A Kick and Lost Circulation Monitoring Method Combining Bi-GRU and Drilling Conditions
-
摘要:
现有根据钻井液池体积和钻井液出口流量变化监测溢流与井漏的方法,未考虑开、停泵工况对出口流量和钻井液池体积变化的影响,易导致误报。为了降低误报率,分析了钻井工况与钻井液池体积和钻井液出口流量之间的相关关系,提出了一种结合钻井工况与双向门控循环单元(bidirectional-gated recurrent unit, Bi-GRU)的溢流与井漏智能监测方法。利用23口井的溢流与井漏监测数据,对提出的模型与现有典型模型分别进行了测试,结果表明:基于Bi-GRU的溢流与井漏智能监测模型的识别准确率为94.25%,优于其他模型;与未考虑钻井工况的Bi-GRU模型相比,误报率由12.52%降至1.12%。研究表明,该方法能够消除溢流与井漏监测时因开、停泵导致的风险误报,能为安全钻井提供技术支持。
Abstract:The existing kick and lost circulation monitoring methods using pot volume and outlet flow of drilling fluids do not consider the influence of the pump on and off on the outlet flow, and pot volume of drilling fluids. So it can easily lead to false alarm. In order to address this problem, the correlation of drilling conditions with pot volume and outlet flow of drilling fluids was established, and an intelligent kick and lost circulation monitoring method combining a bidirectional-gated recurrent unit (Bi-GRU) and drilling conditions was proposed. The proposed model and other representative models for kick and lost circulation monitoring were tested by using the data collected from 23 wells. The experimental results show that the identification accuracy of the proposed model achieves 94.25%, which is superior to those of the other models. Compared with that of the Bi-GRU model without considering the drilling conditions, the false alarm rate of the proposed model drops from 12.52% to 1.12%. The proposed method reduces the false alarms caused by pump on and off states during kick and lost circulation monitoring, and these findings can provide technical support for safe drilling.
-
Keywords:
- kick /
- lost circulation /
- monitoring /
- drilling condition /
- artificial intelligence /
- risk recognition /
- risk false alarm
-
-
表 1 数据集的样本数量与编码方式
Table 1 The sample size of the dataset and encoding method
类型 训练集
样本数验证集
样本数测试集
样本数样本
总数独热
编码正常 1 440 360 350 2 150 100 井漏 600 160 215 975 010 溢流 480 120 200 800 001 表 2 混淆矩阵的定义
Table 2 Confusion matrix definition
实际工况 判别为正常 判别为风险 正常 TP FN 风险 FP TN -
[1] 张晓诚,霍宏博,林家昱,等. 渤海油田裂缝性油藏地质工程一体化井漏预警技术[J]. 石油钻探技术,2022,50(6):72–77. ZHANG Xiaocheng, HUO Hongbo, LIN Jiayu, et al. Integrated geology-engineering early warning technologies for lost circulation of fractured reservoirs in Bohai Oilfield[J]. Petroleum Drilling Techniques, 2022, 50(6): 72–77.
[2] MAO Youli, ZHANG Peng. An automated kick alarm system based on statistical analysis of real-time drilling data[R]. SPE 197275, 2019.
[3] 孙伟峰,李威桦,王健,等. 基于C#与Python混合编程的钻井溢漏风险智能识别平台[J]. 实验技术与管理,2021,38(11):166–172. SUN Weifeng, LI Weihua, WANG Jian, et al. Intelligent identification platform of drilling kick and loss risk based on mixed programming of C# and Python[J]. Experimental Technology and Management, 2021, 38(11): 166–172.
[4] 戴永寿,岳炜杰,孙伟峰,等. “三高” 油气井早期溢流在线监测与预警系统[J]. 中国石油大学学报(自然科学版),2015,39(3):188–194. DAI Yongshou, YUE Weijie, SUN Weifeng, et al. Online monitoring and warning system for early kick foreboding on ‘three high’ wells[J]. Journal of China University of Petroleum (Edition of Natural Science), 2015, 39(3): 188–194.
[5] YI M, ASHOK P, RAMOS D, et al. Natural language processing applied to reduction of false and missed alarms in kick and lost circulation detection[R]. SPE 206340, 2021.
[6] 杨传书,李昌盛,孙旭东,等. 人工智能钻井技术研究方法及其实践[J]. 石油钻探技术,2021,49(5):7–13. YANG Chuanshu, LI Changsheng, SUN Xudong, et al. Research method and practice of artificial intelligence drilling technology[J]. Petroleum Drilling Techniques, 2021, 49(5): 7–13.
[7] 袁俊亮,范白涛,幸雪松,等. 基于朴素贝叶斯算法的钻井溢流实时预警研究[J]. 石油钻采工艺,2021,43(4):455–460. YUAN Junliang, FAN Baitao, XING Xuesong, et al. Real-time early warning of drilling overflow based on naive Bayes algorithm[J]. Oil Drilling & Production Technology, 2021, 43(4): 455–460.
[8] 邓正强,兰太华,林阳升,等. 川渝地区防漏堵漏智能辅助决策平台研究与应用[J]. 石油钻采工艺,2021,43(4):461–466. DENG Zhengqiang, LAN Taihua, LIN Yangsheng, et al. Research and application of intelligent assistant decision making platform of lost circulation prevention and control in Sichuan-Chongqing Area[J]. Oil Drilling & Production Technology, 2021, 43(4): 461–466.
[9] 李中. 中国海油油气井工程数字化和智能化新进展与展望[J]. 石油钻探技术,2022,50(2):1–8. LI Zhong. Progress and prospects of digitization and intelligentization of CNOOC’s oil and gas well engineering[J]. Petroleum Drilling Techniques, 2022, 50(2): 1–8.
[10] 王茜,张菲菲,李紫璇,等. 基于钻井模型与人工智能相耦合的实时智能钻井监测技术[J]. 石油钻采工艺,2020,42(1):6–15. WANG Xi, ZHANG Feifei, LI Zixuan, et al. Real-time intelligent drilling monitoring technique based on the coupling of drilling model and artificial intelligence[J]. Oil Drilling & Production Technology, 2020, 42(1): 6–15.
[11] SEABE P L, MOUTSINGA C R B, PINDZA E. Forecasting cryptocurrency prices using LSTM, GRU, and bi-directional LSTM: A deep learning approach[J]. Fractal and Fractional, 2023, 7(2): 203. doi: 10.3390/fractalfract7020203
[12] 刘汉桥. 基于数据挖掘的海洋钻井井涌早期预测方法研究[D]. 青岛: 中国石油大学(华东), 2020. LIU Hanqiao. Study on early prediction for offshore drilling well kick based on data mining[D]. Qingdao: China University of Petroleum(East China), 2020.
[13] SCHUSTER M, PALIWAL K K. Bidirectional recurrent neural networks[J]. IEEE Transactions on Signal Processing, 1997, 45(11): 2673–2681. doi: 10.1109/78.650093
[14] WANG Yanting, ZHENG Dingkun, JIA Rong. Fault diagnosis method for MMC-HVDC based on Bi-GRU neural network[J]. Energies, 2022, 15(3): 994. doi: 10.3390/en15030994
[15] KIM J. Finding the best performing pre-trained CNN model for image classification: Using a class activation map to spot abnormal parts in diabetic retinopathy image[J]. American Journal of Biomedical and Life Sciences, 2021, 9(4): 176–181. doi: 10.11648/j.ajbls.20210904.11
[16] NIE Qi, LI Yun, XIONG Wenying, et al. Health recognition algorithm for sports training based on Bi-GRU neural networks[J]. Journal of Healthcare Engineering, 2021, 2021: 1579746.
[17] LIU Xun, YOU Junling, WU Yulei, et al. Attention-based bidirectional GRU networks for efficient HTTPS traffic classification[J]. Information Sciences, 2020, 541: 297–315. doi: 10.1016/j.ins.2020.05.035
[18] MATEUS B C, MENDES M, FARINHA J T, et al. Comparing LSTM and GRU models to predict the condition of a pulp paper press[J]. Energies, 2021, 14(21): 6958. doi: 10.3390/en14216958
[19] WANG Guangbin, CHEN Jinhua, ZHONG Zhixian, et al. Multi-source heterogeneous fusion entropy ratio distance feature of bearing performance degradation based on DTW[J]. Vibroengineering Procedia, 2021, 39: 17–23. doi: 10.21595/vp.2021.22269
[20] TANG Hewei, ZHANG Shang, ZHANG Feifei, et al. Time series data analysis for automatic flow influx detection during drilling[J]. Journal of Petroleum Science and Engineering, 2019, 172: 1103–1111. doi: 10.1016/j.petrol.2018.09.018
[21] 刘翔,张立华,戴泽源,等. 一种无输入参数的强噪声背景下ICESat-2点云去噪方法[J]. 光子学报,2022,51(11):354–364. LIU Xiang, ZHANG Lihua, DAI Zeyuan, et al. A parameter-free denoising method for ICESat-2 point cloud under strong noise[J]. Acta Photonica Sinica, 2022, 51(11): 354–364.
-
期刊类型引用(10)
1. 赵金省,宋语桐,张庆祝,徐洋,胡海文,李斌,许明勇,居迎军. 致密砂岩油藏CO_2吞吐孔喉结构变化规律研究. 非常规油气. 2025(02): 54-63 . 百度学术
2. 刘琼. 中原油田CO_2吞吐技术及增产应用. 精细石油化工进展. 2024(05): 32-35 . 百度学术
3. 黄千慧,李海波,杨正明,邢济麟,陈波,李杰,薛伟,姚兰兰,杜猛,孟焕. 页岩(致密)油藏注CO_2吞吐作用距离实验. 大庆石油地质与开发. 2024(06): 128-135 . 百度学术
4. 宫厚健,张泽轲,于越洋,张欢,吕威,徐龙. 二维核磁共振技术评价CO_2吞吐对不同状态页岩油的动用率. 实验室研究与探索. 2024(11): 10-15 . 百度学术
5. 王哲,曹广胜,白玉杰,王培伦,王鑫. 低渗透油藏提高采收率技术现状及展望. 特种油气藏. 2023(01): 1-13 . 百度学术
6. 李凤霞,王海波,周彤,韩玲. 页岩油储层裂缝对CO_2吞吐效果的影响及孔隙动用特征. 石油钻探技术. 2022(02): 38-44 . 本站查看
7. 廖松林,夏阳,崔轶男,刘方志,曹胜江,汤勇. 超低渗油藏水平井注CO_2多周期吞吐原油性质变化规律研究. 油气藏评价与开发. 2022(05): 784-793 . 百度学术
8. 鞠斌山,于金彪,吕广忠,曹伟东. 低渗透油藏CO_2驱油数值模拟方法与应用. 油气地质与采收率. 2020(01): 126-133 . 百度学术
9. 张东,刘显太,刘彦东,刘启玲. CO_2驱合理注入量计算方法. 油气地质与采收率. 2020(01): 107-112 . 百度学术
10. 周元龙,赵淑霞,何应付,王锐. 基于响应面方法的CO_2重力稳定驱油藏优选. 断块油气田. 2019(06): 761-765 . 百度学术
其他类型引用(11)