Pressure Surge

[KGBCHEM]

Dear All,

I’ve got a liquid pipeline fed by centrifugal pumps. Following a surge event (outlet ESD valve rapidly closed, pumps left running), is it possible for the line to have a final or settle-out pressure greater than the pump shut-off pressure? If the surge pressure wave is still bouncing up and down the pipe after it has packed and reached pump shut-off, does the pressure eventually attenuate or can it remain - giving the high pressure?

 

[Homayun]

Yes, if the ESD closes rapidly, surge pressures greater than the pipe design pressure can be developed. There are methods to calculate the surge pressure. I know that Flowmaster software can be used to calculate the surge pressures.

But what about pump minimum flow recycle line? Is there a minimum flow recylce line installed? If yes, then surge may not be of that problem because when the ESD closes, the the minimum flow recycle line should open so that no surge will develop.

 

[KGBCHEM]

Thanks for the reply Homayun. I understand that the instantaneous pressure can easily be above design pressure (from simple Joukowski eqn), but my question is that once the pressure oscillations have died down, can the final or settle-out pressure be greater than the pump shut-off pressure? I`m wondering if the surge waves eventually attenuate/dissipate as they reflect up and down the pipe, leaving a pressure that is same as pump shut-off pressure? Any thoughts?

 

[SNORGY]

KGBCHEM:

I think you would need all of:

(1) a pretty high flow velocity
(2) a substantially long piping system, and
(3) a very rapid valve closure time

in combination in order to impart enough fluid inertia to "overpack" (compress) the fluid parcel between block points enough to raise the settle-out pressure (the sustained pressure reached after transient oscillations have damped out) significantly above the pump-shut-off pressure.

I am simply making a judgement here; I have not run a simulation to prove it, but it appears likely to me that unless you had something like a swing check valve "slam" you probably would never get a rapid enough closure time out of an actuated valve to produce the effect that you are asking about. The speed of sound in the fluid is likely such that in order to produce this effect you would already be into larger line sizes of such lengths that the corresponding valves would be too big to be able to close that fast.

I am not saying it is impossible (because it *is* possible), but I am saying that it would strike me as unusual unless you are dealing with "pipeline-scale" sizes and lengths.

Regards,

SNORGY.

 

[Homayun]

KGBCHEM,
I am of course not an expert in this area, but I would imagine if by the time the osilations have died out, then the pressure spike must have disappeared too. Remembe, it is this rapid osilation and the liquid wave that causes the pressure spike.
One more thing, don't forget that in this case the pump shut-off pressure is probably the design pressure of your pipe.

 

[jawilke]

I don't have a copy of your piping schematic, but is the ESD valve downstream of a check valve? Could be the high pressure is trapped them.