Risk assessment for gas pipelines using fuzzy sets

Most gas pipelines have been buried for many years and it is hard to assess their safety. In some locations, gas pipeline breaks have resulted in severe economic and social consequences. The severity of such consequences increases with the number of people exposed. For these reasons, risk assessment for gas pipeline systems should consider both the prediction of damage occurrence and the estimation of the economic and social impacts. In this paper, a new measure denoted H. Index is introduced to assess the magnitude of gas accidents. In order to examine the correlation between the magnitude of accidents and the pipeline attributes, including the environment, fuzzy component analysis is developed. This paper presents a procedure based on linguistic variables to evaluate the main characteristics for each accident. Actual accidents data in Japan are used to illustrate the approach.