Pump sucking in air

[gaswizzard]

If you were to open a hot outlet below level of cylnder (kitchen tap) the water would be syphoned out of the cylinder due to the absense of an air break , hence the cylinder would collapse , i've seen it happen a number of times , one being in my own house when I was a kid , the cold feed/vent froze in loft and the cyinder collapsed.

I'll always remember the old man (he was a plumber) taking the cylinder out and feeding an air line down it to straighten it out.

 

[dcawkwell]

to be honest although i am sure someone will correct me i cant see the point of the expansion pipe since even if this wasnt there any expansion would simply go back up the cold feed pipe so i would keep the valve closed if that is working for you

...........and if the cold feed got blocked/shut off/frozen? , cylinder would collapse.

how? if cold feed is frozen then there can be no flow of water except the few litres in the pipes below the cylinder. cant see that collapsing a cylinder. even then the movement of expansion in coldfeed due to heating is likely to make this less likely
lets revise it and say have the valve open just ever so slightly
that should cure the problem. but whatever works.
oh for modern systems instead of dark ages stuff

Hmmm. Next conversion I do I'm going to cap both feed and expansion
and then try it.

 

[misterdubya]

Try isolating the cold feed to the cylinder and then open a hot tap with a non-return valve in the open vent watch the cyliner colpase -still its easy to transport it to the scrapyard afterwards!!

Or how to create an oversized egg timer.

 

[Onetap]

Hmmm. Next conversion I do I'm going to cap both feed and expansion
and then try it.

It does happen, particularly on unvented water cylinders, which have a cold feed but no open vent. Isolating the cold feed and draining an unvented system through a tap on a lower floor is liable to cause strange noises in the airing cupboard.

Some (but not all!!!!) unvented cylinders had a certain type of T&P relief valve which had a useful additional feature of acting as a vacuum breaker valve. The cylinders have great strength under pressure but, having no internal formers or bulkheads, have virtually no strength to resist a vacuum. They will collapse like a plastic coke bottle under a slight vacuum; it's the same principle; you won't burst a plastic pop bottle by blowing into it, but you can suck it flat very easily.

PS
http://www.rwc.co.uk/public/training/tgmar04p2.pdf

See the paragraph starting with "The fifth function......" on the right hand side of page 2.

 

[Onetap]

The problem is the pump is pulling air in from the open vent.

The simple answer is that you should not fit any valves to the open vent; you should increase the available pressure at open vent tee by;
a) raising the tank or
b) reducing the pressure losses due to friction along the flow path from tank, to cylinder cold feed to cylinder outlet, etc., and/or
c) reducing the pressure losses due to friction by reducing the flow rate by throttling the pump's discharge (or not turning the shower on full-blast).

Regarding b above, there are pressure losses due to friction along all the pipe and all the fittings that the water flows through. You can reduce the pressure losses by using larger pipe, using pulled bends rather than elbows,using full bore ball valves (rather than cheap service ball valves or gate valves), using copper rather than plastic (larger bore, no inserts at joints), etc.. Any restriction causes pressure losses. You can find tables for these pressure losses at a given flow rate and can calculate the pressure at any point in the pipe.

The pressure at the open vent tee is equal to the static pressure/head MINUS the frictional pressure losses. In your case, the frictional pressure losses are too great and/or the head is too small; the pressure at the open vent tee is less than atmospheric, and air enters the pipework.

You could try running the shower and gradually increasing the flow rate to see at what point air is sucked in. This would give you an idea of how severe the problem is and whether it could be fixed by replacing a few fittings or whether it need a whole rethink.

PS Fitting a non-return valve in the open vent would stop air being sucked in; you'd also get cavitation in the pump which would wreck the plastic impeller.

 

[Whitespirit66]

oh for modern systems instead of dark ages stuff

Yes, thankfully we're out of the dark ages, and there's nothing to go wrong any more.

Hang on a minute......................

 

[coytech]

OH DEAR ME

 

[stevethejoiner]

The problem is the pump is pulling air in from the open vent.

The simple answer is that you should not fit any valves to the open vent; you should increase the available pressure at open vent tee by;
a) raising the tank or
b) reducing the pressure losses due to friction along the flow path from tank, to cylinder cold feed to cylinder outlet, etc., and/or
c) reducing the pressure losses due to friction by reducing the flow rate by throttling the pump's discharge (or not turning the shower on full-blast).

Regarding b above, there are pressure losses due to friction along all the pipe and all the fittings that the water flows through. You can reduce the pressure losses by using larger pipe, using pulled bends rather than elbows,using full bore ball valves (rather than cheap service ball valves or gate valves), using copper rather than plastic (larger bore, no inserts at joints), etc.. Any restriction causes pressure losses. You can find tables for these pressure losses at a given flow rate and can calculate the pressure at any point in the pipe.

The pressure at the open vent tee is equal to the static pressure/head MINUS the frictional pressure losses. In your case, the frictional pressure losses are too great and/or the head is too small; the pressure at the open vent tee is less than atmospheric, and air enters the pipework.

You could try running the shower and gradually increasing the flow rate to see at what point air is sucked in. This would give you an idea of how severe the problem is and whether it could be fixed by replacing a few fittings or whether it need a whole rethink.

PS Fitting a non-return valve in the open vent would stop air being sucked in; you'd also get cavitation in the pump which would wreck the plastic impeller.

Very useful reply, thank you.
The reason I fitted the Turner pump was due to poor pressure on the hot flow. The original system was a combi boiler and all run in 15mm pipework supplying one shower and kitchen sink for our farm office and obviously with mains pressure the shower was fine. We are now using it as a temporary dwelling and when the boiler packed up I removed it and fitted the wood stove. I put the header tank as high as possible, but as it is a single storey bungalow type building the tank is not as high as I would like. Full bore lever valves have been used. I ran 22mm cold feed from header to the copper open vented cylinder. I also ran 22mm cooper pipework wherever I could but had to reduce down to the original 15mm copper as I could not get access to the pipes without major renovation, which was not in the plan.

Straight away the shower was crap, if I held the head down onto the shower tray it would be just about passable, but back up on the hook the water just trickled down the shower head handle and the flexi pipe, so it is a head pressure issue combined with too small bore supply pipe. I cannot increase the head so the pump was fitted.

Fitting a Surrey flange has been suggested and I am tempted to do this, but if I am correct, one would usually install a Surrey/Essex flange if you required a separate shower take off so as not to cause water fluctuations on shower flow when another tap is opened, which in my case is not an issue.
But if my problem is due to the 15mm pipework restricting flow rate, would a surrey flange make any difference?

 

[Onetap]

But if my problem is due to the 15mm pipework restricting flow rate, would a surrey flange make any difference?

The Surrey/Essex/Warix flange is fitted to prevent air bubbles being entrained in the water flow to the pump, where it will airlock the pump, reducing the flow. This air is dissolved in the cold water and is released in the cylinder when the water is heated.

You should have a Surrey, or similar, flange but it is there for a different reason; it has nothing to do with the current problem of air being drawn into the open vent.

 

[Dan Robinson]

Don't get pressure and flow muddled.

Fit the flange as suggested and your problems will be gone.

 

[alex444]

A tragic turn in the thread has led to the discussion of why we have an open vent pipe, oh dear oh dear.

A) To keep the system at atmospheric pressure
B) To provide a clear route of expansion for hot water/steam to escape

B is regardless of whether or not there is a cold feed

This is like level 2 c&g stuff for f@cks sake!

 

[Onetap]

Fit the flange as suggested and your problems will be gone.

It MIGHT well do so, but you can't guarantee it. It is in effect, lowering the junction between the pump inlet pipe and the open vent pipe to a point at the base of the Surrey flange.

He still has a problem with excess pipe resistance and the open vent is controlling that problem by admitting air and so reducing the pump's water flow rate.

If he did stop the air getting in, by just fitting a Surrey flange, the flow rate and the pressure losses will increase. He still has his problem, but it is much more likley that it will appear as cavitation in the pump (if it's not already happening). That would cost him a pump. The correct solution, IMHO, is to fit a Surrey, Essex or Warix flange AND re-do the pipework as per the pump manual.

PS The OP doesn't show how the cold feed to the (twin-impeller?) pump is routed; just in case, it should not be taken off the cold feed pipe serving the HW cylinder.

 

[Onetap]

A tragic turn in the thread has led to the discussion of why we have an open vent pipe, oh dear oh dear.

Only for dc.

 

[Dan Robinson]

Fit the flange as suggested and your problems will be gone.

The correct solution, IMHO, is to fit a Surrey, Essex or Warix flange AND re-do the pipework as per the pump manual.

PS The OP doesn't show how the cold feed to the (twin-impeller?) pump is routed; just in case, it should not be taken off the cold feed pipe serving the HW cylinder.

Yep... Essex flanges a re a PITA to install, but they certainly work once in.

 

[alex444]

Thats why I said surrey too, didn't want to open another can of worms

I'm sure the OP said its a single impeller pump, though in the past on a regular system I usually fit another 25 gal cistern providing the cold supply direct to the pump.

Alex