Anda di halaman 1dari 19

This article was downloaded by: [HEAL-Link Consortium]

On: 9 May 2011


Access details: Access Details: [subscription number 786636650]
Publisher Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-
41 Mortimer Street, London W1T 3JH, UK

Experimental Heat Transfer


Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713770473

A Study on Heat Release Rates of Furniture Under Well-Developed Fire


W. K. Chowa; S. S. Hana
a
Department of Building Services Engineering, The Hong Kong Polytechnic University, Hong Kong,
China

To cite this Article Chow, W. K. and Han, S. S.(2006) 'A Study on Heat Release Rates of Furniture Under Well-Developed
Fire', Experimental Heat Transfer, 19: 3, 209 — 226
To link to this Article: DOI: 10.1080/08916150600616949
URL: http://dx.doi.org/10.1080/08916150600616949

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf


This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
Experimental Heat Transfer, 19:209–226, 2006
Copyright © Taylor & Francis Group, LLC
ISSN: 0891-6152 print/1521-0480 online
DOI: 10.1080/08916150600616949

A STUDY ON HEAT RELEASE RATES OF FURNITURE


UNDER WELL-DEVELOPED FIRE

W. K. Chow and S. S. Han


Department of Building Services Engineering, The Hong Kong Polytechnic
University, Hong Kong, China

With the increasing interest in well-developed fires, the burning behavior of furniture
under flashover should be studied. Full-scale burning tests on the heat release rates of
furniture under flashover with oxygen calorimetry will be reported in this article.
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Nine tests on sofas with and without the treatment of fire retardant, wood desks,
and the fire source itself to start the burning were carried out. The heat release rate,
oxygen concentration, thermal radiative heat fluxes at floor level, and air temperatures at
some selected positions were measured.
Response of furniture under small accidental fires and bigger pool fires to onset
flashover will be discussed. Obviously, burning furniture under a flashover fire will give
a much higher heat release rate. Sofas treated with fire retardant might burn even more
vigorously. This point should be considered carefully in designing fire-safe furniture.

Keywords heat release rate, flashover, furniture, building fire

1. INTRODUCTION
The number of big building fires started from burning furniture in the Far East
appears to be increasing. In Hong Kong, the big fires in the Garley building [1] and
Mei Foo Sun Chuen [2] led to significant revisions of the building fire safety codes.
Consequently, foam sofa furniture has to be treated with fire retardants. Fire safety has
become an even bigger concern after the World Trade Center (WTC) incident, the Korean
train fire, and Hong Kong train fire incidents. Fire safety should not only be limited to
protecting against accidental fires as before, but also arson and terrorist attack fires should
be considered. There is no excuse to have such a big fire during the refurbishment of a
building such as the Garley building fire [1] which was without adequate fire protection.
Although there had been numerous studies on furniture fires in the literature [3],
including the Combustion Behaviour of Upholstered Furniture (CBUF) project [4] in
Europe and others in USA [5–7], not much work focused on local furniture samples.
Design, technology, and materials used in different places would change the fire behavior
of furniture. Very little information is available for heat release rates in burning local
furniture. Such information has to be measured for fire hazard assessment. Wood and
foam furniture commonly used were selected to study their fire behavior.

Received 10 August 2005; accepted 31 October 2005.


Address correspondence to W. K. Chow, Department of Building Services Engineering, The Hong Kong
Polytechnic University, Hong Kong, China. E-mail: bewkchow@polyu.edu.hk

209
210 W. K. CHOW AND S. S. HAN

NOMENCLATURE
EHC effective heat of combustion (MJ/kg) tfp time to first peak value of heat
Q(t) heat release rate (kW) release rate (s)
QCG (t) heat release rate of the combustible tig ignition time from start (s)
with gasoline (kW) NetTHR net total heat released from Q(t)
QG (t) heat release rate of gasoline (kW) for full-scale tests (MJ)
QtB average heat release rate per unit area THR total heat released per unit area for
over the burning time (kW/m2 ) bench-scale tests (MJm−2 )
t time (s) m mass loss (kg)
tB burning time (s) Q(t) net heat release rate (kW)

In response to having so many arson fires in small enclosures such as karaoke


boxes [8–10] and train vehicles [11, 12], it is essential to study the burning behavior
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

under well-developed fire after flashover. Assessing furniture under small accidental fires
of 100 kW to 300 kW as in the gas ignitor for standard test ISO 9705 [13, 14] appears
to be insufficient.
Full-scale burning tests on studying the heat release rates of selected furniture sam-
ples under accidental and flashover fires were carried out. Experiments were conducted
in a room calorimeter [15] with an exhaust hood and a fan-duct system constructed in a
remote area in Northern China.
Nine sets of tests on foam sofas, desks, and a fire source were arranged. The heat
release rate, oxygen concentration, radiative heat fluxes at floor level, and air temperatures
at some selected positions were measured with experimental details reported before [16].
Carbon monoxide (CO) and carbon dioxide (CO2 ) were not measured due to resources
limitation. Analysis on the data and further interpretation of the results will be reported
in this article.
Furniture samples were tested with a cone calorimeter to better understand the
results from full-scale burning tests. The principle of superposition, a technology com-
monly used in fire hazard assessment in the Far East, was applied to analyze the heat
release rate of several combustible components. The room air temperature was compared
with the standard BS 476 [17] temperature/time curve to assess the potential hazard of
burning furnitures in an enclosure.
An accidental fire is started from burning a small object such as a litter bin. The
fire is localized with some combustibles ignited. It would take time to have the whole
room involved in fire. Once a fire is well developed after flashover, furniture would be
exposed to higher external heat fluxes, making it easier to burn. Higher thermal radiation
feedback will give a higher pyrolysis rate and faster chemical reactions. The heat release
rate will be higher to give a more hazardous environment. It is necessary to understand
how furniture is burnt after flashover, not just ignite a small part by a small heat source.

2. FULL-SCALE BURNING TESTS


The room calorimeter [15] is shown in Figure 1 including the locations of thermo-
couples for measuring the air temperature. Nine tests were carried out on three sets of
four-seater sofas and two wood desks as in Figure 2.
FURNITURE UNDER FIRE 211
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Figure 1. The room calorimeter.

The nine testing arrangements are:

• Burning under an accidental fire with 0.5 liter of gasoline in a 0.2 m diameter
pool.
Test T1: Testing sofa sample SF1
Test T2: Testing sofa sample SF2 with less amount of foam than SF1
Test T3: Testing SF1 under a lower exhaust rate by reducing the fan power
Test T4: Testing SF1 treated with fire retardant
Test T5: Testing a wood desk
Test T6: Testing SF1 and a wood desk
• Testing under flashover condition
Test T7: Testing SF2 by setting up a bigger pool of diameter 1 m with 12 liters
of gasoline to onset flashover in the room first.
212 W. K. CHOW AND S. S. HAN
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Figure 2. Furniture samples tested. (a) Sofa. (b) Desk.

As too much dark smoke was generated and flew out of the room, it was difficult
to justify flashover using the criterion of flame coming out from the opening. Therefore,
gas temperature measured near to the ceiling was used as the criterion for determining
flashover. The response time of the thermocouples and data acquiring system are small
enough to give almost simultaneous real-time measurement. The value was taken to be
600◦ C in this study. Typical examples of the measured gas temperatures are shown in
Figures 3 and 4.
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

213
Figure 3. Vertical temperature profiles.
214 W. K. CHOW AND S. S. HAN

• Measuring the heat release rates of the gasoline pool fires


Test T8: Testing a 1 m pool fire of 12 liters of gasoline to onset flashover.
Test T9: Testing with a pool fire of 0.5 liter of gasoline, used as a small starting
fire for tests T1–T6 to simulate an accidental fire.

In tests T1–T6, the furniture was put near to the rear wall. The gasoline pool fire
was placed adjacent to the furniture. For test T7, a bigger pool of 1 m diameter with
12 liters of gasoline was put at the center of the room to onset flashover first.
The surface of the sofa sample in test T4 was protected by a commercial fire
retardant coating commonly used in the Far East including China and Hong Kong. This
is an intumescent type of fire retardant without releasing information on the chemical
composition. However, a thicker charring layer would be formed to protect the interior.
The appearances of the samples with fire retardant before and after the bench-scale
tests are shown in Figures 5a and 5b. This was a preliminary test with the objective
to assess the effect of applying such a fire retardant coating on the burning behavior
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

of furniture.
Sofa samples selected were comprised of wood frame, foam filling, and leather
coating. However, the thick foam fillings burnt most vigorously. The desk samples were
made of oak. These types of furniture are commonly used in local offices and residential
buildings in China. There might be slight differences in the material composition or
surface treatment, giving some changes in their fire behavior. Consumer products in
advanced countries are required to pass a series of strict assessments by following codes
and standards (e.g., [18]). However, most of the furniture commonly used in the Far
East, including Hong Kong, were not assessed so rigorously. Consequent to several big
fires in burning lots of furniture [1] or starting from igniting furniture foam [2], some
more fire safety standards on materials were specified by the Hong Kong government
[19–22]. However, there is no demonstration that such requirements were set up with
extensive development. More importantly, there is no database on fire testing results of
local furniture. Systematic fire tests on assessing local furniture should be carried out to
get a more reliable database before setting up new regulations.
In the 1980s, fire behavior of furniture was investigated by Babrauskas [5, 6]. The
heat release rate and other parameters on burning furniture were measured in furniture
or room calorimeters. Furniture testing fires developed up to flashover were studied.
However, most of the tests started from a smaller ignition source. There were not many
studies on burning furniture in a flashover fire. Note that burning furniture might lead to
flashover in a small enclosure. Therefore, studies on furniture exposed to a flashover fire
should be carried out. Preliminary studies will be reported in this article.
The testing procedure was described clearly before [16], basically following stan-
dard procedures as in ISO 9705 [13] and ASTM E1537 [23]. The procedures were
changed slightly to suit local constraints and for setting up flashover fires. The smaller
gasoline ignition source gave an average heat release rate of 20 kW with a duration of
3 min. This was different from the standard ignition source of 100 kW in 10 min and
300 kW for another 10 min in ISO 9705, as well as a constant 19.3 kW standard ignition
source with a duration of 80 s in ASTM E1537. This smaller ignition source was used
to understand igniting furniture under an accidental fire. Another bigger ignition source
with a peak heat release rate of 2.4 MW was used to onset flashover. The fire behavior
of furniture was then assessed under flashover.
FURNITURE UNDER FIRE 215
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Figure 4. Room air temperature. (a) Average temperature rise. (b) Maximum temperature rise.
216 W. K. CHOW AND S. S. HAN
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Figure 5. Effect of fire retardant. (a) Sample treated with fire retardant. (b) Appearance of fire retardant after
burning. (c) Effect of fire retardant on the heat release rate.

A thermal radiative heat flux meter was placed at the floor level as shown in
Figure 1. Thermocouples were put in positions labeled in Figure 1 as:

C: Corner of the wall near the room opening


M: Center of the room
T: Near to the ceiling
FURNITURE UNDER FIRE 217
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Figure 6. Lower exhaust rate.

Rc (R1 to R6): Near to the rear wall corner


Rm (R7 to R12): Near to the middle of the rear wall
D1 to D4: Room opening
D5: A point at the top of exhaust hood near to the duct

This study gave more temperature profiles than other standard tests [13, 23], though the
smoke and toxic gases were not measured.
Test 3 was conducted at a lower exhaust rate to check whether flue gas can be
extracted completely. Under such a low ventilation rate, combustion products might flow
out of the hood as in Figure 6. The exhaust fan-duct system should be set at sufficient
fan power to avoid leakage of flue gases.
There is in fact a concern in measuring the heat release rate of burning a large
amount of combustibles. Burning a smaller fire, say up to 1 MW, will not give out so
much smoke. It is relatively easier to design the fan-duct system. Further, losing 10%
of flue gas for a 1 MW calorimeter only gives up to 0.1 MW uncertainty. Such a slight
change would not affect the fire safety design. However for a 20 MW calorimeter, this
would give 2 MW uncertainty. The impact to the final fire safety design is significant.

3. RESULTS
Results on heat release rate curves, oxygen consumption rates, heat fluxes at floor
level position, and air temperature were measured and reported before [16]. Curves on
the heat release rate and radiative heat flux as shown in Figures 7 and 8 are taken out
for further analysis.
A summary of the key information such as the burning time, tB , peak heat release
rate, pkHRR, time to pkHRR, tfp , and peak heat flux is shown in Table 1. The vertical
temperature profiles and average room air temperatures in tests T1–T7 were measured and
are shown in Figures 3 and 4, respectively. The measured average room air temperatures
218 W. K. CHOW AND S. S. HAN
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Figure 7. Heat release rate.

Figure 8. Heat flux.


FURNITURE UNDER FIRE 219

Table 1. Summary of full-scale burning tests

Test Accidental fire Pool fire to onset flashover

Parameter T1 T2 T3 T4 T5 T6 T7 T8 T9

Mass/kg 22 20 22 23 31 53 20 9.5 0.4


tb /s (min) 1664 2000 2330 1643 2033 1576 1715 317 718
(28) (33) (39) (27) (34) (26) (29) (5) (12)
pkHRR/MW 1.05 0.63 1.12 1.05 0.64 2.06 3.01 2.45 0.04
tfp /s 405 465 496 541 360 365 290 270 475
NetTHR/MJ 353 402 483 411 375 868 451 418 18
EHC /MJkg−1 17 19 21 20 12 17 22 44 44
Peak heat flux/ 2.93 0.67 1.71 2.92 1.07 9.52 12.72 18.6 —
kWm−2

are compared with the standard temperature/time curve of BS 476 [17] as shown in
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Figure 4.
The following key points were observed:

• Room air temperature rise induced by the furniture fire was high for quite a long
time under flashover. Even without flashover, the maximum temperature might
reach a dangerous level and affect the building structures.
• Treating sofa materials with this selected commercial sample of fire retardants
would only delay the ignition time, tig , by several minutes under an accidental
small fire source. Once ignited, the materials burnt as unprotected sofa in test T1.
No obvious improvement was observed for the foam treated with fire retardant
under a flashover fire, except delaying the ignition time. This point is verified
by the bench-scale cone tests under heat fluxes over 20 kWm−2 as shown in
Figure 9.
• Under flashover condition in test T7, SF2 was ignited quickly with most of the
combustibles burnt up. Note that the amount of gasoline used was only to onset
flashover as shown in the results for test T8. Once flashover was achieved, almost
all the gasoline was consumed, but the sofa kept burning.

4. ANALYSIS ON HEAT RELEASE RATE


The following points are observed on the heat release rate of burning furniture:

• Burning the sofa foam SF1 without the treatment of fire retardants in test T1
would give a pkHRR over 1 MW as shown in Figure 7. Burning it with fire
retardant applied on the surface in test T4 did not reduce the pkHRR. However,
the time to peak heat release rate tfp was delayed by about 136 s. This is clearly
shown by plotting the two heat release rate curves for tests T1 and T4 together
in Figure 5. Applying a fire retardant coating over the sofa foam might delay the
heat release. This will allow a longer time for evacuation, which is very useful
in dealing with crowd movement and control.
• The principle of superposition, i.e., adding up the heat release rate curve of
each combustible item to give the total heat release rate of the arrangement,
220 W. K. CHOW AND S. S. HAN
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Figure 9. Bench-scale tests on heat release rate per unit area. (a) Furniture wood. (b) Furniture foam.
FURNITURE UNDER FIRE 221

might be applicable under smaller fires. This is demonstrated by adding the heat
release rate curves for T1 and T5, and comparing the resultant curve with the one
measured in test T6 plotted in Figure 10. These two curves gave similar patterns
with some differences. A possible reason is that combustion was enhanced to
release more heat when different combustibles are burnt together as in test T6.
Higher radiative heat fluxes emitted from adjacent combustibles might give more
vigorous burning process. The resultant transient heat release rate of burning the
two items might be estimated by adding their individual transient heat release
rate curves for smaller fires with lower thermal radiation. However, the ignition
time, tig , should be considered carefully in combining the transient curves of
individual items.
• Burning the sofa SF2 under a flashover fire (i.e., test T7) is very different from
burning it in an accident as in test T2. This is clearly illustrated by the net peak
heat release rates of 1.2 MW in test T7 and 0.7 MW in test T2, as shown in
Figures 11 and 12, though the net total heat released did not increase much.
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Therefore, testing furniture under a flashover fire will give a better assessment
on fire safety.

Furthermore, the following is observed for the case with a lower exhaust rate as
shown in Figure 6: Burning the sofa foam without adequate ventilation will give a longer
burning duration, but roughly the same value of pkHRR. The time to peak heat release
rate, tfp , can be extended to 91 s.

Figure 10. Superposition of heat release rates.


222 W. K. CHOW AND S. S. HAN
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Figure 11. Net heat release rate.

Figure 12. Net total heat released.


FURNITURE UNDER FIRE 223

5. TOTAL HEAT RELEASED AND EFFECTIVE HEAT OF COMBUSTION


The net heat release rate Q(t) of burning the combustibles can be computed from
the heat release rate QCG (t) of the combustible with a gas pool and the heat release rate
of a gas pool QG (t) only:

Q(t) = QCG (t) − QG (t). (1)

The results are plotted in Figure 11. It is observed that most single items of furniture
gave a net heat release rate higher than 1 MW with burning duration several minutes in
an accidental fire. The value will be higher in a post-flashover fire. The net peak heat
release rate for sofa SF2 in test T7, represented by curve G, is 0.568 MW higher than
that in test T2 as represented by curve B. If there are more adjacent combustibles, such
flaming fires would ignite them to give a much higher heat release rate. The net peak
heat release rate curve F in Figure 10 is 2.02 MW instead of 1.02 MW, due to igniting
an adjacent wood desk by the burning sofa. Note that the net peak heat release rate was
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

only 0.61 MW for burning only one wood desk.


The total heat released THR(t) (in MJ) can be obtained by integrating the heat
release rate curve Q(t) (in kW) from ignition time, tig to time t:
 t
THR(t) = 10−3 Q(t)dt. (2)
tig

The net total heat released can be calculated by the net heat release rate Q(t) using
this equation. The results for the seven tests are shown in Figure 12.
It is observed that there are great differences between burning one sofa alone and
burning the sofa together with other combustibles, especially under a post-flashover fire.
The net THR in burning SF1 and a wood desk in test T6 was 868 MJ. The value is
higher than 728 MJ by adding up the net THR values of 353 MJ for SF1 in test T1 and
375 MJ for a wood desk in test T5. The net THR of SF2 was 451 MJ in test T7, but
only 402 MJ in test T2. More complete combustion in using up more fuel under higher
temperatures gave higher heat release rate.
The effective heat of combustion of combustibles can be calculated by the total
heat released THR and the mass consumed m:
THR
EHC = . (3)
m
The computed values of EHC in the full-scale burning tests are shown in Table 1. The
values of EHC were measured to be 17–21 MJkg−1 for most samples, except being 12
MJkg−1 for the desk.
The values of the effective heat of combustion for wood-framed furniture samples
tested by Babrauskas [5, 6] lay mainly between 14.6–18.4 MJkg−1 . Although there are
slight differences among those results due to different components and materials used in
the furniture, agreement is still acceptable.
In addition, the samples were tested in a cone calorimeter. The furniture foam
and wood samples were cut into 10-cm squares. Radiative heat fluxes of 20 kWm−2 ,
35 kWm−2 , 50 kWm−2 , and 70 kWm−2 were adjusted. The results on the key parameters
such as ignition time, tig , peak heat release rate per unit area, pkHRR, time to first peak
224 W. K. CHOW AND S. S. HAN

Table 2. Summary of bench-scale burning tests

Group Furniture foam

Parameter No fire retardant With fire retardant Furniture wood

Heat flux/kWm−2 70 50 35 20 70 50 35 20 70 50 35 20
Mass/g 10 10 10 9.6 10 10 10 11 79 80 78 76
tig /s 1 2 3 4 4 4 10 15 10 24 73 614
pkHRR/kWm−2 574 500 430 362 821 758 560 367 256 199 167 232
tfp /s 356 55 57 71 26 35 29 60 14 23 21 174
THR/MJm−2 22 24 24 24 24 24 23 23 110 121 124 94
EHC /MJkg−1 21 23 24 25 23 23 22 21 14 15 16 13
tB /s 111 114 130 135 108 111 130 140 1200 1400 1600 1900
QtB /kWm−2 203 214 196 178 219 214 192 179 91 93 91 83

of heat release rate curve, tfp , total heat released per unit area, effective combustion heat,
EHC, burning time, tB , and average heat release rate per unit area over the burning time,
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

QtB , are shown in Table 2.


From the cone calorimeter tests, burning the furniture wood under heat fluxes higher
than 35 kWm−2 and all the furniture foam would give almost complete combustion. The
values of EHC were 21–25 MJkg−1 for furniture foam and 13–16 MJkg−1 for furniture
wood. It appears that the values of EHC for furniture in tests T1 to T6 are lower than
those samples in bench-scale tests due to the items were exposed to lower radiative heat
fluxes. The values of EHC measured in full-scale and in bench-scale tests are similar
under similar heat fluxes, such as the value of EHC is 22 MJkg−1 for sofa SF2 in test
T7 under flashover conditions.

6. CONCLUSIONS
Full-scale burning tests on different furniture arrangements were carried out in
a facility for measuring the heat release rate by the oxygen consumption calorimetry.
Wood and foam materials used in the furniture were tested with a cone calorimeter
under radiative heat fluxes up to 70 kWm−2 . The burning behavior of furniture under a
well-developed fire was studied.
This is a preliminary study on the fire behavior of local furniture. It is not possible
to burn a large sample of furniture due to resources limitation. Typical furniture with
common characteristics were selected. However, more bench-scale tests were carried out
to compare with the full-scale testing results.
Combustibles of the selected sofa samples are wood frame, foam filling, and leather
coating. The desk samples were made of oak wood. Furniture samples of sofa and desk in
different arrangements were tested. Because of resources limitation, only one test for each
arrangement was conducted to compare the differences. The testing procedure basically
followed the standard procedures as in ISO 9705 and ASTM E1537. There were slight
changes in using gasoline as the ignition source.
From the studies both on full-scale burning tests and bench-scale burning tests,
burning up all furniture under a flashover fire will give a much higher net heat release
rate and total heat released, compared with burning only part of the combustibles in an
accidental fire without flashover. This point should be considered carefully in designing
fire safe furniture when there is possibility to have a well-developed fire.
FURNITURE UNDER FIRE 225

Furniture treated with fire protective coating appears to be safe while testing under
small accidental fires as in the standard without flashover [23]. However, it might burn
even more vigorously under a flashover fire. Furniture samples treated with fire retardant
should be tested under a flashover fire.
The principle of superposition [24, 25], and combination of full-scale and bench-
scale testing results are proposed in assessing the fire hazard. However, superposition
is valid for smaller fires. There, the thermal radiation effects from other items are rela-
tively small. The interaction between different combustibles should be studied further to
estimate the total heat release rates.

ACKNOWLEDGMENT
The project is funded by a PolyU research studentship with account number
G-W 136.
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

REFERENCES
1. Garley Fire, South China Morning Post, Hong Kong, 21 November 1996.
2. Mei Foo Sun Chuen Fire, South China Morning Post, Hong Kong, 17 April 1997.
3. V. Babrauskas and S. J. Grayson, Heat Release in Fires, Elsevier Applied Science, London
and New York, 1992.
4. B. Sundstöm, Fire Safety of Upholstered Furniture—The Final Report on the CBUF Research
Programme, Interscience Communication Ltd., London, 1995.
5. V. Babrauskas, J. Lawson, W. Walton, and W. Twilley, Upholstered Furniture Heat Release
Rates Measured with the Furniture Calorimeter, NBSIR 82-2604, National Bureau of Standards,
Gaithersburg, Maryland, 1982.
6. V. Babrauskas and J. F. Krasny, Fire Behavior of Upholstered Furniture, NBS Monograph 173,
National Bureau of Standards, Gaithersburg, Maryland, 1985.
7. J. F. Krasny, W. J. Parker, and V. Babrauskas, Fire Behavior of Upholstered Furniture and
Mattresses, Noyes Publications, New York, 2001.
8. W. K. Chow, Review on Heat Release Rate of Burning Furniture, Int. J. Eng. Performance-
Based Fire Codes, vol. 4-2, pp. 54–59, 2002.
9. W. K. Chow, Assessment on Heat Release Rate of Furniture Foam Arrangement by a Cone
Calorimeter, J. Fire Sci., vol. 20-4, pp. 319–328, 2002.
10. 15 Die as Fire Rips through Karaoke Bar, South China Morning Post, Hong Kong, 26 January
1997.
11. Taegu’s Subway Line Has Been Beset by Disaster, Main section, p. 10, South China Morning
Post, Hong Kong, 19 February 2003.
12. 14 Injured in Peak-Hour MTR Arson Attack, Editorial, p. A1, South China Morning Post,
Hong Kong, 6 January 2004.
13. ISO 9705: 1993(E), Fire Tests—Full-Scale Room Test for Surface Products, International
Standards Organization, Geneva, Switzerland, 1996.
14. T. G. Cleary, T. J. Ohlemiller, and K. M. Villa, Influence of Ignition Source on the Flaming
Fire Hazard of Upholstered Furniture, Fire Safety Journal, vol. 23, pp. 79–102, 1994.
15. W. K. Chow, G. W. Zou, H. Dong, and Y. Gao, Necessity of Carrying out Full-Scale Burning
Tests for Post-Flashover Retail Shop Fires, Int. J. Eng. Performance-Based Fire Codes, vol. 5-1,
pp. 20–27, 2003.
16. W. K. Chow and S. S. Han, Full-Scale Burning Tests on Heat Release Rates of Furniture, Int.
J. Eng. Performance-Based Fire Codes, vol. 6-3, pp. 168–180, 2004.
226 W. K. CHOW AND S. S. HAN

17. BS 476, Fire Tests on Building Materials and Structures—Part 8: Test Methods and Criteria
for the Fire Resistance of Elements of Building Construction, British Standards Institution,
London, 1972.
18. UL 1056, Standard for Fire Test of Upholstered Furniture, 3rd ed., Underwriters Laboratories
Inc., Northbrook, IL, USA, 2000.
19. Codes of Practice for the Provision of Means of Access for Firefighting and Rescue Purposes,
Buildings Department, Hong Kong, 1995.
20. Codes of Practice for Fire Resistance Construction, Buildings Department, Hong Kong, 1996.
21. Codes of Practice for the Provision of Means of Escape in Case of Fire, Buildings Department,
Hong Kong, 1996.
22. Codes of Practice for Minimum Fire Service Installations and Equipment and Inspection,
Testing and Maintenance of Installations and Equipment, Fire Services Department, Hong
Kong, 1998.
23. ASTM E 1537-02a, Standard Test Method for Fire Testing of Upholstered Furniture, ASTM
International, West Conshohocken, PA, USA, 2002.
24. W. K. Chow and H. W. Au Yeung, On the Superposition of Heat Release Rate for Polymeric
Downloaded By: [HEAL-Link Consortium] At: 12:54 9 May 2011

Materials, Architectural Sci. Rev., vol. 46-2, pp. 145–150, 2003.


25. W. K. Chow, Support on Carrying out Full-Scale Burning Tests for Karaokes, Int. J. Eng.
Performance-Based Fire Codes, vol. 3-3, pp. 104–112, 2001.