hydraulic analysis of pipeline with discharge at high elevation

[Ed_edm]

This might be a basic question but google isn't helping and nobody I asked gave me a decent response yet. I'm doing hydraulics of pipeline that goes from plant A at elevation 0m, over a dyke at elevation 20m, feeding a pond at elevation 0m.

I want to calculate what flow can i get at different pump discharge pressures, to evaluate upsizing the pump.

If the pipe ends directly at the pond, at elevation 0m, this is pretty straightforward. Using AFT Fathom, the outlet node is a pressure node at the pond elevation.

Now my question is what if the pipe ends above the pond, say at elevation 10m, then falls into the pond ? How do I model this ? How do i characterize the outlet node ? If I simply put a higher elevation, it seems to give a lower flow as if the pond were at the higher elevation.

 

[bimr]

You have a simple question, but there is no method to answer your question without knowing some of the details. One has to know the piping material, the fluid velocity, the pipe length, and the pipe diameter.

If you would like to have your question answered, I suggest you prepare a rough sketch with the details and post it.

 

[LittleInch]

Flow down vertical risers which is your 10m drop can be many things depending on flow and pipe size.

If the flow is small and the pipe big then your flow in the vertical drop will be open chanel type flow like a waterfall and hence the end elevation is essentially at 10m.

The other extreme is that the flow is big and the pipe small such that the pressure drop over your 10m is higher than the vertical fall. This is unlikely. But then the end of your pipe is at 0m

Our somewhere in between which is where it gets difficult.

So plug in 10m at the end for your analysis and it won't be any worse than that and might be a bit better.

See what the effect is of you say the end point is at 5m and you're probably closer to reality

But some details such as flow, velocity, pipe size etc would help.

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

 

[1503-44]

I'd recommend sizing the pump for the 20+m elevation with a pipe sized for the 0m elevation discharge flow. If you don't size the pump for 20m elevation and you turn the pump off, you will lose the water in the down slope, and the siphon effect will no longer work to your advantage to increase your flow rate, as it is doing now. Once that down slope water is lost, you may have trouble filling it again, if you don't have a pump with 20+m head capability.

If you design it that way, the system will pump at the 20m elevation flow rate until the downsloped pipe is full, then the flow rate will increase to equal the full rate you see for the 0m discharge elevation.

 

[cvg]

your dike is 20 meters high. in order to fill the pipe and get flow over the dike, you need to size everything to at least handle 20 meters of static head plus friction loss. once you get the pipe filled, than you might be able to rely on some siphon effect to increase the flow, but only if you can maintain a full pipe flow.

 

[katmar]

It seems that AFT Fathom is assuming that you have slack flow in the downleg after the dyke, and is taking no credit for head being recovered in the downleg. This is the worst case scenario. If you raise the discharge point to 10 m above the pond then the level in the downleg might back up the same 10 m and the pump would not "see" the extra pressure caused by the raised discharge point.

You should install an air-release/vacuum-breaker valve at the top of the dyke to enable the maximum recovery of pressure in the downleg after the dyke. If necessary, do a search here and on the web in general for "slack flow" to learn more about these concepts.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[1503-44]

If you size your pump for 20+m head and have turbulent flow in your pipe, why do you need an air release vacuum breaker valve? Any air will be forced out the end of the pipe. A vacuum breaker is only needed if your pipe wall is too thin to withstand external pressure.

 

[LittleInch]

Ed. How long is this pipe and where along this pipe is this 20m hunk?

Then we can all stop theorising and work out if what you're doing is realistic or not.

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

 

[1503-44]

What theory is needed?

Pump water up a hill.
Pump at 0m elv and a pipe going over a dike.
If outlet elv is 0, there is no net head, +20-20 =0m and flow is max.
If outlet is at +10m, net head = +20-10 = 10m, flow is less than max.
If outlet is at +20, net need = +20-0 = 20m, flow is even lesser.
If the pump does not have 20m + pipe friction head, water will not go over the dike.
If you get the water up, it flows down by itself.
Doesn't get any simpler than that.

I do the same thing every other day with my garden hose and I don't need vacuum breakers, or
air release valves. Ed is going to have to have a very big story attached to get any more complex than that.

 

[katmar]

@1503-44 The air will not necessarily be forced out. It all depends on the flow rate and pipe size. The vacuum breaker has little to do with protecting the pipe from external pressure. The breathing of air in and out has to do with the control of slack flow. I know you have written here about slack flow in the past (even if it was under a prior pseudonym). Why do you ask this question as though you do not understand the problem?

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[katmar]

Further to LittleInch's questions. We need to know the flow rate and the pipe size if we are going to properly understand this problem.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[1503-44]

Katmar, please. If internal pressure drops below liquid vapor pressure, you have slack flow. A vacuum breaker allows air in. So now you still have slack flow, but instead of liquid and liquid vapor alone, you have liquid, liquid vapor plus air and the pressure at the valve increased to 0psig. What have you accomplished? Yes, No vacuum, but perhaps even faster velocities downslope, but I rather doubt that. Vapor forms nearly as fast, or even faster than the valve can let air in. You have just made a mixture in your pipe. Good thing you apparently do not do any oil pipeline design. If the pipe will handle external pressure differential, vacuum valves are not required. Google your topic and read the return links.

LittleInch, how many air valves and vacuum breakers have you seen on an oil pipeline?

Why has this become a big data problem?
Ed explained that he has less flow when his outlet elevation increases.
ED THAT IS A TOTALLY EXPECTED RESULT.
Increasing resistance always leads to lesser flow.
Do you have another question?

 

[katmar]

Indeed, I have never seen an air vent on an oil pipeline and I doubt whether LittleInch has ever seen one either. But who said this is an oil pipeline? An open discharge into a pond sounds a lot more like water than oil.

Designing oil pipelines and water pipelines are two entirely separate disciplines. Water and oil have different densities, viscosities, vapor pressures, commercial values, flammabilities and toxicities. Just because your favorite tool is a hammer it does not mean that every problem is a nail. You have to use the right tool for the job. In oil pipelines it is common to overcome the boiling problems caused by slack flow by introducing significant back pressures. In water pipelines this would just be a waste of energy and the problem is simply, reliably and safely solved with a combination air release/vacuum breaker valve placed at the high point.

I tried Google as you recommended and the very first entry explained why air release and vacuum breakers are required. Have a look at

https://help.dripdepot.com/support/...-vent-vacuum-relief-in-your-irrigation-system

Scrolling down the search results page showed MANY similar results. What I have recommended is common practice in water transport but maybe you don't see it in oil pipelines.

It is important to note that the OP did not say that when he increased the outlet elevation the flow decreased. He said that when he increased the elevation of the outlet in the computer model the calculated flow rate decreased. These two statements are not the same. He has every right, and indeed a duty, to question the output of the program.

Unfortunately it is the nature of engineering problems that you need (relevant) information to be able to solve them. LittleInch's and my signature taglines both say this in different words:
LittleInch - "Remember - More details = better answers"
katmar - "An undefined problem has an infinite number of solutions"

Using the right data is so much better than making wild assumptions, although the way so many of the queries are formulated here on Eng-Tips a fair bit of assuming is inevitable.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[LittleInch]

We don't know enough to say whether this is a slack flow issue or not.

It depends on the hydraulic gradient and where along the line this 20m hump appears.

Given that the OP said when he puts in 10m as the end point that the flow goes down tells me that this is not slack flow and hence this hump is not relevant in this case as the pressure at that high point is > 0barg

Water pipe design is indeed a bit different and tends to be at much lower pressures where even the 1 bar difference that slack flow makes is relevant.

But in the main for any type of pipeline I try to avoid slack flow as if you control the back pressure to get the pressure as low as possible at the high point the difference in flow rate to that compared to slack flow is often very low. It is just human nature to think that if I keep opening this valve I will get more and more water out. But if you have a large hill in the way you don't.


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

 

[katmar]

Given that the OP said when he puts in 10m as the end point that the flow goes down tells me that this is not slack flow and hence this hump is not relevant in this case as the pressure at that high point is > 0barg

No, what it tells us is that the computer program is not interpreting this as slack flow. We need the additional information that we have listed so that we can confirm whether or not this is true.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[1503-44]

Katmar, the ONLY thing you said is better data is better. The rest is rubbish.

Liquid is liquid. Yes, properties vary, but they all have density, viscosity, vapor pressures, commercial value, flammabilities and toxicity, of which only the first 3 are relevant to flow. We carefully account for the varying properties of all liquids when we accurately model them, including wax content and variation of all their properties with pressure and temperature and even time, as it moves along thousands of kilometers of pipeline. Reasonably pure water is probably the easiest and most well studied liquid to model. Try a hot waxy crude, heavy bitumen/water emulsion, or a molton sulfer pipeline, or better still a multiproduct pipeline with gasoline, diesel, jet fuel and crude batches moving along the pipeline all at the same time.

The specific link you point to now is absolute manufacturer's propaganda. They make them, so of course they will tell you that you must have one or, better a thousand of them.

Why is there cause to question the output of a program that agrees with a reality that all those with experience can to so attest. I have 30 years experience modeling flow of many diverse fluids, many with extreme properties using one of the best pipeline simulation programs ever developed, including transient thermal and transient hydraulic effects. I am very familiar with how these analysis tools work. I know very well what to expect when the elevations of pipeline outlets are increased or decreased, free flowing or backpressured. I have designed and operated pipelines over 1000 miles long. I have designed and operated pipelines with over 3000m elevation differences, going up, coming down and both up and down. I have designed several pipelines each one crossing 3 mountain ranges with elevations as high as 2700 to 3300m. I fully understand partially flowing pipelines in slack flow and pipelines in full flow conditions. AFT Fathom does not model slack flow to my knowledge. It might be able to loosly approximate it, if at all, but I do not use that program. Please do not pretend to lecture me in the finer points of pipeline design any further. If I have something I think I need to know more about, I will be the first to admit it and I will ask you for help if you are so inclined to assist, but only if LittleInch first comes up with a blank look on his face. You still have not explained any reason for your vacuum valves air release valves are necessary, not that I need to hear any more mfgr propaganda.

Now if you care to lecture others here about how many vacuum valves and air release valves you must have in a couple hundred meters of water pipeline going 20m over a dike into a pond, please feel free to do so. Without any more substantial information forthcoming, all of us have said more than enough about this huge problem that nobody can possibly reach any conclusions about until we have LIDAR drawings.

Ed sorry you got mixed up in a passing contest. I was just trying to answer your simple question within the information limits that you left us with. I get frustrated when people think they can't give a simple answer to a simple question. Sorry. I do hope you got the answer you needed. If you do need more help, of course more data will be necessary, but until then ... I'm sure you understand that I can find better things to do than argue with katmar all day.

 

[katmar]

Well, 1503-44 clearly doesn't like his 30 years of experience questioned so I won't respond to his calling my advice "rubbish".

We can wait and see if we get the necessary information from the OP to come to a proper conclusion. Then we will know where this problem fits in between 1503-44's garden hosepipe and his multi-thousand km wax-saturated oil pipeline.

It has happened too often in the past that I have spent valuable time explaining my proposals, only to find that the OP never even bothers to come back and read what we have written. If he comes back and we get the necessary information then I will discuss whether using breather valves in his application will be beneficial.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[1503-44]

With the minimum info the OP provided, how can you possibly even think that ANY VALVE IS REQUIRED? Ever watched Gold Rush? Pump, pipe goes up, pipe goes down, water comes out, no valves required.