Journal of Loss Prevention in The Process Industries: Harshad Shrigondekar, Arindrajit Chowdhury, S.V. Prabhu

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Journal of Loss Prevention in the Process Industries 71 (2021) 104448

Contents lists available at ScienceDirect

Journal of Loss Prevention in the Process Industries


journal homepage: http://www.elsevier.com/locate/jlp

Performance of water mist system with base injection in extinguishing


small container fires
Harshad Shrigondekar, Arindrajit Chowdhury, S.V. Prabhu *
Department of Mechanical Engineering, Indian Institute of Technology Bombay, India

A R T I C L E I N F O A B S T R A C T

Keywords: The purpose of the present work is the experimental investigation on the extinguishment of diesel container fires
Diesel container fires by a water-mist system with base injection through multiple nozzles placed symmetrically at various positions
Water mist system around the rim of the container. Experimental parameters selected are the orifice diameter of the nozzle, the
Base injection
number of nozzles, position of nozzles (angle of orientation, radial and vertical distance of the nozzle from the
Pressure-swirl nozzle
edge of the container), and injection pressure. Nozzles used in the present work are pressure-swirl nozzles with Z-
Angle of orientation
type swirl-insert with solid-cone spray. Desirable properties of the system such as the non-clogging operation of
the nozzle and uniform mass flux density in the spray are tested for the nozzles. The optimized solution is
provided considering water consumption and the physical stability of the fuel surface in the container.
Displacement of oxygen is observed as a predominant extinction mechanism due to the effective application of
spray in the base region of the fire. The present work provides the parametric study of the extinction establishing
the feasibility of the water mist system with a base injection for extinguishing tank fires. This may serve as
guidelines for the preliminary design of the system with the extension for larger tank fires.

Indian Oil Corporation, Jaipur, which caused 11 tanks with various fuels
viz. diesel, petrol, and kerosene to burn continuously for 11 days.
1. Introduction Sharma et al. (2013) evaluated the same accident for vapor cloud ex-
plosions and reported the lessons. Rodante (2005) investigated a
Accidents such as storage tank fires in the process industries can naphtha storage tank fire at a refinery located on the island of Pulau
cause significant economic loss and human suffering if not solved in Merlimau, Singapore in 1988. The damage was restricted to only one
time. Planas-Cuchi et al. (1997) conducted a detailed survey of various dike; however, all three tanks in the dike were lost. It is extremely
fire accidents and concluded that around 59% of 2398 entries were difficult for any of the firefighting techniques available to combat such a
caused due to ignition of liquid fuel. Pool fires and tank fires share giant fire. A mechanism of automatic detection and application of fire
almost equal contributions of around 4% each if 2761 entries were suppressant in the early stage of the fire may be crucial for the successful
categorized by the general type of fire. Chang and Lin (2006) reviewed extinction of such fires. Shuai et al. (2012) conducted a risk-based in-
242 accidents related to storage tanks occurring worldwide over 40 spection for large-scale crude oil tanks. They have reported the corrosion
years, among which 74% occurred in petroleum refineries, oil terminals, state of the tanks, comparison of various standards of inspection.
or storage, where ingredients such as crude oil, gasoline, and oil prod- Tank fires are widely studied in the literature for the characterization
ucts such as fuel oil, diesel, kerosene, and lubricants were involved. The of fire. Koseki and Yumoto (1989) measured characteristics of 2.7 m
maximum number of accidents (52%) occurred in external floating top square dike heptane fire with four open-top tanks of 0.8 m diameter and
tanks. Zheng and Chen (2011) reviewed 50 storage tank fires during 50 also without any tank. They experimentally measured burning rate,
years in China. 50% of the accidents were with crude oil and other oil flame temperature, vertical hot gas velocity, radiation, and gas con-
products such as gasoline and diesel. 64% of the accidents occurred in centration. Guan et al. (2013) experimentally studied the influence of
petrochemical plants, refineries, and oil depots. top opening on the characteristics of oil tank fire such as mass loss rate,
There is a need to ensure the safety of storage tanks with an open or flame shape, puffing frequency, flame height, flame radiation, and tank
closed roof from accidental fires to prevent the loss in the process in- shell temperature. Diesel oil fire with a container of 0.3 m diameter and
dustries. Roy (2011) provided a presentation on the fire that occurred at

* Corresponding author. , Department of Mechanical Engineering, Indian Institute of Technology, Bombay, Powai, Mumbai, 400 076, India.
E-mail addresses: hhs999@gmail.com (H. Shrigondekar), arindra@iitb.ac.in (A. Chowdhury), svprabhu@iitb.ac.in (S.V. Prabhu).

https://doi.org/10.1016/j.jlp.2021.104448
Received 5 June 2020; Received in revised form 26 January 2021; Accepted 22 February 2021
Available online 5 April 2021
0950-4230/© 2021 Elsevier Ltd. All rights reserved.
H. Shrigondekar et al. Journal of Loss Prevention in the Process Industries 71 (2021) 104448

Nomenclature V average discharge velocity through the nozzle (m s−1)

Letter symbols Greek letters


Cd discharge coefficient (dimensionless) α angle of orientation (◦ )
D diameter (m or mm) θ azimuthal angle between the nozzles (◦ )
Dv,a diameter (μm) below which (a × 100) % of the spray’s ρ density of water (kg m−3)
volume resides, 0 ≤ a ≤ 1 μ dynamic viscosity of water (Pa-s)
ΔP injection pressure (bar or Nm−2) Subscript
G type of the thread act actual
H height of the nozzle tip from the edge of the container (cm) o orifice
L length (mm) T total
M mass of water consumed (g) t thread
ṁ mass flow rate (kg s−1) th theoretical
R radial distance of the nozzle tip from the center of the
container (cm) Abbreviation
r radial distance of the nozzle tip from the edge of the BSPT British standard pipe taper
container (cm) ( ) DSLR digital single-lens reflex
Re Reynolds number Re = ρVDo 4ṁact
μ = πμDo (dimensionless) FDS fire dynamics simulator
t extinction time (s)

0.25 m height is considered which is geometrically proportional to the Mawhinney and Eng 1993; Mawhinney et al. 1994; Grant et al. 2000; Liu
oil tank of 3000 steres. Water mist may decrease the risk of the probable and Kim 2000) in detail. Extinction performance of the water mist sys-
‘boil over’ process by reducing the heat feedback to the fuel surface and tem with the top injection of the spray is widely studied in the literature
also due to the insignificant amount of water depositing at the bottom of (Hsieh et al. 2006; Cong and Liao 2009; Cong et al. 2009; Jenft et al.
the container. Drysdale (2011) discussed the estimation of radiant flux 2014; Liang et al. 2015; Liu et al. 2017). Performance of the
received in case of storage tanks in a petrochemical plant from the pressure-swirl nozzles with X-type swirl-insert and full-cone spray from
adjacent fire such that water mist system might be designed to prevent different heights in extinguishing diesel and heptane fires has been
the rise of temperature. studied by Shrigondekar et al. (2018; 2020). The portable water mist
Water possesses desirable properties such as high specific heat and system is also reported to be employed for suppression of Class B fires of
high latent heat of vaporization. It is also readily available and various sizes due to non-suitability of fixed water mist system with top
environment-friendly for the application of fire suppression. Thus, water injection, or to study the effectiveness in different fires or to study the
has emerged as an effective substitute for Halons after the Montreal effect of additives conveniently (Huang et al. 2007; Liu et al. 2007;
protocol (Grant et al. 2000). Water mist systems are widely studied for Xiaomeng et al. 2010). However, experimental investigation with base
the suppression of class B fires (flammable liquids, oils, etc.). Typical injection is reported solely by Ndubizu et al. (2000). They have reported
sprinklers may not be useful due to possible splashing or spillage caused extinction performance in terms of suppressed burning rate instead of
by the droplets with higher momentum (Grant et al. 2000). In a con- extinction time, with the only experimental parameter being selected is
ventional sprinkler system Dv,0.99 (99% volume diameter for the the orientation of the nozzles on 0.5 m JP8 and heptane pool fires.
flow-weighted cumulative volumetric distribution of water droplets) Suppression performance with 45◦ base injection is reported to be more
may be of the order of 5000 mm (Back 1995). As per the standard NFPA efficient than top injection and bottom injection. Similar results are
750 (2019), water mist is defined as a water spray for which the Dv,0.99 is reported by Prasad et al. (1999) with the numerical study of suppression
less than 1000 μm within the nozzle operating pressure range. The of small-scale methanol liquid pool fires. Prasad et al. (1998) reported
theory of various methodologies used to produce water mist is reported the results of the numerical simulation of the interaction between water
in detail (Khavkin et al. 2003; Nasr et al. 2013; Lefebvre and McDonell mist and methane-air diffusion flames stabilized above Wolfhard–Parker
2017). Water mist nozzles usually have a small orifice diameter. Clog- burner geometry. The water–fuel ratio required for extinguishment is
ging of the nozzle may cause the failure of the system. Thus, the inner reported to reduce with decreasing droplet size for base injection.
structure of the nozzle should be such that the nozzle should operate Mahmud (2016) conducted a numerical simulation using a fire dy-
without clogging. Shrigondekar et al. (2018; 2020) have studied in namics simulator (FDS) to predict the effect of the location of fires,
detail the characterization of the spray from various simplex nozzles obstructions, the number of nozzles, and the size of droplets on the
with X-type swirl-insert. Similar work for various nozzles is reported in suppression of fire with pinewood, char, and gypsum as fuel using the
the literature (Tate 1960; Lan et al. 2014). However, the uniform dis- water mist spray. Mishra (2018) studied the influence of volume
tribution of mass flux density is not reported in the literature for water blockage ratio on explosion overpressure generation numerically for a
mist nozzles. Solid-cone simplex nozzles generally possess the peak in model plant of Indian Oil Corporation Limited, Jaipur accident of the
the mass flux density at the center and reduction in it with the radial year 2009.
distance (Jain et al. 2014; Shrigondekar et al. 2018, 2020). Uniform Thus, there is a need for further experimental investigations on the
mass flux density of water in the spray may be desirable over the fuel extinction performance of a water mist system with a base injection
surface to mitigate lingering small flames at the boundary of the spray which may be more efficient than a top injection. This may serve as the
area or the edge of the container in case of top injection. Thus, nozzles motive for the design and development of a water mist system aimed at
with a novel Z-type swirl-insert are studied in the present work for the the extinction of tank fires of various sizes. Present work is aimed at
desirable properties such as non-clogging operation and uniform dis- studying the feasibility of the water mist system with the base injection
tribution of flux density in the spray ensuing from the nozzle which are for the extinguishment of the container fires aiming at small/medium
crucial for the application of fire suppression. tank fires with the emphasis only on the extinction performance.
Required principles for the preliminary design of water mist systems Objectives of the present work are.
for the suppression of class B fires are reported (Rasbash 1986;

2
H. Shrigondekar et al. Journal of Loss Prevention in the Process Industries 71 (2021) 104448

2. Experimental setup and methodology

This section presents the geometry of the nozzles employed, the


methodology for producing the spray through multiple nozzles, mea-
surement of the discharge coefficient of the nozzles, and the distribution
of mass flux density in their sprays at various injection pressures. It also
presents the methodology for the application of a water mist system with
base injection in extinguishing diesel container fires. The range of in-
jection pressure (2–10 bar) in this study is divided into 3 categories for
convenience – small (2–4 bar), intermediate (5–7 bar), and large (8–10
bar) injection pressures.

2.1. Geometric details of the nozzle

Geometric details of the solid-cone simplex nozzle with Z-type swirl


insert are shown in Fig. 1 and Table 1. Nozzles with orifice diameters of
0.85 and 1.2 mm are termed nozzle I and II respectively.

2.2. Production of spray through multiple nozzles

Experimental setup for producing water spray through multiple


Fig. 1. Geometric details of the solid-cone simplex nozzle with Z-type swirl nozzles is shown in Fig. 2. Plain tap water is pressurized by a centrifugal
insert (Lechler manual). pump and supplied through multiple hydraulic lines. A pressure trans-
mitter with an indicator is used to measure the injection pressure.
1. To characterize the pressure-swirl nozzles with Z-type swirl-insert for
extinguishing diesel container fires
2. To provide an efficient solution for a given fire by investigating the
performance of the water mist system with base injection in extin-
guishing diesel container fires of two different sizes with various
experimental variables such as orifice diameter of the nozzle, num-
ber of nozzles, their position (angle of orientation, radial and vertical
distance of the nozzle from the edge of the container), and injection
pressure of the nozzle

Fig. 3. Experimental setup for measurement of actual mass flow rate.

Table 1
Geometric details of the nozzles used in the present work.
Nozzle Geometric parameter

Part number G * All dimensions are in mm.

Total length (LT) Total length (Lt) Diameter (D) Hex Orifice diameter (Do)

I 490.368.1Y.CA 1/8′′ BSPT 18.00 6.50 10.00 11.00 0.85


II 490.408.1Y.CA 1/8′′ BSPT 18.00 6.50 10.00 11.00 1.20

Fig. 2. Experimental setup for producing water spray through multiple nozzles.

3
H. Shrigondekar et al. Journal of Loss Prevention in the Process Industries 71 (2021) 104448

2.4. Distribution of mass flux density in the spray

Mass flux density in the spray at various locations in a plane at an


axial distance of 2 m is measured by using a mechanical patternator. It is
ensured that the nozzle is placed vertically above the center of the
patternator. The patternator is developed such that water is collected in
all 4 quarters of the circular area instead of 1 or 2 quarters (assuming
symmetric spray) as reported in the literature (Cong and Liao 2009).
Thus, this design of the patternator can capture the possible asymmetry
in the spray, if it exists. The schematic of the patternator designed is
shown in Fig. 4.
Tubes and cups are accommodated into a Polyethylene foam sheet
for collecting water. They are placed vertically such that their inlet
planes are horizontal. 55 tubes of an inner diameter of 27 mm and a
volume of 50 ml are used at the central portion of the patternator with a
small pitch of 40 mm. 60 cups of an inner diameter of 51 mm and a
volume of 100 ml are used in the outer region. The pitches are alter-
nately placed 65 mm and 75 mm. Water is sprayed for different dura-
tions (2–15 min) depending on the injection pressure. Tubes and cups
Fig. 4. Details of the mechanical patternator (Dimensions in mm).
are sufficiently deep to avoid splashing of water from inside. Mass flux
density is measured instead of volume flux density to avoid a larger
uncertainty in the measurement due to parallax error. Water collected in
2.3. Discharge coefficient
each tube and cup is weighed by an electronic weighing scale with a
least count of 10 mg. Swaying of the spray due to the breeze of air is
The discharge coefficient of the nozzle is evaluated by measuring the
eliminated by carrying out experiments in a closed space.
actual mass flow rate (ṁact ) and calculating the theoretical mass flow
rate (ṁth ). Experimental setup for the measurement of the actual mass
2.5. Water mist system with the base injection for extinguishing diesel
flow rate is as shown in Fig. 3. The actual mass flow rate of water is
container fires
measured by catching water in a container for a specified time. The
container (30 L) is placed on the weighing scale (make: Contech) of
Experiments are carried out for the extinguishment of the diesel
capacity 30 kg and sensitivity 0.2 g. The time required to collect a mass
container fires by the water mist system with base injection through
of 5–10 kg water (depending upon the injection pressure) is noted by
multiple nozzles. Experimental rig depicting a water mist system with
using a stopwatch with a least count of 0.01 s.
base injection through multiple nozzles (3 in this case) is shown in Fig. 5.
A water mist system is a deluge system in which all nozzles discharge
water simultaneously. Circular mild steel containers with a depth of

Fig. 5. Experimental rig depicting water mist system with base injection by multiple nozzles (3 in this case).

4
H. Shrigondekar et al. Journal of Loss Prevention in the Process Industries 71 (2021) 104448

Fig. 6. Overview of the present work.

0.08 m and the inner diameters of 0.3 and 0.5 m are filled with fuel with
an initial lip height of 10 mm. The lip height is the depth of the fuel
surface from the edge of the container. The level of the fuel in the
container is maintained constant in the literature (Alexander et al. 1982;
Ndubizu et al. 2000) with a self-adjusting liquid level apparatus, espe-
cially to measure the burning rate. However, the level of the fuel in the
container decreases with the progress of the fire in the present work to
simulate the condition of real tank fires. Initial lip height of 10 mm is
selected such that the fuel spillage after the sudden application of mist is
avoided and also the effect of the applied mist on the stability of the fuel
level can be observed. The container is placed over thermal insulation on
the horizontal platform. Fuel is placed over the thickness of water in the
container by Ndubizu et al. (2000) to study the effect of the application
of mist on the burning rate application of mist. However, the present
work is aimed only to study the feasibility of the water mist system with
base injection in extinguishing liquid pool fires with emphasis on the
extinction performance.
Another platform is provided for housing multiple nozzles with the
hole at the center for accommodating the container such that different
nozzles can be applied to the fire symmetrically. There is a provision for
fixing various nozzle holders on the platform to avail multiple nozzles at
various radial distances from the center, azimuthal angles between the
nozzles (θ), and orientation angles (α). The container can be adjusted
vertically to avail different positions of the nozzles with respect to the
container.
Fuel is ignited manually with the help of a blow torch (for diesel, a Fig. 7. Comparison of the mass flow rates from different nozzles and their
small amount of heptane is added initially). Fuel is allowed to burn discharge coefficients.
freely for 4 min (pre-burn time) for both 0.3 and 0.5 m fire such that it
reaches the steady-state of burning rate. Roof ventilators are utilized for the instant of the activation of the spray to the instant at which the last
the exhaust of combusted products. Careful attention is provided to part of the flame disappears. The instant at which the spray is activated
avoid swaying of the flame and the spray because of the movement of air is captured accurately to increase the accuracy of the measurement of
by enclosing the entire system with vertical mesh screens. The spray is the extinction time. This is accomplished with the help of the necessary
activated by utilizing a solenoid valve. The entire extinction phenome- electric connections which turn on a bulb simultaneously as the solenoid
non is studied by recording a video by digital single-lens reflex (DSLR) valve is activated. Each experiment is repeated thrice for ensuring
camera at 60 frames per second. Extracted images from the captured repeatability. Various configurations can be attained by varying the
video are studied extensively to examine the extinction process and the radial distance of the nozzle from the outer edge of the container (r),
mechanisms involved. Extinction time is measured as the duration from height from the top edge of the container (H), and angle of orientation

5
H. Shrigondekar et al. Journal of Loss Prevention in the Process Industries 71 (2021) 104448

Table 2 presents the extinction performance of the water mist system with the
Correlation for mass flow rate through nozzles. base injection for 0.3 m and 0.5 m diesel container fires with various
Nozzle, Do (mm) Equation Number experimental parameters. It is interesting to compare the extinction
performance by both nozzles for various experimental variables
I, 0.85 ṁact = 8.76(ΔP) 0.33 (2)
considering the consumption of water. This work aims to optimize the
II, 1.2 ṁact = 11.38(ΔP)0.46 (3) set of experimental parameters for a given fire, considering the extinc-
Shrigondekar et al. (2018), 1.2 ṁact = 12.51(ΔP)0.38 (4) tion time, physical stability of the fuel surface inside the container, and
water consumption.

(α). Overview of the present work is presented in Fig. 6.


3.1. Discharge coefficient
3. Results and discussion
Comparison of the mass flow rates from the nozzles in the present
work with the nozzle with X-type swirl-insert with an orifice diameter of
This section presents the results of various measured characteristics
1.2 mm studied previously by Shrigondekar et al. (2018) is presented
such as discharge coefficient and mass flux density of the nozzles. It also
and variation in their discharge coefficients with Reynolds number is

Fig. 8. Polar contours of mass flux density in spray for various injection pressures at distance of 2 m from nozzle I.

6
H. Shrigondekar et al. Journal of Loss Prevention in the Process Industries 71 (2021) 104448

Fig. 9. Polar contours of mass flux density in spray for various injection pressures at distance of 2 m from nozzle II.

shown in Fig. 7. coefficient which denotes the influence of the efficient flow area at the
Reynolds number is calculated as given below, nozzle orifice as described by Lan et al. (2014). The discharge coefficient
may be assumed to be constant over the entire range of Reynolds
ρVDo 4ṁact
Re = = (1) number in the case of nozzle II.
μ πμDo

where, Do is the orifice diameter, Vis the average velocity through the 3.2. Patternation of the spray
orifice area obtained from the actual mass flow rate (ṁact ), ρis the
density of water, and μ is the dynamic viscosity of water at 25 ◦ C. Patternation of the spray resulting from both nozzles for various
Correlations developed for mass flow rate as a function of injection injection pressures (ΔP: 2–10 bar) is measured. Polar contours of the
pressure are shown in Table 2. The index in the correlation (power-law) mass flux density of the spray at various injection pressures for nozzle I
provided for the mass flow rate through nozzle II is close to 0.5 (ideal, and II are displayed in Figs. 8 and 9 respectively. The magnitude of mass
non-viscous theory) whereas it is significantly smaller for nozzle I. flux density in the spray is larger for nozzle II (Do = 1.2 mm) compared
Variation in the discharge coefficient is significant for nozzle I compared to nozzle I (Do = 0.85 mm) at a given injection pressure owing to a larger
to nozzle II, especially for larger Reynolds numbers. This may be mass flow rate through nozzle II. It increases monotonously with the
attributed to the reduction in the variable component of the discharge injection pressure for both the nozzles. The distribution of mass flux

7
H. Shrigondekar et al. Journal of Loss Prevention in the Process Industries 71 (2021) 104448

Table 3 work with a smaller variation of mass flux density with the radial
Summary of the parametric study of the extinction performance. distance.
Fire Parameters Parameters Conclusions Corresponding Thus, it may be concluded that the distribution of the mass flux
size Fixed varied link density in the spray at a 2 m plane is more uniform for the nozzles with
(m) Z-type swirl-insert compared to those with X-type swirl-insert.
0.3 Position of Number of 2 nozzles fail to 3.3.1.1
the nozzle nozzles (2 and 3), extinguish, Fig. 10
holder and α: Do (0.85 and 1.2 extinction times 3.3. Extinguishment of diesel container fires
−15◦ hence mm), are small, no
position of ΔP splashing of the
the nozzle (r: (2,3,4,5,6,7,8,9, fuel droplets,
The procedure of evolving an efficient solution for a given fire is
5 cm, H: 11.4 and 10 bar) however, based on altering various experimental parameters one after another,
cm) disturbance of with each parameter guiding the selection of the subsequent set of pa-
the fuel layer rameters. The selection of the parameters with the conclusions is sum-
0.3 Position of α: 0◦ (r: 6.5 cm Extinction time 3.3.1.2
marized in Table 3.
the nozzle and H: 13.5 cm), is slightly larger Fig. 12
holder, α: +15◦ (r: 9 cm for configuration
number of and H: 15.2 cm), with an 3.3.1. Diesel fire of 0.3 m size
nozzles: 3 Do (0.85 and 1.2 orientation
mm), angle of +15◦ ,
3.3.1.1. Effect of number of nozzles and effect of orifice diameter at a given
ΔP however, it is
(2,3,4,5,6,7,8,9, recommended position of the nozzle. The position of the nozzle holder is held constant
and 10 bar) due to negligible for all nozzles such that the radial distance of the nozzle tip from the
disturbance of outer edge of the container is 5 cm and the height of the nozzle tip from
the fuel layer
the top edge of the container is 11.4 cm with the orientation of −15◦ .
0.3 Number of ΔP (2–10 bar in Small splashing 3.3.1.3
nozzles: 3, increments of 1 is observed for Fig. 17
The maximum number of nozzles attempted is three which covers the
Position of bar) both the nozzles entire area of the pan effectively. Two nozzles are also attempted, which,
the nozzle at almost all however, fail to extinguish the 0.3 m fire. Variation in extinction time
holder (closer injection and water consumed by both nozzles with injection pressure with three
to the pressures, closer
nozzles is shown in Fig. 10. Large extinction time at small injection
container at a position of the
small height) nozzle holder is pressures may be considered as failure in extinguishing the fire. How-
and also α: discarded ever, the same may not be termed as a failure if only fire suppression is
+45◦ , hence aimed. Since the aim of the water mist system in the present work is fire
the position extinction and not fire suppression, extinction time above 20 s for at
of the nozzle
(r: 6.75 cm,
least one of the three trials is considered as failure and not shown in the
H: 4 cm) graph.
Nozzle position for efficient solution of 0.3 m fire: Number of nozzles: 3, (r: 9 cm, H: Extinction time is large for small injection pressures for both the
15.2 cm), α: +15◦ , efficient nozzle Do : 0.85 mm, efficient ΔP: 6–8 bar nozzles. Although there is a trend of a decrease in the extinction time at
0.5 Nozzle Number of The frequency of 3.3.2.1 small injection pressures, extinction times are comparable at interme-
position (r: 9 nozzles: 3, 4, and splashing is Fig. 18 diate and large injection pressures. However, extinction time by both the
cm, H: 15.2 5,Do (0.85 and larger for fewer
nozzles is almost constant and equal for injection pressures of 6–8 bar.
cm, α: +15◦ ) 1.2 mm) nozzles, For 5
nozzles, Extinction times by both nozzles are almost comparable over the range
extinction times of injection pressures studied with those achieved by nozzle I being
are comparable slightly larger. Significant variation in the water consumption by both
for both nozzles the nozzles with the comparable extinction times at the constant injec-
at all except
small injection
tion pressures (6, 7, and 8 bar) or comparable water consumption with
pressures the considerably different extinction times at injection pressures of 4, 5,
9, and 10 bar is attributed to the significant difference in the mass flow
For a given position of the nozzle, efficient solution: ΔP: 5–7 bar, Do : 0.85 mm,
rate from both the nozzles at a given injection pressure as depicted in
and number of nozzles: 5.
Fig. 7 (a). There is no splashing of the fuel droplets observed for any of
the experiments. However, physical disturbance of the fuel layer in the
density in the spray is not uniform at a given plane for both nozzles; its
container with the application of the spray is observed in almost all
variation in the radial direction depends on the injection pressure.
cases. One such observation recorded after the flame is extinguished for
However, it can be observed that mass flux density is almost uniform
both nozzles is shown in Fig. 11. Disturbance of the fuel layer may have
within the radial distance of 20 cm for both nozzles which is true at
occurred in the presence of flame also, however, it is not registered due
injection pressures of 2–7 bar and 2–4 bar for nozzle I and II respec-
to the presence of smoke, vapors, and water droplets. Thus, the hori-
tively. Also, the distribution of the mass flux density is not axisymmetric.
zontal position of the nozzles and larger (positive) orientation angles are
Overall, the shape of the spray pattern is elliptical for both the nozzles
attempted with the same position of the nozzle holder.
with the position of a major axis remaining almost constant with an
increase in the injection pressure. Since the azimuthal position of the
3.3.1.2. Effect of orientation angle and orifice diameter for a given position
nozzle is un-changed in all the experiments, this asymmetry may be
of the nozzle holder. Orientation angles of 0◦ (horizontal nozzle) and
related to the distinctive structural design of the inner portion of the
+15◦ are attempted with the position of the nozzle holder being un-
nozzle (Z-type swirl-insert with 2 inlet ports). There are dual peaks of
changed in anticipation of less physical disturbance of the fuel layer in
mass flux density observed at larger injection pressures for nozzle II
the container with the application of the spray. The radial distance of the
whereas single peaks are observed for nozzle I. These nozzles (I and II)
nozzle tip from the outer edge of the container (r) is 6.5 cm and 9 cm for
with Z-type swirl-insert are compared with the nozzles with X-type
orientation angles of 0◦ and +15◦ respectively. However, the height
swirl-insert studied previously by Shrigondekar et al. (2018). It can be
from the top edge of the container (H) is 13.5 cm and 15.2 cm for
observed that the difference between the largest and the smallest mass
orientation angles of 0◦ and +15◦ respectively. Variation in the extinc-
flux density is smaller for the nozzles (I and II) studied in the present
tion time and the water consumed for three nozzles with injection

8
H. Shrigondekar et al. Journal of Loss Prevention in the Process Industries 71 (2021) 104448

Fig. 10. Variation in extinction time and water consumed by both nozzles with injection pressure.

Fig. 11. Disturbance of the fuel layer recorded after flame is extinguished.

pressure for three different orientation angles is shown in Fig. 12. high boiling points) at a positive orientation angle. However, extinction
Extinction time varies with the injection pressure for both the nozzles time for the configuration with an orientation angle of +15◦ is not
and all three configurations in a similar fashion. Extinction time is substantially larger compared to those for the configurations with an
considerably large at small injection pressures, decreasing with an in- orientation angle of −15◦ and 0◦ . This signifies that fuel cooling which is
crease in the injection pressure. However, a further increase in the in- less effectively achieved for positive orientation angle may not be the
jection pressure does not necessarily decrease the extinction time and prominent mechanism of the extinction in the case of base injection.
extinction time oscillates slightly around a constant value. Large Instead, the prominent extinction mechanism for base injection is
extinction times or failure at the smallest injection pressure may be observed to be oxygen displacement. Comparison of the effect of the
attributed to both small mass flow rates and large droplets which lead to orientation angle on the oxygen displacement and consequently the
ineffective flame cooling, fuel cooling, and oxygen displacement. This is extinction time is representatively shown for one injection pressure (6
significant for nozzle I for the configuration with an orientation angle of bar) for nozzle I and II in Fig. 13 and Fig. 14 respectively. Considerable
+15◦ . Extinction times for the configurations with −15◦ and 0◦ are enlargement of the fire is observed after the application of the spray for
significantly closer for a given nozzle at almost all injection pressures. negative orientation angle compared to other angles. Oxygen displace-
This implies that there is no significant advantage in reducing the ment as the prominent extinction mechanism is discussed in detail later
orientation angle and aiming at the fuel surface. In contrast, agitation of on in this article with the observed visuals in section 3.4. Smoke and
the fuel surface is observed for the negative orientation angle as dis- water vapors formed due to the interaction of the spray and the flame
cussed above. Thus, the negative orientation angle is not recommended are observed to travel down and displace the oxygen in the area near the
for the container fire with small lip heights. Extinction time is larger for fuel surface which can be observed in Fig. 13 (m)-(n)–(o) and Fig. 14
configuration with an orientation angle of +15◦ compared to configu- (m)-(n)-(o). Thus, it may be anticipated that a further increase in the
rations with that of −15◦ and 0◦ for the given nozzle and injection orientation angle would increase the extinction time accounting for the
pressure. This may be attributed to the smaller contribution of fuel inefficient displacement of oxygen and travel times of smoke. The
cooling (one of the extinction mechanisms, prominent for fuels with concept of extinction due to a combination of mechanisms such as fuel

You might also like