Loft water tank - overflow pipe problems.

[bri160356]

Hi guys, I have recently replaced an old dripping ‘float valve & ball’ in my loft ‘Domestic Hot Water’ feed tank with a complete new assembly and whilst testing the ‘soundness’ of the connection through the plastic wall of the tank by holding the ‘ball’ submerged and filling the tank above the level of the valve, I was horrified to note that the overflow pipe/assembly was not coping with the water ingress at all. The tank rapidly filled, and even with the outlet completely submerged the water ‘outflow’ was very sedate ! – the tank continued to fill rapidly. My immediate reaction was that the pipe was blocked and I duly replaced with a new 90 degree tank connector and 21.5mm standard solvent weld pipe. However it made no difference. Further investigation showed that if I lightly used the palm of my hand as a ‘plunger’ against the overflow pipe (when it was completely submerged) then it appeared to release a slight air lock and the water would gush out like a fire hose. The water receded rapidly until the level allowed air back into the overflow pipe at which point it would revert back to its very sedate outpouring and the tank would again rapidly start to fill.
The overflow pipe and the float valve connection are both about 8cm below the rim of the tank. The overflow pipe is a straight run (no bends or angles except the 90 degree connector) approx 3m long, across the roof space approx 30 degrees below the horizontal and it emerges outside under the eaves.
This overflow pipe arrangement is original (25yrs from the house being new) as is the plastic tank, so I guess this overflow issue was a potential flood disaster waiting to happen.
How can I solve this problem? Any help from you wonderfully knowledgeable people would be gratefully received and many thanks in advance.
Regards

 

[Nige F]

The overflow pipe is a straight run (no bends or angles except the 90 degree connector) approx 3m long, across the roof space approx 30 degrees below the horizontal and it emerges outside under the eaves.

Below horizontal should be above horizontal so it falls from tank to eaves - and supported at least once in the middle of 3m. run

 

[bri160356]

Hi, many thanks for the reply but I sincerely believed that my description that the pipe runs from tank to eaves at an angle of 30 degrees 'below' the horizontal was correct.

But many apologies if it is my terminology that is incorrect or confusing; I should probably have said "the pipe runs 'downhill' from tank to eaves at an angle of approx 30 degrees from floor level"
The pipe is supported at 2 separate points in its length.
Given the info above any thoughts that may solve my problem?

Regards

 

[Burnerman]

Bit of a bizarre problem for sure....are you sure there's no blockage / kinking / compression where the pipe heads through the wall or eaves?
The overflow should actually be below the water valve inlet position.
John

 

[bri160356]

Hi, many thanks for your interest; the pipe is new; absolutely no blockages or kinks. As I said in my OP once the water starts to flow (after my intervention)it absolutely flies out like a firemans hose but returns to its sedate flow once air re-enters the pipe.

 

[bri160356]

Bit of a bizarre problem for sure....are you sure there's no blockage / kinking / compression where the pipe heads through the wall or eaves?
The overflow should actually be below the water valve inlet position.
John

Hi again, as I said in my OP the overflow and inlet are both approx 8cm below the rim.
If I lowered the outlet point to just above the nominal water level do you think the extra pressure would be enough to dislodge any small air-pockets and get the water flowing out properly?
Is it a trapped air pocket, combined with the shallow angle of the outlet pipe that is stopping the water flow ??
Regards

 

[Burnerman]

I'm certainly not clued up enough on fluid dynamics or Eulers laws, but when you block the pipe end and then release it you are inducing a self syphoning effect of the water.....
The same effect could possibly be introduced by allowing the overflow to fall vertically first, therefore increasing the water velocity initially before then flowing out to the eaves but there again I might be talking my usual bovine excrement
John

 

[bri160356]

Hi, great input – certainly not bovine derived. . So are you saying that the pipe should be effectively a dog-leg of maybe 30cm vertically downwards from the outlet position and thence outward in a straight run to the eaves?

Or would my suggestion of lowering the outlet point be another solution? (this would be preferable for me as no extra kit required)
Regards

 

[Burnerman]

My points are purely guesswork Bri - but certainly the outlet should be below the fill valve, but I can't see that having any effect on the flow issue!
I'd keep the vertical drop a bit longer than 300mm if possible.
You'll need a blanking plug to close off the old hole if you do reposition.
Lets know how it goes!
John

 

[bri160356]

Hi John, many thanks; a trip to the merchants beckons.
Not sure which option I'll try first but I'll let you know the outcome.
Regards

 

[bri160356]

My points are purely guesswork Bri - but certainly the outlet should be below the fill valve, but I can't see that having any effect on the flow issue!
I'd keep the vertical drop a bit longer than 300mm if possible.
You'll need a blanking plug to close off the old hole if you do reposition.
Lets know how it goes!
John

Hi John, as you said that the outlet should be lower than the inlet I took your advice and tried that option first. I measured the original ‘overflow’ outlet and the centre of the hole was only 5.5cm below the rim of the tank. I drilled a new hole centred 11.5cm below the rim and re-fitted the bent ’90 degree’ tank connector and overflow pipe.(I’ve fitted a blanking plug in the original hole as you advised).
To my (initial) disappointment, as I began to fill the water tank, the outflow rate of the newly positioned connector was exactly the same as before i.e. very sedate and was certainly not matching the rate of ingress.
However, once the level of water rose to be 6.5cm above the overflow connector centre (I measured it carefully) it appears that the siphon effect came into force and the water level receded rapidly until it was low enough to allow air back into the overflow pipe at which point the sedate outflow would be resumed – until it was once again 6.5cm underwater and the siphon effect would start again. This cycle of sedate outpouring followed by a rapid outpouring would continue indefinitely until the water ingress was stemmed. So in essence problem solved!
I guess there must be some correlation between the ‘bore’ of the overflow pipe (21.5mm), the downward angle of the pipe (in my case about 25 degrees) and the pressure of water at the outlet point (6.5cm below the surface) before ‘full’ water flow commences.
That would explain why my original connector position (5.5cm below the rim) was totally inadequate and a potential flood disaster waiting to happen.
I don’t profess to understand the dynamics of exactly what was going on but I’m sure someone does! (I’m not a plumber; but I guess that’s obvious!) ; I’m just pleased it’s sorted.
Many thanks for your help and guidance John; it’s greatly appreciated.
Regards

 

[Agile]

I dont understand what is going on but there is another and better fix!

You should have a stopcock or isolating valve feeding the float valve.

Adjust this so the inlet flow is just below the sedate rate and so it will never run over that.

Make sure the fill level is oly about 80-100 mm above the base of the tank.

Tony

 

[Burnerman]

Either way, I'm sure we appreciate the feedback, although its certainly been a strange one!
So long as your water consumption from that tank doesn't allow the thing to run dry you can restrict the water inlet....another method to prevent a descending ceiling would be to fit another overflow!
I wonder just how many installations have been tested like this.
John

 

[bri160356]

...
I wonder just how many installations have been tested like this.
John

Hi John, I have no way of knowing but my two neighbours live in detached houses exactly the same as mine, so no doubt the plumbing was installed and tested by exactly the same bricklayer.
Regards

 

[Peet]

Firstly thankyou for a decent and very descriptive post!

Once the inlet of the overflow pipe is fully covered with water the siphon should start, if the pipe run is good. The fact that you have to raise the water past this, and submerge it more, suggests that the pipe run is not all it seems.

The overflow pipe should always be falling towards its end and not have any level sections or any sections that start to go back up in it. Air locks are one reason for this but you also don't want any stored water in it in case it freezes. Once the overflow pipe is not in use anymore it should be dry.

Also, and this is important, the joints should not only be water tight but also air tight otherwise once a siphon starts it effectiveness will reduce. I've seen a few where the solvent weld joints are simply pushed together.

The siphon is important as it enlarges the effective capacity of the overflow pipe. While the water is partially covering the inlet of the overflow it makes a gurgling noise (which if the overflow pipe falls a long way can be very loud) and acts as a warning to what's going on. If the water can continue to rises to where it covers the inlet completely then it goes quiet but the flow increases markedly. Its flow is then depenedent on the pipe's bore and also vertical fall. For this reason some of the overflows I have fitted, once clearing the building, continue vertically downwards. Some fall by as much as several feet before ending. In small volume tank (like toilet cisterns) this ability to suck is important as there's a lot of systems around with an overflow that cannot take the full flow of the ball valve. Like Agile says turning down the inlet valve can be good but in a cold water tank sometimes you dont want to do that.

The vertical distance from the inlet of the overflow to the top of the tank gives you a buffer between the overflow occuring and the tank overflowing. Where the ball valve is is down to you and its requirements. Usually the ball valve mounting is above the normal working water line of the tank.