Development of a human reliability assessment framework addressing specific needs and requirements of maritime operations

  • Sung Il Ahn

Student thesis: Doctoral Thesis

Abstract

The safety of life at sea is a top priority in maritime operations. Therefore, essentialprocedures and regulations are enforced to prevent loss of life. However, human error is one of the main contributors to accidents in safety-critical industries. For a human reliabilityassessment in the maritime domain, the main question is how we correctly understand thehuman factors in the maritime situation practically. This research study aims to develop practical frameworks, which are different optimal combinations of analysis methods corresponding to the different research scope, to evaluate human errors in maritime operations that can more cleverly be identified, quantified, and integrated into the probabilistic risk assessment. Thus, achieving the aim is expected to improve overall safety within the maritime domain. This research proposes four human reliability assessment frameworks corresponding to maritime systems' different complexities and interactions. Firstly, this study introduces the Bayesian CREAM framework, which is a method of determining the contextual mode for overall human error probability estimation. This method aims to determine the need for more specific HRA research and to support the quick analysis required by providing a simple and imminent calculation method. Second, the CREAM-based framework is developed to extend human errors to various human activities. This method minimises the expert’s subjectivity while achieving a quantified human failure probability with a systematic and logical approach. Third, the SPAR-H-based framework is proposed to integrate human errors into a probabilistic risk assessment framework. This framework offers a new approach to human reliability, assessed through a customised reliability block diagram analysis to provide a new risk model. Finally, for a complex modern socio-technical system, a hybrid method is proposed to assess system reliability by combining the System Theoretic Process Analysis (STPA) and the Success Likelihood Index Method (SLIM). The proposed frameworks were applied to emergency response operations, including emergency steering, engine room fires, man overboard, and emergency shutdown system for the LNG bunkering process. The various frameworks established for human reliability evaluation will contribute to the wider utilisation of human reliability assessment, ranging from simple and convenient analysis performed by ship's crew to complex analysis performed by expert human reliability analysts. Eventually, this study will enhance maritime safety by analysing human errors, identifying further problems, and adopting safety measures.
Date of Award4 Aug 2022
Original languageEnglish
Awarding Institution
  • University Of Strathclyde
SponsorsUniversity of Strathclyde
SupervisorRafet Kurt (Supervisor) & Osman Turan (Supervisor)

Cite this

'