应用单相关分析和多元回归快速预测储层敏感性

Rapidly predicating the reservoir sensitivity with single correlation and multiple regression analytic method

敏感性研究是分析储层损害机理的前提,对于优化钻探和开发过程中的各个作业环节,以及制定系统的油气层保护技术方案具有非常重要的意义。对近年来发展起来的各种储层敏感性预测方法进行了探讨,发现利用单相关分析和多元回归预测储层敏感性是一种较理想的、快速的预测敏感性的新方法。在常规岩心分析和敏感性矿物分析的基础上,提取与各敏感性有关的信息,利用单相关分析和多元回归预测储层敏感性,其准确率能够达到85%,基本上满足了油气层敏感性预测的需要。相对单一的预测方法,二者结合的预测结果有明显改善,并且该方法计算简单、适用性强、物理意义明确。

Sensitivity study is the premise of reservoir damage mechanism analyzing which is extremely significant to optimizing each drilling and development work process, and is equally important to draw up scientific reservoir protection solutions. By discussing various reservoir sensitivity prediction methods developed in recent years, we found that using single correlation analysis and multiple regressions to forecast reservoir sensitivity is a fairly ideal, fast and new method. On the basis of conventional cores analysis and sensibility mineral analysis, we extract useful information relating to various sensitivity factors, and predict the reservoir sensitivity using the new method above mentioned. The accuracy of forecasting results can reach 85%, which basically meet the needs of reservoir sensitivity predicting. Compared with the single method, the predicted results by combining single correlation analysis and multiple regression are obviously improved, so the new method is simple, widely applicable and with explicit physical significance.