夏文鹤, 潘硕, 孟英峰, 李永杰. 基于音频信号的气体钻井返出岩屑量监测方法研究[J]. 石油钻探技术, 2017, 45(3): 121-126. DOI: 10.11911/syztjs.201703021
引用本文: 夏文鹤, 潘硕, 孟英峰, 李永杰. 基于音频信号的气体钻井返出岩屑量监测方法研究[J]. 石油钻探技术, 2017, 45(3): 121-126. DOI: 10.11911/syztjs.201703021
XIA Wenhe, PAN Shuo, MENG Yingfeng, LI Yongjie. The Returned Cuttings Monitoring Method for Gas Drilling Based on Audio Signals[J]. Petroleum Drilling Techniques, 2017, 45(3): 121-126. DOI: 10.11911/syztjs.201703021
Citation: XIA Wenhe, PAN Shuo, MENG Yingfeng, LI Yongjie. The Returned Cuttings Monitoring Method for Gas Drilling Based on Audio Signals[J]. Petroleum Drilling Techniques, 2017, 45(3): 121-126. DOI: 10.11911/syztjs.201703021

基于音频信号的气体钻井返出岩屑量监测方法研究

The Returned Cuttings Monitoring Method for Gas Drilling Based on Audio Signals

  • 摘要: 为了监测钻井过程中的井壁坍塌、井底岩爆等井下工况信息,进行了基于音频信号的气体钻井返出岩屑量监测方法研究。该方法利用音频采集系统采集返出岩屑在排砂管中运移所产生的音频信号,根据短时能量确定声音段的起止点并计算特征参数,建立声音的特征参数库,再利用神经网络算法排除干扰声音,然后分析排砂管内不同大小岩屑的声音特征,利用动态时间弯折算法识别岩屑的大小,计算岩屑流量,进而判断气体钻井携岩状态及井下工况。双探7井现场试验结果表明,该方法对干扰声音的分类成功率达到96.8%,对不同粒径岩屑的识别率达到85.0%。研究结果表明,基于音频信号的气体钻井返出岩屑量监测方法可以监测返出岩屑流量变化情况,有效判断气体钻井的井下工况,从而降低气体钻井作业风险。

     

    Abstract: To find the best way to monitor the process of well drilling (e.g.wellbore collapse and bottom hole rock burst),the method of returned cuttings monitoring in gas drilling based on audio signals was studied.In this method,an audio acquisition system is used to collect the audio signals generated from the migration of the returned cuttings in the clearance pipe.The starting and ending points of sound segments are determined and correlatedto the short-time energy by which their characteristic parameters can be calculated.After a sound characteristic parameter database is created,interference sound is eliminated through a neural network algorithm.Then,the acoustic characteristics of cuttings with different size in the clearance pipe are analyzed.The cuttings sizes can be differentiated by using the dynamic time warping after which the flow rate of cuttings then is calculated.Accordingly,the carrying status of the cuttings and downhole working condition of gas drilling can be primarily determined.This method was tested on site in Well Shuangtan 7 and results demonstrated that the classification success rate of interference sounds reached 96.8% and the identification rate of cuttings size reached 85%.Results also indicate that this method of monitoring the amount of returned cuttings in gas drilling based on the audio signal can also be used to monitor the variation trend of cuttings flow rate and effectively determine downhole working conditions and thus reduce risk in drilling gas wells.

     

/

返回文章
返回