基于Rosenbluthe改进方法的地层坍塌压力不确定性分析

Uncertainty analysis of borehole collapse pressure based on an improved Rosenbluthe method

1.中国石油大学(华东)储运与建筑工程学院,山东青岛266580;
2.青岛理工大学理学院,山东青岛266520;
3.南京理工大学基础前沿交叉中心,江苏江阴214443

1. College of Pipeline and Civil Engineering,China University of Petroleum (East China),Qingdao 266580,China;
2. College of Science,Qingdao Technology University,Qingdao 266520,China;
3. Interdisciplinary Center for Fundamental and Frontier Sciences,Nanjing University of Science and Technology,Jiangyin 214443,China

密度、声波时差、自然伽马等测井参数的不确定性使得基于测井数据的地层坍塌压力预测存在不确定性。基于此,首先确定了密度、声波时差、自然伽马等测井参数的不确定性特征;其次引入改进的Rosenbluthe方法,理论推导了岩石力学参数的不确定性表达式,并结合水平地应力和地层压力的不确定性表达式,提出了井眼坍塌压力的不确定性分析方法,最终得到井眼坍塌压力的概率分布情况。以SY1井为例,利用不确定性分析方法和蒙特卡罗方法分别预测了井眼坍塌压力的概率分布情况并对实际钻井过程中井壁失稳风险进行了预测,结果表明不确定性分析方法的预测结果与蒙特卡罗方法吻合度超过了95%,不确定性分析方法预测井壁失稳风险率为68.4%,蒙特卡罗方法预测井壁失稳风险率为67.4%,实际钻井过程中该井深位置发生了井壁坍塌,预测结果与实际情况相符,且所提出的不确定性分析方法预测的井壁失稳风险率略大于蒙特卡罗方法的井壁失稳风险率,对现场施工具有重要的参考价值。

The uncertainties associated with log parameters,such as density,acoustic time,and natural gamma,lead to inaccurate prediction of formation collapse pressure.We take following steps to address this issue.The first step is to identify the uncertainty characteristics of logging parameters,including density,acoustic time,and natural gamma.The second step is to use an improved Rosenbluthe method and derive the uncertainty expressions of rock mechanical parameters,which are combined with the uncertainty expressions of horizontal geostress and formation pressure to obtain the probability distribution of borehole collapse pressure for its uncertainty analysis.In a case study of Well SY1,the uncertainty analysis method and the Monte Carlo method are used separately to assess the probability distribution of borehole collapse pressure and the risk of borehole instability.The results indicate a high level of agreement over 95% between two predictions.,in view of a borehole instability risk of 68.4% from our method and 67.4% from the Monte Carlo method.Our predictions were validated by the drilling process,during which hole collapse did happen.Furthermore,the instability risk predicted using the proposed method is slightly higher than using the Monte Carlo method.These findings provide essential reference value for field operation.

山东省自然科学基金(ZR2021ME024)和中央高校基本科研业务费专项基金(19CX02034A)共同资助。

10.12431/issn.1000-1441.2025.64.01.015