24/04/2026
THE IMPACT OF CLOSING BRADFORD ON AVON FIRE STATION ON RESPONSE TIMES - MODELLED VS. ACTUAL DATA
In the public consultation reports, DWFRS has analysed 577 real incidents that occurred between 1 April 2019 and 31 March 2024 where Bradford on Avon fire station "would provide the nearest pumping appliance" (this should largely correspond to our station ground). They have then used their response time model to calculate how long it would have taken the first appliance to arrive at each of these incidents, if all appliances at all stations had been available 100% of the time. They then re-ran the model to exclude Bradford on Avon's appliance, and compared the modelled average first appliance response times, as follows:
With the BoA appliance (mm:ss) 10:04
Without the BoA appliance (mm:ss) 12:46
Difference (mm:ss) + 2:42
So the DWFRS model suggests that, if Bradford on Avon fire station is closed, it would have taken, on average for those 577 incidents, an extra 2 minutes and 42 seconds for an appliance to arrive from a different fire station. DWFRS states, in its report, that this increase in response time is "considered, from a professional advice perspective, to be tolerable". We'll revert to this point in our next post.
Under a Freedom of Information request, we have obtained data for actual first appliance response times to 1,223 actual incidents in the Bradford on Avon station ground between 1 April 2016 and 31 December 2025. We have enough data (1,141 incidents) to show average response times for Bradford on Avon, Winsley, Limpley Stoke and Westwood, and to compare the BoA appliance with appliances from other fire stations.
The data highlights two issues with the DWFRS modelling. Firstly, their model significantly understates actual first appliance response times, which in the real world are much longer than 10 minutes and 4 seconds. Secondly, while the 2 minutes 42 seconds modelled response time impact is in the right ball-park for Westwood and Limpley Stoke (potentially explained by their relative proximity to Trowbridge and Bath fire stations which accounted for, respectively, 54% and 47% of the attending first appliances for incidents in those villages), the response time impact for Bradford on Avon and Winsley is a lot longer.
Any decision to close fire stations must be based on robust data, with any modelled results reconciled to real-world data to verify the accuracy of the model. DWFRS should follow the example set by Oxfordshire FRS and publish the results of its validation exercise.