NPSH Curve

[Vic123]

I want to test the pump and plot the NPSH curve.
Please advice how to go about the process. I have no idea as what to measure and how to go about the process

 

[Bjegovic]

1) set a test rig - you need to recirculate fluid, measure relevant pressures and flow, and power at pump shaft in varying conditions

2) set ctiteria for cavitation (is it incipient, such as formation of first bubble - to be detected visually or by stroboscope, or you will detect loss of power or head drop as a sign of developed cavitation)

never done this, but seens to me like a serious task; try consulting books and standards

 

[JJPellin]

In the pump shops, they have test rigs set up for this. They run the pump at a fixed flow rate with full suction. They measure the Total Differential Head based on suction and discharge pressures. They calculate 97% of that value. They pinch down on the suction in small increments. After each suction adjustment they readjust to get the flow back to the original value. They keep doing this until they get down to the 97% of TDH that they calculated. This is the point where the pump is cavitating enough to demonstrate a 3% head loss. Once they reach the 97% value, they calculate the net positive suction head available and that defines one point on the curve. Then they change the flow rate and do the entire process again. They do not test below recommended minimum continuous flow. At 3% head loss, there can be a considerable amount of cavitation. I would prefer to see a curve for insipient cavitation. But this can be hard to define.

 

[rmw]

We had a similar procedure as described by jjpellin when I was a test engineer for an evaporator manufacturer. We weren't trying so much to develop an NPSH-R curve as we were trying to verify the manufacturers curve especially at the higher flow rates where the curve began to increase sharply.

We found that lots of manufacturers either didn't do it accurately or maybe they cheated. It mattered with our type of equipment. We also found out who had it right and whose curves could be trusted.

I recently spouted off in a meeting at a clients office regarding this topic and called the names of several trusted pump companies (Carver pump among them) based on our test results, and they then introduced me to someone in the meeting who I hadn't previously met and who turned out to be the local Representative of Carver Pump. He was naturally beaming at my recommendation of his pump.

rmw

 

[vanstoja]

jjpellin's description is generally correct for a closed, pressurized system NPSH test described in the Hydraulic Institute Standard or ASME Power Test Code, but "pinching down on the pump suction" is a rather vague term for the process of stepwise reduction in suction pressure to the pump. When reducing pressure stepwise it is necessary to start at a high enough pressure to differentiate a 3% pressure reduction from the highest to lowest flowrates to be tested since the the slopes of the pressure versus head curves will vary substantially with the flowrate. Ideally, at least three or more flowrates should be used including maximum, design rated and minimum stable operating flowrates. To discern the 3% head dropoff point effectively, it may require many pressure reduction steps at each flowrate to offset measurement errors and flow effects such as alternate vane cavitation which causes step changes in pump head level. We used to run as many as 20 pressure steps at each flowrate to get accurate results for a high (4100 RPM-GPM-Ft.)specific speed pump.