Analysis of complex urban lifeline systems

• Main Author: Teo, Wen Lynn
• Other Authors: Cheung Sai Hung
• Format: Final Year Project
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Civil engineering
• Online Access: http://hdl.handle.net/10356/72974
• Institution: Nanyang Technological University

Urban infrastructure systems such as water, power, and transportation supply are absolutely vital for a fully-functioning society given that they form the critical backbone. This study reveals the interdependencies and complexity of the pivotal power lifeline network through the assessment of cascading failure when the system is subjected to the triggering seismic event. Although the overall cascading failure risk can be quantified with respect to the current studies, the research is limited in the simulations of cascading infrastructural failure and its corresponding behavior. In this dissertation, the cascade phenomenon is established through the observation of general trends and sequence of propagating failure in the system. The simulation of each triggering event considers the effect of varying magnitude of an earthquake on correlated and uncorrelated system respectively. This incurs different area of initial failures as well as differing types and number of initial failed nodes which have probable influence over the cascade trends. A multi-paradigm coding language is adopted to simulate and model the real-time network system behavior and the influence on performance loss and power redistribution under the triggering events. With respect to the results concluded, the cascading behavior displays a significant difference from the expected train of failure. In spite of the considerable influence on the initial phase of failure of the system, differing seismic magnitude on both correlated and non-correlated systems generally impose similar cascading spread where collapse propagates in an irregular manner. The trend defies the intuitive aspect of propagation with failure occurring not only in affected areas but also along the vicinity and at random parts of the system. Regardless, the triggering event would result in an ultimate catastrophic collapse of close to 90% of the overall network. These conclusions render control mechanisms and restoration management strategy to be of vital importance on a global system level instead of implementing it in the locality of initial failure so as to provide a safety blanket for the system when subjected to an external perturbation.