孤家子—后五家户气田地震反演与储层建模综合研究

地震反演和储层建模是储层横向预测的两种方法。地质统计学地震反演是一种基于模型的地震反演，采用地质统计学空间插值技术建立模型，将测井曲线或测井解释成果曲线演化为“假阻抗”来合成地震记录，寻找子波，再将地震道直接反演成储层属性。这种方法避免了波阻抗转换为储层属性所造成的误差。储层建模横向预测中，应用协克里金算法联合地震反演与原始测井两种数据，综合建立地质模型。在孤家子—后五家户气田实际应用中，选取具有代表性的R2.5梯度曲线进行反演，反演结果反映了砂体分布。结合各井点测井解释数据，进行了储层建模，提高了横向预测精度。

Geostatistical seismic inversion and reservoir modeling study in Gujiazi-Houwujiahu region

Seismic inversion and reservoir modeling are two methods to characterize reservoir distribution. The seismic inversion method used in this paper is a geostatistical inversioa It is a model-based seismic inversion, which uses geostatistical spatial interpolation to construct model, and uses pseudo-impedance derived from logging information (original or interpreted) to build seismogram. By directly inverting seismic traces into reservoir properties, this method can minimize the errors caused by transferring wave impedance into reservoir properties. In the prediction of lateral reservoir variations, co-Krige method is used to combine seismic inversion results and original log data to form an integrated geological model. In Gujiazi-Houwujiahu gas field, we use representative R2. 5 gradient curve to invert sand distribution, and use interpreted log data to build reservoir model. Satisfactory results have been achieved.