Pipeline Head vs Pump Shut off Pressure 1

[Tang49]

Good afternoon all,

I'd like your help to understand some hydraulics.
Let's consider a centrifugal pump for clear water, installed at z=0m, with a 10 km discharge pipeline, pumping into a tank at z=100m, along a linear profile.
Let say that the losses are roughly 50m, hence a TDH of 150m under normal conditions for a given flow. Let's consider the shutoff pressure at Q=0 at around 120% of TDH, that's 180m.

Without considering any transient, what would the internal pressure of the pipe if a valve is closed at 5km or 10km ?

My understanding would be, for example at pk 5km (z=50m) 180m-50m = 120m ; and at pk=10km (z=100m) 180-100=80m. In other words, the piezometric line is flat at the shutoff pressure along the pipe profile while the valve is closed, i.e. the fluid is laminated = the pump cannot provide more "lifting energy" than Pshut-off.
Therefore, looking at this parameter only (no transient), the pressure rating of the pipeline should satisfy these pressure, i.e. derating of the thickness along the profile.

The sector concerned being waterworks, often being less stringent than petroleum or equivalent industries.

Thank you very much for your help.
And thanks to this forum, mostly ending here when looking for engineering details..

 

[LittleInch]

That looks good to me.

Yes you can change the design pressure due to increase in height as you go along a pipeline in direction of flow. It's sometimes called "telescoping".

However it is rarely worth the hassle of having different wall thickness at different locations and making sure you use the right pipe in the right place.

The bit lots of people forget is that if the pipe ends up lower or has a low point lower than the pump you need to increase the design pressure compared to the start.

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

 

[TugboatEng]

With now flow there is no friction loss. The distance from the pump is irrelevant. The head at any point in the pipe will be the shut-off head minus the change in elevation.

Provided the later sections of pipe are properly protected by their own relief valves try can be rated closer to normal operating pressure as opposed to shut-off head.

 

[1503-44]

You can design the pipe for pump maximum discharge head - elevation head at any point on the pipeline and do so without providing a pressure relief valve, as no higher pressure is possible to attain. The Baku, Azerbijan to Ceyhan, Turkey (BTC Pipeline) is 40" diameter and used five different wall thicknesses as pressure drops during the climb over two mountain ranges in Turkey, both reaching altitudes of over 2000m. As LittleInch says, keeping the different wall thicknesses from being welded in at the wrong elevations required a lot of concentration and considerable logistical effort. Try to keep to a minimum number of wall thicknesses as possible while still getting some economy. The moving of pipe to proper locations has a hidden cost. Don't think that it's all money saved.

A black swan to a turkey is a white swan to the butcher ... and to Boeing.

 

[georgeverghese]

What is the max possible suction head at the pump? Add that to the total dead head of 180m to get to the max operating pressure of the line. Suggest adding another 10% to this MOP to get to the the minimum required MAWP.

 

[1503-44]

Why the extra 10%. In a long pipeline that could easily add 10% to the cost of the entire project. It is not by any means small potatoes. Fortunately no pipeline project manager will let him do that.

A black swan to a turkey is a white swan to the butcher ... and to Boeing.

 

[LittleInch]

The pipeline codes specifically allow (other than B31.8) a 10% allowance for occasional overpressure above design pressure. Unlike B 31.3 it doesn't define "occasional"....

But I do agree that you need to use the highest possible inlet pressure/head (e.g. full tank) into the pump plus the no flow head to get your max pump pressure / head.


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

 

[Tang49]

Thank you all for your feedback.
We sometimes have to work with a range of dia from 800 to 2000mm, and I have seen the telescoping you are describing in some studies, but not in real life.
Besides, the pressure rating in water works is rarely above 25 or 40 bars, so the rule of t > D / 120 often dedicates the minimum thickness, as I understood.

 

[LittleInch]

Well I've been designing pipelines for longer than I care to think too much about and can number the pipelines which I have done this on one hand.

One issue on longer lines is that this stops you from increasing flow on liquid lines by putting in a booster pump in the future as the pipe is then not thick enough.

On a recent long pipeline through hilly ground, I used an initial guide that wall thickness / design pressure shouldn't change unless there was more than 200m difference in elevation for longer than 20km. Now that was just an initial cut and in the end we went quite a bit further and ended with only three different wall thicknesses down from 5 or 6. The feedback from those with experience was that the logistics issues were severe if you had too many changes and the risk of putting the wrong pipe in the wrong place increases exponentially with increasing numbers of wt changes. You often need to introduce an external colour coding system to easily check.

I think you mean D/t > 120? That's pretty thin, but water lines at large diam are a different beast to hydrocarbon lines.

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

 

[1503-44]

D/t should be < 100 (B31.4- 403.2.5)

to avoid special measures being taken to prevent ovalization during transport, handling and installation.

A black swan to a turkey is a white swan to the butcher ... and to Boeing.

 

[LittleInch]

Mr 44 - he's talking water. They use paper thin pipes....

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

 

[1503-44]

Yeah I know that, which is why I posted the B31.4 reference.
It's also that the problem does not exactly get washed away with water, right?

A black swan to a turkey is a white swan to the butcher ... and to Boeing.