vfd pump operations - 1 pump vs 2

[elsoulman]

I have been tasked with developing an interim operations protocol for an existing booster pump station that was designed incorrectly. The appropriate mods will be made to the PS after the peak season has passed in several months, but in the mean time, we need to try and operate this station with what we have. There are three pumps in the existing station, all three-stage vertical turbine pumps with VFD's, and all the same head and flow. They were designed incorretly in that the original engineer did not take in to account the existing pressure on the suction side of the pumps, and designed around a pump head of 220 feet rather than 150 feet. Our final mods will be to remove one stage of the existing pumps, and things should work smoothly.

We have tried running the PS with just one pump (w/ VFD) turned all the way down, and there is too much pressure on the suction side of the pump (from wells) to be able to run the pumps (really don't need the BPS, but it is being requested from managment). For the interim operations, a senior engineer mentioned that he thought the PS should be run with two pumps turned all the way down rather than just one turned all the way down, and that this operations will allow the pumps to operate on their curve. WHY??? Just trying to figure out why two pumps with VFDs turned down will operate better than one pump with its VFD turned all the way down with the case that I have. Any help would be greatly appreciated!

 

[Artisi]

The idea of running 2 pumps to get back on the curve will depend on what is happening on the discharge side of the station.
The senior enginer is probably thinking that if you run 2 units, the flow thru the station will increase meaning a higher discharge head from the increased friction loss in the pipeline (assumming a common line) which in turns means the pumps will see a higher total head which MIGHT put them back onto the curve.
Other than the above I can see no other reason for running 2 units and if the pipeline friction amounts to little then you will have 2 pumps running off the curve - so be cautious.

 

[elsoulman]

We specifically talked about running the same flow on the discharge side, but he implied that it would be better to run two VFD's rather than one, and for some reason that the station would perform better - and I'm still puzzling as to why!

 

[ccfowler]

elsoulman,

You will need to study the performance curves of your pumps to understand how they may operate in your expected mode. If you don't have actual test stand generated curves for your pumps, the published generic curves for your pump model should serve adequately for your estimating purposes. Remember to key all of your affinity law calculations to actual pump shaft speed (not drive frequency).

You apparently have lots of NPSHa, and that is always a very good thing. It should provide some "cushion" for uncertainties in your estimates of pump performance.

If these are reasonably new pumps (which I am presuming from your initial posting), the pump manufacturer will probably be willing to provide some helpful advice. Even if these are older pumps, the manufacturer may still be willing to be helpful.

From my experience, pump manufacturers usually are very responsive to reasonable, coherent, concise requests for advice on their pumps. If you don't expect them to do your work for you, they will usually share information based on a wealth of related practical experience. Be sure that you have a thorough understanding of your system's characteristics (head vs. flow curve, piping configuration, transient conditions, ...) before asking them for help.

 

[Artisi]

Unless I'm missing something and without a lot more detail - ie, pump curves and pipeline friction data I can't see any benefit in running 2 units unless you have provision to run them even slower than you are doing now.

 

[Artisi]

As a second thought - have you looked at running the pumps with partially closed discharge valves, not an option that I would recommend but as a stop gap until as such times you can shut the station down for modification it might be the only option.

 

[elsoulman]

Artisi - your suggestion of running the pumps with a partially closed valve on the discharge side is the one I have been a proponent of, but this senior engineer (nationally renowned) said that running the 2 vfds as compared to one could get us back on the curve... still confused on his suggestion! Just trying to get educated!!

 

[BigInch]

Will twice the flow (2 pumps turned all the way down) pull down the suction head enough such that eventually getting back on the curve is possible?

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[BigInch]

Sorry, might have read the OQ backwards.

Otherwise I'm with Artisi and can see no apparent benefit from running 2.

What's the system curve look like?


BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[Artisi]

I think it about time your "nationally renown" engineer explains why - running 2 pumps will get you back onto the curve.
Following this logic, would 3 pumps be even better?