Time series data analysis for automatic flow influx detection during drilling

Abstract Automatic and early detection of flow influx during drilling is important for improving well-control safety. In this paper, a new method that can automatically analyze real-time drilling data and detect the flow influx event is presented. The new method combines the physics-based dimension reduction and time-series data mining approaches. Two kick indicators are defined, representing the drilling parameter group (DPG) and flow parameter group (FPG), respectively. Additionally, two real-time trend-analysis methods, the divergence of moving average (DMA), and the divergence of moving slope average (DMSA) are applied to quantify trend evolutions of the two indicators. The kick event is identified based on the anomalous trends held by the two kick indicators. A final kick-risk index (KRI) is calculated in real time to indicate the probability of kick events and to trigger the alarm. The method is tested against four offshore kick events. With KRI threshold setting as 0.8, the average detection time is 64% less than the reported detection time. The application of DPG kick indicator allows the early kick detection without additional downhole sensors or costly flow meters.

[1]  William Cox,et al.  SMART Kick Detection: First Step on the Well-Control Automation Journey , 2015 .

[2]  C. A. Johancsik,et al.  Torque and drag in directional wells-prediction and measurement , 1983 .

[3]  Don Reitsma.  Development of an Automated System for the Rapid Detection of Drilling Anomalies using Standpipe and Discharge Pressure , 2011 .

[4]  Brian Tarr,et al.  Next-Generation Kick Detection During Connections: Influx Detection at Pumps Stop (IDAPS) Software , 2016 .

[5]  O. J. Shirley,et al.  Application of Drilling Performance Data to Overpressure Detection , 1966 .

[6]  Shuai Zhang,et al.  A new pattern recognition model for gas kick diagnosis in deepwater drilling , 2018, Journal of Petroleum Science and Engineering.

[7]  Benoit Daireaux,et al.  Toward Drilling Automation: On the Necessity of Using Sensors That Relate to Physical Models , 2014 .

[8]  N. P. Brown,et al.  Slimhole Early Kick Detection by Real-Time Drilling Analysis , 1997 .

[9]  Mark Smith,et al.  Stuck-Pipe Prediction by Use of Automated Real-Time Modeling and Data Analysis , 2017 .

[10]  Stephen Lai,et al.  Enhanced Kick Detection with Low-Cost Rig Sensors Through Automated Pattern Recognition and Real-Time Sensor Calibration , 2015 .

[11]  Dev Kumar,et al.  Automated Trend-Based Alerting Enhances Real-Time Hazard Avoidance , 2017 .

[12]  Tak-Chung Fu,et al.  A review on time series data mining , 2011, Eng. Appl. Artif. Intell..

[13]  G. Chilingar,et al.  Chapter 6 Drilling parameters , 2002 .

[14]  Phil Griffin.  Early Kick Detection Holds Kill Pressure Lower , 1967 .