Advice needed for hot water feed to new shower pump

[Grundfos1]

the vent pipe is just out of shot and is t'd off the copper hot water feed on the right.

The gunky stuff is some sealant, there is no leak, I just put it there as those compression joints are directly above the element, didn't want a leak, if it were to appear to drip into the housing and short out the 240v.
An unnecessary safety precaution if you like.

 

[swidders]

That brass gate valve needs to be removed - can't have one on the vent pipe here. If you must have one, it should be below the tee to the vent that you describe.

 

[Grundfos1]

OK, will move it , however that's not the cause of the original problem I guess

 

[mogget]

I'd check the feed from the cold cistern to the cylinder, for any blockages or restrictions.

 

[Grundfos1]

Other pics are here

Untitled

• Grundfos1
• 2 Mar 2009
• 1

Untitled

• Grundfos1
• 2 Mar 2009
• 1

 

[seco services]

you haven't got a cylinder stat have you ?
hot water over 65 degrees can cause problems with pumps cavitating.

 

[Grundfos1]

I was wondering that, the water is very hot, I have turned the immersion down to 60 (it was up at 80), but obviously that only applies when its in use. The boiler has no temp setting for the hot water.

Thanks all for your great help... gives me many avenues to attack.

 

[Lower]

Have you checked the levels in the cold water tank when the pump starts to get air in it? Is one tank feeding the other sufficiently quickly?

 

[Grundfos1]

Ok, so far today I have done the following tests to test out some of the suggestions you guys have made:

I turned all water heating off and ran off any stored hot water, so all that was present in the hot water tank was cold water, selected the "hot side" of the mixer and ran the pump..... all worked well for approx 1 min before it did exactly the same cough, splutter and the pump increased in noise. Turning the pump off and waiting for a few mins then starting again, the same occurs. I conclude from this that the water temp is not causing the issue (Thanks Seco Services for the suggestion re the tank thermostat).

After the above test I isolated the main water in to the float valve for the cold water storage and popped up in the loft, the tanks had barely had anything drained out of them.

Re-supplying the tanks with water, I turned the mixer fully to the cold setting and ran the pump, this chugged along happily until I got bored, over 5 mins, the cold water storage tanks happily fed it for the enitre duration. As a point to note the 22mm cold water supply to the pump is slightly higher in the tank than the 22mm supply to the hot water tank, so if any were to run out first it would be the cold, not hot. (Thanks Lower for your suggestion)

This definitely seems to happen after running the pump when the mixer is on fully hot for approx 1 min, or longer if the mixer is set halfway between cold and hot.

As per Moggets suggestion (thank you) is the cold supply to the hot water tank restricted? is this the last possible problem? and how on earth do you check that?

 

[Grundfos1]

Anybody know any good plumbers in Haverhill? on the Suffolk/Essex/Cambrdige border

 

[mogget]

As per Moggets suggestion (thank you) is the cold supply to the hot water tank restricted? is this the last possible problem? and how on earth do you check that?

Things I'd be looking for include low quality valves with narrow waterways, debris in the pipe etc.

Also if the cylinder feed has a long horizontal run directly from the storage cistern, this could cause problems, as there's not much pressure at this point to overcome the pipe friction.

There's another test you can do if you haven't already (too lazy to re-read it all). Start up the pump with the shower on maximum hot, wait until the aeration happens, and try drawing some water from another hot tap - on the next floor down if possible. Leave the hot tap on long enough for any problems to show. What happens?

 

[Grundfos1]

Thank you mogget, When I carried out your test, the hot tap on the floor below had zero flow out of it approx 30 seconds before the pump struggled.

So this has to prove the hot water tank is not being filled up quick enough.

The pictures below show the HWT feed (the 22mm Copper) The stop valve half way down doesn't turn, other than that there is little too it, although there is a strange T piece in the attic going to a dead end.

I have enough 22mm Speedfit left to redo the whole run, although due to tight space I'll join it onto the copper on the horizontal part near the tank.

Just out of interest do I need a bleed nipple in it (as currently fitted) and would the horizontal part going into the tank be better off with a slight drop in it?

Thanks again for everyone's advice.

Untitled

• Grundfos1
• 3 Mar 2009
• 1

Untitled

• Grundfos1
• 3 Mar 2009
• 1

Untitled

• Grundfos1
• 3 Mar 2009
• 1

Untitled

• Grundfos1
• 3 Mar 2009
• 1

 

[mogget]

Thank you mogget, When I carried out your test, the hot tap on the floor below had zero flow out of it approx 30 seconds before the pump struggled.

So this has to prove the hot water tank is not being filled up quick enough.

That's right.

The pictures below show the HWT feed (the 22mm Copper) The stop valve half way down doesn't turn,

I think this gate valve is a likely suspect! It may not be fully open. You may be able to free it, or it may need replacing - possibly with the one you need to take out of the vent pipe Alternatively, a full bore lever valve would do the job.

While the cold feed is empty, it would be convenient to cut and cap off the dead leg, close to the tee.

If the old gate valve was restricted, I'd put it back together and test it at this point.

I have enough 22mm Speedfit left to redo the whole run,

Unless parts of it are in 15mm, it would probably best to upgrade the feed to 28mm, if it still needs doing once the gate valve is on.

Just out of interest do I need a bleed nipple in it (as currently fitted) and would the horizontal part going into the tank be better off with a slight drop in it?

Yes, the drain point is there in case you need to empty the cylinder. It won't matter that the pipe is level, as the water is flowing through under pressure rather than just 'running downhill'.

 

[jas05]

Ok, so far today I have done the following tests to test out some of the suggestions you guys have made:

I turned all water heating off and ran off any stored hot water, so all that was present in the hot water tank was cold water, selected the "hot side" of the mixer and ran the pump..... all worked well for approx 1 min before it did exactly the same cough, splutter and the pump increased in noise. Turning the pump off and waiting for a few mins then starting again, the same occurs. I conclude from this that the water temp is not causing the issue (Thanks Seco Services for the suggestion re the tank thermostat).

After the above test I isolated the main water in to the float valve for the cold water storage and popped up in the loft, the tanks had barely had anything drained out of them.

Re-supplying the tanks with water, I turned the mixer fully to the cold setting and ran the pump, this chugged along happily until I got bored, over 5 mins, the cold water storage tanks happily fed it for the enitre duration. As a point to note the 22mm cold water supply to the pump is slightly higher in the tank than the 22mm supply to the hot water tank, so if any were to run out first it would be the cold, not hot. (Thanks Lower for your suggestion)

This definitely seems to happen after running the pump when the mixer is on fully hot for approx 1 min, or longer if the mixer is set halfway between cold and hot.

As per Moggets suggestion (thank you) is the cold supply to the hot water tank restricted? is this the last possible problem? and how on earth do you check that?

To prevent scalding the cold feed to the cylinder should be the higher of
the two connections so the hot water runs out first

 

[Grundfos1]

OK.... have replaced the tank feed to the hot water cylinder, to just above the drain point in the original photo. still have exactly the same problem!!

Any other ideas?

Untitled

• Grundfos1
• 3 Mar 2009
• 1