基于叠前地质统计学反演的高分辨河道识别及薄层预测——以东海B气田为例

High-resoulution Channel Identification and Thin Bed Prediction Based on Pre-stack Geostatistics Inversion——A Case Study of B Gas Field in the East China Sea

东海盆地B气田主力储层为浅水三角洲分流河道沉积砂体,储层埋藏深、横向变化快,薄储层、薄夹层发育,常规反演技术无法清晰识别河道和准确分辨薄层,开发井实施面临风险,为此采用叠前地质统计学反演技术来开展储层精细刻画研究。通过对岩相概率和变差函数的精细化分析求取了可靠的地质统计参数,降低了反演的不确定性。反演结果准确预测出了砂体内部7~8 m厚的薄夹层分布,刻画出了4期河道砂体的沉积变迁过程,据此指导了主力层多口水平分支井的实施并获得70.4%~92.3%的高砂岩钻遇率,从而证实了该技术对中深层薄层预测的有效性和准确性。

The main reservoir of B gas field in the East China Sea are sands deposited by distributary channels of the shallow water delta and it has the characteristics of large burial depth of reservoir,quick lateral variation,thin reservior and developing inter layer.The conventional inversion technique cannot describethe river channe land distinguish thin layers,and there are great risks in the implementation of development wells.Therefore,the pre-stack geostatistics inversion technique is used to characterize the reservoir finely.Reliable geological statistical parameters are obtained by fine analysis of Litho facies probabilities and variograms,reducing the uncertainty of inversion.The result of inversion has accurately predicted the distribution of thin inter layer whose thickness is 7~8 m in the internal of the sand body and characterized the sedimentary evolution process of sand bodies on the 4 channel.Several horizontal multilateral wells are implemented successfully and high sandstone drilling ratio from 70.4%to 92.3%are achieved,which shows the validity and accuracy of the technique to predict thin bed of mid-deep reservoirs.