1. Trang chủ
  2. » Tất cả

CFD based exploration of the dry low NOx hydrogen micromix combustion technology at increased energy densities

10 6 0

Đang tải... (xem toàn văn)

THÔNG TIN TÀI LIỆU

Thông tin cơ bản

Định dạng
Số trang 10
Dung lượng 1,67 MB

Các công cụ chuyển đổi và chỉnh sửa cho tài liệu này

Nội dung

CFD based exploration of the dry low NOx hydrogen micromix combustion technology at increased energy densities Q2 Q1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31[.]

Trang 1

H O S T E D B Y

ORIGINAL ARTICLE

hydrogen micromix combustion technology

at increased

A Haj Ayeda,n, K Kusterera, H.H.-W Funkeb, J Keinzb, D Bohnc

Q1

a

B&B-AGEMA GmbH, Jülicher Str 338, Aachen 52070, Germany

b

FH Aachen University of Applied Sciences, Hohenstaufenallee 6, Aachen 52064, Germany

c

RWTH Aachen University, Templergraben 55, Aachen 52062, Germany

Received 21 May 2016; accepted 22 December 2016

KEYWORDS

Micromix combustion;

Hydrogen gas turbine;

Dry-low-NOx(DLN)

combustion;

Hydrogen combustion;

High hydrogen

combustion

Abstract Combined with the use of renewable energy sources for its production, hydrogen represents a possible alternative gas turbine fuel within future low emission power generation

Due to the large difference in the physical properties of hydrogen compared to other fuels such

as natural gas, well established gas turbine combustion systems cannot be directly applied for dry-low-NOx (DLN) hydrogen combustion Thus, the development of DLN combustion technologies is an essential and challenging task for the future of hydrogen fuelled gas turbines

The DLN micromix combustion principle for hydrogen fuel has been developed to significantly reduce NOx-emissions This combustion principle is based on cross-flow mixing

of air and gaseous hydrogen which reacts in multiple miniaturized diffusion-typeflames The major advantages of this combustion principle are the inherent safety againstflash-back and the low NOx-emissions due to a very short residence time of reactants in the flame region of the micro-flames

The micromix combustion technology has been already proven experimentally and numerically for pure hydrogen fuel operation at different energy density levels The aim of the present study is to analyze the influence of different geometry parameter variations on the flame structure and the NOxemission and to identify the most relevant design parameters, aiming to provide a physical understanding of the micromixflame sensitivity to the burner design and identify further optimization potential of this innovative combustion technology while increasing its energy density and making it mature enough for real gas turbine application

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112

http://ppr.buaa.edu.cn/

www.sciencedirect.com

Propulsion and Power Research

http://dx.doi.org/10.1016/j.jppr.2017.01.005

2212-540X & 2017 National Laboratory for Aeronautics and Astronautics Production and hosting by Elsevier B.V This is an open access article under the

CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).

n Corresponding author.

E-mail address: ayed@bub-agema.de (A Haj Ayed).

Peer review under responsibility of National Laboratory for Aeronautics and Astronautics, China.

Propulsion and Power Research ]]]];](]):]]]–]]]

Trang 2

The study reveals great optimization potential of the micromix combustion technology with respect to the DLN characteristics and gives insight into the impact of geometry modifications

onflame structure and NOxemission This allows to further increase the energy density of the micromix burners and to integrate this technology in industrial gas turbines

& 2017 National Laboratory for Aeronautics and Astronautics Production and hosting by Elsevier B.V.

This is an open access article under the CC BY-NC-ND license

( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).

1 Introduction

Aviation and power generation industry has need of efficient,

reliable, safe and low-pollution energy conversion systems in the

future Gas turbines will play a decisive role in long-term high

power application scenarios, and hydrogen has great potential as

renewable and sustainable energy source derived from wind- or

solar power and gasification of biomass substituting the limited

resources of fossil fuels[1] Hydrogen impacts the operation of

common gas turbine systems due to its high reactivity requiring

combustion chamber modifications to guarantee efficient, stable,

safe and low NOxcombustion Besides optimized combustion

technology and related exhaust gas emissions, modifications of

the gas turbine control and fuel metering system have to be

applied to guarantee safe, rapid and precise changes of the

engine power settings[2–7] Against this background the Gas

Turbine Section of the Department of Aerospace Engineering at

Aachen University of Applied Sciences (AcUAS) and

B&B-AGEMA GmbH work in the research field of low-emission

combustion chamber technologies for hydrogen gas turbines and

related topics investigating the complete system integration of

combustion chamber, fuel system, engine control software and

emission reduction technologies The hydrogen gas turbine

research at AcUAS started during the European projects

EQHHPP [8] and CRYOPLANE [9] where the low NOx

micromix hydrogen combustion principle was invented When

hydrogen is burned as fuel with air, only NOxemissions occur,

but Refs [2,3,10] and [11] have shown that the combustion

process has to be modified and optimized in order to achieve

low NOx emissions Because of the large difference in the

physical properties of hydrogen compared to other fuels such as

kerosene and natural gas, well established gas turbine

combus-tion systems cannot be directly applied for dry-low-NOx(DLN)

combustion Thus, the development of DLN hydrogen

combus-tion technologies is an essential and challenging task The DLN

micromix combustion principle for hydrogen is being developed

and optimized for years to significantly reduce NOx-emissions

by miniaturizing the combustion zone, reducing the residence

time of reactants in the combustion zone, and enhancing the

mixing process using a jet in cross-flow design A review of the

previous research activities at AcUAS is presented in Ref.[12]

Especially theflame anchoring – mostly dominated by the

resulting recirculation zones and vortices within the micromix

burner geometry[10]and by the momentumflux ratio of the

jet in cross-flow[11]– is most essential to the micromix low

NOx characteristics Based on previous investigations a

micromix combustion chamber with about 1600 miniature injectors (Fig 1) was designed for a small size Auxiliary Power Unit APU GTCP 36-300 and successfully tested[13] The GTCP 36-300 requires about 1.6 MW thermal energy converted to shaft power generating electrical and pneumatic power up to 335 kW The combustion section consists of an annular reverseflow combustion chamber in which the micromix combustor is to be integrated

The micromix hydrogen combustion research is done using

an interactive optimization cycle including experimental and numerical studies on test burners, full scale combustion

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103

Fig 1 Micromix prototype combustor for gas turbine Honeywell/Garrett Auxiliary Power Unit APU GTCP 36-300.

Fig 2 Interactive optimization cycle of micromix combustor research and development for APU GTCP 36-300.

Trang 3

chamber investigations and the feasibility is proven in real gas

turbine operation (Fig 2)

Based on these studies the impact of different geometric

parameters onflow field, flame structure and NOxformation

are identified and the micromix combustion principle is

continuously optimized

Within previous studies the influence of combustion

modeling and burner design parameters on flow field,

temperature distribution and flame structure has been

studied for a low energy density burner configuration

having a fuel injector diameter of 0.3 mm [14,15] The

study discussed in Ref [15]has shown potential to reduce

NOxemission of the burner by controlling lateral cool air

flows around the flame, which are established at given

geometric parameters Thefindings of Ref.[15]have been

applied to design a high energy density burner with an

injector diameter of 1 mm (increasing the heat rate per

injector by more than 11 times) This burner has been

analyzed numerically and experimentally in Ref [16]and

has proven low NOxability at increased energy density

Within the present study, the impact of further geometric

parameter variations of the high energy density micromix

burner (1 mm injector diameter) on itsflame structure and NO

emission is studied numerically in order to reveal possible

further optimization potential The main driver of this study is

the fact that the high energy density of the burners leads to

increased flame thickness and flame length These would

increase the peak temperatures and the residence time of

nitrogen and oxygen in the flame, which promotes NOx

formation Thus, the flames need to be optimized in shape

and position to minimize their NOx emissions (note that the

flame position decides on its interaction with neighboring

flames) Thereby, a step-by-step variation of two major

geometric dimensions is performed within a geometrically

feasible range 3D CFD simulations of the reactingflow have

been performed for the different micromix burner variations in

order to evaluate the resultingflow field, flame structure and

NO emissions and understand the influence of the single

parametric variations on the complex reactiveflow field of the

micromix burner The observations resulting from this study

will reveal optimization potentials of the micromix combustion

technology in terms of NO emissions, especially with regards

to increased energy densities

2 Micromix hydrogen combustion 2.1 Micromix description

Gaseous hydrogen is injected through miniaturized injec-tors perpendicularly into an air cross-flow through small air guiding panel (AGP) structures This leads to a fast and intense mixing, which takes place simultaneously to the combustion process As a result, miniaturized microflames develop and anchor at the burner segment edge downstream

of the injector nozzle Multiple micro flames instead of large scale flames lower the residence time of the NOx

forming reactants and consequently the averaged molar fraction of NOx can be reduced significantly as has been shown in Ref [6] The main influence on the low NOx

characteristic can be ascribed to the key design parameters blockage ratio BR of the air guiding panel AGP (Fig 3) and injection depth y of the fuel into the oxidizer cross-flow (Fig 3(b)) The blockage ratio BR is represents the ratio between the air guiding height and the height of the air guiding panel (AGP) (both indicated inFig 3)

The blockage ratio influences shape, position and size of the flame stabilizing vortices downstream of the air guiding panel

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 Fig 3 (a) Aerodynamic flame stabilization principle, and (b) hydrogen

injection depth de finition.

Nomenclature

A area (unit: mm2)

BR blockage ratio

d diameter/inner diameter (unit: mm2)

D outer diameter (unit: mm2)

ED energy density (unit: MW/(m2·bar))

ṁ massflow (unit: kg/s)

p pressure (unit: bar)

T temperature (unit: K)

Greek letters

Φ equivalence ratio

Subscripts

3 combustor inlet

4 combustor outlet AGP air guiding panel crit critical

fuel fuel/hydrogen

H2-seg hydrogen burner

Trang 4

and the burner segment The jet-in-cross-flow mixing of fuel

and air stabilizes the low NOx emission characteristics of the

combustion principle as long as the injection depth y (Fig 3(b))

is not penetrating the shear layer of the AGP-vortex (critical

injection depth ycrit) A recirculation of the fuel/air mixture into

the AGP-vortex leads to raised NOxemissions[13]

Within the present study, the air guiding height is kept

constant and the height of the air guiding panel is varied, as will

be explained inSection 3 The injection depth y is kept constant

2.2 Test burner configuration for numerical study

The presented numerical study investigates the

computa-tional model of the atmospheric test burner with a hydrogen

injector diameter of dH2¼1 mm, which has been presented by

the authors in Ref [16] This burner configuration was

established by increasing the energy density per fuel injector

to more than 11 times, compared to the first developed

micromix burners with an injector size of 0.3 mm, that have

been investigated by the authors in Ref.[15]

Experimental tests and numerical analyses have been

performed for the burner in question at different

equiva-lence ratios and are explained in Ref [16] Thereby, the

micromix flames were found very stable and well in

accordance with the typical micromix structure, despite of

the increased energy density Fig 4 shows the

experimen-tally observedflame structure at an equivalence ratio of 0.4

(design point)

A 3D movable exhaust gas probe is located behind the

combustor and extracts exhaust gas samples that are

supplied to the analysis modules of the continuous gas

analysis system ABB Advanced Optima AO2020 by heated

tubing designed to avoid concentration changes of the

different components within the exhaust gas sample and

condensation of water in the tubing that could influence the

analysis results The gas samples are directed through a gas

dehydrator to each analyzing module by heated tubes and

hoses under controlled pressure conditions The Advanced

Optima exhaust gas analysis system determines the amount

of unburned hydrogen (ABB Caldos 27) and the

concentra-tion of O2 (ABB Magnos 206) For the determination of

NOx (i.e NO and NO2), an Eco Physics CLD 700 EL is

used and directly connected to the hot exhaust gas sample

Internal hot tubing and particle filters in the device allow analyses without pre-processing of the gas sample and prevent water condensation The cross-sensitivity to the remaining water vapor in the sample is below 0.5% of the measured value The measurement accuracy is 70.1 ppm (applied measuring range 0–10 ppm) Before each test campaign, all exhaust gas analyzing devices are calibrated using zero-point calibration gases and defined reference-point calibration gases

The exhaust gas samples are extracted at different positions downstream of the burner The measured species concentra-tions at different posiconcentra-tions are averaged to get a representative exhaust gas composition and burner emissions

The corrected NOx emissions (@ 15% O2) obtained experimentally and numerically are given against the equivalence ratio in Fig 5 At the design equivalence ratio

ofΦ¼0.4 the NOxemission was measured to approx 2 and calculated to approx 1.4 ppmv @ 15% O2, which proves the low NOxability of the micromix test burner[16] This burner is considered as reference case within the present study

3 Parametric study and numerical exploration

3.1 Simulation approach For the numerical simulation of the different burner varia-tions a simplified numerical approach is applied It uses the 3D numerical simulation of the flow field within the test-burner based on a RANS solver, reduced combustion reaction mechanism and thermal NO formation models to analyze flow-field-structures, temperature distribution, tendencies of flame-anchoring, flame-structure and emission behavior In this section of the paper, the simplified numerical approach is described, and its application to calculate the reactive flow in the burners is presented The aim of the numerical analysis is to understand the basicflow phenomena and qualitatively identify tendencies of the different design parameter influences with respect toflow- and flame-structure, and resulting thermal NO emissions The emission calculation includes only thermal NO,

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 Fig 4 Optical flame appearance of established micromix flamelets at

design point Φ¼0.40–1 mm injector burner [16]

Fig 5 Measured and calculated NOx/NO emissions at different equivalence ratios for the 1 mm injector burner [16]

Trang 5

because it is a good and fast indicator of the burner

configuration emission behavior and very useful for the

numerical prediction of the test-burner emission characteristics

prior to testing Therefore, the calculated NO emissions are

expected to be generally slightly below the real values, but

provide an excellent qualitative evaluation possibility for the

intended numerical design exploration of the high energy

density micromix combustion technology

3.2 Computational domain

The numerical analysis has been carried out using a

commercial CFD code [17]and has been based on simplified

geometric models derived from the different burner con

figura-tions to be investigated The geometric model is shown inFig 6

and covers a longitudinal burner slice, which makes use of the

symmetric nature of the burner in both lateral and vertical

directions The symmetric boundaries along the lateral direction

are set on the cross section through the center of one hydrogen

injection hole and on the cross section between two hydrogen

injection holes, respectively Along the vertical direction the

symmetry planes are set on the center section through one air

guiding panel and on the center section through one hydrogen

segment Thus, the slice model contains one half of a hydrogen

injection hole and one half of an air guiding gate

3D steady RANS calculations have been performed The

realizable k, ε turbulence model with all yþ wall treatment

has been applied The wall treatment is decided depending

on the local dimensionless wall distance yþ values For

high yþ values the wall function approach is used For low

yþ values (below or not much larger than 1) no wall

function is required, since the boundary layer is well

discretized by the numerical mesh The hydrogen

combus-tion process has been simulated based on a reduced

hydrogen combustion reaction model including one step

hydrogen combustion reaction, where the reaction rate has

been calculated by the hybrid EBU combustion model

described in Ref.[11] This model combines the turbulent

mixing driven reaction rate and the chemical kinetic

reaction rate (finite chemistry) The turbulent mixing driven

reaction rate is calculated via the EBU (Eddy Break Up)

combustion model formulation, which assumes that

reac-tants are directly burnt after mixing The chemical kinetic

rate is calculated based on the Arrhenius formulation and considers the chemical time scale needed to burn reactants when they are fully mixed By application of the hybrid EBU approach both reaction rates (turbulent mixing driven rate and chemical kinetic rate) are calculated and compared

The smallest rate is assumed as reaction limiting The chemical reaction rate is calculated according to the following Arrhenius formulation:

r¼ ATnexp Ea

RT

H2

½ αO 2

The parameters of the Arrhenius formulation for the global reaction mechanism are selected in accordance to Fernández-Galisteo et al [18], where A¼2.05E14, n¼0,

Ea¼1.67E8, α¼2 and β¼0 The unit of the resulting reaction rate is kmol/(m3s)

By applying the reduced hydrogen combustion reaction model the calculation time is reduced significantly to 4 days per case and large number of parameter variations can be achieved within a reasonable numerical effort If detailed hydrogen combustion reaction mechanism was considered, the calculation time would exceed several weeks for the used calculation mesh The application of the reduced hydrogen combustion reaction model reduced combustion model is found reasonable and sufficiently accurate in quality and quantity as has been found in Ref.[16], especially in terms of predicted NOx emissions The use of a RANS solver with turbulence modeling might lead to deviations of calculated mixing and turbulence-chemistry interaction in terms of quantity However, the applied model allows a reasonable study of the qualitative behavior and trend of different micromix configurations, which helps advan-cing this technology within reasonable engineering time scales

The application of higherfidelity modeling, e.g with LES and detailed chemistry modeling, would be considered for validation and further study of the accurateflame structure in the future

Fig 5shows measured and calculated NOxemissions for the high energy density burner with an injector diameter of 1mm The calculated values are based on the reduced hydrogen combustion reaction model and show good agreement with the measured values

Thermal NO formation has been considered by applica-tion of the extended Zeldovich NO formaapplica-tion mechanism

A corresponding numerical model is provided by the applied CFD code Its activation adds NO to the transported species within the solution domain This allows the evaluation of NO distribution within the reaction zone and the full hot gas path as well as the evaluation of NO concentrations at the burner outlet boundary (calculate the

NO emission)

The spatial discretization has been performed using the STAR-CCMþsurface remesher and polyhedral mesher resulting in an unstructured polyhedral mesh

The polyhedral cell shape is especially advantageous as

it helps minimizing the total number of cells while maintaining mesh resolution quality and thus, helps saving calculation time and cost Progressive mesh

refinement has been performed along the reaction and

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 Fig 6 Computational domain.

Trang 6

hot gas zone starting from the hydrogen injection

surrounding There the smallest volume cell size has

been selected to get a sufficient resolution inside the

mixing and the reaction zone The refinement process

has been performed iteratively within a reference

calculation until a mesh independent solution could be

obtained The final mesh includes approx 900,000

volume cells in total

3.3 Boundary conditions

The fuel and the air jet are introduced separately into the

burner model via two inlet boundaries as shown inFig 7 The

inlet boundaries are set far enough from the air guiding panel

and the fuel injection hole in order to avoid any boundary

influence on the key flow phenomena in the mixing and

combustion regions No-slip wall boundaries represent the air

guiding panel and the hydrogen segment walls

Since contact with hot gas is limited to the front surface

of the H2segment and the surfaces surrounding the reaction

and exhaust gas zone are symmetry planes, heat transfer

from the hot gas into the burner wall has been neglected and

has not been considered within the numerical simulations

for all burner configurations

The air and fuel inlet parameters have been defined

according to the experimental conditions for the test burner

configuration The air inlet pressure is 1 bar according to the

test rig design The inlet air is preheated to 560 K to simulate

the APU inlet condition The fuel inlet temperature is 300 K

The fuel mass flow has been selected according to the

design operating point (Φ¼0.4)

3.4 Numerical results and parametric study

The micromix burning principle is characterized by distinct

reaction zones, anchoring near the edge of the H2segment and

stabilized by the inner and the outer vortex pairs as shown in

Fig 8 The inner vortex pair results from the air recirculation

downstream the air guiding panel after contraction in the air

gate The outer vortex pair is created by recirculating hot gas

downstream the H2 segment Due to the axial shift in the

position of the H2segment front face and the air guiding panel,

an inclined shear layer is established in-between the vortices

and combustion reaction takes place and is stabilized along this

inter-vortex shear layer

The structure and orientation of the micromix flame is depending on the structure of the mentioned shear layer, which is in turn defined by the size, position and intensity

of the stabilization vortices

Fig 9 shows the calculated temperature distribution (bottom part) and the calculated thermal NO mass fraction (top part) in the reference micromix burner (reference geometry) The micromixflames are clearly separated from each other and well anchored and stabilized according to the micromix burning principle

Looking to the temperature distribution, two peak tem-perature regions can be distinguished The first is found along the first flame fragment, which is stabilized in-between the inner and outer recirculation vortices along the inter-vortex shear layer This zone is thin, but shows a significant temperature gradient across the flame, which is typical for this kind offlames

The second peak temperature zone is found downstream

of the inter-vortex shear layer (as marked inFig 9) Here, the remaining fuel that was not burnt along the first flame fragment starts to burn and the last heat release of the injected fuel takes place In this zone, a higher peak

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 Fig 7 Computational domain, close up to fuel injection region.

Fig 8 Typical recirculation and vortex structure of the micromix burning principle.

Fig 9 Calculated temperature (bottom) and NO mass fraction (top) distributions for the reference burner (x ¼0, Δk¼0).

Trang 7

temperature is found and the high temperature zone is found

thicker than thefirst fragment, indicating a concentration of

heat release

Thisflame structure is not typical for the micromix burning

principle, which aims to burn all the injected fuel along the thin

inter-vortex shear layer and thus, avoid high fuel

concentra-tions, high temperature peaks and thus, reduce NOxemissions

This newflame structure is due to the increased energy

density of the considered burner (note that the injector size

of the burner in question is 1 mm, which leads to an 11

times higher energy density compared to the originally

invented burner, having an injector size of 0.3 mm) Since

the overall burner dimensions are not scaled with the same

factor as the injectors (due to combustor integration issues),

the length of the inter-vortex shear layer becomes not

sufficient to accomplish all the heat release (or to

accom-modate the wholeflame) The fuel that could not be burnt

along the shear layer starts to burn further downstream,

building the aforementioned secondflame fragment

The calculated NO mass fraction for the reference burner

(shown in the top part ofFig 9) reflects the flame structure

pretty well and clearly shows two distinct high NO zones:

inside thefirst flame fragment and inside the second flame

fragment Thereby, a clear NO mass fraction peak and

concentration is found in the second flame fragment This

means that the new flame structure, which is dividing the

flame into a “shear layer” and a “post shear layer” part, has

a negative influence on the NOx emission level of the

burner It is expected to reduce the burner's NOxemissions

by reducing the extent of the“post shear layer” part of the

flame This could be achieved by increasing the shear layer

length, so that more heat release can take place within the

thin“shear layer” part of the flame This could be achieved

by increasing the air guiding panel (AGP) height, which

increases the size of the small inner vortices and

conse-quently enlarges the inter-vortex shear layer

A further measure that could reduce NOxemissions of the

burner is to increase the mixing path length (way between

injection andflame anchoring) by shifting the flame anchoring

edge away from the fuel injector (seeFig 10)

Starting from the reference geometry, 2 different

para-meter variations have been studied according to the above

considerations:

1) Variation of the anchoring point distance from the fuel

injector: variation of the x parameter from x¼0 mm

(reference case) to x¼2 mm (variation X2) and x¼4 mm

(variation X4) This variation aims to increase the

distance of fuel and air mixing before the flame is

established, leading to lower peak flame temperature

along the flame anchoring path

2) Variation of the AGP (air guiding panel) height k by

Δk¼1 mm (variation K1), Δk¼2 mm (variation K2) and

Δk¼4 mm (variation K4) This variation aims to increase

the size of the inner recirculation vortex and thus, increase

the length of the shear layer between both stabilization

vortices and provide more space for the flame to burn without merging with neighboringflames

Figs 11 and 12 show the calculated temperature and thermal NO distribution for the burner variations X2 and X4 Two major phenomena can be observed:

– The NO mass fraction along the shear layer flame part and also inside the post shear layerflame part is reduced gradually from the reference case to larger x values

– The peak temperatures in both flame fragments are gradually reduced from the reference case to larger x values, despite of the shorter shear layer length

– The NO emission of the burners is gradually decreased from 1.38 ppm for the reference burner to 0.58 ppm for the X4 burner variation This is a significant NO emission reduction

The achieved reduction in NO emission is due to the reduction of peak temperature in bothflame fragments This

is due to the enhanced mixing (longer mixing path) between the fuel injection point and the flame anchoring point, leading to lower equivalence ratio peaks at theflame zones

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112

Fig 10 Variation of the burner's mixing distance by variation of the

x parameter.

Fig 11 Calculated temperature (bottom) and NO mass fraction (top) distributions for the burner variation X2 (x ¼2 mm, Δk¼0).

Trang 8

The study did not include further x value increase in

order to maintain the manufacturing and integration

feasi-bility of the burner head

The variation of the air guiding panel height (k) has been

performed starting at the X4 burner variation to maintain

the improvement achieved within the first variation study

Fig 13shows the variation of the AGP height (k)

Figs 14–17show the calculated temperature distributions and NO distributions for the burner variations X4, K1, K2 and K4

Following the increase of the k parameter, the tempera-ture and size of the “post shear layer” flame fragment decrease gradually Finally, at k¼4 mm, the typical and preferredflame structure is obtained, which burns nearly all the fuel within thefirst (shear layer) flame fragment

This change inflame structure is achieved thanks to the larger size of the inner recirculation vortex, which provides

a longer shear layer with the outer recirculation vortex and gives the flame more space to burn before reaching the

“summit” of the inner vortex (seeFig 17) This observation

is interesting: it means that influencing the inner vortex size allows direct control of NO emissions Controlling the inner vortex size is in turn possible by selecting the geometric burner parameters adequately

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112

Fig 12 Calculated temperature (bottom) and NO mass fraction (top)

distributions for the burner variation X4 (x ¼4 mm, Δk¼0).

Fig 13 Variation of the AGP (air guiding panel) height by variation of

the k parameter.

Fig 14 Calculated temperature (bottom) and NO mass fraction (top)

distributions for the burner variation X4 (x ¼4 mm, Δk¼0) with

modi fied NO mass fraction scale.

Fig 15 Calculated temperature (bottom) and NO mass fraction (top) distributions for the burner variation K1 (x ¼4 mm, Δk¼1 mm).

Fig 16 Calculated temperature (bottom) and NO mass fraction (top) distributions for the burner variation K2 (x ¼4 mm, Δk¼2 mm).

Trang 9

The NO concentration shows the same trend: the high NO

concentration zone downstream of the shear layer part becomes

smaller with increasing k value Finally, the major NO formation

zone is found within the shear layer flame part Further, the

highest NO mass fraction is decreased from 3 105for the

reference geometry to 1.4 105for the K4 variation.

Thanks to these improvements inflame structure and heat

release zone extent, the total NO emission of the burner has

been reduced from 1.38 ppm for the reference geometry to

0.28 ppm for the K4 variation, which is an emission

reduction by nearly 5 times

Although both emission values are very low (already at a

single digit level), the significant NO emission reduction

provides the possibility to increase the energy density

further, e.g when operating at higher pressures or further

increasing the fuel injector diameter

4 Conclusion

The micromix test burner with an injector diameter of 1 mm

has been tested successfully under atmospheric conditions and

has proven its dry-low-NOx ability over a wide operating

range, despite of its increased energy density Due to the

increased energy density, the micromixflames become thicker,

longer and develop a“post shear layer” flame fragment, where

NO formation is increased due to higher temperatures

A parametric study and numerical exploration of the high

energy density micromix burner revealed that it is well possible

to positively influence the flame shape by influencing the

stabilization vortices and/or the mixing path length An

adequate selection of the burner geometric parameters allows

adjusting mixing length, flame length and inter-shear layer

length to suppress the NO rich “post shear layer” flame

fragment This has been found to significantly decrease the

NO emissions of the burner in question by approx 80% This

offers a great potential of further increasing the micromix

energy density while maintaining low NOx emissions

Espe-cially the consideration of elevated pressure conditions (for

integration in real gas turbine combustors) leads to thicker and longer micromixflames The design of adequate burners for real gas turbine applications can make use of the present findings to balance the design requirements in terms of energy density, manufacturability, stability and emission behavior

Acknowledgements The numerical flow and combustion simulations pre-sented in this paper have been carried out with the STAR-CCMþ Software of CD-adapco Their support is gratefully acknowledged

References

[1] T Lieuwen, V Yang, R Yetter, Synthesis Gas Combustion:

Fundamentals and Applications, CRC Press Taylor & Francis Group, Boca Raton, 2010

[2] G Dahl, F Suttrop, Engine control and low-NOxcombustion for hydrogen fuelled aircraft gas turbines, Int J Hydrog

Energy 23 (1998) 695–704

[3] G Dahl, R Dorneiski, Low NOx-potential of hydrogen-fuelled gas turbine engines, in: Proceedings of the 1st International Conference on Combustion Technologies for Clear Environ-ment, Villamoura, Portugal, 3–6 September, 1991

[4] H.H.-W Funke, S Börner, P Hendrick, E Recker, Mod-ification and testing of an engine and fuel control system for

a hydrogen fuelled gas turbine, in: Progress in Propulsion Physics, Vol III, 2009

[5] H.H.-W Funke, S Börner, P Hendrick, E Recker, Control system modifications for a hydrogen fuelled gas-turbine, in:

Proceedings of the 13th International Symposium on Trans-port Phenomena and Dynamics of Rotating Machinery, 2010

[6] H.H.-W Funke, S Börner, P Hendrick, E Recker, R Elsing, Development and integration of a scalable low NOxcombustion chamber for a hydrogen fuelled aero gas turbine, in: Proceedings

of the 4th European Conference for Aeronautics and Space Sciences, accepted for: Advances in Propulsion Physics, 2011

[7] S Börner, H.H.-W Funke, F Falk, P Hendrick, Control System Modifications and Their Effects on the Operation of A Hydrogen-Fueled Auxiliary Power Unit, in: Proceedings of the 20th International Symposium on Air Breathing Engines, 2011

[8] F Shum, J Ziemann, Potential Use of Hydrogen in Air Propulsion, Euro-Québec Hydro-Hydrogen Pilot Project (EQHHPP), European Union, Contract No 4541-91-11 EL ISP PC, Final Report, 1996

[9] A Westenberger, Liquid Hydrogen Fuelled Aircraft – System analysis, CRYOPLANE, European Commission Final Technical Report No GRD1-1999-10014, 2003

[10] H.H.-W Funke, S Börner, J Keinz, P Hendrick, E Recker, Low NOxhydrogen combustion chamber for industrial gas turbine applications, in: Proceedings of the 14th International Symposium on Transport Phenomena and Dynamics of Rotating Machinery, Honolulu, Hawaii, 2012

[11] H.H.-W Funke, S Börner, J Keinz, K Kusterer,

D Kroniger, J Kitajima, M Kazari, A Horikawa, Numerical and Experimental Characterization of Low NOx Micromix Combustion Principle for Industrial Hydrogen Gas Turbine Applications, ASME Turbo Expo 2012, GT2012-69421, Copenhagen, DK, 2012

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 Fig 17 Calculated temperature (bottom) and NO mass fraction (top)

distributions for the burner variation K4 (x ¼4 mm, Δk¼4 mm).

Trang 10

[12] H.H.-W Funke, S Börner, A Robinson, P Hendrick,

E Recker, Low NOx H2 combustion for industrial gas

turbines of various power ranges, in: Proceedings of the

5th International Conference the Future of Gas Turbine

Technology, ETN-2010-42, Brussels, Belgium, 2010

[13] H.H.-W Funke, E Recker, S Börner, W Bosschaerts, LES

of jets in cross-flow and application to the micromix

hydrogen combustion, in: Proceedings of the 19th

Interna-tional Symposium on Air Breathing Engine,

ISABE-2009-1309, Montreal, Canada, 2009

[14] A Haj Ayed, K Kusterer, H.H.-W Funke, C Striegan,

D Bohn, Experimental and numerical investigations of

the DLN hydrogen micromix combustion chamber of an

industrial gas turbine, Propuls Power Res 4 (3) (2015)

123–131

[15] A Haj Ayed, K Kusterer, H.H.-W Funke, C Striegan,

D Bohn, Improvement study for the dry-low-NOx hydro-gen micromix combustion technology, Propuls Power Res 4 (3) (2015) 132–140

[16] H.H.-W Funke, J Keinz, K Kusterer, A.H Ayed,

M Kazari, J Kitajima, A Horikawa, K Okada, Experi-mental and Numerical Study on Optimizing The DLN Micromix Hydrogen Combustion Principle for Industrial Gas Turbine Applications, ASME Turbo Expo 2015, GT2015-42043, Montréal, Canada, 2015

[17] CD-adapco, User Guide Star-CCMþ 9.02 CD-adapco, 2014

[18] D Fernández-Galisteo, A.L Sánchez, A Linán, F.A Wil-liams, One-step reduced kinetics for lean hydrogen-air deflagration, Combust Flame 156 (2009) 985–996

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16 17 18 19 20 21 22 23 24 25 26 27 28 29

Ngày đăng: 19/11/2022, 11:43

Nguồn tham khảo

Tài liệu tham khảo Loại Chi tiết
[1] T. Lieuwen, V. Yang, R. Yetter, Synthesis Gas Combustion:Fundamentals and Applications, CRC Press Taylor & Francis Group, Boca Raton, 2010 Sách, tạp chí
Tiêu đề: Synthesis Gas Combustion:Fundamentals and Applications
Tác giả: T. Lieuwen, V. Yang, R. Yetter
Nhà XB: CRC Press
Năm: 2010
[3] G. Dahl, R. Dorneiski, Low NO x -potential of hydrogen-fuelled gas turbine engines, in: Proceedings of the 1st International Conference on Combustion Technologies for Clear Environ- ment, Villamoura, Portugal, 3–6 September, 1991 Sách, tạp chí
Tiêu đề: Low NO x -potential of hydrogen-fuelled gas turbine engines
Tác giả: G. Dahl, R. Dorneiski
Nhà XB: Proceedings of the 1st International Conference on Combustion Technologies for Clear Environment
Năm: 1991
[4] H.H.-W. Funke, S. Bửrner, P. Hendrick, E. Recker, Mod- ification and testing of an engine and fuel control system for a hydrogen fuelled gas turbine, in: Progress in Propulsion Physics, Vol. III, 2009 Sách, tạp chí
Tiêu đề: Modification and testing of an engine and fuel control system for a hydrogen fuelled gas turbine
Tác giả: H.H.-W. Funke, S. Bửrner, P. Hendrick, E. Recker
Nhà XB: Progress in Propulsion Physics, Vol. III
Năm: 2009
[5] H.H.-W. Funke, S. Bửrner, P. Hendrick, E. Recker, Control system modi fi cations for a hydrogen fuelled gas-turbine, in:Proceedings of the 13th International Symposium on Trans- port Phenomena and Dynamics of Rotating Machinery, 2010 Sách, tạp chí
Tiêu đề: Control system modifications for a hydrogen fuelled gas-turbine
Tác giả: H.H.-W. Funke, S. Bửrner, P. Hendrick, E. Recker
Năm: 2010
[6] H.H.-W. Funke, S. Bửrner, P. Hendrick, E. Recker, R. Elsing, Development and integration of a scalable low NO x combustion chamber for a hydrogen fuelled aero gas turbine, in: Proceedings of the 4th European Conference for Aeronautics and Space Sciences, accepted for: Advances in Propulsion Physics, 2011 Sách, tạp chí
Tiêu đề: Development and integration of a scalable low NO x combustion chamber for a hydrogen fuelled aero gas turbine
Tác giả: H.H.-W. Funke, S. Bửrner, P. Hendrick, E. Recker, R. Elsing
Nhà XB: Proceedings of the 4th European Conference for Aeronautics and Space Sciences
Năm: 2011
[7] S. Bửrner, H.H.-W. Funke, F. Falk, P. Hendrick, Control System Modi fi cations and Their Effects on the Operation of A Hydrogen-Fueled Auxiliary Power Unit, in: Proceedings of the 20th International Symposium on Air Breathing Engines, 2011 Sách, tạp chí
Tiêu đề: Control System Modifications and Their Effects on the Operation of A Hydrogen-Fueled Auxiliary Power Unit
Tác giả: S. Bửrner, H.H.-W. Funke, F. Falk, P. Hendrick
Nhà XB: Proceedings of the 20th International Symposium on Air Breathing Engines
Năm: 2011
[8] F. Shum, J. Ziemann, Potential Use of Hydrogen in Air Propulsion, Euro-Québec Hydro-Hydrogen Pilot Project (EQHHPP), European Union, Contract No. 4541-91-11 EL ISP PC, Final Report, 1996 Sách, tạp chí
Tiêu đề: Potential Use of Hydrogen in Air Propulsion
Tác giả: F. Shum, J. Ziemann
Nhà XB: European Union
Năm: 1996
[9] A. Westenberger, Liquid Hydrogen Fuelled Aircraft – System analysis, CRYOPLANE, European Commission Final Technical Report No. GRD1-1999-10014, 2003 Sách, tạp chí
Tiêu đề: Liquid Hydrogen Fuelled Aircraft – System analysis
Tác giả: A. Westenberger
Nhà XB: CRYOPLANE
Năm: 2003
[10] H.H.-W. Funke, S. Bửrner, J. Keinz, P. Hendrick, E. Recker, Low NO x hydrogen combustion chamber for industrial gas turbine applications, in: Proceedings of the 14th International Symposium on Transport Phenomena and Dynamics of Rotating Machinery, Honolulu, Hawaii, 2012 Sách, tạp chí
Tiêu đề: Proceedings of the 14th International Symposium on Transport Phenomena and Dynamics of Rotating Machinery
Tác giả: H.H.-W. Funke, S. Bửrner, J. Keinz, P. Hendrick, E. Recker
Năm: 2012
[2] G. Dahl, F. Suttrop, Engine control and low-NO x combustion for hydrogen fuelled aircraft gas turbines, Int. J. Hydrog.Energy 23 (1998) 695–704 Khác

TÀI LIỆU CÙNG NGƯỜI DÙNG

TÀI LIỆU LIÊN QUAN

🧩 Sản phẩm bạn có thể quan tâm