Improve the Drilling Operations Efficiency by the Big Data Mining of Real-Time Logging

Abstract

Today's data is tomorrow's oil and gas. Only the data can tell us right or wrong, but not the experience or feel. The real-time logging data conforms to the 6V features of the Big Data (Velocity, Variety, Volume, Veracity, Visualization and Validity). As a result, the drilling operations efficiency is significantly improved by the Big Data mining of real-time logging. The Big Data mining helps recognize the drilling operations automatically and identify the invisible non-production time (INPT). Firstly, the real-time logging data is acquired by the comprehensive logging unit. Secondly, the drilling operations are recognized by applying restrictions to drilling parameters. Thirdly, the daily time and the total time is breakdown based on the above logging data and operations categories. Finally, the INPT is identified by setting the target value based on the Big Data mining and learning curve. The savings potential, which is determined by the average and target, is critical to improve the operations efficiency. On one hand, the Normal Distribution is established by setting the specific operation time (such as slips connection time, etc.) to the X-axis and the operation count to the Y-axis. The average and covariance of the Normal Distribution is calculated. On the other hand, the target value is based on the Big Data mining by the Bayesian network model and the total-time learning curve of batch wells. As a result, the real-time drilling efficiency can identify the best-performing operations, crews and rigs. The crew-based operational performance comparisons are effective to identify the best-performing crews and to indicate where best to focus training and crew supervision efforts in the future. And the real-time drilling efficiency can measure the rig performance in order to make INPT visible. At last, the real-time drilling efficiency yields cost and time savings on both deep-water and complex wells. The method is successfully applied to BD gas field in Indonesia which is HTHP and LW oil field in South China Sea which is in deep-water. Application shows the INPT represents about 32%. The Big Data mining of real-time logging significantly improve the drilling operations efficiency, detect and minimize the INPT. As a result, the Big Data mining yields cost and time savings for tomorrow's market.

Human Centric Digital Transformation and Operator 4.0 for the Oil and Gas Industry

“…The invention, improvements and application of new data recording tools (e.g., seismic acquisitions devices, channel counting, fluid front monitoring geophones, carbon capture and sequestration sites, logging while drilling), measurement and data formats have made it even more applicable to employ big data analytic tools in drilling operations [100]. In this area, big data analytics can be deployed to improve the drilling performance [105], find invisible non-production time [106], and reduce the risks associated with drilling operations, the risks of drilling failures, and can also lower drilling development costs [107], [108]. A report done by B. Marr, [109] noted that Shell uses big data analytics in near realtime to detect failures.…”

Procedure of Analysis and Digital Processing of Large Data Array in Real Time upon Intellectualization of Roller Drilling

“…Faults of control system of this process can be caused by minor oscillation of the variables. In order to solve this problem, it is required to apply Data mining technologies aiming at detection of usable data in large data arrays, detection of regularities and trends in the existing data applicable to roller drilling [7][8][9], as well as the technologies of predictive analytics aimed at achievement of reliable forecasts on the basis of processed and structured digital information about rock properties using intelligent system of roller drilling control and estimation of interrelation between various factors, their interpretation and estimation of risks for prevention of faults in operation of control system. Using Data mining technologies for digital processing of information about rock properties by control system of roller drilling would allow to reveal regularities and trends in data, which usually cannot be detected using conventional analysis of data due to complex bonds or excessive data amount.…”

Drilling Conditions Classification Based on Improved Stacking Ensemble Learning

“…YIN Qishuai et al [2] proposed a program based on working condition recognition method, which effectively eliminated the influence of personnel on drilling condition recognition. This method, however, did not take into account the relationship between the data before and after or the potential fluctuation of drilling data because it was based on programming language.…”