Data-driven depth calibration for distributed acoustic sensing

doi:10.1190/tle35070610.1

Data-driven depth calibration for distributed acoustic sensing

2016-07-01

0

PRE

AI

DOI：10.1190/tle35070610.1

原文链接

原文求助

摘要

En 中文

Distributed acoustic sensing (DAS) can make use of an ordinary telecom fiber as a continuous array of acoustic sensors to acquire, for example, downhole seismic. One challenge when using DAS is assigning correct depth to a given part of the DAS data record. Usually, depth is assigned by linear interpolation between reference points in the record for which the depth is known. We present a case where this approach did not work satisfactorily due to a fiber accumulation of unknown length between the reference points in the top and bottom of the well. The data we discuss were acquired using Silixa's iDAS technology retrofitted to previously installed fiber-optic cables. During installation, the cables were clamped to the tubing at each coupling between tubing sections. We found a pattern in the iDAS data that could be related to the clamping points. The pattern is particularly clear for frequencies in the 100–200 Hz band and may be caused by eigenmode vibrations set up by the clamping points. Based on this observation, a method for data-driven depth calibration was developed by matching the observed pattern to the clamp positions known from the tubing-tally information. An algorithm was designed to do the optimization. With regularly spaced clamps, several solutions are output, and some knowledge of the approximate depth is needed to pick the appropriate one. Our data-driven depth calibration was successfully applied to iDAS data from six wells.

Data-driven depth calibration for distributed acoustic sensing

摘要

AI总结

期刊

机构

引用论文