Creating System Curve for Pumping System

[hceng11]

Hello. I am creating a system curve for a force main that is approximately 2 miles long, and conveys water from one pump station to another (see attached sketch). The pump system is located in the mountains, so there are sections of both pressure and gravity in same line. My question is whether I need to account for friction losses within the pipe for the entire distance of the force main (Sta 0+00 to Sta 96+15), or just for the section between the first pump station and controlling high point (Sta 0+00 to Sta 51+50)?

Thanks

 

[bimr]

You have not provided enough information to understand your project. However, if you pump the entire length, then yes, you will have friction losses on both the upward and downward segments. The pump will have to be sized to pump over the hill on startup, but will then operate with less power as the energy is recovered on the downhill segments.

 

[fel3]

hceng11…

Your ground profile dictates the shape of hydraulic grade line you need to get flow from the first pumping station to the second one. It needs to look something like the attached sketch. You will need air release/vacuum (ARV) valves at the high points (not just the two major ones) and blow-offs at the low points. The HGL at the high points need to be sufficiently higher than the ARVs at all times to prevent valve cycling.* Unless you HGL profile is considerably higher than the ARVs, I suggest a surge analysis to check for transient low pressures at these points.

BTW, if there was no uphill segment ahead of the second pumping station, you could pump to the second high point and let it gravity flow from there.

Fred

* Years ago, I investigated a 1.25-mile-long, 12-inch-diameter agricultural drainage force main that had a problem with ARVs cycling during pumping. Upon seeing the ARVs spit water and hearing them suck in air on my first visit to the site, I immediately concluded that the design engineer had goofed up the hydraulic calculations such that the hydraulic grade line was fluctuating around the elevation of the valves. My business partner was skeptical because the design engineer, who had since left the company, was *very* experienced (in fact, he had taught one of my college classes). However, when I examined his calculations, I found a note that minor losses in the 4-inch lift station piping was "insignificant, by inspection," or something like that, and would be ignored. When I calc'd the minor losses, I discovered that they exceeded the head loss in the 12-inch force main. His hydraulic grade line was a nice straight line that was above the ARVs, but mine dropped sharply in the lift station, then nearly hugged the ground the rest of the way. The last few downstream ARVs were above my hydraulic grade line, which caused them to open and close every few seconds, which in turn caused the flow in the force main to surge, which created a feedback loop that caused additional ARVs to open and close. Interestingly, because this force main was had a flat (not level) and slowly rising profile, the ARVs weren't really needed and certainly not every 200 feet. My project was to lengthen the force main to reach a more distant evaporation pond, so my new larger pump solved the problem.

==========
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

 

[bimr]

You might search this site for the problems that others have posted in regards to forcemains. The problems seem to comes up quite frequently. When you pump over the elevations, you have to ensure adequate velocity in the downhill segments to clear the line of air. Otherwise, you will encounter air bound pipes or two phase flow situations. The two phase flow situations result in higher headloss because of the reduced flow area that the fluid is passing through. Unless the forcemain is cleared of air, the forcemain will not have the expected capacity. A low velocity of 2-3 ft/sec in your application is a recipe for a forcemain problem.

 

[hceng11]

Thank you both for your feedback.