EPA/600/A-97/1 07
A Review of Nitrous Oxide Behavior in the Atmosphere, and in Combustion
and Industrial Systems
William P. Linak
Air Pollution Technology Branch, MD-65
Air Pollution Prevention and Control Division
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711 USA
John C. Kramlich
Department of Mechanical Engineering, FU-10
University of Washington
Seattle, Washington 98195 USA
For presentation at
Air Pollution in the 21st Century:
Priority Issues and Policy Trends
April 13-17,1997
Noordwijk, The Netherlands
blinak @ inferno.rtpnc .epa. gov
kramlich @ u. washington.edu.
l
-------
ABSTRACT
Tropospheric measurements show that nitrous oxide (N2O) concentrations are increasing
over time. This demonstrates the existence of one or more significant anthropogenic sources, a
fact that has generated considerable research interest for several years. The debate has
principally focused on (1) the identity of the sources, and (2) the consequences of increased N2O
concentrations. Both questions remain open, to at least some degree.
The environmental concerns stem from the suggestion that diffusion of additional N2O
into the stratosphere can result in increased ozone (O3) depletion. Within the stratosphere, N2O
undergoes photolysis and reacts with oxygen atoms to yield some nitric oxide (NO). This enters
into the well known O3 destruction cycle. N2O is also a potent absorber of infrared radiation and
can contribute to global warming through the greenhouse effect
In combustion, the homogeneous reactions leading to N2O are principally NCO + NO ¦-+
N2O + CO and NH + NO -~ N2O + H, with the first reaction being the more important in
practical combustion systems. During high-temperature combustion, N2O forms early in the
flame if fuel nitrogen is available. The high temperatures, however, ensure that little of this
escapes, and emissions from most conventional combustion systems are quite low. The
exception is combustion under moderate temperature conditions, where the N2O is formed from
fuel nitrogen, but fails to be destroyed. The two principal examples are combustion in fluidized
beds, and in applications of nitrogen oxide (NGX) control by the downstream injection of
nitrogen-containing agents (e.g., selective non-catalytic reduction with urea). There remains
considerable debate on the degree to which homogeneous vs. heterogeneous reactions contribute
to N2O formation in fluidized bed combustion. What is clear is that the N2O yield is inversely
correlated with bed temperature, and conversion of fuel nitrogen to N2O is favored for higher-
rank fuels.
Formation of N2O during NOx control processes has been confined primarily to selective
non-catalytic reduction. Specifically, when the nitrogen-containing agents urea and cyanuric
acid are injected, a significant portion (typically > 10%) of the NO that is reduced is converted
•into N2O. Hie use of promoters to reduce the optimum injection temperature appears to increase
the fraction of NO converted into N2O. Other operations, such as air staging and rebuming, do
not appear to be significant N2O producers. In selective catalytic reduction, the yield of N2O
depends on both catalyst type and operating condition, although most systems are not large
emitters.
Other systems considered include mobile sources, waste incineration, and industrial
sources. In waste incineration, the combustion of sewage sludge yields very high N2O
emissions. This appears to be due to the very high nitrogen content of the fuel and the low
combustion temperatures. Many industrial systems are largely uncharacterized with respect to
N2O emissions. Adipic acid manufacture is known to produce large amounts of N2O as a
byproduct, and abatement procedures are under development within the industry.
2
-------
INTRODUCTION
Nitrous oxide (N2O) was long neglected as a pollutant species Ln comparison to the
attention given to the other nitrogen oxides (NOx). Unlike other NOx species, N2O is known to
be extremely inert in the troposphere. This inertness suggested that little environmental
consequence was associated with N2O emissions, so it was relatively easy to consider it a non-
pollutant.
In a more practical vein, the difficulty involved in obtaining N2O measurements probably
contributed to its neglect for many years. At times when NOx were being measured routinely on
many sources, N2O was still a specialty measurement requiring return of batch samples to the
laboratory, followed by a difficult gas chromatographic analysis.
Interest in N2O emissions was largely started by atmospheric chemists, who observed
that the tropospheric concentration was increasing with time at a rate of approximately 0.25%/yr.
This increase suggested the existence of at least one unknown, substantial anthropogenic source.
It also triggered interest in the consequences of this increased tropospheric N2O burden.
Examination of the global N2O budget shows that, while its source (natural and
anthropogenic) is through ground level emissions into the troposphere, its primary sink occurs
through diffusion to the stratosphere. Here, the N2O is finally destroyed by either photolysis or
reaction with singlet oxygen atoms. The result is that a portion of the N2O is converted into
nitric oxide (NO), which enters the ozone (O3) destruction cycle. Thus, increased tropospheric
N2O concentrations can lead to increased O3 removal rates. (It is critical to remember that
considerable N2O is made naturally, and this represents a major contribution to natural O3
destruction in the stratosphere. Hie concern is that increased anthropogenic N2O will accelerate
this natural rate. This differs from the chlorofluorocarbon (CFC) problem where the natural
tropospheric concentrations of CFCs are zero.)
In addition to its impact on stratospheric O3, N2O contributes to global warming. The
N2O molecule is a strong absorber of infrared radiation at wavelengths where carbon dioxide
(C02) is transparent. Although the concentration of N2O is much less than that of CO2, it is a
much stronger absorber on a molecule-by-molecule basis. This suggests that increased N2O
concentrations in the troposphere could lead to more retention of long wavelength radiation
emitting from the surface of the Earth.
The search for the anthropogenic sources has concentrated on (1) industrial processes that
may emit globally significant quantities of N2O, and (2) biological processes that may produce
N20 on a widespread basis. Although there has been extensive work in both of these areas, the
work has been hampered by measurement difficulties. Nonetheless, a substantially improved
picture of the global N2O budget has emerged. Based on this understanding, steps are being
taken to modify the processes that generate anthropogenic N2O.
Certain features have come to be recognized as contributing to N2O emissions from
combustion systems. First and foremost is the oxidation of fuel nitrogen under relatively low-
temperature conditions. This allows N2O to form, and to avoid subsequent destruction. Thus,
any system in which nitrogen in a combined form is oxidized under relatively low temperatures
can lead to N2O emissions. Practical examples include combustion fluidized beds, and NOx
control processes that involve the downstream injection of nitrogen-containing compounds, such
as urea. In most combustion systems, however, the flame temperature is sufficiently high that
any N2O formed in the flame zone is destroyed before the gases are emitted. Thus, most
combustion systems do not emit much N2O.
This paper describes atmospheric behavior and the behavior of N2O in combustion
systems including pulverized coal and fluid bed combustion systems, and thermal waste
remediation. The paper concludes with a review of information on N2O behavior during NOx
control activities, and N2O from both mobile and other industrial sources. Notably missing is
any discussion of N2O from biological activities. Although this is an important component of
the global N2O budget, it falls outside of the scope of the paper.
ATMOSPHERIC CHEMISTRY AND ENVIRONMENTAL CONSEQUENCES
Role of N?Q in Global Warming
3
-------
With an effective surface temperature of approximately 6000 K, most of the sun's
radiation is emitted within a spectral range of 100 to 3000 nm. These wavelengths include the
visible and portions of the ultraviolet and infrared spectra. The Earth's atmosphere is transparent
to most of this incident radiation and, as this radiation reaches the Earth, it is either reflected
back to space or absorbed to heat the surface. To maintain constant temperatures, heat gained by
the sun's incident radiation must be balanced by heat losses through re-radiation. With an
average temperature of approximately 300 K, the Earth emits most of its radiation at infrared
wavelengths above 3000 nm. Unlike incident solar radiation, the Earth's atmosphere is not
entirely transparent to outgoing infrared radiation. Atmospheric gases such as water (H2O),
CO2, methane (CH4), N2O, O3, and more than a dozen synthetic gases such as CFCs and
chlorinated solvents, absorb the Earth's radiation. These gases then re-emit this energy. A
portion is radiated toward space at cooler atmospheric temperatures, and another portion is
radiated back to Earth's surface where it results in additional surface heating. The net result is
increased surface temperatures. This "greenhouse" effect is necessary for the existence of life on
Earth, and accounts for a temperature enhancement from 253 K (-4 °F), the calculated average
surface temperature without a greenhouse effect, to 288 K (59 °F), the Earth's current average
temperature.1-2 Without the greenhouse effect, the Earth would be covered with ice.
While H2O vapor absorbs radiation across the entire spectrum, other predominant
greenhouse gases absorb radiation in distinct bands. These absorption bands are for CO2 (13000
to 17000 nm), CH4 (7000 to 8000 nm), N2O (8000 to 8500 nm), and O3 (9000 to 10000 nm). In
the pre-industrial atmosphere, nearly 80% of the radiation emitted by the Earth was in the
spectral range of 7000 to 13000 nm. This region was referred to as the "window" because of its
relative transparency to outgoing radiation.1 However, as the natural balances of these
atmospheric gases are changed (i.e., increased) through human activities, and as previously
unknown synthetic greenhouse species, such as CFCs, and chlorinated solvents, are introduced
into the atmosphere, this balance is upset, resulting in increased absorption in the CO2, CH4,
N2O, and O3 bands, and new absorption by synthetic gases which absorb strongly in the window
region. This increased absorption decreases the Earth's heat loss, and increases the greenhouse
effect and net global warming. Since surface temperatures also drive climate and hydrological
cycles, the excess energy now available powers changes to weather and precipitation patterns.
These effects cannot be easily predicted, however, because of the nonlinear interactions
involved.1
Table 1 summarizes key greenhouse gas concentrations in the atmosphere.
Comparison is made between pre-industrial concentrations, determined through polar ice core
analysis, and current ambient concentrations. The data indicate a trend of increasing atmospheric
concentrations for these species. Ramanathan1 suggests that atmospheric increases in CH4 and
carbon monoxide (CO), such as those seen during the past century, may have increased
tropospheric O3 concentrations by 20%. Levine4 identifies the relative contribution of several
atmospheric gases to global warming: CO2 (49%), CH4 (18%), CFCs (Refrigerant-11 and -12)
(14%), N2O (6%), and other trace gases (13%).
Role of frbO in Stratospheric €h Chemistry
Approximately 85% of the Earth's atmosphere, including almost all of its H2O vapor, is
associated with the troposphere, which extends from the surface to about 15 km. The remaining
15% is associated with the stratosphere, which extends from approximately 15 to 50 km above
the Earth's surface. Over 90% of the atmospheric O3 resides in the stratosphere. Stratospheric
O3 shields the earth from biologically lethal ultraviolet (UV) radiation (wavelengths below 310
nm). Most importantly, stratospheric O3 shields the earth from UV^B radiation with incident
waVelengths from 280 to 310 nm. These wavelengths are especially harmful because they lie in
a regime where the solar spectrum and DNA (biological) susceptibility overlap.5
In 1929, Chapman6 identified a "classical" mechanism to describe the formation and
destruction of stratospheric O3. According to this mechanism, the chemical production of O3 is
initiated by the photodissociation of molecular oxygen by solar radiation with wavelengths of
242.3 nm or less:
4
-------
02 + hv -+ O + O X<242.3nm ' (1)
Once dissociated, atomic oxygen may combine with molecular oxygen (O2) and a third body, M
[usually nitrogen or O2], to form O3:
O + O2 + M-+O3 + M (2)
O3 destruction can also occur through photodissociation:
03 + hv —~ O + 02 (3)
or through reaction with atomic oxygen:
O3+O-+2O2 (4)
At the time, these reactions were thought to fully describe the global stratospheric O3
balance. However, over the past 20 to 30 years three other destruction routes were discovered
involving reactions with hydroxyl radical (OH):
0H + 03-»H02 + 02 (5)
H02 + O3 -» OH + 202 (6)
HO2 + 0-+ OH + O2 (7)
chlorine (CI):
and NO:
CI + O3 —»CIO + 02 (8)
CIO + O -» CI + 02 (9)
NO + O3 N02 + 02 (10)
NO2 + O-+NO+O2 (U)
Most importantly, these three mechanisms are catalytic in nature, resulting in the destruction of
O3, without the net destruction of the OH, CI, or NO reactanL Thus, these species are recycled
and remain available for numerous O3 destruction steps.
N2Q/NO Chemistry
The stratospheric formation of NO is the result the photolysis of N20 and reaction with
excited singlet-D oxygen, O^D), via the reaction set:
N20 + hv-»N2 + 0(1D) (12)
N20 + 0(lD) ~+ 2NO (13)
N20 + 0(lD) N2 + 02 (14)
Levine2 identifies the photolysis reaction (12) as being responsible for approximately 90% of the
N20 destruction, while Reactions 13 and 14 each accounts for about 5% of its destruction. With
an atmospheric lifetime of approximately 150 years, N20 is extremely long-lived. N20 is also
very stable in the troposphere. Its destruction takes place only after its diffusion into the
stratosphere.2 Reaction 13 leads to the production of stratospheric NO and to the subsequent
chemical destruction of stratospheric O3 through the reaction set described by Reactions 10 and
11, with the net result:
03 + O -~ 202 (4)
Note that NO is not destroyed during this mechanism, and is recycled for further reaction with
O3. Levine2 identifies this catalytic NO cycle as responsible for about 70% of the global
chemical destruction of stratospheric O3. However, a large portion of this contribution is from
natural sources of N20. With the exception of relatively minor emissions from high flying
aircraft, Reaction 13 is the only known source of stratospheric NO. Other N0X species released
into the troposphere as a consequence of combustion and other industrial activities are quickly
removed and do not have the atmospheric lifetimes necessary to reach the stratosphere. Table 2
summarizes the atmospheric concentrations of major gases, selected trace gases (including OH,
CFCs, and NO), and trace nitrogen species compiled by Levine.2-7 It should be noted that N20
is the second most abundant nitrogen species in the atmosphere after molecular nitrogen.
Tropospheric Measurements of N2O
Since the mid 1970s, systematic tropospheric measurements of N20 have been made at
5
-------
locations worldwide.11 These data, summarized by Khalil and Rasmussen12 and presented
in Figure 1, show the atmospheric concentration of N2O to be currently increasing at an average
rate of approximately 0.80 ± 0.02 ppbv per year or approximately 0.27 ±0.01% per year .12
Based on these data, current atmospheric concentrations are estimated to be 313.7 ppbv. It is
also interesting to note that Weiss8 has determined that N2O is unequally distributed between the
northern and southern hemispheres, with the northern hemisphere higher by 0.83 ±0.15 ppbv at
the time that those measurements were made (1976-1980). This may be indicative of the larger
land'mass or larger population centers and industrial activity in the northern hemisphere.
Data by Pearman et at,13 Khalil and Rasmussen, 12»14 Ethridge eta!.,15 and Zardini et
a/.16 examining N2O concentrations in air bubbles trapped in polar ice core samples suggest that
this temporal increase is a relatively recent phenomenon. In polar regions, where yearly snow
falls do not melt, air associated with the snow is trapped in tiny bubbles as subsequent
accumulation and pressure convert older snow to ice. Analysis of the air within these bubbles
can yield information concerning the composition of gases in the atmosphere hundreds,
thousands, or tens of thousands of years ago. These data, summarized for the past 1000 years by
Khalil and Rasmussen12 and presented in Figure 2, include a composite set of measurements
from 0 to 1820 A.D.,14 1600 to 1966 A.D.,15 1600 to 1900 A.D.,16 and 1800 to 1900 A.D.12
Understandably, these data are subject to much more uncertainty compared to the precise
measurements of the past 15 years. Evident from Figure 2 is the absence of any significant trend
between 0 and 1500 A.D. At that time, however, concentrations were seen to suddenly drop and
then rise again. Khalil and Rasmussen12 suggest that this phenomenon was the result of the
"little ice age" which reportedly occurred at this time and which may have resulted in reduced
biological activity. For the period 1880 to 1960, the trend shows a steady increase of 0.07 ± 0.01
ppbv per year. For comparison, the most recent atmospheric data (1976-1988) (see Figure 1 and
Table 1) have also been included, and show an even more expanded rate of increase (0.08 ± 0.02
ppbv). Khalil and Rasmussen12 point out that the ice core data are often imprecise, ambiguous,
and subject to potential errors. Moreover, the most recent ice core data (late 1800s - early 1900s)
are subject to even greater uncertainties due to problems associated with resolving air bubble
formation in ice over short time intervals. The use of ice core data to resolve historical trends of
greenhouse gases is reviewed by Raynaud et a/.17
Tropospheric N2O Balance
Major uncertainties exist concerning the identification and apportionment of the global
sources of N2O. It is known, however, that these global sources must balance the global rate of
atmospheric destruction plus the rate of atmospheric accumulation. Rate data suggest that
Reactions 12,13, and 14 destroy approximately 10.5 ± 3.0 teragrams of nitrogen (in the form of
N2O) per year (Tg N per year). Also, the atmospheric accumulation described above requires the
production of another 3.5 ± 0.5 Tg N per year. As a result, total global production of N2O must
be approximately 14 ± 3.5 Tg N per year to balance these destruction and accumulation terms.2
Table 3 presents estimates of the global sources of NoO published by several groups in
recent years.2,12,18,19,20,21,22,23 The contributions from natural and anthropogenic sources have
been grouped separately for comparison. Natural sources include nitrification and denitrification
of nitrogen species in soils and oceans. Denitrification involves the chemical transformation of
soil nitrate (NO3") to molecular nitrogen (N2) and N2O. Almost all of the N2O produced by
denitrification escapes into the atmosphere. Nitrification involves the oxidation of reduced soil
nitrogen species [such as ammonium (NH4+)] to nitrite (NO2") and NO3" with N2O as an
intermediate product These same processes are responsible for N2O production in oceans.
However, it is uncertain whether denitrification in oxygen-deficient deep waters or nitrification
in oxygen-rich surface waters is responsible.2 Table 3 indicates general agreement among the
four recent studies which have presented natural source data. The single exception is the soils
source published by Khalil and Rasmussen12 which is notably higher than the other estimates.
Table 3 suggests that the global estimates of anthropogenic sources are highly variable,
with values ranging from 1.0 to 9.7 Tg N per year. In addition, the different studies often do not
6
-------
include the same set of sources in their anthropogenic estimates. For example, Khalil and
Rasmussen12 include a sewage source that the others do not IPCC20 also includes estimates
from adipic and nitric acid production which are notably significant. Several data sets12-18'23
also suggest indirect N2O formation through atmospheric transformations from NOx precursors,
or heterogeneous mechanisms involving atmospheric nitrates, although these sources are not well
quantified. Of particular interest is the potential source identified by Khalil and Rasmussen,12
which identifies climatic feedback and accelerated biogenic activity from CO2 increases and
global wanning as being responsible for approximately 0.2 Tg N per year. This estimate was
taken from ice core data and N2O trends seen during the "little ice age." Evident from Table 3 is
that not all of the global sources of N2O have been identified, and that those that have been
identified are subject to large error as indicated by the large range of estimates presented.
However, if we neglect values presented by de Soete18 who summarized anthropogenic sources
only, we see that, while the magnitude of total global sources ranges widely (5.2 to 19.2 Tg N
per year), they bracket the sum of the destruction and accumulation terms determined
independently (14 ± 3.5 Tg N per year).
Table 3 indicates that N2O emissions from fossil fuel combustion sources contribute a
relatively small portion of the total anthropogenic source. However, this was not always
believed to be true. Only recently, fossil fuel combustion, especially coal combustion, was
believed to be the major contributor to the measured increases in ambient N2O concentrations.
These increases also seemed to track measured increases in ambient CO2 concentrations.
Previous research24 presented data indicating direct N2O emissions from coal combustion
exceeding 100 ppm, and an approximate average N20-N:NOx molar ratio of 0.58:1. These data
seemed to confirm earlier suggestions25'26 that combustion of fossil fuels (and coal in particular)
represented a dominant factor in the observed increase of N2O. In addition, emissions factors
generated using N2O stack concentrations of 100 to 200 ppm were adequate to close the global
anthropogenic mass balance.
Additional combustion measurements, gathered by a number of research groups, ,
however, did not always confirm the early results. These numerous studies often used various
N2O sampling and analytical methodologies including samples measured on-line and samples
extracted into containers for subsequent analysis in a laboratory environment. An explanation
for the resulting growing scatter in the data was proposed by Muzio and Kramlich27 who
suggested the presence of a N2O sampling artifact They presented evidence that indicated that
N2O was produced in sampling containers awaiting analysis. They further hypothesized a
mechanism for this formation involving NO, SO2, and H2O, This evidence questioned the
validity of all existing data which involved container sampling. Additionally, since secondary
reactions converting NO to N2O in the sample containers were found to occur easily at room
temperature,27'28'29'30'31 -32 a new indirect relationship between anthropogenic NO emissions
(including combustion) and global N2O increases was suggested (see Table 3).12-18 Although
. not yet characterized or quantified, this may include NO conversion in plumes, in the
troposphere, or on surfaces.33
Since the discovery of the sampling artifact, new research has sought to characterize N2O
emissions from fossil fuel combustion sources, and determine the effect that modifications used
to control NOx has on N2O emissions. For conventional stationary combustion sources
(including coal combustion), recent measurements indicate average N2O emissions less than 5
ppm. These values are more than a factor of 20 times less than emissions believed to be
produced several years ago. Interestingly, fluidized bed combustors, and several of the thermal
DeNOx and catalytic processes developed for NOx control, also seem to contribute to increased
levels of N2O. At present, these technologies are not in widespread application, and the
associated increases in N2O emissions do not add significantly to the global flux. However, as
these technologies are further developed, and their use becomes more common, they have the
potential of affecting global emissions. Several of these technologies will be discussed further in
later sections.
N20 IN GAS-FIRED COMBUSTION
7
-------
Kinetic data show that N2O can be formed as an intermediate in the combustion of fuel
nitrogen. Once the fuel nitrogen is consumed, however, fast destruction reactions ensure that
little escapes the flame. Higher N2O emissions can occur only if the flame is quenched, or the
fuel nitrogen is introduced downstream of the flame zone. In both cases, N2O destruction
reactions are diminished.
Although the qualitative features of the flame behavior of N2O are understood,
quantitative prediction remains uncertain. In particular, recent measurements of the rate of the
critical NCO + NO reaction indicate that (1) its rate is lower at high temperatures than previously
thought, and (2) potentially, only a portion of the reaction branches to N2O. In addition, recent
measurements also suggest that the destruction reaction N2O + OH is approximately 10 times
slower than the rate used in most modeling studies. Since most kinetic modeling has used both
the earlier, higher NCO + NO rate with a branching ratio into N2O of unity, and the higher rate
for N2O + NO, these findings have the potential to upset current views of N2O formation from
cyano species.
Elementary Reactions of N2O Relevant to Gas Flames
Reactions thought to be important to the behavior of N2O in combustion systems include:
N20 + M-4N2 + 0 + M (15)
N20 + O -4 N2 + 02 (16)
—> NO + NO (17)
N20 + H-*N2 + OH (18)
N20 + OH N2 + H02 (19)
NCO + NO-+ N20 + CO (20)
-~N2 + CO2 ¦ (21)
-> N2 + CO + O (22)
NH + NO—» N20 + H (23)
¦The reactions governing NCO concentrations are also important to the problem. These include
reactions forming NCO, and alternate destruction pathways that compete with Reaction 20 for
available NCO. Rates for these reactions must be carefully selected with an awareness that the
understanding is rapidly evolving. In contrast, the reactions and rates governing the
concentration of the other precursor, NH, are more firmly established.
Reaction of Fuel Nitrogen
As noted above, the addition of fuel nitrogen to laminar flames leads to the appearance of
N2O as an intermediate. Significant emissions occur, however, only for low temperature flames
(obtained either through a low adiabatic flame temperature or through high heat extraction at the
burner). Also, addition of fuel nitrogen to large turbulent flames yields no more N2O in the
exhaust than if no fixed nitrogen was included with the fuel.34-35 Modeling efforts suggested
that Reaction 18 is sufficiently fast to be capable of removing all the N2O formed in the flame
zones, even if the N2O formation rate is artificially augmented by unrealistically rapid char
production rates. If N2O is to be emitted from any fossil fuel systems, what could be the source?
One clue is provided by early data on the oxidative pyrolysis of fuel-nitrogen compounds.
Researchers have recognized that, although nitrogen is bound into complex organic structures in
coal, it appears in devolatilization products almost exclusively as HCN and NH3.36'37'38 It is
now known that the HCN and NH3 arise from secondary tar cracking reactions after primary
devolatilization, although the exact mechanism remains an area of research. Early studies on the
pyrolysis39 and oxidative pyrolysis40-41'42'43 of model fuel-nitrogen compounds showed that,
under certain oxidizing, moderate-temperature regimes, compounds such as cyanogen, pyridine,
and HCN could yield large amounts of N2O.
Kramlich et al.^ were able to generate large exhaust concentrations of N2O in a tunnel
furnace by the downstream injection of cyano species. The primary natural gas flame of this
furnace was designed to generate 600 ppm NO. The post-flame gases were quenched by heat
extraction at a rate of 350 K/s. A side-stream injector was used to introduce HCN into the
8
-------
furnace at various locations and temperatures. The results, shown in Figure 3, indicate that at
between 1100 and 1500 K a significant fraction of the HCN was converted to N2O. Similar
results were obtained for an acetonitrile spray, with the maximum conversion temperature offset
somewhat due to the time required for the evaporation of the spray. NH3, however, generated
very little N2O under these conditions.
Application of a plug-flow reaction kinetics model (with distributed side-stream addition)
to these results reproduced the major features of Figure 3: a peak in N2O emissions as a function
of injection temperature, and the lack of N2O production from NH3. Sensitivity analysis showed
that the N2O behavior is governed by:
HCN + Q -~ NCO + H (24)
NCO + NO-* N20 + CO (20)
NH + NO -*¦ N20 + H (23)
N2O+-H-+N2 + OH (18)
Above the favorable temperature window, N2O removal via Reaction 18 was rapid, and alternate
pathways for the oxidation of HCN and its intermediates were opened. At lower temperatures,
HCN failed to react within the time available. In the case of NH3 injection, competing oxidation
reactions within the temperature window prevented Reaction 23 from generating significant
N2O. One feature of the model was the over-prediction of the conversion of HCN to N2O. Use
of a recommended reduced branching ratio44 of 40% was entered into the model, which resulted
in much more realistic N2O predictions. This lends global support to the fractional branching
ratios recently reported.45'46
For such a mechanism to explain N2O emissions in pulverized coal flames, a means of
transporting volatile HCN to these cooler environments must be proposed. While late
devolatilization or turbulent mixing limitations could provide some HCN within the appropriate
temperature window, neither of these is likely to act as a major source in practical systems. This
is consistent with the low N2O emissions reported from oil- and coal-fired furnaces, as will be
discussed shortly.
Industrial Gas Flame Data
The emission of N2O from industrial gas flames has always been found to be quite low.
Figure 4 provides a compilation of data from large-scale, turbulent gas
flames.24'25'32-34'35,47'48'49'50,51'52 The left-hand half of the plot compares NOx and N2O data,
while data in the right-hand panel did not have accompanying NOx data. In general, emissions
are so low as to be of little environmental consequence when it is remembered that the
atmosphere contains approximately 0.3 ppm N2O. Two of these data were obtained with NH3
doping into the fuel,34,35 but this failed to generate significant N2O. The highest concentrations
noted in Figure 4 are associated with (1) a Swedish home heating furnace,49 (2) early data by
Hao et al.,24 and (3) early data by Pierotti and Rasmussen.25
The home heating furnace likely involves lower overall flame temperatures and greater
opportunity for quench of the flame. There is ample evidence that the quench of flames
containing fuel nitrogen will produce higher levels of N2O,53 although it has not been shown that
this is true when fixed nitrogen species are absent from the fuel.
The data of Hao et al.24 were obtained by collection in 50 cm3 syringes that had an
opportunity to age before analysis. Since SO2 was not present in these gas flames to a significant
extent, it is unlikely that a large amount of N2O was generated in the sample.27'29-30 However,
there is evidence that the formation of N2O within sample containers can proceed to a limited
extent, and at a slower rate, in the absence of one of the necessary ingredients: SO2 or
moisture.28 Although this question cannot be conclusively settled in retrospect, the large
quantity of recent data from Austria50 and Japan51-54 suggest that most industrial gas equipment
does not produce more than 2 ppm N2O. Reported concentrations in excess of these values must
be carefully examined to determine if special combustion conditions exist which give rise to *
emissions above the anticipated level.
9
-------
n2o behavior in oil and pulverized coal flames
Pulverized coal flames were among the first industrial combustion sources to be
characterized for N2O emissions. The widespread use of pulverized coal, coupled with the
relatively high amounts of fuel nitrogen contained in the fuel, suggested that this class of sources
was worthy of attention.
* The initial measurements on utility boilers were reported in 1976 25<26 This was followed
by a limited number of studies through 1986-87. It was later found that large amounts of N2O
could be formed within the flasks that were used for sample storage in many studies. Subsequent
studies either modified the storage technique to avoid N2O formation, or used on-line methods
where the samples had little time to age. These later studies have shown very low emissions
from pulverized-coal- or oil-fired units, generally less than 5 ppm. There have been, however,
several studies of time-resolved N2O behavior in coal flames, which indicate that much higher
N2O concentrations exist early in the flame. These studies are valuable for their mechanistic
insight
Early Coal Studies
Pierotti and Rasmussen25 reported three measurements (32.7, 32.8, and 37.6 ppm) from a
pulverized-coal-fired university power plant They used 6 liter, electropolished stainless-steel
flasks for their samples, and the subsequent analysis was by gas chromatograph/clectron capture
detection (GC/ECD). Nominally, these samples would have been subject to the sampling
artifact, although the concentrations measured were much lower that those reported in later
studies. One reason for this might have been the fact that the source and the analytical laboratory
were in close proximity, and the time between sampling and analysis may have been short.
Alternatively, the stainless-steel surface may have moderated the pH of the material absorbed on
the walls and helped to slow the conversion of NO to N2O.
Weiss and Craig26 used 2 liter Pyrex sample flasks to collect stack gas from the Mohave
coal-fired station in Nevada. These were preconcentrated by freezing in a liquid-nitrogen bath.
The concentrated sample was then evaporated and dried. These concentrated gases (containing
mainly N2O and CO2) were analyzed by an ultrasonic phase-shift detector. Thus, these samples
would have also been subject to in-container N2O formation. The reported N2O emission was
25.8 ppm, which is also low compared to subsequent studies where the sampling artifact was,
active.
Kramlich et al.^ measured N2O emissions from a small-scale coal-fired tunnel furnace.
The goal was to identify whether the use of air staging for NOx control would lead to enhanced
N2O emissions. These samples were obtained by an on-line preconcentration procedure similar
to that of Weiss and Craig (Le., the samples withdrawn from the reactor were immediately frozen
in a liquid nitrogen trap and then evaporated, dried, and analyzed on-site by thermal conductivity
gas chromatography). Although the results were insensitive to the stoichiometry of the rich zone,
increased reactor cooling increased N2O emissions, and premixed coal flames yielded more N2O
than diffusion coal flames. Although the elevated N2O emissions (20-90 ppm) could be due to
sample system chemistry, it is possible that the reduced combustion temperatures in this small
facility caused more of the volatile HCN to oxidize in the 1200-1500 K window where N2O
emissions can be formed.34
Castaldini etalexamined a series of sources associated with NOx abatement
operations. Hao et a/.24 surveyed nine oil-fired and three coal-fired furnaces in the northeastern
United States, as well as measurements on a pilot-scale furnace.56 For both Castaldini and Hao a
correlation was noted in which the apparent N2O emissions (obtained using sample flasks and
remote analysis) were about 25% of the NOx emissions measured on-line at the site. This
apparent proportionality was due to the conversion of NOx within the sample flasks to N2O at an
approximate 4:1 ratio. Following the identification of this artifact, both measurement methods
development and rctesting of combustion sources have been priority activities.
Recent Database on Oil and Pulverized Coal Emissions
10
-------
Extensive recent surveys of N2O emissions from pulverized-coal- and oil-fired furnaces
have been reported. Table 4 summarizes sources of these data.23,32,47,50,51,54,57,58,59,60,61 it ^
clear from reviewing this very large database that the majority of the measurements from
industrial combustion systems yield very little N2O. Emissions in excess of 5 ppm are very rare,
with the exception of combustion fluidized beds, which are discussed in the next section. One
general trend is that higher emissions tend to be associated with oil-fired units, although the
emission levels are still quite low. The reason for this is not known.
At these very low emission levels, the care taken in measurement becomes critical.
Simple drying or SO2 scrubbing will prevent the formation of large amounts of N2O. However,
measurement studies show that residual moisture, SO2, and/or long sample lines will still allow a
few ppm of N2O to form. Thus, the preponderance of the data suggest very low values, and the
few outliers warrant close examination to determine if unusual combustion conditions exist, or if
sampling procedures were adequate.
COMBUSTION FLUIDIZED BEDS
Among fossil fueled combustion systems, combustion fluidized beds have consistently
shown the highest N2O emissions in field measurements. Although this unwelcome finding has
become widely recognized only during the last several years, it has spawned a large research
effort. The work has focused on (1) formation mechanisms, (2) emissions as a function of
combustion parameters, and (3) control strategies.
Field Data
Field measurements on various full-scale fluidized bed systems have been reported.
Table 5 summarizes some of the features of these measurements 47-49'50'59'60'62,63,64'65'66'67'68
It must be borne in mind that the measurements reported at the third international workshop on
N2O emissions49 were made before the sample container artifact was discovered, although the
vast majority of these measurements were made under procedures that would minimize the in-
container generation of N2O (i.e., most of the samples were dried before storage). The field tests
consistently indicate emissions substantially above those found from pulverized coal flames or
fuel oil flames.
Attempts to draw general conclusions from the field data are very difficult due to the
wide variation in combustor configurations, operating parameters, and fuel types. In spite of this,
certain trends are clear from the data. First, N2O emissions uniformly decrease with increasing
bed temperature. At the same time, NOx emissions increase. In general, lower rank fuels deliver
lower N2O emissions. Most other operating parameters (excess air, addition of limestone)
appear to have a weak influence on N2O. Carefully obtained sub-scale data have been the means
of advancing our understanding beyond that available from field data.
Summary of Maior Trends
The following presents the major observations reported in the reviews by Mann et al,23
and Hayhurst and Lawrence.69 First, it cannot be overemphasized how difficult it is to extract
mechanistic information from fluid bed experiments. In general, a wide variety of experimental
designs and scales have been used to study N2O formation in fluidized beds. In addition, enough
information is frequently not provided to fully rationalize these data.69 In fluidized beds many of
the parameters are coupled in actual experiments (e.g., excess air and bed temperature) which
makes it difficult to cleanly extract the influence of a single variable. These observations show
the highly empirical nature of the fluidized bed database, and suggest that much care must be
taken in data analysis to develop general conclusions. In spite of these difficulties, considerable
progress has been made in understanding this complex phenomenon.
The most pronounced trend is that of temperature. Reduced bed temperatures almost
universally cause higher N2O emissions, as the measurements presented on Figure 5 show.66
11
-------
Figure 5 also shows that NO increases with bed temperature, suggesting that higher operating
temperature is not the best means of controlling N2O.7® In addition, higher operating
temperatures can cause sintering of added limestone sorbents and reduce SO2 capture.
The influence of fuel type is less significant than that of temperature. In general, lower
rank fuels tend to yield lower N2O emissions. This has been attributed to the tendency of the
lower rank fuels to favor NH3 release over HCN release.36,37 The HCN is acknowledged to be
more efficiently converted to N20.34,69 Alternately, the higher surface area of lower rank fuels
may "promote more complete heterogeneous N2O destruction.70
Other correlations have been observed, including an inverse relationship between the O/N
ratio of the fuel and N20,23<71and an increase with the carbon content of the coal.®3 Although
the specific mechanisms underlying these correlations are unclear, both the O/N ratio and coal
carbon content can, within limits, be indirect indicators of coal rank.
Excess air has been an unusually difficult parameter to distinguish from temperature
because they are so closely coupled in fluidized bed operation. In experiments where the
temperature effect was removed, higher excess air generally increases N2O emissions. Mann et
a/.23 examined the coupling between temperature and excess air, and found that, at higher
temperatures, excess air has less of an effect on N2O emissions, as shown in Figure 6.
Clearly, more experimental work and detailed modeling would help identify the
controlling mechanisms in practical fluidized bed combustion. Many results and conclusions are
contradictory, and subject to more than one interpretation. Nonetheless, fundamental data have
identified two viable mechanisms for N2O formation in fluidized beds:
• Devolatilization of fuel nitrogen as HCN and NH3, followed by oxidation to N2O.
• Oxidation of char nitrogen to NO, followed by the reaction of this NO with char
nitrogen to yield N2O.
On the surface, homogeneous chemistry seems to be capable of explaining many of the
major trends. The release of volatile nitrogen as HCN under fluidized bed conditions yields a
known N2O precursor into an environment where N2O formation is favored. Lower rank coals
are known to yield more of their fuel nitrogen as NH3, which does not convert to N2O as
efficiently under fluidized bed conditions. This tentatively explains the lower emissions
observed with lower rank fuels. Alternately, low rank coals may experience greater
heterogeneous N2O reduction due to their different ash composition and morphology.
Homogeneous chemistry is capable of reproducing the most prominent characteristic of N2O
behavior in fluidized beds, the decrease in emissions with bed temperature. It does so by
showing that the key intermediate, NCO, becomes increasingly converted to NO rather than N2O
as bed temperature increases.
Heterogeneous reactions are also capable of generating N2O. The reduction of NO at a
char surface to yield N2O does not appear to occur if oxygen is not available to expose fresh char
nitrogen 72 In the presence of oxygen, the effective mechanism appears to be the reaction of NO
with exposed char nitrogen to yield N2O rather than the absorption of NO on the surface,
followed by reaction with a second NO.73*74
Extrapolation of these results to practical fluidized beds is more difficult. For example,
the data of Tullin et alJ3 show that heterogeneous N2O formation increases with NO doping, a
feature used to imply that NO reduction at the char surface is the source of the N2O. Other data
for practical beds show no increase in N2O with NO doping.75 This implies that strong N2O
destruction reactions in the bed are active and capable of removing any additional N2O that may
be generated by the reaction of char nitrogen with NO. Since the bed does generate N2O
emissions in spite of this strong reduction reaction, the actual source flux for N2O must be many
times that represented by the emission. The key question is the identity of this source. The
hypothesis is that, in a realistically loaded bed, a large amount of volatile nitrogen passes through
N2O as an intermediate. The emission is only a fraction of the total amount of N2O formed. The
yield of N2O from char nitrogen may be masked by this active volatile chemistry in a
realistically loaded bed.
An important message from these data is that both the volatile combustion and char
oxidation processes involved in N2O formation are coupled in systems operating under practical
conditions. Experimental systems that seek to decouple the process by moving away from
12
-------
practical conditions (e.g., batch processes, light loading of an otherwise inert bed, use of char
instead of coal) are expected to generate sound fundamental data. These results cannot, however,
be directly extrapolated to describe trends in full-scale units. To do so requires a fully coupled
model that correctly integrates all of the fundamental steps.
BEHAVIOR OF N20 DURING NOx CONTROL PROCESSES
.The goal of NOx control procedures is to convert NO into N2 by modifying the
combustion environment, introducing a selective agent, or combining a selective agent with a
catalyst. Since all of these processes involve nitrogenous intermediates, there is an opportunity
for a portion of these intermediates to react and form N2O. This possibility was recognized
early, at least with respect to combustion modifications.35.55
This section will review the influence of the major N0X control technologies on N2O
emissions, including:
Air staging
* Reburning or fuel staging
* Selective non-catalytic reduction (SNCR), including the Thermal DeNOx process
(NH3 injection), the urea injection process, the RapreNQx process (cyanuric acid
injection), and the use of advanced or promoted agents
* Selective catalytic reduction (SCR)
In addition, we will briefly review the steps that have been taken to specifically control N2O
from combustion fluidized beds.
Staged Combustion
Air staging generally refers to the division of the combustion air into at least two streams
such that the fuel is initially processed through a region of reduced oxygen availability. Under
these conditions, conversion of fuel nitrogen to N2 is improved. The second air stream
completes fuel burnout. This basic strategy is executed in a number of ways, including fuel
biasing and "burners out of service." The low-NOx burners that are now available from most
manufacturers are based on providing staged environments within the burner. The division of
the air into two or more streams has been practiced in fluidized beds to create a fuel-rich zone for
improved N0X control.
It is well established that reducing conditions leads to lower N2O emissions from
pulverized coal combustion. Thus, one might suspect that the early formation of N2O would be
reduced during staged combustion. This, however, does not always appear to be the case based
on results from brown coal.76 In spite of the overall fuel-rich conditions in the primary zone,
free oxygen still persists early in the flame, and with it, early N2O.
At the point where the secondary air is added, the fixed-nitrogen species are nominally
distributed among NO, HCN, and NH3. Thus, the oxidation of at least the HCN might be
expected to form N2O. However, the data76 show no N20 at the staging point. One probable
reason for this is the relatively high temperature, above 1000'C. Another is the concurrent
burnout of the CO from the primary zone, which will generate H atoms via CO + OH -~ CO2 +
H, Thus, at the staging point, any N2O which is formed would likely not survive. Earlier work
in a smaller furnace35 failed to find an influence of staging. The emission levels in this facility
were elevated (20-90 ppm) which was attributed to the relatively cool combustion temperature.
Another study examining air staging in laboratory-scale furnaces32 found similar N2O emissions
compared to unstaged operation. TTiis study reported N2O emissions from the staged combustion
of natural gas, No. 2 fuel oil, and No. 5 fuel oil all less than 1 ppm. Coal combustion emissions
ranged from 1 to 5 ppm.
Air staging can be approached in circulating fluidized bed combustors (CFBCs) by
dividing the air injection location. Mann et ai2^ performed a brief examination and found no .
influence on N2O emissions. Likewise, Hiltunen et al.^ found no direct influence of air staging
beyond that attributable to temperature changes, Jahkola et al11 found a weak increase in N2O
with staging, while Shimizu et al78 and Bramer and Valk79 all saw a concurrent decrease in
13
-------
NOx and N2O with increased staging. Hayhurst and Lawrence69 attribute the latter results to
both increasing temperature in the flue gas and the creation of a rich zone at the bottom of the
bed. It is clear from the varying results that staging appears to have only a weak influence on
N2O, and its effect is difficult to separate from other parameters.
Rebuming
. Rebuming or fuel staging involves the addition of a second fuel stream after the primary
fuel burnout is completed. For example, a low-NOx burner can be used to complete the burnout
of the primary fuel. Secondary fuel is added above these burners to create a moderately fuel-rich
zone. Within this zone, radicals generated by the secondary fuel decomposition attack NO to
produce N2, HCN, and NH3. A final air stream is added to burn out the secondary fuel and
convert any remaining reduced nitrogen to NO or N2. Both coal (containing fuel nitrogen) and
natural gas (nitrogen-free) have been proposed as rebuming fuels. Process descriptions and
development history are available in the literature.8®
For coal rebuming, the fuel is introduced under a reduced temperature compared to a
normal industrial coal flame. This suggests that higher conversions of volatile nitrogen to N2O
may occur than would be normally expected. At present, only very limited measurements have
been reported in the open literature. In one subscale study, application of coal rebuming to a
gas-fired primary flame increased N2O emissions from less than I to 11-13 ppm,34 Although in
terms of concentration this increase appears to be small, it represents a 6.5% conversion of fuel
nitrogen to N2O. This is about an order of magnitude greater than the conversion found in the
coal-fired primary flame (0.7%). This is a little surprising because, according to the discussion
on air staging, one would expect that the N2O would be destroyed at the final air staging point
Whether this trend for increased N2O formation extrapolates to large scale is yet to be seen. In
contrast, rebuming with natural gas over a coal-fired primary yielded a greater than 50% N2O
reduction.
In their kinetic modeling study, Kilpinen et al.81 studied natural gas rebuming. They find
•no N2O formation in the fuel-rich zone. If, however, the final air addition temperature is reduced
below 1200 K then some of the HCN from the rich zone is irreversibly converted to N2O. Such
an air injection temperature is, however, too low for practical boiler operation since CO burnout
times would become unaeceptably long. They did not attempt to simulate coal rebuming, in
which such a temperature is far too low to provide adequate char bumout. They do find that the
performance is strongly transport-influenced, so it is difficult to extrapolate the findings to coal.
It is, however, clear that gas rebuming should be a good tool against primap' zone N2O, and that
coal rebuming may either form or destroy N2O, depending on the initial primary zone
concentration.
An approach similar to rebuming has been attempted in fluidized beds, where natural gas
was injected into the cyclone of a CFBC.67 At substantial firing rates (of the order of 10% of the
heat input), N2O reductions of the order of 50% were achieved, compared with kinetic
predictions of 90%. The difference was attributed to the effects of imperfect or finite-rate
mixing, possibly with a contribution due to heat loss. Alternately, the recent data of Glarborg et
al.82 suggest that the rate of the critical N2O + OH destruction reaction may be much slower than
that used in the model. This may account for the discrepancy.
Selective Non-Catalvtic Reduction
Selective non-catalytic reduction (SNCR) had its origins in the observation that NH3
wojild selectively react with NO under appropriate temperatures to yield N2.85 Many years of
work have resulted in an excellent understanding of this process, which is summarized by Miller
and Bowman.84 The following reactions are important:
NH3 + OH -> NH2 + H20 (25)
NH2 + NO -~ N2 + H20 (26)
NH2 + NO-* NNH + OH (27)
14
-------
NNH + M->N2 + H + M (28)
H + O2 —~OH + O (29)
This reaction sequence is self-propagating under the correct conditions. That part of the NH2 that
is consumed by Reaction 27 leads to the generation of the OH radicals (via both Reactions 27
and 29) needed to facilitate Reaction 25. The only significant acknowledged means for
generating N2O is the reaction:
NH + NO —* N20 + H (23)
Both modeling and experimental studies indicate that, while some N2O is formed, it is a very
minor product.34*85'86'87 Faster oxidation reactions effectively compete for the NH under these
conditions.
Alternate agents have been proposed to avoid the handling problems associated with NH3
and to improve NOx removal performance. The principal competitor for NH3 is urea,
CO(NH2)2.88 The urea is injected as either an aqueous solution or a dry powder. Some .
controversy has surrounded the products of the initial thermal decomposition reaction.
Arguments based on consistency between data and modeling suggest that the products are NH3 +
HNCO,65-85 which has been confirmed by experiment.86
Once released into the gas phase, the HNCO reacts primarily according to the following
sequence:89-90
HNCO + OH -+ NCO + H20 (30)
NCO + NO -» N20 + CO (20)
HNCO + H -+ NH2 + CO (31)
In the absence of other reactions, the chain branching associated with NH2 and N20
consumption must support the decomposition of the HNCO. Other reactions (e.g., concurrent
wet CO oxidation) can also provide the radicals needed to drive Reaction 30. The key difference
between urea and NH3 is that urea can generate HNCO, NCO, and N20 as major products of
reaction, NH3 does not. Thus, as is well-known, urea generates substantially more N2O
emissions when used as a SNCR agent.85-87 Typically, less than 5% of the NO reduced is
converted to N20 when NH3 is used. This compares to conversions greater than 10% for urea.
The RapreNOx process is another SNCR process based on the use of cyanuric acid as an
agent,91-92 The cyanuric acid thermally decomposes to yield HNCO, which reacts according to
Reactions 30, 20, and 31 to destroy NO, in the course of which N2O is formed as a byproduct93
Comparison of urea and cyanuric acid as agents shows that urea generates less N2O under
equivalent conditions. This is expected since only half of the nitrogen contained in the urea
becomes associated with HNCO following injection. The other half forms NH3 which does not
yield significant N20. In the case of cyanuric acid injection, all of the nitrogen initially becomes
HNCO, and thus final N20 yields are increased.85
A major limitation of SNCR is the relatively narrow temperature window over which the
agents are active at removing NOx. Also, in large-scale facilities, the NO removal at the
optimum temperature is not complete. Thus, a considerable research effort has been expended to
enhance performance. One approach is the co-injection of combustible compounds (e.g., CO or
H2) with the agents. This has the effect of shifting the optimum temperature window to lower
temperatures. Depending on the amount of free oxygen present and the amount of combustible,
the performance at the new optimum temperature can be either better or worse than the original
unpromoted system. The temperature shifts because the oxidation of the combustible generates
excess free radicals that are needed to initiate the reaction of the agent. Without the reaction of
the combustible, the agent would not react at the lower temperature because it cannot supply
sufficient radicals through chain branching.
. : Kinetic modeling shows that the combustible fuel acts to generate free radicals. These
radicals promote agent decomposition: HNCO + OH —~ NCO + H20. .Next, N20 forms via
NCO + NO. At higher temperatures, the N20 is destroyed by N20 + H. At lower temperatures,
the H-atom is still available, but the relatively high activation energy prevents the N2O + H
reaction from being effective. Thus, as the temperature window for NO removal is shifted to
lower temperatures by the combustible, most of the reactions "follow" the window due to their
weak temperature sensitivity. The N20 + H reaction is the exception due to its strong
temperature dependence, and the N20 produced from the agent reaction fails to react further.
15
-------
The conclusion is that the presence of a combustible may or may not widen the SNCR
window for urea injection, but that it appears to (I) reduce the temperature at which peak N2O
emissions are observed, and (2) increase these peak emissions. Note the similarity between this
situation and that within a FBC. In a FBC, fuel nitrogen in the form of HCN is released by the
coal. This reacts in the presence of oxidizing coal volatiles. Thus, the combustibles within the
volatiles can be viewed as "promoters," which tend to reduce the optimum temperature for NO
reduction and N2O formation. With sufficient combustibles present, as would be the case in a
FBC, this optimum temperature would be reduced below 70G"C. Thus, as temperature is
increased, N2O emissions would be expected to be reduced, and NO emission would increase,
which is the normal characteristic of FBCs.
Recent experimental work suggests that sodium additives may be one effective means of
reducing N2O formation in SNCR. Chen et a/.94 performed tunnel furnace experiments in which
a variety of sodium compounds were co-injected with urea. Figure 7 shows NOx reduction and
N2O emissions as a function of urea injection temperature. The figure shows that urea promoted
by monosodium glutamate gave both an increase in NOx removal performance and a substantial
reduction in N2O emissions. Other sodium compounds, such as Na2C03, also reduced N2O
emissions, although SO2 tended to reduce effectiveness. This could be due to the formation of a
non-reactive sulfate coating on the sodium particles, or through suppression of sodium volatility
due to sulfate formation.
Use of NH3 for NOx reduction in FBCs has been extensively examined. Hie results
indicate that little N2O is generated until NH3 is added in high stoichiometric excess over
NO.65,77-95 Urea injection, however, is strongly correlated with increased N2O emissions.65
However, fixed bed studies using quartz, clay, and ash show that, even with urea injection, N2O
yields are sharply depressed.96 This suggests that appropriate inorganic surfaces can be used to
suppress N2O formation when urea is used as a SNCR agent
Selective Catalytic Reduction
In the present context, selective catalytic reduction (SCR) refers w the reduction of NO
by added NH3 over a catalyst. This distinguishes it from processes involving other agents, such
as CO, H2, and CH4. The process is, at present, applied only to stationary sources of NOX) with
mobile sources being dominated by direct catalytic reduction without use of an agent like NH3.
Selective catalytic reduction is presently being applied to industrial systems in Japan and
Germany, and is coming to increasing use in other parts of the world.
The formation of N2O during SCR was noted in the early 1970s (e.g., Otto et a/.).97 A
very detailed review of the fundamentals of SCR, including the problem of N2O production, is
available.98 In addition, a general review of the application of catalysts to environmental
problems is also available, which includes SCR and other topics.®®
Most SCR systems are based on either noble metal catalysts or vanadium in combination
with other metals and various substrates. Laboratory work suggests that N2O can be a major
product of SCR over noble metal catalysts.100 The amount of N2O formed depends on the state
of the surface, and also on the nature of the substrate.98 The formation appears to be due to a
reaction through a Langmuir-Hinshelwood mechanism between two adsorbed NO molecules.97
The nature of the platinum surface seems to have an effect, with single crystals not yielding
significant N2O. We were unable to find published field measurements on noble metal catalysts,
although unpublished information suggests that a significant portion of the reduced NO can be
converted into N2O in practical installations on gas turbine sets.101
; : In addition to the general review of Bosch and J an ss en,9 8 the problem of N2O formation
over vanadium has been specifically reviewed by Odenbrand et a/.102 In earlier work, the
proposed mechanism involved the reoxidation of the vanadium surface by adsorbed NO to yield
reduced N2O.98 Recent work suggests that N2O arises directly from the reaction of NH3 and NO
at low temperatures, and from direct NH3 oxidation at high temperatures. Because of this
mechanism, the minimum for NOx emissions falls at a lower temperature than the maximum in
N2O emissions.9^.103 The practical consequence is that, below about 30CTC, only negligible
16
-------
N2O is formed, while the formation becomes significant at higher- temperatures due to NH3
oxidation.102 This is supported by long-term, pilot-scale testing of a wide number of vanadium
catalysts.104 Here, no significant NjO was noted, at least to 40CTC. Since catalyst sintering
begins to become a problem above these temperatures,102 this is unlikely to be a common
operating condition. A survey of N2O emissions from 22 SCR installations in Japan indicated no
emissions exceeding 1 ppm.54
Review of the data suggests that, within the broad bounds outlined above, the actual yield
of N2O is highly variable. It depends on catalyst type, and on catalyst treatment (e.g., crystal
size), contamination, support, and background gas composition. It will likely depend on catalyst
age. The results do suggest that vanadium catalysts do not generate significant N2O under their
normal operating conditions, but that noble metal catalysts may.
Summary
In general, NOx control procedures have led to significantly increased N2O if they
promote the reaction of cyano species under reduced temperature conditions. Thus, coal
rebuming may, under some conditions, lead to enhanced N2O due to the release of fuel nitrogen
.under reduced temperatures. The N2O yields are reduced somewhat, however, by the concurrent
oxidation of the volatiles, which leads to N2O scavenging and competitive oxidation of the NCO
intermediate.
The downstream injection of urea and cyanuric acid both lead to N2O formation.
Preliminary data suggest that concurrent injection of combustible promoters (e.g., CO or H2)
leads to increased N2O formation as agent injection temperatures are reduced. The application of
SCR suggests that N2O emissions are negligible from vanadium catalysts if they are operated at
their nominally low temperatures. Noble metal catalysts, however, can convert significant
quantities of NO into N2O during SCR.
THERMAL WASTE REMEDIATION
To date, a limited number of field measurements have appeared describing emissions
from thermal waste remediation activities. Almost all of these have dealt with municipal solid
waste (MSW) or dried sewage sludge. The limited number of results available at present support
only a qualitative description of the trends. Nonetheless, some significant differences between
waste incineration and coal combustion are apparent.
Most of the measurements are on MSW units. The following data have been reported:
• Iwasaki et al.;48 10 units (8 stokers, 1 fluidized bed, 1 batch)
• Yasuda and Takahasi:105 5 units (2 stokers, 3 fluidized beds)
• Hiraki et al. :106 At least one unit
• Watanabe etal.:lQ7 12 units (5 stokers, 5 fluidized beds, 2 rotary kilns)
As illustrated in Figure 8, the most striking variation is the decrease in emissions with
combustion temperature. In spite of the low combustion temperatures, however, the emissions of
N2O rarely exceed 20 ppm. This appears be due at least partly to the relatively low nitrogen
content of the fuel; Iwasaki et a/.48 report fuel-nitrogen contents of about 0.5% and emission
factors of approximately 70 g N20/metric ton waste. This corresponds to a fractional conversion
of fuel nitrogen to N2O of approximately 1%. Interestingly, none of the data suggest a consistent
influence of combustor configuration (fluidized bed vs. stoker/grate) on emissions.
A smaller number of sludge incinerators were also examined, including:
• Iwasaki et a/.:48 4 units (2 multi-stage and 2 fluidized beds)
• Yasuda and Takahasi:105 5 units (4 fluidized beds and 1 rotary grate)
• Hiraki et aZ.:106 At least one unit
These units showed much higher emissions: in the case of Yasuda and Takahasi105 up to 600
ppm. This appears to be primarily a response to the high nitrogen content of the sludge. Iwasaki
et at48 report 5-8% fuel nitrogen, with emission factors corresponding to 400 g N20/metric ton
sludge. This still represents only a 0.5% conversion of fuel-nitrogen. Yasuda and Takahasi105
17
-------
evaluated a sufficient number of units at various temperatures to suggest that higher temperatures
in fluidized beds can reduce the very high N2O emissions associated with sludge combustion.
Interestingly, the NOx emissions were low enough not to be influenced in a significant
way by the change in temperature. For MSW incineration the NOx levels were much
higher.4*5'105 Mixed MSW and sludge incineration appeared to take on the characteristics of
"diluted" sludge incineration {i.e., increased N2O emissions in proportion to the increased
amount of fuel nitrogen).
- Very little work has been reported on other waste treatment activities. Emissions from
liquid injection incineration of high nitrogen wastes have not been reported. The high
temperatures that are typical of these units are not expected to support high N2O emissions.
Fume incinerators, however, may operate at much lower temperatures. Frequently, the fume
represents a relatively inert stream containing dilute fuel-nitrogen compounds. Many of these
fumes have low heating values that must be supplemented by gas fuel to obtain a stable flame.
For economic reasons, the gas usage is minimized, which can yield a low-temperature flame.
Such an environment may favor N2O emissions.
MOBILE SOURCES
While limited, the historical N2O database for mobile sources appeared not to have been
impacted by the sampling artifact. Although mobile source measurements using chassis
dynamometers were often made by sample extraction using Tedlar sampling bags and seldom
performed on-line, SO2 concentrations in mobile source vehicle emissions are many times
smaller those from stationary coal and heavy oil combustion. Thus, the sampling artifact which
dominated measurements from coal-fired utility boilers did not seem to affect measurements
from mobile sources. Mobile source emissions levels established in the early and mid 1980s,
before the artifact issue was brought to light, compare favorably with measurements made in
later years by researchers who were well aware of the potential sampling problems and took care
to avoid the sampling artifact. Table 6 presents a comparison of N2O emission rates for several
. classes of vehicles. 108,109,110 of particular note is the good agreement among the values
reported by the three groups, and the fact that catalytically equipped vehicles emit up to 20 times
more N2O than comparable non-catalyst equipped vehicles.
Based on the results above, a conservative estimate of 62.1 mg N20/km (100 mg
N20/mile), and the 1982 estimate for the U.S. vehicle fleet size (115 x 106) and distance traveled
(2.6 x 1012 km, 1.6 x 1012 miles),108 the total U.S. production of N2O from mobile sources is
approximately 1.6 xlO5 metric tons/yr (1.6 x 1Q11 g KjO/yr, 1,0 x 1011 g N/yr, or 0.1 Tg N/yr).
Assuming that the world fleet size and distance driven per year are three times those of the U.S.,
then worldwide mobile N2O emissions are approximately 0.3 Tg N/yr. This value compares
favorably with the values given in Table 3 and constitutes approximately 8.5% of the total
anthropogenic flux. However, in addition to the uncertainties regarding fleet size and distance
driven used in the analysis above, many research issues remain including the applicability of
different driving cycles to actual use, engine/emission control malfunction or non-optimal
operation, quantification of the number of vehicles that use catalysts (including type of catalyst),
and the influence of ambient conditions. -
EMISSIONS FROM INDUSTRIAL SOURCES
Few industrial sources have been identified as potential emitters of significant N2O. One
receiving recent attention is the manufacture of adipic acid, used primarily in the manufacture of
nylon. By one report111 the manufacture of adipic acid accounts for about 10% of the
anthropogenic flux to the troposphere, based on adipic acid production. This inventory failed to
take into account existing abatement within the industry, and an improved estimate is 5-8% of
the anthropogenic flux.112
Adipic acid is formed by the reaction of cyclohexanone and cyclohexanol under nitric
acid oxidation. The reaction produces an off-gas that contains 1 mole of N2O byproduct for each
mole of adipic acid produced, along with some NOx. This stream, which can contain 30-50 mole
18
-------
% N2O, is usually passed through an absorber to recover the NOx, and then vented to the
atmosphere. Some plants incinerate the stream in process boilers to reduce the NOx, which
coincidentally destroys the N2O. In 1990, approximately 32% of the off-gas streams were abated
in this manner.112
Since the recognition of adipic acid as a significant atmospheric source of N2O, the
industry has launched several cooperative projects to evaluate abatement options,112 Some of
these options have the goal of simply eliminating the N2O from the exhaust stream at the lowest
cost.1 Others focus on converting the N2O into NOx, which can then be used as a nitric acid
feedstock. Approaches currently under evaluation include:
* N2O decomposition over a catalyst to yield N2 or NOx for byproduct recovery.
* High temperature N2O thermal decomposition to yield NOx as a recovered
byproduct.
* Thermal destruction in boilers.
It is recognized that no one technology is likely to be applicable to all plants because of site
specific technical and economic factors.
Other industrial sources that involve the oxidation of nitrogen compounds under
moderate temperature conditions are candidates as N2O emitters. One example mentioned in the
literature are catalytic cracker regenerators.^ These units are used in gasoline manufacture to
regenerate the catalyst used to crack feedstock after it has become coated with a nitrogen-rich
coke. The coke is burned off the catalyst in a fluidized bed. Temperatures are moderate in the
bed, fuel nitrogen levels are high, and the volatile content of the coke is low. Thus, many of the
factors that contribute to high N2O emissions in fluidized bed coal combustion are present. To
date, however, no emissions measurements are reported.
CONCLUDING REMARKS
This review has attempted to bring together the widely scattered literature on the
relatively new problem of N2O emissions from energy conversion and industrial equipment, and
the influence of these emissions on the environment. Some of these results are summarized in
Table 7.18 This is still an evolving area, where changes in both quantitative and qualitative
interpretations are likely. Far from being the last word, this paper will likely act only as a
starting point for future work.
Within homogeneous chemistry, the principal issues include the products of the NCO +
NO reaction, and the rates and products of the reactions consuming HCN and HNCO that give
rise to NCO and other species in moderate temperature combustion. A significant amount of
modeling effort has used rates for NCO + NO that are likely too high. Thus, the adequacy of
homogeneous chemistry to explain N2O yields in processes such as combustion fluidized beds
and in selective non-catalytic reduction needs to be revisited.
A considerable amount of effort has gone into defining the influence of basic operating
parameters on N2O emissions from fluidized beds. While basic operating trends are now known,
a clear mechanistic understanding is not yet complete. The relative roles of homogeneous vs.
heterogeneous N2O production are shrouded by the fact that char behavior is strongly dependent
on the degree of devolatilization. Since fluidized beds contain chars of widely varying ages, the
overall behavior represents an ensemble average. This has clearly complicated the task of
identifying mechanistic information from actual fluidized bed data. Well defined mechanistic
experiments are needed, and these have begun to appear. Particularly useful studies include the
examination of simulated differential fluidized bed elements. Another approach is to
characterize bed response to perturbations in the gas-phase environment {e.g., the addition of
N2O or HCN into the feed air), or in the solid phase (e.g., the spiking of a well characterized
char into a combustor).
ACKNOWLEDGEMENTS/DISCLAIMER
Portions of this work were conducted under EPA Purchase Order 2D1449NAEX with J.
Kramlich. The research described in this article has been reviewed by the Air Pollution
Prevention and Control Division, U.S. Environmental Protection Agency, and approved for
19
-------
publication. The contents of this article should not be construed to represent Agency policy nor
does mention of trade names or commercial products constitute endorsement or recommendation
for use.
references
1. Ramaiiatiian, V., Tbe greenhouse theory of climate change: a test by an inadvertent global experiment,
Science 24,293 (1988).
2. Levine, J. S., The global atmospheric budget of nitrous oxide, and supplement global change: atmospheric
and climatic, 5th International Workshop on Nitrous Oxide Emissions, Tsukuba, Japan (1992),
3. Houghton, J. T., Jenkins, G. J., and Ephraums, J. J., Climate Change: The IPCC Scientific Assessment,
Cambridge University Press, Cambridge, UK (1990).
4. Levine, J. S., Nitrous oxide: sources and impact on atmospheric chemistry and global climate, Presented at:
LNETI/EPA/IFP European Workshop on the Emission of Nitrous Oxide, Lisboa. Portugal (1990).
5. Janetos, A., Overview and update of EPA's stratospheric ozone and global climate program, Presented at:
LNETI/EPA/IFP European Workshop on the Emission of Nitrous Oxide, Lisboa, Portugal (1990).
6. Chapman, $., A theory of upper atmospheric ozone, Quart. J. Ray. MeteroL Soc. 3,103 (1930).
7. Levine, J. S., Impacts of N2O and other trace gases on stratospheric ozone, EPA/IFP European Workshop
on the Emission of Nitrous Oxide from Fossil Fuel Combustion, Rueil-Malmaison, France, June 1988,
EPA-600/9-89-089 (NTIS PB90-126038) (1989).
8. Weiss, R, F„ The temporal and spatial distribution of tropospheric nitrous oxide, J. Geophys. Res. 86,7185
(1981).
9. Rasmussen, R. A., and Khalil, M. A. K., Atmospheric trace gases: trends and distributions over the last
decade. Science 232,1623 (1986).
10. National Oceanic and Atmospheric Administration (NOAA), Geophysical Monitoring for Climate Change,
Vol. 16, B.A. Bodhaine and R.M. Rosson, eds., pp. 67, U.S. Dept. of Commerce, Washington, D.C. (1988).
11. Prinn, R., Cunnold, D., Rassmussen, R. A., Simmonds, P., Alyea, R., Crawford, A., Fraser, P., and Rosen,
R„ Atmospheric emissions and trends of nitrous oxide deduced from 10 years of ALE-GAGE data, J.
Geophys. Res. 95(18), 369 (1990).
12. Khalil, M. A. K., and Rasmussen, R. A., The global sources of nitrous oxide, J. Geophys. Res. 97(D13),
14,651(1992).
13. Pearman, G. L, Etheridge, D., de Silva, F„ and Fraser, P. J., Evidence of changing concentrations of
atmospheric CO2. N2O and CH4 from air bubbles in Antarctic ice, Nature 320(20), 248 (1986).
14. Khalil, M. A. K„ and Rasmussen, R. A., Nitrous oxide: trends and global mass balance over the last 3000
years, Ann. Glaciol. 10, 73 (1988).
15. Ethridge, D. M,, Pearman, G, I., and de Silva, F., Atmospheric trace gas variations as revealed by air
- trapped in an ice core from Law Dome, Antarctica, Ann. Glaciol. 10,28 (1988).
\
16. Zardini, D„ Raynaud, R„ Scharffe, D., and Seller, W„ N2O measurements of air extracted from Antarctic
ice cores: implication on atmospheric N2O back to the last glacial-interglacial transition, J. Atmos. Chem.
8, 198 (1989).
17. Raynaud, D„ Jouzel, J., Bamola, J. M., Chappellaz, J., Delmas, R. J., and Lorius, C., The ice record of
greenhouse gases, Science 259,926 (1993).
20
-------
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33,
34,
35.
36
37,
de Soete. G. G., General discussion, conclusions, and need for future work, Presented at: LNETl/EPMFP
European Workshop on the Emission of Nitrous Oxide, Lisboa, Portugal (1990).
Intergovernmental Panel on Climate Change (IPCC, 1990), taken from Mlnami, K., N2O emissions from
soils and agro-environment, 5th International Workshop an Nitrous Oxide Emissions, Tsukuba, Japan
(1992).
Intergovernmental Panel on Climate Change (IPCC, 1992), taken from Minami, K„ N2O emissions from
soils and agro-environment, 5th International Workshop on Nitrous Oxide Emissions, Tsukuba, Japan
(1992).
Elkins, J. W„ State of the research for atmospheric nitrous oxide (N2O) in 1989, Intergovernmental Panel
on Climate Change, National Oceanic and Atmospheric Administration (1989).
Elkins, J. W., Curient uncertainties in the global atmospheric nitrous oxide budget. Presented ac
LNETI/EPA/IFP European Workshop on the Emission of Nitrous Oxide, Lisboa, Portugal (1990).
Mann, M. D„ Collings, M. E„ and Botros, P. E., Nitrous oxide emissions in fluidized-bed combustion:
fundamental chemistry and combustion testing, Prog. Energy Combust. ScL 18,447 (1992).
Hao, W. M., Wofsy, S. C„ McElroy, M. B„ Be6r, J. M„ and Toqan, M. A., Sources of atmospheric nitrous
oxide from combustion, J. Geophys. Res. 92(D3), 3098 (1987).
Pierotti, D., and Rasmussen, R, A., Combustion as a source of nitrous oxide in the atmosphere, Geophys.
Res. Lett. 3,265(1976).
Weiss, R. F,, and Craig, H., Production of atmospheric nitrous oxide by combustion, Geophys. Res. Lett. 3,
751-753(1976).
Muzio, L. J„ and Kramlich, J. C., An artifact in the measurement of N2O from combustion sources,
Geophys. Res. Lett. 15,1369 (1988).
de Soete, G. G., Parametric study of NjO formation from sulphur oxides and nitric oxide during storage of
flue gas samples, Institut Franyais du Pdtrole, Report No. 36 732-40 ex. (1988).
Muzio, L. J., Teague, M. E., Kramlich, J. C„ Cole, J. A., McCarthy, J, M., and Lyon, R. K„ Errors in grab
sample measurements of N2O from combustion sources, /. Air Pollut. Control Assoc. 39, 287 (1989).
Lyon, R, K., and Cole, J. A., Kinetic modelling of artifacts in the measurement of N2O from combustion
sources, Combust. Flame 77, 139 (1989).
Lyon, R. K., Kramlich, J. C., and Cole, J. A., Nitrous oxide: sources, sampling, and science policy,
Environ. Sci. Technol. 23, 392 (1989).
Linak, W. P., McSorley, J. A., Hall, R. E., Ryan, J. V., Srivastava, R K., Wendt, J. O. L., and Mereb, J. B„
Nitrous oxide emissions from fossil fuel combustion, J. Geophys. Res. 9S(D6), 7533 (1990).
Khalil, M. A. K„ and Rasmussen, R. A., Nitrous oxide from coal-fired power plants: experiments in the
plumes, J. Geophys. Res. 97(D13), 14,645 (1992).
Kramlich, J. C, Cole, J. A., McCarthy, J. M., Lanier, W. S., and McSorley, J. A., Mechanisms of nitrous
oxide formation in coal flames, Combust. Flame 77,375 (1989).
Kramlich, J. C., Nihart, R. K„ Chen, S. L„ Pershing, D. W„ and Heap, M. P., Behavior of N2O in staged
pulverized coal combustion, Combust. Flam 48,101 (1982).
Chen, S. -L„ Heap, M. P., Pershing, D. W,, and Martin, G. B„ Fate of coal nitrogen during combustion,
Fuel 61,1218 (1982).
Chen, S. -L., Heap, M. P., Pershing, D. W., and Martin, G. B„ Influence of coal composition on the fate of
volatile and char nitrogen during combustion, 19th Symposium (International) on Combustion, p. 1271, The
21
-------
Combustion Institute, Pittsburgh (1983).
38. Kramlich, J. C„ Seeker, W. R., and Samuelsen, G. S„ Chemical effects accompanying the. thermal
decomposition of pulverized coal particles, Fuel 67, 1182 (1988).
39. Axworthy, A. E,, Dayan, V. H„ and Martin, G. B., Reactions of fuel-nitrogen compounds under conditions
of inert pyrolysis, Fuel 57,29 (1978)
40. * Houser, T. I., and Lee, P. K., Oxidation of cyanogen, Combust. Sci. Technoi 23,177 (1980).
41. Houser, T. I., McCarville, M. E., and Houser, B. D„ Kinetics of oxidation of pyridine in a flow system.
Combust. Sci. Technoi. 27,183 (1982).
42. Houser, T. J., McCarville, M. E., and Zhuo-Ying, G., Nitric oxide formation from fuel-nitrogen model
compound combustion, Fuel 67,642 (1988).
43. Axworthy, A. E., Kahn, D. R., Dayan, V. H., and Woolery, D. 0., Fuel decomposition and flame reactions
in conversion of fuel nitrogen to NOx, EPA-60Q/7-81-158 (NTIS PB82-108358) (1981).
44. Zahniser, M. S„ McCurdy, K„ and Kolb, C. E„ Data reported in Miller, J. A., and Bowman, C. T.f
Mechanism and modeling of nitrogen chemistry in combustion, Prog. Energy Combust. Sci. 15,287
(1989).
45. Cooper, W. P., and Hershberger, I. F,, Measurement of product branching ratios of the NCQ + NO reaction,
/. Phys. Chem. 96, 771 (1992).
46. Becker, K. H„ Kurtenbach, R., and Wiesen, P., Investigation of the N2O formation in the NCO + NO
reaction by Fourier-transform infrared spectroscopy, 5th International Workshop on Nitrous Oxide
Emissions, Tsukuba, Japan (1992).
47. Muzio, L. J., Montgomery, T. A., Samuelsen, G. S„ Kramlich, J. C, Lyon. R. K., and Kokkinos, A.,
Formation and measurement of N2O in combustion systems, 23rd Symposium (International) on
Combustion, p. 245, The Combustion Institute, Pittsburgh (1990).
48. Iwasaki, Y„ Tatsuichi, S., and Ueno, H., N2O emissions from stationary sources, 5th International
Workshop on Nitrous Oxide Emissions; Tsukuba, Japan (1992)."
49. Ryan, J. V., and Srivastava, R. K., EPA/IFP European Workshop on the Emission of Nitrous Oxide from
Fossil Fuel Combustion, Rueil-Malmaison, France, June 1988, EPA-600/9-89-089 (NTIS PB90-126038)
(1989).
50. Vitovec, W.. and Hackl, A., Pyrogenic N2O emissions in Austria -- measurement at 45 combustion sources,
5th International Workshop on Nitrous Oxide Emissions; Tsukuba, Japan (1992),
51. Yokoyama, T„ Nishinomiya, S., and Matsuda, H., N2O emissions from fossil fuel fired power plants. 5th
International Workshop on Nitrous Oxide Emissions; Tsukuba, Japan (1992).
52. Gaydon, A. G., and Wolfhard, H. G„ Spectroscopic studies of low-pressure flames, 3rd Symposium
Combustion, Flame and Explosion Phenomena, p. 504, Williams and Wilkins, Baltimore, MD (1949).
53. Martin, R. J„ and Brown, N. J„ Nitrous oxide formation and destruction in lean, premixed combustion,
Combust. Flame 80,238 (1990).
54.. • Yokoyama, T., Nishinomiya, S., and Matsuda, H., N2O emissions from fossil fuel fired power plants,
v Environ. Sci Technoi. 25,347 (1991),
55. Castaldini, C., DeRosier, R„ Waterland, L. R„ and Mason, H. B., Environmental assessment of industrial
process combustion equipment modified for low-NOx operation, In Proceedings of the 1982 Joint
Symposium on Stationary Combustion NOx Control, Volume II, EPA-600/9-85-022b (NTIS PB85-235612)
(1985).
22
-------
56. Hao, W. M., Wofsy, S, C, McElroy, M. B., Farmayan, W. F„ Toqan, M. A., Be£r, J, M., Zabniser, M. S., i
Silver, J. A,, and Kolb, C. E., Nitrous oxide concentrations in coal, oil and gas furnace flames, Combust,
Set. Techno!. 55, 23 (1987).
57. Sloan, S. A, and Laird, C. K., Measurements of nitrous oxide emissions from PP. fired power plants,
Atmos. Environ. 24A, 1199 (1990).
58. Laird, C. K., and Sloan, S. A., Nitrous oxide emissions from U.K. power stations, Atmos. Environ. 21 A,
1453 (1993).
Persson, K., N2O emissions from coal fired boilers during thermal NOx reduction tests: Experiences from
SWEDCO-supported projects, Presented at: LNETI/EPA/IFP European Workshop on the Emission of
Nitrous Oxide, Lisboa, Portugal (1990).
Sage, P. W„ Nitrous oxide emissions from coal-fired plant: An update on the Joule collaborative project,
5th International Workshop on Nitrous Oxide Emissions; Tsukuba, Japan (1992).
Soelberg, N. R., Characterization of NOx in utility flue gases, Test Report for Canadian Electrical
Association, Energy and Environmental Research Corp., Irvine, CA (1989).
Amand, L. -E., Leckner, B., Andersson, S., and Gustavsson, L., N2O from circulating fluidized bed boilers-
present status. Presented at LNETl/EPA/IFP European Workshop on the Emission of Nitrous Oxide, Lisboa,
Portugal (1990).
Amand, L.-E. and Leckner, B., Influence of fuel on the emission of nitrogen oxides (NO and N2O) from an
8-MW fluidized bed boiler, Combust. Flame 84,181 (1991).
Amand, L. -E., Leckner, B., and Andersson, S„ Formation of N2O in circulating fluidized bed boilers, 2nd
Nordic Conference on Control of SO 2 and NOx Emissions from Combustion of Solid Fuels, Lyngby,
Denmark (1990).
' 65. Braun, A., Bu, C., Renz, U., Drischel, J., and Koser, H. J. K., Emission of NO and N2O from a 4 MW
fluidized bed combustor with NO reduction, Fluidized Bed Combustion: ASME1991, 709 (1991).
66. _ Brown, R. A., and Muzio, L. J., N2O emissions from fluidized bed combustion, Fluidized Bed Combustion:
ASME 1991, 719(1991).
67. Gustavsson, L., and Leckner, B., N2O reduction with gas injection in circulating fluidized bed boilers,
Fluidized Bed Combustion: ASME 1991, 677 (1991).
68. Hiltunen, M„ Kilpinen, P., Hupa, M., and Lee, Y„ N2O emissions from CFB boilers: experimental results
and chemical interpretation, Fluidized Bed Combustion: ASME 1991, 687 (1991).
69. Hayhurst, A. N., and Lawrence, A. D., Emissions of nitrous oxide from combustion sources, Prog. Energy
Combust. Sci. 18, 529 (1992).
70. W6jtowicz, M. A., Oude Lohuis, J. A., Tromp, P. J. J., and Moulijn, J, A., N2O formation in fluidizcd-bcd
combustion of coal, Fluidized Bed Combustion: ASME 1991,1013 (1991).
71. Alio, M. J„ and Rantanen, J. T., Emissions of nitrogen oxides in pulverized peat combustion between 730
and 900*C, Fuel 68, 586 (1989).
72. de Soete, G. G., Heterogeneous N2O and NO formation from bound nitrogen atoms during coal char
\ combustion, 23rd Symposium (International) on Combustion, p. 1257, The Combustion Institute, Pittsburgh
(1991).
73. Tullin, C. J., Sarofim, A. F., and Be£r, J. M., Formation of NO and N2O in coal combustion: the relative
importance of volatile and char nitrogen, J. Inst. Energy 66,207 (1993).
74. Tullin C. J., Goel, A., Morihara, A., Sarofim, A. F„ and Be6r, J. M., NO and N2O formation for coal
59.
60.
61.
62.
63.
64.
23
-------
combustion in a fiuidized bed: effect of carbon conversion and bed temperature, Energy A Fuels 7,796
(1993).
75. Moritomi, H„ and Suzuki, Y„ N2O emissions from circulating fiuidized bed combustion, 4th China-Japan
Fluid Bed Symposium., Tsukuba, Japan (1991).
76. Hein, K. R. G., N2O emissions from the combustion of low calorie coals-research and large scale
, experience, 5th International Workshop on Nitrous Oxide Emissions; Tsukuba, Japan (1992).
77. Jahkola, A., Lu, Y., and Hippinen, I., The emission and reduction ofNOx andN20 in PFB-combustion of
peat and coal, Fiuidized Bed Combustion: ASME1991, 725 (1991).
78. Shimizu, T., Tachiyama, Y., Souma, M., and Inagaki, M., Emission control of NOx and N2O of bubbling
fiuidized bed combustor, Fiuidized Bed Combustion: ASME 1991,695 (1991).
79. Bramer, E. A., and Valk, M.t Nitrous oxide and nitric oxide emissions by fiuidized bed combustion,
Fiuidized Bed Combustion: ASME 1991, 701 (1991).
80. Chen, S. -L„ McCarthy, J. M„ Clark, W. D., Heap, M. P., Seeker, W. R., and Pershing, D. W., Bench and
pilot scale process evaluation of rebuming for in-furnace NOx reduction, 21st Symposium (International)
on Combustion, p. 1159, The Combustion Institute, Pittsburgh (1987).
81. Kilpinen, P., Glarborg, P., and Hupa, M„ Rebuming chemistry: A kinetic modeling study, Ind. Eng. Chem.
Res. 31,1477 (1992).
82. Glarborg, P., Johnsson, J. E., and Dam-Johansen, K„ Kinetics of nitrous oxide decomposition, Combust.
Hame, 99,523-532 (1994).
83. Lyon, R. K„ Method for the reduction of the concentration of NO in combustion effluents using ammonia,
U.S. Patent 3,900,544 (1975).
•84. Miller, J. A., and Bowman, C. T.f Mechanism and modeling of nitrogen chemistry in combustion, Prog.
Energy Combust. Sci. IS, 287 (1989).
85. Muzio, L. J., Mart/., T. D., Montomery, T. A., Quartucy, G. C., Cole, J. A., and Kramlich, J. C., N2O
formation in selective non-catalytic NOx reduction processes, American Flume Research Committee 1990
Fall International Symposium (1990).
86. Caton, J. A., and Siebers, D. L„ Comparison of nitric oxide removal by cyan uric acid and ammonia,
Combust Sci. Technal. 65,277 (1989).
87. Kilpinen, P., and Hupa, M., Homogeneous N2O chemistry at fiuidized bed combustion conditions: A
kinetic modeling study. Combust. Flame 85,94 (1991).
88. Arand, J. K„ Muzio, L. J., and Setter, J. G„ Urea reduction of NOx
-------
sodium, 1993 ACS Meeting, Denver, CO (1993).
95. Amand, L. -E„ and Andersson, S„ Emission of nitrous oxide (N2O) from fluidized bed boilers,
Proceedings 10th Int. Conf. on FBC, San Francisco, CA (1989).
96. Wall man, P. H., and Carlsson, R. C. J., NOx reduction by NH3: the effects of pressure and mineral
surfaces, Fuel 72,187 (1993).
97. * Olio, K„ Shelef, M., and Kummer, I. T.f Studies of surface reactions of NO by isotope labeling. II.
Deuterium kinetic isotope effects in the ammonia-nitric oxide reaction on a supported platinum catalyst, J.
Phys. Chem. 75, 875 (1971).
98. Bosch, H., and Jans sen, F„ Catalytic reduction of nitrogen oxides: A review on the fundamentals and
technology, CataL Today 2,369 (1988).
99. Armor, J. N„ Environmental catalysis: a review, Appl. CataL B 1,221 (1992).
100. Meier, H„ and Gut, G„ Kinetics of the selective catalytic reduction of nitric oxide with ammonia on a
platinum catalyst, Chem. Engr. Sci. 33, 123 (1978).
101. Muzio, L. J., Fossil Energy Research Corp., personal communication (1993).
102. Odenbrand, C. U. I., Gabrielsson, P. L. T„ Brandin, J. G. M„ and Andersson, L. A. H„ Effect of water
vapor on the selectivity in the reduction of nitric oxide with ammonia over vanadia supported on silica-
titania, Appl. CataL 78,109 (1991).
103. de Soete, G. G„ Nitrous oxide formation and destruction by industrial NO abatement techniques, including
SCR, Institut Franfais du Pdtrole, Report No. 37 755-10 ex., Presented at: Spring 1990 AFRC Meeting,
Tucson, AZ (1990).
104. McGrath, T., Energy and Environmental Research Corp., personal communication (1993).
105. Yasuda, K,, and Takahasi, M., Emissions of nitrous oxide from waste management system, 5th
International Workshop on Nitrous Oxide Emissions: Tsukuba, Japan (1992).
1067 Hiraki, T., Shoga. M., Tamaki, M., and Ota, S., N2O emission factor of stationary combustion sources and
N2O removal effect by ordinary flue gas treatment, 5th International Workshop on Nitrous Oxide
Emissions; Tsukuba, Japan (1992).
107. Watanabe, I., Sato, M„ Miyazaki, M., and Tanaka, M., Emission rate of N2O from municipal solid waste
incinerators, 5th International Workshop on Nitrous Oxide Emissions; Tsukuba, Japan (1992).
108. Sigsby, J. E., N2O emission rates from mobile sources. Session III, EPA Workshop on N2O Emissions from
Combustion, Durham, NC, February 1986, EPA-600/8-86-035 (NTIS PB87-113742) (1986).
109. Prigent, M. F„ and de Soete, G. G., Nitrous oxide (N2O) in engine exhaust gases. A first appraisal of
catalyst impact, 1989 SAE International Congress and Exposition (890492), Detroit, MI (1989).
110. Dasch, J. M., Nitrous oxide emissions from vehicles, J. Air Waste Manage. Assoc. 42(1), 63 (1992),
111. Thiemens, M. H., and Trogler, W. C., Nylon production: An unknown source of atmospheric nitrous oxide,
Science 251,932 (1991).
112.- Reimer, R. A., Parrett, R, A., and Slaten, C. S„ Abatement of N2O emissions produced in adipic acid
manufacture. 5th International Workshop on Nitrous Oxide Emissions; Tsukuba, Japan (1992).
25
-------
Table 1. Summaij of key greenhouse gases (adapted from Ramanathan1, Levine2, and
Houghton etal.1) ¦
C02
CII4
Refrigerant-
11
CCI3F
Refrigerant-
12
CCI2F2
CH3CQ3
CCI4
N2O
Pre-Indus trial
atmospheric
conc. (1750-
1800)
275-280
ppm
0.7-0.8 ppm
0
0
0
0
285-288
ppb
Appro*, current
atmospheric
conc. (1985-
1990)+
345-353
PPm
1-.72 ppm
220-280 ppt
380-484 ppt
130 ppt
120 ppt
304-310
ppb
Current rate of
annual
atmospheric
accumulation
1.8 ppm
(+0.46-
0.5%)
0.015 ppm
(40.9-1.1%)
9.5 ppt
(+4.0-10.3%)
17 ppt
(+4-10.1%)
+15.5%
+2.4%
0.8 ppb
(+0,25-
0.35%)
Projected
atmospheric
conc. mid 21st
century^
400-600
ppm
2.1-4.0 ppm
700-3000 ppt
2000-4800
ppt
350450
ppb
.Atmospheric
lifetime^
50-200 yr
10 yr
65 yr
130 yr
-
-
150 yr
*% = percent by volume, ppm = parts per million by volume, ppb = parts per billion (10^) by volume, ppt = parts
per trillion (10^) by volume.
11990 concentrations are based on extrapolation of measurements reported for earlier years.
$Mid 21st century concentrations are based on current annual rate of atmospheric accumulation, and do not consider
activities aimed at reducing emissions.
§ Atmospheric lifetime is the ratio of atmospheric content to the total rate of removal. CO2 lifetime is a rough
indication of the time necessary for CO2 concentrations to adjust to changes in emissions.
26
-------
Table 2. Selected trace gases, and trace nitrogen species in the atmosphere (adapted from
Levine2-^).
Major gases
Concentration*
< Nitrogen (N2)
78.08%
Oxygen (O2)
20.95%
Argon (Ar)
0.93%
Selected trace gases
Concentration
Water {H2O)
0 to 2%
Carbon dioxide (CO 2)
353 ppm
Ozone (O3)
Tropospheric
0.02 to 0.1 ppm
Stratospheric
0.1 to 10 ppm
Methane (CH4)
1.72 ppm
Refrigerant-12 (CO2F2)
0.48 ppb
Refrigerant-11 (CCI3F)
0.28 ppb
Hydroxyl (OH)
Tropospheric
0.15 ppt
Stratospheric
0.02 to 0.3 ppt
Trace nitrogen species
Concentration
Nitrous oxide (N2O)
310 ppb
Ammonia (NH3)
0.1 to 1,0 ppb
Nitric oxide (NO)^
Tropospheric
0 to 1 ppb
Stratospheric
up to 0.02 ppm
Nitric acid (HNO3)
50 to 1000 ppt
Hydrogen cyanide (HCN)
200 ppt
Nitrogen dioxide (NO2)
10 to 300 ppt
Nitrogen trioxide (N03)*
100 ppt
Peroxyacetyl nitrate
50 ppt
(CH3CO3NO2)
Dinitrogen pentoxide (N205)* .
lppt
Pemitric acid (H02N02)^
0.5 ppt
Nitrous acid (HNO2)
0.1 ppt
Nitrogen aerosols
Ammonium nitrate (NH4NO3)
10 ppt
Ammonium chloride (NH4CI)
0.1 ppt
*% = percent by volume, ppm = parts per million by volume, ppb = parts per billion (10^) by volume, ppt = parts
per trillion (10^) by volume,
tExhibits strong diurnal variation with maximum concentration during the day.
$ Exhibits strong diurnal variation with maximum concentration during the night.
27
-------
Table 3. Estimates of global sources of N2O (Tg N per year).*
de Soete^t Levine2
IPCG19'20 Elkins22 and
Marui et
g/23t
Khalil and
Rasmussen*2
Soils.
Tropical soils
wet forests
dry savannas
Temperate soils
forests
grasslands
Ocean?
Total natural sources
NATURAL SOURCES
3.7
0.01-1,5
1.4-2.6
5.1-7.8
2.2-3.7
0.5-2.0
0.05-2.0
NK§
1.4-2.6
4.2-10.3
3,7
<0.5
1.4-2.6
5.6-6.8
7.6
1.9
9.5
Biomass burning
Sewage
Agriculture
Cattle operations
Irrigation (ground water
release)
Use of nitrogen fertilizers
, . on agricultural fields
leaching into
groundwater
Land use changes
(deforestation)
Fossil fuel combustion
Stationary sources
Mobile sources
Industrial activities
Adipic acid (nylon)
Nitric acid
Global warming
Atmospheric formation
(indirect)
Dry and wet deposition
Total anthropogenic sources
ANTHROPOGENIC SOURCES
0.5(0.4-0.6) 0.1-1.0 0.2-1.0
1.0(0.7-1.3)
0.2 (0.1-0.3)
0.4 (0.2-0.6)
0.01-1.1
0.5-1.1
0.8-1.3
0.1-0.3
NK
2.1 (1.4-2.8) 1.54.8
0.01-2.2
0.1-0.3
0.2-0.6
0.4-0.6
0.1-0.3
1.0-5.0
0.02-0.29
0.5-1.1
0.015-1.4
0.8-1.3
<0.5
0.13-5.0
2.0-9.6
1.0(0.1-1.9)
1.0 (0.2-2.0)
0.3 (0.2-0.6)
0.5 (0.5-1.3)
0.6 (0.3-1.9)
0.4
0.0(0.0-0.1)
0.5 (0.1-1.3)
NK
0.2 (0,0-0.6)
NK
4.5 (1.8-9.7)
Total
2.1 (1.4-2.8) 6.6-12.6 5,2-15.3
7.6-16.4
14.0(11.3-19.2)
*1 Tg = 1 x 106 tonnes = I x 1012 g.
f Anthropogenic sources only.
iData presented in the review by Mann et al,2^ was taken from Elkins.2'
§NK - not known.
28
-------
Table 4. Survey of recent N2O data from combustion systems.*
Andersson et alj
Dahlberg etalJ
Electric Power Research Institute*
Laird and Sloan^
Linak et al^-
Muzio ei
Persson^
Sage®®
Sloan and Laird^
Soelberg®*
Vitovec and Hackl-^
Yokovama et at
Several European plants
17 combustion plants
14 utilities
3 comer-fired, 2 opposed-fired {1 low-NOx)
3 wall-fired coal, 3 wall-fired oil (2 low-NOx)
6 coal-fired
2 oil-fired
7 coal-fired
1 coal-fired
2 stokers
4 wall-fured, 3 comer-fired,
both normal burners and low-NOx burners
11 plants
9 coal-fired plants, including pulverized coal,
lignite, stoker-firing, and briquettes
11 oil-fired plants
4 petroleum-coke plants
7 coal-fired
21 oil-fired
~
Pulverized-coal fired unless otherwise indicated.
^ As reported in Mann et at. ^
-------
Table 5. Summary of field data for combustion fluidized beds.
Am and and Leckner^
Amand et aL^M
Braun et al
65
Brown and Muzio^
Gustavsson and Leckner^?
Hiltunen et alP$
Muzio et al
. Persson®
Ryan and Srivastava^
60*
Sage1
Vitovec and Hackl-^*
One 8 and one 12 MW CFB. Detailed parametric
variations of excess air, lime addition, and char loading.
One 4 MWt AFBC used for plant heating. Examined
influence of bed temperature, fuel-type, and NOx control
strategies.
One AFBC and one CFBC were examined in detail.
Variations included bed temperature, excess air, and
sorbent feed. Also more limited parametric variations at
three cogeneration CFBC sites.
Examined gas injection for N2O control at a 12 MWj
CFBC. Emissions range from 80 to 250 ppm as a function
of temperature. Examined a number of parameters.
Measurements from E CFBCs. Results correlated in terms
of mean furnace temperature and fuel type. Emissions
range from 10 to 140 ppm.
Three CFBCs with N2O ranging from 26 to 84 ppm at 3%
O2. One unit at 3 loads: 55,75,100% of full load, yielding
126,93, 84 ppm N2O, respectively.
One 15 MW PFBC and one 40 MW CFB. Varied bed
temperature in the larger unit and looked at NOx control
agents.
Summary of Third International N2O Workshop; data
presented on 4 Swedish units and 6 Finnish units.
Two AFBCs: 30,22-77 ppm; one CFBC: 91 ppm.
Austrian measurements. Four CFBCs, three AFBCs.
CFBC results showed strong dependence on fuel type:
bituminous coal = 24, lignite = 7.5, bark/sewage
sludge/bituminous coal = 3,3, bark/sewage sludge = 0.8 ppm.
Converted from mg/rtr to ppm.
30
-------
Table 6. Mobile source N2O emission rates (mg N20/km).
Sigsby108t Prigent and Dascb110
de Soete^
Non-catalyst auto 3.1-3.7(5-6)* 2.9(4.8) 1.5-3.0(2.4-1.8)
Catalyst auto 4.3-85.1(7-137) 9.3-62.1 (15-100) oxidation 1.9-41.0 (3-66) oxidation cat
or 3-way cat 16.2-59.0 (26-95) dual bed cat
8.1-62.8 (13-101) 3-way cat.
Diesel trucks/buses 19.3-91.3(31-147)
Gasoline tracks 29.8-60.3 (48-97) 55.3 (89) light duty, 3-way cat
*Numbers in parentheses have units of mg N2O/11U.
^Compilation of test data from several sources.
31
-------
Table 7. N2O emission from fossil fuel combustion (adapted from de Soete^).
Uncontrolled combustion
N2O emissions
Conventional stationary combustion
1-5 ppm
(coal, oil, gas)
Fluidized bed combustion
20-150 ppm
(depends on temperature, oxygen cone,.
physical/chemical properties of fuel)
Diesel engines
0.03 g N per km
(value given for small passenger ears; heavy duty
engine emissions may be b igher)
Gasoline engines
0,01-0.03 g N per km
NOx control technology
Effect on K jO emissions
Fuel staging for conventional stationary
Up to 10 ppmv increase over uncontrolled
combustion
combustion
Thermal DeNOx controls (SNCR)
NH3 injection
3-5% of NOx reduction converted to N2O
Urea or cyan uric acid injection
10-15% of NOx reduction converted to N2O
Catalytic processes
Selective catalytic reduction (SCR)
Limited laboratory studies indicate increased
emissions from some catalysts. No field data
available.
3-way catalysts (gasoline engines)
New catalysts
3-5 times the uncontrolled emissions
Medium aged catalysts
10-16 times the uncontrolled emissions
32
-------
310
jO >¦
£ 302
O
CM
294
1975
1980 1985
Year (A.D.)
1990
Figure 1. Tropospheric measurements of N20 from four independent sets of long term
investigations: D Weiss,8 ° NOAA/GMCC,10 + ALE-GAGE (Prinn et al.)!1, A Khalii and
Rasmussen12 (adapted from Khalii and Rasmussen12).
-------
310
Year (A.D.)
Figure 2, Composite set of atmospheric concentrations of N20 over the last 1000 years
as determined from ice core air bubbles. Horizontal bars indicate time span over which
data are averaged. Recent tropospheric measurements (as annual averages) are
included in upper right corner (adapted from Khalil and Rasmussen12).
-------
250
200
150
100
50
0
1100 1200 1300 1400 1500 1600 1700 1800
Injection temperature (K)
Figure 3. Emissions of N20 from a tunnel furnace with side-stream injection
of: ~ NH3, A HCN, and O acetonitrile at various temperatures
(adapted from Kramlich et al.34).
-------
25
20
15
e
Q.
CL
o
z1 10
0
~ Hao et a/.24
O Muzio et al.47
A Kramlich ef a/.34
+ Kramlich ef a/.35
¦ Linak ef a/,32
• Iwasaki ef a/.48
V Swedish furnace52
~ Swedish home furnace49
x Pierotti and Rasmussen25
A Vitovec and Hackl50
OYokoyama ef a/.51
+
~
OO
0
200
400 600
NOx (ppm)
800
1000
Figure 4. Compilation of field data on N20 emissions from gas-fired equipment.
-------
100
90.
<5* 80.
CO
@ 70.
£
Q.
•9s 60.
50.
40
260
240
CM
o
% 200
@
E 180
r>
-S
O* 160
Z
140
120
860 870 880 890 900 910 920 930
Average bed temperature (°C)
Figure 5. Comparison of NOx and N20 emissions for various
combustion temperatures from a full-scale combustion fluidized bed:
Nuncia 4 56 MW, ~ 57 MW, o 110 MW (adapted from Brown and
Muzio66).
~!
Q
9
00
6
i ¦ 1 > 1 ' 1 ¦ r
-------
350
300
250
« 200
<§> •
£
Q.
& 150
O
-------
Injection temperature (K)
Figure 7. Tunnel furnace results for NO reduction and N20 formation for injection of
urea and urea + monosodium glutamate (M93). Conditions: NOj = 300 ppm, urea
injection ratio = 1.5 (expressed as N/ NO), and M9G equivalent to 50 ppm Na in the
reactor gas (adapted from Chen et al.94).
-------
30
Q.
Q.
o
CM
400 500 600 700 800
Furnace temperature (°C)
900
1000
Figure 8. Relationship between furnace temperature and emissions for a
MSW incinerator (adapted from Iwasaki et al.48).
-------
MDMDI oct-d "D 1>7a TECHNICAL REPORT DATA „ ... .... . ...
INKMrC-L- Jbt I ir~ I c 14 (Please read Instructions on the reverse before completing'
— .......
i
J
1.REPOB ! 2.
| EPA/600/A-97/107
3. RE
4. TITLE AND SUBTITLE
A Review of Nitrous Oxide Behavior in the Atmos-
phere, and in Combustion and Industrial Systems
5. REPORT'DATE" "
6, PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
W. P. Linak (EPA) and J. C. Kramlieh (Univ. of WA)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
University of Washington
Department of Mechanical Engineering (FU-10)
Seattle, Washington 98195
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
EPA Purchase Order
2D1149NAEX
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Published paper; 10^91-9/94
14. SPONSORING AGENCY CODE
EPA/600/13
15.supplementary notcs APPCD project officer is William P. Linak, Mail Drop 65, 919/
541-5792. Presented at Conference, Air Pollution in the 21st Century: Priority
Issues and Policy Trends, Noordwiik, The Netherlands, 4/13-17/97.
16. abstract paper reviews the behavior of nitrous oxide (N20) in both the atmos-
phere and combustion and industrial systems. In the stratosphere, N20 undergoes
photolysis and reacts with oxugen atoms to yield some nitric oxide (NC). This enters
into the well known ozone (03) destruction cycle. N20 is also a potent absorber of in-
frared radiation and can contribute to global warming through the greenhouse effect.
In combustion, the homogeneous reactions leading to N20 are principally NCO + NO
yields N2C + CO and NH + NO yields N20 + H, with the first reaction being the more
important in practical combustion systems. During high-temperature combustion,
N20 forms early in the flame if fuel nitrogen is available. The high temperatures,
however, ensure that little of this escapes, and emissions from most conventional
combustion systems are quite low. The exception is combustion under moderate
temperature conditions, where the N2C is formed from fuel nitrogen but fails to be
destroyed. The two principal examples are combustion in fluidized beds, and in
applications of nitrogen oxide (NOx) control by the downstream injection of nitrogen-
containing agents (e.g., selective non-catalytic reduction with urea). The N20 yield
is inversely correlated with bed temperature, and conversion of fuel nitrogen to N20
is favored for higher-rank fuels.
17. KEY WORDS AND DOCUMENT ANALYSIS
a, DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Mobile Equipment
Nitrogen Oxide (N20) Waste Treatment
Atmospheres Incinerators
Combustion Sewage
Industries Sludge
Ozone Adipic Acid
Pollution Control
Stationary Sources
Global Warming
13B 15E
07B
14G
2 IB
05C 07A
07C
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report}
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Fotm 2220-1 (9-73)
------- |