A Shift-Level Decision Framework for Troubleshooting Process Upsets in Oil Sands Extraction and Upgrading Operations

Abstract

Oil sands extraction and upgrading operate through highly interconnected process systems in which disturbances can propagate rapidly across separation, froth treatment, upgrading, hydrogen, utilities, and supporting infrastructure. This review examines how shift personnel can diagnose and manage such disturbances more consistently under conditions of uncertainty, time pressure, competing alarms, and incomplete information. The study adopts a structured narrative review approach, synthesizing literature on oil sands processing, abnormal situation management, process monitoring, fault diagnosis, risk prioritization, operational analytics, human factors, resilience, and decision support published up to 2019. The review identifies that effective troubleshooting depends on distinguishing initiating events from propagated symptoms, validating instrumentation, interpreting multivariable process behaviour, and integrating field observations with historical and real-time operating data. It further finds that response quality improves when actions are prioritized according to personnel safety, containment, safeguard integrity, equipment protection, environmental compliance, process stability, and recovery potential rather than production urgency alone. Human factors, including situational awareness, communication, workload, shift handover, procedural discipline, and escalation practices, are shown to be decisive in determining whether technical information is translated into effective action. Drawing these findings together, the study develops an integrated shift-level framework comprising deviation recognition, immediate stabilization, causal diagnosis, response prioritization, corrective intervention, verification, escalation, controlled recovery, and organizational learning. The review concludes that structured decision logic can strengthen consistency, transparency, and resilience in troubleshooting complex process upsets. It recommends embedding the framework within site procedures, operator training, alarm-management practices, simulation exercises, and digital decision-support platforms, while future studies should validate its effectiveness using plant data, simulator trials, and post-event performance analysis.