Closing the Loop with the Subsurface: First Deployment of Autonomous Well Placement Following Reservoir Boundaries

Abstract

Abstract Industry-wide initiatives towards autonomous execution of drilling activities are maturing (Forshaw et al, 2023), and construction of wellbore sections is possible without pro-active human control. Engineers are now rather tasked to monitor the execution of the process and intervene in case of specific exceptions. While automatic drilling optimization primarily focuses on the control of rig operations, and automatic directional drilling focuses on drilling the predefined well plan, autonomous wellbore positioning focuses on maintaining wellbore position within the geologic and petrophysical environment without human intervention. It is about to become an established service by major oilfield service providers. Use of automated directional drilling in various environments (Hansen et al, 2020) has demonstrated that the system can steer a wellbore reliably and fully automatically according to plan. The automation of optimum well placement while dynamically following the reservoir is the ultimate challenge to maximize hydrocarbon recovery and production. This paper presents the results from the first deployment of an autonomous well placement system, designed to automatically navigate the reservoir. The service consists of automatic reservoir boundary mapping, automatic initiation of navigation advice (decision-making) followed by automatic well plan updates (within given constraints), and automatic calculation of optimized set points for steering within the reservoir. The control loop is closed by sending downlinks automatically to the rotary steerable system (RSS) to execute the updated trajectory without human interaction. The underlying system concept thus integrates reservoir navigation, real-time trajectory updates, and directional drilling control. Development of this well placement service was enabled by evaluation and optimization of system behavior using sophisticated, integrated model-in-the-loop simulations. These simulations are also an effective means in pre-job planning through understanding the expected system behavior for a given set of pre-well reservoir models and parameters used for navigation advice triggers and wellbore design parameters, such as dogleg severity limits. The holistic system simulation assures finding the "right" drilling parameters for maximum contact with the reservoir, allowing safe and smooth drilling operations. The first field results successfully prove the technical viability of this complex process automation approach. They reveal that the biggest added value lies in the significantly shortened response times between detecting a changing reservoir trend and actual steering action. The shortened response time in turn enables maximized reservoir contact for additional production. While legacy, manually operated workflows require frequent communication and alignment between three key stakeholders (E&P Operations Geologist, Reservoir Navigation Supervisor, Directional Driller), the integrated system enables a simple and fast check and agreement on system proposals. Of course, a manual modification of a system suggestion is always possible in case of any exceptions or undesired outcomes.