what algorithm can be used to find leakage in underground pipes?

[Darren000]

Could someone please give me some hints?

I am working on a simulation project. how can I detect leakage in pipe based on the feekback from pressure transmitters?

Assume that outlet and inlet are fully closed, pressure should remain constant if there is no leak.

However, If there is a leak, what formula can I find out the rate of leaking based on the pressure decrease?

Is it possible to find leak if outlet and inlet are open? how? install flow meters at outlet and inlet?

Please Kindly help me. Thanks in advance.

 

[BigInch]

Discharge rate of a liquid through an orifice,
Q = Cd A (2 [ρ] [Δ]P g/144)^0.5
Q flow lbs/s
Cd discharge coefficient = 0.61
A area of orifice in2
[ρ] density of liquid pcf
[Δ]P pressure drop across the orifice
g = lbf to lbm = 32.2

If the line is closed in, internal pressure will fall in a near logrithmic manner.
You will need some means to relate volume of fluid in the pipe at any given time to the internal pressure; an equation of state for your fluid. You must assume your pipe is rigid, or you will also have to relate pipe volume to internal pressure as well.

If your line is flowing and you know pressure and flow at two places, calculate the pressure drop in your pipeline at those two places, start at the static head point above each known flowrate point, draw the pressure loss/distance energy grade lines, and where they intersect, you will find your leak.


"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek

 

[bimr]

It is not practical to detect leakage in pipes based on the output from pressure transmitters. There is little precision in the typical pressure gauge so that it is not possible to determine if there is leakage. This method is only useful if there is a catastrophic pipe failure.

Many of the leak detectors on the market are based on acoustic technologies.

 

[BigInch]

?

"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek

 

[bimr]

Acoustic leak detection technologies:

http://web.mit.edu/parmstr/Public/NRCan/nrcc48357.pdf

 

[Darren000]

Thank you BigInch and bimr for your kind responds.

This is for a simulation purpose, so the leaks are actually controlled with valves.
So what I'm trying to find out is just the idea behind this. Lets say if there are flow meters installed on both inlet and outlet, so we know the flow rate at both ends of the pipe. I can say that (OUT rate - leak rate = IN rate) right??

I'll attach a drawing of this project on Monday.
Thank You BigInch and bimr, really appreciated

 

[Darren000]

I am a Control system technician, We have to write a PLC program for this project so we can find/monitor leaks through out the pipe network, based on the values from pressure transmitters and flow meters. I have no idea why they want to do this simulation thing with so much hardcore MATH involved. Like you said, Acoustic leak detection technologies is so much better.

Anyways, Thank you guys

 

[BigInch]

As I said above,

If your line is flowing and you know pressure and flow at two places, calculate the pressure drop in your pipeline at those two places, start at the static head point above each known flowrate point, draw the pressure loss/distance energy grade lines, and where they intersect, you will find your leak.

It's more difficult when the system is closed, but certainly can and is possible and practical.



"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek

 

[Darren000]

Hi BigInch and bimr, please take a look at the attachment. This is the pipe network. Water comes in from supply and goes out from point E if no other outlet are open.

1)By closing A and C, I should be able to measure leak 1 and leak 3 if no other leaks occur right?

2)Known leak 1 and leak 3 rate, now I can open A, D,C and close B to find out other leak point on the left of A. right?

Thanks for reply.

 

[BigInch]

You have to know both flowrate and pressure at two points on any given run of pipe with a leak. One point upstream of the leak and the other point downstream of the leak.

With only one leak, you can get pretty close to where it is, however with more than one concurrent leak, the difficulty is compounded a lot. You can get an idea of where the combined leakage is; shall we say kind of where the centroid of the combined leak is.
For leak 1-3,
Close off C-D-A. With a known flowrate at C, or E and a known flowrate at B, given lengths, diameters and roughness for C-B, you should be able to find the centroid of the leak point. Calculate the head loss per unit length at both points of known flowrate. Start at the measured static head at each point and draw the slopes of those calculated head losses/unit length. Where they intersect will be the centroid of the combined leak. If the two leak points are close, that may tell you something usefull, but if they are far apart, you won't learn much without isolating one first.

Then close off that C-E-B line and do the same for C-D-A.


"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek