Optimization of parallel variable-speed-driven centrifugal pumps operation

[rudiawan .]

dear all..
i am designing the Sea Water Intake pump which the 4 pumps will be operated parallel, and 1 pump will be as stand by pump
The pump data as follows :
Type : vertical turbine pump
Flow : 5000 m3/hr
Head : 20 meters
Electric Motor : 2500 KW, the motor provided with VSD (variable speed drive)
The operation pump speed at normal condition is 740 rpm

I am developing to prepare the operation procedure for the pump operation, may i get the advise how to ramp up the pump speed parallel and safely, from 0 rpm to operation speed (740 rpm)
your advise will be appreciated



regards
rudiawan

 

[Artisi]

Having been involved with pumps of this size, my suggestion is to employ a qualified pump engineer experienced in the start-up and operation of multiple pumps of this flow, head and power. You are not talking about a couple of low power process pumps, this is major equipment that requires a careful study, start-up and operation plan.

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.)

 

[LittleInch]

What's the system resistance? Or system curve?

Is this just lifting seaware into a pond or into a big pipe?

Does each pump effectively work individually or do they join together into a header?

What's the static head?

You need to understand what happens to the back pressure as the flow rises.

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

 

[rudiawan .]

hi littleinch ..
The 4 pumps supply sea water into the pipeline which the diameter is 60 inch, into the sea water tank
the length of the pipeline about 6 km
the pipeline provided with surge control valve to maintain surge pressure due to back pressure

regards
rudiawan

 

[LittleInch]

Post a P&ID or PFD or sketch including any control valves on your pumps and also what is the static lift between sea level and the water level in the "tank".

I would do this one pump at a time, but you need to know individual pump flow rate.

So start one pump, but put the VFD control on pump max flow of 5,000, let the flow stabilise then start pump 2 and do the same thing. Pump 1 will then increase in speed in order to maintain flow. When both pumps are running at more or less the same flow and speed then start the third.

You do need to know individual pump flow rate although total flow rate might work as well and just divide the flow by the number of pumps you have running to feed back into the VFD controllers.

You are trying to avoid going off the end of the curve which you could if you just set a speed.

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

 

[TenPenny]

Remember that when one pump is running, the second pump to kick on will pump zero until the discharge pressure meets that of the first pump.

 

[EdStainless]

Are your VDFs set up as master/slave?
How many flow or pressure sensors do you have?
Can you measure the true power of each motor?
Are there critical frequencies which you must avoid?
I presume that each pump has a check valve?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[georgeverghese]

What is the master controller here that resets the pump speed ? Is it discharge pressure ?
If master is PIC, it would seem logical to auto ramp up pressure at the PIC from zero to control setpoint pressure over say 2-3minutes ?
Startup each pump separately at leave at min speed
Startup other pumps to min speed also
Then auto ramp on P_discharge pressure
You can program all this as a single auto startup sequence

But how do you prime the pump suction line ? Does each pump have a built in priming pump ? If it auto primes, then it is taken care of. Else manual priming at field will be required.

 

[rudiawan .]

The pumps is vertical turbnine and the suction us submerged, hence no issue regarding priming
individual flow meter not available, but speed of VFD are controllabe by PLC / DCS
(see on the attachment)

 

[EdStainless]

Given the variation in both the pumps and the motors we used power to match the pumps just speed.
Yes we had tolerance bands on both speed and power.
Bring pumps up to speed one at a time, we would often use a 'typical min speed' that was maybe 80% of full speed.
Then bring each additional pump up to speed and allow to stabilize, making sure that outlet pressures match.
Once they are all running you can lock them to the master and adjust them as a group.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed