System low head reqd + Gravity flow

[Jhon Moncaleano]

Hi,

 wanted to check the following with you; at the moment I'm working with a system that pumps water from a dam through a hill ~7m elevation for 200m, and from then on it becomes gravity flow going down to a level 25m below the highest point for ~600m. 

I notice that when I run the simulation using DN250 pipe, it shows cavitation in pipes with gravity flow portion, but if I divide the analysis into 2, excluding the gravity flow portion and running the initial pipe, the simulation runs well.

I can also make the system work fine if I do the gravity portion as DN200 (doesn't work with bigger diams), but I feel like I'm getting an incorrect message as there would not be cavitation in the gravity portion, rather a high water column pushing down. 

My question is, would I need to adjust any setting for gravity flows? I would imagine the warning refers to negative pressure or possible siphoning, instead of cavitation?

Thanks (file attached)

Pump example partially gravity.psm

[Martin Griffith]

Hi Jhon, it's because you have negative pressure in the pipe as the water goes down. Pumpsim assumes the pipe is full with water, but in reality perhaps the pipe would never entirely fill with water and would instead behave more as a channel. 

[Jhon Moncaleano]

Thanks Martin, agree, there would be negative pressure, although this would not be cavitation as the warning indicates. on the highest points I would be using air valves (double valves) would there be a way to simulate this? or would you suggest to include channels in a downward portion?

[Martin Griffith]

It is difficult for Pumpsim to model this.

If you have an air valve that is releasing all of the air at the highest elevation then your pipe would be full and then you would see the negative pressures in the downwards portion of the pipeline. These may be sufficient to generate cavitation, but it may manifest ultimately as an entry of air at the pipe exit and then the final portion would be behaving as a channel. Where this begins though is difficult to know. 

You can make an estimate or observe the real system and then change the latter portion to a channel.

I hope this helps, happy to talk more with you about the best way forward.

 

[Jhon Moncaleano]

Hi Martin, I think you were right on this one, we really would be looking at two systems, and one of them would not fully filled. A way I have decided to attack this and I think it works well, is by adding to the end of the line a pressure sustaining valve, which increases the HGL line above the negative value and allows to operate with a "full pipe system", see an example below, Let me know what your thoughts are. 

[Martin Griffith]

Hi Jhon, yes I think this is ok, I think the common use of the pressure sustaining valve idea is to raise the pressure in such a pipeline, reducing flowrate, but avoiding negative pressures and cavitation.