Evaluating network criticality of interdependent infrastructure systems: applications for electrical power distribution and rail transport

[摘要] Critical infrastructure provides essential services of economic and social value. However, the pressures of demand growth, congestion, capacity constraints and hazards such as extreme weather increase the need for infrastructure resilience. The increasingly interdependent nature of infrastructure also heightens the risk of cascading failure between connected systems. Infrastructure companies must meet the twin-challenge of day-to-day operations and long-term planning with increasingly constrained budgets and resources. With a need for an effective process of resource allocation, this thesis presents a network criticality assessment methodology for prioritising locations across interdependent infrastructure systems, using metrics of the expected consequence of an asset failure for operational service performance. Existing literature is focused mainly upon simulating the vulnerability of national-scale infrastructure, with assumptions of both system dynamics and dependencies for simplicity. This thesis takes a data-driven and evidence-based approach, using historical performance databases to inherently capture system behaviour, whilst network diagrams are used to directly identify asset dependencies. Network criticality assessments are produced for three applications of increasing complexity from (i) electricity distribution, to (ii) railway transport, to (iii) electrified railway dependencies on external power supplies, using case studies of contrasting infrastructure management regions. This thesis demonstrates how network criticality assessments can add value to subjective tacit knowledge and high-level priorities both within and between infrastructure systems. The spatial distribution of criticality is highlighted, whilst the key contribution of the research is the identification of high-resolution single points of failure and their spatial correlation across systems, particularly within urban areas. Service-level metrics also have a broad applicability for a range of functions, including incident response, maintenance and long-term investment. The role of network criticality within a holistic and systemic decision-making process is explored, for risk assessment and resilience interventions. The limitations of the research, regarding sample-size caveats and the definition of system boundaries within performance databases, lead to recommendations on cross-system fault reporting and the improvement of information systems.

[发布机构] University:University of Birmingham;Department:School of Engineering, Department of Civil Engineering

[关键词] G Geography. Anthropology. Recreation;G Geography (General)