Computational & Technology Resources
an online resource for computational,
engineering & technology publications
PROCEEDINGS OF THE TWELFTH INTERNATIONAL CONFERENCE ON COMPUTATIONAL STRUCTURES TECHNOLOGY
Edited by: B.H.V. Topping and P. Iványi
Forecasting the Failure Consequences of Oil Pipelines
L. Parvizsedghy and T. Zayed
Department of Building, Civil and Environmental Engineering, Concordia University, Montreal, Canada
L. Parvizsedghy, T. Zayed, "Forecasting the Failure Consequences of Oil Pipelines", in B.H.V. Topping, P. Iványi, (Editors), "Proceedings of the Twelfth International Conference on Computational Structures Technology", Civil-Comp Press, Stirlingshire, UK, Paper 102, 2014. doi:10.4203/ccp.106.102
Keywords: oil, gas, pipelines, ANFIS, inference system, neuro-fuzzy, failure, forecast..
While pipelines are considered to be the most effective and safe way of transporting the hazardous liquids there is probability of failure with monetary and safety consequences. Over 7,300 failures were recorded from 1986 to 2013 in the United States of America which has resulted in almost 2.9 billion dollar property damages and the leakage of around 4.1 million barrels of hazardous liquids in the environment. Also, records prove there were 60 fatalities as well as 2,150 serious injuries which demands significant attention. Accordingly, failures of pipelines specially the ones carrying hazardous liquids has become the subject of interest for the study reported in this paper. Most of the oil pipelines are buried underground and this makes the estimation of their failures more difficult. Although, there are inline inspection tools to assess the condition of pipelines, these tools are very expensive to run regularly. As a result, there is a certain need for a risk assessment tool to forecast the risk of failures on these pipelines. Risk is assessed through the forecasted probability multiplied by the consequences of failure. This paper reports on the development of a model to forecast the consequences of failures in oil pipelines. The consequences of different scenarios of pipeline failures are evaluated. Data has been obtained from the Department of Transportation of the United States of America. Data includes attributes of the pipelines as well as the inspection quality and consequences of the failures on oil pipelines in the United States of America. Neuro-fuzzy is identified as method of pattern recognition while considering the uncertainty of the impact of input variables on the consequences. Historical data were embedded into a neuro-fuzzy system in order to recognize the existing pattern between input and output variables. Input variables include the factors that are predictable before failure. The final model, evaluated the monetary consequences in case of happening of various scenarios after failure including ignition and, or explosion. Finally model rules to forecast the monetary consequences of the failures with respect to the input factors in oil pipelines were generated. This tool will help the operators of oil pipelines to prioritize the pipelines of their network for inspection and maintenance.
purchase the full-text of this paper (price £20)