A risk assessment framework considering uncertainty for corrosion-induced natural gas pipeline accidents

Xinhong Li*, Jingwen Wang, Rouzbeh Abbassi, Guoming Chen

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

9 Citations (Scopus)


Corrosion degradation is one of the main causations of natural gas pipeline failure, which poses a severe threat to human life, assets, and the environment. This paper develops a methodology to assess the risk of corrosion-induced natural gas pipeline accidents considering the uncertainties in incident escalations. This approach can capture both the uncertainties in corrosion failure likelihood and the impact of pipeline leak accidents. A pair of limit state functions are established to estimate the corrosion failure probability of the pipeline. The pipeline accident from gas release to vapor cloud explosion (VCE) is modelled using empirical models. Subsequently, the uncertainties in gas pipeline corrosion failure accidents, reflected by some uncertain parameters, e.g., basic pipeline parameters, corrosion defect and environmental conditions et al., are identified. These identified parameters are discretized and described using a set of probability density functions. Eventually, Monte Carlo (MC) method is utilized to solve the established models. Besides, sensitivity analysis is conducted to study the effect of uncertain parameters on the likelihood and impact of a pipeline accident. A practical case is used to test the methodology, proving a valuable tool for risk assessment of corrosion-induced natural gas pipeline accidents.

Original languageEnglish
Article number104718
Pages (from-to)1-11
Number of pages11
JournalJournal of Loss Prevention in the Process Industries
Publication statusPublished - Feb 2022


  • Natural gas pipelines
  • Corrosion failure
  • VCE
  • Risk assessment
  • Uncertainty


Dive into the research topics of 'A risk assessment framework considering uncertainty for corrosion-induced natural gas pipeline accidents'. Together they form a unique fingerprint.

Cite this