End of Curve Operation requirement for Auto start pumps 4

[SENGUTTUVAN]

Hi
I would like to know that for all the AUTO START centrifugal pumps ( Open valve start up) , whether it should be designed for End of Curve operation? That is , NPSHR should be less than NPSHA even even at its end of curve. I am guided to select a vendor who can satisfy the above condition. We have two pumps ( 1 running + 1 standby) . If one pump trips the other pumps should come in line in auto. we have individual NRV for these pumps . When I should propose to the vendors that my requirement of a centrifugal pump should be designed for the End of curve operation where the NPSHR is less than NPSHA at its end of the curve. Can U please throw some light on this.

 

[Artisi]

You need to draw a system head curve on the pump H/Q performance sheet- as it is possible that the pump will not reach the end of curve on start-up as the system resistance may intersect the H/Q prior to end of curve.

If the pumps can not run to end of curve there is no need for the NPSHr to be less than the NPSHa - calling for this in a contract will only result in larger and more expensive pumps which will then run too far left of their BEP in normal operation.

However, if they can run to end of curve you need to look at the running period before they run up the H/Q curve to the normal operating point - a short period of cavitation is usually not a problem.

 

[BigInch]

Even if your system curve will let you get to the end of curve range, but you don't have to go there for any reason, add a high flow or low suction pressure shutdown.

**********************
"Pumping systems account for nearly 20% of the world’s energy used by electric motors and 25% to 50% of the total electrical energy usage in certain industrial facilities." - DOE statistic (Note: Make that 99.99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[SENGUTTUVAN]

I do agree with your view on the short period of cavitation is usually not a problem. But how to find it out? The pumps i was mentioning is intermittent run pumps and gets started based on level . In my case the NPSHA is 4 metres and at duty point the NPSHR is 3 metres. But at the end of curve operation the NPSHR is 7.5 metres and hence I have been asked to increase either the NPSHA ( by elevating the suction vessel ) or to go for a different pump model which will staisfy the above condition.

Hi Biginch
I am not able to understand your answer of " add a high flow or low suction pressure shutdown" Pl explain

 

[BigInch]

You are asking if the pump needs to be specified to run to end of curve. That's a process question. If you need those end of curve flowrates, the answer is yes. If you don't need those end of curve flowrates, and as Artisi has already mentioned, the answer is no.

If your system curve and pump curve can intersect at high end of curve flowrates, where suction pressures drop below NPSHr, BUT you don't need to pump those high flows, just install a low suction pressure switch, or a high flowrate switch to deavtivate the pump motor, or to control a VFD, or install a flow control valve to limit flows to acceptable ranges, where suction pressure is always higher than NPSHr.

**********************
"Pumping systems account for nearly 20% of the world’s energy used by electric motors and 25% to 50% of the total electrical energy usage in certain industrial facilities." - DOE statistic (Note: Make that 99.99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[SENGUTTUVAN]

Hi Biginch

Now I have understood Ur reply. But in my case , the duty point requirement / flows are well below the end of curve and have an NPSH margin more than 1 metre. Only that during initial start up of the pump, when the discharge lines are empty , whether this pump will go to its end of curve operation ? At this condition from the performance curve we see that the NPSHR at the end of curve is 7.5 M as against the avl NPSHA of 4 metres. Our concern is , during the pump start up ( by keeping discharge valve open) because it may operate upto its end of curve , should we ask the vendor to suggest a different model pump.

 

[TenPenny]

I don't see how the pump would start up under end of curve conditions.

Typically, you want a pump to start up against a closed valve, isn't that so? And in this particular case, you have a non-return valve, correct?

 

[BigInch]

With open discharge and no static head, the pump will quickly reach end-of-curve. Question is will it come back to BEP-normal flow area quickly or not? If it comes back into your NPSHa range quickly, the brief excursion into < NPSHr shouldn't be a problem. If its going to be a long time before you return to BEP-normal flow range, you might want to think about installing a flow control valve.

**********************
"Pumping systems account for nearly 20% of the world’s energy used by electric motors and 25% to 50% of the total electrical energy usage in certain industrial facilities." - DOE statistic (Note: Make that 99.99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[JJPellin]

I agree with TenPenny. If one pump is on line and running and the other pump is in stand-by with a non-return valve (check valve), then it should never reach end of curve operation. The stand-by pump has to be flooded. There is full process pressure on the other side of the non-return valve. When the pump auto-starts, it will be running blocked in until it builds enough pressure to open the check valve. Even if the downstream pressure was reduced by the time the spare pump auto-started on low level, there should not be a great expanse of empty pipe to flood. A very, very short transient while the spare pump builds pressure should not be a concern unless you are talking about a very high energy pump or a vertical turbine pump with many stages. It sounds to me like you don't have a problem.

Johnny Pellin

 

[BigInch]

If the system curve when the other pump is running says he can't get there, then there obviously isn't a problem, and he doesn't need to spec everything to full runout. If he ever needs to start with both pumps off, then the system curve for that condition will apply.

**********************
"Pumping systems account for nearly 20% of the world’s energy used by electric motors and 25% to 50% of the total electrical energy usage in certain industrial facilities." - DOE statistic (Note: Make that 99.99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[SENGUTTUVAN]

I should thank all of you for keep replying. Here as Biginch was explaining , what is that transient time within which the pump will come back to its duty point. How to find it out? Within this transient time the pump should not start to cavitate. I am also stressing the same point what JJpellin was explaining to my process executives. It is strongly emphasised that For ALL OPEN VALVE START UP pumps , the pumps to be selected for END OF CURVE operation . How to go about further in this situation? Another question. If there is always a 3.5 bar back pressure downstream of the NRV in the system how will the NPSHR curve and the H-Q curve of the pump change eventhough this is good for this Open valve start up conditions.

 

[BigInch]

If I may, JJPellin was not thinking that the transient time would be long, which would generally be the case with one pump running and the other winding up. I mentioned the case where both pumps were off and there was no initial system pressure as another possibility where a longer time at higher flows might be possible. That can only be consiered by examining your system. What exactly are you pressuring up? If its a small vessel and short segments of pipe in a process unit, it might not be very long, but if its 50 miles of pipeline, that would take quite a bit of time. If it is a simple system, pumping to a tank for example, you can get an idea of the time it would take just by considering the flowrates produced by the pump against the backpressure provided by the increasing tank level. A more complex system might require a start-up simulation to get an accurate picture.



**********************
"Pumping systems account for nearly 20% of the world’s energy used by electric motors and 25% to 50% of the total electrical energy usage in certain industrial facilities." - DOE statistic (Note: Make that 99.99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[JJPellin]

It can also be pointed out that designing the pump for end of curve operation is not a free option. At that point, the system would have a lower NPSH(available) and the pump would have a higher NPSH(required). Selecting for this condition would tend to drive the selection to a pump with a lower NPSH(r) and a resulting higher suction specific speed. High suction specific speed can result in a pump with poor performance at lower flows. Only select a pump for conditions that it will actually experience for longer than a few minutes. Other examples have come up at my plant. In order to switch pumps from the main to the spare, both pump need to be running at the same time for a few seconds. But, we do not design our piping or our pumps for parallel operation of both pumps. During unit start-up or shut-down, there may be very brief periods of operation at very low or very high flow. For most services, this is not a problem and the pump design should not be changed drastically to be optimum at these unusual operating points. As I already noted, this might not be true if the pump is a Sundyne running at 20,000 rpm or a 30 stage vertical turbine. For a typical single stage pump, short duration transients are usually not a problem.

Johnny Pellin