川渝地区防漏堵漏智能辅助决策平台研究与应用

Research and application of intelligent assistant decision making platform of lost circulation prevention and control in Sichuan-Chongqing area

邓正强 兰太华 林阳升 何涛 黄平 罗宇峰 王君 谢显涛

DENG Zhengqiang LAN Taihua LIN Yangsheng HE Tao HUANG Ping LUO Yufeng WANG Jun XIE Xiantao

中国石油集团川庆钻探工程有限公司钻井液技术服务公司

Drilling Fluid Technology Service Co., Ltd., CNPC Chuanqing Drilling Engineering Company Limited, Chengdu, Sichuan 610056, China

针对川渝地区井漏严重、井漏治理主要以经验为主、一次堵漏成功率低等问题，提出了采用大数据/人工智能技术，对漏失特征、裂缝性质进行大数据分析诊断，对正钻井漏失风险几率及漏失分布进行井漏预判，并智能推送防漏堵漏方案。为此，研发了基于大数据挖掘技术的防漏堵漏智能辅助决策平台。首先将大量已钻井漏失数据按聚类和关联规则算法进行自主聚类、关联、分析，同时根据挖掘出的规律对钻井漏失进行诊断，获取真实漏失关键参数，并对正钻井同一类井段的井漏倾向提出风险预测和预警，对堵漏施工提出针对性处理方案，从而提高防漏堵漏效果。使用防漏堵漏智能辅助决策平台对5口已完钻井和3口正钻井进行堵漏方案推送验证，其中推送方案与完钻井实际堵漏方案符合率60%，与正钻井实际堵漏方案符合率50%，表明推送的堵漏方案对实际堵漏方案制定具有一定的指导性。该软件在页岩气、高磨区块累计应用17井次，一次堵漏成功率达52.9%，未使用该软件时的一次堵漏成功率仅39.1%，取得了良好的应用效果。

In Sichuan-Chongqing area, lost circulation is serious, lost circulation treatment is mainly based on experience and the success rate of primary lost circulation control is low. To solve these problems, it is proposed to apply big data/artificial intelligent technology to carry out big data analysis and diagnosis on lost circulation characteristics and fracture properties, forecast lost circulation risk probability and lost circulation distribution of being drilled wells and intelligently recommend lost circulation prevention and control scheme. To this end, this paper researched and developed an intelligent assistant decision making platform of lost circulation prevention and control based on big data mining technology. Plenty of lost circulation data of drilled wells are autonomously clustered, correlated and analyzed by means of clustering and association rules algorithm, and then lost circulation is diagnosed according to the recognized laws, to obtain the key parameter of actual lost circulation. In addition, risk prediction and prewarning are proposed for the lost circulation tendency in the similar hole section of being drilled well, and the specific treatment scheme of lost circulation operation is prepared, so as to improve the lost circulation prevention and control effect. The intelligent assistant decision making platform of lost circulation prevention and control has already been used to recommend and verify the lost circulation control schemes of 5 drilled wells and 3 being drilled wells. The coincidence rate of the recommended scheme with the actual lost circulation control scheme of drilled wells is 60% and that with the actual lost circulation control scheme of being drilled wells is 50%, indicating that the recommended lost circulation control schemes play a certain guiding role in preparing the actual lost circulation control schemes. This software has been applied 17 well times in total in shale gas and Gaomo Block, with the success rate of primary lost circulation control up to 52.9%. Before the application of this software, however, the success rate of primary lost circulation control is only 39.1%. Obviously, the application effect of this software is remarkable.

10.13639/j.odpt.2021.04.008