Journal of Environmental Protection, 2011, 2, 1095-1100
doi:10.4236/jep.2011.28126 Published Online October 2011 (http://www.scirp.org/journal/jep)
N2O Formation in Selective Non-Catalytic NOx
Reduction Processes
Crisanto Mendoza-Covarrubias1, Carlos E. Romero2, Fernando Hernandez-Rosales3*, Hans Agarwal4
1The Faculty of Mechanical Engineering, Universidad Michoacana de San Nicolás de Hidalgo, Morelia, México; 2The Energy Re-
search Center, Lehigh University, Bethlehem, USA; 3The Centro de Ingeniería y Desarrollo Industrial, Santiago de Querétaro,
Querétaro, México; 4Project Engineer at Foster Wheeler North America Corp, Livingston, USA.
Received July 14th, 2011; revised August 23rd, 2011; accepted September 27th, 2011.
ABSTRACT
Nitrous oxide is not an environmentally regulated species in the U.S., but it does participate in the stratospheric ozone
chemistry and contributes to the greenhouse effect. Nitrous oxide has been found to be a by-product of the selective
non-catalytic reduction process. Chemical kinetic calculations demonstrated that the formation of nitrous oxide in the
urea-based selective non-catalytic reduction process is linked to the conversion of NO by cyano species released from
the process parent compounds. This conversion occurs within in temperature window between 850 and 1050˚C. With
urea injection, nitrous oxide emissions represent up to 20 percent conversion of the NOx reduced. The amount of nitrous
oxide formed depends primarily on the process temperature, the amount of chemical injected, the initial NOx level, and
the carbon monoxide level in the gas stream. These observations, which were based on the chemical kinetics of the
process, should be considered in designing selective non-catalytic reduction systems to minimize nitrous oxide by-
product formation.
Keywords: Selective Non-Catalytic Reduction Process, Nitrous Oxide Emissions Chemical Kinetics
1. Introduction
While global climate changes have been associated with
increased levels of carbon dioxide (CO2) in the atmos-
phere, there is growing concern about the role of other
trace gas species such as methane (CH4), chlorofluoro-
carbons (CFC) and nitrous oxide (N2O). In fact, the trace
gases are expected to equal the global warming effect of
CO2 within 50 years. Nitrous oxide affects the atmos-
phere at the troposphere level as well as at the strato-
sphere level. N2O is a relatively strong absorber of infra-
red radiation and is therefore implicated as an important
contributor to the “Greenhouse Effect.” It is estimated
that a mere doubling of its concentration will result in a
0.28˚C contribution to global warming [1]. Since N2O is
chemically stable in the troposphere, it can be trans-
ported to the stratosphere where it has also been respon-
sible for ozone (O3) depletion (NO + NO2 emitted into
the troposphere are converted into nitrate and removed
by rainfall). In the stratosphere, N2O is converted to NO,
which is the primary species responsible for establishing
the equilibrium stratospheric O3 concentration. One es-
timate suggests that doubling atmospheric N2O concen-
tration would result in a 12 percent decrease in total
column of ozone [2].
Mean global concentrations of N2O are approximately
300 ppb (about one-thousandth that of CO2) and have
been increasing at a rate of 0.2 to 0.4 percent per year [1-
3]. The latest Energy Information Administration –
Emissions of Greenhouse Gases Report reports 2006 U.S,
emissions of nitrous oxide of 375.7 million metric tons
carbon dioxide (CO2) equivalent [4]. Although, the in-
crease of N2O in the atmosphere has been attributed to
anthropogenic sources, uncertainties in the source, sink,
and accumulation terms have prevented a clear estima-
tion of the importance of the various sources. References
in the literature have suggested that combustion of fossil
fuels results in N2O emissions which can account for the
observed increase in N2O [1-5]. Measurements of nitrous
emissions from electric utility power plants have indi-
cated insignificant emissions from gas-fired flames (less
than 10 ppm). On the other hand, field measurements of
coal- and oil-fired power plants have shown a wide
variation of N2O emissions, from 10 to above 200 ppm
[6]. Some data have suggested that N2O molar emissions
could be estimated as about 25 to 40 percent of the NOx

NO Formation in Selective Non-Catalytic NO Reduction Processes 1096 2 x
levels [7].
However, the characterization of N2O emissions from
combustion sources is a complex issue due to reported
findings of a major sampling artifact on measurements
made on grab samples [8]. Reference [8] has demon-
strated that processing combustion products containing
SO2, NOx and water vapor can lead to N2O concentra-
tions several hundred parts per million higher than those
originally formed in the combustion process. The con-
version of NO into N2O occurs by a reaction whose
overall time constant is of the order of 2 hours. N2O emi-
ssions from combustion sources should be low given that
gas-phase chemical reactions such as N2O + radicals =
N2 + radicals and stable species are extremely rapid at
furnace temperatures (1500˚C - 1700˚C). Combined ex-
perimental results and kinetic modeling of sample com-
bustion systems have shown that N2O formed early in the
flame zone, is rapidly destroyed before it can reach the
furnace exhaust [9].
Measurements at several full scale utility combustion
sources, using continuous infrared measurement tech-
niques, have associated utility combustion systems (fir-
ing natural gas, oil, and gas) with direct N2O emission
levels below 10 ppm [10]. Furthermore, there was no
correlation between the level of N2O emission and the
NOx level in the flue gas.
While N2O emissions from conventional combustion
equipment are nowadays recognized to be low, a number
of other sources have been identified as responsible for
significant N2O emission levels which could explain the
increase on N2O in the atmosphere. Sources of N2O in-
clude denitrification of chemical fertilizers, biomass
burning, and advanced combustion and emission control
systems. The selective non-catalytic reduction (SNCR)
process for NOx control has been associated with resi-
dual pollutant emissions. Possible by-products of the
SNCR process included N2O, ammonia slip (NH3), am-
monium sulfate salts, and increased CO emissions. The
contribution of N2O from SNCR is expected to be small.
Reference [11] has estimated that even if 50 percent of
all fossil fuel burned by the U.S. utility industry used
SNCR, the N2O produced would constitute only 0.5 to
2.0 percent of the total global production of N2O. How-
ever, it is prudent to be aware of the potential for N2O
emissions from the SNCR process and to know the ex-
tent to which N2O is a process by-product, as well as, to
know the process parameters and mechanisms leading to
its emission. This information would be important for the
development of control strategies for SNCR-produced
N2O. This paper includes a discussion on N2O formation
in SNCR processes from a chemical kinetics point of
view, and it reports the results of parametric calculations
of N2O emissions using a chemical kinetics model. Em-
phasis is placed on the urea-based SNCR process.
2. Selective Non-Catalytic NOx Reduction
Processes
The SNCR process for NOx removal has been exten-
sively applied in a range of industrial applications.
SNCR is recognized as a NOx control method in the U.S.
Environmental Protection Agency (EPA) NOx rules [12].
It involves the gas-phase reaction of a nitrogen-based
reagent (typically ammonia (NH3) or urea) with NOx at
elevated temperatures to form nitrogen and water. Se-
veral versions of this process have been developed. An-
hydrous ammonia injection, the first generation SNCR
process, works well in the temperature range of 930 to
1035˚C (1700 to 1900°F). In chemistry terms, one mole
of NH3 is required to react with one mole of NOx.
Aqueous ammonia SNCR is functionally similar to an-
hydrous NH3. The major difference is that the reagent is
a water solution of NH3 instead of a gas. In a urea-based
SNCR system, an aqueous solution of urea compound (or
amine salt) is injected into the oxygen-rich upper furnace
and/or high temperature convective section of the boiler
to promote the NOx reduction. The urea-based SNCR
process takes place roughly at a temperature window
between 870˚C to 1150˚C. Cyanuric acid and cyanic acid
injection have also been used as reagents for SNCR
processes. The temperature window for these acids is
similar to the range of temperature in a urea injection
system, 930˚C to 1120˚C.
3. Selective Non-Catalytic NOx Reduction
Theoretical Model
3.1. Chemical Kinetics Model
A chemical kinetics model was developed to describe the
time evolution of the urea-NOx process under homoge-
neous conditions. The numerical model describes an
ideal, one-dimensional plug flow, with the temperature
history approximated by a linear profile. The model as-
sumes the reagent has already been atomized and the
droplets are fully evaporated. Urea decomposition was
modeled as a rapid and instantaneous one step break-
down to NH3 and isocyanic acid (HNCO). The solution
to this problem is simplified because the temperature
history is known and it can described by a set of ordinary
differential equations of the following form:
d
d
kkk
wW Y
t

 (1)
where Yk is the mass fraction of the kth species, t is time,
wk is the molar production rate of the kth species by an
elementary reaction, Wk is the molecular weight of the
kth species, and ρ is the mixture density. The net chemi-
cal production rate wk (the difference between the for-
C opyright © 2011 SciRes. JEP

NO Formation in Selective Non-Catalytic NO Reduction Processes1097
2x
ward and reverse reaction rates) of each species results
from a competition among all chemical reactions in-
volving that species. Each reaction proceeds according to
the law of mass action and the rate coefficients (k) are
configured using a modified Arrhenius form:
exp
nE
kA
TRT






(2)
where the activation energy, E, the temperature exponent,
n, and the pre-exponential constant, A, are parameters in
the model formulation.
The model consists of a detailed chemical kinetic
scheme involving NH3 and HNCO oxidation reactions
and the wet-CO oxidation reactions. The scheme is com-
posed of 105 reversible reactions and 24 species. It com-
prises three complete major reaction subsets: 1) rea-
ctions of N/H/O species that are important in ammonia
oxidation; 2) reactions important in the oxidation of CO;
and 3) reactions involving HNCO and NCO.
Numerical solutions were calculated using Sandia’s
CHEMKIN Collection. The rate constants for reverse
reactions were computed from the equilibrium constants,
calculated using the CHEMKIN Thermodynamic data-
base coefficients [13]. The method of solution used
SENKIN [14] to predict the time-dependent chemical
kinetics behavior of a homogeneous gas mixture. SEN-
KIN uses the DASAC [15] software to solve the nonli-
near ordinary differential equations which describe the
species mass fractions for a system where the tempera-
ture is a specified function of time. The first-order sensi-
tivity coefficients are computed with respect to the ele-
mentary reaction rate parameters.
The kinetic model was validated with experimental
data from a 80 MW coal-fired tangential boiler. Figure 1
shows the comparison between measure and predicted
NOx levels as a function of NSR (NSR is defined as the
actual model ratio of urea to NOx divided by the theo-
retical stoichiometric ratio) for the unit operated at full
load conditions. The data corresponds to single zone urea
injection. Input to the theoretical model included CEM
measurements of NOx, CO, CO2 and O2 in the econo-
mizer; urea injection flow; and temperature information
obtained with an optical temperature probe. No measur-
able ammonia slip occurred at these temperatures and
NSRs.
3.2. N2O By-Product Formation
Since the SNCR process involves chemical activity be-
tween NOx and nitrogen species in a temperature win-
dow between 870˚C to 1150˚C, it is possible for N2O to
be present in the NOx chemical reduction process. Fig-
ure 2 shows the overall chemical path leading to the re-
duction of NO and to the formation of N2O. This simpli-
Figure 1. Comparison between measured and predicted
NOx levels.
Figure 2. SNCR chemical kinetics path.
fication was based on sensitivity and rate-of-production
analyses (evaluation of the order of magnitude of the
reaction rates). The methods used in performing the sen-
sitivity analysis have been described in detail in Refer-
ence [16]. Sensitivity coefficients are displayed as:
 
ik
ik
ki
AX
x
x
XA


 (3)
where βik(x) is the sensitivity coefficient for changes in
the mole fraction of the kth species, the derivate of Xk,
corresponds to changes in the temperature-independent
factor of the ith reaction coefficient, Ai.
Figure 2 shows that N2O is an intermediate step in the
reduction of NO by the species formed from urea. There
is a temperature window in the region from 870 to
1230˚C, where the formation of N2O by the reaction of
cyano species (HNCO, NCO) and NO occurs in high
temperature combustion gases [17]. The N2O formation
path (right-hand-path in Figure 2) involves the formation

NO Formation in Selective Non-Catalytic NO Reduction Processes 1098 2 x
of NCO which subsequently reacts with NO to form N2O
as follows:


2
2
OHHNCONCOH OR1
N
CONONOCOR2

 
At relatively higher temperatures (greater than 1000˚C),
any N2O formed will tend to be rapidly removed by the
reaction:

222
NO OHNHOR3
However, typical applications and SNCR target per-
formance preclude the operation of the SNCR process at
those higher temperatures. As a result, the forward reac-
tions between NO and the cyano species evolving from
reagent breakdown and species oxidation are not kineti-
cally limited. Thus, any N2O formed in the chemical
process can persist in the exhaust gases.
4. Results and Discussion
A series of chemical kinetic calculations was performed
to predict N2O formation from a utility boiler SNCR in-
stallation. The conditions used to generate the results in-
cluded in Figure 5 are typical 100 MW, NOx 523 ppm,
CO 100ppm, CO2 15%, H2O 8%, O2 4%, NSR 1 and
temperature gradient 250 C/sec. The results of these cal-
culations are included in Figure 3. The NOx reduction vs.
temperature curve of Figure 3 is characterized by what is
referred to as left- and right-side of curve. At the left-side
of the curve, NOx decreases with increasing temperature
because reaction rates and OH radical concentrations
increase with temperature, which in turn, increases NOx
reduction. At the right-side, the oxidation reactions of
NH3 and HNCO to NOx become a significant path for the
urea-based reagent and they compete with the NOx re-
duction reactions. These oxidation reaction rates increase
exponentially with temperature. For an NSR of one, the
model indicates that N2O formation occurs in a tempera-
ture window (850˚C to 1050˚C) which is narrower than
the temperature window needed for NOx reduction. This
window coincides with the appearance of NCO, which
confirms the formation of N2O in the urea-SNCR process
through the right-hand-path in Figure 2. N2O peaks at
about 930˚C, which is consistent with the modeling and
pilot scale results reported in Reference [11]. The model
indicates that, although no N2O is present at temperatures
above 1050˚C, significant levels of N2O are present
when the SNCR system is operated on the left-side of the
NOx reduction vs. temperature curve.
The model results were reploted as percent of N2O
converted from NOx (ΔN2O/ΔNOx). The calculations
indicate a peak NOx to N2O conversion of approximately
20 percent (Figure 4). Again, these results coincide with
typical results reported in the literature [18]. However,
reductions of up to 40 percent have been reported in
Figure 3. Effect of NSR on NO x reduction and N2O formation.
Figure 4. NOx reduction and NH3 and N2O emissions vs.
temperature.
Figure 5: Normalized NOx conversion to N2O.
full-scale, utility boiler applications, which suggest that
mixing has an effect on the N2O formation.
4.1. Parametric Calculations
Figure 6 shows the effect of baseline NOx on N2O for-
mation. The urea flow required to achieve stoichiometric
conditions (NSR = 1.0) is also included in the figure.
C opyright © 2011 SciRes. JEP

NO Formation in Selective Non-Catalytic NO Reduction Processes1099
2x
Figure 6. Urea usage and N2O emissions vs. baseline NOx.
Larger urea flows are required to maintain an NSR of 1.0
for increased baseline NOx levels. For the range of base-
line NOx calculated, reduced NOx levels were in the
range of 73 ppm ± 3ppm. N2O emissions from the urea-
based SNCR process increases almost linearly with base-
line NOx. At a release temperature of 980˚C, reducing
baseline NOx by half (from 1000 to 500 ppm) can reduce
N2O emissions by up to 50 percent. In this case, this ef-
fect is kinetically related to the NOx concentration avail-
able for reaction R2.
The effect of carbon monoxide is shown in Figure 7.
The concentration of CO in the flue gas inhibits the NOx
reduction at relatively higher temperatures [19] (greater
than 900˚C). Increased CO levels also have the same
effect on N2O formation. The critical step in CO oxida-
tion is CO + OH → CO2 + H. Because the urea-based
compounds compete for the OH radical, larger concen-
trations of CO slow down the right-hand reaction branch
in Figure 2.
5. Conclusions
Nitrous oxide is not an environmentally regulated species
in the U.S., but it does participate in the stratospheric
ozone chemistry and contributes to the greenhouse effect.
N2O has been found to be a by-product of the selective
non-catalytic process for NOx control for combustion
processes. The SNCR process contribution of N2O to the
atmosphere is expected to be small - less than 2 percent
of the total global production. However, for particular
operating conditions discussed in this work, the process
can increase N2O emissions by almost tenfold, as com-
pared to that produced by the combustion source.
Full-scale applications of the SNCR [20-23] have veri-
fied the formation of significant levels of N2O, and indi-
cated that N2O by-product emission is very closely asso-
ciated with NOx reduction. The literature reports that for
utility size applications provided with urea injection,
N2O emissions are typically 10 to 20 percent of the NOx
reduced, although they can be as high as 30 to 40 percent.
Figure 7. Effect of CO on NOx reduction and N2O formation.
Chemical kinetic calculations demonstrated that the
formation of N2O in the urea-based SNCR process is
linked to the conversion of NO by cyano species released
from the SNCR parent compounds. This conversion oc-
curs within a temperature window of 850 and 1050˚C,
which coincides with the temperature range where NCO
(an intermediate species in the decomposition of urea) is
produced. Maximum N2O production was found to take
place at a temperature of 930˚C. This reiterates that the
N2O production is a problem at the low-temperature side
of the SNCR NOx reduction vs. temperature curve. The
amount of N2O formed depends primarily on the process
temperature, the amount of chemical injected (i.e., NSR),
the initial NOx level, and the CO level in the gas stream.
An explanation for these observations was given based
on the chemical kinetics of the process. With urea injec-
tion, N2O emissions represent up to 20 percent conver
sion of the NOx reduced.
To date, no techniques have been offered commecially
to control N2O emissions from the chemical processes
used in the SNCR technology. The results of this work
indicate that controlling SNCR operation could help to
minimize SNCR-N2O formation. Operation at the right-
hand side of the NOx reduction vs. temperature curve
will translate in lower N2O and NH3 emissions. Reduc-
tion of baseline NOx level will contribute to reductions in
N2O formation and reagent usage.
REFERENCES
[1] R. Weiss, “The Temporal and Spatial Distribution of
Tropospheric N2O,” Journal of Geophysical Research,
Vol. 86, No. C8, 1981, pp. 7185-7195.
doi:10.1029/JC086iC08p07185
[2] R. Lyon, J. Kramlich and J. Cole, “N2O: Sources, Sam-
pling, and Science Policy,” Environmental Science and
Technology, Vol. 23, No. 4, 1989, pp. 392-393.
doi:10.1021/es00181a002
[3] D. Tirpak, “The Role of N
2O in Global Climate and
Stratospheric Ozone Depletion,” Symposium on Station-

N2O Formation in Selective Non-Catalytic NOx Reduction Processes 1100
ary Combustion NOx Control, EPRI CS-5361, 1987.
[4] Energy Information Administration, “Emissions of Green-
house Gases Report,” 2008.
http://www.eia.doe.gov/oiaf/1605/ggrpt/nitrous.html
[5] W. Hao, S. Wofsy, M.McElroy, J. Beer and M. Toqan,
Journal of Geophysical Research, Vol. 92, No. D3, 1987,
pp. 3098-3104. doi:10.1029/JD092iD03p03098
[6] J. Kramlich, J. Cole, J. McCarthy and Wm. Lanier,
“Mechanism of N2O Formation in Coal Flames,” Com-
bustion and Flame, Vol. 77, No. 3-4, 1989, pp. 375-384.
doi:10.1016/0010-2180(89)90142-9
[7] C. Castaldini, R. DeRosier, L. Waterland and H. Mason,
“Environmental Assessment of Industrial Process Com-
bustion Equipment Modified for Low-NOx Operation,”
1982 Joint Symnposium on Stationary Combustion NOx
Control, EPA-600/9-85-022v, 1985.
[8] L. Muzio, M. Teague, J. Kramlich, J. Cole, J. McCarthy,
and R. Lyon, “Error in Grab Sample Measurements of
N2O from Combustion Sources,” Journal of the Air Pol-
lution Control Association, Vol. 39, 1989, pp. 287-293.
[9] R. Roby and C. Bowman, “Formation of N2O in Laminar,
Premixed, Fuel-Rich Flames,” Combustion and Flame,
Vol. 70, No. 1, 1988, pp. 119-123.
doi:10.1016/0010-2180(87)90163-5
[10] L. Muzio, T. Montgomery, G. Samuelsen, J. Kramlich, R.
Lyon and A. Kokkinos, “Formation and Measurement of
N2O in Combustion Systems,” Symposium (International)
on Combusttion, Vol. 23, No. 1, 1990, pp.245-250.
[11] L. Muzio, T. Montgomery, G. Quartucy, J. Cole and J.
Kramlich, “N2O Formation in Selective Non-Catalytic
NOx Reduction Processes,” 1991 Joint Symposium On
Stationary Combustion NOx Control, EPRI GS-7447, 1991.
[12] EPA, “Proposed Rules, Acid Rain Program: Nitrogen
Oxides Emission Reduction Program, 40CFR, Part 76,”
Federal Register, Vol. 57, No. 228, 1992.
[13] R. Kee, F. Ripley and J. Miller, “The CHEMKIN Ther-
modynamic Database,” Sandia Report SAND87-8215,
1987.
[14] A. Lutz, R. Kee and J. Miller, “SENKIN: A Fortran Pro-
gram for Predicting Homogeneous Gas Phase Chemical
Kinetics with Sensitivity Analysis,” Sandia Report
SAND87-8248, 1988.
[15] M. Caracotsios and W. Stewart, “DASAC: Sensitivity
Analysis of Initial Value Problems Including ODE’s and
Algebraic Equations,” Computers and Chemical Engi-
neering, Vol. 92, No. 4, 1985, pp. 359-365.
doi:10.1016/0098-1354(85)85014-6
[16] F. Mauss and N. Peters, “Reduced Kinetic Mechanism for
Premixed Methane-Air Flames,” In: N. Peters and B.
Rogg, Eds., Lectures Notes in Physics, Springer Verlag,
1995, p. 58.
[17] J. Kramlich, J. Cole, J. McCarthy, J. Lanier and J.
McSorley, “Mechanisms of N2O Formation in Flames,”
Fall Meeting, Paper 1A-006, Western States Section, The
Combustion Institute, 1987.
[18] M. Jodal, C. Nielsen, T. Hulgaard and K. Dam-Johansen,
“Pilot-Scale Experiments with NH3 and Urea as Reduc-
tants in SNCR,” 23rd Symposium International on Com-
bustion, Orleans, 22-27 July 1990, p. 2370.
[19] C. Romero and V. Ciarlante, “Sensitivity of the SNCR
Process to Furnace Process Variables,” 1st U.S. DOE
Conference on SCR & SNCR for NOx Control, Pittsburgh,
15-16 May 1997.
[20] T. Hunt, “SNCR Demonstration at Public Service Com-
pany of Colorado’s Arapahoe Station,” 1992 EPRI NOx
Controls for Utility Boiler Seminar, Cambridge, 1992.
[21] D. Teixeria, C. Lin, D. Jones, J. Steinberger, R. Himes, R.
Smith, L. Muzio and S. Okazaki, “Full-Scale Test of
SNCR Technology on a Gas-Fired Boiler,” 1992 EPRI
NOx Controls for Utility Boiler Seminar, Cambridge, 1992.
[22] M. Berg, H. Bering and R. Payne, “NOx Reduction by
Urea Injection in a Coal-Fired Utility Boiler,” 1992 EPRI
NOx Controls for Utility Boiler Seminar, Cambridge, 1992.
[23] G. De Michele, M. Graziadio, P. Massei and M. Cioni,
“SNCR Application on a Tangentially Oil-Fired Boiler,”
ASME, New Orleans, 28 November-3 December 1993.