Waste Management 22 (2002) 443451

www.elsevier.com/locate/wasman

Low NOx burnersprediction of emissions concentration

based on design, measurements and modelling
Ladislav Bebara, Vit Kermesa, Petr Stehlika,*, Josef Canekb, Jaroslav Oralc
a

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).

nitrogen oxides formation by reactions of oxygen and

nitrogen proposed by Zeldovich et al. (1947) is the only
one possible way of nitrogen oxides formation. In principle, nitrogen oxides can be formed by the following
reactions (Beer, 1994):

Reaction of the atmosphere (molecular) nitrogen
and single atom oxygen, the latter formed by
splitting of molecular oxygen under high temperature in an oxidation atmosphere (so called
thermal or Zeldovich NOx).

Reaction of the atmosphere (molecular) nitrogen
and hydrocarbons or hydrocarbon fragments that
originated in thermal decomposition in a reducing
atmosphere (prompt NO or Fenimore NO).

Oxidation of nitrogen compounds included in
the fuel (fuel NO).
In the case of combustion of fuels that do not contain
nitrogen bound in organic material the main source of
nitrogen oxides in an oxidizing atmosphere are the
reactions (R1) and (R2).

0956-053X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0956-053X(02)00028-4

444

L. Bebar et al. / Waste Management 22 (2002) 443451

TTFT
V

Nomenclature
C
c
Fg
K1
Kv
M
[NO2]

QB
r
s

t
T
Teq

molar concentration (mol cm3)

molar concentration (mol mol1)
ue gas ow-rate (mol s1)
reaction rate constant (cm3 mol1 s1)
equilibrium constant of oxygen radicals
formation
molecular weight
concentration of NOx converted into
NO2 in dry ue gas having the
reference concentration of 3% O2
(mg NO2 m3
N )
burner duty (MW)
reaction rate (mol.cm3 s1)
burner design parameter which denotes
the volumetric part of the fuel fed into
the rst stage of the burner (stage ring)
temperature ( C)
thermodynamic temperature (K)
equivalent temperature of NOx formation (K)

N2 O2 NO N

R1

N O2 NO O

R2

The reaction (R1) requires the presence of oxygen

radicals. Oxygen radical can be formed by splitting
oxygen molecule or by reactions of oxygen with another
molecules or substances present in the combustion
chamber.
O2 O O

R3

O2 M O O M

R4

Much broader spectrum of reactions leading to NOx

formation and reactions of radicals was described by for
example Li et al. (1999) who also presented reaction
velocity constants of the reactions in question. In all
contributions devoted to description of formation of
NOx, the authors indicate an important role of the
temperature and ow elds in the combustion chamber
if substances are present prone to form nitrogen oxides in contact with other products of thermal
decomposition. At the same time it is necessary to

N as a subscript in unit m3N means description of normal

conditions (0  C, 101.32 kPa), i.e. normal cubic meter
1

theoretical ame temperature (K)

combustion chamber volume (m3)

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

take into consideration the design parameters of the

burner and the mode of combustion of the fuel.
That is why the task of NOx calculation in the concrete thermal equipment encompasses the need of
simultaneous solving a few problems:

temperature distribution in the combustion
chamber by means of the zonal or an other
eligible model;

ow eld distribution by means of a hydrodynamic model; and

kinetic model development enabling to determine
the products of the series of parallel or consecutive chemical reactions.
These mathematical models are based mostly on the
solution of transport equations for momentum enthalpy,
chemical species concentration and turbulent quantities.
The NOx formation can be considered in a post-processor routine.
Software FLUENT (1993) belongs to the most
popular commercially available Computational Fluid
Dynamics (CFD) packages. CFD analysis complements traditional testing and experimentation and provides the user with added insights and condence in his
design. In addition to predicting uid ow, heat and
mass transfer etc., the software comprises the FLUENT NOx module which provides users with capability
to model thermal, prompt and fuel NOx formation in
combustion systems.

L. Bebar et al. / Waste Management 22 (2002) 443451

The possibility to develop a mathematical model

based on experimental data from a testing facility
equipped by a specic type of burner for predicting NOx
formation was shown for example by Barreiros et al.
(1993).
Mathematical models developed for complex solution
of the problem are rather sophisticated and they ask for
entering number of constants and parameters many of
which are directly dependent on geometry of the combustion chamber and the built-in heat transfer surfaces.
A series of kinetic parameters must be chosen with
respect to the number of reactions leading to NOx formation or to formation of its chemical intermediates. It is
necessary to keep in mind that for all models mentioned
the results obtained by their application are preferably
valid in relation to the space geometry presumed during
the model development. Such models can be very well
utilized for the simulation of conditions leading to lower
nitrogen oxide formation as a consequence of a change
of operating conditions in of a particular equipment.
The dependence of NOx emissions concentration on
various factors is generally known. It was shown that
the main factors which to a great extent inuence NOx
formation are the excess air and the temperature of preheated air (Carvalho et al., 1990, 1992; Burd, 1992).
NOx increases with increasing combustion temperature
and with increasing average furnace chamber temperature. NOx decreases as a function of the amount of ue
gas re-circulated and water injection can also reduce the
NOx emissions (Wood, 1994). The inuence of excess
air on the NOx formation is intricate. With increasing
excess air the oxygen concentration in ue gas becomes
higher but the ame temperature decreases. Therefore
the curve for NOx emissions shows a maximum at certain value of the coecient of excess air (Burd, 1992).
Dierent strategies to reduce NOx formation using
well known technologies like staged combustion, ue
gas re-circulation and ameless oxidation were shown,

445

for example by Flamme and Kremer (1993) and

Flamme et al. (1995). Anyway, technologies like these
have been developed to reduce temperature peaks in
oxygen rich part of the ame.
Low-NOx burners (which are used in the process
industries) are designed to achieve the staging eect
internally. The air and fuel ow elds are partitioned
and controlled to achieve the desired air/fuel ratio which
contributes to reduced NOx formation and results in
complete burnout within the furnace. Low-NOx burners
are practically applicable to all combustion devices with
circular burner designs (Wood, 1994). However, the
burner design is a specic matter and it depends rst of
all on manufacturers know-how and experience.

2. Description of the testing facility

Experimental data have been obtained from testing
low-NOx burner at the research facility intended for
verifying the design of burners for process furnaces and
other industrial combustion chambers.
This facility consists of three cylindrical combustion
chambers: (1) horizontal: 2.5 MWI.D.=1.0 m, length
of 6 m; (2) horizontal: 6.0 MWI.D.=1.6 m, length of
6.0 m; (3) vertical: 8.0 MWI.D.=2.0 m, length of 7.0
m. These combustion chambers are water-cooled and
air-tight. Partial shielding of the internal surface of
combustion chambers by a thermal insulation layer
contributes to better simulation of real operating conditions. Air can be preheated up to 350  C. Measured
data given later were obtained at the horizontal chamber [case (1)see earlier)]. Natural gas was used as a
fuel. There was a possibility to measure concentrations
of NO, NO2, CO, O2 by ue gas analyzer. Other measured data are as follows: ue gas temperature at the
point of ue gas analysis (approximately ue gas outlet
temperature), temperature of the pre-heated air and
burner heat duty.
Experimental data presented further were measured
during testing of a gas burner with forced inow of
combustion air. Nominal duty of the gas burner was
1.18 MW, the ratio of burner duty control was 1:5. The
optimum air excess for the burner was 1.05. Gas injector
stabilizing burner is used for ame stabilization. The
sketch of the burner is shown in Fig. 1. It is obvious that
gas combustion by this burner can be carried out stagewise by making use of the secondary fuel inlet jets.

3. Experimental data

Fig. 1. Gas burner.

A review of experimental data can be found in

Table 1. Tests were carried out within limits of the burner duty control, approx. within 0.51.4 MW range.
Three regimes of the process of combustion dis-

446

L. Bebar et al. / Waste Management 22 (2002) 443451

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

[NO2] is the concentration of NOx converted into the NO2 to be able

to compare it with permissible limit. Basis for data: dry ue gas having
the reference concentration of 3% O2.

tinguished by dierent ratios of the secondary fuel inlet

(the parameter s shown in the Table 1 denotes the
volumetric ow-rate of fuel through the rst stage of the
burner) were considered. Another quantities evaluated
and shown in Table 1 are: excess air a, temperature of
air (and/or preheated air) tair, temperature of exhaust
ue gas tg, burner duty QB, concentration of NOx converted into NO2 in dry ue gas having the reference
content of 3% O2 (NOx).
Basic series of experimental data was obtained from
testing a burner used as a conventional one with the
intention for further design changes aimed at the stage
fuel combustion. All the fuel input is realized through
one nozzle, and the related burner design parameter is
as follows: s=1.00. However, results of testing showed
that using preaheated combustion air in this conventional burner arrangement reects in an excessive NOx
concentration. [NOx)=230310 mg NO2/m3N]. This level

of concentration is not acceptable due to emission limits

in the Czech Republic (which is more or less in agreement with that in EU countries) where the maximum
concentration of nitrogen oxides in ue gas from combustion chambers, boilers, furnaces, etc. is below 200
mg NO2/m3N. Two principal improvements of the burner
design consisting in installation of four additional nozzles and the change in dimension of the main nozzle are
performed. The diameter of nozzles was determined in
such a way so that the burner duty was maintained (i.e.
the overall ow-rate of fuel through all the nozzles)
while the dimensions of nozzles and their number were
selected in order to achieve 40% and/or 60% ow-rate
from the total one through additional (second stage)
nozzles (staged ring). Therefore the related values of
the burner design parameters (which denote the owrate through the mainrst stagenozzle) are as follows: 0.60 and 0.40, respectively. The burner duty
within the interval 0.351.4 MW and excess air within
the interval 1.051.20 enabled to acquire data for real
conditions of operation while an emphasis was laid on
regimes with preheated air. Relative narrow range of
excess air values was given by the purpose and aim of
testingto develop and manufacture a low NOx gas
burner for commercial industrial application.
If we look at results in Table 1 we can see that we
achieved low concentration of nitrogen oxides. By utilizing the technology of staged fuel combustion with this
burner under process conditions, the achieved decrease
of NOx emissions reached down to values corresponding to about 50 mg NO2/m3N in the reference dry ue
gas. Series of newly produced burners were used for
modernization of burner system of a process reaction
furnace of the steam reforming. Burners like these are
also suitable for using in units for the thermal treatment
of wastes.
Other industrial natural gas, oil and combined (and/
or dual) burners were also tested in the earlier experimental facilities. The burner duties were within the
range of 15 MW and the tested burners are being now
used for various industrial purposes (e.g. combined oilgas burners with oil atomization by superheated steam
are also applicable for incinerators). Thanks to the system of staged ring, very low NOx concentration in ue
gases was achieved.

4. Simple mathematical model for NOx emissions prediction

The mathematical model is based on some simplifying
assumptions adopted:
1. combustion chamber, which in fact represents a
reactor, was assumed to be a continuous wellstirred reactor; and

L. Bebar et al. / Waste Management 22 (2002) 443451

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

Considering a uniform temperature in the combustion

chamber as a simplifying assumption we can say that a
hypothetical temperature of NOx formation (Teq) should
be somewhere between the theoretical ame temperature
(TTFT) and the temperature of ue gas at the combustion
chamber outlet (Tg). Teq can be considered to be an
equivalent temperature of NOx formation.
The results of experimental measurements were evaluated in such a way that a corresponding equivalent

447

temperature (Teq) was found for each value of nitrogen

oxides formation measured under a set of combustion
conditions in a given combustion chamber. Reaction
rate constant k1 and equilibrium constant Kv of oxygen
radicals formation were evaluated subsequently for each
equivalent temperature (Teq).
Taking into account the simplifying assumption (1)
previously stating that the whole space of the combustion chamber can be viewed as a continuous well stirred
reactor and owing to a negligible content of NOx at the
combustion chamber inlet, the increase of nitrogen oxides in the combustion chamber can be expressed by the
equation:
Fg CNO Vr

For the purpose of comparison it is necessary to

recalculate the quantity of nitrogen oxides evolved into
the corresponding concentration of NO2 in the reference
dry ue gas having 3% of oxygen according to the well
known equation:
NO2

CNO MNO2 106

0:18



cO2
1  cH2 O 22:414 0:21 
1  CH 2 O

It is also necessary to select parameters that inuence

the equivalent temperature Teq. These parameters
should be measured, given or calculated. We speculated
that three principal parameters inuence the equivalent
temperature in the combustion chamber:
1. temperature of combustion;
2. heat released in the combustion chamber: and
3. design of the burner.
Theoretical ame temperature (TTFT) is considered to
be the temperature that would be achieved in the case of
combustion of the fuel under current excess air and
adiabatic conditions in the combustion chamber.
Therefore, this quantity encompasses: the properties of
the fuel (its heating value, composition) and those of
excess air (a) and the temperature of air (and/or preheated air Tair). The heat load of the combustion
chamber can be characterized by the ratio of the heat
transferred from the ue gas into the cooled walls of the
chamber to the overall heat input of the chamber
(Z=Qabs/Qin). The inuence of the technology of fuel
combustion on the NOx formation is taken into account
via the burner design parameter s expressed as the
volumetric ow-rate of the fuel through the primary jet.
We strived to have a minimum number of parameters,
but Teq can also be a function of other parameters.
A simplied algorithm for the calculation of equivalent temperature from experimental data is shown in
Fig. 2. The calculation of the equivalent temperature is

448

L. Bebar et al. / Waste Management 22 (2002) 443451

Table 2
Measured and calculated values

Fig. 2. Procedure for NOx concentration evaluation.

of iterative nature and it aims to the value at which the

calculated concentration of NOx is in agreement with
the experimental one.
By applying software for statistic calculations we can
determine the following polynomial equation for the
evaluation of Teq.
2
Teq 1196:7 0:139103 TTFT
 32:663109
3
TTFT
 

49:337
s

5. Results and discussion

5.1. Comparison of measured and calculated values of
NOx concentration
The dierence between the NOx emissions measured
at the testing facility and those calculated using the
mathematical model for NOx formation prediction
described above is obvious from Table 2. The average

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

dierence is roughly 9.1%, the maximum one is 25%.

However, the dierence is mostly less as 10% relatively.
If we look at Table 2 we can see that the NOx emissions concentration exceeds the permissible limit in case
of measurements 6 to 10 where the temperature of preheated air is too high. The model mostly under-predicts
measurements. However, this is the result of simplifying
assumptions.
Further, it is necessary to make a remark that the
measured values were not obtained from numerous systematic measurements (as it is a common practice in
research, e.g. by changing step by step only one parameterlet us say e.g. excess air in our casewhile the
other ones are maintained constant). Data utilized were
obtained from various measurements realized with the
aim to test a burner prototype for a further development and installing within the industrial scale. From
this for example only a narrow range of excess air values
was considered even if from the mathematical modeling
and models validating points of view it would have

L. Bebar et al. / Waste Management 22 (2002) 443451

been useful to acquire data from a much wider interval

of operational regimes. However, the data available
satised our aim consisting in developing a methodology for prediction the NOx emissions concentration.
Another comments can be found in Section 7.
5.2. Parametric sensitivity
Computer program for the evaluation of NOx emissions concentration (including its recalculation to the
reference conditions: 3% O2 on a dry ue gas basis)
enables comparison with the emission limit. Flue gas
composition, heat supplied and absorbed, rate of the
reaction, residence time, etc. can be considered as additional results of simulation. If we change values of
selected parameters we can investigate the parametric
sensitivity of the model, i.e. we can evaluate the extent
in which these parameters inuence some output parameters of the model (e.g. residence time, combustion
chamber eciency, theoretical ame temperature,
equivalent temperature of NOx formation and nitrogen
oxides concentration).
The values of excess air a and the technology of
combustion (expressed by the parameter s which characterizes the proportion between primary and secondary
fuel fed into the burner) are the key factors of burner
operation inuencing the NOx formation. This statement is documented in Fig. 3 that showsas for nitrogen oxides formationa qualitative dierence between
combustion of gas fuel in a conventional burner and in
the burner with staged combustion.
Experience from the research and development of the
burners proved that applying the stage combustion in
the same burner can bring a benet reecting in reduction

449

of NOx emissions concentration by 5070% relating to

the original conventional burner design.
The results achieved, thanks to the application of
mathematical model, conrm that the maximum level of
NOx is found in cases of fuel combustion with excess air
ranging approximately from 1.05 to 1.15. At lower
values of excess air the eect of low oxygen concentration in ue gas comes across, at higher values of excess
air the eect of decreased ame temperature comes
across.
A favorable inuence on decreasing nitrogen oxides
levels has the feeding of inert gas (water vapor or recycled ue gas can be inert gas) and this inuence is portrayed in Fig. 4.
The inuence of the temperature of preheated air on
nitrogen oxide emissions formation is illustrated in
Fig. 5 and the data there conrm the empirical knowledge that increasing temperature of preheated air brings
about fuel savings but simultaneously causes an increase
of nitrogen oxides formation by 100% or even more.
However, the inuence of preheated air temperature
on nitrogen oxides concentration has to be taken into
consideration from various points of view because by
increasing the temperature of air we can achieve
reduced fuel consumption. Nevertheless, from the NOx
emissions formation point of view it can be stated that
increasing the air temperature up to 250 to 300  C
brings about 10% saving, but on the other side results in
increasing NOx emissions concentration. Unfortunately,
this penalty is a prevailing factor in this case. For the
burner discussed above the temperature of air preheating oscillating around 200  C proved itself to be acceptable since the NOx emission concentration did not
exceed 200 mg/m3N in the reference ue gas.

Fig. 3. NOx emissions concentration vs. air excess and burner design.

450

L. Bebar et al. / Waste Management 22 (2002) 443451

Fig. 4. Inuence of steam injection on NOx concentration.

Fig. 5. NOx emissions concentration vs. air excess and temperature.

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.

Utilizing the simple mathematical model it is possible

to use this approach for prediction how various regimes
of operation (characterized by changing parameters of
the combustion) inuence the nitrogen oxides formation
in case of testing a concrete type of burner installed in
one of experimental combustion chambers.
The simple semi-empirical mathematical model after
its validation contributes to an evaluation of results of
further testing, to identifying reasons of an unstable
state of the testing device, to decreasing the number of
measured operational regimes (it results in saving cost
fuel and manpower).
By an industrial implementation of the model it has
been conrmed that results of calculations are in relatively good agreement with measures values. Results
from the operation of a steam-reforming furnace where
approximately 100 natural gas burners were installed
proved even lower level of nitrogen oxides emissions
than expected based on experimental research and
mathematical modeling. The resulting concentration of

L. Bebar et al. / Waste Management 22 (2002) 443451

NOx was well below the permissible limit (200 mg NO2/

m3N) as stated earlier). Therefore the newly developed
burners can be used in various countries and satisfy the
related regulations.

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

Carvalho, M.G., Semiao, V.S., Coelho, J.P., 1992. Modelling and

optimization of the NO formation in an industrial glass furnace.
ASME J. Engineering for Industry 114, 514523.
Carvalho, M.G., Semiao, V.S., Lockwood, F.C., Papadopoulos, C.,
1990. Prediction of nitric oxides emissions from an industrial glassmelting furnace. Journal of Inst. of Energy March, 3947.
El-Sherif, A.S., 1998. Eects of natural gas composition on the nitrogen oxide, ame structure and burning velocity under laminar premixed ame conditions. Fuel, 77 14, 15391547.
Flamme, M., Haep, J., and Kremer, H., 1993. NOx Reduction Potential for Glass Melting Furnaces. The International Glass Congress,
Prague, 1012 April.
Flamme, M., Kremer, H., 1993. Reduction of the NOx Output from
Industrial Burners Firing a Mixture of Fuel Gas Combustion Air
Enriched or Oxygen, Second International Conference on Combustion Technologies for a Clean Environment. Lisbon, Portugal. pp.
1922 July.
Fluent, 1993. Computational Fluid Dynamics Software. Fluent Europe, Ltd, Sheeld, UK.
Garg, A., 1994. Specify better low-NOx burners for furnaces. Chemical Engineering Progress January, 4648.
Garg, A., 1992. Trimming NOx. Chemical Engineering November,
122128.
Glassman, I., 1997. Combustion. Academic Press, Inc. Ltd.:London,
United Kingdom Edition.
Hill, S.C., Smooth, J.D., 2000. Modeling of nitrogen oxides formation
and destruction in combustion systems. Progress in Energy and
Combustion Science 26, 417458.
Kunz, R.G., Smith, D.D., Adamo, E.M., 1996. Predict NOx from gasred furnaces. Hydrocarbon Processing November, 6579.
Kuo, K.K., 1986. Principles of Combustion. Wiley Inter-Science, New
York.
Li, S.C., Williams, F.A., Gebert, K., 1999. A simplied, fundamentally
based method for calculating NOx emissions in lean premixed combustors. Combustion and Flame 119, 367373.
Munzio, L.J., Quartucy, G.C., 1997. Implementing NOx control:
research to application. Progress in Energy and Combustion Science
23, 233266.
Turns, S.R., 1995. Understanding for NOx Formation in non-premixed ames, experiments and modeling. Progress in Energy and
Combustion Science 21, 361385.
Wood, S.C., 1994. Select the right NOx control technology. Chem.
Eng. Progress 90, 3238.
Zeldovich, Z.B., Sadovnikov, P.Y., Kamenetskii, F.D.A., 1947. Oxidation of nitrogen in combustion (trans. by M. Shelef). Academy of Sciences of USSR, Institute of Chemical Physics, Moscow, Leningrad.