基于神经网络模型的生物扰动碳酸盐岩储集层识别与孔隙度预测——以塔里木盆地塔河油田奥陶系生物扰动碳酸盐岩储集层为例
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:

本文为国家自然科学基金资助项目(编号:41472104,41102076),国家重点研究发展计划(973计划;编号:2011CB201001),国家油气重大科技专项(编号:2011ZX05014-002- 002,2008ZX05014-002-002)和河南省自然科学基金资助项目(编号:202300410185)的成果


Identification of the bioturbated carbonate reservoir and their porosity prediction based on conventional well logging data using artificial neural networks——Take the Ordovician bioturbated carbonate reservoir in Tahe oilfield,Tarim Basin, as an example
Author:
Affiliation:

Fund Project:

单位:
  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    塔里木盆地塔河油田奥陶系生物扰动碳酸盐岩储集层非常发育,但利用常规测井数据识别生物扰动储集层发育段和准确预测孔隙度难度较大。本文在对研究区16口取芯井奥陶系岩芯上生物扰动区域扰动等级划分的基础上,通过岩性标定测井,优选常规测井参数,基于BP神经网络模型分别建立了适合研究区生物扰动碳酸盐岩储集层识别和孔隙度预测的模型,并对建立的模型进行了有效性检验。结果表明:① 选择自然电位、自然伽马、井径、深侧向电阻率、浅侧向电阻率、补偿中子和密度等常规测井数据作为生物扰动碳酸盐岩储集层识别模型输入层的参数值,生物扰动指数(Bioturbation Index, BI)作为输出结果;选取rprop、sigmoid symmetric和sigmoid stepwise函数分别作为训练函数、隐含层和输出层的激活函数,建立节点数为3、层数为3的神经网络识别模型,识别效果好,适用于研究区奥陶系生物扰动碳酸盐岩储集层的识别。② 选择自然电位、自然伽马、井径、声波、补偿中子和密度值等常规测井数据作为输入层的参数值,对应深度上岩芯柱塞孔隙度测试结果和利用孔隙度样品检验模型计算得出的孔隙度结果作为输出结果,选取incremental、gaussian和sigmoid分别作为训练函数、隐含层和输出层的激活函数,建立节点数为4,层数为3的生物扰动碳酸盐岩储集层孔隙度预测模型,预测效果良好,适用于研究区奥陶系生物扰动储集层孔隙度的预测。该研究对定量表征研究区生物扰动储层特性、储量估算、油藏描述和储层地质建模等具有重要的借鉴意义。

    Abstract:

    Suitable sedimentary environment, ecological conditions, favorable spatiotemporal sediment matching and abundant organism—substrate interaction, the Ordovician multiple burrows in the Yingshan Formation and the Yijianfang Formation in the Tahe oilfield,Tarim Basin, overprinted to form a large- scale bioturbated carbonate rock with lateral continuity and vertical connectivity. These bioturbated carbonate rocks have positive porosity and permeability and are potential oil and gas reservoirs. However, it is difficult to identify bioturbated reservoirs using conventional logging data due to the characteristics of strong reservoir inhomogeneity, heterogeneous hydrocarbon- bearing properties, and small differences in logging responses between hydrocarbon- bearing and water- bearing formations. In this paper, the identification model and porosity prediction model for bioturbated carbonate reservoirs in the study area are established by calibrating the well logging data with core data, selecting the most relevant conventional logging parameters. Methods: Based on defining the bioturbation index of the bioturbated Ordovician carbonate cores from 16 wells, calibrating the well logging data with core data, the conventional logging data, such as self- potential, natural gamma, caliper logs, deep laterolog, shallow laterolog, neutron logs and density logs, are selected as the input parameters, and bioturbation index is used as the output results. The rprop, sigmoid symmetri and sigmoid stepwise functions are selected as the training function, the activation function of the hidden layer and the output layer, respectively. The identification model with 3 nodes and 3 layers, is established to identify the Ordovician bioturbated carbonate strata in the study area. The conventional logging data, such as self- potential, natural gamma, caliper logs, sonic logs, Neutron logs and density logs, are selected as the input parameters, and the porosity values from the measured plugs and the Porosity Calculation Sample Inspection Model is used as the output results. The incremental, Gaussian and sigmoid are selected as the training function, the activation function of hidden layer and output layer, respectively. The porosity prediction model with 4 nodes and 3 layers, is established to predict the bioturbated carbonate reservoir in the study area. Results:The identification model established with 3 nodes and 3 layers, is applied to identify the Ordovician bioturbated carbonate strata in the study area and has acquired a positive identification effect. The porosity prediction model established with 4 nodes and 3 layers, is applied to predict the bioturbated carbonate reservoir in the study area and has also acquired a positive prediction effect. Conclusions: The identification model established for bioturbated carbonate reservoirs meet the requirement of accuracy, and are applicable to identify the bioturbated Ordovician carbonate reservoirs in the study area. The prediction model for the porosity of bioturbated carbonate reservoirs meet the requirement of accuracy, which is applicable to the prediction of porosity of the Ordovician bioturbated reservoirs in the Tahe oilfield,Tarim Basin. This study has important reference significance for quantitative characterization of reservoir characteristics, reserve estimation, reservoir description and establishment of reservoir geological model.

    参考文献
    相似文献
    引证文献
引用本文

牛永斌,赵佳如,钟建华,王敏,徐资璐,程梦园.2021.基于神经网络模型的生物扰动碳酸盐岩储集层识别与孔隙度预测——以塔里木盆地塔河油田奥陶系生物扰动碳酸盐岩储集层为例[J].地质论评,67(6):1898-1909,[DOI].
NIU Yongbin, ZHAO Jiaru, ZHONG Jianhua, WANG Min, XU Zilu, CHENG Mengyuan.2021. Identification of the bioturbated carbonate reservoir and their porosity prediction based on conventional well logging data using artificial neural networks——Take the Ordovician bioturbated carbonate reservoir in Tahe oilfield, Tarim Basin, as an example[J]. Geological Review,67(6):1898-1909.

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2021-02-24
  • 最后修改日期:2021-07-22
  • 录用日期:
  • 在线发布日期: 2021-11-19
  • 出版日期: 2021-11-15