Pipeline pump station pipe burst modelling

James3849

Hi all,


I am currently working on a water pipeline project and am investigating a pipe burst scenario at a pump station with the aim of producing a flow rate entering the pump room in the event of a burst.

For simplification and what has been used previously on similar jobs is to model the burst as an orrice of a certain size dependant on how large the mainline pipe is.

Being a bi-directional pipeline, this pump station could experience either gravity flow dwon from a reservoir, or pumped flow from another pump station.

My question has to do with modelling the head loss accross the orifice. The pipe under investigation is a DN600 ductile iron pipe, to be investigated with a DN250 orifice in a burst scenario. Previously the orifice as been modelled as a minor loss with a K factor of approximately 2.7. But I am questioning the accuracy of this as I cannot find reference to such a value.

Can anyone provide any justifcation for this assumption. I will be happy to provide any additional information if needed.

tyroman

If your K variable is the discharge coefficient for an orifice, it should never be greater than 1.0

The following thread may help;
http://www.physicsforums.com/showthread.php?t=361268

The orifice flow formula used there is from Cameron Hydraulic Data book.

You say you are "modelling the head loss accross the orifice" and have "the aim of producing (calculating?) a flow rate entering the pump room in the event of a burst".

You will need to establish some assumptions (worst case?) to calculate the flow rate.
Case 1.) local pumps operating
Assume the burst is downstream of the pump(s) at the station
Assume the pump(s) continue operating (check NPSHr at runout)
Assume the pipe failure is complete? partial?
Case 2.) remote pumps operating
Assume the pipe failure is complete? partial?

Depending on your system and assumptions, the burst flow rate becomes a matter of orifice flow calculations at the operating point of your pump(s).

If the "head loss" you refer to is the differential head across the orifice, then that would be the difference between the system head at the burst location and atmospheric... for two cases; 1.) local pump(s) operating and 2.) remote pump(s) operating.
.

FredGarvin

The OP is just treating the orifice as a minor loss. The K value of 2.7 is the frictional loss term, not the Cd.

In the ASME specs, the method of calculating the expected delta P is stated as

[tex]\frac{\sqrt{1-\beta^4}-C \beta^2}{\sqrt{1-\beta^4}+C \beta^2} \Delta P[/tex]

Where
[tex]C[/tex] = Discharge coefficient
[tex]\beta[/tex] = Beta ratio (d2/d1)