13 replies to this topic

[vinod]

Sir

I have across a situtaion whereby i need to size and specify a positive displacement pump as a booster to a centrifugal pump(in the downstream of the centrifugal).

I have the following doubts:
1)What type of control system should i give to maintain a constant flow from the centrifugal pump?
2)If somehow my centrifugal pumps less flowrate than the rated capacity of the PD pump,what is the consequence?Will it happen?What can i do to avoid this?
3)If the discharge of the PD pump is closed,should i trip the centrifugal pump?


Thank you
Regards

Vinod

[JoeWong]

Do you mind to advise us what service is this arrangement for ? Why a positive displacement pump as a booster to a centrifugal pump ?

Just wants to see if the combination is good before going into detail.




JoeWong

[vinod]

Hi Sir


The servive fluid is a polyol...........viscoity is almost 400 cp........................

[pleckner]

OK, this is a new one for me, putting a PD pump in series with a centrifugal. This is a recipe for disaster and I can't see why it would be necessary. Pull the centrifugal out and put in a proper PD pump.

You will need to give a lot more information/details before any of us can give you any better guidance.

[Padmakar Katre]

Dear Vinod,
Let me clear you first the purpose of the booster pump is to provide the NPSH required/incresed suction pressure for the second pump.I have never seen this kind of combination rather you go for the two centrifugal pumps one as booster pump with low TDH and the second one as your main pump.
The way you asked this question its more of confusing to the readers.The three questions/doubts you have asked at the end of your query are really very hard to understand to me personally.

In your third query you ahve asked about closing the PD pump discharge what would be the consequence on the Centrifugal pump , forget about this consequence you will damage your PD pump by closing its discharge.

If you can present a schematic drawing of the same with some information like normal flows,pressures at different points (suction pressures,discharge pressures of the pumps), service and the temperature may we can give you some ideas.

Waiting for your reply.

[vinod]

Sirs

My client wants the existing centrifugal pump rto be there ,and since he is planning to pump to a new destination,the fluid being very viscous is not reaching the destination,so he wants to boost it with a PD pump.

While i close the discharge of the PD pump ,i will put a recylce line with a PCV to ensure that the PD pump and the line blocked is not damaged.

The flow required is about 50 m3/hr and the head required is almost 200 m.

Please advise on the control schemes required.

Thank you

Vinod

[JoeWong]

Sirs

My client wants the existing centrifugal pump rto be there ,and since he is planning to pump to a new destination,the fluid being very viscous is not reaching the destination,so he wants to boost it with a PD pump.

While i close the discharge of the PD pump ,i will put a recylce line with a PCV to ensure that the PD pump and the line blocked is not damaged.

The flow required is about 50 m3/hr and the head required is almost 200 m.

Please advise on the control schemes required.

Thank you

Vinod

What do you think if you put PD pump at centrifugal pump discharge ?

JoeWong

[vinod]

Sir

I am junior engineer and this is the first i am coming across such a case.

As far as i know,i will have to ensure a constant flow from the centrifugal pump to the PD pump(as it has a single rated discahrge asssuming it to be running at a single speed).

And that i have to maintain an NPSH for the PD pump(including the acceleration head).

Also that i need to provide a safety valve in the discharge of the PD pump,and arecycle line with PCV connecting back to the suction tank.

Please help me out

Thank you

Vinod

[pleckner]

Go back to the client and tell them that this is a recipe for disaster!

One possible way to get this to work would be to revise the piping so the centrifugal pump goes into a circulation loop and the feed to the PD pump is taken from that loop. You would/should recirculate th flow of the centrifugal pump back to the source. I'm not saying this is practical for your installation because I don't see your details but this is one way to simplify the control.

[vinod]

With due to respect......let me tell that still i dont understand why this is disastrous?

Thank you

Regards

Vinod

[pleckner]

A PD pump is a constant volume device and its curve is almost vertical. It will continue to push the constant volume of fluid no matter what the downstream looks like and will build up tremendous amount of pressure to accomplish its goal, even if it gets blocked in or the flow is impeded for any reason, i.e. block valve, centrifugal pump. The PD pump is also more or less a pulsating device, it does not produce a smooth, constant flow. You will get variations in the flow and line pressure during this pulsating motion.

The best way to control the flow is via speed or stroke length (for those that operate via some piston type of capability). You can use a flow meter and control valve but the design must be more carefully thought out and postion of these devices is more important than they are for centrifugal pumps.

A centrifugal pump is NOT a constant volume device but will change flow with line/sytem pressure changes. Therefore, it is not a good idea to mix the two. Your safety issues are more complicated as will your ability to maintain a good control point to operate at.

[djack77494]

Interesting topic. Though I think this arrangement is VERY unusual and note that it has negative "features", still it may be workable. I'm reading that the centrifugal pump will supply the majority of the hydraulic work, but that it has a high NPSH requirement. The system you propose will provide a single (average) volumetric flowrate, since that is the characteristic of PD pumps. To feed the centrifugal, you MUST remove or dampen the fluctuations associated with most types of PD pump operation. Thus you essentially make the average flowrate = instantaneous flowrate. The centrifugal will always have a given flowrate, and so will operate at a single point on its curve. Upstream and downstream pressure will need to adjust to these requirements, which doesn't seem to be a difficult problem. One other warning is that throughout the range of these adjustments, the centrifugal's suction pressure must never drop below the required NPSH at that flowrate.
HTH,
Doug

[JoeWong]

Phil,
Good explanation...

Doug,
Looks unusual arrangement but may work if properly engineered.
Agreed that providing a dampener at recip. pump discharge may help to dampen the fluctuation but not sure the dampener performance with centrifugal pump downstream...i guess it still works.

Vinod,
Personally my first concern is the NPSHr of the recip. pump. There is acceleration head associate with recip. pump which significant increases the NPSHr for recip. pump. From NPSH perspective, i would think to locate the centrifugal pump first, follow by recip. pump.

Secondly, the centrifugal pump is existing. I presume it works well with existing condition without NPSH issue. Now with new condition (pump to new destination), i presume the head is insufficient. I would think we should locate the recip. pump downstream of existing centrifugal pump. This will not affect the existing arrangement and yet the centrifugal pump will provide necessary NPSHr for the new recip. pump. More over, recip. pump can provide a constant flow to "new destination.".

JoeWong

[vinod]

Thank you sirs for all your help.......................