• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Technology advances >Using Neural Networks for Prediction of Formation Fracture Gradient
【24h】

Using Neural Networks for Prediction of Formation Fracture Gradient

机译：使用神经网络预测地层破裂梯度

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

Accurate formation fracture gradient prediction is an essentialrnpart of well planning. Erroneous fracture gradient estimatesrnmay jeopardize the entire drilling operation and result inrnserious well problems, the least of which are lost circulationrnand kick leading to blowout. Accurate fracture gradient valuesrnplay an important role in the selection of proper casing seats,rnprevention of lost circulation and planning of hydraulicrnfracturing for the purpose of increasing well productivity inrnzones of low permeability. Furthermore, a good knowledge ofrnthe fracture gradient is of great importance in areas wherernselective production and injection is practiced. In such areasrnthe adjacent reservoirs consist of several sequences of densernand porous zones such that, if a fracture is initiated duringrndrilling or stimulation, it can propagate and extend,rnestablishing communication between hydrocarbon reservoirsrnand can extend to a nearby water-bearing formation.rnFracture gradient depends upon several factors includingrnmagnitude of overburden stress, formation stress within thernarea and formation pore pressure. Any prediction methodrnshould incorporate most of the above factors for a realisticrnprediction of the fracture gradient.rnThis paper presents an artificial neural network model thatrnyields reasonably accurate values of the fracture gradient. Therninput training data are actual field data. The results obtainedrnfrom the model are compared with those obtained fromrncorrelation. The comparison shows that the method isrnpromising and under some circumstances it is superior to thernavailable techniques.
机译：准确的地层裂缝梯度预测是井眼计划的重要组成部分。错误的裂缝坡度估算值可能会危害整个钻井作业，并导致严重的井问题，其中最少的问题是井漏和井喷导致井喷。准确的裂缝坡度值在选择合适的套管座，防止漏失和规划水力压裂方面起着重要作用，以增加低渗透率井的产量。此外，在实行选择性生产和注入的地区，对裂缝梯度的了解非常重要。在这样的区域中，相邻的储层由数个致密且多孔的层序组成，这样，如果在钻井或增产过程中引发裂缝，裂缝就可以传播和扩展，在油气储层之间建立连通并可以扩展到附近的含水层。几个因素包括上覆应力的大小，热区内的地层应力和地层孔隙压力。任何预测方法都应结合上述大多数因素，以实现对裂缝梯度的现实预测。本文提出了一种人工神经网络模型，可以合理地得出裂缝梯度的准确值。输入的训练数据是实际的现场数据。将模型获得的结果与相关性获得的结果进行比较。比较表明该方法是有前景的，在某些情况下它优于可用技术。

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