6 replies to this topic

[R NESAMANI]

We have a fire water network. Total water volume of the network is 410 m3.
Due to leakage in the pipe, Pressure of the network decreses, such that pr.drop of 1 Kg/Cm2 occurs in 40 minutes. Can we find out the quantity of water lost in 40 minutes from this pr.drop ?

[tarafdar]

Hi,

Water is an incompressible fluid, its volume doesn't change mutch with pressure.So calculating water loss due to reduction in pressure will not be accurate.Still we can do an approximate calculation.First we assume that during the mentioned 40 minutes there is no temp. change .
In case of water for 1Kg/cm2 change in pressure ,there will be 0.05Kg/m3 change in density(approx).There will be a change in volume by 0.00005 th of original volume.For a 410 m3 network, water loss is only 20 lit to get 1Kg/cm2 pressure drop.For temp. change volume change is significant.

Thanks

Edited by tarafdar, 05 October 2011 - 09:55 AM.

[R NESAMANI]

Thank you Tarafdar.
How do you say that for water 1 kg/cm2 pr. drop change in density is 0.05 kg/cm2 ....?
Can you give me any formulae or link or table for this...?

[tarafdar]

I get the information from http://www.engineeri...sure-d_309.html .

[kkala]

Mech E: As we are doing the hydrostatic test of this drum,
we inject water into it through this manual pump. Pressure
is increasing, what is the cause of it?
Chem E: More and more water comes into the constant
volume drum, so water is compressed and pressure goes up.
Mech E: Water is incompressible, pressure is increasing
because vessel volume increases and the vessel steel is
internally "elongated".
Two practicing students, summer 1969.
My friend was right, because main result of injected water into a vessel is to increase its volume, which brings increase in pressure. But tarafdar is also right, because main result of injected water into a pipe (isolated) is to slightly decrease its unit mass volume, which also brings increase in pressure.
In the inverse sense: When you take a small amount of water out of the vessel, vessel volume is slightly decreased and this is the main reason of resulting lower pressure. Small quantity of water taken out of a pipe (isolated) results in slightly lower water specific volume, main reason for lower pressure.
Steel stresses govern the vessel case, water compressibility the pipe case, due to different size of diameter in the two cases.
Attached <elasticity.xls> estimates water loss for 1 Bar pressure decrease, according to my understanding. Both steel stress and water compressibility are taken into account. Two specific cases are presented, one of 14" Sch 40 pipeline and another of 5 m diameter bullet, both of 410 m3 volume. Water loss for 1 Bar pressure decrease is estimated at 26 lt for the pipeline and 68 lt for the bullet. Second document attached, <pV.pdf> (Pressure vessels by D. Roylance, MIT), is merely relevant info on pressure vessels.
A pipeline can be also considered as a long pressure vessel. The query concerns only a pipeline, the example of a vessel is also presented to show the difference. The two cases are quite similar.
Hope the above can be helpful, any comment welcomed.

Attached Files

•  elasticity.xls   27KB   92 downloads
•  pv.pdf   191.37KB   59 downloads

[R NESAMANI]

Thank you very much Mr. Kkala.
Its very helpfull for me. I can conclude that we are losing approx. 20 lit/40 min of water from the network.

[kkala]

Yes, the water "escape" is approximately 20 lt per 1 bar pressure loss, as also estimated by tarafdar. Pipe diameter has some (rather minor) contribution to the result, mainly depending on water compressibility (inverse of bulk modulus).
More precisely, in case of 14" Sch40 steel pipe (of such a length that internal volume is 410 m3), "elasticity.xls" estimates water escape at 26 lt for pressure loss = 1 bar, of which ~19 lt due to water decompression (by 1 bar) and ~7 lt due to (small) reduction of steel pipe diameter.
The latter ~7 lt has some dependence on pipe diameter size. You can place actual external pipe diameter and thickness into "elasticity.xls" and see the estimated result.