Selection of Pump

[bharath1950]

For Heater charge , we have two pumps ( both are motor driven) . One will be running and another as standby which will be kept in autostart mode . The standby pump starts in autostart mode when the discharge header experiences low low pressure.We have sufficient NPSH margin ( more than 2 metres) at rated conditions and nearly 0.6 meter at 120% of the rated flow. The rated point is nearly 82% of its BEP.
But at the end of the curve the NPSH margin is negative. My query is, when the pump starts in auto, since the NPSH margin is negative , whether the pump will experience cavitation? If so should I have a relook on the selection of the pump? The pumps are driven by 500 kw motors and the motors are selected for operating at end of curve operation.
Thanks to clarify.

 

[JJPellin]

I would not disqualify a pump selection based on a transient condition that might last for a couple of seconds and only during an unplanned auto-start scenario. I am more concerned about the basis for the NPSH margin at rated conditions. I don't know if everyone does this the same. Our specifications require that the pump have an NPSH margin at rated flow of about 1 meter (for hydrocarbon streams). But, the calculation for NPSH available is based on the suction vessel level at bottom tangent or bottom nozzle. This makes the calculation inherently conservative. Since the normal operating liquid level might be more than 2 meters higher than this reference, we have an actual NPSH margin much higher than the calculation would suggest.

Even if you are using actual running level for the NPSH margin, I would not worry about a couple of seconds of low margin (or negative margin) operation during an unusual and rare operating condition.


Johnny Pellin

 

[1gibson]

So in the scenario that disch pressure becomes low (from pump 1 being at end of curve, or disch pressure maybe reduced even further due to pump 1 cavitating) then pump 2 kicks on.

Since they are now two 50% pumps in parallel, system resistance throttles them back to lower flows where less NPSH is required.

Sounds pretty much inherently safe to me, you have almost immediate correction in the event of pump 1 operation beyond rated flow, or cavitating.

I don't think pump 2 is in any danger at all, at least not compared to pump 1. Unless it could be subject to excessive start/stops within a short time frame (check with motor manufacturer.)

 

[stanier]

Suggest you download the following free ware and model the various scenarios to check your NPSHa and compare to the NPSHr from the pump manufacturer.

http://www.pumpsystemsmatter.org/content_detail.aspx?id=110

If this is only a transient event then a risk analysis will determine if there is any concern based on the frequency and length of the events occurrence.

 

[firavia]

selecting a pump is not an easy taks there are so many factors that should be taken into consideration before doing that.
I found so many documents online that explains the steps that are taken in order to select an appropriatye pump.

http://www.firavia.com/mecheng.html

you have to be carefull it is hard to use a system curve for parallel branched piping network.

 

[vpl]

Bhrath1950

This thread provides a link to a pump selection software (that I claim no knowledge of). More importantly, it also offers a link to the McNally Institute which has a lot of information on pump selection: thread407-280011.

Firavia

Seems like you've put together quite a library.

Patricia Lougheed

******

Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of the Eng-Tips Forums.

 

[BigInch]

You'll probably pick up system resistance fast enough that you'll not reach runout flows and NPSH will stay within operation range. A pump start-up transient analysis would most likely confirm it ... or not, but that is the best way to find out.

"I am sure it can be done. I've seen it on the internet." BigInch's favorite client.

"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.live.com

 

[BigInch]

In any case, as JJ mentions, such limited times as that typically required to start up motor driven pumps would likely not present any cause of concern. Sustained operation in that range would be a different story.

"I am sure it can be done. I've seen it on the internet." BigInch's favorite client.

"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.live.com