Event Detection for Managed-Pressure Drilling: A New Paradigm

Abstract

Abstract Effective event detection has traditionally been a largely manual process, exclusively in the realm of experienced personnel, who not only continuously monitor drilling parameters but also identify the current state of operations based on this data. Owing to the large number of variables and inherent uncertainties in drilling operations today, this approach still provides one of the most reliable methods. However, as a result of the advent of managed-pressure drilling and the need to carefully control and manipulate bottomhole pressures within a specified "window," the proper diagnosis and management of well events has become even more critical. Furthermore, with the popularization of an "automated" control philosophy, this can only be fully realized with an integrated event-detection system. Because a manual approach to event detection does not provide the mechanisms required to support such a philosophy, a real-time automated event-detection system is essential. A number of approaches have been tried in the industry to implement such an event-detection system with varying, but limited, success. The approach has typically been based on a collection of rigid and archaic logic-based statements, which attempt to overly simplify and abstract operational complexity, leading to a superficial, limited, and inefficient system. This paper presents a new, novel, and efficient approach to a real-time, automated event-detection system for managed-pressure drilling operations. Events will be discussed and followed by a detailed description of the system, algorithms, and approaches.