A risk analysis study for chemical cargo tank cleaning process using Fuzzy Bayesian Network

Yunus Emre Senol*, Fatma Yasli

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

35 Citations (Scopus)

Abstract

Chemical cargoes are of extremely high purity and could be contaminative, reactive or incompatible with each other. Delivering the cargo in a condition that is as pure as possible at the loading port constitutes the main goal of chemical cargo transportation. After discharging a cargo, it is absolutely critical to make the tanks free of all possible contaminants and make them ready for the next cargo to be loaded. Despite tank cleaning procedures and contaminant detection methods, costly cargo contamination may still be encountered due to dirty remained ship's tanks. In this study, a comprehensive risk assessment is carried out in order to develop risk prevention strategies by increasing the efficiency of tank cleaning operations. A “Dirty Tank Model” is constructed with a Bayesian Network to obtain factors causing dirty tank and investigate them with their causal relationships. Due to insufficient data for the study, expert opinions are used as a mandatory data source utilising Fuzzy Set Theory. The root causes related to the dirty tank are identified following comprehensive reasoning inquiries performed with those experts. The results provide effective information for developing appropriate risk strategies and tailoring them depending on the different conditions related to the tank cleaning processes.

Original languageEnglish
Article number109360
JournalOcean Engineering
Volume235
DOIs
Publication statusPublished - 1 Sept 2021

Bibliographical note

Publisher Copyright:
© 2021 Elsevier Ltd

Keywords

  • Bayesian network
  • Chemical cargo contamination
  • Dirty tank
  • Fuzzy set theory
  • Risk analysis
  • Tank cleaning process

Fingerprint

Dive into the research topics of 'A risk analysis study for chemical cargo tank cleaning process using Fuzzy Bayesian Network'. Together they form a unique fingerprint.

Cite this