Plunger Pump and Multistage Centrifugal Pump opperating in Parallel 1

[curtis26]

I just learned of a pump opperator that wants to run a multistage centrifugal pump in parallel with a possitive displacement plunger pump.

Has anyone seen or heard of this being done before? If so what would be the justification from an operational standpoint for doing this. What measures have to be taken in sizing the pumps and design the system's discharge piping, etc. to make this work?

 

[BigInch]

I could envision the PD running as an injection pump. It would tend to match the CP's head, provided the power was there to lift whatever the PD's flowrate was at the time. Too much volume from the PD might tend to take over control of the discharge line and actually increase its discharge pressure, at which point the CP would slow down its flow and back up on its curve until it reached an equal head, provided that shutoff head was still greater than the PD's discharge head at the time.

Put a check on the CP discharge and run the PD's speed on flow control. You might need a recycle line off the CP's discharge (takeoff before its check valve), in case the PD started forcing the CP's flow below minimum. A power limit, or a pressure relief on the PD itself, or downstream to be sure it can't overpressure the discharge.

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso

 

[micalbrch]

You do not know the specific application, don't you? I can assume that both pumps run in parallel up to a certain discharge pressure (which means that pressure changes over the time), that the centrifugal pump is shut off at a certain point and the PD pump continues at higher pressure. A typical application where centrifugals and PD pumps can run in parallel is filter press feed. The pressure to feed the press increases while the flow is reduced. In WWTPs and mining there is pretty often a high volume fill pump (centrifugal pump) which is turned off at a certain pressure. The PD pump runs in parallel and continues to finish the filtration cycle when the centrifugal pump has been stopped. But I must confess that I have never seen a plunger pump for this application.

 

[zdas04]

The only time I've ever seen this tried in real life used a duplex pump (quite a bit of macro pulsation) and there was a check valve on the CP discharge. The check valve slammed several times a minute and lasted less than a year before the flapper came apart.

I won't do it except in applications like micalbrch mentioned where they are sequential and don't run in parallel very often or for very long.

I've seen a lot of people try to run centrifugal compressors in parallel with recips and I've never seen that work very well. Control of the two machines is difficult and the first time the surge override on the centrifugal backs it off the recip takes over and the dynamic machine coasts from that point forward.

David

 

[bimr]

The Hydraulic Institiute recommends against using pumps with unequal shut off heads in a parallel application.

This is a general guideline, but where the static elevation is the primary system component, they may operate successfully.

The pumps must be started in proper sequence and the system conditions must allow both pumps to operate at or near BEP.

 

[BigInch]

zdas, That's what I meant by "Too much volume from the PD might tend to take over control of the discharge line". The PD must not put enough flow into the discharge line to raise the pressure enough to close the CP's check. I would do it only using the PD as an injection pump for a relatively small amount of flow. This is done successfully, in injection of flow enhancers, corrosion inhibitors and in the reinjection of diesel/gasoline interfaces at small controlled flowrates, so as not to go over contamination limits, into a mainstream gasoline stream feed by a large centrifugal.

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso

 

[zdas04]

I've seen that too, but I didn't think of it as "parallel" more as "injection".

David

 

[BigInch]

Yes a good point. Parallel would imply more or less an equal flow and head... a bad idea as that would almost always imply a high pressure transient each time the PD forced flow into the discharge.

Only put off until tomorrow what you are willing to die having left undone. - Pablo Picasso

 

[curtis26]

Thanks guys! All comments were greatly appriciated.