Report 1: Link Suncorp Residential Fire Insurance Claim with Fire and Rescue NSW Incident and Investigation Data for Insight and Product Delivery

This study is a collaboration between FRNSW and Suncorp, with the purpose of linking Suncorp's general insurance claim, and insurance assessor's data with FRNSW AIRS and FIRU data to determine if there is a pattern to the causes and impacts of residential fires that may be averted or reduced. This is the first time such a study has been undertaken linking insurance data with emergency response data in NSW on house fires. This report provides descriptive statistics on FRNSW AIRS data and Suncorp claimants' data" both separately and linked. While each data set was rich the potential for linked data analysis was limited by the mismatch in the time periods covered by the different data sets to be linked. Although the findings are limited in scope at the moment they provide an important proof of concept and precedent for extending this work as the data available for linkage becomes richer. Clear insights from this report include: "¢ In line with Australian and international research, winter is the time of year with the highest rate of fire incidents, July in particular. "¢ Fires are most common in functional areas such as sleeping areas, assembly areas including lounges, family rooms as well as storage areas including closets, rubbish areas, garages and car ports. "¢ Unattended cooking is a major ignition factor in residential fires, the precipitating circumstance in one-fifth of incidents. "¢ Heat from overloaded electrical equipment is a major contributor to fire but heat from properly functioning electrical equipment is also implicated in one-third of all fire incidents reported. "¢ Almost 2,000 incidents had a smoke alarms that failed to alert the occupants due to the batteries being missing or disconnected; battery discharge or dead or defective; or due to hard wire power failure. Future research would entail econometric modelling on FRNSW AIRS data to identify predictors of fire and fire spread. Modelling exercises on Suncorp data linked with FRNSW AIRS data would be possible upon the provision of eAIRS data for the period 2015 to 2018. This would increase the number of claimants and incidents analysed as the current link is on 362 claimants. In addition, modelling on factors such as the availability of operational smoke alarms (by type) and sprinkler systems or the geographical factors may be undertaken with larger data points.