基于DDMP-GWO融合的气侵早期监测方法  

A Method of Early Gas Kick Monitoring Based on DDMP-GWO Fusion Algorithm

作　　者：王建龙 于志强 郭云鹏 邓兵 杨洋洋 杨建永 郑锋 WANG Jianlong;YU Zhiqiang;GUO Yunpeng;DENG Bing;YANG Yangyang;YANG Jianyong;ZHENG Feng(Bohai Drilling Engineering Technology Research Institute,CNPC,Tianjin 300450;Bohai Drilling Engineering Company Limited,CNPC,Tianjin 300450;Bohai NO.1 Drilling Engineering Company,CNPC,Tianjin 300280;China Petroleum Technical Service Corporation,CNPC,Beijing 100007)

机构地区：[1]中石油渤海钻探工程技术研究院,天津300450 [2]中石油渤海钻探工程有限公司,天津300450 [3]中石油渤海钻探第一钻井工程分公司,天津300280 [4]中国石油集团油田技术服务有限公司,北京100007

出　　处：《钻井液与完井液》2024年第6期747-754,共8页Drilling Fluid & Completion Fluid

基　　金：中国石油集团公司科技项目“钻井安全控制软件”(2024ZZ46-05)。

摘　　要：气侵早期监测对于保障钻井安全至关重要,一旦监测不及时,可能造成井喷、人员伤亡等重大钻井事故。针对目前传统泥浆池增量法预警滞后性强,井下随钻监测法时效性高,但功能单一的现状,提出了一种能够同时实现气侵早期定性识别和定量解释的新方法。分析了气侵时环空流体流速变化规律,构建了气侵风险指数(Gas kick risk index,KRI),并推导了KRI与气体体积分数间的映射关系;进一步基于井下双测点压差(Difference between downhole dual measurement points pressure,DDMP),结合灰狼优化算法(Grey wolf optimization,GWO),构建了环空流体流速实时计算方法;利用气侵模拟场景,分析了气侵早期监测方法的稳定性和有效性。研究结果表明,气侵时环空流体流速整体呈增加趋势,可作为检测气侵的关键特征参数,环空气体体积分数与KRI呈线性关系。环空流体流速计算误差随着两测点距离的逐步增加呈现先减小后增大的趋势,受压力和温度测量误差的影响较小。该方法气侵预警滞后时间为13.8 min,环空气体体积分数反演误差小于10%,不仅能够较早检测到气侵,还能够为井控工艺设计提供环空含气率关键参数,而此时泥浆池增量仅为0.017 m^(3),泥浆池液位无明显变化。Measurement of mud gains in mud pits is a conventional method of detecting downhole gas kick and is still in use in present.Using this method,the detection of a gas kick sometimes remarkably lags the occurrence of the gas kick.Another gas kick detecting method is the monitoring-while-drilling method with which gas kick can be timely detected,but the functions of this method are very limited.In this study a new method with which a gas kick can be qualitatively detected in an early time and quantitatively explained is presented,the changes of the flowrates of the fluids in the annular space while a gas kick is encountered are analyzed,a gas kick risk index(KRI)is designed,and the mapping relationship between the KRI and the volume fraction of the kicked gas is derived.Based on the difference between downhole dual measurement points pressure(DDMP),using the grey wolf optimization(GWO)algorithm,a real-time method for calculating the flow velocities of the fluids in the annular space is constructed.Using a simulated gas kick scenario,the stability and effectiveness of the method for early detection of gas kick are analyzed.The study shows that when a gas kick occurs,the flowrates of the fluids in the annular space are increasing,and this can be used as a key characteristic parameter for gas kick detection.The volume fraction of the gas in the annular space has a linear relationship with KRI.Errors made in calculating the flowrate of the fluids in the annular space first decrease and then increase as the distance between the two measurement points increases,and are less affected by the errors made in pressure and temperature measurement.Using this new method,the lag time for detecting a gas kick is 13.8 min,and the inversion error of the gas volume fraction in the annular space is less than 10%.This method is not only able to detect gas kick earlier,it also provides key parameters for well control design such as the gas fraction of the fluids in the annular space at a mud gain in the mud pits of only 0.017 m^(3),a volu

关 键 词：气侵早期监测 井下双测点压差 灰狼优化算法 环空流体流速 气侵模拟 

分 类 号：TE254.2[石油与天然气工程—油气井工程]