Reliability Engineering and System Safety 206 (2021) 107304 Available online 10 November 2020 0951-8320/© 2020 Elsevier Ltd. All rights reserved. Identifying key factors affecting the performance of team decision-making based on the analysis of investigation reports issued from diverse industries Dong-Han Ham a , Won-Jun Jung a , Jinkyun Park b, * a Chonnam National University, Republic of Korea b Korea Atomic Energy Research Institute (KAERI), Republic of Korea A R T I C L E INFO Keywords: Probabilistic safety assessment Human reliability analysis Team decision-making performance Key factors Event investigation report ABSTRACT Operational experience of sociotechnical systems has revealed that the degradation of their safety is attributable to human error. Accordingly, various kinds of human reliability analysis (HRA) techniques have been proposed over the past several decades for safety enhancement. The Fukushima accident, however, stressed that existing HRA techniques have a limitation in estimating the human error probability (HEP) of safety critical tasks that should be conducted under challenging circumstances (e.g., insuffcient or misleading information) originating from the characteristics of a severe accident condition. This means that the very frst step to properly estimate HEPs in severe accident conditions is to identify a catalog of key factors related to the performance of team decision-making tasks. This study therefore suggests a conceptual model based on signifcant factors pertaining to the performance of team decision-making tasks. Event investigation reports issued from diverse sociotechnical systems were analyzed along the proposed model, and as a result, 14 key factors were identifed that could be a good starting point to scrutinize the performance of team decision-making tasks. 1. Introduction From the perspective of the sustainability of sociotechnical systems, the crucial requirement is to maintain, to the greatest extent possible, operational risk within a minimum level. In the case of a nuclear power plant (NPP), as a representative sociotechnical system with a particular emphasis on safety over its entire lifetime, this requirement is more stringent than in other industries. Probabilistic safety assessment (PSA) or probabilistic risk assessment (PRA) techniques have accordingly been widely used for several decades to evaluate the operational risk of NPPs. Without loss of generality, the PSA technique is useful for estimating the risk level of a target NPP by answering a couple of basic questions such as ‘What could go wrong?’, ‘How likely is it?’, and ‘What would the consequence be?’ [4]. In this light, the PSA technique provides a sys- tematic approach to identifying a catalog of plausible accident scenarios that lead to adverse consequences with associated frequencies (e.g., the frequency of core damage or the frequency of radioactive material release to the environment). To this end, it is important to identify the key systems and/or components in which failure directly results in the adverse consequences. Since the operating experience of sociotechnical systems continuously reveals that human error is one of the determinant contributors to the occurrence of incidents and accidents [1–4], it is indispensable to consider the likelihood of human errors that are able to engender the failure of the key systems and/or components (i.e., safety critical tasks). Accordingly, many kinds of human reliability analysis (HRA) techniques have been proposed for several decades, of which the major role is to estimate the human error probabilities (HEPs) of human operators who have to conduct the safety critical tasks [20]. These HRA techniques allow us to theoretically estimate HEPs by considering diverse contextual factors that affect the performance of human opera- tors. These contextual factors have slightly different names such as performance shaping factors (PSFs), infuence factors, common perfor- mance condition, and error producing conditions. From these concerns, PSAs have been traditionally performed at three levels (Level 1, Level 2, and Level 3 PSA), each with a dedicated purpose and scope. In brief, Level 1 PSA focuses on accident scenarios leading to core damage, while Level 2 PSA deals with accident scenarios that challenge the containment integrity. Level 3 PSA emphasizes public health and other societal risks (e.g., the contamination of land or food) following the potential release of radioactive materials [34]. After the Fukushima accident, however, the nuclear industry faced with several issues that were not properly resolved (or not even * Corresponding author: 989-111 Daedeokdaero, Yuseong-Gu, Daejeon, 34057, Republic of Korea. E-mail address: kshpjk@kaeri.re.kr (J. Park). Contents lists available at ScienceDirect Reliability Engineering and System Safety journal homepage: www.elsevier.com/locate/ress https://doi.org/10.1016/j.ress.2020.107304 Received 5 April 2019; Received in revised form 7 September 2020; Accepted 7 November 2020