In a changing climate, how can we use modelling to better understand the vulnerability of our infrastructure assets to natural hazards and use this in emergency management decision making?
Researcher A/Prof Matthew Mason from the University of Queensland and end-user Matthew Dyer from the Queensland Fire Department turned the spotlight on a project that fills a major gap in Australia's disaster risk toolkit: the Natural Hazards Research Australia (the Centre)-funded Modelling impacts of natural hazards on interconnected infrastructure networks project.
The team’s research increases our understanding of how natural hazards damage infrastructure networks communities rely on and how that damage cascades through interconnected systems.
About the project
The research team wanted to better understand the direct and cascading impacts natural hazards have on infrastructure, develop damage estimation tools to feed into broader modelling and create a framework to support infrastructure resilience planning and investment decisions.
How the model works
A/Prof Mason walked through the project's four core tasks:
- Map the networks – simplify complex real-world infrastructure (roads, electricity distribution and potable water) down to the components most vulnerable to hazards, identified through literature review and consultation with infrastructure operators in southeast Queensland.
- Develop damage models – build "direct damage" functions (how badly an asset is affected by a given hazard intensity) and recovery-time functions (how long it takes an asset to return to functionality, not necessarily full repair). Where data existed, models were empirically derived or adapted from overseas research; where it didn't, expert judgement from industry partners filled the gaps.
- Build a functionality cascade model – a hierarchical Bayesian model that links assets within a network (for example, substations and the feeders connected to them) and across networks (for example, water pumping stations that depend on power).
- Run utilisation case studies – a Brisbane flood scenario, a southeast Queensland tropical cyclone scenario, and a Sunshine Coast bushfire.
What the case studies showed
The Brisbane flood case study modelled four return periods (1-in-20 through to 1-in-2000-year events) and demonstrated that damage and network disruption escalate sharply once critical nodes such as substations are affected, with the power network consistently taking the longest to recover due to knock-on effects through connected feeders. The tropical cyclone case study, based on a scenario similar in track to ex-Tropical Cyclone Alfred but more intense, showed that as much as 80 per cent of power network assets could sustain some level of damage, with cascading effects extending recovery well beyond the footprint of direct damage.
An end user's perspective
Mr Dyer offered an end-user perspective on why the research matters operationally. He noted that disasters rarely affect a single asset in isolation, and that the project gives QFD a clearer, systems-level way of understanding how disruption spreads across power, water, road and communications networks.
He highlighted three ways the research supports operational decision-making: identifying critical cross-network dependencies, prioritising which assets to protect or restore first, and strengthening scenario planning and partnership discussions with infrastructure owners. He also pointed to natural synergies with QFD's existing Critical Infrastructure Disaster Risk Assessment (known internally as SYNINDRA), suggesting future iterations could draw on the project's dependency-mapping and Bayesian network methods.
Mr Dyer also reflected on the appeal of Bayesian methods to emergency management practice, noting the approach echoes how emergency managers already update their expectations in real time as new information comes to hand – with the model helping to formalise and add rigour to that lived experience rather than replace it.
Recently completed, the project brought together researchers from The University of Queensland, James Cook University, The University of Technology Sydney and Risk Frontiers and end-users from the Queensland Fire Department and the Queensland Reconstruction Authority.
The project also included input from infrastructure operators including Powerlink, Energy Queensland, Queensland Urban Utilities, Transport and Main Roads, City of Gold Coast and Optus.