首页> 外文会议>SPE Middle East Oil Gas Show and Conference >Identification of High-Porosity Reservoir Sands From 3D-Seismic Attributes Using Neural Network Technique in South Umm Gudair Field, Kuwait
【24h】

Identification of High-Porosity Reservoir Sands From 3D-Seismic Attributes Using Neural Network Technique in South Umm Gudair Field, Kuwait

机译:基于南翁甘桥南翁甘地技术的三维地震属性高孔隙储层砂的识别

获取原文

摘要

The South Umm Gudair Field is a multi-accumulation structure with the most prolific production of oil from a Ratawi Oolite reservoir (Neocomian). The historic ultimate seal held by the thick Ratawi Shale Member, displays embedded multilayered oil bearing reservoir sands. Sands were sedimented as thin tidal bars anastomosed with distributary and tidal channels deposited under pro-deltaic to delta front environment. To delineate this complex reservoir, a neural network technique was applied to estimate effective porosity integrating well information and 3D multi-attributes between seismically mapped sequences. This method is more efficient than conventional estimation with the ability to build a non-linear relationship between seismic traces and target porosity logs for interpolation. The combination of 20 seismic attributes including impedance derived by model based inversion was selected on the basis of ranking of error factors. The network was trained with effective porosity available at well locations. It subsequently recognized not only that pattern, but also similar patterns by generalization in the study area. The effective porosity cube was generated by the processing of 3D seismic data with this trained network. The representative sections extracted from this volume highlight the limited extent of reservoir sands within the sequences which has also been supported by the simulations of lithofacies with Sequential Gaussian method using well data. The trajectory of the proposed lateral wells are designed in sequences-2 & 4 to intersect the porous sand at the optimum location to ensure good drainage of these reservoir units. The reservoir sand corresponding to upper sequence-4 shows widespread development in the western part of the area of interest with better connectivity, but is shaling out eastward of the field. The paper focuses on the methodology of identification of thin lenticular high porosity reservoir sands to plan for side-tracking existing deeper wells and drilling appraisal & delineation wells.
机译:南美古甘地领域是一种多累积结构,具有来自大提琴鲕粒水库(新科医生)的油脂生产。厚实的ratawi页岩会员持有的历史终极密封,展示了嵌入式多层油储层砂岩。砂岩沉积为薄片散,赋予沉积在Pro-Deltaic下的分布式和潮汐通道沉积到Delta前环境。为了描绘这种复杂的储存器,应用神经网络技术来估计地震映射序列之间的有效孔隙率集成井信息和3D多属性。该方法比传统估计更有效,具有在地震迹线与目标孔隙日志之间构建非线性关系的能力,用于插值。基于误差因子的排名选择包括基于模型的反转导出的阻抗的20个地震属性的组合。网络训练,具有有效的孔隙度可在井位置提供。随后不仅认识到该模式,还通过研究区域中的泛化来识别。通过使用该训练网络的3D地震数据处理产生有效的孔隙率立方体。从该体积中提取的代表性部分突出了序列内的储层砂内的有限程度,这也得到了使用井数据的顺序高斯方法的锂离样的模拟。所提出的横向孔的轨迹在序列-2和4中设计,以在最佳位置与多孔砂相交,以确保这些储存器单元的良好排出。对应于上序列-4对应的储层沙子在兴趣区域的西部具有广泛的开发,具有更好的连接,但在该领域的东方呼出。本文侧重于薄晶状体高孔隙率砂岩的识别方法,计划侧面跟踪现有更深的井和钻探评估井。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号