Risk Analysis of Sweet Natural Gas Pipelines: Benchmarking Simple Consequence Models

Abstract

Over the past decades, risk assessment of industrial facilities perceived to present significant potential hazards has become widespread, and in some jurisdictions it has become a regulatory requirement or a societal norm. The application of risk concepts to transportation pipelines has become commonplace, and while qualitative techniques may be sufficient for such purposes as the establishment of maintenance priorities and preliminary route selection, increasing need has been seen for fully quantitative approaches. A considerable range of quantitative methods and models has been developed, but many of them have a high degree of complexity and most are proprietary. There is a need for simple and transparent methods that still yield results that are sufficiently accurate at least for generic assessment; it is recognized that the more complex approaches will usually be required for detailed, site-specific analysis. One source of complexity, for gas transmission pipelines, is the need for a suite of relatively sophisticated, numerical models to estimate accurately the consequences of major gas release incidents, which are transient in nature. To overcome this difficulty, relatively simple, closed-form consequence estimation schemes, based on steady-state approximations, have been proposed and applied. The current study benchmarks the results of one such method against those obtained using PIPESAFE, a proprietary software tool containing a suite of interlinked models developed and validated specifically for gas transmission pipelines. Within certain limits, the simple approach was found to give reasonable, and generally slightly conservative, estimates of safety consequences, expressed in both individual and societal terms.