When operators are faced with well-integrity problems, a variety of methods
may be used to detect the source of annular communication. Methods for
detecting downhole leak points include spinners, temperature logs, downhole
cameras, thermal-decay logs, and noise logs. However, many of these methods are
ineffective when dealing with very small leaks and can result in collected data
that require a significant amount of logging finesse to interpret.
Ultrasonic listening devices have been used for a number of years to detect
leak sources effectively in surface production equipment. Ultrasonic energy has
some properties that, when compared to audible-frequency energy, make it ideal
for accurate leak detection (Beranek 1972; Povey 1997; Evans and Bass 1972).
Like audible-frequency energy, ultrasonic energy can pass through steel.
However, ultrasonic energy propagates relatively short distances through fluids
when compared to equal-energy audible-frequency sound. Thus, when an ultrasonic
signal of this nature is detected, the detection tool will be in close
proximity to the energy source.
On this premise, an ultrasonic leak-detection tool was developed for
downhole applications to take advantage of the unique properties of
ultrasonic-energy propagation through various media. Data-acquisition equipment
and filtering algorithms were developed to allow continuous logging conveyed on
standard electric line at common logging speeds. Continuous logging has proved
to be significantly more efficient in locating anomalies than static logging
techniques commonly used in noise-logging operations.
During development, the tool was shown to be effective in locating leaks as
small as 0.026 gal/min with an accuracy of 3 ft in production tubing, casing,
and other pressure-containing completion equipment. Leaks also have been
detected through multiple strings of tubing and casing. The tool has proved to
be effective in locating leaks that other diagnostic methods were unable to
© 2009. Society of Petroleum Engineers
View full textPDF
- Original manuscript received:
27 June 2006
- Meeting paper published:
24 September 2006
- Revised manuscript received:
22 October 2008
- Manuscript approved:
30 October 2008
- Published online:
1 May 2009
- Version of record:
1 May 2009