Definition of Risk Project webinar 2

Below is a number of questions and answers raised by Fire and Rescue Service (FRS) delegates who attended the Definition of Risk live webinar held 19th May 2021, presented by ORH.

1. Interested to hear how you can also consider new infrastructure developments into the assessment of community risk?  New transport infrastructure can add risk but also improve response times.  New housing and commercial developments add risk and need to be calculated into future CRMP / IRMPs.

An output of the likelihood model is a list of factors that influence incident volumes. These typically include factors relating to deprivation, social mobility, income, geography, etc. New housing developments can be assessed against this list to provide information about their likely risk. 

The assessment of expected response times (and how these would potentially change with new infrastructure) is outside the scope of this work. 

2. Are considerations going to be made or identified to show how FRSs differ in how they record in the IRS? 

When using the IRS data to examine severity, we are mindful of possible recording differences. The accompanying report will caveat the results in light of this. 

From the data alone, it is difficult to assess how FRSs use the IRS system differently. We cannot know if any differences exist because incidents are different or because the recording is different. To determine this would require interviews with information providers at each FRS. That is outside the scope of this work. 

3. LSOA is 1-3K population.  Targeting every household in LSOA is not necessarily practical.  Is there going to be output which assists to identify household characteristics to target? 

An output of the likelihood model is a list of factors that influence incident volumes. These typically include factors relating to deprivation, social mobility, income, geography, etc. This information can be used to target individual properties. Local FRSs will often possess more granular data that allows for individuals or households to be targeted; such data is not available nationally to form part of this study. 

4. Will response times be included as an influencing factor in consequence analysis?

Yes, however previous research has shown that the likelihood of injuries/fatalities is linked to time of detection rather than response, and this data is not available consistently. 

5. Is there any consideration made for population fluctuations and temporary versus permanent residents for tourist destinations? 

This is a difficult question to assess. It is likely that we will be able to identify seasonal patterns to incidents but associating this with population fluctuations would require additional data. The model uses LSOAs as its base, for each of these we would require information about how the population changes at different points of the year. This would also have to be collected on a national level to allow for a fair comparison. We will investigate the level of seasonality before proceeding. 

6. Will the project be reviewing IRS categories, and will it be looking at creating consistent definitions for incidents for example wildfires to allow better comparisons? 

No, the review of IRS categories is not part of this work, but this will be covered by other NFCC projects. 

The Proof of Concept stage for this project will produce a detailed gap analysis, identifying any shortcomings in the data requirements. The Project engages closely with the Home Office Fire Statistics Team and will raise findings with them to identify potential solutions to this problem. 

7. Will deprivation be considered in the societal factors? Once this process has been developed, how often is it envisaged it would be refreshed? And assured? 

The Indices of Multiple Deprivation (IMD) dataset is included in the list of factors to assess for their links to incident volumes. Additionally, other factors associated with deprivation such as income, access to public facilities and benefit payments are available to assess. 

We will be developing a methodology that will be accessible to all FRSs, some of this can be updated easily when new data is available. In addition, some of the data used in the model is from the 2010 census. There is potential to revisit the work when the 2021 census data is released over the next few years. 

The Community Risk Programme’s Technical Working Group will assure the process that ORH uses in the analysis and modelling.  

8. Identifying ‘place’ requires UPRN. Will these be included?

The model will assess likelihood of risk at an LSOA level. UPRNs may be used to determine volumes of building types within an LSOA, subject to available data. Assessment of individual UPRNs is outside of the scope of this project. 

9. If response times are going to betaken into account in relation to consequence, will this include time to meaningful firefighting action? If response times are taken into account, then it would follow that building type would need to be taken into account as well. 

The building type will be used in the assessment of risk. There is an IRS data field for meaningful firefighting action (‘Firefighting Delay Description’); this will be included in the model if possible.