Accelerometer for leak detection

[mssaleh]

Hi,
I'm going to measure the acceleration from two points on steel pipeline, and build a leak locator by correlation in the first step to develop that for farther signal processing researches next.
I want to know what kind of accelerometers usually used for conventional leak detection by means of correlation and what is the required specifications (Structure\Sensitivity\Freq Range).
I found some vague information toward high sensitivity accelerometer(15V/g) with approx 10Hz-5kHz freq range.
Please advise me with some model and brand name.
Thanks alot.

 

[TPL]

Unless you have a unique application, why don't you just use ultrasoound for leak detection?

 

[mssaleh]

Hi,

Thanks for your reply, actually large diameter buried pipeline inspection is the aim, up to my knowledge there is no ultrasonic leak detection technique for large diameter pipelines for range of 200m. the long range inspection is the main advantage of acoustic leak detection.
Our mission is to find leaks during hydrotest of 58 inch steel pipeline and the water pressure is 1050 psi, so it seems just a very small leakage can be a good source of noise.

Please let me know, If there any special ultrasonic technique for leak detection in large diameter pipeline.

Thanks again for your reply.

 

[Twoballcane]

Ok, this is new to me, how do you detect a leak with accelerometers? Are you comparing vibration, noise, displacments...etc?

Tobalcane
"If you avoid failure, you also avoid success."

 

[Strong]

This accelerometer might meet your requirements:

http://www.wilcoxon.com/prodpdf/996LD spec (98547).pdf

Wilcoxon Model 996LD High sensitivity, leak detection accelerometer
Dynamic Sensitivity........... 12 V/g, ±3 dB
Frequency response, ±3 dB..... 10 - 4,000 Hz
Resonance frequency, nominal...... 14 kHz
Acceleration range, nominal..... 0.10 g peak
Shock limit......... 1,000 g

I do not use this product or represent the company.

Walt

 

[mssaleh]

Hi,

Thank you Strong fot your suggestion, I have seen this accelerometer before. I think (and don't figure on) other detailed specifications are critical too, like as "residual noise".Please let me know if you have any idea about importance of residual noise in the precise measurements.

Acoustic Leak Detection has been developed since 1970 for leak detection and localization of water distribution network's leaks. The main concept is to use the cross correlation of special quantity measured from two points on the pipeline (witch bracket the leak)to determine the time delay and phase shift and then the relative distance of noise source. In this method measurement belong to acoustic surface waves released by leaks. The correlation can be done on any quantity but the pressure (by hydrophones) and acceleration (by accelerometer) have the best results yet. Some new methods use Modal Analysis and the professional Leak Correlator Systems use the coherent function to specify the most appropriate frequency features.

 

[Strong]

mssaleh,

"Residual noise" can be in sensor or the actual background vibrations or sounds from non-leak sources. Most leak tests are done at night and with machinery off to minimize background. Periodic noise (from pumps and compressors, etc.) can really mess up a cross-correlation analysis. Background noise will be your biggest problem! If you don't have any practical experience using this test method, then you should start by experimenting with smaller pipes in controlled situations. It is not as easy as the product brochures indicate! You might want to consider seismic velocity sensors (geophones), since they can have greater sensitivity and lower background noise than accelerometers. Good luck with 58" steel pipes over 200 meters distance!

Walt

 

[mssaleh]

Hi,

Thank you Strong(Walt) for your reply and valuable information. I have some detailed case studies for leak detection by acoustic method in the middle east that were successful for leak detection in 48 inch pipelines for pressure drop of 8psi/hour and 215m sensor's spacing with accuracy of 15m and the environmental condition was stated as desert (porous soil) and very silent surrounding, However the signal processing method was not stated as correlation.

our test bed configuration contains 2" insulated pipes and a pump with pressure of 70 psi. The field tests can be done by means of artificial leaks on the hydrotest header during leak testing, the orifices were designed according to "A reference standard for the development of acoustic emission pipeline" EPA-Document the test bed will be ready in near future.

Please let me know if you have any more information, experiences or technical sources that give more detailed information regarding to seismic velocity sensors.

 

[Strong]

Mssaleh,

It appears that you already have instruments and test setup nearly ready. It would be wise to do your own testng to find out what issues are important. There are so many variables (some uncontrollable) to leak testing that one cannot rely on another person's success or failure.

Walt

 

[SylvestreW]

fwiw, they seem to do what you are suggesting.

http://www.echologics.com/