SHIPP methodology: Predictive accident modeling approach. Part II. Validation with case study

System Hazard Identification, Prediction and Prevention (SHIPP) is a systematic methodology to identify, evaluate, and model the accident process, thereby predicting and preventing future accidents in a process facility. In this methodology, process hazard accidents are modeled using safety barriers rather than causal factors. The fault tree and event tree analysis techniques enhance the accident model to represent a holistic picture of the cause-consequence mechanism of the accident process. Quantitative analysis has two aspects: updating and prediction. The Bayesian theory updates failure probability and consequence occurrence probability when a new observation arrives. The predictive model forecasts the probability of a number of abnormal events occurring in the next time interval. The qualitative and quantitative capabilities of the SHIPP methodology help to design and implement safety strategies into a process system. This methodology was presented in Part I of a series of two papers. This paper (Part II) presents the results of a case study carried out on a liquefied natural gas (LNG) facility. It is conducted with the sole purpose of testing the SHIPP methodology. Gas production and liquefaction were considered as a boundary for accident analysis. This paper illustrates accident model development and cause-consequence analysis for the LNG facility. The updated results demonstrate that the probabilities of abnormal events dramatically change over time as new information is observed. Using accident precursor data for the year 2009, the predictive model estimates that the mean of abnormal events occurring in the next time interval is 22. The qualitative and quantitative analyses provide insight to identify critical safety areas and functions, and to determine the likelihood of failure of these measures. Combining management oversight and engineering analyses, the SHIPP methodology provides a comprehensive, systematic approach to manage a system risk. © 2010 The Institution of Chemical Engineers.