1 Introduction
2 Literature Review
2.1 Common Guidance-Based Design Values
2.2 The Maximum Heat Release Rate as a Function of Fire Damage Area
2.3 Fire Growth Rate
References | Distribution type | Mean/mode (kW/s2) | SD (kW/s2) | Minimum (kW/s2) | Maximum (kW/s2) |
---|---|---|---|---|---|
Holborn et al. [20] | Lognormal | 0.0060 | 0.0450 | – | – |
Baker et al. [21] | Triangular | 0.0326 | – | 0.0000 | 0.4120 |
Nilsson et al. [22] | Lognormal | 0.0110 | 0.0170 | – | – |
Deguchi et al. [23] | Lognormal | 0.0524 | 0.0600 | – | – |
3 The Dwelling Fires Dataset
3.1 Fire Damage Area
Dwelling type | Number of incidents | Room type | Mean (m2) | SD (m2) |
---|---|---|---|---|
Houses | 75,610 | Kitchen | 2.47 | 5.04 |
9511 | Bedroom | 4.36 | 7.18 | |
9810 | Living room | 3.54 | 6.61 | |
94,931 | All rooms | 2.77 | 5.51 | |
Apartments | 38,810 | Kitchen | 2.02 | 4.36 |
3610 | Bedroom | 3.98 | 8.66 | |
3470 | Living room | 3.61 | 6.56 | |
45,890 | All rooms | 2.30 | 5.06 | |
All dwellings | 140,821 | All rooms | 2.62 | 5.37 |
3.2 Fire and Rescue Service Arrival Time
4 Calculation Methods
4.1 Maximum Heat Release Rate
4.2 Fire Growth Rate
5 Results
5.1 Maximum Heat Release Rate
Dwelling type | Room type | Mean (kW) | SD (kW) | 95th percentile (kW) | Equivalent guidance percentiles (lognormal) | |
---|---|---|---|---|---|---|
BD 2410 (1000 kW to 2500 kW) | SCA guide (2000 kW to 6000 kW) | |||||
Houses | Kitchen | 1100 | 2200 | 4000 | 71.2–90.0 | 86.6–97.6 |
Bedroom | 1900 | 3300 | 6700 | 50.4–78.8 | 72.8–94.0 | |
Living room | 1600 | 3000 | 5600 | 59.9–83.9 | 79.1–95.5 | |
All rooms | 1200 | 2400 | 4400 | 58.0–83.5 | 78.4–95.6 | |
Apartments | Kitchen | 900 | 2000 | 3300 | 77.2–92.4 | 89.7–98.2 |
Bedroom | 1800 | 3700 | 6400 | 58.2–82.0 | 77.1–94.4 | |
Living room | 1600 | 3000 | 5700 | 58.8–83.4 | 78.4–95.4 | |
All rooms | 1000 | 2300 | 3800 | 74.3–91.0 | 87.9–97.7 | |
All dwellings | All rooms | 1200 | 2400 | 4200 | 70.2–89.2 | 85.7–97.2 |
5.2 Fire Growth Rate
Dwelling type | Room type | Mean (kW/s2) | SD (kW/s2) | 95th percentile (kW/s2) | Equivalent percentiles (lognormal) | ||
---|---|---|---|---|---|---|---|
Slow (0.0029) | Medium (0.0117) | Fast (0.0469) | |||||
Houses | Kitchen | 0.0026 | 0.0071 | 0.0099 | 79.3 | 96.2 | 99.7 |
Bedroom | 0.0032 | 0.0083 | 0.0099 | 73.9 | 95.1 | 99.6 | |
Living room | 0.0026 | 0.0077 | 0.0100 | 79.1 | 96.1 | 99.6 | |
All rooms | 0.0026 | 0.0071 | 0.0098 | 81.5 | 96.1 | 99.5 | |
Apartments | Kitchen | 0.0022 | 0.0132 | 0.0100 | 85.6 | 96.8 | 99.5 |
Bedroom | 0.0034 | 0.0116 | 0.0100 | 76.1 | 94.5 | 99.3 | |
Living room | 0.0029 | 0.0075 | 0.0109 | 76.4 | 95.8 | 99.7 | |
All rooms | 0.0023 | 0.0119 | 0.0089 | 84.9 | 96.5 | 99.5 | |
All dwellings | All rooms | 0.0025 | 0.0088 | 0.0098 | 81.5 | 96.1 | 99.5 |
Dwelling type | Room type | Mean (kW/s2) | SD (kW/s2) | 95th percentile (kW/s2) | Equivalent percentiles (lognormal) | ||
---|---|---|---|---|---|---|---|
Slow (0.0029) | Medium (0.0117) | Fast (0.0469) | |||||
Houses | Kitchen | 0.0058 | 0.0144 | 0.0217 | 58.3 | 88.9 | 98.6 |
Bedroom | 0.0088 | 0.0168 | 0.0314 | 39.2 | 80.8 | 97.6 | |
Living room | 0.0062 | 0.0140 | 0.0228 | 54.6 | 87.9 | 98.5 | |
All rooms | 0.0060 | 0.0145 | 0.0225 | 56.4 | 88.3 | 98.5 | |
Apartments | Kitchen | 0.0062 | 0.0335 | 0.0234 | 69.7 | 90.0 | 97.8 |
Bedroom | 0.0079 | 0.0141 | 0.0278 | 40.2 | 82.7 | 98.1 | |
Living room | 0.0062 | 0.0101 | 0.0211 | 45.6 | 87.3 | 99.0 | |
All rooms | 0.0063 | 0.0329 | 0.0240 | 68.7 | 89.7 | 97.8 | |
All dwellings | All rooms | 0.0061 | 0.0209 | 0.0235 | 63.0 | 88.9 | 98.1 |
6 Limitations
- While the DFD is the largest quantity of recent fire event data in England, there is a lack of granularity in data for the fire damage area, as well as for the ignition to discovery and discovery to call times used to estimate the FRS arrival time.
- There is a large variability observed in the results (i.e. large standard deviations and coefficients of variation), which may in part be due to the lack of granularity in the input data. To reduce this variability, it would have been beneficial to specify the maximum HRR and growth rate parameters as a function of the total room area, but this information is not provided in the DFD.
- There are limitations in the assumption that fire damage represents a direct correlation to fire area. While this relationship appears reasonable, literature and quantitative evidence on the topic is lacking.
- The selected HRRPUA distribution has been applied uniformly and does not consider the different types of combustibles which may be present in different rooms, for example where kitchens will typically include different items to living rooms, etc.
- In determining the distributions for fire growth rate, the analyses assume that, in all instances, FRS intervention immediately begins upon first appliance arrival, that this intervention stops fire growth and that the growth from ignition to this time of intervention follows a t2 relationship.