ABSTRACT
In modern infrastructures every system depends on other systems up to a certain extent, either directly or indirectly, with the result of having no hard separation among domains that, in spite of providing different services, need each other in order to function. In such a system of systems, a crucial issue is the understanding of response at failures, which propagate across the system domains and spatially across regional and national borders, resulting in long service outages and severe accident scenarios. In this paper we propose a methodology to analyze the system response at a disturbance. The methodology is based on a representation of the infrastructure as a network of systems that are dynamically interacting. The system response at failure is characterized in terms of two indicators: the buffering capacity to resist an input disturbance up to a certain extent, and the recovery from failure. These measures are given an availability function, to evaluate the probability of system response. The integration of resilience measures into the risk analysis represents a significant innovation in response to the need for a comprehensive modeling framework for the design of critical infrastructures. The methodology is exemplified on a simple case study.
