&EPA
United States
Environmental Protection
Agency
Industrial Environmental Researc
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-81-158 Dec. 1981
Project Summary
Fuel Decomposition and
Flame Reactions in
Conversion of Fuel
Nitrogen to NOx
A. Axworthy, D. Kahn, V. Dayan, and D. Woolery
4
An experimental and analytical re-
search program was conducted to
provide information on the chemical
phenomena involved in the conversion
of air and fuel nitrogen to NOx- The
program was divided into three tasks.
Under Task 1, Fuel Decomposition,
the early (preflame) reactions of fuel
nitrogen species were investigated
using both one- and two-stage reac-
tors. Additional inert pyrolysis experi-
ments were carried out with several
fuel oils. The oxidative pyrolysis ex-
periments involved model compounds
(pyridine, benzonitrile, NH3, and HCN),
fuel oils, and coals. The Task 2 study
(Combustion Kinetics) was carried out
in a low-pressure flat-flame burner.
The formation of "thermal," "prompt,"
and "fuel" NO was investigated in
CH4-O2-N2 flames doped with small
amounts of NH3 or HCN. In addition,
the formulation of fuel NO was in-
vestigated in a Hz-CO-Oz-Ar-NHs
flame. This simulated the combustion
of future low-Btu fuel containing
some NH3 as an impurity. Calculations
were made under Task 3, Data Anal-
ysis, on the H2-CO-O2-Ar-NH3 flame
using a detailed kinetic-diffusion
model.
This Project Summary was devel-
oped by EPA's Industrial Environmen-
tal Research Laboratory, Research
Triangle Park, NC, to announce key
findings of the research project that is
fully documented in a separate report
of the same title (see Project Report
ordering information at back).
Introduction
An experimental and analytical re-
search program was conducted to. pro-
vide additional information on the
chemical phenomena involved in the
conversion of air and fuel nitrogen to
NO, in the combustion of fossil fuels.
The experimental techniques developed
under the previous program (EPA Con-
tract 68-02-0635) were improved to
permit the chemistry to be studied in
more detail.
The program was divided into three
tasks. The pyrolysis and oxidative
pyrolysis of fuels and model fuel nitrogen
compounds was investigated in quartz
microreactors in Task 1. Low-pressure
flat-flame burner studies were conducted
under Task 2, and in Task 3 an attempt
was made to fit some of the burner
results to a preliminary kinetic-diffusion
model. The flat-flame burner study was
extended to investigate the formation of
thermal (prompt) NO and fuel NQ'from a
methane flame and the formation of fuel
NO in a simulated low-Btu fuel (CO and
H2) flame.
-------
Task 1 - Fuel Decomposition
Additional inert pyrolysis experiments
were conducted under this task but the
major effort involved oxidative pyrolysis.
Inert Pyrolysis
HCN yields were measured from
coals, fuel oils, and alternative fuels
under inert conditions. Most of the inert
pyrolysis experiments were carried out
in a two-temperature-zone reactor and
it was established that most of the HCN
is not produced directly in the primary
vaporization-pyrolysis process but forms
later in the secondary reactions of
' volatile nitrogen species. The alternative
liquid fuels derived from coal and oil
shale produced strikingly larger per-
centage yields of HCN. When the model
fuel nitrogen compounds benzonitrile
and quinoline were added to oils, the
fraction of the model compound con-
verted to HCN was about the same as
was obtained with the model compound
alone.
Oxidative Pyrolysis and
Oxidation
In the oxidative pyrolysis experiments,
the products were measured as a func-
tion of temperature for two model
compounds (pyridine and benzonitrile),
three oils, and one coal. Two-stage
reactors were employed in some of the
oxidative studies to investigate again
the role of secondary reactions.
The model compounds were decom-
posed at a high oxygen concentration
(75 percent) over a range of tempera-
tures. The product distributions were
quite similar in general with large
amounts of HCN, NH3, and CO forming
around 700°C from both pyridine and
benzonitrile. At temperatures above
750°C, the major oxidation products are
C02, N2, N20 and, in the case of
benzonitrile, NO. The formation of large
yields of N2O (up to 45 percent) was
quite unexpected.
Two-stage oxidative pyrolysis experi-
ments established that the N2 and N20
observed at the higher temperatures
formed in the secondary reactions of the
NH3, HCN, and volatile tars formed in
the primary reaction process. The
oxidation of NH3 and HCN alone was
also investigated and N20 was found to
form in large yields from HCN (up to 56
percent) and as a minor product from
NH3 (about 10 percent). The addition of
NH3 to HCN was found to stabilize the
HCN with respect to oxidative decom-
position. .
In the two-stage oxidation of fuel oils,
only small amounts of N20 (on the order
of 5 percent) formed, but the oxygen
concentrations were lower than in the
model compound experiments. Minor
amounts of NzO (~2 percent) even
formed in the oxidative pyrolysis of coal.
In general, the presence of oxygen only
increased slightly the HCN yields from
fuel oil over those obtained under inert
conditions. It was found that about 10
percent of the volatile nitrogen from a
No. 6 fuel oil is in an extremely reactive
form that, in the presence of some
oxygen, can be converted to HCN at
temperatures as low as 460°C. Only
moderate amounts of NO formed in the
oxidative pyrolysis of fuel oil.
Task 2 - Combustion Kinetics
Low-pressure flat-flame burner stud-
ies were conducted to investigate
further the mechanisms and kinetics
involved in the conversion of fuel-
nitrogen intermediates (such as NH3
and HCN) to NOX in methane flames. In
addition, the rates and mechanisms for
the formation of thermal and fuel NO in
the flame front were compared directly,
and the formation of fuel NO from NH3
(which could be present as an impurity)
was investigated in a simulated low-Btu
fuel (CO + H2).
An experimental technique was
developed in which NH3 and HCN could
be measured accurately at low concen-
trations in these 0.1 atm flames by
oxidizing the NH3 and HCN to NO and
measuring the NO in a chemilumines-
cent analyzer. By employing a quartz
probe and various combinations of
selective catalytic converters and a
chemical trap for HCN, each nitrogen
species could be measured at various
distances above the burner. Coated
Pt/Pt-Rh thermocouples, used to obtain
temperature profiles in the methane
flames, were found to be unsatisfactory
in the CO-H2-O2-Ar flames.
The rates and mechanisms of forma-
tion of thermal (prompt) and fuel NO
were compared by studying CH4-02-N2
and CH4-02-Ar-NH3 flames under iden-
tical conditions. Under fuel-rich condi-
tions, both flames generated relatively
high concentrations of HCN and both
the prompt and fuel NO appeared to
form through an HCN intermediate. The
HCN profiles were very similar. Only
about 25 percent as much prompt NO
formed under fuel-lean conditions,
indicating that prompt NO forms via the
reaction of fuel fragments with N2
resulting in the formation of the HCN
intermediate (the Fenimore mechanism).
The CO-H2-02-Ar-NH3 flame, which
was studied in detail only under fuel-
rich conditions, gave some very unex-
pected results. The ammonia was
converted to NO in high yield and then,
just above the flame front, much of the
NO was consumed and a small amount
of the consumed NO reformed slowly far
above the flame front. These results
indicate that some very unusual nitrogen
species form from NH3 in this flame. It
was shown that most of the NO
consumption at the top of the liminous
zone is promoted by nitrogen species.
Very long-lived nitrogen species are
also formed as evidenced by the slow
formation of NO more than 40 mm
above the burner and the observation
that the added NH3 changes the color of
the flame even farther above the burner.
When NO and NH3 were added to the
feed gas simultaneously, the rapid
consumption of both the NO and NH3
occurred below the luminous zone.
Apparently a direct reaction occurs that
is similar to that involved in the DeNO,
process.
Task 3 - Data Analysis
An attempt was made to fit the results
obtained with the CO-H2-02-Ar-NH3
flame to a 59-reaction mechanism
using a kinetic-diffusion model. The
predicted concentration profiles for both
the major species and the nitrogen
species were quite different from those
measured experimentally. This probably
resulted both from the uncertainty in
certain physical parameters, such as
the temperature profile between the
burner and the luminous zone, and from
deficiencies in the reaction rate set.
Interestingly, the kinetic model predicts
that moderate concentrations of 02 are
present above the flame front of this
fuel-rich flame and this was observed
experimentally.
I
-------
A. Axworthy. D. Kahn, V. Dayan, andD. Wooleryare with Rockwell/Rocketdyne
Division, 6633 Canoga Avenue, Canoga Park, CA 91304.
G. Blair Martin is the EPA Project Officer (see below).
The complete report, entitled "Fuel Decomposition and Flame Reactions in
Conversion of Fuel Nitrogen to NO*," (Order No. PB82-108 358; Cost: $21.50,
subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC27711
U.S GOVERNMENT PRINTING OFFICE; 1981 — 599-017/7407
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
V°1H? PROTECTION AGENCY
O IU
I
------- |