Risk consequences assessment of gas pipeline failure incorporating local loss factors

Abstract

Risk consequence assessment of gas pipeline damage is normally conducted to determine the losses of a failure event such as human, asset, production, public necessities and environmental loss. The current practice of pipeline risk consequences assessment is considered imprecise due to exclusion of the local loss factors, which led to the deterioration of the quality of estimated risk. As a result, the calculated consequences generate an equal risk value to any areas of the buried pipeline throughout the country regardless of the area's unique loss factors and consequences values. This study presents three separate risk consequences models to assess seven different sites which cover rural and urban areas. Similar frameworks are used for all the models but differ in terms of analyses and procedures in the assessment, to generate the risk ranking. Model1 involved quantification of direct summation of all possible losses in terms of monetary value which is highly demanded by the industry. Model2 deployed the use of the Fuzzy Analytic Network Process, Super Decision software and Complex Proportional Assessment (COPRAS) analysis procedure. Lastly, Model3 utilised a series of Survey Analyses that provides outcome as priority vectors for each loss factors. In order to assess the validation of the developed models, the overall risk ranking category was calculated by comparing the obtained results of all models with the existing technical standard which is Pipeline Technical Guideline (PTG11.36.04). Results show percentage difference of 28.6 %, 57.1% and 17.14% for Model1, Model2, and Model3, respectively. The difference can be observed spectacularly on sites with high scores of the public loss value, environmental loss factors considered in the assessment as well as classification of the selected areas. The finding shows significant differences of risk between the existing technical standard and the proposed models. However, Model3 provided the lowest percentage difference can be considered as the most comprehensive and representative model because it involves prioritization of each loss factor in every loss category in monetary form. Through structured model validation and result verification process, the findings indicated that all models are considered comprehensive, fulfilled the objective, reliable, well-defined and practical. In conclusion, this research outcome is possible to be merged with existing technical standards towards the development of automated intelligent Pipeline Integrity Management System (i-PIMS). Consequently, these models are capable of prolonging the long-term integrity of pipeline assessments and simultaneously securing the pipeline owner’s annual profit margins.