水平井随钻阵列感应测井环境动态建模及参数估计

Geological steerable drilling characterized by underground closed-loop control requires fast logging response forecasting, and fast estimation of the reservoir geological characteristics and the observed state of the instrument, and its core is a logging forward and inverse problem. Now the forward and inversion of horizontal well is mainly based on the complex numerical calculation, which is not convenient for the real applications. Morover, the inversion algorithm doesn't take into account the prior information such as the formation structure and borehole trajectory parameters included in horizontal well design programes. This application is for the domestic array induction logging tool MIT, which studies the forward and inversion of its response while drilling based on online identification and estimation theory in complex system. The evolution model of the reservoir strata structure along the horizontal wellbore is constructed based on Markov switching system. The mechanism of array induction logging response in complex formation structural is figured out through the geometric factor, and the physical mechanisms is analyzed using multi- physics coupling software. Then the the instrument forward model can be established. The nonlinear estimation in the forward and inversion of array induction logging is studied using particle filter. Then, the adaptive mechanism of particle filter in Markov switching system is presented and the adaptive strategies of particle sampling and resampling is designed. These proposed methods should be able to enhance the interpretation of horizontal well and improve the rate of reservoir drilling coincidence in theory.

以井下闭环控制为特征的地质导向钻井要求快速的测井响应预报以及油藏地质特征、仪器观测状态估计,其核心是测井正反演问题。而目前水平井测井正反演均以复杂的数值计算为基础，算法的高复杂度限制其实用性，且反演算法未充分利用水平井设计方案中关于地层结构、井眼轨迹参数等先验信息。本申请书以国产阵列感应测井仪MIT为研究对象，基于复杂系统的在线辨识与估计理论研究水平井随钻阵列感应测井正反演问题。拟以马尔可夫切换系统为基础，建立沿水平井井眼轨道油藏地层结构演变模型；拟从几何因子研究复杂地层结构感应测井响应机理，借助多物理场耦合软件进一步分析物理机制，建立仪器正演模型；拟从粒子滤波入手研究阵列感应测井响应正反演的非线性估计问题，揭示Markov切换下粒子滤波的自适应机理，并研究粒子采样与重采样的自适应策略。本项目研究将为提升水平井解释技术及储层钻遇符合率提供有效的理论支撑。

在水平井和大斜度井中，地层电各向异性明显，其电测井响应特征与直井具有明显的差异，基于常规电测井响应的测井储集层油气评价是当前水平井解释的难点。对旋转导向钻井而言，随钻电磁波测井响应的合理解释对于井底钻具组合着陆在合适位置，避免在上覆非生产层中钻进较长的的水平段或在水平段顶部形成剩余阁楼油，从而使井筒与储层接触范围的最大化，也具有重要意义。本项目在分析水平井电测井影响因素的基础上，探究水平井常规测井与随钻电磁波测井的响应特征，开展随钻电磁波测井快速正反演研究，识别流体界面、断层以及井筒外地层的变化，帮助地质导向人员最大程度地扩大井筒与储层的接触范围，实现储层与井轨迹的实时联合优化，同时开展基于MicroCT岩芯宽频电磁特性研究，考虑复杂孔隙结构、组分、各向异性及非均质性对电导率及介电常数的影响，借助岩芯MicroCT图像，综合利用数值模拟得到的电导率及介电常数，提高含油饱和度的预测精度，为油气评价提供理论依据。在该项目资助下，培养了硕士研究生6名，开展与美国休斯敦大学工程学院电磁浅探测工程中心的合作，发表论文十余篇，其中SCI收录1篇、EI收录4篇，完成发明专利9个，在投文章2篇，正在撰写专著1本。同时将研究成果应用在与中国石油测井有限公司的合作课题“随钻电磁波电阻率测井正反演方法研究”、“井旁感应成像特征分析应用”中，经实测数据验证，算法进一步优化，鲁棒性提升，复杂度减小，随钻电磁波正反演的研究成果提升了水平井施工的准确性，有效提高了目的层的钻遇率。基于MicroCT岩芯宽频电磁特性表征研究成果成功应用在某低阻油气田测井资料二次解释中，极大提高了油气评价的准确性。