In keeping with its pioneering spirit and convinced that digital technology can significantly improve drilling performance, the Drilling & Wells Métier has launched the DrillX (Drilling eXperts) program. A cross-disciplinary team of drillers, data scientists and developers is working in Agile mode to deliver and deploy new monitoring and advanced analysis solutions, including predictive analysis, using proprietary machine learning algorithms.
Real-time data gathering started at the end of 2015 with the inauguration of the Real Time Support Center (RTSC) in Pau, a facility for the remote monitoring of operations. This center has been gathering data and supporting operations for more than 200 wells over the past four years.
For Total to ensure both its lead in technology and deliver wells at optimal cost and in complete safety, it is essential to have modern, integrated and automated tools for processing and gathering data from operations in real time (monitoring algorithms, reports, alerts, predictive analyses, risk predictions).
"Smart Operations": Automatic & Intelligent Real-Time Monitoring of Drilling Operations
Drilling rigs operated by Total are now all connected online and being monitored by the RTSC. This center gathers raw data continuously from various sensors (surface and downhole) on our drilling rigs and tools. The DrillX functionality then cleans, transforms and stores this data on the Total Cloud, where artificial intelligence algorithms use it to create meaningful insights.
These algorithms continuously analyze all relevant data to predict the risk of occurrence of each type of operational event. This Machine Learning approach is particularly well suited to:
- Giving an immediate overview of all past events, thereby providing our Supervisors with the Group’s operational “memory”.
- Obtaining a systematic automated analysis of a significant number of different parameters in order to identify any correlations, deviations and/or small anomalies that might be indicators of the possible occurrence of a more serious event with the potential to affect our drilling operations.
Improving Safety and Performance In Drilling Operations
Deployed and utilized on all our drilling rigs, the DrillX tools will further help to prevent the risks associated with our operations, which have the potential to cause major industrial disasters. An unidentified and uncontrolled influx of reservoir fluid during drilling may lead to a blowout. The DrillX artificial intelligence algorithms learn from previous well incidents data. We are therefore able to predict any increase in the risk of occurrence of these events, with the ultimate objective of preventing them, rather than solving them after the event has occurred.
This deployment will help to improve drilling performance and will reduce our operating costs. In 2019, the Group’s Drilling & Wells expenditure totaled about $4 billion. The development of an algorithm that is able to automatically identify differences in the efficiency of two drilling operations will enable an immediate comparison of the procedures used that can then be optimized in order to increase operating efficiency and achieve significant savings.
Operations teams will be able to visualize their wells' status at all times, monitor drilling performance in real time and reliably anticipate risk situations a number of hours in advance, enabling better informed decisions to be made more rapidly.
DrillX Drillers Trained in the Agile Method
13 use cases for DrillX were identified during the ideation phase. The Group selected four of them as initial priorities:
- performance monitoring and benchmarking;
- real-time engineering simulations;
- monitoring the hydrodynamic balance of wells;
- predicting the risk of stuck events of drilling assemblies.
To begin with, several months were spent on preparing the data sets for past incidents, a vital prerequisite for the launch of the first data science work.
Then, starting in April 2019, the teams ran two projects in parallel, working in Agile mode. This close collaboration between different métiers (drilling, IT, data science) enabled the rapid delivery of the first two applications. Since September 2019, they have been used daily on five drilling rigs to predict the risk of influxes during drilling and the risk of stuck drilling assemblies.
All monitoring and predictive algorithm services already developed will be made available to the teams at the end of July 2020 in an initial “industrial” version of the DrillX integrated application platform. With effect from this date, new services and algorithms covering other use cases will continue to be integrated as and when they are delivered.
Drilling & Wells
Drilling Wells Safely & Cost-Effectively
Drilling & Wells
Enhance Well Risk Management
Drilling & Wells
Drilling in an HP/HT Environment in Brunei