for a minimum of 28days following tiling.


apart from that everything is ok....

 

[Perfect Tiling]

Thanks Ajax...thats what I thought. As mentioned I didn't sand the floor as screed company said not to...I did try to sand and scrape a part but nothing came off...no dust...nowt. If moisture is the cause who is to blame? I don't do a bomb test and relied on dates given by customer and screed company who now say 120 days not 75 as previously. I told customer not to put up heating but he did so just my word against his. The heating to the travertine floors has been physically shut off for now so that if moisture levels are too high the floor can still breathe and hopefully be ok. Failed floors are porcelain. This type of screed obviously has some benefits...but not as regards tiling onto!!!!

 

[bugs183]

No one knows these screeds like Ajax, Perfect. He'll sort you out.

 

[Ajax123]

Thanks Ajax...thats what I thought. As mentioned I didn't sand the floor as screed company said not to...I did try to sand and scrape a part but nothing came off...no dust...nowt. If moisture is the cause who is to blame? I don't do a bomb test and relied on dates given by customer and screed company who now say 120 days not 75 as previously. I told customer not to put up heating but he did so just my word against his. The heating to the travertine floors has been physically shut off for now so that if moisture levels are too high the floor can still breathe and hopefully be ok. Failed floors are porcelain. This type of screed obviously has some benefits...but not as regards tiling onto!!!!

it should not really be 75mm deep. That is an error in my opinion as it n needs to be 50mm n underfloor heating or even 40mm with some systems. That would have lopped off a good 50days from the drying time.

now then who is at fault... Well if you have asked if the floor is dry enough and been told categorically yes then you cannot really be held to account. I pf you had been told it was wet and carried on regardless you would be negligent. If you have carried out the tiling without any sort of test or info (which from your original post you have not) then you would probably have a negligence case to answer.moistrue testing is the responsibility of the man contractor normally. Certainly material selection are not at fault as the APD is fine. In terms of sanding this does 2things. 1removes dirt and loose debris and 2 helps to produce a key for the primer to key to. Not sure what sort of project or contract it is though so cannot really predict lines of resposnibility.

 

[Perfect Tiling]

Thanks again Ajax. A moisture ckeck was done by a carpenter prior to laying an oak floor but probably not a true reading. This will probably be a case of everyone blaming someone else. I thought I had done everything as informed. I could have just left it and not said anything but was being honest as I knew it would just get worse as time went on. Forgot to say earlier...the screed seemed soft to me as I could dig into it with a screwdriver and rather than coming out as lumps it was just like grains of sand. The rep from BAL said that if there was a skin the primer would likely have this stuck to it when lifted but it was clean. He thought moisture being trapped by the heating being put on too soon was probably to blame seeing that the travertine is still solid and had to be taken off with a hammer...in pieces (I removed a cutting of travertine that had been down for a week).

 

[Ajax123]

Soft screed suggests that it was segregated which would generally be a result of excessive water addition either at batching or on site. This causes the screedmtomsegregate such that you get a lot of chalky power at the surface with the harder screed underneath. It is usually in patches. Something else that can cause it to be soft is if there has been insufficient reaction due perhaps to freezing but in September this is improbable even in Scotland. This is not really skin as that is a separate sort of entity. Lack of sanding will certainly lead to poor penetration of the primer though.

 

[Perfect Tiling]

Crikey...you know your stuff. The first coat of primer was 1:1 mixed with clean water and was like coating to paper as it soaked right in and didn't puddle at all. Second coat was put on neat at 90 degrees to first. I tried scraping the primer today to see if it came off as a skin but it didn't.

 

[Mark S]

Hi sorry to hear the tiles are lifting,
Not sure whose screed was laid, but at Gyvlon we always recommend the screed is lightly abraided/sanded before the primer is applied. The fact the wood floor is fine would indicate the floor may well be dry.
At 75mm thick the screed is in our opinion 20mm too thick, the optimum thickness with UFH is generally 50-55mm.
As the tile you lifted had primer on the back it would suggest the primer has not penetrated the screed surface, Before commencing any re installation test the floor for moisture using a hygrometer box, or carbide bomb test, nothing else is accurate on anhydrite screeds. The moisture level does need checking.
There are some adhesives on the market that do not require priming, Kerakoll H40 Ideal is one.
hope this helps

 

[Perfect Tiling]

Ajax...forgot to ask. Can this problem not be related to thermal shock due to the heating being turned on too high and too soon. The tiles under the toilet etc where there are no pipes are solid and a perimeter tile has cracked but is still sounding solid. Put this with the fact that the travertine tiles in the rest of the house are solid, leads back to the only tiles going loose are those that have had heat introduced. Floors were fine on Fri last week (I was on my knees siliconing) and thermostats were down at 10 and not calling for heat, today 5 days later they are at 20 with water going throught the manifold at around 45 degreesC. Customer moved in at the weekend so must have turned it up.

 

[Perfect Tiling]

Hi Mark. Thanks for the info. The wood floor is not stuck to the screed, it is an engineered oak floating floor. The tile I lifted was hard to get up and broke into 3 pieces the first part leaving the adhesive on the screed rather than on the tile. The tile had adhesive on it level with the high part of the pattern. Never doing screeds again...not worth the worry...hardibacker or wedi boards all the way!!!

 

[Perfect Tiling]

Mark. If company literature says not to sand then is this a get out clause for me if this is the problem????

 

[Ajax123]

Ajax...forgot to ask. Can this problem not be related to thermal shock due to the heating being turned on too high and too soon. The tiles under the toilet etc where there are no pipes are solid and a perimeter tile has cracked but is still sounding solid. Put this with the fact that the travertine tiles in the rest of the house are solid, leads back to the only tiles going loose are those that have had heat introduced. Floors were fine on Fri last week (I was on my knees siliconing) and thermostats were down at 10 and not calling for heat, today 5 days later they are at 20 with water going throught the manifold at around 45 degreesC. Customer moved in at the weekend so must have turned it up.

yes this is certainly a possibility. The timescales would suggest moisture. Thermal shock could easily be contributory. It s difficulties diagnose at a distance obviously.

Is the trav heated?

 

[Ajax123]

Mark. If company literature says not to sand then is this a get out clause for me if this is the problem????

Yes it would be a you have followed manufacturers guidelines. It does depend on wording though.

 

[Gazzer]

Hi Mark. Thanks for the info. The wood floor is not stuck to the screed, it is an engineered oak floating floor. The tile I lifted was hard to get up and broke into 3 pieces the first part leaving the adhesive on the screed rather than on the tile. The tile had adhesive on it level with the high part of the pattern. Never doing screeds again...not worth the worry...hardibacker or wedi boards all the way!!![/QUOTE]


Uh ?????

 

[Perfect Tiling]

Ajax. Literature says "the screed does not require grinding for subsequent floor coverings or vinyl application." Later on it also states "eliminates the need for subsequent grinding". Travertine floor will be heated but it is isolated just now so customer can't turn it on incase thermal shock is a cause.
Sir Ramic...what I meant was that if I do smaller rooms with flooring where I put down hardibacker or insulation boards for electric UFH I have never had a problem with loose tiles...ever. I could have made as much out of 4-5 standard rooms than what i get from this big job

 

[tommyzooom]

I recently lost a job, as I said it needed to be sanded, the screed Co. (Larsen) said it didn't. Homeowner was adament that it shouldn't be sanded, and implied I didn't know what I was talking about. I assume it has now been tiled on top of a heavy dusting of emulsion paint, as he had all the walls spray painted:mad2:

 

[Perfect Tiling]

Cheers Tommy...its a maze out there. This floor was nice and clean as it had been protected from paint etc. I knew about sanding but screed company and literature said not to...so I followed their instructions...printed ones so I have evidence I hope.

 

[DHTiling]

If the trav is solid and just the porcs came up after heating, then i think it is a strong possibility that it was heated to quickly.. if the heating was ran for 3 weeks then i doubt much moisture was left in the screed..

porcs will react a lot different to thermal shock than stone as porcs rely on polymers to hold the tiles where as stone with hold in the capillaries/structure of the stone backs.. so the thermal stress will make porcs pop easier if subjected to heat to fast as the adhesive would not be at full structural strength till at least 28 + days.

So don't be quick to blame moisture / sulphate delamination till you see how quick the floor was heated up.

 

[Ajax123]

Cheers Tommy...its a maze out there. This floor was nice and clean as it had been protected from paint etc. I knew about sanding but screed company and literature said not to...so I followed their instructions...printed ones so I have evidence I hope.

You do in my opinion. The advice, whilst it flies in the face of the industry guidelines, is manufacturer advice and as such they will stand or fall by it. This does assume that te lack of sanding is a major contributing factor. It might be nothing to do with it but you need to be prepared to be told by lots of "experts" that lack of sanding is the cause. Based on time scales and having read the thread fully my money is on moisture and thermal activity.

 

[Perfect Tiling]

Cheers Dave. The floor appears to have been heated up from 10 degrees to 20 degrees within days. As said earlier, when I left on Fri it was at 10...this morning it was at 20 (setting on air sensor stats) but they were moving in on Sat so it could have been done overnight. They also have the settings on a timer so it would not be a consistent temp but on in morn then off then on again at night.

 

[Mark S]

The wording can be confusing.
The 'skin or laitence' was a few years ago a distinct 'layer, or skin. The screeds manufactured now do not have the same characteristics, however it is still recommended to sand or abraide the surface to open the screed to allow for primer penetration and for a good key for the adhesives.
It is much the same with power floated concrete, no visible 'skin', but the surface is allways scabbled to allow for a good key.
I would always recommend, if in doubt, sand it.
I hope this experience has not totally put you off screeds

 

[bugs183]

Hiya Mark.
I think you'll find there are an a great deal of tilers put off by these screeds.
You've arrived on this Forum a bit late to witness a load of us guys trying to make sense of this system that has been forced upon us.
I took it upon myself to try and have a good go and get to the bottom of these screeds, and to be honest i've given up on them as they are NOT worth the risk for us end trades, or at least ones who really care about their finished product. I'll explain as well as possible:
I've been tiling over 20 years and i kinda know when to tile a floor on concrete that it will stay put,. I know the recommended drying times, the ideas behind adhesives and their performance, blah blah. I do a floor, I get paid and the customer and myself are happy, for years, as the floor does what it's supposed to, it's stayed put.
So along come these screeds, ''ooh this is flat, i dunno what it is but its nice to tile too."
"The builder poured it really fast and it can be tiled in a few days", great stuff, we think, he says it fine to tile to.
Then (having not tiled too many i may add),we actually find out it's not cement based but gypsum, ooh thats odd, why use gypsum for a floor as it's not waterproof??
So then further research shows the drying are nothing like the ones we've been told, and can and ARE even slower than cement screeds.
Then we hear of failures," you should have sanded", "no one told us", "oh you should use epoxy primers", "that's not what BAL say, they spec the same as cement based adhesives".

So back on the phone:
Moisture testing, conflicting conversations from EVERYONE. Lafarge are vague as too whom is responcible, and what method.
One method recommended by one guy is poo poo'ed by another. Testing manufacturers say they don't agree with recommended practise.

Preparation: Who is responcible for sanding, if it even needs it, some say it does, some not.

Adhesives: People like Bal say prime and cement based, other like Kerakoll etc say gypsum friendly, others gypsum based adhesives.
Weren't we taught NOT to use Gypsum based adhesives as they don't stick porcelain, hence the move to modify adhesives when porcelain came out ten years ago?

I've tested all these adhesives with very mixed and mainly unsatisfactory results. Most gypsum adhesives i found didn't stick well at all to porcelain, and then have to be put through the rigours of lateral expansion with underfloor heating.
The best i found was creative impressions GBTA, and it cost £750 to tile a 63m2 slate floor to anhydrite, twice as much as cement based.
How the hell do you quote well for that and then programme it into your schedule to allow the vaugeries of it's drying times???

I've found two ways to approach these screeds:
Sand, prime, GBTA, Ditra Mat and then cement based adhesives. And even then Schluter aren't too happy with Gysum adhesive and Ditra, and the customer is certainly not happy with the price.
The second way is too say, "sorry, i'm not tiling this".
So i have been FORCED by your cack screeds to stop doing a job that i like, and make money at, to change the way i work, simply because NO ONE is helping out us tilers, the dunces taking the flack at the end of the job.

I know Ajax will read this, and he is a legend, and he has tried. But i'm a tiler and i have no confidence in this system at the present time, i am not alone, and the nimber will grow as these failures happen more and more.

This site is having more and more posts from failures from this screed and it needs to be dealt with by the people selling the screeds.
Yes you've sold loads for years in Europe, then bring the standards over here NOW.
Rant over.

 

[Gazzer]

And ...Breathe ! :lol:

 

[Mark S]

Quite a rant, and it appears justified.
However, as you no doubt will gleen, Ajax and myself are singing from the same hymn sheet.
The wording from manufacturers is confusing and does not help. What I would say, that if there is good dialogue with all parties at an early stage, (before screeds are installed), and the floor finish type identified, then steps can be taken to ensure correct preparation methods are undertaken.
When the screed is correctly specified, ie thickness etc, and all aspects of drying are undertaken the screeds do work.
The argument over priming and cement based adhesives over Gypsum or similar one will rage on long after this thread, there is a good argument for both, and down to personall preference.
Some good dialogue would not go amiss!!

 

[bugs183]

Hiya, don't worry i'm not having a crazy rant, but I really feel that we are tilers are simply not being heard, and sorry to say this but your answer has only echoed the conversations i and may other have had or are having.
You and Ajax see the screed form a perspective that we don't.
On many jobs we get called in and are asked to tile a floor, it normally has already been proepared, the screeders are long gone, the builder isn't interested and the customer just wants their job done.
We rarely get to have any input before a job is done:
I did with a local builder, i recommended these floors and everyone has had dryind problems, he wants them all ripped out, i fell like a bit of a pudding.
They are not working as you expect.
We as tilers cannot afford to have dialogue over which adhesive to use, this is OUR jobs here being messed with, and it happening now,we need firm and qualified answers from you guys, that we trust.
If you want dialogue then get Mapei, Bal, Lafarge, Kerakoll etc and a handful of well meaning tilers to sit down in a room so we can be heard.

 

[Gazzer]

I am glad you acknowledge our issues Mark.
I reckon we have heard it all about these screeds but I dont think you will find many fans here, or soft flooring guys !
The problem arises not so much on the large jobs but the smaller jobs. Like a normal house floor. The builder is a general builder and is under pressure to get done on time and looking at costs, screeder comes along and says " Here mate have this, job done in 4 hours, jobs a good un! "
I havent seen one flat floor from this type of screed yet ! I have witnessed soft topping, hard topping. The client nor builder want to pay for special primers, moisture tests ( which in their self prove inconsistent) , They dont want to pay for sanding/abrading of floor "but screeder says its fast track mate " IS IT COCO !
I have just turned my 1st Gyvlon floor down this week as its being pushed too fast and no one will do it right....at the end of the day when it fails I know where the finger will point.

While we sit here trying to make head or tail out of which way to tackle these screeds, one thing is missing and thats info to the client....They are being sold screeds that they dont really need and not given all info only what suits the screeder.

 

[bugs183]

To add, i find it very odd that massive companies such as Mapei have not bought over their Adhesives from Europe that work and have been proven to work for years.
I understand there is a British Standard issue here, but surely it can be achieved and i would prefer that to a Sub Standard job that i feel i do onto these screeds.
I simply do not trust Bal and Mapeis recommendation to tile these screeds using an Acrylic primer such as APD and follow that with cement based adhesives.
Maybe you guys should set up a sample floor laid to you specifications, and get some tilers in to tile it, using all the 'approved' adhesives and priming methods. I think you'll quickly see why we've got a monk on over these screeds.

 

[Mark S]

Sir Ramic,
I do acknowledge there needs to be more dialogue.
As manufacturers of the binder that makes the screed we are not always in front of the customers/clients till after the screed is in, if there are still 'screeders' doing a half truth sales job, we need to be aware of this, something I'm sure Ajax will back me up on.
One of our main roles is the correct specification of the screed, thickness, and to assist with advice re preparation for floor finishes.
I would be happy to sit round a table to discuss how best to tackle these 'issues'.
There are 000's of m2 of Anhydrite screeds installed and floor finishes applied with a hitch, but all the failures get high profile!!

 

[Ajax123]

What is interesting - well it is to me anyway - is that the thread, although entitled anhydrite is actually about a hemi hydrate screed. Mark S and myself work for Manufacturers of Anhydrite, Mark for Gyvlon and I for Francis Flower. Our materials are ostensibly the same, and in fact until recently Francis Flower made Gyvlon binder for them and even more incestuously Mark took over my role at Gyvlon after I moved to Francis Flower a few months ago. Our respective materials perform ostensibly identically in all aspects, drying, thermal properties, optimum depths, formation of laitance etc etc. All of us in the construction industry are after an edge I guess and one of the perceived problems with anhydrirte screeds is the need to sand them to remove laitance or skin. Actually sanding the screed at an early sage is beneficial in a number of ways. The formation of laitance is common to ALL screeds as I have said on numerous occasions. Hemi Hydrates, which incidentally are also supplied by Larsen work slightly differently in that the chemical reaction tends to tie up the laitance in the surface of the screed whereas with anhydrites it sits on top. The Hemi boys seem to think that this means that the screeds dont need to be sanded. Lafarge readymix (as opposed to Gyvlon - different companies) some years ago now began to use admixtures in order to prevent the formation of this skin thus mimicing the hemi hydrates. This was done primarily to redcue the need for sanding because no one wants to sand the screed as this is an extra operation and extra cost and the are practical difficulties on some sites. The draw back of these closed surface systems is that they are much more dense and much less easy to get primers and the like to soak into and adhesives to key to.

What I find interesting is that the anhydrite boys (Mark S and mself) say sanding yet it appears that the hemi boys do not. If you think about the characteristics of the screed and then think about the nature of a general building site which gets dirty you will understand the need for sanding. Whose responsibility is it depends on how the contract is set up. My own advice reflects Marks... if in doubt - sand it. There is a distinction to be drawn here as well MArk and I work for the binder manufacturers. Neither of us make or sel screeds. That is down to our customers the likes of Bardon, Cemex, Tarmac etc etc. We would always hope that our own guidelines are best practice and are mirrored by our customers and in turn their customers. This does not always happen of course as we deal with human beings with alll their frailties. However you can rest assured that a simple call to me or any of my colleagues on the binder side will help you to get it right.

The argument is never as simple as "there is a lack of information given to clients" or "the screeders dont do their bit" or even " the manufacturers dont pass on the information". All of the information about these screeds is in the public domain and available free of charge in official data sheets should it be requested from us as manufacturers.


Harping back to many threads to which I have contributed prep is really straight forward, sand it vac it, prime it (if your adhesive needs a primer) then tile it.

These screeds are here to stay and as I have said many times before you will come across them more and more. There are two options as I see it. Learn about them, learn how to deal with them and get the job done or walk away. It is not for me to decide which is right and which is wrong. That is a business decision for you guys to make. Some have gone one way and some the other.

I dont think the failure in this instance really has anything to do with lack of sanding though.

Anyway that is me done for now as I am off out soon and need my tea 🙂