How to prime a booster pump? 1

[gte682n]

Hello, I have a project that will be installing a set of centrifugal booster pumps in the middle of a feed train. These pumps will take suction from the first set of pumps after the fluid goes through several heat exchangers. My question pertains to the start up sequence of the new booster pumps. I currently foresee starting up the upstream pumps and establishing flow through the first half of the system. What needs to happen when I get to the booster pumps?

Is it advisable to leave the suction and discharge valves open to allow liquid to flow through it before starting the motors?

How else can I ensure that I have a sufficient liquid source to the booster pumps to avoid running them dry?

Thank you,

 

[LittleInch]

It's not normally a good idea to flow through apump when it isn't being powered.

Normal procedure is to start the pump against a closed outlet valve with the inlet valve open.

How you control start-up pressure and flow is down to you and your system to gradually introduce the higher pressure at such a speed that the rest of the system can adjust accordingly until you reach steady state.

It is far from unknown for no one to properly think this through and hence when you start a booster pump with a high inlet pressure, the pump trips either on high outlet pressure, excess flow / amps or low inlet pressure /flow. It often needs some sort of control valve or similar downstream the booster pump to prevent this. A recirculation valve is also sometimes used.

You provide virtually zero data so it's not possible to see what levels of protection this pump has or what is possible / going to happen when you start the pump.

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

 

[gte682n]

I have attached a schematic of the new set up. I should note that both the existing pumps and the new booster pumps will be on variable speed drives. The first pump will be on discharge pressure control, and the booster pumps will be on flow control set by the motor speed. I am thinking that we start he first pump up with the discharge of the booster pump closed. Once we have filled the system up to the booster pump with liquid, then we open the booster pump discharge valve and start the booster pump with a flow setpoint.

A PHA has yet to be completed, so pump protections have not yet been defined.

Thanks for the comments so far.

 

[LittleInch]

If this system is designed to operate only with both pumps in operation then that's Ok, but you call it a "booster" pump. This implies that the system can operate with just the first pump operating?

Then you usually introduce a no return valve around the booster pump to let flow through then simply start the pump when flow is developed with a set point higher than the flow established by the upstream pump

With your plan I would start the booster pump first against a closed valve then open the discharge valve.

I would fill the system first before turning the pumps on. Where is the vapour going to go?

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

 

[gte682n]

The system currently operates with only the first pumps. However, the pressure profile on the first half of the system can exceed the pressure ratings of some of the equipment in the line when exchangers get fouled. The new booster pumps will allow us to lower the upstream pressure and maintain our desired flow. The displaced vapor would push through the system via a bypass line around the booster pumps into an open distillation tower. Once the system is full of liquid at that point, we would start the booster pumps as you mentioned above.

I am curious about your statement to set the booster pump flow setpoint "higher than the flow established by the upstream pump"? Wouldn't this create a flow imbalance? This is why we plan to maintain a discharge pressure setpoint on the first pump, and flow setpoint on the second pump. As the booster pump calls for more flow, the upstream pressure will drop until the first pump catches up and reaches its previous pressure target.

Thanks,

 

[LittleInch]

My point is that unless you set the flow requirement of the booster pump more than what the first pump is doing, then it won't start. It needs to try and achieve some higher flow which then draws the pressure down in the upstream section and forces the flow through the pump and then shuts the non return valve.

Doing it slowly with a VFD is the best way to avoid shocks and trips. You could always stick in a manual control into a low selector block and simply manually or automatically ramp the booster pump up which the flow controller then automatically takes over as the flow gets up to your set point.

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

 

[lilliput1]

Put automatic air vent on high point of line. If vapor need to be collected run vent line to distillation tower. Start booster pump and open its discharge valve when desired pressure at booster pump suction is reached. Control booster pump VFD to maintain set flow.

 

[lilliput1]

Also when suction pressure drops 5 psig (adjustable) from pump start pressure set point shut off booster pump and close its discharge valve in sequence. Provide check valve at pump discharge and manual valve to be always open except for maintenance at suction of booster pump.

 

[georgeverghese]

There should be a single min flow recycle line that serves both sets of pumps - not shown on this scheme. Agree with your process control scheme to run the upstream pumps on min discharge pressure. Setting overall demand purely on FIC on the second set may be okay if there is no need for LIC override from V1002. With an overall min flow recycle loop with FCV wide open on startup, you should be able to start both pumps simultaneously on auto sequence on manual control at min speed, and then slowly load up on auto through the FIC. Agree this high point at 65ft looks troublesome and should be autovented each time on startup - what is T1 at the high point?

 

[gte682n]

georgeverghese,

Just to confirm my understanding, would the min flow recycle take off from the discharge of the booster pump and go back to the suction of the upstream pumps? Also, how would the pumps be able to start simultaneously, as there would be some lag time to fill the system up to the booster pump?

The level control for V-1002 is accomplished by a control valve on its inlet. This stream is fed from upstream storage tank pumps.

T1 is a set of internal condensing coils at the top of a tower. We pre-heat our feed in this pump system. At T1 we are generating internal reflux with this heat recovery.

Thanks,

 

[lilliput1]

VFD can turn down pumps down to 20% rpm. Below this is when you need to recirculate or shut off the pumps. Consider also putting in a low level switch in V-1002 to shut off the pumps and another level switch set at higher level to turn on the main pump. The booster pump setup would be as I had described above.

 

[georgeverghese]

The overall min flow recycle line should be taken off the booster pump discharge and return to V1002. Its FIC setpoint would be the higher of the min flow permissible for the 2 sets of pumps when they are both at min speed.

If the vapor pressure of the fluid is less the pressure at the 65ft elevation point, then there would be no risk of vapor breakout on the suction line to the boosters on startup, and it should be acceptable (but not ideal) to start both pumps simultaneously, since pressure and temp to the boosters should normalise in a short while.

Alternatively, you could delay the booster pump startup and run an automated bypass across the boosters on a timer to allow pressure and temperature to stabilise somewhat, and then close off this bypass/ autostart the boosters at min speed.

Whether you start the boosters simultaneously or delay it ( by using this startup booster bypass) will depend on the volatility at the 65ft point / and or viscosity of this fluid without much preheat.

Though this is not the topic of discussion here, also check that the shaft seal cooling system for each of these pump sets will work okay when the 2 pump sets are at min speed, especially if you are relying on an internal pumping ring to keep seal fluid circulation going - often overlooked.