利用测井资料预测地层压力的误差处理方法

在分析利用测井资料预测地层压力可能产生误差的各种因素的基础上，介绍了未确知有理数、滑动平均、深度平差等方法。通过引入未确知有理数法，提出了一种有效消除干扰信息、提高地层压力预测精度稳定性的方法；通过引入深度平差法，解决了新井借用邻井测井资料时相同层位深度不同的问题，消除了由此产生的预测结果在深度上的错动，形成了一种邻井资料向目的区块合理移置的方法。此外，用实测点直接对预测结果校正可以得到更加准确的预测结果。应用表明，利用所建立的方法可使地层压力的预测误差稳定地控制在10％以内。

New Data Processing Methods to Reduce Errors of Pore-Pressure Prediction Based on Well Logging Data

On the basis of analyzing various factors that may contribute to errors of pore-pressure predlction using well logging data, several data processing methods are presented in this paper, including method of uncertain rational number, techniques of running mean, and depth adjustment. A method is brought forward to effectively eliminate interference and improve prediction accuracy of pore pressure by employing the method of uncertain rational number. The pore pressure prediction method integrated with depth adjustment resolves the problem of different depth of same layers when referring to data from offset wells, and removes the depth offset of predicted results. Thus, amethod has been developed to reasonable referring to offset data in different formations. In addition, more accurate predictions can be achieved after revising the theoretical data with actual ones. Applications show that formation-pressure-prediction errors can be controlled less than 10% using foregoing methods.