Some help here...

[car86]

Greetings...

In our company, we´re facing a curious trouble with the heavy naphta pumping system (2 paralell centrifugal pumps 1 on duty and the another one on spare). From 6 months ago ´til 1 week ago, our pump "A" has been working after an overhaul (new babbit in bearing housing, the mainshaft and the impeller were repaired, brand new mechanical seal system were installed). The pump started to work giving 140KBPD @ 135 Psi (motor current consumption 56 amp/58amp nominal current). The pump "B" were put on spare since pump "A" began working, pump "B" has several damages, impeller excessive wear out, volute wear out as a result of agressive corrosion, causing high vibrations values and low efficiency, ´til now we are waiting for the purchase of a new pump, however we can use it for a short time if the plant need it. So, a week ago our plant were shutted down in order to make some minor repairs, course, pump "A" were shutted down and pump "B" still on spare. When pump "A" were started again, the new values were 120KBPD @ 130psi (motor current consumption 57 amp), insufficient flow of heavy naptha for the process and the consumption with a flow valve with a 19% opening (if flow valve is opened more than 19% the motor consumption rises over the nominal motor ampere). Our team decides doing an experiment which consist on working Pump A and B in parallel in order to achieve the required heavy naphta flow for the process trying to open the flow valve slowly from 19% open ´til the motors current consumption rises the top. The result of this was 176KBPD @ 120psi with 32% flow valve open. Once the pump "B" were stopped, obviously, the flow thru the flow valve decreased, but right the same pump "A" gives a new flow (165KBPD @120psi) with 28% fv open and the motor current consumption keeps under desing value.
May somebody give us some lights about this situation?
I can provide you more information if is needed... Tnks..

 

[LittleInch]

Might be a bit obvious, but what was the inlet pressure, temperature, density and viscosity for all these different events?

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[MikeHalloran]

It appears that you're talking about two different system curves as seen at the pump. It can't climb to the desired flow from a cold start with the 'flow valve' set at 19 pct, but it's working against a very different system curve when the flow valve is at 28 pct. The question is, can pump A get to that second operating point from a cold start without the help of pump B?

What is the flow valve supposed to do, where is it in the system, why is it there at all, and why are you twiddling it?

At this point, I'd get a really big blowup of the entire set of curves for pump A, with the current or power curves, and the same curves for a worn pump (pump B), and start sketching in the system curves as modified by the flow valve at various settings, and try to map the actual operating points you've reached, and then try to find a way to get to that high OP using only pump A.

Sounds like fun, but you'll need a lot of crayons.


Mike Halloran
Pembroke Pines, FL, USA

 

[rotw]

Might be that your iso-power line (max corresponding to your motor rating) is a sort of diagonal crossing the pump curve at two points.

When you are between the two points your motor cannot handle the amps.

When you approach via the right of the curve (as I understand decreasing the flow by switching pump B off) your flow valve opening is set such as to not trip the motor and flow/pressure value is then locked on high flow side : 165KBPD @120psi.

On the opposite, if you approach the curve via the left (conventional start up with pump A only) flow valve opening is set to not trip the motor but the flow/pressure is on low flow side : 120KBPD @ 130psi.

In between the two points there is a region in the curve that you cannot pass through
Could be that evidencing the above was the reason for your supervisor to do the experiment.

Goes without saying that all this is speculative and just a brainstorming exercise.

Need clearly more details like requested previously to figure out.

 

[Artisi]

As Mike Halloran has said some pump performance curves it might help is sorting out the problem. Even the standard performance curve for these pumps is a good place to start.

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.)