1. ## question pump

I have a pump. The NPSH curve is:
5280 cmd - 3m
8160cmd - 4.9
10080cmd - 6.7m

The available NSPH and the flow are shown in the figure.
The available NSPH has variation of 0.017 m, but the flow are almost 2500 cmd.

The pump is set up to design point.
Why the flow has such a big variations?

A little additional information would be helpful in understanding your exact question. What are the conditions you are seeing variations? What does the system look like? Is this pump station supplying all the system demand or only part of it? Is this a Variable speed or full speed pump?

Additionally while design point curves are very useful to the modeler, one must always be aware that they are based on an assumed shape of a pump curve. They use a three point curve by assuming the shutoff head is at 2 x the design head and that the high flow is 2X the design flow at a head of zero. This means, that the accuracy of the pump curve will likely depend on the pump operating near the selected design point. If the system conditions you chose for different runs creates such a variation in flow, generally this means that the system curve for each condition is different enough to cause the difference in flow. Recall the pump ONLY operates where the pump curve intersects the system curve. So where the model has the pump operate is based on the system conditions for each run and this will impact where the pump operates.

This presentation link on system curves may be helpful if you need further explanation of how they work: http://blog.innovyze.com/2017/01/24/...onsiderations/

One last note, if you happen to use a PRV downstream of a pump to simulate a VFD, please note this type of condition can see very large variations in pump station flow based on slight changes to the PRV setting as well. This is because the PRV induces headloss and this can allow for great variability in where the pump will intersect the system curve that would not normally occur if the pump is using an implicit VSP control.