How to downrate a circulator pump.

[chrishutt]

I've set up a secondary return to an unvented cylinder powered by a bronze circulator (Circulating pumps SE50B). The circulator is controlled by a timer and a pipestat fixed to the furthest point of the circuit.

Even though the pump control is set to the lowest power (1 of 3) it's still much more powerful than it needs to be (causing a noticeable noise when running) and I'd like to know if there's a way of downrating it.

I thought about putting a resistor in series (what type and rating?), but I suspect that some of you might know of something a little more sophisticated.

 

[Dan Robinson]

It is possible - Gledhill did it, as have several boiler manufacturers - but would it not cause issues for future maintencance? Surely this is a pipe work problem more than anything.

I'm afraid my electrical knowledge is not going to strech muh further on this one.

 

[Gasguru]

I assume the pump has a permanent split capacitor type motor. AFAIK this type can be controlled (to a limited extent) by voltage however the start torque is very low on these motors and you have to be careful not to allow it to stall (they will overheat or cutout on the thermal overload stat).

Otherwise you could open up the pump and carefully prize off the outer shroud on the impeller. They are normally ultrasonically welded onto the vanes and can easily be free'd.

or remove the impeller and have it machined to a smaller diameter

or talk to circulating pumps....they may have a smaller impeller (they have supplied me in the past many years ago)

or a bypass across the pump

 

[Dan Robinson]

My little grey cells are ticking - I remember a more knowledgable post on another forum a long time ago saying that a "slightly" modified Grundfos would work perfectly well at 80V ac in a Gledhill.

 

[ChrisR]

Just put a smaller capacitor in it. It's probably about 2.5 - 3 microfarad. For a test you could put 2 in series, which would halve the value.

Old pumps go too slow when the cap gets down to about 1uF. Especially Wilo's!

 

[chrishutt]

Thanks for that everyone. ChrisR's smaller capacitor seems to be the answer, although I'm clueless as to how that would work to slow down the pump - anyone like to offer a simple explanation (in plumbing terms if possible - e.g. capacitor is analogous to an accumulator, etc.)? Would a smaller capacitor also reduce the power consumption of the pump?

In general is it worth checking the impedance (?) of old pump capacitors with a view to just replacing the capacitor rather than the whole pump?

 

[Agile]

The way to check a pump capacitor is to measure the capacitance. Most better meters have a capacitance range or can be bought for about £20 in Maplins.

They reduce with age and this is a common cause of failure particularly on the long shaft Wilo which are badged Samson or some similar name.

Because they are no longer available its quicker to replace the capacitor if thats the problem.

The capacitor shifts the phase in a winding to create a rotating magnetic field which produces torque in the rotor. Putting a lower value capacitor will reduce shaft speed but as chris says could give start up problems.

In this application I wonder if there is any scope for part closing the pump valves?

Tony

 

[breezer]

what would be wrong with fitting a "tap" on the flow?

by tap i mean either ball valve or gate valve

or a pressure reducing valve?

 

[chrishutt]

In this application I wonder if there is any scope for part closing the pump valves? Tony

Yes, I could always increase circuit resistance to reduce flow rate, but that might be noisier still and would consume the same energy. I remember that an adjustable resistance valve was pretty standard on circulating pumps 30 years ago, before power rating switches appeared. It's certainly the simplest thing to do.

I originally hoped the return loop would work on gravity circulation, so I installed it in 22mm (same as existing flow loop). However the horizontal runs are much greater than the vertical runs (2.4 metres) and although it was easy to get the gravity circulation started it fizzled out after a while. I had hoped it would keep going very slowly (which is all that is required to keep the far end of the circuit hot). The return circuit has to have a non-return valve and I used a "gravity" swinging door type because I thought this would present the least resistance to flow. All to no avail.

 

[Balenza]

Would one of these not work ?.

http://www.aer.co.uk/acatalog/Variable_Voltage_Speed_Controllers.html

Have never fitted one to a circulator but have one fitted to a fan in my workshop .

 

[ollski]

I don't really get involved in component design but isn't this a scenario the alpha pumps were designed for?

 

[chrishutt]

isn't this a scenario the alpha pumps were designed for?

Not as I understand it. The secondary DHW circuit is a long loop of 22mm pipe with a NRV downstream of the pump on the return to prevent DHW flow back towards the taps from the secondary return connection on the cylinder. Resistance to flow would not vary, which is I think what Alpha pumps respond to?

As for the switch identified by Balenza, nice idea but it would cost as much as the bronze pump did. I was hoping for a cheap and cheerful solution.

Anyway, thanks for the interesting ideas.

 

[Balenza]

The Alpha's are also designed for CH systems and not potable water !.
Might sound too obvious but has all the air been eliminated from the circuit. I fitted one on a 15mm circuit a while back and set the pump on speed 1 and it operated quitely. It was a wilo if I remember correctly though cannot remember the exact model.

 

[Softus]

My little grey cells are ticking - I remember a more knowledgable post on another forum a long time ago saying that a "slightly" modified Grundfos would work perfectly well at 80V ac in a Gledhill.

Gledhill appliances commonly use unmodified UPS 15-50s. I don't know if they get run on as little as 80VAC, but I know for a fact they go as low as 110VAC.

 

[Softus]

Just put a smaller capacitor in it. It's probably about 2.5 - 3 microfarad. For a test you could put 2 in series, which would halve the value.

Er, I thought the capacitor was just to assist the motor in starting up from rest.