Overarching themes
The Panel is developing a methodology to assess the risk load serviced across the state by FRV and the CFA.
The following themes have guided the development of the proposed risk methodology:
- safety and wellbeing: protecting life, property and environment
- clear, transparent, evidence-based methodology
- fire service capacity and capability
- time frame for change
- community expectations and needs
- impact and consequence analysis
- legacy projects and issues
- aligned with the State Emergency Management arrangements and other overarching policy and management principles
The high-level schematic diagram (Figure B.1 in Appendix B), which represents changes in the risk profile over time, broadly references these themes. The State Emergency Management Priorities are also relevant and will be considered.
What is risk?
The risk of fire can be assessed by considering both its likelihood and impact (Appendix A: Theoretical risk overview). The likelihood of fire is examined as a function of past fire history. The impact of fire is understood by considering the elements of impact (hazard, exposure and vulnerability). The Panel addresses hazard by means of the fire services’ capacity and capability to suppress the fire hazard. Exposure refers to the location information and attributes of dwellings, communities, buildings, structures, infrastructure assets, agricultural commodities, environmental assets and business activity. Vulnerability refers to the impact a hazard has on people, infrastructure and the economy. Vulnerability can be classified according to four types: physical, social, economic and system.
Fire incident likelihood (probability) is independent of severity of impact of fire.
Assessment of risk
The assessment of the risk serviced by FRV and the CFA is complex. As these organisations respond not only to fire but also to other emergencies, they encounter several types of risk (e.g., fire, hazardous material, road accidents) that require different equipment and skill sets. Furthermore, these risks vary over space (e.g., population density, socioeconomic factors, land use) and time.
Theoretical risk (likelihood x impact), as outlined in Appendix A, can be translated into an operational context (illustrated by Figure B.2 in Appendix B). The likelihood of fire will be examined using historic FRV and CFA incident data from 2010–2021. This data will be classified according to the type of incident (residential fire, non-structural fire etc.) and used to develop statistical models, and subsequent risk layers for each type of incident. This representation of risk across Victoria will be subject to further spatial analysis (Figure B.3 in Appendix B).
The process of building these risk models for each type of incident requires an understanding of the possible factors that explain the spatial and temporal patterns of fire risk. In the case of a residential fire, they can often be attributed to human factors and socioeconomic circumstances. The composite Index of Relative Socioeconomic Advantage and Disadvantage (IRSAD) integrates these characteristics. Exposure information, including type of dwelling and tenure status, will also considered in the model. Further, spatial indicators such as residential mobility, residential density and level of proficiency in English can be used to explore how readily information will diffuse within a local population, which has implications for the temporal distribution of risk.
Other types of incidents, such as industrial fires, accidents and hazardous material, cannot be explained as simply using socioeconomic data. The drivers behind these incidents are less apparent, and specific neighbourhood characteristics of the incident are unavailable or deficient. In such cases, land use spatial datasets, such as transportation (e.g., airports, railways, roads and harbours) and industrial developments, can be used in combination with exposure and vulnerability datasets to understand the distribution of risk.
The Panel will endeavour to develop as many incident-specific models as required; however, as fire is the Panel’s primary focus, fire models will be developed first.
Much of the academic research in fire risk assessment has addressed solely the likelihood of fire1,2,3, largely due to an absence of detailed impact data (exposure and vulnerability), such as costs of fires, construction materials, or the presence of invalids, elderly persons or children in a building. The Panel will endeavour to integrate impact data in the modelling as the work progresses in phases.
Phasing
Due to the magnitude and complexity of the proposed risk assessment, the process will be undertaken in phases.
Phase 1 will consider likelihood, where the incident data will be integrated with a range of variables that drive risk. Hazard suppression information (FRV and CFA capacity and capability data) will also be integrated in Phase 1 through spatial analysis that examines travel impedance (i.e., distance or time), facility busyness, coverage, and the ratio between the supply and demand of services (Figure B.3, Appendix B).
Phase 2 will incorporate additional exposure information, while Phase 3 will draw on risk information from the Safer Together program4 and attempt to integrate state-wide fuel management risk data. This is an important element of the state’s risk quotient, requiring yearly updates and of critical importance, particularly in the peri-urban fringe, which is vulnerable to bushfire and thus a major concern for regional fire managers.
Population growth and land use planning
As Melbourne and regional centres across the state become more populated and infrastructure is established to accommodate this, the quantum of risk changes. These changes in risk will be assessed (as described in Section 4.3 above) in conjunction with population forecasts, and future land use and development planning (including community, commercial, industrial and residential development).
Knowledge and datasets from Land Use Victoria and the Planning Division of the Department of Environment, Land, Water and Planning (DELWP) will be incorporate in the risk assessment.
Climate change
Climate change will potentially result in variations in the distribution of risk across the state. It may alter weather patterns, environmental conditions and ecosystem function, with implications for the sector’s ability to undertake established disaster mitigation activities. For example, specific conditions are required to ensure that prescribed burning on private land to reduce fuel load can be carried out safely and effectively. Rising temperatures, increasing fuel availability (through drying), growing awareness around smoke issues and CO2 emissions, and less-predictable wind conditions may all contribute to reduced safe planned burning opportunities (AFAC 2018)5. Resourcing pressure, due to increases in frequency and intensity of bushfires, also impacts the overall ability of the fire services to mitigate risk in a bushfire.
The Panel aims to work closely with DELWP, through Safer Together6, to understand how climate change is impacting fuel accumulation and, subsequently, risk across the state.
Adjustment of the FRV fire district boundaries according to risk analysis
Risk layers, generated from the proposed risk models, will be input into spatial accessibility and location allocation‑ modelling (Figure B.3, Appendix B).
Spatial accessibility is a measurement of the spatial connection between the supply and demand of services. Spatial accessibility analysis is used to examine how accessible fire services are to the community. This knowledge is essential for ensuring an efficient operational response and reducing injuries and deaths. Accessibility involves two elements: regional availability and regional proximity. Regional availability means the ratio of supply to demand for each demand location (the physical location where risk is quantified). Regional proximity accounts for the spatial interactions between supply and demands sides. For the spatial accessibility analysis, in the Victorian context, the capacity of the fire services represents the supply, while the quantum of risk is the demand. Areas of high fire risk make a greater demand on fire services than areas of low fire risk.
Location-allocation modelling is a process designed to find the optimal location for service provision across a city or region given the spatial distribution of demand for that service. In the context of the fire services, location-allocation can be used to maximise the coverage of fire stations such that the maximum number of demand areas are covered by the specified response time standard.
Similar to spatial accessibility, supply and demand are integral to the process.
Footnotes
- https://www.researchgate.net/publication/228770569_Locating_fire_stations_in_Belgium_An_integrated_GIS_approach
- https://www.researchgate.net/publication/329948437_Modeling_Spatial-Temporal_Dynamics_of_Urban_Residential_Fire_Risk_Using_a_Markov_Chain_Technique
- https://www.researchgate.net/publication/259129919_Discovering_spatio-temporal_relationships_in_the_distribution_of_building_fires
- Safer Together is the approach to reducing the risks of bushfire in Victoria that combines stronger community partnerships with the latest science and technology to more effectively target our actions. More information is available at https://www.safertogether.vic.gov.au/understanding-risk
- Australasian Fire and Emergency Service Authorities Council 2018. Climate Change and the Emergency Management Sector: Discussion Paper. AFAC, Melbourne, Australia.
- Safer Together is the approach to reducing the risks of bushfire in Victoria that combines stronger community partnerships with the latest science and technology to more effectively target our actions. More information is available at https://www.safertogether.vic.gov.au/understanding-risk
Updated