the influence of hydrogen on the combustion ... · additive to other traditional hydrocarbon fuels...
TRANSCRIPT
The Influence of Hydrogen on the Combustion
Characteristics of Lean Premix Swirl CH4/H2 Flames
Liu Xiaopei*, Chen Mingmin , Duan Dongxia, Zhang Hongwu and Enrico Gottardo
Shanghai Electric Gas Turbine Co.,Ltd, CHINA.
(E-mail: [email protected], [email protected] )
Institute of Engineering Thermophysics, Chinese Academy of Sciences, CHINA.
(E-mail: [email protected], [email protected])
Ansaldo Energia S.p.A., ITALY
(E-mail: [email protected])
ABSTRACT
Combustion characteristics of lean premix swirl CH4/H2 flames have been studied by
numerical simulation and experiment in a model combustor which was designed for natural
gas combustion. The volume concentration of hydrogen in the mixture was varied from 0 to
20%. The effects of hydrogen on the flow field structure was studied by numerical steady
state simulations. A cold condition numerical simulation showed that the variation of
hydrogen concentration has no obvious influence on the shape and strength of the CRZ
(central recirculation zone). In a hot condition numerical simulation however, the hydrogen
concentration changes have an influence on the temperature field, and hence, the hot flow
field does show considerable differences. The diameter of the CRZ decreases with the
increase of hydrogen concentration, while the recirculation becomes weaker. In addition to
that, the flame becomes shorter with the increase of hydrogen concentration, while the NOx
emissions increases. Based on the experimental results and frequency spectrum analysis, it
was found that thermoacoustic oscillations became stronger with the presence of hydrogen.
Yet, the frequencies of the thermoacoustic oscillations were hardly influenced by the
hydrogen.
Keywords: hydrogen; thermoacoustic; swirl.
1. Introduction
Higher combustibility associated with hydrogen has received increased attention as an
additive to other traditional hydrocarbon fuels for extending the lean combustion
flammability limits and achieving more stable combustion. Alternative fuels such as
SNG and biomass syngas are getting more and more attention recently. These fuels
typically contain a significant fraction of hydrogen and they could be further mixed
with natural gas in the pipelines before the point of use. In such cases the composition
of the fuels can be subject to fluctuations especially related to hydrogen concentration.
In order to develop an enhanced fuel flexible gas turbine combustor capable to
withstand a wider fuel composition range it is important to deeply understand the role
of hydrogen in the combustion of such hybrid fuels.
There are significant differences in physical and chemical characteristic between
traditional hydrocarbon fuel and hydrogen/hydrocarbon hybrid fuel. Unlike the
hydrocarbon fuel, most significant effects of hydrogen/hydrocarbon hybrid fuel are
more depended on hydrogen, such as flame speed, heat release ratio, adiabatic flame
temperature[1][2][3]. With the high reactivity of hydrogen, the burning velocity was
improved, and preventing local flame out [4][5]. The change of those micro features
resulted in the variations of flame shape, the position of flame center (center of heat
release), the area of the flame which will affect the mechanism of thermoacoustic
oscillation [6]. In the lean-premixed swirl combustion system, the recirculation play a
significant role in flame stability which was influenced by the addition of hydrogen
into methane [7][8].
According to the time delay model of thermoacoustic oscillation, the key parameter,
the phase between acoustics and combustion, was influenced by turbulent flame speed
which was affected by the parameters such as the density of the fuel, consumption
ratio and the area of flame. While the composition changing, those parameters
become different, finally resulted in the change of the phase between acoustics and
combustion which will enhance or weaken the oscillation [9][10]. On the other hand, the
phase was also influenced by the convection time and chemical time which was
highly depended on the fuel composition [11].
Among the reports and studies on methane flame characteristic, not much is known
about the characteristic of hydrogen/hydrocarbon hybrid fuel [12][13][14]. The most
issues with hydrogen/hydrocarbon hybrid fuel is associated with the significant
variation in their fuel compositions that changes the combustion characteristics such
as flame speed, heat release ratio, local fuel consumption rate and flame instability
mechanisms. Those variation play important role in macro phenomenon such as
flashback, NOX emissions, auto ignition. The objective of the research was to
investigate the role of hydrogen addition to methane fuel in lean-premixed swirl flame
in a model combustor which was designed for methane fuel. The volume
concentration of hydrogen in the mixture was varied from 0 to 20%. The detailed flow
field with different amount of hydrogen was analysed by numerical steady state
simulation. The role of different hydrogen concentration on the thermoacoustic
oscillation and emissions were examined by normal pressure experiment with the
same combustor.
2. Model combustor
The model combustor was mainly including radial swirler with the swirl number is
1.015, fuel injector, pilot nozzle, premixing section and flame tube as shown in Figure
1. The injected fuel from the injector was added to the combustion air from the
annular air inlet at the downstream of the flame tube, and then mixed and swirled in
the swirler as it is passed into the combustion zone.
Figure 1.The model combustor
3. Numerical simulation
3.1 Method and process of the simulation
The focus of the simulation is the detailed flow field in the flame tube, meanwhile the
combustor is a periodical symmetrical structure, so the calculation region was
simplified for reducing the computation load as shown in Figure 2. The angle of the
fan-shaped calculation region is 60°, moreover the mesh of fuel injector, swirler and
the pilot nozzle is refined.
Figure 2.The mesh of the calculation region
The size of the grid is significant important for the result of the simulation as if the
mesh is too coarse may result in an inaccurate outcome, if the size of the mesh is too
small will make the results difficult to converge and a waste of time. There are two
sets of mesh, 3.50 million and 7million, to be selected as the suitable mesh for
simulation. The axial velocity distribution along the radial direction of the two sets of
mesh at different axial position (the detail position as shown in Figure 3) was
compares as shown in Figure 4.
Figure 3.Detail position
The axial velocity along the radial direction of the two sets mesh keep almost the
same. From the perspective of time and computing resource, it is better to choose the
3.5 million mesh as the final mesh for simulation.
0.00 0.01 0.02 0.03 0.04 0.05 0.06-10
0
10
20
30
40
50
3.5 million mesh-L1
7.0 million mesh-L1
Axia
l velo
city(
m/s)
r(m) 0.00 0.01 0.02 0.03 0.04 0.05 0.06
-10
0
10
20
30
40
50
Axi
al ve
loci
ty(
m/s)
r(m)
3.5 million mesh-L2
7.0 million mesh-L2
(a)L1 (b) L2
0.00 0.02 0.04 0.06 0.08 0.10 0.12-20
-10
0
10
20
30
40
50 3.5 million mesh-L3
7.0 million mesh-L3
Axi
al v
elo
city(
m/s)
r(m)
(c)L3
Figure 4.The comparison of axial velocity of the two sets mesh
According to the applicable conditions and calculation precision of different turbulent
model and combustion model, finally the realized k-ε model was chosen as the
turbulence model, the combustion model was finite-rate/eddy-dissipation model for
prevention of the immediately ignition when the fuel mixed with the air.
3.2 The result and discussion of the numerical simulation
In this section results from numerical simulation of different amount of hydrogen flow
field characteristics are presented. The intention of the simulation is to examine the
role of hydrogen on the flow field and velocity profile. In the simulation, the
equivalent ratio is kept constant at 0.583 and, the air mass flowrate is kept constant at
266g/s. To change the composition of the fuel the concentration of hydrogen is
changed
As shown in Figure 5 for different hydrogen concentration, it can be observed that the
profile of the axial velocity along the radial direction at different axial position in cold
condition, at different hydrogen concentration, is almost the same, except for a little
difference at position L2. which is the outlet of the premix section, where the axial
velocity increase with the increase of hydrogen concentration. Those results can be
attributed to the increase of fuel volume flowrate due to the change of fuel
composition. When the air mass flowrate is 266g/s, and the equivalence ratio is the
0.583, increase the hydrogen volume concentration by 10%, the total volume flowrate
will increase about 0.4%. According to that, it can be found the variation in volume
flowrate is relatively small due to the change of fuel composition in the examined
conditions that can’t make significant difference in axial velocity profile.
0.0 0.1 0.2 0.3 0.4 0.5-10
0
10
20
30
40
50
Axi
al v
elo
city(
m/s)
r/R
XH2
=0%
XH2
=10%
XH2
=20%
0.0 0.1 0.2 0.3 0.4 0.5-10
0
10
20
30
40
50
XH2
=0%
XH2
=10%
XH2
=20%
Axia
l ve
loci
ty(
m/s)
r/R
(a)L1 (b)L2
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0-20
-10
0
10
20
30
40
50
XH2
=0%
XH2
=10%
XH2
=20%
Axi
al v
elo
city(
m/s)
r/R
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0-20
-10
0
10
20
30
40
Axi
al v
elo
city(
m/s)
r/R
XH2
=0%
XH2
=10%
XH2
=20%
(c)L3 (d)L4
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0-20
-10
0
10
20
30
Axia
l velo
city(
m/s)
r/R
XH2
=0%
XH2
=10%
XH2
=20%
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
-10
0
10
20
Axia
l ve
locity(
m/s)
r/R
XH2
=0%
XH2
=10%
XH2
=20%
(e)L5 (f)L6
Figure 5.The axial velocity along radial direction at different axial position in cold condition
Sometimes, it could use some parameters that related to recirculation such as the
maximum diameter of the CRZ (the central recirculation zone, the area enclosed by
axial velocity equal to zero), bmax, the length of the CRZ, L, the distance between the
axial position of bmax and swirler, lmax, and the quantity of recirculation of high
temperature product gas to characterize the flow characteristics in the flame tube.
Furthermore, the CRZ is very important for flame stability. The CRZ position in cold
condition is shown in Figure 6. From this figure, it can be seen that there is a little
difference in the bmax which is decrease with the increase of hydrogen concentration.
The phenomenon can be attributed to the small increase in volume flowrate due to the
change of fuel composition that make the increase of axial velocity which can
inhibited the expand of the CRZ.
0.15 0.20 0.25 0.30 0.350.00
0.01
0.02
0.03
0.04
0.05
Radia
l direct
ion p
ositio
n(
m)
Axial direction position(m)
XH2
=0
XH2
=10%
XH2
=20%
Figure 6.The CRZ position for different hydrogen concentration in cold condition
The axial velocity along the radial direction in hot condition at different axial position
is shown in Figure 7. The axial injection velocity increase with hydrogen addition, but
the back flow velocity decrease with hydrogen addition. And also, it can be seen the
difference between the 0 and 10% of hydrogen volume concentration is smaller than
the difference between the 10% and 20% of hydrogen volume concentration.
0.0 0.1 0.2 0.3 0.4 0.5-10
0
10
20
30
40
50
60
XH2
=0
XH2
=10%
XH2
=20%
Axia
l velo
city(
m/s)
r/R 0.0 0.1 0.2 0.3 0.4 0.5
-20
0
20
40
60
XH2
=0
XH2
=10%
XH2
=20%
Axia
l ve
locity(
m/s)
r/R
(a)L1 (b)L2
0.0 0.2 0.4 0.6 0.8 1.0-20
0
20
40
60
XH2
=0
XH2
=10%
XH2
=20%
Axia
l velo
city(
m/s)
r/R 0.0 0.2 0.4 0.6 0.8 1.0
-20
0
20
40
60
XH2
=0
XH2
=10%
XH2
=20%
Axia
l ve
locity(
m/s)
r/R
(c)L3 (d)L4
0.0 0.2 0.4 0.6 0.8 1.0-20
-10
0
10
20
30
40
50
60
70
XH2
=0
XH2
=10%
XH2
=20%
Axia
l velo
city(
m/s)
r/R 0.0 0.2 0.4 0.6 0.8 1.0
-20
0
20
40
60X
H2=0
XH2
=10%
XH2
=20%
Axi
al v
elo
city(
m/s)
r/R
(e)L5 (f)L6
Figure 7. The axial velocity along radial direction at different axial position in hot condition
The Figure 8 shows the CRZ position for different hydrogen concentration. The
hydrogen addition shifts the axial position of bmax which is indicated by the dotted line
in the figure to downstream. The bmax, which is nondimensionalized by dividing the
diameter of the flame tube, is 0.3812, 0.3803, 0.3792 respectively, at hydrogen
concentration 0, 10%, 20%. The bmax is decreased with the increase of hydrogen
concentration. The difference of the diameter in other position is more obvious. The
length of the CRZ increase with the increase of hydrogen concentration. The
temperature distribution in the combustion zone is shown in Figure 9, as the
temperature shown a little increase with the increase of hydrogen concentration. The
increase of the temperature makes the increase of axial injection velocity,
consequently the expansion of the CRZ is inhibited by the higher and higher axial
injection velocity. Due to the decrease of the back flow velocity and higher
temperature results in the reduction of the recirculation flow.
0.20 0.25 0.30 0.35 0.40 0.450.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
r/R
Axial direction position(m)
XH2
=0
XH2
=10%
XH2
=20%
Figure 8.The CRZ position for different concentration at different axial position
0.0 0.2 0.4 0.6 0.8 1.0
400
600
800
1000
1200
1400
1600
1800
Tem
pera
ture(
K)
r/R
XH2
=0
XH2
=10%
XH2
=20%
0.0 0.2 0.4 0.6 0.8 1.0
400
600
800
1000
1200
1400
1600
1800
Tem
pe
ratu
re(
K)
r/R
XH2
=0
XH2
=10%
XH2
=20%
(a)L4 (b)L5
Figure 9.Temperature along the radial direction at different position
4. Experiment study
4.1 Introduction of the test rig
An effective simulation method is still lacked for the thermoacoustic oscillation
especially for the intensity of oscillation which is a main problem that gas turbine
faced. In order to examine the role hydrogen in thermoacoustic oscillation, the
experiment was carried out.
As shown in Figure 10, the experimental station mainly consists of air system which
was supplied by the air compressors, fuel system which was supplied by gas cylinders,
exhaust system, cooling system, measurement system and the model combustor. The
fuels that from the different gas cylinders were depressurized, go through the mass
flow controller, mixing with each other in the mixer, finally, the hybrid fuel was
formed. There are two fuel lines in the head of the burner, the pilot fuel line that used
for ignition, the premix fuel line that supply the hybrid fuel as shown in Figure 11.
The mix process between the combustion air and hybrid fuel is consistent with that
described in the numerical simulation section. The model combustor used in
simulation and experiment is the same.
Figure 10.Atmospheric experimental station
Two dynamic pressure measuring points were arranged in the model combustor, onein
the flame tube and the other in the annular air inlet, and used for monitor the pressure
during the experiment process to analysis the characteristics of the thermoacoustic
oscillation. The type of dynamic pressure sensor is Kulite XCS-190(M)-15D. The flue
gas analyzer is Testo 350. A camera was put at the end of the combustor to monitor
the combustion process in the flame tube. All the measurement signals were
integrated in the collection cabinet for storage.
Figure 11.The model combustor in the station
4.2 Results and discussion of the experiment
The CRZ which acts as a source to stabilize the flame was influenced by the hydrogen
concentration. Some fundamental characteristics such flame speed, ignition delay
time were all changed due to the addition of hydrogen to methane. Those parameters
are all important for the thermoacoustic oscillation. The objective of the experiment
was to examine the role of hydrogen on thermoacoustic oscillation and NOX
emissions. The combustion air was supplied at 250℃,the mass flowrate was 160g/s.
The equivalence ratio was 0.583. The air conditions and equivalence ratio keep
unchanged during the experiment. The hydrogen volume concentration was varied
from 0 to 20%.
The intensity of the oscillation as affected by different amount of hydrogen addition to
the methane fuel is shown in Figure 12. A key parameters that decided the
characteristics of thermoacoustic oscillation is the phase between the pressure
oscillation and heat release rate oscillation. While changing the composition of the
hybrid fuel, the flow field as described in simulation section, the shape of the flame,
and the local dynamic of the flame were all changed, those changing can led to the
variation of the phase which can make the couple between the pressure oscillation and
heat release rate oscillation more and more strengthen results in the intensity in the
flame tube increase with the addition of hydrogen.
It can be seen that the intensity of oscillation in both flame tube and annular air inlet
increase with the increase of hydrogen concentration. Theoretically, the intensity of
the oscillation in the annular air inlet should keep unchanged because of the air
condition was keep unchanged. The strengthen phenomenon can be attributed to the
oscillation in the annular air inlet was influenced by the oscillation in flame tube. The
pressures oscillations in the flame tube could spread upstream across the unchocked
fuel nozzles.
0 5 10 15 200.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
flame tube annular air inlet
osc
illa
tion in
ten
sity(
%)
hydrogen volume concentration(%)
Figure 12.Hydrogen addition effects on intensity of the oscillation
The frequency in the flame tube and annular air inlet remain almost the same with the
change of hydrogen concentration as shown in Figure 13. The dynamics pressure
power spectrum with different hydrogen concentration is shown in Figure 14. There
are three evident peaks in the spectrum, the frequency of the peaks are 170Hz, 500Hz,
750Hz respectively. Those are the three first modal of the oscillation. The modal of
oscillation with different hydrogen concentration keep almost the same, because of
the frequency of the oscillation with all hydrogen concentration is around 170Hz. The
addition of hydrogen to the methane fuel has little influence on the frequency of the
oscillation.
0 5 10 15 20140
150
160
170
180
190
200
flame tubeannular air inlet
oscill
atio
n f
req
ue
ncy(
Hz)
hydrogen volume concentration(%)
Figure 13.Hydrogen addition effects on frequency of oscillation
0 200 400 600 800 1000
90
100
110
120
130
140
Dyn
am
ic p
ressu
re p
ow
er
sp
ectr
um
/dB
re
20
μP
a
frequency(Hz)
flame tube annular air inlet
0 200 400 600 800 1000
90
100
110
120
130
140
Dyna
mic
pre
ssu
re p
ow
er
sp
ectr
um
/dB
re
20
μP
a
frequency(Hz)
flame tube annular air inlet
0 200 400 600 800 1000
90
100
110
120
130
140
Dyn
am
ic p
ressu
re p
ow
er
sp
ectr
um
/dB
re
20μ
Pa
frequency(Hz)
flame tube annular air inlet
Figure 14.Dynamic pressure power spectrum with hydrogen concentration 0, 10%, 20%
respectively
The Figure 15 shows hydrogen addition effects on the emissions. The NOx emissions
increase with the increase of hydrogen concentration. It can be attributed to the
increase of temperature of the combustion zone induced by the addition of hydrogen.
The emission of CO remain very low during the experiment conditions.
0 5 10 15 200
2
4
6
8
10
12
14
16
NOX CO
hydrogen volume concentration(%)
NO
X(
ppm
@1
5%
O2)
0
2
4
6
8
10
12
14
16
C
O(
ppm
@1
5%
O2)
Figure 15.Hydrogen addition effects on emissions
5. Conclusions
The results of the simulation and experiment show:
In hot condition, with the addition of hydrogen the flow field became different, with
the increase, the diameter of the CRZ decrease, the length of the CRZ increase, the
axial injection velocity increase and the back flow velocity decrease. The net
dominate result is the quantity of the recirculation decrease and the intensity of the
recirculation was weakened. According to the results of the experiment, the addition
of hydrogen made the intensity of the oscillation enhanced which influenced the
oscillation in the annular air inlet. The frequency of the oscillation is almost not
influenced by the addition of hydrogen. Because of the increase of temperature of the
combustion zone induced by the addition of hydrogen, the NOX emission was
increased.
The move from 0% to 10% of hydrogen is less effective than the move from 10% to
20% indicating a nonlinear behavior.
Reference
[1] Zhang Q, Noble D R, Lieuwen T. Characterization of fuel composition effects in
H2/CO/CH4 mixtures upon lean blowout. Journal of Engineering for Gas
Turbines and Power, 129:688–694, 2007.
[2] Schefer R W. Hydrogen enrichment for improved lean flame stability [J].
International Journal of Hydrogen Energy, 2003, 28:1131-1141.
[3] Schefer R W, Wickall D M, Agrawal A K. Combustion of hydrogen-enriched
methane in lean premixed swirl stabilized burner. Proceedings of the Combustion
Institute 2002; 29:843–51.
[4] Gauducheau J L, Denet B, and Searby G. A numerical study of lean CH4/H2/air
premixed flames at high pressure. Combustion Science and Technology, 137:81–
99, 1998.
[5] Jackson G S, Sai R, Plaia J M, et al. Influence of H2 on the response of lean
premixed CH4 flames to high strained flows. Combustion and Flame, 132:503–
511, 2003.
[6] Wicksall D M, Agrawal A K. Acoustics measurements in a lean premixed
combustor operated on hydrogen/hydrocarbon fuel mixtures [J]. International
Journal of Hydrogen Energy, 2007, 32:1103-1112.
[7] Zhu S R, Acharya S. Effects of Hydrogen Addition on Swirl-Stabilized Flame
Properties GT2010-23686.
[8] Kim H S, Arghode V K, Linck M B. Hydrogen addition effects in a confined
swirl-stabilized methane-air flame. International journal of hydrogen energy 34
(2009)1054 – 1062.
[9] Lieuwen T C. Unsteady combustor physics. Cambridge University Press; 2012.
[10] Speth R L. Fundamental Studies in Hydrogen-Rich Combustion: Instability
Mechanisms and Dynamic Mode Selection[J]. Massachusetts Institute of
Technology, 2011.
[11] Lieuwen T, Mcdonell V, Petersen E, et al. Fuel Flexibility Influences on
Premixed Combustor Blowout, Flashback, Autoignition, and Stability[J]. Asme
Turbo Expo Power for Land Sea & Air, 2008, 130(1):601-615.
[12] Lorenzo F, Lee J G, Bryan D Q, et al. The effects of fuel composition on flame
structure and combustion dynamics in a lean premixed combustor [C]. ASME
Conference Proceedings, Montreal, Canada,2007.
[13] Taamallah S, Labry Z A, Shanbhogue S J, et al. Thermo-acoustic instabilities in
lean premixed swirl-stabilized combustion and their link to acoustically coupled
and decoupled flame macrostructures[J]. Proceedings of the Combustion
Institute, 2015, 35(3):3273-3282.
[14] Speth R L, Ghoniem A F. Using a strained flame model to collapse dynamic
mode data in a swirl-stabilized syngas combustor. Proc Combust Inst
2009;32(2):2993–3000.