The State Recognition And The Time-dependent Analysis Based On Mud Logging Data In Drilling Operations

In recent years,with the rapid development of big data,artificial intelligence and other information technology,machine learning has been widely used in the field of petroleum engineering.Based on a large number of mud logging data from offshore batch wells,this paper establishes a state recognition model based on artificial neural network and a time-dependent analysis method based on the continuous improvement "Plan-Do-Study-Act" cycle in drilling operations,and develops the corresponding software system.This system consists of state recognition subsystem and time-dependent analysis subsystem in drilling operations.The state recognition subsystem includes real-time mud logging data transmission and state recognition module,the mud logging data transmission interface based on the well site information transmission specification and well site information standard markup language is developed,which realizes the real-time sharing of well site data;based on the mud logging data,an artificial neural network algorithm is used to establish the state recognition model.Based on the state recognition,the time-dependent analysis subsystem is developed,including the automatic statistics drilling operation time module and the continuous improvement drilling efficiency module: the automatic statistics drilling operation time module counts the start time,end time and duration of different drilling operations;the continuous improvement drilling efficiency module divides the whole drilling process into specific drilling operations,and establishes the key performance indicators of each drilling operation.Comparisons of the actual drilling operation time and key performance indicators are used to find invisible lost time in drilling operation.In addition,it is necessary to carry out the learning of standardized drilling operations,to reduce unnecessary drilling operation time,improve drilling efficiency and shorten invisible lost time.The state recognition and the time-dependent analysis system have been successfully applied to the offshore batch drilling in an Oil and gas field,Bohai Sea.The accuracy of the artificial neural network model established to do state recognition in this paper can reach 91% and continuous improvement time-dependent analysis method has shortened 36.01% of invisible lost time.Through the accurate state recognition and the continuous improvement time-dependent analysis of this system,the refined management of the drilling process is achieved,the invisible lost time in the drilling process is shortened,the well construction period is reduced,the drilling cost is saved,and drilling quality and efficiency is improved to a certain extent.