EPA-650/2-75-001-G
JANUARY 1975
Environmental Protection Technology Series
^^$^:':^^SS^ $=&
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U. S. Environ-
mental Protection Agency, have been grouped into series. These broad
categories were established to facilitate further development and applica-
tion of environmental technology. Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields. These 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
9. MISCELLANEOUS
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to
develop and demonstrate instrumentation, equipment and methodology
to repair or prevent environmental degradation from point and non-
point sources of pollution. This work provides the new or improved
technology required for the control and treatment of pollution sources
to meet environmental quality standards.
-------
EPA-650/2-75-001-0
ASSESSMENT OF CATALYSTS
FOR CONTROL OF NOX
FROM STATIONARY POWER PLANTS,
PHASE 1 VOLUME I -
FINAL REPORT
by
E.P. Koutsoukos, J.L. Blumenthal,
M. Ghassemi (TRW), and G. Bauerle (UCLA)
TRW Systems Group
One Space Park
Redondo Beach, California 90278
Contract No. 68-02-0648
ROAP No. 21ADF-003
Program Element No. 1AB014
EPA Project Officer: J .B . Wingo
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
January 1975
-------
EPA REVIEW NOTICE
This report has been reviewed by the National Environmental Research
Center - Research Triangle Park, Office of Research and Development,
EPA, and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
This document is available to the public for sale through the National
Technical Information Service, Springfield, Virginia 22161.
11
-------
ABSTRACT
This two volume document summarizes the investigations performed by TRW and
the UCLA School of Engineering and Applied Sciences on the technical and
economic feasibility of catalysts for nitrogen oxide control from power
generating plants. The objective of the program was to assess the potential
of utilizing catalytic processes in power plant nitrogen oxide emission
abatement.
The approach taken to meet the objective involve a literature survey and the
development of a data bank on pertinent articles and patents, experimental
screening tests on selected promising catalysts, and preliminary design and
cost analyses on candidate processes adapted to new and/or existing power
plants.
The stepwise selection and prioritization of catalysts led to the conclu-
sion that at least two types of catalytic nitrogen oxide control processes
should be adaptable to power generating plants. These are: selective re-
duction of nitrogen oxides with ammonia on non-noble metal catalysts and
simultaneous nonselective reduction of nitrogen and sulfur oxides with coal
derived reductants on non-noble metal catalysts.
Volume I of this report presents our assessment and conclusions. Volume II
is comprised of three data bank citation indices.
This report is submitted in fulfillment of Contract No. 68-02-0648 under the
sponsorship of the Environmental Protection Agency. Work was completed in
in May 1974.
-------
IV
-------
CONTENTS
Page
Abstract i.,- j
List of Figures , vl--j
List of Tables ix
Acknowl edgements X1- -,-
Conclusions xjjj
Recommendations xv
Sections
1. Introduction 1
2. Nature of the Problem and Approaches to Its Solution 5
3. Development of a Data Bank on NO Catalysts and Catalytic 15
Processes (Task 1)
3.1 Catalytic NO Decomposition 18
X
3.2 Selective Catalytic Reduction of NO 36
7\
3.3 Nonselective Catalytic Reduction of NO - Simultaneous
NO -SOV Abatement 68
A A
3.4 Catalytic Oxidation of Nitric Oxide 89
3.5 Task 1 Conclusions and Candidate Catalyst Selection 92
3.6 Potential Hazardous Products of Catalytic NO Abatement
Schemes 96
4. Catalyst Screening and Proof-of-Principle Experiments 105
(Task 2)
4.1 Catalysts Selected for Screening 105
4.2 Catalyst Preparation 109
4.3 Catalysts Screening Test Conditions 114
4.4 Catalyst Screening Test Results 120
4.5 Parametric Investigations on Platinum, Vanadia, and 144
Iron-Chromium Oxide Catalysts
4.6 Prioritized Listing of NO Abatement Catalysts Based on 174
Task 2 Investigations
-------
CONTENTS (CONTINUED)
Page
5. Evaluation of the Cost Effectiveness in the Use of Catalysts 177
to Reduce NO Emissions from Power Plants (Task 3)
X
5.1 Preliminary Design and Cost Analysis of Simultaneous 179
NO -SOV Reduction Schemes
A A
5.2 Preliminary Design and Cost Analysis of Selective NO 202
Reduction with Ammonia Schemes
5.3 Summary of Preliminary Design and Cost Analysis Results 208
on Five NO Abatement Schemes Adapted to 800 MW Power
Plants x
01 c
6. Recommendations for Further Action (Task 4) .
APPENDICES 223
Appendix A 225
Appendix B 237
VI
-------
FIGURES
No_. Page
1 Equilibrium Chemical Composition of CH.-Air Flames as a 7
Function of Mixture Ratio
2 Adiabatic Flame Temperature as a Function of Oxidizer-Fuel 7
Mixture Ratio for Methane-Air Flames
3 Decomposition of NO on Pt Catalysts 35
4 Reduction of NO and 02 on Fe^ as a Function of Temperature 41
5 Selective NOX Reduction by NHs on Pt as a Function of 49
Temperature and Space Velocity
6 Selective NOX Reduction by NHs on Pt as a Function of NH?/NO 49
and Space Velocity
7 Reduction of NO with NH3 on Fe-Cr Catalyst at 400°C, 14% CO, 59
5% HeO, 3% 02 Present in N2 Carrier
8 Effect of Temperature on Selective NO Reduction by NH-,; 61
Fe-Cr Oxide Mixture Catalysts
9 Reduction of NO with NH3 on V205 Catalyst at 400°C, 14% C0? 63
5% H20, 3% 02 Present in N2 Carrier
10 Temperature and Oxygen Effect on Selective NO Reduction by 65
Ammonia; Vanadia Catalysts
11 Reduction of NO with NHa on Pt Catalysts, 250°C, Synthetic 67
Flue Gas (14% C02> 5% H20, 3% QZ in N2).
12 COS Production as a Function of S02 Reduction on CuO Catalyst 77
13 Residual Sulfur Species in Flue Gas Reduced on CuO Versus 77
Extent of S02 Reduction
14 TRW Simultaneous Catalytic NO-SO Reduction by Coal Process 79
A A
15 NO-SO Abatement by the Chevron Hitachi Process 85
A A
16 NO-SO Abatement by the TRW Sulfide Process 87
A A
17 Continuous Catalytic NO Absorption Process 91
A
18 TRW "OXNOX" Oxi dative Scrubbing Process 93
19 Catalyst Screening Test Apparatus 115
20 Decomposition of NO on Pt Catalyst (NA-2)
vii
-------
FIGURES (CONTINUED)
No.. Page
21 Reduction of NO with NH~ on Pt Catalyst at 250°C (14% C09> 157
5% H20, 3% 02 in NZ) J *
22 Effect of Og Concentration and Temperature on NO Reduction 159
with NH3 on V205 Catalyst (Harshaw)
23 Reduction of NO With NHs on VzOs Catslyst at 400°C (14% C0?, 165
5% H20, 3% 02 in N2)
24 Reduction of NO with NH- on Fe-Cr Catalysts at 400°C (14% C0?, 173
5% H20, 3% 02 in N2) J
25 NO -SOW Catalytic Reduction Process Adapted to New Power Plants 183
x(800 MW)
26 Monsanto 's Multi -Section Catalytic Reactor
27 NO -SO Catalytic Reduction Process Adapted to Existing Power 195
xPlafits (800 MW)
28 Sulfide Process Adaptation to an 800 MW Power Plant 199
29 NO Reduction with Ammonia Scheme (Non-Noble Metal Catalyst) 205
/\
30 NOV Reduction with Ammonia Scheme-Platinum Catalyst 209
/\
vm
-------
TABLES
No. Page
1 Equilibrium Concentration of Nitric Oxide which can Form 11
From a Starting Mixture of 80 Vol. % N2, 10 Vol. % 03,
and 10 Vol. % He as a Function of Temperature
2. Keywords Used in Catalytic NOV Document Control System 17
A
3 Results of IIRI's Study of Nitric Oxide Decomposition Catalysts-. 21
Commercial Catalysts
4 Results of IIRI's Study of Nitric Oxide Decomposition Catalysts- 23
Laboratory Prepared Catalysts
5 Decomposition of NO at Low Concentrations 27
6 Summary of Task 2 Catalytic NO Decomposition and Oxidation Data 29
7 Decomposition of Nitric Oxide on Various Catalysts - Reaction Order 33
with Respect to NO
8 Relative Effectiveness of Supported Catalysts for the CO-NO and 39
C0-02 Reactions
9 Reduction of NO with H2 on Catalysts PZ-1-168 (Supported Pt, UOP) 42
in the Presence of Oxygen
10 Summary of Task 2 Data on Selective Reduction of NO with CO and H2 45
11 Selective Treatment of Nitric Acid Plant Tail Gas 51
12 Published Data on Selective Reduction of NO with NHg in Simulated 55
or Actual Flue Gas
13 Summary of Task 2 Data on Selective Reduction of NO with NH3 57
14 Simultaneous Reduction of S02 and NOx by Catalyzed Reaction with 71
CO: Copper-on A1203 Catalyst, 538°C (1000°F), Reactants in N2
15 Data on Reduction of S09 and NO by carbon Monoxide on Supported 75
CuO
16 Typical Steady State Non-Optimized Results on TRW's NO -SO 81
Catalytic Reduction Process
17 Catalytic NO -SOY Reduction by CO on NYU Catalysts 83
X A
IX
-------
TABLES (CONTINUED)
No. Page
18 List of Catalysts Subjected to Screening Tests 107
19 Approximate Composition of Rare Earth Oxide Mixtures 111
20 Catalyst Screening Test Conditions 117
21 Results of Tests with Empty Reactor 119
22 Catalyst Screening Test Results (Pt and Pt-Mo Catalysts)
23 Catalyst Screening Test Results (Mo Based Catalysts) I25
24 Catalyst Screening Test Results (Rare-Earth Oxide Based 129
Catalysts)
25 Catalyst Screening Test Results (Copper and Copper-Lead 133
Catalysts)
26 Catalyst Screening Test Results (Tungsten Oxide and Vanadia 135
Catalysts)
27 Catalyst Screening Test Results (Iron-Chromium Oxide Catalysts)
28 Catalyst Screening Test Results (Iron Oxide and Iron Graphite 143
Catalysts)
29 Decomposition of NO on Pt Catalyst (NA-2) 147
30 Parametric Effects on Platinum Catalysts Used in the Selective 151
Reduction of NO by NH3
31 Temperature, Space Velocity and Short-Term S02 Effects on the 161
Reduction of NO with NH3 on Harshaw Vanadia (NA-25)
32 Effect of NO and NH3 Concentration and Space Velocity on the 163
Reduction of NO with NH3 on Harshaw V205 Catalyst (NA-25)
33 Long-Term SOg Effect on the Catalystic Activity of Vanadia 167
for the Selective NO-NH^ Reaction
34 Parametric Investigations on Fe-Cr Oxide Cayalysts Employed 169
in Selective Reduction of NO with NH3
35 Reduction of NO with NH3 on Fe-Cr Catalyst (NA-28) in the 171
Presence of S02
36 Catalyst Ranking Based on Task 2 Data 175
-------
TABLES (CONTINUED)
No. Page
37 Assumed Chemical Composition for the Reductant Coal, Weight 180
Percent
38 Assumed Power Plant Flue Gas Composition (Volume %) 181
39 Effect of Flue Gas Diversion Rate on Coal Feed Rates and 189
Mass Balance Results
40 Breakdown of the Estimated Capital Cost for the NO -SO 193
Catalytic Reduction Process for New Power Plants
41 Operating Cost Breakdown for the NO -SO Catalytic Reduction 193
Process for New Power Plants
42 Breakdown of the Estimated Capital Cost for the NO -SO 197
Catalytic Reduction Process for Existing Power Plants
43 Operating Cost Breakdown for the NO -S0x Catalytic Reduction 197
Process (Existing Power Plants)
44 Breakdown of Capital Cost for the Sulfide NO -SO Reduction 201
Scheme for New Power Plants
45 Operating Cost Breakdown for the Sulfide NO -SO Reduction 201
Scheme for New Power Plants
46 Breakdown of the Esitmated Capital Cost for NO Reduction by 207
Ammonia Scheme - Non-Noble Metal Catalysts
47 Summary of the Breakdown of the Estimated Operating Cost 207
for NO Reduction by Ammonia Scheme - Non-Noble Metal
Catalytts
48 Breakdown of the Estimated Capital Cost for the NO Reduction 211
by Ammonia Scheme - Platinum Catalysts
49 Breakdown of the Estimated Operating Cost for the NOX Reduction 211
by Ammonia Scheme - Platinum Catalysts
50 Capital and Operating Cost Estimates for the Adaptation 213
of Catalytic NO and NO -SO Schemes to an 800 MW Power
Plant - SummaryxTable x x
XI
-------
ACKNOWLEDGEMENTS
The authors wish to acknowledge their indebtness to Professor Ken Nobe
for his valuable inputs to the project and to the long list of academic
and industrial researchers as well as manufacturing firm sales representa-
tives for their eagerness to supply needed information.
Special thanks are due to Messrs. S. C. Wu and R. A. Orsini for their
assistance in the experimental work at UCLA, to Mr. J. Riley who assisted
in the data bank development, and to Ms. M. Ramirez and Ms. M. Jennings
who are principally responsible for the preparation of the manuscript.
Last, but not least, the authors like to express their indebtness to Messrs,
L. Garcia and R. D. Stern, the Environmental Protection Agency's Program
Monitors, for their constant cooperation and valuable comments during the
duration of this program and to Mr. J. Wingo, L. Garcia's replacement, for
his valuable comments on the final report.
XII
-------
CONCLUSIONS
The Phase I investigations on the "Technical and Economic Assessment of
Catalysts for Control <
following conclusions:
Catalysts for Control of NO from Stationary Power Plants" led to the
A
• Nitrogen oxide control from power plants by catalytic
processes appears both technically and economically
feasible on the basis of laboratory-scale data and
preliminary engineering analysis.
Two types of catalytic NO abatement processes exhibited
A
the best potential for power plant adaptation: selective
NO reduction with ammonia and nonselective simultaneous
A
NO -SO reduction with coal derived reductants (CO and
A J\
Hp). A nitric oxide decomposition on platinum process
indicated promise for 50-60 percent NO abatement.
A
Platinum based catalysts and a number of non-noble
metal catalysts were identified as having medium to
high activity in the promotion of the selective NO
A
reduction with ammonia. The platinum on alumina cata-
lysts indicated the highest activity in this process
with an SO^-free flue gas. An immediate and severe
drop in platinum activity occurred as a result of sulfate
deposition when a flue gas containing 1000 ppm SOg was
used. The principle product of nitric oxide reduction
with ammonia on platinum was N20. The active non-noble
metal catalysts for the same reaction did not promote N2°
production (NO was reduced to nitrogen) and their acti-
•A
vity was not affected by the presence of S02- Platinum
was active in the 200-250°C range; non-noble metal cata-
lysts indicated high activity in the 350-450°C range.
xiii
-------
• The catalytic selective reduction of nitrogen oxides
with ammonia appears to be a potential near term solu-
tion to the NO emission problem from existing power
/\
plants. Platinum based catalysts are indicated for
utilization by power plants fired with sulfur free
fuels. The high activity of these catalysts at
relatively low temperatures facilitates the process
adaptation to power plants. Non-noble metal cata-
lysts, especially the iron-chromium oxide and vanadia
catalysts, are the indicated promoters for the
majority of existing power plants because of their
resistance to S02 poisoning.
• - The simultaneous nonselective NO -SO reduction schemes
A /\
on non-noble metal catalysts are indicated as the de-
sired approach to air pollution abatement for new power
plants. Potentially, this approach represents the long
term solution to NO -SOV emissions from existing power
J\ A
plants.
• Catalysts for the selective reduction of nitrogen oxides
by hydrogen or carbon monoxide were not identified.
Additional data is required before selection of the optimum NO abatement
^
scheme for a particular type of power plant can be made.
xiv
-------
RECOMMENDATIONS
As a result of the investigations performed under Phase I of this program,
the TRW-UCLA Team recommends that the Phase I effort be followed by a second
phase program which should concentrate on the! bench-pilot scale development
of the schemes identified in Phase I as promising candidates for NO and
A
NO -SO abatement from power plants. Specifically, the team recommends
A A
the following technical effort:
• Engineering design data generation on the selective
reduction of nitrogen oxides with ammonia on iron-
chromium oxide catalysts. The product of this in-
vestigation should include empirical rate expressions
for the process, as well as pilot or demonstration
plant design curves; it should also include an assess-
ment of the intermediate to long-term stability of
catalytic activity under the conditions of intended
use.
0 Bench-scale development of the reductant generator
to be used with either of two simultaneous NO -SO
A A
reduction processes identified as promising abate-
ment schemes for power plant adaptation. This task
should include engineering design data generation.
• Engineering design data generation on a single stage
catalytic NO -S0¥ reactor. The NYU catalyst is
xv /\
recommended as the prime candidate since proof-of-
principle tests indicated that it does not promote
COS and H2S production. The task should include
design data generation on the integrated reductant
generator-catalytic reactor scheme.
xv
-------
Bench-scale development of the catalytic-regenerative
NO -SO reduction process (the Sulfide Process). The
x\ /\
product of this effort should include complete scheme
definition for new power plant adaptation, assessment
of potential for existing plant adaptation, and ade-
quate data for pilot plant design.
• Assessment of long-term platinum activity toward NO
X
decomposition from SO^ containing flue gases.
• Proof-of-principle investigations on a total pollu-
tant abatement process recommended as a second gen-
eration air pollution abatement process for power
plants. Prime candidate is the TRW "OXNOX" Process
which in principle is capable of removing NO , SO ,
Ai A
and trace elements from power plant flue gases
through an oxidative wet scrubbing scheme.
The bench-scale scheme development tasks should include: (a) complete
definition of reactor emissions and an assessment of their impact on the
environment (even for nonregulated components); (b) proposed scheme im-
pact on current pollution control equipment efficiency; and (c) capital,
energy, and materials requirements and impact on supply if the scheme
were to be universally adapted by power plants.
xvi
-------
1. INTRODUCTION
The role of nitrogen oxides in smog formation and the hazardous effect of
nitrogen dioxide to the respiratory system have been well documented and
are virtually universally accepted. Whether present nitrogen oxide levels
in the U.S. atmosphere as a whole warrant concern has been debated. Very
few experts doubt, however, that if this air pollutant continues unabated
its deleterious effects will be felt by a large portion of the U.S. and
the world population in the future. Recent shortages in high grade (re-
latively clean) fuel availability and its high cost will bring future
hazards nearer. Low grade fuels, with substantial nitrogen constituents,
will soon be called upon to meet the worlds energy requirements. Proposed
methods of combustion modification will have little effect on nitrogen
oxide generation from the combustion of such fuels. Flue gas treatment
appears to be the answer and the catalytic approach to flue gas treatment
looks the most attractive in theory.
TRW in association with the UCLA School of Engineering and Applied Science
were awarded by the Environmental Protection Agency a 15 month, two man-
years program (Contract No. 68-02-0648) to assess the technical and eco-
nomic feasibility of catalysts for NO control from power generating plants.
A
The award represented Phase I of a two phase program aimed at identifying
and testing one or more catalytic NO abatement processes suitable for
A
power plant adaptation. Sufficient test and engineering analysis data were
to be generated to permit pilot scale process design.
The objective of Phase I was to identify candidate catalytic NO abatement
A
processes potentially suitable for power plant adaptation and to rank them
in terms of technical feasibility and cost effectiveness. The approach
taken to meet the objective involved a literature survey and the development
of a data bank on pertinent articles and patents, catalyst screening tests
on selected promising catalysts, and preliminary design and cost analyses on
candidate processes adapted to new and/or existing power plants. Each step
of this approach was aimed at the selection and prioritization of catalysts
1
-------
and catalytic process schemes in terms of their potential utilization in NO
A
abatement from power plants.
The Phase I Program was divided into four tasks:
• Task 1 - Development of a Data Bank on NO.. Catalysts
A
and Catalytic Processes involved the establishment of
a data bank containing the latest state-of-the-art
information on NO catalysts and catalytic processes.
A
Data accumulation included: (a) a comprehensive re-
view of TRW/UCLA's extensive in-house data on NO
A
catalysts and catalytic processes dating back to
1955, (b) literature survey, (c) a review of current
research grants and contracts in the field of NO con-
A
trol, and (d) interviews with officials of catalyst
and carrier material manufacturing firms, of chemical,
petroleum, and automotive industries, of electrical
utilities, and of institutions and companies actively
engaged in air pollution abatement research. The
information and data obtained in the Task 1 effort
was utilized to generate a list of candidate catalysts
that indicated potential for use in the control of NOV
/\
from stationary sources but had not yet been evaluated
under representative flue gas conditions.
• Task 2 - Catalyst Screening and Proof-of-Principle
Experiments involved the experimental evaluation of
candidate catalysts selected in Task 1. These cata-
lysts were evaluated on synthetic power plant flue
gas and under conditions which closely simulated
those of actual flue gases. Nitric oxide decom-
position and/or oxidation catalysts, selective
nitric oxide reduction catalysts, and nonselective
NO -SO reduction catalysts were investigated. The
A A
highest ranked potential NO abatement catalysts were
A
-------
subjected to parametric scans in order to establish
their sensitivity to variations in power plant flue
gas conditions.
Task 3 - Evaluation of the Cost Effectiveness of the
Use of Catalysts to Reduce NO Emissions from Sta-
^
tionary Power Plants. Under this task a first level
of detail process design and cost analysis was per-
formed on NO abatement catalytic processes ranked
/\
highest as potential candidates for power plant
adaptation as a result of Task 1 and Task 2 investi-
gations.
Task 4 - Recommendations for Further Action. This
task summarizes the recommended further action on
catalytic approaches to NO abatement from power
X
plants. A research and development plan is out-
lined which indicates the sequence of tasks to be
undertaken in the development of one or more cata-
lytic NO abatement schemes to the demonstration
X
level of testing. The recommendations are based
on data developed in Tasks 1 through 3 of the
Phase I Program.
The ensuing sections of this two volume report summarize the effort, results,
conclusions, and recommendat'
abatement from power plants.
conclusions, and recommendations of the Phase I Program on catalytic NO
A.
-------
2. NATURE OF THE PROBLEM AND APPROACHES TO ITS SOLUTION
It has been recognized for a number of years that oxides of nitrogen
emanating primarily from high temperature combustion sources are a major
reactant in the formation of photochemical smog. Nitric oxide (NO) which
is formed in combustion processes usually converts in the atmosphere to
the more hazardous nitrogen dioxide (M^). Concentrations of nitrogen
dioxide as low as 10-20 ppm have been shown to cause persistent pathologic
changes in animals. Thus, in addition to its role in the formation of
photochemical smog, nitrogen dioxide, by itself, is a hazardous air pollu-
tant. As our urban areas become larger it therefore becomes increasingly
important to limit the emissions of oxides of nitrogen to the atmosphere.
Fossil fuel fired power plants represent the largest single stationary
source of nitrogen oxide pollution in this country, accounting for approxi-
mately 23% of the total oxides of nitrogen emissions. It is currently
estimated that NOV emissions from U.S. power plants total some 4.5 million
1
tons per year. Up to the present time the great majority of research and
development on oxides of nitrogen pollution abatement (particularly in terms
of catalytic abatement approaches) has been concentrated on the problem of
mobile source emissions (estimated at 8 million tons/year) and relatively
little work has been aimed directly at the distinctly different problem of
controlling NO emissions from power plants. With the pending removal of
A
lead from gasolines, technically viable approaches to controlling NO
A
emissions from mobile sources appear to be at hand (although at current
fuel prices the economic feasibility has been questioned). Analogous
effort for NO abatement in stationary sources has been much less and,
A
therefore, the required technology has not arrived.
Nitric oxide forms in high temperature flame environments through the direct
combination of nitrogen and oxygen. In evaluating technical approaches for
eliminating or removing oxides of nitrogen from fixed combustion sources,
it is necessary to first consider the equilibrium thermochemistry involved.
That is not to say that the chemical kinetics (both homogeneous and heter-
ogeneous) are not of key importance, but equilibrium thermochemistry sets
-------
the limits on what can occur kinetically. Figures 1 and 2 show the adiabatic
flame composition and temperature (at one atmosphere pressure) as a function
of air-fuel ratio for a methane-air flame. From the thermochemical maps for
the methane-air flame which are depicted in Figures 1 and 2 the following
general observations can be drawn.*
• The maximum concentrations of nitric oxide which can form
by nitrogen fixation in the methane-air flame occur at air-
fuel ratios near to where maximum flame temperatures occur.
• The equilibrium chemical composition for fuel rich flames
(air/ CH^ wt. ratio <16) is considerably different from the
flame composition for oxidizer rich flames (air/CH. wt.
ratio >16). In particular, the carbon monoxide and hydro-
gen concentrations in the fuel rich flame are appreciable
while the oxygen concentration is essentially zero. Just
the opposite occurs in oxygen rich flames. The overall
stoichiometry of fixed combustion sources (considering
both the primary and secondary combustion zones) gen-
erally lies on the oxidizer rich side, and hence the com-
bustion flue gases contain, in addition to NO, excess
oxygen (3% typical) and essentially no reductants (CO
and H«). Therefore, in order to use a catalytic reduction
process for removing nitric oxide (e.g., 2NO + 2CO -* N« +
ZCOg), it is necessary to either grossly change the fuel
to air ratio of current power boilers or add to the flue
gas stream a particular reductant (e.g., NHg) which will
react selectively with NO.
• On the oxidizer rich side of Figure 1 where the concentra-
tions of oxygen and nitrogen are essentially constant, the
These basic conclusions relative to nitrogen fixation are generally valid
for coal and liquid hydrocarbon flames as well as for natural gas-air
f1ames.
-------
NITRIC OXIDE
CONCENTRATION
RANGE FOR POWER
PLANT STACK GASES
AIR /CH. WT RATIO
4
Figure 1. Equilibrium Chemical Composition of CH.-Air Flames
as a Function of Mixture Ratio
2300
2100
1900
1700
-
'-
1500
1300
1100
\
10 20 30 40 50 60 ?0 30 90 100
Figure 2. Adiabatic Flame Temperature as a Function of
Oxidizer-Fuel Mixture Ratio for Methane-Air Flames
-------
equilibrium concentration of nitric oxide decreases rapidly
as the adiabatic flame temperature decreases. Correspond-
ingly, the equilibrium (but not necessarily the actual)
concentrations of nitric oxide will decrease as the com-
bustion gases from a fixed source are cooled.
• Over the entire mixture ratio range, the equilibrium con-
centrations of nitrogen dioxide (N02 or N^O^) which can
form in the flame are negligible compared with the equili-
brium nitric oxide concentration. That is not to say, how-
ever, that nitric oxide which is formed in the high tem-
perature flame will not oxidize to nitrogen dioxide either
in the exhaust vent or the atmosphere.
In general, NO production in the high temperature combustion zone is some-
y\
what lower than that predicted from thermochemical calculations (kinetic
limitation due to short residence times). By the same token slow kinetics
are responsible for the observed lack of NO decomposition or oxidation dur-
ing cooling of the flue gases. Thus, the nitrogen oxide constituents of
power plant flue gases at the point of atmospheric discharge equal those
generated in the high temperature zone and they are the same composition
(virtually all nitric oxide); the typical composition of nitrogen oxides
in power plant stack gas is 90-95% NO and 5 to 10% N09. Flue gas NO abate-
£, J\
ment approaches must, therefore, be capable of controlling NO.
In addition to the formation of NO in flame environments by the fixation
/\
of nitrogen (Ng + Og ->• 2NO), organic nitrogen compounds present in coal and
fuel oils can be oxidized to nitric oxide at temperatures well below those
123
required for the direct formation of NO from oxygen and nitrogen. ' '
This source of oxides of nitrogen is particularly important in low tem-
perature combustion processes involving coal or fuel oil (such as fluid-
ized bed combustion of coal) but becomes of lesser importance (with respect
to nitrogen fixation) as the flame temperature of the combustion process
increases.
8
-------
Nitrogen oxide volume concentrations in power plant flue gases range from
0.02 to 0.15% depending on fired fuel and mode of combustion.
Technical approaches to the problem of preventing NO emissions from sta-
A
tionary combustion sources can generally be divided into two major cate-
gories:
1. Approaches which prevent the formation of NO in the flame
A
environment (combustion modifications).
2. Approaches which chemically or physically remove oxides of
nitrogen from the cooled combustion gases prior to venting
to the atmosphere.
The first category of approaches have generally involved three types of
modifications to the combustion system:
• Increasing the fuel-to-air ratio which lowers both the
thermodynamic and kinetic potential for NO formation.
A
• Lowering the peak flame temperature by recirculating
flue gases, staged combustion, or injecting steam or
water into the combustion zone. This type of approach
also lowers both the thermodynamic and kinetic poten-
tial for NO formation.
A
0 Modifying the shape and size of the combustion zone
to minimize the residence time of reactants in the
peak temperature zones. This approach lowers the
kinetic potential for NO formation.
A
Combustion modification approaches to NOV control in stationary power sta-
A
tions are most attractive from an economic point of view. In principle
they have the potential of being the lowest cost approach to NO control.
A
However, their applicability is limited in that they are not capable in
-------
preventing all NOV formation in the combustion zone and they can not influ-
A
ence NO formation from the oxidation of fuel nitrogen. In addition, these
. . X
approaches appear difficult to apply to coal -fired boilers. For natural
gas and low nitrogen oil-fired boilers, combustion modification looks
attractive, especially if complete NO removal is not required.
/\
In the second category of technical approaches (chemical or physical re-
moval of oxides of nitrogen) are included:
• Direct catalytic decomposition of nitric oxide
(NO C&' 1/2 N2 + 1/2 02)
• Selective catalytic reduction of nitric oxide in
the presence of excess oxidants. Proposed selective
reductant systems are:
Ammonia (6NO + 4NH3 m 5N2 + 6H20)
Methane (4NO + CH3 ^at- 2N2 + C02 + 2H20)
Carbon Monoxide (2ND + 2CO C^' N2 + 2C02)
Hydrogen (2NO + 2H2 ****' N£ + 2H20)
Hydrogen Sulfide (2NO + 2H2S C*t' 25 + N2 + 2H20)
The oxygen in the flue gas should not be affected
by nor should it affect the catalyst.
• Nonselective catalytic reduction of nitric oxide in
the presence of excess oxidants (e.g.,
NO + CO + excess 02 ' C02 + Ng)
0 Catalytic oxidation of NO to N02 followed by liquid
scrubbing.
10
-------
• Absorption (scrubbing)
• Adsorption
Nitric oxide formed in the high temperature flame zone of a combustion
process becomes thermodynamically more and more unstable as the combustion
gases cool (Table 1) (however, the kinetic stability of NO increases as the
temperature of the combustion gases decreases). Thus, in principle, in the
presence of the proper catalyst, it should be possible to effect the low
temperature decomposition of the nitric oxide present in the stack gases.
This is a very appealing approach since all that would be required is to
pass the combustion gases through a bed of catalysts prior to entering the
stack and the nitric oxide would decompose to nitrogen and oxygen.
Table 1. EQUILIBRIUM CONCENTRATION OF NITRIC OXTDF WHICH CAN
FORM FROM A STARTING MIXTURE OF 80 VOL._ % N-, 10 VOL.
% 02, AND 10 VOL. % He AS A FUNCTION OF TEMPERATURE
Temperature
°K
1500
1400
1300
1200
1100
1000
900
800
700
600
Pressure
atm
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
Equil ibrium
Concentration
of Nitric Oxide,
ppm
919
548
301
150
66
25
7.3
1.6
0.23
0.017
n
-------
Next to direct catalytic decomposition of nitric oxide, the simplest cata-
lytic approach to nitric oxide removal is selective reduction. In this
approach a reductant is added to the flue gas stream in just sufficient
quantities to react with the nitric oxide and a selective catalyst is chosen
which will promote the nitric oxide reduction but will not promote the re-
duction of oxygen present in the combustion gas stream. The catalyst sys-
tem must be highly selective for NO since oxygen may be present in the com-
bustion gases in concentrations two orders of magnitude higher than the NO.
In principle, this approach is very attractive since the cost of adding
just enough reductant to react with the nitric oxide can be very low due
to the low quantity of NO present.
In the nonselective NO reduction approach sufficient reductant (CH., CO,
X *f
or Hg) is added to the combustion gas stream (or boiler) to completely re-
duce the residual oxygen and S02 (to sulfur) as well as the nitric oxide.
The catalyst system must be active for NO (and SOg) reduction but need not
be selective.
The oxidative NOX scrubbing approach involves catalytic oxidation of NO to
N02 (or NO-N02 mixtures) followed by absorption of the oxidized products in
scrubbing solutions.
As indicated in the introductory section of this report, the objective of
Phase I of this program was to assess the feasibility of adapting catalytic
NO abatement processes to power plants. Fulfillment of the objective
A
required; (a) the thorough review of pertinent literature in order to esta-
blish the state-of-the-art on catalytic NO abatement, (b) the evaluation
of the available data for compatibility to power plant conditions, and (c)
the generation of experimental, design, and economic data where needed.
The next section describes the developed data bank on catalytic NO abate-
ment and summarizes the state-of-the-art on catalytic processes potentially
adaptable to power plants.
12
-------
REFERENCES FOR SECTION 2
1. Bartok, W. et al. Systems Study of Nitrogen Oxide Control Methods
for Stationary Sources. Esso Research and Engineering Company.
Final Report, Vol. 2. Publication Number 6R-2-M, PS-69. November
20, 1969. (0178N)*
2. Shaw, J. T., and A. C. Thomas. Oxides of Nitrogen in Relation to the
Combustion of Coal. Paper presented at Conference on Coal Science.
Praque. 10 June 1968.
3. Smith, W. S. Atmospheric Emissions from Fuel Oil Combustion. Public
Health Service Publication No. 999-AP-2. 1962.
Document Retrieval System Accession Number.
13
-------
3. DEVELOPMENT OF A DATA BANK ON NOY CATALYSTS
/\
AND CATALYTIC PROCESSES (TASK 1)
Technical and economic assessment of the use of catalysts for NO abatement
J\
in stationary power plants requires the assembly of a data bank containing
state-of-the-art information. The data bank, stored at TRW, consists of
literature pertinent to the subject which was derived from: (a) a literature
survey, (b) a review of current research grants and contracts including pri-
vate communications with principal investigators, (c) a review of TRW/UCLA
in-house data on NO catalysis, and (d) interviews with representatives of
A
catalysts and catalyst support manufacturing firms. Information on recent
NO abatement efforts in Japan was obtained from a group of scientists and
/v
utility executives who recently visited the USA to exchange views on NO
1
abatement technology and from the review article by Ando.
The assembly of references from the open literature was concentrated pri-
marily on the period 1969-1974. For the years prior to 1969 the bibliographies
on catalytic NOX abatement in'systems studies and review articles by Bartok
et al.2 and Shelef and Kummer3 (1969) and by NAPCA4 (1970) served as the
primary reference sources. A computerized search of the literature pertain-
ing to NO abatement and related subjects for the period 1920-1970 revealed
A
that the combination of the three cited bibliographies contain an essen-
tially complete listing.
Hand searches of the literature were performed for the period 1969-mid 1973
concentrating primarily on Chemical Abstracts, Air Pollution, Air Pollution
Index, Engineering Index, and Pollution Abstracts. In addition, computerized
searches were made of Chemical Abstracts ("CA-Abstracts"), Engineering Index
("Compendex"), APTIC files, and IFI Plenum patent files. Listings on on-going
NO R&D was partially derived from the Smithsonian Abstracts.
A
The documents (articles, patents, reviews, etc.) selected as pertinent to the
program were procured for review, classified and stored in the data bank file.
Unpublished data from TRW/UCLA research and from private communications with
other groups engaged in NOV abatement R&D were also included in the data bank.
X
15
-------
Upon receipt, the documents were keyworded, assigned accession numbers, and
entered into TRW's computerized document retrieval system. The system enables
document retrieval from the central file by author, keyword, or accession num-
ber. In excess of 250 documents have been entered into this retrieval system.
A complete listing of them by the above three categories comprises Volume II
of this report.
The documents selected for inclusion in the data bank contained information from
the following areas of NO and NO -SO R&D:
/\ A A
t Decomposition and oxidation of NO .
A
• Selective and nonselective catalytic NO reduction with actual or
/\
synthetic flue gases and ICE (internal combustion engine) exhaust.
t Basic studies on catalytic NO decomposition, oxidation, and reduction.
A
t Basic studies on ammonia decomposition and oxidation.
0 Preparation procedures for NO abatement catalysts.
A
• Simultaneous NO -SO abatement
A A
• Economic and engineering analyses on NO -SO control processes pro-
A A
posed for power-plant or related source utilization.
The selected documents were classified and keyworded to reflect both broad
and narrow categories for easy accession to very general and very specific
information. For example, an article describing NO abatement by catalytic
J\
reduction on CuO catalysts was keyworded under "Stationary Source Control"
and under "Copper Based Catalysts". The 28 keywords used in document classi-
fication are listed in Table 2 below.
The information retrieved through the described procedure was thoroughly re-
viewed in order to assess the state-of-the-art on catalytic NO control and to
A
identify and assess catalysts and catalytic processes potentially adaptable to
power plants. The ensuing paragraphs in this section summarize pertinent data
in NO decomposition, selective and nonselective reduction, and oxidation. The
/\
cited bibliography is listed at the end of the chapter in the order of appear-
ance; data bank accession numbers are also given. More detailed information in
these fields can be found in the articles listed in the data bank bibliography,
Volume II of this report.
16
-------
Table 2. KEYWORDS USED IN CATALYTIC NOV DOCUMENT CONTROL SYSTEM
A
Keyword
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24
25.
26.
27.
28.
BASE-METAL-CAT
CATALYT-DECOMP
CATALYT-OXID
CATALYT-REDXN
CU- BASED- CATAL
FLUE-GAS-CONTR
MOBILE-SOURCE
NOBLE-METAL-CAT
OTHER-NOX-CAT
OXID-W/OTHER
OXID-W/02
PATENTS
POWER-PLANTS
RARE- EARTH-CAT
RED-W/GAS-MIXT
RED-W/SULF-COMP
REDXN-W/CO
REDXN-W/FUEL
REDXN-W/HC
REDXN-W/H2
REDXN-W/NH3
REDXN-W/OTHER
REVIEW-ARTICLES
SELECTIVE- CATAL
SIM-NOX-SOX-CON
STATION- SOURCE
TRANS-METAL-CAT
REL-MATL
Refers To
All non-noble metal catalysts
Catalytic decomposition of NO
A
Catalytic oxidation of NOV
X
Catalytic reduction of NOV
A
Copper-based catalysts
Data or information given for actual flue gases
Automotive NOV control
X
Noble metal catalyst
Catalyst other than noble, transition, and heavy
metals and rare earths (e.g., Alkaline earths)
Oxidation with oxidant other than oxygen
Oxidation with oxygen
Patented catalysts or systems
References specific for power plants
Catalysts containing rare earths
Data on information given for synthetic gas
mixtures
Reduction with sulfur compounds
Reduction with carbon monoxide
Reduction with liquid hydrocarbons
Reduction with gaseous hydrocarbons
Reduction with hydrogen
Reduction with ammonia
Any NO catalytic reduction not covered above
A
Articles presenting a view of the field of NO
A
catalysis or containing extensive bibliographies
Selective reduction in the presence of 02
Simultaneous control of NO and SO
X X
Stationary source control
Transition metal catalysts
Additional data related to NO processes
A
17
-------
3.1 CATALYTIC NOY DECOMPOSITION
/\
Nitric oxide abatement from power plant flue gas by decomposition into ele-
mental nitrogen and oxygen (NO -> 1/2 N2 + 1/2 02) is theoretically the sim-
plest, possibly the least expensive, and therefore the most desirable
approach. The obvious advantage of this approach is that flue gas additives
are not required. The predominant nitrogen oxide in power plant flue gas is
NO which is thermodynamically unstable at stack gas temperatures. Unfortu-
nately, its homogeneous decomposition rate is immeasurably low. Thus, the
emphasis on NO decomposition has centered on catalytic approaches.
The attractiveness of this approach has lured a number of investigators
into the study of NO decomposition and their conclusions have been both
pessimistic and optimistic. The disagreement among investigators concern-
ing the feasibility of this NO abatement approach for application to
X
stack and exhaust gases may be largely due to the variety of conditions
under which the individual studies were performed, rendering comparisons
difficult, and to the requirement for severe data extrapolation before
conclusions on approach applicability could be drawn. In fact, data gen-
erated with actual or simulated stack gases are virtually non-existant;
there is some data generated with auto exhaust on a limited number of
catalysts but at too high temperatures and space velocities to be
directly relatable to power generating plants. This lack of data led to
the decision to screen under this program a number of catalysts for NO
decomposition potential; simulated power plant flue gas was used in these
tests.
The data available in the literature as well as that generated under Task
2 suggests to us that the NO decomposition approach should not be dis-
regarded as a potential NO abatement process for power plants, especially
X
if partial NO removal (50 to 60%) could be considered adequate.
X
The ensuing paragraphs summarize the results of a number of important
investigations on NO decomposition; complete list of citations on this
subject appear in the data bank printout.
18
-------
5
• Howard and Daniels report that NO-catalyst mixtures kept
sealed in tubes at ambient temperature over a half cen-
tury revealed no NO decomposition.
• A large number of catalysts (commercial and laboratory
prepared) were screened for NO decomposition potential
by Reisz et al. at IITRI (Illinois Institute of Techno-
logy Research Institute). The tests were conducted with
2000 ppm NO in nitrogen, at 500-700°C (932 to 1292°F),
and at actual space velocities exceeding 100,000 hr
(>30,000 hr STP) with only one exception (V00K). Their
2
data, as tabulated by Bartok et al., are presented in
Tables 3 and 4. The data in these tables indicate
that at high space velocities only the platinum on
asbestos catalyst exhibited a slight NO decomposition
activity (10%). The only catalyst (7% vanadia on alumina)
tested at low space velocity (800 hr STP) promoted
NO decomposition to 20% at 500°C (932°F); however, the
low conversion and low space velocity cast doubt on
the potential of this catalyst.
2
0 Bartok et al present additional NO decomposition
data generated at Southern California Edison's El
Segundo Power Station by Haagen-Smit and at Battelle
Memorial Institute in Columbus, Ohio by Walling; the
data were obtained through private communication.
According to the report, Haagen-Smith investigated
over 20 materials for NO decomposition activity with
particulate-free, actual power plant flue gas in the
temperature range of 66 to 760°C (150-1400°F) and at
space velocities in the range of 500-3500 hr" . The
list included a number of support materials (alumina,
silica, molecular sieves and steel wool) as well as
copper, chromium, vanadium, iron, and platinum in
19
-------
Table 3. RESULTS OF IIRI's STUDY OF NITRIC OXIDE
DECOMPOSITION CATALYSTS - COMMERICAL CATALYSTS
Catalyst
Magnesia-alumina spinel (27-70%)
Silica-alumina (88-12%)
Silica-phosphoric acid (70-30%)
Vanadia-silica potassium sulfate
(10-65-23%)
Vanadia-alumina (7-93%)
Hopcalite
Chromia-alumina (19-81%)
Ferric oxide
Magnetite
Nickel oxide (NiO-Si02)
Cobalt-molybdate
(Co-Mo03-Al203)
Copper oxide-alumina (10-90%)
Zinc oxide (99%)
Mossy zinc
Molybdena-alumina (10-90%)
Palladium-alumina (0.5-99.5%)
Silver oxide-alumina (20-80%)
Tungstic oxide-alumina (10-90%)
Platinum-asbestos (5-95%)
Platinum-alumina (0.5-99.5%)
Platinum-alumina (0.6-99.4%)
Catalyst
Manufacturer
Norton, LMA-520
Universal Oil Products
Universal Oil Products,
No. 2
Davison Chemical , 903
Harshaw Chemical, V-X-
L-533
Mine Safety Appliances
Harshaw Chemical CR-0205
Girdler G3, carbon mono-
xide shift
Tennessee Valley Author-
ity, ammonia synthesis
University Oil Products,
hydro generation
Harshaw Chemical, Co-Mo-
0401 desulfurization
Harshaw Chemical , Cu-0801
Harshaw Chemical, Zn-040'
Baker
Harshaw Chemical, Mo-1202
Baker
Harshaw Chemical, Ag-030'
Harshaw Chemical, W-0101
Baker
Baker
Baker, dehydrogeneration
Decomposition Temperature
500°C (932°F)
sva
1880
1490
1940
1825
38
1450
135
1955
1930b
2000
1820
1840
930
1255
1870
2140
1795
1900
1635
1680
2000
1770
2140
Decomp.
3
4
0
0
20
0
0
0
2
0
7
0
0
3
0
4
7
1
0
0
18
5
0
600°C (1112°F)
SV
2075
2025
1785
2140
1935
1670
1870
2045
1550
1620
2000
2080
1770
Decomp.
4
0
0
0
3
0
5
2
3
0
10
0
0
700°C (1292°F)
SV
1785
2100
1910
1430
2140
2085
1890
Decomp.
0
0
1
0
7
0
0
Volumes of gas per minute per volume of catalyst.
250°C (482°F).
21
-------
Table 4. RESULTS OF IIRI's STUDY OF NITRIC OXIDE DECOMPOSITION CATALYSTS - LABORATORY-PREPARED CATALYSTS
ro
CO
Sodium oxide-alumina
Potassium oxide-alumina
Potassium oxide-alumina
Chromium oxide-alumina
Manganese oxide-alumina
Iron oxide-alumina
Cobalt oxide-alumina
Zinc oxide-alumina
Strontium oxide-alumina
Silver oxide-alumina
Cadmium oxide-alumina
Barium oxide-alumina
Lead oxide-alumina
Oxide
Wt. %
10
5
15
10
10
10
10
10
10
10
10
2.5
5
Alumina0
Wt. %
90
95
85
90
90
90
90
90d
90
90
90
97.5
95
Decomposition Temperature
500°C (932°F)
SV
1835
1840
2250
1835
2045
1690
1860
1955
1910
1665
1740
2140
1940
% NO
Decomp.
1
3
2
0
0
0
0
1
0
0
0
0
0
600°C (1112°F)
SV
1970
1815
2090
1870
1685
2000
1860
2090
1585
1815
2205
1845
1845
% NO
Decomp .
0
0
1
0
0
0
0
1
0
0
0
0
4
700°C
SV
1835
1820
2045
2060
2130
2090
1955
2140
1585
2220
1955
1860
(1292°F)
% NO
Decomp.
1
0
1
0
0
0
0
1
0
0
0
5
Volumes of gas per minute per volume of catalyst.
250°C (482°F)
Harshaw Chemical, Al-0401
Alcoa, XH-151
-------
various forms. "Of the materials tested only copper
oxide showed promise as a decomposition catalyst, pro-
viding a little more than 10% decomposition at test
conditions". Walling tested a number of commercial
catalysts, including platinum, with synthetic flue
gas containing 500 ppm NO and 3% oxygen at about
3400 hr~ space velocity and in the temperature range
of 149-427°C (300-800°F). He observed some activity
with Ni, Co, and Pt catalysts at 316°C (600°F) which
diminished with exposure time; the initial activity
was attributed to sorption.
Sakaida et al, working with approximately 4000 ppm
NO in N2 and using commercial 0.1% Pt-3% Ni catalyst
(A1203 carrier) found conversion levels of NO as high
as 30% at atmospheric pressure and at least 60% at 15
atm during decomposition at 538°C (1000°F). Space
velocity was about 300 hr~ , which is low for stack
gas use particularly with a platinum catalyst.
Q
Sourirajan and Blumenthal studied cobalt and copper
oxide catalysts for decomposition of 300 to 2100 ppm
NO over the temperature range 300-1000°C (572-1832°F),
Copper oxide on silica gel was found to be the most
active catalyst. AT 510°C (950°F) approximately 69%
decomposition was observed.
g
Shelef et al were not able to duplicate the con-
versions reported by Sourirajan and Blumenthal; how-
ever, Shelef used a much more concentrated NO gas
mixture and data comparison may not be valid. In
addition, Shelef's comment that the Sourirajan data
may not have been taken at steady-state conditions
does not appear warranted in view of the over 300
hour activity constancy reported by the Sourirajan-
Blumenthal team.
24
-------
• Table 5 summarizes test conditions and NO conversions
on catalysts suggested in the above citations as having
NO decomposition activity. In fact, this table summari-
zes all the "promising" NO decomposition data generated
with low NO concentration mixtures which was uncovered
during the literature search performed for this program.
There are several claims in the literature, especially
patent literature, of catalytic systems active in NO
decomposition, but insufficient or no data are presented
to support the claims. For example, Stephens claims
that when a "small amount" of NO was injected into a
helium carrier gas and passed over a praseodymia zinc
oxide catalyst nitric oxide was decomposed to nitro-
gen and oxygen. However, actual test data was not
furnished in the patent.
• Table 6 presents the NO decomposition data generated
under Task 2 of this program.* The data was generated
with simulated power plant flue gas (14% CO,,, 3% Op,
5% H20, 1000 ppm NO, and balance N2) but without S02-
The S02 effect was to be investigated only on proven
active decomposition catalysts. Under the conditions
of these screening tests (400°C, 20,000 hr STP space
velocity) only platinum containing catalysts revealed
NO decomposition activity meriting further investigation.
One of the lead-doped copper oxides indicated 13% NO de-
position, all others were below 10%. The two batches of
Pt on alumina (NA-1 and NA-2) showed different activity
even though according to the manufacturer they were of
the same composition; this discrepancy has not been ex-
plained. The data derived from the Pt-Mo catalysts
*
In a few cases the 02 present in the simulated flue gas led to oxidation
of NO to N02; these data are also presented in the table. The decom-
position of NO in such tests was calculated on the basis of the remain-
ing unoxidized NO.
25
-------
Table 5. DECOMPOSITION OF NO AT LOW CONCENTRATIONS
Authors
Reisz et al .
Sakaida et al .
Sourirajan and
Blumenthal
Catalyst
7% V205 on A1203
5% Pt on asbestos
0.5% Pt on A1203
0.1% Pt, 3% Ni on
A1203 (Girdler
6-43)
CuO-Si02 (3056/70%)
Space Velocity
(Hr-1) STP
800
42,000
37,200
37,100
1,200
600
600
300
340
340
1,320
Temperature
500
500
600
500
538
538
427
538
510
380
510
Decomposition
of NO (%)
20
18
10
5
12
20
6
30
72
80
69
NO Concentration
(ppm)
2000
2000
2000
2000
^4000
^000
^4000
•x4000
1290
892
892
Reference
6
7
8
ro
-------
Table 6. SUMMARY OF TASK 2 CATALYTIC NO DECOMPOSITION AND OXIDATION DATA
Catalyst3
NA-1. (0.5% Pt on A1203, Engelhard)
NA-2. (0.5% Pt on A1203> Engelhard)
NA-3. (22.22 Mo, 0.1% Pt on A1203, 1/16 in spheres)
NA-4. (22.2% Mo, 0.1% Pt on A12C2, pressed)
NA-5. (27% Mo on A1203, pressed)
NA-6. (14.7% Mo on A1203, calcined in air)
NA-7. (14.7% Mo on A1203, reduced 20 hrs at 480°C in Hg)
NA-8. (14.7% Mo on A1203, reduced 4 hrs at 700°C in H2)
NA-11. (15% rare earth cobalt oxide on AIJX,, pressed)
NA-12. (NA-10 + 10% Pb, Pb impregnated)
NA-13. (15% 6d2(Mo04)3 on A1203, pressed)
NA-14. (15% Gd Mo03 on A1203> pressed
NA-15. (15% GdV03 on A1203, pressed)
NA-17. (16.5% CuO-Al203)
NA-21. (13.1% W on A1203, 20-30 mesh)
NA-22. (13.1% W on A1203> pressed)
NA-23. (10% W03 on A1203, Harshaw)
NA-24. (10% V205 on A1203, Filtrol)
NA-28. (10% mixture of 83% Fe203, 17% Cr203 on A1203
NA-29. (same formulation as NA-28)
NA-30. (same formulation as NA-28, pressed)
NA-35. (Fe on graphite chips) d
NA-36 (Fe on graphite chips) e
NO Oxidationb
(%) ' .
0
0
0
0
0
0
0
0
0
0
0
0
6.7
0
0
0
0
0
0
0
0
0
0
NO Decomposition
(*)
45.2
14. Oc
27.5
22.8
24.8
2.8
0.7
1.7
5.6
8.4
0.4
0.7
5.9
1.6
11.9
13.0
0.5
6.1
8.9
7.3
7.6
0
2.5
Catalyst impregnated on 1/8 x 1/8 inch Al203 pellets unless specified. Catalyst calcined
in air and not prereduced unless specified.
Feed contained 1000 ppm NO, 14% C0?, 5% H20, 3% 02 (unless specified). Reaction temperature
was 400°C (752°F) and space velocity was 20,000 hr (STP).
0 0.5% 02 present in this test.
Impregnated from organic solution of Fe(N03)3.
e Impregnated from aqueous solution of Fe(N03)3-
29
-------
(NA-3 and NA-4) suggests that the NO decomposition
value generated from catalyst NA-2 may be more repre-
sentative of low Pt content catalysts. Table 6
also shows that lowering the oxygen content of the
flue gas from 3% to 0.5% lowers the NO decomposition
efficiency of the catalyst.
t The effect of oxygen on NO decomposition has been a
somewhat controversial subject among investigators
engaged in this area of work. The opinions range
from negative effect to no effect to positive effect.
Most reported work considers Og as nitric oxide decomposi-
tion inhibitor; the retarding effect has been assigned to
11 12 13
either adsorbed molecular oxygen ' ' or adsorbed
atomic oxvqen.14 However, both Vetter and Wise and
Freeh * found that oxygen had an enhancing effect
on the decomposition of NO in a thermal flow reactor.
The former considered that a chain reaction occurred
involving oxygen atoms reacting with NO in an ini-
tiation step; the latter found that below 1000°K
the decomposition of NO was heterogeneous and uni-
18
molecular. Lawson found that NO will decompose
catalytically in dry air but not in moist air, sug-
gesting that the presence of water vapor and not
excess Op is a limiting factor. In a recent study
specifically designed to investigate the effect of
19
$2 on the decomposition of NO, Amirnazi et al.
observed an inhibiting effect by Q^ using 1-5 to
15% NO and from 0 to 5% 02 on several catalysts in-
cluding supported platinum; they considered the
equilibrium chenrisorption of oxygen on sites re-
quired for the rate-determining process of NO
chemisorption as the inhibiting step. Our data
indicated that oxygen enhances NO decomposition
30
-------
under the conditions of the Task 2 screening experiments
(NA-1, 3 and 0.5% 02 tests). The test conditions may be
the responsible parameter in this controversy.
• Differences of opinion have also surfaced on the assign-
ment of reaction orders with respect to NO. Table 7
illustrates these differences if one assumes the data
from the individual tests can be compared. The investi-
gations represented in Table 7 were performed with pure
NO or with high NO concentration gas mixtures; rate data
derived from them can not be safely extrapolated to
power plant flue gas NO concentrations. The only
exception is the work by Sakaida et al.7 presented
earlier (Table 5).
On the basis of the data presented, the platinum catalyst was selected
as the only one meriting further investigation. Thus, platinum was
scanned under Task 2 for temperature and space velocity effects on its
activity toward NO decomposition. The data derived from batch NA-2 is shown
in Figure 3. Both temperature and space velocity influenced the extent
of NO decomposition. A maximum was obtained during the temperature effect
tests which appears to indicate that further variation of this parameter
will not improve decomposition; lower space velocities, however, may in-
crease NO conversion further provided economics permit them. The 52.5%
NO decomposition at 300°C and 10,000 hr"1 (STP) S.V., if proven valid in
long-term tests,* may be adequate for certain applications even if not
improvable. The effect of S02 was not investigated during these tests,
but if extrapolations from the S02 effect on the ammonia -N0x-Pt system
are valid, S02 should not affect this catalyst at decomposition temperatures
exceeding 300°C (572°F).
*
These data were taken at steady-state conditions at an approximate rate
of one point per hour; all data was reproduced. These tests were run
long enough to preclude the possibility that NO decomposition was
really NO sorption.
31
-------
Table 7. DECOMPOSITION OF NITRIC OXIDE ON VARIOUS CATALYSTS. REACTION ORDER WITH RESPECT TO NO
Authors
Fraser and Daniels
Yur'eva et al.
Shelef et al.
Winter
Backman and Taylor
Zawadski and Perlinsky
Green and Hinshelwood
Sakaida et al.
Amirnazmi et al .
Amii liuzfiii et al .
Harding
Catalyst
Metal oxides
Transition metal oxides
Supported Pt and oxides
Oxides
Pt wire
Pt-Rh wire
Pt wire
Supported Pt-Ni
Oxides
Supported Pt
A1203
Reactor
Flow
Recycl e
Flow
Recycle
Batch
Batch
Batch
Flow
Flow
Flow
Flow
Gas Mixture
10% NO in He
100% NO
4-100% NO in
He
100% NO
100% NO
100% NO
100% NO
0.404 and
0.432% NO
in N£
1.5-15% NO
0-5% 0? in
He c
1.5-15% NO
0-5% 0, in
He i
10-15% NO in
He
Tempera-
ture
range,
°C
740-1040
250-750
279-938
330-870
1210
860-1060
882-1450
427-538
450-1000
450-1000
644-807
Pressure
Torr
760
100-380
760
50-400
201-479
100
200-500
1-15 atm
780-960
780-960
760
Reaction
order with
respect to
NO
0
2
1
2
1
1
2
1
1
2
Reference
14
20
21
22
12
23
11
7
19
19
24
CO
CO
A commercial Pt catalyst also tested in this study resulted in NO order of 0.5.
-------
60
£40
9A
o 20
fL
2
O
o
Ul
o
O-20.000 hf-'S.V.
.-I
- 10,000 hr'S.V
_L
200
250 300 350
TEMPERATURE (°C)
400
Figure 3. Decomposition of NO on Pt Catalysts
35
-------
3.2 SELECTIVE CATALYTIC REDUCTION OF NO
Selective catalytic reduction of NO rates second only to decomposition in
desirability as an NO abatement process. Since NO in power plant flue gas
coexists with from one to two orders of magnitude larger quantities of oxygen
it is highly desirable to selectively reduce NO without affecting or being
X
affected by the oxygen present. A number of investigators have been pursu-
ing this avenue of NO abatement and the results appear promising with some
A
systems (e.g., NH7).
The principal reductants proposed for NO selective reduction are carbon
monoxide, hydrogen, and ammonia. Others include sulfides, various amines
hydrazine and urea. Experimental data in the latter group is virtually non-
existant and none was generated in Task 2. The reasons for treating these
latter reductants as unattractive range from cost (hydrazine, urea) to
toxicity (HpS). In the subsections that follow representative investigations
on selective NO reduction by CO, H~ and NH^ on noble and non-noble metal
catalysts are presented; a more complete citing of selective catalytic NO
reduction investigations can be found in the data bank printout.
3.2.1 Selective NO.. Reduction by Carbon Monoxide and Hydrogen
A review of the pertinent literature as well as the data generated under
Task 2 failed to identify a promising catalyst for the selective reduction
of NO by either CO or H?, especially one that could be used for power
/\ ^
plant NO abatement. Under special conditions (low space velocity, low
^
02 to NO ratio, dry flue gas) iron oxide exhibited some activity for NO
reduction with CO in the presence of 02 at approximately 150°C. Also,
certain noble metal catalysts promoted the selective reduction of NO with
H2 at approximately 300°C provided sulfur and a number of metallic com-
pounds were not present in the flue gas stream. Neither of these cata-
lytic systems can be presently recommended as meriting further study to-
ward utilization in power plans, but one can not rule out a future cata-
lytic scheme involving selective NOX reduction by CO or HZ; the search for
a selective catalyst for these reactions continues.
36
-------
Two approaches have been employed in the search for an active and selective
NO-CO or NO-Hp catalyst. The first involves the identification of reaction
parameters or catalyst constituents that favor the NO reduction reaction
J\
over that of oxygen reduction; the second involves attempts to retard
("poison") the oxygen reduction process so that NO reduction takes over.
A
The ensuing paragraphs outline representative investigations on selective
N0x reduction with CO and H2 (catalytic) and summarize the data from the
screening tests performed under Task 2.
• Early work in this area was performed at the Franklin
p/r
Institute by Taylor who showed that 02 had a detri-
mental effect on the reduction of NO by CO when the
CO concentration was less than that required to re-
duce both oxygen and NO. In fact, he showed that
CO reacted preferentially with 02 on various
chromites and chromite promoted iron.
8 27
t Sourirajan and Blumenthal * confirmed the non-
selectivity of CO, concluded that oxygen was not a
catalyst "poison" but a competing reactant, and
suggested the two-stage catalytic converter approach
to auto-exhaust purification rather than continued
search for selective NO-CO catalysts.
28
0 Shelef, et al. investigated several transition
metal oxides for selective activity toward the NO-
CO reaction in the presence of oxygen; but none of
the oxides indicated selectivity. However, when
the CO-NO and C0-02 reactions were studied sepa-
rately, the NO reduction reaction exhibited a higher
rate than the 02 reduction rate on supported
Fe203 and Cr03 at low temperatures; representative
data is summarized in Table 8. According to these
investigators the average oxidation state of these
catalysts was lower when oxygen was not present;
37
-------
Table 8. RELATIVE EFFECTIVENESS OF SUPPORTED CATALYSTS
FOR THE CO-NO AND C0-02 REACTIONS 28
Supported Catalyst
Fe2°3
Cu20
Cr203
NiO
Pt
Co304
Bare Support
MnO
V2°5
Temperature
CO-NO
145
155
175
220
250
285
350
425
435
560
(°C) of 50% CO Removal
co-o2
180
115
140
265
220
215
115
365
195
405
Flow rate 1400 cc/min, catalyst volume 80 cc.
Inlet gas composition in CO-NO reaction ^1.2% CO; 2% NO.
Inlet gas composition in C0-02 reaction ^1.2% CO; 1.2% 02.
39
-------
it was, therefore, implied that the lower valence state
of the catalysts favored NO reduction. In another study
3
Stielef and Kummer showed that at a very narrow tem-
perature range (near 170°C) and very low space velocities
Fe203 reduced NO to N20 in the presence of oxygen; the
data is shown in Figure 4. Even lower dips in reactor
29
outlet NO concentration were observed by Bauerle et al.
with near stoichiometric oxidant-reductant gas mixtures
but on noble metal catalysts (platinum, rhodium). It
appears questionable whether the indicated "dips" in NO
concentration represent true selectivity or merely paral-
lel reactions at nearly equal rates. In any case, the
above experiments were performed under conditions which
can not be extrapolated to power plant flue gas conditions
(high N0/02 ratios, no water vapor). It is conceivable
that a catalyst exists whose "light off" temperature for
the CO-NO reaction is substantially lower than that of
the CO-Op reaction so that the desired selectivity at
adequate rates would materialize; the data reviewed has
not hinted such a catalyst.
Sorensen and Nobe30 investigated the NO-CO and 02-CO
reactions on lead doped CuO catalysts. They found
that lead retarded the 02-CO reaction while it
appeared to enhance the NO-CO reaction. In the
current program we tested the above reactions
simultaneously on several lead doped CuO catalysts
in an attempt to induce NO-CO selectivity by
"poisoning" the 02-CO reaction; under the test
conditions utilized the attempt failed.
40
-------
a
o
2
*
O
LJ
O
§
O
UJ
I-
o
INLET NO
INLET CO
100 200 300
TEMPERATURE, °C
Figure 4. Reduction of NO and 02 on Fe203 as a Function of Temperature 3
400
-------
The data on selective catalytic NOX reduction with H2
is very similar to that described for CO. Partial
3
selectivity has been reported on certain Pt and Pd
catalysts at low temperatures and with sulfur-free
25
gas mixtures. Jones ,et al. report that "under cer-
tain conditions, many supported noble metal catalysts
are capable of promoting the removal of NO from auto-
mobile exhaust, even in the presence of large amounts
of oxygen The catalysts which exhibit the
selective nitric oxide reduction are very sensitive
to poisoning, particularly by sulfur". The tests were
performed on commercial noble metal catalysts. Table
9 summarizes some of the data which may be considered
pertinent to this program. The data indicates that NO
conversion strongly depends on NO inlet concentration.
Conversion of NO to NH3 was not high, but most of the
NO was reduced to N«0 (not indicated in the table).
Transition metal oxides did not indicate activity toward
the NO-Hp reaction in the presence of oxygen.
Table 9. REDUCTION OF NO WITH HZ ON CATALYST PZ-1-168
(SUPPORTED Pt, UOP) IN THE PRESENCE Of OXYGEN 25
NO Inlet
Cone.
(ppm)
3450
1900
885
500
245
NO
Reacted ,
(*)
82.5
83.5
77.5
64.0
32.6
NH3
Formed ,
(ppm)
95
69
28
19
13
Inlet NO
Converted to
NH3, %
2.75
3.63
3.16
3.80
5.30
Reacted NO
Converted to
NhU, ppm
3.33
4.35
4.08
5.94
16.30
Carrier gas: N2; H2 inlet cone.: 1.43%; 03 inlet cone.: 0.9-1.0%;
Space velocity: 20,000 hr"1; reaction temperature: 200°C (392°F)
42
-------
• Table 10 summarizes the data generated in Task 2 as a
part of the catalyst screening test matrix. The tests
were performed with simulated flue-gas to which 1000 ppm
of CO or H2 were added (the flue gas contained 1000 ppm
NO and 3% 00). The test temperature was 400°C and the
^ _i
space velocity 20,000 hr . The data strongly resembles
that obtained during the NO decomposition tests (Table
6); the implication appears to be that the observed
NO conversions are not due to selective NO reduction by
CO or Hp but to NO decomposition. In view of the Jones
data, which indicated that the optimum temperatures for
these selective reduction reactions were in the 150 to
300°C range, the Task 2 tests were performed at higher
than optimum temperature; it is conceivable that at a
lower temperature some NO reduction with CO or Hp could
have taken place, but it is very improbable that the
extent of reaction would have been such as to alter our
negative conclusion with respect to practical appli-
cation or potential of these selective processes.
There have been several additional studies on NO-CO, NO-hydrocarbon, and
NO-Hp reactions in the presence or absence of oxygen which have been in-
cluded in the data bank citations (Volume II of this report). The majority
of these investigations represent basic studies on the above reactions with
no hint of selectivity; their discussion is beyond the scope of this report
since the data can not be related to potential application for NO abate-
/\
ment from power plants.
3.2.2 Selective NO.. Reduction by Ammonia
*™ """ "•—••- L X
Ammonia is generally considered to be the only relatively inexpensive,
truly selective NO reductant in the presence of oxygen. In fact, many in-
vestigators of the catalytic reduction of NO by NH3 have shown that oxygen
enhances the reduction rate up to a certain temperature which depends on
31
the catalyst used. Markvart and Pour have suggested that NH3 dissociation
on the catalyst surface is the controlling step in the NH3-NO reaction; NO
43
-------
Table 10. SUMMARY OF TASK 2 DATA ON SELECTIVE REDUCTION OF NO WITH CO AND H,
Catalyst3
NA-1. (0.5<$ Pt on A1203, Engelhard)
NA-3. (22.25 Mo, 0.1'. Pt on At203, 1/16 in spheres)
NA-4. (22. 2S Ho, O.H Pt on A1203, pressed)
NA-5. (27' Ho on AljOj, pressed)
NA-6. (14.7' Ho on A1203)
NA-7. (14.7" Mo on A1203, reduced 20 hrs at 480°C
in H2)
NA-11. (15: rare earth cobalt oxide on A1203> pressed^
NA-12. (NA-10 + 10'; Pb)
NA-13. (15? Gd2V03 on AljOj, pressed)
NA-15. (15%- GdV03 on AljOj, pressed)
MA-17. (16.5* CuO-Al2Oj)
NA-18. (With IS Pb NA-17)
NA-19. (With K Pb NA-17)
NA-20. (With 10X Pb NA-17)
NA-21. (13.1* V. on Al^, 20-30 mesh)
NA-24. (102 V205 on A1203, Filtrol)
NA-26. (Girdler G3A, Iron Chromium)
NA-28. (10;! mixture of 833S Fe203. m CrjOj on
A1203)
HA- 29. (Same formulation as NC-28)
NA-30. (Same formulation as NC-28, pressed)
NA-35. (Fe on graphite chips)0
NA-36. (Fe on graphite chips)d
0, in Feedb
2 <;:)
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
3
0.5
Apparent NO Reduction By
CO (1000 ppm)
46.0
22.6
25.0
14.9
30.8
14.3
1.3
0.5
2.6
2.2
1.8
0.8
0
0
13.1
3.9
3.3
6.6
1.5
10.4
0
4
11.0
9.4
0.4
2.5
2.3
3.0
3.8
0
0
1.0
3.8
2.0
4.1
1.4
4.4
1.1
0
0
1.0
1.1
H2 (1000 ppm)
46.6
19.4
26.9
12.3
28.1
18.9
2.6
2.4
1.4
0.5
2.9
0.6
4.9
3.7
8.4
9.9
2.1
2.2
8.0
3.9
8.0
3.9
7.8
5.6
3.6
2.7
2.7
4.8
13.5
22.3
3.4
1.1
4.8
3.0
4.0
2.7
3.0
1.7
5.7
1.9
1.2
18.0
0.7
0
9 Feed also contained 1000 ppm NO, 14£ C02> 5% H20, 1000 ppm HZ in NZ> 400°C. 20,000 hr'1 space velocity.
Catalyst impregnated on AUO, pellets unless specified. Catalyst calcined in air and not prereduced
unless specified.
c Impregnated from organic solution of Fe(NOj)3.
* Impregnated from aqueous solution of Fe(N03)3-
45
-------
reacts with most of these fragments to form Np and/or N^O. When 02 is pre-
sent, ammonia fragment consumption is higher (including fragments that do
not react with NO); thus, the catalyst surface is freed faster and both the
NO reduction rate and extent of reduction increase (equilibrium shift). A
op oo 0
more sophisticated NO-NH3 reaction mechanism is offered by Otto et a-|.oc>OJ>°
in a series of papers involving Cu, Ru, and Pt catalysts; however, the in-
ferences concerning the rate determining step and the oxygen effect appear
to be the same.
The literature survey performed in this program as well as the UCLA investi-
gations (Task 2) strongly suggested that a catalytic process involving the
selective NHg-NO reaction is potentially adaptable to power plants for
efficient NO abatement. The above sources suggested both noble and non-
noble metal catalysts for this purpose. Non-noble metal catalysts are less
active and therefore less efficient than noble metal catalysts, but they are
also much less expensive and possibly more resistant to poisoning (especially
sulfur). Promising catalysts include: Pt, vanadia, Fe-Cr oxide mixtures,
Mo, Cu-Pb and La-Cu-Zr. Most of the data on the catalyst-NH^-NO -Op process
has been obtained on platinum in connection with nitric acid plant tail
gases; data on non-noble metal catalysts was principally derived from patents
and Task 2 investigations. The next few paragraphs present a summary of
representative investigations on selective NO reduction by NH~ derived from
y\ O
the open literature and Task 2 tests.
35
• Newman and Rose studied the oxidation of ammonia on
platinum and concluded that the production of nitrogen
was the result of a secondary reaction between product
nitric oxide and ammonia. They also reported that in
the system Pt-NO-NH3-02 the product N2 below 500°C is
derived from the NO-NH3 reaction while at temperatures
above 500°C from the 02-NH3 reaction. The same obser-
vation was made during later studies (including Task
2), but the temperature where the oxygen takes over
appears to be lower than 500°C; there is also a tem-
perature range (probably between 300-400°C) where
both NO and 02 contribute to N2 production.
46
-------
The majority of work on selective NO reduction by ammonia
J\
on noble metals (Pt, Pd, and Ru) was performed by Andersen
qc
and co-workers, who also investigated cobalt and nickel
They report that Pt is by far the most efficient catalyst
in the above group for selective NO reduction by ammo-
A
nia. With an NH3 to NO ratio of one in a gas containing
10 moles of 02 per mole NH3, catalytic action was suffi-
ciently selective to reduce NO to 10 ppm from approximately
3000 ppm. A summary of the data is presented in Figures
5 and 6. The first set of data indicates that space velocity
has no effect on NO reduction in the temperature range of
J\
200 to 350°C and that maximum conversion occurs at 200 +_ 50°C.
The data in Figure 6 indicates that reactor pressure, water
vapor content (up to 1.6%), and the NH3 to NO ratio (above
one) have no effect on NO reduction. Salt formation (mostly
ammonium nitrate) was observed in the catalyst reactor ef-
fluent at "low temperatures"; neither the quantity nor the
exact composition of these salts were reported.
The Pt catalyst was tested for 3 months without signs
2
of activity deterioration. However, Bartok et al.
state that according to information received from
Andersen the noble metal catalysts are easily poisoned
by sulfur and their use is restricted to gases contain-
ing less than 1 ppm sulfur compounds. NOp present in
the catalytic reactor feed deactivated most of the Pt
catalysts, also (Andersen); a specially prepared Pt
catalyst (MPS 900) alleviated the latter problem. The
improved Engelhard catalyst has been successfully in-
corporated into monolith forms; Table 11 presents data
taken on nitric acid plant tail gas with this catalyst.
38
t Environics, Inc., is currently investigating, at a
pilot plant scale, the selective reduction of NO with
/\
NH3 on Pt under EPA contract. The flue gas is drawn
47
-------
100
a
LU
g
£
80
20
10,000
60
40
90,000 Hr"1 Space Velocity
140 180 220 260 300 340 380
INLET TEMPERATURE °C
420 460 500
Figure 5. Selective NOX Reduction by NH3 on Pt. as a Function
of Temperature and Space Velocity 36
(Catalyst: 0.5% Pt on A1203; Inlet gas: 3% 02, 0.3% NO,
0.3% NHa, 0.8% H20; pressure 100 psig)
3000
1000
300
o 100
-------
Table 11. SELECTIVE TREATMENT OF NITRIC ACID PLANT TAIL GAS37
Catalyst: NCM-S900 (Honeycomb)
Fuel: NH3
Operation Pressure: 100 psig
Tail Gas Composition:
0.3% NO + N02
3.0% 02
0.9% H£0
Balance N
Number
1
2
3
4
5
6
Percent
of Fuel
0.46
0.46
0.37
0.42
0.37
0.46
Percent
Fuel Over
Stoichiometric
50
50
20
40
20
50
Space
Velocity
(Hr-1)
20,000
50,000
100,000
100,000
150,000
150,000
Inlet
Temperature
(°c)
259
255
223
226
239
232
Residua
N02
(ppm)
0
0
0
50
20
20
1 M-OxIdes
NO + 1/2N02
(ppm)
31
82
100
86
171
193
Effluent
NH3
(ppm)
10
10
198
--
308
318
Assuming equal amounts of NO and N02 to be present.
-------
from a 60 MW gas fired power plant unit. The catalyst,
0.3% Pt on a monolith, is mounted on a rotating heat
exchanger unit similar to units used in air preheaters.
The disc rotates between flue gas and inlet air stream.
Flue gas enters the unit at approximately 300°C and
exits at 255°C while the counter flow air is heated to
about 150°C. The rotating catalyst bed is approxi-
mately 0.22 M3 (8 ft3); only half of the catalyst is
in the flue gas stream at any time. At space
velocities up to 50,000 hr~ , inlet NO concentra-
/\
tions averaging 200 ppm, and NhL to NO ratios be-
O X
tween 2 and 4, 90% or better NO conversions have
X
been consistently achieved during six months of
operation with sulfur-free flue gas. A substantial
portion of the reduced NO appeared as N70 (private
f\ £
communication information).
Other investigations involving noble metal and
non noble metal mixtures include the following:
39
Gajewski et al. successfully used a Pt on alumina
catalyst for the selective reduction of NO with NH-
3
at a semicommercial size HNO, plant (30 m /hr tail
AO J
gas). Jones and Weaver " tested successfully Pt and
CuO oxide catalysts on auto exhaust to which ammonia
had been added for the selective reduction of NO;
amines and ammonium salts are suggested as substi-
41
tutes for ammonia. Griffing et al. received a
patent on the use of CuO-Pt and CuO catalysts for
the selective reduction of NO by NH_ in auto ex-
haust; they recommend 316-427°C (600-800°F) with
the CuO-Pd catalyst and 371-649°C (700-1200°F)
with CuO.
52
-------
A number of non-noble metal catalysts have been pro-
posed for use in the selective reduction of NOX with
ammonia in the presence of oxygen, but data on them
are limited. The available data, derived principally
from patent literature, is summarized in Table 12
(this table also includes data on Pt not presented
earlier). Non-noble metal catalysts suggested or
claimed (patents) to be efficient promoters of the
NO -NH, selective reduction reaction include the
A o
following: copper oxide, iron base NH3 decomposition
catalysts, vanadia, manganese dioxide, moly, tungsten
trioxide, iron-chromium oxide mixtures, cobalt oxide,
and zirconium promoted lanthanum cuprate. The
majority of these catalysts have been proposed for
nitric acid tail gas treatment or auto exhaust NO
J\
abatement, but the data suggests that they may also
be candidates for power plant use. Several of the
Table 12 catalysts, or catalysts of similar composi-
tion, were screened under Task 2 of this program
with synthetic power plant flue gas. Alumina
supported vanadia and certain mixtures of iron-
chromium oxides proved to be efficient selective
NOX-NH3 catalysts. CuO and W03 exhibited little
or no activity. Mo, Fe, Cr, and Co based catalysts
varied from low to medium activity. Catalyst com-
position and method of preparation appeared to have
a pronounced effect on catalytic activity with cer-
tain groups, especially the Fe-Cr catalysts.
Table 13 lists the catalysts screened under Task 2
for NH3-NO-02 activity and the percent NO reduction
attained with each of them. The tests were con-
ducted with synthetic flue gas (1000 ppm NO, 3% 02,
14% C02 5% H20, balance N2) to which near stoichio-
metric quantities of NH3 with respect to NO were
53
-------
Table 12. PUBLISHED DATA ON SELECTIVE REDUCTION OF NO WITH
NH3 IN SIMULATED OR ACTUAL FLUE GAS
Author
Gajewskl, et al.
Environics, Inc.
Jones and Weaver
Jones and Weaver
Griff ing, et al. '
Atroshchenko,
et al.
Atroshchenko,
et al.
Atroshchenko,
et al .
Nonnenmacker
and Xartte
Schmidt and
Schulze
Jaros and
Krlnzek
Kudo, et al.
Catalyst
0.2XPt-Al,0,
4 3
Pt-Al 0
0.3XPt-Al?03
5XCuO-Al203
5%Cu-0.5%SiO,-
A1.0,
2 3
Fe-based NH3
synthesis
catalyst
Fe-based NH.
synthesis
catalyst
V2°5
*°2
6.8X VzOs
Corundum
8.9X V205-S10-
6.8X V205-
6.8X V205-
10X MoOj-
10X H03-A1203
85X Fe203
10X Cr203
2X Crt>3
The above +3%
S102. Alkali
metal or alka-
line earths
10X 00304 on
A1203
0.5X Pt on
A1203
Zr-promoted
La Cu03
on Al£03
(Unspecified
concentration)
NH /NO S'MCe
iwyiw velocity Temperature
Inlet (hr-'J (°C)
1.5-2.0 40,000 250
4 45,000 250
0.8-1.0 2 present, no content shown in-
cludes N02; space velocity calcu-
lated assuming a density of unity
for the catalyst
2-4X 02, No C02 present.
3X 0- present.
3% 0, present
£
IS 0., 12X CO.. 12X H.O. 1000 ppm
SO, balanced,
2 £
eference
39
38
40
41
42
43
42
44
45
46
47
55
-------
Table 13. SUMMARY OF TASK 2 DATA ON SELECTIVE REDUCTION
OF NO WITH NH3
Catalyst3
NA-1 (0.5S Pt on A1203)
NA-2 (0.5X Pt on AljOj)
NA-3 (22. U Mo, 0.1% Pt on AljOj, 1/16-in. spheres)
NA-4 (22.2% Mo, 0.1", Pt on A1203, pressed)
NA-5 (27% Mo on AljOj, pressed)
NA-6 (14.7% Mo on A1203)
NA-7 (14.7% Mo on Mfiy reduced 20 hours at 480°C in H2)
NA-8 (14.7% Mo on A1203, reduced 4 hours at 700°C in HZ)
NA-9 (2.7% Co304, 15% MoO on AljOj, Filtrol)
NA-11 (Rare earth cobalt oxide, pressed)
NA-12 (NA-10 with 10% Pb)
NA-13 (15% Gd2(t1o04)3 on A1203> pressed)
NA-14 (15% GdMo03 on Al^, pressed)
NA-15 (15% GdV03 on AljOj, pressed)
NA-17 (16.5% CuO on A1203)
NA-18 (NA-17 with 1% Pb)
NA-1 9 (NA-17 with 5% Pb)
NA-20 (NA-17 with 10% Pb)
NA-22 (13.1%W on AlgOj, pressed)
NA-23 (10% W03 on A1203, Harshaw)
NA-24 (10% V205 on A1203> Filtrol)
NA-25 (10% V20g on A^Oj, Harshaw)
NA-26 (Girdler G3A, Iron-chromium)
NA-27 (Girdler G3A - 2nd sample)
NA-28 (10% mixture of 83% Fe20j, 17% Cr203 on A1203)
NA-29 (Same formulation as NA-28)
NA-30 (Same formulation as NA-28, pressed)
NA-31 (15% Cr203 on ^z°3' Pressed>
NA-32 (15% Fe203 on A1203, pressed)
NA-33 (15% Fe203 on AljOj)
NA-34 (20% Fe203 on AljOj, Harshaw)
NA-35 (Fe on graphite chips)
NA-36 (Fe on graphite chips)d
Reduction of N0b
31.8
9.8
38.0
46.1
46.4
39.7
45.6
45.2
14.6
35.3
4.7
6.0
11.7
61.2
37.2
52.0
53.7
51.9
16.0
3.2
60.3
63.7
17.4
2.7
67.1
61.6
32.2
36.9
45.7
49.2
15.7
12.0
5.8
a The catalysts were prepared by impregnation of 1/8 x 1/8 inch alumina cylinders
and calcined in air unless otherwise specified.
b Simulated feed flue gas composition: 1000 ppm NO, 700-1200 ppm NHj, 14% CO?,
3% 02, 5% H20 in N2; reactor temperature: 400°C, space velocity 20,000 hr~'.
c Impregnated from organic solution of Fe(N03)3
57
-------
added; the reaction temperature was 400°C (752°F) and the
space velocity 20,00 hr"1 (STP). The vanadia based cata-
lysts (NA-15, 24, 25) and the UCLA prepared (impregnated)
Fe-Cr oxide mixture catalysts (NA-28 and 29) promoted the
highest percent NO reduction (>60%). The lead doped CuO
(NA-18, 19, 20) promoted 52-54% NO reduction but unleaded
CuO only 37% (NA-17). Moderate catalytic activity, 35 to
49% NO reduction, was exhibited by Mo-Pt catalysts (NA-3
and 4), supported molybdena (NA-5, 6, 7, and 8), a rare
earth-cobalt oxide mixture (NA-11), supported chromia
(NA-31), and supported iron oxides (NA-32and 33). Approxi-
mately one-half of the catalysts screened showed little or
no activity including the commercially prepared Fe-Cr
oxides and W03. The Pt-Al203 catalysts (NA-1 and 2) were
subjected to the same screening test for reference; as
expected they were not very active at 400°C. High NO
reduction was obtained with platinum in the 200-250°C
(392-482°F) range.
On the basis of the data presented in Tables 12 and 13 several catalysts
could be labeled potential candidates for the selective NO^NH^ reaction
meriting additional experimental evaluation for ultimate use in power
plants. Two of them, iron-chromium oxide mixture and vanadia, appeared to
be the best and they were, thus, selected for a brief parametric investi-
gation which included the following: NH3 to NO ratio, NO concentration,
space velocity, temperature, oxygen and SOp effects. Platinum was also
subjected to the same investigation for reference. These experiments,
including the screening tests, and the data derived from them are described
in detail in Section 4. A brief summary of the results and conclusions from
the parametric investigations is presented below.
Figure 7 depicts data taken on UCLA prepared Fe-Cr oxide catalysts at
five NH3 to NO ratios, three space velocities, three inlet NO concentra-
tions, and two catalyst bed sizes. Substantial effects on NO conversion
were observed with the NH3 to NO ratio and with space velocity; the other
58
-------
too
80
60
8
I
40
O
u
20
€>
Q
O
Inlet No.
PPm
1000
1000
1000
1000
1000
1000
1000
500
250
Space Velocity
Hr"' (STP)
5000
10000
15000
20000
10000
20000
10000
20000
20000
0.6
Q8 1.0
NH3/NO RATIO
1.2
Catalyst Wt.
grams
28
28
28
14
14
14
14
14
14
L4
Figure 7. Reduction of NO with NH3 on Fe-Cr catalyst at 400°C,
14% CO, 5% H20, 3% 02 present in N£ carrier.
59
-------
parameters (inlet NO concentration, catalyst bed size at constant space
velocity) had no discernable effect on NO conversion. The NH3 to NO ratio
effect appears to level off at a ratio value of 1 (the stoichiometric
value is 0.67) and approaches zero at a value of about 1.2. The effect of
space velocity appears to diminish at about 15,000 hr'1. NO reduction in
excess of 90% was reproducibly obtained in the presence of 3% oxygen at
400°C, 10,000 hr"1 (STP), and greater than 1 NH3 to NO ratios. The tem-
perature effect is shown in Figure 8; the optimum reaction temperature
is shown to be 400 +_30°C. These catalysts did not promote N20 production
within the experimental range tested and under certain conditions, they de-
composed all of the excess ammonia used. Thus, a flue gas stream which was
free of NO, HgO and NH3 exited the catalyst bed. The SO effect on NO con-
version was also investigated. These catalysts were subjected for over 50
hours to synthetic flue gas containing in excess of 1000 ppm S02- Selective
NO reduction by NH3 was not affected by the presence of S02 at 400°C; the
S02 was also unaffected.
Figures g and 10 present a summary of the parametric investigations per-
formed with commercial vanadia on alumina catalysts (Harshaw, Filtrol).
The results are very similar to those obtained with the Fe-Cr oxide cata-
lysts. A slightly larger space velocity effect was observed with these
catalysts than with the Fe-Cr oxides and the excess NH3 was not decomposed
on them. They also did not promote N20 production and they were not affected
by the presence of S02 in the flue gas stream. In addition to the para-
meters, the effect of oxygen on selective NO reduction by NH3 was investi-
gated on this catalyst; the results are indicated in Figure 10. Apparently
the oxygen has a positive effect on NO reduction at low temperatures and
oxygen concentrations which diminishes as both these parameters increase.
On the basis of the above data and the preliminary cost analysis presented
in Section 5, the Fe-Cr oxides and vanadia catalysts appear to be definite
candidates for utilization in power plants as the means of NO abatement.
A
It should, however, be noted that this conclusion was based on very pre-
liminary data; substantial experimental work is needed before these cata-
lysts are considered ready for power plant adaptation (especially long-
term data on physical and chemical stability and scale-up performance).
60
-------
100 r-
80
4>
£60
UL
O
g 40
CO
-------
100
80
i 60
fc
40
§
o
20
o
•
3
€
O
O
O
Inlet NO
ppm
1000
1000
1000
1000
IOOO
1000
500
350
250-300
250
Space Velocity
Hr"1 (STP)
5000
10000
20000
15000
5000
10000
10000
5000
IOOOO
20OOO
0.6
OB 1.0
NH /NO RATIO
1.2
Catalyst Wt.
grams
28
28
14
14
14
14
28
28
28
14
1.4
Figure 9. Reduction of NO with NH3 on V205 catalyst at 400°C,
14% C02» 5% H20, 3% 02 present in N2 carrier.
63
-------
100
80
60
o
CO
(K
UJ
> 40
o
o
20
O
a
o
CONCENTRATIONS
(ppm)
NO NH3 02
1000 H90 1000
1000 1190 3000
1000 1190 9100
1000 1190 0
200
300 4OO
TEMPERATURE (°C)
500
Figure 10. Temperature and Oxygen Effect on Selective NO
Reduction by Amnonia; Vanadia Catalysts
65
-------
The parameters investigated on the two catalysts discussed above were also
studied with 0.5% Pt on alumina; pertinent data is given in Figure 11
The effect of NH, to NO ratio was similar to that observed with non-noble
-1
metal catalysts. Space velocity up to 20,000 hr (STP) had no effect on
NO reduction; NO conversion was also not affected by NO concentration in the
range of 250-1000 ppm. The temperature effect (not shown) was similar to
that observed with the non-noble metal catalysts except that the maximum
in NO reduction occurred at approximately 250°C (150 degrees below that of
the non-noble metal catalysts). Substantial quantities of N«0 were pro-
duced with this catalyst and the excess ammonia was not decomposed although
some of it was converted to N20. The difference in behavior with respect to
NpO production and NH3 decomposition or oxidation between Pt and non-noble
metal catalysts may be due to the reaction temperature (250 vs 400°C).
This latter parameter appears to be responsible for the S02 effect observed
with platinum. An immediate loss of activity for NO reduction was observed
with Pt when 1000 to 3000 ppm SOp was added to the synthetic flue gas
stream at 209°C. At 250°C a gradual loss of catalytic activity was observed
over an 18 hour period. In both cases the catalyst was completely regenerated
by passing air through it at 400°C. During the early stages of regeneration
SO, evolution was observed indicating that salt deposition had previously
occurred on the catalyst.
The low temperature and high space velocity render Pt an attractive catalyst
for the selective NO reduction by NH7. Its cost and to a lesser extent the
X «3
N20 production are drawbacks that could possibly be tolerated; its deactivation
by S02, however, excludes its use with flue gases containing S02- It is, of
course, conceivable that if the flue gas contains only a few ppm S02 the Pt
catalyst can remain sufficiently active for several days or even weeks be-
fore it requires regeneration; it is, however, believed that the frequency
of regeneration will be higher than that for power plant maintenance. This
shutdown implies the need for a second catalyst bed, or a part of it. The
latter requirement could render Pt unattractive because of cost.
66
-------
100
80
60
co
G:
UJ
O
o
40
20
O-
D
A
Inlet NO
ppm
1000
500
1000
250
800
Space Velocity
Hr-1 (STP)
20000
20000
10000
20000
2000
I I
0.4
0.8
1.2
1.6
2.0
2.4
NH3/ NO RATIO
Figure 11. Reduction of NO with NH3 on Pt catalysts, 250°C,
Synthetic Flue Gas (14% C02> 5% HgO, 3% 02 in N2)
67
-------
47
Very recently, Hitachi Ltd. of Japan has begun operation of an NO abate-
J\
ment pilot plant utilizing the selective NH3-NO process promoted by a
zirconium doped lanthanum-copper oxide catalysts (Table 12). The pilot plant
o
operates at a power plant flue gas flow rate of about 100 m /min (3500 SCFM)
and at a space velocity of approximately 25,000 hr~ . Preliminary results
indicate that maximum NO reduction occurs at 300°C with S02 free flue gas;
reportedly, higher temperatures are required when S02 is present. Actual
NO conversion data was not available at this writing. From what is known
A
to date, this catalyst appears to be even more promising than the Fe-Cr
oxides and vanadia catalysts (higher space velocity operation).
3.3 NONSELECTIVE CATALYTIC REDUCTION OF NO - SIMULTANEOUS NO -SOV ABATEMENT
/\ J\
Nonselective catalytic NO reduction has been investigated in conjunction
J\
with NO abatement of three sources: automobiles (1C engines), nitric acid
A
plants, and power generating plants (large combustion sources). The large
majority of these investigations relate to the first two sources; however,
the data and conclusions derived from them can be of value to NOX control
in power plant flue gas provided proper extrapolations are made.
Nonselective NOV reduction implies that sufficient reductant is present in
A
the flue or exhaust gas streams to reduce all the oxidant constituents of
the stream, principally 02> S02 and NOX- The reductants are either generated
in the burner or engine by fuel rich combustion or are added to the com-
bustion stream; cost consideration limits the choice of additives to H2, CO,
and hydrocarbons. Reductant generation during combustion is possible with
1C engines and probably with natural gas-and oil-fired burners. Reductant
must be added (separately generated) to coal-fired power plant flue gas
and nitric acid plant tail gas.
Nonselective NO reduction catalysts, especially non-noble metal catalysts,
A
behave differently with reductant rich gas generated at the combustion
source than with gases of the same reductant-oxidant stoichiometry (i.e.
rich) to which the reductant was added later. NO reduction in the former
X
gas streams occurs at lower temperatures and somewhat higher space velocity
than in the latter. The principal reason is the difference in oxygen
68
-------
concentration of the catalytic reactor inlet stream. Rich or near stoi-
chiometric combustion permits only low quantities of oxygen into the com-
bustion gases, part or all of which reduces to C02 prior to reaching the
catalyst. Lean combustion gases and nitric acid plant tail gases contain
substantial quantities of oxygen which normally do not react with the
added reductant prior to reaching the catalyst bed. Thus, care must be
exercised in data extrapolations from automobile exhaust NO control in-
vestigations to power plant application. Also, in a reducing environment
sulfur compounds can be severe catalytic activity poisons and data on NO
reduction derived from sulfur-free gas mixtures should not be extrapolated
to combustion gases from sulfur containing fuels. It is principally for
these reasons that catalysts which proved successful for NO abatement in
/\
1C engine exhausts or nitric acid plant tail gas are not as efficient or
they are inappropriate for power plant use.
Control of NOX by nonselective reduction on noble metal catalysts has been
well developed for nitric acid tail gas. Such schemes are presently in use
and could be adaptable to power plants using sulfur-free fuel. Very low
sulfur fuel may be used if a small excess of oxygen is present in the reactor
(oxygen inhibits platinum poisoning). However, the present cost of sulfur
free or nearly sulfur-free fuel, the scarcity of natural gas and the cost of
noble metal catalysts mitigate against this approach. Selective reduction
by ammonia on non-noble metal or even noble metal catalysts would be a more
preferable approach; non-noble metal catalysts in the selective process
would probably be operated at the same temperature and space velocity values
as Pt or Pd in the nonselective process for the same extent of NO reduction.
A
The use of the nonselective process in nitric acid plants was justified be-
cause of the relatively low volume of gases to be treated, high NO concen-
y\
trations, and clean fuel availability at low cost (in the past). Even for
these sources, however, serious thought is given to conversion to the se-
lective ammonia reduction process. Further discussion of the nonselective
reduction processes as applied to tail gases (Pt, Pd catalysts) is not
warranted. Several descriptions of these schemes and data on them can be
found in the papers cited in the subject index printout under the keywords
"noble metal catalysts" or "nonselective reduction"; a brief discussion of
2
these processes is also given by Bartok et al.
69
-------
Non-noble metal catalysts have been proven active promoters of nonselective
NO reduction, even in the presence of oxygen and sulfur, provided the
A.
reductant-oxidant stoichiometry ratio in terms of equivalents was at 1.
Most of the investigations, which are substantial in number, were performed
in conjunction with NO control in automobile exhaust. No attempt is being
/\
made here to review the individual studies on catalytic NO reduction.
2 3 48
Data summaries on many of them are given by Bartok, Shelef, Yolles,
Caretto, and Perrine and Limin; a comprehensive index of the indivi-
dual studies can be found in the printouts of Volume II of this report.
The aim here is to summarize the types of catalysts claimed to be efficient
in nonselective NO reduction, to outline the conditions under which activity
/\
was evident, and to present in more detail catalysts and schemes that appear
promising for simultaneous NO -SO reduction. This approach is taken from
/\ /\
the conviction that nonselective reduction schemes adapted to power plants
should be capable of substantially removing both NO and SO .
A /\
A host of non-noble metal catalysts have been claimed efficient NO-CO,
NO-Ho, or NO-hydrocarbon reaction promoters at temperatures as low as 200-
L. I
300°C and space velocities in excess of 10,000 hr (STP). Most of these
claims are based on NO-reductant-diluent systems. As the gas mixture con-
stituents approach power plant flue gas composition (COg. ^0, Og and SOg
present) the NO reduction temperature rises, space velocities drop, and
the promising catalyst list becomes shorter. Copper based catalysts, rare
earths, certain transition metals and mixtures of the above head the list
of active catalysts for nonselective NO reduction. Some of them, e.g.,
A
copper based catalysts, promote the simultaneous N0x-S0x reduction to N2
and S , respectively.
The operating conditions of the catalyst (temperature, space velocity) de-
pend on the type and source of reductant, the reductant-oxidant ratio, and
flue gas composition (C02» HpO, S0p» and especially the quantity of 0^ to
be reduced on the catalyst). Investigations in ICE-exhaust control reveal
that a number of non-noble metal catalysts can reduce NO in near stoichio-
/\
metric exhaust at temperatures below 427°C (800°F) and space velocities in
excess of 15,000 hr (STP); higher temperatures were required for complete
70
-------
conversion. Under the present contract (Task 2) over 30 catalysts were
screened for nonselective NO reduction activity with H2 and CO. An S02-
free synthetic flue gas was used at 20,000 hr"1 space velocity and 400°C.
All the non-noble metal catalysts proved inactive; the Pt and Pt-Mo cata-
lysts showed good activity. These tests and the data derived from them
are presented in Section 4 of this report. It is doubtful that lower space
velocities (e.g., 10,000 hr" ) would have proven any of the non-noble metal
catalysts efficient at reducing NOV nonselectively at temperatures at or
A
below 400°C. In fact it does not appear that an efficient non-noble metal
catalyst has been identified for use in nonselective simultaneous NO -SO
reduction at about 10,000 hr"1 (STP) and below 538°C (1000°F). At lower
space velocities non-noble metal catalysts have been reported efficient near
427°C (800°F), which appears to be the upper desired temperature limit for
power plant application.
Ryason and Harkins (Chevron Research) were one of the first teams to in-
vestigate simultaneous NO -SOV reduction with CO on non-noble metal cata-
XX -I
lysts at low space velocities (2,000 to 4,000 hr STP) in the 500°C tem-
perature regime. Typical data obtained with copper oxide on alumina cata-
lysts and with dry synthetic flue gas are given in Table 14 below.
Table 14. SIMULTANEOUS REDUCTION OF S02 AND NOX BY CATALYZED REACTION WITH
CO: COPPER-ON A1203 CATALYST, 538°C (1000°F),REACTANTS IN N2
Reactant
so2,
%
0.47
0.47
0.47
Concentration
CO, C02,
%
1.
1.
0.
%
3 8
3 8
97 5.8
NOX,
ppm
-v.125
VI 25
•\/l60
Residence
time, sec
0.57
0.46
0.57
Percent
Reduction,
so2
98
98
95
Apparent
Percent
Reduction,
N0x
78-100
31-100
100
71
-------
Part of the S02 in these experiments was reduced to COS; the majority was
reduced to elemental sulfur. Using thermodynamic data these investigators
calculated that the maximum elemental sulfur recovery possible in flue gases
similar to the above at 538°C would be 77% of the S0« present and it will
occur at CO concentrations which are stoichiometric with respect to NO and
rt
S02 in the gas mixture. The above investigators attempted to obtain the
same type of data with flue gas from an oil-fired boiler operated on the
fuel rich side of stoichiometry. These attempts failed to generate
quantitative data because of mechanical difficulties; the presence of ele-
mental sulfur, however, indicated that S02 conversion was taking place.
Ryason and Harkins concluded that simultaneous NO -SO reduction by CO on
r\ X
Cu, Ag, and Pd catalysts was technically feasible; they suggested that the
COS production, though not desirable, should not be a problem because of
the low concentrations involved when near stoichiometric quantities of CO
are used. There was no mention of HpS formation. According to Ryasen and
Harkins the required reductant (CO) can be generated at the burner by
slightly rich fuel/air operation. This suggestion appears impractical and
possibly undesirable for power plant adaptation. In general, it is diffi-
cult to operate burners near stoichiometric air-to-fuel ratios with the re-
quired degree of stability to produce CO within a narrow ppm range (exces-
sive generation of CO will cause the production of large quantities of COS,
plus the fact that CO is a pollutant itself); certain fuels, e.g., coal, can
not be burned efficiently without excess air; finally, generation of re-
ductant rich flue gas at burner temperatures may cause partial reduction of
S02 to sulfur species which can then react with boiler tubing to form sul-
fides, thus reducing their life-span. It would appear that reductant
addition or reductant generation in a two stage type combustion is indi-
cated for this type of process adaptation to power plants.
Similar but more extensive catalytic NO -SO reduction studies with CO on
A /\
non-noble metal catalysts have recently been performed at the University of
Massachussetts, Chemical Engineering Department Laboratories (partially
co
EPA sponsored project). Quinlan et al. reported on data obtained with a
commercial alumina supported copper oxide catalyst (Harshaw 0803) and with
72
-------
synthetic flue gases. The data was generated at the 400-425°C (753-796°F)
temperature range and at space velocities of 6,000 to 8,000 hr'l (STP);
larger than stoichiometric quantities of CO were used in all the tests.
Typical results are shown in Table 15.
The data indicates virtually complete NO reduction to nitrogen but only a
maximum of 62% conversion of S02 to elemental sulfur when both pollutants (S0«
and NO) were present. Higher S02 conversions to elemental sulfur were
attained in the absence of NO. Between 15-20% of the reduced SO,, was con-
verted to COS. Attempts to increase S02 reduction by increasing the re-
ductant to oxidant ratio (CO concentration) or by increasing residence
time (lower space velocities) increased the production of nonrecoverable
gaseous sulfur species, principally COS. Figures 12 and 13 illustrate
this point. Figure 12 shows COS production as a function of S02 reduction;
Figure 13 shows the noncollectable sulfur species (other than elemental
sulfur species in the catalytic reactor effluent) as a function of S02
reduction. These investigators report that neither H2S nor NH3 were
detected in the reactor effluent during these investigations.
The same groups have been working with other catalysts most of which are
proprietary. During a telephone communication with Professor Kittrel
(head of the group) we were informed that one of the proprietary cata-
lysts they are working on proved to be much more efficient in the pro-
motion of simultaneous NO -SO reduction than CuO; neither the nature
X /v
of the catalyst nor data generated on it has become available to us. It
is, however, our understanding that even this catalyst promotes COS pro-
duction and possibly FLS.
The Kittrel group also investigated Fe-Cr oxide catalysts one of which
was Girdler G3A. Reportedly, it was one of the more active NOX~SOX re-
duction catalysts. The catalyst screening studies performed during this
program did not indicate so. Copper oxide showed higher activity than
G3A. Neither catalyst was very active under screening test conditions
(400°C, 20,000 hr'1) for nonselective NO reduction with CO and H2. It is
true, however, that test conditions to which Kittrel and UCLA subjected
73
-------
Table 15. DATA FOR REDUCTION OF S02 AND NO BY CARBON MONOXIDE ON SUPPORTED CuO
52
Temperature
°C
425
425
425
410
410
410
410
401
401
401
400
Upstream
CO
6140
6060
5750
6060
6060
5830
5830
6320
6160
6624
6560
Compos i
so2
2190
2250
2190
2250
2250
2210
2210
2160
2255
2200
2200
tion, ppm
NO
293
990
0
995
1005
352
0
0
400
908
0
Contact Time,
Sec
0.178
0.178
0.230
0.228
0.228
0.228
0.228
0.228
0.229
0.229
0.229
Conversion, %
so2
72.0
40.9
92.4
36.9
41.7
63.9
73.1
82.9
62.1
43.1
73.7
NO
97.7
94.5
-
83.2
92.0
97.0
-
-
100.0
96.6
-
% COS
Production
10.5
9.2
11.6
11.6
12.0
11.5
12.2
11.0
11.3
12.6
11.8
-------
CM
O
co
O —1
LU z
O HH
a u.
oo
CO LU
O O
o o;
60
50
40
30
20
10
SYSTEM VARIABLES
CATALYST: HARSHAW Cu 0803
PARTICLE SIZES' 20/30 MESH
AND 1/8 IN
TEMPERATURE' 700-980°F
CONTACT TIME:.2K)-.6!0 SECS
CO RATIO-'.80-15
NO LEVELS'EOO-IOOO ppm
WATER LEVELS:.8-12 °/<
0 10 20 30 40 50 60 70 80 30100
S02 REDUCED, PERCENT OF INLET S02
Figure 12. COS Production as a Function of S02 Reduction on CuO Catalyst
52
90
CD
i-l CM
Z O
i—I CO
LU
CO i-l
a
z: Li-
ii) o
2-
o o
a:
Q£ LU
=> O-
co
80
70
60
50
40
30
20
10
0 20 40 60 80 100
S02 REDUCED, PERCENT OF INLET S02
Figure 13. Residual Sulfur Species in Flue Gas Reduced
on CuO Versus Extent of S02 Reduction 52
77
-------
G3A were substantially different that comparison may be unfair. Actual
data from the Kittrel group on G3A was not made available to us.
In 1970, TRW investigated the Harshaw 0803 catalyst (10% CuO on alumina)
for activity in NO -SO reduction with CO; this is the same catalyst re-
rt ^
ported by Quinlan et al. Under the TRW test conditions (higher temperature
and space velocity) both H«S and COS were produced in the synthetic flue
gas stream to which CO was added in stoichiometric amounts with respect to
oxidants present (02, NO, S02); COS production, however, was substantially
lower than that observed by Quinlan. It is possible that this catalyst
promoted the water-gas shift reaction at the higher temperatures TRW used
(600-700°C), but did not promote it at the temperatures used by the
University of Massachusetts groups (^00°C). Otherwise, the data can be
considered comparable when extrapolated.
TRW selected the above catalyst as a potential candidate for the simultaneous
catalytic NO -SOV process conceptually depicted in Figures 14a and 14b. In
X j\
Figure 14a, which depicts the process as envisioned adapted to new power
plants, a portion of the flue gas (30-40%) at the secondary superheater is
diverted through a coal-fed reductant generator. The reductant rich (CO, H2)
generator effluent returns to the boiler where complete oxygen reduction
occurs homogeneously and the generated heat is absorbed by the boiler. Under
proper conditions of reductant generator operation (size, residence time,
temperature) the boiler flue gas at this point should only contain the quanti-
ty of CO required to reduce the S02 and NOX constituents of the flue gas. The
entire flue gas steam is passed over the selected NOX~SOX reduction catalyst
where NOX is reduced to N2 and S02 to elemental sulfur; then it goes through
the economizer, sulfur collector, air preheater and to the stack.
Even though a preliminary review of the scheme in Figure 14a by Combustion
Engineering Company suggested it to be feasible for existing power plants,
the scheme in Figure I4b is more likely to be the preferred one for existing
power plants. This scheme differs from that of Figure 14a in the following
aspects: the flue gas is treated downstream of the air preheater. Over 50%
of the flue gas is diverted through the CO generator, which is operated at
78
-------
a. NEW PLANTS
EXISTING PLANTS
SULFUR
RECOVERY
f
CATALYST
700°C
S0
COAL
v
CO
GENERATOR
900°C
i
"
37% F.G.
CO + 1/202 •*• C02
PRIMARY
SUPERHEATER
i
SECONDARY
SUPERHEATER
.3% 02,0.2% SO
0.1% NO
BURNER
WASTE HEAT
Rnn FR
ii
750*C
CATALYST 1
BED |
157% F.G. |_
43%
™J,
F.G/
COAL
760°C
CO
GENERATOR
STACK
540°C
BOILER
BURNER
SV = 1,500 HR"1 (STP) IN CO GENERATOR; 9,000 HR"1 (STP) IN CATALYST BED
NOY REDUCTION: COMPLETE
A
SOV REDUCTION:
X
Figure 14. TRW Simultaneous Catalytic NO-SOV Reduction By Coal Process
A A
-------
lower temperatures. Part of the oxygen in the undiverted flue gas stream
would probably convert to C02 on the catalyst. The energy from the coal
fed to the CO generator must be absorbed by either a waste heat boiler or
a gas turbine for the production of additional power.
Proof-of-principle tests on the scheme depicted in Figure 14a were performed
under TRW funding in 1970. Typical data is shown in Table 16. A prelimi-
nary cost analysis on both schemes was performed under Task 3 of this pro-
gram and the results are presented in Section 5. The capital costs for
adopting these schemes to 800 MW plant were estimated at 7.3 and 15 million
for new and existing power plants, respectively; the operating costs were
less than one mil per KWH in both cases. In the above cost estimates it
was assumed that a single-stage catalytic treatment would be sufficient to
meet clean air requirements. This presupposes either that the catalytic
reactions can be optimized to minimize H«S and COS production or a differ-
ent catalyst can be identified which would not promote the production of
these pollutants.
Reportedly a catalyst capable of simultaneously reducing NO -SO by CO with-
J\ i\
out FLS or COS production has been identified by researchers at NYU (New
York University). The catalyst (AL 30873) is proprietary and its composition
was not revealed to us. However, Professors Hnatow and Happel,* principal
investigators on the process, sent us the data shown in Table 17. The data
was generated on two catalysts. One of the catalysts (AL 21773) promoted
the production of H2S but not COS; the second (AL 30873) did not promote
either.
If additional testing, especially long term tests, validates the data in
Table 17, simultaneous catalytic NO -SO reduction could prove to be the
rt A
most desirable abatement process for power plant adaptation.
Presently not with NYU.
80
-------
Table 16.
TYPICAL STEADY STATE NON-OPTIMIZED RESULTS ON TRW'S NO -SO
rt
CATALYTIC REDUCTION PROCESS
00
DATA SE
INPUT
FLUE »
GAS
DATA SE
INPUT
FLUE
GAS
T NO. 1
/ N,, CO, 70%
\ 5% H20
J 3.3% 02 r am u COAL
) 800 ppm NO BED
3300 ppm S0?
T NO. 2
N2, C02 50%
5% H-0
2
3.3% 02 »- 50%,, COAL ,
800 ppm NO BEO
3100 DDtn SO,,
e. «„-...,.-..,, ...i
N?, C0?, H?0, S0?, NO
Oos rn
M 25 X STOIfH )
1 N2. C02, H20, S02, NO
1.2% CO
(1.7 X STOICH.)
_j
q;
i-
o
<
t—
o
<
•MM
INVERSIONS
61% . SULFUR
J6% fcH2S (510 ppm)
---5*., «-rn<; (lnn ppm)
?Q% •. SO, (RfiO nnm)
-90% * IJ2
-10* ^
""3
CONVERSIONS
t ''•f »- SULFUR
*" H2S (720 ppm)
-^ — ^COS (150 ppm)
-^ •"SO, (NONE)
2 l '
k. N
»u nz
-90%
~m ^ fill
^ rn /i-^nn nnm^
CONDITIONS:
CATALYST TEMPERATURE
RESIDENCE TIME
PRESSURE
CATALYST
CATALYST BED SIZE
COAL TEMPERATURE
COAL BED SIZE
FRACTION OF F.G.
THROUGH COAL BED
1280 +30°F
- 0.40 SECONDS AT S.C.
- ATMOSPHERIC
- 10% COPPER ON ALUMINA
- 18.7 CC (CORRESPONDING
TO 110 FT3/10&FT3F.G./
HR)
- 1740°F
- 43 CC(CORRESPONDING
TO 260 FT3/105FT3F.G./HR)
- 30%
CONDITIONS:
SAME AS FOR EXAMPLE NO. 1 EXCEPT
• COAL TEMPERATURE - 1690°F VS 1740°F
COAL BED SIZE
FRACTION OF F.G.
THROUGH COAL BED
- 45 CC (CORRESPONDING TO
260 FT3/106FT3F.G./HR)
VS 43 CC
- 50% VS 30%
-------
Table 17. CATALYTIC NO -SOV REDUCTION BY CO ON NYU CATALYSTS
A A
Catalyst
No.
AL 21773
AL 30873
Feed Composition, Mole %
so2
1.0
1.0
NO
0
1.0
CO
1.9
2.7
H20
3.0
3.0
He
Balance
Balance
Space
Velocity,
HH
5,250a
10,500a
21,000a
10,500a
Temp
°C
520
520
520
520
Length
of Run,
Hours
12
12
12
7
Conversions
% of Feed
so2
95b
95b
95b
85b
NO
-
-
100
COS/SO,
% *
0
0
0
0
H9S/S09
to/ t
h
40
37
30
0
oo
OJ
These are actual space velocities; the STP values are approximately 65% lower.
Maximum possible conversion; CO completely reacted.
-------
A modified version of the simultaneous catalytic NO -SO reduction process
A A
has been recently advanced to the pilot plant scale in Japan by the Hitachi
Ship Building Company. The pilot plant is capable of treating 170 SM3/min
(5900 SCFM) flue gas drawn from the Sakai Refinery, Kansai Oil Company.
Complete NOX reduction to N2 is claimed and the S02 concentration in the
catalytic reactor effluent is reportedly less than 10 ppm. Figure 15 is
a block diagram of the process which was originally developed by Chevron
Research Company.
Flue gas is divided into two portions. The larger portion of the split
stream is heated to 850°C. The hot gas is used to heat manganese-iron
sulfite, formed in the S02 absorption step. Manganese ferrite (regenerated
absorbent) is sent to the S02 absorber (dotted line). The S02 expelled is
used for manufacture of H2S04. The flue gas stream is mixed with the re-
generation bypass stream and passed through a CO generator containing coke.
After dust elimination and S02 absorption, the gas is passed through a
catalytic converter containing cupric oxide catalyst.
The process as shown could be adapted to either existing or new power plants
and it does not have the H2$-COS production problem; therefore, it appears
promising. However, proper process assessment requires additional opera-
tional data which was not available to us as of this writing. A process
drawback may be the high regeneration temperature of the absorbent.
As an extention to its NO -SOV catalytic reduction process TRW has conceived
A J\
a modified NO -SO scheme, labeled the TRW Sulfide Process. The process
J\ A
can be considered as a catalytic-regenerative scheme capable of complete
and simultaneous NO -SO abatement. The process concept is depicted in
/\ /\
Figure 16. In this scheme simultaneous NOX-SOX reduction occurs in the CO gen-
erator; therefore, a second stage reactor (e.g., CuO catalyst) is not needed.
The entire flue gas is diverted at the secondary superheater to the coal bed
(CO generator) where the SOp of the flue gas is reduced to elemental sul-
fur, H2S, and possibly COS. The reduced sulfur species react with iron,
fed to the CO generator with the coal, to form sulfides. The latter are
removed with the ash for possible regeneration. Simultaneously, the NO is
A
84
-------
03
cn
P
SO 2
CONVERTER
SCRUBBER
CONC. H2S04
FLUE
GAS ^-J
FUEL
-*• HEATER
t '
AIR
CLEANED
i
GA
->•
1
i
REGENERATOR
j
L ^
b..
^
I
CO
GENERATOR
HEAT
RECOVERY UNIT
i
1
|
r |^ DUST
** COLLECTOR
•
NOX
REDUCER
so2
^ ABSORBER
!
Figure 15. N0 ~so Abatement by the Chevron Hitachi Process 1
-------
ORE (Fe) + SO2 —
MSn (FeSn)
NO — N2 AND NH-
COAL + O2-CO
f* r^ A 1 _L U f~\
LvJAL + ruU -»
CO + H2
AIR
TO PREHEATER
COAL
+ ORE (MO
FLL
COAL
+ ORE
OR
SCRAP Fe
IOOO°C
ASH
1
i
KJ
E G
1
* /*
AS
i
J
\AS
AIR!
Y/~\ 1
TO 1
CO + O9 — CO0
NH3 + °2 ~* N2
+ H2O
H2 + 02-H20
N/-/-V /-S
2 / ^^2 ' U2
SO2 , NO , H2O
BURNER
"O BOILER
•»e"^*v//~i c: /A j* C AJI/^\ \
-------
reduced to N2 (it has not been established if any of these reactions are
catalyzed by the ash or iron present in the CO generator). The CO genera-
tor effluent gas, rich in reductants (CO and H2) and free of 02, N0x, and
sulfur compounds, is returned to the boiler at approximately the same tem-
perature it left (the CO generator operates nearly isothermally because of
competing endothermic, e.g., C02 reduction, and exothermic, e.g., oxygen
and water reduction, reactions with coal). Preheated air is added to the
boiler for the complete oxidation of the reductants to C02 and water; the
generated heat is absorbed by the boiler in the normal manner.
Limited proof-of-principie data with synthetic flue gas (14% C02> 5% H20,
3.3% 02, 3000 ppm S02> 800 ppm NO) revealed total S02 and NO removal with
this process. Also the oxidation of the reductants in a simulated boiler
was complete. The experiments were performed at 920 +. 20°C (^1700°F) and
1,500 hr" space velocity. The coal bed size was approximately 50 cc; the
coal was mixed with 2.5% of its weight iron in the form of wire. Iron con-
sumption was approximately equal to 15 wt. % of the coal consumption in the
CO generator indicating FeS formation rather than FeS,
'2*
The process in Figure 16 was costed on the basis of the scan proof-of-
principle data described above (two tests were performed). The estimated
capital cost for process adaptation to a new 800 MW power plant was esti-
mated at $5 million; the operating cost was estimated at 2 mils per KWH.
Details on this cost analysis are presented in Section 5 of this report.
A large fraction of the operating costs resulted from the iron consumed in
the process ($100 per ton) and not regenerated. It is believed that less
expensive reactants than iron can be used in the CO generator (e.g., iron
oxide, dolomite, or even coal ash) as sulfur getters. It is also believed
that the majority of these sulfur-getters can be regenerated, if desirable.
The Sulfide Process as depicted in Figure 16 is adaptable to new power plants
only since the required boiler changes for adaptation to existing plants
will be substantial. However, a scheme similar to that presented in
Figure 14b may render this process adaptable to existing power plants.
-------
3.4 CATALYTIC OXIDATION OF NITRIC OXIDE
Oxidation of NO to N02 is desirable because N02 can be easily removed from
power plant flue gas in a variety of wet-scrubbing processes as nitrites,
nitrates, or complex salts. Equimolar mixtures of N0-N00 are also re-
2
movable by the above processes; thus, complete NO to N02 conversion is
not required.
The nitrogen oxides in power plant flue gases are virtually all NO so that some
degree of oxidation is required before the wet-scrubbing processes can be
utilized for NOX abatement. Homogeneous oxidation of NO at the flue gas
or wet-scrubber conditions is impractically slow. Two catalytic approaches
have been tried in an effort to improve NO scrubbing efficiency. One
A
involves gas phase NO oxidation, the other liquid phase oxidation. Neither
has proven successful to date, at least not for power plant flue gas treat-
ment. Some promise, however, has been indicated by at least two processes
involving liquid phase catalytic oxidation. The available data, or at
least the data available to us, are not sufficient for firm process assess-
ment; thus, the labeling of these processes as promising is only tentative.
One of these processes is the Continuous Catalytic Absorption Process de-
picted in Figure 17 as applied to nitric acid plant tail gas. Mayland
53
and Heinze describe the process as a catalytic oxidative absorption
operation. The nitric acid tail gas is passed countercurrently with dilute
nitric acid over a catalytic packing which promotes the oxidation of NO to
nitric acid; the acid is recovered. According to the authors this process
is more efficient and more cost-effective that the nonselective gas phase
catalytic processes presently used (Section 3.3}
Attempts to obtain information concerning the catalyst and the process
operating conditions failed to yield results. According to the authors
(telephone communication) additional information on the process will be
released upon completion of pilot plant and full scale plant development
presently underway. A pilot plant unit is scheduled to operate on an
ordnance nitration plant tail gas. Before a definite conclusion can be
drawn concerning potential applicability of this process to power plant
89
-------
CATALYTIC
ABSORBER
TAIL GAg
H20
STEA
^RECOVERED
ACID
COOLING
WATER
PURIFIED
STRIPPER
TAIL GAS
Figure 17. Continuous Catalytic NO Absorption Process
53
-------
flue gas it is necessary to know at least the space velocity required for
efficient process operation and if N0? must be present in the tail gas prior
to processing.
A second process with potential NO scrubbing capabilities is the TRW
s\
"OXNOX" Process. This is also a catalytic oxidative scrubbing process
which utilizes hypochlorides (or chlorine) to convert NO to nitrates. Key
to the process efficiency and cost effectiveness is the presence of N02 in
the flue gas prior to entering the scrubber. This constituent is the pro-
cess catalyst which promotes oxidation in the scrubber. Proof-of-principle
experiments indicated that it must represent at least 5% of the NOV pre-
X
sent in the flue gas prior to entering the scrubber. Virtually complete
NO scrubbing was attained under practical operating conditions when 10
n
ppm N02 was present in a synthetic flue gas containing 420 ppm NO.
Figure 18 depicts the envisioned power plant adaptation scheme for the
TRW "OXNOX" Process. The catalyst for N0-S02 oxidation step in the
diverted flue gas stream has not been identified as yet. Neither the
literature survey nor the screening tests performed under this program
(Task 2) revealed an efficient catalyst for NO oxidation. However, cer-
p
tain charcoals and rare earth-vanadium oxide mixtures (Task 2) may be
sufficiently active for the "OXNOX" Process because of the small fraction
of the flue gas requiring catalytic treatment in the gas phase (very low
space velocities become practical).
It should be noted that the "OXNOX" Process is being proposed as a "total
pollutant abatement process" (NOX> S02, Hg, Sb, PNA, etc.).
3.5 TASK 1 CONCLUSIONS AND CANDIDATE CATALYST SELECTION
The information reviewed in Task 1 led to the conclusions and candidate
catalyst selections presented below:
92
-------
VD
5% OF
GAS STREAM
TOTAL GAS
STREAM
CATALYTIC
OXIDATION OF
NO TO NO*
ANDSOgTOSq
95% OF GAS STREAM
OXIDATIVE
SCRUBBER
Figure 18. TRW "0XNOX" Oxidative Scrubbing Process
-------
• NO Decomposition. The available data in the literature
indicates limited success in the identification of an
efficient catalyst for NO decomposition. Platinum,
A
copper oxide, and vanadia have shown some potential
for NO decomposition at low space velocities, but
they have not been tested on power plant flue gas.
These three catalysts were selected as prime candi-
dates for screen-testing with simulated power plant
flue gas in Task 2.
t Selective NO.. Reduction. No catalysts were identi-
^_^^H«^B«_H_MB_^BM^^^^^ ^_^K_
fied as efficient promoters of the selective reduc-
tion of NO with either H« or CO. Several catalysts
have been shown or claimed to be effective promoters
of the selective reduction of NO with NH3; however,
data on them generated with actual or simulated
power plant flue gas are virtually nonexistent. The
following catalysts were selected for screening
selective NO reduction potential under Task 2: Pt,
Pt-Mo, Mo, rare-earth oxides, W03, VgOg, Fe and Cr
oxides, and Fe on graphite.
• Nonselective NO Reduction. Noble metals appear to
be the most effective nonselective NO reduction cata-
X
lysts, but only in sulfur-free flue gases. Non-noble
metal catalysts (e.g., CuO, Fe and Cu chromates) re-
quire at least 500°C temperature for effective
activity; however, these catalysts can be efficient
in the presence of sulfur containing flue gas. Hydro-
gen, carbon monoxide, and hydrocarbons are suggested
as reductants (gaseous fuels). The nonselective re-
duction approach would be desirable for power plant
adaptation only as a simultaneous NO -SO abatement
X X
process (reductant and adaptation costs would be
unjustifiably high to be used as an add-on to a
94
-------
desulfurization process or to a power plant using
sulfur free fuel).
Available data on simultaneous NO -SO catalysts are
A A
adequate for preliminary engineering analysis, pro-
vided the assumption is made that a second reactor
is not required for abatement of product HgS or COS
(gaseous S02 reduction products formed on most non-
selective reduction catalysts).
A representative simultaneous NO -SO catalytic re-
A A
duction process was selected for preliminary design
and cost analysis. Alsos CuO and a number of transi-
tion metal based catalysts were selected for screen-
ing with H2 and CO containing synthetic power plant
flue gas with the objective of identifying a non-
selective NO reduction catalyst effective at tern-
A
peratures below 400°C.
NO Oxidation. It was not possible to identify an
effective NO oxidation catalyst during the literature
review. Certain types of charcoal were suggested as
promising NO oxidation catalysts at low space veloc-
ities. Such catalysts would be inappropriate for
utilization in high gas volume NO sources, e.g.,
A
power plants. The required large bed volumes and
the potential for high attrition render these cata-
lysts undesirable. Rare-earth oxides, vanadia,
and tungsten oxide have also been suggested as
possible NO oxidation catalysts, but the available
data on them was insufficient for assessment;
they were selected for screening under Task 3.
95
-------
3.6 POTENTIAL HAZARDOUS PRODUCTS OF CATALYTIC NOV ABATEMENT SCHEMES
J\
Experience has shown that catalytic activity and catalyst integrity depend
greatly on the chemical nature of the catalyst, the method of its prepa-
ration, the chemical composition of the environment which it is being
subjected, and the reaction parameters. In the air pollution control area,
loss in activity can cause excessive emissions in the pollutant being con-
trolled as well as emissions of any air polluting additives injected into
the stream for the catalytic conversion of the original pollutant. Loss
of catalytic integrity (physical or chemical) can result in particulate
emissions or in emissions of hazardous catalyst-flue gas reaction products.
In addition, a catalyst may promote side reactions under certain operational
conditions whose products may be detrimental to the environment.
Catalysts proposed for NO abatement may not be immune to the problems
^\
described above. As indicated earlier, concern has been raised over side
reactions producing HpS and COS. Production of metalic carbonyls by the
reaction of CO with certain catalysts (monel, nickel, iron, manganese,
and even noble metals) have been mentioned as concerns. Particulate
generation due to catalyst errosion or attrition is always a concern.
Sulfur poisoning of certain catalysts is a potential problem.
The validity of the above concerns can not be assessed unless detailed
data are available on the particular catalytic scheme proposed for power
plant adaptation. In the ensuing paragraphs an attempt is made to in-
dicate potential problems that may be encountered with each of the pro-
posed principal methods of NO abatement (decomposition, reduction, oxida-
J\
tion), to assess probabilities of problem occurance, and, when possible,
to present potential remedies. This analysis is based on extrapolated and
in most cases incomplete data and it should be considered only tentative.
• NO Decomposition Schemes. The biggest concern with this
-A
scheme is low catalyst efficiency under practical tempera-
ture and space velocity conditions. Platinum appears to be
96
-------
the most effective catalyst. Platinum is subject to sulfur
poisoning in reducing atmospheres, but there has not been
any indication that it is poisoned by SCL in an oxidizing
atmosphere at NOX decomposition conditions. Decomposition
catalysts could, in principle, promote NO oxidation to N02
or N204 if the reactor is operated at low-to-moderate tem-
peratures (<400°C); however, platinum does not appear to
promote N02 formation at NO decomposition temperatures and
we know of no other catalyst which is effective in NO de-
composition at temperatures where N02 is stable. Should
N02 be formed, it can easily be scrubbed in an alkali
scrubber.
Selective NO reduction with NHL (other reductants are not
considered here because of low effectiveness). The primary
concern with this group of catalysts is long term effective-
ness (emissions of'residual NOV and NHL). Platinum may pre-
X o
sent a problem with SO containing flue gas. This catalyst
yx
appears to be "poisoned" by ammonium sulfate decomposition;
however, it can easily be regenerated by hot gas (approx.
300°C gas). Its use with sulfur containing flue gas depends
on economics and the latter depend on the concentration of
S02 in the flue gas. Sulfur dioxide does not appear to be
a problem with the non-noble metal catalysts proven effec-
tive in selective NOV reduction; however, these catalysts
X
are efficient at 400°C while platinum has optimum activity
at 200°C. Platinum promotes N20 formation, but this nitro-
gen oxide is not considered a pollutant at present. Forma-
tion of NO or N204 does not appear to be promoted by any
of the catalysts proposed for this scheme. Formation of
HpOp is a question mark, although its formation is highly
unlikely at 400°C.
Nonselective NO.. Reduction. The most likely reductant to be
— /v "-•----
used in this scheme is CO or a combination of CO and H2 (coal
derived reductants). This scheme could present severe side
reaction problems. A number of catalysts proposed for this
97
-------
scheme promote the production of COS or H2S or both. These
gases are highly toxic and difficult to abate at the con-
centrations produced. This, of course, is a potential
problem only with S02 containing flue gases; its magnitude
depends on the SOg concentration in the flue gas, the parti-
cular catalyst used, and the reactor operating conditions.
Even though the H2S/COS problem has surfaced with the majority
of the catalysts tested for simultaneous NO /SO reduction,
NYU has reported preliminary data on an effective SO /NO
reduction catalyst which may not promote COS or HpS formation
(see Section 3.3 of this report).
Metal carbonyls have also been mentioned as a concern in
schemes involving NOX reduction by CO. Metal carbonyls may
be formed from the reaction of CO with the active metal of
the catalyst, with the walls of the reactor, or with trace
metals in the flue gas; the biggest concern is reaction
with the catalyst. Nickel, iron, manganese, and possibly
noble metal carbonyls could potentially form under certain
flue gas environments. Nickel carbonyl is the most toxic
of the metal carbonyls (0.3 ppb tolerance level in ambient
air), but it does not form above 200°C and no nickel con-
taining catalyst has been suggested as an effective NO
^\
reducing catalyst at such low temperature. The concern
over nickel carbonyl arose from the proposed application
of monel catalysts for NO control in auto-exhaust. It
J\
was feared that nickel carbonyl could form during engine
start-up (reaction of the CO produced during engine start-
up with the catalyst in the afterburner). There is no
analogous cituation in a power plant; but even if it were,
54
tests at ESSO proved the concern unfounded. The other
three carbonyls are a question mark. They are not as toxic
as nickel carbonyl and to our knowledge a tolerence level
has not been established for them. In addition, it is not
98
-------
certain that they will form at all or that they would be
stable in the atmosphere; however, they should not be dis-
missed without further investigation.
Incomplete elemental sulfur collection and unreacted CO
emissions are additional potential problems of this scheme.
Elemental sulfur collection efficiency can be improved by
the use of electrostatic precipitators and CO can be
catalytically oxidized in a second stage reactor, if neces-
sary. However, both these units will increase the cost of
the process.
• Oxidation of NO to NOo. Hazardous by-products of this scheme
£.
have not been identified principally because an effective
catalyst for this scheme has not been found.
In addition to the specific scheme and specific catalyst pollutant genera-
tion potential there is also the particulate generation problem which
could arise from any of the catalysts proposed. Physical and/or chemical
degradation of the catalyst could result in unacceptable levels of fine
particulates or even mists. However, it should be noted that if the
problem is not severe enough to affect catalytic activity, particulate
collection can easily be alleviated by electrostatic precipitators.
The candidate catalyst selections for NO abatement from power plants pre-
•^
sented in the previous section of this report were based on data indicating
that these catalysts would probably promote the conversion of NO to non-
•^
polluting species under practical operating conditions. The available
data were insufficient for assessment of catalyst efficiency, catalyst
stability, and extent of possible side reactions. This type of information
must be generated during bench scale and pilot plant testing of the pro-
mising catalyst selected for use in a specific NO abatement scheme.
r\
During the catalyst screening studies described in the next section, very
preliminary data were generated on catalyst efficiency and stability so
as to establish potential; detailed assessment of performance has been
scoped for Phase II of this program.
99
-------
REFERENCES FCR SECTION 3
!• Ando, J., and H. Tohata. NOX Abatement Technology in Japan. Faculty of
Science and Engineering Report. Chuo Univ. Kasuga, Bunkyo-Ky Tokyo.
May 1973. (0186N)*
2. Bartok, W. et al. Systems Study of Nitrogen Oxide Control Methods
for Stationary Sources. Esso Research and Engineering Company.
Final Report, Vol. 2. Publication Number 6R-2-M, PS-69. November
20, 1969. (0178N)
3. Shelef, M., and J. T. Kummer. The Behavior of Nitric Oxide in
Heterogeneous Catalytic Reactions. Chemical Engineering Progress
Symposium. Series Vol. 67. Publication Number 115. 1971. p. 74-92.
(0134N)
4. Nitrogen Oxides: An Annotated Bibliography. National Air Pollution
Control Administration. U.S. Department of Health, Education, and
Welfare. Raleigh, N.C. Publication Number AP-72. August 1970.
(0231N)
5. Howard, C.S., and G. Daniels. Stability of Nitric Oxide Over a
Long Time Interval. Journal of Physical Chemistry. 62(3):360-1,
1958. (0232N)
6. Riesz, C. H., F. L. Morritz, and K. D. Franson. Catalytic Decom-
position of Nitric Oxide. Air Pollution Foundation. Publication
Number 20. 1957. 22 p. (0240N)
7. Sakaida, R.R., R. 6. Rinder, U. L. Wang, and W. J.Corcoran. Cata-
lytic Decomposition of Nitric Oxide. Journal of Amer. Inst. of
Chemical Engineering. Vol. 7. Publication Number 4. December
1961. p. 658-663. (0081N)
8. Sourirajan, S., and J. L. Blumenthal. Catalytic Decomposition of
Nitric Oxide Present in Low Concentrations. Proc. 2nd Int. Congress
Catalysis. Paris. Publication Number 131. 1960. (0024N)
9. Shelef, M., K. Otto, and H. Gardhi. The Heterogeneous Decomposition
of Nitric Oxide on Supported Catalysts. Atmospheric Environment.
3:107-122, 1969. (0001N)
10. Stephen, R. E. Method of Decomposing Nitrogen Oxides. U.S. Patent
3,552,913. January 5, 1971. 4 p. " (0063N)
*
Document Retrieval System Accession Number.
100
-------
11. Green, T. E., and C. N. Hinshelwood. The Catalytic Decomposition of
Nitric Oxide at the Surface of Platinum. Journal of Chem. Soc. 1926.
pp. 1709-13. (0233N)
12. Backman, P. W., and 6. G. Taylor. Decomposition of Nitric Oxide by
Platinum at Elevated Temperatures and Its Retardation by Oxygen.
Journal of Phys. Chem. 33:447-55, 1929. (0234N)
13. Brennan, J. A. Removal of Nitrogen of Nitrogen Oxides from Combustion
Gases. U.S. Patent 3,015,369. May 23, 1960. (0250N)
14. Fraser, J. M., and F. Daniels. The Heterogeneous Decomposition of
Nitric Oxide With Oxide Catalysts. Journal of Physical Chemistry.
62:215-222, February 1958. (0008N)
15. Vetter, K. Kinetics of the Thermal Decomposition and Formation of
Nitric Oxide. Z. Electrochem., Part I and II. 53:369-80, 1959.
(0235N)
16. Wise, H., and M. T. Trech. Kinetics of Decomposition of Nitric Oxide
at Elevated Temperatures - I. Rate Measurements in a Quartz Vessel.
Journal Chem. Physics. 20(l):22-4, 1952. (0236N)
17. Wise, H., and M. T. Treach. Kinetics of Decomposition of NO at Elevated
Temperatures. II. The Effect of Reaction Products and the Mechanism of
Decomposition. Journal Chem. Physics. 20(11):1724-7, 1952. (0237N)
18. Lav/son, A. A Low Temperature Catalystic Approach to NOX Control.
Journal of Catalysis. 24:297-305, 1972= (0205N)
19. Amirnazmi, A., J. E. Benson, and M. Boudart. Oxygen Inhibition in the
Decomposition of NO on Metal Oxides and Platinum. Journal of Cata-
lysis. 30:55-56, 1973. (0180N)
20. Yureva, T. M., V. V. Popovski, and G. K. Boreskov. Catalytic Properties
of Metal Oxides of Period IV of the Periodic System with Respect to
Oxidation Reactions. Translated from Kinetika I Kataliz, V. 6. Publi-
cation Number 6. Nov.-Dec. 1965. p. 1041-1045. (0206N)
21. Roth, J. T. Process for Catalytically Treating Exhaust Gases. U.S.
Patent 3,493,325. February 3, 1970. 9 p. (0057N)
22. Winter, E. R. S. The Catalytic Decomposition of Nitric Oxide by
Metallic Oxides. Journal of Catalysts. 22:158-170, 1971. (0007N)
23. Zawadzki, J., and G. Perlinski. La Decomposition du Bioxyde d Azote
par les Catalyseurs de Platine. Comptes Rendus. 198:260, 1934.
(0245N)
24. Harding, J. W. Kinetics of Catalytic Decomposition of Nitric Oxide.
Univ. Microfilms Inc. 1969. 76 p. (0072N)
101
-------
25. Jones, J. H., et al. Selective Catalytic Reaction of Hydrogen with
Nitric Oxide in the Presence of Oxygen. Environmental Science and
Technology. September 1971. p. 790-798. (0011N)
26. Taylor, T. R. Catalytic Removal of Nitrogen Oxides, Carbon Monoxide
and Hydrocarbons from Combustion Exhaust Streams. Franklin Institute
Laboratories. Philadelphia. 1959. 18 p. (0176N)
27. Sourirajan, S., and J. L. Blumenthal. The Application of Copper-Silica
Catalyst for the Removal of Nitrogen Oxides Present in Low Concentra-
tions by Chemical Reduction with Carbon Monoxide or Hydrogen. Inter-
national Journal of Air Water Pollution (London). 5(l):24-33, 1961.
(0169N)
28. Shelef, M., K. Otto, and H. Gandhi. The Oxidation of CO by 0? and NO
on Supported Chromium Oxide and Other Metal Oxide Catalysts. Journal
of Catalysis. 12(4):361-375, 1968. (0167N)
29. Bauerle, 6. L., G. R. Service, and K. Nobe. Catalytic Reduction of
Nitric Oxide with Carbon Monoxide. I and EC Product Research and
Development. 11(1):54-58, March 1972. (0016N)
30. Sorenson, L. L. C., and K. Nobe. Nitric Oxide Reduction with CO and
C, Hydrocarbon Oxidation on CU-A1203-Effect of Lead. I and EC Product
Research and Development. ll(4):423-425, December 1972. (0026N)
31. Markvart, M., and V.Pour. The Influence of Oxygen on the Catalytic
Reduction of Nitric Oxide by Ammonia. Journal Catalysis. 7:279-281,
1967. (0175N)
32. Otto, K., M. Shelef, and J. T. Kummer. Studies of Surface Reactions
of NO by Isotope Labeling; II, Deuterium Kinetic Isotope Effect in the
Ammonia - Nitric Oxide Reaction of a Supported Platinum Catalyst.
Journal of Physical Chemistry. 75(7):875-879, April 1971. (0114N)
33. Otto, K., M. Shelef, and J. T. Kummer. Studies of Surface Reactions
by Nitrogen - 15 Isotope Labeling I. The Reaction Between Nitric Oxide
and Ammonia Over Supported Platinum. Journal Phyp. Chem. 74(13):2690-
2698, 1970. (021 IN)
34. Otto, K., and M. Shelef. Studies of Surface Reactions of Nitric Oxide
by Isotope Labeling. IV, Reduction of Nitric Oxide by Ammonia and
Hydrogen Over Supported Ruthenium. Journal Phy. Chem. 85(5-6):308-
322, 1973. (0212N)
35. Neumann, B., and H. Rose. The Catalytic Oxidation of Ammonia Into Nitric
Acid. A. Angew. Chem. (Weinheim). (Trans, from German). 1:45-48,
51-55, February 24, 1920. (0150N)
36. Andersen, H. C., W. J. Green, and D. R. Steele. Catalytic Treatment of
Nitric Acid Plant Tail Gas. Industrial and Engineering Chemistry.
53:199-204, March 1961. (0210N)
102
-------
37. Andersen, H. C., P. L. Romeo, and W. J. Green. A New Family of Cata-
lysts for Nitric Acid Tail Gases. Nitrogen. 50:33-36, Nov.-Dec. 1967.
(0071N)
38. Catalytic Reduction of N0x-Pilot Plant Operation. Progress Report.
Environics Inc., Huntington Beach, Calif. July 1973. (0248N)
39. Gajewski, A., S. Kupiek, and J. Zygadlo. Investigation of Catalytic
Reduction of Exit Nitric Oxides by Ammonia. Przemysl Chemical (Polish).
51(1):44-46, 1971. (0149N)
40. Jones, J. H., and E. E. Weaver. Exhaust Gas Purification. U.S. Patent
3,599,427. August 17, 1971. 6 p. (0136N)
41. Griffing, M. E., F. W. Lamb, and R. E. Stephens. Method of Controlling
Exhaust Emission. U.S. Patent 3,449,063. June 10, 1969.
42. Atroshenko, V. I., A. P. Zasorin, and 0. N. Kulish. Catalysts for the
Interaction of Nitrogen Oxides With Ammonia Studied to Purify Waste
Gases from Nitric Acid Production. Katal. (Russian) No. 9. 1972.
p. 26-30. (0051N)
43. Kulish, 0. N., A. P. Zasorin, and A. V. Atroschenko. Kinetics of the
Contract Interaction of Ammonia with Nitric Oxide and Oxygen. Izv.
Vyssh. Ucheb. Zaved., Khim. Khim. Tekhnol. (Russian) 15(6):955-957,
1972. (0042N)
44. Nonnenmacher, H., and K. Kartte. Selective Removal of Oxides of Nitro-
gen from Gas Mixtures Containing Oxygen. U.S. Patent 3,279,384.
October 18, 1966. 4 p. (0038N)
45. Schmidt, K. Method for Removing Nitrogen Oxides from Gases Through
Catalytic Reduction of These Substances to Nitrogen. West German
Patent 1,259,298. January 25, 1958. 2 p. (German). (0090N)
46. Jaros, S., and J. Krizek. Catalytic Reduction of Waste Nitrogen Oxides.
Chem. Prumysl (Prague). 17/42(11):581-586, 1967. (Czech.) (0164N
47. Kudo, T., T. Manabe, T. Gejo, M. Seki, and K. Yoshida. New Oxide Cata-
lyst with Perovskite-Related Structure for Reduction of NO with NH3.
Presented at 167th Meeting of Am. Chem. Society, Division of Water, Air,
and Waste Chem., Los Angeles, Calif. April 1974. (0253N)
48. Voiles, R. W., H. Uise, and L. P. Berriman. Study of Catalytic Control
of Exhaust Emissions for Otto Cycle Engines. Stanford Research Institute.
Final Report. Irvine, Calif. April 1970. (0251N)
49. Coretto, L. S., M. W. McElroy, J. L. Nelson, and P. D. Venturini.
Project Clean Air. California Univ. Berkeley. Dept. of Mechanical
Eng. Task Force 1. Vol.1. 68 p. Sept 1, 1970. (0116N)
50. Perrine, R. L., and H. Limin. Power and Industry; Control of Nitrogen
Oxide Emissions. Project Clean Air. Calif. Univ. Berkeley. Task Force
5. 1(9). Sept. 1970. (0220N)
103
-------
51. Ryason, P. R., and J. Harkins. Studies on a New Method of Simultaneously
Removing Sulfur Dioxide and Oxides of Nitrogen from Combustion Gases.
Journal of the Air Pollution Control Assoc. Vol. 17. No. 12. p. 796-799
Dec. 1967. (0030N)
52. Quinlan, C. W., V. C. Okay, and J. R. Kittrell. Simultaneous Catalytic
Reduction of Nitric Oxide and Sulfur Dioxide by Carbon Monoxide. Ind.
Ing. Chem. Process Des. Develop. Vol 12. No. 3. p. 359-365. 1973.
(0185N)
53. Mayland, B. J., and R. C. Heinze. Continuous Catalytic Absorption for
NOX Emission Control. Chemical Engr. Progress. Vol. 69. No. 5.
p. 75-76. May 1973. (0187N)
54. Berstein, L. S., et al. Application of Catalysts to Automotive NO
Emissions Control. Presented at the Automotive Engineering Congress,
Detroit, Jan. 11-15, 1971. Paper No. 710014. (0133N)
104
-------
4. CATALYST SCREENING AND PROOF-OF-PRINCIPLE EXPERIMENTS (TASK 2)
The objective of this task was to screen candidate NO abatement catalysts
X
for applicability to power generating plants. These catalysts were select-
ed as potential candidates on the basis of information assembled in Task 1.
Only catalysts on which available data was inadequate for either technical
or preliminary engineering evaluation were tested.
Catalyst screening and proof-of-principle experiments were performed with
simulated, S02~free, power plant flue gas nominally at 400°C (752°F) and at
20,000 hr"1 (STP) space velocity.*
Catalysts which indicated promise as efficient promoters of N0x abatement in
any of the above processes (decomposition, oxidation, reduction) were fur-
ther tested for sensitivity to important parameters. Principal parameters
varied included temperature, space velocity, and important flue gas com-
ponent concentrations. In addition, the S02 effect on these catalysts was
investigated.
The ensuing sections list the screened catalysts and detail catalyst pre-
paration, screening procedures, screening results, and the results of para-
metric investigations on the promising candidate catalysts.
4.1 CATALYSTS SELECTED FOR SCREENING
The catalysts selected for screening are listed in Table 18. The catalysts
are grouped on the basis of predominant active metal and not necessarily in
the chronological order of testing; thus, the numbering system differs from
that used in interim monthly reports. Suggestions for selection were derived:
(a) from previously published work where these or similar catalysts were sub-
jected to gases containing important components of power plant flue gas, (b)
from patents where claims of N0x abatement were made, and (c) from previous
research data at TRW and UCLA.
Unless otherwise indicated, space velocities throughout this report are
given as the ratio of flue gas volume flow rate at standard conditions per
hour to the volume of the catalyst used.
105
-------
Table 18. LIST OF CATALYSTS SUBJECTED TO SCREENING TESTS
Catalyst
No.
NA-1
NA-2
NA-3
NA-4
NA-5
NA-6
NA-7
NA-8
NA-9
NA-10
NA-11
NA-12
NA-13
NA-1 4
NA-1 5
NA-1 6
NA-17
NA-1 8
NA-1 9
NA-20
NA-21
NA-22
NA-23
NA-24
NA-25
NA-26
NA-27
NA-28
NA-29
NA-30
NA-31
NA-32
NA-33
NA-34
NA-35
NA-36
Principal
Active
Element(s)
Pt
Pt
Mo, Pt
Mo, Pt
Mo
Mo
Mo
Mo
Mo, Co
R.E.
R.E., Co
R.E., Pb
Gd, Mo
Gd, Mo
Gd, V
La, Co
Cu
Cu, Pb
Cu, Pb
Cu, Pb
W
W
W
V
V
Fe, Cr
Fe, Cr
Fe, Cr
Fe, Cr
Fe, Cr
Cr
Fe
Fe
Fe
Fe, C
Fe, C
Composition (Typea)
0.5;,; Pt on alumina
New batch of NA-1
22.2;:. Mo> Q.lZ Pt on alumina (1/16 inch spheres)
NA-3 pressed into cylinders6
27^ Ho on alumina (pressed )
14. T, Mo on alumina
NA-6 reduced in H2 at 480°C
NA-6 reduced in H, at 700°C
15" Mo03, 2.7i Co304 on alumina
Rare earth oxides (refined ore, pressed)
15% R.E. cobalt oxide on alumina (pressed)
NA-11 doped with 10% Pbc
15S Gd2 (MoO^), on alumina (pressed)
15S Gd M03 on alumina (pressed)
15% Gd V03 on alumina (pressed)
15% LaCo03 on alumina
16.5% CuO on alumina
NA-17 doped with IS Pbc
NA-17 doped with 5~~ Pbc
NA-17 doped with 10% Pbc
13. 1* W on alumina (20 x 30 mesh)
NA-21 pressed into pellets •
10% W03 on alumina
10% V205 on alumina
10% V20g on alumina
80% Fe203, 7°; Cr203, 1.6? graphite, 0.4% MgO, 0.2. Si02d
Similar composition to NA-26 (second sample)
10% Fe203-Cr203 (83Z-17%) on alumina
NA-28 composition (different salts)
56.7? Fe203, 6.7% Cr^m 1.4% Cr03> 35.2" A1203 (pressed)
15% Cr203 on alumina (pressed)
15% Fe203 on alumina (pressed)
15% Fe-0, on alumina
20% Fe203 on alumina
Fe on graphite chips (organic impregnation)
Fe on graphite chips (aqueous impregnation)
Source
Engelhard
Engelhard
TRW
TRW
TRW
UCLA
UCLA
UCLA
Filtrol
Molycorp
UCLA
UCLA
UCLA
UCLA
UCLA
UCLA
UCLA
UCLA
UCLA
UCLA
TRW
TRW
Harshaw
Filtrol
Harshaw
Girdler
Girdler
UCLA
UCLA
UCLA
UCLA
UCLA
UCLA
Harshaw
UCLA
UCLA
a Unless otherwise stated the catalysts were prepared by impregnation of performed 1/8 x 1/8 inch
alumina cylinders.
Pressed catalysts made by dry-mixing active material with alumina and forming into 1/8 x 1/8 inch
cylinders.
c Pb-doped catalyst made by impregnation of base catalyst with Pb(N03)2 solution and calcining.
Balance unspecified.
107
-------
The active catalyst elements belong principally to the transition metal and
rare earth groups with platinum, copper, lead and graphite being the only
other elements screened. Every attempt was made to include in the screen-
ing task representative catalysts from those classes for which NO abate-
x\
ment potential had been reported, provided the following criteria were met:
• The candidate catalyst indicated potential for utili-
zation at temperatures below 427°C (800°F).
• The available data indicated that the catalyst could
physically survive the power plant flue gas conditions
(e.g., flow rates at practical space velocities, tem-
perature, impurities).
• The available data were not sufficient to infer the
catalyst's potential for utilization in power plant
NOX abatement. If adequate data were available or
the available data could be safely extrapolated, the
catalyst was not screened.
It is apparent that the criteria used are somewhat subjective and as a re-
sult the above list may not be as comprehensive as intended. One group of
catalysts intentionally restricted, because of cost, was the noble metals;
the only catalyst investigated from this group was platinum. One batch
of platinum (NA-1) was screened in order to generate baseline
data for comparison purposes. Platinum was also tested for S02 effects on
its activity and parametrically scanned for NO decomposition potential.
In addition, two preparations of a molybdenum-platinum catalyst (NA-3 and
NA-4) were screened to determine the effect of Mo on the Pt activity in the
selective reduction of NO by NH, (temperature of maximum conversion). The
temperature of maximum NO reduction by NH3 acquires special importance when
SOg is present in the flue gas stream (platinum activity is severely inhi-
bited by sulfate deposition at temperatures below approximately 300°C).
108
-------
As a rule, the individual catalysts in Table 18 were selected as potential
candidates for one of the outlined NO abatement processes (decomposition,
A
oxidation, selective £nd nonselective reduction); however, most of them
were subjected to the eight tests (Section 4.3) designed to screen cata-
lytic activity toward all processes envisioned as practical for power plant
use.
4.2 CATALYST PREPARATION
A number of the catalysts screened for NO abatement potential were pre-
7\
pared by TRW or UCLA as indicated in Table 18. The following paragraphs
detail the methods of preparation of the individual catalysts (those not
commercially available).
In general, the screened catalysts were prepared by either impregnation of
carrier material with salts of the desired active ingredients or by pressing
the powdered carrier (AlgO-j) with the active metal oxide or oxides. The
following list outlines the procedure used with each catalyst.
NA-3 - This catalyst was prepared by soaking 35 g of a UOP
platforming catalyst containing approximately 0.1% Pt on
AlpOg (1.6 mm spheres) with a 75 ml solution of ammonium
molybdate (41 g/100 ml). The unused solution was decanted
off and the catalyst was dried and then reduced in H^ for
4 hours at 700°C.
NA-4 - A sample of NA-3 was ground with 5% stearic acid
(used as a die lubricant) and pressed into 3.2 x 3.2 mm
(1/8 x 1/8 in) cylinders. The stearic acid was burned
out by calcining in air at 500°C for 16 hours.
NA-5 - This catalyst was prepared by the NA-3 procedure
except that 115 g of Reynolds 14-20 mesh AlgOg and 175 ml
of the ammonium molybdate solution were used.
109
-------
NA-6 - This catalyst was prepared by the NA-3 procedure
except that 35 gm of Filtrol Grade 86 A1203 (3.2 x 3.2 mm
cylinders) was used as carrier. Instead of reduction the
catalyst was calcined in air at 480°C for 18 hours.
MA-7 - A sample of NA-6 was reduced at 480°C in H0 for 20
c.
hours.
NA-8 - A sample of NA-6 was reduced at 700°C in H0 for 4
——~~ c.
hours.
NA-10 - Fifteen grams of bastnasite, a mixture of rare
earth oxides (Molycorp; see Table 19) was ground with 85 gm
of A1203 (Filtrol Grade 90) and 5 gm of stearic acid and
pressed into 3.2 x 3.2 mm cylinders. The pellets were cal-
cined at 500°C in air for 16 hours.
NA-11 - A mixture of rare earth oxides (American Potash
and Chemical Corporation; see Table 19) was ground with
cobalt carbonate in a preparation such that the molar ratio
of cobalt to each rare earth was unity. The mixture was cal-
cined in air for 20 hours at 1000°C. Compound formation with
the general perovskite formula, (R.E.)'Co03, has been veri-
fied for La, Gd, Pr, and Nd in individual tests. Fifteen
grams of the resultant mixed oxide was mixed with 85 gm
A1203 and pressed into cylinders as described for NA-10.
NA-12 - Twenty-five grams of NA-11 were impregnated (to
total solution take up) with 7.5 ml of a solution contain-
ing 4 gm of Pb (N03)2- The catalyst was calcined in air
at 500°C for 16 hours.
110
-------
Table 19. APPROXIMATE COMPOSITION OF RARE EARTH OXIDE MIXTURES
Percentage as Oxides
Ce02
La2°3
Nd203
Pr6°ll
Sm2°3
6d203
Y2°3
Other Rare Earth Oxides
Total Rare Earth Oxides
so3
P2°5
Na20 + 1^0
CaO + MgO
Fe203 + A1203
Fluorine
Sr02
BaO
SrO
Si02
Rare Earth Oxide
(American Potash
and Chemical Corp.)
45.6
22.8
16.2
4.7
2.8
1.9
0.2
0.8
95.0
2.0
0.5
0.1
1.0
1.5
-
0.1
-
-
-
Bastnasite
(Molycorp)
27.6
41.3
11.6
4.3
0.4
0.3
0.3
86.2
86.2
-
0.5
1.0
0.8
0.1
6.0
-
1.5
0.9
3.0
111
-------
NA-13 - A mixture of 4.6 gm GeLO., and 6.7 gm
~~"^^~~~~ C, O
(NH4)6Mo7024«4H20 was dehydrated and then heated at
1000°C for 16 hours. Fifteen grams of the resultant
compound Gd2(Mo04)3 was mixed with 85 gm AUOg and
pressed into cylinder as per NA-10.
NA-14 - A mixture of 3 gm Gd203, 0.8 gm Mo, and 1.5 gm
(NH4)6Mo7024«4H20 was reacted at 800°C in He for 16 hours.
A 15/85 (with Al^) was pressed into cylinders as per
NA-10.
NA-15 - A sample of V203 was made by reducing V,,05 in H2
at 800°C for 16 hours. A mixture of 1.5 gm V203 and 3.5 gm
Gd203 was heated in He at 900°C for about 72 hours. The
material was pressed into cylinders with 85% A1203 as per
NA-10.
NA-16 - A 1:1 (molar) mixture of La90, and CoCO, was cal-
~~~~~~~ L. O 3
cined in air for 16 hours at 1000°C. The compound LaCo03
was verified by X-ray diffraction. A 15/85 mixture with
A1203 was pressed into cylinders as per NA-10.
NA-17 - Forty-five ml of a solution containing 90.6 gm
Cu(N03)2«3H20 was impregnated on 150 gm A1203 pellets
(3.2 x 3.2 mm), dried and calcined at 500°C in air for
16 hours.
NA-18 - Thirty-six grams of NA-17 were impregnated with
0.48 gm Pb(N03)2 (total take up) and recalcined as per
NA-17.
NA-19 r Thirty-five grams of NA-17 were impregnated with
2.4 gm Pb(NO,)o and calcined as per NA-17.
O £
NA-20 - Thirty-three grams of NA-17 were impregnated with
4.8 gm Pb(N03) and calcined as per NA-17.
112
-------
NA-21 - This catalyst was prepared by soaking 25 gm
Harshaw 20-30 mesh A1203 in a 30 ml solution contain-
ing 7.62 gm ammonium metatungstate. The excess solu-
tion was decanted off and the catalyst was reduced in
H2 at 700°C.
NA-22 - A sample of NA-21 was pressed into 3.2 x 3.2
mm cylinders following the procedure described for
NA-4.
NA-28 - Impregnated 100 gm A1000 pellets (3.2 x 3.2 mm)
————— ^ j
with a solution containing 53.8 gm Fe(N03)3-9H20 and 2.8 gm
CrO-. The pellets were dried and calcined in air at 500°C
for 16 hours.
NA-29 - Prepared as per NA-28 except solution contained
25.3 gm (Fe(N03)3-9H20 and 5.70 gm Cr(N03)3'10.8 HgO;
50.4 gm A1203 pellets were used.
NA-30 - Pressed as per NA-10 a mixture of 34 gm Fe203,
4 gm cr203 and 0.8 gm cr03 with 21.2 gm A1203.
NA-31 - Pressed as per NA-10 a 15/85 mixture of Cr203 and
A1203.
NA-32 - Pressed as per NA-10 a 15/85 mixture of Fe203 and
NA-33 - This catalyst was prepared by total impregnation of
32 gm A1203 pellets (3.2 x 3.2 mm cylinders) with a solu-
tion containing 15.2 gm Fe(N03)3-9H20.
NA-35 - This catalyst was prepared by impregnation of 15 gm
graphite chips ( 4 mm diameter x 2 mm thick) with a solution
of 12 gm Fe(N03)3-9H20 isopropanol. Isoamyl alcohol was
113
-------
added and the solution boiled for 4 hours. The chips were
washed quickly in acetone and dried in air at 180°C.
NA-36 - This catalyst was prepared by impregnating 25.7
graphite chips with a saturated aqueous solution of
Fe(N03)3-9H20. The catalyst was dried in air at 180°C.
4.3 CATALYSTS SCREENING TEST CONDITIONS
Figure 19 is a schematic diagram of the test apparatus used in this phase
of effort. A glass manifold enabled mixing of N2 carrier gas, NO, C02. H2,
air (02), and CO. Each gas stream was equipped with individual metering
valves and bead-type flow meters. For precise measurement of the flow rates
of the species used at low concentrations, flow could be diverted prior to
entry into the mixing manifold through a soap-bubble flow meter.
The gas mixture passed through a 4 m preheater coil (6.4 mm diameter stain-
less steel tube) and into the reactor. Both reactor and preheater were
immersed in an electrically heated, fluidized bed furnace containing powder-
ed alundum. The temperature of the fluidized heater was held at the desired
level with a meter-relay using a chromel-alumel thermocouple. With the pro-
per air flow rate, the fluidization was sufficient to reduce temperature
gradients in the bath to below 5°C.
Water vapor was added to the gas mixture by vaporization of a measured flow
of pumped liquid water in a vaporizer tube (1.4 cm diameter x 30 cm long)
located immediately upstream of the preheater coil. The vaporizer was heat-
ed with electrical tapes to 320-380°C.
Ammonia was fed to the gas stream, when necessary, at the exit of the
vaporizer; an analyzed supply of an NHo-N2 mixture (-^5% NH-) was used. In
tests for the investigation of sulfur effects, S02 (anhydrous) was injected
into the gas stream near the point of water admission.
Tests were conducted at a pressure of 4 mm Hg gage which was maintained by
regulation of the vent valve. Samples of the reactant effluent streams
could be taken through a valving arrangement.
114
-------
en
H20
VAPORIZER
MANOMETER M
PREHEATER-
NO
NH3-N2
REACTOR
AIR CO, CO
H2
VENT
AIR
ELECTRICALLY HEATED
FLUIDIZED BATH
3
SAMPLE
TO ANALYZERS
DRAIN
TO W. T. METER
DIAPHRAGM
_l PUMP
ABSORBER
Figure 19. Catalyst Screening Test Apparatus
-------
Analyses were performed as follows:
NO Nondispersive infrared (NDIR) (Beckman Model 315A)
C02 NDIR (MSA LIRA 300)
CO NDIR (MSA LIRA 300)
S02 NDIR (MSA LIRA 200)
N02 Visible colorimetry (Beckman Model 77 flow colorimeter)
N20 Gas chromatography. Perkin-Elmer Model 990 with 3-
meter column of Porapak Q and T.C. detector.
NH~ Absorption in a bubbler of a known quantity of gas
(measured with a wet test meter) in 2% boric acid
followed by titration with 0.03N HC1 using bromo-
cresol green indicator. All lines from the reactor
to the bubbler were heated to prevent water con-
densation. In tests in which S02 was present,
which interfered with the wet analysis, adsorbed
NH~ was analyzed using a specific ion electrode
(Orion).
The reactor used in the screening tests was constructed of stainless steel
tubing (1.4 cm diameter x 9 cm long). The catalyst loading was ^14 gm.
Standard total flow rate was 283 1/hr (STP) resulting in a space velocity
of 20,000 hr"1 (STP). The screening tests were performed at 400°C (752°F).
The standard screening test conditions are listed in Table 20 (in a few in-
stances additional isolated points were taken at temperatures other than
400°C and space velocities othei
the test matrix was as follows:
400°C and space velocities other than 20,000 hr ). The rationale behind
Test 1 - To determine the potential of the catalyst for pro-
moting the decomposition and/or oxidation of NO in a typical
stack gas. In general, Test 1 was always performed first;
since an overall oxidizing atmosphere existed in the flue
gas, prereduction of the catalyst was considered unnecessary.
116
-------
Table 20. CATALYST SCREENING TEST CONDITIONS
Test
No.
1
2
3
4
5
6
7
8
NO
(ppm)
1000
1000
1000
1000
1000
1000
1000
1000
°2
(?)
3
3
3
0.5
0.5
3
0.5
0.5
CO?
(XT
14
14
14
14
14
14
14
14
HeO
(1)
5
5
5
5
5
5
5
5
NH3
(ppm)
0
667
0
0
0
0
0
0
CO
(ppm)
0
0
0
0
0
1000
1000
11000
H2 ,
(ppm)
0
0
1000
1000
11000
0
0
0
Catalyst Screened for NO
A
Oxidation and decomposition (lean combustion)
Selective reduction (NH3) (lean combustion)
Selective reduction (H2) (lean combustion)
Selective reduction (H2) (near stoichiometric com-
bustion)
Nonselective reduction (H2) (near stoichiometric
combustion)
Selective reduction (CO) (lean combustion)
Selective reduction (CO) (near stoichiometric com-
bustion)
Nonselective reduction (CO) (near stoichiometric
combustion)
Nominal Conditions: 400°C, 20,000 hr (STP) space velocity.
-------
Test 2 - To determine the catalyst's potential for pro-
moting selective reduction of NO with a near stoichiometric
amount of NH3 in the presence of oxygen.
Test 3 - To investigate the catalyst's potential for
selectively promoting the reduction of NO with H2 in
large excess of 02 (lean combustion).
Test 4 - A second NO-H2 selective reduction test, but with
a smaller excess air (near stoichiometric combustion).
Test 5 - To determine the catalyst's potential for non-
selective NO abatement with H9.
A £•
Tests 6, 7, and 8 - Analogous to Tests 3, 4, and 5 with
CO as the reductant.
Tests 5 and 8 were performed with flue gas containing 0.5% 02, instead of
3%, in order to avoid large quantities of heat release in the catalyst bed
and therefore nom'sothermal testing.
The flue gas composition, temperature, and space velocity for these tests
were selected because it is believed that they represent realistic process
adaptation conditions to power plants while at the same time they facilitated
data acquisition and accuracy. It was felt that catalysts which would not
exhibit any activity under the selected values of temperature and space
velocity had a very remote chance of being active at any other value of
these parameters compatible to power plant conditions; a possible exception
is NO oxidation. NO oxidation and decomposition tests were performed
simultaneously and the selected temperature for them was a compromise. The
NO content of the test flue gas was set at a higher value than that normally
present in power plant flue gas in order to ascertain measurement accuracy,
especially that of products. S02 was not included in the screening tests in
order to avoid undue complications in data acquisition; the S02 effect on the
activity of promising catalysts was investigated during the parametric tests.
118
-------
Extensive variations from nominal conditions were employed with promising
candidate NO abatement catalysts (Section 4.5). Space velocity variations
/\
were accommodated by changes in the standard flow rate, but in certain cases
a larger catalyst bed was used (^28 gm) in a somewhat larger stainless steel
reactor (1.9 cm i.d. x 9.8 cm long) than the one previously described.
Prior to catalyst screening tests the empty reactor and preheater coils
were checked for catalytic activity. The data in Table 21 indicates that
at the nominal screening test conditions the input flue gas constituents
were not materially affected by either homogeneous reaction or the empty
reactor and preheater coil.
Table 21. RESULTS OF TESTS WITH EMPTY REACTOR
Test
No.
1
2
3
5
8
Inlet Gas Composition
NO
(ppm)
974
974
910
932
903
NH3
(ppm)
0
>5000a
0
0
0
CO
(ppm)
0
0
0
0
1087
H2
(ppm)
0
0
931
9780
0
Outlet Gas Composition
NO
(ppm)
967
972
891
931
889
NgO
(ppm)
0
0
0
0
0
N02
(ppm)
0
0
0
0
0
CO
(ppm)
0
0
0
0
1073
Conv.
of NO
(*)
0.7
0.2
2.1
0.1
1.6
a Initially, stoichiometric NHs was admitted. After no reaction was
observed the NHs content was increased to the indicated large excess.
The catalyst screening test data was generated under steady state conditions.
Every effort was made to insure that the data was not affected by transient
surface adsorption or surface reaction flue gas interactions with the cata-
lyst. Sufficient quantities of flue gas were passed through the reactor to
insure system saturation by even the most dilute component. This procedure
also insured saturation of the analytical instrument manifold. Each data
point was taken after at least one hour of unchanging reactor effluent com-
position.
119
-------
4.4 CATALYST SCREENING TEST RESULTS
The data from the screening tests has been organized into six groups. Data
summaries and evaluation for each group are presented in separate subsections.
Catalysts considered as meriting further investigation were subjected to
parametric and S02 effect studies; data from these studies are presented
in Section 4.5. The criteria used for labeling a catalyst as a promising
candidate for power plant utilization, therefore meriting further investi-
gation, varied somewhat with the process in which the activity was exhibited
and with the products of the reaction. In general, at least 50% NO con-
version was required as a minimum under the nominal screening test condi-
tions (Section 4.3). For NO decomposition somewhat lower conversions were
considered acceptable for further testing. NO oxidation conversions as low
as 10-20% were also considered acceptable because the screening test tem-
perature was relatively high for this reaction (the decomposition and oxi-
dation tests were performed simultaneously, therefore, a compromise tem-
perature was used).
4.4.1 Platinum and Platinum-Molybdenum Catalysts
Table 22 summarizes the catalyst screening test data taken with SOg-free
synthetic flue gas on alumina supported Pt (NA-1) and Pt-Mo (NA-3 and NA-4)
catalysts. The NA-1 sample was a commercial catalyst prepared by Engelhard.
The NA-3 and NA-4 samples were prepared at TRW on a commercial support
(Section 3.2). The NA-3 catalyst was in the form of 1.6 mm (1/16 inch)
spheres; the other two catalysts in this table and those in the subsequent
tables of this section were 3.2 x 3.2 mm (1/8 x 1/8 inch) cylinders. The
NA-4 sample was prepared from NA-3 by reshaping it into the cylinder form.
Data from Test No. 1, the decomposition and oxidation screening test, indi-
cates that both the Pt and the Pt-Mo catalysts exhibited potential in NO
decomposition but no activity toward NO oxidation. However, when the nominal
3% oxygen concentration in the synthetic flue gas was reduced to 0.5%, the
NO decomposition activity of Pt was severely reduced. Similar oxygen effect
was observed during the selective NO reduction tests with H2 (compare NO con-
versions in Test Nos. 3 and 4) and with CO (Test Nos. 6 and 7). Since
120
-------
Table 22. CATALYST SCREENING TEST RESULTS (Pt AND Pt-Mo CATALYSTS)
Catalyst
No. and
Type
NA-1 (0.5% Pt
on alumina)
HA-3 (22.2% Mo
O.lt Pt on
alumina)
NA-4 ( Pel-
let i zed
NA-3)
Test
No.
1
1
2
3
4
5
6
7
8
1
2
2C
2d
2e
3
4
5
6
7
8
1
2
3
4
5
6
7
79
8
Feed Gas Composition9
02 NO NH3 \\2 CO
(%) (ppm) (ppm) (ppm) (ppm)
3 1023 000
0.5 1077 000
3 1061 707 0 0
3 1063 0 1030 0
0.5 1059 0 1048 0
0.5 1022 0 11000 0
3 1081 0 0 995
0.5 1050 0 0 995
0.5 1036 0 0 10355
3 1008 000
3 1065 768 0 0
3 968 760 0 0
3 968 760 0 0
3 968 760 0 0
3 1098 0 1030 0
0.5 1047 0 1030 0
0.5 1064 0 11000 0
3 1040 0 0 1048
0.5 1020 0 0 1073
0.5 1026 0 0 11000
3 898 6 0 0
3 931 793 0 0
3 932 0 1012 0
0.5 988 0 1012 0
0.5 988 0 11000 0
3 879 0 0 1003
0.5 997 0 0 1003
0.5 1007 0 0 1030
0.5 1059 0 0 11000
Monitored Gas Constituents
in Reactor Effluent b
NO N02 N20 CO NH3
(ppm) (ppm) (ppm) (ppm) (ppm)
561 0 0
926 0 N.A.
724 0 150 0
568 0 0 - 0
854 0 0 - 0
14 0 0-515
523 0 0 29 -
813 0 0 0 -
0000-
778 0 0
660 0 290 0
328 0 340 - 0
346 0 385 - N.A.
400 0 356 - N.A.
803 0 0 - 0
918 0 0 - 0
0 0 0 - 548
780 0 0 0 0
868 0 0 0 0
0 0 0 200 0
675 0 0
502 0 318 - 0
670 0 0 - 0
801 0 0 - 0
185 0 0 - 545
608 0 0 0 0
854 0 0 0 0
886 0 0 0 0
114 0 0 700 230
NO
Reduct.
or
Decomp.
(*)
45.2
14.0
31.8
46.6
19.4
98.6
46.0
22.6
100
22.8
38.0
66.1
64.3
58.7
26.9
12.3
100.0
25.0
14.9
100.0
24.8
46.1
28.1
18.9
81.3
30.8
14.3
12.0
89.2
NO
Oxida-
tion
(«
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Excess
NHs
sage
(«)
.
15f
-
-
-
_
-
-
_
184f
76f
83f
100f
-
-
-
-
-
-
177f
-
-
-
-
-
-
-
In addition to the indicated gas components the synthetic flue gas contained 142 CO- and 5% water vapor
with the balance being nitrogen.
Effluent hydrogen was not monitored.
e 14.000 hr"1 space velocity (nominal 20,000 hr"1)
d Reaction temperature 263°C (nominal 400°C)
' Reaction temperature 300°C
f Calculated on the basis of the desired reaction 6NO + 4NH3 •* SNg + 6H20.
' Repeat test after overnight air exposure at 400°C.
121
-------
neither H2 nor CO selectivity for NO reduction appears evident (NO con-
versions during Test Nos. 1, 3, and 6 are nearly identical), the oxygen
effect observed during the selective tests must be also due to NO decom-
position. The Pt-Mo catalysts (NA-3 and NA-4) exhibited the same behavior
except that the extent of NO decomposition on these catalysts, under nominal
flue gas Op concentration (3%), was lower than that obtained on Pt. Thus,
the tentative conclusion was drawn that NO decomposition increases with
increasing 0« concentration in the range of 0.5 to 3%.
Comparison of data derived from Test Nos. 1, 3, and 6 for Pt and Pt-Mo
catalysts reveals that the activity of platinum toward NO decomposition is
twice that of the latter catalysts. However, this difference in activity
was not verified; the NA-1 sample was accidentally discarded before any
reproducibi1ity tests on NO decomposition could be performed. Attempts to
reproduce the 45.2$ NO decomposition on a new batch of 0.5% Pt on alumina,
acquired from Engelhard, failed. A reproducible value of about 27% NO de-
composition was obtained with the new catalyst. This NO conversion value
is closer to that generated on NA-3 and NA-4 catalysts. The 27% value was
assumed to be more reliable even though the 45% decomposition attained in
Test No. 1 of NA-1 was indirectly reproduced in Test Nos. 3 and 6 performed
on the same sample.
The 27% NO decomposition value was considered sufficient to subject platinum
to parametric investigation (Section 4.5).
As expected (Section 3.2) selective NO reduction by NH3 on Pt at 400°C was
inefficient (Test No. 2). The data on NA-1 indicates only 32% NO conversion;
this value is lower than that obtained in Test No. 1 which implies that NH3
inhibits NO decomposition on platinum. However, even if it is assumed that
the true NO decomposition value is 27%, the use of Pt to promote the
selective NO-NHg reaction at 400°C is not justified. Of course, Pt is a
very efficient catalyst for the above reaction at lower temperatures, but
only with S02-free flue gas. As is indicated in Section 4.5 (parametric
investigations), S02-poisoned Pt catalysts at 200-300°C can be regenerated
at 400°C by air. The implication of this finding is that if Pt were to
122
-------
provide the NO-NH3 reduction reaction at 400°C its use with SCL flue gases
would have been possible.
The objective of screening the NA-3 and NA-4 catalysts was to investigate
the effect of Mo on Pt activity (especially its activity for the NO-NH3
reaction). The data in Table 22 indicates that the Pt promoted Mo cata-
lysts behaved very similar to the supported Pt catalyst (NA-1). NO re-
duction by ammonia was low at 400°C, but increased with decreasing tem-
perature. Also, both catalyst types promoted the production of N20 in
quantities nearly equal to NO reduced (on per mole basis).* Thus, Pt
activity does not appear to be materially affected by the large presence
of Mo.
The last column in Table 22 indicates the percent excess usage of ammonia
during Test No. 2. This excess value was calculated from the reactor "in"
and "out" values of NO and NH3 and on the basis of the reaction
6NO + 4NH3 + 5N2 + GHgO
According to their reaction, the quantity of ammonia consumed in Test No. 2
should equal 67% of the NO reduced (mole or volume basis); additional
ammonia consumption is labeled "excess usage".
As expected, the Pt catalyst was a very efficient promoter of nonselective
NO reduction with both Hg and CO (Test Nos. 5 and 8). However, with H2 as
the reductant NH3 production was higher than expected at 400°C under very
nearly stoichiometric oxidant-reductant conditions. The Pt-Mo catalysts indi-
cated activity equal to Pt for these reactions (the slightly lower activity
exhibited by NA-4 must be due to surface area reduction during preparation).
In addition to generating baseline data on Test Nos. 1 through 8, the screen-
ing tests on Pt and Pt-Mo catalysts furnished the following information:
To our knowledge, N20 production during the NO-NHg reduction reaction on
Pt in the presence of 02 had not been reported prior to being reported in
this program's interim reports.
123
-------
• Pt was identified as a potential NO decomposition cata-
lyst meriting further investigation.
• The Pt-Mo catalysts behaved very similar to the Pt cata-
lyst in the Table 22 experimental matrix. In view of
the lower Pt content of these catalysts, they may be
preferable to Pt because of cost considerations. How-
ever, the role of Mo must be established more precisely
and longer-term activity tests must be performed prior
to such substitution.
4.4.2 Molybdenum-Based Catalysts
Table 23 presents the screening test data generated on four molybdenum-on-
alumina catalysts (NA-5 through 8) and one alumina supported molybdenum-
cobalt oxide catalyst (NA-9). These catalysts were prepared by TRW and UCLA
as described in Section 4.2; the first four differ from each other either in
Mo content or in pretreatment prior to testing. In general, catalysts were
not activated prior to testing (other than exposure to flue gas until steady
state was reached), but NA-7 and NA-8 were reduced in H2 prior to screen
testing.
The first three catalysts of Table 23 were subjected to the entire screen-
ing test matrix; the last two were only tested for NO decomposition or
oxidation potential (Test No. 1) and for activity in Test No. 2 (NO-NH3
selective reduction). One of the catalysts (NA-7) was subjected to addi-
tional off-nominal temperature and space velocity testing in order to com-
pare the derived data to that generated on the Pt-Mo catalysts (Table 22).
The Mo catalysts were selected primarily for testing their potential in the
selective reduction of NO with ammonia. Their selection was based on the
activity they exhibited toward hydrazine decomposition (TRW patent).
The data in Table 23 verified the expectation that Mo catalysts would pro-
mote NO reduction by NH, in the presence of oxygen (Test No. 2 data). Under
l
the screening test conditions (400°C, 20,000 hr ) about 46% NO reduction
was achieved; on NA-6 NO reduction was 40% while on NA-9 (the Co-Mo oxide
124
-------
Table 23. CATALYST SCREENING TEST RESULTS (Mo BASED CATALYSTS)
Catalyst
No. and
Type
NA-5 (27% Mo
on A1203)
NA-6 (14.7%
Mo on
A1203)
Test
No.
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
Feed Gas Composition
02 NO NH3 H2 CO
(%) (ppm) (ppm) (ppm) (ppm)
3 996 0 0 0
3 1045 752 0 0
3 995 0 1020 0
0.5 1031 0 1020 0
0.5 1047 0 11000 0
3 1050 0 0 1003
0.5 1012 0 0 1003
0.5 1026 0 0 11000
3 1036
3 977 628
3 1055 - 1003
0.5 1020 - 1003
0.5 1020 - 11000
3 1007 - - 1003
0.5 986 - - 1003
0.5 993 - - 11000
Monitored Gas Constituents
In Reactor Effluentb
NO N02 N20 CO NH3
(ppm) (ppm) (ppm) (ppm) (ppm)
968 0 0
560 0 0 - 250
969 0 0 - 0
1006 00-0
950 0 0 - 0
1036 0 0 550 0
1007 0 0 865 0
1013 0 0 N.A. 0
1028 00--
589 0 0 358
1040 .00-0
1015 00-0
1007 0 0 - 15
981 0 0 836 0
964 0 0 897 0
989 0 0 ^,11000 °
NO
Reduct.
or
Decomp.
(*)
2.8
46.4
2.6
2.4
9.3
1.3
0.5
1.3
0.7
39.7
1.4
0.5
1.4
2.6
2.2
0.4
NO
Oxida-
tion
(X)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Excess
NHs
Usage
(%)
55
-
-
-
-
-
-
_
143
-
-
-
-
—
t\>
171
a In addition to the indicated gas components the synthetic flue gas contained 14% C02 and 5% water
vapor with the balance being nitrogen.
Effluent hydrogen was not monitored.
-------
Table 23. (Continued) CATALYST SCREENING TEST RESULTS (Mo BASED CATALYSTS)
ro
Catalyst
No. and
Type
NA-7 (14.7*
Mo on
A1203)C
NA-8 (14.7%
Mo on
A1203)J
NA-9 (2.756
co3o4
15* Mo03)
Test
No.
1
2
2d
2e
2
2f
29
2h
2
2
2
21
3
4
5
6
7
8
1
2
1
2
Feed Gas Composition
02 NO NH3 H2 CO
(*) (ppm) (ppm) (ppm) (ppm)
3 1022 000
3 975 729 0 0
3 914 718 0 0
3 916 718 0 0
3 931 740 0 0
3 931 740 0 0
3 931 740 0 0
3 932 740 0 0
3 937 873 0 0
3 916 1112 0 0
3 937 866 0 0
3 933 1140 0 0
3 1052 0 1038 0
0.5 1036 0 1038 0
0.5 1036 0 10355 0
3 1048 0 0 1000
0.5 1022 0 0 1012
0.5 1023 0 0 10355
3 910 0 0 0
3 950 734 0 0
3 1029 000
3 992 812 0 0
Monitored Gas Constituents
in Reactor Effluent
NO N02 N20 CO NH3
(ppm) (ppm) (ppm) (ppm) (ppm)
1005 00--
530 0 0 - 246
684 0 0 - 447
532 0 0 - 299
501 0 0 - 242
522 0 90 - 192
540 0 66 - 144
744 0 0 - 515
445 0 0 0 400
412 0 0 0 585
441 0 0 0 381
276 0 76 0 259
1021 00-0
1030 00-0
955 0 0 - 0
1029 0 0 1000 0
1014 0 0 1000 0
1014 0 0 9690 N.A.
905 0 0 0 0
521 0 0 0 335
1023 0000
847 0 0 0 627
NO
Reduct.
or
Decomp .
(*)
1.7
45.6
25.2
41.9
46.2
43.9
42.0
20.2
52.5
5b.O
53.0
79.0
2.9
0.6
7.8
1.8
0.8
0.9
0.6
45.2
0.6
14.6
NO
Oxida-
tion
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
. 0
0
0
0
0
0
0
Excess
NH3
Usage
(*)
_
62
77
64
74
100
129
80
44
57
47
80
-
-
-
-
-
-
-
40
-
91
Catalyst reduced for 20 hours at 480°C in H2
Reaction temperature 300°C
Reaction temperature 350°C
Reaction temperature 450°C
Reaction temperature 500°C
Reaction temperature 500°C
Reaction temperature 275°C
1 10,000 hr'Vpace velocity (STP) (all others at
20,00 hr~')
J Catalyst reduced at 700°C for four hours in H.
-------
catalyst) it was only 15%. The extent of NO reduction did not appear to be
influenced either by the Mo content of the catalyst in the 14.7 to 27% range
or by catalyst pretreatment (Hg reduction). However, a slight improvement
in activity was observed with exposure time to NH., containing flue gas (NA-7),
The tests on NA-7 were performed in the presented sequence; the data from
the last three runs of Test No. 2, performed under nominal conditions, indi-
cate NO reduction in the 53-55% range versus 46% observed earlier.
Nitric oxide reduction improved dramatically with decreasing space velocity
(79% NO reduction at 10,000 hr , NA-7). Maximum NO reduction appeared to
occur at 400 +_ 50°C; at lower temperatures dropped off severely (only 20%
at 275°C). This behavior is contrary to that exhibited by the Pt-Mo cata-
lysts; it supports the conclusion drawn in the previous section that the
Pt-Mo catalyst behaves more like Pt than Mo catalyst. The same conclusion
is drawn from the fact that the Mo catalyst did not exhibit appreciable
activity in Test Nos. 1, 5 and 8 (NO decomposition and nonselective re-
duction).
The Mo catalyst must be considered as a potential candidate for NO abate-
j\
ment by means of the selective NH, reduction process. They were not investi-
gated further, however, because they were ranked third and only the top two
ranked catalysts underwent parametric investigation. As indicated above,
NA-7 was subjected to a temperature scan and two values of space velocity.
4.4.3 Rare-Earth-Based Catalysts
Table 24 presents the data derived from screening tests performed on seven
rare earth and rare-earth-transition metal oxide mixtures (NA-10 through
NA-16). NA-10 was a commercial rare-earth oxide mixture (refined ore); all
the other catalysts in this group were prepared at UCLA (refer to Section
4.2 for method of preparation). These catalysts were screened primarily
for NO oxidation-decomposition potential.
The data in Table 24 indicates that under the nominal screening test con-
ditions these catalysts failed to promote any of the NO abatement reactions.
A
The only exception was NA-15 (15% 6dV03 on alumina) which appeared to promote
128
-------
Table 24. CATALYST SCREENING TEST RESULTS (RARE EARTH OXIDE BASED CATALYSTS)
Catalyst
No. and
Type
NA-11 (Rare
earth-
cobalt
oxides on
alumina)
NA-12 (NA-
10*Pb)
Test
No.
1
1
2
3
4
5
6
7
8
1
2
3
4
5
b
7
8
Feed Gas Composition3
02 NO NH3 H2 CO
(%) (ppm) (ppm) (ppm) (ppm)
3 1022 0 0 0
3 1011 000
3 1011 735 0 0
3 1011 Q 1012 Ov
0.5 1011 0 1012 0
0.5 1011 0 11000 0
3 1011 0 0 1003
0.5 1011 0 0 1003
0.5 1011 0 0 11000
3 1012 000
3 1012 700 0 0
3 1012 0 1057 0
0.5 1012 0 1067 0 '
0.5 1012 0 11000 0
3 1012 0 0 1012
0.5 1012 0 0 1012
0.5 1012 0 0 10355
Monitored Gas Constituents
in Reactor Effluentb
NO N02 N20 CO NH3
(ppm) (ppm) (ppm) (ppm) (ppm)
1004 0 - -
620 ^150 0
654 0 .0 - 325
961 0 0 - 0
974 0 0 - 0
1038 00-0
1018 0 0 318 0
1020 0 0 737 0
1007 0 0 3045 0
658 V| 40 * 0
964 0 0 - 0
927 0 0 ' 0
913 6 0 .- 0
1029 0 0 - 0
879 0 0 60 0
973 0 0 75 0
913 0 0 2500 0
NO
Reduct.
or
Decomp .
1.8
23.8
35.3
4.9
3.7
0
0
0
0
21.1
4.7
'8.4
9.9
0
13.1
3.9
9.9
NO
Oxida-
tion
0
14.8
0
0
0
0
0*
0
0
13.8
0
0
0
0
0
0
0 -
Excess
NH3
Usage
(%)
-
72
-
-
_
-
-
-
-
>800
-
-
-
-
-
"
ISS
a In addition to the indicated gas components the synthetic flue qas contained 14% C02
and 5% water vapor with the balance being nitrogen.
Effluent hydrogen was not monitored.
-------
Table 24. (CONTINUED) CATALYST SCREENING TEST RESULTS (RARE EARTH OXIDE BASED CATALYSTS)
Catalyst
No. and
Type
NA-13 (15%
Gd2(Mo04)3
ft'2 3
NA-14 (15%
on
A1203)
NA-15 (15Z
GdVOs
on A1203)
NA-16 (15%
LaCo03 on
A1203)
Test
No.
1
2
3
4
5
6
7
8
1
2
1
2
3
4
5
6
7
8
1
Feed Gas Composition3
°2 NO NH3 H2 CO
W (ppm) (ppm) (ppm) (ppm)
3 1014 000
3 997 792 0 0
3 1021 0 978 0
0.5 1000 0 978 0
0.5 936 0 9515 0
3 993 0 0 962
0.5 995 0 0 962
0.5 950 0 0 9780
3 954 0 0 0
3 907 904 0 0
3 1002 000
3 1004 764 0 0
3 1009 0 978 0
0.5 971 0 978 0
0.5 955 0 10059 0
3 1007 0 0 1030
0.5 972 0 0 1030
0.5 869 0 0 9780
3 1017 00 0
Monitored Gas Constituents
in Reactor Effluentb
NO N02 N2o CO NH3
(ppm) (ppm) (ppm) (ppm) (ppm)
1010 0000
937 0 0 0 730
1000 0000
978 0 0 0 0
917 0 0 0 0
960 0 0 962 0
929 0 0 940 0
941 0 0 9620 0
947 0 0 0 0
801 0 0 0 743
876 67 0 0 0
390 0 0 0 405
876 52 0 0 0
933 0000
900 0 0 0 0
902 0 0 986 0
957 0 0 986 0
841 0 0 N.A. 0
912 0 0 0 0
NO
Reduct.
or
Decomp.
(X)
0.4
6.0
2.1
2.2
2.0
3.3
6.6
0.9
0.7
11.7
5.9
61.2
8.0
3.9
5.8
10.4
1.5
3.2
10.3
NO
Oxida-
tion
(X)
0
0
0
0
0
0
0
0
0
0
6.7
0
5.2
0
0
0
0
0
0
Excess
NH3
Usage
(%)
55
_
_
_
_
-
.
127
-14
-
-
-
-
-
-
-
In addition to the indicated gas components the synthetic flue gas contained 14% C02 and 5% water
vapor with the balance being nitrogen.
Effluent hydrogen was not monitored.
-------
NO oxidation (Test No. 1). The produced N02 corresponded to less than 10%
NO oxidation, but at 400°C this extent of conversion may be significant.
The same catalyst promoted the selective reduction of NO with NH3, but to
no greater extent than vanadia on alumina alone did. These catalysts were
not investigated further.
4.4.4 Copper Oxide and Lead Doped Copper Oxide Catalysts
The catalysts in this group were prepared by UCLA. The group includes one
copper oxide on alumina catalyst (16.5% CUO, 83.5% alumina), NA-17, and three
lead-doped CuO catalysts (1%, 5%, and 10% lead) prepared from NA-17. These
catalysts were selected for screening as potential promoters for the selective
reduction of NO by NHj and CO. As indicated in Section 3.2, CuO has been
suggested as a promising Test No. 2 catalyst (selective NO-NH3 reaction),
but convincing data was not available. Earlier experimental work at UCLA
had suggested that lead doped CuO may prove to be an active catalyst for
the selective reduction of NO with CO by inhibiting the C0-02 reaction. The
latter expectation did not materialize, but copper oxide did exhibit some
activity for the selective reduction of NO with NH3. The activity of the
lead doped CuO in the NO-NH, reaction was significant. The latter catalysts
indicated higher activity than CuO in the nonselective NO reduction with H,,,
also. Table 25 presents the screening test data generated on this group of
catalysts.
The data indicates that lead improves the activity of CuO in virtually all
the reactions included in the catalyst screening test matrix. Significant
NO reduction was obtained in Test Nos. 2 and 5 (selective reduction with
NH3 and nonselective reduction with Hg). In Test No. 2 performance, the
lead doped CuO could be ranked equal to the Mo catalysts; toxicity con-
siderations, however, should rank the Mo catalysts higher. The NO con-
versions in Test No. 5, even though appreciable, do not justify the use of
these catalysts in nonselective NO abatement by reduction, but they may be
A
good candidates for simultaneous NO -SO abatement by this process especially
A A
if they do not promote HS production.
132
-------
Table 25. CATALYST SCREENING TEST RESULTS (COPPER AND COPPER-LEAD CATALYSTS)
Catalyst
No. and
Type
NA-17 (16. 5%
CuO on
alumina)
NA-18 (NA-17
+ H Pb)
NA-19 (NA-17
+ 5Z Pb)
NA-20 (NA-17
+ 10J Pb)
Test
No.
1
2
2C
3
4
5
6
7
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
Feed Gas Composition3
0- NO NH, H- CO
(%) (ppm) (ppm) (ppm) (ppm)
3 946 0 0 0
3 946 557 0 0
3 1021 750 0 0
3 1029 0 1049 0
0.5 1029 0 1049 0
0.5 1029 0 11700 0
3 1029 0 0 1021
0.5 1029 0 0 1021
0.5 1029 0 0 11000
3 1019 000
3 1019 723 0 0
3 1019 0 1048 0
0.5 1019 0 1067 0
0.5 1019 0 11000 0
3 1030 0 0 1057
0.5 1030 0 0 1020
0.5 1030 0 0 11000
3 1020 000
3 1020 761 0 0
3 1020 0 1029 0
0.5 1020 0 1067 0
0.5 1020 0 10355 C
3 1020 0 0 1067
0.5 1020 0 0 1067
0.5 1020 0 0 11000
3 1048 000
3 1048 651 0 0
3 1048 0 1030 0
0.5 1048 0 1030 0
0.5 1048 0 11000 0
3 1048 0 0 1030
0.5 1048 0 0 1030
0.5 1048 0 0 11000
Monitored Gas Constituents
In Reactor Effluent13
NO N02 N20 CO NH-
(pptn) (ppm) (ppm) (ppm) (ppm)
931 0 0
594 0 0 125
776 0 0 - 533
923 0 0 120
951 0 0 - 33
871 0 0 14
1027 0 0 548 0
988 0 05
-------
These catalysts were not investigated further again because of ranking and
not because of lack of potential.
4.4.5 Tungsten Oxide and Vanadia Catalyst
Several investigators have proposed the use of these catalysts for NO
^
abatement by selective reduction with ammonia (Table 12, Section 3.2).
Insufficient data on claims made on these catalysts led to the decision to
screen under this program the catalysts presented in Table 26. The NA-21
and NA-22 tungsten oxide catalyst were prepared by TRW; they represent the
same catalyst in two shapes. As it was the case with the Mo catalysts,
NA-21 and NA-22 were active in hydrazine decomposition. The other three
catalysts in Table 4.9 represent commercial preparations (one tungsten
oxide and two vanadia catalysts). Only one tungsten oxide (NA-22) and one
vanadia catalyst (NA-24) were subjected to the entire test matrix. The
other three catalysts, which were similar to or were derived from the first
two, were only subjected to tests where potential activity had been indicated
earlier.
The data in Table 26 indicates that regardless of claims, the WOo catalysts
did not appreciably promote the selective reduction of NO by NH~ under screen-
ing test conditions; this is at least true for the catalyst preparations
tested under this program. Some activity was exhibited toward NO decom-
position and nonselective NO reduction with H2» but not sufficient to
suggest additional investigation under this program.
The vanadia catalysts (NA-24 and NA-25) promoted substantial NO reduction
under Test No. 2 conditions. These catalysts were selected for parametric
and SOp effect investigations; the generated data is presented in the next
section. It should be noted that the vanadia catalysts did not promote N^O
production and that the excess ammonia consumption was negligible.
4.4.6 Iron and Chromium Based Catalysts
In Section 3 of this report it was indicated that iron, chromium, and iron-
chromium catalysts, especially the oxides, have been investigated and to a
lesser extent suggested for NOV abatement utilization. Predominantly,
J\
134
-------
Table 26. CATALYST SCREENING TEST RESULTS (TUNGSTEN OXIDE AND VANADIA CATALYSTS)
CO
Ol
Catalyst
No. and
Type
NA-21 (13.1%
W on
alumina)
20-30 mesh
NA-22 (13.1%
W on
alumina)
1/8 inch
cylinders
NA-23 (10«
W03 on
Harshaw
NA-24 (10*
V205 on
A1203)
mtrol
NA-25 (10%
V205 on
A1203)
Harshaw
Test
No.
1
3
4
1
2
3
4
5
6
7
8
1
2
3
4
5
1
2
3
4
5
6
7
8
2
Feed Gas Composition3
02 NO NH3 H2 CO
W (ppm) (ppn) (ppm) (ppm)
3 1034 000
3 1034 0 1030 0
0.5 960 0 1030 0
3 1029 000
3 1042 768 0 0
3 1033 0 978 0
0.5 1064 0 1003 0
0.5 1053 0 10669 0
3 1038 0 0 1003
0.5 1057 0 0 1003
0.5 1061 0 0 11000
3 1014 000
3 1023 942 0 0
3 1037 0 1000 0
0.5 1119 0 1000 0
0.5 1010 0 11000 0
3 953 0 0 0
3 946 751 0 0
3 969 0 978 0
0.5 1050 0 978 0
0.5 1048 0 10670 0
3 1031 0 0 946
0.5 966 0 0 978
0.5 1014 0 0 11000
3 969 670 0 0
Monitored Gas Constituents
in Reactor Effluentb
NO N02 N20 CO NH3
(ppm) (ppm) (ppm) (ppm) (ppm)
911 0 0 - -
894 0 N.A. - N.A.
746 0 0 - N.A.
895 0 0 - -
876 0 180 - 422
987 0 0 - 0
1046 00-0
658 0 0 - 217
996 0 0 475 0
1034 0 0 793 0
952 0 0 10355 0
1009 0000
959 0 0 0 845
1004 0000
1024 0000
1004 0000
895 0 0 - -
318 0 72 - 215
936 0 0 - 0
1038 00-0
841 0 0 - 0
975 0 0 760 0
974 0 0 826 0
976 0 0 11000 0
352 0 0 0 75
NO
Reduct.
or
Decomp .
(*)
11.9
13.5
22.3
13.0
16.0
4.8
1.7
37.5
4.0
2.2
10.3
0.5
6.2
3.2
0
0.6
6.1
66.4
3.4
1.1
19.8
3.8
0
3.7
63.7
NO
Oxida-
tion
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Excess
NH3
Usage
W
-
-
.
786
_
_
.
.
.
-
127
-
-
-
.
28
-
-
-
-
.
-
45
In addition to the indicated gas components the synthetic flue gas contained 145! C02 and
5% water vapor with the balance being nitrogen.
Effluent hydrogen was not monitored.
-------
their use has been proposed for selective NOX reduction with NH3 and for
nonselective NO reduction with CO. It was, therefore, decided to screen
several of them under the conditions of the screening test matrix.
Table 27 presents the data derived from five mixed iron-chromium oxide
catalysts and one chromium oxide catalyst. The first two Fe-Cr oxide mix-
tures (NA-26 and NA-27) were commercial preparations containing magnesia
and silica in addition to the iron and chromium oxides (Girdler catalysts).
The other three mixed oxide catalysts were prepared at UCLA and were com-
posed of iron and chromium oxides supported on alumina. The chromium oxide
on alumina catalyst was also prepared at UCLA.
The data in Table 27 shows that activity of any significance was exhibited
only during Test No. 2 by this group of catalysts. It is also evident from
the data that catalyst composition and possibly catalyst preparation play a
substantial role in the activity of these catalysts. The catalysts with high
iron oxide content (NA-26, 27 and 30) exhibited low activity in Test No. 2
compared to that shown by NA-28 and NA-29. The 15% chromia on alumina cata-
lyst showed moderate activity during the same test (selective NO-NhU re-
duction). The two UCLA-prepared, low active metal content catalysts (NA-28
and NA-29) exhibited the same activity as the vanadia catalysts for the
selective reduction of NO by NH3 in the presence of 3% oxygen; thus, they
were also selected for parametric and S02 effect studies (Section 4.5). It
should be noted that N20 was not produced on these catalysts and that excess
NH3 utilization was negligible.
Table 28 presents the data from the last group of catalysts screened under
this program. They are three iron oxides on alumina and two iron on graphite
catalysts. The latter two catalysts were investigated because of reported
activity in ammonia synthesis. The iron oxides on alumina were screened in
order to compare data derived from them to that generated on the chromia
catalyst (NA-31) and on the mixed Fe-Cr oxide catalysts (NA-28 and NA-29).
The iron oxide catalysts exhibited lower activity than the mixed Fe-Cr oxide
catalysts under Test No. 2. conditions. As indicated above, the chromia on
136
-------
Table 27. CATALYST SCREENING TEST RESULTS (IRON-CHROMIUM OXIDE CATALYSTS)
Catalyst
No. and
Type
NA-26
(Girdler
G3A)
8Q% Fe203,
7% Cr2&3,
balance
MgO, Si02
and
graphite
NA-27
(Girdler
Fe-Cr
similar
to NA-26)
Test
No.
1
2
3
4
5
6
7
8
2
Feed Gas Composition9
02 NO NH3 H2 CO
(%) (ppm) (ppm) (ppm) (ppm)
3 1029 000
3 1029 683 0 0
3 1029 0 1029 0
0.5 1029 0 1048 0
0.5 1029 0 11000 0
3 1029 0 0 1030
0.5 1029 0 0 1030
0.5 1029 0 0 11000
3 1004 693 0 0
Monitored Outlet.
Gas Composition
NO N02 N20 CO NH3
(ppm) (ppm) (ppm) (ppm) (ppm)
953 0 0 - -
850 0 400 - 95
980 0 0 0
998 0 0 0
958 0 0 0
1031 0 0 654 0
1018 0 0 697 0
991 0 0 5400 0
977 0 334 0 86
NO
Reduct.
or
Decomp.
7.4
17.4
4.8
3.0
6.9
0
1.0
3.7
2.7
NO
Oxida-
tion
0
0
0
0
0
0
0
0
0
Excess
NH3
Usage
394
_
-
_
-
-
-
3270
CO
-------
Table 27. (Continued) CATALYST SCREENING TEST RESULTS (IRON-CHROMIUM OXIDE CATALYSTS)
Catalyst
No. and
Type
NA-28 (10%
Fe2°3'Cr2°3
on A1203
NA-29 (NA-28
prepared
from
different
salts)
Test
No.
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
Feed Gas Composition3
02 NO NH3 H2 CO
(X) (ppm) (ppm) (ppm) (ppm)
3 960 0 0 0
3 940 767 0 0
3 947 0 954 0
0.5 936 0 954 0
0.5 935 0 10355 0
3 960 0 0 995
0.5 960 0 0 995
0.5 992 0 0 10059
3 931 0 0 0
3 954 780 0 0
3 1016 0 1030 0
0.5 996 0 1030 0
0.5 985 0 10340 0
3 1026 0 0 995
0.5 1002 0 0 995
0.5 1002 0 0 10340
Monitored Outlet.
Gas Composition
NO N02 N20 CO NH3
(ppm) (ppm) (ppm) (ppm) (ppm)
874 0 0 00
309 0 0 0 79
909 0 0 00
911 0 0 00
867 0 0 00
924 0 0 898 0
941 00 888 0
977 0 0 8835 0
863 0 0 00
366 0 0 0 461
986 0 0 00
979 0 0 00
886 0 0 0 0
984 0 0 774 0
988 0 0 811 0
964 0 0 10025 0
NO
Reduct.
or
Decomp.
(X)
8.9
67.1
4.0
2.7
7.3
3.8
2.0
1.5
7.3
61.6
3.0
1.7
1.0
4.1
1.4
3.8
NO
Oxida-
tion
(X)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Excess
NH3
Usage
(X)
63
_
-
-
-
-
-
mi
-18
_
-
-
-
_
-
CO
<£>
-------
Table 27. (Continued) CATALYST SCREENING TEST RESULTS (IRON-CHROMIUM OXIDE CATALYSTS)
Catalyst
No. and
Type
NA-30
(56.7%
Fe-,0,,
2 3
6.7%
Cr203,
1.4%
A12°3
Pressed
NA-31
(15%
(>203 on
Pressed
Test
No.
1
2
3
4
5
6
7
8
2
Feed Gas Composition
02 NO NH3 H2 CO
(%) (ppm) (ppm) (ppm) (ppm)
3 958 0 0 0
3 1022 800 0 0
3 1018 0 978 0
0.5 998 0 978 0
0.5 1007 0 10059 0
3 1064 000
0.5 1015 000
0.5 1015 000
CHROMIUM OXIDE CAT/
3 1018 598 0 0
Monitored Outlet.
Gas Composition
NO N02 N20 CO NH3
(ppm) (ppm) (ppm) (ppm) (ppm)
885 0 0 00
693 0 105 0 296
960 0 0 00
979 0 0 00
969 0 0 00
1017 0 0 978 0
1004 0 0 978 0
997 0 0 10059 0
\LYST
642 0 106 0 113
NO
Reduct.
or
Decomp.
7.6
32.2
5.7
1.9
3.8
4.4
1.1
4.8
36.9
NO
Oxida-
tion
0
0
0
0
0
0
0
0
0
Excess
NH3
Usage
(%)
130
-
-
_
-
-
-
93
In addition to the indicated gas components the synthetic flue gas contained 14% C02 and 5% water vapor
with the balance being nitrogen.
Effluent hydrogen was not monitored.
-------
Table 28. CATALYST SCREENING TEST RESULTS (IRON OXIDE
AND IRON GRAPHITE CATALYSTS)
Catalyst
No. and
Type
HA" 32
(15J Fe203
on A1203)
pressed
M-33 (15%
Fe203 on
A1203)
Impregnated
M-34 (20%
Fe2°3
on A1203)
Harshaw
H-35 (Iron
on
graphite)
M-35 (Iron
on
graphite)
Test
No.
2
1
2
2
1
2
1
lc
2
2C
2C
3
4
5
6
7
8
1
2
3
4
5
6
7
8
Feed Gas Composition3
02 NO NH3 H2 CO
(%) (ppm) (ppm) (ppm) (ppm)
3 1015 675 0 0
3 1046 000
3 988 756 0 0
3 1007 940 0 0
3 1018 000
3 975 852 0 0
3 1003 000
3 1091 000
3 975 698 0 0
3 1041 1040 0 0
3 1037 852 0 0
3 1036 0 978 0
0.5 997 0 1067 0
0.5 989 0 11000 0
3 1036 0 0 995
0.5 1006 0 0 1030
0.5 997 0 0 10355
3 1015 000
3 1015 748 0 0
3 1034 0 962 0
0.5 988 0 962 0
0.5 1021 0 11000 0
3 1051 0 0 997
0.5 1007 0 0 997
0.5 1015 0 0 11220
Monitored Outlet
Gas Composition b
NO N02 N20 CO NH3
(ppm) (ppm) (ppm) (ppm) (ppm;
551 0 00 309
993 53 0 0 0
502 0 00 275
480 0 00 595
1011 0000
822 0 00 753
1013 0 0 - -
1103 0 0 - -
890 0 0 - 66?
1010 0 0-771
1034 0 0-826
1023 00-0
817 0 0 79
852 0 0 - N.A.
1037 0 0 825 -
1031 0 0 980 -
963 0 0 10250
990 0 0 - -
956 0 0 - 738
1027 00-0
988 0 0-0
901 0 0 - 30
1041 0 0 423 -
966 0 0 793 -
872 0 0 9662
NO
Reduct.
or
Decomp.
W
45.7
0
49.2
52.3
0.7
15.7
0
0
12.0
3.0
0.3
1.2
18.0
13.9
0
0
3.3
2.5
5.8
0.7
0
11.8
1.0
1.1
14.1
NO
Oxida-
tion
<«)
0
5.1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Excess
NH3
Usage
(«)
18
48
-2
-3
.
-
632?
269
26
-
-
-
-
-
-
_
0
-
-
-
-
-
-
In addition to the indicated gas components the synthetic flue gas contained 14% C0? and 5% water
vapor with the balance being nitrogen.
Effluent hydrogen was not monitored.
C 14,000 hr"1 (STP) space velocity (nominal 20,000 hr"1)
143
-------
alumina catalyst had also shown lower activity than the mixed oxides for
the same reaction. The tentative conclusion appears to be that a synergistic
effect takes place when the two oxides are mixed. However, additional data
is required on the effect of catalyst composition on its activity before the
above conclusion becomes unequivocal. The exhibited activity by chromia and
by the iron oxide catalysts toward the selective NO reduction by NH3 must be
considered promising. These catalysts were not investigated further.
The iron on graphite catalysts were in all respects inferior to the other
iron based catalysts investigated in this program. The small activity to-
ward the selective reduction of NO with H2 (Test No. 4) exhibited by NA-35
does not appear real in view of the data generated under Test No. 6 (non-
selective reduction).
4.5 PARAMETRIC INVESTIGATIONS ON PLATINUM, VANADIA, AND IRON-CHROMIUM
OXIDE CATALYSTS
Several catalysts emerged from the screening tests (400°C, 20,000 hr" ) as
potentially useful in NO abatement in power plants and, therefore, meriting
A
additional investigation. They were as follows:
• NO decomposition: Pt on alumina; Pt-Mo on alumina, and
possibly W03.
• NO oxidation: possibly GdV03 on alumina.
• Selective NO reduction
(a) With NH3: Pt (250°C), V^, GdVO-j, and Fe^-Cr^
on alumina (high activity); Pt-Mo, Mo, lead pro-
moted CuO, Fe^O., on alumina (substantial activity;
rare earth-cobalt oxide, CuO, and Cr203 on alumina
(medium activity).
(b) With H2 or CO: none.
144
-------
• Nonselective NO reduction with H2 and CO: high activity
Pt and Pt-Mo on alumina; several catalysts exhibited
low to medium activity for the H2~NO reaction.
Program schedule and budget considerations necessitated the selection of
only three of the above catalysts for parametric and S02 effect investi-
gation. These were Pt, V205, and Fe203'Cr203.
Platinum was selected principally for NO decomposition studies and for the
S02 effect on its activity toward the NO-NH., reaction. However, some para-
metric investigations were performed on the NH.,-NO reaction on Pt in order
to generate baseline data on this process and to probe N20 production.
The vanadia on alumina catalyst was selected for its high activity in the
promotion of the NO-NH3 reaction in the presence of oxygen.
The iron-chromium oxide mixture catalyst was selected for the same reactions
as the vanadia catalyst.
4.5.1 Platinum Catalysts
As indicated earlier, 45 and 27% NO decomposition was obtained on two batches
of presumably identical composition 0.5% Pt on alumina Engelhard catalyst
(NA-1 and NA-2) when an S09-free synthetic power plant flue gas was passed
^ _1
over the catalyst at 400°C and 20,000 hr (STP) space velocity. Even
though the two NO conversion values differ substantially, both values were
considered as indicating promise for use in NO abatement by the most
A
desirable method (decomposition). Thus, the NO decomposition reaction on
Pt was investigated at several temperatures and at two space velocities.
The generated data are presented in Table 29 and Figure 20. The following
tentative conclusions can be drawn from these data:
• Decomposition of NO as a function of temperature goes
through a maximum at or near 300°C.
145
-------
Table 29. DECOMPOSITION OF NO ON Pt CATALYST (NA-2)
Inlet Gas
Composition
NO 02
(ppm) (%)
1003
1023
1023
1091
1091
1091
1089
1089
1107
1000
1000
1054
1080
3
3
3
3
3
3
3
3
3
3
3
3
3
Space
Velocity
(hr"1 STP)
10,000
10,000
10,000
10,000
10,000
10,000
20,000
20,000
20,000
20,000
20,000
20,000
20,000
Temperature
(°C)
400
400
350
300
250
308
305
350
400
255
304
356
400
Outlet Gas
NO
(ppm)
620
635
553
518
636
520
608
708
817
646
584
667
783
Composition
N20
(ppm)
138
33
0
0
0
0
0
0
0
0
0
0
0
Decomposition
of NO
(*)
38.2
37.9
49.5
52.5
41.7
52.3
44.2
35.0
26.2
35.4
41.6
36.7
27.5
Feed also contained 14% C02 and 5% H20 in nitrogen.
-------
60
540
O 20
Q.
2E
O
o
UJ
O- 20,000 hr''S.V.
• - 10,000 hr"!S.V
200
250 300 350
TEMPERATURE (°C)
400
Figure 20. Decomposition of NO on Pt Catalyst (NA-2)
149
-------
• NO decomposition increases with decreasing space
velocity in the range of 10,000 to 20,000 hr"1 (STP).
• NO decomposition decreases with decreasing oxygen
concentration in the flue gas in the range of 0.5
to 3% oxygen (single point data, Table 22).
• The highest NO decomposition, 52.5%, was attained
at 300°C and 10,000 hr"1 (STP).
The data appears to warrant the conclusion that Pt on alumina has exhibited
sufficient NO decomposition potential to be seriously considered for power
plant utilization, especially when complete NO abatement is not needed
A
It should, however, be pointed out that this conclusion is based on small
scale, short time testing; thus, additional experimental work is needed.
The S02 effect should also be examined, although indirect evidence from
the S02 effect experiments on the NH3-NO-Pt process indicates that this
flue gas constituent may not be a problem above 300°C.
Table 30 presents the data generated from experiments probing the effects
of temperature, NH3/NO ratio, NO and S02 concentrations, and space velocity
on the activity of platinum in the selective reduction of NO by NH3. The
tests were initiated on the NA-1 catalyst (Engelhard 0.5Pt on alumina), but
were switched to the NA-2 sample (same composition) when NA-1 was acciden-
tally lost. Two samples of the latter catalyst were used. The data is
presented in the sequence generated.
The data generated on the NA-1 catalyst at 250°C indicate that an immediate
drop (approximately 15%) in catalytic activity occurs when 1000 ppm S02 is
introduced to a flue gas containing near stochiometric quantities of ammonia
(870 ppm) with respect to NO present. Increase of the NH3-to-NO ratio
(1110 ppm NH3) restored the catalytic activity. The data appear to suggest
that the observed decrease in NO reduction in Run 2 was due to NH, depletion
150
-------
Table 30. PARAMETRIC EFFECTS ON PLATINUM CATALYSTS USED IN THE SELECTIVE REDUCTION OF NO BY NH-
en
Run
1
2
3
4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1
2
3
4
5
6
7
8
9
10
n
12
13
14
15
Temp.
(°c)
250
250
250
250
255
200
262
300
250
212
209
209
209
260
300
400
400
250
205
250
250
250
250
205
205
257
205
203
247
247
250
250
250
250
Inlet Gas Composition8
NO NH3 S02
(ppm) (ppm) (ppm)
NA-1 CATALYST
1016 870 0
1062 870 1000
1055 1110 1000
1004 1110 0
NA-2 CATALYST
976 814 0
971 814 0
1047 846 0
1047 846 0
1046 846 0
1046 846 0
993 799 0
1043 783 1000
1043 783 1000
1043 783 1000
1043 783 1000
1043 783 1000
988 858 0
988 858 0
988 858 0
NA-2 CATALYST
1046 797 0
1046 754 0
1046 978 0
1054 1302 0
1054 1302 0
1045 773 0
1051 771 0
1051 771 0
1051 1027 0
1051 1027 0
1051 1128 0
449 491 0
449 575 0
1058 1094 0
1003 1232 0
Outlet Gas Composition
NO NgO NH3 S02
(ppm) (ppm) (ppm) (ppm)
271 571 N.A.b N.A.
385 490 N.A. N.A.
280 670 N.A. N.A.
256 670 N.A. N.A.
218 665 ' 0
200 652 0
290 625 0
516 558 0
358 765 0
290 N.A. 0
238 689 0
663 278 345 80
804 318 389 320
537 415 108 450
516 458 N.A. 520
860 347 13 0
891 290 0
292 682 0
184 808 0
285 .654 0 -
456 389 0
278 803 0
202 912 0
86 1405 0
304 734 0
304 591 0
230 700 0
87 872 0
205 672 0
176 717 0
87 187 0
79 266 0
129 736 0
119 787 0
NH3/NO
Inlet
Ratio
0.86
0.82-
1.05
1.11
0.83
0.84
0.81
0.81
0.81
0.81
0.80
0.75
0.75
0.75
0.75
0.75
0.87
0.87
0.87
0.76
0.55
0.93
1.24
1.24
0.74
0.73
0.73
0.98
0.98
1.07
1.09
1.28
1.03
1.23
Conv.
of NO
(*)
73.3
63.7
73.5
74.5
77.7
79.4
73.2
50.7
65.8
72.2
76.0
36.4
22.9
48.5
50.5
17.5
9.8
70.4
81.4
72.7
56.4
73.4
80.8
91.8
70.9
71.0
78.0
91.7
80.5
83.3
80.6
82.4
87.8
88.1
Space Velocity
(Mr-1 x ID"3)
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
10
10
S02 Exposure
Time, Hours
.
1
2
-
.
_
_
_
-
.
_
1
2
3
4
5
0
_
-
_
-
.
-
-
-
.
_.
_
-
-
_
_
-
-
Inlet gas also contained 5% ^0 and 14% C02 in NZ; 14 grams catalyst was used.
NA = not available (data was not or could not be taken).
Catalyst heated to 400°C for three hours prior to this test.
-------
Table 30. (Continued) PARAMETRIC EFFECTS ON PLATINUM CATALYSTS USED IN THE SELECTIVE REDUCTION OF NO BY NH,
Run
16
17
18
19
20
21
22
23
24C
25
26
27
28
29
30
31
32
33
34
35
Temp.
(°C)
250
250
250
250
250
250
250
250
255
255
255
255
255
255
243
243
250
250
250
250
Inlet Gas Composition9
NO NHa S0£
(ppm) (ppm) (ppm)
1003 1232 1200
1003 1232 1200
1083 1261 1200
1064 1261 1200
1064 1261 0
1064 1261 0
989 1189 0
989 1189 0
893 1250 0
961 1514 0
503 886 0
259 304 0
259 520 0
259 600 0
272 497 0
277 568 0
809 839 0
809 810 0
809 880 0
811 1230 0
Outlet Gas Composition
NO N20 NH3 S02
(ppm) (ppm) (ppm) (ppm)
329 N.A. N.A. 76
296 N.A. 55 200
296 N.A. 54 296
633 N.A. 600 N.A.
198 N.A. N.A.
105 N.A. 120
124 787 86
105 879 N.A.
116 796 0
141 893 42
78 363 N.A.
57 131 0
29 195 0
29 N.A. 0
41 180 0
41 180 0
251 304 0
210 418 0
186 442 0
131 508 0
NH3/NO
Inlet
Ratio
1.23
1.23
1.16
1.19
1.19
1.19
1.20
1.20
1.39
1.58
1.76
1.17
2.01
2.32
1.83
2.05
0.91
1.00
1.09
1.52
Conv.
of NO
(%)
67.1
70.5
72.7
40.5
81.4
90.1
87.5
89.4
87.0
85.3
84.4
77.9
88.8
88.8
84.9
84.9
68.9
74.0
77.0
83.8
Space Velocity
(HH x 10-3)
10
10
10
10
10
10
10
10
10
20
20
20
20
20
20
20
20
20
20
20
S02 Exposure
Time, Hours
0.5
1.0
1.5
19.5
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Ol
CO
Inlet gas also contained 5% H?0 and 14% COp in hL; 14 grams catalyst was used.
NA = not available (data was not or could not be taken).
Catalyst heated to 400°C for three hours prior to this test.
-------
rather than loss of catalytic activity.* Thus, short time catalyst ex-
posure to S02 at 250°C did not prove detrimental to its activity.
At 209°C the drop in NO conversion due to the presence of 1000 ppm S02 was
also immediate and much more severe than at 250°C (NA-2, Sample A, Run 7
versus Run 8). NO conversion dropped from 76% to 36.4% after one hour
exposure to S02 (Run 8) and to 22.9% after two hours (Run 9). Increasing
the temperature to 400°C in three steps increased NO conversion until, at
temperatures above 300°C, NH3 oxidation by oxygen took over (Runs 10 through
12). Subsequent to removal of the S02 from the flue gas and temperature
reduction to 250°C (Run 14) and to 205°C (Run 15) NO conversion was fully
restored. In these experiments the NH3 in the reactor effluent gas was
measured. The data indicate that ammonia depletion was not the reason for
the observed severe drop in NO conversion; it must have, therefore, been
due to loss of catalytic activity which was fully restored by increasing
the temperature to 400°C. The loss of catalytic activity for NO reduction
may have been due to a competing reaction (e.g., S02 oxidation) or to
physical blockage of catalyst active sites by a salt deposition on the
catalyst. Evidence that either or both of these mechanisms of Pt poison-
ing could have occurred was furnished by the observed substantial S03
evolution during the temperature excursion to 400°C.
Longer-term S02 effect studies were performed on the second NA-2 sample
(Sample B). In these experiments large excess of NH~ and lower space
velocity was used. As it was observed with NA-1, short-term (1.5 hours)
catalyst exposure to S02 had no apparent effect on the activity of NA-2
at 250°C, provided excess NH3 was present (Runs 16, 17, and 18); longer-
term exposures, however, caused severe drop in activity (19.5 hours).
Catalyst poisoning was again reversible (Run 20).
The above data clearly indicates that platinum is not recommended as a
promoter for the NO-NH3 selective reduction reaction if the flue gas to be
At the time, NH3 determination in the effluent in the presence of SO;?
could not be made. Later, a specific ion electrode for NHs was acquired
to perform these measurements.
154
-------
treated contains S02 in quantities of 1000 ppm or more. The NO conversion
dip because of S02 is immediate and severe in the temperature range where
Pt is effective for NO reduction with NH3; the required frequency of cata-
lyst regeneration could be impractically high. It may be technically feasi-
ble to utilize Pt to promote the above reaction with flue gas containing
only a few ppm S02 (1-10 ppm) because the frequency of regeneration would
be lower. It should be noted, however, that a second catalytic reactor
would be required which may render the process economically unfeasible. The
true upper limit of S02 catalyst tolerance should be determined experimentally
if economics and Pt availability permit the use of a second reactor. The
planned tests at Environics (Dec. 1974) with 0.5% sulfur distillate oil
(approx. 200 ppm S02 in the flue gas) should shed additional light on this
subject.
Table 30 summarizes also the data generated by the variation of additional
process parameters listed earlier (NH3/NO, No concentration, temperature and
space velocity). The following conclusions were drawn from these data:
• Temperature had a pronounced effect on NO reduction by
NH3 on Pt. Optimum conversion occurred in the 200-250°C
range; NO conversion dropped off appreciably on both sides
of this range (data obtained at temperatures below 200°C
are not shown in Table 30).
§ NO concentration in the range of 250 to 1000 ppm had
little or no effect on NO conversion to N2<
• The value of the NHo-to-NO ratio had a pronounced effect
on NO reduction in the range of 0.67 (stoichiometric
value) to 1.2; at values higher than 1.2 the effect
appeared to level off.
t Space velocity in the 5,000 to 20,000 hr'1 (STP) range
had only a minor effect, if any, on NO reduction.
The NhL-to-NO ratio, NO concentration, and space velocity effects are il-
lustrated in Figure 21.
155
-------
100
? 80
fe 60
o
CO
UJ
o
o
40
20
0.4
0.8
0
O-IOOO
• - 500
O-IOOO
O- 250
A- 800
SPACE VELOCITY
(hr1, STP)
20000
20000
10000
20000
20000
1.2
1.6
2.0
NH3 /NO RATIO
2.4
Figure 21. Reduction of NO with NH3 on Pt catalyst at 250°C,
(14% C02, 5% H20, 3% 02 in NZ)
157
-------
It was indicated in the previous section that a substantial quantity
of the NO reduced by NH3 on Pt catalysts was apparently converted to
N20 instead of Np- The parametric data verified ^0 production and under
certain conditions indicated that the NpO produced exceeded NO consumption.
A number of experiments were performed to investigate N^O production and
the effect of oxygen on it. The generated data and inferences drawn from
them are presented in Appendix A of this report. The important conclu-
sion drawn from the data is that NpO production on Pt can not be avoided
as long as oxygen is present in the flue gas. In addition to oxygen, N20
production was influenced by temperature and the NH3-to-NO ratio.
4.5.2 Vanadia Catalysts
The catalyst screening tests (Section 4.4) revealed that both samples of
commercial vanadia catalysts (NA-24, Filtrol and NA-25, Harshaw) appeared
to be equivalent in activity for the selective reduction of NO with NH3.
Better than 60% NO conversion was attained on both catalysts at 400°C
with no N20 production and with insignificant excess NH3 consumption. At
400°C these catalysts performed substantially better than Pt in the selective
reduction of NO with NH3. Thus, they were ranked top candidates, along with
the Fe-Cr oxide catalysts, for further investigation.
Initially, both the Filtrol and Harshaw catalysts were subjected to para-
metric investigation, but inconsistencies in activity and physical in-
stability eliminated the Filtrol catalyst from further testing. The Harshaw
vanadia (NA-25) was subjected to temperature, oxygen and NO concentration,
NH3-to-NO ratio, space velocity, and S02 effect studies and proved a very
promising candidate for utilization as a promoter of the NH,-NO reaction
<3
in the presence of oxygen and S02.
Figure 22 presents data on the temperature and oxygen concentration effects
generated with dry, C02-free flue gas on NA-25 at 20,000 hr"1 (STP). Water
vapor and C02 were not used in these tests in order to separate the oxygen
effect from that of H20 and C02; the flue gas was also SOg-free. The data
in Figure 22 indicate that NO reduction increases with increasing oxygen
158
-------
100 _
o
Q
LU
c:
20 .
200
Coneentration (ppm)
300 400
TEMPERATURE (°C)
500
NO
1000
1000
1000
1000
NH3
1190
1190
1190
1190
°2
1000
3000
9100
0
Figure 22. Effect of 0? Concentration and Temperature on
NO Reduction with NH3 on V205 Catalyst (Harshaw).
159
-------
concentration, but that this effect diminishes as the oxygen approaches
typical flue gas oxygen concentration values and as the temperature reaches
or exceeds the optimum reaction temperature. The data also indicates that
the optimum temperature range for NO reduction by NH3 on vanadia is 350-
400°C.
The same optimum reaction temperature range was observed when the full flue
gas was used (H20, C02 and S02 present), but NO reduction was slightly
lower (70 versus 80%) at the same NH3-to-NO ratio when water vapor and C02
were present.
Table 31 summarizes some of the data on temperature, space velocity, and
short-term S02 effects on the activity of vanadia for the selective NO re-
duction with NH3. Table 32 presents additional space velocity effect
data as well as data generated on the effects of NO and NH3 concentration.
The following conclusions can be drawn from these data concerning the
selective NO reduction with ammonia on the Harshaw V,,0,- catalyst (NA-25):
c. 3
0 The optimum reaction temperature appears to be at or
near 400°C (Runs 1 through 5, Table 31).
• NO reduction increased with decreasing space velocity
in the range of 20,000 hr"1 to 10,000 hr"1 (STP);
additional space velocity reduction, e.g., to 5,000
hr~ , did not appear to influence NO reduction. The
space velocity effect was observed at all NO concen-
trations and NH3-to-NO ratios used (Run 7, Table 31
indicated abnormally low NO conversion when compared
to data obtained with similar NH3-to-NO ratios at
10,000 hr"1, Table 32).
• NO concentration in the range of 250 to 1000 ppm had
little or no effect on NO reduction to N2 (Table 32).
160
-------
cr>
Table 31. TEMPERATURE, SPACE VELOCITY AND SHORT-TERM S02 EFFECTS ON THE
REDUCTION OF NO WITH NHg ON HARSHAW VANADIA INA-25)
Run
No.
1
2
3
4
5
6
7b
8C
9
10
11
12
13
14
15
16
Temperature
Inlet Gas Composition3
NO NHs S02
(ppm) (ppm) (ppm)
A. TEMPERATURE AND SPACE VELOCITY EFFE
400
310
445
485
400
400
400
400
969 670 0
969 670 0
969 670 0
969 670 0
966 650 0
960 982 0
973 944 0
975 950 0
B. S00 EFFECT:
400
400
400
400
400
400
400
400
979 668 0
979 668 1000d
979 410 1000
992 860 1000
928 720 0
936 720 1000
936 1060 1000
936 482 1000
Outlet Gas Composition
NO N20 NHa
(ppm) (ppm) (ppm)
CTS:
352 0 75
464 0 146
352 0 87
364 0 35
390 0 87
325 0 168
241 0 28
259 0 124
352 0
352 0
532 0
311 0
380 0
380 0
287 0
508 0
NO
Reduction
(X)
63.7
52.1
63.7
62.4
59.6
66.1
75.2
73.4
64.0
64.0
45.7
68.6
59.1
59.4
69.3
45.7
Excess NHq
Usage J
45
56
42
57
47
92
88
73
-
-
-
-
-
-
Feed gas also contains 5% H20, 14% C02, 3% 02> in N,:.
-1
Space velocity was 10,000 hr (STP) rather than the nominal 20,000 hr
(STP) rather th
Each S02 test was of at least 1 hour duration.
-1
Space velocity was 15,000 hr (STP) rather than the nominal 20,000 hr"1
-------
Table 32. EFFECT OF NO AND NH3 CONCENTRATION AND SPACE VELOCITY ON THE
REDUCTION OF NO WITH NH3 ON HARSHAW V205 CATALYST (NA-25)
Run
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Inlet Gas Composition9
NO NHs
(ppm) (ppm)
340 320
924 1048
268 285
281 181
479 336
474 380
517 483
951 1370
973 1224
285 154
258 154
258 297
1078 827
1128 983
Space
Velocity
Hr-1 x
TO-3 (STP)
5
10
10
10
10
10
10
20
20
20
20
20
5
5
Outlet Gas Composition
NO N20 NH3
(ppm) (ppm) (ppm)
00 0
20 0 123
00 0
103 0 0
180 0 19
91 0 N.A.
00 26
259 0 501
296 0 373
157 0 0
135 0 0
87 0 N.A.
261 0 N.A.
126 0 N.A.
Conv.
of NO
(*)
100
97.8
100
63.3
62.4
80.8
100
72.8
69.6
44.9
40.3
66.2
75.8
88.8
Catalyst
Weight
(gm)
28
28
28
28
28
28
28
14
14
14
14
14
14
14
Excess NH,
Usage *
(%}
41
54
50
52
59
N.A.
32
89
112
81
88
N.A.
N.A.
N.A.
CO
Inlet gas also contains 14% C02, 5% H20, 3% 02 in N2; 400°C (752°F),
-------
a The value of the NFL-to-NO ratio in the range of 0.5 to
1.2 had a dramatic effect on the extent of NO reduction;
the latter increased with increasing value of this ratio
approximately linearly at low space velocities. This
effect leveled off at ratio values exceeding 1.2.
• Virtually complete NO reduction to N2 was attained at
400°C, 10,000 hr'1 (STP), and an NH3-to-NO ratio of 1.2.
• The short-term S0« effect on NO conversion was negligible
or nonexistent (Table 31, Runs 9 through 16).
Figure 23 illustrates most of the above effects and the justification of
the conclusions drawn.
Long term SOp effects were investigated at low space velocities in order to
reduce synthetic flue gas consumption. The same sample of catalyst (NA-25)
was subjected to SOp containing flue gas (1200 to 1500 ppm) for over 50 hours.
No detectable catalytic activity deterioration was observed during these
tests. The data is summarized in Table 33.
4.5.3 Iron-Chromium Oxide Catalysts
Several iron-chromium oxide catalysts were screened for activity in the
selective reduction of NO by NH.,. By far the most promising proved to be
two UCLA catalysts prepared by impregnation of alumina with aqueous solu-
tions of iron and chromium nitrates and from solutions of iron nitrate and
chromium trioxide, respectively. Whether their superiority over similar
commercial catalysts was due to composition differences or to method of
preparation is not known.
Under screening tests conditions (400°C, 20,000 hr ) these catalysts (NA-28
and NA-29) promoted NO reduction by nearly 70%. This extent of NO conversion
represented the highest value attained on any of the 36 catalysts screened.
Thus, NA-28 and NA-29 were ranked as the prime candidates for additional
testing.
164
-------
100
80
60
SPACE VELOCITY
Hr"1 (STP)
40
s
§
o
20 -
d °
O
Q
€
O
O
0
•
o
1 1
1000
1000
1000
1000
1000
1000
50O
350
250-300
250
i
5000 2
IOOOO 2
20000 1
15000
5000
IOOOO
IOOOO J
5000 ;
IOOOO 2
20000 1
1 1 — 1
0.6
OB
1.0
1.2
NH3/NO RATIO
CATALYST WT.,
GRAMS
28
28
14
14
14
14
28
28
28
14
14
Figure 23. Reduction of NO With NH3 on V205 Catalyst at
400°C; 14% C02, 5% H20, 3% 02 in N2)
165
-------
Table 33. LONG-TERM S02 EFFECT ON THE CATALYTIC ACTIVITY
VANADIA FOR THE SELECTIVE NO-NH3 REACTION
OF
Run
No.
1
2
3b
4C
5
6
7
8
9
10
lld
12
13e
Inlet Gas Composition9
NO NHs S02
(ppm) (ppm) (ppm)
1084 946 0
1084 946 1500
1175 1040 1500
1130 840 1500
1130 840 0
1040 698 0
1014 847 0
1018 1096 0
1016 881 0
1043 841 0
1043 841 1200
1026 965 0
1040 1035 1200
Space
Velocity
Hr'1 x 10~5 (STP)
5
5
5
5
5
10
10
5
5
5
5
5
5
Outlet Gas Composition
NO N20 NHs
(ppm) (ppm) (ppm)
52 0 0
99 0 0
166 0 0
152 0 0
131 0 0
332 0 0
221 53 0
8 62 66
155 0 0
185 0 0
185 0 0
28 0 0
38 0 0
Conv.
of NO
95.2
90.9
85.9
86.5
88.4
68.1
78.2
99.2
84.7
82.3
82.3
97.3
96.3
Excess NH3
Usage
38
44
55
29
26
48
50
54
53
47
47
45
55
b
c
d
e
Inlet gas also contained 3% 0?, 5% H?, 14% C0? in N?. All runs were performed at 400°C.
28 gm catalyst used.
4 hour test (normally data points were taken after one hour of steady-state operation).
6 hour test.
19 hour test.
25 hour test.
-------
The parametric investigations performed on these catalysts involved tem-
perature, NO concentration, NH3-to-NO ratio, space velocity, and S02 effects
on NO reduction by NH3; long-term S02 effects were performed only on NA-28.
The data are summarized in Tables 34 and 35.
The data in these tables indicate that the two Fe-Cr oxide catalysts showed
the same activity for the selective NO reduction and the same sensitivity,
or insensitivity, to important reaction parameters. The data generated on
the Fe-Cr catalysts is also practically identical to that generated on
vanadia. Thus, the conclusions drawn on the vanadia apply to these cata-
lysts also. Important parameters to NO conversion were: (a) temperature,
with the optimum value being near 400°C, (b) NH^-to-NO ratio, with its
optimum value being between 1.0 and 1.2, and (c) space velocity, with an
optimum value of about 10,000 hr (STP).
The NA-28 catalyst was exposed for over 70 hours to SO^ containing flue gas
(1,000 to 1,500 ppm) under a number of different experimental conditions
with no discernable effect on its activity for the NFU-NO reaction (Table
35).
Figure 24 presents the 400°C parametric study data on these catalysts and
graphically illustrates the similarity of these catalysts. Furthermore,
comparison of Figures 23 and 24 clearly shows the similarity between the
Fe-Cr oxide and vanadia catalysts. As was the case with vanadia, these
catalysts did not promote N^O production either.
Their was one exception to the apparent complete similarity between these
two catalysts and vanadia which could be potentially significant. The NA-28
catalyst at low space velocities (10,000 hr" or lower) appeared to com-
pletely decompose (or oxidize) the excess ammonia in the system, regardless
of the value of the NFL-to-NO ratio. This can be interpreted as a dis-
advantage because excess ammonia consumption is higher than with NA-29 or
vanadia where most of the ammonia not used by the NO is present in the
reactor effluent gas. To be of value, however, the ammonia must be re-
covered from the flue gas stream and such recovery is highly unlikely to
168
-------
Table 34. PARAMETRIC INVESTIGATIONS ON Fe-Cr OXIDE CATALYSTS
EMPLOYED IN SELECTIVE REDUCTION OF NO WITH NH.
UD
Run
No.
1
2
3
4
5
6
7
8
9
10
11
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Inlet Gas Composition3
NO NH3
(ppm) (ppm)
CATALYST NA-28
931 915
952 553
870 1093
879 759
879 759
907 942
907 942
907 942
907 890
893 1110
966 900
CATALYST NA-29
915 1107
915 1107
915 1107
945 980
970 1147
519 693
266 337
998 730
967 863
979 635
965 1081
886 1107
1004 725
1004 1120
Space Velocity
(Mr-1 x 10-3 (STP)
20
20
20
20
20
20
20
20
20
10
20
10
10
10
20
20
20
20
20
10
10
10
10
20
20
Outlet Gas Composition
NO NH3 N20
(ppm) (ppm) (ppm)
223 97 0
529 68 0
228 352 0
446 370 0
323 112 0
223 119 0
318 257 0
233 0 0
372 0 0
46 0 0
200 0 N.A.
554 871 0
90 286 0
1 24 39 0
256 159 0
242 237 0
119 N.A. 0
60 73 0
326 189 0
128 158 0
261 124 0
95 312 0
68 203 0
320 N.A. 0
190 N.A. 0
Conv.
of NO
(%)
76.0
44.4
73.8
49.2
63.3
75.4
64.9
74.3
59.0
94.8
76.6
39.5
90.2
86.4
72.9
75.1
77.1
77.4
67.3
86.8
73.3
90.2
92.3
68.1
81.1
Excess NH3
Usage
(*)
73
72
73
35
75
80
74
110
150
97
N.A.
-2
49
103
79
87
N.A.
93
21
26
7
33
66
N.A.
N.A.
Reaction
Temperature
(°C)
400
400
400
325
400
400
350
450
500
400
400
260
350
450
400
400
400
400
400
400
400
400
400
400
400
Inlet gas also contains 14% C02> 3% 02, 5% H20 in N2-
-------
Table 35. REDUCTION OF NO WITH NH3 ON Fe-Cr CATALYST
(NA-28) IN THE PRESENCE OF S02
Run
No.
1
2
3
4
5
h
6D
7
8
9
10C
11
12
13
14d
15
16e
17f
189
19h
20 1
21J
22
Inlet
NU
(ppm)
1041
1055
1055
1055
929
969
981
981
969
1040
996
1015
1015
1061
1089
1089
982
977
977
907
907
940
Gas Composition3
NH3
(ppm)
887
1231
987
770
915
915
938
1129
673
960
858
1012
898
898
740
740
840
840
840
868
868
767
bU2
(ppm)
0
0
0
0
0
1200
0
0
0
1200
0
0
0
1000
0
1400
1400
1400
1400
0
1500
0
Space
Velocity
(hr'T STP)
10
5
5
5
5
5
5
5
5
5
10
15
15
15
5
5
5
5
5
5
5
20
Outlet
NU
(ppm)
176
18
86
245
95
99
57
0
325
100
132
77
170
168
220
241
156
174
172
34
57
309
Gas Composition
NH3
(ppm)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
79
(ppm)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Conv.
of NO
83.1
98.3
91.8
76.8
89.8
89.8
94.2
100
66.5
90.4
86.7
92.4
83.3
84.2
80.0
77.9
84.1
82.2
82.2
96.3
93.7
67.1
5% H90, 14% CO, in N7. All tests were performed
b
c
d
e
f
g
h
i
j
Feed gas also contains 3% 07
at 400°C. " "
7 hour test.
4 hour test.
5 hour test.
2 hour test.
After 20 hours exposure to SOp-
After 23 hours exposure to 502-
After 42 hours exposure to S02-
After overnight exposure to air at 400°C.
After 7 hours exposure to S02 in this test. Total exposure in series, 49 hours.
Total exposure of catalyst to S02 in all tests, 71 hours, including preliminary
qualitative tests.
171
-------
100
80
60
4O
O
O
20
€)
O
O
•
A
CATALYST
NO.
NA-28
NA-28
NA-28
NA-28
NA-28
NA-29
NA-29
NA-29
NA-29
CAT.
WT.
GRAMS
28
28
28
14
14
14
14
14
14
SPACE VELOCITY
Hr"1 (STP)
5000
10000
15000
20000
10000
20000
10000
20000
20000
0.6
08
1.0
1.2
NH3/NO RATIO
NO CONC.
ppm
1000
1000
IOOO
1000
IOOO
IOOO
IOOO
500
250
1.4
Figure 24. Reduction of NO with NH^ on Fe-Cr Catalysts
at 400°C (14% C02, 5% H20, 3% 02 in Ng)
173
-------
prove cost effective. But the ammonia in the reactor effluent may have to
be recovered for air pollution control reasons. Should this be the case
the NA-28 must be considered as the most desirable catalyst for NO abate-
X
ment in power plants through selective NO reduction. No other catalyst
tested in this program or reported in the literature combined efficient NO
reduction with ammonia in the presence of oxygen and S00 and complete de-
" £
composition of the excess ammonia.
4.6 PRIORITIZED LISTING OF NO ABATEMENT CATALYSTS BASED ON TASK 2
INVESTIGATIONS x
The experimental investigations on NO abatement catalysts performed under
A
this program identified one catalyst, Pt, potentially useful in NO decom-
position and several catalysts which promoted the selective NO reduction by
NH3 in the presence of oxygen and sulfur dioxide. Only catalysts that indi-
cated substantial activity in one of the catalytic NO abatement processes
A
at temperatures of 400°C (752°F) or lower were considered. These catalysts
are listed below (Table 36) in the order ranked, with the top ranked catalyst
listed first. The catalysts are ranked separately for utilization in power
plants whose flue gas contains S02 and for utilization with virtually S02-
free flue gas. The most probable temperature and space velocity for optimum
catalyst performance are indicated also.
The top ranked catalysts in the two groups were selected as promising candi-
dates for use in power plant NO abatement, especially for existing power
A
plants. Thus, a preliminary design and cost analysis of their adaptation
to power plants was performed in Task 3.
Efficient catalysts for selective or nonselective NO reduction with H2 or
CO were not identified during the screening tests (400°C, 20,000 hr"1) of
Task 2.
174
-------
Table 36. CATALYST RANKING BASED ON TASK 2 DATA
Cat.
Rank
1
2
3a
4a
1
2
3
4
Cat. No. and
Active Components
Intended
Process
Expected Operation
Temp., °C
S.V., Hr-i (STP)
A. UTILIZATION WITH SO.-CONTAINING FLUE GAS
NA-28, Fe-Cr
Oxides
NA-29, Fe-Cr
Oxides or NA-25,
V2°5
NA-7, Mo-Oxide
or NA-18, CuO-Pb
or NA-33, Fe-
Oxide
NA-2, 0.5% Pt
or NA-4, Pt-Mo
NH3 Red.
NH, Red.
NH3 Red.
NH3 Red.
Decomp.
Decomp.
350-400
350-400
350-400
350-400
300
350
10,000-15,000
10,000-15,000
10,000-15,000
MO, 000
MO, 000
MO, 000
B. UTILIZATION WITH SO,, - FREE FLUE GAS
NA-2, 0.5% Pt
NA-28, Fe-Cr
Oxide
NA-29, Fe-Cr
Oxide or
NA-25, V205
NA-2, 0.5%Pt
NH3 Red.
NH3 Red.
NH3 Red.
Decomp.
200-250
350-400
350-400
350-400
300
^20,000
MO, 000
MO, 000
MO, 000
MO, 000
Expected
NO Conv., %
>««
>90%
>80%
>50%
>50%
>90%
>90%
>90%
>90%
>50%
Not tested with SO^-containing flue gas, but indirect evidence leads
to expectation of negligible SO^ effect on catalytic activity.
175
-------
5. EVALUATION OF THE COST EFFECTIVENESS IN THE USE OF CATALYSTS
TO REDUCE NOV EMISSIONS FROM POWER PLANTS (TASK 3)
X
The objective of this task was to conduct a preliminary design and cost anal-
ysis on catalytic NO abatement processes assessed as potentially adaptable
J\
to power plants on the basis of information collected in Tasks 1 and 2
(Sections 3 and 4 of this report).
Review of the literature, including unpublished work (Task 1) and of the experi-
mental data generated under this program (Task 2) suggested the following:
0 Truly efficient catalysts that can potentially perform
in power plant flue gas environment have only been iden-
tified for nonselective NO reduction and for selective
/\
NO reduction with ammonia.
J\
0 For the majority of power plants (power plants fired with
sulfur containing fuels) a simultaneous NO -SO abatement
A J\
scheme is the most desirable. The nonselective NO -SO
X X
reduction schemes appear to be the most promising in this
category. Preliminary data presented in Section 3.3 indi-
cate that simultaneous NO -SO reduction is feasible. In
X X
one case, NO was completely reduced to nitrogen while
J\
approximately 80% of the SOV was simultaneously reduced to
X
elemental sulfur without measurable production of HgS or
COS (NYU catalyst); in a second scheme, NOX and SOX were
virtually completely reduced to nitrogen and recoverable
metallic sulfides in a single reactor (TRW Sulfide Process).
In both cases coal can be used as the reductant. The
major question marks with these schemes concern the diffi-
culty and cost of adaptation to existing power plants.
0 For the small number of power plants expected to continue
to use sulfur-free or very low sulfur fuels, the selective
NO reduction with NH-, on Pt process may be more desirable.
X <3
This process has demonstrated high efficiency at low tem-
peratures (^200°C)
177
-------
• For existing power plants (fired with sulfur containing
fuels) for which a nonselective catalytic process is not
desirable because of design incompatibility or because the
power plant is already equipped with SO scrubbers, or for
A
power plants located in areas where fuel is scarce of too
expensive, the selective NO abatement by ammonia on non-
A
noble metal catalysts is indicated (e.g., Fe-Cr oxides or
V2o5).
Current and projected fuel costs suggest that the candidate nonselective
NO -SO abatement process must be capable of utilizing the lowest possible
A A
grade of fuel as the process reductant and that the process adaptation scheme
should be such that complete and efficient energy utilization is possible.
Reductant generation at the boiler burner by fuel-rich combustion has not
been possible with fuels other than natural gas, to our knowledge. But even if
possible with some fuels, it would be difficult to control the oxidant-
reductant ratio at the precise value required by the simultaneous catalytic
NO -SO process.
A- A
The use of piped natural gas, CO, or H2 could be the most desirable reduc-
tant source technically, but the scarcity and cost of these fuels render it
unfeasible for general adaptation.
Reductant generation on the power plant site, preferably as an integral
part of the process, by a relatively inexpensive fuel, such as coal, appears
to be the desirable reductant source. Thus, a scheme involving two-stage
combustion would be desirable.
On the basis of the above input, the following five processes were selected
for preliminary design and cost analysis:
1. One simultaneous catalytic NOX~SOX reduction process,
utilizing coal as the reductant, for adaptation to
new power plants.
2. One simultaneous NO -SO catalytic reduction by coal
A A
process for adaptation to existing power plants.
178
-------
3. One simultaneous NO -SOV reduction process, catalytic
A X
with respect to NO and either throwaway or regenerative
X
with respect to S02, for new power plants (TRW Sulfide
Process).
4. One selective NO reduction with NH~ on non-noble metal
X 3
catalysts process for either new or existing power plants.
5. One selective NO reduction with NH~ on platinum process
X *3
for new or existing power plants.
The next two sections of this report present the bases for the design and
cost evaluation of these five NO abatement processes and summarize the
X
engineering analysis results. The processes are treated in two groups,
simultaneous NO -SO reduction processes and selective NO reduction with
XX X
ammonia processes. Specifics on individual processes are presented in
separate subsections of the appropriate section.
5.1 PRELIMINARY DESIGN AND COST ANALYSIS OF SIMULTANEOUS NO -SO REDUCTION
SCHEMES
Processes 1 and 2 above were assumed adapted to power plants by the TRW
schemes, conceptual diagrams of which are presented in Figures 14a and 14b
(Section 3.3). The TRW schemes were selected because they are believed to
be very efficient schemes, which can utilize any type of inexpensive fuel
as the reductant source, and because they are the only schemes on which
data were available. It was also assumed that a single-stage catalytic
reactor would be sufficient for adequate NO -SOV reduction. This assumption
X X
implies that H~S and COS production on the catalyst does not take place or,
if it does, it is at levels acceptable for atmospheric discharge; such
catalyst has been identified (Table 14, Section 3.3) by NYU. The cata-
lytic reactor was sized on the basis of TRW generated data on a Harshaw
CuO catalyst (Table 16, Section 3.3) because the NYU data was insufficient
and it also arrived late. However, the deductions drawn from the engineer-
ing analysis of these processes would not be affected materially by a cata-
lyst switch such as the one contemplated here (both catalysts are non-noble
metal catalysts). The catalytic reactor design was based on Monsanto's
reactor design for the CATOX Process.
179
-------
Process 3, the "TRW Sulfide Process" which was presented in Section 3.3
(Figure 16), does not require a catalytic reactor. The catalyst for the
NO reduction is the sulfur product getter (if not the coal), which for the
X
purpose of this analysis was iron (it is believed that iron oxide or other
metal oxides may be used). Thus, the reductant generator and the catalytic
reactor are a single unit in this simultaneous NO -SO abatement approach.
/\ A
The product iron sulfide was treated as a waste because of lack of data on
a regenerative process for the sulfur getter. The process engineering anal-
ysis on this scheme was based on limited proof-of-principie data generated
earlier at TRW (Section 3.3, page 67).
The analyses were based on process adaptation to an 800 MW power plant (320
tons/hr coal consumption on boiler burners and reductant generator combined)
except for Process 2 (NO -SO catalytic reduction scheme for existing plants),
A A
where the electric generating capacity exceeded 800 MW by the quantity of
power produced from the reductant generator coal.
Table 37 lists the assumed composition of the reductant coal.
Table 37. ASSUMED CHEMICAL COMPOSITION FOR
THE REDUCTANT COAL, WEIGHT PERCENT
Carbon 70
Hydrogen 5
Oxygen 10
Sulfur 3
Ash 11
Nitrogen 1
Total 100% (Dry Basis)
Moisture Content 4%
Water oxygen = (4) (16/18) = 3.6%
Water hydrogen = (4) (2/18) + 0.4%
Non-water hydrogen =5-0.4 = 4.6%
Non-water oxygen = 10 - 3.6 = 6.4%
180
-------
Based on the Dulong formula* a coal of this composition would have a heat
content of 12,628 Btu/lb.
The flue gas production in the primary burner was assumed to be 281,250
standard cubic feet (60°F and 1 atmosphere) per ton of coal consumed (90 x
10 SCFH from a power plant consuming 320 tons per hour of coal). Table 38
presents the assumed power plant flue gas composition.
Table 38. ASSUMED POWER PLANT FLUE GAS COMPOSITION (VOLUME %)
N2
co2
H20
°2
so2
NOV
X
Fly ash, etc.
Total
75
14.5
7.0
3.0
0.2
0.1
0.2
100%
Because of the extraordinarily large size of some of the required equipment
(e.g., fans, reaction vessels) and the uniqueness of some of the applications
(e.g., internal insulation of large ducts to carry high temperature flue gas),
many of the capital cost items could not be estimated from the data reported
in the literature (e.g., by Guthrie ) for smaller size industrial units.
Accordingly, the estimated capital costs for many of the process scheme com-
ponents had to be obtained from equipment manufacturing firms and supply
houses. Table B-l in Appendix B of this report contains a partial listing
of the various companies contacted.
* Btu/lb = 14,544 C + 62,028 (H - 0/8) + 4050 S where C, H, 0, and S are
in fractional weights.
181
-------
The f.o.b. equipment costs were adjusted to include estimates of installation
costs, project indirect costs, and contingencies. Except where the equip-
ment manufacturers suggested a lower value, the installation cost was esti-
mated at 58% of the f.o.b. equipment cost. This is in line with the data by
Guthrie which indicate that labor costs for equipment installation and for
erection of field materials range from 54 to 66% of the equipment cost, with
the "norm" being close to 58%. The installation costs for fans and multi-
clones were estimated at 30% of the equipment cost. In accordance with the
data reported by Guthrie, the project indirect costs and contingencies were
estimated at 34% of the installed equipment costs and 18% of the project cost,
respectively. The estimates of the total operating cost include the follow-
ing: a 10-year straight line depreciation cost; an 8% of the capital invest-
ment per year allowance for maintenance, insurance, local taxes, etc.; labor
cost with 100% overhead; heat losses and power consumption costs; and, where
applicable, costs for the two NO -SO catalytic reduction processes and for
/\ /\
the modified process for SO removal, no credit was allowed for the recovered
A
by-products (sulfur and ferrous sulfide, respectively).
The next three subsections present the design and cost data generated on each
of the three simultaneous NO -SO processes. Process 1, catalytic NO -SO
AX XX
process for existing power plants, is discussed in considerable detail; the
other two schemes, which contain a number of similar elements to Process
1, are discussed in less detail. Details and support data on the analysis
of these schemes are presented in Appendix B of this report.
5.1.1 Simultaneous Catalytic NO -SO Reduction Scheme-New Power Plants
Figure 25 presents a schematic flow diagram for the application of the NO -
/\
SOX Catalytic Reduction Process to new power plants (gas flow rates, etc.,
shown on the figure will be discussed later). A portion of the hot boiler
flue gases, generated at the power plant burners (primary combustion), are
diverted from the secondary superheater boiler region to a coal bed
("Reductant Generator") which is also fed with a fraction of the total coal
to be consumed by the power plant in the generation of the 800 MW power.
The oxygen in the flue gas reacts with the coal in the coal bed to generate
carbon monoxide,
182
-------
00
CO
1700°F, 28.87 MM SCFH
1900°F
\J
1860°F, 30.77 MM SCFH
PRIMARY
SUPERHEATER 1300°F 77.49 MM SCFH
COAL
46 TONS/HR-
SECONDARY
SUPERHEATER
FIREBOX
2400°F
COAL
274 TONS/HR
1300°F 77.49 MM SCFH
1
ECONOMIZER
CATALYST BEDS—<
77.49 MM SCFH
AIR PREHEATER
X~i j ,.
•^ II. M I/
rn [JVWF J
I.D. FANS
REDUCTAHT
GENERATOR
T
ASH
MULTICLONE
COLLECTOR
ASH
F.D. FANS
17.38 MM SCFH
94.87 MM SCFH ^ TQ ELECTROSTAT|C
PRECIPITATOR AND STACK
94.87 MM SCFH
T
MULTICLONE
COLLECTOR
SULFUR
Figure 25. NOX-SOX Catalytic Reduction Process Adapted to New Power Plants (800 MW)
-------
C + 1/2 02 -> CO (1)
Other major reactions taking place in the reductant generator include
oxidation of the organically-bound hydrogen and (some) reduction of water
and carbon dioxide,
2 (-H) + 1/2 02 + H20 (2)
C + H20 -»• CO + H2 (3)
C + C02 ->- 2 CO. (4)
The oxygen-free reductant-rich effluent gas from the CO generator is mixed
in the primary superheater section of the boiler with the undiverted por-
tion of the flue gas. Through proper system design the two gas streams can
be so proportioned that the mixing would result in the reduction of all the
oxygen and in the oxidation of all the hydrogen and carbon monoxide, except
for a small quantity of CO needed in a subsequent catalytic step for the
stoichiometric reduction of sulfur and nitrogen oxides. This would eliminate
possibilities for carbon monoxide discharge to atmosphere and for catalyst
poisoning by excess oxygen. The reactions involved are as follows:
CO + 1/2 02 •*• C02 Homogeneous oxidation (5)
of carbon monoxide and
hydrogen and removal of
H2 + 1/2 -»• H20 oxygen (in the boiler)(6)
2CO + S02 -»• 2 C02 + 1/n S. Catalytic reduction (7)
of nitrogen and sul-
fur oxides (in the
CO + NO -» 1/2 N2 + C02 catalytic reactor) (8)
Except for small quantities of heat lost to the surrounding and discharged
in the ash, the gaseous effluent from the reductant generator contains
essentially all the heat value of the coal fed to this unit. This heat,
which is represented by the small rise in the diverted gas temperature and
by the fuel value of CO and H2 carried by this gas, is extracted by the
boiler heat exchangers (superheaters, etc.).
184
-------
As indicated in Figure 25, following the reduction of nitrogen and sulfur
oxides in the catalytic reactor, the flue gas is diverted through the
economizer and air preheater (for further extraction of heat) to a sulfur
recovery system. In this system the sulfur vapor is condensed by direct
air injection (cooling) and recovered in cyclone collectors.
The scheme design criteria were established from the data generated by TRW
in the apparatus depicted schematically in Figure B-l (Appendix B). The
pertinent information derived from the proof-of-principie data, taken with
synthetic flue gas on packed coal and catalyst beds, may be summarized as
follows:
• Efficient generation of reductants requires coal bed
temperatures in excess of 900°C (1650°F), although
reductant generation is possible at lower temperatures.
The utilized space velocity range was 1,000-1,500 hr~ .
0 The quantity of reductants generated in the coal bed
can easily be controlled by the size of the fraction
of the flue gas diverted to it, the space velocity of
the feed flue gas (residence time in the bed) and the
coal bed temperature.
• At temperatures up to 1650°F, the principal source of
reductants is coal oxidation by flue gas oxygen and
water (Reactions 1 and 3 above). At higher temperatures,
a substantial portion of the CO is produced from the
reaction of flue gas C02 with coal (Reaction 4). At a
temperature of 1700°F the estimated percent contri-
butions to the total CO production from Reactions 1,
3 and 4 are 45, 29, and 26, respectively.
• The NO and SO in the diverted flue gas are totally
reduced in the coal bed. The nitrogen oxides are re-
duced to N2 and NH- with the latter decomposing to N2
185
-------
increases with the coal feed rate, and since only stoichiometric quantities
of CO and H0 would be required for reaction with oxygen, NO , and SO in
t XX
the undiverted portion of the flue gas, the fraction of the gas diverted to
the coal bed would, in turn, depend on the coal feed rate for the CO
generator. Only when the diverted and undiverted fractions of the flue gas
are at a right proportion will not excess oxygen reach the catalyst (to
"poison" it) and no excess carbon monoxide (an air pollutant) be dis-
charged in the stack gases.
Table 39 presents a summary of the calculated data (coal feed rates and
expected level of excess oxygen or CO discharge) for assumed flue gas
diversion fractions of 30,33.33, 37.50 and 40%. Items 7 and 8 of this
table indicate that the optimum fraction of diverted flue gas is approxi-
mately 37.5% of total (near zero oxygen and no CO in stack gas). The
other three values of diverted flue gas result either in excess oxygen,
which would poison the catalyst, or excess CO, which is a pollutant.
Thus, 37.5% flue gas diversion was selected as the design value (28.87 MM
SCFH diverted gas flowrate). This flow rate requires a coal feed rate of
46 tons/hr. to the reductant generator. Since an 800 MW consumes 320 tons/
hr. coal, the coal feed rate to the primary burners is 274 tons/hr.
Table B-2 (Appendix B) presents detailed mass balance calculations for the
37.5% diverted flue gas fraction.
Heat balance calculations, based on 37% flue gas diversion, indicated that
8
the heat release in the reductant generator would be 1.85 x 10 Btu/hr
which would cause an increase in the temperature of the diverted gas from
927°C (1700°F) to 1016°C (1860°F); heat transfer losses were considered
negligible. Analogous calculations on the catalytic reactor revealed that
no appreciable change in temperature is expected in this unit. The quantity
of air necessary to cool the flue gas from 177°C(350°F) to 138°C (280°F) in
the elemental sulfur recovery unit was estimated at 17.38 MM SCFH (60°F).
Changes in flue gas flow rates (minor) due to change in flue gas composition
(Reactions 1 through 8) are indicated in Figure 25.
Unit sizing was not optimized, but an attempt was made to select dimen-
sions that did not result in high pressure drops. The reductant gen-
erator was designed as four circular units, each 30 feet in diameter
186
-------
and H2 or oxidizing to N2 and H20 in the boiler.
The sulfur oxides are reduced to elemental sulfur,
H2S and COS at the approximate ratio of 6:3:1, but
they reoxidize to S02 in the boiler prior to enter-
ing the catalyst. The inorganic sulfur content of
the coal in the reductant generator remains as
pyrite or sulfide in the ash; the organic sulfur
of this coal is converted to reduced gaseous sul-
fur compounds which are carried to the boiler by
the diverted flue gas where they oxidize to SCL.
t The NO and SO of the combined flue gas stream
A A
are reduced on the catalyst bed (Harshaw CuO) to
N2 and SR (60-70%), H2S (15-25%, and COS (3-5%),
respectively. Recommended operating conditions:
482-704°C (900-1300°F) and 8,000 to 10,000 hr"1
space velocity.
The following space parameters were selected for this engineering analysis:
Reductant Generator Catalytic Reactor
Temperature 927°C (1700°F) 704°C (1300°F)
Space Velocity 1,500 hr"1 9,000 hr"1
The NO and SO conversions on the catalyst bed were considered adequate to
A X '
meet clean air standards so that further treatment of the flue gas, other
than elemental sulfur recovery, was not considered necessary. The optimum
fraction of flue gas to be diverted to the reductant generator under the
selected conditions of temperature and space velocity had to be calculated.
It is apparent from the proof-of-principle results that at operating tem-
perature, the coal feed rate for the CO generator is determined by the quanti*
ty of the flue gas (i.e., the amount of oxygen) which is fed to the unit.
Since the quantity of reductants (CO and H2) generated in the coal bed
187
-------
oo
to
Table 39. EFFECT OF FLUE GAS DIVERSION RATE ON COAL FEED RATES AND MASS BALANCE RESULTS
(Operating Temperature for Reductant Generator 927°C (1700°F))
1.
2.
3.
4.
5.
6.
7.
8.
Flue Gas Diverted to Reductant Generator, % ->
Coal feed rate to reductant generator, tons/hr
Coal feed rate to primary burner, tons/hr
CO produced in the reductant generator, Ib-mole/hr
Hydrogen produced in the reductant generator, Ib-mole/hr
Oxygen required ror oxidation of H2, H2S, S, and excess CO
in the reductant generator effluent, Ib-mole/hr
Oxygen available in the undiverted flue gas, Ib-mole/hr
Excess oxygen reaching catalyst, Ib-mole/hr (Item 6
minus Item 5)
Excess CO discharged in stack gases (Item 5 minus
Item 6 multiplied by 2)
30
38
282
5095
1478
2953
4388
1435
0
33.33
42
278
5206
1633
3313
4262
949
0
37. 5a
46
274
6168
1788
3699
3804
105
0
40
50
270
6705
1944
4456
3600
0
912
Table B-2 (Appendix B) presents mass balance calculations for this fraction of
diverted flow.
-------
and 15 feet in total height and containing 7 feet of 1/2-inch coal
particles (packed beds).* The estimated operating pressure drop through
the bed was calculated to be 10 inches of water.
The catalytic reactor is a 30 feet diameter by 130 feet cylindrical
unit depicted in Figure 26. The reactor was based on the Monsanto multi-
section design. Each section is ten feet high and contains one foot of
catalyst (total catalyst volume 8760 cubic feet). The total pressure drop
through the reactor was estimated at 2 inches of water (based on 1/2-inch
catalyst cylinders and a bed void fraction of 0.5).
The multiclones selected for the removal of fly ash from the flue gas
feed to the catalyst reactor consist of 8 units each with a capacity
of 600,000 cfm. The manufacturer's estimate of the pressure drop through
the multiclones is 3.1 inches of water. The units are expected to effect
95% fly ash removal (assuming about 5% minus 10 micron particle size).
Similar type multiclones (5 units) were assumed for sulfur recovery.
Duct cross sections were determined from assumed linear gas velocities
of 2,000 and 3,000 fpm for fly ash laden and fly ash free flue gas,
respectively.
Four induced draft and two forced draft fans are used in this scheme.
Induced draft fans were recommended by several manufacturers for moving
large volumes of hot gas. Each induced draft fan has a capacity of 500,000
cfm working against a total estimated scheme pressure of 15 inches of water
at 177°C (350°F). Each of the two forced draft fans, selected for air
injection in the sulfur recovery section, has a capacity of 169,000 cfm
working against an estimated pressure drop of 2 inches of water.
Although the detailed design for a system (equipment) to divert the
hot flue gas through the CO generator has not been worked out, discussions
with one manufacturer of flow control equipment indicated that the fabrica-
tion of a "non-tight" diversion damper for the proposed application should
not pose any major engineering difficulty.
* A fluidized bed may be a more efficient design.
190
-------
FLUE GAS , T^
INLET ^
^_
r x
="-=~-=--^-=-T-=_=-SJ=.-=
«»«»»«
^SECTION NO. 3,
ETC.
IECTION NcxT
SECTION NO 1
j
».
1
-
>FLUE GAS
OUTLET
Figure 26. Monsanto's Multi-Section Catalytic Reactor
191
-------
Details on equipment size, type of construction, and unit costs are
given in Table B-3 (Appendix B).
Summaries of the estimated capital and operating costs for this scheme
(Figure 25) are presented in Tables 40 and 41, respectively. The
total capital cost was estimated at $7,305,000 ($9.13 per kw) and the
annual operating cost at $2,529,000 (0.040 CO + H2, respectively) would be the principal
sources of reductants. For the actual operating temperature of the reduc-
tant generator (1000°-1420° range), reactions 1 and 3 were assumed to
account for 95% and 5% of the total CO produced, respectively. The coal
feed rate to the reductant generator was estimated at 54 tons per hour and
the fraction of the flue gas diverted to this unit was calculated to be
0.57. Pertinent computations are presented in Table B-5 (Appendix B).
192
-------
Table 40. BREAKDOWN OF THE ESTIMATED CAPITAL COST FOR THE NO -SO
CATALYTIC REDUCTION PROCESS FOR NEW POWER PLANTS x x
(Table B-3 in Appendix B presents details)
CO Generation Section
1) Flue gas diversion damper
2) Coal feeder
3) CO generators
4) Ductwork
Catalytic Reaction Section
1) Ductwork
2) Multiclones
3) Catalytic reactor
Cost ($000)
250
15
480
900
1,645
1,200
1,210
1,225
Induced Draft Fans (and Motors)
Sulfur Collection Section
1) Ductwork
2) Multiclones
3) Forced draft fans
3,635
750
TOTAL
1,275
7,305
(or $9.13/KW)
Table 41. OPERATING COST BREAKDOWN FOR THE NOX-SOX CATALYTIC
REDUCTION PROCESS FOR NEW POWER PLANTS
(Table B-4 in Appendix B presents details)
Cost ($000 per year)
1) Depreciation
2) Maintenance, insurance, taxes, etc,
3) Labor
4) Electric power
5) Energy losses
6) Catalyst replacement
730
584
240
522
168
285
TOTAL
2,529
or 0.040 */KWH
193
-------
38.70 MMSCFH
UD
en
1400°F
Ih
MULTI-
/CLONES
X
u.
U
to
5
8
8
u.
0
o
00
r—
• HEAT
EXCHANGER
1000°F
-_ WASTE HEAT
BOILER
~L
Ul
CO
f-
01
U
I
U
Y«
ASH
1000°F
52.93
SCF
,970'F
WLTI-
LONE
)LLEC-
TOR
M
H_
f
REDUCTANT
V
J
51 .30 MM SCFH
AMBIE
COAL
54 TONS/HR
ASH
FLY ASH
350°F
F.D.FANS $$""
TO ELECTROSTATIC
_»>PRECIPITATORAND
STOCK
280°F
90.40 MM
SCFH
MULTICLONE
COLLECTOR
110.98 MM
SCFH
SULFUR
I.D. FANS
Figure 27. NO -SO Catalytic Reduction Process Adapted to Existing Power Plants (800 MW)
X X
-------
A second difference between this scheme and that designed for new power
plants involves the energy utilization from the reductant generator coal.
In the previous scheme the effluent gas from this unit was returned to
the superheater section of the boiler where the energy was absorbed in
the normal power plant manner; thus, the coal used in the reductant
generator was considered a part of the 120 tons per hour consumption of
the 800 MW plant. In this scheme, the energy from the coal consumed in
the reductant generator is available for recovery at approximately 538°C
(1000°F) after the flue gas exits the heat exchanger. Energy recovery
was assumed to take place in a waste heat boiler. This unit was neither
sized, nor costed. It was assumed that the cost of the coal fed to the
reductant generator (adjusted for losses) and the cost of recovering
its energy will equal the credit from the power derived from it. This
power was not considered as part of the 800 MW plant capacity; thus, 120
tons per hour coal (or equivalent fuel) was fed to the boiler burning in
this scheme.
The procedures used for computing flow rates and gas temperatures at various
points in the scheme and for the selection and sizing of equipment are
identical to those described for the adaption of the same process to new
power plants (Figure 25). Details on equipment sizing, construction
materials, and cost are presented in Table B-6 (Appendix B).
Tables 42 and 43 summarize the breakdown of the estimated capital and
operating costs for the scheme shown in Figure 27. The data in Table 42
indicate that the heat exchanger unit accounts for a significant portion
(48%) of the estimated total capital cost. For this preliminary design,
no attempt was made to optimize the exit gas temperature for the unit or
to investigate alternate methods for heating the flue gas feed for the
reductant generator. Also, the cost of the waste heat boiler was not
considered and no credit was taken for the estimated 70 MW of additional
electricity generated in this unit (estimate based on an assumed heat
extraction efficiency of 20%). Similarly, in estimating the total operating
cost, the cost of the 54 tons/hr of coal fed to the CO generator was not
considered as an operating expense (refer to earlier discussion). The
estimated total capital cost for the process is $14,935,000 (or $18.67
per KW of generating capacity for the main plant boiler) and the estimated
196
-------
Table 42. BREAKDOWN OF THE ESTIMATED CAPITAL COST FOR THE NOv-SOv
CATALYTIC REDUCTION PROCESS FOR EXISTING POWER PLANTS
(Table B-6 in Appendix B presents details)
Reductant Generation Section
1) Multic!ones
2) Coal feeder
3) CO generators
4) Ductwork
Catalytic Reaction Section
1) Multiclones
2) Catalytic Reactor
3) Ductwork
Sulfur Recovery Section
1) Ductwork
2) Forced draft fans
3) Multiclones
Heat Exchanger
Induced Draft Fans
Cost ($000)
250
15
745
900
1,910
3,635
TOTAL
1,335
7,120
935
14,935
($18.67/KW)
Table 43. OPERATING COST BREAKDOWN FOR THE N0x-S0y CATALYTIC
REDUCTION PROCESS (Existing Power Plants)
(Table B-7 in Appendix B presents details)
1) Depreciation
2) Maintenance, insurance, taxes, etc.
3) Labor
4) Electric power
5) Energy losses
6) Catalyst replacement
Cost ($000/Year)
(0.0621 per KWH)
197
-------
annual operating cost is $3,998,000 (or 0.062
-------
vo
10
COAL
220 TONS/HR
61.82 MM SCFH
UJiu
•V
£
F *
1
f
I
= L
L
lr
i
i
||
1
1700°F
350°F AI
15.63 MM
-^^•PR
•*r
350°F
"
78.34
1910°F
65.99 MM SCFH
PREHEATER
H
78.34
SCFH
MULTICLONE
COLLECTOR
Fe 14 TONS/HR
COAL
100 TONS/HR
ASH
Fe-SULFIDE
TO ELECTROSTATIC
PRECIPITATORS
AND STACK
FANS
FLY ASH
Figure 28. Sulfide Process Adaptation to an 800 MW Power Plant
-------
Table 44. BREAKDOWN OF CAPITAL COST FOR THE SULFIDE NO sn
REDUCTION SCHEME FOR NEW POWER PLANTS x
(Table B-8 in Appendix B presents details)
Cost ($000)
Coal/iron feeder 25
Reductant generators 960
Ducts 2j850
Multic!ones 495
Induced draft fans 745
5,075
(or $6.35/KW)
Table 45. OPERATING COST BREAKDOWN FOR THE SULFIDE NO -SO
REDUCTION SCHEME FOR NEW POWER PLANTS x x
(Table B-9 in Appendix B presents details)
Cost ($000/Year)
Depreciation 508
Maintenance, insurance, taxes, etc. 406
Labor 160
Electric power 512
Energy losses 168
Iron scrap 7,728
9,482
(or 0.15£ per KWH)
201
-------
adaptation to new power plants (800 MW) was estimated at 5,075,000 (or
$6.35 per KW); the annual operating costs were estimated at 9,482,000 (or
O.ZOtf per KWH). Details are presented in Tables B-8 and B-9 (Appendix B).
It was pointed out earlier (Section 5.1) that iron ore or other metallic
ores may be used instead of iron and that these ores may be regenerable.
If iron ore were used in the above computations as a non-regenerable
reactant, the annual operating costs would be reduced by $5 million, even
with no credit for the sulfide product. These computations were based at
an iron ore price of $12.16 per ton f.o.b. Lake Superior Ports (Iron Age,
January 21, 1974) and an iron content in the ore of 51.5%. The estimated
new annual costs were $4,398,000 or 0.069
-------
an existing plant were not estimated. The multi-stage catalytic reactor
depicted in Figure 26, Section 5.1, was selected for both schemes. A
simpler reactor may be preferable for the platinum promoted scheme, e.g.
the type used by Environics (Section 3.2.2), but design and cost data
on such a reactor were not available.
Heat and mass balance computations, equipment sizing and costing, and
operating cost determinations were made as described for the simultaneous
NOX-SOX reduction processes.
Catalyst replacement was assumed to occur once per year; platinum was
assumed reclaimed. Ammonia was priced at 12
-------
REDUCING VALVE
COAL
320 TONS/HR
GAS
OMPRESSE
LIQUID
NH
I.D. FANS
4020 LBS/HR NH3
MULTICLONE
COLLECTOR
TO ELECTROSTATIC
-*-PRECIPITATORS
AND STACK
Figure 29. NOX Reduction with Ammonia Scheme - (Non-Noble Metal Catalyst)
-------
Table 46. BREAKDOWN OF THE ESTIMATED CAPITAL COST FOR NOY
REDUCTION BY AMMONIA SCHEME - NON-NOBLE METAL X
CATALYSTS (Table B-10 in Appendix B presents details)
Cost ($000)
Ducts 1,685
Multic!ones 905
Catalytic reactor 1,505
Ammonia storage tank and feeding system 170
Induced draft fans 650
4,925
(or $6.15/KW)
Table 47. SUMMARY OF THE BREAKDOWN OF THE ESTIMATED OPERATING COST
FOR NO REDUCTION BY AMMONIA SCHEME - NON-NOBLE METAL
CATALYSTS
Cost ($000)
Depreciation 492
Maintenance, insurance, taxes, etc. 394
Labor 160
Electric power 521
Energy losses 17
Catalyst replacement cost 195
Ammonia consumption 3,859
5,638
(or 0.088tf per KWH)
207
-------
5.2.2 NO Reduction by Ammonia on Platinum Catalysts - Power Plant
Adaptation Scheme
Except for a lower operating temperature and the use of a platinum
catalyst, this process is essentially identical to the ammonia process
(described above) using a non-noble metal catalyst. A schematic flow
diagram for the process is shown in Figure 30. The sizing of the cata-
lytic reactor and ammonia feeding equipment and the calculation of ammonia
consumption rate were based on laboratory test data (Task 2) which had
indicated a maximum NO removal efficiency of 91.7% for the following
^
operating conditions: temperature 204°C (400°F); ammonia-to-NO volume
ratio 0.977; and space velocity 20,000 hr .
Summaries of capital and operating costs for the process are presented
in Tables 48 and 49, respectively. Because of the high cost of platinum
($35.55/Kg for a catalyst containing 0.5% platinum by weight), the
initial catalyst charge represents a significant capital investment
(44,350,000 or 56% of the total capital cost). The processing (rejuvena-
tion) of catalyst was assumed to be on a once-per-year basis at an esti-
mated cost of $11.13/Kg. The data in Table 49 indicate that catalyst
restoration and ammonia consumption costs account for 18.4 and 51.2% of
the total operating cost, respectively. The total estimated capital and
operating costs for the process correspond to $9.64 per KW and 0.115tf per
KWH, respectively. Details are presented in Tables B-ll and B-12 in
Appendix B of this report.
5.3 SUMMARY OF PRELIMINARY DESIGN AND COST ANALYSIS RESULTS ON FIVE
NOV ABATEMENT SCHEMES ADAPTED TO 800 MW POWER PLANTS
A
Table 50 presents a summary of the costs associated with the adaptation
of selected simultaneous SO -NO reduction and selective NO reduction
XX X
with ammonia processes to an 800 MW power plant. This Table also
indicates the most cost sensitive items for each scheme.
The five schemes summarized in Table 50 represent three NO abatement pro-
rt
cesses. The NO -SO simultaneous catalytic reduction process was designed
X X
and costed for adaptation to new and existing power plants (800 MW). The
simultaneous NO -SO reduction Sulfide Process (catalytic-throwaway or
X X
regenerative) was adapted to new power plants (800 MW). The selective NO
x
208
-------
ro
COAL
320 TONS/HR-
D
AIR PREHEATER'
(MODIFIED)
90 MM SCFH
TO ELECTROSTATIC
PRECIPITATORS AND STACK
CATALYST
PRESSURE
REDUCING
VALVE
90 MM SCFH
FLOW
LBS/HR
MULTICLONE
COLLECTOR
400° F
T
ASH
Figure 30. NO Reduction with Ammonia Scheme - Platinum Catalyst
/\
-------
Table 48. BREAKDOWN OF THE ESTIMATED CAPITAL COST FOR THE NO
REDUCTION BY AMMONIA SCHEME - PLATINUM CATALYSTS. x
(Table B-ll in Appendix B presents details)
Cost ($000)
Ducts 850
Multiclones 755
Catalytic reactor (including catalyst charge) 5,285
Induced draft fans 650
Ammonia feeding equipment 170
7,720
(or $9.64 per KW)
Table 49. BREAKDOWN OF THE ESTIMATED OPERATING COST FOR THE NO
REDUCTION BY AMMONIA SCHEME - PLATINUM CATALYSTS
(Table B-12 in Appendix B presents details)
Cost ($OOoT
Depreciation 772
Maintenance, insurance, taxes, etc. 618
Labor 160
Electric power 521
Energy loss 172
Catalyst restoration 1,361
Ammonia consumption 3,782
7,386
(or 0.115tf per KWH)
211
-------
Table 50. CAPITAL AND OPERATING COST ESTIMATES FOR THE ADAPTATION OF
CATALYTIC N0¥ AND NO -SOV SCHEMES TO AN 800 MW POWER PLANT
A XX
SUMMARY TABLEa
Costs
Capital Cost ($)
Total
Per KW
Operating Cost ($)
£3 Annual
00 Per KWH, i
Major Cost Items
NO -SOV Cat. Red. Process
X y\
Power Plant
New
7,305,000
9.13
2,529,000
0.040
Ducts,
Catalytic
Reactor
Power PI ante
1 Existing
14,935,000
18.67
3,988,000
0.062
Heat
Exchanger,
Ducts
Sulfide
Process
(NOX-SOX
Reduction) ,
New Plants
5,075,000
6.35
9,482,000
0.15
Scrap Iron
NOX Reduction by
Ammonia Process
Non-Noble
Metal
Catalyst
4,925,000
6.15
5,638,000
0.088
NH3
Noble
Metal
Catalyst
7,720,000
9.64
7,386,000
0.115
NH3, Catalyst
a Costs do not include working capital and return on investment. Ten year straight-line depreciation
was used. Maintenance, insurance, taxes = 8% of capital (annually). Power cost: 1.4$ per KWH.
Catalyst replacement (regeneration of Pt) once per year. Costs based on 1973 prices.
Sulfur-product credit was not taken.
c Capital costs of waste-heat boiler and coal used in CO generator were not costed: however, credit
for power generated in this unit was not taken either.
Scrap iron was costed at $100 per ton. The iron is not regenerated.
-------
reduction with ammonia on non-noble metal and platinum catalysts process
was adapted to new power plants (800 MM), but the computed costs should
very closely relate to scheme adaptation to existing power plants. The
ammonia process on platinum was designed for S0«-free flue gas.
Comparison of costs and scheme simplicity considerations suggest the follow-
ing process ratings:
t For existing or new power plants fired with sulfur-free
fuel, the selective NO reduction with ammonia on
platinum is indicated.
• For existing power plants fired with sulfur containing
fuels, the selective NO reduction with ammonia on non-
noble metal catalysts is indicated, provided an S02
abatement process is available and compatible with the
ammonia process; the Fe-Cr oxide catalyst is recommended
with a second choice being vanadia.
• For new power plants fired with sulfur containing fuels
cost comparison indicates that the preferred process is
the simultaneous catalytic NO -SOY reduction process.
A A
Scheme simplicity and degree of NOX~SOX abatement
suggest the sulfide process as the most preferable.
It should be noted that the presented engineering analyses were based on
conceptual schemes supported with only a few points of proof-of-principle
data; consequently, they represent a first level process design and cost
effort. The generated cost estimates should be considered as yardsticks
of the relative merit of the five schemes rather than as absolute scheme
adaptation costs, even though every effort was made to utilize as accurate
information as possible in costing them. Previous experience has indicated
that preliminary process cost estimates tend to escalate substantially with
process development.
214
-------
6. RECOMMENDATIONS FOR FURTHER ACTION (TASK 4)
Program Tasks 1 through 3 suggested that catalytic NO abatement from power
plants should be technically and economically feasible. Our assessment of
the data reported in the literature and that generated under this program
can be summarized as follows:
t The simplest and probably least expensive means of power
plant NO abatement would be through NO decomposition
X X
on platinum, provided 50-60% abatement could be con-
sidered adequate. It is not certain, however, that
this scheme is adaptable to power plant flue gases con-
taining so2.
• For existing power plants operating with sulfur free
fuels (virtually S02-free flue gas), the selective re-
duction of NO with ammonia on platinum is indicated.
X
Recommendation of this process is based principally on
its efficient low temperature operation (air preheater
temperatures) which renders simplicity to its adaptation
to power plants. It is believed that the Environics
Corporation data (pilot plant scale) will verify this
assessment.
§ For the majority of existing power plants operating with
sulfur containing fuels, the selective NO reduction
X
with ammonia on non-noble metal catalysts (Fe-Cr oxides,
VpOg) process appears to be the simplest approach. This
assessment presupposes the use of a second process for
S02 abatement.
A simultaneous catalytic NOX-SOX process is potentially
adaptable to this type of plant , but additional bench-
scale data is required before the technico-economic fea-
sibility of such a scheme can be considered preferable
215
-------
to a series type of NO -SO abatement scheme (e.g.,
X X
selective NO reduction followed by SO scrubbing).
X A
A key question on the catalytic NO -SOV process is
X X
the efficiency of the catalyst to simultaneously
abate both pollutants in a single-stage reactor
(the single point data available to us from NYU
are considered promising, but insufficient for
efficiency assessment). Additional data are also
required in order to establish the efficiency of
the TRW scheme (Section 5.1) as applied to exist-
ing power plants (reductant generation efficiency
at temperatures below 800°C (1472°F), efficiency
of homogeneous oxygen reduction at 760°C (1400°F),
and efficiency of energy absorption at 538°C
[1000°F]). Potentially, the scheme appears pro-
mising, but its evaluation was based on extrapolated
data and on the assumption of the existence of an
efficient single-stage catalyst.
For new power plants, a simultaneous, nonselective
NO -SO reduction scheme is recommended. In new
X X
plant adaptation such a scheme can become an inte-
gral part of the plant, instead of an add-on
retrofit, and, therefore, it can be much more
efficient (Section 5.1) than when adapted to
existing plants. The catalytic NO -SO scheme
X X
appears very cost effective, provided a single-
stage catalytic reactor proves efficient. The
Sulfide Process is simpler, but its demonstrated
efficiency with iron must be verified with iron
oxide or dolomite to be considered a high ranked
candidate process.
216
-------
t For total pollutant abatement (NO -SO , Hg, As, Sb,
A A
PNA, etc.) from either existing or new power plants,
a catalytic oxidative wet scrubbing process may be
the desired approach, but additional data is required
for even a preliminary assessment of the technico-
economic feasibility of such processes.
It should be noted that the above scheme assessment is based on very prelimi-
nary data and the ranking of the schemes is tentative. However, every effort
was made to utilize data in the individual scheme analyses which closely re-
presents power plant conditions, especially coal fired power plants. Thus,
the described schemes are believed to be adaptable to real power plants.
Their performance and their impact on the efficiency of the power plant can
not be precisely determined at this stage of their development. In general,
their impact should not exceed that of S02 scrubbers. For example, the
selective NO reduction processes are similar and yet simpler than the CAT-OX
A
desulfurization process and are expected to be more compatible with and to
have less effect on power plant operation than the CAT-OX process. There is
greater uncertainty on the impact of the simultaneous NO -SO reduction
•A -A
schemes on the efficiency of existing power plants. In principle, they
should have no impact because they do not require but generate energy which
can be tapped independently of the main power plant.
The apparent conditional assessment and ranking of the above NO abatement
A
schemes do not imply doubts on the potential of these processes; they merely
suggest that additional effort is required before the optimum NO abatement
A
scheme(s) for power plant adaptation can be selected. The ensuing para-
graphs present the recommended course of action (technical plan) for
selection and development of the optimum process or processes and the
corresponding adaptation of the scheme(s) to the demonstration-scale level.
The recommended program is presented in the following four tasks.
Task 1 - Bench-Scale Development of the NO..-NH0-Non-Noble Metal Catalyst Scheme
" ~'J ~""J A J -.-... -i- — — . —r **_
It is proposed that the iron-chromium oxide catalyst (NA-28) which promoted
the oxidation of excess ammonia to the prime choice for this scheme with
217
-------
vanadia as the alternate candidate. It is recommended that 0.5 to 2.0
kilogram catalyst beds be tested on appropriately modified actual flue gas
(this size of system is considered to be the most cost effective since it
permits the generation of data that can be safely extrapolated without re-
quiring large capital expenditures for experimental set-up).
The following investigations are recommended as a minimum effort for this task:
• Determination of process parametric effects (S.V., tem-
perature, oxygen, water vapor, nitric oxide concentration,
NH3/NO ratio, S02). These parameters are to be investi-
gated for quantitative data generation in the range of
process operability determined in this phase of the
program.
• Development of empirical NO reduction rate expressions.
A
• Generation of process design curves adequate for pilot
or demonstration plant engineering design.
• Update of engineering analysis performed in this pro-
gram phase.
• Determination of long-term catalyst stability (at
least 200 hours of operation with a single batch of
catalyst, which should include several temperature
and environment cycles).
• Assessment of current nonregulated emissions resulting
from NO catalyst schemes.
J\
• Total power plant impact of NO control: i.e., cost,
•A
effluent changes, temperature effects, and effects on
current emissions control equipment efficiency, dura-
bility, and operating costs.
• Projected demand, cost, and availability of the active
catalyst materials.
218
-------
Task 2 - Bench-Scale Development of a Simultaneous Nonselective NO -SO
Reduction Scheme
It is recommended that the objective of this task be the investigation of
the sulfide process and the catalytic NO -SO process for both new and
A X
existing power plants on a single bench-scale set up. The task is comprised
of two major subtasks. The first subtask involves the investigation of the
sulfide process as well as reductant generation for the catalytic NO -SO
X A
process. The second subtask involves parametric investigations on the most
promising nonselective NO -SO reduction catalyst (e.g., the NYU catalyst).
X A
It is proposed that the first subtask involves the following investigations:
• Proof-of-principle tests on iron oxide or dolomite
utilization in the sulfide process.
• Proof-of-principle tests on iron oxide or dolomite
(calcium-magnesium oxide) regeneration.
• Determination of parametric effects on coal gasifier
operated as a reductant generator for the catalytic
NO -SO processes and as a sulfide process reactor.
X X
• Determination of reductant generation rates for new
power plants H700°F) and for existing power plants
H400°F).
• Determination of iron oxide or dolomite consumption
and regeneration rates for efficient NO -SO abate-
X X
ment (if proof-of-principle tests prove successful).
• Determination of homogeneous oxygen reduction rates
at 1700 and 1400°F (927 and 760°C) in simulated
boiler (catalytic NO -SO processes). Determination
X X
of excess reductant oxidation by air in boiler
(sulfide process).
219
-------
• Update of engineering analysis on sulfide process
(if proof-of-principle tests on iron oxide or dolo-
mite prove successful).
• Determination of long-term recyclabi1ity of iron
oxide or dolomite (at least one ten cycle test).
The effort in the second subtask should involve the following:
• Determination of optimum operating conditions for
selected NO -SO,
X /
investigations.
selected NO -SOV catalyst through parametric
X X
Determination of empirical rate expressions for
NO -SO reduction.
X X
Update of engineering analyses on the nonselective
catalytii
plants).
catalytic NO -SO schemes (existing and new power
X X
• Determination of long-term catalyst stability (at
least 200 hours of operation with a single batch
of catalyst, which should include several tem-
perature and environment cycles).
Task 3 - Nitric Oxide Decomposition on Platinum
This task may be investigated as an extention of Task 1; therefore, the
same experimental apparatus can be used for both tasks. Three month schedule
at a 0.5 man year level should be sufficient for this task, if performed in
conjunction with Task 1. The recommended effort should involve the investi-
gation of the S0« effect on catalytic activity, the effect of NO and oxygen
£ X
concentrations on decomposition rates, and long term catalyst stability
tests.
220
-------
Task 4 - Proof-of-Principie Tests and Preliminary Engineering Analysis on
Total Pollutant Abatement Scheme (TRW OXNOX Process)
Proof-of-principle tests on this process should involve: (a) catalytic oxi-
dation of NO to N0« at low space velocities and temperatures, and (b) total
pollutant process scrubbing efficiency. In addition, a preliminary engineer-
ing analysis, similar to those presented in Section 5, should be performed
on this process.
The effort is estimated at the one man year level with a six month schedule.
The proposed program recommends the bench pilot scale development of more
than one NO abatement process. The rationali
X
recommendation is predicted on the following:
than one NO abatement process. The rationale behind the multi-process
X
Nitric oxide abatement requirements on existing
power plants vary because of fuel used, geographical
location, S02 abatement schemes used or expected to
be used. Thus, a single NO abatement process is
A
not expected to be universally desirable or even
acceptable. Reductant availability and cost,
grounds availability, and power plant design are
additional variables that can influence selection
of the desirable NO abatement scheme. Thus, more
J\
than one scheme should be available for existing
power plants.
Nitric oxide abatement schemes for new power plants
should be integrable with other pollutant abatement
schemes, preferably into a single unit. Since S02
is the other major pollutant emitted by power plants,
a process that simultaneously and efficiently con-
trols both pollutants must be developed for these
plants. With existing power plants, the integrated
approach to air pollution control is not easy or may
221
-------
not be possible; with new power plants It appears
feasible (at least for NO -SO abatement). Ideally,
A ^
simultaneous total pollutant abatement is desirable.
• Near term and distant solutions to NO pollution may
/\
differ. Development of proven feasible NO abatement
/\
processes should not be reason for discontinuing proof-
of-principle tests on potential total pollutant abate-
ment processes to be used as a second generation
approaches.
222
-------
APPENDICES
223
-------
APPENDIX A
I N20 PRODUC
OXIDE REDUCTION WITH AMMONIA ON PLATINUM CATALYSTS
PARAMETRIC EFFECTS ON N20 PRODUCTION DURING NITRIC
A number of tests were performed to investigate the effect of NH3 and oxygen
concentrations and the NFL-to-NO ratio on N20 production during NO reduction
with ammonia on platinum catalysts. Several tests were performed with a
helium carrier in order to obtain a meaningful nitrogen mass balance. The
NpO, 02, and N2 flue gas components were monitored chromatographically in
these tests (Porapak Q and molecular sieve 5A columns at 25°C).
Table A-l summarizes the data taken on the NA-1 catalyst (0.5% Pt on alumina).
It is noted in the table that the obtained nitrogen mass balance (columns 16
and 17) is as good as can be expected from the analytical instruments used
and the number of species (NO, N02, NJD, NH,, N2) and quantities involved.
The first four experiments in Table A-l were performed using typical, sulfur
free, simulated flue gas containing 14% C02, 5% H20, and 3% oxygen by volume.
The utilized ratios appear in column 5 and lie on both sides of stoichiometry
(the value of the ratio at stoichiometry is 0.667) for the desired reaction
6NO + 4NH3 + 5N2 + 6H20 (1A)
AT 250°C (Experiments Nos. 3 and 4) N20 production was substantial, regard-
less of NH3 to NO ratio in the feed flue gas; however, the quantity of N20
produced doubled when this ratio was increased from 0.482 (">40% below stoi-
chiometry) to 1.02 (^20% above stoichiometry). Since NO conversion (3rd
column from the right) increased only by 30%, it is assumed that at least
some of the N20 produced is the result of ammonia oxidation by oxygen.
According to the literature, ammonia oxidizes to NO on noble metal catalysts
at temperatures above 300°C and to N20 at lower temperatures. Experiment
Nos. 3 and 4 indicate that NHL concentration had a large effect on NJD pro-
duction in the presence of oxygen; thus, complete process selectivity toward
Reaction 1A was not verified.
225
-------
Table A-l. PARAMETRIC EFFECTS ON.NgO PRODUCTION ON Pt-Al203 (NA-1)
Exp,
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Catalyst
Temp.
°C
Reactor Feed Gas Composition
NO NH3 NH3a 02 H20 . C02 Diluent
(ppm) (ppm) NO (ppr) (%) (%) Used
A. EFFECT OF NH, CONCENTRATION
345
345
250
250
245
250
250
250
250
250
250
250
J
1027 120 .117 30,000 5 14 N?
982 473 .482 30,000 5 14 N-
982 473 .482 30,000 5 14 N2
982 1000 1.02 30,000 5 14 N2
934 >3000 >3.2 -.50 0 0 He
1070 1742 1.63 ^50 0 0 He
1090 1420 1.30 .,50 0 0 He
980 1284 1.31 '..50 0 0 He
1000 799 .799 -.50 0 0 He
980 564 .576 45 00 He
940 558 .594 -,,50 0 0 He
965 385 .399 ->-50 0 0 He
B. EFFECT OF OXYGEN CONCENTRATION
268
253
250
250
250.
250
994 1170 1.18 70 0 0 He
918 1050 1.14 80 00 He
918 1019 1.11 630 0 0 He
918 1050 1.14 780 0 0 He
918 1043 1.14 6250 0 0 He
918 1104 1.20 10,510 0 0 He
Monitored Reactor Outlet
NO NH3 N20 N2b 02
(ppm) (ppm) (ppm) (ppm) (ppm)
673 0 50 N.A. N.A.
600 60 285 N.A. N.A.
391 N.A. 420 N.A. N.A.
218 N.A. 850 N.A. N.A.
0 >3000 0 735 N.A.
0 980 0 810 N.A.
0 657 0 890 N.A.
0 393 0 855 N.A.
0 42 80 690 N.A.
18 40 240 365 130
0 22 279 495? N.A.
312 0 235 195 N.A.
0 195" 0 820 130
0 191 0 740 170
115 59 710 75 230
131 39 715 100 440
150 33 560 230 5,800
140 42 560 175 10,100
Total System
Nitrogen
Balance, ppm
in uut
N.A. N.A.
N.A. N.A.
N.A. N.A.
N.A. N.A.
N.A. N.A.
2812 2600
2510 2437
2264 2103
1799 1582
1544 1268
1498 1570?
1350 1172
2166 1835
1968 1671
1937 1744
1968 1800
1961 1763
2022 1652
NO „ c ,MH d
fnrw AN£ AW13
UUPIV » *•
33 N.A. .27
39 N.A. 1.08
60 N.A. N.A.
78 N.A. N.A.
100 .79 N.A.
100 .76 .71
100 .82 .70
100 .87 .91
100 .69 .76
98 .37 .54
100 .53 .57
68 .30 . 59
100 .82 .98
100 .81 .94
87 .09 1 . 20
86 .13 1.28
84 .30 1 . 07
85 .22 1.13
CO
f\)
Ammonia to nitric oxide mole ratio in the feed; the stoichiometric value of this ratio for the desired reaction
(6NO + 4NH3 > 5N2 + 6H20) is 0.667.
Concentration of nitrogen produced during the catalytic reduction of NO with NH,.
0
Mole ratio of nitrogen produced to nitric oxide consumed (experimental value).
Mole ratio of ammonia consumed to nitric oxide consumed; the theoretical value of this ratio for the
above reaction is 0.667.
-------
In order to isolate the NhU concentration effect from that of oxygen,
Experiment Nos. 5 through 12 were performed in which the NH3 to NO mole
ratio in the catalyst feed was varied from >3.2 to 0.399. In addition,
helium was used as the diluent gas instead of nitrogen; thus, the quantity
of nitrogen produced from the reduction of NO was determined. It is apparent
that in the absence of oxygen, N20 production decreased with increase in the
NhL concentration of the feed, all other parameters remaining constant,
while nitrogen production increased. These observations are illustrated in
Figure A-l. The exact shape of the curves may not be those drawn, due to
data scatter, but the trends are real.
The data from Experiment Nos. 5 through 8 verified Reaction 1A above, when
ammonia concentration in the feed stream was equal to or exceeded stoi-
chiometry and oxygen was not present. This is evident from the values of
the two ratios tabulated in the last two columns of Table A-l. Within experi-
mental uncertainty, the values of these ratios corresponded to those expected
from Reaction 1A. The theoretical values for AN2/ANO and ANhL/ANO are 0.833
and 0.677, respectively. When the NH3 to NO ratio in the catalysts feed
dropped below stoichiometry (Experiment Nos. 9 through 12), the ratios in
the last two columns of Table A-l were reduced in value, indicating a
different overall reduction reaction than that represented by Reaction 1A.
Two reactions that may take place during substoichiometric operation of the
process are:
4NO + 2NH3 -»• N20 + 2N£ + 3H20 (2A)
and 8NO + 2NH3 -> 5N20 + 3H20 (3A)
Thus, Experiment Nos. 5 through 12 indicate the following:
• In the absence of oxygen and water vapor Reaction 1A
accurately represents the reduction of NO by NH^ on
Pt at 250°C provided at least stoichiometric amounts
of NH3 are present.
228
-------
400
000
E
Q.
O.
o
UJ
O
O
O
£
200.
500 1000
NH, CONCENTRATION (ppm)
1500
Figure A-l.
Effect of NH3 Concentration on N,,0 and N2
Production. 0.5% Pt on Alumina Catalyst,
1000 ppm NO (nom) in He, 20,000 hr'1 (STP),
250-268°C.
229
-------
• When substoichometric quantities of NH3 are present in
the catalyst feed stream (to as low as 40% below stoi-
chiometry), NO reduction remains near 100% and part of
the NO is reduced to N^O.
• The nitrous oxide, N20, is not significant in the cata-
lyst effluent when NH3 in the feed exceeds stoichiometry,
in the absence of $2» an<* NO conversion is complete.
Experiment Nos. 14 through 18 were performed in order to more quantitatively
determine the oxygen effect. The reactor (catalyst) temperature was kept at
250°C, space velocity at 20,000 hr"1 (STP), and the feed NH3 to NO ratio
nearly constant at approximately 60% above sotichiometry (1.11 to 1.20).
The oxygen concentration was varied from 0 to 10,000 ppm (1%).
The data from Experiment No. 14 (80 ppm oxygen in the feed gas) indicates
compliance with Reaction 1A. Thus, N20 production is zero, N2 production
is very near to the expected value, and the AN2 to ANO ratios are close to
the theoretical values. The only discrepancy appears in the ANH~ to ANO
ratio which is about 20% higher than expected. No explanation is available
for this discrepancy. In Experiment No. 15, the oxygen level in the feed
gas was raised by approximately one order of magnitude (from 80 to 630 ppm).
The data indicated a dramatic oxygen effect on NO reduction. N20 production
increased from 0 to 710 ppm, N2 production decreased from 740 to 75 ppm, NO
conversion decreased from 100 to 87%, the AN2 to ANO ratio value was reduced
to 0.09 from 0.81, and the ANH3 to ANO ratio reached 1.20 from 0.94. The
change in the last quantity indicates excess NH3 utilization which can only
be explained through oxygen oxidation. One possible reaction is
2NH3 + 202 -*• N20 + 3H20 (4A)
Reaction 4A apparently takes place in parallel to Reactions 1A, 2A, or 3A,
above. It is possible that other reactions, not postulated here, take place.
However, whatever the reactions and reaction mechanisms, the fact remains
that under the conditions of described experiments, the presence of oxygen
230
-------
in the flue gas at a concentration of at least 630 ppm, influences both the
extent of NO conversion and catalyst selectivity. According to the data
from Experiment Nos. 16 through 18, the oxygen effect levels off at higher
than 630 ppm oxygen concentration in the feed (within experimental uncer-
tainty).
The fact that introduction of 02 into tne test 9as mixtures caused a de-
crease in the conversion of NO was somewhat disturbing in view of reports
in the literature that 02 enhances the reduction of NO with NH3 on Pt
(Section 3.2.2). In addition, the tests conducted had not determined that
an NO conversion maximum exists. Additional tests were performed with
variable 02 concentration over a wider range in temperatures using a second
sample of Pt-Al203 catalyst (NA-2).
Table A-2 shows the results of a series of experiments in which COp and FLO
were not incorporated in the test gas mixture. Oxygen concentration was
varied from 0 to 3000 ppm.
As before complete conversion of NO occurred with about 1100 ppm NH3 in the
absence of 02 at about 250° (see Runs 10 and 11 in this table and Run 8 in
Table A-l). Very little N20 was produced with excess NH~ in the absence of
02 as described before.
Figure A-2 shows the effect of temperature on conversion and N20 production
more clearly. The reason for the discrepancy between reported enhancement
and the observed retardation by 02 at 250°C, as indicated earlier, is now
apparent. Above 220°C 02 indeed has a retarding effect while below 220°C
the effect is one of enhancement. The effect of 02 on N20 production is
quite pronounced in the 0-800 ppm range; however, there appears to be little
influence due to 02 concentration between 800 and 3000 ppm on either NO con-
version or NpO production. In fact, at 250°C, the conversion indicated for
the 02 tests is in close agreement with that obtained earlier (Table A-l,
Run 4) for tests that included C02 and H20 in the mixture, indicating that
neither of these constituents is particularly influential to the selective
NO reduction by ammonia.
231
-------
Table A-2. EFFECT OF 02 ON REDUCTION OF NO WITH NH3 ON Pt-Al203 (NA-2)
Run
No.
1
2
3
4
5
6
7
8
9
111
11
12
13
14
15
16
Temperature
(°C)
290
260
230
190
167
147
126
125
205
240
285
410
256
195
178
162
Inlet Gas
Composition (ppm)
NO 02 NH3
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
3000
3000
3000
3000
3000
3000
3000
0
0
0
0
784
784
784
784
784
1141
1141
1141
1141
1141
1141
1141
1141
1141
1141
1141
1137
1137
1137
1137
1137
Outlet Gas
Composition (ppm)
NO N20 NH3
200
145
100
75
40
62
468
1000
277
25
0
360
103
65
92
123
520
648
764
725
807
657
176
0
60
0
N.A.
235
671
823
756
613
0
0
0
0
0
52
608
1042
735
534
432
29
12
14
48
98
NO
Conversion
(°/\
\/o)
80.0
85.5
90.0
92.5
96.0
93.8
53.2
0
72.3
97.5
100
64.0
89.7
93.5
90.8
87.7
a Only NO, 02 and N2 carrier gas were present. Space velocity ^20,000 hr~
233
-------
1000 ppm NO; 1140 ppm NH3 in N2; 20,000 hr'1 (STP) space velocity;
O • Without 02
O • 3000 ppm 02
A A 784 ppm 02
V V D^a taken with 3% 02, 5% H20, 14% C02 also present
Dotted line denotes maximum possible N20 due to NO-NH? reaction alone.
100 r-
-ilOOO
O.
a.
o
LU
o
13
O
O
ce
a.
o
CVI
100
200 300
TEMPERATURE (°C)
400
Figure A-2. Effect of 0^ on Selective Reduction of NO with NH3
and Production of N20 on Pt-Al 0- (NA-2)
235
-------
Up to the temperature for which a maximum in NO conversion occurred in the
presence of 02, the N20 produced corresponds closely with the maximum
possible due to Reaction 3A alone, based on NO conversion observed (as
shown by the dotted line in Figure A-2). Above that point N20 must be a
product of the reaction between NhU and Op-
The type of data taken in this project cannot be used to draw mechanistic
conclusions for either the relative participation of NO and NH-, in the
formation of N«0 or the shift in the NO conversion curve to lower temper-
atures in the presence of Og. Studies to determine such mechanisms, per-
haps through the use of isotope labelling, are highly recommended.
236
-------
APPENDIX B
Table B-l. PARTIAL LISTING OF CONTACTED EQUIPMENT MANUFACTURERS AND SUPPLIERS
(Contact was made with the local representatives and/or engineering
personnel at the headquarters)
Fans and Motors
Buffalo Forge Company
Zurn Air Systems Divsion
Westinghouse Electric Corporation
Aerovent, Inc.
Garden City Fans and Blower Company
General Electric Company
Refractory Lining and Insulation
Babcock & Wilcox Refractories Division
Catalysts/Chemicals
The Harshaw Chemical Company, Catalyst Department
Engelhard Industries Division, Engelhard Minerals and Chemical Corporation
U.S. Fuel
USS Agri-Chemicals Division
Frank Davis Company, Subsidiary of Rockwood Industries
Bethlehem Steel Corporation
Cyclones
Western Precipitation Division, Joy Manufacturing Company
Envirotech, Air Pollution Control Group
Control Damper
North American Manufacturing
Zurn Air Systems Division
Feeders and Associated Equipment
Green!ee Engineering Company
Wallace and Tiernan Division, Penwalt Corporation
American Meter Company, A Division of Singer
Heat Exchanger ("Preheater")
Bigelow-Liptak Corporation
Air Preheater Company
Electric Rates
Southern California Edison Company
CO Generator Vessel, Chemical Storage Tanks and Duct Material
Peabody Engineering and Supply Company
Buffalo Tank Division, Bethlehem Steel Corporation
American Bridge Division, U. S. Steel
Chicago Bridge and Iron
Production SLeel
237
-------
h I
c. ••
SULFUR
TRAP
WET TEST
FLOW METER
NO ANALYZER
SOAP
BUBBLER
CO GENERATOR "BOILER"
(O2 REDUCER]
GAS
CHROMATOGRAPH
VENT
VENT
VENT
- FLOWMETERS
- ELECTRICALLY HEATED
AND INSULATED
(J) - 3 WAY VALVES ® - 2 WAY VALVES
Figure B-l. Schematic Diagram of Bench Scale NOX-SOX Reduction Process Apparatus
-------
Table B-2. SUMMARY OF MASS BALANCE CALCULATIONS FOR THE NO -SO
CATALYTIC REDUCTION PROCESS FOR NEW POWER PLANTS
(37.5% of flue gas diverted to reductant generator
at a temperature of 1700°F)
Let a = coal feed rate for CO generator, tons/hr
(320-a) = coal feed rate for primary burners, tons/hr
Flue gas generated at primary burners = (320-a) (281,250 SCFH/ton), SCFH
Flue gas diverted to reductant generator = (320-a) (281,250) (0.375)
= (105,467) (320-a), SCFH
= 278 (320-a) Ib-mole/hr
Oxygen in flue gas CO generator feed (3%) = 8.34 (320-a) Ib-mole/hr
Reactions in the reductant generator
C + 1/2 02 = CO (1)
2 (-H) +1/2 02 - H20 (2)
C + H20 = CO + H2 (3)
C + C02 = 2CO (4)
Moi sture content of coal: 4%
Total hydrogen and total oxygen content of coal 5 and 10% respectively
Non-water hydrogen: 5-4 (2/18) = 4.6%
Non-water oxygen: 10-4 (16/18) = 6.4%
Non-water hydrogen in coal available for reaction (2):
(a tons/hr) (2000 Ibs/ton) (0.046) = ^ & lb/mole/hr
Non-water oxygen in coal: ^a^20°°)(°-064)= 4a Ib-mole/hr
Assuming oxidation of all the non-water hydrogen, oxygen consumed in
reaction (2): 46a 0. ,. , ..
' -j- = 23a Ib-mole/hr
Flue gas oxygen available for reaction (1): 8.34 (320-a) + 4a - 23a
= (2667 - 27.33a) Ib-mole/hr
At 1700°F, contributions from reactions 1, 3, and 4 to the total CO pro-
duction are 45, 29 and 26%, respectively. Therefore, for 1 Ib-mole of CO
produced, 0.45 Ib-mole originate from reaction (1), 0.29 Ib-mole originate
from reaction (3) and 0.26 Ib-mole originate from reaction (4). These
reactions would consume a total of 0.87 Ib-mole of carbon (0.45, 0.29, 0.13
Ib-mole in reactions 1, 3 and 4, respectively). The fraction of carbon con-
sumed in reactions (1), (3) and (4) are, therefore 0.517, 0.333, and 0.150,
respectively.
241
-------
Table B-2.(Continued) SUMMARY OF MASS BALANCE CALCULATIONS FOR THE NOX-SOX
CATALYTIC REDUCTION PROCESS FOR NEW POWER PLANTS
Carbon consumed in reaction (1): 2(2667-27.33a) Ib-mole/hr
= 24(2667-27.33a) Ibs/hr
= 0.012 (2667-27.33a) tons/hr
Total coal consumed [70% carbon, 51.7% in reaction (1)]
(88.4 - 0.91a) ton/hr
Coal supplied to reductant generator (a) = coal consumed in reductant generator,
therefore, a = 88.4 - 0.91a
a = 46 tons/hr
Coal feed rate to reductant generator = 46 tons/hr
Coal feed rate to primary burner: 320 - 46 = 274 tons/hr
Carbon in the 46 tons/hr of coal fed to reductant generator = 5367 Ib-mole/hr
Ib-mole/hr CO produced: |^- = 6168
Ib-mole/hr H2 produced: (6168)(0.29) = 1788
Ib-mole/hr oxygen in the undiverted portion (62.5%) of flue gas: 3804
Ib-mole/hr S02 in the undiverted flue gas (0.2%): 254
Ib-mole/hr NO in the undiverted flue gas (0.1%): 127
Ib-mole/hr sulfur (H^S and S) in the CO generator effluent gas:
from undiverted flue gas 152
from coal (50% of sulfur released) 43^
Total 195
Assuming 60% elemental sulfur, 40% HgS, total oxygen consumed in "boiler"
for conversion to SO^: 234 Ib-mole/hr
Amount of SOp produced: 195 Ib-mole/hr
Total amount of S02 to be reduced catalytically: 195 + 254 = 449 Ib-mole/hr
CO required for catalytic S02 reduction (S02+2CO 2C02 + 1/n $n): 898 Ib-mole/hr
CO required for catalytic NO reduction (NO + CO C02 + 1/2 N2): 127 Ib-mole/hr
"Excess" CO to be oxidized by oxygen in the undiverted flue gas:
6168 - (898 + 127) = 5143 Ib-mole/hr
Oxygen required for oxidation of "excess" CO: 2571 Ib-mole/hr
Oxygen required for oxidation of FL: 894 Ib-mole/hr
Oxygen required for oxidation of H2S and S: 234 Ib-mole/hr
Total amount of oxygen required = 2571 + 894 + 234: 3699 Ib-mole/hr
Total amount of oxygen available: 3804 Ib-mole/hr
Excess oxygen: 105 Ib-mole/hr
243
-------
TABLE B-3. DESIGN AND CAPITAL COST DATA FOR THE NOX-SOX CATALYTIC REDUCTION
PROCESSES FOR NEW POWER PLANTS
Item
Description
Est. Cost
($000)
l.Reductant Generation Section
a) Diversion Damper
b) Coal Feeder
c) Reductant Generator
Vessels
d) Ductwork
2. Catalytic Reaction Section
a) Ductwork
b) Multiclones
c) Catalytic Reactor
3. Induced Draft Fans and
Motors
4. Sulfur Collection Section
a) Ductwork
b) Multiclones
c) Forced Draft Fans
and Motors
22.5' x 22.5', based on flue gas velocity of 4000 ACFM
Chain type bucket elevator 130 ft long
4 vessels, 30 ft diameter, 15 ft high, containing 7 ft of
coal (bed volume based on a space velocity of 1500 V/V/hr
Field erected from carbon steel sheets (1/4-in thick at
bottom and first ring, 3/16-in thick at other places).
Expected pressure drop 10 in of water (based on coal particle
size of 1/2-in and a porosity of 50S)
Insulation and refractory lining: 1 layer of firebrick, 1 layer
of Insulating brick. (Total material cost ?TMP sq ft)
250 ft 30 x 30, 1/4-in carbon steel ($10.20 Ib/sq ft $900/ton)
Internal insulation with 3 in vacuum formed ceramic fiber wool
(S6.00/sq ft), 1.5 in mineral wool backing ($0.63/sq ft),
abrasion resistance ceramic coating ($.17/sq ft), studs
(S0.62/sq ft); total estimated material cost at S7.42
500 ft, 40' x 40', 1/4-in carbon steel (10.20 Ib/sq ft S900/ton)
Internal insulation (see item Id above)
Eight units, Western Precipitation Type, 12V •„ 35 sixe 270-15
(quoted fabrication price $482,000)
External insulation (for estimated external surface area of
3070 sq ft per unit) with 1.5-in 8-15 ceramic fiber blanket
(1.90/sq ft), 1-in mineral wool backing ($0.63/sq ft), studs
($1 sq ft); total estimated material cost $3.53/sq ft.
30 ft diameter, 130 ft high, consisting of 13 sections
(Konsanto multi-stage design). Monsanto 1965 estimated
installed cost S2500/ft of height. Estimated 1973 price
based on assumed escalation rate of 6° per year
($3985/ft of height)
Estimated pressure drop through bed 2 in H_0
Externsl insulation (same as for multiclones - item 2b, above)
Catalyst charge 570,000 Ibs, 1/2-in tablets, 50. porosity,
65 Ib/cu ft bulk density, S0.50/lb
Catalyst depth per section 1 ft (based on space velocity of
9000 V/V/hr)
250
15
185
295
345
555
460
740
990
220
820
120
285
Fans, 4 units, each 500,000 cfm capacity; total system
standard static pressure of 20 in water (price for Buffalo
Forge Fans Model *1780 L39, 885 RPM, 1400 HP, S61,000/fan f.o.b.) 500
Motors 4 units (G.E. type K motors, 900 RPM, 1500 HP, $30,000/unit) 250
250 ft 25' x 25'; 1/4-in carbon steel (10.20 Ib/sq ft $900/ton)
250 ft 30' x 30'; 1/4-in carbon steel (10.20 Ib/sq ft $900/ton)
5 units (same as item 2-b above, no insulation)
2 units, each capable of handling a volume of 169,000 cfm
against pressure of 2 in water (price for Aerovent
vaneaxial for Model VW 849B, 727 RPM with 125 HP motor
56,300 per unit)
285
345
620
25
1,645
3,635
750
1,275
7,305
245
-------
Table B-4. OPERATING COST DATA FOR THE NOX-SOX CATALYTIC REDUCTION
PROCESS FOR NEW POWER PLANTS
Item
Description
Est. Cost
($000)
Per Year
1. Depreciation
2. Maintenance, etc.
3. Labor
4. Electric power
5. Energy losses
6. Catalyst replacement
10 year straight line (10% of capital
investment per year)
Maintenance, insurance, taxes, etc.;
estimated at 8% of capital investment
per year
Labor at $5/hr, 3 operating positions,
100% overhead, 8000 working hours per
year
Estimated power consumption 6250 HP, 8000
operating hours per year, 1.4^/KW-hr
(based on Southern California Edison
Company schedule A-8 for large electric
users)
Equivalent of 3 tons/hr of coal; (esti-
mate includes allowance for heat loss
from insulated surfaces, heat re-
jected in ash discharged from CO
generator and multiclones, and heating
value not recovered due to CO consump-
tion in the catalytic reactor); estimated
price of coal (70% carbon content) $7/ton
Catalyst cost $0.50/lb, once per year re-
placement of 570,000 Ibs of catalyst
730
584
240
522
168
285
2,529
247
-------
Table B-5. SUMMARY OF MASS BALANCE CALCULATIONS FOR THE NO -SO
CATALYTIC REDUCTION PROCESS FOR EXISTING POWER PLANTS
Coal feed rate for primary burner = 320 tons/hr
Flue gas generated at primary burner = 90 x 106 MM SCFH
let x = fraction of flue gas diverted to CO generator
(1 - x) = undiverted fraction of flue gas
Y = coal feed rate for CO generator, tons/hr
Carbon feed for reductant generator = (YH°-7°) 200° = 117Y Ib-mole/hr
Non-water hydrogen in reductant _ (Y)(0.046)(2000) .cv ,. .. .,
generator feed 2 46Y lb-™>le/hr
Non-water oxygen in reductant _ (Y)(0.046)(2000) .v ,. , ..
generator feed ' 32 = 4Y lb-™le/hr
Oxygen in flue gas feed for reductant (90 x 106)(x)(0.03) ,,,<- ,. , ..
generator 379 = 7116X Ib-mole/hr
Reactions in reductant generator and estimated per cent contribution from
each reaction to the total CO production:
C + 1/2 02 = CO 95% (1)
2(-H) + 1/2 02 = H20 - (2)
C + H20 = CO + H2 5% (3)
C + C02 = 2 CO 0% (4)
One Ib-mole of C consumed (0.95 Ib-mole in reaction 1, 0.05 Ib-mole in
reaction 3) produces 1 Ib-mole of CO (0.95 Ib-mole from reaction 1,
0.05 Ib-mole from reaction 3)
Oxygen consumed in reaction 2 = 23Y
Oxygen available for reaction 1 = 7116x - 23Y + 4Y
= 7116x - 19Y
Carbon consumed in reaction 1 = 14,232x - 38Y
Carbon consumed in reductant = _> (14>232x - 38Y) = 14,981x - 40Y, therefore
generator u-yb
117Y = 14981X - 40Y and Y = 95.4x
249
-------
Table B-5. (Continued) SUMMARY OF MASS BALANCE CALCULATIONS FOR THE NOx~SOx
CATALYTIC REDUCTION PROCESS FOR EXISTING POWER PLANTS
For total oxygen consumption in reductant generator:
CO produced = 117Y Ib-mole/hr
H2 produced = (117Y)(0.05) = 5.85Y 16-mole/hr
Oxygen-consuming reaction in subsequent gas mixing and catalytic reduction
CO + 1/2 02 + C02 (5)
H2 + 1/2 02 -> H20 (6)
2CO + S02 +2 C02 + 1/n Sn (7)
CO + NO + C02 + 1/2 N2 (8)
S02 content of undiverted flue gas = 474.4 (T-x) Ib-mole/hr
NO content of undiverted flue gas = 237.2 (1-x) Ib-mole/hr
CO required for S02 reduction 948.8 (1-x) Ib-mole/hr
CO required for NO reduction 237.2 (1-x) Ib-mole/hr
Total CO required for catalytic reduction = 1186 (1-x) Ib-mole/hr
CO to be oxidized in reaction 5 = 117Y - 1186 (1-x) Ib-mole/hr
Oxygen required for reaction 5 = 58.5Y - 593 (1-x) Ib-mole/hr
Oxygen required for reaction 6 = 2.92Y - Ib-mole/hr
Total oxygen required for reactions 5 and 6 = 61.42Y - 593 (1-x) Ib-mole/hr
Oxygen required = oxygen available
61.42Y - 593 (1-x) = 7116 (1-x)
Y = 125.5 (1-x)
Y = 95.4x (from above), therefore
95.4 x = 125.5 (1- )
X = 0.57
Y = 54 tons/hr
251
-------
Table B-6. DESIGN AND CAPITAL COST DATA FOR THE WX-SOX CATALYTIC
REDUCTION PROCESS FOR EXISTING POWER PLANTS
Item
Description
Est. Cost
(SOOO)
1. Reductant Generation Section
a) Multiclones
b) Coal Feeder
c) Reductant Generator
Vessels
d) Ductwork
2.
4.
5.
Catalytic Reaction Section
a) Ductwork
b) Multiclones
c) Catalytic reactor
3. Heat Exchanger
2 units (see item 2b, Table B-3)
Chain type bucket elevator, 30 ft long
4 vessels, 40 ft diameter 15 ft high, containing 7 -t of coal
(bed volume based on a space velocity of 1500 V/V hr) (see
item Ic, Table B-3 for descriptive details); expected pressure
drop 10-in of water insulation (see item insulation and item
Ic, Table B-3)
To carry flue gas feed for CO generator, 250 ft 30' x 30 (see
item Id, Table B-3 for descriptive detail)
Internal insulation (see item Id, Table B-3)
To carry undiverted flue gas to catalyst, 250 ft 20' x 20' (see
item Id, Table B-3)
External insulation (at $3.53 material cost/sq ft - see item 2b,
Table B-3)
Ductwork to carry flue gas to waste heat boiler, 250 ft 40 x 40
(see item Id, Table B-3)
Internal insulation (see item Id, Table B-3)
5 units (see item 2b, Table B-3)
External insulation (see item 2b, Table B-3)
30 ft diameter 150 ft high 115 sections (see item 2c, Table B-3)
External insulation (see item 2c, Table B-3)
Catalyst charge, 655,000 Ib (see item 2c, Table B-3)
Expected pressure drop through bed 2-in H-0
Depth of bed per section 1 ft (based on space velocity of 9000 V/V/hr)
250
15
455
345
555
1.910
230
40
740
620
135
950
135
325
3,635
20 units 12' x 22' x 13', containing 528 tubes 4-1/2" O.D. (stain- 6,870
less steel and carbon steel tubes; hot side 1 pass coal sic'e 3 passes)
Estimated heat transfer rate 625 MM Btu/hr
External insulation (at $3.53 sq ft material cost - see item 2b,
Table B-3)
Induced Draft Fans and Motors 5 fans (see item 3, Table B-3)
5 motors (see item 3, Table B-3)
Sulfur Collection Section
a) Ductwork
b) Multiclones
c) Forced Draft Fans
500 30' x 30', no insulation (see item 4a, Table B-3)
5 units (see item 4b, Table B-3)
2 units (see item 4c, Table B-3)
250
7,120
625
310
935
690
620
25
1,335
14,935
253
-------
Table B-7. OPERATING COST DATA FOR THE N0x-S0x CATALYTIC
REDUCTION PROCESS FOR EXISTING POWER PLANTS
Item
Description
Est. Cost ($000)
Per Year
1. Depreciation
2. Maintenance, etc.
3. Labor
4. Electric power
5. Energy losses
6. Catalyst
10 year straight line (10% of capital)
Maintenance, insurance, taxes, etc.
estimated at 8% of capital per year
Labor at $5/hr, 3 operating positions,
100% overhead, 8000 working hours per
year
Estimate power consumption 7750 HP, 8000
working hours per year; 1.4^/KW-hr (see
item 4, Table B-4)
Equivalent of 3 tons of coal/hr (see item
5, Table B-4)
Catalyst cost $0.5/lb, once per year
replacement of 655,000 Ibs of
catalyst
1,493
1,195
240
575
168
327
3,998
255
-------
Table B-8. DESIGN AND CAPITAL COST DATA FOR THE
SULFIDE NOV - SOV REDUCTION
A A
Item
Description
Est. Cost
($000)
1. Coal/Iron Feeder
2. Reductant Generator
Vessels
3. Ductwork
4. Multiclones
5. Induced Draft Fans
and Motors
Chain type bucket elevator, 25
30 ft long
8 units, 30 ft diameter, 15 ft high,
containing 7 ft of coal (see
item lc, Table B-3)
Vessel cost 370
Insulation and refractory lining 590
250 ft of 50' x 50' (see item Id,
Table B-3)
Duct 575
Internal insulation 925
250 ft of 45' x 45' (see item Id,
Table B-3)
Duct 515
Internal insulation 835
4 units (see item 2b, Table B-3) 495
no external insulation
4 fans (see item 3, Table B-3) 500
4 motors (see item 3, Table B-3) 245
5,075
257
-------
Table B-9. OPERATING COST DATA FOR THE SULFIDE NO -SOY REDUCTION SCHEME
y\ A
Item
Description
Est. Cost ($000)
Per Year
1. Depreciation
2. Maintenance, etc.
3. Labor
4. Electric power
5. Energy losses
6. Iron Scraps
10 year straight line (10% of capital
investment per year)
Maintenance, insurance, taxes, etc.
(8% of capital investment per year)
Labor at $5 per hour, 2 operating
positions, 100% overhead, 8000
working hours per year
Estimated power consumption 6000 HP,
8000 operating hours, 1.4<£/KW-hr
(based on Southern California Edison
Company schedule B-8 for large
eletric users)
Equivalent of 3 tons/hr (see item 6,
Table B-4)
14 tons/hr, $69/ton (price for ferrous
scrap in Los Angeles, Iron Age,
January 24, 1974)
508
406
160
512
168
7,728
9,482
259
-------
Table B-10.
DESIGN AND CAPITAL COST DATA FOR NOX ABATEMENT
BY REDUCTION WITH NH3 - NON-NOBLE METAL CATALYSTS
Item
Description
Est. Cost
($000)
1. Ductwork
2. Multic!ones
3. Catalytic
Reactor
4. Ammonia Feeding
System
a) Storage Tank
b) Pressure Re-
ducing Valve
and Flow
Meter
5. Induced Draft
Fans and
Motors
250 ft 40' x 40' (see item Id, Table B-3) 460
Internal insulation with 1.5 in of vacuum
formed ceramic fiber ($3.00/sq ft), 1.5-in
mineral wool backing ($0.63/s ft), ceramic
fiber coating cement ($0.17/ sq ft) and studs
$0.e2/sq ft); total insulation material $4.42/
sq ft 440
250 ft 35' x 35' (see item Id, Table B-3) 400
Internal insulation (same as above) 385
6 units (see item 2b, Table B-3) 745
External insulation (see item 2b, Table B-3) 160
30 ft diameter 170 ft high, 17 sections,
estimated actual pressure drop 2.7-in
water, (see item 2c, Table B-3) 1,070
Catalyst charge 300,000 Ibs, 1/2 in tablets, 195
50% porosity, 65 Ib/cu ft bulk density,
$0.50/lb, catalyst depth per section 1/2
ft (based on space velocity of 15,000 V/V/Hr)
External insulation (at $3.53/sq ft material
cost: - see item 2c, Table B-3)
30 ft diameter spherical tank, 5 day ammonia
demand storage capacity (for 250 psi inter-
nal pressure), field erected
2 in axial flow stainless steel controller
12 in turbine meter (maximum capacity
150,000 cu ft/hr)
5 fans, 500,000 cfm capacity each, total
system standard static pressure of 9 in
H20 (prices for Buffalo Forge Model #1780,
L-39 705 RPM, 1100 BMP $47,500/unit)
5 motors (GE type K motors, 720 RPM, 1250
HP $28,500 per motor)
240
165
490
170
4,925
267
-------
Table B-ll.
DESIGN AND CAPITAL COST DATA FOR NOX ABATEMENT BY REDUCTION
WITH NH3 - PLATINUM CATALYST
Item
Description
Est. Cost
($000)
1. Ductwork
2. Multiclones
3. Catalytic
Reactor
4. Ammonia Feeding
System
a) Storage Tank
b) Pressure Re-
ducing valve
and flow
meter
5. Induced Draft
fans and
Motors
300 ft 35' x 35' (see item Id, Table B-3)
External insulation at $3.53/sq ft mate-
rial cost (see item 2c, Table B-3)
5 units (see item 2b, Table B-3)
External insulation at $3.53/sq ft mate-
rial cost (see item 2c, Table B-3)
30 ft. diameter, 130 ft high, 13 sections
estimated actual pressure drop 2.6 in
water, (see item 2c, Table B-3)
External insulation at $3.53/sq ft mate-
rial cost (see item 2c, Table B-3)
Catalyst charge 270,000 Ibs, 1/8 in
tablets, 50% porosity, 60 Ib per cubic
ft bulk density, catalyst depth per
sectio Nl/2 ft (based on space velocity
of 20,000 V/V/hr), $35.55 per Kg ($26.20
platinum cost, $9.35 manufacturing cost)
See item 4a, Table B-10
See item 4b, Table B-10
5 fans (see item 5, Table B-10)
5 motors (see item 5, Table B-10)
480
370
620
135
815
120
4,350
165
5
490
170
7,720
263
-------
Table B-12.
OPERATING COST DATA FOR NOX ABATEMENT BY REDUCTION WITH
NON-NOBLE METAL AND PLATINUM CATALYSTS
Est. Cost Per Year ($000)
Item
Description
Non-Noble
Metal
Catalyst
Platinum
Catalyst
1. Depreciation
2. Maintenance,
etc.
3. Labor
4.
5.
Electric
Power
Energy
Losses
6. Ammonia
Consumption
7. Catalyst
Replacement
10 year straight line (10% of
capital investment per year)
Maintenance, insurance, taxes, etc.
(8% of capital investment/year)
Labor at $5 per hour, 2 operating
positions, 100% overhead, 8000
working hours per year
Based on 6250 HP, 1.4tf/KW-hr (see
item 4, Table )
For non-noble metal catalyst; equi-
valent of 0.3 tons/hr coal
For noble metal equivalent of 3.2
tons/hr of coal (it includes
energy loss due to gas diversion
at 400°F instead of 350°F)
Non-noble metal catalyst: 4020 Ibs/hr
NH3 at 0.12
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse-before completing)
1. REPORT NO.
EPA-650/2-75-001-a
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Assessment of Catalysts for Control of NOx from
Stationary Power Plants, Phase 1, Volume I
5. REPORT DATE
January 1975
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S),
E. P. Koutsoukos, J. L. Blumenthal, and
M.Ghassemi (TRW), and G.Bauerle (UCLA)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
TRW Systems Group
One Space Park
Redondo Beach, CA 90278
10. PROGRAM ELEMENT NO.
1AB014; ROAP 21ADF-003
11. CONTRACT/GRANT NO.
68-02-0648
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
EPA, Office of Research and Development
NERC-RTP, Control Systems Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIC
Final; Through 5/74
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
is. ABSTRACT rp^g report summarizes results of investigations to assess the technical
and economic feasibility of utilizing catalytic processes in power plant nitrogen
oxide (NOx) emission abatement. The investigations involved a literature survey
and the development of a data bank on pertinent articles and patents, experimental
screening tests on selected promising catalysts, and preliminary design and cost
analysis of candidate processes adapted to new and/or existing power plants. The
stepwise selection and prioritization of catalysts indicated that at least two types of
catalytic NOx control processes should be adaptable to power generating plants:
selective reduction of NOx with ammonia on non-noble metal catalysts; and
simultaneous nonselective reduction of NOx and sulfur oxides with coal-derived
reductants on non-noble metal catalysts.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Nitrogen Oxides
Catalysis
Feasibility
Ammonia
Sulfur Oxides
Air Pollution Control
Stationary Sources
Non-Precious Metals
Coal-Derived Reduc-
tants
13B
07B
07D
14A
8. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
283
2O. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
267
------- |