Automation of well construction combines process and machine automation to deliver cost savings and efficiency gains, alongside safer operations and faster collaborative decision making. Integration of well planning and execution improves performance, minimizes risk, and creates the framework for batch control of well construction. Generating intuitive and standardized insights from historical and live data streams enables reducing uncertainty and driving technical limit performance at every stage, for the safest and most economical well delivery possible.
Remote operations have become standard practice in well construction and drilling automation is rapidly growing. Machine-learning based models can predict hazards and best operating parameters. The paper describes how these elements are combined to easily analyze offset well data from multiple sources for performance benchmarking, deep technical analysis, and risk management. Integration with physics-based models (the digital-twin) establishes a coherent execution roadmap, a real-time digital recipe that is directly used in rigsite automation. During execution, model assumptions are replaced by live sensor data. Those models are then re-calculated in real-time for automated process control and become available for future planning.
The viability of integrated well planning and automation systems is no longer disputed, and industry focus has shifted towards proving the potential of this approach. Pertinent case examples from drilling operations are examined in the paper, including optimizing performance through harnessing analytics of large data sets from offset wells, and on exploring the integration of AI techniques for risk prediction and mitigation. The insights gained, coupled with the digital well plan, are instrumental for optimization and automation of tripping equipment in and out of hole, and of drilling ahead operations.
Integrating the digital well plan with "lessons learned" from the analysis of all available data and using a digital-twin concept with physics-based and data driven models, provides the foundation for the next step in process automation: the creation of digital procedures for process automation of well construction. This capability extends the cost savings and efficiency gains realized in recent years through remote operations.