Developing a Set of Human Factors Barriers for Deepwater Drilling Risk Assessment

Abstract

Abstract Objectives/Scope Human error and human factors are contributing causes in over 50 percent of well control incidents, according to SINTEF data. While straight-forward cognitive errors and wrong actions contribute to 15 percent of incidents, the larger span of human factors includes communication failures, workload, management practices, and information availability and presentation problems that lead to mistakes, such as imperfect kick detection or underbalanced wells. Yet human factors is often treated as an issue of human reliability estimates and the need to define better procedures that design human error out of the system. There are, however, many shortcomings to the focus on procedures and human reliability estimates, for example, variations among installations, situations that lie outside existing procedures and estimates, and analyses of rote procedure-following that exclude situation variability and expert problem solving. What is needed is a deeper, human factors-centric, approach that looks beneath the procedures for underlying systemic causes: operator situation awareness and decision making, team common ground and collaboration, and resilient socio-technical organizational processes. Our project is taking this human factors approach with a comprehensive analysis of well control during deepwater drilling. Methods, Procedures, Process Our novel approach is to marry the risk management framework of threats and barriers with the human factors concept of the decision cycle. The decision cycle concept outlines the cognitive steps involved in decision making from detection, to interpretation, deciding, and acting. Our innovation is that these concepts can be brought together to identify a set of 117 "human factors barriers" for effective detection of an influx. These human factors barriers support team situation awareness, communication, and decision making. These barriers include the availability and representation of data, user vigilance, support for critical thinking, crew resource management, shift change hand-off processes, system status availability, and change management processes. These human factors barriers can be applied readily to other decisions within well control. Results, Observations, and Conclusions We have created a database of BSEE/BOEM incident reports and reviewed these incidents against the human factors barriers. We have identified a number of "hot spots" or trends where specific human factors barriers were frequently compromised. These trends include sensor placement for situation awareness, shift hand-off processes and team communication, and decision support for critical thinking and diagnosis of uncertain situations. Novel/Additive Information These human factors barriers and the identified trends can be used a) pre-incident to review and audit drilling rigs and well plans to identify and then avoid potential problems, b) post-incident to investigate and report on incidents from a human factors perspective and thereby better inform the industry concerning these topics, and c) to identify key topics for further research and development of improved systems and processes.