| 用地球物理测井方法识别碳酸盐岩储集层的岩性及孔隙结构--以巴西深海J油田案例 |
| |
| 引用本文： | 胡瑶,李军,苏俊磊.用地球物理测井方法识别碳酸盐岩储集层的岩性及孔隙结构--以巴西深海J油田案例[J].地球物理学进展,2020(2):735-742. |
| |
| 作者姓名： | 胡瑶 李军 苏俊磊 |
| |
| 作者单位： | 中国石油化工股份有限公司石油勘探开发研究院 |
| |
| 基金项目： | 国家科技重大专项基金资助项目“典型油气藏岩相及储层参数测井定量评价”(2016ZX05033-003-001);国家科技重大专项基金资助项目“不同类型缝洞储集体测井精细评价”(2016ZX05014-002-001)联合资助. |
| |
| 摘 要： | 位于巴西深J油田的盐下碳酸盐岩岩性及孔隙结构的识别对于区域储层的分布规律十分重要,但由于海外资料的获取量有限,如何最大限度地利用已有资料进行精准有效的岩性识别成为难题.本文在已有的各项地质信息资料的基础上,先利用成像测井资料提取出储层内主要岩性典型的结构标准图版,然后探索各类岩性在常规测井曲线上的响应规律,并提取出识别岩性的敏感参数,建立起常规测井参数的交会图法识别规律.确认其应用在全井段识别中取得效果后,进一步利用数据挖掘软件中的决策树技术,组合成交会图-决策树模型法.将此模型应用于实际井资料的处理,利用薄片资料和成像图像进行验证,得到了更高的岩性识别符合率,证实了模型在仅有常规测井资料中的适用性.
|
| 关 键 词： | 岩性识别 成像测井图像 常规测井资料 敏感参数 交会图-决策树模型 |
Identification of lithology and pore structure of subsalt carbonate rocks by geophysical logging method:Brazil deep sea J oilfield case |
| |
| HU Yao,LI Jun,SU Jun-lei.Identification of lithology and pore structure of subsalt carbonate rocks by geophysical logging method:Brazil deep sea J oilfield case[J].Progress in Geophysics,2020(2):735-742. |
| |
| Authors: | HU Yao LI Jun SU Jun-lei |
| |
| Institution: | (Sinopec Petroleum Exploration and Production Research Institute,Beijing 100083,China) |
| |
| Abstract: | The identification of sub-salt carbonate lithology and pore structure in deep J oil field in Brazil is very important to the distribution of regional reservoirs. However, due to the limited acquisition of overseas data, how to maximize the use of existing data to identify lithology accurately and effectively becomes a problem. Based on the existing geological information data, this paper firstly uses the imaging logging data to extract the typical structural standard plots of the main lithology in reservoir, then explores the response laws of each lithology in the conventional logging curves. The sensitive parameters of the identification of the lithology are extracted to establish the regularity by crossplot method of the conventional logging parameters. After confirming that the application of this method has achieved the effect in the identification of whole sections in a well, the decision tree technology of the data mining software is further utilized to combine the crossplot-decision tree model method. Applying this model to the processing of actual well data, using CT thin-section data and imaging logging images to verify the results, a higher coincidence rate in lithology identification is obtained, which confirms the applicability of the model under the conditions of only conventional logging data available. |
| |
| Keywords: | Lithology identification Imaging logging images Conventional logging data Sensitive parameters Crossplot-decision tree model |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
