• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>The 4th Saudi International Conference on Information Technology: Big Data Analysis 2016 >A data-driven smart proxy model for a comprehensive reservoir simulation
【24h】

A data-driven smart proxy model for a comprehensive reservoir simulation

机译：数据驱动的智能代理模型，用于全面的油藏模拟

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

One of the most important tools for studying fluid flow behavior in oil and gas reservoirs is reservoir simulation. It is constructed based on a comprehensive geological information. A comprehensive numerical reservoir model has tens of millions of grid blocks. Therefore, it becomes computationally expensive and time consuming to run the model for different reservoir simulation scenarios. There are many efforts have been made to reduce the computational size using the proxy models. Proxy models are the substitute to the complex numerical simulation by producing a meaningful representation of the complex system in a very short time. The conventional proxy models are either statistical or mathematical approaches. These conventional approaches are still limited to the complexity of the reservoir and the number of the numerical simulation runs needed to build the proxy model. In this study, a smart proxy model that is based on artificial intelligence and data mining is presented. A grid based smart proxy model is developed to reproduce the dynamic reservoir properties of a full- field numerical simulation in few seconds. A comprehensive spatio-temporal database is built using the conducted numerical simulation run. The data from the database is trained, calibrated, and verified throughout the development of the smart proxy model. Smart proxy model is able to produce pressure and saturation at each reservoir grid block accurately and with a significantly less computational time compared to the numerical reservoir simulation model.
机译：研究油气藏中流体流动行为的最重要工具之一是油藏模拟。它是基于全面的地质信息而构建的。一个综合的数值储层模型具有数千万个网格块。因此，在不同的油藏模拟场景下运行模型变得计算量大且费时。已经进行了许多努力来使用代理模型来减小计算量。代理模型通过在非常短的时间内生成复杂系统的有意义的表示，从而替代了复杂的数值模拟。传统的代理模型是统计方法或数学方法。这些常规方法仍然受限于储层的复杂性和建立代理模型所需的数值模拟运行次数。在这项研究中，提出了一种基于人工智能和数据挖掘的智能代理模型。开发了基于网格的智能代理模型，以在几秒钟内重现全油田数值模拟的动态油藏属性。使用进行的数值模拟运行，建立了一个综合的时空数据库。在智能代理模型的整个开发过程中，都会对数据库中的数据进行培训，校准和验证。与数字储层模拟模型相比，智能代理模型能够在每个储层网格块处精确地产生压力和饱和度，并且计算时间大大减少。

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