ZigZag Centrifugal Pumps not holding prime. 1

[TonyVov]

Looking for some ideas in troubleshooting and fixing a natural pool system.
I have a system with 2 centrifugal pumps sucking from a pond and pushing water 30m uphill to circulate back down through various ponds and gardens.
Pumps are 1.5kw and 2kw multistage at 4-6 M3/hr
To help pump life pumps are set up zig zag one running at a time. Single suction pipe 2inch with one fooot check valve and outlet check valve on each pump..
Uphill pipe 3inch.
The problem i have is the system is often losing prime (only 1.5m suction head) and the pumps run dry without warning, resulting in rapid damage to pump systems.
I suspect single suction line input to 2 pumps is higher risk as the 'off' pump sees the inlet suction up to outlet check valve. So we will modify input to 2 pipes and 2 foot valves.

I also suspect that the on off zig zag pump operation results in water hammer and higher starting currents. We plan to add 'delay off' relay to zig zag system to allow pump one to keep running while pump two is starting.

My problem and question is to have the unmanned system to shutdown if the pumps run dry. Our observation is that the pressure can build to 30m 3 bar level but still have no flow and run dry so a pressure switch does not work
I am looking for some way to identify if there is flow or not and in the absence of flow (for example after 2-3 minutes operating) to shut the system down for operator inspection.
I just cannot seem to find any simple device to detect if a centrifugal pump is actually flowing our not.

 

[hydtools]

Sense motor current or water flow instead of pressure.

Ted

 

[TonyVov]

Ted thanks for reply, our controls at the moment are simply timers and contactors. I have been unable to find a simple sensor that detects if flow ia present or not. This is why I came to this forum with the question. I am looking currently in the catalogs of Schneider electrical and Grundfos Pumps.

 

[FMJalink]

Hi TonyVov,

If I understand the situation correctly, you have on the suction a single suction line, wbich if for both pumps. The single suction line is with a foot valve. On the discharge each pump has a non-return valve.
When you stop a pump, the water in the part down stream the pump up to the non-return valve will fall back down to the suction. Pending on the height/length of the suction leg this can be considerble. Specially when the other pump is sucking from the suction side with a combined suction line.

You could increase the volume between pump and non-return valve. Then after stopping a pump line will stay longer up to above the pump filled with water.

The use of separate suction lines will most likely be beneficial. This depends on suction pressure of the running pump. When lower then in discharge of the idle pump, the running pump will reduce the water quantity in the discharge line part of the idle pump.

Do not whether you have also problems with air entrainment at the foot valve. Is so you could consider a small return line/opening from the additional volume at the discharge to the pump suction. With proper sizing this makes your pump self priming, though wit a reduction of the net through flow,
If required a level detection could be placed in the pump discharge line. This should be abetter and easier method than a flow detection.

Hope this helps with the challenge,
FMJalink

 

[LittleInch]

A diagram always helps a lot, but I think you need first to figure out why the pumps are not able to switchover.

Does the sump ever run dry? or start to vortex down?

What is the minimum submergence of the pipe going into the sump?

How do you maintain water level? some will evaporate.

Maybe the standby pump will collect air in its system?

Check valves are pretty notorious for failing to seal, especially small cheap ones. Use actuated valves or solenoid valves on the pump outlet.

What is your change over period and why? Why not run them for a month at a time and then just be there to change over?

Simple flow switches are available, but you need a timer on them to avoid spurious triping.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[3DDave]

Separate suction lines should help.

If they are plumbed together then the running pump is pulling a vacuum on the inlet to the non-running pump maximizing the chances it will lose prime. Likewise, alternating them increases the time for the prime to be lost.

Just running one pump with the other as a ready-to-go backup is the better choice. When one pump fails, switch over, and then you have the most time to replace the failed pump. As it is, with equalized operation there is equalized wear and they could equally fail at or nearly the same time.

 

[Rputvin]

Always curious to see smaller suction lines on a pump vs. the discharge. Unless the discharge pipe is oversized to account for a long run, you usually expect the opposite when trying to minimize suction losses.

Add your switch after the collected discharge and check valves. Pressure switch set somewhat below standard operating pressure, but above 0 will do the trick, as well as a paddle flow switch. Either device can be <$50 itself.

As LI mentioned, a timer in the neighborhood of 5 seconds will prevent errant trips of the fault circuit - you need to allow the pumps to spool when the circuit is energized, and account for anything like a bubble passing that could interrupt flow/pressure for a moment.

Not sure what a "zig zag" pattern is. Agreed that more details might present a recognizable issue that could be solved and prevent the need for controls intervention (although a fault circuit to energize the backup is still a good plan if the process is worth investing in some extra controls for a better guarantee of continuous service)

 

[Artisi]

First step, why is it losing time?
Faulty foot valve?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)