ABSTRACT
ABSTRACT Systems under disaster conditions pose stringent operational conditions to components and subsystems. This study uses the classic relationship between demand curves and capacity curves: both are derived from the failure analysis of mechanical components and structures. The testbed that is used to present the model is a chemical industrial complex in Ulsan, Republic of Korea. While the model can accept any natural disaster as input, earthquake was the chosen one within the framework of maximum likelihood estimation and probability of occurrence of earthquakes. The annual probability of occurrence of earthquake is used to determine hazards curve which serves as input to the model the seismic risk curve that: provides the analyst with decision indicators (DI), and leads to an improved benefit/cost analysis for any change and improvement to the system, such as: retrofit of the components and base isolation. The vulnerability of the subsystems due to damage is obtained with a quantitative and indicative failure criterion. The gain, when there is a preventive retrofit, is also obtained according to the state of the system. Furthermore, the gain, when it is a mitigating procedure, can also be obtained accordingly. The output of the study is an indicator for measuring the level of economic implication which will result from damage to components under consideration. The further away the curve is from the abscissa and ordinate, the higher the risk; thus, preventive or mitigating retrofit is employed. Eventually, the result helps analyst and government in planning, maintenance and design of reliability assessment program for major infrastructure like chemical plants, nuclear power plants, thermal plants and other complex systems. The novelty of the study is the deployment of parallel algorithm to expediting the analysis.
