Autonomous Directional Drilling Using the RSS Simulator for the Drillbotics 2020-2021 Virtual Competition

Abstract

Abstract The automation of directional drilling is a recent development that is expanding with both physical and virtual concepts. In the same way, SPE's Drilling Systems Automation Technical Section (DSATS) supports the creation of new automated directional drilling models and ideas with its international competition Drillbotics®. This paper describes the design, algorithms, graphical user interface (GUI), and features of the rotary steerable system (RSS) Simulator V2.4, developed by the winning team of the 2020-2021 Drillbotics Virtual Rig Competition. The RSS Simulator, developed in Python, simulates a drilling trajectory of a directional well based on target coordinates, dogleg severity (DLS) constraints, and cubic Bezier curves. The simulator has two general modules: the RSS Model and the Trajectory Control Optimizer (TCO). The first one simulates the behavior of the bit and bottom hole assembly (BHA) tool (e.g., force on the bit, inclination, azimuth) considering the uncertainties at survey stations using the RSS Model developed by the University of Stavanger. The TCO autonomously creates the well plan, controls the bit's deviation, and corrects the trajectory when it has diverged from the planned trajectory. The simulator can drill either 2D or 3D trajectories. In the case study, one 3D trajectory example provided by the Drillbotics committee is illustrated. The results obtained show differences among the original (given), the planned (optimal, before drilling), and the simulated (during drilling) trajectories. Moreover, the comparison between the simulation with/without the TCO module reveals the necessity and importance of autonomous and real-time trajectory control. Besides, the BHA tool activation (due to the autonomous control of the simulator) is perceived in the simulation's results, which influences the rest of the parameters like the DLS, the force acting over the bit, and the rate of penetration (ROP) in the axial axis, inclination axis, and azimuth axis calculated with the RSS Model. The work presents a novel way to predict the bit position and control the trajectory, considering bit position uncertainties, drilling parameters uncertainties, DLS, and tortuosity calculations in an automated way. It will help users to develop further automated algorithms and mathematical models for reaching autonomous directional drilling in the future.