Megaflow / water hammer

[dom]

Long time off this forum, but hoping the plumbing brains still frequent thi s place!

PROBLEM:

When Central Heating is set to hot water *only*, and when the cylinder is i n the last few minutes of it's heat-up cycle (before the boiler cuts out) - water hammer occurs close to (or in) the Megaflo cylinder.

Water hammer never occurs if CH is completely off (no matter how the hot wa ter is used or not used), nor does water hammer occur if CH set to radiator s only.

Water hammer occurs occasionally and to a lesser degree, again in the last few-minutes of cylinder heat-up, if both cylinder heating and radiators are on.

By touch, the water hammer seems to be affecting both the CH inlet and outl et pipes to the cylinder equally (and sometimes also the cold mains water i nlet to the cylinder, but not the hot water outlet).

The water hammer can almost always be stopped by simultaneously switching o n heating to the radiators (and temporarily stopped by drawing off some hot water).

Gas CH, modern house (c. 12 years), Megaflow, S-plan, cold water mains pres sure up to 6 bar. Seems to be a standard megaflow configuration with genuin e Heatrae Sadia parts.

SO FAR:

All pipes seems reasonably well clipped. Bled air from rads. CH water seems clean. Checked S-plan (2 port) valves have correct direction of flow. CH system when cold pressure 1.2 - 1.5 Barr, when hot pressure 1.8-2.0 Barr . Mains water cold 4.5 - 6.0 Barr, hot water 3.0 - 4.5 Barr. Have checked that the 8 Barr temp/pressure discharge valve, and 10 Barr saf ety valve both manually operate, and discharge to tundish. Have regenerated air pocket as per instructions on the Megaflow cylinder.

I haven't yet tried turning off the cold water inlet to the cylinder, and l etting the boiler do a how-water warm-up cycle with the Megaflow depressuri sed - to attempt to prove the problem is definitely CH side, and not mains- water side (incidentally the quarter-turn valve just before pressure-reduci ng valve is frozen in the fully-on position, so have to use the under-sink main cold-water stopcock to isolate the hot-water system).

Could it be pump overun at the boiler, and the pressure-operated CH-circuit bypass valve not operating correctly? My impression (but not certainty) is the that the 2-port valve to the cylinder is still open when water hammer occurs.

[dom]

Clarification - I should have made clear that the water hammer (when it occ urs in a nearly-fully-heated cylinder) is not the few-Hz bang-bang-bang, bu t more more like more-or-less continuous-tone vibration noise that randomly rises and falls in loudness, randomly dropping to barely audible for a few seconds only to increase again, and continue until the boiler finally cuts out.

[John Rumm]

So this is a hammer in the CH primary system, not in the DHW take off from the cylinder?

Has it always done this, or is it new?

What is the set point temp on the cylinder stat?

What kind of circulator pump is it? (and what speed is it set to?)

[dom]

Thanks for your interest, John.

1) I *think* it's in the CH primary - hence me planning to try depressuring the hot water side and let it go through a heat up cycle to be certain.

I would say for definite that the cylinder HW outlet is not affected by vibration. Trouble is the noise is of a pitch, and reflected around in an airing cupboard, that makes it hard to localise - except by touch.

By touch I can say CH inlet and outlet to the cylinder affected equally, cold feed affected sometimes.

2) The noise issue has (I'm told) re-occurred from time-to-time over the 12 years (approx) of usage. Allegedly no problems in the early years, but has come and gone spontaneously after that. 3) Cylinder stat set to its mid-point and what I would describe as a normal domestic HW temperature at the tap (sorry, no thermometer immediately available). 4) Circulator pump is built into the boiler (Ideal icos) and is non-adjustable single-speed.

[Tim+]

Do you have a bypass valve that allows water to circulate back to the boiler after the HW cylinder has reached its set temperature and the motorised valve has closed?

I'm guessing that the noise relates to this.

Tim

[dom]

Yes, there is one, and I'm wondering if that's the issue too.

However it isn't my impression that the motorised valve has closed - but I will check this again to be certain.

[tabbypurr]

Water hammer can be solved with an upward pointing blanked off bit of pipe. It fills with dissolved gas, acting like a capacitor.

NT

[dom]

Yes, there is one of those in a piping "top loop" near the cylinder.

[Tim+]

It can also be caused by too much "dead leg" volume. Might be worth bleeding all the air out of it.

Tim

[John Rumm]

Could it actually be bearing noise or vibration from the pump itself that is just "ringing" the coil in the cylinder?

[dom]

Yes - because of the pattern of vibration I had indeed wondered about 'ringing' of the coil.

One theory I entertained was that as the tank coils got hotter that they were expanding and becoming loose in whatever restrains them (I've never sawn a megaflow in half to determine all its secrets).

However Tim+ suggests that it relates to the bypass valve, and that would rather fit with an ill-defined pattern of occurrence of noise dependent on the state of each individual TRV.

Perhaps a combination of the two - a coil prone to resonance, and a bypass valve issue that triggers it?

[harry]

The best way to pinpoint a noise is with a listening stick. This can be a big screwdriver. You put the blade on the suspected source and press the handle against your ear (or rather the area just in front of your ear, not the actual ear hole.)

Try throttling one of the pump isolation valve (the down stream one).

Try opening and closing radiator valves.

You can get "flutter" on the obturating device on any valve due to wear. Zone valves are specially suspect. I think this last may be your problem as you say the noise depends on shutting heating on/off. This is usually done by means of an "electric/zone valve".

Using listening stick, check the pressure relief valve inside the boiler. (This maintains constant pressure in the system by opening as the radiator valves close.)

It is on a pipe linking flow and return pipes. (Adjusting this might help with the problem)

[Fredxxx]

Last few minutes suggests the boiler may be in bypass mode. There should be a bypass valve and I would suspect that. Some are adjustable and if this is too I would be tempted to alter it's position to see if the hammer stops or gets worse.

The other possible is your 2-port valve. This could be 'bouncing' internally and causing the hammer. Most valves have a lever to manually operate it. If you open it during a hammer episode and it goes away it confirms it could be this valve, but unfortunately doesn't rule out the bypass valve.

[John Rumm]

You could wait for the noise and then go have a listen / feel of the pump to find out.

Spring loaded bypass valve would also be possible, adjusting its setting while the noise is there would probably influence it.

If the noise happens when the rads are off, then the TRV settings ought not come into play.

Yup certainly could be.

[Dave W]

Clarification - I should have made clear that the water hammer (when it occurs in a nearly-fully-heated cylinder) is not the few-Hz bang-bang-bang, but more more like more-or-less continuous-tone vibration noise that randomly rises and falls in loudness, randomly dropping to barely audible for a few seconds only to increase again, and continue until the boiler finally cuts out.

Is there anything as simple as a ball-c*ck in a cold-water header tank in your system? I only ask because I had just such a water-hammer sound as you describe, like a low cello note. This was just after I repaired the plastic ball which had got half full of water, but was now so light it vibrated. I cured it by clipping a submerged metal plate to the arm of the ball c*ck by a clothes peg to act as a damper.

[dom]

No, it's an entirely sealed system. CH has filling loop, and expansion tank in boiler. Hot water system is mains pressure with an air pocket in the top of the Megaflow cylinder.

[dom]

in the last few minutes of it's heat-up cycle (before the boiler cuts out) - water hammer occurs close to (or in) the Megaflo cylinder.

water is used or not used), nor does water hammer occur if CH set to radiat ors only.

t few-minutes of cylinder heat-up, if both cylinder heating and radiators a re on.

tlet pipes to the cylinder equally (and sometimes also the cold mains water inlet to the cylinder, but not the hot water outlet).

on heating to the radiators (and temporarily stopped by drawing off some h ot water).

essure up to 6 bar. Seems to be a standard megaflow configuration with genu ine Heatrae Sadia parts.

afety valve both manually operate, and discharge to tundish.

letting the boiler do a how-water warm-up cycle with the Megaflow depressu rised - to attempt to prove the problem is definitely CH side, and not main s-water side (incidentally the quarter-turn valve just before pressure-redu cing valve is frozen in the fully-on position, so have to use the under-sin k main cold-water stopcock to isolate the hot-water system).

it bypass valve not operating correctly? My impression (but not certainty) is the that the 2-port valve to the cylinder is still open when water hamme r occurs.

Some follow-up - that has unfortunately left me none the wiser:

I've tested the cylinder warm-up with the hot water-side isolated from the mains and depressurised. Nada, nothing, zilch. No water hammer.

Repeated the test immediately (after venting some hot to get another heat-u p cycle), this time with the cylinder pressurised - this time got water ham mer.

BUT, using Harry's suggestion of a listening stick, the strongest vibes see m to be at the CH inlet to the cylinder, and slightly less at the CH outlet . Virtually none at the cold water inlet, and definitely none at the HW out let.

I'm also pretty sure the water hammer is happening at the end of the cylind er heat-up cycle, but *before* the the cylinder 2-port valve closes.

I'm mystified.

New data suggests it isn't the boiler circuit bypass valve (or even the CH side at all).

That pretty much leaves the pressure reducing valve on the cold water inlet side, and/or the >8-bar/>90c safety valve.

I believe that safety valve also contains a non-return valve.

Could the water hammer be some sort of interaction between the air-pocket i n the cylinder and a defective non-return valve?

I'm getting to the point where "replace every component near the tank" is s ounding a sensible course of action.

New ideas/theories very welcome.

[John Rumm]

The one on my Vaillant does not appear to have any additional non return on the side mounted over temp / pressure valve - it relies on the tundish for that. Its basically a rubber faced piston held against a flat brass mating surface by spring tension.

Have you checked there is no let by into the tundish? I noticed with mine that if one turns taps off abruptly, it can shock a tiny amount of water past the side mounted over temp/pressure valve. Over time that built up some scale on it and then it started letting by continuously. I had to descale it and polish the brace facing part to get it to resel properly.

You may find some of the controls rather pricey alas!

Change in the volume of the air pocket? (Is it in a bladder on the megaflow?[1]) - you could try adding a bit more air charge to it and see if that makes a difference.

[1] The unistor has a separate expansion vessel externally.

[dom]

========================= ================\

========================= ================/

Ah - sorry, I was being unclear.

According to the Megaflow documentation, the non-return valve is inline wit h the cold water feed into the cylinder, i.e. pressure-reducing-valve - > safety-valve (with non-return on inlet sid e) -> tank-inlet.

I'm guessing the intent here is that as the pressure rises inside the cylin der, it can't create back-pressure on the pressure reducing valve.

My (current) theory is that this non-return valve isn't effective, and the water-hammer is some sort of resonance between the cylinder air-pocket and back into the cold supply.

The tundish here had been wet recently, but now seems to be completely dry since I regenerated the air-pocket in the cylinder as per the instructions on the megaflow.

(the air pocket AIUI is simply the top portion of the cylinder, with the ho t outlet pipe dipping sufficiently deep into the cylinder that the air has no way out.This would tally with their instructions to turn off the cold su pply, open the farthest hot tap, and then hold open the safety valve - unti l it stops gurgling - to regenerate the air pocket)

[harry]

is in the last few minutes of it's heat-up cycle (before the boiler cuts ou t) - water hammer occurs close to (or in) the Megaflo cylinder.

t water is used or not used), nor does water hammer occur if CH set to radi ators only.

ast few-minutes of cylinder heat-up, if both cylinder heating and radiators are on.

outlet pipes to the cylinder equally (and sometimes also the cold mains wat er inlet to the cylinder, but not the hot water outlet).

ng on heating to the radiators (and temporarily stopped by drawing off some hot water).

pressure up to 6 bar. Seems to be a standard megaflow configuration with ge nuine Heatrae Sadia parts.

safety valve both manually operate, and discharge to tundish.

nd letting the boiler do a how-water warm-up cycle with the Megaflow depres surised - to attempt to prove the problem is definitely CH side, and not ma ins-water side (incidentally the quarter-turn valve just before pressure-re ducing valve is frozen in the fully-on position, so have to use the under-s ink main cold-water stopcock to isolate the hot-water system).

cuit bypass valve not operating correctly? My impression (but not certainty ) is the that the 2-port valve to the cylinder is still open when water ham mer occurs.

e mains and depressurised. Nada, nothing, zilch. No water hammer.

-up cycle), this time with the cylinder pressurised - this time got water h ammer.

eem to be at the CH inlet to the cylinder, and slightly less at the CH outl et. Virtually none at the cold water inlet, and definitely none at the HW o utlet.

nder heat-up cycle, but *before* the the cylinder 2-port valve closes.

H side at all).

et side, and/or the >8-bar/>90c safety valve.

in the cylinder and a defective non-return valve?

sounding a sensible course of action.

Does the water hammer stop instantly when you draw some hot water? Does it stop instantly when the CH circulation pump is switched off? Water hammer is associated with water flowing through the pipework. One sort is cause by something in the flow flapping (usually a valve part.) The other sort is caused by trapped air/vapour in the pipework. Can you see any high points in the pipework where air could be trapped? Are you sure the expansion vessel is sound? ie has air in it at the right p ressure. You could try messing round with this pressure (and the system pressure to see if it makes any difference.

Very rarely in domestic installations you get water hammer caused by cavita tion. High water temperatures and low pressure (caused by the pump) makes the water to turn to steam, the hammer occurs when the steam bubble collap ses. This usually happens when system pressure is too low.

The safety valve has twin functions. One is over pressure. On later ones there is also an over temperature function.

Any non-return valve is to stop heated water getting back into the mains wa ter if the mains water is turned off (at the stop tap). So the NRV only actually opens when you are drawing hot water.