Fuel 162 (2015) 313322

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Adaptive air distribution in an ejector burner for the utilisation

of methanol-mixed fuels
Jun Shi a, Jingyu Ran a,b,, Changlei Qin a, Li Zhang a,b
a
b

Key Laboratory of Low-grade Energy Utilisation Technologies and Systems, Chongqing University, Ministry of Education, Chongqing 400030, China
College of Power Engineering, Chongqing University, Chongqing 400030, China

a r t i c l e

i n f o

Article history:
Received 5 December 2014
Received in revised form 1 September 2015
Accepted 1 September 2015
Available online 14 September 2015
Keywords:
Methanol-mixed fuels
Ejector burner
Adaptive air distribution
Molar entrainment ratio
Combustion efficiency
Structure optimisation

a b s t r a c t
Numerous studies have focused on the utilisation of alcohol fuels because of the energy crisis. However,
the combustion efficiency and stability of alcohol fuels are unacceptable. This study proposed an ejecting
combustion method for utilising methanol-mixed fuels and numerically investigated the characteristics
of adaptive air distribution in an ejector burner. The geometrical parameters of the ejector burner were
optimised and validated by an experiment. Results show that the suction effect of negative pressure in
the mixing chamber and the entrainment effect of fuel jet flow both play important roles for an ejector
burner to draw air. The positions of ejector nozzle exit locating at the suction chamber axis and low operating pressure are beneficial for obtaining a stable air distribution. Molar entrainment ratio (MER) rapidly
increases with an increase in parameter a, which is defined as the ratio of throat diameter to nozzle exit
diameter, but declines with increasing ejector back pressure. In the experiment, the changing rate of MER
is less than 6.4%, and combustion efficiency is higher than 99.2% in the load range of 20120%, which is
highly consistent with that of the simulation. The optimised burner could automatically distribute air
supply and facilitate stable combustion.
2015 Elsevier Ltd. All rights reserved.

1. Introduction
Energy utilisation has recently become a public concern
because of the shortage in fossil fuels, the increase in oil price
and the severity of environmental pollution. A number of clean
and renewable energies are under development [14]. Methanolmixed fuel, which mainly consists of methanol, propane and a
small amount of additives, is a potential substitute for gasoline
because of their similarities in physical property. However, with
only half of the low calorific value of gasoline, methanol-mixed
fuel has a high latent heat of vaporisation, which is more than
thrice that of the former fuel. Aside from the difficulties in ignition
and combustion stability, combustion efficiency would thus be low
if methanol-mixed fuels are utilised by using traditional methods.
This study thus presents a novel ejecting combustion method [5]
for the utilisation of methanol-mixed fuels. This method is characterised by the vaporisation of liquid methanol-mixed fuels and the
automatic adjustment of air supply using an ejector. The adaptive
air distribution and combustion performance of an ejector burner
must be studied, because the efficiency in the automatic
Corresponding author at: No. 174 Shazhengjie, Shapingba, Chongqing 400044,
China. Tel.: +86 23 6510 2107; fax: +86 23 6511 1832.
E-mail address: ranjy@cqu.edu.cn (J. Ran).
http://dx.doi.org/10.1016/j.fuel.2015.09.004
0016-2361/ 2015 Elsevier Ltd. All rights reserved.

adjustment of air supply is poor when the ejector burner load

changes in a large scope.
Ejector is widely used in industrial sectors, such as water desalination, geothermal power and chemical plants, particularly in
refrigeration systems [6,7]. Yang et al. [8] analysed the influence
of different geometrical structures of ejector nozzles on drawing
quality, and circular nozzle exit was found to have the best performance. Varga et al. [9], Ruangtrakoon et al. [10] and Yan et al. [11]
indicated that the ratio of ejector throat area to nozzle exit area
significantly affects the drawing quality of ejector. Nozzle position
also affects entrainment efficiency. Pounds et al. [12] and some
other investigators [1315] studied the influence of nozzle position
on entrainment performance, and found an optimal nozzle position
could produce a maximal coefficient of performance. Pianthong
et al. [16] and some investigators [1719] studied the effects of
mixing chamber length, suction chamber angle and diffuser chamber length on entrainment performance. In addition, Chunnanond
and Aphornratana [20], Yan et al. [21] and Chong et al. [22]
investigated the static pressure along the ejector axis at different
operating conditions. Opgenorth et al. [23] investigated the effect
of back pressure on entrainment ratio to improve the efficiency
of ejectors in refrigeration systems. Sriveerakul et al. [24,25] and
other researchers [26,27] discussed the mixing process of fuel flow
and air in mixing chamber. However, most of the aforementioned

314

J. Shi et al. / Fuel 162 (2015) 313322

Nomenclature
Qv
Qh
Nh
Nf
Mk
Mf
Ph
Th
Pdy
Pjy
R
S
PO
Pb

qf
qh

volume flow of liquid fuel (m3 s1)

volume flow of mixing gas (m3 s1)
molar flow rate of mixing gas (mol s1)
molar flow rate of fuel (mol s1)
molar mass of air (kg mol1)
molar mass of fuel corrected by mass fraction
(kg mol1)
absolute pressure of mixing gas (Pa)
temperature of mixing gas (K)
total pressure of mixing gas (Pa)
static pressure of mixing gas (Pa)
gas constant, R = 8.314 (J k1 mol1)
ejector throat area (m2)
operating pressure (Pa)
back pressure (Pa)
liquid fuel density corrected by mass fraction (kg m3)
density of mixing gas (kg m3)
flow velocity of mixing gas (m s1)

studies have focused on ejectors used in refrigeration systems, and

to our knowledge, no work has been reported on the adaptive air
distribution characteristics of ejector burners.
In this work, a CFD simulation method was used in this study to
investigate the adaptive air distribution characteristics of an ejector burner with varying geometrical parameters. As the air distribution characteristics are determined by its geometrical
parameters and operating parameters, the effects of four key
parameters on MER were studied, including the fuel nozzle exit
location (NEL), the ratio a of the ejector throat diameter to the fuel
nozzle exit diameter, the operating pressure and the ejector back
pressure. The structure of the ejector burner was optimised to realize automatic adjustment of air supply when the burner load
changes in the range of 20120%. Finally, an experimental investigation was conducted to validate simulation results and the performance of the optimised ejector burner. This work could contribute
to the design of ejector burners with adaptive air distribution and
promote the practical application of methanol-mixed fuels.
2. Research methods
2.1. Physical structure
The ejector burner was designed on the basis of ejecting combustion technology. A small amount of heat coming from the combustion chamber vaporises the liquid methanol-mixed fuels in the
spiral vaporiser. The gaseous fuel is then fed to the fuel nozzle from
the fuel outlet and jetted into the suction chamber with a high
speed. Owing to the entrainment effect, a large amount of air could
be drawn into the suction chamber and then mixed with the gaseous fuel in the mixing chamber and the diffuser. Finally, combustion occurs after the mixture is sprayed into the combustion
chamber. Fig. 1 shows the geometrical structure of the ejector burner studied in this work. The ejector burner was designed to have a
maximum load of 100 kW. The preliminary dimensions are listed
in Table 1.
2.2. CFD models
2.2.1. Governing equations
In this study, computation was performed using the package of
FLUENT 6.3. The governing equations used in the numerical computation are momentum, energy, continuity, mass transport and

L
D

length (mm)
diameter (mm)

Greek

a
g
b

ratio of throat diameter to nozzle exit diameter

combustion efficiency
fuel mass fraction

Subscripts
f
pure fuel
k
air
h
mixing gas of air and fuel
m
mass
i
inlet
o
outlet
Abbreviations
MER
molar entrainment ratio
NEL
nozzle exit location

ke equations [28]. The expression of these governing equations

in a cylindrical coordinate system is given as follows:

@
@
@
quu r rqv u r
qwu
@x
@r
r@h






@  @ u
@ 
@ u
@  @ u

r Cu

Cu  su
Cu
@x
@x  r@r 
@r  r2 @h
@h

where u denotes different variables, Cu denotes generalised diffusion coefficient and Su denotes generalised source. Different values
of u, Cu and Su have different expressions [29,30].
2.2.2. Grids
As shown in Fig. 2, the grids of calculation model were divided
into two parts by the combustion chamber outlet, namely, the ejector burner domain and the combustion domain. The combustion
domain is an axial symmetric cuboid meshed into approximately
0.25 million hexahedral elements. The ejector burner domain
was meshed in the same manner (approximately 0.15 million elements). However, the size of grids in the ejector domain is relatively smaller than that in the combustion domain because of the
comparatively smaller dimension and faster flow velocity of gaseous fuel. A total of 0.4 million computing elements were used,
which was verified by a grid with a total of 0.76 million
elements. For two grids, the difference of static pressure at mixing
chamber exit is less than 0.05%, which illustrates that the grid with
0.4 million elements has a satisfying computational accuracy.
2.2.3. Solution methods
Based on practical operating conditions, fuel entrance and air
entrance were set as pressure inlet boundary condition, and combustion domain exit was set as pressure outlet boundary condition
in the computational model. The pressure in air entrance and combustion domain exit was 0 Pa (gauge pressure). The adiabatic
boundary condition was used for the walls. According to fuel composition, the mass fractions of methanol and propane at the fuel
entrance were set as 0.96 and 0.03, respectively. Nitrogen was used
to replace the nonreactive additive, and the mass fraction was 0.01
at the fuel entrance. The mass fraction of oxygen was set as 0.23 in
the air entrance. The initial temperature was 450 K at the fuel
entrance.
The standard ke turbulence model was adopted in this work,
and the constants used in this model were C1e = 1.44, C2e = 1.9,
rk = 1.0 and re = 1.2 [29,30]. Segregated solver was used in the

315

J. Shi et al. / Fuel 162 (2015) 313322

Fig. 1. Geometrical structure of ejector burner for methanol-mixed fuels, (a) is 3-D geometrical structure, and (b) is structural diagram. 1 Suction chamber; 2 Ejector
mixing chamber; 3 Ejector diffuser; 4 Contraction section; 5 Combustion chamber.

Table 1
Preliminary and determined geometrical parameters of ejector burner.
Items

Preliminary
value (mm)

Determined
value (mm)

Items

Preliminary
value (mm)

Determined
value (mm)

Nozzle exit diameter DJ

Ejector throat diameter DH
Diffuser outlet diameter DD
Ejector exit diameter DP
Combustion chamber diameter DC

4.2
30
46
36
106

4.5
24
46
36
136

Suction chamber length LX

Mixing chamber length LH
Diffuser length LD
Contraction section length LP
Combustion chamber length LC

15
140
140
20
150

12
100
100
20
200

calculation, and SIMPLE algorithm [28] was employed to couple

the pressure and velocity with a relaxation factor of 0.7 for an easy
convergence. The species transport model was also selected to
handle the mixing of gaseous fuel and air. Combustion simulation
was conducted using the volume reaction, where the finite-rate/
eddy-dissipation model was selected to deal with the interaction
of turbulence and chemistry reactions. The P1 radiation model
was used to obtain accurate temperature distribution.

2.3. Experimental system and methods

2.3.1. Experimental system
The schematic of the experimental system used in this work is
shown in Fig. 3. The optimised burner was installed at the front of a
furnace. A liquid flow meter, a thermocouple, a pressure gauge and
a pitot tube were used to monitor the volume flow Qv, the temperature Th, the pressure Ph and the flow velocity v of the mixing gas,

316

J. Shi et al. / Fuel 162 (2015) 313322

Fig. 2. The grid structure of computational model in this study.

respectively. A portable gas analyser was installed at the furnace

exit to analyse the flue gas composition, and a draft fan was also
used to connect the furnace and a chimney to keep the furnace
in a slightly negative pressure. An electric heater was used to
vaporise the liquid fuel at the beginning, which would be shut
off after the burner went into normal operation.
2.3.2. Experimental analysis methods

Nk
MER
Nf

AU 2 BU C 0
2

Molar entrainment ratio (MER) defined as Eq. (2) is the ratio

between the molar amount of ejected air Nk and the molar amount
of fuel Nf, and it represents the capability of the ejector burner to
adjust the air distribution. This study proposed an experimental
method to measure MER of the ejector burner indirectly. Based
on the definition of MER, it could be calculated according to Eq.
(3) in which a variable U was used to denote MER. The fuel molar
flow Nf was calculated according to Eq. (4) by measuring the volume flow of liquid fuel Qv. The mixing gas molar flow Nh could
be obtained following Eq. (5) based on the gas state equation.
The volume flow of mixing gas Qh was calculated as Eq. (6) in
which the flow velocity of mixing gas v was monitored by using
a pitot tube and computed via Eq. (7). Finally, an equation for the
calculation of mixing gas density qh was provided as Eq. (8).

Nh  Nf
Nf
Q v qf
Nf
Mf
Ph Q h
Nh
RT h
Qh vS
1
Pjy qh v 2 Pdy
2
Q q Nf UMk
qh v f
Qh
U

pressure of mixing gas in the mixing chamber; Th is the temperature

of mixing gas; R is the gas constant; S is the area of ejector mixing
chamber; Pdy is the total pressure of mixing gas; Pjy is the static
pressure of mixing gas; Mk is the molar mass of air.
According to the transformation of Eqs. (3)(8), the Eq. (9) can
be deduced. The solution of Eq. (9) is the final result of MER for
the ejector burner.

3
4
5

In Eq. (9), A Q v qf N f M k RT h , B Q v q2f RT h Q v qf N f M k RT h and

2

C Q 2v q2f RT h  2P h Pdy  Pjy S2 M f .

Combustion efficiency was calculated following Eq. (10) on the
basis of the analysis results of flue gas composition at the furnace
exit, particularly the percentage of carbon monoxide and hydrocarbon. Besides, the temperature of combustion flame was measured
by an armoured thermocouple installed in the furnace, as shown in
Fig. 3.

g1

COs Cx Hy s
CO2 l

10

where [CO]s is the percentage of carbon monoxide in the flue gas,

[CxHy]s is the percentage of hydrocarbon in the flue gas and [CO2]l
is the theoretically determined percentage of carbon dioxide in
the flue gas under the condition of complete combustion.
2.3.3. Operating conditions
The experiment was designed to study the automatic adjustment of air distribution in the optimised ejector burner when its
load was changed. Thus, six operating conditions were selected
with the load changing from 20% to 120%. The operating conditions
are summarised in Table 2.

6
7

3. Results and discussion

3.1. Effect of NEL

where qf is the liquid fuel density corrected by mass fraction; Mf is

the molar mass of fuel corrected by mass fraction; Ph is the absolute

NEL at the axial direction has a significant influence on the air

distribution of ejector burners. Therefore, based on the primary
structure of the ejector burner, this study investigated the

317

J. Shi et al. / Fuel 162 (2015) 313322

Fig. 3. Schematic diagram of combustion system with optimized burner.

Table 2
Operating conditions for experiment (gauge pressure).
Load

20%

40%

60%

80%

100%

120%

Operating pressure P (MPa)

Mass flow rate Qm (kg s1)

0.008
1.08  103

0.015
2.16  103

0.04
3.24  103

0.065
4.32  103

0.08
5.43  103

0.1
6.52  103

variation of MER when the nozzle exit was located at different

positions with the load of ejector burner being 20%, 50%, 100%
and 120%. The back pressure of ejector was set to be standard
atmosphere pressure. To illustrate the movement of NEL in the
horizontal coordinate, a zero point was determined at the suction
chamber entrance, as shown in Fig. 1b. The influence of NEL on
the air distribution of ejector burner was studied by varying NEL
from 5 mm (the left side of the suction chamber entrance) to
22 mm (the right side of the mixing chamber entrance).
As shown in Fig. 4, MER remains almost unchanged when NEL is
located in the segment from 1 mm to 12 mm, but it is fluctuant
when NEL is located in the segment of 5 mm to 1 mm and
17 mm to 22 mm. Many previous references [13,14] have reported
that a large amount of air could be drawn into the fuel jet flow by
entrainment effect. Additionally, this work finds that the suction
effect of negative pressure in the mixing chamber also plays an
important role in drawing air. As shown in Fig. 5a, as NEL is equal
to 5 mm, the velocity attenuation of fuel jet is fast, which means
that a large amount of air is drawn into fuel jet and the entrainment effect is strong, and Fig. 5d illustrates that the static pressure
in the mixing chamber is high, which implies that the suction
effect of mixing chamber is weak and little air is inhaled. On the
contrary, as NEL is equal to 22 mm, Fig. 5c shows that the fuel
jet diffuses slowly and Fig. 5f shows that the static pressure in
the mixing chamber is low, which implies that the entrainment

effect of fuel jet flow is weak, but the suction effect of mixing
chamber is enhanced. Therefore, two factors mentioned above
result in an unstable amount of drawing air and fluctuant MER
when NEL is located at the positions outside of ejector or in the
mixing chamber of ejector.
As shown in Fig. 5b, as NEL is equal to 8 mm, the stable and
homogeneous air flow field in the suction chamber makes a stable
amount of air drawn by the entrainment effect, and Fig. 5e shows
that the static pressure of mixing chamber is reasonable, which
indicates that the amount of air drawn by the suction effect is also
stable. So NEL locating in the suction chamber can make MER to
keep stable, and the position of middle of suction chamber
(NEL = 7 mm) was selected to conduct these studies in following
parts.
3.2. Effect of ratio a
The influence of ratio a of the ejector throat diameter to the
nozzle exit diameter on MER was studied as well. The ratio a
was varied from 4.29 to 11.43, and the load of burner was chosen
to be 20%, 50%, 100% and 120%. NEL was fixed at 7 mm according to
the conclusions in Section 3.1. The back pressure of ejector was set
to be standard atmosphere pressure. As shown in Fig. 6, MER
increases rapidly with an increase in ratio a when a is below 8.5,
but it keeps constant when a is higher than 8.5. As a high a leads

318

J. Shi et al. / Fuel 162 (2015) 313322

Fig. 4. Simulation result for MER with different NELs under the condition of 20%,
50%, 100% and 120% burner loads.

to a large throat area which is beneficial for the suction effect of

negative pressure in the mixing chamber. Thus, an increasing
MER is obtained before a reaches 8.5. However, as shown in
Fig. 7, a large amount of air increases the negative pressure of
the mixing chamber, which can weaken the suction effect of negative pressure. Therefore, the increasing rate of MER reduces with
an increase in the ratio a. Fig. 7 shows that the pressure is almost
up to zero when the ratio a is equal to 10 and 11.43, so MER almost
keeps constant when the ratio a is larger than 8.5.
3.3. Effect of operating pressure
Operating pressure determines air resistance in the ejector and
thus affects MER stability of the ejector burner. In this section, the
influence of operating pressure (gauge pressure) ranging from
0.01 MPa to 0.15 MPa on MER was investigated when the back
pressure of ejector was set to be standard atmosphere. Four

geometrical structures were selected to perform this study, with

the nozzle exit locating in the middle of suction chamber
(NEL = 7 mm) and the ratio a being 4.29, 5.71, 7.14 and 8.57.
Fig. 8 shows that MER has a downward tendency with an increase
in operating pressure for the four selected geometrical structures.
In this section, as the ejector structure was fixed, the negative
pressure in the mixing chamber is the key factor that affects
the value of MER. Fig. 9 shows that the increase in operating pressure reduces the static pressure in the mixing chamber, and the
decrease rate keeps unchanged. Thus, the suction effect should
be enhanced linearly. According to the gas flow characteristics in
a nozzle, the mass flow rate of gas fuel also should increase with
an increase in operating pressure. So if there is not some flow
resistance, the MER should keep a constant. However, for a high
operating pressure, a large amount of air has high resistance to
flow in an ejector with fixed structures. Thus, MER has some
decrease in the measured range of operating pressure. From the
discussion above, it is clear that low operating pressure is
beneficial for keeping a stable amount of drawing air.
3.4. Effect of ejector back pressure
The influence of ejector back pressure varying from 300 Pa to
300 Pa (gauge pressure) on MER was also investigated. The load of
burner was set as 20%, 50%, 100% and 120%, and the nozzle exit was
fixed at NEL = 7 mm. To obtain a reasonable MER, ratio a was set to
be 5.7 according to the results in Section 3.2. Fig. 10 shows that a
large ejector back pressure would result in a small MER, especially
when the burner has a low load. It can be explained from Fig. 11,
the static pressure both in the mixing chamber and diffuser raises
with an increase in back pressure of ejector, which causes a high
flow resistance and weakens the suction effect of mixing chamber.
As high flow resistance hinders the air flow in the ejector,
particularly under a low load, abnormal operation of ejector burner
occurs when the back pressure is higher than 200 Pa. The ejector
does not work normally because of the failure of mixed gas passing
through the ejector exit.

Fig. 5. Distribution of fuel velocity and static pressure under the condition of NEL = 5 mm, NEL = 8 mm and NEL = 22 mm.

J. Shi et al. / Fuel 162 (2015) 313322

Fig. 6. Simulation result for MER with different ratios a under the condition of 20%,
50%, 100% and 120% burner loads.

319

Fig. 9. Static pressure along the ejector centre with different operating pressures
and the ratio a of 7.14.

Fig. 7. Static pressure along the ejector centre with different ratios a under the
condition of 100% burner load.
Fig. 10. Influence of ejector back pressure on the MER with the burner load of 20%,
50%, 100% and 120%.

Fig. 8. Simulation result for MER with different operating pressures and the ratio a
of 4.29, 5.71, 7.14 and 8.57.
Fig. 11. Static pressure along the ejector centre with different back pressures under
the condition of 100% burner load.

3.5. Structure optimisation

As aforementioned, air distribution significantly affects combustion efficiency. Thus, the geometrical structure of the primary
ejector burner needs to be optimised. According to simulation
results, firstly, NEL was selected at 0.5LX. Secondly, the operating

pressure of the ejector burner ranged from 0.01 MPa to 0.08 MPa
to reduce air resistance, and the fuel nozzle exit diameter was
recalculated to be DJ = 4.5 mm. Finally, ratio a was determined to
be 5.3 on the basis of the optimal MER for complete combustion,

320

J. Shi et al. / Fuel 162 (2015) 313322

Fig. 12. Comparison of experimental and simulation results under different load conditions. (a) MER, (b) combustion efficiency and (c) the temperature of combustion flame.

Fig. 13. The combustion flames of methanol-mixed fuel in the optimized ejector burner.

and the throat diameter of the ejector was set to be DH = 24 mm.

For a high drawing quality, the diameter of the combustion chamber was expanded to be DC = 136 mm to keep the low pressure
environment at the ejector exit. The optimal structure parameters
are summarised in Table 1. Based on the optimised structure
parameters, the proportions of every part were obtained as
follows: DX: DD: DP: DC = 2: 1: 1.9: 1.5: 5.7, LH: LD: LP: LC = 5: 5: 1:
10, and the length of suction chamber LX = 0.5DH.
A numeral simulation was conducted to validate optimal structure with burner load being set as 20%, 40%, 60%, 80%, 100% and
120%. As shown in Fig. 12a, results show that the simulated MER

is highly consistent with that of complete combustion. An

insignificant downward tendency of MER is found because of the
increasing air resistance in the ejector, but the changing rate is
below 2.3%. Under all simulation conditions, the combustion efficiencies are higher than 99.9% (Fig. 12b), which were calculated
via the Eq. (11). The highest temperature is approximately
2000 K (Fig. 12c), and the combustion of methanol-mixed fuels is
stable in the wide range of load.

bi  bo
 100%
bi

11

J. Shi et al. / Fuel 162 (2015) 313322

where g is the combustion efficiency. bi is the fuel mass fraction at

the burner inlet, and bo is the mass fraction of the unreacted fuel at
the exit of the computational domain.
3.6. Experimental validation
Comparison of experimental and simulation results is shown in
Fig. 12a. The experimental MER is in reasonable agreement with
that of the simulation, although the value of the former is slightly
higher than that of the latter. The changing rate of experimental
MER is less than 6.4% in the measured load range. Although
approximately 4% higher than the numerical simulation results,
the experimental MER could still ensure that the burner works efficiently and stably, and the combustion efficiency is above 99.2% at
the tested points. Thus, MER changing rate is acceptable.
Comparation of combustion efficiencies obtained in experiment
and simulation is shown in Fig. 12b. The combustion efficiency
acquired in experiment is in good agreement with that of numerical study. The experimental combustion efficiency insignificantly
declines, but it can still maintain a value above 99.2%. The main
reason is that the experimental MER decreases with an increase
in burner load, which leads to a slight deviation of the best air
supply of the burner.
The photograph of combustion flames was captured and shown
in Fig. 13. The combustion flame can keep horizontal at all measured conditions. As the premix and backflow of high temperature
gas in the combustion chamber, the efficiency and stability of combustion has been improved. The temperature of combustion flames
was measured. The comparations of temperatures measured in
experiments and those obtained in simulations were shown in
Fig. 12c. As a result of increasing combustion intensity, the temperature of combustion flame increases with an increase in burner
load when the load is below 40%. When the load is higher than
40%, the flame temperature could maintain at approximately
1650 K. The experimentally measured temperatures are lower than
those in numerical simulation. The reason is that adiabatic boundary condition was used in numerical studies, but the heat loss of
combustion chamber was inescapable. Although the distinction
for measured temperature exists, the combustion temperature
have been improved, which is beneficial to promoting the practical
application of methanol-mixed fuels.
4. Conclusions
To study adaptive air distribution in an ejector burner and
improve the combustion efficiency and stability of methanolmixed fuels, numerical simulation and structure optimisation were
conducted. The adaptive air distribution and combustion characteristic were also experimentally studied with the load changing
in a wide range to verify simulation results.
Simulation results show that the suction effect of negative pressure in the mixing chamber and the entrainment effect of fuel jet
flow both play important roles for an ejector burner to draw air.
As the suction effect and entrainment effect are stable when the
nozzle exit is located in the suction chamber at the axial direction,
MER keep unchanged, and the positions of nozzle exit locating in
the suction chamber are beneficial to facilitate stable air
distribution. A defined a (the ratio of the ejector throat diameter
to the fuel nozzle exit diameter) has an important effect on the
drawing quality of the ejector burner. MER rapidly increases with
an increase in ratio a as a result of the increasing air flow area.
However, a large amount of air increases the negative pressure of
the mixing chamber and weakens the suction effect of negative
pressure in the mixing chamber. So MER keep steady when a is
higher than 8.5. A high operating pressure can form a low negative

321

pressure in mixing chamber and enhance the suction effect, so the

amount of drawing air can increase with an increase in mass flow
rate of fuel gas lineally. An increase in operating pressure can
reduce MER slightly, because of the increasing flow resistance.
Therefore, low operating pressure is beneficial for keeping a stable
amount of drawing air. The ejector back pressure is also a key
parameter for the ejector burner to work normally, and a high back
pressure can raise the pressure in the mixing chamber and ejector
diffuser, which would weaken suction effect and hinder air flow.
For the optimised ejector burner, the experiment indicates that
the changing rate of MER is less than 6.4% in the burner load range
of 20120%, which is approximately 4% higher than that in numerical simulation. But it can be found the optimised burner could distribute air supply automatically. The combustion efficiency
measured in experiment is lower than that in simulation, but its
value is still above 99.2% when the changing rate of MER is 6.4%
under the condition of 120% load. In the experiment, air supply
can facilitate the complete combustion of this fuel, and the change
of MER is acceptable. The combustion of this fuel in the optimised
burner is stable and these experimental flame photographs imply
that the combustion efficiency and stability are improved. As the
heat loss of combustion chamber, temperature of combustion
flame measured in the experiment is 1650 K when the burner load
is higher than 40%, which is lower than that obtained in simulation
with the adiabatic boundary condition. But the combustion temperature still has some improvement, which is beneficial to promoting the practical application of methanol-mixed fuels.
Acknowledgments
Experiments were conducted in the Low-grade Energy
Technology and System Laboratory (Chongqing University). The
authors gratefully acknowledge the financial support by the Key
Natural Science Foundation of Chongqing (cstc2013jjB90003,
cstc2013jcsf90001).
References
[1] Balat M, Balat H. Recent trends in global production and utilization of bioethanol fuel. Appl Energy 2009;86:227382.
[2] Demirbas A. Bio-fuels sources, bio-fuel policy, bio-fuel economy and global
bio-fuel projections. Energy Convers Manage 2008;49:210616.
[3] Chum HL, Overend RP. Biomass and renewable fuels. Fuel Process Technol
2001;71:18795.
[4] Kamarudin SK, Shamsul NS, Ghani JA, Chia SK, Liew HS, Samsudin AS.
Production of methanol from biomass waste via pyrolysis. Bioresour Technol
2013;129:4638.
[5] Shi J. Investigation on auto-adaptive air distribution characteristics and
structure optimization of ejecting burner with biomass methanol
fuel. Chongqing: Chongqing University; 2013.
[6] Li XC, Wang T, Day B. Numerical analysis of the performance of thermal ejector
in steam evaporator. Appl Therm Eng 2010;30:270817.
[7] Chunnanond K, Aphornratana S. Ejectors: applications in refrigeration
technology. Renew Sust Energy Rev 2004;8:12955.
[8] Yang X, Long X, Yao X. Numerical investigation on the mixing process in a
steam ejector with different nozzle structures. Int J Therm Sci 2012;56:
95106.
[9] Varga S, Oliveira AC, Diaconu B. Numerical assessment of steam ejector
efficiencies using CFD. Int J Refrig 2009;32:120311.
[10] Ruangtrakoon N, Thongtip T, Aphornratana S, Sriveerakul T. CFD simulation on
the effect of primary nozzle geometries for steam ejector in refrigeration cycle.
Int J Therm Sci 2013;63:13345.
[11] Yan J, Cai WJ, Li YZ. Geometry parameters effect for air-cooled ejector cooling
systems with R134a refrigerant. Renew Energy 2012;46:15563.
[12] Pounds DA, Dong JM, Cheng P, Ma HB. Experimental investigation and
theoretical analysis of an ejector refrigeration system. Int J Therm Sci
2013;67:2009.
[13] Varga S, Oliveira AC, Diaconu B. Influence of geometrical factors on steam
ejector performancea numerical assessment. Int J Refrig 2009;32:1694701.
[14] Zhu YH, Cai WJ, Wen CY, Li YZ. Numerical investigation of geometry
parameters for design of high performance ejectors. Appl Therm Eng
2009;29:898905.
[15] Liu F, Groll EA, Li DQ. Investigation on performance of variable geometry
ejectors for CO2 refrigeration cycles. Energy 2012;45:82939.

322

J. Shi et al. / Fuel 162 (2015) 313322

[16] Pianthong K, Sheehanam W, Behnia M, Sriveerakul T, Aphornratana S.

Investigation and improvement of ejector refrigeration system using
computational fluid dynamics technique. Energy Convers Manage 2007;48:
255664.
[17] Chen WX, Chong DT, Yan JJ, Liu JP. Numerical optimization on the geometrical
factors of natural gas ejectors. Int J Therm Sci 2011;50:5541561.
[18] Li XC, Wang T, Day B. Numerical analysis of the performance of a thermal
ejector in a steam evaporator. Appl Therm Eng 2010;30:270817.
[19] Bartosiewicz Y, Aidoun Z, Mercadier Y. Numerical assessment of ejector
operation for refrigeration applications based on CFD. Appl Therm Eng
2006;26:60412.
[20] Chunnanond K, Aphornratana S. An experimental investigation of a steam
ejector refrigerator: the analysis of the pressure profile along the ejector. Appl
Therm Eng 2004;27:31122.
[21] Yan JJ, Shao SF, Liu JP, Zhang Z. Experiment and analysis on performance of
stean-driven jet injector for diatrict-heating system. Appl Therm Eng
2005;25:115367.
[22] Chong DT, Yan JJ, Wu GS, Liu JP. Structural optimization and experimental
investigation of supersonic ejectors for boosting low pressure natural gas. Appl
Therm Eng 2009;29:2799807.

[23] Opgenorth MJ, Sederstrom D, McDermott W, Lengsfeld CS. Maximizing

pressure recovery using lobed nozzle in a supersonic ejector. Appl Therm
Eng 2012;37:396402.
[24] Sriveerakul T, Aphornratana S, Chunnanond K. Performance prediction of
steam ejector using computational fluid dynamics: Part 2. Flow structure of a
steam ejector influenced by operating pressures and geometries. Int J Therm
Sci 2007;46:82333.
[25] Sriveerakul T, Aphornratana S, Chunnanond K. Performance prediction of
steam ejector using computational fluid dynamics: Part 1. Validation of the
CFD results. Int J Therm Sci 2007;46:81222.
[26] Singh G, Sundararajan T, Bhaskaran KA. Mixing and entrainment
characteristics of circular and noncircular confined jets. J Fluids Eng
2003;125:83542.
[27] Javier G, Jose M, Francisco C. A one dimensional model for the determination of
an ejector entrainment ratio. Int J Refrig 2012;35:77284.
[28] Fluent Manual, version 6.3.23. Fluent Inc; 2006.
[29] Leschziner MA, Rodi W. Computation of strongly swirling axisymmetric free
jets. AIAA J 1984;22.
[30] Sloan DG, Smith PJ, Smoot LD. Modeling of swirl in turbulent flow systems.
Prog Energy Combust Sci 1986;12:163247.