[1]趙元良,樊兆亞,葛盛權,等.基于大數據分析的流度預判及MDT優化[J].測井技術,2019,43(06):652-656.[doi:10.16489/j.issn.1004-1338.2019.06.019]
 ZHAO Yuanliang,FAN Zhaoya,GE Shengquan,et al.Fluidity Prediction and MDT Optimization Based on Big Data Analysis[J].WELL LOGGING TECHNOLOGY,2019,43(06):652-656.[doi:10.16489/j.issn.1004-1338.2019.06.019]
點擊復制

基于大數據分析的流度預判及MDT優化()
分享到:

《測井技術》[ISSN:1004-1338/CN:61-1223/TE]

卷:
第43卷
期數:
2019年06期
頁碼:
652-656
欄目:
解釋評價
出版日期:
2019-12-15

文章信息/Info

Title:
Fluidity Prediction and MDT Optimization Based on Big Data Analysis
文章編號:
1004-1338(2019)06-0652-05
作者:
趙元良1樊兆亞2葛盛權1陳輯超2
(1.中國石油塔里木油田分公司,新疆庫爾勒841000;2.斯倫貝謝公司,北京100020)
Author(s):
ZHAO Yuanliang1 FAN Zhaoya2 GE Shengquan1 CHEN Jichao2
(1. PetroChina Tarim Oilfield Company, Korla, Xinjiang 841000, China; 2. Schlumberger China, Beijing 100020, China)
關鍵詞:
巖石物理測井大數據決策樹回歸分析神經網絡流度預測庫車山前區域
Keywords:
petrophysical logging big data decision tree regression analysis neural network fluidity prediction the front of the Kuqa mountain
分類號:
P631.84
DOI:
10.16489/j.issn.1004-1338.2019.06.019
文獻標志碼:
A
摘要:
庫車山前區域庫車坳陷白堊系巴什基奇克組砂巖經過了漫長的埋藏過程和復雜的成巖演化,形成了一套超深低孔隙度特低滲透率儲層。模塊化電纜地層測試器(MDT)是測量地層流體性質及地層流度信息最直接、經濟的方法,然而由于庫車山前區域儲層的復雜性,MDT作業成功率不足50%。通過對該區域30口MDT作業井的800多個測壓點數據及巖石物理測井結果進行訓練決策樹致密性預判模型,形成了基于巖石物理測井結果的單元線性回歸、多元線性回歸、神經網絡的方法進行流度定量預測模型,建立了完整的流度預測流程,利用流度預測結果可指導MDT作業可行性分析并優化儀器。
Abstract:
After a long burial process and complex diagenetic evolution, the Cretaceous Bashiqiki Formation sandstone evolved to ultra-deep reservoirs with low porosity and ultra-low permeability in the Kuqa depression in the front of the Kuqa mountain. A modular wireline formation tester (MDT) is the most direct and economic tool to measure the properties and fluidity of formation fluid. However, due to the complexity of the reservoir in the front of the Kuqa mountain, the success rate of MDT operation is less than 50%. Based on more than 800 pressure points and petrophysical logging data of 30 MDT wells in this area, this paper establishes a tight prediction model of decision tree, then develops unit linear regression, multiple linear regression and neural network based on petrophysical logging data for building a quantitative fluidity prediction model, and finally establishes a completely process of fluidity prediction The fluidity prediction results can guide feasibility analysis of MDT operation and optimization of tools.

參考文獻/References:

[1]MDT reference manuals [Z]. Schlumberger, 2001.
[2]丁次乾. 礦場地球物理 [M]. 東營: 中國石油大學出版社, 2003.
[3]李航. 統計學習方法 [M]. 北京: 清華大學出版社, 2012.
[4]李漢林, 趙永軍, 王海起. 石油數學地質 [M]. 東營: 中國石油大學出版社, 2008.

相似文獻/References:

[1]申波,毛志強,羅興平,等.應用因子分析方法識別熱液蝕變作用類型[J].測井技術,2015,39(04):518.[doi:10.16489/j.issn.1004-1338.2015.04.025]
 SHEN Bo,MAO Zhiqiang,LUO Xingping,et al.Identification of Hydrothermal Alteration Type by Using Factor Analysis Method[J].WELL LOGGING TECHNOLOGY,2015,39(06):518.[doi:10.16489/j.issn.1004-1338.2015.04.025]

備注/Memo

備注/Memo:
第一作者:趙元良,男,1968年生,從事石油勘探開發研究工作。E-mail:[email protected](修改回稿日期: 2019-01-29本文編輯王小寧)
更新日期/Last Update: 2019-12-15
脉动时空彩游棋牌游 福彩开奖结结果 基金配资地址 90888九龙高手论坛精选 武汉赖子麻将官方网站 青海11选5中奖规则 历年英超排名 弈乐贵州麻将手机版 天津快乐十分开奖号码 中奖一千万多久到账 北京快3形态一定牛 棋牌游戏开发? 体彩p3和值遗漏 三分pk拾怎么稳赢不亏 互联网投资理财排行榜 澳门小神仙2020彩图 天才麻将少女真人版顺序