Hot water pump cavitting

[GeoffJ]

Hi

I installed a brass pump to circulate hot water from the top of the HW tank to the bottom 18 month ago. This is part of a SolarPV system. The surplus electricity is directed to the hot water heating element. When the top of the tank is hot and the element thermostat kicks-in a relay activates a circulating pump so that hot water is pushed to the bottom of the tank. In this way the whole HW tank can be brought up to thermostat temp.

A picture of the pump has been loaded

At first there was cavitation and advice from this site made it clear that a surrey flange is needed on the tank. This was installed and it all appeared to work well for the first summer (or I failed to test it properly). Certainly I can hear bubbles in the tank and the pump keeps working (whereas previously the pump would run and then cavitate after 30 seconds or so)

Now when the pump starts it will normally cavitate - bleeding the air gets on the pump gets it working quite easily and it will work find until it then needs to next restart. Being linked to the availability of surplus power the pump can stop and restart quite often.

Any ideas on how to stop it cavitating?

My thoughts are
a) Relocate the pump so that it is on a vertical piece of pipe. So small amounts of air can collect above the pump and the pump can remain primed
b) Obtain a self priming brass pump (if there is such a beast)

Many thanks
Geoff

 

[Madrab]

A surrey flange will remove most of the air bubbles but not all. Usually a pump is located at the bottom of a vertical run, therefore any bubbles that may collect when pump not in use will collect at the top of the pipe therefore the pump is still primed before it starts.

Your setup has the pump at the highest point of the loop therefore any bubbles that may find their way in will sit exactly where your pump is. Your idea of placing it on a vertical run would be a good place to start IMO.

 

[GeoffJ]

Hi Madrab
Thank you for the guidance.
I put the pump at the top of the pipe work because it meant that I could get the air out of the pipework using the air bleeding screw on the pump. I guess that was not such a smart idea?

If the pump is moved to a vertical pipe (on the UP side) - do you think I would also need to insert vent at the top of the pipework? Could I install some sort automatic air bleed vent?
Thanks
Geoff

 

[Madrab]

I would lend more towards placing it into the down leg of the pipework, the pipe to the right of the pump. Yes you could tee off and put a short length of pipe in and then pop in a bleed valve, tho I'd try it without first see how it performs.

The bleed screw on the pump is really only for initial bleeding and shldn't really used as a regular bleed point.

 

[GeoffJ]

Hi
Are you sure you would put the pump on the down-side?
My thoughts are
1) I am getting some air in the pipework (to be expected)
2) It will build up at the highest point in the loop (just behind the pump)
3) Once the air pocket reaches the pump it won't prime
Regards
Geoff

 

[Nozzle]

You want your pump low down, as it adds static head. Take some hot water, drop the pressure of it (in the eye of the pump) and what you might end up with, if close enough to saturation temperature, is flashing off which will knacker your pump in no time at all. (This has nothing to do with trapped air)

Nozzle

 

[Madrab]

As nozzle suggest though from a trapped air position, if pump is on the down side of the pipe low down, you will have air free water and any head will help in the startup of the pump. Once the pump is running any air subsequently drawn through should clear itself.

At the moment you'll have air collecting at the pump before it starts, therefore no chance of the pump running and therefore needs to be bled.

That's why I mentioned to try moving the pump first see if it deals with any air it catches after startup and keeps running. If still having issues then it's easy enough to add a vertical leg for venting.

 

[GeoffJ]

Thanks Madrab an Nozzle

One final question before a reach for a wrench!! There is a non-return valve in the circuit. Any guidance for where that should be placed in relation to the pump and the tank connections?

 

[Nozzle]

If it's a flap type, it should be positioned as per the manuf. manual. If it's a ball in a cage type, I don't think it matters

Nozzle

 

[GeoffJ]

Hi,
Pleased to report that with the pump relocated to near the floor the pump now runs nicely with no air.

(It took longer than I thought to test this as I needed some good sunshine.)

Thanks for your guidance.
Regards
Geoff