Waves in header tank

[mc2fool]

Says it all really....

I've been investigating a loud grinding/chuntering/dunno what noise (difficult to describe!) from our header tank during and immediately after periods of prolonged high water usage.

So, I went up, took the cover off and manually played with the ballcock (steady there, this is Building and DIY!) and everything is working fine.

So, I set up the conditions under which the noise happens and it all starts off fine but then after a while waves appear in the tank! Not exactly huge ones but enough for the float ball to bobble up and down enough to alternate the inlet valve between gushing and almost off, and on an around a quarter of a second cycle, and the rapid on-off is the noise.

So ... well I figure I'm going to have to put some baffles in the tank, to stop the waves forming (maybe simply a bottomless plastic bottle around the inlet, with a slot cut out for the ballcock arm?), but I thought I'd (ahem) float the problem here to see if anyone has any other ideas.

[kiloran]

Can you bend the arm of the ballcock so that the water is at a slightly lower level? This change in the conditions might be enough to sort it.

--kiloran

[DrFfybes]

Check the ballcock isn't taking in water. It should float on the surface, but if it is partially sunk it might 'bob' more than it should.

Paul

[9873210]

So, I set up the conditions under which the noise happens and it all starts off fine but then after a while waves appear in the tank! Not exactly huge ones but enough for the float ball to bobble up and down enough to alternate the inlet valve between gushing and almost off, and on an around a quarter of a second cycle, and the rapid on-off is the noise.

One solution is to add more hysteresis. The water level where the valve opens must be below the water level where the valve closes. In between these levels the valve stays in whatever position it was already in. As the tank is filling the valve will stay open until water reaches the higher level, and then stay closed until the water level reaches the lower level. If this dead band is larger than the waves there will be no problem.

At least some valves allow you to set the two levels independently, either with two set screws, or if the float moves freely on a vertical shaft by moving the snap rings at the ends of motion.

I should add that if you set up the outflow to be very slightly less than the inflow (with the valve open) you will always get oscillations. The best you can do is increase the period by adding hysteresis as above or by increasing the surface area of the tank.

[Mike4]

Says it all really....

I've been investigating a loud grinding/chuntering/dunno what noise (difficult to describe!) from our header tank during and immediately after periods of prolonged high water usage.

So, I went up, took the cover off and manually played with the ballcock (steady there, this is Building and DIY!) and everything is working fine.

So, I set up the conditions under which the noise happens and it all starts off fine but then after a while waves appear in the tank! Not exactly huge ones but enough for the float ball to bobble up and down enough to alternate the inlet valve between gushing and almost off, and on an around a quarter of a second cycle, and the rapid on-off is the noise.

So ... well I figure I'm going to have to put some baffles in the tank, to stop the waves forming (maybe simply a bottomless plastic bottle around the inlet, with a slot cut out for the ballcock arm?), but I thought I'd (ahem) float the problem here to see if anyone has any other ideas.

Firstly, nowadays we have to call them "float valves". Ballcox is reseverved now for the stuff some people post in the "Current Affairs and News" board.

Secondly this is a common problem, resonance in the cold water cistern. I first encountered it aged about 7, when it happened in my mum and dad's house. Dad let me up into the loft where I observed this wave effect develop first hand. Fascinating, and partly why I became a plumber!

Anyway dad fixed it by adding some damping, in the form of one of those (then) newfangled plastic flowerpots. He drilled a hole in the centre of the base, removed the (then) ball cock float and puting the flower pot on and replacing the float. Now the flowerpot sat partially submerged and had a suitably damping effect to stop the standing waves resonating.

Nowadays the fix it to fit an 'equilibrium' float valve. These decouple the effect by closing two or three seconds after the setpoint water level is reached.

[monabri]

Does the inlet water pipe to the header tank vibrate (like Billy O so you'll see it and if you put your hand on the pipe you'll feel it!)?

[mc2fool]

Thanks for the replies to date.

@kiloran: the arm is probably bendable but I'm not sure I see how that'll make any difference.

@DrFfybes: I'll check but it does float on the surface and I don't think the ball is taking in water, but in any case I'd have though that if it had done so that'd actually dampen its bobbing.

One solution is to add more hysteresis. The water level where the valve opens must be below the water level where the valve closes. In between these levels the valve stays in whatever position it was already in. As the tank is filling the valve will stay open until water reaches the higher level, and then stay closed until the water level reaches the lower level.

No, it's not like that. As the water gets higher and the arm moves up the valve just closes gradually and the water flow gets slower and slower. It's a traditional ballcock, like one of these:


https://www.stevensonplumbing.co.uk/ballcock-valves.html

Firstly, nowadays we have to call them "float valves". Ballcox is reseverved now for the stuff some people post in the "Current Affairs and News" board.

Well the image and link above still calls them ballcocks and even Screwfix has their float valves under a heading of BALLCOCKS.

Anyway dad fixed it by adding some damping, in the form of one of those (then) newfangled plastic flowerpots. He drilled a hole in the centre of the base, removed the (then) ball cock float and puting the flower pot on and replacing the float. Now the flowerpot sat partially submerged and had a suitably damping effect to stop the standing waves resonating.

Ha! Interesting! I've got a few of those lying around, do you suggest a pot just larger than the ball or much larger? It's an easy shot so worth a try, although may not have the same effect as in your mum & dad's house's tank, as ours is is actually the communal header tank for our small block of six flats*, so is quite a big tank; the ballcock only extends into about a fifth of its length.

(* Before anyone pipes up that it's the landlord's problem, the block is a collectively enfranchised self managed one, which means mostly me!)

@monabri: No vibration that I'd noticed, certainly not like Billy O but I'll check.

[kiloran]

@kiloran: the arm is probably bendable but I'm not sure I see how that'll make any difference.

My reasoning is that it's some form of closed-loop feedback mechanism which has a number of variables. Bending the arm will change one or more variables and perhaps stop the feedback. Crude but might be effective.

--kiloran

[AsleepInYorkshire]

I think the problem maybe that the beer you've been fermenting in there is attracting a few rats and when you look in the tank you can't see them as they have very strong lungs and are sat at the bottom of the tank holding their breath

I'd second Mike's advice and look into equilibrium valves

AiY(D)

[88V8]

It's a traditional ballcock, like one of these:

https://www.stevensonplumbing.co.uk/ballcock-valves.html.

Ahh, a proper brass valve.

OK, easy then.
Buy some brass or copper tube and jam it into the outlet to make a 'silencer'. That will prevent the standing wave.

You may find what you need at a car accessory shop, it won't be a standard plumbing size, that got scuppered when they made silencers non-kosher a few years ago.

Edit... or perhaps plastic https://www.aqualine.co.nz/store/ballcocks/silencer-ballcock/

V8

[mc2fool]

Ahh, a proper brass valve.

OK, easy then.
Buy some brass or copper tube and jam it into the outlet to make a 'silencer'. That will prevent the standing wave.

You may find what you need at a car accessory shop, it won't be a standard plumbing size, that got scuppered when they made silencers non-kosher a few years ago.

Edit... or perhaps plastic https://www.aqualine.co.nz/store/ballcocks/silencer-ballcock/

V8

If you are talking about a bit of pipe that goes from the outlet downwards to below the water level, that is of course illegal.

However, there is quite a big air gap in our configuration and a bit of pipe going from the outlet downwards to above the highest water level, leaving still the minimum required air gap, might well reduce the splash, and with it the impetus for waves.

Hmmm, thinks ... or maybe the other way round; a large-ish diameter pipe suspended (somehow) below the outlet, with the top of the pipe an inch or two above the highest water level and the bottom of the pipe and inch or two above the bottom of the tank, so that all of the splash is contained within the pipe....

[Mike4]

Or just do the job properly and fit something like this:

https://www.amazon.co.uk/Fluidmaster-PR ... r=8-1&th=1

[mc2fool]

Or just do the job properly and fit something like this:

Yeah, that's fine in theory but I suspect the current setup has been there since the flats were built (1978) and there's bunches of encrustation around the joints and you know how it goes ... turn off supply and try to undo nut ... what nut?

Wire brush encrustation to try and discover nut. Encrustation not moving, try scraping it off with a knife. Some progress but not a lot. Apply white vinegar (we're in a hard water area), then wire brush and scrape again, rinse and repeat until nut finally discovered. Try to undo nut with wrench, nut not moving. Apply WD40. Try again, still not moving ... angle grinder time? etc!

And there's always the risk in trying to undo old things that are seized up that something breaks ... so I prefer one of the more ad hoc solutions described that can be applied without touching the existing parts, which if I did could end up causing me more grief that it's worth!

[88V8]

Ahh, a proper brass valve.
OK, easy then.
Buy some brass or copper tube and jam it into the outlet to make a 'silencer'. That will prevent the standing wave.

If you are talking about a bit of pipe that goes from the outlet downwards to below the water level, that is of course illegal.

Indeed, because if one's cold tank is not covered and hence full of deceased mice and the odd starling, and if the water pressure drops in the feed pipe such that the silencer might suck a little of the water out of the tank into the mains, then the mains might get polluted. But if the tank is covered..... and after all, it is probably the water you brush your teeth with and you still seem to be alive....

V8

[pochisoldi]

@kiloran: the arm is probably bendable but I'm not sure I see how that'll make any difference.

My reasoning is that it's some form of closed-loop feedback mechanism which has a number of variables. Bending the arm will change one or more variables and perhaps stop the feedback. Crude but might be effective.

--kiloran

Three more ways to avoid oscillation triggered by waves making the valve bounce open and closed:

1. If there's a movable outlet nozzle (more likely in a toilet cistern than a header tank), move it so it points somewhere else so that the waves radiate in a different direction or the waves+reflections in the area of the float cancel each other out.

2. Use a piece of plastic string to tie an old CD to the ballcock arm, so that the CD normally sits horizontally in the water.
When the float tries to quickly bob up and down the CD will drag in the water, dampening the reaction of the ballcock arm to the (relatively) rapidly changing input. It will have no effect on how fast the tank refills (i.e. a slow steady rate of change of level), and will only come into play if the ballcock arm starts moving quickly up and down)

A third method is to fit something in the pipework to reduce water hammer - the usual (el cheapo) method is to have a dead end stub pointing upwards which is usually filled with air. (had one of those in a flat I owned, never had a water hammer issue).
The alternative is to buy and fit a water hammer arrestor.
These both work by absorbing pressure pulses in the pipework.
If the noise from the ballcock is not affected by turning the kitchen (or any directly mains fed) tap on and off, then I'd doubt a water arrestor would help.
It's also the most awkward/expensive way to fix the issue - last resort IMHO, given that methods 1 and 2 don't involve compression joints or soldering.


PochiSoldi

[jfgw]

Where the ball-cock tail passes through the side of the tank, there should be a back-plate to provide additional support. It is the black plastic component with two holes in it here, https://www.toolstation.com/byelaw-30-kit/p40164. (As an aside, there should be a screened breather in this kit—hopefully, it is just missing from the photo.) The two holes allow it to be used with 1/2" or 3/4" ball-cocks. Obviously, you don't need one of these in a galvanised or asbestos cement tank as they are stiff enough without.

Make sure that the pipe going to the ball-cock is securely clipped.

Replace the ball-cock washer, or replace the entire ball-cock. A normal BS1212 part 2 (brass) ball-cock should be ok. Those old part 1 valves are not used much now.*
Part 1: https://www.screwfix.com/p/flomasta-brass-part-1-float-valve/913fy
Part 2: https://www.screwfix.com/p/flomasta-brass-part-2-float-valve/358fy

If you are still having trouble, try replacing the ball with a bigger one. They are normally 4 1/2" but a 5" or 6" one will screw on.


Julian F. G. W.


*If you replace a ball-cock, you should comply with the current regs. Normally, this means that an old part 1 valve should be replaced with a part 2 (or 3 or 4) valve, and that a means of isolation (such as an isolation valve or stop-cock) should be fitted if there is not one already.