A hierarchical Bayesian approach to modelling fate and transport of oil released from subsea pipelines

Ehsan Arzaghi, Mohammad Mahdi Abaei, Rouzbeh Abbassi, Vikram Garaniya, Jonathan Binns, Christopher Chin , Faisal Khan

Research output: Contribution to journalArticlepeer-review

33 Citations (Scopus)


The significant increase in global energy demand has drawn the attention of oil and gas industries to exploration of less-exploited resources. Arctic offshore region is reported to hold a great proportion of un-discovered oil reserves. While this can be a promising opportunity for the industry, more exploration activities will also increase the possibility of oil spill during the entire process including production and transport. A comprehensive risk assessment based on Ecological Risk Assessment (ERA) method is then required during the planning and operation stages of future Arctic oil production facilities. In the exposure analysis stage, ERA needs an evaluation of the oil concentration profile in all media. This paper presents a methodology for predicting the stochastic fate and transport of spilled oil in ice-infested regions. For this purpose, level IV fugacity models are used to estimate the time-variable concentration of oil. A hierarchical Bayesian approach (HBA) is adopted to estimate the probability of time to reach a concentration (TRTC) based on the observations made from a fugacity model. To illustrate the application of the proposed method, a subsea pipeline accident resulting in the release of 100 t of Statfjord oil into the Labrador Sea is considered as the case study.
Original languageEnglish
Pages (from-to)307-315
Number of pages9
JournalProcess Safety and Environmental Protection
Publication statusPublished - Aug 2018


  • Fugacity
  • Fate and transport
  • Bayesian inference
  • Hierarchical Bayesian modelling
  • Oil spill


Dive into the research topics of 'A hierarchical Bayesian approach to modelling fate and transport of oil released from subsea pipelines'. Together they form a unique fingerprint.

Cite this