Multiple pumps - start up against closed butterfly control valves 2

[JohnWeal]

Good Morning,

Are there any recommended opening / closing procedures for bringing a pump on line or closing one down by utilising the pump control calve located in the discharge pipe for each pump prior to joining the manifold?

We have 9 pumps, 7 of which are on line at any one time (2 dedicated standby units)

The pumps are all FIXED SPEED submersible canister type. (motor is with pump and not above floor level - the whole pump is lowered to the bottom of the 1100mm canister)

Diameter is 800mm and the manifold is 1400mm.

Fluid: Raw water

Would like to know if there are recommendations for the start-up procedures and any advice for correct installation considerations would be very much appreciated. Cavitation, surge, duration of closed head etc.

Best Regards
John

 

[JJPellin]

Our standard procedure for starting up a single stage centrifugal pump would call for pinching on the discharge block valve for start-up and then opening the valve once the pump was up to full speed and pressure. This is based on the normal use of a gate valve as the discharge block valve. And, this procedure is generic to cover a wide range of situations. If there are multiple pumps discharging into a common header and if each pump has a check valve in the line, then pinching is not necessary in most cases. The check valve will remain closed until the pump gets up to speed and builds enough pressure to overcome the pressure on the other side of the check valve. I would also tend to be more concerned with pinching on a butterfly valve as comparted to a gate valve.

Unless there is some process reason why the pumps need to be started up against a pinched or closed discharge valve, I would start them up with the valves open in this situation. Possible process drivers for pinching on the valve could include the need to have all changes in flow and pressure occur gradually. Some processes are prone to upset or unit trip if the flow or pressure changes quickly. If that is not the case, I don’t see a need to close a valve for start-up.

You described a situation where each pump has an individual control valve of a butterfly configuration. I assume that each pump also has an individual check valve (non-return valve). Please confirm that is the case.


Johnny Pellin

 

[JohnWeal]

Many thank JJ.....

YEs we have NRVs in the discharge line. Agree the 'back pressure' on the downstream side of the NRV is at the common manifold pressure agree, unless the flow / pressure is equivalent, ie reaches the duty point then the NRV doesn't open. I guess it is more of a problem say on an individual well pump if it is an independent system and the rising main may have a low pressure or indeed be empty. But in that case we have small bypass valves fitted to the mainline valves to control pipe filling rate.

Regards
John

 

[EdStainless]

And with multiple pumps on a manifold it is good to have voltage sensing so that you can tell if a pump/motor is rotating backward.
A leak in the check/control valve could be spilling flow back through a pump. If it is turning fast it may cause damage if started.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[QualityTime]

I assume you have a check valve and a motorized butterfly valve on the discharge of the pump. Make sure the butterfly motorized valve operator has enough internal limit switches (fully closed, 5%, 7%, fully open). We usually ask for 6 internal limit switches. Make sure you get a very slow moving valve operator that takes, at least, around a minimum 45 second – 60 second close to open time. As you can see the motorized valve limit switches control the pump. The pump is never directly wired to energize and de-energize itself.

1. Starting a Pump

[ol a]
[li]The motorized valve moves from fully closed to the fully open position[/li]
[li]When the valve is 5% open position (approx.) the valve limit switch is interlocked to the pump motor circuit. The pump motor energizes. Basically the pump energizes against an almost closed valve[/li]
[li]The valve continues to the fully open position.[/li]
[li]This procedure allows a smooth introduction of flow into the system without the sudden undesirable system pressure bump[/li]
[/ol]


2. Stopping a Pump

[ol a]
[li]The motorized valve moves from fully open to the fully closed position direction[/li]
[li]When the valve is 7% open position (approx.) the valve limit switch is interlocked to the pump motor circuit. The pump motor de-energizes. Basically the pump de-energizes against an almost closed valve[/li]
[li]The valve continues to the fully closed position.[/li]
[li]This procedure allows a smooth lowering of flow into the system to prevent sudden surging that occurs when a pump is pumping full blast into the system and then the pump is suddenly de-energized.[/li]
[/ol]


3. What Happens in the middle of a Power Failure?

[ol a]
[li]The pump motor will de-energize. It will not restart up by itself when power is resumed. This will prevent the pump creating a surge/pressure bump into the system[/li]
[li]The motorized valve must be moved to the closed position (manually).[/li]
[li]Restart the pump as per (1) above[/li]
[/ol]

 

[JohnWeal]

Some great answers thankyou.
QualityTime, for 3b, why is the valve manually closed? Is it not a case of witing for the power to be resumed and the valve closes prior to pump energizing? Or is it to ensure no backflow? (Even with an NRV before it). Ie no pump can start unless its control valve is fully closed.
And with sensors to confirm no backward rotation.

 

[QualityTime]

It has been a long time for me but, but as you said, you may be able to initiate the valve to close remotely.

I am not sure how a voltage sensor will work if the motor is turbining due to backwards flow. As long as you have a check valve I would not worry about water running backwards through the pump. I have done a lot of big pumps and I think this concept might be a bit overboard. Remember that for water to flow backwards BOTH the motorized valve and the check valve must be having problems. This is highly unlikely. Even if does happen, remember there is always maintenance program that has to be done on pumps. An operator can easily determine from sound if there is a leak. That is what an operator is for. Time to fix the problem.

Motor contacts or a current sensor is all you need to determine if a pump is running or not.