How to Account for Positive Suction Pressure in Pump-System Curve? 1

[KernOily]

You think you know something, then a question comes at you forcing you to re-evaluate what you know, or at least think you know...

Question: How do you predict the operating point when a pump takes suction from a pressurized line?

Does it act like two pumps in series, i.e. add the heads at every Q?

Would I construct a 'virtual' series operation pump curve, i.e. the pressurized line acts like the first pump (a virtual pump), and the second pump in series is the actual pump? Then, overlay this 'series' curve over my system curve to obtain the operating point?

Thanks guys!!!

 

[BoomerSooner7]

Not a guru by any means, but we do similar tests at our facility with a charge pump feeding our test pump. We measure the differential pressure (discharge minus suction) of the test pump and plot that against the flow rate, which is obviously constant through the system at any given operating point.

 

[JJPellin]

The curve for a centrifugal pump represents the differential pressure provided by the pump. It is valid for any suction condition as long as the pump is not starved for suction and cavitating. So, the pump curve is unchanged by the presence of positive pressure in the suction line. The system curve is also a representation of the pressure differential between the suction and discharge sides of the pump. A system curve could go negative at low flows if the static pressure in the suction system is higher than the static pressure in the discharge system. If there is a significant dynamic component to the system curve, it would tend to go positive at higher flows. If the system does not have enough dynamic component to go positive at higher flows, then you don't need a pump, you just need a valve to throttle the higher suction pressure into the lower discharge pressure.

A pressurized suction does not change the pump curve. It does change the system curve. The pump will operate at the intersection of the pump curve and the system curve.

Let me know if I misunderstood your question.


Johnny Pellin

 

[BigInch]

KernOily, You only got one pump, right?

Just add the pump's differential head at any given flowrate to the head at its intake at the same given flowrate.



"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

 

[KernOily]

OK. Thanks for the great replies.

I attached the pump and system curve.

This system has one pump that can either take suction from an atmospheric tank OR a pressurized line at 150 psig. The destination is 10,000' away and 300' uphill.

Note the system curve is negative when the system takes suction from the pressurized line, just as JJ said.

I created a fake 'series pump' curve by doing what Big Inch said: adding the heads at every Q.

Now: When the pump is taking suction from the 150 psig line, i.e. the system curve is the olive green line, which is the correct operating point: Point 1, or Point 2?

 

[BigInch]

That certainaly doesn't look right.

According to how I read your description, each discharge system curve (from either source) has a static head at 0 gpm of 300 ft. You calculate and draw the discharge system curve separately from the suction head + pump differential head beginning those curves at 0 gpm with the static head in that pipeline.

The suction line graph starts with the tank pressure head equivalent or the pipeline pressure head equivalent.

As I see it, you are pumping to the same destination using the same flowrates, so the discharge system curve is the exact same for both.

I made a corrected spreadsheet for you.

BTW, the NPSHR curve is relative to absolute pressure. I don't know if those values are already corrected to 0 psig reference or not. If the 6 to 9 ft are NPSHR values not corrected to 0 psig, they would appear as a nearly flat line at -24 to -27 feet on my diagram, which is referenced to 0 psig = 0 ft head. In other words, a suction pressure of 0 psig is actually 33 "feet absolute" of NPSHA.

"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

 

[KernOily]

BI - Awesome. Thanks for taking the time to make the spreadsheet. I glanced at it and didn't get illuminated, so I need to spend some quality time with it. Thanks again. Pete

 

[BigInch]

The only part that might be difficult to follow is where I enter your calculated discharge pipeline system curve numbers to get the discharge pipeline head losses vs Q <cells G4 to L11>, which I then added to the 300 ft static head at 0 flow you need just to first get the top of the hill wet. I didn't recalculate friction losses, since I didn't know the diameter or roughness of your pipeline.

After beginning with that 300 ft head, my curves are the same as yours, except they plot on top of each other. Those curves would be different only if you were pumping up different hills, which would change the biginning static pressure of each pipeline, if you had two pipelines of different diameters, which would change the slopes of the curves, or were pumping two products of different viscosities, or densities from your tank or pipeline source, which would change both static head and curve slopes.

I ignored NPSHR, since that doesn't affect any of the system curve values; just nice to confirm that the suction heads (33 ft and 346 + 33 = 379 ft) are always above R.

I do have one question though. Typically your suction heads will drop as flowrate increases due to friction losses in the suction piping, or may change as your tank fluid level goes down, but yours appear to stay the same at all flow rates. As long as you're happy with that, I've no problem.

You know what to do if you have more questions.

"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