Physical description of Forest Fires

Transcription

1 Physical description of Forest Fires Albert Simeoni UMR 6134 CNRS UMR 6134 Univerità di Corsica BP Corte, France

2 1 Introduction aim to understand the behaviour of forest fires and predict their spreading This research started 50 years ago in Australia, USA and USRR The involved mechanisms are very complex The operational objectives are to predict the fire spread and to estimate the impact of the fire The predictions are done thanks to fire spread models inserted in simulators The impacts are estimated thanks to modelling and experiment

3 Fire Spread Modelling Where does the difficulty in the modelling of forest fire lay? The Physical laws are known GIS and weather models provide the Environmental data Wind (O 2 ) Wind, Topography Difficulty to model the time scales Combustion Heat transferts Difficulty to model the space scales Reaction (m) Fuel particule (mm to cm) Vegetation species (20 cm to 20 m) Plot scale in field experiments (100 m) Fire scale (km) Difficulty in modelling the huge variable variability (fuel gases ) 2

4 3 Fire Spread Modelling The problems in the modelling of forest fires are due to the numerical solving of mixed complex problems at different scales Three kinds of modelling: Empirical (Mc Arthur, 1966; simulator = Fire grass meter) Semi-empirical (Rothermel, 1972; simulator = Farsite or Behave) Physical Simplified (Albini, 1981; simulator in development by our team) Detailed (Grishin, 1996; simulator = Firetec or Firestar)

5 Fire Spread Modelling Two ways to describe fire spreading with physical modelling: - Detailed models that describe as finely as possible the mechanisms Drawbacks: small times and lengths Advantages: detailed description of the phenomena Knolwedge and prevention - Simple models that only take into account the main mechanisms involved in fire spreading Drawbacks: less accurate and less general than detailed models Advantages: possibility to implement them at the field scale Simulation 4

6 Fire Spread Modelling Detailed models A multiphase model Fuel layer Solid phase Gas phase Control volume Gas Phase Transfers Solid Phase Convection Turbulence Gas mixture Combustion Radiation Soot Mass Heat Momentum Energy Heat transfer Drying Pyrolysis Combustion Radiation 5

7 6 Fire Spread Modelling For each phase: Mass, species, momentum and energy balances Example Mass balance: g g. g gvg t k surf pr k k M k M k t M Interface relationships Example Interface equation for mass: surf M M M pr gk Detailed models A multiphase model k Sub-models Example Arrhenius type laws Radiative Transfer Equation L g L n. e ds a L a e. L g g g g g g g g g 0 k p S pk k gk

8 7 Fire Spread Modelling Simplified models A semi-physical model Assumptions: One single equivalent phase Assumed flame geometry Losses Medium equivalent to the Radiation fuel Burned area Combustion Heat release area Advection Unburned area Radiation Radiation

9 8 Fire Spread Modelling Energy balance T t k v V g. T k ( T T ) K T a Q t k R Radiative transfer fl fldv v fl fl 2 S r Mass conservation V g, x x a Vg, Buoyancy V g x cos cos 4 v BT g x fl T 2 T V g, z g, z a Perfect gas g T a T a 1 g cos ds 1 sl g fl ds v t k

10 9 Fire Spread Modelling Meeting Combustion ESF Theory Exploratory and modelling Workshop Improved Institute Quantitative of Physics Combustion Fire Description Physics With Group Multi- Species Inversions London, Of Observed 24 June Plumes 2009

11 Fire Spread Modelling September June 24, 15, Theoretical and Experimental Physical description Tools for of Combustion forest fires in Forest Fires 10 4

12 Fire Spread Modelling Improved Quantitative Fire Description With Multi Modelling Heterogeneity Non flammable areas burned unburned empty burned unburned empty 11

13 Fire Impact Improved Quantitative Fire Description With Multi Experiments under controlled conditions (heat flux, flow) Fire intensity: 0% - Ph 26% - Ph 63% - Ph 0% - Pp 26% - Pp 63% - Pp 12 HRR (kw) Time (s) Fire emissions: CO (ppm) CO2 (ppm) BRE Centre for Fire Safety Engineering Exhaust CO2 Molar Fraction Exhaust CO Molar Fraction Time(s) 12

14 13 Fire Impact Smouldering

15 Improved Quantitative Fire Description With Multi Fire Impact Smouldering Fire Spread and Severity BRE Centre for Fire Safety Engineering Emissions 100 CO CO Yield (%) Yield (%) CO CO Moisture content (%) Heat flux (KW/m²) 14

16 Fire Impact Improved Quantitative Fire Description With Multi Thermal impact (but also influence on vegetation and wind on fire spread) 15

17 Fire Impact Thermal impact (but also influence on vegetation and wind on fire spread) Limits of exposure Firefighter: 7000 W/m² (90 s) Unprotected person: 6400 W/m² (8 s) heat flux (W/m²) st spread - 5 m 1st spread - 10 m 2d spread - 5 m 2d spread - 10 m time (s) 16

18 17 Conclusions and perspectives Physical modelling of forest fires is a young science Two kind of tools arte developed: Fire spread simulation Fire impact estimation Many improvements are necessary but simple answers can be provided A strong perspective is the coupling between fire and atmosphere: Integration of satellite data to evaluate the fire Using fire simulations to detect fires Characterize the fire as a source of gases

19 Physical description of Forest Fires Albert Simeoni UMR 6134 CNRS UMR 6134 Univerità di Corsica BP Corte, France