A New Combustion Methodology for Low Emission Gas Turbine Engines Arvind G. Rao1, Yeshayahou Levy2 1
Faculty of Aerospace Engineering Technische Universiteit Delft, 2629HS Delft, The Netherlands 2 Faculty of Aerospace Engineering Technion: - Israel Institute of Technology, Haifa 32000, Israel. E-mail address of the corresponding author [
[email protected]]
Abstract: In the present paper, a novel combustion methodology for gas turbine combustors operating in the flameless combustion (FC) regime is conceptualised. The methodology has been designed with an objective of reducing O2 concentration and combustion temperatures, thus creating an environment suitable for FC. Preliminary analysis shows that unlike in the conventional gas turbine combustors, transferring heat from primary combustion zone to secondary (annulus) cooling air can help in reducing O2 concentration in the reactants, increasing the concentration of H2O and CO2 in the reactants and in reducing the combustion temperature, thus reducing NOx formation further. The new proposed methodology with internal conjugate heat transfer is compared vis-à-vis to other existing schemes and the benefits are brought out explicitly. The proposed methodology, which calls for transfer of heat from the primary combustion zone to alternative air streams, will have a major impact on modern gas turbine combustors design. Keywords: Clean Energy, Flameless Combustion, Flameless Oxidation, Gas Turbine Combustion, Low NOx Combustion
I. INTRODUCTION Emissions produced by aircraft engines have evolved as one of the important sources of pollution with respect to the Green House gases, and hence leading to Global Warming. Since aircraft emit their pollutants in upper troposphere or lower layers of the stratosphere, the impact of these pollutants in deteriorating the environment is much more as compared to the land-based pollutants. The regulations will become more stringent in the near future, with the consequent need to reduce pollutants level drastically, especially NOx by more than 70% [1].
Some of the latest combustion Dry Low NOx Combustor (DLN) technologies developed for the land-based gas turbines includes; Lean Premixed (LP) for gas fired units and Lean Premixed Prevaporized Combustion (LPP) for liquid fuelled units. Additional methods are Staged Combustion, Rich burn Quick quench Lean burn (RQL), Catalytic Combustion and more. However, most of these technologies have met only limited success as far as their pollution control norms coupled with operational stability is concerned. Moreover, most of the abovementioned technologies are difficult to be ported to an aircraft jet engine. The inherent difficulties in applying these above-mentioned technologies have been discussed in detail by Arfi [2]. Many parameters are responsible for the formation of pollutants in a gas turbine combustor, the important being combustion process and operating conditions. Lefvebre [3] gives a comprehensive review of pollutants formed in a gas turbine combustor and their characteristics. As pointed out by many researchers, thermal NO is the main component within NOX pollution. Since temperature is responsible for production of NO, reducing combustion temperature and minimising hot spots is a crucial point to reduce NOx emission from gas turbines. Flameless Combustion (FC) is a recently discovered combustion regime that forms uniformly distributed combustion (and hence temperature) with a low visible flame and low combustion oscillations. The definition of FC and its characteristics are often shared by other types of combustion technologies that are
Rao and Levy
referred to in the literature as FLameless OXidation (FLOX®) [4], High Temperature Air Combustion (HiTAC) [5, 6], Moderate and Intense Low oxygen Dilution (MILD) combustion [7, 8], and heat-recirculating combustion [9]. The operating regime for flameless combustion is shown below in Fig. 1. Recirculation Ratio 4
16 8
2
1
0.5
0
s la m e
es la m
1500
H ot F
Flameless Combustion
hF
1200
600
Non Combustible zone
O xy
Ignition Boundary
- r ic
Auto Ignition Temp 900
N orm al C om bustion
T em p. of R eactants (K )
1800
L if te d fl ames
2100
300 0
3
6
9
12
15
18
21
24
27
30
79
76
73
70
%O2 in reactants 100
97
95
92
88
85
82
% Dilutants (N2+CO2+H2O)
Fig. 1: Different Regimes of Combustion
Figure 1 is an attempt to show different combustion regimes schematically in terms of O2 concentration and initial temperature of reactants. The ignition boundary defines the boundary below which combustion is not possible. It is found from Chemkin® simulations that auto-ignition temperature does depend on O2 concentration to some extent. It can be seen from Fig. 1 that the required initial temperature for ignition increases rapidly as the O2 concentration reduces in the reactants, however, it stabilizes when the reactant’s initial temperature approaches its auto-ignition temperature. When the O2 concentration is quite low (≈ 12% or below) the combustion is possible only if the reactants are above the auto-ignition temperature of the mixture. This unique combustion regime, which occupies top-left corner of Fig.1, is termed as Flameless Combustion. The FC is characterised by low flame temperature and distributed combustion, very different operational regime than the normal
2/11
combustion regime, and hence holds the key for future developments in combustion systems. The recirculation ratio for a typical furnace application is also shown in Fig.1 for reference. The high temperature and low O2 concentration in reactants is achieved by dilution of fuel air mixture with vitiated combustion gases. High temperature air enhances thermal field uniformity and flame stability, and produces smaller temperature gradients relative to conventional combustion [5]. The velocity of air in FC is generally very high as compared to that in a conventional combustion in order to create a recirculation zone as well as to assure premixing of fuel and air before combustion. Because of the high oxidizer temperature (above auto ignition temperature of fuel/air mixture) and high turbulence and mixing, the Damköhler number for the reactions are low and therefore the reactions are more stable as compared to a conventional gas turbine combustor. Hence, FC can be characterized by • Highly distributed combustion zone with uniform temperature, a prerequisite for reduced NOx production • Spontaneous auto-ignition when the fuel air mixture reaches above the autoignition temperature • Highly transparent flame with low thermo-acoustic oscillations • Recirculation of combustion products also increases the chemical reaction time and therefore lowers the Damköhler number, a required condition for practically obtaining FC in a combustor.
II. FLAMELESS COMBUSTION TURBINE COMBUSTOR
IN
GAS
Application of FC has been successfully demonstrated in industrial furnaces [4, 9]. In these furnaces, high oxidizer temperature is obtained by either preheating the inlet air using recuperation
Rao and Levy
the same. The NOx emission levels were very low. However, the design was not optimized and therefore the CO emission level was very high. Luckerath et al [12] have demonstrated flameless combustion in a straight flow cylindrical combustor, at high pressure (up to 30 bars). The combustor had 12 equally spaced circumferential nozzles with air and fuel injected from each nozzle at a high velocity. The NOx and CO levels were low. Nevertheless, the operating range is relative narrow and the heat density of the combustor is low. The basic mechanism used by Luckerath et al [12] was to shield the fuel stream with cold air until there is a mixing between recirculated combustion gases and the cold shielding air. Guoqiang et al [13] and Overman et al [14] have used similar methodology. It should be noted that since fuel is injected in such a way that fresh air is shielding the fuel, and because these systems operate under lean conditions, it is expected that O2 concentration within the combustion zone would be relatively high (>15%). Hence, it seems that such combustion systems do not fall strictly into the FC regime (of low O2 content) as depicted by Fig. 1., and therefore can be characterised as a premixed distributed lifted flame that is ignited when the mixture reaches its auto ignition temperature (due to mixing with recirculated gases). Even though such systems have shown lower emission level as compared to the conventional gas turbine combustors, such combustors have to operate near their lean burn out limit (LBO) for producing desirable combustion properties and therefore may fail to perform satisfactorily while meeting the dynamic requirements of an aircraft gas turbine engine.
or regeneration (to recover heat from the hot furnace exhaust gas), or by direct mixing of the fresh air with recirculated exhaust gases (internal recirculation) [4]. Investigations show that the thermodynamic process and operational parameters of gas turbine combustors are quite different from those of industrial furnaces operating in the FC mode. Gas turbine combustion chambers differ from industrial furnaces by being smaller, being adiabatic, by operating at elevated pressures and temperatures, and by maintaining a higher level of oxygen in the recirculated combustion products [10]. Several teams are working worldwide in design and development of gas turbine combustor operating in the FC regime. Levy et al [10, 11] have been working on a reversed flow combustor for small and medium size gas turbine engines. Schematic of the flow process for the prototype combustor, FLOXCOM, designed at the Technion - Israel Institute of Technology is shown in Fig. 2.
Fig. 2. Schematic of the Combustion Process in the FLOXCOM Combustor [11]
The FLOXCOM combustor was successful in achieving very low levels of NOx and was amongst the first to demonstrate application of FC to gas turbines. The fuel was injected after diluting fresh air with recirculated combustion gases. The combustor was able to achieve stable combustion even in very lean mixtures (air fuel ratio by mass from 17 to 80). The main principle was to dilute the incoming fresh air with the combustion gases before injecting fuel in
3/11
Rao and Levy
b) The operating temperature and pressure are high, hence increasing the chemical reaction rate c) The operational range of FC combustors is narrow d) Volume required by an FC combustor is large
III. CHALLENGES FOR FC IN GAS TURBINE APPLICATION Initial studies on FC / FLOX® were performed on non-adiabatic type combustion systems, like the industrial furnaces. The reduced temperature level due to heat extraction from within the furnace allows operation at near stoichiometric conditions with the consequence of very low O2 concentration in the combustion gases. In such systems, the flame temperature can be controlled by varying equivalence ratio, the preheated air temperature, the recycling rate of combustion gases and the amount of internal heat absorption. Contrary to this, in an adiabatic combustor (such as gas turbine combustor), the thermodynamic parameters of the recirculating and the exhaust gases are determined only by the aerodynamic and thermodynamic processes inside the combustor, thus giving less control parameters. The low O2 concentration within recirculated gases of an industrial furnace is favourable for FC. This increases the flame size and lift-off distance, thus forming a well-stirred distributed flame, and hence decreasing flame luminosity (increasing flame transparency). Achieving complete mixing between the fuel, fresh air, and recirculated combustion products is critical for proper operation in the FC regime, especially for gas turbine applications. Higher O2 concentration in the combustion products requires higher recirculation ratios to dilute the oxidizer in order to form the extended reaction zones required to achieve the FC mode. Thus, innovations are required in preheating fresh incoming air by mixing it with the combustion products. Hence, some of the main difficulties in applying the FC principles to a gas turbine engine can be summarized as: a) High recirculation rates are required to lower the O2 concentration in the reaction zone
IV. A NEW METHODOLOGY FOR GAS TURBINE COMBUSTORS As described earlier, the key for achieving FC mode in gas turbines is to have a relatively high temperature (higher than auto-ignition temperature) and low O2 concentration within the reactants, just prior to their combustion. In order to meet the above-mentioned stringent demands, a novel combustion methodology / scheme is proposed herein, schematically shown in Fig. 3. The junctions have been marked in diamonds whereas the stations are marked in ellipses. Every junction has inlet or exit stations. The combustor inlet air (from the compressor exit) is split into primary air and secondary air at junction “1”. The primary air is then injected into the combustor at junction “2”, where it is mixed with the hot recirculated combustion products. The main combustion takes place in the path between junction “2” & “3”. At junction “3”, the combustion gases are split into two streams, one is directed to the recirculation zone, while the other towards the exhaust (where it is mixed with the secondary air). In the recirculation zone, combustion products drawn from junction “3” are cooled to some extent because of heat transfer from recirculation zone to the secondary air stream. The fuel is added within the recirculation zone at junction “4” because the O2 concentration within this zone is very low, thus the fuel and the recirculated combustion gases can mix properly. Because of the low O2 concentration, only a certain part of the fuel added gets burnt in the recirculation zone (between junction “4” & “2”,
4/11
Rao and Levy
“4” & “2”) of the recirculation zone and partially in the main combustion zone (between junction “2” & “3”). Thus limiting the maximum temperature rise below the critical temperature (≅ 2000K), above which thermal NO formation would increase exponentially. • The distribution of combustion within the precombustion and the main combustion zones makes the scheme a virtually staged combustor without physically separating the combustion zones or adding an extra combustor. Even though, conventional gas turbine combustors are adiabatic in global term of no heat transfer from the combustor to the surroundings, it is found from the present exercise that transferring heat from the main combustion zone to the secondary air has many benefits, as illustrated from the analysis in subsequent sections.
precombustion zone), while rest of the fuel is burnt in the main combustion zone between junction “2” & “3”. The secondary air is heated due to the heat transferred from the recirculation zone. The heated secondary air is mixed with part of the high temperature combustion products at the exit (junction “5”) before exiting the combustor. The ratio of fuel burnt in precombustion zone (between junction “4” & “2”) is dictated by: I) the temperature and O2 concentration at junction “4”, which will determine the chemical reaction time, II) the average velocity and the physical distance between junction “4” & “2”, which will determine residence. Thus by adjusting these two factors, the quantity of fuel burnt in the precombustion zone can be controlled.
Primary Air
(α)
21
2
23
24 Inlet
Recirculation Zone
12 1
Main Combustion (1-β)
(β) Pre-combustion
15
42
32
x
Junction
xx
Station
3
35
34
4
Exit
V. THE THERMODYNAMIC MODEL
51
43
(1-γ)
53 5
Fuel (γ)
(1-α)
The governing equations for the abovedescribed system (shown in Fig. 3) can be written by using the heat and mass balance equations at every junction in the system. Junction - 2 (mixing of fresh air and recirculated gases) In this junction, the recirculated gas (after precombustion) mix with the fresh primary air
Heat exchanger Secondary Air
Fig. 3: Schematic of the Proposed New FC Methodology
The salient features of this novel scheme as compared to the other proposed contemporary gas turbines combustors are: • Fuel is injected into the O2 deficient recirculation zone, thus creating an optimum environment in which the fuel and the recirculated combustion gases can be mixed. • Certain amount of heat is transferred from primary combustion zone to the secondary cooling air, thus reducing the overall temperatures in the primary combustion zone and hence limiting the NOx formation. • Even though the fuel is injected at one place, the energy from the fuel is added in two steps, partially in the precombustion region (between junction
m1⋅CP1⋅T1⋅α + (K⋅α⋅m1 +( K+1)⋅mf)⋅ CP24⋅T24 = CP23⋅T23⋅(K+1)⋅ (α⋅ma+ mf)
(1)
where, the recirculation ratio is defined by, K=
mR
(α ⋅ m1 + m f ) ;
(2)
the fraction of air going into the primary combustion zone is given by, α = m12/m1
(3)
and the mass of fuel burnt (for initial guess) is calculated by , mf =
m1 ⋅ C P , av ⋅ (T5 − T1 ) + m f ⋅ C P , f ⋅ (T5 − T f ) Qf (4)
5/11
Rao and Levy
where CP,av is the average specific heat at constant pressure for air in the temperature range (T5 –T1). The above equation has to be computed iteratively until convergence is achieved because the specific heat, CP, changes at every point in the cycle with the gases composition and temperature.
temperature and species concentration. The Method of Successive Substitution [16] is used to solve the governing equations and all variables are recalculated and successively substituted until satisfactory convergence is achieved and all the constraints are met. For the present analysis, air is considered to be composed of only four species, N2, O2, CO2 and H2O. The molar concentration of all four species is calculated at every junction. Only onestep global reaction is considered for calculating the species concentration. The CP for the gas mixture is evaluated by using the relationship given by Mc Bride
Line 2-3 (main combustion):
CP23⋅T23⋅(K+1)⋅ (α⋅ma+ mf) + (1-β)⋅ mf ⋅ Qf = CP32⋅T32⋅ (K+1) ⋅ (α⋅ma+ mf)
(5)
where β is the fuel fraction burnt during the precombustion (β⋅ mf) and the fraction of fuel burnt in the main combustion is given by (1-β)⋅ mf (the recirculated flow is fuel rich) Line 3-4 (heat extraction): CP34⋅T34⋅K⋅ (α⋅ma+ mf) - γ ⋅ Qex = CP43⋅T43⋅K⋅ (α⋅ma+ mf) ; where CP34 = CP32 ,& T34 = T32
[17],
(6)
Nsp
∑ Yi ⋅ C Pi
(11)
i =1
Where NSP is the number of species (four in this case) and Yi is the molar fraction of each specie. McBride [17] has given the variation of CP with temperature CP,i / ℜ = a1,i + a2,i⋅T + a3,i⋅(T)2 + a4,i⋅(T)3 + a5,i⋅(T)4
Junction - 4, (Fuel Injection):
K⋅ (α⋅ma + mf)⋅ CP43 ⋅ T43 + mf ⋅ CPf ⋅ (T42-Tf) = (K⋅α⋅ ma + (Κ+1)⋅ mf)⋅ CP42 ⋅T42 (7)
where CPf is the specific heat of fuel at const. pressure and Tf is the temperature of fuel Line 4-2 (pre-combustion & heat extraction):
(9)
Where a1i, a2i,….a5i, are the coefficients describing the polynomial fit of CPi for the ith specie.
(K⋅α⋅ ma + (Κ+1)⋅ mf) ⋅ CP42 ⋅T42 + β⋅ mf ⋅ Qf - (1-γ)⋅
Qex = (K⋅α⋅ ma + (Κ+1)⋅ mf) ⋅ CP43 ⋅T4
C P ,mix =
(8)
VI. RESULTS AND DISCUSSION
where β is the ratio of fuel burnt during precombustion and (1-γ) is the fraction of heat extracted from the recirculated combustion gases to the secondary cooling air Line 1-5 (heat absorption): Heat transfer from the recirculated combustion gases (stream 3 → 4 → 2) and by-pass line (stream1→5):
The methodology described above is used to study the thermodynamic characteristics of the new proposed scheme. The focus is to study the variation of temperature and specie concentrations (O2, CO2, N2 and H2O) at all junctions and stations. The above calculations are carried out for an inlet mass flow rate of 1kg/s. The mass flow ratio between primary and secondary air is dictated by the maximum permissible temperature within the combustor (for NOx limitation), which is limited at 2000K for this study. As elaborated earlier, the main objective is to have a more uniform temperature distribution throughout the combustor and to have a low O2 and higher CO2 concentration in the combustion zone. It has been proved by Arfi [2] that higher concentration of CO2 in the reactants decreases NO
(1−α) ⋅ m1⋅ C P1⋅T1 + Qex = (1- α) ⋅ m1⋅ C P51⋅T51
(9)
Junction - 5 (mixing between combustion gases and secondary air): (α⋅ m1 + mf)⋅ CP35 ⋅T35 + (1- α)⋅ m1⋅ C P51⋅T51 = (m1 (10) + mf)⋅ C P5⋅T5
where CP35 = CP32, & T35 = T32 It should be noted that in every equation presented above, the value of specific heat at constant pressure (CP) is calculated iteratively with respect to
6/11
Rao and Levy
conventional combustor (the numbering correspond to junctions in Fig. 3). There are two small temperature peaks for the proposed combustor, unlike a single large temperature peak in the conventional combustor. This is a proof that the new proposed scheme can reduce formation of NO by a large margin as compared to other contemporary combustion technologies. The variation of O2 concentration within the proposed combustor as compared to a conventional combustor is in Fig. 6. The proposed combustion methodology is much better than the conventional combustion because the mass fraction of the O2 before onset of the main combustion and precombustion is much less as compared to that of a conventional combustor.
formation substantially, e.g. PSR studies carried out by Arfi [2] show that total fixed Nitrogen (TFN) formation is reduced by two orders of magnitude when reactants are diluted with 10% CO2 for a stoichiometric mixture at 300 K and with a residence time of 100 ms. Following the previously described procedure, the temperatures and concentrations of O2 and CO2 at every location in the combustor with a heat extraction of 100 KJ, results are shown in Fig. 4. It should be noted that the analysis has been carried out with Methane (CH4) as the fuel, recirculation ratio of 1.0, and a precombustion ratio (β) of 0.25. The effect of various parameters like heat extraction, recirculation ratio, etc, on the proposed combustion methodology is described in the subsequent sections.
21 2 23 24
15
Station
2700
Fuel=CH4
T42 = 1785K O2 = 5.8% CO2 =10.5%
Fuel
42 4 43
53 5 51
T5 = 1400K O2 = 14.5% CO2=5.6%
2100
Exit Exit T51 = 765K O2 = 23% CO2=0.0%
3 1800
1500
4
Heat extraction = 100KJ
5
2
1200
Heat exchanger Secondary Air
Conventional combustor New combustor (main flow) New combustor (recirculation)
2400
3 35 34
T24 = 2000K O2 = 2.6% CO2=12.8%
Pre combustion
32
Junction
900
1
Fig. 4: Calculated Chemical Equilibrium Temperature and Specie Concentrations at various Junctions (K = 1, β = 0.25, γ = 0.5)
600 0.0 0
1
2
3
0.25
4
5
0.50
6
7
1.0
0.75
Fig. 5: Schematic representation of the temperature variation along combustor
It can be seen from Fig. 4 that while the maximum equilibrium temperature is limited to 2000K, the O2 concentration in recirculation zone is below 6%, and the CO2 concentration is more than 10%. Such combination of temperature and species concentration at the entrance to the combustion zone is suitable for FC and it is expected that the flame would be of distributed type. The schematic of temperature variation along the combustor for this case, as compared to a conventional gas turbine combustor, for the same inlet and outlet conditions is shown in Fig. 5. It can be seen that peak temperature and temperature gradient for the proposed combustion methodology are quite low as compared to that of a
25
1
15
2 10
on l Combusti Conventiona
20
O 2 Mass Fraction (% )
T1 = 600K O2 = 23% CO2=0.0%
12 1
Main Combustion
x xx
Temperature (K)
Primary Air
Inlet
T32 = 1879K O2 = 6.1% CO2 =11%
T23 = 1248K O2 = 12.7% CO2=6.5%
O2 conven Conventional combustor New combustor O2 Flox(main flow) New combustor (recirculation) O2 FLOX Recirculated
5
Mai n co mbu stio n
5
Pre 0
0 0.0
1
2 0.25
3
on busti com
4 0.50
3
4
5
6 0.75
7
8 1.0
Non-Dimensionalized Combustor Length
Fig. 6: Schematic representation of O2 variation
Effect of heat extraction The heat extraction from recirculation zone to the secondary air has a significant effect on the thermodynamics of the cycle
7/11
Rao and Levy
in terms of reducing O2 concentration and increasing the concentration of CO2 & H2O, thereby suppressing the formation of NO and creating an environment in which FC can be sustained. Heat extraction also enables to lower the primary air mass flow rate, thus increasing the equivalence ratio. This increase combustion stability an essential requirement for gas turbine engines, especially for aircraft applications. Increase in CO2 concentration reduces temperature rise within the combustion chamber because of its high heat capacity and hence reduces the thermal gradients. In addition, the kinetic effect of CO2 may favour recombination reactions of HCN species to N2, leading to lower NOx formation within the combustion system. Extraction of heat also enables more fuel to be burnt in the precombustion zone without exceeding the temperature limit. This results in a better temperature distribution throughout the combustor and a lower temperature difference between the two streams at junction 5 (exit), giving a better pattern factor at the combustor exit. The variation of CO2 and O2 concentration at various junctions with heat extraction is seen in Fig. 7.
reaction time. When compared to the physical mixing time, it lowers the Damköhler Number. The effect of K on mass fraction of CO2 and O2 at the various junctions within the combustor, with heat extraction of 100 kJ/kg is shown in Fig.8. It can be seen that for a recirculation ratio of more than 2, the mass fraction of CO2 is more than O2. The effect of recirculation ratio on the temperature distribution within the proposed combustor is shown in Fig.9. It is seen that the temperature gradients (T32-T23), (T24-T23), (T42-T34) within the combustor are reduced and the temperature become comparatively uniform throughout the entire combustor. This can reduce the number of hot spots and the consequent thermal fatigue of the combustor liner, thereby increasing the reliability of the entire combustion system. 18
CO2@23 PCO2_2, CO2@32 PCO2_3, CO2@24 PCO2_42, O2@23 PO2, PO3, O2@32 PO42, O2@24
Concentration (%, mass)
15
12
9
6
3
0 0.5
1
1.5
2
2.5
3
3.5
4
Recirculation Ratio
15
CO2 @24 O2 @23
Fig. 8: Effect of K on O2 and CO2 concentration (β = 0.25, QR = 100 kJ/kg, γ = 0.5)
CO2 @32 9
1200
T3-t2 (T32-T23)
CO2 @23
O2 @32
(T24-T23) T42-T2 1000
6 PO2, O 2 @23 O2 @24 PO3, O 2 @32 PO42, O 2 @24 PCO2_2, CO 2@23 CO PCO2_3, 2@32 CO PCO2_42, 2@24
3
0 0
30
60
90
120
150
Heat Extracted (kJ/kg of inlet air)
Fig. 7: Effect of Heat Extraction on O2 and CO2 concentration (K = 1, β = 0.25, γ = 0.5)
Temperature difference (K)
Mass Fraction (%, (%) Concentration
12
(T42-T34) T42-T3
800
600
400
200
0 0.5
Effect of Recirculation Ratio (K) The Eq. 2 defines the recirculation ratio. Increasing K increases the dilutants (CO2, H2O) concentration and reduces O2 in the main combustion zone. This slows the reaction and increases the chemical
1
1.5
2
2.5
3
3.5
Recirculation Ratio
Fig. 9: Effect of K on Temperature Gradients (β = 0.25, QR = 100 kJ/kg, γ = 0.5)
8/11
4
Rao and Levy
to its ejector action and will also reduce the Damköhler number. Since there is no need for anchoring the flame, there are no swirling loses. The mixing loses are also expected to be less as compared to a conventional combustor because the temperature difference between the primary and the secondary air is lower. Hence, it can be said that the proposed combustor exploits all facets of FC, i.e. low O2 concentration, high concentration of CO2 and H2O in the reactants, distributed combustion and even temperature distribution. One of the other main constraints in a gas turbine combustor is CO quenching at the cold walls due to excessive cooling or improper residence time. However, in the proposed combustor the CO quenching phenomena is less likely to occur because the average temperature within the combustor is higher than the auto-ignition temperature. In addition, having a dedicated specified physical recirculation zone in the combustor helps to increases the combustion stability substantially. The unique combination of these favourable characteristics is expected to give a new direction to the philosophy of gas turbine combustion. It should be noted that the combustor shown in Fig. 10 is generic in nature and can be ported to can / can – annular or an annular type of combustor.
A flameless gas turbine combustor A schematic of a gas turbine combustor operating on the proposed combustion methodology is conceptualised and illustrated in Fig. 10. The recirculation zones which are formed due to the ejector action of the high velocity inlet air is schematically shown in the figure. The fuel is injected within these toroidal recirculation zones. In the conventional gas turbine combustors, secondary air is typically fed and mixed with combustion gases over the combustor length for film cooling, thus prohibiting the formation of an oxygen deficient zone within the combustor. Hence, in order to establish FC, it is essential to replace film cooling by alternative a heat transfer mechanism for cooling the combustor liner, without air mass exchange. For this particular design, the heat transfer mechanism is shown schematically. However, in an actual combustor the longitudinal (along the combustor length) composite metallic fins or jet impingement cooling could be used to enhance the overall heat transfer coefficient. Preliminary analyses of the heat extraction process in the proposed combustor suggest that around 10 % of the total energy released can be transferred to the secondary air. Even though the use of longitudinal fins for heat transfer enhancement would cause disruption of flow in the primary and secondary streams, it is anticipated that the pressure drop would be of the same order as it is for the conventional combustor because of the following reasons: The total pressure loss in a conventional combustor is due to diffusion, mixing, swirling and friction. In the proposed combustor, it is not necessary to diffuse the compressed air to very low inlet velocities. In fact, higher air velocity will enhance the recirculation ratio due
Heat Transfer Surface
ar nd co Se
Fuel
ir ya
4 ion Zone Recirculat
Primary air
Combustion
2
1
Zone Recirculat ion
Se co nd ary
Mixing
3
5 Zone
Zone
4 air Fuel
Fig. 10: Schematic of a proposed FC gas turbine combustor
9/11
Rao and Levy
VII. CONCLUSIONS A new FC methodology for gas turbine combustors is presented. The new scheme employs heat transfer from recirculating gases to the secondary cooling air and hence, is “non-adiabatic” as far as the primary combustion zone is concerned. It is observed that heat transfer enables the designer to reduce O2 concentration substantially and helps to create an environment in which FC can be established. For the proposed combustor, the O2 concentration at various junctions within the combustor is much lower than in a conventional combustor, or any other contemporary concepts. Also CO2 and H2O concentration in the reactants are observed to be high. Hence, it is expected that the proposed combustor will be able to meet the stringent demands of lower emissions and increased reliability posed to the future gas turbine combustors. A practical design of a combustor operating on the newly proposed methodology is also presented. The proposed methodology presents an attempt to build a combustor operating on the FC mode for gas turbines while exploiting all the facets of FC, which include low O2 concentration, high concentration of CO2 and H2O in the reactants and an even temperature distribution. It is found that due to the internal heat extraction, the combustion can take place at nearer stoichiometric conditions and hence will exhibit superior stability characteristics, further enhanced because the combustor has a dedicated physical recirculation zone. This unique combination of favourable characteristics is expected to give a new insight to gas turbine combustion, which is poised for a major technological change.
VIII. NOMENCLATURE CP D
K m Qex Qf ℜ
recirculation ratio [-] mass flow rate [kg/s] heat extraction transfer rate [W/kg] calorific value of the fuel [MJ/kg] universal gas constant [ 8.3144 J⋅mol-1⋅K-1] T temperature [K] Greek scripts α mass flow ratio [-] β precombustion ratio [-] γ heat extraction ratio [-] Subscripts f fuel i ith species R refers to recirculated gases ACKNOWLEDGEMENT
The authors are thankful to the Israel Council for Higher Education, Govt. of Israel for supporting the research.
IX. REFERENCES [1].
[2].
[3]. [4].
[5].
[6].
[7].
[8].
[9].
specific heat at constant pressure [J/kg-K] diameter of duct [m]
[10].
10/11
Szodruch, J., “Technological Targets for Future Air Transportation in Europe”, Panel Presentations, ISABE 2005, Munich, September 2005. Arfi, P., Reduction of NOx Emissions from Gas Turbines using Internal Exhaust Gas Recirculation, PhD Thesis, Technion-Israel Institute of Technology, 2002. Lefebvre, A. H., Gas Turbine Combustion, 2nd ed., Taylor & Francis, Philadelphia, 1999. Wünning, J.A., and Wünning, J.G., “Flameless Oxidation to Reduce Thermal NO-Formation”, Prog. Energy Combust. Sci., Vol. 23 (1997), 8194. Katsuki, M., and Hasegawa, T., “The Science and technology of combustion in highly preheated air”, Proc. Combust. Inst. 27 (1998) 3135-3146. Tsuji, H., Gupta, A. K., Hasegawa, T., Katsuki, M., Kishimoto, K., and Morita, M., “High Temperature Air Combustion: from Energy Conservation to Pollution Reduction”, CRA Press, Boca Raton, (2003). Dally, B., Riesmeier, E., and Peters, N., “Effect of Fuel Mixture on Moderate and Intense Low Oxygen Dilution Combustion”, Combust. Flame 137 (2004) 418-431. Cavaliere, A., and Joannon, M. D., “Mild Combustion”, Prog. Energy Combustion Sci., 30 (2004), 329-366. Milani A., and Saponaro, A., “Diluted Combustion Technologies”, IFRF Combustion Journal, 2001. Levy. Y., “Principle and Practice of the Low NOx FLOXCOM Gas Turbine Combustor,”
Rao and Levy
[11].
[12].
[13].
[14].
[15].
[16]. [17].
Flameless Combustion Workshop, Lund, Sweden, June 2005.. Levy, Y., Sherbaum, V., and Arfi, P., “Basic Thermodynamics of FLOXCOM, the Low-NOx Gas Turbines Adiabatic Combustor,” Applied Thermal Engineering, Vol. 24 (2004), 1593-1605. Luckerath, R., Shutz, H., Noll, B., and Aigner, M., “Experimental Investigation of FLOX Combustion at High Pressure,” Flameless Combustion Workshop, Lund, Sweden, June 2005. Li, G., Gutmark, E. J., Overman, N., Cornwell, M., Stankovic, D., Fuchs, L and Vladimir, M., “Experimental Study of a Flameless Gas Turbine Combustor” GT2006-91051, ASME Turbo Expo 2006: Power for Land, Sea and Air, May, 2006, Barcelona, Spain. Overman, N., Cornwell, M., and Gutmark, E.J., “Application of Flameless Combustion in Gas Turbine Engines,” AIAA 2006-4920, 42nd Joint Propulsion Conference & Exhibit, California, July 2006. Levy. Y., Sherabaum, V., Erenburg. V., “Fundamentals of Low-NOx Gas Turbine Adiabatic Combustor,” GT 2005-68321, Proceedings of ASME Turbo Expo 2005. Stoecker, W. F., Design of Thermal Systems, McGraw–Hill, New York, 1989, pp. 111–126. McBride, J.B., “Coefficients for Calculating Thermodynamic and Transport Properties of Individual Species” NASA TM-4513, Oct. 1993.
11/11
