Sonic well logging for alteration detection

    公开(公告)号:US06526354B2

    公开(公告)日:2003-02-25

    申请号:US09741573

    申请日:2001-02-01

    IPC分类号: G01V140

    CPC分类号: G01V1/44

    摘要: A method for determining alteration of a region of an earth formation surrounding an earth borehole, comprising the steps of providing a logging device that is moveable through the borehole; transmitting sonic energy into the formation and receiving, at a plurality of transmitter-to-receiver spacings, sonic energy that has traveled through the formation, and producing signals representative of the received sonic energy for the plurality of transmitter-to-receiver spacings; determining sonic transit times and differential transit times for the respective transmitter-to-receiver spacings; deriving a test statistic from the differential transit times; and determining the presence of alteration of a region of the formations from the test statistic. An associated apparatus for carrying out the method is also described.

    Methods for downhole waveform tracking and sonic labeling
    3.
    发明授权
    Methods for downhole waveform tracking and sonic labeling 有权
    井下波形跟踪和声波标记方法

    公开(公告)号:US06625541B1

    公开(公告)日:2003-09-23

    申请号:US09591405

    申请日:2000-06-12

    IPC分类号: G01V100

    CPC分类号: G01V1/48 G01V1/288

    摘要: Methods for downhole waveform tracking and sonic labeling employ “tracking algorithms” and Bayesian analysis to classify STC waveforms. More particularly, according to the tracking part of the invention, a probability model is built to distinguish true “arrivals” (e.g. compressional, shear, etc.) from “false alarms” (e.g. noise) before the arrivals are classified. The probability model maps the “continuity” of tracks (slowness/time over depth) and is used to determine whether sequences of measurements are sufficiently “continuous” to be classified as tracks. The probability model is used to evaluate the likelihood of the data in various possible classifications (hypotheses). Prior and posterior probabilities are constructed for each track using the tracking algorithm. The posterior probabilities are computed sequentially and recursively, updating the probabilities after each measurement frame at depth k is acquired. The classification part of the invention determines the hypotheses with the maximum “posterior probability” according to Bayes' Theorem.

    摘要翻译: 井下波形跟踪和声波标记的方法采用“跟踪算法”和贝叶斯分析来分类STC波形。 更具体地,根据本发明的跟踪部分,构建概率模型以在到达之前分类真实的“到达”(例如压缩,剪切等)与“假警报”(例如噪声)。 概率模型映射轨道的“连续性”(慢度/深度上的时间),并且用于确定测量序列是否足够“连续”以被分类为轨道。 概率模型用于评估数据在各种可能分类(假设)中的可能性。 使用跟踪算法为每个轨道构建先验和后验概率。 顺序和递归地计算后验概率,在深度k获取每个测量帧之后更新概率。 本发明的分类部分根据贝叶斯定理确定具有最大“后验概率”的假设。