Pumping over a hill yet again 3

[TomaszKruk]

Hello,

I read all the other topics, but english is not my first language so I would like to be 100% sure.

I'm working on an industrial water system where I am to use existing pumping station and pump the water to a new plant some 60 kms away. There are local high points on the way, but the hills are of lower elevation than the receiving point. So my pump has to provide sufficient pressure to pump over those hills. The classic siphon will not occur.

Do I still need to worry about possible vaporization? I read somewhere that the velocity difference on the upward and downward slope could cause flashing, and disturb the flow. Is this right? If so - how to calculate it?

If it's as obvious as I think it is (that no - it will not be a problem) then sorry for spamming. Internet is a great resource but sometimes the data is so fragmented it gets you scared without real reason.

I've attached an epanet's profile of pressure to visualize what I'm talking about. Generally you can see the pressure falling, since the terrain elevation and head losses decrease it. The local minimums display points of high elevation I'm worried about.

 

[LittleInch]

So long as your pipeline is

A) full of water and
B) Kept at a pressure of > 0 barg at all points

then you don't need to be concerned about the intermediate hills.

Your pressure plot indicates positive pressure at all locations so you are all OK.

The issue you talk about is "slack line" pumping where the flow can only just get to the top of the hill then falls away creating vapour and flow disturbance on the downstream section. This doesn't happen here.

Can you also plot the profile and the hydraulic gradient?

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

 

[TomaszKruk]

thank you for your answer. The "just able to get to the top" thing got me worried - I saw around 5 forum topics about difference in velocity when pumping over a hill, and only one answer clearly stating that hydraulic gradient high enough means there's no risk of vapor creation.

From what I understand the "just able to get to the top" part means that the risk is in the start up, shutting off and, later, in system aging?

Here's the profile of hydraulic head and elevation. It's only partial, as the pipeline is quite long, and I woud either have to plot every 10th junction or so or make the graph hard to read.

 

[katmar]

There are local high points on the way, but the hills are of lower elevation than the receiving point. So my pump has to provide sufficient pressure to pump over those hills. The classic siphon will not occur.

Is it not worth considering some form of air removal or flushing so that you can take the siphon effect of pressure recovery on the downlegs into account?

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[LittleInch]

You only seem to be giving us half of the profile, but I think you've got the correct design IMHO

What you really need to do is superimpose the Profile of head graph with the ground elevation profile.

Depending on how much pressure / head you need at the end of your pipeline you can then see what head you need to maintain a positive head pressure at all points.

If you have a large dip right at the end then you may need to maintain a back pressure to keep the pressure positive.

I don't agree with Katmar - you get very little advantage from a slack line system and end up with lots of issues plus if the hills are lower elevation than the end you simply can't get a syphon effect, nor do you really want it.

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

 

[TomaszKruk]

Thank you LittleInch for excellent answers and Katmar for the input.

The graphs I added are misleading in that they show only the part of the pipeline I was worried about. As I wrote in the first post - the end of the pipeline is way above the intermediate hills so siphon would not work and is not necessary (thankfully - since I doubt anyone around would like to touch that).

As for slack lines in general - I read only a few posts about it but everyone seems to avoid them like fire. I don't have personal experience with it but yeah - opinions seem to be consistently negative. As with everything - it's up to the person who's going to take the responsability for the whole thing.

 

[katmar]

@LittleInch - I am not proposing using a slack system. I was actually warning against it. In your first reply you stated as an assumption that the line must be full of water and I agree that should be the goal. If it is full of water then you don't have to worry about pumping over intermediate hills (except at start-up) because you get the pressure recovery from the siphon affect on the downleg after the hill. But in slack flow the downlegs are full of air and you get no pressure recovery. This results in higher pressure pumps being required.

Unfortunately the "full of water" condition does not happen automatically. Either air vents must be provided to remove the air, or the line size must be correct to allow the air to be flushed out by the water.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[LittleInch]

katmar - OK I misread your post, my apologies.

At 60km long I normally work on being able to pig it and use that during first fill to sweep all the air out, but if not then fair enough, you end up with a load of high point air valves which are more common on water supply systems. I get what you say about the syphon, but the impact is at best a few metres and difficult to control.

The problem with slack line operation is that it is possible and it does work, but doesn't in fact get you very much benefit most of the time and results in a lot of flow disturbance and vibration. As far as possible you should design to avoid it. IMHO.



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

 

[cvg]

pumped water always contains dissolved and entrained air and it needs to be continually removed as it accumulates, unless you have high enough velocity to sweep it out. higher velocities require more energy to pump. with the long pipeline and high points, that will almost certainly require combo air/vac valves

 

[itsmoked]

I was surprised to find the 48" Alyeska pipeline to be slack.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[LittleInch]

So was I. They apparently had issues with ground vibration caused by the slack flow.

I've never really understood why they do this when the advantage is very small. They should create a back pressure just enough to create 10m positive head at the high point and then use an energy recovery turbine.

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

 

[itsmoked]

It sure decoupled the sonic signals from the Valdez end for me.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[bimr]

What capacity does the pipeline have? Issues that need to be concerned with are flow velocity and the potential for pipe collapse if a vacuum is caused by flow separation.

 

[TomaszKruk]

around 1.5 m/s