Risk and uncertainty analysis of gas pipeline failure and gas combustion consequence

Abstract

Taking into account a general concept of risk parameters and knowing that natural gas provides very significant portion of energy, firstly, it is important to insure that the infrastructure remains as robust and reliable as possible. For this purpose, authors present available statistical information and probabilistic analysis related to failures of natural gas pipelines. Presented historical failure data is used to model age-dependent reliability of pipelines in terms of Bayesian methods, which have advantages of being capable to manage scarcity and rareness of data and of being easily interpretable for engineers. The performed probabilistic analysis enables to investigate uncertainty and failure rates of pipelines when age-dependence is significant and when it is not relevant. The results of age-dependent modeling and analysis of gas pipeline reliability and uncertainty are applied to estimate frequency of combustions due to natural gas release when pipeline failure occurs. Estimated age-dependent combustion frequency is compared and proposed to be used instead of conservative and age-independent estimate. The rupture of a high-pressure natural gas pipeline can lead to consequences that can pose a significant threat to people and property in the close vicinity to the pipeline fault location. The dominant hazard is combustion and thermal radiation from a sustained fire. The second purpose of the paper is to present the combustion consequence assessment and application of probabilistic uncertainty analysis for modeling of gas pipeline combustion effects. The related work includes performance of the following tasks: to study gas pipeline combustion model, to identify uncertainty of model inputs noting their variation range, and to apply uncertainty and sensitivity analysis for results of this model. The performed uncertainty analysis is the part of safety assessment that focuses on the combustion consequence analysis. Important components of such uncertainty analysis are qualitative and quantitative analysis that identifies the most uncertain parameters of combustion model, assessment of uncertainty, analysis of the impact of uncertain parameters on the modeling results, and communication of the results’ uncertainty. As outcome of uncertainty analysis the tolerance limits and distribution function of thermal radiation intensity are given. The measures of uncertainty and sensitivity analysis were estimated and outcomes presented applying software system for uncertainty and sensitivity analysis. Conclusions on the importance of the parameters and sensitivity of the results are obtained using a linear approximation of the model under analysis. The outcome of sensitivity analysis confirms that distance from the fire center has the greatest influence on the heat flux caused by gas pipeline combustion.