Think Like a Well, Act Like a Facility: The Autonomous Fusion of Subsurface and Surface

Abstract This paper presents a field-driven enhancement to ROBOWELL®, ADNOC Offshore's Autonomous Well Control system, designed to optimize oil and gas production operations. The objective of this study is to address a recurring operational challenge: optimizing well performance without exceeding the handling capacity of surface facilities, particularly the slug catcher. During ROBOWELL® deployment, the system demonstrated effective well optimization; however, instances were observed where increased gas production exceeded slug catcher limits, leading to surface instability. To mitigate this, a control logic module was developed to incorporate real-time slug catcher level data into the autonomous decision framework. This logic continuously monitors downstream buffer capacity and dynamically constrains well adjustments based on available processing headroom. The new module was integrated into the existing Advanced Process Control (APC) framework, enabling closed-loop coordination between reservoir inflow and surface processing readiness. As a result, the system now modulates its optimization behavior in response to downstream conditions, holding back when slug catcher levels are high and resuming optimization when capacity is available. Field results demonstrated a significant reduction in operational upsets and manual interventions, with sustained production optimization that respected downstream thresholds. This advancement redefines the role of autonomous well control from a localized optimizer to a facility-aware orchestrator, marking a step change in digital oilfield automation. The integration of upstream and downstream logic within a single autonomous system represents a novel approach in field operations. It validates the potential for control systems to make facility-aware decisions without human input, paving the way for fully connected, autonomous production environments that maximize value across the hydrocarbon value chain. This work contributes to the broader vision of intelligent field automation by demonstrating how real-time data integration across production domains can enhance system responsiveness and stability. The approach offers a scalable model for other facilities seeking to align subsurface optimization with surface constraints and sets a precedent for future developments in autonomous oilfield control systems.