摘要
地层压力的准确预测是油气资源高效开采的前提,目前已形成了多种地层压力随钻监测方法,并取得了广泛的应用。通过测井数据开展地层压力预测是一种常用的地层压力监测方法,然而目前通过该方法求取地层压力过程中还存在伊顿常数等关键参数难以准确获取的问题,导致在部分区域内误差较大。该文基于声波测井地层压力评价方法及密度测井地层压力评价方法,结合广义回归神经网络模型,构建了一种新的基于多随钻测井数据的地层压力监测方法,通过多测井数据计算得到的地层压力与实测地层压力数据间的耦合影响,对地层压力预测值实现快速迭代优化。研究成果表明相较于单独通过声波测井数据及密度测井数据得到的地层压力评价结果,通过该文方法能够更快速地得到地层压力预测值,且误差更小。与dc指数法及岩石强度法等目前常用的地层压力预测方法相比,通过该文方法得到的地层压力评价结果误差最小。研究成果通过浙江油田ZT区块相关数据进行了对比验证,在该地区具有较好的适用性。
The accurate prediction of formation pressure is the prerequisite for efficiently exploiting oil and gas resources—various methods for monitoring formation pressure while drilling have been developed and widely applied.Predicting formation pressure through logging data is commonly used to monitor formation pressure.However,due to the difficulty in accurately obtaining key parameters such as Eaton constants,the method currently fails to meet the demand for high-precision prediction of formation pressure in some areas.Based on the acoustic logging formation pressure evaluation method and the density logging formation pressure evaluation method,combined with the generalized regression neural network model,this paper constructs a new formation pressure monitoring method based on multiple logging data while drilling.The coupling effect between the formation pressure calculated from multiple logging data and the measured formation pressure data is used to optimize the formation pressure prediction value rapidly.The research results show that compared to the formation pressure evaluation results obtained through acoustic and density logging data alone,the formation pressure prediction value obtained through this method can be obtained more quickly with more minor errors.Compared with the commonly used formation pressure prediction methods,such as the dc index and rock strength techniques,this paper’s formation pressure evaluation results have the slightest error.The research results were verified by comparing relevant data from the ZT block of Zhejiang Oilfield and have good applicability in this region.
作者
王铁成
卫乾
熊伟
王强强
张宁俊
史亚会
杨兵
王新华
WANG Tiecheng;WEI Qian;XIONG Wei;WANG Qiangqiang;ZHANG Ningjun;SHI Yahui;YANG Bing;WANG Xinhua(Kunlun Shuzhi Technology Co.,Ltd.,Beijing 100010,China;Tracy Energy Technology Hangzhou Co.,Ltd.,Hangzhou 310000,China)
出处
《应用声学》
2025年第3期744-750,共7页
Journal of Applied Acoustics
基金
中国石油天然气集团有限公司前瞻性基础性战略性技术攻关课题“勘探开发一体化协同研究软件研发”(2021DJ7001)。
关键词
地层压力
随钻监测
随钻声波测井
随钻密度测井
神经网络
Formation pressure
Monitoring while drilling
Acoustic logging while drilling
Density logging while drilling
Neural network
作者简介
王铁成(1976-),男,河北涿州人,硕士,高级工程师,研究方向:油气工业互联网及油田数字化、智能化;通信作者:王强强,E-mail:wangqiangqiang@tracyenergy.cn。
