Numerical investigation of combustion ...

3 downloads 0 Views 2MB Size Report
Power generation capacity. Ref: «Fuel Conditioning System for Ammonia-Fired Power Plants » By Arif Karabeyoğlu , Brian Evans, 2012 ...
2015 International Conference on Power and Energy Systems (ICPES 2015)

Ammonia as a renewable energy source and development of reduced chemical mechanisms

Hadi Nozari and Arif Karabeyoğlu Department of Mechanical Engineering Koç University İstanbul, Turkey



Why ammonia?



Challenges



Study Scope



Some Results

• Extensive use of fossil fuels • Major problems: Human health and welfare, environmental issues • A hot concern: Replacing current energy carriers • The Intergovernmental Panel on Climate Change (IPCC) report: Atmospheric CO2 levels rose almost twice as fast in the first decade of this century

«THE WORLD MUST RAPIDLY MOVE AWAY FROM CARBON-INTENSIVE FUELS»

Clean

Energy Dense

Ideal Energy Carrier

Safe

Cost Effective

NH3 as a Green Energy Source

SUSTAINABLE, RENEWABLE, NO CARBON INCLUDED FUEL

WELL DEVELOPED

AND EFFICIENT SYNTHESIS PROCESS

Haber-Bosch Process

EXISTING PRODUCTION, STORAGE AND TRANSPORT INFRASTRUCTURE

• High energy density

Ref: «Ammonia and related chemicals as potential indirect hydrogen storage materials», By R. Lan et al., 2012

Power generation capacity

Ref: «Fuel Conditioning System for Ammonia-Fired Power Plants » By Arif Karabeyoğlu , Brian Evans, 2012

Drawbacks also exist!



Slow kinetics (low flame speed)



NH3 , a source of fuel NOx in flames

Flame Speed (cm/sec)

200

150

145 H2

100 50

CH4 NH3

40 10

0 Standard Temperature and Pressure Air as oxidizer

«Ammonia – It’s Transformation and Effective Utilization» AIAA 2008 Ref: Miller, J. A. et al. , Combust. Sci. Tech., 34:149–176 (1983)

A simplified reaction mechanism

8

x 10

-3

2100 Temerature 2000

4

1900

2

1800

Effect of H2 addition to the mixture 0



Constant Inlet T (400°C) • Thermal NOx and Fuel NOx; the only players?

0

20

40

60

1700 100

80

%E H2 8

x 10

-3

Constant Flame T 7



Decoupling thermal NOx from total NOx

[NOx Level @Const. Inlet T of 400℃ ] – [NOx Level @ Const. Flame T of 1730 K] × 100 = 𝑻𝒉𝒆𝒓𝒎𝒂𝒍 𝑵𝑶𝒙 𝑺𝒉𝒂𝒓𝒆 [NOx Level @ Const.Inlet T of 400℃ ]



Thermal NOx share increase with increasing H2 % (Flame T)

NOx Mole Fraction

Constant Flame T (1730 K) • General expectation: decrease in Total NOx (decreasing fuel bond N) • Total NOx still increasing (despite const. thermal NOx)! • Effect of H and HNO accumulation → Increasing ROP of some key reactions • NO2+H↔NO+OH • HNO+H↔NO+H2 • HNO+OH↔NO+H2O

5 4 3 2 1 0

a 0

20

40

60

100

90 80 70 Extra H2 to be added

60 50 40 30 20 10

b 0

0

20

40

60 %E H2

ɸ=0.5

80

%E H2

% Thermal NOx Share in Total NOx



Constant Inlet T

6

80

100

Flame Temperature (K)

NOx Formation

Total NOx Mole Fraction

NOx 6

NOx Formation 0.01 Pure NH3 0.009

Effect of equivalence ratio variation

80% NH3 60% NH3

0.008



Increasing / Decreasing trend

Two opposite effects: 1) Increase in thermal NOx by increase in adiabatic flame T

NOx Mole Fraction

20% NH3 0.007

10% NH3 Pure H2

0.006 0.005 0.004 0.003 0.002

2) Decreasing fuel NOx by decreasing O/F ratio 𝑁𝐻𝑖 + 𝑂𝑋 → 𝑁𝑂 + 𝐻𝑖 𝑋

0.001 0 0.4

0.6

0.8 1 Equivalence Ratio

OX: Oxygenated species

Noticeable reduction in NOx emission under the rich conditions P=17 bar, T=673 K

1.2

1.4

Reduced Mechanism NOx Emission Prediction

P=17 bar, T=673 K

Reduced Mechanism Comparison with Experiments

Full Konnov Red. Mech.1 Red. Mech.2 Red. by Duynslaegher et al. Exp. Van Wonterghem et al. Exp. Duynslaegher et al.

140

Flame Speed (cm/s)

120 100 80 60 40 20 0

10

20

DuynslaegherC., PhD dissertation, Universit_ecatholique de Louvain; 2011. Wonterghem JV,Tiggelen AV., Bulletin des SocietesChimiquesBelges1955; 64:99-121

30 Dilution (%N2) STP Condition

40

50

60

Reduced Mechanism Efficiency Chart Full Konnov Mech.

1

Red. Mech.1 Red. Mech.2 Duynslaegher et al.

• •

CPU Time • Reference: Konnov Mech. (unity) Precision • Reference: Experimental Data

Normalized CPU time

0.8

0.6

0.4

0.2

0

4

5

6 7 8 9 10 Average deviation from experimental data (%)

11

0.02





Ignition delay time not captured accurately

The final NOx emission very well predicted by the two reduced mechanisms

NO Mole Fraction



0.015

0.01

0.005

Concentration time profiles very well predicted (once the time scale is translated to account for the differences in the delay times between the reduced mechanism and the full Konnov mechanism)

Konnov Red. Mech. #1 Red. Mech. #2

0 8

0 -3 x 10

1

2

3

t (sec)

-4

x 10 Konnov

Needed

OH Mole Fraction

Red. Mech. #1 Red. Mech. #2

6

4

2

Reduced mechanism applicable to non-steady conditions 0

0

1

2 t (sec)

ɸ=0.5, P=17 bar, 80 %E NH3

3 -4

x 10



Controlling role of radicals in laminar flame speed and autoignition process with OH as the most influential radical



Adding H2 to the fuel mixture improved laminar flame speed and autoignition process



Despite improving flame speed and autoignition process, adding H2 does not necessarily decrease the NOx level



Total NOx formation is highly dependent on the competition between fuel NOx and thermal NOx levels



Localized rich combustion seems to minimize NOx. Ammonia slip may be the compromise



The obtained reduced mechanisms give reliable predictions for flame speed and exhaust emissions



About the same accuracy of Konnov mechanism with nearly 5 times less CPU time expense

Thank You

Objective •

To burn NH3 in a power generation system in efficient and environmentally friendly way: –

Low NOx emission



Low NH3 slip

Approach •

Chemical kinetics study of NH3/H2/Air under practical combustion conditions



Developing a reduced mechanism capable of predicting combustion characteristics with an acceptable accuracy



CFD simulation of combustion



Experimental study to examine the validity of results

NH3/H2/Air

𝑥𝑁𝐻3 + 𝑦𝐻2 + 𝑎𝑠𝑡𝑜𝑖 𝑂2 + 3.76 𝑁2 → 𝑏 𝐻2 𝑂 + 𝑐 𝑁2

P=17 atm ɸ=0.5-1.8

Studied Cases % E of NH3

NH3

H2

O2

N2

100

0.1228

0

0.1843

0.6929

80

0.0958

0.0382

0.1819

0.6840

60

0.0701

0.0745

0.1797

0.6757

40

0.0456

0.1091

0.1776

0.6677

20

0.0223

0.1421

0.1755

0.6600

Reactants’ Composition

%𝐸 𝑁𝐻3 =



Major Combustion Properties

Steady – –



𝑋𝑁𝐻3 × 𝐿𝐻𝑉𝑁𝐻3 × 100 𝑋𝑁𝐻3 × 𝐿𝐻𝑉𝑁𝐻3 + 𝑋𝐻2 × 𝐿𝐻𝑉𝐻2

Burning velocity (laminar flame speed) Emission (NOx formation)

Non-steady –

* A.A.Konnov, Combust. Flame 156 (2009) 2093–2105

Ignition delay time

Konnov Mechanism*

Elements: N/H/O Species: 30 Reactions: 240

Ammonia decomposition analysis Aim: To identify the contribution of each reaction in molar conversion of NH3

80%NH3

100% NH3

Study Cases

ɸ = 0.5, P=17 bar, T= 673 K

60%NH3

The most important reactions and their contribution (%)

Fuel mixture

NH3+OH↔NH2+H2O

NH3+O↔NH2+OH

NH3+NH2↔N2H3+H2

Pure NH3 80%NH3 60%NH3

95%

4.64%

0.23%

94.8%

5.04%

0.11%

94.5%

5.45%

ignorable

Importance of OH radical in flame speed

Correlation

ɸ = 0.5, P=17 bar, T= 673 K

Average Flame Speed Prediction Error (%)

Reduced Mechanism • •

2 Versions Obtained Applicable to steady state conditions

Red. Mech. 1

Red. Mech. 2

Red. Belgium

0.5

8.4

5.6

30

1

4.8

6

19.6

1.1

4.9

5.9

18.3

1.2

4.6

6

19.4

1.5

4.4

6.7

18.3

Konnov (full)

Belgium Red.*

Red. Mech. 1

Red. Mech. 2

1.8

4.2

7.7

16.8

# elements

3

3

3

3

3

7.8

8.8

4.7

# species

30

19

21

18

# reactions

240

80

91

80

Reduction based on: -

ɸ

Contribution of rections in form/decomp of species Tracking form./decomp. of imp radicals in flame speed (OH,H,O, …) Tracking form./decomp. of imp radicals in NOx emmision (NO,NH,NH2, …)

P=17 bar, T=673 K * Duynslaegher C, Contino F,Vandooren J, Jeanmart H., Combust Flame 2012.

Average NOx Formation Prediction Error (%) ɸ

Red. Mech. 1

Red. Mech. 2

Red. Belgium

0.5

7.5

5.6

10.6

1

2.8

1.8

2.5

1.1

5

2.8

5.5

1.2

3

2.7

2.8

1.5

3.2

4.5

10

1.8

3.2

12.8

29.7

3

7.5

23.6

51.4

Reduced Mechanism Flame Speed Prediction

P=17 bar, T=673 K

Reduced Mechanism Flame Speed Prediction Error (%)

Experiments of Duynslaegher etal.

Experiments of Van Wonterghem et al.

Comparison with Experiments

STP Condition

% N2

Konnov

Red. Mech.1

Red. Mech.2

Red. By Duynslaegher et al.

0

1.2

0.4

0.8

10.3

5

1.5

0.6

1.3

11.2

10

3.6

2.9

3.5

13.1

15

4.7

3.8

4.8

13.8

20

7.1

6.3

7.1

15.6

25

11.3

10.0

10.7

19.2

30

15.1

14.7

15.2

22.8

35

1.4

1.1

1.9

7.9

40

5.9

6.0

4.5

1.4

45

9.5

9.1

10.5

12.6

50

3.8

3.6

5.8

4.7

55

1.9

1.2

3.6

3.1

60

12.1

14.0

15.5

12.8

61.7

3.9

5.6

7.8

5.6

Autoignition

Importance of radicals in autoignition and ignition initiation

100%NH3

Accumulation of influential radicals close to the ignition time ɸ=0.5, T=1300 K, P=17 bar

80%NH3

Autoignition -

Effect of initial mixture T

-

Effect of NH3 content

ɸ=0.5, P=17 bar