Testing individual pump performance when pump is in parallel? 3

[mdsupabro]

Hello,

I have been in heated discussion with a vendor who recently tested overhauled and tested a boiler feedpump for me. I attempted to explain that you can not measure the performance of an individual pump performance (head and flow) when the pump in parallel. Because if the discharge head of one pump is higher then the other, the stronger pump's discharge head will be what is actually measuring. Even if the dishcharge pressure is measured after before the the discharge lines combine to a common header. I certaint this to is true, however I cannot remember what physics law this falls under. Can any one help.

 

[ccfowler]

Individual pump performance can indeed be tested while pumps operate in parallel. The important issues involve the system configuration, suitable provisions for measuring pressures, temperatures, flows, and power in addition to the accuracy of instrumentation being used. Good data can be acquired, but don't expect it to be cheap.

 

[JJPellin]

Since the two discharge lines are tied together with no imposed pressure drop between them, the two pumps must be running at the same discharge head. This does not make it impossible to measure performance of an individual pump. The problem is not pressure. The problem is flow. It is very rare to have individual flow measurements at each pump. You cannot take the total flow and simply divide by two. This would assume that both pumps are exactly equal. If you have motors, you can take the amp draw and attempt to correlate that with flow. But, if you use the original HP curve for the pump, you would be assuming that the pump is in "like new" condition. If it is in degraded condition, the assumption would provide a false result. If you are fortunate enough to have individual flow measurements that can be trusted for accuracy, your vendor is correct and you can measure the performance of an individual pump. In our plant, out of 1400 centrifugal pumps, we have perhaps 6 pumps that have this capability. As the previous reply noted, this level of instrumentation is costly.

Johnny Pellin

 

[mdsupabro]

Thank you all for replying. So I'm hearing that when measuring TDH of two pumps in parallel, if measured, will not be the same and you CAN be measured while both pumps are in parallel. Secondly, the vendor measured suction flow of the pumps, which have separate lines coming into the pump.

 

[JJPellin]

Actually, no. The TDH of the two pumps running in parallel must be the same unless there is some imposed pressure drop between them. If you have a control valve or a pinched block valve or orifice between the two pumps, they could be operating with different pressures. Otherwise, with wide open lines, each pump would have the same suction pressure, the same discharge pressure and thus, the same TDH. It is the flow which may be different based on differences in the condition of the two pumps.

Johnny Pellin

 

[EdStainless]

If you measured two pumps in parallel and knew the common outlet head, inlet suction on each, total flow and power draw from each motor......
And then re-built one of them and re-tested you could infer how much it changed.
But by nature your second test will not be at the same flow as the first. Combined with the fact that you don't really know the pump curve for either pump. No, you can't measure much.


= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[mdsupabro]

Thanks again last question, What physics principle is this, that I may reference?

 

[BigInch]

First law of thermodynamics,

http://www.grc.nasa.gov/

http://WWW/K-12/airplane/thermo1f.html

From the hydraulics of flow and differential head of each pump, you know the output power of the pump; From the electrical power consumed, you know the power input.
Input_power - Output_power - frictional & conversion losses ==0

http://virtualpipeline.spaces.msn.com

 

[ClicketyClack]

I would also apply the "Law of Kharma" and apologize to the vendor. He obviously knows what he is doing and you obviously need his help

 

[quark]

I would read the discharge pressure of the pump in focus, immediately after the discharge flange and before any intentional or unintentional; partial or full flow restrictors.

Then I would disconnect it from the discharge header, simulate the same discharge pressure as that of parallel arrangement and then read the flow readings.

As this is a boiler feed pump, you should have approximate idea about the flowrates (by boiler capacity, by fuel consumption etc. etc). Now check whether the test pump capacity is nearer to 50% of the total flowrate and that is all.