Predicting human reliability for emergency fire pump operational process on tanker ships utilising fuzzy Bayesian Network CREAM modelling

Muhammet Aydin, Sukru Ilke Sezer, Seher Suendam Arici, Emre Akyuz*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

An emergency fire pump is a critical equipment which assists the ship crew in handling extreme emergency situations involving fire on-board tanker ships. In case the main fire pump becomes ineffective during a fire, the emergency fire pump is used to extinguish the fire. In this case, it is necessary for the ship's crew to operate and use the pump without any failure. The aim of this study is to systematically predict human (crew) reliability for emergency fire pump operational processes on tanker ships since human dependability plays a significant role in safer shipment. To achieve this goal, fuzzy and BN (Bayesian Network) CREAM (Cognitive Reliability and Error Analysis Method) modelling is applied. CREAM is used to methodically estimate the probability of human error within a fuzzy set that accounts for uncertainties of CPC (Common Performance Condition), while BN is competent in estimating control modes in CREAM's reliability analysis. The paper's findings show ship crew reliability (9.08E-01) for the emergency fire pump operation process on tanker ships. These findings are expected to provide valuable information to prevent human errors and improve safety on tanker ships during firefighting, thereby reassuring the maritime industry of the potential for increased safety.

Original languageEnglish
Article number119717
JournalOcean Engineering
Volume314
DOIs
Publication statusPublished - 15 Dec 2024

Bibliographical note

Publisher Copyright:
© 2024 Elsevier Ltd

Keywords

  • Bayesian network
  • CREAM
  • Emergency fire pump
  • Human reliability
  • Operational safety

Fingerprint

Dive into the research topics of 'Predicting human reliability for emergency fire pump operational process on tanker ships utilising fuzzy Bayesian Network CREAM modelling'. Together they form a unique fingerprint.

Cite this