predicting fire suppression by water spray with numerical codes: model development and validation...
TRANSCRIPT
Predicting fire suppression by water spray with numerical codes: model development
and validation
Alexandre JENFT1, Armelle MULLER1, Grégoire PIANET1,
Arnaud BRETON1 Pascal BOULET2, Anthony COLLIN2,
1 CNPP, Route de la Chapelle Réanville, BP 2265, F-27950 Saint Marcel - FRANCE
2 LEMTA, 2 Avenue de la Forêt de Haye - TSA 60604 - 54518 Vandoeuvre-lès-Nancy cedex – FRANCE
Contents
• Suppression mechanisms by water spray
• Experimental study
• Numerical results
• New suppression model writing
• New suppression model results
• Conclusions / current work
2
Suppression mechanisms
• Gas phase cooling;
• Oxygen displacement and fuel vapor dilution;
• Fuel surface cooling and wetting;
• Radiative transfer attenuation;
• Kinetic effects.
3
Experimental setup
Metrology:• 18 thermocouples• Gas analyser for O2, CO2
and CO• Load cell • Video camera
Water mist characteristics:• Pressure = 10 bars• Flow rate = 6.3
l/min/nozzle• Injection angle = 130° • D32 = 112 µm 4
Suppression observationTest 14tapp = 1 min
a) t0+10 s b) tapp-1 s
c) tapp+5 s d) tapp+10 s e) tapp+20 s
5f) tapp+30 s g) tapp+60 s h) tapp+65 s
Experimental results: fuel oil
6
N° Dpool
[cm]T0
[°C]tapp
[s]HRRapp
[kW]TA2,290,app
[°C]dtvid
[s]dtgas
[s]dtpool
[s]
13 35 12 38 28 21 40 35 35
14 35 11 64 48 32 65 61 65
20 35 7 94 52 42 65 58 78
21 35 6 126 56 51 70 57 78
31 25 10 38 7 17 31 29 35
32 25 10 64 12 23 30 30 62
33 25 10 99 19 29 69 71 104
34 25 10 128 24 33 99 108 126
35 25 10 186 32 39 106 95 126
36 25 10 305 40 47 105 103 125
37 25 10 544 46 50 126 131 156
Time - [s]0 50 100 150 200 250 300 350 4000
50
100
150
200
250
300
350
400
450
500
TA1,50 - [°C]TA1,150 - [°C]TA1,250 - [°C]TA1,290 - [°C]
tgas
tapp
Time - [s]0 50 100 150 200 250 300 350 400
100
200
300
400
500
600
700
Tpool,1 - [°C]Tpool,2 - [°C]Tpool,3 - [°C]Tpool,4 - [°C]Tpool,5 - [°C]
tpool
tapp
Numerical model
7
Main parameters:
• Cell size : 5 cm x 5 cm x 5 cm;
• Power increase defined as a ramp following the actual measured curve until stationary regime;
• After mist activation, HRR guided toward a reduction through suppression model.
Suppression model:
with
dttk
ff emtm )(''
0,'' )(
)()( '' tmatk w
8
N° 13 14 20 21
HRRapp – [kW] 28 48 52 56
a – [m²/kg/s] 0.2 0.15 0.4 0.2
N° 31 32 33 34 35 36 37
HRRapp – [kW] 7 12 19 24 32 40 46
a – [m²/kg/s] 1 1 0.24 0.26 0.42 0.74 0.24
It is impossible to predict the value of ‘a’ for a test which has not been carried out prior to the simulation.
This model does not allow predictive simulations.
Fire suppression model
9
Pyrolysis rate reduction during water application is linked to fuel surface temperature. The model is written:
B and E are empirical coefficient which can be easily determined, even with no preliminary real test.
The model is based on Arrhenius law:
ignfuelpyro
ignfuelfuel
ignfuelpyro
TTtm
TTtRT
ETtTBtm
if 0)(
if ))(
exp()()(
''
''
))(
exp()(''
tRT
EAtm
fuelpyro
Tfuel - [°C]
mpyro-[kg/s]
100 150 200 250 3000
0.2
0.4
0.6
0.8
1
1.2
1.4
ExperimentsModel
x 10-3
.
10
How does it work ?
1. Simulate the part before water application with a specific interest in HRR (or pyrolysis rate) evolution and fuel surface temperature;
2. Identify B et E on this part;
3. Put optimal values for B and E in simulation input file;
4. Simulate the whole test.B = 0.00122 kg/m²/K0.5/s
E = 3751 J/mol
Results on suppression time
11
N° 13 14 20 21 31 32 33 34 35 36 37
tsup,exp
(s)40 65 65 70 31 30 69 99 106 105 129
tsup,num
(s)22 39 50 66 8 17 32 43 57 74 86
Gap(s)
18 26 15 4 23 13 37 56 49 31 43
Gap(%)
45 40 23 6 74 43 54 57 46 30 33
• The new model predicts suppression by fuel cooling in every tests, just like in real tests;
• Suppression time prediction still needs improvement.
12
Conclusions
1. An experimental study has been carried out to understand suppression mechanisms;
2. For the “fuel cooling” cases, a new suppression model has been developed and integrated to FDS;
3. This model allows predictive simulations for fire suppression by water spray.
13
Current work
1. Improve model results by improving fuel temperature
calculation through particles / fuel exchanges modeling;
2. Validate the model on other configurations;
3. Determine FDS capability to determine extinction by flame cooling and inerting effects in FDS 6.