Case Study of Weak Signals and Confirmation Bias in Drilling Operations

Abstract

Summary The paper presents a case study of the circumstances leading up to a situation in which a supposedly vertical well ended up at 14° before the deviation was detected and a correction could be made. This case study illustrates how a relatively routine activity could end up off the plan, and also highlights the impact of weak signals and confirmation bias on the situation. Process safety has become a focus of attention, with much work performed to ensure that organizations are educated and trained in the techniques necessary to avoid catastrophic incidents. Human error has come under the spotlight and “human factors,” including nontechnical skills, are being included in training programs, such as well control. However, the language around human factors and the psychological terms used, such as “cognitive bias,” “chronic unease,” and “weak signals,” are abstract and unfamiliar to the drilling community. The recent emphasis on human factors has mostly been placed on well-control training, suggesting that nontechnical skills are only required during abnormal events. Yet, as this case study will show, developing these skills, in general, brings benefits in routine operations to avoid what might be regarded as “normal” nonproductive-time incidents. By describing a real and very recent well incident, the paper shows how the same set of circumstances can be investigated from several different perspectives and how one can draw different conclusions about root cause, depending on the agenda of the investigator. There is no doubt that no single perspective fully explains what happened in this incident, and why, but traditionally there is little evidence of a human-factors perspective being adopted during incident investigations beyond the use of the label of “human error.” The paper shows how, without the benefit of hindsight, the rig team made a perfectly reasonable set of decisions under the circumstances with which they thought they were being confronted. There was evidence in some of the data to suggest that the well was deviating from the vertical, but there were competing pressures on the team for them to disregard the warnings. Although there are certain obvious procedural and technical lessons to be drawn from the event, deeper lessons relate to the difficulties associated with overcoming the tendency to see what is expected and the central role played by leaders to support and commit to the time and cost of precautionary actions. It is hoped that the paper will encourage the drilling community to look beyond the apparent interpretations of an event to consider why it was that otherwise sensible people did what they did. There are many incidents of this type from which the community can learn at a relatively low cost, particularly focusing on the human-factors aspects. Yet, the learning from these incidents and the training to avoid them will, ultimately, develop the skills that will be instrumental in helping teams identify and react to conditions that might otherwise develop into a catastrophe.