Accepted Manuscript

Experiment study on the burning rates of ethanol square pool fires affected by

wall insulation and oblique airflow

Pei Zhu, Xishi Wang, Changfa Tao

PII: S0894-1777(14)00285-4

DOI: http://dx.doi.org/10.1016/j.expthermflusci.2014.11.006

Reference: ETF 8349

To appear in: Experimental Thermal and Fluid Science

Received Date: 25 August 2014

Revised Date: 21 October 2014

Accepted Date: 7 November 2014

Please cite this article as: P. Zhu, X. Wang, C. Tao, Experiment study on the burning rates of ethanol square pool

fires affected by wall insulation and oblique airflow, Experimental Thermal and Fluid Science (2014), doi: http://

dx.doi.org/10.1016/j.expthermflusci.2014.11.006

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers

we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and

review of the resulting proof before it is published in its final form. Please note that during the production process

errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Experiment study on the burning rates of ethanol square pool firesaffected by wall insulation and oblique airflow

Pei Zhu1, Xishi Wang1,2*and Changfa Tao1,3

1State Key Lab. of Fire Science, University of Science and Technology of China, Hefei, 230026, China2 Anhui Province Center of Collaborative Innovation for City Public Security, Hefei, 230026, China

3Hefei General Machinery Research Institute, Hefei, 230031, China

*Corresponding author: Tel.: 86-551-63606437, Fax: 86-551-63601669, Email: wxs@ustc.edu.cn

Abstract: The effects of pool wall insulation condition and oblique air flow on the

burning rates of ethanol square pool fires have been studied experimentally in a

small-scale wind tunnel, which focused on the heat transfer process of the fuel pool

system due to flame tilt and wall insulation. The results showed that the mass burning

rate of a pool fire decreased when the pool wall was insulated, especially in the cases

with larger airflow speeds. The maximum value of the mass burning rate without wall

insulation reached approximately 2.3 times of that with wall insulation at an air speed of

2.93 m/s and a tilt angle of 0°. For both cases with or without wall insulation, the mass

burning rates decreased with increase in pool sizes and increased with increase in airflow

tilt angles. However, for the cases with wall insulation, the differences in the mass

burning rates with different pool sizes are relatively small as the airflow speeds exceed

1.5 m/s, while the differences are relatively large for the cases without wall insulation. In

addition, the turbulence of the flame fluid would be weakened when the pool wall was

insulated.

Keywords: Pool fire; Burning rate; Wall insulation; Oblique airflow; Heat transfer

2

NomenclatureL pool edge length (cm)A fuel surface area (m2)

pC specific heat (kJ kg−1K−1)m fuel mass (g)

'm& mass burning rate (gs-1)''m& mass burning rate per unit area (gm-2s-1)m²D & difference between with and without wall insulation (gm-2s-1)tiltedm²& mass burning rate per unit area with oblique airflow (gm-2s-1)

levelm²& mass burning rate per unit area with horizontal airflow (gm-2s-1)k value of tiltedm²

& / levelm²&

q tilt angle (°)gH gasification heat of the fuel (kJ kg−1)

vH latent heat of evaporation (kJ kg−1)evpQ& net energy to evaporate the fuel (kW)evp withoutQ -& net energy to evaporate the fuel without wall insulation (kW)

evp withQ -& net energy to evaporate the fuel with wall insulation (kW)

inQ& total heat transfer from the flame and surroundings to the fuel (kW)outQ& energy loss from the fuel to the surroundings (kW)fradQ& radiative heat transfer from flame to the fuel surface (kW)

fconvQ& convective heat transfer from the flame to the fuel surface (kW)

w fuelQ -& total heat transfer from the pool wall to the fuel (kW)

wfQ& heat transfer from the flame to the pool wall through the top surface (kW)

radQ& radiative heat transfer from the tilt flame to the pool wall outside surface (kW)convQ& convective heat transfer from the tilt flame to the pool wall outside surface (kW)condQ& conductive heat transfer from the tilt flame to the pool wall outside surface (kW)

1lossQ& heat loss from the fuel to the surroundings through the pool wall (kW)2lossQ& heat loss from the fuel surface to the surroundings through reradiation (kW)

v airflow velocity（m/s）T wall temperature with wall insulation(℃)

'T wall temperature without wall insulation (℃)t time (s)

Subscriptswith with wall insulationwithout without wall insulationf flamew pool walll leeward rims streamside rimw windward rim

3

1. Introduction

The fire hazards of liquid hydrocarbon fuels have been a serious safety concern, which

is attributed to liquid hydrocarbons being ubiquitous throughout society, including in oil

tanks, etc. A considerable number of studies [1-7] on both the practical and the

fundamental aspects of pool fires have been conducted, which focused on the burning

rate, the flame height and pulsation, the thermal radiation, the total kinetic energy, and

the soot microstructure. However, the above studies mainly considered the cases of

burning in a quiescent ambient condition. In practical scenes, most pool fires may burn

under airflow environments, such as ambient wind in storage plants, etc.

Some reports had been focused on the effects of airflow on pool fires [8-12]. For a

large pool fire, the influence of wind speed on the burning rate was negligible up to 2 m/s,

but it had a certain effect at wind speeds higher than 2 m/s [8]. The results [9] of

methanol square pool fires with transverse air speeds from 0 to 5.5 m/s showed that the

burning rate of the smallest square pool (7.5×7.5 cm) monotonously increased with

increased wind speeds, while the burning rate of the largest square pool (30×30 cm) was

essentially invariant to this range of air speeds, and the intermediate-sized square pools

showed a non-monotonic increase. The experiment of a small pool fire in a 1/20

reduced-scale tunnel [10-12] showed that the burning rates of methanol square pool fires

decreased, while those of acetone and heptane square pool fires increased with increased

air speeds.

It should be noted that in most previous studies, the burning pools were embed into the

floor, so that the pool wall was not exposed to the surroundings and the tilted flame could

4

not touch the pool wall. But in most practical scenes, the burning pools may be placed

above the floor, such as the oil tanks in petrochemical enterprises, etc. In these scenarios,

the tilted flame can touch the pool wall and this will enhance the heat transfer from pool

to the fuel and then accelerate the fuel evaporation.

However, most of the above studies focused on horizontal cross airflow, but few

considered the effect of tilt wind on fire behaviour. Several studies [13-14] had been

investigated the effect of tilt wind on fire behaviors, but only focused on the flame spread

rate, flame shape of solid fuel fires. The studies [15-16] on the effects of oblique air flow

on burning rates of square ethanol pool fires showed that the mass burning rate increased

with the increase of tilt direction and the pool wall heat effect became more obvious

under oblique airflow, but they did not consider the effect of the pool wall insulation,

although it indicated that the insulation may mainly affect heat transfer [17, 18].

The current experimental study focuses on examining if the wall insulation conditions

would influence the burning characteristics of the pool fire under oblique airflow due to

flame tilt and the extent of these effects compared to that without wall insulation. Then

attempt to provide method on how to reduce the hazards of a liquid pool fire under such

conditions. However, there are few relevant reports on this topic. The results may be

helpful for estimating the mass loss and heat transfer of a pool fire under the coupling

effects of oblique airflow and pool wall insulation.

2. Experimental setup and methodology

2.1. Experimental facility and condition

Fig. 1 shows the schematic diagram of the experimental setup. The cross-section area

5

is 42×60 cm. The fuel used in the experiments was ethanol-C2H5OH, its properties are

shown in Table 1. An ethanol flame generally produces little soot when combusting under

normal atmospheric conditions; therefore, the flame emits relatively low levels of

thermal radiation compared to yellow soot-producing flames. The stainless steel square

pools with edge lengths, L, of 4, 6, 8, and 10 cm, a wall thickness of 1 mm and a depth of

1.5 cm were used based on the size in comparison with the cross-sectional area of the

model.

As shown in Fig. 1, airflow was produced using a 0.75 kW wind fan positioned at one

end of the wind tunnel, which generated downward wind with air speed up to 5 m/s. The

angle θ represents the wind tilt direction. A KA12 four-channel anemometer with

resolution of 0.01 m/s was used to measure the airflow speed inside the wind tunnel. In

this work, airflow speeds ranging from 0 to 3.0 m/s were considered to provide relatively

steady airflow. Turbulence intensity of the airflow was found to be less than ±6% at all

test points.

Table 1 The properties of the fuel

Fuel Density Boiling point Heat of combustion Specific heat at constant pressure

Ethanol 0.79 g/cm3 78.4℃ 1365.5 kJ/mol 2.4 kJ/(kg×℃)

The burning pool was placed on a horizontal support frame. An asbestos board with a

thickness of 1 cm between the pool and the support frame was used to avoid the pool

bottom heated by the flame and reduce heat loss from pool bottom. An electric balance

with sampling intervals of 0.1 s and resolution of 0.01 g was positioned below the

support frame to record the mass loss history of the fuel. Images of the flames were

6

recorded with a 3.3 Mega pixels SONY HDR-PJ10 digital camera.

Fig. 1. Schematic diagrams of the experimental apparatus

The experiments were carried out in a laboratory under the following ambient

conditions: the temperature was 18 ± 2 , the pressure was 101 ± 5 KPa, and the

humidity was 25±15%. All of the test cases are specified in Table 2. Each test case was

carried out at least three times. In each test, the door and the window were closed to

avoid the influence of ambient wind, and the fan was switched on after all of the

equipment and the fuel were set to the prescribed conditions. The fan and the data

acquisition system were activated at the same time, then after approximately 30 s, the

fuel was ignited. Each repeated test was conducted after the wind tunnel and fuel pan

returned to the initial ambient conditions.

Table 2 Experimental test cases

Test cases initial fuelmass(g)

Tilt angle of airflow(θ) Airflow speedv(m/s)

L(cm) Pool wallinsulation

Case1 17 0°10°20°30°

0/0.28/0.49/0.72/1.02/1.

25/1.74/2.31/2.93

4 With and without

Case2 35 0°10°

0/0.28/0.49/0.72/1.02/1.

6 With and without

7

20°30°

25/1.74/2.31/2.93

Case3 70 0°10°20°30°

0/0.28/0.49/0.72/1.02/1.

25/1.74/2.31/2.93

8 With and without

Case4 95 0°10°20°30°

0/0.28/0.49/0.72/1.02/1.

25/1.74/2.31/2.93

10 With and without

2.2. Heat balance of the fuel without pool wall insulation

Fig. 2 shows the schematic diagram of the heat balance in a liquid fuel pool fire

without wall insulation under oblique airflow. Three K-type thermocouples with bead

diameters of 1 mm and a resolution of 0.1 were soldered at the middle of the wall

surfaces to measure the wall temperatures, which are according to the leeward wall

temperature, 'lT , stream side wall temperature, '

sT , and windward wall temperature, 'wT ,

respectively. The fuel evaporating rate is determined by the net gain energy, evpQ& , of the

fuel from flame and surroundings, which is the difference between the total heat inQ&

into the fuel and the total heat loss outQ& from the fuel. The heat loss through the bottom of

the pool was ignored due to the asbestos board. Thus, the heat balance in the fuel can be

represented as:

win fconv fuelfradQ Q Q Q -= + +& & & & (1)

w wfuel f rad cond convQ Q Q Q Q- = + + +& & & & & (2)

1 2out loss lossQ Q Q= +& & & (3)

where fradQ& is radiative heat transfer from flame to fuel, fconvQ& is convective heat

transfer from flame to fuel, w fuelQ -& is the total heat transfer from pool wall to fuel, wfQ& is

the conductive heat transfer from flame to wall top surface, 1lossQ& is the heat loss from

fuel to the surroundings, 2lossQ& is the heat loss through pool wall to the surroundings.

radQ& , condQ& , convQ& are radiation, conductive, convective heat transfer from flame to the

8

pool walls, which include two stream side walls and the leeward wall due to flame tilt.

Therefore, the net gain energy can be expressed as:

w 1 2evp without in out fconv f rad cond conv loss lossfradQ Q Q Q Q Q Q Q Q Q Q- = - = + + + + + - -& & & & & & & & & & &

without gm HA· ·²= & (4)

gH is the effective heat of gasification, which is defined as:

pvg TH dH c= + ò (5)

where vH is the fuel latent evaporation heat and pdTcò is the sensible heat of fuel

warming up.

Fig. 2. Schematic of the pool structure and layout without wall insulation

In Eq. (4), the radiative heat transfer fradQ& from flame to fuel due to the flame tilt,

should be very small, which has been discussed by Hu et al. [20]. wfQ& , 1lossQ& , 2lossQ& are

also relative small comparing to other terms. The forced airflow may induce surface

circulation of the liquid, and thus improve the convective heat transfer fconvQ& at the

gas-liquid interface. In addition, the flame wrapping will lead to the pool walls heating

due to radQ& , condQ& , convQ& , where the most of the heat absorbed by the pool walls being

considered transfer to the liquid fuel through convective heat transfer at the solid-liquid

evpQ&

9

interface eventually. So all of these convective heat transfer will be the main factors that

affect the heat flux rate received by the fuel in the cases without wall insulation.

2.3. Heat balance of fuel with pool wall insulation

Fig. 3 shows the schematic diagram of the heat balance in a liquid fuel pool fire with

wall insulation under oblique airflow. The pool wall insulation was realized by wrapping

a layer of thermal insulation material with 5-mm thickness outside the pool. The thermal

insulation material being used here is Aerogel Blanket, which is non-flammable, and its

properties are shown in Table 3. Similarly, the three wall temperatures were also

measured, which are lT , sT and wT respectively. Due to the thermal insulation layer, the

heat transfer of radQ& , convQ& , condQ& and 1lossQ& will be ignored. Therefore, the heat balance for

the fuel can be represented as:

(6)

w wfuel fQ Q- =& & (7)

2out lossQ Q=& & (8)

Thus, the net gain energy can be represented as:

w 2fradevp with in out fconv f lossQ Q Q Q Q Q Q- = - = + + -& & & & & & &with gm HA· ·²= & (9)

Due to the wall insulation, the heat of radQ& , condQ& , convQ& transferred from flame to pool

wall were weakened obviously, while the convective heat transfer fconvQ& was still

important due to the forced airflow.

Therefore, the difference in the net gain energy of the fuel evpQD & between the cases

without and with wall insulation can be represented as:

2evp evp without evp with rad cond conv loss gQ Q Q Q Q HAQ Q m · ·²

- -= - = + +D - = D& & & & & & & & (10)

win fconv fuelfradQ Q Q Q -= + +& & & &

evpQ&

10

where m²D & = without withm m² ²-& & , and withoutm²& is the mass burning rate without wall insulation

and withm²& is the mass burning rate with wall insulation. Comparing to the heat of radQ& ,

condQ& , and convQ& transferred from the flame to the pool wall, the heat loss 2lossq& is small.

Fig. 3. Schematic of the pool structure and layout with wall insulation

Table 3 The properties of the pool and thermal insulation material

Material Density(g/cm3) Thermal conductivity (w/m×k)

Stainless steel 7.79 16(20 ) 23(600 )

Aerogel Blanket 0.22 0.018(25 ) 0.045(650 )

3. Experimental results and discussion

3.1. Flame behaviour of a pool fire

Fig. 4 and Fig. 5 show the typical flame images of a 10×10 cm pool fire with an

increase in airflow speeds and tilt angles without and with wall insulation, respectively.

In the quiescent condition, the flame shapes generally depend on the relative strength of

the buoyancy of the combustion products in drawing ambient air horizontally over the

pool’s base to participate in the combustion. As the airflow speeds increased, even by a

small amount, under the quiescent conditions (e.g., 0.49 m/s), the flame shapes were

distorted considerably, with the main part of the flame shifting to the downstream side,

11

which is because the buoyant force has not persisted in the flow for a long time, and the

horizontal air momentum dominates in the near field of the pool. In our experiments, the

fuel pan was only put on a support frame whose top was a 10×10 cm flat plate, and the

leeward edge of the fuel pan was aligned with the downstream edge of the flat plate.

It is found that with a tilt angle of 20° or 30°, the flame was tilted horizontally and

even downward, where the tip of the tilted flame was lower than the bottom of the pool at

high airflow speeds. This is a particular characteristic of the flame with the effect of

oblique airflow as shown in Fig. 4 and Fig. 5. It is also found that the turbulence intensity

and the volume of the flames with wall insulation weakened and reduced compared to

those without wall insulation. For the cases with pool wall insulation, the wrapping of the

flame around the fuel pan side rim was not obvious when the tilt angle was 0° and 10°,

but obvious for other cases.

Fig. 4. Flame images of pool fires under different airflow speeds and directions(L=10 cm, without pool wall insulation)

The detachment of the flames from the windward rim of the pools occurred at

relatively large air speeds and tilt angles, which was obvious at all tilt angles when the

airflow speed exceeded 1.74 m/s without wall insulation. In addition, the flame length

12

was also detected based on the flame images, which showed that the flame length first

increases and then decreases with increased airflow speeds. From these phenomenon, it

can be preliminary deduced that the fuel evaporation and fuel vapour diffusion rate from

the fuel surface decreased with wall insulation. Other explanations will be combined with

the measured mass loss and temperature data in the following sections.

Fig. 5. Flame images of pool fires under different airflow speeds and directions(L=10 cm, with pool wall insulation)

3.2. Typical temporal evolution of fuel mass and pool wall temperature

A typical temporal evolution of fuel mass loss m and mass loss rate &'m with and

without wall insulation are shown in Fig. 6, which corresponds to a pool size with L=10

cm, a tilt angle of 30°, and an airflow speed of 2.31 m/s. It can be seen that the total

burning time and steady burning time with wall insulation are all longer than those

without wall insulation. Generally, as shown in Fig. 6, the variation of the burning rate

can be divided into four stages. In the cases with pool wall insulation, they are the first

stage with rapid increase, the second stage with slow increase, the third stage with a long

time of steady combustion, and the last stage with rapid decaying. However, in the cases

without pool wall insulation, rapid increase occurred for the second stage, and the steady

stage remained for a relatively short time. These can be explained that as the heating

13

effect of the three rims of the pool by flame would decrease remarkably due to wall

insulation. However, in the cases without wall insulation, the pool wall will be heated

directly by the wrapping flame, which will enhance the fuel evaporation rate. Based on

the requirement of data analysis, only the steady burning stage was in a quasi-steady state

system. During this stage, the fuel liquid level will decrease with ongoing burning, but

the temperature readings were still steady at this stage in the current study for both of the

cases.

0 100 200 300 400 500 600

0

20

40

60

80

100without&'m

withoutmwithm

mas

s(g)

t(s)

with&'m

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

（）

mas

sbu

rnin

gra

teg/

s

steady stage

Fig. 6. Typical temporal evolution of the mass loss and mass burning rate of pool fires with andwithout wall insulation (θ=30°, v=2.31 m/s, L=10 cm)

Figure 7 shows the typical temporal evolution of the wall temperature, which

corresponds to the cases in Fig. 6. In the steady burning stage, the wall temperature for

both cases basically remains a constant value. This will further account for the stability

of the mass burning rate at this stage.

In addition, it can be found that the windward rim temperature is lower than the other

rim temperature because the flame will be detached from the windward rim at an airflow

speed of approximately 2.31 m/s (as shown in Fig. 4 and Fig. 5), while the temperature of

the side and leeward rims are relative large due to the heating of the wrapping flame. In

14

the last decay stage, the rapidly rising rim temperature is caused by the fuel burning out,

and the fuel pan inside can be heated directly by the flame. Therefore, in the following

sections, the time-averaged mass burning rates per unit area &''m at the steady stage

were selected as the final mass burning rate in the analysis.

0 100 200 300 400 500 600

20406080

100120140160180200

Without wall insulationT'lT'sT'w

T()

t(s)

With wall insulationTlTsTw

steady stage

Fig. 7. Typical temporal evolution of the pool wall temperature with and without wall insulation(θ=30°, v=2.31 m/s, L=10 cm)

3.3. Burning rate of a pool fire without and with wall insulation

The burning rates without&''m of a pool fire without wall insulation under different

airflow speeds and directions are presented in Fig. 8. It is obviously shown that the mass

burning rates increase with increased air speeds and decrease with increased pool size,

and it can be expressed as µ& ''m 1/ L . The results related to the horizontal airflow agree

well with the results of a small square methanol pool fire presented by Wood et al. [10].

For a small tilt angle, as shown in Fig. 8(a) and (b), the burning rates increase

monotonically, and rapid increase occurs approximately 1.5 m/s. However, for relative

larger tilt angles, as shown in Fig. 8(c) and (d), the burning rates vary non-monotonically

with the increase of airflow speeds, which involves short-lived stability. When the tilt

15

angle is relative large, the side and leeward rim of the pool will be heated by the tilt

flame, which cause initial sharp increases in the burning rate. The short-lived stability

may be caused by steady net energy gain from the flame and the surroundings of the

fuel.

0.0 0.5 1.0 1.5 2.0 2.5 3.010

20

30

40

50

m''(

g/m

-2s-1

)

v(m/s)

4cm 6cm 8cm 10cm

θ=0°

（ ）a

0.0 0.5 1.0 1.5 2.0 2.5 3.010

20

30

40

50

m''(

g/m

-2s-1

)

4cm 6cm 8cm 10cm

v(m/s)

θ=10°

（ ）b

0.0 0.5 1.0 1.5 2.0 2.5 3.0

10

20

30

40

50 θ=20°

v(m/s)

m''(

g/m

-2s-1

)

4cm 6cm 8cm 10cm

（ ）c

0.0 0.5 1.0 1.5 2.0 2.5 3.0

10

20

30

40

50 θ=30°

v(m/s)

m''(

g/m

-2s-1

)

4cm 6cm 8cm 10cm

（ ）d

Fig. 8. Burning rate of square pool fires without wall insulation for different airflow speeds and pool sizes

It is also found that as air speeds increase from 1.25 m/s to 1.74 m/s, the mass burning

rates increases rapidly at different wind directions and pool size. As shown in Fig. 4, the

flame detachment occurred at the air speed of 1.74 m/s, but it did not happen at the air

speed of 1.25 m/s. When the flame was detached, the turbulence of the flame was

obviously enhanced, and the convective heat transfer between the fuel and the flame at

the gas-liquid interface would be improved, therefore the combustion process would be

accelerated. When the air speed exceeds 1.74 m/s, the increase in the mass burning rates

is reduced. Overall, the various paths for heat transfer from the flame to the fuel lead to

evpQ&

16

the complex coupling effects on fuel evaporation and fuel combustion. The key point in

steady state is that the net rate of energy to the fuel by all of these mechanisms balances

the rate of energy consumed for fuel evaporation.

17

The burning rates with&''m of pool fires with wall insulation under different airflow

speeds and directions are presented in Fig. 9. Similar to Fig. 8, the burning rates decrease

with the increase of pool size, but the burning rates decrease obviously compared to the

cases without wall insulation. In addition, the differences in the burning rates with

different pool sizes are relatively small compared to the cases without wall insulation. In

which, the differences are relatively large when the airflow speeds exceed 1.5 m/s. For a

small tilt angle (θ=0° and θ=10°), as shown in Fig. 9(a) and (b), the burning rates with&''m

first decrease and then increase with increased air speeds, except for the cases of 4×4 cm

pool fires.

For small tilt angles and low air speeds, flame tilting would cause the view factor from

the flame to the fuel surface to decrease, which would reduce the radiative heat transfer

fradQ& . In addition, due to the small vertical components of the airflow speeds, the

convective heat transfer fconvQ& from the flame to the fuel is relative small. On the other

hand, due to wall insulation, the heating effect by flame wrapping is approximately

eliminated. Meanwhile, the flame is vertical in the quiescent condition, and the energy

from the flame to the fuel will be relatively large compared to the cases at low air speeds.

All of these lead to the first decrease in the burning rates. With the continued increase in

air speeds, the convective heat transfer will obviously increase, which will lead the

further increases in the burning rate.

fconvQ&

18

0.0 0.5 1.0 1.5 2.0 2.5 3.0

12

16

20

24

28m

″(gs

-1m

-2)

v(m/s)

4cm 6cm 8cm 10cm

θ=0°

（ ）a

0.0 0.5 1.0 1.5 2.0 2.5 3.0

12

16

20

24

28 θ=10°

m″(g

s-1m

-2)

4cm 6cm 8cm 10cm

v(m/s)

（ ）b

0.0 0.5 1.0 1.5 2.0 2.5 3.0

12

16

20

24

28 θ=20°4cm 6cm 8cm 10cm

m″(g

s-1m

-2 )

v(m/s)

（ ）c0.0 0.5 1.0 1.5 2.0 2.5 3.0

12

16

20

24

28 θ=30°

m″(g

s-1m

-2)

L=4cm L=6cm L=8cm L=10cm

v(m/s)

（ ）d

Fig. 9. Burning rate of square pool fires with wall insulation for different airflow speeds and pool sizes

19

When the tilt angle further increases up to 20° and 30°, the burning rates will

always increase with increased airflow speeds, except for the cases where the airflow

speeds are below 0.28 m/s. The understanding of this phenomenon is that the effects of

the oblique airflow on the flame shape and the fuel become more obvious due to the

increase in the vertical components of the airflow speeds. For instance, the convective

heat transfer was obviously enhanced compared to the one with small tilt angles

of airflow. For the cases with larger pool sizes, especially the 8×8 cm and 10×10 cm

pools, the lowest value of the burning rate occurred when the airflow speed was

approximately 0.28 m/s. The reasons are similar to the cases with small tilt angles, which

was discussed above.

However, for a small pool size of 4×4 cm with all of the tilt angles, the mass burning

rate increased with increased airflow speeds. It can be interpreted that the heat feedback

per unit area from the flame to the fuel is relatively large compared to the large pool size.

For all of the pool sizes, there is also a critical value of air speed leading to a relatively

fast increase in the burning rate, which corresponds to the flame detachment. In addition,

from Fig. 9, for all of the pool sizes and oblique airflow angles, a steady or very slow

state of increase in the mass burning rate with increased airflow speeds generally appears,

which can also be interpreted as the net rate of energy gain to the fuel by all of these

mechanisms and balances the rate of energy consumed for fuel evaporation.

As shown in Table 4, a parameter /without withm m² ²& & is introduced to quantitatively indicate

the difference in the mass burning rate between the cases with and without wall

insulation. The values correspond to the ratio of the mass burning rates with different

with&''m

fconvQ&

20

airflow tilt angles and pool sizes at the maximum airflow speed of 2.93 m/s. The

maximum value of is approximately 2.3 for the 4×4 cm pool with a tilt

angle of 0°, while the minimum value is approximately 1.43 for the 10×10 cm pool with

a tilt angle of 30°. Therefore, it can be said that the effects of wall insulation on the flame

burning rate are very notable for the cases with relatively large airflow speeds and small

pool sizes.

Table 4 The ratio of the mass burning rate per unit area with different pool sizes and airflow tilt angles

4×4 cm 6×6 cm 8×8 cm 10×10 cm

0° 2.32 2.3 2.17 2.0110° 1.86 2.02 1.77 1.620° 2.17 1.62 1.44 1.4330° 2.02 1.56 1.43 1.42

/without withm m² ²& &

/without withm m² ²& &

21

3.4. Effects of airflow directions on mass burning rates

To indicate the influence of airflow directions on the burning rate of pool fire more

clearly, the burning rates with and without wall insulation for different airflow tilt angles

were plotted in Fig. 10 and Fig. 11. The data of burning rates for different pool sizes

under different tilt angles and airflow speeds are shown in Fig. 10. It can be seen that the

mass burning rates increase with increased tilt angles for all of the cases. However, the

differences in the mass burning rate with different tilt angles are relatively large in the

cases with wall insulation when airflow speed exceeds 1.5 m/s. In addition, the difference

of burning rate between the cases without and with wall insulation would decrease with

the increase of pool size. To describe the variation of HRR with ventilation velocity,

Carvel et al. [20] defined a coefficient, /vent natk Q Q= , where ventQ is the HRR with forced

ventilation and natQ is the HRR with natural ventilation. Similar to this, we define the mass

burning rate enhancement coefficient, k, as:

'' ''/tilted levelk m m= & & (11)

where ''tiltedm& is the mass burning rate with oblique airflow, ''

levelm& is the mass burning

rate with horizontal airflow.

22

0.0 0.5 1.0 1.5 2.0 2.5 3.010

20

30

40

50m

''(g/

m-2

s-1)

With wall insulation θ=0° θ=10° θ=20° θ=30°

Without wall insulation θ′=0° θ′=10° θ′=20° θ′=30°

v(m/s)

L=4cm

(a)

0.0 0.5 1.0 1.5 2.0 2.5 3.010

20

30

40

50

m''(

g/m

-2s-1

)

With wall insulation θ=0° θ=10° θ=20° θ=30°

Without wallinsulation θ′=0° θ′=10° θ′=20° θ′=30°

v(m/s)

L=6cm

(b)

0.0 0.5 1.0 1.5 2.0 2.5 3.010

20

30

40

50

m''(

gs-1m

-2)

With wall insulation θ=0° θ=10° θ=20° θ=30°

Without wall insulation θ′=0° θ′=10° θ′=20° θ′=30°

v(m/s)

L=8cm

(c)

0.0 0.5 1.0 1.5 2.0 2.5 3.010

20

30

40

50

m″(g

m2 s-

1 )

v(m/s)

With wall insulation θ=0° θ=10° θ=20° θ=30°

Without wall insulation θ′=0° θ′=10° θ′=20° θ′=30°

L=10cm

(d)

Fig. 10. Mass burning rate with and without wall insulation for different airflow speeds and directions

Fig. 11 and Fig. 12 show the results of k with and without wall insulation for different

airflow speeds and pool sizes, where withk and withoutk represent the data of the pool fire

with or without wall insulation, respectively. It can be seen that there are significant

differences in the effect of airflow directions on the mass burning rate between the two

cases. In addition, the values of k increase with an increase in the tilt angle for both cases.

For the cases without wall insulation, as shown in Fig. 11, the values of k first increase

until a peak value is reached and then decrease with an increase in airflow speed.

Combined with Fig. 10, it can be concluded that the effect of the airflow direction on the

mass burning rate is only prominent at relatively lower airflow speeds (e.g., 0.49-1.25

m/s). As shown in Fig. 4, at this range of airflow speeds, the variation of flame shape is

obvious with the increased airflow tilt angle. For instance, with an air speed of 0.72 m/s,

the flame tilt gradually increased with the airflow tilt angle increasing from 0° to 30°. At

23

θ=0°, the flames deflected but did not reach the horizontal, and the flame did not

completely wrap the pool wall. However, at θ=30°, the flame deflected to the horizontal

and would touch the pool wall, such that the energy of radQ& , condQ& , and convQ& from the

flame to the pool wall would enhance, thus leading to an increase in the net energy gain

evpQ& .

Under large airflow speeds, the flame would be totally deflected to the horizontal or

even downward, the side and leeward rims of the fuel pan would all be wrapped by the

flame, and the differences in the energy from the flame to the wall will be not large with

different airflow tilt angles at the same airflow speed. Therefore, the differences in the

burning rate with different airflow directions are mainly caused by the convective heat

transfer fconvQ& from the flame to the fuel. However, the energy in this part is relatively

small compared to wall heated effect on fuel heating. Therefore, the effect of the airflow

directions on the mass burning rate is not obvious for the cases with large airflow speeds.

0.0 0.5 1.0 1.5 2.0 2.5 3.00.9

1.0

1.1

1.2

1.3

1.4

1.5 Without wall insulation

k with

out

v(m/s)

θ=10° θ=20° θ=30°4cm

（ ）a

0.0 0.5 1.0 1.5 2.0 2.5 3.00.8

1.0

1.2

1.4

1.6

k with

out

v(m/s)

θ=10° θ=20° θ=30°6cmWithout wall insulation

（ ）b

0.0 0.5 1.0 1.5 2.0 2.5 3.00.8

1.0

1.2

1.4

1.6

1.8

k with

out

v(m/s)

θ=10° θ=20° θ=30°8cmWithout wall insulation

（ ）c0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.8

1.0

1.2

1.4

1.6

1.8

k with

out

v(m/s)

θ=10° θ=20° θ=30°10cmWithout wall insulation

（ ）d

Fig. 11. A plot of k via airflow speed without wall insulation for different airflow tilt anglesand pool sizes

24

0.0 0.5 1.0 1.5 2.0 2.5 3.0v(m/s)

θ=10° θ=20° θ=30°4cmWith wall insulation

（ ）a0.9

1.0

1.1

1.2

1.3

1.4

1.5k w

ith

0.0 0.5 1.0 1.5 2.0 2.5 3.00.8

1.0

1.2

1.4

1.6

k with

v(m/s)

θ=10° θ=20° θ=30°6cmWith wall insulation

（ ）b

0.0 0.5 1.0 1.5 2.0 2.5 3.00.8

1.0

1.2

1.4

1.6

1.8

k wit

h

v(m/s)

θ=10° θ=20° θ=30°

8cmWith wall insulation

（ ）c

0.0 0.5 1.0 1.5 2.0 2.5 3.00.8

1.0

1.2

1.4

1.6

1.8

k wit

h

v(m/s)

θ=10° θ=20° θ=30°

（ ）d

10cmWith wall insulation

Fig. 12. A plot of k via airflow speed with wall insulation for different airflow tilt anglesand pool sizes

For the cases with wall insulation, as shown in Fig. 12, the values of k first increase

and then remain steady at a certain value with an increase in airflow speeds. Due to the

insulation of the pool wall, the effect of the wall heating will not exist. The effects of the

oblique airflow on the convective heat transfer fconvQ& from the flame to the fuel would

be the main part that would lead to the change in the burning rates with different airflow

tilt angles. Therefore, the differences in the burning rates with different tilt angles are

relatively large with wall insulation, compared to that without wall insulation under large

airflow speeds.

4. Conclusions

The effects of oblique airflow on the burning rate of square ethanol pool fires with and

without wall insulation have been studied experimentally based on a small-scale wind

tunnel. The variation in the fuel mass loss and wall temperature were measured to

25

analyse the effect of wall insulation and oblique airflow on the pool fire burning rates.

The following conclusions can be drawn:

(1) The mass burning rates are obviously reduced due to the insulation of the pool wall,

especially in the cases with larger airflow speeds.

(2) The mass burning rates decrease with increased pool size. However, in the cases

with wall insulation, the differences in the mass burning rates are relatively small, while

for the cases without wall insulation, the differences are relatively large, especially when

the airflow speeds exceed 1.5 m/s.

(3) The burning rates increase with increased airflow tilt angles for both cases with or

without wall insulation. However, in the cases with wall insulation, the change in the

mass burning rates with different airflow tilt angles are relatively large when the airflow

speed exceeds 1.5 m/s.

Acknowledgments

The authors appreciate the support of the China National Key Basic Research Special

Funds Project (Grant No. 2012CB719704) and the Natural Science Foundation of China

(Grant No. 51323010).

References

[1] Nasr A, Suard S, El-Rabii H, Garo JP, Gay L, Rigollet L, Heat feedback to the fuel

surface of a pool fire in an enclosure, FireSafetyJ. 60(2013)56-63.

DOI: 10.1016/j.firesaf.2012.12.005

[2] Alireza V, David SN, Larry WK,Effects of altering the liquid phase boundary

conditions of methanol pool fires,Exp. Therm.Fluid Sci. 44 (2013) 786–791.

DOI: 10.1016/j.expthermflusci.2012.09.023

[3] Philippe Dagaut, Abderrahman El Bakali, Alain Ristoridoi, The combustion of

kerosene: Experimental results and kinetic modelling using 1- to 3-component

26

surrogate model fuels, Fuel 85(2006)944-956. DOI:10.1016/j.fuel.2005.10.008

[4] Benjamin DD, John L, Thomas KB, Marcos C, Robert GBJ, Sergey BD, Pool

fires-An empirical correlation. Combust. Flame160(2013)2964-2974.

DOI: 10.1016/j.combustflame.2013.06.020

[5] Novozhilov V, Koseki H, CFD prediction of pool fire burning rates and flame

feedback. Combust. Sci. Technol. 176(2004)1283-1307.

DOI: 10.1080/00102200490457484

[6] Sudheer S, Prabhu SV, Measurement of flame emissivity of gasoline pool

fires,Nucl.Eng. Des. 240 (2010) 3474–3480.DOI: 10.1016/j.nucengdes.2010.04.043

[7]Chen B, Lu SX, Li CH, Kang QS, Lecoustre V. Initial fuel temperature effects on

burning rate of pool fire. J. Hazard. Mater. 188(2011)369-74.

DOI: 10.1016/j.jhazmat.2011.01.122

[8] Chatris JM, Quintela J, Folch J, Planas E, Arnaldos J, Casal J, Experimental study of

burning rate in hydrocarbon pool fires. Combust. Flame 126(2001)1373-1383.

DOI: 10.1016/S0010-2180(01)00262-0

[9] Woods JAR, Fleck BA, Kostiuk LW, Effects of transverse air flow on burning rates of

rectangular methanol pool fires, Combust. Flame 146(2006)379-390.

DOI: 10.1016/j.combustflame.2006.02.007

[10] Roh JS, Ryou HS, Kim DH, Jung WS, Jang YJ, Critical velocity and burning rate in

pool fire during longitudinal ventilation, Tunn. Undergr. Sp. Technol. 22(2007)

262-271.DOI: 10.1016/j.tust.2006.08.003

[11] Roh JS, Yang SS, Ryou HS, Tunnel fires: experiments on critical velocity and

burning rate in pool fire during longitudinal ventilation, Journal Fire Sciences 2007;

25: 161-176.DOI: 10.1177/0734904107067300

[12] Roh JS, Yang SS, Ryou HS, Yoon MO, Jeong YT, An experimental study on the

effect of ventilation velocity on burning rate in tunnel fires-heptane pool fire case,

Build. Environ. 43(2008) 1225-1231.DOI:10.1016/j.buildenv.2007.03.007

[13] Boboulos M, Purvis MRI, Wind and slope effects on ROS during the fire

propagation in East-Mediterranean pine forest litter, Fire Safety J. 44(2009)

764-69.DOI: 10.1016/j.firesaf.2009.03.006

27

[14] Perez Y, Pastor E, Agueda A, Planas E, Effect of wind and slope when scaling the

forest fires rate of spread of laboratory experiments. Fire Technol.

47(2011)475-89.DOI: 10.1007/s10694-010-0168-7

[15] Tao CF, Wang XS, Zhang XN, Effects of transverse air flow on mass loss rate of

alcohol pool fires in an inclined wind tunnel, J. Civ. Eng. Manag. 2014 (In press).

[16] Tao CF, He YP, Li Y, Wang XS, Effects of oblique air flow on burning rates of

square ethanol pool fires, J. Hazard. Mater. 260(2013) 552-562.

DOI: 10.1016/j.jhazmat.2013.06.015

[17] Chou HM, Wong KL, Heat transfer characteristics of an insulated regular polygonal

pipe by using a wedge thermal resistance model, Energ. Convers. Manage. 44(2003)

629-645. DOI: 10.1016/S0196-8904(02)00064-X

[18] Wong KL, Chou HM, Li YH, Complete heat transfer solutions of an insulated

regular polygonal pipe by using a PWTR model, Energ. Convers. Manage.

45(2004)1705-1724. DOI: 10.1016/j.enconman.2003.09.025

[19] Carvel RO, Beard AN, Jowitt PW, Drysdale DD, Variation of heat release rate with

forced longitudinal ventilation for vehicle fires in tunnels, Fire Safety J.

36(2001)569-96.DOI: 10.1016/S0379-7112(01)00010-8

[20] Hu LH, Liu S, Wu L, Flame radiation feedback to fuel surface in medium ethanol

and heptane pool fires with cross air flow, Combust. Flame 160 (2013) 295–306.

DOI: 10.1016/j. combust flame.2012.10.016

28

Highlights

l Turbulence and volume of the flame reduce significantly as pool wall is

insulated.

l The mass burning rates obviously decrease as the pool wall is

insulated.

l The reduction of mass burning rate is sensitive to pool size and airflow

speeds.

l Mass burning rate with wall insulation is less sensitive to pool size as

v>1.5 m/s.

l Mass burning rate are quite different under different airflow directions.