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

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 real-time 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.