Académique Documents
Professionnel Documents
Culture Documents
www.elsevier.com/locate/wasman
Technical University of Brno, Institute of Process and Environmental Engineering, Technicka 2, 61669 Brno, Czech Republic
b
VUCHZ a.s. Brno, Brno, Czech Republic
c
EVECO Brno, Ltd., Brno, Czech Republic
Abstract
This paper describes possible ways of prediction of nitrogen oxides formation during combustion of hydrocarbon fuels. Mathematical model based on experimental data acquired from the testing facility has been developed. The model enables to predictat a
high probability measurethe extent of nitrogen oxides emissions. The mathematical model of nitrogen oxide formation relies on
the application of simplied kinetic equations describing the formation of nitrogen oxides at so-called equivalent temperature. It is
a semi-empirical model that comes out of experimental knowledge. An important role played by the burner design itself has been
emphasized and therefore an important supplementary parameter of the model is the characteristic of the burner design. It has been
established that there was a good agreement between experimental data and those calculated by the application of the model to
various conditions marked out by dierent combustion parameters in the combustion chamber. The results obtained by application
of the model respect the inuence of parameters validated by industrial practice that control the formation of nitrogen oxides in the
course of fuel combustion. Such parametersrst of alltare the temperature in the combustion chamber and the concentration of
the substances taking part in the reaction. By application of the model, it is possible to assess the consequence of, for example the
surplus of combustion air, the increase of temperature of combustion air, the supply of inert gas, etc. on the nitrogen oxides emissions of the operating burner under evaluation. Ecient combining of experience and sophisticated approach together with
importance of thus access for an improved design are shown. # 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
The diculty of predicting the quantity of NOx emissions from fuel combustion under certain conditions has
its origin in complex mechanism of nitrogen oxide formation and in many circumstances inuencing it.
Many contributions have been devoted to the analysis
of conditions of formation of nitrogen oxides during
combustion of various types of fuels. Problems like
these have been dealt with in a great detail (Kuo, 1986;
Carvalho et al., 1990, 1992; Garg, 1992, 1994; Azevedo
et al., 1993; Tums, 1995; Kunz et al., 1996; Glassman,
1997; Munzio and Quartucy, 1997; El-Sherif, 1998; Li et
al., 1999; Hill and Smooth, 2000). Beer (1994) described
the mechanism of nitrogen oxide formation by thermal
decomposition and by reactions of various nitrogen compounds and radicals formed in the course of combustion
and he has shown that the conventional mechanism of
* Corresponding author. Tel.: +420-5-4114-2373; fax: +420-54114-2373.
E-mail address: stehlik@kchz.fme.vutbr.cz (P. Stehlik).
0956-053X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0956-053X(02)00028-4
444
TTFT
V
Nomenclature
C
c
Fg
K1
Kv
M
[NO2]
QB
r
s
t
T
Teq
N2 O2 NO N
R1
N O2 NO O
R2
R3
O2 M O O M
R4
Greek symbols
excess air coecent
1
the term proportional to the relative
contribution of the reverse reaction
Z=Qabs/Qin ratio of the heat absorbed by the
cooled wall of the combustion chamber
and heat input
Indices
g
H 2O
NO
NO2
O2
O
N2
abs
air
in
ue gas
steam
nitrogen oxide
nitrogen dioxide
oxygen
oxygen radical
nitrogen
absorbed
air
inlet
445
3. Experimental data
446
Table 1
Measured data from testing gas burner
Measurement
No.
s ()
QB
(MW)
a()
tair
( C)
tg,out
( C)
NO2
(mg/m3N)
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
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.534
0.522
0.685
0.900
1.020
0.534
0.522
0.685
0.900
1.020
0.429
0.900
0.686
0.837
1.020
1.230
1.350
0.990
1.430
1.130
0.870
0.580
0.676
1.018
1.022
1.430
0.870
0.363
0.978
1.156
1.306
1.065
1.000
1.193
1.20
1.20
1.10
1.15
1.10
1.16
1.20
1.11
1.15
1.11
1.15
1.15
1.12
1.13
1.12
1.19
1.08
1.15
1.06
1.06
1.06
1.05
1.21
1.12
1.15
1.06
1.06
1.15
1.16
1.20
1.11
1.09
1.12
1.13
20
20
20
20
20
330
320
345
330
340
10
10
10
10
10
330
330
310
250
250
250
235
5
10
10
10
10
345
340
325
345
175
265
340
730
720
780
860
880
760
660
790
880
910
600
850
760
850
880
945
970
880
940
880
820
740
700
850
870
960
840
590
800
900
940
880
860
890
109
110
108
113
117
283
232
287
309
312
61
71
74
71
61
118
147
143
106
119
119
122
21
30
32
30
34
71
60
63
66
51
52
60
2. NO is the major constituent for industrial furnaces and combustion chambers, therefore thermal or Zeldovich NO formation by the reaction
mechanism (R1), (R2) shown earlier was
assumed.
Overall rate of nitrogen oxides formation by the
reactions (R1) and (R2):
r 2k1 Co CN21
where k1 is rate constant for chemical reaction according to the Zeldovich mechanism of NOx formation
(Carvalho et al., 1990; FLUENT, 1993; Zeldovich,
1997). Reaction velocity equation takes into account
concentration of oxygen radicals (CO) the presence of
which is necessary to keep the reaction (R1) running.
The oxygen radicals are formed by thermal decomposition of molecular oxygen in accordance with (R3) or by
the reaction of molecular oxygen with other molecules
of substances present in the combustion chamber
according to (R4). In agreement with the studies of
Carvalho et al. (1990) we accepted the assumption that
the concentration of oxygen radicals (CO) will satisfy
the requirement of the equilibrium constant of the
reaction (R3), therefore:
Co Kv Co2
1=2
The term (1) is proportional to the relative contribution of the reverse reaction and it represents so
called remoteness of the system from its equilibrium
state. It is possible to make sure that under the described experimental conditions the said remoteness from
equilibrium of the overall reaction of nitrogen oxides
formation is sucient enough to justify disregarding the
term. Since the reaction of nitrogen oxides formation is
slower than the reaction of hydrocarbon combustion, it
is possible to consider the NOx formation to be a consecutive reaction with respect to fuel combustion reactions. Rate equation, where oxygen and nitrogen
concentrations are expressed in their dependency on the
conditions of fuel combustion, was used to express the
rate of nitrogen oxides formation as follows:
r 2k1 Kv Co2
1=2
CN2
447
448
49:337
s
Measurement
No.
[NO2]
measurd
(mg/m3N)
[NO2]
calculated
(mg/m3N)
Error
absolute
(mg/m3N)
Error
relative
(%)
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
109
110
108
113
117
283
232
287
309
312
61
71
74
71
61
118
147
143
106
119
119
122
21
30
32
30
34
71
60
63
66
51
52
60
97
100
122
98
109
270
240
282
243
273
55
54
66
67
63
117
151
123
106
110
118
129
17
28
24
31
34
70
49
52
75
49
59
66
12
10
14
15
8
13
8
5
66
39
6
17
8
4
2
1
4
20
0
9
1
7
4
2
8
1
0
1
11
11
9
2
7
6
11
9
13
13
7
5
4
2
21
12
10
24
11
5
3
1
3
14
0
7
1
6
19
6
25
2
0
1
18
17
13
4
13
10
449
Fig. 3. NOx emissions concentration vs. air excess and burner design.
450
6. Conclusion
In spite of introducing some simplifying assumptions,
the results obtained by using the simple mathematical
model for predicting NOx emissions of a gas burner are
in good agreement with the measured data from a testing facility. The agreement of results for a gas burner is
relatively acceptable. Application of the mathematical
model described in this paper enables to nd the optimum parameters of operation of gas burners, incinerators and process furnaces in general and that can
bring about the minimum nitrogen oxides emissions and
fuel consumption.
The methodology described in this paper can be
applied to various types of burners. We applied it to a
gas burner and dual (gas/oil) burner. The mathematical
model and related computer program are quite open,
therefore it is possible to extend the methodology and
the computational tool e.g. for considering fuel NOx
in addition to the thermal ones.
7. Future work
Research and development in the eld described earlier continues. Mathematical model created (based on
concrete input data from testing a burner) can be used
for just for this type of burner. However, the methodology developed can be applied for dierent types of
burners. At present a new type of low NOx dual burner
(natural gas/low caloric gas) been developed and is
installed at research facility.
Future work will consist on as many measurements as
possible with the aim of further improving the methodology of NOx emissions concentration prediction. The
newly developed burner described earlier will be installed at the unit for thermal treatment of sludge from pulp
production in the nearest future, and thus we will have a
unique opportunity to acquire data from real industrial
operation. This will also contribute to the model validation and improving.
References
Azevedo, J.L.T., Carvalho, M.G., Coelho, J.P., Coimbra, C.F.M.,
Nogueira, M., 1993. Modeling of combustion and NOx emissions in
industrial equipment. Pure & Appl. Chem. 65, 345354.
Barreios, A., Carvalho, M.G., Costa, M., Lockwood, F.C., 1993.
Prediction of the near burner region and measurements of NOx and
particulate emissions in heavy fuel oil spray ames. Combustion and
Flame 92, 231240.
Beer, J.M., 1994. Minimizing NOx emissions from stationary combustion. Reaction engineering methodology. Chem. Eng. Science 49,
40674083.
Burd, V., 1992. Squeezing Clean Energy. Chemical Engineering
March, 145150.
451