Вы находитесь на странице: 1из 17

Essam Eldine Mouguib et al.

/ (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES


Vol No. 5, Issue No. 2, 212 - 228

Numerical Simulation of Warehouses


Fire Suppression
Essam Eldine Mouguib Mahmoud Ahmed Fouad
M.Sc, Mech. Power Dept. Professor, Mech. Power Dept.
Faculty of Engineering, Cairo University Faculty of Engineering, Cairo University
Giza, Egypt Giza, Egypt
emouguib@hotmail.com dr.mahmoud.fouad@gmail.com

Abstract— A CFD simulation has been conducted to study


Warehouses represent a unique fire challenge to both fixed fire
the rack storage fires and suppression means in a pharmaceutical suppression systems and the manual firefighting forces that
warehouse. Simulations have been carried out for different fire are called upon to deal with a fire. Modern warehouses and
locations and rack storage geometries, to predict fire growth rate storage occupancies are especially subject to rapidly
developing fires of great intensity, because complex

T
and spread. Also, the activation time periods of in-rack and
ESFR sprinklers, fire growth control and fire suppression have configuration of storage and building layout are usually
been simulated. The use of the foam-water sprinkler system has conducive to fire spread, presenting numerous obstacles to
been also considered. Simulations results showed that, the in- manual fire suppression efforts. The only proven method of
rack sprinkler would actuate faster than the ESFR ceiling controlling a warehouse fire is within properly designed and
mounted sprinklers. The successive operation of the adjacent
maintained automatic sprinkler systems. If sprinkler protection
nearby in-rack sprinklers has a great effect on the control on the
is not provided, the likelihood of controlling a fire in a
ES
fire growth. Also, the in-rack sprinklers have extinguished the
fire faster than ESFR sprinklers, due to the fast control of fire
growth. The foam-water sprinkler system has controlled the fire
growth in such time slightly more than the in-rack sprinklers and
considerably more than the ESFR sprinklers. The foam-water
warehouse is minimal. Some critical elements must be
considered when developing a comprehensive risk mitigation
strategy to protect various facilities. These elements include
commodity classification, common storage configurations,
sprinkler system has the fastest suppression, compared to other various protection schemes, hazards associated with some of
cases, due to the great effect of the foam solution on the fire the common types of warehouses and loss prevention
spread. Also, the foam-water sprinkler system does not destroy guidelines for minimizing the frequency and severity of a loss.
product, due to the lower water content. They have limited smoke
damage, and because of the detergent properties of the foaming
Warehouses can range from several hundred to more than a
agent, they provide a self-cleaning effect. When studying the million square feet and can include among other occupancies
effect of the rack storage geometry, it is found that the narrow storage garages, refrigerated storage facilities, isolated storage
vertical and horizontal flues have a great effect on the fire buildings, underground storage locations, and air-supported
A
growth, as they do not allow the fire spread to adjacent surfaces, structures. A variety of commodities is displayed and stored
which facilitate the sprinklers job to control the fire. Also, the within these facilities, including soft goods, clothes,
narrow vertical and horizontal flues have a great effect on the furnishings of all types, bedding materials, paints, home repair
fire suppression. The storage height has a strong impact on the and building materials, chemicals, and plastics. Moreover, big
sprinkler activation. Upon the obtained results, the best sprinkler box retail spaces often have ceiling or roof heights in excess of
activation was dedicated to the in-rack sprinklers. The best
suppression period was dedicated to the foam-water sprinkler
16 feet and, in many cases, as high as 35 to 40 feet. Using rack
storage configurations, these types of retail stores will
IJ

system. To get a better suppression performance for high bay


warehouses fires, in-rack sprinklers can be used along with foam- typically display products at lower elevations and use the
water sprinkler system. But this configuration has a remarkable higher elevations for product storage.
impact on the economic-wise criteria. So, in order to have a
reasonable optimal configuration, in-rack sprinklers can be NOMENCLATURE
installed along with ESFR ceiling sprinklers.
A area of the plume at a given height
b width of the storage boxes.
Keywords-component; in-rack sprinklers; ESFR sprinklers; Cb non-dimensional constant and it can be taken as 3.4 according to
foam-water sprinkler system; activation time; fire growth Heskestad [12]
control; fire suppression. CD drag coefficient of the droplets
Cp specific heat
I. INTRODUCTION CT non-dimensional constant and it can be taken as 0.12 according to
Heskestad [12]
Among the most challenging occupancies from a property Cu non-dimensional constant and it can be taken as 9.1 according to
loss control viewpoint are warehouses, distribution centers and Heskestad [12]
large retail businesses referred to as ―
big box‖ establishments. d plume diameter
Ds drag of the spray

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 212


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

g acceleration of gravity 2.2 Descriptions of Warehouse Geometrical Configuration


h height of the horizontal flue
Hc heat of combustion per unit mass of fuel The warehouse layout consists of one level building of
Kd diameter constant, ranges from 0.25 to 0.5 total floor area of 3700 m2 and total floor to ceiling height of
l longitudinal length of the rack 13.7 m. Fig.1 describes the warehouse plan layout. The
Lfi mean flame height commodity is stored in single and double rows racks with
mi mass flow rate of air at the ith tier slatted shelves. The rack height is of 1.5 m and has an area of
mn mass flow rate of air at the highest tier
me total mass-entrainment rate of air below the level of the flame tip 2.8 m x 1.2 m. The flue space was 0.2 m and the aisle width
mf fuel flow rate was 1.2 m. The commodity stored in the warehouse is
Mp plume momentum composed of pharmaceutical commodities and chemical
n total number of drops in the spray materials. According to NFPA code, these commodities are
Pn,a static pressure at the top of the rack storage classified as Class IV commodity. They can include pills,
Q total heat release rate from the burner
Qc convective heat energy released from the burner
powders, and chemicals, put in glass or plastic bottles placed in
Qci,a released convective energy at the top of the ith box corrugated carton boxes. The carton sheets which used to form
Qci,b released convective energy at the base of the ith box the boxes have a thickness of 10 mm with the following
Qmax max heat release rate thermal data: heat conductivity 0.12 W/m oC, specific heat 800
r plume radius J/kg and density 700-800 kg/m3
RTI response time index of sprinkler
s height of the boxes.
t time in sec 2.3 Model Description
tg fire growth time to reach 1055 kW (1000 btu/sec) The model, as shown in fig.2, consists of a two-
tc ramp-up time from ignition associated with the transport of hot air

T
from fire to the sprinkler location
dimensional model of 13.7 m height and 6.4 m width. The
tv virtual time commodity is arranged in 18 boxes, each have a height of 1.5
T gas temperature m and a surface area of 2.4 m x 1.2 m. The maximum storage
Tmax max gas temperature height is 10.8 m. There are four air outlets placed on both
Ti,a air temperature at the top of the box i
Ti,b air temperature at the base of the box i
sides of the model, each of 1.5 m height and are located at
T air ambient temperature altitude of 2.02 m and 7.48 m respectively. The probability of
fire hazard is considered in four locations inside the model, as
T
Tr
Ti,b
Ti,l
To
temperature rating of the sprinkler
gas temperature at the base of each box
bulk temperature
centerline excess gas temperature
ES
average hot air temperature at the sprinkler location
shown in fig.2 and the fire pool has a surface area of 1.4 m 2.
The sprinkler system consists of seven in-rack sprinklers
located on two levels and two ceiling sprinklers and it is
distributed according to NFPA requirements.
u average hot air velocity at the sprinkler location

v velocity of the plume


velocity of the droplets
vd ‫ﻩﺏﻥﻱ ﺍﻝﺡﺭﻙﺕ‬ ‫ﺥﺱﺍﻯ‬

w width of the vertical flue


‫ﻁﻑﺍء‬. ‫ﻭﺍ‬ ‫ﺍﻝﻭﻱﺍﺓ‬

z height from floor level


z0 height of virtual origin
 fire growth rate = [1055/tg2]
water
A
‫ﺍﻝﻭﻙﺕﺏ‬ 24.90

 Descending rate of HRR.


water

‫ﺍﻝﻉﻝﻭﻱ‬
9.30 2.20

11.70
‫ﺍﻝﻉﻕﻱﻭﺕ‬

 Sprinkler location index


‫ﺍﻝﻭﻁ‬
‫ﺏﻉﺕ ﻭﺍﻝﻉ‬

 air density at ambient air


‫ﺍألﻕﺭﺍﺹ‬
‫ﻱﺍﺩﺓ‬

4‫ﻕ‬ ‫ﺍﻝﻑﻯﺍﺭ‬
‫ﺡﺫﻱﻕﺕ‬

47.65

48.85
‫ﻩﻯالﺩﺍﺙ ﺍﻝﺱﻯالﺭ‬

p density of the plume ‫ﺍﻝﻭﻉﺍﻩﻝ‬


‫ﺍﻝﻯﺭﺵﺕ‬

‫ﺍﻝﺥﻝ‬

‫ﺍﻝﺱﻯﺍﺉﻝ‬
‫ﻱﺕ‬

o sprinkler actuation time 1‫ﻕ‬


‫ﺍﻝﻍال‬

‫ﻩﺡﻯالﺙ ﺍﻝﻙ ﻩﺭ‬


‫ﻍﺭ‬
‫ﻱﺕ‬

‫ﻑﺕ ﺍﻝﻙ ﻩﺭ‬

water

IRS In-rack sprinkler


‫ﺏﺍء‬

‫ﺏﺍء‬

‫ﻩﻥﺍﻭﻝﺕ ﺍﻝﻩﻯﺍء‬

w.c
CS
water

Ceiling sprinkler ‫ﻩﺏﻥﻱ ﺍﻝﻭﺽﺍﺩ ﺍﻝﺡﻱﻯﻱ‬


w.c
IJ

3‫ﻕ‬
‫ﻩﺏﻥﻱ‬ water

‫ﺕﻍﻱﻱﺭ‬
w.c

‫ﺍﻝﻭالﺏﺱ‬

II. MATHEMATICAL MODELING


w.c
‫ﻩﻥﻁﻕﺕ ﺍألﺱ‬

‫ﻩﻥﻁﻕﺕ ﺍﻝﻯﺯﻯﺏﺍﻝﻭﺥﺍﺯﻯ‬ ‫ﻩﺏﻥﻱ ﺍﻝﻭﺥﺍﺯﻯ ﺍﻝﻕﺫﻱﻭﺕ‬


‫ﺡﺫﻱﻕﺕ‬
‫ﺕالﻡ‬

water

‫ﻩﻙﺍﺕﺏ‬

2.1 Introduction ‫ﺍﻝﻭﺥﺍﺯﻯ‬


w.c

water

2‫ﻕ‬
w.c

The present work has used the FLUENT model version ‫ﻩﺏﻥﻱ ﺍﻝﻭﻁﻉﻥ‬
w.c

water

‫ﻩﺏﻥﻱ‬
‫ﺍﻝﻭﺥﺍﺯﻯ ﺍﻝﺝﺫﻱﺫﺓ‬

6.3.26, to built-up the numerical CFD models describing the


various case studies discussed in this work. The GAMBIT
Elec

w.c w.c
water water
Elec

preprocessor geometry builder version 2.2.30, developed by


Fluent, was used to built-up the models geometry and mesh
generating. 55.65

Fig.1 Warehouse Plan Layout

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 213


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

HRR
F# Fire hazard probability
A # Air outlet window

Q= Qma x

Q= ( t ) ² Q= ( t - t 4 ) ²

t 1 t 2 t 3 t 4
t ime

Fig. 3 Develop Curve of HRR of a Building Fire

characterized by their rate of heat release, measured in terms


of the number of kW (Btu/sec) of heat liberated. Previous
researches have shown that most fires grow exponentially and
can be expressed by what is termed the ― power law of fire
growth model,‖ which follows:

T
Q tp (1)
where: p equals 2.
In fire protection, fuel packages are often described as
having a growth time (tg). It is the time necessary after the
ignition with a stable flame for the fuel package to attain a
heat release rate of 1055 kW (1000 Btu/sec). The following
ES equations describe the growth of design fires:
1055
Q  2 t2
tg
for SI units (2)

Equation (3.2) can be generally expressed as:


Q  t 2 (3)
Fig. 2 Model Description
2.4.1 Fire Plume Generation
Four ignition sources were mounted at several locations,
2.4 Fire Modeling as shown in fig.2. Each ignition source consisted of a square
The ignition in a warehouse initially occurs at certain burner nozzle (25cm x 25cm) located at the floor area, where
location and then grows up at a rate depending on several the gas fuel is injected. The fire plume is created when the fuel
A
factors such as the type of fuel, oxygen access, type of (methane) injected from the burner burns in the presence of
commodity and the configuration of storage. Subsequently, the oxygen, high temperature and minimum concentration of the
developed fire transitions to the flashover period, which is a reactants. A single-step irreversible chemical reaction is
rapid transition from the growth period to a fully developed assumed:
fire period, leading the total surface of the combustible CH4 + 2O2  CO2 + 2H2O (4)
materials to be involved in the fire. At the fully developed fire
The combustion reaction and airflow can be described by
IJ

stage, the heat release rate (HRR) and average temperature


the conservation equations of mass, momentum, energy and
reach their peaks while the fire is rapidly spreading to other
species along with the sub-models describing the turbulence
locations through various paths. If the initial fire in the initial
location is not discovered and suppressed in the first place, it and combustion. The standard K- model is used to estimate
the turbulence characteristics of the gas phase flow, by solving
will eventually spread to the whole warehouse. It is assumed
the equations of turbulence kinetic energy and the dissipation
that all buildings have approximately the same fire-
rate, so as to calculate the turbulent effective diffusion
development process which consists of five stages: ignition,
coefficient.
flashover, full-development, collapse and extinguishment.
Fig.3, illustrates the fire growth behavior and the 2.5 Governing Equations
development of air temperature and heat release rate, where t 1 The theoretical model is used to calculate the fire
is the time from ignition to flashover, t 2, time from flashover parameters in a two-dimensional rack storage configuration.
to full-development, t3, time from full-development to collapse The model predicts the air temperature and velocity and the
and t4, time from collapse to extinguishment. Fires can be flame diameter in the flues. Gas temperature and velocity are

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 214


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

represented by a mean value across the cross-section of the T 


1/ 2

flue. The input parameters used in the model are: bT  C b  0  z  z 0  (11)
 The longitudinal length of the rack (l).  T 
 The height (s) and the width (b) of the boxes.
 The width of the vertical flue (w) 2.7 Plume Width and Flame Height
 The height of the horizontal flue (h). As a plume rises, it entrains air and widens. Generally the
 The convective heat energy released from the burner (Qc) total plume diameter and height can be estimated as:
d  kd z (12)
2.5.1 Continuity Equation L f  3.73w  0.343Q 2 / 5 (13)
By using the continuity equation, we can find that:
m1  2m 0, 0 2.8 Modeling of Water Spraying
. To a better understanding of the fire suppression, it is
(5) useful to consider the reaction of the flame and fire plume to
m i  m i 1  2m 0,i
the droplet spray and to consider this situation as a
competition between the downward momentum of the spray
m n  m n 1  2m n , n and the upward momentum of the fire plume. If the downward
spray is strong enough to balance or overpower the upward
where i = 1, ………, n (n = 6) momentum of the fire plume, the structure of the fire plume

T
changes. The momentum of the fire plume, Mp, is used to
characterize the fire size. However, to characterize the
2.5.2 Conservation of Energy strength of the spray, the drag of the spray, Ds, is used since
When the mass flow rate in the vertical flue of each tier is this is the physical mechanism of the interaction between the
known, while the mass flow rate is constant between the base droplets and the gas of the plume. Thus, a spray that has a very
and the top of each box, the temperature at the corresponding large effect on a fire plume does so by creating a large drag on
height can be calculated using the following expressions:
Qci, a  mi c p (Ti , a  T )
Qci, a  m i c p (Ti , a  T )
Accordingly, the temperatures at the corresponding
ES (6)
the fire plume. The ratio of the drag of the spray to the
momentum of the plume, Ds/Mp, is a non-dimensional
parameter that characterizes the effect of the spray on the
dynamics of the fire. To calculate this parameter, the drag of
the spray, Ds, can be expressed as:
locations can be calculated as shown in the following section. 1
Also, by using the relationship for the mass flow rate: Ds  n p C D Ad (v vd ) v  vd (14)
2
m  uwl (7)
The drag coefficient for the droplet, CD, depends primarily on
and the ideal gas law : the Reynolds number based on the droplet slip velocity:
T    T (8) U V d
Re   (15)
where the influences of pressure changes and gas composition

are neglected, the velocity at the top and the base of each tier
A
can be calculated as shown in the following section. The plume momentum can be calculated from the plume
velocity profile and width as
2.6 In-Rack Temperature and Velocity M p   v(r 2 )dA (16)
Turbulent buoyant axi-symmetric fire plumes with a large A

density defect or temperature rise relative to the surrounding 2.8.1 Calculation of the Sprinkler Actuation Time
are known as strong plumes, while plumes with a small The heat flow into a sprinkler heat sensing element occurs
IJ

density defect or temperature rise are known as weak plumes. over a period of time. The thermal response coefficient is
Above axi-symmetric buoyant turbulent diffusion flames, the needed to accurately predict the heat sensing element
centerline values of excess temperature and velocity and the response. A measure of the speed with which heat transfer
plume radius obey the following relationships: occurs is currently called the detector time constant (0). The
 
1/ 3 time constant is a measure of the sensitivity of the sprinkler
T0 1 Qc2 / 3 (9)
 CT  2 2 2  sensing element. Upon calculating the air temperature and
T   z  z 0 5 / 3
 gc p   T  velocity at the sprinkler location, the sprinkler actuation time
(0), can be obtained using the following equation:
 g 
1/ 3
Qc1 / 3 RTI  T  T  (17)
u 0  Cu   (10) o  ln    tc
c  T   
 p    z  z 0  1/ 3
u  T  Tr 

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 215


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

2.9 Matrix of Experiments 3.1 Fire Simulation


The conducted CFD calculations have been classified At the beginning, the fire is growing exponentially with
according to the following criteria: time and the heat release rate (HRR) takes a "t squared" shape
1. The consideration of four probabilities of fire profile. The phenomena of air temperature growing with time
hazards, as shown in fig.2. have been developed by using user-defined functions written
2. The operation of the following systems to suppress in C++ code. When exporting these user-defined functions
each of these fire hazards: into the model, we can get a prediction of the behavior of the
 In-rack and standard ceiling sprinklers. unsteady temperature rise with time. Figures 4.a, 4.b, 4.c and
 Early-Suppression-Fast-Response (ESFR) 4.d show the simulation results for the contours of the total
sprinklers. temperature, obtained after several times, along the
 Foam-Water Sprinkler system. symmetrical axis of the flame (fire hazard vertical center line)
3. The effect of rack storage geometrical configuration for the four fire hazard probabilities studied for case study A.
on fire suppression The fire reaches the fully-developed stage, where the heat
release rate and air temperature reach its maximum values.
TABLE 1. MATRIX OF EXPERIMENTS Fig.4.e shows the contours of total temperature of the fully-
developed fire obtained along the symmetrical axis of the
Rack Storage flame.
Geometrical Suppression System Used
Name of Fire Probability

Configuration 3.1.2 Effect of Rack Storage Geometrical Configuration on


Case Study Name

T
Fire Growth
Vertical Flue Space

Foam-Water
In-Rack and
Suppression
Horizontal Flue

Sprinklers

Sprinklers

Sprinkler
Standard

Cases B1-1, B2-1 and B3-1 represent the three fire hazard
Without

Ceiling

System
ESFR
Space

probabilities studied for case study B with the vertical flue


width equals to 30 cm. Fig.5 represents the simulation
results of total temperature for case study B3-1.
Sub-Case Study Name Cases C1-1, C2-1 and C3-1 represent the three fire hazard

A
v
(cm)

20
h
(cm)

32
fire 1
fire 2
fire 3
fire 4
A1-1
A2-1
A3-1
A4-1
A1-2
A2-2
A3-2
ES A1-3
A2-3
A3-3
A4-3
A1-4 probabilities studied for case study C with the vertical flue
width equals to 40 cm. Fig.6 represents the simulation results
of total temperature for case study C3-1.
Cases D1-1, D2-1 and D3-1 represent the three fire hazard
v h fire 1 B1-1 B1-2 probabilities studied for case study D with the horizontal flue
B (cm) (cm) fire 2 B2-1 B2-2 width equals to 48 cm.
30 32 fire 3 B3-1 B3-2
v h fire 1 C1-1 C1-2
C (cm) (cm) fire 2 C2-1 C2-2
40 32 fire 3 C3-1 C3-2
v h fire 1 D1-1 D1-2
A
(cm) (cm) fire 2 D2-1 D2-2
D
fire 3 D3-1 D3-2
20 48
fire 4 D4-1 D4-2
v h fire 1 E1-1 E1-2
(cm) (cm) fire 2 E2-1 E2-2
E
fire 3 E3-1 E3-2 60 sec 360 sec 480 sec
20 64
fire 4 E4-1 E4-2
IJ

Fig. 4.a Simulation Results of Total Temperature for Case A1-1

III. CFD SIMULATION


The CFD simulations and case studies performed to
predict the activation times of in-rack sprinklers with ceiling
sprinklers and the ESFR sprinkler system as well as the Foam-
Water sprinkler system and the suppression efficiencies of
such systems are presented herein after. The activation times
will be simulated by using a deterministic fire-water
interaction model. The model illustrates the behavior of four
cases of fire hazard probabilities in the warehouse
compartment. Each case is subjected to different suppression 60 sec 360 sec 540 sec
systems including the use of the different suppression systems
illustrated above. Fig. 4.b Simulation Results of Total Temperature for Case A2-1

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 216


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

60 sec 90 sec 510 sec

Fig. 4.c Simulation Results of Total Temperature for Case A3-1


60 sec 90 sec

T
60 sec 180 sec 360 sec 360 sec 660 sec

Fig. 5 Simulation Results of Total Temperature for Case B3-1

case study D3-1.


ES
Fig. 4.d Simulation Results of Total Temperature for Case A4-1

Fig.7 represents the simulation results of total temperature for

Cases E1-1, E2-1 and E3-1 represent the three fire hazard
probabilities studied for case study E with the horizontal flue
width equals to 64 cm. Fig.8 represents the simulation results
of total temperature for case study E3-1.
Cases A4-1, D4-1 and E4-1, represent the three fire
Case A1-1 Case A2-1 hazard probabilities to study the effect of rack storage height
of 10.8m, 11.60m and 12.4m respectively
A
13.5 min 16 min
Case A3-1 Case A4-1
IJ

60 sec 180 sec

13.5 min 10.5 min 600 sec 780 sec

Fig. 4.e Simulation Results of Total Temperature for Fully-Developed Fire Fig.6 Simulation Results of Total Temperature for Case C3-1

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 217


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

60 sec 180 sec 60 sec 180 sec

T
360 sec 540 sec 240 sec 320 sec
Fig.7 Simulation Results of Total Temperature for Case D3-1 Fig.9 Simulation Results of Total Temperature for Case D4-1
ES
60 sec 180 sec 60 sec 180 sec
A
IJ

360 sec 570 sec 240 sec 290 sec


Fig.8 Simulation Results of Total Temperature for Case E3-1
Fig.10 Simulation Results of Total Temperature for Case E4-1

Fig.10a shows the contours of total temperature of the fully- sprinkler (IRS-14) will actuate at t = 59 sec. The second in-
developed fire obtained along the symmetrical axis of the rack sprinkler (IRS-13) will actuate at t = 62 sec. The third in-
flame. rack sprinkler (IRS-15) will actuate at t = 64 sec. The
sprinklers totally control the fire after 170 sec. The total
3.2. Fire Suppression by Water Sprinkler Systems suppression of the fire will occur after approximately 4.75
3.2.1 In-Rack Sprinklers with Standard Ceiling Sprinklers min.
For case A3-2, as shown in fig.11, the first in-rack

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 218


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

Case B3-1 Case C3-1 Case D3-1

59 sec 62 sec 64 sec


18 min 23 min 14 min
Case E3-1 Case D4-1 Case E4-1

T
170 sec 4.75 min

16 min 10 min 8.5 min Fig.11 Total Temperature Contours for Case A3-2
ES
Fig. 10.a Simulation Results of Total Temperature for Fully-Developed Fire

3.2.2 Early-Suppression-Fast-Response (ESFR) Sprinklers


For case A3-3, as shown in fig.12, the first ESFR
sprinkler will actuate at t = 140 sec. The second ESFR
sprinkler will actuate at t = 142 sec. The sprinklers totally
control the fire after 305 sec. The total suppression of the fire
will occur after approximately 7.4 min.

3.3 Foam/Water Sprinkler Systems


140 sec 142 sec 305 sec
For case A1-4, as shown in fig.13, the first sprinkler will
A
actuate at t = 135 sec. The second sprinkler will actuate at
t = 140 sec. The sprinklers totally control the fire after 255
sec. The total suppression of the fire will occur after
approximately 5 min.

3.4. Effect of Rack Storage Geometrical Configuration on Fire


Suppression
IJ

For case B3-2, as shown in fig. 14, the first sprinkler


(IRS-14) will actuate at t = 68 sec. The second sprinkler
(IRS-15) will actuate at t = 72 sec. The third sprinkler (IRS-13) 360 sec 7.4 min
will actuate at t = 78 sec. The sprinklers totally control the fire
after 170 sec. The total suppression of the fire will occur after Fig.12 Total Temperature Contours for Case A3-3
approximately 4.5 min
For case C3-2, as shown in fig.15, the first sprinkler approximately 5 min. For case D3-2, as shown in fig.16, the
(IRS-14) will actuate at t = 72 sec. The second sprinkler first sprinkler (IRS-14) will actuate at t = 68 sec. The second
(IRS-13) will actuate at t = 79 sec. The third sprinkler (IRS-15) sprinkler (IRS-13) will actuate at t = 74 sec. The third
will actuate at t = 83 sec. The sprinklers totally control the fire sprinkler (IRS-15) will actuate at t = 82 sec. The sprinklers
after 170 sec. The total suppression of the fire will occur after will totally control the fire after 180 sec The total suppression
of the fire will occur after approximately 6.5 min

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 219


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

135 sec 140 sec 255 sec 72 sec 79 sec 83 sec

T
270 sec 300 sec 170 sec 300 sec
Fig.13 Total Temperature Contours for Case A1-4 Fig.15 Total Temperature Contours for Case C3-2
ES
68 sec 72 sec 78 sec 68 sec 74 sec 82 sec
A
IJ

170 sec 270 sec 180 sec 390 sec

Fig.14 Total Temperature Contours for Case B3-2 Fig.16 Total Temperature Contours for Case D3-2

For case E3-2, as shown in fig.17, the first sprinkler C.1 Effect of Rack Storage Height on Fire Suppression
(IRS-14) will actuate at t = 69 sec. The second sprinkler
(IRS-13) will actuate at t = 76 sec. The third sprinkler (IRS-15) For case A4-3, as shown in fig.18, the first ESFR sprinkler
will actuate at t = 85 sec. The sprinklers totally control the fire (CS-1) will actuate at t = 30 sec. The sprinkler totally controls
after 190 sec. The total suppression of the fire will occur after the fire after 90 sec. The total suppression of the fire will
approximately 7.5 min. occur after approximately 2.5 min.

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 220


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

69 sec 76 sec 85 sec 26 sec 75 sec 85 sec

T
190 sec 450 sec 180 sec 210 sec

Fig.17 Total Temperature Contours for Case E3-2 Fig.19 Total Temperature Contours for Case D4-2
ES
30 sec 60 sec 90 sec 22 sec 30 sec 75 sec
A
IJ

120 sec 150 sec 81 sec 180 sec

Fig.18 Total Temperature Contours for Case A4-3 Fig.20 Total Temperature Contours for Case E4-2

For case D4-2, as shown in fig.19, the first sprinkler For case E4-2, as shown in fig.20, the first sprinkler
(CS-1) will actuate at t = 26 sec. The sprinkler totally controls (CS-1) will actuate at t = 22 sec. The sprinkler totally controls
the fire after 85 sec. The total suppression of the fire will the fire after 81 sec. The total suppression of the fire will
occur after approximately 3.5 min. occur after approximately 3 min.

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 221


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

IV. RESULTS AND DISCUSSIONS Flame


vertical
4.1. Results for Fire Simulation velocity (m/s)
14
At the beginning of ignition, the fire simulation showed an
exponentially fire growing with time and the Heat Release
12
Rate (HRR) takes a "t squared" shape profile. For sub-case
A1-1, as shown in fig.4, the intensity and speed at which the
vertical spread accelerates in the vertical flue space, allows the 10

flames to reach the top of storage within 68 sec from ignition.


The fire will continue to grow in intensity, involving new 8

burning surfaces of the rack storage and releasing higher heat


rates, until reaching the flashover at approximately 8 min from 6

ignition. Results of the simulations are shown in fig.21, 22 and


23, where the excess in air temperature, flame vertical velocity 4
and heat release rate (HRR) are plotted against time. As shown
in fig.21, the excess in air temperature increases with time, for 2
all cases. Based on the temperature profiles presented in
section 4, the width of the thermal plume can be determined. 0
The results are plotted in fig.24, where the thermal plume 0 200 400 600 800 1000

T
width, bT, is plotted against time. If no suppression happens, t (sec)

the fire will spread to the nearby stock and additional stock is Fig. 22 Variation of flame vertical velocity with Time for Case Study A1-1
consumed and the fire is getting out of control and the rack
storage collapses, spreading the fire over large areas of the Q (KW)
3,000
warehouse. The fire will be on his way to develop to
encompass the whole warehouse building. At this stage, the
ES
temperature and HRR of the building reach their peaks, T max
and HRRmax, at approximately 13.5 min from ignition and the
fire has a strongest ability to spread outside the warehouse
building. Then the fire exhibits an approximately fully-
2,500

2,000

developed behavior, as shown in fig.21 and fig.22. Once the


fire duration reaches the fire proof limit of the structural 1,500
materials, the building is able to collapse. After collapse, the
rack storage is totally consumed and with the decline of fire
1,000
intensity, the ability of fire out-spreading gradually declines.

500
A
o
T ( K)
1,600 0
0 200 400 600 800 1000
1,400
t (sec)

1,200 Fig. 23 Variation of Heat Release Rate (HRR) with Time for Case Study A1-1

1,000 4.1.2 Effect of Rack Storage Geometry


IJ

800 Fig.25 shows the temperature distributions along the


vertical axis of the flame for sub-cases A1-1, A2-1, A3-1,
600 B1-1, B2-1, B3-1, C1-1, C2-1 and C3-1 respectively. As the
vertical flue width varies, the air entrainment into the rack
400 behaves differently and consequently, the flame pattern is
greatly influenced.
200
As the vertical flue width increases, the air entrainment inside
0
the flame increases, leading to more rapid fire development.
0 200 400 600 800 1000 On the other hand, as the vertical flue width decreases, there
t (sec) will be less entrained air inside the flame, leading to higher
flame heights.
Fig. 21 Variation of Gas Temperature with Time for Case Study A1-1

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 222


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

b T (m)
0.50 4.2 Results for Fire Suppression Simulation
0.45 4.2.1 In-rack Sprinklers vs. ESFR Sprinklers
0.40
The results obtained from the CFD simulations indicated
0.35
that, for case A1-2, two sprinklers have been actuated, the first
0.30 sprinkler (IRS-11) actuated at 55 sec, and the second sprinkler
0.25
(IRS-12) actuated at 60 sec. For case A2-2, three sprinklers
have been actuated, the first sprinkler (IRS-12) actuated at 75
0.20
sec, the second sprinkler (IRS-13) actuated at 80 sec and the
0.15 third in-rack sprinkler (IRS-14) actuated at 83 sec. For case
0.10
A3-2, three sprinklers have been actuated, the first sprinkler
(IRS-14) actuated at 59 sec, the second sprinkler (IRS-13)
0.05
actuated at 62 sec and the third sprinkler (IRS-15) actuated at
0.00 64 sec.
0 200 400 600 800 1000
The results obtained from the CFD simulations indicated
that, for case A1-3, the ESFR sprinkler (CS-1) actuated at 130
t (sec)

Fig.24 Variation of Thermal Plume Width with Time for Case Study A3-1 sec. For case A2-3, the first ESFR sprinkler (CS-1) will
actuate at 150 sec. The second ESFR sprinkler (CS-2) actuated

T
Figure 26, also, shows the temperature distributions along at 156 sec. For case A3-3, the first ESFR sprinkler (CS-1) will
the vertical axis of the flame for sub-cases A1-1, A2-1, A3-1, actuate at 140 sec. The second ESFR sprinkler (CS-2) actuated
D1-1, D2-1, D3-1, E1-1, E2-1 and E3-1 respectively. As the at 142 sec. For case A4-3, the ESFR sprinkler (CS-1) actuated
horizontal flue width increases, more air is entraining inside at 30 sec. The sprinkler totally controls the fire after 90 sec.
the flame increases, leading to more unsymmetrical flame and The total suppression of the fire will occur after approximately
rapid fire development in the horizontal direction and 2.5 min.
consequently, incorporating of horizontal flues can have a
ES
reverse effect on the stability and the symmetry of the flames.
As the vertical and horizontal flue become larger and
larger, the flame height will become more similar to open fire
When comparing the simulation results of the ESFR
sprinklers to those obtained for the in-rack sprinklers, it can be
noticed that, the activation time for the in-rack sprinklers is
much less than for the ESFR sprinklers. The in-rack sprinklers
plumes, which can lead to even more rapid fire growth. This are somehow near to the flame tips and consequently are being
being the most common known cause of fire outbreak for the faster thermally influenced.
storage of bulk materials.

Temperature Sprinkler
Location Index
(oC)
(
1,400 1.20
A
A1-1
Case A1-2, In-Rack Sprinkler, v = 20 cm, h = 32 cm
A2-1 Case A2-2, In-Rack Sprinkler, v = 20 cm, h = 32 cm
A3-1 Case A3-2, In-Rack Sprinkler, v = 20 cm, h = 32 cm
Case A1-3, ESFR Sprinkler, v = 20 cm, h = 32 cm
1,200
B1-1
IRS-15 Case A2-3, ESFR Sprinkler, v = 20 cm, h = 32 cm
1.00
B2-1 Case A3-3, ESFR Sprinkler, v = 20 cm, h = 32 cm
B3-1 Foam Sprinkler, v = 20 cm, h = 32 cm
C1-1
IRS … In-Rack Sprinkler
C2-1 CS …. Ceiling Sprinkler
1,000 C3-1

0.80 IRS-14 IRS-14

800
IJ

0.60 IRS-13 IRS-13

600

0.40
400 CS-2
CS-2

CS-2

IRS-12 IRS-12

200 0.20
CS-1 CS-1 CS-1

CS-1
IRS-11
0
0.00
0 5 10 15
0 30 60 90 120 150 180
Vertical Position (m)
Activation Time (sec)

Fig. 25 Comparison of the Gas Temperature along the vertical Fig.26 Activation Time for In-Rack and ESFR Sprinklers for Case studies A
Centerline of flame for different vertical Flue width

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 223


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

Sprinkler 4.2.2 Effect of Vertical Flue Space


Location Index
( The results obtained from the CFD simulations indicated
0.60 that, for case B1-2, two sprinklers have been actuated, the first
Cases A, In-Rack Sprinkler, v = 20 cm, h = 32 cm Cases A, In-Rack Sprinkler, v = 20 cm, h = 32 cm in-rack sprinkler (IRS-1) actuated at 71 sec. The second in-rack
Cases A, In-Rack Sprinkler, v = 20 cm, h = 32 cm Case A, ESFR Sprinkler, v = 20 cm, h = 32 cm sprinkler (IRS-2) actuated at 76 sec. The two operating
Case A, ESFR Sprinkler, v = 20 cm, h = 32 cm Case A, ESFR Sprinkler, v = 20 cm, h = 32 cm sprinklers fight the fire growth and succeed to control the fire
Case A1-4, Foam Sprinkler, v = 20 cm, h = 32 cm
growth after 165 sec. The total suppression of the fire will
occur after approximately 6 min For case B2-2, three sprinklers
0.40 have been actuated, the first in-rack sprinkler (IRS-12) actuated
at 78 sec. The second in-rack sprinkler (IRS-13) actuated at 84
sec. The third in-rack sprinkler (IRS-14) actuated at 89 sec. The
sprinklers totally control the fire after 170 sec. Total
suppression of the fire will occur after approximately 4.5 min.
For case B3-2, three sprinklers have been actuated, the first
0.20 sprinkler (IRS-14) actuated at 68 sec. The second sprinkler
(IRS-15) actuated at 72 sec. The third sprinkler (IRS-13)
actuated at 78 sec. The sprinklers totally control the fire after
170 sec. The total suppression of the fire will occur after
approximately 4.5 min.

T
For case C1-2, two sprinklers have been actuated, the first
0.00 in-rack sprinkler (IRS-1) actuated at 75 sec. The second in-
0 100 200 300 400 500 rack sprinkler (IRS-2) actuated at 81 sec. The sprinklers
totally control the fire after 165 sec. Total suppression of the
Time (sec)
fire will occur after approximately 6.5 min. For case C2-2,
three sprinklers have been actuated, the first in-rack sprinkler
Fig.27 Fire Control Growth For In-Rack And ESFR

Regarding the control of the fire growth and the fire


extinguishment, the in-rack sprinklers control the fire growth
ES (IRS-12) actuate at 75 sec. The second in-rack sprinkler
(IRS-13) actuated at 80 sec. The third in-rack sprinkler
(IRS-14) actuated at 86 sec. The sprinklers totally control the
fire after 155 sec. Total suppression of the fire will occur after
in someway faster than the ESFR sprinklers. This can be due approximately 6 min.
to the successive operation of the adjacent nearby in-rack For case C3-2, three sprinklers have been actuated, the
sprinklers which has a remarkable effect on the fire growth first sprinkler (IRS-14) actuate at 72 sec.
control.
Sprinkler Sprinkler
Location Index Location Index
( (
0.60
A
0.60 Cases A, In-Rack Sprinkler, v = 20 cm, h = 32 cm
Cases A, In-Rack Sprinkler, v = 20 cm, h = 32 cm Cases A, In-Rack Sprinkler, v = 20 cm, h = 32 cm Case A, ESFR Sprinkler, v = 20 cm, h = 32 cm
Cases A, In-Rack Sprinkler, v = 20 cm, h = 32 cm Case A, ESFR Sprinkler, v = 20 cm, h = 32 cm Case A1-4, Foam Sprinkler, v = 20 cm, h = 32 cm
Case A, ESFR Sprinkler, v = 20 cm, h = 32 cm Case A, ESFR Sprinkler, v = 20 cm, h = 32 cm Cases B, In-Rack Sprinkler, v = 30 cm, h = 32 cm
Case A1-4, Foam Sprinkler, v = 20 cm, h = 32 cm Cases C, In-Rack Sprinkler, v = 40 cm, h = 32 cm
Cases D, In-Rack Sprinkler, v = 20 cm, h = 48 cm
Cases E, In-Rack Sprinkler, v = 20 cm, h = 64 cm

0.40 0.40
IJ

0.20
0.20

0.00
0.00
0 40 80 120 160 200 240
0 5 10 15
Time (sec)
Time (min)

Fig.28 Fire Suppression for In-Rack and ESFR Sprinklers for Case studies A Fig.29 Sprinklers Activation for Different Rack Storage Geometries

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 224


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

Sprinkler Sprinkler
Location Index Location Index
( (
1.20 0.60
Case A1-2, In-Rack Sprinkler, v = 20 cm, h = 32 cm Cases A, In-Rack Sprinkler, v = 20 cm, h = 32 cm
Case A, ESFR Sprinkler, v = 20 cm, h = 32 cm
Case A2-2, In-Rack Sprinkler, v = 20 cm, h = 32 cm
Case A1-4, Foam Sprinkler, v = 20 cm, h = 32 cm
Case A3-2, In-Rack Sprinkler, v = 20 cm, h = 32 cm
Cases B, In-Rack Sprinkler, v = 30 cm, h = 32 cm
Case B1-2, In-Rack Sprinkler, v = 30 cm, h = 32 cm Cases C, In-Rack Sprinkler, v = 40 cm, h = 32 cm
1.00 Case B2-2, In-Rack Sprinkler, v = 30 cm, h = 32 cm Cases D, In-Rack Sprinkler, v = 20 cm, h = 48 cm
Case B3-2, In-Rack Sprinkler, v = 30 cm, h = 32 cm Cases E, In-Rack Sprinkler, v = 20 cm, h = 64 cm
Case C1-2, In-Rack Sprinkler, v = 40 cm, h = 32 cm
Case C2-2, In-Rack Sprinkler, v = 40 cm, h = 32 cm
Case C3-2, In-Rack Sprinkler, v = 40 cm, h = 32 cm
0.80 Case D1-2, In-Rack Sprinkler, v = 20 cm, h = 48 cm
0.40
Case D2-2, In-Rack Sprinkler, v = 20 cm, h = 48 cm
Case D3-2, In-Rack Sprinkler, v = 20 cm, h = 48 cm
Case E1-2, In-Rack Sprinkler, v = 20 cm, h = 64 cm
Case E2-2, In-Rack Sprinkler, v = 20 cm, h = 64 cm
0.60 Case E3-2, In-Rack Sprinkler, v = 20 cm, h = 64 cm

0.40
0.20

0.20

0.00
0.00

T
0 30 60 90 120 150 180
0 5 10 15
Activation Time (sec)
Time (sec)

Fig.31 Fire Suppression for Different Rack Storage Geometries


Fig.30 Fire Growth Control for Different Rack Storage Geometries

The second sprinkler (IRS-13) actuated at 79 sec. The third Total suppression of the fire will occur after approximately 6
sprinkler (IRS-15) actuated at 83 sec. The sprinklers totally min. For case D3-2, three sprinklers have been actuated, the
control the fire after 170 sec. Total suppression of the fire will
occur after approximately 5 min
When analyzing the above results, it can be noticed the
ES first sprinkler (IRS-14) actuated at 68 sec. The second sprinkler
(IRS-13) actuated at 74 sec. The third sprinkler (IRS-15)
actuated at 82 sec. The sprinklers totally control the fire after
180 sec. Total suppression of the fire will occur after
remarkable effect of the vertical flue width on the activation
time of sprinklers. As the vertical flue width increases, the approximately 6.5 min.For case E1-2, two sprinklers have been
vertical flame spread is slowed down, which make some actuated, the first in-rack sprinkler (IRS-11) actuated at 66 sec.
retardation on the sprinklers activation. Regarding the control The second in-rack sprinkler (IRS-12) actuated at 77 sec. The
of the fire growth and the fire extinguishment, the effect of the sprinklers totally control the fire after 180 sec. Total
vertical flue width is very remarkable on the fire growth control suppression of the fire will occur after approximately 6 min.
and fire extinguishment. As the vertical flue width increases, For case E2-2, three sprinklers have been actuated, the first in-
more air is entrained inside the flame, leading to more rapid fire rack sprinkler (IRS-12) actuated at 82 sec. The second in-rack
A
development and consequently, imposing more difficulty for sprinkler (IRS-13) actuated at 90 sec. The third in-rack
the sprinklers to fight the fire growth. sprinkler (IRS-11) actuated at 98 sec. The sprinklers totally
control the fire after 185 sec. Total suppression of the fire will
4.2.3 Effect of Horizontal Flue Space occur after approximately 6 min. For case E3-2, three
sprinklers have been actuated, the first sprinkler (IRS-14)
The results obtained from the CFD simulations indicated actuated at 69 sec. The second sprinkler (IRS-13) actuated at 76
that, for case D1-2, two sprinklers have been actuated, the first sec. The third sprinkler (IRS-15) actuated at 85 sec. The
IJ

in-rack sprinkler (IRS-11) actuated at 65 sec. The second in- sprinklers totally control the fire after 190 sec. Total
rack sprinkler (IRS-12) actuated at 74 sec. The two operating suppression of the fire will occur after approximately 7.5 min.
sprinklers fight the fire growth and succeed to control the fire When analyzing the above results, it can be noticed the
growth after 185 sec. Total suppression of the fire will occur small effect of the horizontal flue height on the activation time
after approximately 6.5 min. For case D2-2, three sprinklers of sprinklers. As the horizontal flue height increases, some
have been actuated, the first in-rack sprinkler (IRS-12) actuated flame are spread in the horizontal flue space to other adjacent
at 82 sec. The second in-rack sprinkler (IRS-11) actuated at 89 surfaces, which impedes the vertical flame spread in the
sec. The third in-rack sprinkler (IRS-13) actuated at 95 sec. The vertical flue space, yielding to somehow slower activation.
sprinklers totally control the fire after 170 sec.

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 225


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

Sprinkler of the foam-water sprinkler system is slightly more than close


Location Index to the ESFR sprinklers activation.
( Regarding the control of the fire growth and comparing
0.60 these results with the results obtained for the cases A1-2 and
Sprinkler Activation, Case A4-3, ESFR Sprinkler, v = 20 cm, h = 32 cm
A1-3, it is noticed that the foam-water sprinkler system
controls the fire growth in such time slightly more than the in-
Sprinkler Activation, Case D4-2, ESFR Sprinkler, v = 20 cm, h = 48 cm

Sprinkler Activation, Case E4-2, ESFR Sprinkler, v = 20 cm, h = 64 cm

Fire Growth Control, Case A4-3, ESFR Sprinkler, v = 20 cm, h = 32 cm


rack sprinklers and reasonably more than the ESFR sprinklers.
Fire Growth Control, Case D4-2, ESFR Sprinkler, v = 20 cm, h = 48 cm Also, regarding the fire extinguishment and comparing
Fire Growth Control, Case E4-2, ESFR Sprinkler, v = 20 cm, h = 64 cm these results with the results obtained for the cases A1-2 and
0.40 Fire Suppression, A4-3, ESFR Sprinkler, v = 20 cm, h = 32 cm A1-3, it is noticed that the foam-water sprinkler system
Fire Suppression, Case D4-2, ESFR Sprinkler, v = 20 cm, h = 48 cm
extinguishes the fire in such time slightly less than the in-rack
Fire Suppression, Case E4-2, ESFR Sprinkler, v = 20 cm, h = 64 cm
sprinklers and reasonably less than the ESFR sprinklers.

4.3 Calculated Activation Time

0.20
The activation time of sprinklers can be calculated using
the equation 17, knowing the air temperature and velocity
around each sprinkler:
RTI  T  T 
o  ln    tc

T
 
u  T  Tr 
0.00 Fig.33, exhibits the calculated values of sprinkler activation
0 50 100 150 200 times compared to the simulated ones, for case studies A.
Time (sec)
V. CONCLUSIONS
Different Rack Storage Heights

4.2.4 Effect of Storage Height


ES
Fig.32 Sprinklers Activation, Fire Growth Control and Fire Suppression for
5.1 Conclusions for Fire Simulation
The fire simulation showed an exponentially fire growing
with time and the Heat Release Rate (HRR) takes a "t
squared" shape profile. The intensity and speed at which the
The results obtained from the CFD simulations indicated
vertical spread accelerates in the vertical flue space, allows a
that, for case D4-2, the sprinkler (CS-1) has been actuated at
fast reaching of the flames to the top of storage. The fire will
26 sec. The operating sprinkler fights the fire growth and
continue to grow in intensity, involving new burning surfaces
succeeds to control the fire growth after 26 sec. Total
of the rack storage and releasing higher heat rates, until
suppression of the fire will occur after approximately 3.5 min.
reaching the flashover phase. The fire will spread to the
For case E4-2, the sprinkler (CS-1) has been actuated at 22
nearby stock and additional stock is consumed and the fire is
sec. The sprinkler totally controls the fire after 81 sec. Total
getting out of control and the rack storage collapses, spreading
A
suppression of the fire will occur after approximately 3 min.
the fire over large areas of the warehouse. The fire will be on
Comparing these results with the results obtained for case A4-
his way to develop to encompass the whole warehouse
3, it is clear that the storage height has a strong impact on the
building. When considering the rack storage geometry, the
sprinkler activation. As the rack storage increases, the
narrow vertical flue width entrains less air inside the flame,
activation will be faster. Regarding the control of the fire
leading to higher flame heights. The wider vertical flue width
growth and the fire extinguishment and comparing these
entrains more air inside the flame, leading to more rapid fire
results with the results obtained for case A4-3, the
development. In the same way, the longer horizontal flue
IJ

improvement of the fire growth control and the fire


height will have a reverse effect on the stability and the
extinguishment due to the increase of the rack storage, can be
symmetry of the flames. As the vertical and horizontal flue
noticed.
become larger and larger, the flame height will become more
4.2.5 Foam-Water Sprinkler System similar to open fire plumes, which can lead to a very rapid fire
growth.
The results obtained from the CFD simulations indicated
that, for case A1-4, two sprinklers have been actuated, the first 5.2 Conclusions for Fire Suppression Simulation
sprinkler actuated at 135 sec. The second sprinkler actuated at
140 sec. The two operating sprinklers fight the fire growth and 5.2.1 Conclusions for Sprinkler Activation Time
succeed to control the fire growth after 255 sec. Total Although some trends have discussed the economical use
suppression of the fire will occur after approximately 5 min. of the in-rack sprinklers in such fire suppression and the use of
Comparing these results with the results obtained for the ESFR ceiling mounted sprinklers in warehouses in place of in-
cases A1-2 and A1-3, in which in-rack sprinklers and ESFR rack fire sprinkler systems, the simulation results showed that
sprinklers are used respectively, it is noticed that the activation

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 226


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

Sprinkler job to control the fire. Regarding the control of the fire growth
Location Index for the foam-water sprinkler system, it is noticed that these
( systems control the fire growth in such time slightly more than
1.20 Case A1-2, In-Rack Sprinkler, v = 20 cm, h = 32 cm
the in-rack sprinklers and considerably more than the ESFR
Case A2-2, In-Rack Sprinkler, v = 20 cm, h = 32 cm
sprinklers.
Case A3-2, In-Rack Sprinkler, v = 20 cm, h = 32 cm
1.00 5.2.3 Conclusions for Fire Suppression
Case A1-3, ESFR Sprinkler, v = 20 cm, h = 32 cm

Case A2-3, ESFR Sprinkler, v = 20 cm, h = 32 cm As the same way discussed above, the in-rack sprinklers
Case A3-3, ESFR Sprinkler, v = 20 cm, h = 32 cm have extinguished the fire faster than ESFR sprinklers, due to
0.80 Calculated Sprinkler Activation,Case A1-2, In-Rack the fast control of fire growth, due to the successive actuation
Sprinkler, v = 20 cm, h = 32 cm
Calculated Sprinkler Activation, Case A2-2, In-Rack of the adjacent nearby in-rack sprinklers. Also, the narrow
Sprinkler, v = 20 cm, h = 32 cm
Calculated Sprinkler Activation, Case A3-2, In-Rack vertical and horizontal flues have a great effect on the fire
Sprinkler, v = 20 cm, h = 32 cm
0.60 Calculated Sprinkler Activation, Case A1-3, ESFR suppression. The foam-water sprinkler system has the fastest
Sprinkler, v = 20 cm, h = 32 cm
Calculated Sprinkler Activation, Case A2-3, ESFR suppression, compared to other cases, due to the great effect of
Sprinkler, v = 20 cm, h = 32 cm
Calculated Sprinkler Activation, Case A3-3, ESFR the foam solution on the fire spread. Also, the foam-water
Sprinkler, v = 20 cm, h = 32 cm
0.40 sprinkler system does not destroy product, due to the lower
water content. They have limited smoke damage, and because
of the detergent properties of the foaming agent, they provide
0.20 a self-cleaning effect.

T
5.3 Best Results Obtained:
0.00
The best sprinkler activation was dedicated to the in-rack
0 30 60 90 120 150 180 sprinklers. The best suppression period was dedicated to the
Activation Time (sec) foam-water sprinkler system. To get a better suppression
Fig. 33 Calculated Values of Sprinkler Activation Times Compared to
Simulated Ones, for Case Studies A

the in-rack sprinkler will actuate faster than the ESFR ceiling
ES performance for high bay warehouses fires, in-rack sprinklers
can be used along with foam-water sprinkler system. But this
configuration has a remarkable impact on the economic-wise
criteria. So, in order to have a reasonable optimal
mounted sprinklers. When the flames start somewhere in the configuration, in-rack sprinklers can be installed along with
rack storage, the in-rack sprinklers are somehow near to the ESFR ceiling sprinklers.
flame tips and consequently are being faster thermally
influenced and actuated. The earlier actuation of the in-rack 5.4 Recommendations for Future Work
sprinklers may have a great effect on the fire growth control. Many research points seem to be essential as an extension
Although the expensive installation of in-rack sprinklers and to the present work for the rack storage fire suppression. The
their problems arising from the operational problems, which study of the effect of using fire-resistant materials, or fire-
might prevent layout improvements from being made, in-rack
A
proof materials on the rack storage fire spread is essential.
sprinklers still might be recommended for fire fighting of the Also, a fully developed rack storage model (engineering
extra-hazards occupancies and class IV commodities rack models and/or CFD models) dedicated to predict the
storage. In the present study, the in-rack sprinklers orientation competition between the downward momentum of the water
was according to NFPA13, where only two rows of in-rack spray and the upward momentum of the fire plume may be of
sprinklers are placed in the rack storage as described in a great importance. Also, the use of Glycerin as an antifreeze
chapter 3, so as to eliminate the installation cost and the for weatherproofing residential and commercial fire sprinkler
IJ

operational problems as much as possible. Regarding the rack systems can be studied by CFD models. The Glycerin, have
storage geometry, the narrow vertical flue width has a very many advantages due to its low toxicity and its low ability to
remarkable effect on the activation time of sprinklers. corrode the plastic pipes and fittings.
Meanwhile, the narrow horizontal flue height can improve the
sprinklers activation. The storage height has a strong impact REFERENCES
on the sprinkler activation. [1] Beard Alan N., "Flashover and Boundary Properties", Fire Safety
Journal, Vol. 45 (2010), pp 116–121.
5.2.2 Conclusions for the Control of Fire Growth [2] Beard Alan N, "Dependence of Flashover on Assumed Value of the
Discharge Coefficient", Fire Safety Journal, Vol. 36 (2001), pp 25-
By observing the above results, it is found that many 36.
[3] Chen Tao, Hongyong Yuan, Guofeng Su, Weicheng Fan, "An
adjacent nearby in-rack sprinklers have been actuated in a
Automatic Fire Searching and Suppression System for Large
successive manner, which allowing more control on the fire Spaces", Fire Safety Journal, Vol. 39 (2004), pp 297–307.
growth. Furthermore, the narrow vertical and horizontal flues [4] Cooper Leonard Y., "The Interaction of an Isolated Sprinkler Spray
have a great effect on the fire growth, as they do not allow the and a Two-layer Compartment Fire Environment Phenomena and
Model Simulations", Fire Safety Journal, 15 (1995), pp 89-107.
fire spread to adjacent surfaces, which facilitate the sprinklers

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 227


Essam Eldine Mouguib et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES
Vol No. 5, Issue No. 2, 212 - 228

[5] Delichatsiosa M.A., Liub X., Brescianinib C., "Propagation of [33] Persson B., Lonnermark A., Persson H., Mulligan D., Lancia A. and
Axisymmetric Ceiling Jet Front Produced by Power Law Time Demichela M., "Large-Scale Foam Application—Modelling of Foam
Growing Fires", Fire Safety Journal, Vol. 38 (2003), pp 535–551. Spread And Extinguishment", Swedish National Testing and
[6] Foley M., Drysdale D. D., "Heat Transfer From Flames Between Research Institute, 2001, ISBN 91-7848-856-7
Vertical Parallel Walls", Fire Safety Journal, Vol. 24 (1995), pp 53- [34] Poreh Michael, Garrad Gordon, "A Study Of Wall And Corner Fire
73. Plumes", Fire Safety Journal, Vol. 34 (2000), pp 81-98.
[7] Grosskopfa K.R., Kalbererb Jennifer, "Potential Impacts of Ultra- [35] Rho J.S., Ryou H.S., "A Numerical Study of Atrium Fires Using
High-Pressure (UHP) Technology on NFPA Standard 403", Fire Deterministic Models", Fire Safety Journal, Vol. 33 (1999), 213-229.
Safety Journal, Vol. 43(2008), pp 308–315. [36] Robert G. Zalosh, "Industrial Fire Protection Engineering", Wiley,
[8] Hamins Anthony, Johnsson Erik, Donnelly Michelle, Maranghides New York, 2003, pp. 387, ISBN 0-471-49677-4
Alexander, "Energy Balance in a Large Compartment Fire", Fire [37] Rufino Paolo, diMarzo Marino, "The Simulation of Fire Sprinklers
Safety Journal, Vol. 43 (2008), pp 180–188. Thermal Response in Presence of Water Droplets", Fire Safety
[9] Hansell G.O., "Heat And Mass Transfer Processes Affecting Smoke Journal, Vol. 39 (2004), pp 721–736.
Control In Atrium Buildings", Ph.D. thesis. South Bank University, [38] Rusch D., Blum L., Moser A., Roesgen T., "Turbulence Model
London, 1993. Validation for Fire Simulation by CFD and Experimental
[10] Hasofer Abraham M. & Beck Vaughan R., "A Stochastic Model for Investigation of a Hot Jet in Crossflow", Fire Safety Journal, Vol. 43
Compartment Fires", Fire Safety Journal, Vol. 28 (1997), pp 207- (2008), pp 429–441.
225. [39] Schwille John A., Lueptow Richard M., "The Reaction of a Fire
[11] Hauptmanns U., Marx M. , Grunbeck S., "Availability analysis for a Plume to a Droplet Spray", Fire Safety Journal, Vol. 41 (2006), pp
fixed wet sprinkler system", Fire Safety Journal 43 (2008) 468– 476 390–398.
[12] Heskestad Gunnar, "Scaling The Interaction of Water Sprays and [40] Seattle Fire Department, "High-Piled Combustible Storage", October
Flames", Fire Safety Journal, Vol. 37 (2002), pp 535–548. 2007, Fire Dept., City of Seattle
[13] Holborna P.G., Nolana P.F., Goltb J., "An Analysis of fire Sizes, fire [41] Sherman C.P. Cheunga, Richard K.K. Yuena, G.H. Yeohb, S.M.,
Growth Rates And Times Between Events Using Data From fire "Low Sensitivity Study on Three Different SN Order Schemes of the

T
Investigations", Fire Safety Journal, Vol. 39 (2004), pp 481–524. Discrete Ordinates Method for Two-Compartment Enclosure Fire",
[14] Ingason Hauker & de Ris John, "Flame Heat Transfer in Storage Fire Safety Journal, Vol. 40 (2005), pp 736–744.
Geometries", Fire Safety Journal, Vol. 31 (1998), pp 39-60. [42] Sjolin Vilhelm, Evans David D., and Jason Nora H., "First
[17] Ingason H., "Modeling of a Two-Dimensional Rack Storage Fire", International Conference On Fire Suppression Research", United
Fire Safety Journal, Vol. 30 (1998), pp 47-69. States Department of Commerce, National Institute of Standards and
[18] Ingason H., "Plume Flow in High Rack Storages", Fire Safety Technology, May 2004.
Journal, Vol. 36 (2001), pp 437-457. [43] Soonil Nam, "Actuation of Sprinklers at High Ceiling Clearance
[19] Ingason H., "Investigation of Thermal Response of Glass Bulb Facilities", Fire Safety Journal, Vol. 39 (2004), pp 619–642.

[20]
30 (1998), pp 71-93.
ES
Sprinklers Using Plunge and Ramp Tests", Fire Safety Journal, Vol.

Jayaweera T.M., Yu H.-Z., "Water Absorption in Horizontal


Corrugated Boards under Water Sprays", Fire Safety Journal, Vol. 41
(2006), pp 335–342.
[44]

[45]
Soonil Nam, "Development of a Computational Model Simulating
the Interaction Between a Fire Plume and a Sprinkler Spray", Fire
Safety Journal, Vol. 26 (1996), pp 1-33.
Underwriters Laboratories Inc., "Report of In Rack Sprinkler
Comparison Fire Tests", Conducted for HSB industrial risk insures,
[21] Kellera A., Loepfeb M., Nebikerb P., Pleischb R., Burtschera NC1838 97NK31099, May 1998
H.,"On-Line Determination of The Optical Properties of Particles [46] United States Fire Administration, "Sprinklers Control Arson Fires in
Produced by Test Fires", Fire Safety Journal, Vol. 41 (2006), pp Rack-Storage Warehouse Mt. Prospect, Illinois", Technical Report
266–273. Series, Federal Emergency Management Agency, United States Fire
[22] Lattimera Brian Y., Christopher P. Hanauskaa, Joseph L. Scheffeya, Administration, National Fire Data Center, Nov 2001.
Frederick W. Williams, "The Use of Small-Scale Test Data to [47] Vytenis Babrauskasa, Richard D. Peacockb, Paul A. Renekeb,
Characterize Some Aspects of Fire Fighting Foam for Suppression "Defining Flashover for fire hazard calculations: Part II", Fire Safety
Modeling", Fire Safety Journal, Vol. 38 (2003), pp 117–146. Journal, Vol. 38 (2003), pp 613–622.
[23] Lattimer Brian Y., Trelles Javier, "Foam Spread Over a Liquid Pool", [48] "Warehouse Sprinkler Design Configurations Not Covered by NFPA
A
Fire Safety Journal, Vol. 42 (2007), pp 249–264. 13", Fire Protection Engneering, Winter 2006.
[24] "Lessons Learned from Understanding Warehouse Fires", Fire [49] White D., Beyler C. L., Fulper C. & Leonard J., "Flame Spread on
Protection Engineering, Winter 2006. Aviation Fuels", Fire Safety Journal, Vol. 215 (1997), pp 1-31.
[25] Magrabi S.A., Dlugogorski B.Z., Jameson G.J., "A Comparative [50] Widmann John F., Jason Duchez, "The Effect of Water Sprays on
Study of Drainage Characteristics in AFFF and FFFP Compressed- Fire Fighter Thermal Imagers", Fire Safety Journal, Vol. 39 (2004),
Air Fire-Fighting Foams", Fire Safety Journal, Vol. 37 (2002), pp pp 217–238.
21–52 [51] Widmann John F., "Phase Doppler Interferometry Measurements in
[26] Magnusson S. E., "A Proposal for a Model Curriculum in Fire Safety Water Sprays Produced by Residential Fire Sprinklers", Fire Safety
IJ

Engineering", Fire Safety Journal Vol. 25 (1995), pp 1-88. Journal, Vol. 36 (2001), pp 545–567.
[27] Marbach Giuseppe, Loepfe Markus, Brupbacher Thomas, "An Image [52] William D. Walton, Nelson Bryner and Nora H. Jason, Editors,
Processing Technique for Fire Detection in Video Images", Fire "Building and Fire Research Laboratory", National Institute of
Safety Journal, Vol. 41 (2006), pp 285–289. Standards and Technology Gaithersburg, MD 20899-8644, July 2000
[28] Miles S. D., Cox G., "Prediction of Fire Hazards Associated with Fire Research Needs Workshop Proceedings Emmitsburg, Maryland,
Chemical Warehouses", Fire Safety Journal, Vol. 27 (1996), pp 265- October 20, 1999.
287.
[29] NFPA code 2002 Edition
[30] Novozhilov V., Harvie D. J. E. & Kent J. H., "A Computational
Fluid Dynamics Study of Wood Fire Extinguishment by Water
Sprinkler", Fire Safety Journal, vol. 29 (1997), pp 259 282.
[31] P&G Fire Protection Standard, "Sprinkler Protection Design for High
Bay Warehouses", No. 618, March 1, 2007, pp 1-9.
[32] Peacock Richard D., Reneke Paul A., Forney C. Lynn & Kostreva
Michael M., "Issues in Evaluation of Complex Fire Models", Fire
Safety Journal, Vol. 30 (1998), pp 103–136.

ISSN: 2230-7818 @ 2011 http://www.ijaest.iserp.org. All rights Reserved. Page 228