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

Real-time intelligent drilling monitoring technique based on the coupling of drilling model and artificial intelligence

钻井过程中钻井复杂监测对于减少事故发生、降低钻井成本意义重大。在实际钻井过程中,钻井复杂的分析和判断主要靠人工完成,难以保证预警效率。为此,将动态钻井物理模型与人工智能、数据挖掘算法相结合,提出基于实时录井数据的实时钻井监测及事故预警技术。该技术以钻井施工过程中的综合录井数据作为输入,利用模型算法来实时准确呈现钻井过程中的井下工况条件,预测即将发生的复杂风险。从实时井眼清洁及水力学监测、实时卡钻预测、实时井涌监测3个方面对实时钻井监测及预警技术进行了详细分析。该技术可以实时准确模拟井下工况条件,识别并降低钻井事故发生概率,帮助钻井工程师提早发现问题,减少钻井事故的发生,并减少对自然环境以及人员安全的影响,为现场施工提供辅助决策,降低非有效生产时间。

Drilling complexity monitoring in the process of drilling is of great significance to reduce accidents and drilling costs.In the process of actual drilling,drilling complexity analysis and discrimination is mainly implemented artificially,so the prewarning efficiency can be hardly guaranteed.In this paper,the real-time drilling monitoring and accident prewarning technique based on realtime mud logging data was put forward by combining dynamic drilling physical model with artificial intelligence and data mining algorithm.In this technique,the composite mud logging data in the process of drilling construction is input,the model algorithm is used to accurately present the downhole conditions in the process of drilling in real time to predict the impending complexity risk.Then,this technique was analyzed in detail from three aspects,i.e.,real-time hole cleaning and hydraulic monitoring,real-time pipe sticking prediction and real-time well kick monitoring.It is indicated that this technique can accurately simulate the downhole conditions in real time,identify drilling accidents,reduce its occurrence probability and assist drilling engineers to find out the problems in time to reduce the drilling accidents.In addition,it can diminish the impact on natural environments and personnel safety,provide the auxiliary decisions for field construction and reduce the non-effective production time.