Abstract

Early detection and localization of downhole leaks are essential to maintain well integrity, reduce cost, and minimize downtime. New technology has been developed to detect leak locations in a well quickly and to characterize the flow profile of the leak by using an array of hydrophones. The technology uses advanced modeling and beamforming algorithm to map out the flow pattern in a 2D image within the well’s completion structure. However, during continuous logging, the leak signal may be contaminated by guided wave noises such as the road noise from the tool string, and the logging results will be compromised. This paper demonstrates a method to estimate and remove guided-wave noise to enhance the leak detection answer products.

The data from continuous logging may be contaminated with significant road noise due to equipment contacting the casing or borehole which produces Stoneley or tube waves. For single and dual hydrophone tools, additional runs may be needed to stop these tools at selected locations to record data without this contamination, but this approach prolongs the acquisition time and limits the vertical resolution.

In order to obtain depth-continuous and high-resolution leak information, an advanced array signal-processing technique has been developed to enhance the signal quality. Extensive studies on field data were conducted to extract the features and characteristics of the leak noise, even when those features overlap in time or frequency with contamination noise. The processing method employs multiple steps that analyze the hydrophone array to remove the contamination noise in the time or frequency domain, leaving the leak noise for flow and leak location analysis.

The proposed method has successfully identified high noise activity at certain depths as road noise in continuous logging data. Road noise may increase in amplitude within a limited depth due to a momentary change in logging activity. The elevated noise generated can be identified as guided-wave noise instead of a potential leak. The method can be implemented in real-time and the results will save additional rig time conducting further stationary logging at the non-leak depths. Field data results also suggest that the proposed method improves the signal-to-noise ratio of the continuous logging data significantly and delivers quality noise spectrum and leak location logs for the industry.

The proposed method has been proven to be effective in identifying and enhancing leak signals and removing contaminating signals due to guided wave noises. It has greatly enhanced the quality of the detection, resolution, and location of leaks in wellbore tubulars produced from continuous logging data. These high-quality continuous logging results will help field engineers to make more accurate decisions quickly during logging operations and could avoid costly and time-lengthy stationary logging programs.

This content is only available via PDF.
You can access this article if you purchase or spend a download.