Sizing Pump & Developing system Curve for Negative static head

[ubique345]

Hello,

I'm currently helping with the design of a pumping system where there overall static head is negative, and i've never dealt with a problem like this before.
I was hoping you guys could point out considerations for this type of problem, and also help explain how a system curve would be plotted for such a system.

The lift station and water source is about 659 m at x(horizontal distance)=0
It pumps over a hill at 670m elevation at x=1400m
The final destination is at 585m at 6000m

Now, the total static head is less than the total friction head, so how would one go about plotting a system and pump curve for this? Would I need to add a valve at the end to increase friction loss, forcing back pressure to bring my curve above zero?

When sizing the pump, would I just use the static head from the start point to the high point as my static head, and add the friction of the full 6000m to get my totel required head? I'm looking at ~6000 gpm for a target Q
I know air/vacuum valves will need to placed throughout the peaks and at some other key points in the system, but are there any other important
factors for consideration, or items that are typically overlooked when dealing with this type of problem?

Thanks

 

[LittleInch]

OK, A lot to deal with here.

What you're looking at is what is often termed slack flow. Search for this on this site and elsewhere and you'll find a lot of info, e.g.

http://www.eng-tips.com/viewthread.cfm?qid=369908

For the system curve, all you need to look at for starters is the first section to your high point. Thus you have a static head to achieve (11m) before you get any flow, then as far as your pipe is concerned, as soon as it get there the water basically just flows away faster than you're pumping it in, hence the system curve will be only for the first 1400m.

However this only lasts until the overall friction losses don't exceed the fall from your high point to the end point, which if I get this right is 85m over 4.6km, 18.5m/1000m. This is only valid if this hydraulic line doesn't hit any other high points between KP 1.4 and KP6.0. A profile is required to make sense of this.

Your system curve would therefore start at 11m, have a curve based on 1400m of losses up to a certain point as noted above) then start again from that point with a steeper curve as you are then looking at 6000m of pipe. Your current flow sounds like it won't get to that point of change.

Slack line flow makes it difficult to get steady state flow in your pipe downstream of the high point. You get high velocities, surges, sometimes vibrations, you can't really do leak detection on it and it makes calculations difficult. If you stop the flow at the far end but keep pumping you can get very high surge pressures as the vacuum inside the pipe collapses and the water columns hit each other.

You can insert a back pressure valve at the far end to keep pressure at all points >atmospheric pressure. The most efficient system is to install a pressure TX at the high point and control the back pressure valve on that point, but you can do it manually.

You can install smaller pipe in your downhill section to raise the steady state friction drop to achieve positive pressure at the high point.

Without doing anything which could raise friction losses, though this is a good idea to prevent all sorts of operation issues, the pump then needs to raise at least 11m, plus the friction losses from KP0 to KP 1.4 on the basis that you believe the friction losses form high point to end point are less than the static head.

As said - get the profile and start drawing hydraulic lines on it, i.e. pump end add the height in m you need for static head (11m plus your friction losses and see where the line ends)

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[ubique345]

Thank you littleinch, your information was a great help.