Pump over a hill

[djcbgn]

My hydraulic model consists of a raw water system with a high point approximately 3,000 feet from the pump station and the ultimate discharge point about 40,000 feet from the pump station. After the high point the pipe runs downhill only about 5,000 feet and then the pipe is flat for the remaining length to the discharge location. There are also several stream crossings where the pipe may drop about 20 feet and then comes back up. The disharge elevation is lower than the water surface in the pump station so basically we are pumping over a hill, then flat the rest of the way.

My approach was to use an air/vacuum valve at the high point and maintain positive pressure throughout the system. Will the HGL at the air/vacuum valve be equal to the elevation of the valve? I really want the HGL to be above the elevation of air/vac valve, because I have a long run of pipe after the hill and friction loss of the downhill section of pipe is greater than the static head difference between the high point and discharge elevation. Again , I do not want the negative pressures in the system. I am modeling using WaterCAD.

 

[77JQX]

Will the HGL at the air/vacuum valve be equal to the elevation of the valve? - Don't know: you haven't given flow, diameter, or elevations.

I am modeling using WaterCAD - It seems you have the tools to answer this yourself. Stick a node at the high point and see what WaterCAD tells you.

 

[MiketheEngineer]

Once you get it started - will it not act like a siphon??

 

[coloeng]

1. Determine the flow you want to model through the pipeline.
2. Determine the friction loss from the pump to the top of the hill at this flow.
3. Select/Develop your pump curve so that the head at the pump for that flow equals the (elevation of the hill minus the elevation of the pump station) + (friction loss from pump to hill) + (additional head required at the hill).

 

[stanier]

Beware such systems are liable to have waterhammer challenges on pump trip. You could design the system such that the gravity section is designed as such. So instead of an air valve a break tank could be used.

You could put a hydraulically actuated control valve at the pipe exit to keep the HGL above the pipeline elevation. However this would add to the energy bill. If the grades are suitable abarometric leg could be installed at the pipeline exit.

Is direction drilling through the high point possible?

Can you provide long sections of the high point and flow rates, pipe diameter, material????

Watercad is based on on the Epanet engine and is perfectly good at modelling the steady state. however you need software such as Impulse, Hytran, Flowmaster, KyPipe etc to model the transient conditions.

"Sharing knowledge is the way to immortality"
His Holiness the Dalai Lama.

http://waterhammer.hopout.com.au/

 

[BigInch]

You must design the system so that the HGL is above the top of the hill, or you won't be able to pump water over it!

The air valve will let out air travelling up the hill and yes, the HGL must still be higher than the high point of the pipeline.

If you shut off the pump, the HGL will eventually fall and intersect the elevation of the high point. Then water between the high point and the end of the pipeline will run out, if you don't have a backpressure valve at the end of the pipeline holding it in. The liquid column of water will break at the high point, forming water vapor behind it as it runs out the outlet. Water vapor at normal temperature is around 1/2 psiA, so a vacuum will form and the vacuum valve will then open to allow air at atmospheric pressure to go into the pipeline to replace the water running out. If the valve is sized properly, no vacuum is then created.

A backpressure valve at the end of the pipeline uses no more energy than you are already using just to pump the water to the top of the hill. If you didn't have the valve there, water would just accelerate and run out faster emptying the pipe, causing you to need a vacuum valve. Water must use up all the available potential energy it had at the top of the hill by the time it gets to the bottom. A necessary constraint, since low pressure water can't store potential energy; its potential energy is equal to the elevation that it has at any given moment, any excess potential energy at any given time is expended by a change in its velocity. A backpressure valve just holds the water back in the pipe (at a backpressure equal to holding it's static head pressure up to the top of the hill), rather than simply letting it run out.

From "BigInch's Extremely simple theory of everything."

 

[stanier]

Biginch,

I understand your point but here are circumstances whereby the pipeline from the high point to the exit is not a consistent slope. There may be other high points and valleys. Then the pipeline is not going to completely drain.

Not all water pipelines operate as you suggest with the GL above the elevation. It is desirable to potable water pipelines as you dont want contaminated air drawn into the pipe on pump trip. It does make for a complex system where air valves are relied upon to exhaust air on pump start up. It is common.

An alternative solution to what you suggest is a surge tank with dipping tube at the high point. When the level in the surge tank reaches the bottom of the dipping tube it behaves as a gas accumulator. It provides pressure to overcome the friction in the pipeline. These are useful in overcoming transient events.

The original posting refers to the high point 3000ft into a 40000ft pipeline. The pipeline is mostly flat after the hill. This will remain full. The pump will accelerate the fluid and the pipeline will act with siphon effects. Hence differential head seen by the pump will be the difference between suction and discharge levels plus friction. The pipeline is not going to empty.

"Sharing knowledge is the way to immortality"
His Holiness the Dalai Lama.

http://waterhammer.hopout.com.au/

 

[BigInch]

I don't do siphons. They're not really worth it when you consider that air will eventually be trapped there and it breaks the siphon effect and you have to start them over and over and over. I prefer to keep HGL above the high point. You can only go lower by 30 ft or so before the siphon stops working, so does it really matter. The alternative stuff sound like a real PIA. Not something I would do when something so simple is possible, probably at less cost.

Ya I know that you very seldom ever have a constant slope from high point to exit, but that's isn't really important to solving this problem either. The idea is that if there is no backpressure valve, he either loses most all static head (the line may drain completely or not), or is left with a near vacuum in the parts of the line that drain, or if he has an vacuum valve, no vacuum and air in the line. I'll let him decide which one he wants. Don't forget that when starting up again, all those broken fluid columns add to the net head required, so the pump might just need extra head than before to restart flow. Backpressure valve sound better now?

Since the pump needs to start the siphon in your scenario, the pump must have the capability to reach at least the top of the hill. When the siphon effect is "ON", the pump would need less head to continue flow, but its probably going to continue with the same head as before, so the pipeline's flowrate will increase a little instead. I wouldn't suggest a VFD or even a pressure control valve to keep it at siphon flow rate; I'd just let make some extra head and let the pipeline flow a bit faster. Again, I'll let him decide what he wants to do about it.

From "BigInch's Extremely simple theory of everything."