Page 1
Proceedings of the 2nd European Symposium on Fire Safety Science
Edited by:
Prof Bart Merci, Ghent University - Faculty of Engineering and
Architecture
Dr Georgios Boustras, Center for Risk, Safety and the Environment
(CERISE), European University Cyprus
EUC Cultural Center, Nicosia June 16th – 18th 2015
Page 2
Acknowledgements
The editors would like to thank the organizing committee, the program committee and all those
that worked to bring this symposium to success. In particular we would like to thank Mr Chris
Bachtsetzis for the valuable help throughout and for helping us enormously with the
proceedings.
Nicosia, June 2015.
ISBN 978-9963-2177-0-0
Page 3
Table of Contexts
Keynote Lectures
“Fire hazards with new energy carriers”, G.Marlair, A. Lecocq, B. Truchot, ………………p.1
“Fire Extinguishment in Large Facilities – A Case Study from Research to Engineering
Application”, Hong-Zeng Yu ………………………………………………………………...p. 10
“Fire Research for the Fire Service”, Stefan Svensson ……………………………………p. 18
“Forest Fire Safety Research”, D. X. Viegas ……………………………………………....p. 26
Papers submitted and accepted for presentation
“A comprehensive approach for the fire safety design of tall automated warehouses”, P.
Lhotsky, C. Mastroberardino, G. Hadjisophocleous …………………………………………..p. 31
“Experimental Study of Upward and Lateral Flame Spread on MDF Boards in Corner
Configurations”, D. Zeinali, T. Beji, G. Maragkos, J. Degroote, B. Merci
………………………………………………………………………………………………….p. 35
“Application of FDS to an Under-Ventilated Enclosure Fire with External Flaming”, G.
Zhao, T. Beji, and B. Merci ………………………………………………………………...…p. 41
“Numerical Simulation on Mass Flow across the Smoke Layer Interface in a Room Fire
under Mechanical Extraction”, L. Yi, Y.Z. Li, R. Huo and W.K. Chow
………………………………………………………………………………………………….p. 47
“A Study on Building-Integrated Photovoltaic System Fire with Double-Skin Façade by
Scale Modeling Experiment”, Nadia C.L. Chow, S.S. Han ………………………………...p. 52
Page 4
“Full-scale Experiments to Investigate the use of a Water Curtain over Openings to
Prevent Fire Spread to Adjacent Properties”, Matt Turco, Paul Lhotsky, and George
Hadjisophocleous ……………………………………………………………………………...p. 57
“Geographic Reasoning on Multi-Sensor Smoke Spread Data”, F. Vandecasteele1, T. Beji, B.
Merci, S. Verstockt …………………………………………………………………………………p. 63
“Fire Exposure Assessment Inside Large Buildings”, M. Chaos …………………...……p. 69
“Model for Estimating Property Loss Risk in Building Fires”, A.Paajanen, T.
Hakkarainen, K. Tillander ……………………………………………………………………..p. 75
“Large eddy simulations of a one meter diameter methane fire plume”, G. Maragkos, B.
Merci …………………………………………………………………………………………..p. 80
“The assessment of meteorological risk for wildfires in the Adriatic region of Croatia”,
Tomislav Kozarić, Marija Mokorić, Lovro Kalin …………………………………………….p. 86
“Large Eddy Simulations of a Ceiling-Jet Induced by the Impingement of a Turbulent Hot
Air Plume on a Horizontal Ceiling”, S. Ebrahim Zadeh, G. Maragkos, T. Beji, B. Merci
………………………………………………………………………………………….……....p. 91
“Modeling of thermal events in Lithium-ion batteries”, P. Andersson, J. Anderson, F.
Larsson, B-E. Mellander ………………………………………………………………….…...p. 97
“Fire risks with Lithium-ion batteries – Measurements of gas emissions”, Petra Andersson,
Per Blomqvist, Fredrik Larsson, Anders Lorén ……………………………………………...p. 102
“Experimental Study on the Two-Dimensional Spread of Smoldering Peat Fires”, Xinyan
Huang, Francesco Restuccia, Michela Gramola, Guillermo Rein
………………………………………………………………………………………..p. 107
Page 5
“Flame Spread Monitoring and Estimation of the Heat Release Rate from a
Cable Tray Fire using Video Fire Analysis (VFA)”, T. Beji, S. Verstockt, P. Zavaleta, B.
Merci ………………………………………………………………………………...p. 113
“Modeling the Bench-Scale Fire Experiments of Medium Density Fiberboard”, Kaiyuan
Li, Xinyan Huang, Jie Ji …………………………………………………………………....p. 119
“Probabilistic Study of the Resistance of a Generic Concrete Structure according to
Eurocode Natural Fire”, Mohammad Heidari, Fabienne Robert, Guillermo Rein
…………………………………………………………………………………………..…….p. 125
“Numerical simulation of cable tunnel fire suppression with water mist”, K. Laakkonen, J.
Vaari ……………………………………………………………………………………...…..p. 131
“Experimental study of soot deposition during an enclosed pool fire”, Louis Decoster, Axel
Bellivier, Olivier Vauquelin, Fabien Candelier, Herve Bazin
……………………………………………………………………………………………...…p. 137
“The Effect of Wind on Burning Rate”, S. McAllister, M. Finney ……………………....p. 142
“Evaluate efficiency of water mist suppression of Class A fire by ISO room and bench
scale tests”, Qiang XU, Cong JIN, Greg GRIFFIN, Yong JIANG ………………………....p. 148
“A preliminary thermal and mechanical simulation study of load-bearing cold-formed steel
drywall systems exposed to fire”, Ilias D. Thanasoulas, Iason K. Vardakoulias, Dionysios I.
Kolaitis, Charis J. Gantes, Maria A. Founti ……………………………………………….....p. 154
“Experimental investigation of Externally Venting Flames using a medium-scale
compartment-façade configuration”, Eleni K. Asimakopoulou, Dionysios I. Kolaitis and
Maria A. Founti ……………………………………………………………………………....p. 160
“Design Tool for the Definition of Thermal Barriers for Combustible Insulation
Materials”, J. P. Hidalgo, Welch, S., J. L. Torero ………………………………………..…p. 166
Page 6
“Experimental characterization of the water density pattern from sprinkler nozzles
arrangement in a confined and mechanically ventilated compartment”, H. Pretrel and L.
Audouin ……………………………………………………………………………………....p. 171
“Comprehensive Aerial System Coupled with Sensor Network for Early Fire Detection,
Prevention, Monitoring, Prediction and Fighting”, I.Gkotsis, P. Michalis, G. Leventakis,
L.Georgaklis, K.Votis, S.Rogotis, D.Tzovaras
……………………………………………………………………………………..…p. 177
“Radiative heat flux in a pre-flashover compartment fire”, N. Johansson & J. Wahlqvist
………………………………………………………………………………………..p. 178
“Development of a CFD Model for Large-Scale Rack-Storage Fires of Cartoned
Unexpanded Plastic Commodity”, Ankur Gupta , Karl Meredith, Gaurav Agarwal, Sai
Thumuluru, Yibing Xin, Marcos Chaos and Yi Wang ………………………………………p. 183
“Parametric CFD study of an air curtain for smoke confinement”, L.-X. Yu, T. Beji, S.
Ebrahim Zadeh, F. Liu, B. Merci …………………………………………………………….p. 189
“A framework for evaluating the thermal behavior of carbon fiber composite materials”,
J.P. Hidalgo, P. Pironi, R.M. Hadden, S. Welch ………………………………………...…p. 195
“Fire Safe Design of Concrete – For or Against Spalling?”, C. Maluk and L. Bisby
………………………………………………………………………………………………...p. 201
“Experimental Study on the Thermal Decomposition of C6F-ketone”, Renming Pan, Kuang
Qin, Liying Cao, Qiang Xu, Pin Zhang ……………………………………………………...p. 207
“Improvement of the Firefighters' Training: Assessment of the Constraints During
Compartment Fire Behavior Trainings in Shipping Container”, S. Roblin, B. Batiot, T.
Page 7
Rogaume, F. Richard, J. Baillargeat, M. Poisson , A. Collin, Z. Acem, A. Marchand, M.
Lepelletier …………………………………………………………………………………....p. 213
“Modeling the fire behavior of polymer composites for transport applications”, A.Matala,
A. Paajanen, T. Korhonen, S. Hostikka ………………………………………….….p. 219
“An experimental investigation of improvised incendiary devices used in urban riots: The
‘Molotov cocktail’ ”, Dionysios I. Kolaitis ……………………………….………….p. 224
“Determination of the moisture content of Nordic spruce wood with an integral model”, P.
Mindykowski, M. Jørgensen, S. Svensson, G. Jomaas …………………………………………….p. 230
“Fire performance of plasterboard containing Phase Change Materials”, M. S.
McLaggan, R. M. Hadden, M. Gillie ………………………………………………...………p. 234
“Predictions of external radiative heat flux using various modeling approaches vs.
experimental data”, Piotr Tofilo, Vladimir Mozer ………………………………….……...p. 240
“Study of Fumes Cooling Efficiency with the 3D Firefighting Method”, S. Brohez, C.
Delvosalle, C. Fourneau …………………………………………………………………..….p. 245
“Developing and Post-Flashover Fires in a Full Scale Room. Thermal Environment, Toxic
Emissions and Effects of Fire-fighting Tactics”, Abdulaziz A. Alarifi, Herodotos N.
Phylaktou, Gordon E. Andrews, Jim Dave2 and Omar A. Aljumaiah
………………………………………………………………………………………..p. 251
“Modeling of large-scale syngas jet fires”, Z.S. Saldi, J.X. Wen …………….……...p. 257
“Developing and Managing a System for Prevention, Detection and Control of Forest
Fires”, George Papageorgiou, Kostas Papageorgiou, George Hadjisophocleous
………………………………………………………………………………………………...p. 263
Page 8
“Characterization of water mist systems using full-scale tests and computer modeling”, D.
Alvear, M. Lázaro, J.I. Melgosa, F. Arnáiz ………………………………………………….p. 272
“Predicting the heat release rate of liquid pool fires using CFD”, T. Sikanen, S. Hostikka
………………………………………………………………………………………………...p. 278
“Threat level assessment of smoke emissions from compartment boundaries”, Vladimir
Mozer, Miroslav Smolka, Piotr Tofilo ……………………………………………...………..p. 284
“On the possibility of assessing fire protection levels”, Vladimir Mozer ………………p. 289
“Uncertainties in material thermal modeling of fire resistance tests”, K. Livkiss, B. Andres,
N. Johansson, P. van Hees ………………………………………………………….……..…p. 294
“Numerical Study of the Thermal Behavior of a Thermo-Structural Aeronautical
Composite under Fire Stress”, N.Grange, K. Chetehouna, N. Gascoin , S. Senave
…………………………………………………………………………………...…...p. 300
“Consensus decision making and group cohesion in evacuees”, A.Cuesta, M. Lázaro, O.
Abreu and D. Alvear ………………………………………………………………...p. 306
“Experimental and numerical simulation of the Fire Spread across a Fuel Break in a
Ridge”, Raposo, J.R., Viegas, D. X., Almeida, M., Crissantu, I., Salis, M ; Oliveira, R
…………………………………………………………………………………………..…….p. 311
“The Fire Destroying 3 Historic and 37 Modern Buildings in Lærdalsøyri January 2014”,
T. Log ……………………………………………………………………………………...…p. 315
“Examination and Modeling of Thermal Runaway Issues Pertaining to New and Aged Li-
ion Batteries”, Sara Abada, Martin Petit, Amandine Lecocq, Guy Marlair, Valérie Sauvant-
Moynot, François Huet ………………………………………………………………………p. 321
Page 9
“Analysis of current investigation protocols to identify whether improved data collection
and sharing provides a better understanding of the circumstances surrounding fire
fatalities”, Iain S. Gavin ……………………………………………………………….……p. 327
“Stereo PIV investigation of the bidirectional natural convection flow at a
horizontal vent”, K. Varralla, H. Pretrela, S. Vauxa, O. Vauquelin ……………………..…p. 334
“Sensitivity of Structural Fire Safety Design according to Eurocode 1-1-2 Localised Fire”,
Mohammad Heidari, Fabienne Robert, Guillermo Rein ……………………………………..p. 340
“Failure of Shear Walls in Fire Condition”, A.T. Kassem ………………………...……p. 346
“Mathematical modeling of the impact of crown forest fire on buildings and structures”,
V. Perminov ……………………………………………………………………………….…p. 352
“Experimental Study of the Effect of Ceiling Vent on Fuel Mass Loss Rate”, A.Chang
Liu, B. Qiang Li, C. Yang Jiang ………………………………………………….…p. 357
“Fire risk assessment of Ethylene storage tanks using Dow Fire and Explosion Index
(F&EI)”, Nihal Anwar Siddiqui, Prasenjit Mondal, Bikrama Prasad Yadav and Junim Akthar
……………………………………………………………………………………………..….p. 362
“Seasonal forecast maps of meteorological fire danger (FWI): Case study for the
municipality of Eastern Attika-Greece”, Athanasios Sfetsos, Vassiliki Varela, Diamando
Vlahogianni , Nikolaos Gounaris …………………………………………………….………p. 367
Page 10
On the possibility of assessing fire protection levels
Vladimir Mozer
Faculty of Security Engineering, University of Zilina, 1. maja 32, Zilina 010 26, Slovak republic
Abstract The possibility of direct comparison of fire safety levels achieved through the implementation of various fire
protection measures and systems has long been a topic of interest in the field of fire engineering. The difficulty
usually lies with the selection of a comparability factor which would allow for a direct quantitative comparison. This
paper deals with one aspect of the problem – comparing the economic impact of a fire based on the area damaged.
By the utility of probabilistic fire modelling, a set of fire scenarios are evaluated with 16 different levels of fire
protection. The fire safety measures considered are fire alarm system, portable fire extinguishers, sprinkler
protection and compartmentation. Firstly, the area damaged by the fire was established the impact was evaluated in
relation to the expected occurrence of fire, i.e. what yearly fire-related damage may be expected for each fire
scenario/level of fire protection. Subsequently, mathematical relationships for fire protection system justification
based on yearly cost of fire protection and fire loss reduction were established.
Corresponding author: [email protected]
Proceedings of the 2nd
IAFSS European Symposium of Fire Safety Science
1. Introduction
In the process of fire safety design of a building a
number of fire protection options may arise. Although
the minimum required level is usually set in legislation
and standards, there may be various approaches in
addressing them or stakeholder(s) place an additional
performance objective. This often includes property
protection, business continuity and heritage protection.
The fire safety engineer is then faced with a task of
finding such a combination of fire protection systems
which minimise potential fire loss and threat to life. The
second selection criterion in the process is the cost of
such a combination.
The economic justification of a fire safety design
alternative may be a relatively complicated task which
depends on an array of input parameters, some of which
are not readily available [1]. Having said that, there are
approaches which can be utilized to establish, on a
probabilistic basis, how a given set of fire protection
measures is expected to perform. Most often the event
tree analysis (ETA) is applied to this type of problem as
it is relatively simple and useful when little data is
available on the outcomes of concern [2].
This paper will examine the utility of probabilistic
fire modelling, namely the aforementioned ETA, in
ranking fire safety levels for various fire protection
alternatives.
2. Calculation approach
The modelled situation is represented in a series of
nodal events ordered in a sequence. From the initiating
event the nodal events are “branching” towards the
individual outcomes, each representing a specific
scenario.
A general form of an event tree is shown in Figure 1.
The frequency of each of the outcomes Fx is then
expressed as:
xx PFF . (1)
where F is the frequency of the initiating event – a fire
starting in a given type of occupancy, and Px represent
the probabilities of nodal events occurring.
Figure 1 General form of an event tree [3]
The problem with this type of analysis, however, is
the limited availability of statistical data of required
detail and structure, confirmed by recent studies of
Slovak fire statistics [4] [5]. Whereas the data for
deterministic fire models may be acquired via various
methods of testing in relatively shorts periods of time
(e.g. [6] [7]), gathering the necessary statistical data is a
long-term process. Engineering judgement and
approximation have therefore often to be used.
3. Outcome interpretation
Since an event tree results in a number of potential
outcomes it is very important how these results are
interpreted and accounted for in the final analysis. There
are two alternatives:
1. selecting the most probable outcome and its
occurrence interval as the representative value;
2. accounting for each of the outcomes identified
with respect to their occurrence intervals.
289
Page 11
The first alternative appears to be applicable only for
cases when one of the outcomes has a significantly
higher occurrence frequency compared to the others.
If fire damaged area is used as representative of an
outcome, then it is possible to express the second
alternative simply as:
n
i i
i
F
SS
1
F,d
d (2)
where Sd is the average expected fire damage
representative of a particular set of fire protection
measures (nodal events) [m2.yr
-1], Sd,Fi is fire damaged
area associated with the i-th outcome [m2], and Fi
represent the frequencies of the individual outcomes
[1.yr-1
].
The outcome is therefore expressed in m2 of fire
damaged area of the building per year for the purposes
of this paper. It is also possible to express the outcome
as a probability of a fire fatality or injury etc.
4. Model scenario definition
Lets consider a fabricated building which has a total
floor area of 4000 m2 and is divided into four identical
fire compartments, each having a floor area of 1000m2.
There are four fire protection measures available for
this particular building: compartmentation, fire alarm
system, fire extinguishers and sprinkler protection.
These measures yield 16 potential alternatives of fire
protection as listed in Table 1.
Table 1. Fire protection alternatives considered
No. Fire alarm Fire exting. Sprinklers Compartment
1 N N N N
2 Y N N N
3 N Y N N
4 N N Y N
5 N N N Y
6 N Y N Y
7 Y N Y N
8 N Y Y N
9 Y Y N N
10 Y N N Y
11 N N Y Y
12 Y Y Y N
13 Y Y N Y
14 Y N Y Y
15 N Y Y Y
16 Y Y Y Y
Obviously, for a real building the minimum
requirement for each of the above measures would be
set in legislation or standards, depending on the building
type, size etc. This would lead to rejection of certain
combinations from Table 1, however, for demonstration
purposes they are all considered in our example.
5. Specification of nodal events and interactions
Having specified the building and fire protection
levels an event tree was constructed; please, refer to
Figure 1 at the end of the paper. The individual nodal
events and associated probabilities are discussed below.
It should be pointed out that the probabilities and
other data included are for demonstration purposes only,
despite being extracted mostly from peer-reviewed or
official sources. The purpose of their inclusion was
avoidance of full use of fabricated values which could
lead to skewed results.
Because the development of a fire is not solely
driven by fire protection measures in place but also by
fuel and ignition source configuration and other
parameters the first nodal event after ignition was
specified as spread beyond the first item ignited.
Various sources [3] [8] [9] indicate that a relatively
large proportion, approx. 40%, of fires actually never
grow beyond the first item ignited. To err on the side of
safety, due to high variability, the probability of fire
spread from the first item ignited was selected to be 0.8,
i.e. twice as probable as the studies indicate.
The second nodal event was automatic fire detection
(Fire alarm), i.e. what is the probability that the
detectors will activate and raise alarm to building
occupants. Successful detection was assigned a
probability of 0.85, from the interval of 0.8 – 0.9, as
indicated in [3] [10] [11], and is representative of smoke
detectors.
No automatic notification of the fire brigade was
assumed, however, if the fire alarm system was not
considered, the delay in discovering a fire was
translated into a lower probability of successful fire
suppression by portable fire extinguishers and a higher
probability of a fire involving the entire compartment
due to delayed fire brigade attendance (called by
occupants).
The third nodal event was manual fire suppression
by portable fire extinguishers (Fire exting.). The rate of
successful fire suppression varies significantly from
25% to 95% [12]. The discussion provided in [12],
which reviews a number of sources, led to a decision to
discard the extreme values, leaving a range of 40 – 85%.
From this the probability of successful suppression by
fire extinguishers was taken as 0.6 when automatic
detection was present and activated. When there was no
detection or its activation failed, the suppression success
probability was decreased by 50% to 0.3 due to the
likely delay in discovering the fire resulting in a
prolonged growth period.
The fourth nodal even was specified as sprinkler
suppression. Sprinkler systems are well documented and
the data available from various sources [3][11][13][14]
indicate that the probability of successful fire
extinguishment is approximately 0.9. This value is used
in the event tree for cases where sprinkler protection is
assumed.
The penultimate nodal event represents the
capability of fire compartmentation to contain the fire
within the compartment of origin for fire-fighting
purposes. There is also an alternative of burnout, which
was considered as an alternative fire extinction for when
the fire brigade failed to extinguish the fire. The
290
Page 12
probability that compartment boundaries will contain
the fire inside the compartment of origin was taken to be
0.8, deriving it from the values of 0.7 – 0.9 indicated in
[3][15][16].
The final nodal event represents the success of fire-
fighting operations performed by the fire brigade. There
are four alternative probabilities specified, arbitrarily
adjusted basing on [2][3][12][17][18]. These are 0.8,
0.6, 0.6 and 0.5, depending on the performance of fire
detection and compartmentation.
When any of the fire protection means were not
assumed in the calculation their probability of
successful operation was set to 0.
The following areas were specified for the individual
outcomes, mainly deriving from[2] [3][11]:
1 m2 – fire contained to 1
st item ignited
2.5 m2 – fire suppressed by fire extinguisher
10 m2 – fire suppressed by sprinkler system
500 m2 – compartment and fire-brigade successful
1000 m2 – compartment burnout
2000 m2 – compartment failed / fire-brigade successful
4000 m2 – complete burnout
The above difference between 10 m2 and 500 m
2
may seem rather large, however, it would be difficult to
specify further segmentation as no specific compartment
layout – rooms – is considered in this study.
Following a review [19] of various sources of fire
occurrence probabilities it was decided a value of 0.005
as the probability of the initiating event. The selected
value represents an average absolute probability of a fire
starting derived from 10 building occupancy types.
6. Results and discussion
The results of the above described ETA calculations
are summarised in Table 2. For each level of fire
protection an expected fire damage per year was
determined. The yearly expected fire damage ranges
from 0.2 m2 for full protection to 12m
2 for no fire
protection.
The results are divided into two groups sprinklered
and nonsprinklered. For this particular case the yearly
fire damage would indicate that having the building
protected with a sprinkler system offers a very similar
level of protection than having the building divided into
fire compartments, fitted with a fire alarm system and
portable fire extinguishers. Of course this result must
not be generalised as it is valid only for the particular
fabricated case.
It may be also concluded that without
compartmentation the expected damage remains high;
this is of course relative to the value density discussed
below. This conclusion also correlates with the fact that
sprinkler protection is often required in large
uncompartmented buildings.
For the scenarios with sprinkler protection the
decrease of fire damaged area achieved through the
provision of additional fire protection measure is less
pronounced. In relative terms, the fire damage for the
case with all the fire protection measures in place is six
times lower than for sprinkler protection alone,
however, in absolute terms, the difference is about 1 m2.
Table 2. Results of ETA calculations – fire damage for
various levels of fire protection No. Fire alarm Fire exting. Sprinklers Compartment Damage
[m2/year]
1 N N N N 12
2 Y N N N 11.28
3 N Y N N 8.4
9 Y Y N N 4.88
5 N N N Y 4.64
10 Y N N Y 4.21
6 N Y N Y 3.25
13 Y Y N Y 1.83
4 N N Y N 1.24
7 Y N Y N 1.17
8 N Y Y N 0.87
12 Y Y Y N 0.51
11 N N Y Y 0.5
14 Y N Y Y 0.46
15 N Y Y Y 0.35
16 Y Y Y Y 0.2
One of the factors which affects the installation of
any particular fire protection measure or their
combination is the average value of protected property
per unit of area; value density – Vd [€.m-2
].
Furthermore it is not just the direct loss, expressed
as the product of Vd and Sd, but also an array of losses
which could be for the purposes of this paper summed
as indirect loss - Li [€.yr-1
]. These would include
business interruption, cost of fire brigade operation,
environmental impact, etc.
The final factor is the cost of fire protection per year
– Cp [€.yr-1
]. Considering solely the property protection
objective, the increase in the yearly cost of fire
protection measures should never exceed the expected
reduction in yearly loss associated with fire. The
justification criterion for the inclusion of a particular
fire protection measure could be expressed through an
efficiency factor ce:
).().().( p1p2ei2d2d2i1d1d1 CCcLVSLVS (3)
From Equation (3) the efficiency factor ce can be
expressed as:
)(
).().(
p1p2
i2d2d2i1d1d1e
CC
LVSLVSc
(4)
ce ≤ 1 inclusion not justified
ce > 1 inclusion justified
Where the subscript 1 indicates a design
configuration without a particular fire safety measure
and the subscript 2 indicates a design configuration in
which the fire safety measure has been included.
The economic efficiency of a fire protection measure
grows proportionally with ce.
291
Page 13
For selection of an appropriate level of fire
protection, when a number of systems or measures are
being considered in a particular building design, the
combination with the highest value ce of should be
adopted, from the property protection point of view.
Such a combination of fire protection systems has the
highest economic efficiency of funds invested.
It should be reminded again, that property protection
is not and, for most cases, cannot be the sole
performance objective criterion to be considered when
selecting appropriate level fire protection. In a real
situation, some of the fire protection levels (measures
combinations) would not be considered if they did not
meet the minimum legislative or standard requirements
for fire safety, which are usually concerned with life
safety.
7. Conclusion
The purpose of this work was to analyse the
potential of probabilistic fire modelling, using the event
tree analysis approach, to assist the fire safety engineer
in selecting an appropriate level of fire protection when
multiple design alternatives are available.
For a model building 16 various level of fire
protection were analysed, ranging from no protection to
full protection, comprising a fire alarm system, portable
fire extinguishers, sprinkler protection and
compartmentation.
The focus of the comparison was property protection
and the criterion evaluated was the expected yearly fire-
damaged area.
As expected sprinkler protection was identified to
have the greatest impact on reducing the fire damage,
however, the combination of the other three measures
offered a similar degree of damage reduction.
Analyses such as the one presented in this paper
provide useful output for cost-benefit assessment. The
final selection of an appropriate level of fire protection
should be based on minimum life safety requirements
and the most economically efficient combination of fire
protection measures; the highest value of efficiency
factor ce.
Acknowledgement
This work was supported by the Slovak Research and
Development Agency under the contract No. APVV-
0727-12.
References
[1] RAMACHANDRAN, G. The economics of fire
protection. London: E & FN Spon, 1998. ISBN
0419207805.
[2] RAMACHANDRAN, G., CHARTERS, D. A.
Quantitative risk assessment in fire safety.
London, New York: Spon Press, 2011. ISBN
9780419207900.
[3] PD 7974-7. Part 7 Probabilistic risk assessment.
London: BSI, 2003.
[4] CONEVA, I., PANÁKOVÁ, J.: Požiare a ich
následky na základe kategorizácie budov In:
Advances in fire and safety engineering 2014:
recenzovaný zborník: Trnava: AlumniPress, 2014.
- ISBN 978-80-8096-202-9.
[5] CONEVA, I., PANÁKOVÁ, J.: Požiarovosť podľa
kategorizácie stavebných objektov za obdobie
2008-2012 In: Bezpečnost, spolehlivost a rizika
2014 – Zborník konferencie, Liberec: Technická
univerzita, 2014. - ISBN 978-80-7494-110-8.
[6] XU, Q., ZACHAR, M., MAJLINGOVÁ, A., JIN,
C., JIANG Y.: Evaluation of plywood fire
behaviour by ISO tests in: European Journal of
Environmental and Safety Sciences Vol. 1(1)
(2013)
[7] MARTINKA, J., CHREBET, T., BALOG, K.: An
assessment of petrol fire risk by oxygen
consumption calorimetry in: Journal of Thermal
Analysis and Calorimetry Vol. 115(3) (2014)
[8] HASOFER, A. M., BECK, V. R., BENNETTS, I.
D. Risk analysis in building fire safety
engineering. Amsterdam, London: Butterworth-
Heinemann, 2007. ISBN 075068156X.
[9] MÓZER, V., WILKINSON, P., SMOLKA, M.,
TOFILO, P. Probabilistic-deterministic modelling
of fire spread [online]. European Journal of
Environmental and Safety Sciences. 2014, 2(2),
79-83. Available from: http://www.sci-
institute.com/2014_volume_2_issue_2/9a_mozer.p
df.
[10] BUKOWSKI, R. W., BUDNICK, E. K.,
SCHEMEL, C. F. Estimates of the Operational
Reliability of Fire Protection Systems. Third
international conference on fire research and
engineering,. Conference proceedings Chicago,
Ill., October 4-8, 1999. Bethesda, Md: Society of
Fire Protection Engineers, 1999, pp. 87-98.
[11] RASBASH, D. Evaluation of fire safety.
Chichester, England, Hoboken, NJ: John Wiley,
2004. ISBN 0471493821.
[12] GHOSH, B. Assessment of the benefits of Fire
Extinguishers as fire safety precautions in New
Zealand Buildings. Christchurch, New Zealand,
2008.
[13] BUKOWSKI, R. W., BUDNICK, E. K.,
SCHEMEL, C. F. Estimates of the Operational
Reliability of Fire Protection Systems. Third
international conference on fire research and
engineering,. Conference proceedings Chicago,
Ill., October 4-8, 1999. Bethesda, Md: Society of
Fire Protection Engineers, 1999, pp. 87-98.
[14] FITZGERALD, R. W. Building fire performance
analysis. Chichester: Wiley, 2004. ISBN 0-470-
86326-9.
[15] TYLDESLEY, A., REW, P., HOULDING, R.
Benefits of Fire Compartmentation in Chemical
Warehouses. Process Safety and Environmental
Protection. 2004, 82(5), 331-340.
[16] HOULDING, R. C., REW, P. J. Benefits of fire
compartmentation in chemical warehouses.
Research report 152. Norwich, 2003.
292
Page 14
[17] MELINEK, S. Fire Fighting and Extent of Spread.
FRN 912. Borehamwood, 1972. Fire Research
Notes.
[18] SÄRDQVIST, S., HOLMSTEDT, G. Correlation
Between Firefighting Operation and Fire Area:
Analysis of Statistics. Fire Technology. 2000,
36(2), 109-130.
[19] MÓZER, V., WILKINSON, P., TOFILO, P. Use
of Statistical Data in Fire Engineering Design – an
International Comparison. Zeszyty Naukowe SGSP.
2015, 53(1), 167-181.
Figure 1
Ignition Spread Fire alarm Fire exting. Sprinklers Compartm. Fire brigade Damage
0.6 2.5 m2
Y
0.9 0.9 10 m2
Y Y
0.8 N
0.8 500 m2
Y 0.4 0.8 Y
5.00E-03
Y N 1000 m2
Y
N 0.2
0.1 0.6 2000 m2
N Y
0.2 N 4000 m2
0.4
0.3 2.5 m2
Y
N 0.9 10 m2
0.1 Y
N
0.6 500 m2
0.7 0.8 Y
Y N 1000 m2
N 0.4
0.1 0.5 2000 m2
N Y
0.2 N 4000 m2
0.5
N 1 m2
0.2
N 0 m2
293
Page 15
Center for Risk, Safety and the Environment
(CERISE)
ISBN 978-9963-2177-0-0