Workshop at Indian Institute of Science

1
Fire Safety Engineering Structures in Fire

• Workshop at Indian Institute of Science
• 9-13 August, 2010
• Bangalore
• India

Fundamentals of Fire Dynamics Session JT3 to JT8
Organisers CS Manohar and Ananth Ramaswamy Indian Institute of Science
Speakers Jose Torero, Asif Usmani and Martin Gillie The University of Edinburgh
Funding and Sponsorship
2
Evacuation

• te tde tpre tmov
• te total egress time
• tde detection time
• tpre pre-movement time
• tmov movement or displacement time

3
Movement

• Egress is formulated on the basis of displacement
  velocities

V m/seg
1 m/seg
D people/m2
4
Movement time(tmov)

• Calculations are based on empirical data (SFPE
  Handbook of Fire Protection Engineering)

5
Pre-Movement (tpre)

• Purely statistical
• Large error bars
• Could potentially be the longest time

6
Detection time (tde)

• Calculated as a function of
• Technology used
• Growth of the fire
• Compartment geometry

7
Detection

• Obvious alarm mechanism
• Types of detectors
• Smoke detectors (ionization, photoelectric)
• CO Detectors
• Temperature detectors
• Multiple inputs (Artificial Intelligence)
• etc.

8
Movement time

• Empirical results are available for
• Doors (q) Personas/m.seg Act like valves
• Q w.q
• Where w is the width of the door and Q the total
  flow rate
• Stairs (q) Personas/m.seg Act like pipes
• Ramps (q) Personas/m.seg Act like pipes
• Corridors (q) Personas/m.seg Act like pipes
• Open spaces (V) velocity in m/s as a function of
  the density (N people/m2)
• Etc.

9
Simple Systems

• Egress time is the displacement time time to go
  through a door Required Safe Egress Time (RSET)
• Untenable conditions give the Available Safe
  Egress Time (ASET)
• Codes transform these times into a maximum
  displacement distance a required minimum door
  width

10
Maximum Egress Distance

• Egress Time (te) (RSET)
• te td tp
• tddMax/Ve
• tp W.Ve,p
• W, dMax given by codes

11
More Complex Scenarios

• Maximum egress distances can not be achieved
• Safe areas need to be generated with fire rated
  walls, doors, etc.
• Example Stairs in high rise buildings
• The design of these safe areas has to withstand a
  fire longer than the RSET
• Egress is directed to safe areas

12
Complex Systems
Street Level
Train Platform

• Generally are multiple entry, multiple exit
  systems
• Requires more complex calculations
• But the principles are the same

16.4
37.6
Restaurant Coffee shop
8.1
Waiting Area
37.9
13
Complete Problem

• To be able to analyze such a system all
  components must be understood
• It is necessary to calculate tf, te y ts
• Uncertainty needs to be established

14
The Fire Time (tf)

• To calculate the characteristic fire times it is
  essential to understand compartment fire dynamics

15
Introduction

• Smoke inhalation is responsible for most of the
  deaths in a fire
• What do we need to know to determine the amount
  of smoke produced by a fire?
• What is in that smoke?
• How is the smoke going to migrate from the room
  of origin to the rest of the building?
• What has to be done to control smoke migration?
• How do we use the smoke for warning? - Detection

16
(No Transcript)
17
)
The Pre-flashover Fire
The Front Room Fire (BRE Video)
18
)
The Pre-flashover Fire

• )

19
)
The Pre-flashover Fire

• )

20
)
The Pre-flashover Fire
21
The Compartment Fire
22
)
The Pre-flashover Fire

• )

23
Flashover
24
The Fire

• Temperatures in Two Zone Fires are controlled by
  fuel burning rates (Fuel Limited)
• Temperatures in fully developed fires are
  controlled by ventilation (Ventilation Limited)

25
Timeline
26
Ventilation Limited
A (Floor Area)
H0
A0 Opening Area H0 Opening Height
27
Ventilation

• A 3000 m2
• A0 100 m2
• H0 3 m
• Ventilation Factor A/A0H01/220

28
The Temperatures

• Empirical Data can be used to estimate fire
  temperatures

29
Duration of the Fire

• Depending on the average compartment temperature
  a mass loss rate can be established
• Total consumption of the fuel defines the longest
  possible fire duration

30
Duration of the Fire

• C.I.B.
• Fuel consumption per unit area per unit time
• Fuel Load
• Duration of the fire

31
Simplest Approach
Slope is defined by losses through the walls
(resistance method) 7oC/min (tgt60 min) 10oC/min
(tlt60 min)
TemperatureC.I.B.
tf
32
Resistance Method
1/Ahr
1/Ahr
Ta
Tf
1/Ahk
1/Ahc
1/Ahc
33
Fuel Limited (Growth-PreFlashover)

• Zone Model Divides the room into two well
  defined zones
• Upper Layer Hot combustion products
• Lower Layer Cold air
• Implies strong simplifications but help
  understand the dynamics of the problem

34
Initial Stages of a Compartment Fire

• Upper Layer - The parameters that need to be
  evaluated are
• The temperature of the upper layer Tu
• The velocity at which the Upper Layer descends

35
Initial Stages of a Compartment Fire

• These parameters can be obtained from, the ideal
  gas law and conservation of mass and energy in
  the Upper Layer

36
Conservation of Energy

• QP
• Unknowns

P
37
Correlations

• The Energy Release Rate
• Mass of air entrained
• Mass Burning Rate Generally obtained from
  empirical correlations

38
References

• Different engineering correlations are proposed
  in the literature
• SFPE Handbook of Fire Protection Engineering
• NFPA-The Fire Protection Handbook
• Karlsson and Quintiere, Enclosure Fire
  Dynamics, CRC Press, 2000.
• Drysdale, An Introduction to Fire Dynamics,
  John Wiley and Sons, 1999
• Cox, Combustion Fundamentals of Fire, Academic
  Press, 1995
• etc., etc., etc.

39
The Energy Release Rate, Q
P

• The effective energy release rate that will be
  transferred to the combustion products is
  unknown.
• The effective energy used to gasify the fuel is
  unknown.

40
Assumptions

• Total Energy
• Feedback is generally assumed to be small
• Radiation is assumed to be a fraction of the
  total energy released

41
Simplifications

• Under these assumptions we can correlate
  everything with Q
• There is no need to calculate QP directly
• How do we calculate Q?

42
The Energy Release Rate
Can be found in tables but generally only for
simple materials, i.e. liquid fuels
Can be found for some particular conditions,
generally difficult to generalize to real
scenarios

• Generally Q is evaluated empirically

43
Standard Test Methods

• The Cone Calorimeter
• Energy Release Rate obtained from Oxygen
  Consumption
• ASTM E 1354 Standard Test Method for Heat and
  Visible Smoke Release Rates for Materials and
  Products Using an Oxygen Consumption Calorimeter
• NFPA 264 Standard Method of Test for Heat and
  Visible Smoke Release Rates for Materials and
  Products Using an Oxygen Consumption Calorimeter
• ISO 5660 Rate of Heat Release of Building
  Products (Cone Calorimeter)
• Ohio State University Calorimeter
  (OSU-Calorimeter)
• Energy Release Rate obtained from temperature
  measurements of the combustion products
• ASTM E906 Standard Test Method for Heat and
  Visible Smoke Release Rates for Materials and
  Products

44
The Cone Calorimeter (ASTM E 1354 )
45
Fundamental Issues

• Heat Release Rate is obtained indirectly by
  measuring O2 consumption
• Mass is obtained real-time allowing a true mass
  loss rate to be obtained
• External Heat Fluxes of 0 to 100 kW/m2 may be
  achieved simulating conditions from incipient
  stages of a fire to post-flashover conditions

46
O2 Consumption

• COMPLETE COMBUSTION
• Main simplifying assumptions
• Energy release per unit mass of O2,constant E
  13.1 MJ/kg of O2 consumed
• Ideal gas law applies
• O2 depletion factor assumes each mole of air
  required for complete combustion is replaced by
  1.105 moles of products

47
Energy Released per kg of O2
48
Example

• The assumption is reasonable but can be improved
  by measuring CO, CO2, soot concentration and
  reconstructing the chemical reaction

49
O2 Consumption
To Blower
Exhaust Duct
Plenum
Hood
control volume
50
Calculations

• Oxygen concentration is measured at the exhaust
• Incoming oxygen concentration is that of air
• Therefore oxygen consumed is given by

51
Calculations

• Using this information, the energy release rate
  can be calculated as

52
Calculations
Density of the exhaust gas
Pressure differential across the exhaust orifice
A Cross sectional area of the exhaust stack
c Orifice coefficient
53
Solution

• How much smoke?
• How much time does it take? H(tO)HO
• The following equations need to be solved

54
Solution
Q comes from experimental data Calorimetry
55
Experimental Results

• Ideal Scenario

Q
t
56
Kerosene
57
Gasoline
58
Naphthalene
59
The Real Scale Application

• Large Scale Calorimeters
• Factory Mutual
• Underwriters Laboratories
• BRE

60
Design Fire

• Simple representation of the HRR

61
t2 parameters

• Incipient heat release rate (Qi)
• Incipient period (to)
• Growth time (tg)
• Growth HRR(Qo)
• Peak HRR (Qmax)
• Total HR (Q)
• Burnout time (tbo)

62
Fire growth characterization
Qat2
63
Loveseat
64
Loveseat
Qmax
Qat2
65
Bunk bed

• Corner ignition of lower bunk
• Data from Fire on the Web (www.bfrl.nist.gov)

66
Mattress
67
HRR data resources

• BFRL / NIST - Fire on the Web
• www.bfrl.nist.gov
• Lund University - Report on initial fires
• www.brand.lth.se
• Many other scattered reports
• Some data included in fire model suites
• CFAST FPETool

68
Initial Stages of a Compartment Fire

• With Q and the empirical correlations we can
  come back and evaluate

69
The Solution

• For most cases the solution to those three
  simultaneous equations has to be achieved
  numerically
• Several codes are available that will solve the
  equations
• Always remember what the assumptions are and
  where correlations were included
• Make sure that the assumptions and correlations
  apply to your particular scenario

70
Smoke Movement
71
Objectives

• How much smoke?
• How much time will it take for the smoke to come
  out of the room of origin?
• What is in the smoke?

72
Solution

• How much smoke?
• How much time does it take? H(tO)HO
• The following equations need to be solved

73
Solution
Q comes from experimental data Calorimetry
74
Design Fire

• Simple representation of the HRR

75
How do we calculate the area?

• A function of the flame spread
• Flame spread is a function of ignition

76
Ignition

• Simplest case
• 1-D
• Constant heat flux

77
Ignition Events

• Flash Point
• Fire Point
• Auto-Ignition
• Piloted Ignition
• Piloted ignition minimizes environmental
  variables-preferred to study the solid phase!

78
Standard Protocols to Assess the Solid

• Introduce many simplifications
• A standard methodology will be described and all
  simplifications and assumptions studied

79
The Lateral Ignition Flame Test (LIFT-ASTM-1321)

• Ignition Test Flame Spread Test

80
The Lateral Ignition and Flame Spread Test (LIFT)
81
Ignition Delay Time
82
Results

• The experimental data is fitted to the
  theoretical predictions and all characteristic
  values are extracted
• The total heat transfer coefficient is evaluated
  (hT)
• Material properties are evaluated (krC, Tig)
• Assumptions
• Semi-Infinite Solid
• Linearized Total Heat Transfer Coefficient
  hThChS,r
• Solid remains inert until ignition

83
Summary
Critical Heat Flux for Ignition
84
Ignition Properties

• Material Tig oC krC Critical
• s.kW2/m4K2 Heat Flux kW/m2
• Douglas Fir 382 0.94 16
• Cedar 402 1.22 18
• Iroko 410 1.30 17
• Polyisocianurate 445 0.02 21
• Polyurethane 390 0.30 16
• PMMA 378 1.02 15
• Acrilic 300 0.42 10

85
Flame Spread

• Propagation Rates are controlled by orientation
• Propagation defines the evolution in size with
  time of the fire

86
Opposed Flame Spread
Thermally Thick
Thermally Thin
87
Thermally Thick

• Solution
• Flame Spread Parameter
• Solution

88
LIFT Test - Flammability Diagram
89
Flame Spread Properties

• Material Minimum Flux F
• kW/m2 kW2/s.m3
• Douglas Fir 6.0 2.3
• Cedar 9.0 1.2
• Particle Board 5.7 2.1
• Polyurethane 0.0 11.7
• Acrilic 2.0 9.9
• PMMA 0.0 14.4

90
Design Fire

• Simple representation of the HRR

91
Smouldering

• Smouldering leads to propagation rates 100 times
  slower than flaming fires
• Therefore is important to establish if the fire
  originated in smouldering

92
Smouldering Limits
93
What is inside the smoke?

• Generally defined by yields (Yp)
• A yield is the fraction of the total mass that
  corresponds to the specific product
• The mass of the specific product is given by

94
Typical Yields
Irritants
From Purser, SFPE Handbook, 1995
These values vary from fuel to fuel and from
burning conditions to burning condition
95
Carbon Monoxide (I)
Flaming Combustion
From Tewarson, SFPE Handbook, 1995

• CO yields for smoldering tend to be much higher,
  i.e. 6 for polyurethane foam

96
Carbon Monoxide (II)
From Purser, SFPE Handbook, 1995

• Time to incapacitation because of CO inhalation

97
Carbon Monoxide (III)

• It is necessary to know the concentration of CO
  within the smoke
• An additional differential equation has to be
  incorporated for each species
• The CO concentration is a direct function of the
  fire size