套管井中子测井校正方法研究

套管井测井是当前测井技术的重要发展方向。套管、水泥环胶结质量及厚度是影响套管井中子测井的主要因素。实际测井资料显示，套管井中子测井效果较好，能指示地层含氢量的相对变化，能用于储层识别。用图版校正法、一元回归分析法、多元逐步回归法、BP神经网络法，对17.78cm套管和24.45cm套管中子测井进行校正。实际资料处理证实，BP神经网络校正方法精度最高，其它3种校正方法精度相当。17.78cm套管中子测井及其校正效果好于24.45cm套管。根据实际测井资料情况，推荐了相应的校正方法：在有井径和套管井中子长短源距离计数率资料情况下，选用BP神经网络校正方法对套管井中子进行校正；在没有井径资料的情况下，用一元回归校正法对套管井中子进行校正。

CORRECTING METHOD OF NEUTRON LOGGING IN CASED WELL

Well logging in cased well is important in the development of current logging technology.The casing diameter and quality of cementation and the thickness of the cement ring are the main factors which influence neutron logging.It is showed by logging data that the effectiveness of corresponding neutron logging is good,it can not only indicate the relative variety of the hydrogenbearing quantity of layers but also be used in identifying reservoir.The emendation of 17.78cm and 24.45cm neutron logging can be processed by the way of plate emendation,oneparameter regression analytical method,multiparameters step regression method and the BP neutron network method,it is proved by real information processing that the BP neutron network emendation method has the highest precision,while other ones the approximate precision.The information got by 17.78cm neutron logging and its emendation effectiveness is better than that 24.45cm neutron logging.According to the real logging information,The corresponding emendation as followed is recommended,if there are well diameter and long and short space neutron counting rate,BP neural network method is used to correct neutron logging data of cased well,and when well diameter is not available,oneparameter regression analytical method is appropriate.