Abstract: Seismic reservoir lateral forecasting,based on the reservoir properties is to converse seismic data into geological information,and to study the lateral variety and distribution of a reservoir among the wells.The restriction of various seismic forecasting methods and diversification of seismic explanation increased the risk apparently due to the lack of well data.In order to correctly converse the data with high precision,the facies-restrained reservoir lateral forecasting method was introduced,which is guided by geology,restrained by geological rule,integrating the geological,seismic and well-logging data,and by advanced computer techniques.It is certified in application that this method is not only based on advanced theories and computer technique,but also a reservoir forecasting method with practical value for drilling effect improvement at an early stage of oil exploration and development.

Key words: facies-restrained, prediction, neural network, sedimentary facies, reservoir

摘要： 地震储层横向预测就是在单井储层特征研究的基础上,将地震信息转化为地质信息,从而得到井间及无井区储层的横向变化特征及分布.在井资料缺乏的情况下,地震资料的多解性以及各种地震预测方法的局限性就愈加明显地提高了预测的风险性.为实现两种信息的高精度有效转化,提出一种以地质思路为指导,以地质规律为约束,高度综合地质、地震和测井资料,应用先进的计算机手段进行储层预测的新思路--相控地震储层横向预测方法.实际应用表明,这种方法不仅具有先进的理论基础和计算机手段,而且具有较高的预测精度,对油田勘探和开发,指导早期勘探部署,提高钻探成功率具有极高的实用价值.

关键词: 相控, 预测, 神经网络, 沉积相, 储集层