Determining volume of fluid dispersed from a broken pipe

[nathan]

I have the pipe size, flow rate, and a duration.

How can I figure out the amount of fluid that was released from the pipe?

 

[Q_Goest]

You have the pressure at the outlet of the pipe (ie: ambient pressure). If you have the geometry of the pipe back to some point where you also have the pressure, you can calculate the flow rate from standard pipe flow equations. All you need is two pressures and the geometry between them (and of course the type of fluid, temperature, etc... ).

 

[SteamKing]

If you have the flow rate and duration, then the quantity discharged is the product of these two figures.

 

[nathan]

Fluid = Water
Pressure = 50 psi
Diameter= 2''

The pipe didnt completely rupture, but was leaking from the threading of a 2'' valve. Any thought for determining the head loss. Assuming the valve threading thickness of around 1/254 (thickness of a piece of paper)

so I guess the area of the outlet would be around

circumfrence = 2''*3.14 = 6.28 ''
Thickness = 1/254 = 0.004''

Area = L*W = 6.28*0.004 = 0.025 in^2 (this would represent the area of the outlet)

any thoughts?

 

[Q_Goest]

Is this pipe thread?

 

[nathan]

Yes.

 

[nathan]

I was trying to come up with an area for the leaking threads. I know this is not exact, but I felt it was a good conservative estimation for what is going on. The leak is from the threads.

Perhaps I am on the wrong path.

 

[Q_Goest]

Tough one... I suppose you're trying to work on an RCA? (Root cause analysis or equivalent) Perhaps a bit more background would help.

Leaking threads are not generally modeled as being a 'gap' that exists, 360 degrees around the pipe fitting. I think that's what you were trying to suggest, that there was a 0.004" gap all the way around the circumference of this pipe and there was a flow, parallel to the axis of the pipe down this annular gap. I don't think that model would be realistic.

In my experience, leakage through pipe threads is thought of as following a spiral path that follows the thread. However, that flow path isn't going to be uniform. Perhaps you could model the leakage path as a pipe with an ID of .004" and a length equivalent to the length of this spiral around the threads. I think that might give you a more accurate model. Still, the 0.004" dimension is completely arbitrary and that value alone could be off by up to a factor of 5 to 10. You might also consider that the spiral leak path is more like a very thin rectangle with dimensions of a few thousandths of an inch times whatever the thread depth is. You could look at it a lot of different ways I suppose.

I don't personally think you can do a reasonably accurate analysis of this in any event. I've been an engineer for over 25 years and never heard of anyone trying to do what you're suggesting. If it helps you might be able to put an upper limit on the leakage rate by doing some kind of analysis like this but it would only be a very small part of the overall evidence for an RCA.