EP A-600/S-89-089
October 1989
EPA/IFP EUROPEAN WORKSHOP ON THE EMISSION OF NITROUS
OXIDE FROM FOSSIL FUEL COMBUSTION
(Rueil-Malmaison, France, June 1-2,1988)
Prepared by:
Jeffrey V. Ryan
and
Ravi K. Srivastava
Acurex Corporation
Environmental Systems Division
Posf Office Box 13109
Research Triangle Park, North Carolina 27709
EPA Contract No. 68-02-4701
EPA Task Officer:
Joseph A. McSorley
Air and Energy Engineering Research Laboratory
Research Triangle Parfc, North Carolina 27711
AIR AND- ENERGY ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711 "
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are;
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the MISCELLANEOUS REPORTS series. This
series is reserved for reports whose content does not fit into one of the other specific
series. Conference proceedings, annual reports, and bibliographies are examples
of miscellaneous reports.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22181.
ii
-------
ABSTRACT
This report summarizes the proceedings ol an Environmental Protection Agency (EPA)/lnstitut
Francais du Petrole (1FP) cosponsored workshop addressing direct nitrous oxide (N2O) emission from
fossil-fuel combustion. The third in a series, the workshop was held at the IFP in Rueii-Malmaison,
France, on June 1-2,1988.
Increasing atmospheric N2O concentrations have been linked to depletion of stratospheric ozone
(O3) and to global climatic warming. The combustion of fossil fuels has been identified as a potential
major anthropogenic source of N2O. This workshop was organized to accomplish two goals. The first
goal was Jo exchange information among various international research and industrial groups that are
involved in N2O chemistry and modeling as well as N2O measurement. The second goal was to develop
a network for coordinating future efforts.
The 5 technical sessions at this workshop addressed the following issues:
1. Stratospheric O3 depletion and global climate change
2. Mechanisms of N2O formation and destruction during combustion
3. N2O measurement techniques
4. Full-scale field data
5. Practical conclusions based on general discussion
A sampling artifact discovered in the course of an EPRI/EPA sponsored research study revealed
that N2O can be generated in a sample container in the presence of nitrogen oxides (NOx), sulfur dioxide
(SO2). and water. This artifact potentially discredits much of the N2O emissions data collected from
bomb samples containing the above compounds. Recent sampling techniques that minimize the artifact
have produced data from stationary sources that indicate that the direct emission of N2O from fossil-fuel
combustion may not be as great as previously believed. However, limited data indicate that some specific
sources (e.g., fluidized bed combustors) may be high N20 emitters. A standardized sampling protocol
needs to be designed and implemented to validate or supercede current data.
Hi
-------
TABLE OF CONTENTS
Section Page
Abstract . .
Acknowledgement. . .
1 Introduction....
, ii
. ix
. ,1
Opening Session , , , 2
2.1 Stratospheric Ozone Depletion and Global Climate Change {S, Seidel, EPA) . , , , , 2
2.2 OECD Perspective (P. Stoipman, OECD) .4
2.3 Discussion 5
2.4 Impacts of N2O and Other Trace Gases in Stratospheric Ozone (J. Levine, NASA) . .6
2.5 Modeling the Effects of N2O and Other Trace Gases on Climate and Ozone Distribution (C.
Bruhl, Max Planck institut) 11
3 Session!: The Mechanisms ot N2O Formation and Destruction During Combustion . ...... .13
3.1 Basic Kinetics (N. Brown, Lawrence Berkeley Laboratory) 13
3.1.1 Gas Phase Kinetics (J. Kramlich, EERC) 24
3.1.2 Heterogeneous Reactions (G. de Soete, IFP) 33
3.2 Overall Chemical Information From Laboratory Combustor Experiments 44
4 Session 2: N2O Measurement Techniques 59
4.1 Operating Procedure for In-House Research 60
4.2 Water Sorbents Used in N2O Measurement .60
4.3 Interferences From Sulfur Dioxide 62
4.4 Continuous Infrared Analyzer for On-Line N2O Measurement. .66
4.5 Optical Interferometry .70
5 Session 3: Full-Scale Field Data 71
5.1 Flue Gas Exhaust Data From Stationary Sources 71
5.1.1 N2O Emission Data From Test Rigs and Full Scale Units 71
5.1.2 German Power Plants 74
5.1.3 Italian Power Plants 74
5.1.4 Emissions of N2O From British Power Stations 80
5.1.5 Emission Data from Swedish Stationary Combustion Sources 80
5.1.6 Excess Air Variation in a Circulating FBC 87
5.1.7 N2O Emissions From Power Plants in Finland 87
5.1.8 Data from Dutch Power Stations 87
5.1.9 Data From U.S. Power Plants .95
5.2 Exhaust Data From Mobile Sources 95
5.2.1 Comparison of Vehicles With and With Out Catalytic Converters .95
5.2.2 Comparison of N2O Emissions from City and Suburban Driving 101
(continued)
iv
-------
TABLE OF CONTENTS (concluded)
Section Page
6 Session 4; General Discussion to Arrive at Practical Conclusions 104
7 Summary and Conclusions .108
8 References 114
Appendix A: European Workshop on N2O Emissions Agenda .115
Appendix B: European Workshop on N2O Emissions List of Participants 118
v
-------
LIST OF FIGURES
fiflym Page
3-1 N2O formation in methane and hydrogen flames.. . .14
3-2 N2O formation versus equivalence ratio. 15
3-3 N2O formation versus dopant concentration 16
3-4 N2O formation versus nitrogen dopant ...........17
3-5 Edge effects on N2O formation . . . . 19
3-6 Comparison of modeling and experimental results with NH3-doped flame 20
3-7 Comparison of modeling and experimental results with NO-doped flame 21
3-8 Comparison of modeling and experimental results with N20-doped flame.. 22
3-9 Comparison of modeling and experimental results with NH3-doped flame 23
3-10 Kinetics used in model calculations 25
3-11 Post-flame modeling results .27
3-12 Post-flame modeling results (artificial HCN introduction) 28
3-13 Post-flame experiment 29
3-14 Mechanism overview .30
3-15 Reaction mechanisms. 32
3-16 NO and CO over metal (JPB 93) catalyst.. . .35
3-17 Heterogeneous formation and destruction of N2O on metal catalyst 36
3-18 Heterogeneous reduction of N2O on char bound carton 38
3-19 Reaction rates for Eschweiler char 39
3-20 Cedar Grove char conversion of nitrogen .40
3-21 02 + HCN on graphite . .41
3-22 Formation of N2O from bound nitrogen during heterogeneous char oxidation. . ......... 42
3-23 Main reaction paths of N2O formation during char combustion. 43
3-24 Rebum condition using natural gas .46
3-25 NO/N2O for Utah bituminous (N2O measured from sample containers) 47
3-26 NO/N2O for Montana sub-bituminous (N2O measured from sample containers)......... .48
3-27 On-line data in relation to other reported data. . 49
3-28 Fluid bed modeling. 51
3-29 Predictions under varied conditions. .52
3-30 Rebuming predictions 53
3-31 NOx and N2O formation from coal 56
3-32 Schematic of experimental flame measurements 58
4-1 Losses of N2O on different absorbents 61
4-2 N2O sampling system 63
4-3 Effect of SO2 on N2O 64
4-4 Synthetic gas mixtures in glass sample flask 65
4-5 Effect of SO2 level on N2O 67
4-6 Effect of NOx level on N2O formation. 68
4-7 Hypothetical chemistry 69
(continued)
vi
-------
LIST OF FIGURES (concluded)
Figure Page
5-1 Comparison with previous results 76
5-2 Extractive flue gas sampling system. 82
5-3 N2O and NOx concentrations from all plants and combined combustion techniques. ...... 92
5-4 N2O versus NOx emissions 96
5-5 N2O versus NOx emissions compared with Boulder Workshop data .97
5-6 EPRI N2O measurements with repeat condition. 98
5-7 City versus suburban driving 102
5-8 Gasoline versus diesel vehicles. . 103
VII
-------
LIST OF TABLES
Table Paoe
2-1 Atmospheric Nitrogen Species ,10
3-1 Experimental Measurements .54
5-1 N2O Data: PFBC Test Rig 72
5-2 N2O Data: SCR Test/Pilot-Scale 73
5-3 N2O Data: Brown Coal Full-Scale RWE 75
5-4 N2O Emissions From German Power Plants 77
5*5 Data From Facility Burning #6 Fuel Oil. ......... 78
5-6 Data From Facility Burning Coal. 79
5-7 Summary of N2O and Total NOx Emissions 81
5-8 Measurement of N2O Under Varied Conditions , . .83
5-9 Combustion Methods and Fuels in Swedish Combustors 84
5-10 Emissions From Various Combustors .85
5-11 FBC Emissions 86
5-12 Excess Air Variation Data 88
5-13 N2O Field Test Data From Various Combustion Plant Sources Using an FBC , ........ .89
5-14 N2O Field Test Data From Various Combustion Plant Sources Using a Conventional Burner. .90
5-15 N2O Field Test Data From Various Combustion Plant Sources Using Grate Combustion .... 91
5-16 N2O Concentrations in Stack Gas From Electricity Power Stations 93
5-17 Catalytic DeNOx (SCR) 94
5-18 Catalyst Vehicle Versus Carburetor Vehicle ,99
5-19 Catalyst Vehicle Data .100
7-1 N2O Emissions Data Summary .110
viii
-------
ACKNOWLEDGMENT
Many people are responsible for the success of the Rueil-Malmaison workshop. Dr. G. de Soete
of the Institut Francais du Petrole (IFP) deserves special praise for locating and contacting participants
worldwide and for soliciting international interest in characterizing N2O emissions Inom combustion
sources. The authors would like to acknowledge IFP for hosting the workshop and the U.S.
Environmental Protection Agency (EPA) for being the leading organizer and co-sponsor. Special thanks
are extended to Mr. J. McSorley of EPA for his efforts in organizing the workshop.
ix
-------
SECTION 1
INTRODUCTION
This report summarizes the European workshop on direct nitrous oxide (N2O) emissions from
fossil-fuei combustion that was held at the Institut Francais du Petrole (IFP), Rueil-Malmaison, France, on
June 1 -2,1988, The workshop was the third in a series on N2O emissions and was co-sponsored by the
U.S. Environmental Protection Agency (EPA) and the Institut Francais du Petrole (IFP). Its primary focus
was to assess and evaluate the role that combustion of fossil fuels plays in directly emitting N2O into the
atmosphere. This subject is significant because the ambient levels of N2O have been reported to be
increasing and because N2O has been linked to stratospheric ozone (03) depletion and global climatic
warming. Another focus was to establish an international network for exchanging information related to
N2O emissions.
The workshop brought together a complementary group of experts involved in the various
aspects of N2O research. Specifically discussed were the role of N2O in stratospheric O3 depletion and
global climate change, the mechanisms of N2O formation/destruction {involving both homogeneous and
heterogeneous reactions), and the status of N2O measurement techniques. Presentations were also
made on emissions data collected from a variety ot stationary, sub- and full-scale facilities, as well as on
data from mobile sources. After these presentations/discussions, conclusions were reached regarding
the current status and future direction of N2O research.
This report documents both the formal and informal material presented at the workshop, and it
inductes vigorous discussion periods that followed formal presentations. It is an account of the
proceedings and summarizes the workshop's content chronologically.
t
-------
SECTION 2
OPENING SESSION
2.1 STRATOSPHERIC OZONE DEPLETION AND GLOBAL CLIMATE CHANGE (S, SEIDEL, EPA)
Mr. S. Seidel presented an overview of projected stratospheric O3 depletion and global climate
change. He suggests that human activities are responsible tor increases in atmospheric gases'
concentrations in recent times. These changes are causing global warming (a result ot the greenhouse
effect) and stratospheric O3 depletion.
Carbon dioxide (CO2) is considered to be one of the major contributors to climatic warming.
From 1950 to 1985, the ambient concentration of CO2 has increased from 315-345 ppm. This increase in
CO2 concentration of approximately 0.5 percent annually may be mostly attributed to burning fossil
fuels [1].
Chlorofluorocarbons (CFCs) are known to participate in stratospheric reactions that destroy O3.
Used primarily in aerosol spray cans, solvents, refrigerants, and foam blowing, the concentration of CFCs
in the atmosphere has increased 5-7 percent annually since 1975 [2].
Methane (CH4) also plays a major role in O3 depletion and globai climate change. Rice paddies,
fermentation, and natural gas leaks resulting from mining and transportation of natural gas are the
possible major sources of CH4 emissions. Over the past 10 years, CH4 emissions are estimated to be
Increasing at the rate of 1.0 percent per year (3j.
Nitrous oxide (N2O) is another trace gas that contributes to the depletion of O3 in the
stratosphere. Ice core data show steady levels of N2O until the start of the industrial revolution, [4] but
currently ambient concentrations of N2O are increasing at an annual rate of about 0.2 percent.
N2O, CFCs, CH4, and CO2 are essentially transparent to solar radiation, but they trap the earth's
infrared radiation, resulting in the greenhouse effect. Increases in the concentrations ot these gases lead
to increased global warming. There is a need to characterize this phenomenon. Various climate change
2
-------
models thai lake into account the greenhouse effect of these gases are able to predict the temperatures
on earth and other planets reliably, which supports the view that these gases are contributors to the
greenhouse effect.
The modeled predictions of global temperature, using recorded data, suggest thai temperatures
may rise as much as 5 °C worldwide by the year 2030. Moreover, by the year 2025, temperatures will
reach their highest point in the past 100,000 years. Global temperatures have already increased 0 63 °C
since 1880, and the 4 hottest years on record have occurred during the 1980s.
Many global features are expected to be adversely affected by greenhouse climatic changes. For
example, the ecosystem will be thrown off balance as organisms attempt, over a relatively short span, Jo
adapt to environmental changes that would normally occur over thousands of years. Global warming will
melt the polar ice caps, causing the seas to rise and to change coastlines. Warmer temperatures will tax
human health and society; infrastructure decisions, such as resource management and distribution, will
become significantly more difficult, owing to an inability to predict supplies.
Whe comparing a sector analysis of proposed sources of N2O, natural sources are thought to
comprise over 50 percent of total emissions, while combustion of fossil fuels and biomass burning are
considered to be major anthropogenic contributors. A breakdown of combustion by sector shows utilities
and industry as being the major sources. Projections of contributions to N2O emissions suggest a
greater rate of increase as economies and populations grow; the relative contribution by the United States
will decrease, while that of the rest of the world will increase with continued development [5],
EPA was designated by the U.S. Congress to conduct 2 studies related to climate change. The
first was to study the impact of global environment change, and the second was to investigate possible
ways to stabilize atmospheric concentrations of greenhouse gases. The stabilizing report had the
following goals: (1) to develop an analytical framework for the study of N2O; (2) to provide information on
the relative effectiveness of control options that may be available; and (3) to highlight future research
issues for analysis.
Currently, the N2O data base is insufficient. Moreover, there is a critical need to foster research
aimed at developing a better inventory of N2O emissions sources and contributions, to obtain an
3
-------
improved understanding of the combustion process related to N20 emissions, and to improve
understanding of control options. Accelerated N2O research is vital, because critical decisions are being
made to control nitrogen oxides, new combustion devices are being designed and buitt, and developing
nations are growing and contributing to the N2O problem—all with insufficient information.
An international cooperation is essential if nations are to respond effectively to global
environmental threats. Until recently, such cooperation has been nonexistent. For example, the first
international meeting to address the potential environmental impact of CFCs was the Vienna Convention
of 1985. The Montreal Protocol of 1987 was the first cooperative, international accord designed to
regulate anthropogenic emissions.
Mr. Seidel summarized the lessons of the Montreal Protocol as follows:
1. A scientific consensus is necessary among all nations that recognizes the seriousness of O3
depletion.
2. Scientists must focus on emission trends throughout the world.
3. Solutions (i.e., control technologies) must be found, and all nations seeking solutions must
have a common direction.
4. Incremental steps must be followed, because O3 depletion involves long-term problems.
5. Active international programs must be developed because the effects of O3 depletion are
international. Before any agreement can be reached, however, research must be
standardized for a common understanding on emission-control options.
2.2 OECD PERSPECTIVE (P. STOLPMAN. OECD)
Dr. P. Stolpman stated that at this time Organization for European Countries
Development (OECD) is only involved in the issue of global climate change. OECD's involvement in this
area is relatively new and has evolved from a meeting in Washington, DC, among 8 OECD countries, the
World Meteorological Organization (WMO), the Commission of European Communities, OECD, and
representatives of EPA. At this meeting, the participants proposed a program to assess the effects of
worldwide climate changes and to examine preventive policy options.
4
-------
OECD intends to work closely with UNEP, WMO, and the International Energy Agency in
imptementing the program over the next 3 years, focusing on the following tasks:
1. Review the assessment projects being undertaken throughout the world.
2. Develop a computer data base. The United States currently employs AP42, a major
emissions data base. Similarly, OECD hopes to develop a data base like AP42 for
greenhouse gases.
3. Develop alternative emission scenarios.
4. Project environmental and socioeconomic changes.
5. Examine policy options that will address the consequences of atmospheric changes and
prevent further changes.
2.3 DISCUSSION
A brief discussion of Mr. Seidel's and Dr. Stolpman's presentations followed. Professor
A. Williams from University of Leeds asked about the accuracy of the climate change models in view of
feeack mechanisms and pointed out that EPA models differ in the range of global temperature
increases predicted. He asked Mr. Seidel to comment on the accuracy of his predictions.
Mr. Seidel responded that feedback mechanisms, such as clouds, changes in the earth's albedo,
and changes in the oceans, were too complex to be included in the simple climate change models.
These models, however, did include the greenhouse effect associated with CO2 and other relevant
gases. Mr. Seidel further stated that the models seem to agree that a global temperature increase of 4 °C
was imminent for doubled CO2 concentration. The 5 °C increase predicted by the models was based on
increases of CO2 and other trace gases.
Mr. S. Jansson of ASEA PFBC asked how the measurements and temperatures were calculated
for the models.
Mr. Seidel responded that the original estimates were based on measurements over land.
Recent improvements include measurements over the seas thus producing a better average.
5
-------
Dr, S. Lanier asked if the models described by Mr. Seidei assume a 2,7 percent annual growth
rate of CFCs as projected above their current base over the next 100 years, or if they assume a constant
influx.
Mr. Seidei responded that his experience in analyzing future trends in CFC emissions in the
United States and other countries showed that in the past 5-6 years, the growth has been 6-7 percent
globally and 10-13 percent In the United States. The 2.7 percent growth rate used in the models is an
approximation that includes about a 4 percent growth rate in the developing world and a somewhat tower
growth rate in the developed world.
2.4 IMPACTS OF N2O AND OTHER TRACE GASES IN STRATOSPHERIC OZONE (J. LEVINE, NASA)
Dr. J. Levine of Langley Research Center, NASA, began his presentation by expressing NASA's
interest over the last 15 years in the relationship between the origin, evolution, and future of the earth's
atmosphere. NASA has developed theoretical computer models to study this relationship, some of which
are used to predict climate changes in the coming decades. NASA's three major areas of environmental
concern are: the ozone layer depletion, global climate change, and increasing N2O emissions.
Levine believes that in the last 100 years the trace gas composition of the atmosphere has
changed faster than in the past 4.5 billion years, owing to increased human activity. Measurements from
NASA's Nimbus 7 Satellite on Octobers, 1979, using TOMS (Total O3 Measurement System) depicting
O3 distribution in Dobson Units (DUs), shows the O3 distribution to be between 250-350 DUs at the
South Pole. Dr. Levine also presented measurements of the same view of the earth taken 4 years after
the first set of measurements. The second measurement set revealed a 50 percent decrease in O3 (to
150 DUs) over the center of the South Pole, which dramatically represents the Antarctic ozone hole.
The phenomenon shown in Dr. Levine's second measurement set only exists for 4-6 weeks,
during which time the O3 starts to disappear from the South Pole stratosphere as spring begins in the
Southern hemisphere. Dr. Levine suggested that the O3 disappearance probably represents the largest
geophysical perturbation ever measured over so short a period. NASA's studies have also shown that
the total column of O3 in the atmosphere has been depleted by 60 percent and, at certain altitudes, by
95 percent from a decade earlier.
6
-------
Ozone may be second only to oxygen as the earth's most important gas, because O3 has made
possible the evolution and survival of life on the earth's surface. NASA believes that ozone, the only gas
that absorbs ultraviolet radiation, developed some 600 million years ago, about the same time that life
evolved from the ocean to land. Ninety percent of the earth's O3 is found in the stratosphere, and
10 percent is found in the troposphere. If the total mass of the atmosphere were brought to the earth's
surface at standard temperature and pressure, the gas layer would be 8 km thick, while the ozone layer
would be only 2.1 mm thick. However, O3 is actually distributed in a layer located between 15 and 50 km.
When the total column of O3 is reduced by 10 percent, the UV radiation reaching the earth increases by
20 percent. Therefore, the rule of thumb is that the percent increase of ultraviolet radiation reaching the
earth increases by twice the percent decrease of O3.
Dr. Levine also discussed the causes of the Antarctic ozone hole phenomenon. A unique
meteorological situation occurs over the South Pole every spring called the South Polar Vortex. For
4-6 weeks, a mass of air is isolated from the rest of the atmosphere and does not physically interact with
it. Because of its low temperature, some interesting atmospheric chemistry takes place that has been
only recently understood.
In the stratosphere, where 90 percent of the O3 resides, the following equations characterize the
production of O3.
O2 + hu -> O + 0(S242 nm) {1)
02 + O + M O3 + M (2)
In 1929, Dr. Sydney Chapman from the University of Alaska stated that O3 may be destroyed in
two ways. The first is by undergoing photolysis by solar radiation.
03 + h\) -> O2 +0 (< 1100 nm) (3)
7
-------
The second is to react with atomic oxygen and become molecular oxygen, as shown in the
following reaction:
O3 + 0 -> 202 (4)
The reactions (1), (2), (3), and (4) are called the Chapman Reactions.
The first satellite measurements of O3 were made in the early 1970s; however, the O3 level was
found to be much lower than predicted by the Chapman reactions.
Beginning in the early 1970s, three O3 destruction mechanisms were found that involve oxides of
nitrogen (N), hydrogen (H). and chlorine (Ci).
The NOx catalytic loss mechanism is:
NO + O3 -» NO2 +O2
NO2 + O N0 + 02
net: O3 + O -»202
The only source of NO in the stratosphere is the following reaction:
N20 + 0(1d)->2N0
Thus the only source of NO in the stratosphere originates with N2O.
The HOx catalytic loss mechanism is:
OH + O3 -> H02 + 02
HO2 + O3 -> OH + 2O2
net: 2O3 -» 302
The major source of OH in the stratosphere is CH4. Very little water vapor diffuses from the
troposphere into the stratosphere. Methane, however, diffuses up and is chemically transformed to water
vapor.
8
-------
The ClOx catalytic loss mechanism is:
CI + O3 —¦* CIO + O2
CI + O -* CI + O2
net: O3 + O -4 20g
When asked why there is such concern about CFCs, Dr. Levine responded that there are no
natural sources of CI in the stratosphere, but there are two sources of CI in the troposphere: votcanos
and sea salt spray. However, neither of these sources penetrate into the stratosphere. Scientists are
interested in CFCs because they are virtually the only sources of CI in the stratosphere and atmospheric
CFC concentrations are increasing at a very rapid rate.
The catalytic cycle that destroys O3 may vary with altitude. World Meteorological Report #16
states that integrated over all altitudes and all seasons, 60-70 percent of the global O3 destruction is
"caused by NOx, aind that the ultimate source of NO in the stratosphere is N2O.
Certain gases in the atmosphere have the ability to trap infrared (IR) radiation, thereby producing
the greenhouse effect. Dr. Levine suggested that the major greenhouse gas is water vapor, which may
cause 90 percent of the natural greenhouse effect. The next major gas is C02- These two gases absorb
and trap IR radiation, which is then reradiated and directed toward the earth's surface. In this context, Dr.
Levine referred to two EPA reports, "Can We Delay a Greenhouse Warming?" and "Projecting Future Sea
Level Rise." The latter report suggests that an increase in the earth's temperature of 4 °C is amplified at
the North and South Poles to 8-12 8C because of feedback caused by melting ice. The increased
temperature at the poles may cause a significant increase in ice-cap melting.
Dr. Levine noted that the atmospheric window is a wavelength band in 8-12 ^m and is so named
because very little IR absorption occurs within it by the 2 major greenhouse gases (water vapor and
CO2). Any gas that has IR absorption between 8-12 nm can greatly influence the climate. CFCs and
N2O have IR absorption in this band.
After molecular nitrogen at 78 percent by volume, N2O is the second most abundant nitrogen
species in the earth's atmosphere at 330 ppbv (see Table 2-1). Major sources ol N2O are soil, biornass
burning, and fossil fuel combustion. Fertilization of agricultural fields is a major source of N2O. Soil
9
-------
TABLE 2-1. ATMOSPHERIC NITROGEN SPECIES (LEVINE)
SPECIES
SURFACE CONCENTRATION QR RANGE
(RV vmiwwni
(BY VOLUME)
MOLECULAR NITROGEN
NITROUS OXIDE (N^l
AMMONIA (NH3)
NITRIC ACID
HYDROGEN CYANIDE (HCNI
NITROGEN DIOXIDE (N02)
NITRIC OXIDE (NO)
NITROGEN TRIOXIDE (NO3)
PAN {CH3CO3NO2)
DINITROGEN PENTOXIDE m$5)
PERNITRIC ACID
NITROGEN AEROSOLS:
0.1 - l.Oppbv
50 - 1000 pptv
~ 200 pptv
10 - 300 pptv
5 -100 miv2
100 pptv4
50 pptv
i pptv3
0.5 pptv2
0.1 pptv
78.08%
330 ppbv
AMMONIUM NITRATE (NH4NO3)
10 pptv
0.1 pptv
AMMONIUM CHLORIDE (NH4CI)
lppbv = PART PER 81LL10N BY VOLUME = 10"9; pptv = PART PER TRILLION BY
VOLUME = 10-12; THE TOTAL ATMOSPHERIC NUMBER DENSITY AT THE
SURFACE = 2.55 x 1019 MOUCULES cm"3.
Exhibits strong diurnal variation with maximum concentration
DURING THE DAY.
3EXHIB ITS STRONG DIURNAL VARIATION WITH MAX I MUM CONCENTRATION
DURING THE NIGHT.
10
-------
undergoes denitritication, during which nitrate is converted to N20, and nitrification, in which ammonia
eventually forms N2O.
Dr. Levine fell that biomass burning is currently underestimated as a major source of N2O.
Analysis of Defense Meteorological Society Project (DMSP) satellite measurements of the earth at night,
suggests that as much as 1-5 percent of the land area of our planet may be burned every year, primarily
in developing countries. Measurements made by NASA of N2O levels from biomass burning yielded
concentrations of several ppm.
laboratory and field measurements suggest that lightning is not a major source of N2O.
Currently, its importance is believed to be 3 orders of magnitude less important than other sources.
Atmospheric sinks, or loss mechanisms, destroy approximately 10.5 x 1012 g of N2O. Assuming an
annual N2O increase of 0.2-0 3 percent, the amount of N2O in the atmosphere would increase by
3.5 x 1Q12 g. The total of the sinks and the increase (14 x 1012 g) equals the total annual source.
Dr. McElroy and Dr. Wofsy from Harvard University and Dr. Seiler and Dr. Conrad from the Max Planck
Institut have estimated sources that generally agree with the numbers reported by Dr. Levine. [6,7]
In conclusion, N2O affects the stratosphere by chemical conversion to NO and subsequent
destruction of O3. Nitrous oxide affects the troposphere by contributing to the greenhouse effect. Further
research investigating N2O in the stratosphere and troposphere is needed. At this point, the key task is
determining exactly the sources of the measured increase in concentration of N2O in the earth's
atmosphere.
2.5 MODELING THE EFFECTS OF N2O AND OTHER TRACE GASES ON CLIMATE AND OZONE
DISTRIBUTION (C. BRUHL, MAX PLANCK INSTITUT)
Dr. C. Bruhl of the Max Planck Institut for Chemistry in Mainz, Germany, discussed a
one-dimensional photochemical climate model [8]. The model is used to examine variations in globally
and hemispherically averaged vertical profiles of temperature and chemical constituents from preindustrial
times up to the present, and for 4 future scenarios.
The model considers the growing concentrations of CO2, CH, N20, CFCs, hatons, CO, and NOx
owing to anthropogenic causes. One case assumes that emissions of CFCs and halons will be affected
11
-------
by adherence to the Montreal Protocol with no exceptions or delays (about 7 ppbv total CI in 2070 in the
upper stratosphere). The other case assumes only a 15 percent reduction in the rate of emissions (about
9 ppbv total CI in 2070 in the upper stratosphere), Each case then studies increased emissions of N2O
from combustion (13.7 x 1012 g N/yr for the year 2050, [5J) for one scenario and constant N2O emissions
on the 1985 value (3.5 x 1012 g N/yr) for the other scenario. The natural other sources of N2O are
estimated to be 6.6 x 1012 g N/yr. Preindustriai concentrations are deduced from ice core data.
In the case of moderate total CI concentrations (effective Montreal Protocol adherence), below
about 5 ppbv in the upper stratosphere, the depletion of O3 is dominated by the NQx catalytic O3
destruction cycle in the middle stratosphere. For the case of higher total CI concentrations (a reduction in
the emission rates of CFCs and Halons by only 15 percent), the dominant effect of transforming active
chlorine into reservoirs, such as CIONO2 and HCI by NOx, formed in the stratosphere from N2O results in
less depletion of total O3 through increasing N2O emissions. However, the increase in N2O
concentration contributes significantly to the greenhouse effect, resulting in higher global temperatures,
as predicted by the model. The global temperature increase by the year 2075 will be about 2.2 K, of
which 0.12 K will be due to the increase in N2O emissions, while an increase in CO2 alone to 590 ppmv
by that year will cause a 1.4 K rise in global temperature, compared to preindustriai times. This model
takes into account heat fluxes into the oceans, but does not include ice albedo or cloud feedback which is
considered in some global climate models. The healing by N2O is largest in the tropopause region.
12
-------
SECTION 3
SESSION 1: THE MECHANISMS OF N2O FORMATION AND DESTRUCTION DURING COMBUSTION
3.1 BASIC KINETICS (N. BROWN, LAWRENCE BERKELEY LABORATORY)
Dr. N. Brown of Lawrence Berkeley Laboratory was the discussion moderator of the first session.
She described her ongoing investigation, assisted by R. Martin, examining the formation/destruction
chemistry of N2O in premixed, laminar, lean, atmospheric-pressure flames. The study involves both
experimental and modeling efforts. For the experimental work, CH^air/nitrogen compound flames and
H2/02/argon/nitrogen compound flames were chosen. The H2 flames were chosen to provide results
without the complexities of caibon chemistry and to allow a comparison of results to model predictions.
These flames were lean to simulate power plant operating conditions. Experimental variables were bulk
flow rate (28-40 Ipm), to include the effects of heat transfer to Ihe burner, equivalence ratio (0.75-0.9),
and nitrogen additive (NH3, NO, N2O, and N2). Ammonia (NH3) was chosen because it simulates typical
nitrogen compounds found in oil flames; NO was chosen to provide information about NO
destruction/N20 formation; N2O was chosen to provide information about N2O destruction; and N2 was
chosen to provide information for an emissions inventory. The N atom concentration for each additive
was varied between 2,100 and 3,300 ppm. Thermochemistry, probe, and quenching effects were also
studied.
Very little difference was found in the amounts of N2O formation in the CH4 and H2 flames
(Figure 3-1). N2O occurred as an early flame intermediate in concentrations less than 20 ppm but was
minimal (<1 ppm) in post-flame regions. N2O formation increased with decreasing flame temperature,
decreasing equivalence ratio (Figure 3-2), and increasing dopant concentration (Figure 3-3). N2O
concentration for a dopant varied in the order NH3 > NO > N2 (Figure 3-4). Nitrous oxide concentrations
were found to be highest at the burner edge, possibly because of an interaction with cool surroundings
13
-------
15
<>o
0
<> I O 1 ^0
ODU/alr
~ HZ/otr
3 2 4 6 8 10
Height
-------
10
8-
D <~
CD
»
E
CL
CL
C
fM
Z
a
+
OPh t - a 76
~ Phi - a 98
+Phi - a 91
UH
~
4 6
Height (mm)
10
Figure 3-2. N2O formation versus equivalence ratio. (Brown)
15
-------
15
A
a
g 1 Oh
E
Q. O
3- op
o
ry
:z:
O 2450 ppn m
0 3300 ppa >M3
png d^> l d 0 | d a
Hq i ght (mm)
Figure 3-3. N2O formation versus dopant concentration. (Brown)
10
16
-------
30
25
| 20
E
g; 15
8 10
z
5
°0 2 4 6 8 10
Height (mm)
Figure 3-4. N2O formation versus nitrogen dopant. (Brown)
17
-------
leading to quenching (Figure 3-5). Conversion of N20 to NO was less than 8 percent, and the ratio
N2O/NO was less than 2 percent.
For modeling efforts, Dr. Brown used Sandia codes Premix with lull chemistry and transport,
Chemkin with full chemistry, and Thermochemistry—1987 Chemkin release. Dr. Brown stressed the
importance of a thermodynamic data base on model predictions. The results of her model predictions and
H2 flame experiments are shown in Figures 3-6 to 3-9. For modeling purposes, actual measured
temperature profiles were used rather than solutions of the energy equation.
Figure 3-6 depicts an NH3 doped flame at temperature of 1,750 K. There is an excellent
agreement (within 2 percent) between model predictions and experimental measurements. Figure 3-7
shows an NO doped flame at about 1,750 K; the agreement between model and experiments is not quite
as good. NO agrees within 6 percent, while measurements of N2 are scattered, possibly because of
measurement errors of the order of 25 ppm over measured N2 values of the order of 75 ppm. Figure 3-8
depicts a N2O doped flame at about 1,750 K; the model overpredicts NO and underpredicts N2
Figure 3-9 shows results from a cooler NH3 llame at about 1,250 K. Changes in flame chemistry at lower
temperatures cause a decrease in modeling ability. Thus the agreement between model and experiments
is lacking for N2O and NO. Also noticeable is a slower decay of N2O with height as opposed to a faster
decay observed in higher temperature flames. Ongoing work by Dr. Brown has shown large chemistry
changes between 1,100 and 1,250 K. Thus temperature is an extremely important parameter in modeling
efforts.
In N2O formation chemistry, the dominant reaction in H2 flames is NH + NO -» N2O + H and is
3,000-6,000 times more dominant than NH2 + NO -4 N2O + H2 reaction. In hydrocarbon flames,
NCO + NO -» N2O + CO seemed dominant. In both types of flames, the destruction mechanism of
N2O + H -> N2 + OH seemed most likely, although N2O + H -» NH + NO is another path. The branching
ratio of above two destruction paths is temperature dependent and is not well understood.
Dr. Brown concluded that agreement between the experiment and modeling was satisfactory near
1,700 K, but the chemistry at temperatures less than 1,100 K needs more clarification. Other conclusions
of Dr. Brown's research are:
18
-------
<1
1 IcP
J p
If
x 1
'~!
I ¦ ,
?¦
~ * ONO (ppra/100)
Jn k *N20 CppD
/
^ B
*T 0
a , **t
I
1 r t-k
3Q 25 20 15 10
Rodius (mm)
Figure 3-5. Edge effects on N2O formation, (Brown)
19
-------
Expected TotoL N (as NO)
Ar-/H2/02/NH3 69/19/12/.22
25. Orr—i ;
\Lfr ~—
, 1
n CT 1
CG 5 — I
i
1 1
, I
20. 0^ ,
17. 5 4-
IS-.
15.0If-til® n ~ D
J
12. 5 - J Flame Coda
I I ' ¦ • •
10. 0 —( | __-_x N20 Cppm)
ON2 Cppm/100)
D NO
X[
5. 0 .
2-5/ \
\
gl I* 1._ 1
0 1 2 3 4 5 6
Height (mm)
Figure 3-6. Comparison of modeling and experimental results with NH3-doped flame. (Brown)
20
-------
Ar/H2/02/N0 69/19/12/.21
U
01
3E
C
O
¦I—I
•P
0
L
xJ
C
01
o
c
o
o
L
o
Ql
*_
3
-P
O
t.
8.
E
0>
25. 0
22.5
20. 0
17.5
15.0
12.5
10.0
7.5
5. 0
2.5
Q
~ ~
7 o
i
, n
Ej
~
Flame Coda
om cppm/io)
Q NO Cppm/100)
x N20 Cppnr)
+ Tempertitur-a
-------
Ar/H2/02/N20 69/19/12/. 11
150
Expected Total N (o© N2>
125
aJ
Ql
100
¦p
o
L
±3
C
ai
o
c
o
u
L
o
ai
L
D
«P
0
L.
01
Q.
E
(V
O
~
Flame Coda
<> N2 Cppm/10)
O NO
-------
250
Ar/H2/Q2/NH3 84/10/6/. 18
225 —
200
4->
-------
* N20 destruction results in N2 production.
* Destruction reactions need to be determined over a wide range of temperatures and product
yields.
* More information on N2O production reactions is needed.
* Probe effects need to be eliminated because they could have a significant effect on N2O
measurements,
3.1.1 Gas Phase Kinetics (J. Kramlieh. EERC)
Dr. J. Kramlieh, of Energy and Environmental Research Corporation (EERC), focused on the
emissions of N2O Irom industrial flames. Specifically discussed were the issues of N2O formation and
emission in coal and oil flames and the effects of firing configuration and pollution corarol strategies on
these emissions. This study has involved both experimental work and modeling.
A significant amount of N2O, well over the 10 ppm level, is emitted from many industrial coal and
oil flames. However, little N2O is emitted from gas flames; the highest level measured so far Irom
industrial gas flames Is 11 ppm, with levels of 2-3 ppm being typical.
A down-fired, tunnel furnace, with a realistic quench environment, was used to establish baseline
N2O concentrations from natural gas/air and coal/air flames. About 70 ppm of NO and an unmeasurable
amount of N2O were measured from the gas flame, and about 1,100 ppm ot NO and about 80 ppm of
N2O were measured from the coal flame. The N2O measurements were made using GC/ECD analysis.
A possible hypothesis at this point was that fuel nitrogen present in coal caused an increase in N2O
emissions. To verify this hypothesis, fuel N was added to the gas flame in the form of HCN and NH3.
These doped gas flames emitted about 900 ppm ot NO but still almost no N2O (<0.5 ppm). The
experiments described revealed that there was something special about the coal flame that caused it to
emit higher concentrations of N2O.
Kinetic calculations using a 5 ms well stirred reactor (to simulate ignition zone) and a 200 ms
plug-flow reactor (to simulate burnout zone) were performed to investigate NO and N2O emissions from
gas flames. The kinetic set used is shown in Figure 3-10. A realistic temperature profile was imposed on
the plug-flow reactor for these simulations. The results indicate that for the two gas flames (with and
24
-------
KINETICS
.... .N?0 + H
NH + NO -> 2
n2 + oh
N2H + OH
NH2 + NO -> N2 + H20
n2o + h2
N,0 + CO
NCO + NO -> 2
n2 + co2
n2 + o + m-> n2o + m
N„0 + H ->
NO + NH
n2 + oh
Figure 3-10. Kinetics used In model calculations. (Kramlich)
25
-------
without fuel nitrogen) N20 was formed in the flame zone, with maximum concentrations ol about
6-7 ppm, but decayed rapidly to insignificant levels. The rapid decay is due possisty to the N2O
destruction reaction, involving a H atom, being sufficiently rapid in the high tempefature flame zone.
These results verified earlier experimental findings and further indicated that, given rapid N2O destruction
in the flame zone, the key to N2O emissions from coal and oil flames was in the post-flame chemistry.
Subsequently, post-flame generation of N2O from coal flames was modeled Taken into account
were heterogeneous effects of char oxidation, ehar-NO reduction, and volatiles chemistry. Generally, the
combustion of volatiles is fairly complete in the flame zone. However, depending on the firing
configuration, fuel type, and stoichiometry, some volatile combustion may take place in the post-flame
regions. The N2O destruction reaction involving a H atom slows down in the post-flame zone because of
lower temperatures and the paucity of H atoms. The results of this modeling effort, shown in Figure 3-11,
indicate that char oxidation and char-NO reduction cannot form sufficient N2O. NH3 volatiles give low
N2O yields (maximum of about 2.5 ppm), but HCN volatiles react to produce N2O levels of about
100 ppm at the exit of the reactor. Production of HCN volatiles in the post-flame zone may be linked to
the N carried by char beyond the flame zone. Model results, shown in Figure 3-12, also indicate that if
HCN is artificially introduced into the system, it will convert to N2O between 1,100-1,400 K and that
maximum N2O yields will be about 700 ppm and 250 ppm for 100 percent and 40 percent conversion,
respectively.
Experiments were then carried out on a down-fired laboratory furnace with a CH4/Air/NO flame
where HCN, NH3, and acetonitrife were injected into the furnace at various locations (temperatures) along
the combustor. The results, shown in Figure 3-13, are congruent with the model predictions obtained
earlier and further indicate thai other cyano species such as acetonitrile will also yield N2O in post-flame
regions.
Figure 3-14 summarizes the study with a mechanism overview and shows that predominant
generation of N2O occurs in the post-flame zone from conversion of HCN to N2O. Presence of HCN in
the post flame can be credited to devolatilization of char-N and to free HCN transported from the edge of
26
-------
POST-FLAME
• USE FCR—III MECHANISM
Q-HPLUG FL0W
i
AIR
10
11 i 111
• HCN AT STAGING POINT OXIDIZED
• CHAR OXIDATON + CHAR-NO REDUCTION
CANNOT FORM SUFFICIENT N20
10
-4
• HCN OEVOLATILIZATION
o
-------
HOMOGENEOUS N20
• USE FCR-lli MECHANISM
^HCN
orp|=R h-»
• GOAL: DOES N20 APPEAR
IF HON IS ARTIFICALLY
INTRODUCED/?
CONCLUSION
• HCN WILL CONVERT TO NoO
BETWEEN 1100 - 1400K
• NH3 WILL NOT CONVERT
NCO + NO - N20 + CO (40%)
¦ Ng + CO 2 (60%)
i i i r
800
cv
O
& 600
o
>
IE
a
400
2
Ql
QL
9,200
z
Figure 3-12. Post-flame modeling results (Artificial HCN introduction). (Kramiich)
28
-------
• OTHER CYANO SPECIES ALSO YIELD N20
• NH3 DOES NOT, UNDER THESE CONDITIONS
Figure 3-13. Post-flame experiment. (Kramlich)
29
-------
IN POST-FLAME
CHAR-N DEVOLATIUZATION HCN —
^ +
nh3
FREE HCN N20
AT EDGE OF VOLATILE FLAME
CHAR CONTAINS CHAR NITROGEN
STRATIFIED FLOW MAY CONTAIN HCN
IN VOLATILE FLAME ZONE
-NO
DEVOLATILIZATIQN * HCN
-
Mechanism overview. (KramJich)
30
-------
volatile flame in stratified flow. Some N20 is produced in the flame zone but is rapidly destroyed to form
N2- The pertinent reactions are shown in Figure 3-15.
An enthusiastic question and answer period followed Dr. Kramlich's presentation. Dr. Zeliinger
asked how much HCN was injected during the tests. Dr. Kramlich replied that starting with 600 ppm NO
in the post-flame environment, sufficient HCN was added so that a total of 900 ppm would be measured if
all of the HCN had been converted.
Mr. M. Morgan from the International Flame Research Foundation (IFRF) commented thai,
although Dr. Kramlich's tests showed conditions in which HCN to N2O formation was favored, it was not
clear where NO formation was favored. He inquired whether any light could be thrown on this issue?
Dr. Kramlich responded that at higher temperatures NO would be favored because at these temperatures
N2O destruction would be very efficient and because precursors to N2O formation would be oxidized. At
somewhat lower temperatures, N2O may be favored, and at even lower temperatures HCN itself may
exist.
Dr. F.C. Lockwood of Imperial College requested clarification of the meaning of "post-flame
region," since his experience was with industrial equipment. Dr. Kramlich described the post-flame region
as being beyond the luminous or volatile flame where the majority of the gas species have essentially
reached equilibrium and the major reaction occurring is heterogeneous oxidation of char.
Mr, S.A. Sloan of Central Electric Research Laboratories was interested in the implications of
NOx control by staged combustion on N2O emissions. Dr. Kramlich responded that, because N2O does
not tend to survive a fuel-rich environment, staged combustion may lower N2O emissions.
Dr. H.J.A. Hasenack of Royal Dutch Shell Laboratory added that Mr. Sloan s concern in regard to
staged combustion technology may be important as data suggest that N2O formation may be possible if
the fuel-rich zone becomes cool enough to fall in the 1,150-1,500 K temperature window. He warned
against blindly applying NOx control techniques to control N2O emissions.
Dr. W.S. Lanier of the EERC emphasized that the goal of the EERC study was not to simulate a
speciJic type of coal flame. The study was designed to identify possible mechanisms involved in N2O
emission, and the conditions used were for that purpose only.
31
-------
HCN + 0 - NCO + H
NCO + NO - N2O + CO
FORMATION
N20 + H - N2 + OH
REMOVAL
HCN + OH - HNCO + H
HNCO + H - NH2 + CO
ALTERNATE ROUTES
Figure 3-15. Reaction Mechanisms. (Kramlich)
32
-------
Dr. de Soete noted that in the model used ail ol the char-bound N was converted to n20 He
asked if Dr. Kramlieh believed that all char-bound nitrogen formed primarily N2O when it reacts during
char oxidation. Dr. Kramlieh replied that reactions producing N2O were favored. However, since no N2O
was seen at the end of the flame zone, this choice is irrelevant.
3.1.2 Heterogeneous Reactions (G. de Soete. IFP)
Dr. de Soete of IFF began by defining a heterogeneous reaction as one between 2 different
phases, as in gas to solid. Heterogeneous reactions implying N2O as a reactant or as a product are
numerous. Three examples are:
1. Formation of N2O from bound N during heterogeneous char oxidation.
2. Heterogeneous reduction of N2O on char-bound C atoms during heterogeneous char
oxidation.
3. Formation of N2O during catalytic NO reduction (coupled with catalytic N2O reduction),
which may play a larger role in N2O emissions from mobile sources.
A heterogeneous reaction rate can be expressed by:
Vj = kjenmPy
where kj is the rate constant, enm (a fraction) expresses the dependence of the covering degree, which is
proportional to active sites in the solid involved in the reaction, and Py is the dependence of partial
pressure of one or more of reactant gaseous species.
Thus the study, focusing on heterogeneous reactions that imply N2O as a reactant or product,
involved 2 tasks. The first was a series of steady-state trials performed under adsorptiorVdesorption
equilibrium conditions aiming at determining of global rates of N2O release and destruction and
identifying of reaction products (establishing N, C, and O balances). The trials had to be performed using
a progressive complexity of the solid reactant (graphite-char) and reactive gas composition. Specifically,
the first task provided identification of the adsorption and desorption reactions composing the
heterogeneous mechanisms and their reaction rates Vjs. The second task involved a time-resolved study
33
-------
of the reactions identified in the lirst phase. This allowed determination ot the fractions leading to
determination of the rate constants, kj's.
Dr. de Soete cautioned that although much can be learned about heterogeneous mechanisms
from time-resolved or transient-state reactions, practical requirements should not be ignored. For
example, exclusion of concomitant interfering gas phase could be accomplished by avoiding pyrolysis of
the solid and utilizing temperature conditions (or which interfering gas phase reactions may be ignored.
The formation of N2O during NO catalytic reduction plays a major role in emissions from mobile
sources using three-way catalysts. A catalyst does not normally work without a reducing agent,
especially a reducing gas agent (i.e., CO or H2). When NO is fed with CO as a reducing agent over a
three-way catalyst, the fractional decrease of NO is a function of the temperature of the catalyst material.
As shown in Figure 3-16 for a metal catalyst, at tower temperatures, as the concentration NO decreases
because of reduction by CO, N2O is formed and participating CO is qualitatively converted into CO2. At
higher temperatures, reduction of N2O takes place.
" Thus, passing N2O and NO over a typical metal catalyst results in:
1. Reduction of NO at a much lower temperature than N2O, resulting in the formation of N2O
as a function of temperature.
2. Intermediate formation of N2O when NO is fed as a function of temperature.
It can be concluded from the above that transformation of N2O into NO is a minor reaction. The
probable reactions and rates of major reactions during heterogeneous formation and destruction of N2O
on metal catalysts are shown in Figure 3-17.
In all circumstances using graphite, N2O will be reduced at a lower temperature than will NO, as
opposed to the situation when a metal catalyst is used with a reducing agent. Since N2O reduces at a
lower temperature than NO, little N2O is found when NO is reduced. Although reduction of NO could
result in formation of N2O, however, this reduction of NO at the temperature where N2O can be formed
occurs so fast that the N2O cannot be seen in the gas. For graphite, this is true in both the presence and
absence of oxidizing gases. The situation is the same during heterogeneous reduction of N2O and NO
34
-------
800
DO OO OO Oq
#• • • • • ^
600
CO
0.1 5 —
ppmv
400
% CO
or
% CO2
0.10—
0 k
600
T/ K
Figure 3-16, NO and CO over metal (JPB 93) catalyst. (deSoete)
35
-------
MAJOR REACTIONS :
REDUCTION OF NO
JNTON2
SCAVENGING OF
OXYGEN ATOMS
NO + 2(-M) (~MO) + (-MN)
2(-MN) N2 + 2(-M)
NO + (-MN) —>N2 + (-MO)
CO + {-MO) (C02) + (-M)
N20 FORMATION {a'O + (-MN) (N20) + (-M)
REDUCTION OF N20 {,V20 + (-M) (N2) + (-MO)
MINOR REACTIONS :
TRANSFORMATION
OF N20 INTO NO
fN20 + {-MO) 2W + (—M)
[Ar20 + (-M) N0 + {-MN)
(probable mechanism)
(hypothetic)
(hypothetic)
(?)
(?)
RESULTS FROM A TYPICAL CATALYST : (Rates in cm3g^s~l)
V,2 = 7.7 exp{-m0/T)PNO Vu = 0.33 exp{-3900/T)PNO V,0 = 1.63 10~1 exp{-3l30/T)Pco
F,3 = 3.7101 eip(—4120/7') V1 = 5-1 105 exp{-lim/T)PNi0
Figure 3-17, Heterogeneous formation and destruction of N2O on metal catalyst, (de Soete)
36
-------
on char-bound carbon atoms. The probable reactions and rates for heterogeneous reduction of N20 on
char-bound carbon are shown in Figure 3-16.
Shown in Figure 3-19 are the experimentally determined overall reaction rates of C oxidation and,
NO, N20, and HON formation as a function of temperature for Eschwetler char. The subscript e in this
figure implies that these reaction rates were obtained under adsorption/desorption equilibrium rates. The
figure suggests a rough proportionality between Vc, Vno, and VN20-
For a Cedar Grove char, plotted in Figure 3-20 are fractions of nitrogen converting to NO and to
N2O as functions of fraction of carbon burnt. These results have been obtained for all fractions of carbon
burnt and for temperatures up to 1,400 K. As a rough approximation, the fractions of N converted to NO
and N2O seem lo be proportional to the fractions of carbon burnt. Thus it seems that a constant fraction
of N is formed into NO and another constant fraction of N is converted into N2O in a heterogeneous way.
This is true for all char burnout fractions and at all temperatures up to 1,400 K.
Therefore, the main reaction path of N2O formation from bound nitrogen, during char combustion,
at temperatures below 1,400 K, is probably not a formation from HON, occurring in the gas phase, but a
heterogeneous reaction directly linked to char oxidation. This is suggested by reaction order with respect
to oxygen of one for Vc (overall combustion rate), Vno (overall NO formation rate}, and Vn20 (overall
N2O formation rate), and similar temperature dependence of their overall rate. Trials where HON is
added to the oxidizer gas during combustion of carbonaceous materials, show that HON is mainly
converted into N2 and NO, instead of into N2O {Figure 3-21).
Thus based on this study, the probable mechanisms involved in formation of N2O from bound
nitrogen during heterogeneous char oxidation are shown in Figure 3-22. The parentheses in this figure
imply solid bound groups. Main reaction paths for these mechanisms are shown in Figure 3-23.
Dr. J. Wendt asked for a characteristic residence time in the reactor, to which Dr. de Soete
responded that it was on the order of 20-50 ms. Dr. Wendt also asked if experiments were conducted
where physical characteristics of particles were changed in order to determine effects of intraparticie
diffusion. Dr. de Soete answered that particle size effects were checked.
37
-------
REDUCTION OF NO INTO N2
FORMATION OF N20 FROK NO
REDUCTION OP N20 INTO N2
FORMATION OF NO PROM NZO
ELIMINATION OF BOUND OXYGEN
NO + 2 {-C)
2 (-CN) —
NO + {-C.N).
NZO + (-C) .
NZO + (-CO)
--w (-CO) + (-C.N)
N2 + 2 {-€)
N20 + (-C)
16
« N2 + (-CO)
N20 + {-CO) -
I?
18
CO + {-CO) "V C02 + <-C)
{-CO) CO + free carbon site
N2 + COZ
2 NO + (-C) 1 SCAVENGING EFFECT
r of N20 mid CO
2(-CO)
C02 +(-C)
C02 ADSORPTION
C02 + (-c) co 4 {-co)
TYPICAL RESULTS for CO, NO and N20 on GRAPHITE {Rates in cm3 g"J s"1) ;
V = 5.44 1 01 exp( - 1 6060/'T)P V = 2.00 101 exp(- 144 00/T) PKt^
o JVU 9 NO
V = !. 65 10= exp(-19400/T) PN2C V ? = 3.6? 103 exp{-15450/T) PN2Q
V = 1.21 10® exp'- 352 00,/T) P__. V = 2.25 105 exp{-1 9600/T)
4 CO b
Figure 3-18. Heterogeneous reduction of N20 on char bound carbon. (deSoete)
38
-------
Figure 3-19. Reaction rates for Eschweiler char, (de Soete)
39
-------
Figure 3-20, Cedar Grove char conversion of nitrogen, (de Soe!e)
40
-------
T/K
Figure 3-21. O2 + HCN on graphite, (de Soete)
41
-------
OXIDATION OF (-C) and (-H)
NITROGEN SPECIES INVOLVING REACTIONS
CARBON REACTIONS :
02 + 2(-C) -i- 21 -CO)
(-CO) iCOt frw site
J(-CO) — €02 + (-C) + f.w vir
CO + ( -CO) C02 + (-C)
C02 + t-C) — CO + (- CO)
HYDROGEN REACTIONS :
2t-CH, — H2 + 2<-C)
H2-M-CO) — H20 + i-Cl
H20 t ! -C) -S. H2 + (-CO)
FORMATION OF NO i
02 + {—C)+(-CN) (-CO)+{-CNO)
(-CNO) NO + (-C)
(-CN) + (-CO) -2, NO + H~C)
FORMATION OF N20 ;
21-CNO) Ji- (-CO) + (-C) + N20
NO + (-CN) -2» N20 + (-C)
<-CNO)+(-CN) — N20 + 2(-C)
HETEROGENEOUS REDUCTION OF
NO ANDN20
NO + 2(-C) -i. (-CO) + (-C N)
2(-C.N) -i. N2 + 2i-C)
NO + (-C.N) N20 +• (-C)
N20 + (-C) -!!• S2 + (-CO)
N20 + (-C0)-^N2 + C02
N20 + (-CO) -!!. 2N0 + (-C)
NO + (-C) + (-CH) — (-CO) + (-CHN)
FORMATION/DESTRUCTION OFHCN :
(-CH) + (-CN) — HCN + (-C)
H2 + (-C) * (-CN) -2L (-CH) + (-CHN)
(-CHN) -- HCN + fr« etc
HCN + (-CN) — N2 + (-CH)
(HCN — NO+ .,.?)
FORMATION/DESTRUCTION OF NIC =
H2 + (-CHN) — NH3 + (-C)
NH3 + (- CO) + (-CN) — N2 + H20 + (CH) + (-C)
(NH3 — HCN ')
(NH3 — NO ?)
Figure 3-22. Formation ol N20 from bound nitrogen Awing heterogeneous char oxidation, (da Soete)
42
-------
N20 FORMATION from BOUND NITROGEN during CHAR combustion
(main reaction paths)
G —CM
#
I
*
a —oh
i
__ Qt j_j
C - C
c-c
I
0 - c
1
O-CN
o-oisr
G — OH
I
02
02
1 4^
C-CNO
O-CN
0 — OH
1
=£» N20
=£> HON
N2
Figure 3-23. Main reaction paths of N2O formation during char contxjstion. (de Soete)
43
-------
Dr. S, Lanier commented that all of the reactions Dr. de Soete had studied were with O2 He
asked if any water (H2O) were included in the stream, because a substantial amount of water is in the
gases that are reacting heterogeneously with the char. Dr. de Soete responded no.
Dr. J. Kramlich asked lor Dr. de Soete's speculation on how much this reaction could contribute
to a coal flame situation. Dr. de Soete replied that he could not provide any quantitative speculation.
3.2 OVERALL CHEMICAL INFORMATION FROM LABORATORY COMBUSTOR EXPERIMENTS
Dr. J. Wendt of the University of Arizona was the discussion moderator of the session in which
data on laboratory-scale coal combustion and integrated measurements were presented.
Laboratory-scale coal combustors are larger than bench-scale facilities and have a firing capacity of
20 kW to 4 MW. The speakers in this session discussed studies involving integrated measurements
taken under practical conditions where both homogeneous a no heterogeneous reactions occurred. Data
were collected from two types of combustors: premixed combustors, which are one-dimensional,
plug-ftew reactors, and turbulent diffusion-flame combustors. Dr. Wendt reminded the audience that this
session would be very informal, that preliminary results would be offered, and that comments and
questions were encouraged.
Dr. Wendt presented the results of DOE-funded research ai the University of Arizona. The work
focused on optimizing reburning configurations for NOx control. A secondary focus of the work was on
N2O measurements. However, Dr. Wendt explained that the research was still in progress and the
resufts were therefore preliminary.
The experimental research at the University of Arizona is being conducted on a premixed,
down-fired combustor at a firing rate of 27 kW to study NO and N2O emissions from the combustion of
Beulati lignite and Utah #2 bituminous coals. N2O was measured using an on-line GC/ECD technique
where the sample gas from the combustor passed through a water quench sampling probe and a
refrigeration condenser for water removal prior to analysis. Nitrous oxide levels for both coals were found
to be insignificant at less than 2 ppm for stoichiometric ratios (SH) ranging from fuel rich to fuel lean.
Measurements of NO and N2O as a function of residence time at fixed fuel lean SRs, with combustion
gas temperatures in the range 850-1,290 K for Beulah coal and 1,050-1,525 K for Utah coal, showed
44
-------
N2O concentrations to be less than 5 ppm and the N2O/NO concentration ratio to be very small.
Figure 3-24 illustrates that there is little N2O formation under reburn conditions using natural gas as the
secondary fuel.
Dr. Hasenack pointed out that the temperature range in the post flame region appeared to be
similar to that allowing N2O to be formed from HCN. Dr. Wendt replied that, although HCN was not
measured in the tests reported here, other tests showed high levels of HCN in the rich post flame region
but tow levels under fuel lean conditions.
Following Dr. Wendt's presentation, Dr. W. Linak, from the Combustion Research Branch (CRB)
of EPA, outlined the results of his experiments with laboratory-scale pulverized coal eombuslion. His
work has focused on characterizing the effect of different coal types, flame shapes, and NOx levels on
N2O emissions.
Dr. Linak presented N2O emission data collected from a down-fired, 60,000 Btu/h
laboratory-scale combustor equipped with a variable swirl burner. Samples were drawn from a
post-combustion region under stack type conditions. The purpose of these experiments was to change
NO emissions through changes in the overall turbulent diffusion flame type and thus to determine if a
correlation between N2O and NO emissions was valid. In an attempt to vary the NOx concentration,
burner swirl and flame shapes were altered by changing axial and radial air ratios and locations of the
injector tube. A Utah bituminous coal and a Montana sub-bituminous coal were used in these
experiments, the first having a higher organic content but lower moisture level than the latter. For each
coal, data were collected at 3 (fuel lean) stoichiometric ratios and 2 different flame shapes—one axial and
one radial. Gas samples were collected in stainless-steel containers after refrigeration drying and were
analyzed for N2O by GC/ECD. The widely scattered data, presented in Figures 3-25 and 3-26, show that
N2O emissions varied widely (10-250 ppm). There seems to be no correlation between N2O emissions
and NOx emissions.
Earlier data (Figure 3-27), generated using on-line GC/ECD analyses and the same two coals,
show very low N2O levels <10 ppm. These data were considered suspect, however, because on-line
analytical procedures had not been verified at the time the samples were measured.
45
-------
SRI =1.1, SR2=0.9, SR3= 1.06 REBURNING
UTAH BITUMINOUS #2
RESIDENCE TIME, SEC
~ NO + N2Q
I "I "1 " ' 1 1
TK 1613 1448 1373 1223 1161
Figure 3-24. Rebum condition using natural gas. (Wenctt)
46
-------
N20 ppm (0% 02)
250
200
150
100
O Axio)/SR=1.0
~ Axial/SR=1.2
A Axial/SR=1.4
• Rodiol/SR-1.0
¦ Radial/SR-1.2
A. Radial/SR= 1.4
250 500 750
NO ppm (10% 02)
1000 1250
Figure 3-25. NO/N20 for Utah bituminous (N20 measured from sample containers). (Linak)
47
-------
N20 ppm (0% 02)
250
200
150
100
50
0
~
O A
¦ ~
A
_l I I I L_J I l_
I ,, I L_l I I I I.I I .1
J t—l L-
O Axial/SR=1.0
~ Axia1/SR = 1.2
A Axiol/SR-1.4
• Radial/SR=1.0
¦ Radial/SR=1.2
A Radial/SR=1,4
0 250 500 750 1000 1250
NO ppm (0% 02)
Figure 3-26. N0/N20 for Montana sub-bituminous (N20 measured from sample containers). (LinaK)
4a
-------
N20 ppm
250
200
150
100
50
~
^ "8
o
*
,^nn ak
\ °*T *
'®*, , , I
250
~
500 750 1000
NOx ppm
* Acurex/EPA
O Harvard
O MIT
A EPRI
• RWE
Utah bit, 4/2/87
A Montana sub, 4/3/87
1250
Figure 3-27. On-line data in relation to other reported data. (Linak)
49
-------
Following Dr. Linak's presentation, Dr. J, Kramlich, ol EERC, described his fluid bed model as a
stirred reactor for bed particles releasing volatiles such as HCN, CO, and so forth, and as a plug-flow
reactor for the gases moving up through the bed. Initial model results, depicted in Figure 3-28, show that
NO concentration initially increased because of a leaner stoichiometry, reached a maximum of about
200 ppm, and then decreased in bed region with relatively less lean stoichiometries. The final
stoichiometry was taken to be 1.25. The results also show that N2O emissions at the end of a residence
time of 0.5 s, corresponding to the beginning of freeboard part, were reasonably high (about 250 ppm).
The model predictions for variations in excess air, bed temperature, and fuel nitrogen speeiation
are shown in Figure 3-29. The stoichiometry in the bed was lean, so an increase in excess air did not
have much effect on N2O emissions, which are very sensitive to temperature in the range 1,000-1,200 K.
NH3 does not produce much N2O; HCN produces reasonably high levels of N2O; and a combination of
NH3 and HCN produces intermediate N2O levels. All these parametric effects, explored by the model,
agree with N2O formation/destruction chemistry understood so far.
Model predictions were also made on N2O emissions from coal and natural gas reburning.
These results are shown in Figure 3-30. The first simulation was that of a gas-fired, NO-doped primary
flame, which produced no N2O. Utah bituminous coal was added downstream of the primary flame as
the reburning fuel. The coal represented a minor contributor to the total reactor fuel rate (stoichiometric
ratio of 0.9 and 0.8) but produced significant N2O (45 and 85 ppm), possibly owing to HCN conversion to
N2O at relatively low post-flame temperatures according to homogeneous kinetics. The next model
considered, the case of a coal primary flame with natural gas added downstream as the reburning fuel.
Coal primary flame produced about 50 ppm N2O. The addition of the natural gas, however, dropped
N2O levels almost to zero, probably because N2O does not survive a fuel-rich zone very well. These
results suggest that application of carefully picked gas reburning conditions may decrease both NO and
N2O emissions.
Dr. A. Williams of Leeds University followed Dr. Kramlich with a summary of his investigations.
His experiments and measurements are shown in Table 3-1.
50
-------
FLUID BED
VOLATILES,
CHAR OXIDE
PRODUCTS
o
» "
¦ o.
COAL.
INERT BEO
MATERIAL
AIR
W.U
« % fM*1*
'St' t* ~•vft
•It.
EXHAUST
mtm
FLUIDIZED 8ED
COMBUSTOR
tffftft
AIR
MODELING
APPROACH
300
200
>
£L
Q.
too
1
-1
I 1
1100K.
2.1% N IN FUEL
SRI -
1.25
•NOV
•n2o
i
•
i I
0.1 0.2 0.3 0.4
TIME, SECONDS
MODELING
RESULTS
0.5
Figure 3-28. Fluid bed modeling. (Kramlich)
51
-------
FLUID BED
>
2
Q.
Q_
•*
o
CM
z
400
300
200
BED
TEMPERATURE
r= s "T
100 -
20 30
PERCENT
1000
1200
FUEL NITROGEN
SPECIATION
Figure 3-29. Predictions under varied conditions. (Kramlich)
\
52
-------
REBURNING
COAL REBURNING
• GAS PRIMARY
• ALL N20 FORMED FROM
REBURNING FUEL
• COAL REBURNING FORMS
SIGNIFICANT N2O
GAS REBURNING
• COAL PRIMARY
• GAS NEARLY ELIMINATES
PRIMARY N20
01 80
*
O
>
-------
TABLE 3-1. EXPERIMENTAL MEASUREMENTS (WILLIAMS)
FueJ-N%
n2o
N2O
CO2
NOx
in the
Combustor Type
ppm
%
ppm
Coal
NOx
A
Fluidized Bed Combustion Run (1)
35
7.9
450
1.2
0.078
(Bed Temperature 900 °C) Run (2)
55
10.5
500
1.2
0.110
B
Drop Tube Furnace {Methane
Flame Supported, 1,110 °C) Run (1)
70
10.8
450 to
1.3
Run (2)
55
7.7
500
1.6
C
Grate (Bed) Combustion
100
4
450
1.6
0.222
D
Char Combustion (ca 1,500 °C)
200
10
0.40
E
Coal Devolatilization inert
very small
54
-------
His measurements reveal that N20 is about 10-20 percent of the total NOx for iluldized bed and
drop-tube combustion, with grate combustion producing higher levels than systems with sustained high
temperature in the post-flame region.
He suggested the mechanism of N2O formation as outlined in Figure 3-3i where coat pyrolizes to
give tar + gases + char, all of which can produce N20.
Dr. Williams' kinetic calculations reveal that the N2O contribution from the interaction of
NH3 + NO was only about 5 ppm. N2O yield from tar combustion was about 10 percent of the NO
produced, and N2O yield from char combustion was also about 10 percent of the NO produced. These
results suggest that at each stage of combustion, either single or staged, the N2O is about 10 percent (or
possibly higher) of the NO (+NO2) produced, and production of N2O is a strong function of temperature.
Dr. Williams noted that the following reaction is crucial in N2O formation and that the NCO
species can only come from HCN:
NCO + NO -» N20 + CO
Dr. de Soete pointed out that in Dr. Williams' experiments N20 was produced by gaseous NCO,
and they did not take into account heterogeneous formation. Dr. Williams replied that this was not
considered because he could not quantify any HCN on the char. Dr. de Soete stated that he had indeed
found HCN in char, supporting heterogeneous formation of N2O.
Dr. Lanier asked where the H atom comes from in the experiments with char, since the dominant
N2O destruction step involves a reaction with a H atom. Dr. Williams answered that since he was
interested only in N2O formation routes and thus high N20 values, no hydrogen was made available lor
N2O destruction.
Dr. J. Smart from the International Flame Research Foundation (IFRF) presented data on his
experiments performed on a 4 MW horizontal, swirled pulverized coal fired furnace. These experiments
have so far yielded only preliminary measurements. They were conducted on a burner operated with no
pollution abatement and therefore yielded high NOx emissions. The fuel was a western-Canadian
55
-------
Figure 3-31. N0X and N2O formation from coal. (Williams)
56
-------
bituminous coal, pulverized lo a standard grind for the experiment. The species of NO, N20, HCN, and
NH3 were measured in the flame in 2 dimensions (axial and radial), as shown in Figure 3-32.
The preliminary results indicated that HCN and NH3 found in the flame at the 0.1 m axial location
were essentially gone by the 0.2 m location. Nitrous oxide values were low al the 0.1 m location,
although at 0.2 m near peak values were achieved. N2O and NOx levels in furnace exhaust (at about
1,070 °C) were 43 ppm and 720 ppm (at 3.6 percent O2) The data suggest N2O formation from
fuel-bound N, but further work is needed to characterize this hypothesis.
57
-------
HP.
horizontal
posiHon (m)
AD. axial distance (m)
o Q20 am aso o,bo 1.00 120 uo 62s
> 1 f 1 f 1 1 1 i
0 - •lv..\ % \ ,a >.;>•> .XiV \ \ v,^>-» *>, % - \ -.{ {•, -i »'i -s,\:i,a
1 i
0.20
0.40
devoWolization
• volatile combustion
particle trajectories
IRZ boundary
— coal * primary air
— swirted secondory air
/
1 '
I t
1
, /'
/ /
/ /
exhaust
Figure 3-32. Schematic of experimental flame measurements. (Smart)
58
-------
SECTION 4
SESSION 2: N2O MEASUREMENT TECHNIQUES
Mr. J. McSorley of EPA, the discussion moderator for the second session, began by addressing
the need to develop standardized sampling techniques, Given the measured increase in atmospheric
N2O levels, it can be concluded that a significant anthropogenic source exists. The relative importance ol
candidate sources such as agriculture, bio mass burning, fossil fuel combustion, and others is unclear at
this time. Therefore, a reliable, accurate method for identifying and analyzing these sources is crucial.
Mr. McSorley presented the following EPA program issues:
1. What is the global contribution of N2O from fossil-fuel combustion versus biogenic and
natural N2O source emissions to stratospheric O3 depletion? Furthermore, are current N2O
global emission inventories reliable with respect to N2O combustion sources?
2. Should N2O emissions be controlled, and if so, will the controls make a difference in
stratospheric O3 depletion?
The related technical issues are stated below:
1. What is the relative importance of N2O emissions from fossil-fuel combustion sources (i.e.,
stationary combustion sources versus mobile combustion sources)?
2. What is the relative importance of N2O emissions from natural and biogenetic sources (i.e.,
agriculture, biomass burning, oceans, and other sources)?
Mr. McSorley identified a crucial need for a standardized N2O sampling and analytical protocol to
verify the reliability of N2O data gathered worldwide. A wide variance in sampling methodology is
evident. Grab samples have been collected in many types of containers, including glass, stainless-steel
canisters, and Tedlar bags. Sample moisture content also varies to a great degree. Furthermore, the
development of an on-line real time N2O analyzer would eliminate long analytical delays in research
programs as well as the problems associated with grab sampling.
59
-------
Limited methods of N20 measurement are available. Quantilicaiion is usually done by gas
chromatography using an electron capture detector (GC/ECD). The ECD has excellent detection limits,
being used1 to measure ppb ambient N2O levels. However, the ECD is also susceptible to many
imerferants found in combustion gases. The thermal conductivity detector (TCD) has also been used
successfully, although detection levels are not as great as those of the ECD. A disadvantage of the GC
analysis is that a true real time measurement is not possible and that a high skill level is required to
perform the analysis.
\
4.1 OPERATING PROCEDURE FOR IN-HOUSE RESEARCH
Mr, J. Ryan of Acurex Corporation, responsible for EPA in-house analytical efforts, presented
information on the development of a recommended operating procedure used for in-house combustion
research. The method was derived from several different procedures designed for N2O GC/ECD
analysis. During procedure development, Mr. Ryan evaluated possible interferences contained in
combustion gas samples and how they could be treated. The result was a technique for the removal of
interfering H2O and organic interferants contained in N20 samples using a P205/activated charcoal
pre-column. The possibility of any C02 interference was investigated by comparing N20 values when
the C02 constituent was eluted both before and after N20 using different columns. No significant
difference in N20 value was observed. The possibility of detector desensitizing from oxygen exposure
was not evident after repetitive introduction over a 7-h period ol a standard gas mixture containing
oxygen. The nonlinear property of the detector was described, demonstrating the need for a multi-point
calibration in establishing accurate quantifiable limits because of the large variance in response factors.
4.2 WATER SORBENTS USED IN N2O MEASUREMENT
Dr. de Soete then gave a brief presentation on the importance of carefully choosing H20
sorbents used in N20 sampling techniques. Many desiccants are commercially available, such as
magnesium perchtorate [Mg(CI04)2], calcium chloride (CaCi2), aquasorb (P205 with indicator) and silica
gel (Si02). Silica gel, however, may retain N20 (Figure 4-1); flushing with argon may release N20,
possibly affecting sample integrity.
60
-------
(t = 5 minutes)
ABSORBENT WHf^1) N20oui/N20:
CaCh 160 1
M9(C10<)2 511 1
h2o (m) i
Silicagel 160 0.064
N20 + C0 + C02 ON SILICA GEL
Figure 4*1. Losses of N2O on different absorbents, (de Soete)
61
-------
4,3 INTERFERENCES FROM SULFUR DIOXIDE
Dr, I. Muzio of Fossil Energy Research Corporation (FER Co.) and Dr. J. Kramiich of Energy and
Environmental Research Corporation (EERC) presented startling results obtained during a collaborative
study of interferences from SO2 with respect to N20, in sample containers under certain conditions,
While conducting natural gas tests using FER Co's combustor, high N2O values were measured while
doping with ammonia (to generate NQx) and SO2.' N2O samples were collected in glass flasks using an
extraction system where moisture removal was not utilized prior to collection (see Figure 4-2). It was
discovered that while doping with 2,500 ppm SO2, N2O concentrations on the order of 300 ppm were
obtained, while the condition without SO2 doping measured only about 1 ppm N2O.
Dr. Muzio then contacted Dr. Kramiich who repealed the tests at EERC, obtaining the same
results (see Figure 4-3), Similar tests were performed with a hydrogen flame in order to eliminate N2O
formation mechanisms associated with carbon chemistry. Again, similar results were observed when
NOx and SO2 levels were duplicated. The SO2 injection location was then varied with no significant
change in results.
At this time, Drs. Muzio and Kramiich began to suspect that a reaction in the sampling system
was the source of the N2O artifact. After changing probe types, NO levels were observed while SO2 was
pulsed, under the premise that the NO was converted to N2O by the SO2. However, after NO levels
were found to be constant, concentration was directed towards the possibility of ihe artifact reaction
occurring in the sample container, and an attempt was made to generate N2O in the sample flask, A
synthetic mixture containing N2, O2, CO2, and NO was placed in flasks, one containing deionized H2O
and no SO2, and the other containing 1,500 ppm SO2 and dilute sulfuric acid. Over 150 ppm N2O was
generated from the S02/sulfuric acid containing mixture, while virtually no N2O was observed in the
mixture lacking these two components (see Figure 4-4). Next, maintaining the same SO2 containing
mixture conditions, the pH of the liquid in the sample flasks was varied. No significant difference in N2O
generation was observed. The test was performed again without the SO2 constituent. This time no N2O
generation was observed, isolating the importance of SO2 in the artifact mechanism. The effects of
varied SO2 levels were then evaluated, A step change in N2O was observed when SO2 exceeded
62
-------
STAINLESS STEEL
QUARTZ LINED
Figure 4-2, N2O sampling system. (Muzb, Kramiich)
63
-------
400
300
A
o. 200
O
«N
z
100
0
FUEL
NO, ppm
SO^i ppm
FFMCa COMBI.'STOR
ch4/nh3
750
0
ch4/nh3
750
2S00
MMJZQmmTQR
ZJ
ch4/nh3
600
1900
H 2
600
2100
Figure 4-3. Effect of SO2 on N2O. (Muzio, Kramlich)
64
-------
BASE GAS MIXTURE:
LIQUID: HzO H2S04 (PH=2)
S02: 0 1500
Figure 4-4 Synthetic gas mixtures in glass sample llask. (Muzio. KramBch)
65
-------
500 ppm, and it seemed to level off at over 1,500 ppm (see Figure 4-5). The effect of NO level variance
was also observed in relation to N2O generation while holding SO2 levels constant. A ratio of N2O to
NOx was obtained for the 2 varied NO concentrations (see Figure 4-6}. The ratios are comparable to
emissions reported from utility boilers. Figure 4-7 shows hypothetical chemistry obtained by a scrubber
literature search that may be responsible for the N2O artifact [9]. The reaction rates shown are consistent
with the observations made to this point.
Tests were designed to determine if the removal of SO2 or moisture could enable a sample of
uncompromised integrity to be collected. The sampling system was again evaluated to determine how
the components could be removed from the sample mixture. Sodium hydroxide (NaOH) and sodium
carbonate (Na2C03) were found to be effective in neutralizing SO2 while having no evident effect on
known N2O levels when added to the sample container prior to collection. The conditioning of the sample
prior to collection was also evaluated. Passing the sample gas through an ice-bath condenser reduced
the NO/SO2 interaction considerably, although some N2O generation was observed. Passing the sample
gas through an impinger containing a NaOH solution was successful in preventing N20 generation, but
also removed C02, another important component, from the mixture as well.
In closing, Dr. Muzio stressed the impetus of the artifact discovery in regard to current emission
data bases and to the contribution from fossil-fuel combustion to the global N2O budget. Furthermore,
since this artifact reaction is evident in sample containers, is it possible for it to happen in the atmosphere
as well?
4.4 CONTINUOUS INFRARED ANALYZER FOR ON-LINE N2O MEASUREMENT
Dr. Muzio also reported on the status of research that he has been conducting with
Ms. T. Montgomery of the University of California, Irvine, involving the development of an on-line N2O
analyzer. Ms. Montgomery and Dr. Muzio presented information on their investigations at the previous
N2O workshop in Boulder, Colorado [10]. The objective of the project, funded by the Electric Power
Research Institute (EPRI), was to develop a continuous infrared analyzer for on-line N2O monitoring of
combustion gases. A prototype production model is being manufactured based on the results of
Dr. Muzio's and Ms. Montgomery's research. The instrument operating principles are based on the ability
66
-------
400
300
E
CL.
o.
O
c*
21
200
100
NOx: 750 ppm
/
/
I
I
I
V ^sESL^BSviia
0
A F
/
500
1000
1500
2000
2500
S02, ppm
Figure 4-5. Effect of SO2 level on N2O. (Muzio, Kramlich)
67
-------
S02: 2500 ppm S02: 2500 ppm
Figure 4-6. Effect of NO* level on N2O formation. (Muzio, Kramlich)
68
-------
S02 ~ H20 - H * HSO
2N02 ~ H20 — 2H ~ N02 ~ NO3 [2.2 MIN]
HSO3 ~ N02 — NO" ~ HSO^ [7.1 SEC 3
pH >4.5
PRODUCT
il2 NO" ~ H+ — HNO
NO"
2HN0 — N20 | ~ H20
Figure 4-7. Hypothetical chemistry. (Muzio, Kramlich)
69
-------
of N20 to absorb infrared radiation at several wavelengths. The work focused on the 7.8 absorption
band because of the ability to eliminate the interference of other gases present in the sample also
absorbing radiation at this wavelength. The interfering gases, NO2 and SO2, can be easily removed by
using a sodium carbonate/sodium sulfite scrubbing solution prior to introducing the gases inio the
analyzer. In the past year, the precision of the instrument has been improved to 5 percent without sample
conditioning and to 2 percent when removing the interfering gases. The production instrument will be
available for field evaluation soon.
4,5 OPTICAL INTERFEROMETRY
Mr. A. Galias of CNS presented information on the development of optical interferometric
methods for measuring gases of environmental concern The methods used involve both ultra violet (UV)
and near infrared radiation absorption properties. Optical interferometry offers several advantages over
conventional IR analyzers in that the Fourier transforms of the interferograms contain information on all of
the wavelengths present and are not limited to 1 wavelength of detection. Therefore, more than 1 gas
can be measured simultaneously. In addition, the precision in wavelengths measured is also greater,
which is important in regard to isolating interferences of other gases absorbing radiation in the same
wavelength region. Mr, Galais has been successful in measuring NO, N02. and S02 simultaneously with
detection levels approaching 1 ppm. Currently, methods for measuring ammonia (NH3) and N20 are
under development.
70
-------
SECTION 5
SESSION 3: FULL-SCALE FIELD DATA
The third session of the workshop included a series of presentations on data collected from both
mobile and stationary sources. The session included data from full-scale stationary utility facilities in the
United States and Europe, The presentations provided N2O emission values measured from different
research facilities as well as industrial equipment to provide a better representation of N2O discharge
levels and to assess the role of fossil-fuel combustion on global N2O levels. Dr. K. Hein, from RWE, was
the session moderator, assisted by Dr. W. Linak of EPA during the stationary-source presentations and
by Mr. A, Douaud of the IFP during the mobile-source presentations. Because of the sampling artifact
discussed in the previous session, each presenter was asked 1o include information regarding sampling
and analytical techniques used, moisture and sulfur content of the sample, and any efforts taken to
remove these constituents.
5.1 FLUE GAS EXHAUST DATA FROM STATIONARY SOURCES
Most of the stationary-source data presented by European scientists provided material for an
excellent comparison with data from the United Stales. European data represented about 70 different
facilities ranging in firing capacities from 0.5-800 MW, with most facilities in the 100-600 mW range. Data
from circulating and stationary fluidized bed combustors (FBCs) were included in the presentations. Data
from coal combustors made up a significant portion of the presentations, although emission data from
peat, distillate oil, wood, refuse, and natural gas combustion were also discussed.
5.1.1 NpQ Emission Data From Test Rias and Full Scale Units
Dr. Hein and Dr. L. Raible, both of RWE, presented data covering emission values from test rigs
and full-scale facilities. Table 5-1 contains data from a PFBC test rig. The high N20 values are
consistent with other data on fluidized bed combustors. Table 5-2 contains data obtained from several
test- and pilot-scale SCR facilities. Different catalysts were compared for NOx reduction and N20
71
-------
TABLES-! N20 DATA: PFBC TEST RIG (HEIN)
probe/
date
power
MWth
P
bar
bed
temperature
°C
°2
%
CO
vpm
*
vpm
N2°
vpm
1/23.3.88
0,75
6
835
6,3
82
173
95
2/2*1.3.88
0,56
6
835
8,83
1*11
203
193
3/2*1.3.88
0/59
6
850
9,16
159
211
168
¥29.3.88
0,71
7
850
8,83
107
203
95
5/31.3.88
0,3*1
3
8*6
7,29
512
255
115
Coal: 9 %
8/ ^ % asli
29 % volatiles
-------
TABLE 5-2. N2O DATA: SCR TEST/PILOT-SCALE (HEIN)
Test unit
catalyst type
NCL
,25
mg/nr
before/behind
SCR
T
°C
vpm
before SCR
NjO vpm
behind SCR
RWE high
dust SCR
T1O2 (Jap.)
molecular sieve
Fe/Cr-oxide
600/200
350
1 - 2
0,31 - 0,W
< 0,65
1
BASF high
dust SCR
TiOj (jap.)
1050/158
380
HA
0/8
low dust SCR
UO2 (Jap.)
1113/167
365
13
2
73
-------
emissions. Table 5-3 contains data collected from full-scale units using various NOx reduction methods
while firing on a European brown coal. The N20 levels presented were relatively low in relation to U.S.
reported data from similar facilities. To demonstrate this, the N20 values were added to a graph
presented at the Boulder, Colorado worttshop. This graph, conlaining the RWE data, is reproduced in
Figure 5-1.
The sampling system, collected the gas in both glass and stainless-steel vessels. Moisture was
removed with a refrigeration condenser prior to sample collection. The sulfur content ol the brown coal
was estimated to be 0.8-0.9 percent. The N2O analysis was done by both GC/ECD and TCD.
Dr. J. Jacobs of VGB presented data collected at two combustion test rigs operated by
Dr. H. Kremer at Bochum University-Rohr. On-line N2O measurements were made on a 50 KWjh
pulverized coal combustor and a 100 KWtn fluidized bed combustor. The on-line measurements were
made using an IR analyzer utilizing a dual wavelength (7.8 nm/8.168 (im) absorption technique. The
maximum values of N2O observed from the coal dust flame were 30 ppm at 2,000 ppm N2O. The
fluidized bed combustor yielded peak values of 100 ppm N2O.
5.1.2 German Power Plants
Dr. H, Koeserof L&C Steinmuller presented data from several large German power plants (see
Table 5-4). The data also show relatively low N2O emissions. The samples were collected while the
utilities were operated under normal conditions, Moisture was removed from the gas stream using a
refrigeration condenser. The samples were analyzed by GC/ECD and GC/TCD at Steinmuller.
5.1.3 Italian Power Plants
Dr. R. Tarli of Ente Nazionale Energia Elettrica (ENEL) presented data that probably best
represented N2O emission values from industrial facilities. His stack gas samples were analyzed by an
on-line GC/ECD method, which produced a near real-time value. Data were collected from both fuel oil
and coal combustion facilities, as shown in Tables 5-5 and 5-6, respectively. Virtually no N2O was
detected from burning a high sulfur (3 percent), #6 fuel oil, while 5 ppm was the maximum value obtained
from burning an Alrican low sulfur (0.6 percent) bituminous coal.
74
-------
TABLE 5-3. N2O DATA: BROWN COAL FULL-SCALE RWE (HEIN)
Prode/ j unit/
Klth/WltfTOlit
X N in coal
°?
00
K)
NjO m
date power
confiustlon
as NO2
vpm
ECD
HO
Wei
3
modification
raw (waf)
%
mg/tr^
(d"y 6X Oj)
mg/n^
(dry,6% 03)
1/25.1.88 j T/300
wltti CM
0,38 <1,05)
2,6
210
170
102
11,1
2/2SH.88i "V75Q0
=4
1
with CM
0,38 <1,05)
2,6
210
170
102
6
5
3/2S4.88 | "1"/30D
with 01
0,37 (0,99)
2,2
210
195
120
5,5
i*
^/2&4,88 ; T/300
1
•»
i
Hitft CM
0,37 (0,99)
2,2
210
195
120
5,2-6,5
5
5/28^.88 I "1*7300
with CM
0,37 (0,99)
2,2
210
195
120
4
6/28A88 j T/300
wltfl CM
0,37 <0,99)
2,2
210
195
120
<
7/06.5.88 j "2"/300
!
with CM
0,
-------
[Fie. 12 from; EPA/NOAA/NASA/USOA HgO-Morkshop; Sept. 87, Boulder, Colorado)
N0X PPM
Figure 5-1, Comparison with previous results. (Hein)
76
-------
TABLE 5-4. NgO EMISSIONS FROM GERMAN POWER PLANTS (KOESER)
firing equipment
else
MW.l
f u « X
NO
vpi
flue gaa
°2
Vol-S
N_G
vpa
wel-bottom boiler
510
gorain
bltualnous ooal
863
4.6
26.7 1 3.8
dry-bottom boiler
465
goraan
bltuatnous coal
196
46
2.9
2.9
25
18 after SCR
dry-botLoa boiler
84
laported
bltualnous ooal
307
75
7.8
8.1
3.9
2 after SCR
42 t/h
aunlolpal rafuaa
101
10
4.7 1 1.1
77
-------
TABLE 5-5. DATA FROM FACILITY BURNING #6 FUEL OIL (TARLI)
UNIT "A'
Nominal rating:
Boiler type:
Burner configuration:
Burner type:
Fuel:
320 MW(e)
Once-through (U.P.). subcrttteal
Opposite firing
Cell burners
#€ oil (3% S)
Flue Gas
Unit
02
Recirculation
Concentration
Load
at ECO
Fans Damper
(ppm)
(MWe)
(%)
Opening (%)
NOx
NO
N2O
310
0.8
100
100
355
335
N.D.
280
1.1
100
100
295
280
N.D.
200
1.6
100
100
235
225
N.D.
310
0.8
0
5
480
460
1
250
1.3
0
25
335
320
1
160
2.5
0
65
220
200
N.D.
78
-------
TABLE 5-6. DATA FROM FACILITY BURNING COAL (TARLI)
UNIT T
Nominal rating: 171 MW(e)
Boiler type: Drum boiler, natural circulation
Burner configuration: Comer firing
Number of burner
elevation: 6
Fuel: Bitiminous coal {from South Africa)
Unit O2
Load at ECO Concentration ppm
(MWe) (%) NOx N2O
130 4 NA 5.1 ±0.2
165 4 386 ± 5 3.3 ±0.4
165 4 386 ± 5 2.1 ±0.2
165-115 4 NA 2.5 ±0.1
79
-------
5.1.4 Emissions of NgO From British Power Stations
Dr, A, Sloan of Central Electricity Research Laboratories (CERl), presented data collected from 2
power plants in England. Dr. Sloan was interested in verifying the existence of a N20/NOx ratio
relationship. Table 5-7 contains data from the Eggborougn and Fiddler's Ferry power plants. While
expecting to see a ratio on the order of 20 percent, ratios of less than 5 percent were obtained. The N2O
data obtained at the Eggborough power station were collected with a sample system using a refrigeration
condenser for moisture removal. The samples were collected in polyvinylidenechioride bags and
analyzed within 24 h by GC/ECD. At Fiddler's Ferry, a more extensive N2O measurement effort was
performed. Both wet and dry grab samples as well as an on-line IR analyzer were employed. Figure 5-2
shows the set up for N2O sample collection and measurement. The data presented in Table 5-8 show
good agreement between both the wet and dry grab samples and the IR measurements. The grab
samples were analyzed on site by-GC/ECD within an hour of collection. SO2 emission levels at both
power plants were estimated to be 1,200 ppm,
5.1.5 Emission Data from Swedish Stationary Combustion Sources
Mrs. K. Dahiberg of the Swedish Environmental Research Institute (SERf) presented data from a
wide variety of combustors and fuels used in Sweden. Table 5-9 lists the type of combustion unit and the
fuel used in 17 different power plants. Table 5-10 lists the N2O emission values from each unit. The -
N2O levels of the FBCs were high relative to the other methods of combustion presented. Grab samples
were collected in 15 L polyethylene coated aluminum bags and analyzed by GC/ECD. Silica gel was
used as a desiccant. Tests were performed both with and without the use of silica gel with no significant
variance in N2O values obtained.
Mr. S. Andersson of Gotaverken Energy Systems, an FBC manufacturer, presented data
collected with the assistance of Mrs. K. Dahiberg of SERI on their FBC units (see Table 5-11). The N2O
values presented were considerably lower than those obtained by Dahiberg at similar facilities. It was
hypothesized that the use of fuel lower in nitrogen content may account for this, since the samples were
collected and analyzed the same way.
80
-------
TABLE 5-7. SUMMARY OF N2O AND TOTAL NO* EMISSIONS (SLOAN)
Station
Unit
Load
(«We)
Ho. of
Samples
Dry at 31 0| dry
Ratio
N,0:NOX
as %
NjO
(ppm)
NO + NOi*
ppm
Eggborough
4
455
4
32 ± 2.0
700**
4.6
Eggborough
1
300
4
26 ± 3.3
-
-
Fiddler's Ferry
1(Low NOx)
485 ± 15
20
2.4 + 1.3
270 ± 25
0.8 ± 0.5
Fiddler's Ferry
3
47S t 25
16
4.5 ± 2.6
385 ± 30
1.1 ± 0.6
* Assumed NOj = 0.05 NO.
** Typical of previous flue gas monitoring.
81
-------
Figure 5-2. Extractive flue gas sampling system. (Sloan)
62
-------
TABLE 5-8. MEASUREMENT OF N20 UNDER VARIED CONDITIONS {SLOAN}
Run
No,
Date
Load
(We)
o»
dry
(%)
NO
dry
(ppm)
NjO dry
NjO dry at 3% 02 d.
18
(ppm)
GC
(PP»)
IB
(ppm)
GC
(ppm)
F1A
9.12.87
450
7.2
260
2.0
5.8
2.6
7.6
FIB
ft
450
7.6
230
2.4
3.7
3.2
5.0
F2A
11
490
6.4
210
2.3
1.1
2.8
1.4
F2B
If
490
6.4
210
2.3
2.2
2.8
2.7
F3A
N
490
6.3
210
2.2
1.6
2.7
2.0
F38
N
490
6.5
210
2.4
2.2
3.0
2.7
F4A
n
490
6.8
225
2.2
0.6
2.8
0.8
F4B
m
490
6.8
225
2.2
2.1
2.8
2.7
F5A
n
490
6.7
225
1.6
1.1
2.0
1.4
FSB
n
490
6.7
225
1.6
1.5
2.0
1.9
F6A
N
490
6.8
235
1.7
1.8
2.1
2.3
F61
H
490
6.8
210
1.7
0.6
2.1
0.8
F7A
It
490
6.5
215
2.4
1.0
3.0
1.2
F7B
*1
490
6.5
210
2.4
0.7
3.0
0.9
F8A
It
490
6.7
210
1.6
0.6
2.0
0.8
F9A
10.12.87
480
6.5
195
-
1.0
-
1.2
F9B
ft
480
6.5
195
-
1.4
-
1.7
F10A
If
480
6.4
195
-
2.9
-
3.6
F10B
If
480
6.4
195
-
1.6
-
2.0
F11A
11.12.87
490
6.5
220
_
3.4
_
4.2
83
-------
TABLE 5-9. COMBUSTION METHODS AND FUELS IN SWEDISH COMBUSTORS (DAHLBERG)
Plane
No. Fuel
01.
02.
03«
04.
Oil
Combustion unit
Small furnace for house heating - 30 kw
Top-mounted pressurized air burner 12 MW
Top-mounted with 2 rotary burners 50 MW
Tangentially placed angle burners - 70 MW
K5. Pulverized coal Front-wall mounted burners
K6.
K7.
K8.
ICS.
Coal
Tangentially mounted burners
Chain-grate, with a steam boiler
producing
Bubbling fluidized bed
circulating fluidized bed
90
125
MW
MW
40 ton/h
16 m
50 MW
F10. Wood chips
Fll.
Stationary inclined grate furnace 5.8 MW
Converted boiler with pre-oven 30 KW
V12. Firewood
Tile stove
T13. Feat
T14. *
Moving grate furnace
Circulating fluidized bed
5.5 KW
42 MW
G15. Natural gas
G16.
Front-wall mounted burners
Small furnace for house heatinc
150 MW
1.17. Rlack Liquor Recovery furnace
84
-------
TABLE 5-10, EMISSIONS FROM VARIOUS COMBUSTORS (DAHLBERG)
Hut
m.
K,0
ppov
. **°
B3/K3
CO,
t
l o£ Mxinai
lotdi&f
61.
1
2
«.a
O.J
0.S
7,1
)
2
».«
2
2
10.0
01.
31
17
12.4
i?
11
11.7
9J.
40
26
11.1
ids
3
i
70
4<
2
14.fi
SO
o«.
2)
12
14. S
38
20
14.S
ICS.
2li/
20
1.7
5
4
11.«
3
2
11.5
60
49
11.0
*#.
7
1
14.5
3
3
10.«
JT7.
1
0.1
11.0
•0
»
7
10.0
90
a
2
9.1
90
5
4
u.o
tl
XI.
137
90
14.3
o.
11
57
14.5
70
79
51
14.4
90
121
••
13.7
to
m
*
»
-
10€
4»
~
•0
no.
s
s
10. S
70
3
a
13.2
IS
4
4
9.9
10
ru.
5
4
13.7
75
5
1
11,$
•0
1
•
U.O
71
via.
0.9
1
4.0
I
2
4.4
10
14
4.0
3
*
2.0
m.
18
It
12.5
IS
«
15.S
9
<
U.O
tn.
52
34
1S.1
015.
2
»
1.0
SO
2
2
1.4
7 J
OK.
9
•
10
3
3
a.o
23
22
7.0
cn.
2
3
ia.o
0.1
0.3
9.3
2
a
10.3
The ¦aapla wsa itortd 3 days 6*(or« t&alytla
85
-------
TABLE 5-11. FBC EMISSIONS (ANDERSSON)
Boiler type
(design load)
Load
Fuel
n2o
C02
N20 at 10 % CO2
CFB (.20 MW)
IB MW
RDF
12 ppnt
10
11,0 \
11,7
1 1 ppm
CFB (SO MW)
St MW
Wood waste
1 9 ppm
14
14,5 \
1 }, 8
1 3 ppm
10
CFB (20 MW)
18 MW
Peat
10 ppra
6
-
-
CFB <40 MW)
26 MW
Peat
75 ppm
76
12,1 \
12,0
62 ppm
63
33 MW
Coal
56 ppm
49
13,0 \
43 ppm
34 MW
Coal +
8 \ Peat
69 ppm
62
12,4 \
13,7
56 ppm
45
86
-------
5.1.6 Excess Air Variation in a Circulating FBC
Mrs. L, Fuller of Studsvik Energy presented data collected on 2 different coals burned in a
40 MWth circulating FBC in which excess air was varied {see Table 5-12). A trend toward higher N2O
emission was observed with a higher stoichiometric ratio. Grab samples were collected in 2 L Tedlar
bags and analyzed by GC/ECD. Moisture removal was accomplished using silica gel. Ms. Fuller
remained that the samples were analyzed on the same day and 24 h after collection with little difference
in final values. Tests were also conducted to compare samples collected with and without silica gel to
ensure sample integrity.
5.1.7 NpQ Emissions From Power Plants in Finland
Dr. G, Hellen of EKONO discussed data collected from 20 power plants in Finland using different
types of combustion techniques and fuels (see Tables 15-13 through 15-15). Again, high N2O values
were found in FBCs, with considerably lower values found in other types of combustion facilities.
Figure 5-3 presents a graphic representation of the N2O and NOx values from all plants and fuels. There
did not seem to be an obvious relationship between N2O and NOx emissions. Dr. Hellen also described
his work in assessing the sampling and analytical equipment used. Samples were collected in
polyethylene lined aluminum bags using silica gel for moisture removal and analyzed by GC/MS A test
was performed demonstrating that the use of silica gel resulted in no losses of N2O. In addition, a
stability study was conducted on an actual stack gas sample. The sample was repeatedly analyzed over
a 2-wk period with no significant change in concentration observed.
5.1-8 Data from Dutch Power Stations
Dr. H. Spoelstra of KEMA presented data he collected while inventorying N2O emissions from
Dutch power stations. Table 15-16 represents data collected from facilities burning natural gas as well as
coal, revealing relatively low N2O emissions. Table 15-17 contains data obtained from a catalytic NOx
reduction facility. N2O samples were collected both before and after the titanium base catalyst. Lower
N2O concentrations were observed at the exit of the catalyst. No FBCs were sampled. Samples were
collected in teflon lined stainless-steel containers downstream of 2 ice-bath condensers and analyzed by
GC'ECD.
87
-------
TABLE 5-12. EXCESS AIR VARIATION DATA (FULLER)
Test no.1
Test no.2
COAL:
KUTZNEZ
AUSTRALIAN
% volatiles (dry)
% nitrogen
% sulphur
37.1
2 . 2
0.37
36 .9
1 .7
0 .58
Load (MW^)
Limestone addition (kg/h)
27
50
30
175
Test
no.
10.
Bed temp.
C
Flue gas
temp. C
CO+H.
ppm
SO,
ppi
3.2
5.0
7.3
870
848
826
125
129
133
41
40
48
146
94
79
3.5
5.3
5.3
7.2
848
840
852
845
127
132
132
137
34
34
35
34
164
164
166
153
I oiyjen
88
-------
TABLE 5-13. N2O FIELD TEST DATA FROM VARIOUS COMBUSTION
PLANT SOURCES USING AN FBC (HELLEN)
Flaat Coabu*-
tioa
tech-
nique
no
rue! ruel Fuel
type aitro^ea coal
coatent content
wei^ht-% wsight-%
dry dry
Circ. Coal
Pluidise
Wood(1/2)
Killed
p»«tC2/3)
'Stat.
Fluid,
B*d(5FB) 4-011(1/3}
Nill«d
peat
Co al
1.9
1.3
1.1
l.S
i.a
1.25
(I
7S
33
Si
55
76.1
Koala*1 Load
power
KWt I
SO
30
60
It
IS
33
• 5
10
100
100
IS
100
»20- Eai»»ien BOx- Emission H20/HO*
eeae. factor eooc. factor (voluae1
H20 BOx
ppav,dry a
-------
TABLE 5-14 N2O FIELD TEST DATA FROM VARIOUS COMBUSTION PLANT SOURCES USING A
CONVENTIONAL BURNER (HELLEN)
Last
Ceatws-
m*i ru*i
Pv«l
Boninal
Mad
¦20—
Balsaion
0
vaight-%
mt
%
ppav,dry if 820/MJppav,dry
¦9 H02/KJ
4
dry
dry
{input)
(input)
1
Bursar
Pulvari- 1.1
55
160
100
20
17
270
240
7
iad paat
2
•
P.p.(3/4 > 1.7
57
200
IS
29
20
240
ISO
12
R.gas(1/4!
3
¦
pulvari- 1.42
53.7
5.3
100
IS
11
440
290
4
3
¦
lad paat 1.42
53.7
5.3
74
21
13
330
220
6
4
Burn«r
Pulvari- 1.6
72
450
100
11
7
630
430
2
4
•
tad coal I.6
72
450
SO
3
2
500
340
1
5
¦
1.9
68
240
100
27
18
620
430
4
6
Burner
Haavy oil 0.5
86
3.2
95
39
23
240
150
16
7
•
0.5
86
3.2
100
73
46
230
150
32
8
Burttar
natural 0.09
SB.9
400
«6
1
1
450
280
«
?ts
90
-------
TABLE 5-15. N2O FIELD TEST DATA FROM VARIOUS COMBUSTION PLANT SOURCES USING
GRATE COMBUSTION (HELLEN)
float
Combus-
f«l
ru«i
Pu«l
Load
*20-
Emission
BOx-
Emission
tion
typ*
nitrog»n
coal
pow«r
cone ¦
factor
cone.
factor
t»ch-
content
content
H20
SO*
niqtx*
BO
vaight-%
KWt
1
ppmv,dry a? B20/HJpp*v,dry
¦f H02/MJ
dry
dry
(input)
(input)
1
Ccat«
Hillsd
1.6
55
15
90
17
13
240
220
p«afc
2
Grat*
Pr**s«d
1.8
55
5
•0
6
S
270
220
p.at
3
Scat*
Wood
0.7
50
14
60
1
1
80
80
chip*
«
*
¦
0.7
50
27
56
1
1
SO
80
5
Grata
coal
1.6
72
35
66
20
ia
150
140
6
6rat«
Coal(1/2)
1.1
SS
50
100
M
26
180
140
(volua*)
1
13
19
Wood(1/2 3
91
-------
TOTAL NUMBER OF PLANTS; 20
NOv - CONCENTRATION (PPM)
Figure 5-3. N2O and N0X concentrations from all plants and combined combustion techniques. (Helien)
92
-------
TABLE 5-16. N2O CONCENTRATIONS IN STACK GAS FROM ELECTRICITY POWER STATIONS
(SPOELSTRA)
Gas
Max cap
Cap
Range
No. Of
NOx
MW
MW
ppm (3% O2, day)
Samples
Cone, ppm
100
100
<1
4
215
100
100
<1
4
190
100
70
-1
4
215
Power Coal
Max cap
Cap
Range
No. of
NOx
MW
MW
ppm (3% O2, day)
Samples
Cone, ppm
120
120
10
1
20-25
2
180
108
3
1
144
10
1
180
10-15
3
400
400
.9
1
(350)
600
600
<1
2
(550)
600
600
5-8
6
(400)
( ) = estimated
93
-------
TABLE 5-17. CATALYTIC DaNOx (SCR) (SPOELSTRA)
Power Coal
Range of N2O Concentration {3% O2, dry)
CAP IN No. of Out No. of
MW ppm Samples ppm Samples
120 20-25 2 6 2
10 13 4
94
-------
5,1.9 Data From U.S. Power Plants
Mr. S. Wilson of Southern Company Services presented N2O data collected for the EPA.
Mr. Wilson was asked by the EPA to collect samples in order to assist in evaluating the GC analytical
system being developed through J. McSorley's in-house program. Samples were collected trom
conventional utility boilers with various firing configurations. Figure 5-4 illustrates the relatively high N2O
concentrations obtained from these coal-fired facilities. Figure 5-5 is a graphic representation of his data
plotted along with results presented at the Boulder workshop. A detailed description ot the sampling arid
analytical background revealed that almost no moisture removal was attempted prior to collection. The
sample canisters contained condensed combustion water, and analysis was performed as late as 2-wk
after collection. Because of the proposed sampling artifact and the use of coal with high sulfur content,
conditions seemed conducive to sample container N2O generation. These sampling and analysis factors
cause the integrity of these data to be highly suspect.
Dr. Muzio presented data that he had collected over the last few years for EPRI where during
sample collection, very little had been done to remove moisture or SO2 Figure 5-6 shows these data
along with a single point under the same firing conditions where both the SO2 and moisture have been
removed from the sample. Where earlier over 100 ppm N2O had been reported, replicating this condition
reveals an N2O concentration of approximately 3 ppm.
5.2 EXHAUST DATA FROM MOBILE SOURCES
Only 2 presentations addressed N2O emissions from mobile sources.
5.2.1 Comparison of Vehicles With and With Out Catalytic Converters
Mrs. A, Lindskog of SERI described her research on automobile 3-way catalytic NOx converters.
The concentration of N2O in automobile exhaust was measured during the Swedish driving cycle. The
fests included catalyst and non-catalyst vehicles under cold-start and warm-start conditions. Table 5-18
indicates that the N2O emissions from a catalyst vehicle are 4-5 times greater than that of a non-catalyst
vehicle, depending on start conditions. Table 5-19 presents N2O concentrations both before and after
95
-------
600
500
400
300
200 -
100
¦
NSPS T-Fired m/ OFA
0
Pre NSPS T—Fired (8 comer)
*
Pre NSPS T-FIred (4 Corner)
+
Pre NSPS Wall-Fired (circular)
A
Pre NSPS Wall-Fired (dual ceil)
~
Pre NSPS Wall-Fired (triple cell)
~
~i 1 1 1 r
0,2 0.4 , v 0.6
(Thousands)
NOx Emissions (ppm 9 3% 02)
y = 0.596x - 105.2
0.8
Figure 5-4. N2O versus NOx emissions. (Wilson)
96
-------
CM
o
N
E
a.
Ql
n
c
o
E
u
o
600
500
400 -
300
200
100
Various Combustion Sources
S. Wilson data
EPA 600/8-88-079
y = 0.414x
y = 0.186x
-T 1 1 1 ¥
0.2 0.4 0.6
(Thousands)
NOx Emissions (ppm @ 3% 02)
—1—
0.8
Figure 5-5. N2O versus NOx emissions, compared with Boulder Workshop data. (WilsonJ
97
-------
500
I ) n2o/nox
400
300
200
100 -
O BOILER TYPE;
SIZE:
FUEL;
a BOILER TYPE:
SIZE:
FUEL:
I—
O (J7)
(.30) |
(J3)
TANGENTIALLY FIRED
800 MW |
WESTERN BITUMINOUS (S: 0.59%, N: 0.87%)
TANGENTIALLY FIRED
400 MW (320 MW CURRENT CAPACITY)
EASTERN BITUMINOUS (S: 0J7te, N: 1.16%)
I I £s».
200
100
02, % DRY
5 6
raicoNDmoNBD sample
(S02 AND H2O REMOVED)
Figure 5-6, EPR1N2O measurements with repeat condition. (Muzio)
98
-------
table 5-18. CATALYST VEHICLE VERSUS CARBURETOR VEHICLE (LINDSKOG)
'
CATALYST VEHICLE
CARBURETTOR VEHICLE
test no
COLD
START
WARM
START
COLD START
WARM
START
NOx
N^O
NO^
n2o
NOk
n2o
NOx
N2O
1
O , 275
O, 13
O, 063
O, 031
1, 191
O, 096
1, 3*43
O . 09*4
' 2
O. 282
O. *41
O, O 75
0.62
1- 163
O, 0*4*4
1 . 321
O, 0*4*4
3-
O, 267
O , 61
O, 11*4
O jr 30
1, 125
O » 061
1,313
C O,57 >
*4
O „ 289
O » 1*4
0,06*4
0,070
1- 078
0,031
1, 2U6
O, 062
MEAN ;
a, 32
O. 25
O, OSS
O, 06 7
99
-------
TABLE 5-19. CATALYST VEHICLE DATA (LINDSKOG)
CONCENTRATION
OF N20 IN THE EXHAUST PIPE
, BEFORE AND
AFTER THE CATALYST (ppMV)
test NO
COLD START
WARM START
BEFORE
AFTER
BEFORE
AFTER
1
19
39
26
11
2
25
14 3
33
25
3
19
SO
2*4
U2
i*
18
29
30
27
THE VEHICLE WAS NOT PREPARED AHEAD OF THE TEST.
100
-------
the catalyst under the 2 start conditions, in regard to the n20 generation artifact, Mrs, Lindskog
encountered a situation where the N2O concentration increased in one of her samples, but she could not
explain a possible cause.
5.2^ Comparison of NpQ Emissions from City and Suburban Driving
Mr. M. Prigent of 1FP presented data, shown in Figures 5-7 and 5-8, reflecting research
conducted at the institute, The data compare N2O emissions proportional to catalyst temperature
produced by simulated city and suburban driving on gasoline engine vehicles. City driving, with lower
exhaust temperatures, generated higher N2O values than suburban driving. The city, or European driving
style, was contrasted with the driving style prevalent in the United States, in which higher exhaust
temperatures are generated and lower N2O values are observed. Tests were also conducted using
diesel engine vehicles, obtaining results generally similar to those of the gasoline engine vehicles.
101
-------
THREE- WAY Pt-Rh CATALYST (MGD 8? 4)
tFTER ZOOM EffClt/e BENCH fUPtD ACIHC
<169
no
it9
1*9
.—S,
2r
tta
?¦*
C
*64
55
N
u
*
II
»»
to
9
1
* '1
A
( .
/ s
! /
X
t
j /
i /
* m d M refi tei r.w m
i qui valence ratio
Steady state i-0 concentration in engine exhaust as a
function of eqaivalenee ratio; inlet catalyst 7 = 450°C
THREE-WaY Pt-Rh CATALYST (MCD 87 4)
ATTEJt 20OH IHCiHt BENCH MAPID ACWC
MUtVALCMCt RATIO - 1.00 , S.V - 90,000 /H
$46 460 tft iM
TEMPERATURE (SEC.C)
CO, HC and NO gross conversion and HO into »20 conversion
as a function of catalyst inlet T at 0 ¦ 1,00
THREE-*aY Pi-P.h CATALYST (MOD 258 A)
AfTEM 200* ESCWE BENCH RAPID ACiSC
MVrrLB FHLTT «M OCC.C S V . so ,000 ,'H
52
5
s
%
8
¦k i
• "x._: /
j*- 1 ;
* ;
:
! •
¦ y^*
; \ i •
V [
1-fc- «
— *c
* •+-#**
j
i/i -
¦ \ !»* HC i
V' i \ ' :
tt\ : \ ]
1 \ \
' \
j
T i
¦/:
J- ~ ~
.... - .
!
! ^
'
• ?
r ,
'*L/
•
4 1
-—*—
(H I,»J IM «.«* IM I si l.M (OJ ».»« 1.0*
rourvt! rvrr RtTm
CO, HC and SO grass conversion and NO into NjO conversion
as a function of equivalence ratio; inlet cata1v«t T«4S0°C
Figure 5-7. City versus suburban driving, (Pr^jeni)
102
-------
VEHICLt tMlbblUN
CftH-iOC t»«tl *** CA70LrSI
V7\
ES
Gasoline vehicle, 1-91 engjne, carburettor, with or
without open-loop 3-way Pt-Rh catalyst
rtm
Load
ms
VPM
*7*
VPM
s
JO00
4/4
m
44.1
imj
3SD0
4/4
41 S
JM
8-SS
<000
4/4
J60
16 .4
4 fj
30CC
3/i
m
U-1
4-09
Diesel engine, 1.91 ¦ steady state NO* and NjO
concentrations in exhaust gas
| u-
ea
ft ^ u«.
Gasoline vehicle, 2.21 engine, EFI^Ot sensor, vith
without closed loop 3-*ay Pt-Rh catalyst
ill 15
/ \
iw.* )H*i
Diesel vehicle, 1.91 engine, NjO global emissions
on ECE and FTP driving cycles
Figure 5-8. Gasoline versus diesel vehicles. (Prigent)
103
-------
SECTION 6
SESSION 4: GENERAL DISCUSSION TO ARRIVE AT PRACTICAL CONCLUSIONS
The remainder ol the workshop was designated for open discussion of material presented during
the past 2 days and for reaching practical conclusions on the direction of future research efforts.
Dr. S, Lanier of EERC, the session moderator, began by presenting 2 key questions representing EPA's
concerns:
1. What is the direct global contribution of N2O from fossil-fuel combustion versus other
sources, and are present global N2O inventories reliable?
2. Should N2O emissions be controlled, and wilt controls make a difference?
The information derived from the previous two N2O workshops resulted in a series of significant
issues that received additional research. These workshops concluded:
1. The need to understand the N2O formation and destruction mechanism.
2. The need to standardize sampling measurements of N2O from flue gases and the need to
develop an on-line analyzer.
3. The need to develop a global N2O combustion emission dala base.
4. The need to address the effect of possible controls.
In an attempt to define the relationship between combuslor sources and the annual increase in
measured N2O levels, Dr, Lanier presented a unique method of back-calculating the average influx of
N2O and the relative contribution of fossil-fuel combustion. Dr, Lanier pointed out that current data
suggest a 0.2-0.3 percent increase in N2O annually or about 0.69 ppb/yr. Dr. Hasenack was not satisfied
that Dr. Lanier's figures show a true increase and questioned their statistical significance. Dr. Levine
responded that the 0.2-0.3 percent increase in 12 different sites over an 8-yr period was valid and was
statistically supported by the investigators who reported the figures.
104
-------
Dr. Lanier accounted for his previous statement by saying that with the current atmospheric
burden and global sinks an estimated 5.5 x 1Q6 metric tons of N2O as nitrogen (N) are added to the
atmosphere annually. His method of back-calculating was sufficiently accurate that the AP42
atmospheric model was not essential. If coal combustion is the major source of anthropogenic N2O, then
given a global coal consumption of 2,4 x 1Q9 metric tons per year, petroleum equivalent, an average
emission of 132 ppm N2O would account for the observed increase in atmospheric N2O concentrations.
Dr. Lanier noted that if direct N2O emissions of coal-fired combustors exceeded 100 ppm, then these
combustors would be a dominant source of N2O. However, if N2O levels are less than 20 ppm, then coal
combustors are probably not a dominant source. Therefore, the validity of the data could establish
whether or not coal-fired utility boilers are significant contributors to the N2O increase.
Finally, mobile source N2O emissions were discussed. Dr. Lanier proposed that with an average
emission rate of 100 mg/mi and an estimated 1 6 x 1012 mi of driving in the United States each year,
American drivers contribute 1.5 x 105 metric tons of N2O to the total influx. If the United States accounts
for 48 percent of worldwide gasoline consumption, then the calculated total N2O emissions worldwide
would be 3.1 x 1QS metric tons, indicating that mobile sources are probably not a significant source of
N2O.
Dr. Lanier offered the following conclusions to the participants for discussion:
1. A critical need exists to develop and validate sampling and analysis procedures,
2. The GC/ECD method of analysis appears adequate, but the development of a continuous
emission monitor for N2O is encouraged.
3. Following validation, repeat sampling of utility boiler and fuel classes, which have previously
indicated high N2O emissions, is urgently needed.
4. Prediction of global N2O increases based on AP42 NOx factors is not scientifically
justifiable.
5. Mobile sources are a minor contributor to the N2O emissions.
6. European data indicate that direct N2O emissions from coal-fired boilers are a minor
contributor of N2O.
105
-------
7. Continued evaluation of N20 emissions from fluidized beds is encouraged.
8. Other combustion sources, such as a petroleum catalytic crackers should be evaluated,
9. No clear consensus exists as to how NOx combustion modifications will affect emissions.
A vigorous discussion period followed Dr. Lanier's conclusions, showing general agreement with
them. Dr. Hasenack, however, was not convinced that NO cannot be converted to N2O in the
atmosphere, and he recommended adding to the list a call for research to explore the question. The
group generally agreed and added that contribution from other combustion sources should also be
evaluated, biomass burning in particular.
The conference participants agreed most strongly that a suitable and valid sampling protocol
should be developed. Dr. de Soete solicited European volunteers to assist in investigating the sampling
protocol. IFP and EPA had already committed additional research to that area. Dr. Smart and
Mr. Morgan offered their services and suggested a collaboration with EPA, since strong ties were already
present between EPA and their organizations, Dr. de Soete and Mr. McSorley were suggested as
designated contacts to facilitate information exchange on sampling protocols.
Dr. Hein summarized the final conclusions reached from the discussion as follows:
1. New information on N2O formation and destruction mechanisms in stack gas samples
precludes earlier coal- and oil-fired boiler data.
2. Assuming that the European data are valid, the direct emission of N2O from coal-fired
boilers is a minor contributor to observed global increases in N2O.
3. Sampling, handling, and measurement techniques must be thoroughly studied to assess NO
conversion to N2O in sample containers characterizing catalytic effects.
4. Repeat and extended measurements of NOx and N2O emissions from different combustion
equipment types, fuel, etc., are needed, with special attention to FBC units as high emitters
of N2O.
5. All other combustion sources, including biomass burning, should be included in the
evaluation of global N2O emissions.
106
-------
In closing the workshop, Mr, P. Eyzat, director ol the IFP, expressed his concern over the
long-term global effects of N2O emissions and the crucial need for collaboration among nations. He
stated that although more information is needed, it is safe to assume that mobile sources have a minor
rote in global increases in N2O. Mr. Eyzat also expressed concern over the dilemma that low NO*
emission systems may have high N2O emissions, citing three-way catalytic mufflers and FBCs as
examples. How should the problem ot the N2O increase be ranked in the priorities ot combustion system
research and development? More research and funding are needed from all nations.
107
-------
SECTION 7
SUMMARY AND CONCLUSIONS
information presented at this workshop suggests that N2O emissions have a significant effect on
both stratospheric O3 depletion and global climate change. Studies indicate an annual increase in N2O
in the atmosphere of 0.2-0.3 percent, and isolating the source of the increase is essential to prevent
further worldwide climate change. There is still some question as to whether the combustion of fossil
fuels is directly responsible for the increase. More research is needed 10 find reliable and accurate
techniques for establishing a data base of N2O emissions from combustion.
The mechanisms and conditions involved in N2O combustion formation and destruction are not
well characterized. Nitrous oxide formation is possible in post-flame, temperature-dependent, gas phase
reactions. Material presented at the workshop suggests that volatile HCN exposed to temperatures
between 1,150 K and 1,500 K can be converted to N2O. The possibility of this is supported by models
and experiments. Heterogeneous N2O formation mechanisms are also possible, but the amount of N2O
they produce is less than observed high stack emissions. The use of catalysts for NOx reduction
methodologies have not been well characterized for N2O formation. The production of N2O is
temperature dependant and does not seem to be significant in these reactions.
Most of the information presented at the workshop was related to N2O measurement lechniques.
Dr. Muzio's presentation on N2O generation in sample containers questions the validity and accuracy of
much of the reported data. A verifiable sampling technique is essential to assess the role of fossil-fuel
combustion in rising concentrations of atmospheric N2O. Dr. Muzio pointed out that NOx, SO2, and H2O
are involved in the N2O artifact scenario. The removal of either of these components could halt the
generation of N2O. However, considerable research is still needed in this area, because these
mechanisms may also occur in the atmosphere. Should this be the case, positive identification of the
reaction mechanism is crucial, which opens an entirely new avenue of research. Although sampling
108
-------
questions still exist, the analytical methodologies employed in quantifying grab samples are adequate.
The development of an on-line continuous N2O monitor would be invaluable in combustion research,
Substantial data were presented on the N2O direct-emission levels from various European
stationary utilities employing several different types of combustion techniques. Fuels other than coal were
included in these presentations. Considerably lower N2O emissions were realized in relation to United
States data. In the presentations, methods of sampling were also described. Much of the European data
include extensive moisture removal in sampling. A refrigeration condenser was used, and desiccants
were employed. Several presentations included on-line measurements, which showed very low N2O
concentration levels on non-FBC combustors. Table 7-1 represents a compilation of N2O emissions data
presented over the course of the workshop. The data presented on FBCs showed significantly higher
levels of N2O emissions, which supports predictions of computer modeling. The emission of N2O from
mobile sources appears to be very dependent on driving styles, The conversion of NOx to N2O is directly
related to catalyst temperature, which is proportional to driving speed or rotations per minute. Lower
temperature catalysts produce higher N2O emissions.
A general discussion period completed the 2-day workshop, put in perspective the information
presented, and enabled participants to set priorities tor research and goals, Standardization of a practical
and accurate sampling protocol was deemed most essential. There is no way to assess N2O emissions
from fossil-fuel combustion unless an accurate measurement technique is available. Once a suitable
sampling protocol has been installed, the retesting of utility boilers, catalytic crackers, and other
combustion sources should be performed. The combustion of fuels other than coal and heavy distillates
should be considered, particularly biomass burning, where limited sampling efforts have shown significant
emissions. A general consensus was reached that direct emissions of N2O from the combustion of fossil
fuels is possibly not the major source of N2O in the atmosphere, but more research should be conducted
to support this conclusion.
109
-------
TABLE 7-1. N20 EMISSIONS DATA SUMMARY
COUNTRY GROUP
COMBUSTOR
SIZE
LOAD
TYPE
MW
MW
FUEL USED
WALL FIRED
450
455
PULVERIZED COAL
WALL FIRED
450
300
¦
TANG. FIRED
500
475
•
TANG. FIRED
500
485
PULVERIZED COAL
CIRC FL BED
80
85%
COAL
-
30
70%
COAL
CIRC FL BED
60
100%
MILLED PEAT/WOOD
STAT FL BED
18
100%
MILLED PEAT/OIL
-
15
75%
, MILLED PEAT
STAT FL BED
35
100%
COAL
CONV BURNER
160
100%
PULVERIZED PEAT
200
85%
PULV PEAT/NAT GAS
5.3
100%
PULVERIZED PEAT
5.3
74%
PULVERIZED PEAT
450
100%
PULVERIZED COAL
450
60%
PULVERIZED COAL
240
100%
PULVERIZED COAL
3.2
95%
HEAVY OIL
3.2
100%
HEAVY OIL
CONV BURNER
400
86%
NATURAL GAS
GRATE
15
90%
MILLED PEAT
¦
5
80%
PRESSED PEAT
-
14
60%
WOOD CHIPS
¦
27
56%
WOOD CHIPS
¦
35
66%
COAL
GRATE
50
100%
COAL/WOOD
SAMPLE
COLLECTION
METHOD
MOISTURE
REMOVAL
METHOD
ANALYTICAL
METHOD
N20
PPM
NOX
PPM
S02
PPM
ENGLAND
CERL
FINLAND
EKONO
PVC BAG
PVC BAG
PVC BAG/(ON-LINE)
PVC BAG/(ON-LINE)
P.E. LINED
ALUMINUM BAGS
CONDENSER
CONDENSER
SILICA GEL
P.E. LINED
ALUMINUM BAGS
GC/ECD
GC/ECD
GC/ECD/(IR)
GC/ECD/(IR)
GC/MS
SILICA GEL
GC/MS
32
26
2.4
4.5
96
80
7
16
43
23
20
29
18
21
11
3
27
39
73
1
17
6
1
1
20
34
-700
270
385
130
100
170
200
220
130
270
240
440
330
630
500
620
240
230
450
280
270
80
90
150
180
-1200
-1200
-1200
-1200
(continued)
-------
TABLE 7-1. N20 EMISSIONS DATA SUMMARY (continued)
COUNTRY GROUP
COMBUSTOR
TYPE
SIZE
MW
LOAD
MW
FUEL USED
SAMPLE
COLLECTION
METHOD
MOISTURE
REMOVAL
METHOD
ANALYTICAL
METHOD
N20
NOX
S02
PPM
PPM
PPM
ND
355
ND
295
—
ND
235
—
1
480
—
1
335
—
ND
220
—
5.1
—
—
3.3
386
—
2.1
386
—
2.5
—
—
-1
215
-1
190
—
-1
215
—
10
—
—
20-25
—
—
3
—
—
10
—
—
10-15
—
—
<1
350
—
<1
550
—
5-8
400
—
3-6
—
ITALY ENEL OPPOSITE FIRING 320
ONCE THROUGH
NETHERLANDSKEMA
OPPOSITE FIRING
ONCE THROUGH
CORNER FIRED
DRUM BOILER
CORNER FIRED
DRUM BOILER
SCR
320
171
171
100
100
100
120
120
180
180
180
400
600
600
120
310
280
200
310
250
160
130
165
165
165-115
100
100
70
120
120
108
144
180
400
600
600
120
#6 FUEL OIL (3% S)
#6 FUEL OIL (3% S)
BITUM. COAL (.6% S)
BITUM. COAL (.6% S)
NAT GAS
NAT GAS
COAL
ON-LINE
CONDENSER
GC/ECD
ON-LINE CONDENSER
TEF LINED SS BOMB CONDENSER
GC/ECD
GC/ECD
COAL
TEF LINED SS BOMB
CONDENSER
GC/ECD
(continued)
-------
TABLE 7-1. N20 EMISSIONS DATA SUMMARY (continued)
COUNTRY GROUP
COMBUSTOR
SIZE
LOAD
TYPE
MW
MW
FUEL USED
TOP FIRED
12
OIL
TOP FIRED ROTARY
50
100%
-
TOP FIRED ROTARY
50
70%
¦
TOP FIRED ROTARY
50
60%
¦
TANG FIRED
70
—
OIL
TANG FIRED
125
—
PULV COAL
WALL FIRED
90
—
PULV COAL
BUBBLING FBC
16
—
COAL
CIRC FBC
50
60%
¦
-
-
70%
•
-
¦
80%
¦
-
-
90%
¦
CIRC FBC
50
—
COAL
STAT GRATE
5.8
70%
WOOD CHIPS
STAT GRATE
•
80%
¦
STAT GRATE
5.8
85%
•
PRE-OVEN
30
75%
¦
PRE-OVEN
30
80%
WOOD CHIPS
MOVING GRATE
5.5
—
PEAT
CIRC FBC
42
—
PEAT
WALL FIRED
150
50%
NAT GAS
WALL FIRED
150
75%
NAT GAS
RECOVERY FURN
—
—
BLACK LIQUOR
CIRC FBC
20
18
RESIDUAL DIST FUEL
¦
20
18
PEAT
-
40
26
PEAT
-
•
33
COAL
¦
40
34
COAL/PEAT
CIRC FBC
50
51
WOOD WASTE
SAMPLE
COLLECTION
METHOD
MOISTURE
REMOVAL
METHOD
ANALYTICAL
METHOD
N20
PPM
NOX
PPM
S02
PPM
SWEDEN
SERI
GES
P.E. LINED
ALUMINUM BAGS
P.E. LINED
ALUMINUM BAGS
P.E. LINED
ALUMINUM BAGS
P.E. LINED
ALUMINUM BAGS
SIUCA GEL
GC/ECD
SILICA GEL
SILICA GEL
GC/ECD
GC/ECD
SILICA GEL
GC/ECD
17.28
40
3
4
23,38
3.7
3-60
137
128
88
106
79
165
5
4
3
5.8
5
9-18
52
2
2
~1-2
10,12
6,10
76
49,56
62.69
14,19
(continued)
-------
TABLE 7-1. N20 EMISSIONS DATA SUMMARY (concluded)
SAMPLE
MOISTURE
COMBUSTOR
SIZE
LOAD
COLLECTION
REMOVAL
ANALYTICAL
N20
NOX
S02
COUNTRY GROUP
TYPE
MW
MW
FUEL USED
METHOD
METHOD
METHOD
PPM
PPM
PPM
S.E.
CIRC FBC
40
27
COAL (.4% S)
TEDLAR BAGS
SILICA GEL
GC/ECD
123-209
61-87
79-146
CIRC FBC
40
30
COAL (.6% S)
TEDLAR BAGS
SILICA GEL
GC/ECD
233-444
36-64
153-166
W. GERMANY RWE
PRESS FBC
0.34
COAL
SS AND GLASS BOMBS
CONDENSER
GC/TCD
115
255
—
-
—
0.56
193
203
—
¦
—
0.59
168
211
—
¦
—
0.71
95
203
—
PRESS FBC
—
0.75
95
173
—
¦—
300
—
BROWN COAL (-.9% S)
5-11
102-120
—
—
300
—
-2
102
—
—
300
—
SS AND GLASS BOMBS
CONDENSER
GC/TCD
5-S
322
—
L&CS
WET BOTTOM
510
BITUM COAL
SS AND GLASS BOMBS
CONDENSER
GC/TCD
27
863
—
DRY BOTTOM
—
465
BITUM COAL
¦
¦
¦
25
196
—
DRY BOTTOM
—
64
BITUM COAL
¦
-
¦
4
307
—
GRATE
—
42(t/h)
MUNICIPAL REFUSE
SS AND GLASS BOMBS
CONDENSER
GC/TCD
5
101
—
VGB
COAL COMBUSTOR
0.05
COAL DUST
ON-LINE
IR
30
2000
FBC
—
0.1
PULV COAL
ON-LINE
—
IR
100
—
—
USA SCS
WALL, TANG. FIRED
—
—
PULV COAL
SS AND GLASS BOMBS
—
GC/ECD
100-400
300-500
—
FERCO/
TANG FIRED
800
800
PULV COAL
GLASS BOMB
GC/ECD
130
EPRI
TANG FIRED
600
800
PULV COAL
GLASS BOMB
*
GC/ECD
3
—
—
EPA
DOWN FIRED
0.029
0.017
PULV BIT COAL
SS BOMB
CONDENSER
GC/ECD
10-220
150-900
DOWN FIRED
0.029
0.017
PULV BIT COAL
ON-LINE
CONDENSER
GC/ECD
< 10
600
—
UA/DOE
DOWN FIRED
0.029
0.027
PULV COAL
GLASS BOMB
CONDENSER#
GC/ECD
<5
400-900
—
NOTE: * BOTH MOISTURE AND S02 WERE REMOVED PRIOR TO SAMPLE COLLECTION
# A WATER QUENCH PROBE WAS USED AT POINT OF SAMPLING
-------
SECTION 8
REFERENCES
1. 1968-1983 Data From C.D. Keeling, Scripps Institution ol Oceanography. 1984 Data From
GMCC/NOAA (Preliminary).
2. M.A. Khalil and R.A. Rasmussen: Tellus 35B. (1983).
3. D.R. Blake and F.S, Rowland, Science. 239.1988.
4. G.I. Pearman, Etheridge, D., De Siiva F., and Fraser, P.J. Nature. 320 (i986).
5. M. Kavanaugh, Atmospheric Environment 21, 463-468 (1987).
6. M.B. McElroy and S.C. Wofsy: Tropical Forests and World Atmospheres. 1985.
7. W. Seiler and R. Conrad: Geophysiology and Amazonia, 1987.
8. Ch. Bruhl, P.J. Crutzen, Climate Dynamics 2,173-203 (1988). 1
9. Chang, S.G. et al. in Flue Gas Desulfurization, edited by J.L. Hudson and G.T. Rocheile, ACS
Symposium Series 188 American Chemical Society, Washington, DC, pp 127-152.1982.
10. Kramlich, J.C. et al., EPA/NGAA/NASA/VSDA N2O Workshop Volume I: "Measurmenet Studies
and Combustion Sources," September 15-17,1987, Boulder, Colorado. EPA-600/8-88-079
(NTIS No. PB88-214911), 1988.
114
-------
APPENDIX A
EUROPEAN WORKSHOP ON N2O EMISSIONS
Organized by EPA and 1FP
Institut Francais du Petrole (Building "le Sequoia")
June 1-2,1988
AGENDA
Wednesday June 1,1988
8:00 a.m Registration of Participants
8:30 a.m Welcome Address (G. de Soete, IFP)
OPENING SESSION
8:45 a.m. 1.0 Stratospheric Ozone Depletion and Global Climate Change
(S. Seidel, EPA)
9:30 a.m. 2.0 OECD Perspective (P, Stolpman, OECD)
Discussion
10:00 a.m. 3.0 Impacts of N2O and Other Trace Gases on Stratospheric Ozone
(J. Levine, NASA}
11:45 a.m. 4.0 Modeling the Effects of N2O and Other Trace Gases on Climate and
Ozone Distribution (C. Bruhl, Max Planck Institut)
Discussion
12:00 p.m. SESSION 1: MECHANISMS OF N2O FORMATION AND
DESTRUCTION DURING COMBUSTION
1.0 Basic Kinetics (Discussion moderator, N. Brown,
Lawrence Berkeley Lab)
1.1 Gas Phase Kinetics (J. Kramiich, EERC)
1:00 p.m. Lunch (Third Floor)
2:00 p.m. 1.2 Heterogeneous Reactions (G. de Soete, IFP)
Discussion
3:00 p.m. Break
3:45 p.m. 2.0 Overall Chemical Information from Laboratory Combustor Experiments
{Discussion moderator, J. Wendt, University of Arizona)
Data Presented by:
W. Linak, EPA (15 min)
J. Kramiich. EERC (10 min)
A. Williams, Leeds University (20 min)
J. Smart, IFRF (15 min)
115
-------
SESSION 2: N20 MEASUREMENT TECHNIQUES
(Discussion Moderator: J. MeSorley, EPA)
Problems (J, MeSorley, EPA)
The Need to Develop Standardized Sampling and Analysis Methods
Ambient Methods
Source Methods
Gas Chromatography with ECD (J. Ryan, Acurex Corporation)
Interferences from Sulfur Dioxide
Continuous On-line Infrared Methods (L. Muzio, EPA)
Optical Interferometry: A Potential Method for N2O Analysis
{A. Calais, CNS)
Coffee (Cafeteria, Ground Floor)
SESSION 3: FULL-SCALE FIELD DATA
(Discussion Moderator; K, Hein, RWE)
Flue Gas Exhaust Data from Stationary Sources (W. Linak, EPA)
Data Presented by:
K. Hein, RWE (10 min)
H. Koeser, Steinmuller (5 min)
R. R.Tarli, ENEL (15 min)
J. Philippe, CERChar (15 min)
A. Sioan, CEGB (10 min)
K. Dahlberg, SERI (15 min)
A. Feugier, IFP (10 min)
L, Fuller, Studsvik Energy (10 min)
G. Hellen, EKONO (30 min)
H. Ten Brink, ENC (5 min)
H. Spoelstra, CHEMA (10 min)
S. Wilson, SCS (10 min)
Discussion
Lunch {Third Floor)
Exhaust Data from Mobile Sources (A. Douaud, IFP)
Data presented by:
A. Lindskog, SERI (15 min)
M. Prigent, IFP (20 min)
SESSION 4: GENERAL DISCUSSION TO ARRIVE AT PRACTICAL
CONCLUSIONS
(Discussion Moderator: S. Lanier, EERC)
Conclusions
Relative N2O Fluxes
Scatter and Uncertainties of the Data
Delineation of Information Gaps
Reemphasis on Importance of Standardized Sampling and
Analysis Methods
Satisfaction with N2O Source Sampling and Analysis Methods?
116
-------
3.0 Needs lor Further Daia Collection
Gaps
Priorities
4.0 Preparation of Written Record of Conclusions
5:15 p.m. CLOSING ADDRESS (P. Eyzal, IFP)
117
-------
APPENDIX B
EUROPEAN WORKSHOP ON N2O EMISSIONS
Co-Sponsored by EPA AND IFF
institui Francais du Petrole
LIST OF PARTICIPANTS
Mr. J. Anderson
ASEA PFBC
S-6120Q Finspong, Sweden
Mr. Sven Andersson
Golaverken Energy Systems
P.O. Box 8734
40275 Goteborg, Sweden
Dr. Aldo BaWacchi
Ente Nazionale Energia Eleftrica
DptSAS
Via A. Pisano 120
1-56100 Pisa, Italy
Mr. Niklas Berge
Studs vik Energy
Studs vik
61182 Nykoping, Sweden
Dr. Richard Borio
Combustion Engineering
1000 Prospect Hill Road
P.O. Box 500
Windsor, Connecticut 06095
Mr. H.M. Ten Brink
Netherlands Energy Research Foundation
P.O. Box 1
1755 ZG Petten, Netherlands
Dr. Nancy Brown
HI Cyclotron Road, Bldg 29C - Room 100
c/o Lawrence Berkeley Laboratories
Berkeley, California 94720
Dr. Christoph BrOhl
Max Planck Institut fur Chemie
P.O. Box 3060
D-6500 Mainz, West Germany (RFA)
118
-------
Madame Michele Chevrier
Regie Nationale des Usines Renault
Direction des Etudes - DLA
8-10 Avenue Emfle Zola
92109 Boulogne Billancourt, France
Mrs. K. Dahlberg
Swedish Environmental Research Institute
S-40258 Goteborg, Sweden
Mr. Andre Douaud
Institut Francais du Petrole
BP 311
92506 Rueil-Malmaison, France
Dr. Robert Epprrley
Fuel Tech. Europe
28A Cadogan Square
London SW1 XOJH, England
Mr. Pierre Eyzat
Institut Francais du Petrole
BP 311
92506 Rueil-Malmaison, France
Mrs. Lisa Fuller
Studsvik Energy
S-61182 NYKOPING, Sweden
Dr. Dian Gaiten
NOAA
USA
Mr. R.E. Hall
U.S. Environmental Protection Agency
Combustion Research Branch (MD-65)
Research Triangle Park, North Carolina 27711
Dr. H.J.A, Hasenack
Royal Dutch Shell Laboratory
Badhuisweg 3
1031 CM Amsterdam, The Netherlands
Prof. K Hein
RWE, Betriebsverwaltung Fortuna
Posttach 1461
5010 Bergheim 4, West Germany (RFA)
Dr. Goran Helien
EKONO Oy
P.O. Box 27
SF-00131 Helsinki, Finland
-------
Dr. AJain Henriel
Peugeot S.A.
Etudes ei Recherches
Direction Technique, Lab. da Chemie
Centre Technique Citroen
78140 Velizy Villacoublay, France
M. Toib Hulggard
Instltuitet for Kemitenknik
Danmarks Tekniske Hojskole
Bygning 229
2800 Lyngby, Denmark
Dr. J. Jacobs
Combustion Department
VGB Technische Vereinigung der
GrosSkiattwerksbetreiber
Kfinkestrasse 27-31
4300 Essen 1, West Germany (RFA)
Mr. Sven A. Jansson
ASEA PFBC
S-61200 Finspong, Sweden
Prof. Karl Knoblauch
Berbau-Forschung BmbH
Franz-Fisher-Weg 61
4300 Essen 13, West Germany (RFA)
Dr. Heinz Koeser
L & C Steinmuelier GmbH
Fabrfcstrasse 1
5270 Gummersbach, West Germany (RFA)
Dr. Hans-Jurgen Krabbe
Vereinigte Elektrizitratswerke
Westfalen AG, Postfach 105056,
Rheindanddamm 24
4600 Dortmund 1, West Germany (RFA)
Dr. John Kramlich
Energy and Environmental Research Company
18 Mason
Irvine. California 92718
Dr. W.S Lanier
Energy and Environmental Research Company
3622 Lyckan Parkway, Suite 5006
Durham, North Carolina 27707
Mr. Dominique Laurent
ELF, Centre de Recherche Solaize
BP 22
69360 Saint Symphorien d'Ozon, France
120
-------
Dr. Joel Levine
NASA
Lang ley Research Center
Hampton, Virginia 23665
Dr. W. P. Linak
U.S. Environmental Protection Agency
Combustion Research Branch (MD-65)
Research Triangle Park, North Carolina 27711
Prof. Oliver Lindkvist
Chalmers University of Technology
Department of Inorganic Chemistry
41296 Goteborg, Sweden
Mr. Jan Lindquist
Ambassade de Suede
Dpt Technique
17 Rue Barbet de Jouy
75007 Paris, France
Mrs. Anne Lindskog
Swedish Environmental Research Institut
P.O. Box 47086
S-40258 Goteborg, Sweden
Dr. F.C. Lockwood
Mechanical Engineering Department
Imperial College
Exhibition Road
London, SW7 2BX, England
Mr. Joseph McSorley
U.S. Environmental Protection Agency
Combustion Research Branch (MD-65)
Research Triangle Park, North Carolina 27711
Dr. Mark Morgan
International Flame Research Foundation
c/o Hoogovens BV, P.O. Box 1000
1970 CA, Ijmuiden, The Netherlands
Dr. Larry Muzio
Fossil Energy Research
23342 Unit C
South Point
Laguna Hills, California 92653
Dr. Heine Nielsen
Petrokraft AB
Box S2090
40025 Gothenburg, Sweden
121
-------
Mr. Jean-Luc Phillip©
CERChar
Rue Aime Dubost, BP 19
62670 Maztngarbe, France
Mr. Michel Prigent
Institut Francais de Petrole
BP 311
Rueii-Malmaison, France
Prof Daniel Puechberty
UA CNRS 230
Coria, BP 118
76130 Mont Saint Aignan, France
Dr. Ludwig Raible
RWE Betriebsverwaltung Forluna
Postfach 1461
5010 Bergheim 4, West Germany
Mr. Jeffrey V. Ryan
Acurex Corporation
Research Triangle Park, North Carolina 27709
Mrs. Kelley Ryan
Acurex Corporation
Research Triangle Park, North Carolina 27709
Mr. S. Seidel
U.S. Environmental Protection Agency
ANR-445
Oflice of Air and Radiation 40 L
M Street, SW
Washington, DC 20460
Dr. Sam A. Sloan
Central Electric Research Laboratories
Kelvin Avenue
Leatherhead, Surrey KT22 7SE
England
Dr. John P. Smart
International Flame Research Foundation
c/o Hoogoven, P.O. Box 10000
1970 CA, Ijmuiden, The Netherlands
Dr. Gerard de Soete
Institut Francais du Petroie
BP 311
92506 Rueil-Malmaison, France
-------
Dr. Paul Slolpman
Head of Pollution Control Division
OCDE
15 Bd de I'Admiral Bruix
15016 Paris, France
Dr. Roberto R. Tarli
Ente Nazionale Energia Elettrica
Opt SAS
Via a; Pisano 120
1-56100 Pisa, Italy
Mr, Jean-Noel Thuillard
Section Pollution et Carburants
Centre Technique Renault
Rue Cochei
91510 Le Lardy, France
Prof. Jost L.O. Wendt
University of Arizona
Geology Building, Room 142
North Campus Drive
Tuscon, Arizona 85721
Prof. Alan Williams
Department of Fuel and Energy
Leeds University
Leeds LS2 9JT, England
Dr. Steven Wilson
Southern Company Services
800 Shades Creek Park
P.O. Box 2625
Birmingham, AL 35202
Dipl. Ing. Gunter Zellinger
Oesterreichische Draukrattwerke AG
Hauptverwaltu ng/BM
Kohldorfestrasse 88
A-9020 Klagenfurt, Austria
-------
TECHNICAL REPORT DATA
{Please read fnitructioiis on the reverse before completing}
1. REPORT NO, 2.
EPA-600/9-89-089 PI
3. RECIPIENT'S ACCESSION-NO. , „ _
,90 1 2 80.58 as.
1. TITLE AND SUBTITLE
EPA/IFP European Workshop on the Emission of
Nitrous Oxide from Fossil Fuel Combustion
5. REPORT DATE
October 1989
6. PERFORMING ORGANIZATION CODE
7, AUTHOR(S)
Jeffrey V. Ryan and Ravi K. Srivastava,
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Acurex Corporation
P. O. Eox 13109
Research Triangle Park, North Carolina 27709
10. program element no,
11, CONTRACT/GRANT NO,
68-02-4701
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Proceedings; 6/88-9/89
14. SPONSORING AGENCY CODE
EPA/600/13
is, supplementary notes^EERL project officer is Joseph A. McSorley, Mail Drop 65, 919/
541~2920. EPA-600/8-88-079 and EPA-600/8-86-035 relate to earlier workshops
ill this-, series.
16, ABSTRAc-T„rrj;.,ie rep0rj. summarizes the proceedings of an EPA/lnstitut Francais du
Petrole (IFP) cosponsored workshop addressing direct nitrous oxide (N20) emission
from fossil fuel combustion. The third in a series, it was held at the IFP in Rueil-
Malmaison, France, on June 1-2, 1988. Increasing atmospheric IS"20 concentrations
have been linked to depletion of stratospheric ozone (03) and to global'climate war-
ming. The combustion of fossil fuels has been identified as a potential major anthro-
pogenic source of N2G. The workshop had two goals: (1) to exchange information
among various international research and industrial groups that are involved in N20
chemistry, modeling, and measurement; and (2) to develop a network for coordina-
ting future related efforts'. The five technical sessions addressed: stratospheric 03
depletion and global climate"-change, mechanisms of N20 formation and destruction
during combustion, N20 measurement techniques, full-scale field data, and practi-
cal conclusions based on general^discussion, A sampling artifact discovered during
an EPRI funded research study revealed that N20 can be generated in a sample con-
tainer in the presence of NOx, S02, and H2C. This artifact potentially discredits
much of the N20 emissions data collected from samples containing the above com-
pounds when stored for some time prior to analysis.
17, KEY WORDS AND DOCUMENT ANALYSIS
a, DESCRIPTORS
b.IDENTIFIERS/OPEN ENOED TERMS
c. COSATI Field/Group
Pollution Chemistry
Nitrogen Oxide (N20) Mathematical
Fossil Fuels Models
Combustion Measurement
Ozone Sampling
Climatic Change
Pollution Control
Stationary Sources
Global Warming
13B 07
07B
2 ID 12 A
21B 14G
14B
04B
19, DISTRIBUTION STATEMENT
Release to Public
19, SECURITY CLASS (This Report/
Unclassified !
2t.NP_OF_PAG ES
20, SECURITY CLASS (Thispage)
Unclassified
22, PRICE
EPA Form 2220-1 (9-73j
i
------- |