EPA-600/7-76-031
U.S. Environmental Protection Agency Industrial Environmental Research EPA-600/7-76
Office of Research and Development Laboratory
Research Triangle Park. North Carolina 27711 October 1976
CATALYTIC REDUCTION OF
NITROGEN OXIDES WITH AMMONIA:
Utility Pilot Plant Operation
Interagency
Energy-Environment
Research and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S.
Environmental Protection Agency, have been grouped into seven series.
These seven broad categories were established to facilitate further
development and application of environmental technology. Elimination
of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The seven series
are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from
the effort funded under the 17-agency Federal Energy/Environment
Research and Development Program. These studies relate to EPA's
mission to protect the public health and welfare from adverse effects
of pollutants associated with energy systems. The goal of the Program
is to assure the rapid development of domestic energy supplies in an
environmentally—compatible manner by providing the necessary
environmental data and control technology. Investigations include
analyses of the transport of energy-related pollutants and their health
and ecological effects; assessments of, and development of, control
technologies for energy systems; and integrated assessments of a wide
range of energy-related environmental issues.
REVIEW NOTICE
This report has been reviewed by the participating Federal
Agencies, and approved for publication. Approval does riot
signify that the contents necessarily reflect the views and
policies of the Government, nor does mention of trade names
or commercial products constitute endorsement or recommen-
dation for use.
This document is available to the public through the National Technical
Information Service, Springfield, Virginia 22161.
-------
EPA-600/7-76-031
October 1976
CATALYTIC REDUCTION OF
NITROGEN OXIDES WITH AMMONIA:
UTILITY PILOT PLANT OPERATION
by
Jules M. Kline, Paul H. Owen, and Y.C. Lee
Environics, Lie.
4101 Westerly Place, Suite 107
Newport Beach, California 92660
Contract No. 68-02-0292
Program Element No. EHE624
EPA Project Officer: Richard D. Stern
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
CONTENTS
PAGE
List of Figures jy
List of Tables v
Acknowledgements vi
Sections
1. Summary 1.1
1.1 Contract Effective Date 1.1
1.2 Program Objectives 1.2
1.3 The Utility Pilot Plant 1.2
1.4 Gas-Fired Testing 1.3
1.5 Oil-Fired Testing 1.4
1.6 Cost Analysis (Gas-Fired Boiler) 1.6
2. Conclusions 2.1
3. Recommendations 3.1
4. Introduction 4.1
4.1 Nitrogen Oxide Emission and Control 4.1
4.2 Objective of Pilot Plant Program 4.2
4.3 Environics1 Previous Work 4.4
4.4 Discussion 4.6
5. Design of Utility Pilot Plant 5.1
5.1 System Design 5.1
5.2 Control System Design 5.6
5.3 Ammonia Injection System Design 5.9
5.4 Analytical Techniques 5.10
6. Laboratory Pilot Plant Testing 6.1
6.1 Long Duration Testing 6.1
6.2 Parametric Testing 6.3
6.3 Trace Products Analysis 6.4
7. Pilot Plant Installation and Checkout 7.1
7.1 Installation 7.1
7.2 Initial Checkout 7.1
7.3 Final Checkout 7.3
8. Gas Fired Pilot Plant Testing/Utility Pilot Plant . 8.1
8.1 Utility Pilot Plant Parametric Tests 8.1
8.2 Pilot Plant NH3 Mixing Tests 8.2
8.3 Pilot Plant Long Duration Tests 8.2
9. Oil Fired Utility Pilot Plant Testing 9.1
10. Oil Fired Laboratory Pilot Plant Testing 10.1
11. Cost Estimate-480 MW Plant 11.1
Appendix A A.I
ill
-------
LIST OF FIGURES
NO. PAGE
4.1 Temperature vs. Removal Efficiency 4.8
4.2 Removal Efficiency vs. Space Velocity 4.9
4.3 Removal Efficiency and Effluent NHo vs.
Inlet NH3 Catalyst 4.10
4.4 Effect of Exposure to 300 ppm S02 on Catalyst. . . 4.11
5.1 Utility Pilot Plant Schematic 5.12
5.2 Calculated Temperature Profile: Air Preheater
7HKX-22 5.13
5.3 Ammonia Mixing Test Results 5.14
5.4 Ammonia Mixing Test Results 5.15
5.5 Conversion Efficiency vs. Temperature for
NH3 Oxidizer 5.16
6.1 Catalyst Life Test in Weeks 6.5
6.2 Effect of Ammonia on Removal Eff 6.6
6.3 Effect of Space Velocity on Removal Eff 6.7
6.4 Effect of Temperature and NHo/NO Ratio on
Removal Eff 6.8
7.1 Photograph of Utility Pilot Plant 7.4
7.2 Photograph of Utility Pilot Plant 7.5
8.1 Optimum Test Conditions-Utility Pilot Plant. ... 8.5
8.2 Inlet Ammonia Concentration Distribution 8.6
8.3 Utility Pilot Plant Long Duration Test Results . . 8.7
8.4 Utility Pilot Plant Long Duration Test Results . . 8.8
8.5 Utility Pilot Plant Long Duration Test Results . . 8.9
10.1 NOx Removal Eff. vs. Inlet Ammonia Cone.
New Catalyst (Pt) 10.6
10.2 NOx Removal Eff. vs. Inlet Ammonia Cone.
(Utility Pilot Plant Catalyst) 10.7
10.3 NOx Removal Eff. vs. Temp. Laboratory Pilot
Plant 10.8
10.4 NOx Removal Eff. vs. Temp. Utility Pilot Plant
Catalyst - Regenerated 10.9
iv
-------
LIST OF TABLES
NO. PAGE
4.1 Catalysts Included in Small Scale Testing .... 4.12
4.2 Summary of Catalyst Performance: NO
(300 ppm) in N2 4.15
6.1 Trace Products Analysis by TRW 6.9
9.1 LADWP Test Report 9.6
11.1 Capital Outlay Schedule 11.4
11.2 Cost Summary 11.5
v
-------
ACKNOWLEDGEMENTS
Throughout the performance of the work reported herein,
many people have contributed their knowledge and know-how
toward the success of this program.
We wish to acknowledge the contribution of the Los
Angeles Department of Water and Power (LADWP), who allowed
the installation and operation of the pilot plant at their
Valley Steam Plant, Unit 4, despite the attendant inconven-
ience. We wish to thank LADWP management and particularly
Messrs. Hans Sonderling, Wesley Pepper, Charles Sun-Woo,
Irvine Tuttle and Lloyd Reeve for their cooperation and assis-
tance.
We also wish to acknowledge the significant contributions
of the Environmental Protection Agency, particularly Messrs.
Richard D. Stern, Roger Christman, Jim Wingo, Luis Garcia,
Kenneth Baker, and Dale Denny.
Some of the small scale work performed outside the scope
of this program, reported in Section 3.3 as background in-
formation, was performed with the support of the Southern
California Edison Company and the San Diego Gas and Electric
Company.
vi
-------
SECTION 1
SUMMARY
1.1 The contract #68-02-0292 was effective, beginning January
1973, although work actually began in October 1972.
1.2 PROGRAM OBJECTIVES
The principal objective of the contract was to treat the
flue gas of a gas fired boiler to demonstrate, on a utility
pilot plant scale, the performance and practical aspects of a
NOx reduction system, using a platinum catalyst with ammonia
as the reductant. Because of the decreasing availability of
natural gas for utility boilers, the program was expanded to
include tests on oil firing.
More specifically, the primary objective of this program
was to demonstrate that the combined emissions of NOx and NH3
could be controlled to less than 50 ppm.
The second objective was to establish the useful life of
the platinum catalyst. Chemical activity of the catalyst could
conceivably have been reduced by poisoning, or its physical
stability could have been impaired by exposure to the flue gas
environment. A projected catalyst life of 12 months or more
was desired.
The third objective was to establish optimum temperatures,
flow rates of NH^, size of required catalyst bed, and system
control parameters, for future use in designing full-scale flue
gas treatment to remove NOx.
The fourth objective was to determine the economic feasi-
bility of utilizing this process to reduce the NOx emissions
from a 500 MW gas-fired utility boiler to below 50 ppm. Capital
investment cost estimates, including the installation and/or
1.1
-------
modification of boiler components, and the initial cost and
replacement cost of the catalyst, and operating cost estimates,
including maintenance, as well as cost of the NH-j used as a
reductant, were desired.
1.3 THE UTILITY PILOT PLANT
Negotiations between Environics and the Los Angeles Depart-
ment of Water and Power began prior to the effective date of the
contract. An agreement was reached with the Los Angeles Depart-
ment of Water and Power (LADWP) in October 1972.
In this agreement, Boiler #4, having 170 MW capability,
at the San Fernando Facility of the LADWP, was designated as
the boiler for the pilot plant. Additionally, LADWP agreed to
furnish engineering services and approve modifications to the
boiler that may be necessary to accommodate the pilot plant.
LADWP was to receive all progress reports and the final report.
Periodic reviews were conducted with LADWP engineering per-
sonnel and operating personnel at the San Fernando facility.
The pilot plant was sized to handle a slipstream equivalent to
1.5 MW output (150,000 SCFH). Its design was based upon
Environics1 previous small scale test work.
1.3.1 Catalyst Selection
Environics' previous small scale tests had indicated that
a 0.37o platinum catalyst, supported on a high surface area
washcoated ceramic honeycomb, would provide high activity for
NOx removal in steam boiler flue gas with a relatively low
pressure drop through the catalyst bed. Thus, selection of the
pilot plant catalyst was limited to such a catalyst, with the
objective of selecting the proper size support, to provide
acceptable levels of activity and pressure drop for a utility
boiler. Testing was performed to determine pressure drop and
activity characteristics of several catalyst/support configurations
1.2!
-------
A realistic space velocity of 50,000 v/v/hr is targeted. The
catalyst volume is therefore 7 cu. ft.
The unique feature of the utility pilot plant is the in-
corporation of the catalyst in the air preheater by replacing
the heat transfer elements inside the air preheater with the
catalyst. This feature makes the use of catalytic reduction
to treat the flue gas simple and practical, particularly in
retrofitting an existing power plant. Because of the closely
coupled arrangement of the power plant components, there is
normally no room for the addition of a separate catalytic con-
verter. For a new plant, a separate catalytic converter to
treat its flue gas is feasible.
1.4 GAS-FIRED TESTING
1.4.1 The gas-fired testing began with small scale-long
duration testing. The small scale test boiler ran for
175 days (4200 hours) with no sign of catalyst deterioration.
Although there were several instances of soot accumulating on
the catalyst, these were considered due to upsets in the
operation of the small boiler, not to its normal function, and
the catalyst was easily regenerated. At the end of the test
period, 90% NOx removal was again obtained, with an (NOx + NHo)
exit concentration less than 50 ppm. The analysis for trace
products found no detectable contaminants from side reactions.
The small scale-long duration testing is discussed in detail
in Section 6.
1.4.2 The results of the utility pilot plant operation on gas
firing were equally encouraging. The pilot plant was
operated on gas firing for over 2000 hours and the catalyst
showed no sign of deterioration. Although steady state con-
ditions were difficult to maintain, it appeared that the pilot
1.3
-------
plant catalyst maintained 85-9070 NOx removal under optimum
conditions of 490°F and 200 ppm NHg.
That the pilot plant could not accumulate more running
time during the gas fired test period was due mainly to
problems with the boiler. These problems included long per-
iods when gas was not available, lengthy major maintenance
shutdowns, and minor maintenance shutdowns to repair tube
leaks and other malfunctions. The pilot plant itself, because
of its relative simplicity, was operated for essentially the
maximum time possible during the test period. The details are
discussed in Section 8.
A summary of the optimum operation conditions and its per-
formance are given below:
Catalyst Pt, 0.03% by weight.
Ammonia inlet 200 ppm.
Ammonia exit Less than 15 ppm.
Temperature 490°F.
Pilot plant flow rate 150,000 SCFH approximately.
Pilot plant power 1.5 MW
(equivalent)
Space velocity 50,000 v/v/hr.
NOx removal efficiency 85-90%
Running time, 2000 hrs.
accumulative
1.5 OIL-FIRED TESTING
1.5.1 Results of the oil fired testing, with the utility pilot
plant and laboratory pilot plant, were not as encourag-
ing as those on gas firing. Indeed, these results were some-
what contradictory to the results of Environics' previous small
scale work.
1.4
-------
The utility pilot plant was operated on oil firing for a
total time of over 400 hours. During this period, the maximum
NOx removal efficiency that was achieved was 65%; the average
NOx removal efficiency was 50%. Because of the difficulties
encountered in maintaining constant temperature in the pilot
plant, it was not possible to conduct accurate parametric
tests. However, it did appear that to obtain maximum NOx re-
moval efficiency, a catalyst temperature between 550 and 600°F
was required. In contrast to this, the previous small-scale
data indicated that peak NOx removal efficiency of 80-95%
could be obtained between 520 and 575°F. It is possible that
higher NOx removal efficiency could have been obtained in the
utility pilot plant if it would have been possible to maintain
a constant, higher temperature. This is because of the rela-
tively long time required to obtain steady state conditions in
the presence of S02 (presumably due to adsorbtion/desorbtion
of some NHo/S02 compound on the catalyst surface). It is also
possible, however, that the relatively poor results were due
to the effect of species in the flue gas other than SO^.
The other problems encountered during oil fired operation
of the pilot plant were catalyst plugging with soot and
ammonium sulfate accumulation. The plugging, observed on the
leading edge of the catalyst, could probably be minimized by
use of a larger cell size honeycomb support, countercurrent
operation of the air preheater, and/or by periodic cleaning
operations using compressed air or water. The white particu-
late accumulation could probably be minimized by use of an
automatic NH~ concentration control loop, to maintain an appro-
priate inlet NHo concentration as power plant load changes.
The details are given in Section 9.
1.5.2 However, the laboratory pilot plant tests did show that
the oil "poisoning," if any, is reversible; i.e., when
1.5
-------
the boiler was switched to gas fuel, the catalyst gradually
recovered its gas-fired performance.
1.6 COST ANALYSIS (GAS-FIRED BOILER)
Based on preliminary cost estimates that were performed,
it was expected that capital cost for the full scale system
would be below $15 per KW, and that operating costs would be
below 0.1 cents per KW hour. Capital cost is quite sensitive
to catalyst life, and thus it is expected that this program
would enable much more accurate cost estimates than have here-
tofore been possible.
Based upon program results, the capital cost for a full
scale catalytic reduction system is estimated to be $11 per KW
plant capacity. The total operating cost associated with such
a system is estimated to be 0.02«i per kw-hr. These estimates
were based on 1974 dollars and a gas fired 480 MW plant.
1.6
-------
SECTION 2
CONCLUSIONS
The following conclusions can be made about the results
of the pilot plant program.
1. Catalytic reduction is a practical method of reducing
NOx emissions from a gas fired (natural or synthetic gas)
steam boiler. Up to 907o reduction can be achieved, without
emission of excessive unreacted NH-j. The catalyst can be
conveniently placed in the air preheater wheel. Thus, the
retrofitting of an operating utility steam boiler with a
catalytic converter by replacing the heat transfer buckets
with catalytic elements without any change of the external
configuration is made simple, practical and inexpensive.
2. The effluent ammonia has consistently been less than
10 to 15 pptn.
3. The catalytic converter when using gaseous fuel con-
sistently yielded 85-9070 conversion at a space velocity of
50,000 v/v/hr.
4. Reliability of the catalytic converter system, using
gaseous fuel, has been demonstrated for more than 4,000 hours
of accumulative testing.
5. The controls associated with the catalytic converter
system can be simple, using a feedback system, which senses
either NOx or ammonia effluent or summation of both to regu-
late ammonia injection.
6. Catalytic reduction does not provide as great a
degree of NOx control in an oil-fired plant as in a gas-fired
plant. Perhaps 65 to 70% control may be achieved with oil.
Installation of the catalyst in the air preheater provides a
2.1
-------
desirable continuous regeneration effect. The soot accumula-
tion problem can be solved by a soot remover located in the
air preheater. The optimum NOx conversion temperature for
oil fired boilers is above 575°F.
7. In a plant fired with gas and oil, alternately, high
performance will return when firing is switched back to gas,
from oil.
8. The capital cost for a full scale catalytic reduction
system is estimated to be less than $11 per KW plant capacity.
The total operating cost associated with such a system is
estimated to be less than 0.024 per KW-hr. These estimates
were based on 1974 dollars and a gas fired 480 MW plant.
2.2
-------
SECTION 3
RECOMMENDATIONS
The following recommendations are made for continued
development of a catalytic NOx reduction system leading to a
large scale prototype pilot plant.
1. Further research should be performed, on a laboratory
pilot plant scale, to optimize the system for oil firing.
Such research should include tests of platinum and transition
metal catalysts, with particular attention to the beneficial
effects of operation in a rotating air preheater wheel. A
practical system must be capable of operating on both oil and
gas fuels.
2. Simultaneous with the laboratory pilot plant research,
the utility pilot plant testing should be continued. If pos-
sible, negotiation with LADWP to relocate the present pilot
plant either at the Valley Steam Plant or other location so
that the utility pilot plant can be operated at higher tempera-
tures .
3. A full scale prototype catalytic converter system
should be initiated for gas-fired plants such as LADWP Scatter-
good #3 or other LADWP's smaller plants. This recommendation
is directly applicable for synthetic gas plants of the future.
4. Long-term need is in the coal-fired plants. Research
should begin in catalytic removal of oxides of nitrogen,
associated with coal-firing.
3.1
-------
SECTION 4
INTRODUCTION
4.1 NITROGEN OXIDE EMISSION AND CONTROL
Furnaces, stationary engines and turbines burning fossil
fuels emit large quantities of nitrogen oxides. The sources
of these emissions are power plants, oil refineries, gas pump-
ing stations, and oil and gas producing facilities.
Combustion modification techniques appear to be the
quickest and least expensive methods to accomplish a first-
order reduction of NOx emissions from most stationary sources.
The techniques which have been used include low excess air
combustion, staged combustion, flue gas recirculation, steam
or water injection, and various combinations of these tech-
niques. A realistic control level using such techniques is
100 ppm NOx, for gas-fired furnaces, and higher levels for
staionary engines which burn natural gas, gasoline, jet fuel,
fuel oil, or coal.
However, the control of NOx emissions by combustion modi-
fication may not be applicable to those special processes
which require very high temperatures (i.e., glass manufacture),
compact furnaces or processes burning fuels containing large
amounts of organically bound nitrogen. Furthermore, it may be
desirable, at least in metropolitan areas, to reduce NOx emis-
sions more than is possible with combustion modification.
One method which has the potential for significant reduc-
tion of NOx emissions from stationary sources is catalytic re-
duction of NOx with ammonia. In this method, a controlled
amount of NH~ is mixed with the boiler flue gas, and the mix-
ture is passed over a catalyst at the proper conditions of
flowrate and temperature. The catalyst promotes the following
4.1
-------
reactions which convert the NOx and NH^ to harmless nitrogen
and water vapor:
6 NO + 4 NH3 -> 5N2 + 6H20 (1)
6 N02 + 8 NH3 ----> 7 N2 + 12 H20 (2)
Ammonia is considered a selective reductant for NOx since
it reacts via Reactions (1) and (2) in preference to the reac-
tions with oxygen given below:
3 02 + 4NH3 -> 2 N2 + 6 H20 (3)
5 02 + 4NH3 -> 4 NO + 6 H20 (4)
(In practice, Reactions (3) and (4) also proceed, but to
a lesser extent, or at a higher temperature, than Reactions
(1) and (2).)
Catalytic reduction of NOx with NHo can be employed as an
alternative to, or in combination with, combustion modification,
and can reduce NOx emissions by as much as 957o.
4.2 OBJECTIVE OF PILOT PLANT PROGRAM
The objective of the work reported herein was to demon-
strate, on a utility pilot plant scale, the practicability of
catalytic reduction of NOx with NH-, using a platinum catalyst.
The utility pilot plant was sized to handle a slipstream from
a utility steam boiler, equivalent to approximately 1.5 MW
output (fO 150,000 SCFH). Its design was based upon Environics'
extensive previous small scale test work.
The pilot plant was installed at the Los Angeles Depart-
ment of Water and Power Valley Steam Plant, Unit No. 4, a 170
MW steam boiler, equipped for natural gas and fuel oil firing.
Initially, it was desired to limit operation of the pilot
plant to gas firing only. Subsequently (in view of the decreas-
ing availability of natural gas fuel for utility boilers), the
4.2
-------
program was expanded to include tests on oil firing, as well.
More specifically, the primary objective of this program
was to demonstrate that the combined emissions of NOx and NEU
could be controlled to less than 50 ppm.
The second objective was to establish the useful life of
the platinum catalyst. Chemical activity of the catalyst
could conceivably have been reduced by poisoning, or its
physical stability could have been impaired by exposure to
the flue gas environment. A projected catalyst life of 12
months or more was desired.
The third objective was to establish and to confirm pre-
viously obtained operating parameters such as optimum tempera-
tures, flow rates of NH^size of required catalyst bed, and
system control parameters, for future use in designing full-
scale flue gas NOx removal equipment. Space velocity goal
was 50,000 v/v/hr.
The fourth objective was to determine the economic feasi-
bility of utilizing this process to reduce the NOx emissions
from a 500 MW gas fired utility boiler to below 50 ppm.
Capital investment cost estimates, including the installation
and/or modification of boiler components, and the initial
cost and replacement cost of the catalyst, and operating cost
estimates, including maintenance, as well as cost of the NHL
used as a reductant, were desired.
Based on preliminary cost estimates that were performed,
it was expected that the capital cost for a full scale system
would be below $15 per KW, and that operating costs would be
below 0.1 cents per KW hour. Capital cost is quite sensitive
to catalyst life, and thus this program would enable much
more accurate cost estimates than have heretofore been possible.
The contract #68-02-0292 was initiated during October 1972.
4.3
-------
Accordingly, the work reported herein was begun on October 1,
1972. Prior to that time, however, Environics had conducted
extensive testing of the catalytic reduction system, on a
small scale (100 - 3000 SCFH). The small scale work under-
taken before contract initiation is summarized as follows.
4.3 ENVIRONICS' PREVIOUS WORK
431 Gas Fired Small Scale Work
Prior to this program, Environics had conducted an exten-
sive literature search, followed by laboratory research on
catalytic reduction of NOx with NH., using numerous catalysts
In several different sized small scale test units. Some of
the catalyst tested are described in Table 4.1. Some small
scale field tests on actual boilers and gas turbines were
also performed. The results of this work are described below.
Initial laboratory work was performed using bottled gas
mixtures of NO in nitrogen (with the addition of 02 in some
tests), and a one inch diameter converter. Flowrates with
this bottled gas system were approximately 100-300 SCFH.
Results of these tests (summarized in Table 4.2) showed
that 90 to 987. NOx removal could be achieved, using either
platinum, copper oxides or vanadia based catalysts. This
work demonstrated the feasibility of application of catalytic
reduction with NH^ to control NOx emission from steam boilers.
Platinum catalysts were significantly more active than vanadia
or copper oxide, giving 951 removal at approximately 40,000
v/v/hr space velocity compared to 95-981 removal at 10,000
v/v/hr for CuO, and 93% removal at 10,000 v/v/hr for V205>
The next stage of the small scale work was initiated in
order to determine whether other components of flue gas (e.g. ,
water vapor. CO and C02) would affect the behavior of the
4.4
-------
various catalysts. In this stage of the program, a gas fired
furnace was used to provide flue gas to a two inch diameter
converter. Thus, the gas stream simulated the flue gas in a
steam boiler.
The small scale test program concentrated on platinum
catalysts, because of their high activity. Transition metal
catalysts, such as copper oxide, manganese oxide, etc., were
also tested extensively, however. Results of work with the
platinum catalysts will be discussed herein.
Experiments were performed to determine the effects of
temperature, space velocity and inlet NOx concentration, on
NOx removal efficiency, with several platinum catalysts.
A typical result showing the effect of temperature on
NOx removal efficiency is presented in Figure 4.1. As can be
seen, peak NOx removal efficiency of as high as 98% was ob-
tained in the temperature range of 400* - 500°F.
Results showing the effect of space velocity on NOx re-
moval efficiency are shown in Figure 4.2.
Experiments were also performed to determine the effect
of inlet NH.J concentration on both NOx removal efficiency and
NH.j effluent concentration. Results are shown in Figure 4.3;
for platinum catalyst, at 500°F.
The next phase of this small scale testing was to determine
the effect of S02 on performance of the various catalysts. This
work is described below.
4.3.2 S0? Experiments
It had been noted in the literature, and subsequently, that
information on the effects of SO, on platinum (and other) cata-
lysts was contradictory. In view of the increasing quantity of
fuel oil, relative to natural gas. being used by electrical
-------
generating stations, it was decided to direct the efforts of
the small scale testing towards determining the effect of SC^
on the NOx reduction system.
This work was performed using the test furnace system
described above, by injecting SC^ into the exhaust gas upstream
of the converter.
Initial SC^ tests were made using a platinum catalyst
operating at other than optimum conditions (450 to 500 F,
50,000 hr SV). The results were that, within several hours,
the catalyst showed a marked decrease in performance, with
NOx removal efficiency dropping to below 2070. It was even-
tually learned that this apparent S02 "poisoning" could be
prevented by increasing the temperature of the catalyst above
575°F.
A typical result of some of these SC^ experiments, with
platinum catalysts, is presented in Figure 4.4, showing NOx
removal efficiency as a function of catalyst age, with tests
at various temperatures as indicated.
4.4 DISCUSSION
The purpose of the following comments is to put the small
scale work described in Section 4.3, above, into the time
perspective of the pilot plant work described in the remainder
of this report.
At the time design of the pilot plant was started (October
1972), the literature search, bottled gas experiments, gas
fired furnace experiments and a portion of the S02 experiments
had been completed. The pilot plant was designed to operate
only on natural gas fuel. Such operation was considered prac-
tical for a utility boiler in view of the fuel availability
situation at that time.
The remainder of the S02 experiments was conducted during
4.6
-------
the period in which the utility pilot plant was being designed
and operated. In view of the results of that work, and the
increasing shortfall in natural gas supply, the pilot plant
was modified and operated on fuel oil firing, towards the end
of the program described herein.
It should be noted that none of the small scale work
described in Section 4.3 was performed as part of the utility
pilot plant program.
The following sections discuss the design, operation, and
results of the utility pilot plant program.
4.7
-------
100
o^-
O
Ul
"i
o
V
ce
x
O
V 500 ppm NH3
O 400ppm NH3
D 3'00ppm NH3
O 2 OOppm NH3
t> lOOppm NH3
< Oppm NH3
Catalyst Pt. Flue gas2%02- NO 230 ppm Sp?ce Velocity 50 000
3 0
400 450 500 550
Temperature in Degrees F.
Figure 4.1 Temperature Vs NOx Removal £ff.
600
650
-------
100
80
60
QJ
40
o
a:
x 20
0
O
Catalyst: ft
Flue Gas- 2.5%09
N0-22?ppm
Ammonia 500 ppm
Q Temperature 450 degree F
49€ " "
50
75 100 125 _
Sp?re Velocity times 10 v/v/hr
Figure 4.2 Removal Efficiency Vs Space Velocity
175
200
-------
Catalyst Pt.
Furnace Flue gas 240 ppm NOx- 2% 0.
100
Space Velocity 50,000 v/v/hr
Temperature 500 degrees F.
80
•^60
o
c
O)
ui 40
o
-------
10
80
0)
'o
I 40
E
O)
o:
X
O
20.
0
0
Catalyst:Pt. Space Velocity: 50,000 v/v/hr
Furnace Flue Gas :225 ppm NO , 2% 0 , 300 ppm S00 NHL 450 ppm
X C. 2 T
QT=550CF
100
AT=600*F
I
D T=575°F
200
300
400
Effect of Exposure to 300 ppm SO^ on Platinum Catalyst in Hours
Figure 4.4
700
650
u_
o
600 ~
I
550 2
&
500 I
500
-------
Table 4.1
Catalysts Included in Small Scale Testing
Designation
Pt02H - 1F2
- 1FA
- 1F5
- 1F6
Pt02H - 1H1
- 1H3
- 1H4
Pt02H - 3A1
Pt02H - 1G1
Pt03H - Al
- A2
Pt06H - Bl
Pt07H - Al
PtOAR - Al
- Bl
- Cl
- Dl
Composition
Pt, 0.3%
Pt, 0.3%
Pt + Ru,
0.37»
Pt, 0.3%
Pt. 0.3%
Pt, 0.3%
Pt, 0.3%
Pt
Support
Ceramic Honeycomb/
Al£03 wash coat,
12 cpi* (American
Lava)
Ceramic Honeycomb/
A12C-3 wash coat, 8
cpi (American
Lava)
Ceramic Honeycomb ,
1/8" cell (Du-
Pont)
Ceramic Honeycomb ,
1/8" cell (Hexcel)
Ceramic honeycomb,
12 cpi (American
Lava)
Ceramic Honeycomb ,
1/8" cell (Du-
Pont)
Ceramic Honeycomb,
silica washcoat
8 cpi (American
Lava)
Metal support
Dl-metal contains
alumina
Manufacturer
Matthey Bishop
Mat they Bishop
Matthey Bishop
Matthey Bishop
Englehard
Du Pont
Matthey Bishop
Matthey Bishop
4.12
-------
Table 4.1 (continued)
Designation
Ptl6G - A
Pt20G - A
CulOH
CullH
Col5T
Col6T
Cu09T
Cu09T - 1
Cu08T
Cu09H - Al
LaOl - P
Composition
Pt, 0.3%
Pt, 0.3%
CuO/Cr90Q,
17% J
CuO
CoO
CuO 6.2%
Co20q 5%
Aluminum
hydroxide
33%
Barium pro-
moted copper
chromite
(Ba/CuO/
Cr203)
CuO, CoO
Lanthanum
oxide ,
Alumina
Support
Zeolite granules
Active carbon
granules
Ceramic Honeycomb,
1/8" cell (Du-
Pont)
Ceramic Honeycomb,
1/8" cell (Du-
Pont)
Tablets, 1/8"
diameter, 1/8"
long
Tablets, 1/8"
diameter
1/8" long
Tablets, 3/16"
diameter
1/8" long
Ceramic Honeycomb,
1/8" cell (Du-
Pont)
Tablets, 1/8"
diameter, 1/8"
long
Manufacturer
Matthey Bishop
Matthey Bishop
DuPont
DuPont
Girdler
Harshaw
Girdler
DuPont
Environics
4.13
-------
Table 4.1 (continued)
Designation
M19H - A
M17H - A
V14E
Composition
MnO and
others
Monel, 100%
V205 16%
Mo03 3%
Support
Ceramic Honeycomb,
1/8" cell (Du-
Pont)
Pure Monel in
honeycomb struc-
ture
Alumina extru-
sions, 3/16"
diameter, 3/16"
long
Manufacturer
DuPont
Hexcel
Harsh aw
cpi = corrugations per inch
4.14
-------
SUMMARY OF CATALYST PERFORMANCE: NO
300 ppm) IN
CATALYST
SPACE VELOCITY
v/v/hr
TEMPERATURE
RANGE (°F)
NOX
REMOVAL
EFFICIENCY
Cu07T (CuO/Cr203)
CuOST (Ba/CuO/Cr2Oc)
PtOlT (Pt)
Cu09T (CuO/CoO)
V14T (V205/Mo03)
Pt02H (Pt)
10,500(1)
10,500(1)
17,500(2)
10,500(1)
10,500(1)
40,000(3)
600-700
600-700
600-700
600
600-650
500
95-98%
95-987o
15-52%
99%
75-93%
90%
(1) Catalyst bed dimensions = 1" dia. x 2.0" deep (pellets)
(2) Catalyst bed dimensions = 1" dia. x 1.2" deep (pellets)
(3) Catalyst bed dimensions = 3/4" dia. x 2.0" deep (honeycomb)
Table 4.2
-------
SECTION 5
DESIGN OF UTILITY PILOT PLANT
Design of the utility pilot plant was conducted in sev-
eral areas. These were (1) system design, (2) control system
design, (3) ammonia injection system design, and (4) analyti-
cal techniques.
5.1 SYSTEM DESIGN
System design included overall design of the ductwork,
selection of the catalyst support from the standpoint of
activity and pressure drop and determination of the proper
location for the catalyst in the air preheater to insure rela-
tively uniform temperature distribution.
5.1.1 Overall System Design and Analysis
A schematic conceptual design of the pilot plant, showing
design values of temperature, flowrate, and static pressure is
presented in Figure 5.1. The flue gas sample point is located
at an access hatch, upstream of the Unit 4 Air Preheater, in a
section of the main duct through which there is flow at all
times. This layout was approved by Valley Steam Plant per-
sonnel of the LADWP.
5.1.2 Catalyst Selection
Environics1 previous small scale tests had indicated that
a 0.37o platinum catalyst, supported on a high surface area
wash-coat, would provide high activity for NOx removal in
steam boiler flue gas, with a relatively low pressure drop
through the catalyst bed. Thus, selection of the pilot plant
catalyst was limited to such a catalyst, with the objective
5.1
-------
of selecting the proper size support, to provide acceptable
levels of activity and pressure drop for a utility plant.
Testing was performed to determine pressure drop and activity
characteristics of several catalyst/support configurations.
5.1.2.1 Pressure Drop Tests. Pressure drop experiments were
performed on the various supports considered suitable for use
in the pilot plant catalyst bed. The supports used were
American Lava, 5 corrugations per inch, 10 mil wall thickness,
split cell, and Hexcel A3-30, alumina, 30 holes per square
inch, L3-50 lithium aluinum silicate, LAS, 50 holes per square
inch, and L3-100, 100 holes per square inch. The experiments
were performed on uncatalyzed support beds 6 inches square,
at several depths (3", 6", 9" and 12" for 5/10-SC and L3-50)
and at several flowrates (CO 180, 240, 350 SCFM for L3-100
and L3-50, and
-------
from the standpoint of pressure drop, would be L3-50. A 9-
inch bed would impose a hot pressure drop of 2.1 inches HoO,
less than the 2.5 inches allowed in the overall design. If
necessary, a 12 inch bed could be used, which would require
only a slightly reduced hot gas flow rate at a pressure drop
of 2.5 inches of water column.
5.1.2.2 Catalyst Activity Tests. To determine the suitability
of Hexcel L3-50 as a support, samples of platinized L3-50
were obtained from Matthey Bishop for evaluation. These
samples, assembled to form a catalyst bed 6" square x 12"
deep, were placed in the 12000 SCFH test boiler reactor. The
catalyst was tested with the following results.
A short test of one hour duration was conducted, at which
time 957o NOx removal efficiency was observed. The system was
shut down for the weekend, after which testing was resumed.
During startup, the boiler was accidentally operated in a
fuel-rich condition, resulting in incomplete combustion and
deposition of carbon on the catalyst surface. Testing was
continued for several days, on the theory that the carbon
would burn or blow off spontaneously. However, at no time
during those several days was removal efficiency in. excess
of 80% observed. The catalyst was therefore removed from the
reactor and placed in a controlled atmosphere kiln, in which
temperature and oxygen concentration were gradually increased
to 750°F and 10% 02, in order to "burn off" the carbon. The
catalyst was then replaced in the reactor and tested continuously
for four days. Immediately after this regeneration procedure,
94% NOx removal efficiency was observed. By the end of the
four day test, this had declined to 92%.
In addition to these observations of removal efficiency,
several samples of the catalyst were removed, and their BET
surface area was measured. It was determined that the new
5.3
-------
catalyst (prior to any testing) had a surface area of approxi-
2
mately 20 m /gm. The catalyst after carbon deposition
(immediately prior to regeneration) had a surface area of
2
approximately llm /gm. The catalyst following regeneration
9
had a surface measurement of 19 m /gra.
Ignoring the decrease in the removal efficiency which
resulted from carbon deposition, this catalyzed L3-50 still
appeared somewhat unstable, i.e., NOx removal efficiency de-
creased from 94 to 92% in only a few days. Furthermore, its
BET surface area, when new, was considerably lower than that
of catalyzed 8/8 SC (American Lava) (20 vs 37 m2/gm). The
conclusion was thus reached that, at that time, it could not
be guaranteed that L3-50 could be catalyzed in a completely
satisfactory manner. Therefore, it was decided to use cata-
lyzed American Lava 8/8 SC as the pilot plant catalyst. (This
catalyst is identical to that used in the long duration-small
scale testing, and to the platinum catalyst used in the pre-
vious small scale work.)
5.1.3 Catalyst Heat Transfer Analysis
A study of heat transfer in the air preheater wheel was
performed, in order to estimate temperature distribution of
air, gas and matrix in the wheel, and to determine whether
catalyst temperature could be controlled as the power plant
load changes. Details of the calculations are given below:
Basis of Calculations
1. Flue gas: Inlet temperature 700°F; flow rate:
50,000 SCFH.
2. Air: Inlet temperature; 70 °F; flow rate: variable.
3. Wheel size: 7HKX-22: Air Preheater Company (standard
speed: 3 RPM).
4. Co-current flow.
5.4
-------
A schematic of the wheel, showing depth of hot end mater-
ial (steel matrix) and cold end material (catalyst matrix) is
presented below.
Catalyst
10" Steel
Flue Gas
Matrix
Matrix
t
Air
Calculations
The two most important variables which control the temper-
ature distribution in the regenerative wheel are the speed of
the wheel and the air flow rate. Therefore, calculations were
carried out for three different cases:
Case I : RPM = 3; air flow = 50,000 SCFH/ft2;
Case II : RPM = .41; air flow = 50,000 SCFH/ft2;
Case III: RPM - 3; flue gas outlet temperature » 475 °F.
(Back calculate the air flow needed.)
The estimated matrix temperature, and average air and gas
temperatures resulting from these calculations, are shown in
Figure 5.2. In this graph, matrix temperature, as a function
of axial position, is shown by lines for the three cases above.
The indicated matrix temperatures are actually the temperature
extremes experienced by the matrix, at the end of its passage
through the flue gas (when it is hottest) , and at the end of
its passage through the air side (when it is coldest) .
Conclusions
1. By changing the wheel speed from 3 RPM to .41 RPM, the
matrix temperature excursions can be varied from less than
5.5
-------
+ 10°F to about + 250°F. Therefore, this wheel can be used
to simulate a wide range of real wheels, which are in use in
utility steam power plants.
2. By changing air flow rate, catalyst matrix temperature
can be controlled to any desired value, between approximately
700°F (air flow - 0) and 400°F (air flow = gas flow = 50,000
SCFH/ft2).
3. In particular, catalyst temperature can be controlled
to the desired value for peak NOx removal efficiency, i.e.,
475°F, by using an air flow rate of 36,000 SCFH/ft . At the
standard 7 HKX-22 wheel speed of 3 RPM, temperature excursion
of the catalyst matrix under these conditions will be ^20°F.
5.2 CONTROL SYSTEM DESIGN
The pilot plant control system was designed to accomplish
four functions: (1) inlet NHg concentration control, (2) flue
gas temperature control, (3) flue gas flow measurement and
control, and (4) air preheater inlet pressure control. The
various functions are described below.
5.2,1 Inlet Ammonia Concentration
Three methods to control NH^ concentration were considered,
The three methods are:
Method 1: Control NH« flow as a function of inlet
flow rate and NOx concentration.
Method 2: Control NH~ flow to a constant concentration
in the inlet stream.
Method 3: Control NH3 flow to minimize the total NOx
+ NH- concentration in the exit stream.
Process control equipment manufacturers were contacted to
determine the nature of the equipment required for the control
5.6
-------
system. Method 3 would be the most desirable. However, the
hardware for this system is relatively sophisticated and was
not readily available.
It was thus decided to simulate the control method 3 to
control the ammonia injection rate manually, using a hand
valve and calibrated rotameter.
5.2.2 Flue Gas Control System
Flue gas temperature is controlled with a temperature
controller measuring gas temperature on the air preheater
flue gas exit stream. The controller adjusts the position
of the air flow control damper operator to provide the air
flow rate required to maintain the set temperature.
In addition, high temperature protection is provided by
a control loop separate from the above temperature controller.
A separate high limit controller monitors the temperature of
the exit gas. If this temperature rises above a set limit
(e.g., 650°F), the controller will turn off the flue gas fan,
close the flue gas inlet damper, and close a solenoid in the
NHo supply line. This controller requires manual reset before
operation of the system can be resumed.
5.2.3 Flue Gas Flow Control System
The flue gas inlet damper is controlled by a modutrol
motor and an open/close switch. The damper is held full open
or full closed in response to this switch position, and a
pressure gauge on the control panel indicates whether the
damper is open or closed.
Flue gas flow rate is measured with a venturi flowmeter.
The pressure drop across the venturi is measured and read on
a Dwyer Photohelic differential pressure gauge on the control
panel.
5.7
-------
Flue gas flow rate is controlled by setting the flue gas
ID fan damper position using a modutrol motor. This control
can be performed either manually, with a panel switch, or
automatically with switches in the differential pressure
gauge. A panel meter indicates the position of the damper
(0 to 100% open).
5.2.4 Air Preheater Inlet Pressure Drop Control
In order to prevent leakage from the preheater air side
to the preheater gas side, or vice versa, it is necessary to
equalize the pressure at the air inlet with that at the gas
inlet. This is accomplished by measuring the pressure drop
between the air and gas inlets with a differential pressure
switch. The output of this switch is used to open or close
the inlet air pressure damper, through a modutrol motor, to
achieve the desired condition.
5.2.5 Installation and Safety Consideration
A safety system will shutdown the pilot plant if the set
point of the high limit temperature controller is exceeded or
if the electrical signal that the plant is firing fuel oil is
received. In either case, the shutoff damper will close, the
ammonia flow solenoid valve will close, the flue gas ID fan
will stop, and a signal will be sent to an annunciator in Unit
4 control room. Valley Steam Plant personnel were instructed
to call Environics personnel if such a shutdown occurred dur-
ing a period of unattended operation.
Pressure taps are V swagelok fillings brazed to the out-
side of the duct and flush with the duct wall on the inside.
Sample taps are %" stainless tubes extending to the center of
the duct. Thermocouples are 1/8" diameter Type J, extending
to the center of the duct. Twelve Type J thermocouples are
5.8
-------
provided, 9 cemented to the catalyst surface and 3 extending
into the gas stream.
5.3 AMMONIA INJECTION SYSTEM DESIGN
5.3.1 Laboratory Tests
Tests on the NHo injection system were conducted with
the small scale test boiler.
The test apparatus size was 6" square; a flow of 50,000
2
SCFH/ft was maintained to simulate the flow regime of the
pilot plant. Samples were taken from the center of each
square in the 9 point sample grid, and analyzed for NH^ by
the oxidation method described below.
NHo was injected through a conical spray pattern orifice,
manufactured by Spraying Systems Co., catalog #1/8 BX0.5,
316SS. NH« concentration was determined at each point of the
nine point sampling array, 14 duct diameters downstream of
the nozzle.
Results of three average complete sets of data are shown
in Figure 5.3. Mixing was fairly uniform, except for a low
average concentration in the bottom row.
On the theory that the injection nozzle was tilted
slightly upward, a fourth scan was made after pointing the
nozzle 20 degrees downward of its initial position. These
results are shown in Figure 5.4. For this test, the concen-
tration in the bottom row was higher than that in the top two.
The following conclusions were reached from these tests:
1. Injection through a simple nozzle will provide satis-
factory mixing for use in the pilot plant.
2. Nozzle alignment is somewhat critical, in order to
obtain uniform mixing.
5.9
-------
3. Use of mixing devices or inert gas dilution will not
be necessary to provide adequate mixing.
5.3.2 Pilot Plant Injection System
On the basis of the small scale tests, the NHo injection
system for the utility pilot plant was designed with sixteen
injection nozzles, one for each 6" square in the 24" square
inlet duct. This nozzle array was located as far upstream of
the catalyst as possible, to take maximum advantage of the
available mixing length.
5.4 ANALYTICAL TECHNIQUES
The following analytic techniques were selected for use
in the pilot plant program.
5.4.1 NOx Analysis
Oxides of nitrogen were analyzed with a Beckman Model 915
chemiluminescence analyzer.
5.4.2 Ammonia Analysis
Ammonia analysis was performed by two techniques. The
bulk of the analyses was performed using catalytic oxidation
of ammonia to NO, followed by analysis of the product NO in
the chemiluminescence analyzer. Periodic checks were made by
wet chemistry. These techniques are described below.
Catalytic Oxidation of NHo
The reactor used to analyze for NHo by oxidation to NO
is described below.
o
The oxidizer uses 50 cm of catalyst (270 Pt on gamma
A190^ pellets, 1/8" diameter) which, with a gas flow of 10
5.10
-------
SCFH, gives a space velocity of 6,000 hr~ . The oxidizer
converts NH~ to NO by the reaction:
4NH3 + 50x > 4NO + 6H20.
In order to determine the temperature required for 100%
conversion of NH^ to NO, the following tests were performed.
Premixed NHo/N2 gas (400 ppm) was passed through a calibrated
rotameter and then mixed with dry air to achieve a 2% Q*
concentration, before entering the reactor. The ammonia con-
centration (NH.j) was determined by the standard wet chemical
method. The product (NO) of this reaction was analyzed by a
Beckman NDIR analyzer.
By varying the temperature of the catalyst bed, a curve
(Figure 5.5) showing conversion of NH^ to NO, as a function of
temperature, was obtained. The results indicated that 1007o
conversion was obtained at a temperature of 1200°F. The same
result was also obtained by Klimisch and Barnes (Environmental
Science and Technology, June 6, 1972).
It was concluded that by operating the oxidizer at the
proper temperature (1200°F), 100% conversion of NH3 to NO could
be obtained, allowing rapid response continuous analysis for
NHo, using a conventional NO analyzer.
This reactor was used in the NH-j mixing studies, allowing
the opportunity to check conversion efficiency over a longer
period of time than was done in the temperature studies, with
satisfactory results.
5.4.3 Oxygen Analysis
Oxygen is analyzed in the pilot plant using a Beckman
Model 715 Process Q Analyzer.
-------
Flue gas inlet
Temp.-700 degree F
Flow rate-200,000 SCFH
Pressure-5" water column
Ammonia
(P=-9.5" water Column
T=675 Degrees F. T= 475 Degrees F f(T=475/NDjwrees ~
P=-6" water column P=-
1
1
[}/\ >\ ! /— x
Shutoff Venturi
Damper Flowmeter
Air inlet ft y j
Flow rate- 1 85, OOOSCtfw r >
Pressure control
Damper
T=80
t
deg
re
es*F
>
HT-^Vt. J
/
!
/
\
. (
k 1
jTem
P=-6" Water column
Air Pre-heater
Co- currentf I flow
\
Control Damper
I.D. Fan
Control Damper
.D. Fan
; 475 Degrees F .
(P=-9.5" Water Column
Figure 5.1 Utility Pilot Plant Schematic
Ul
-------
7001
600
500
400|
u.
ra
2l
300
20
10
\\
\
\
\
\
\
\ sx
\
•• /
V
/
/ /
r //
4 8 12 16 20 24
Air Preheater Wheel Depth in Inches
Steel Matrix +4 Catalyst Matrix
Air Temp. Gas Temp Matrix Temp.
Case I A ?? Case I _,,
Case I I • J> Case II
Case ((!• n Case III
Figure 5.2 Calculated Temp. Profile
Air Preheater 7HKX-22
5.13
-------
6" Dia. Pine
1
4
7
2
5
8
3
6
9
i
Average ( NH.-+ NO) in ppm
1+2+3 428
4+5+6 422
7+8+9 396
1+4+7 415
2+5+8 392
3+6+9 420
Figure 5.3 Ammonia Mixing Test Results
Injection Nozzle Zero Degree to Horizontal
Boiler N0x=45 ppm
-------
^ 6" Diameter pipe
440±IO
435±5
G
5IO±5
435110
420±IO
O
500±5
~-\
410+10
420±IO
5 2 5. +3
Avjrage (Ammonia + NO) ppm 428
Av
;rage ( Ammonia + NO ) ppm 425
Average ( Ammonia +NO) ppm '512
Figure 5.4 Ammonia Mixing Test Results
Ammonia Injection Nozzle negative Angle 20
Boiler Ambient N0x=45ppm
-------
K>€
INLET CONDITIONS:
Ammonia: 620 ppm
Oxygen 1% in Nitrogen
Flow rate: 5 SCFH
50.
o
d-l
en
o
c
_0
"en
o5
o
O
0
-O-
500
600
700
800
900
j;ooo
,100
1200
Reactor Temperature in degrees F
Figure 5.<5
Conversion Efficiency vs. Temperature for NH-j Oxidizer
-------
SECTION 6
LABORATORY PILOT PLANT TESTING
During the design stage of the utility pilot plant pro-
gram, laboratory pilot plant testing was performed to validate
results of previous work and to provide previously unavailable
information on catalyst life and possible formation of trace
exhaust products.
These tests were performed using a 550,000 BTU/hour gas
fired boiler, equipped with an ID fan to direct a portion of
the flue gas through the catalyst bed. Flow through the
reactor was measured using an orifice plate. Analysis for
NOx was done with either a Beckman Model 315 BL Non Disper-
sive Infrared Analyzer (sensitive to NO only) or a. Beckman
Model 971 Chemiluminescence Analyzer. Oxygen concentration
was determined with a Beckman Model 715 Process 02 Monitor.
Ammonia concentration was determined using the catalytic and
wet chemical techniques previously described.
6.1 LONG DURATION TESTING
Long duration testing was conducted to provide an estimate
of catalyst performance under exposure to actual gas fired
flue gas over a long period of time. These tests were also
expected to provide warning of any catalyst deterioration that
might occur in the utility pilot plant, so that steps could
be taken to investigate appropriate regeneration techniques.
Long duration testing was conducted for a total time of
175 days. Conditions during the test were as follows:
Temperature: 490 to 500°F
NH3 concentration (inlet): 200 to 250 ppm
NOx concentration (inlet): 240 ppm
6.1
-------
Flue gas Q£ concentration: 57o
Space Velocity: 50,000 v/v/hr
Catalyst: 0.3% Pt on 8/8 SC support
(MBI), 4"x4"x3" thick.
Results for the entire test period using weekly averages
of the data are shown in Figure 6.1. Figure 6.1 shows NOx
removal efficiency, total removal efficiency and temperature.
It can be seen that NOx removal efficiency remained approxi-
mately 95% and total removal efficiency remained between
80-85% for the entire test.
During the test period on days 151-154, it was discovered
that the (NOx + NHo) exit concentration was unusually high
(100 ppm). It is believed that this was due to an upset in
boiler operation (accidental shutdown of the induced draft
fan) which resulted in poor combustion and exposure of the
catalyst to carbon particles. After the condition had per-
sisted for several days, and could not be corrected by chang-
ing NH^ concentration, it was decided to remove and attempt
to regenerate the catalyst. (It was noted during disassembly
of the converter that the orifice plate, immediately downstream
of the catalyst, was indeed covered with fine soot.)
The regeneration procedure was similar to that used pre-
viously, with the L3-50 catalyst. The catalyst was placed in
a kiln, through which air was passed at a low flow rate
(5 SCFH), and held at 400°F for 30 minutes. The temperature
was gradually increased to 700°F, over a period of 90 minutes,
and maintained there for 16 hours.
After this treatment, the catalyst was replaced in the
converter, and testing resumed. Activity was restored to pre-
vious levels; the observed (NOx + NH~) exit concentration after
regeneration was again approximately 40 ppm.
6.2
-------
6.2 PARAMETRIC TESTING
Parametric testing was performed, using the same gas
fired boiler and catalyst as in the long duration testing,
to determine the effects of space velocity, inlet NHo concen-
tration, and temperature upon removal efficiency. This test-
ing was desired for two reasons: (1) to extend previous
small scale data and (2) to provide scale-up information upon
which design and evaluation of the utility pilot plant could
be based.
6.2.1 Effect of Inlet NH3 Concentration
Two parametric studies of the effect of inlet NH-, concen-
tration on removal efficiency and exit NH~ concentration were
performed. Figure 6.2 shows the results at 490°F. In general,
if inlet ammonia injection exceeded 250 ppm, the NOx removal
efficiency did not increase and the exit ammonia concentration
increased.
6.2.2 Effect of Space Velocity
Experiments were performed to determine the effect of space
velocity on NOx removal efficiency. Results, shown in Figure
6.3, are similar to what has been observed previously in smaller
scale testing. Removal efficiency decreases with increasing
space velocity; however, between 50,000 and 75,000 v/v/hr this
effect is not very pronounced.
6.2.3 Effect of Temperature
Figure 6.4 shows results of experiments performed to deter-
mine the effect of temperature on NOx removal efficiency.
Data is presented at 470 and 500°F, showing NOx removal effi-
ciency and NHo exit concentration as a function of NH~/NOx
ratio in the inlet gas. Removal efficiency is relatively
6.3
-------
unchanged by temperature in this region; however, NHo exit
concentration decreases quite rapidly with increasing tem-
perature. Unfortunately, it was not possible to run any
experiments at higher temperatures. Numerous attempts to
both increase fuel flow and to plug boiler tubes produced
little or no temperature increase. However, previous data
indicate that at 525°F, removal efficiency would be only
slightly lower than at 500°F and NH^ exit concentration would
be significantly lower.
6.3 TRACE PRODUCTS ANALYSIS
It was desired to determine whether any undesirable trace
products were formed as byproducts of the NOx/NHo reaction
over the platinum catalyst. To accomplish this, analysis of
three samples of the exit gas from the long duration test
catalyst was performed by TRW, Inc. Results are shown in Table
6.1.
Several comments may be made about this data:
(1) Environics1 in-house analysis of the exit gas, con-
ducted immediately before these three samples were taken, found
14 ppm NO, zero ppm NO 2 and 13 ppm NH~. The difference between
our analysis and TRWs for NO and N02 cannot be explained
readily. The difference in the NH^ analysis is probably due
to adsorption of the NH~ in the water condensed in the sample
flasks. This water was not analyzed.
(2) The fact that no N20 or HCN was found indicates that
these compounds are not produced by this reaction. Although
this data is not conclusive, it was felt to be sufficient, and
the trace products investigation was considered complete.
6-4
-------
10
on
O\J
c
14—
LU
O
O)
C£
X
O
Z40
2C
P—
S*3**
A A A
O-D O /^—fy^V-^H^M^—n-t^ m J-L j-
v-r^-t>-.___ -\J-~*
-------
100
75
0>
'o
£ 50
LU
CO
O
0)
X
O
25
Temperature 490 F
Space Velocity 50,000 v/v/hr
NOx Inlet Cone. 250 ppm
Catalyst Pt.
O NOx Removal tfficiency
D (NOx+NH3) Exit Concentration
A MM- Exit Concentration
200
150
Q.
Q.
ioo|
4-1
0)
O
O
O
50
100
200
300
400
500
600
700
0
800
Inlet Concentration in ppm
Figure 6.2 Effect of NH-on NOx Removal Efficiency
cr-
-------
00
80
g 60
LU
I 40
X
o
20
Ammonia Inlet Cone.
o
A
O
362 ppm
3fO ppm
260 ppm
NOx Inlet Cone.
Temperature
Catalyst
21 0 ppm
500* F
Pt.
I
25
50
100
125
-3
150
175
Space Velocity in v/v/hr times I 0
Figure 6.3 Effect of Space Velocity on Removal Eff.
-------
100
00
^o
o^
C
•
tt
Ul
o
801
60(
40l
20
NOx Removal Eff. at 472^
QNOx Removal Eff. at 50(fF
QNH3 Exit Cone, in ppm at 472*>F
Exit Cone, in ppm at 500° F
Space Velocity 50,000 v/v/hr
Catalyst Pt.
80
60
o
c
o
O
o
E
X
LJ
20
0
1.0 2.0
Figure^.4 Ratio NH3/NOx Inlet Cone.
Effect of Temperature arid NHVNO Ratio on Removal Eff.
3.0
cx>
-------
Table 6.1
Trace Products Analysis by TRW
Sample #
1
2
3
NO
ppm
32
32
36
N02
ppm
15
0
6
NH3
ppm
0
0
0
N20
ppm
0
0
0
HCN
ppm
0
0
0
co2
01
/o
6
6
6
°2
%
5
5
5
vO
-------
SECTION 7
PILOT PLANT INSTALLATION AND CHECKOUT
7.1 INSTALLATION
Installation of the utility pilot plant was performed
during the months of March and April, 1973. Los Angeles
Department of Water and Power personnel were employed to
lift the air preheater to the work level and to position it in
its proper location. Environics1 personnel performed the re-
mainder of the installation work; i.e., installation of the air
and gas ID fans, ductwork, flow control dampers, ammonia in-
jection system, control panel, and control system components.
A photograph of the completed pilot plant is shown in Figures
7.1 and 7.2.
7.2 INITIAL CHECKOUT
Initial system checkout work was performed before the
catalyst baskets were installed in the air preheater wheel.
Details of these tests are described below.
Sampling System. The sample handling system was found
to function properly, except that it was felt desirable to
reverse the positions of the Hankison dryer and sample pumps,
so that the pumps are upstream of the dryer. After installa-
tion of normally closed solenoid valves on the drain ports of
the dryer, the system automatically blows out condensed water,
when the sampling system is in the ^ position.
It was also found necessary to modify the sampling sys-
tem by adding heating tape to the inlet and exit samples
manifolds, in order to eliminate water condensation and thus
provide stable NH^ analyses through the oxidizers.
7.1
-------
Pressure Readout System. Pressure readout is provided
at the system inlet, gas flowmeter and preheater air-gas inlet.
The Dwyer magnehelic gauges were found to function satisfac-
torily; however, it was felt desirable to replace the 0-1"
W.C. gauges provided for the flowmeter and air-gas inlet with
0-2" W.C. and 5-0-5" W.C., respectively, to provide a wider
range for readout during startup and off-normal conditions.
Temperature Readout System. The 24 point temperature
recorder was found to function satisfactorily.
Manual Damper Control Systems. Results of tests of the
various manual control loops are presented below.
Shutoff Damper: A stiffer linkage rod was required to
obtain satisfactory operation.
Air Pressure Damper: A stiffer linkage rod was required
to obtain satisfactory operation.
Gas Control Damper: Satisfactory operation was obtained
with no significant changes.
Temperature Control System. Initially, severe overshoot
and oscillation was observed in the temperature control sys-
tem. Adjustment of the temperature controller proportional
band and auto reset potentiometers improved, but did not com-
pletely eliminate the problem. Analysis of the situation
indicated that there were two other factors contributing to
this problem. The first was the high degree of leakage within
the air preheater. This was due to the absence of the cold
end basket covers; when these covers were installed, the leak-
age was significantly decreased and temperature control was
improved. The second trouble source was the lack of automatic
control of the air inlet pressure damper. An automatic control
system for this damper was installed and tested.
Air Pressure Damper Automatic Control System. To achieve
automatic control of the air pressure damper, a Dwyer 1640
7.2
-------
differential pressure switch was installed and the system was
rewired to allow the damper to be controlled either by this
switch or manually. This automatic control was found to be
unstable, due to the delay between a change in damper posi-
tion and the resulting response in the pressure switch
diaphragm. It was decided to decrease the speed of the
damper motor, by installing a cam timer in series with the
pressure switch, such that the damper motor is actuated for
only one second in each 10 second time cycle. This had the
effect of slowing damper travel to the point where the pres-
sure switch could respond fast enough to provide stable control
as boiler load or temperature setpoint changed.
7.3 FINAL CHECKOUT
To accomplish the final checkout, the catalyst baskets
were loaded into the air preheater. Asbestos gasket material
was used to prevent gas bypassing the catalyst. Initial test-
ing with catalyst in place indicated that the design flow rate
of 180,000 SCFH could not be reached. Maximum possible flow
was 115,000 SCFH; a flow of 157,000 SCFH is required for
50,000 v/v/hr space velocity. Flowrate was subsequently in-
creased to 140,000 SCFH (45000 hr~ space velocity) by increas-
ing the speed of the flue gas ID fan.
7.3
-------
Figure 7. I Utility Pilot Plant
1-Inlet Duct From Boiler Economizer
2-Catalytic Converter & Air pre-heater
3-Exhaust Ducts
7.4
-------
Figure 7.2 Utility Pilot Plant
1-Exhaust I.D. Fans 2-Exhaust Duct 3~lnstrumentation Panel
7.5
-------
SECTION 8
GAS FIRED PILOT PLANT TESTING
UTILITY PILOT PLANT
The bulk of the utility pilot plant testing was performed
during periods when the steam plant was burning gas fuel.
Following this work, tests on oil fuel were conducted. The
results of the gas-fired tests are described below; results
of the oil fired tests are described in Section 9.
Gas fired testing was conducted in two areas: parametric
tests, with the objective of determining optimum operating
conditions for the utility pilot plant, and long duration
tests, with the objective of determining the long term per-
formance of the catalyst under extended exposure to gas fired
flue gas. In addition, tests were performed to determine whe-
ther the NHo was sufficiently well mixed with the bulk of the
flue gas prior to entering the air preheater, which contained
the catalyst bed.
8.1 UTILITY PILOT PLANT PARAMETRIC TESTS
Tests were performed to determine the effects of tempera-
ture and inlet NH~ concentration on NOx removal efficiency and
total removal efficiency. NOx removal efficiency and (NOx +
NH~) exit concentration, as a function of temperature, at a
constant inlet NH-j concentration were measured.
Inlet NH~ concentrations of 100, 200, 300, 400 and 500
J _i
ppm were probed. Space velocity for these tests was 45,000 hr
This data indicate that 200 ppm ammonia is sufficient to
obtain 85% removal efficiency at the typical NOx inlet concen-
trations encountered (125 ppm) and that higher ammonia
3.1
-------
concentrations do not significantly improve this removal effi-
ciency, but only increase the amount of unreacted ammonia in
the exit flue gas. The data also indicates that the greatest
NOx reduction is obtained at exit flue gas temperatures of
490 - 500°F. Figure 8.1 gives the optimum results.
8.2 PILOT PLANT NH3 MIXING TESTS
Tests were also made to determine NHo distribution at the
inlet of the air preheater. Each of the six probes in the
inlet sample manifold was sampled individually to determine
NHo concentration. Results are shown in Figure 8.2. Average
NHo concentration at that time was 277 ppm. The difference
between the various sample points was less than + 6%.
8.3 PILOT PLANT LONG DURATION TESTS
During operation of the pilot plants, a total time of
2021 hours on gas firing was achieved. This testing was con-
ducted during the period from July 1973 through July 1974.
Although the pilot plant was operated whenever conditions of
operation of Valley Steam Plant Unit 4 permitted, there were
numerous shutdown periods. The pilot plant was shut down for
any of the following reasons: oil firing on Unit 4, unscheduled
repairs to Unit 4 (tube leaks, etc.), scheduled maintenance to
Unit 4, low load on Unit 4, and a strike by Valley Steam Plant
personnel.
Results of the gas fired duration tests are presented in
Figures 8.3 through 8.5. Each figure shows NOx removal effi-
ciency, gas temperature (at the catalyst exit) and NHo inlet
concentration, as a function of catalyst operating time in
hours. These figures include data from the parametric tests
as well as the long duration tests at optimum conditions. It
can be seen that, over the duration of this testing, no
8.2
-------
indication of catalyst deterioration was observed, i.e., at
optimum conditions, NOx removal efficiency of 75-9070 was
achieved throughout the test period.
It has been shown by tests performed by Environics prior
to the contract initiation and confirmed by the long duration
laboratory testing during the contract performance that NOx
removal efficiency is sensitive to the temperature of the flue
gas and the amount of ammonia injected. Figure 8.1 shows
that the maximum removal efficiency occurs at a temperature
between 490° and 500°F. at an ammonia injection of 200 ppm.
Increase of the ammonia injection will increase the NOx re-
moval efficiency; however, the effluent ammonia may also be
increased. Therefore, the utility pilot plant long duration
testing tried to probe the effects of temperature and ammonia
injection while other operating conditions were maintained
constant for a relatively long time.
Figures 8.3 through 8.5 show that:
(1) For approximately 700 hours the NOx removal efficiency
was essentially 85% at temperatures between 460 and 470°F.
Ammonia injection was maintained at 200 ppm. Space velocity
obtained was 45,000 v/v/hr.
(2) Attempts were made to decrease ammonia injection to
125 ppm at lower flue gas temperatures. Under these condi-
tions, the NOx removal efficiency decreased to approximately
607..
(3) Attempts were also made to increase ammonia injection
at higher flue gas temperatures. Under the conditions of 500
ppm ammonia injection and flue gas temperature at 530°F, the
NOx removal efficiency increased to slightly higher than 85%.
Although there was no measurement of effluent ammonia, it is
believed that the effluent ammonia did increase.
8.3
-------
(4) When the system operating parameters were returned
to the initial values, namely, 200 ppm ammonia injection and
flue gas temperature between 460-490°F, the NOx removal effi-
ciency did return to approximately 8570, previously obtained.
It appears that the results of the utility pilot plant
long duration testing did meet the objective of the program.
It has demonstrated that: (1) the ammonia reduction system
can remove NOx at 85% efficiency; (2) the catalyst did have
a life more than 2000 hrs without deterioration; and (3) it
can operate at a space velocity of 45,000 v/v/hr for more than
2000 hours.
8.4
-------
00
8G
60
8=40
£21
0
Boiler NOx Cone.
Ammonia Inlet Cone.
NOx Removal Eff. %
NOx+NB-Exit Conc.rdpm
Spaced elocity
I 25 ppm
200 ppm
o
A
45,000 v/v/hr
20
80
A
o
c
o
O
403
470
480 490 500 510 520
Figure 8.1 Temperature in degrees F.
Optimum Test Conditions-Utility Pilot Plant
530
0
540
00
Ul
-------
+ 10" ^
Top
1
2
3
4
5
6
4
Duct Dimension and Probe Arrangement
Probe Location NH_ Cone, in ppm
1 263
60"
2 273
i
3 286
4 290
5 291
6 258
Bottom
Figure 8.2
Inlet NH., Concentration Distribution
8.6
-------
100
80
60
Ixl
"240
o
o:
X
O
20
Test Conditions in this Period
Ammonia Injection-200 pom *
Flue gas Temperature-460?470F
Space Velocity: 45,000 v/v/hr
I
300 4UU t>UU
100 ZDO30D 41
Test Time in Hours
Figure 8.3 Utility Pilot Plant Long Duration Test Results
ouu
/uu
OO
-------
100
80
c
5 40
UJ
"to
0>
a:
X
O
0
Test Condition in this Period —
Amnonia Injection-App. I25ppm
Temperature-410-420°F
Space Velocity: 45,000 v/v/hr
- Test Conditions in this Period""—H
Ammonia Injection I App. 500 pom
Temperature -530°F
Space Velocity: 45,000 v/v/hr
800
900
,000
.300
1,100 1,200
Test Time in Hours
Figure 8.4 Utility Pilot Plant Long Duration Test Results
,400
,500
00
00
-------
100
80
6C
£ 40
o
-------
SECTION 9
OIL FIRED UTILITY PILOT PLANT TESTING
The original objective of the work described herein was
to design, construct and operate a utility pilot plant NOx
reduction system on gas fired flue gas only. As described
previously, this pilot plant was operated on gas fired flue
gas for 2000 hrs, showing 85-90% NOx removal efficiency.
Because of the success obtained with these gas fired
tests, and in view of the need for a utility NOx removal
process to be compatible with oil firing, the program was
redirected to allow testing on oil fired flue gas. Environics'
previous small-scale work indicated that a temperature of
approximately 550°F was required for optimum NOx removal effi-
ciency in the presence of S02» and that 80% NOx removal effi-
ciency could be achieved. Since the utility pilot plant, as
designed and built, could not provide temperatures above 500
to 525°F, an electric heater was installed in the pilot plant
ductwork, immediately upstream of the air preheater gas inlet.
This heater was a 75 kw Calrod duct heater, expected to in-
crease the gas temperature by 90°F. Results of this work are
described below.
After installation of the flue gas heater, the utility
pilot plant was restarted, on oil firing. Due to the addi-
tional pressure drop in the system, imposed by the heater,
the maximum flue gas flowrate that could be obtained was
110,000 SCFH*(corresponding to a space velocity of 35000 hr ).
The utility pilot plant was operated on oil firing for
approximately 412 hours, bringing the total catalyst age to
2499 hours. (During this time, the pilot plant was also
operated for one 20-hour interval on gas firing.)
9.1
-------
NOx removal efficiency ranged from a low of 19% to a
high of 657o. The 1970 NOx removal efficiency was observed
with a gas temperature of 498°F and inlet NH., concentration
of 150 ppm. The 65% NOx removal efficiency was observed
with a gas temperature of 554°F, and inlet NHo concentration
of 300 ppm.
There were several problems encountered during operation
of the utility pilot plant under oil firing which prevented
obtaining steady state parametric data on the effects of tem-
perature and NHo concentration on NOx removal efficiency.
These problems are described below.
First, there was a problem in obtaining and maintaining
high catalyst temperatures. Although the flue gas heater did
provide the expected temperature increase (100°F), gas tem-
perature at the inlet to the pilot plant system was lower than
expected and varied considerably with plant load. Catalyst
temperature of 598°F could be obtained at full load (168 MW);
however, at low load (40-50 MW), it was impossible to maintain
temperatures above 515°F. Furthermore, because total system
demand did not require high loading, it was impossible to main-
tain a temperature as high as 575°F for even 24 hours, to de-
termine steady state NOx removal efficiency.
Attempts were made to increase the catalyst temperature
by reducing the total gas flow rate through the pilot plant.
Although reducing the gas flow rate increased the temperature
across the gas heater, heat losses to atmosphere through the
system seemed to increase. The net effect was that the air
preheater exit temperature remained approximately the same.
Temperatures were also observed to drop slightly during per-
iods of rainy, cold or windy weather. It thus was concluded
that the present thickness of insulation on the pilot plant
is not quite sufficient for high temperature operation.
9.2
-------
There also appeared to be some air flow through the air
preheater, even when the air fan damper was completely closed.
This was probably due to air leakage through the closed air
inlet damper, induced by the gas ID fan drawing air across the
rotor seals.
These fluctuations made a systematic investigation of the
effect of temperature on NOx removal efficiency nearly im-
possible at the pilot plant.
Problems were also encountered in maintaining constant
gas flow (and, therefore, constant space velocity) through the
pilot plant. During high loading, the draft at the pilot
plant inlet increased, and therefore pilot plant gas flowrate,
and space velocity, decreased. At full load (168 MW), maxi-
mum obtainable flowrate was approximately 90000 SCFH. At low
load, draft at the pilot plant inlet decreased, allowing flow-
rates up to 165000 SCFH. As described above, however, at low
load gas temperature dropped significantly, preventing operation
at optimum temperatures.
A third problem observed during operation of the pilot
plant on oil firing was the accumulation of white powdery mater-
ial in the exit ductwork. This accumulation was primarily
observed after the pilot plant had been run overnight with NH-j
concentrations above 200 ppm. It is believed that during late
evening and early morning operation, when the boiler was run
at low load, the catalyst temperature dropped to the point
where significant amounts of unreacted NH^ passed through the
catalyst. This NHo then reacted with the S02 in the flue gas
to produce (NH,) SO/ or other ammonia/sulfur compounds. The
temperature record indicated that, during low load operation
overnight, catalyst temperature did drop, to below 500°F.
Because of this problem, it was decided that, except for week-
ends, 24 hour operation of the pilot plant was possible to
9.3
-------
provide catalyst aging data, but that the ammonia supply
should be turned off except when Environics' operating per-
sonnel were in attendance. The LADWP subsequently analyzed
a sample of the white material; their results (Table 9.1)
indicate that it is primarily ammonium sulfate.
The fourth problem that was encountered during operation
of the pilot plant on oil firing was fouling of the catalyst
with soot particles. After 350 hours of operation on oil fir-
ing, a decrease in total gas flowrate was observed. At full
load, with an inlet draft of 6" WC, the maximum flow obtain-
able was 50000 SCFH. After operating the pilot plant for
several days under various load conditions, it was decided
to remove and examine the catalyst baskets.
A basket was removed from the air preheater and the lead-
ing edge of the catalyst was found to be fouled, primarily
with black soot. There were also some flakes of iron oxide,
and some white particulate matter. At least one-half of the
honeycomb passages appeared to be blocked. The fouling did
not appear to extend into the honeycomb and, in fact, the
trailing edge seemed to appear as clean as when first in-
stalled. Because the pilot plant gas and air flow are co-
current through the air preheater, whereas in normal use they
would be counter current, it was decided to rotate this basket
180° and reinstall it, thus simulating counter current flow.
It was thought that this would possibly remove some of the
fouling. After running the pilot plant for 20 more hours in
this manner, this basket appeared to be slightly less fouled,
but was still quite plugged.
Two baskets were then removed and returned to the Envi-
ronics laboratory where the plugged leading edge was first
swept with a small brush, and then compressed air was blown
through the catalyst honeycomb from the trailing edge. This
9.4
-------
procedure appeared to satisfactorily remove all the fouling
material from the honeycomb. The remaining baskets were re-
moved, cleaned at the utility pilot plant in a similar manner,
and replaced. Immediately, after re-starting the pilot
plant, a flow rate of 90000 SCFH was obtained, at full load,
with an inlet draft of 6" WC. During the remainder of the
catalyst operating time (100 hours), further fouling of the
catalyst was not observed.
The results of the oil fired pilot plant testing described
above were unexpected in several respects. By comparison with
Environics' previous small scale data, much higher NOx removal
efficiency was expected (80% compared with the average 5070
observed in the pilot plant). The difficulty in obtaining
high and constant gas temperatures, due to low load demand,
was also unexpected. Results of the pilot plant work indi-
cated that the optimum temperature might be above 550°F, and
such temperatures could not be readily obtained. The degree
of soot fouling encountered was also unexpected, as were the
frequent maintenance shutdowns of the boiler.
It was therefore decided to redirect the program to labora-
tory pilot plant testing, on oil fired furnace flue gas. This
would permit more extensive determination of the effect of
temperature than possible in the pilot plant. Results of this
work are described in Section 10.
9.5
-------
Table 9.1
Department of Water & Power
of the City of Los Angeles
Power System
Report No. C-1658
Page 1 of _2
ma
TESTING LABORATORIES
SPECIAL TEST & INVESTIGATION REPORT
DEPOSIT
Environics' Pilot Plant
Valley Steam Plant, Unit No. 4
A sample of a deposit, removed on February 3, 1975 from
the exhaust duct of Environics' Pilot Plant at Valley Steam
Plant, Unit No. 4, was submitted by Mr. Charles Sunwoo for
chemical analysis.
The purpose of the pilot plant is to determine the effec-
tiveness of the reduction of oxides of nitrogen by the addition
of ammonia vapor to the stack gases in the presence of a
platinum catalyst.
The unit had been gas fired for several months without
accumulation of any noticeable deposit. After fuel oil firing
for only a few days a large amount of the submitted material
had collected.
Most of the deposit consisted of flat flakes, with
diameters ranging from 1/16 to 3/8 inches. Many were of a
rusty brown or white to yellow color on both sides, others
were rusty on one side and white or yellow on the other side.
It also contained magnetic particles.
The weight loss of the sample was determined at 105°C.
The loss at this temperature is primarily water of hydration.
The sample was then heated to 815°C and the weight loss deter-
mined. The loss at the higher temperature accounted for over
half the weight of the sample.
When heated to the higher temperature the sample first
Date Completed:3/7/75
Authorization:
Job Card No. 24320-T-12
Copies to: W. W. Pepper
E. L. Morrison
(3)
Test Made by: AEC~
Report by: AEC
Checked by:
(1) Approved: s/W. E. Greninger
9.6
-------
Table 9.1 (continued)
emitted dense white fumes. They were strongly basic and it
was confirmed that ammonia was being liberated. When heating
was continued the reaction of the fumes changed from basis
to acidic, indicating that sulfur trioxide was being liberated.
These two gases are decomposition products of ammonium sulfate.
The ashed residue of the sample was analyzed and was
found to consist primarily of iron oxide. Small amounts of
constituents associated with fuel oil were also present.
However, the large amount of iron, compared to the other ele-
ments present, indicated that most of the deposit is the result
of corrosion.
By convention the metallic constituents are reported as
their respective oxides, but are also present as other com-
pounds . More than half of the sample was soluble in water and
about 99% was soluble in dilute hydrochloric acid. The sulfates
were determined on the acid soluble portion of the deposit.
Data
Determinations on sample as received
Percent by weight
Weight loss at 105 C 3.89
Weight loss at 815 C 53.16
Iron as Fe203 38.94
Silica Si02 0.95
Zinc as ZnO 0.46
Vanadium as V205 0.30
Calcium as CaO 0.25
Magnesium as MgO 0.23
Nickel as NiO 0.22
Sodium as NaoO 0.19
Aluminum as A1203 0.14
Copper as CuO 0.02
Water solubles 54.33
Acid solubles 98.86
Sulfates as 803 38.00
9.7
-------
SECTION 10
OIL FIRED LABORATORY PILOT PLANT TESTING
The laboratory pilot plant tests were conducted using
the 1.2 x 10 BTU/hr boiler, used for the NH3 mixing tests
and catalyst selection tests described in Section 5. The
boiler was modified to burn #2 fuel oil, and to heat and con-
trol the flue gas temperature over a range from 500 to 640°F.
Combustion gases are exhausted from the boiler stack by
an ID fan and pass through the heater section, mixing section,
and catalytic converter. All ductwork downstream of the fan
is 6" square. The heater section contains 6 finned "Calrod"
heaters, rated at 2 kw each. The heaters are wired in pairs,
two of which are switched manually, and the third by a tem-
perature controller. The flue gas temperature at the catalyst
bed inlet can be controlled to any desired temperature between
500°F, with none of the heaters on, and 640°F, with all heaters
on. Ammonia and NO are injected into the center of the boiler
stack, so that they may mix with the flue gas during passage
through the fan and downstream ductwork. (Injection of bottled
NO is required to increase the flue gas NO concentration from
75 ppm produced by the burner, to 125 ppm expected in this
utility pilot plant boiler.)
Total flue gas flowrate is approximately 12,500 SCFH,
and is measured by an orifice plate at the duct exit. Since
the catalyst bed dimensions are 6" x 6" x 12" (catalyst volume
q
is 0.250 ft ), the catalyst operates at a space velocity of
50,000 v/v/hr.
Inlet and exit gas samples are drawn from the center of
the duct upstream and downstream from the catalyst bed.
Samples are analyzed for NOx, 02, and SO2 by a Beckman 951
10.1
-------
Chemiluminescence NOx Analyzer, Beckman 715 Process 02 Monitor,
and EnviroMetries NS200 S02/N0x Analyzer (for S02).
The first tests conducted with the laboratory pilot plant
used a batch of platinum catalyst, 6" square x 12" thick.
The catalyst was ordered from Matthey Bishop, Inc., suppliers
of the pilot plant catalyst, and was to be produced in a man-
ner identical to that of the utility pilot plant catalyst.
First, the effect of inlet NH^ concentration on NOx re-
moval efficiency was determined, at three different temperatures.
Results of these tests are shown in Figure 10.1. As can be
seen, NOx removal efficiency is, in general, approximately the
same at 540 and 580°F, and significantly less at 610°F, at the
NHo concentration tested. Furthermore, NOx removal efficiency
continues to increase with increasing NHo concentration up to
1000 ppm, although the increment decreased beyond 550 ppm.
Next, the effect of continued exposure to oil fired flue
gas on NOx removal efficiency was explored. Using the data of
Figure 10.1, the conditions initially chosen were 500 ppm NH^
and 540°F.
At 540°F, NOx removal efficiency dropped from approxi-
mately 707o to 50% after about 8 hours. At this point, tempera-
ture was increased to 580°F. Removal efficiency increased to
70%, but, after approximately 32 hours, dropped to 6070. At
that point, temperature was again increased, to 610°F. Removal
efficiency remained constant for 20 hours, at 6570, but then
decreased to 40% after 60 hours of operation at 610°F (total
age, 140 hours). The NOx removal efficiency at 580°F was
checked and found to be 20%, compared to 60% when the catalyst
age was 60 hours. Activity at a higher temperature (640°F)
was also tested, and found to be 30%. At that time, oil fired
testing of Catalyst #1 was stopped.
The catalyst was then removed from the converter, and, in
10.2
-------
an attempt to regenerate it, was heated at 700°F, overnight,
in an electric kiln. Following this procedure, the catalyst
was run on gas firing to determine its NOx removal efficiency
as a function of temperature. After 72 hours, the catalyst
reached the steady state performance. The NOx removal effi-
ciency peaked at 85% at a temperature of 525°F. This is
quite different from the performance of the utility pilot
plant catalyst where peak efficiency of 90% at 475*" to 500°F
has been observed. Tests on this catalyst were discontinued
and the catalyst manufacturer was contacted to determine whe-
ther this batch was indeed identical to that supplied for the
pilot plant. It was then discovered that the manufacturing
process was slightly different from that used originally.
Although the resulting catalyst was considered identical,
insofar as precious metal loading and substrate characteristics
are concerned, this may account for the observed discrepancies.
The next step in the laboratory pilot plant work was to
remove some catalyst from the utility pilot plant, for testing
in the laboratory pilot plant. Two catalyst baskets were re-
moved from the air preheater and the catalyst was cut to pro-
vide two 6" square x 12" thick beds for the laboratory pilot
plant.
Catalyst from basket #3 of the utility pilot plant air
preheater was tested first, in the oil fired laboratory pilot
plant, with the results shown by the circles in Figure 10.2.
Figure 10.2 shows the effect of inlet NHo concentration on
NOx removal efficiency, and Figure 10.3 the effect of tempera-
ture on NOx removal efficiency. The effect of NH^ concentra-
tion was first determined at 560°F. Other operating conditions,
constant for all tests, were: space velocity, 50,000 hr and
inlet NOx concentration: 100 ppm. The results were somewhat
surprising, in that NOx removal efficiency continued to in-
crease as NH-j concentration was increased to 1000 ppm. It was
10.3
-------
decided to determine the effect of temperature using an NH^
inlet concentration of 300 ppm. Results of this work are
shown by the circles in Figure 10.3. At this level, peak
NOx removal efficiency of 45% occurred at 600-620°F. It
appeared that the catalyst's performance was rapidly decreas-
ing, compared with the average of 50% observed for several
hundred hours during oil fired operation of the utility pilot
plant. Furthermore, the performance peak occurred at signif-
icantly higher temperatures than observed previously.
The catalyst from basket #2 was then loaded into the
laboratory pilot plant, and tested in a manner similar to the
first, with the results shown by the squares in Figure 10.2.
The performance of basket #2 catalyst was similar to, but some-
what lower than, that of basket #3.
On the theory that operation in the rotating air preheater
in the pilot plant (alternate exposure to flue gas and warm
air) provided a continuous regeneration effect, delaying the
deactivation of the catalyst, basket #3 catalyst was regene-
rated. The regeneration procedure consisted of placing the
catalyst in an electric kiln held at 500°F for 66 hours.
After this procedure, the catalyst was replaced in the reactor
and tested, with the results shown in Figure 10.4.
The first tests performed were to determine the effect of
temperature on NOx removal efficiency, at 300 ppm inlet NH.,
concentration, under oil firing. Results are shown by the
circles in Figure 10.4. It can be seen that the catalyst per-
formance was significantly improved by the regeneration pro-
cedure. Peak NOx removal efficiency increased to 58% (from
45%) and the temperature peak shifted to 580 to 600°F (from
600-620°F). After operating for 14 hours at 580°F, the cata-
lyst performance had dropped slightly, from 54% to 48%. At
this point, it was decided to switch the boiler to gas firing
10.4
-------
in order to determine whether the catalyst performance would
return to that observed on gas firing before exposure to oil
fired flue gas. Therefore, the gas burner was installed in
place of the oil burner on the lab pilot boiler and tests
were made to determine the effects of temperature on NOx re-
moval efficiency under gas firing, with other conditions
identical to those described above. Results are shown by
the squares in Figure 10.4.
It appears that the effects of exposure to oil firing are
reversible; i.e., after a switch to gas firing, the catalyst
gradually reverts to its performance prior to oil firing.
The data from the oil fired laboratory pilot tests on the
utility pilot plant catalyst is, in several respects, contrary
to what has been observed in Environics' previous small-scale
(2" dia.) tests. In these previous experiments, the catalyst,
identical to that used in the current work, exhibited 80% NOx
removal efficiency at 575°F, and 300 ppm NH3, after 450 hours
exposure to furnace flue gas containing 300 ppm S02- The
major difference between the previous and current work, other
than scale, is that the early work was conducted with flue
gas from a gas fired furnace, to which bottled 862 was in-
jected. It is possible that the flue gas in the current oil
fired boiler tests contains compounds other than SC^ (such as
heavy metal compounds) which could "poison" the catalyst.
Although no definite conclusions can be drawn, it does
suggest that the platinum catalyst may be susceptible to
poisoning by some component of oil fired flue gas, other
than 862. Using a platinum catalyst in actual oil fired
flue gas, NOx removal efficiency of 50% can be obtained in
a regenerative system such as an air preheater wheel, at
temperatures between 550 and 600°F. The activity of the
catalyst does appear to be restored when the boiler is re-
turned to gas firing.
-------
00
Oil Fired Laboratory Pilot Plant
Test Conditions: Temperatures D 540 F
O580 F
O6IO F
Catalyst- Pt.(New)
Space Velocity 50,000 v/v/hr
230 400 600 300
Inlet Ammonia Concentration in ppm
FigurelO. I NOx Removal Eff. Vs Inlet Ammonia Cone
New Catalyst (Pt)
.000
10.6
-------
Oil Fired Laboratory Pilot Plant
100 Test Conditions: Temperatures 560, F
80
60
LU
§40
o
E
OJ
C£
X
o
20
Catalyst- Pt. ( Utility Pilot Plant)
Space Velocity 50,000 v/v/hr
Inlet NO* Cone. 100 ppm
0
800
200 400 600
Inlet Ammonia concentration in ppm
Figure 10.2 NOx Removal Eff. Vs Inlet Ammonia Cone.
(Utility Pilot Plant Catalyst)
1.000
10.7
-------
00
60
o
O>
Qi
X
O
20
Test Conditions:
Catalyst: Utility Pilot Plant (Pt)
Space Velocity : 50.000 v/v/hr
Inlet NOx Cone. : 100 ppm
Ammonia Injection- 300 ppm
Basket #3 catalyst
0
500
600
Laboratory Pilot Plant Flue Gas Temp, in Degrees F
FigurelO.3 NOx Removal Eff. Vs Temp.
Laboratory Pilot Plant
700
o
oo
-------
Basket #3 catalyst regenerated, gas firing
80
60
Test Conditions:
D Gas Fired
O Oil "
Space Velocity: 50,000v/v/hr
Inlet NOx Cone. 100 ppm
Ammonia Injection Cone. 300 ppm
Catalyst: Utility Pilot Plant-REGENERATED
Jasket #3 regenerated, oil firing
LU
(0
o
0>
4C
20
0
450
500 550 600
' Laboratory Pilot Plant Flue Gas Temp, in Degrees F.
FigurelO.4 NOx Removal Eff. Vs Temp.
Laboratory Pilot Plant
650
-------
SECTION 11
COST ESTIMATE-480 MW PLANT
It was desired to prepare a cost estimate for a full
scale catalytic NOx reduction system, as part of the pilot
plant program. A preliminary cost estimate for a full scale
gas fired boiler NOx reduction system is therefore presented
below, based upon the following assumptions.
1. The system will be installed on a 480 MW plant,
flowing SOxlO6 SCFH of flue gas.
2. The concentration of NOx in the flue gas will be 125
ppm.
3. The catalyst cost will be $1,000 per cubic foot.
4. Ammonia will cost 18 cents per pound, including
tankage.
5. The catalyst bed will be designed for operation at
50000 hr" space velocity.
6. The catalyst will be installed in the middle baskets
of the existing air preheater.
7. The catalyst bed will be replaced at 5 year intervals;
25% of the replacement catalyst cost can be offset
by reclamation of the precious raetal in the spent bed.
8. The cost of capital is 10%.
Cost of the system can be broken down as follows:
1. Capital Costs
A. Catalyst bed
B. Catalyst containment system
C. Instrumentation and control system
-------
2. Operational Costs
A. Ammonia consumption
B. Maintenance expense
The estimated capital expenditure schedule is shown in
Table 11.1. These estimates are discussed separately, below.
1. Catalyst Bed Cost
The required catalyst bed volume (in the flue gas
stream) is 1000 cubic feet. Because of its location in the
air preheater, only 457» of the catalyst is actually in the
o
gas stream at any given time. Therefore, 2222 ft of catalyst
3
are actually required. A catalyst cost of $1000 per ft has
been quoted for purchase of such quantities. The catalyst cost
is therefore $2,222,000. The best projection that can be
made at this time is that the catalyst bed would have to be
replaced no more often than every five years. It is probable
that 25% of the replacement catalyst cost could be defrayed by
recovery of the precious metal from the spent catalyst. Re-
placement beds would thus cost $1,666,500 each. The present
value of the original catalyst bed and five replacement beds
is thus $4,668,000 over the 30 year life of the plant.
2. Catalyst Containment
The catalyst containment system will hold the catalyst
modules in place in the air preheater wheel, to provide pro-
tection from mechanical shock, prevent gas from bypassing the
catalyst, and allow access to sections of the catalyst bed for
cleaning and/or replacement. Cost of this system is estimated
to be $500,000.
3. Instrumentation and Control System Cost
It is estimated that the instrumentation and control
system will initially cost approximately $20,000 and have a
life of 10 years. The present value of instrumentation is
then $30,700 over the 30 year life of the plant.
11.2
-------
4. Operation Cost
Ammonia will be consumed at the maximum rate of
455 Ib/hr. Ammonia cost is estimated to be 18 cents per
pound, or $82 per hour. Thus, for a 480 MW steam plant,
ammonia cost would amount to 0.017 cents per kilowatt-hour.
Additional operational cost will come from maintenance of
the system. It is estimated that they will require less
than one man full time, or $20,000 per year (0.0005 cents
per kilowatt-hour).
A cost summary for the catalytic reduction system is
presented in Table 11.2. Standard present value methods were
used. The capital cost per KW of plant capacity is $10.83.
The operating cost is 0.0176*i per KW hour.
11.3
-------
CAPITAL OUTLAY SCHEDULE
Years
Catalyst
Catalyst
Containment
Instrumen-
tation
0
$2,222,000
$ 500,000
$ 20,000
5
1,666,500
10
1,666,500
20,000
15
1,666,500
20
1,666,500
20,000
25
1,666,500
30
(555,500)*
*denotes salvage value of last catalyst bid.
Table 11.1
-------
TABLE 11.2
COST SUMMARY
FOR 480 MW STEAM PLANT
1. CAPITAL COST
Present Value of Catalyst Cost
Present Value of Catalyst Containment
Cost
Present Value of Instrumentation Cost
Total Present Value Cost
Total Present Value Cost, per KW of
Plant Capacity
$4,668,000
500,000
30.680
$5,198,680
$10.83
2. OPERATING COST
Total Maintenance Cost (per year)
Maintenance Cost (per KW hour)
Ammonia Cost (per KW hour)
Total Operating Cost per KW hour
20.000
0.00050
0.0171Q
0.01760
11.5
-------
APPENDIX A
CONVERSION FACTORS
Environmental Protection Agency policy is to express
all measurements in agency documents in metric units. When
implementing this practice will result in undue cost or lack
of clarity, conversion factors are provided for the non-
metric units used in a report. Generally, this report uses
British units of measure. For conversion to the metric sys-
tem, use the following conversions:
To convert from
British units
SCFH
v/v/hr
ppm
Inch
Feet
square inches
Inches of water
RPM
MW
°F - it °C + 32
Btu per hour
To Metric
units
NCMH
(Normal Cubic
Meter Per Hr)
v/v/hr
ppm
Centimeter
Meter
square centimeters
mm of Mercury
RPM
%
MW
Ki lo-calories per hour
Multiply by
2.832 x 10"2
1
1
2.540
0.3048
6.452
1.868
1
1
1
0.2520
A.I
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. HKPOflT NO,
EPA-600/7-76-031
2.
4. TITLE ANDSUBTITLE
Catalytic Reduction of Nitrogen Oxides with Ammonia:
Utility Pilot Plant Operation
5. REPORT DATE
October 1976
6. PERFORMING ORGANIZATION CODE
I. RECIPIENT'S ACCESSION NO.
7. AUTHOR(S)
Jules M. Kline, Paul H. Owen, and Y. C. Lee
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environics, Inc.
4101 Westerly Place, Suite 107
Newport Beach, California 92660
10. PROGRAM ELEMENT NO.
EHE624
11. CONTRACT/GRANT NO.
68-02-0292
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final: 1/73-10/76
14. SPONSORING AGENCY CODE
EPA-ORD
is.SUPPLEMENTARY NOTES ffiRL-RTP project officer for this report is R. D. Stern, 919/549-
8411 Ext 2915, Mail Drop 61.
is. ABSTRACT
repOrt describes work to demonstrate, on a utility pilot plant scale, the
performance, reliability, and practicability of reducing nitrogen oxides (NOx) emis-
sions from steam boilers by reduction of NOx with ammonia over a platinum catalyst.
A utility pilot plant treating a slipstream from an operating electric utility boiler,
equivalent to approximately 1. 5 MW output, was designed, installed, and tested on gas
and oil fuel. Laboratory pilot plant testing supplemented the utility pilot plant testing.
Results of gas -firing the utility pilot plant, together with laboratory pilot plant test
results, indicated that the catalytic reduction system has consistently provided 85-90^
NOx removal for over 4000 hours , with no significant performance loss at a 50,000
per hour space velocity. Results of oil-firing the utility pilot plant, together with
laboratory pilot plant tests results , indicated that this system could provide at least
65% NOx removal when the flue gas temperature is above 575 F. Preliminary cost
estimates indicate a capital expenditure of less than #11 per kW plant capacity, with an
operating expenditure of less than 0. 02^ per kWhr for a full-scale system, based on
1974 dollars and a gas -fired 480 MW plant.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Flue Gases
Nitrogen Oxides
Catalysis
Platinum
Reduction (Chemistry)
Ammonia
Boilers
Natural Gas
Fuel Oil
Air Pollution Control
Stationary Sources
Catalytic Reduction
13 B
2 IB
07B
07D
13A
21D
13. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
92
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
A.2
------- |