Help! Floor being poured tomorrow. Lots of air in UFH loops

[oranjeboom]

As usual, things are turning a bit last minute. Manifold is now on, loops are on etc.

 

I have started to bleed my 12 circuits and am finding that there is a lot of air being dispelled in the bucket (had to empty about 6 buckets on one loop already). As concrete is being poured tomorrow first thing I'm feeling a bit anxious. Have had to rig up a temporary water supply to the manifold from an outside temporary tap involving a longish hose pipe and there are a few drips along the way to the manifold supply valve (i.e. at the outside tap etc). Is it possible that the air is being introduced into the ufh set up, rather than within the ufh pipework/manifold? No sign of leaks in pipework yet. I won't have a pressure reading until I have water in all circuits so can't yet determine if there is a pressure fall. Certainly no leaks in the ufh pipework (had a visual check in the rooms done so far).

 

I'm hoping the air is being introduced elsewhere - could that be the case?

 

Does it matter that I may have some air in ufh pipework when the concrete gets poured? I presume as long as it is all under pressure then I should be okay? The air can be expelled at some other distant point I presume.

 

Hoping @Nickfromwales or some other experts can advise!  TIA!

 

 

[Onoff]

My uneducated guess is that as some pressure test systems use air it won't matter a carrot...I could be wrong!

 

The worry would be is if air is getting in due to a leak! But if you can't see a wet patch all's well I imagine.

 

Testing my one room loop was pretty easy tbh I just filled up one end from the garden hose onto the copper then shut it off via the valve. Gauge was reading 8/9bar (mad mains pressure here). I bled via the drain cock and clear plastic tube topping up with mains until no more bubbles came out.

 

 

[Jeremy Harris]

I didn't even bother to fill our UFH before the pour, as we had no water on site at the time and it seemed too much hassle to do.  All went well, and there were no problems with any of the three loops.  I have a feeling that MBC don't usually fill UFH loops before pouring slabs either; others may be able to comment on whether this was true for their slab pour.

 

As for getting air out, then I found it did take a fair time for all the air to come out of all three loops.  I filled the system and ran the pump at full speed, with the auto-bleed valve open, and it took a couple of hours for all the air to find its way out, IIRC.

[richi]

If I read your post correctly, you're filling/flushing the pipework from a garden hose that's dripping. If there's a drip, there's an opening to the atmosphere. If water is flowing past an opening in a pipe, Herr Prof. Dr. Bernoulli says it'll tend to suck air into the pipe.

[Onoff]

3 minutes ago, richi said:

If I read your post correctly, you're filling/flushing the pipework from a garden hose that's dripping. If there's a drip, there's an opening to the atmosphere. If water is flowing past an opening in a pipe, Herr Prof. Dr. Bernoulli says it'll tend to suck air into the pipe.

 

Good point. Best I guess to fit a valve to shut off the leaky hose...and a gauge to see any drop.

[oranjeboom]

Thanks all. I added a PRV into the mix of things which I think may not be helping the various joins prior to the PRV. Very high pressure here like @Onoff.

 

I guess with your slab Jeremy @JSHarris you 'just' had the slab and didn't have too many people on that during the pour. My slab is in the confines of 9 or so rooms, so repeated heavy traffic in some places.

 

Thanks @richi...need to dig out my physics books again.

 

It'll take longer to fix all non-UFH leaks so hopefully as long as I get it all pressurised albeit with a bit of air in the loops then it sounds like it won't matter for the purposes of the concrete pour tomorrow. I just don't like the thought of people trampling over the pipework if there's no pressure at all.

 

Onto loop 3....

[Jeremy Harris]

The pressure won't make any significant difference to the stiffness of the pipe in practice, as the forces on the walls of the pipe from the pressure are very low, even if the pipe is at a couple of bar or so.  The inherent stiffness of the pipe exceeds anything that a bit of modest water pressure will improve, and you can just as easily crush a bit of UFH pipe that's filled with water as a bit that isn't.

[PeterW]

How are you filling the loops ..??

 

The correct way is to put the hose onto the filler connection, close all the loop valves and the drain valve and then turn the hose on. That will show you any leaks in the manifold. 

 

Open the manifold drain half way and then open the first loop full bore. and let the water run - it will spit and bubble but let it run until it's pretty clear of air. Close the valve and move to the next one, again running to a hose and letting it run clear. 

 

It can take 2-3 minutes per loop to get the air out of each loop. Once you've got each loop clear then open them all half a turn and run it for another 2-3 minutes before closing the drain - you will watch the pressure rise in the manifold, take it to 3-4 bar then close the hose connection. 

 

And now wait ......

 

At 3 bar any leak in a pipe will be easy to see. You may find the pressure valve connection on the manifold leaks slightly (mine did) so just watch for a weep there. 

 

Mines been sat at 1 bar for 4 months now and tbh it's never moved. Next time it gets the pressure released will be for the pump and buffer to be fitted so I'm not going to get too worried about the water in it not having antifreeze or inhibitor in it - I've calculated the buffer and loop capacity and will just make the concentration up based on that and wait for it to naturally mix which will only take a few days. 

[oranjeboom]

7 minutes ago, JSHarris said:

The pressure won't make any significant difference to the stiffness of the pipe in practice, as the forces on the walls of the pipe from the pressure are very low, even if the pipe is at a couple of bar or so.  The inherent stiffness of the pipe exceeds anything that a bit of modest water pressure will improve, and you can just as easily crush a bit of UFH pipe that's filled with water as a bit that isn't.

Thank's @JSHarris....not what I wanted to read!!!!  Why isn't there a "Do not like this button" next to the "like this" button???? 

 

Loop number 4 like a jacuzzi until it settled down. Tightening some outside joins seems to help a little at least.

[oranjeboom]

7 minutes ago, PeterW said:

How are you filling the loops ..??

 

 

@PeterW Yes, that IS how I was going to do it. I assumed that I would have been shipped a defacto pressure kit from Wunda, so I released a bit of wind yesterday when I realised they hadn't. Quote included a pressure guage but that was it (i.e. no automated air valve) so I knew there must be some way of using that. I called them and they confirmed they had changed things a bit so now you can pressure test with pump unit in place, replace one of the temp guages with the pressure guage. Test each loop one by one, and then at the end when all return valves are off on the return bar, the manifold itself is under pressure.

 

Method here: http://www.wundatrade.co.uk/downloads/M07 Wunda Manifold.pdf page 4.

[Nickfromwales]

On 10/09/2017 at 11:40, richi said:

If I read your post correctly, you're filling/flushing the pipework from a garden hose that's dripping. If there's a drip, there's an opening to the atmosphere. If water is flowing past an opening in a pipe, Herr Prof. Dr. Bernoulli says it'll tend to suck air into the pipe.

The pressure differentials here would just see mains water squirting out . If air is getting in that way il donate my next donut. 

The reason it takes forever to get the air out is because most folk foolishly try and fill / bleed all the loops at the same time with the manifolds on . That means the air gets to the manifold, can't be ejected quick enough through a tiny bleed point, so guess where it goes? Back into ALL the loops and round and round we go. 

 

To fill / bleed UFH loops you simply connect a hose from the mains to one manifold drain point, and then another hose from the other manifold drain point going to a drain and then purge each loop one at a time. Takes literally seconds that way. Just close the manual actuator caps that comes on each one, ( factory fitted usually ), leaving the first one open. Run the hose into the one loop and go watch the end of the other hose. By the time you get to it you'll have probably already purged 100% of the air out. Repeat closing / opening accordingly. 

A 10 port manifold should be bled in 15-20 mins max. 

Shut all the drain taps, replace the blank covers provided and open the automatic air vent that's provided and go get some beer. 

Bingo bango. 

Edited January 14, 2020 by Nickfromwales
Edited to add detail.

[richi]

7 hours ago, Nickfromwales said:

The pressure differentials here would just see mains water squirting out. 

 

What pressure? It sounded like @oranjeboom is just flushing it through, with a hose at one end and a bucket at the other

[Nickfromwales]

The resistance of the water flowing through the UFH loops vs the cold mains would leave positive pressure in the hose, ( unless it's a ridiculously low pressure supply ). Plus the pipe would need to be constrained at the point of the pinhole. 

Likelihood of Venturi type suction occurring along the hosepipe? Imo, zilch.  

 

The ufh pump / manifold rarely gives a flow rate of more than 5 litres per minute through each circuit so it's vital to use the cold mains purge first to ensure that no pockets of air remain. The loop can airlock with ease when you revert back to the pump only if it's not 'blasted' through. 

Ive serviced a few where one zone has air-locked and the others are working. Sometimes you can get them going again if you turn every other actuator off but some I've had to purge through again, usually in installs where the pipes rise slightly as they move away from the manifold. 

[oranjeboom]

1 hour ago, richi said:

 

What pressure? It sounded like @oranjeboom is just flushing it through, with a hose at one end and a bucket at the other

I was flushing through, getting rid of air and have now pressurised all loops and the manifold. Started at about 3.5 bar, now just above 2.2. Not too concerned as I can see where it's losing pressure. A couple of the manifold connections need a tighten here and there, but will have a play with that before I hit the sack at 3am probably again. Faffing with my mesh chairs at the moment, getting my glazing protected, cutting the last of the 2000 cable ties and checking my datum for the 50th time.

 

How many chairs do I need? Every 2nd (200mm mesh) square? or can I spread them out a bit more?

 

 

[richi]

15 hours ago, Nickfromwales said:

Likelihood of Venturi type suction occurring along the hosepipe? Imo, zilch. 

 

I agree, but that's not what I was suggesting ;-)

[Nickfromwales]

If I've grabbed the wrong end of the stick I'll apologise and put it with all the other sticks  . 

[oranjeboom]

Didn't see venturi in my pipework today. A few drips in manifold but managed to screw it all up tight. Pressure seemed to hold too. So did my pressure, after doing an all nighter prepping everything. Half slab done, and hopefully get it all finished tomorrow if the bloody mixer turns up on time.

 

After 3hours of sleep this weekend, i may actually go to bed now! I'll post some pics soon. Actually better seal the front door b4 bloody wabbit or cat make it in there...

[oranjeboom]

Only had about 7hrs sleep in an 84hr period prior to the slab pour so felt pretty drowsy on the day of the pour. I was still alert enough to spot/rectify certain issues including:

 

- a bit of ufh pipe popping out of the concrete where bit of mesh had lifted. The guy's intention was to just try and cover it up with a few mm of concrete! In then end I placed a bit of quarry tile on top of pipe and then placed a concrete block on top that to push/keep the tile depressed into concrete. I can probably lift the tile later on and 'fill in' with a bit of conceret/latex

 

- builder had been using top of perimeter upstand as datum instead of datum on block wall behind the datum! Potential 20mm difference! Had to scrape quite a bit off and re-level.

 

However, it's all ended up pretty poor and I haven't even needed a level to spot high/low spots - some areas are quite noticeable when just walking across areas. There are a couple of areas that will need to be ground down as they are simply too high. Better that, than have the whole area needing to be latexed later on.

 

Looks like @Barney12 you're not the only one with  an uneven slab! 

 

Not going to get too stressed about this right now as I need to get on with other items. The positive thing is that the ufh pipes seemed to have survived the pour. Manifold pressure dropped down to 1bar since the pour (but stayed stable during the pour itself at around 3.2bar).

 

One of the hallways:

 

kitchen:

 

Tile with UFH pipe below:

 

Lounge room. Various levels! Plenty of latex needed.

 

Builder must have known on the day that things weren't great as he was already coming back about latexing etc (at his expense).

 

1) Should I get him back sooner than later to rectify things? His idea was to do any latexing just prior to my bamboo floor going in to stop plasters crud ending up on his 'perfectly latexed floors'. I can understand that to some extent, but at the same time don't want him to be uncontactable in 4-5months time.

 

2) How easy is it to shave off say 20mm of concrete? 20mm is a guestimate at this stage, hoping it is less. Presume the sooner this is done whilst the concrete is still a bit green, the better. What tools are we looking at?

 

Any other considerations?

 

Thanks, OB

 

[jamiehamy]

What concrete mix did they use,  do you know? I raised a question about slab levels in another thread - keen to find out what mixes the rolling floors used. There's got to be a better solution to this - too many people with uneven slabs! What's the common denominator? 

[Gone West]

12 hours ago, oranjeboom said:

1) Should I get him back sooner than later to rectify things?

 

Definitely sooner if you're confident he can do a good job with the latex. Unless you're lucky they have a habit of forgetting they said it or not answering their phone. You can always put down floor protectors until you lay the bamboo.

[Nickfromwales]

+1. Prime the concrete first with 50/50 PVA and water and absolutely saturate it. Do NOT let the builder just pour it straight on. 

The slab will suck it dry before it's levelled out properly, and make 1m% sure they use 2-part leveller NOT the water mix one.

  

[Onoff]

10 minutes ago, Nickfromwales said:

+1. Prime the concrete first with 50/50 PVA and water and absolutely saturate it. Do NOT let the builder just pour it straight on. 

The slab will suck it dry before it's levelled out properly, and make 1m% sure they use 2-part leveller NOT the water mix one.

  

Does it matter whether it's the ordinary "red" PVA or the more expensive "black", waterproof stuff?

[Nickfromwales]

Just now, Onoff said:

Does it matter whether it's the ordinary "red" PVA or the more expensive "black", waterproof stuff?

I've only ever used the cheapest ones which were also used for prepping for skimming etc. Never lost a patient. yet.  

[oranjeboom]

On 9/18/2017 at 08:38, PeterStarck said:

Definitely sooner if you're confident he can do a good job with the latex. Unless you're lucky they have a habit of forgetting they said it or not answering their phone. You can always put down floor protectors until you lay the bamboo.

I got him to email me his intent on returning to latex etc. Doesn't guarantee that he will of course. He also gets most of his work from a local conservation architect so the builder wouldn't want any negative feedback going back to architect.

 

Yes, he also mentioned the PVA glue mix @Nickfromwales, so maybe he does know about floors. 

 

However, I took the courage and got out the level and laser today. The valleys remind me of South Wales! 

 

Main concern is where the 4.5m slider is. I'd clearly marked for him where the max level of concrete should be and it's ended up well above that. 

 

My bamboo floor was supposed to end up just under the door threshold (as below), but in most places it'll be just above that if it's not ground down.

 

And valleys where I can get my fingers under the level. 

 

Then there's an area of concrete where I can really gouge at it with the end of my measuring tape - will that harden. Hopefully not concrete laitance as that could be disastrous.

 

I guess the concrete is still failrly green (poured a week ago) and patches above, may still harden. Apparently this was C35, but I intend to call the suppliers tomorrow to confirm if that really was the case.

 

And the high spots are probably best to have ground down now before it 'hardens'/cures/sets more and more (excuse the language). I know the process takes years, but the sooner I have him back to sort out his peaks at least (the valleys can wait with latexing a bit later on).

Edited September 19, 2017 by oranjeboom

[Gone West]

14 hours ago, oranjeboom said:

Then there's an area of concrete where I can really gouge at it with the end of my measuring tape - will that harden. Hopefully not concrete laitance as that could be disastrous.

 

I guess the concrete is still failrly green (poured a week ago) and patches above, may still harden. Apparently this was C35, but I intend to call the suppliers tomorrow to confirm if that really was the case.

 

And the high spots are probably best to have ground down now before it 'hardens'/cures/sets more and more (excuse the language). I know the process takes years, but the sooner I have him back to sort out his peaks at least (the valleys can wait with latexing a bit later on).

 

That's looks odd to me. I've never had concrete I can gouge with a tape measure after a week of hardening. Get him back asap and see what he thinks.