Harmonized approach to stress tests for critical infrastructures against natural hazards

Critical Infrastructures (CIs) provide essential goods and services for modern society; they are highly integrated and have growing mutual dependencies. Recent natural events have shown that cascading failures of CIs have the potential for multi - infrastruc ture collapse and widespread societal and economic consequences. Moving toward a safer and more resilient society requires improved and standardized tools for hazard and risk assessment of extreme events, and their systematic application to whole classes o f CIs. Among the most important assessment tools are the stress tests, designed to test the vulnerability and resilience of individual CIs and infrastructure systems in natural disasters. We present the main results o f the STREST project regarding extreme event quantification with focus on extreme earthquakes and extreme earthquake consequences. We show that extremes result from the combination of stochastic, site -specific and/or explicit physical processes. The stochasticity of earthquake risk is represent ed by random phenomena (e.g., random earthquake clusters, spectral acceleration sigma) and model uncertainties. Site -specific aspects include geotechnical properties, near -source effects and ground shaking spatial correlations, which can locally increase t he seismic risk. Finally, physical processes include maximum fault rupture propagation, earthquake interactions (i.e., aftershocks) and associated vulnerability changes, inter -hazard interactions (e.g., tsunamis, landslides), natech interactions (i.e., dom ino effects within the CI system following an earthquake), and additional CI interactions. Combination of all these processes tends to yield more extremes (fattening the risk curve) based upon which the CI stress test is made. The different steps of the ST REST stress test method are presented in a companion paper.