x>EPA
          unuea States      Industrial Environmental Research  EPA-600/7-79-181
          Environmental Protection  Laboratory          August 1979
          Agency        Research Triangle Park NC 27711
Design Criteria
for Stationary
Source Catalytic
Combustion Systems
                  *
Interagency
Energy/Environment
R&D Program Report

-------
                  RESEARCH REPORTING SERIES


Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and  application of en-
vironmental technology. Elimination of traditional grouping was  consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

    1. Environmental Health Effects Research

    2. Environmental Protection Technology

    3. Ecological Research

    4. Environmental Monitoring

    5. Socioeconomic Environmental Studies

    6. Scientific and Technical Assessment Reports (STAR)

    7. Interagency Energy-Environment Research and Development

    8. "Special" Reports

    9. Miscellaneous Reports

This report has been assigned to the 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 sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses 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 environ-
mental issues.
                        EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval  does not 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 recommendation for use.

This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                    EPA-600/7-79-181

                                          August 1979
       Design  Criteria for
Stationary Source Catalytic
     Combustion  Systems
                     by

         J.P Kesselring, W.V. Krill, H.L. Atkins,
            R.M. Kendall, and J.T. Kelly

       Acurex/Energy and Environmental Division
               485 Clyde Avenue
           Mountain View, California 94042
             Contract No. 68-02-2116
           Program Element No. EHE624A
         EPA Project Officer: G. Blair Martin

       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

-------
                             TABLE OF CONTENTS

Section                                                                Page
          ACKNOWLEDGEMENTS                                             Xix
   1      SUMMARY                                                      1-1
          1.1   Catalyst Screening Test Results 	   1-1
                1.1.1   Catalyst Materials Review 	   1-1
                1.1.2   Catalytic Combustor Analysis  	   1-2
                1.1.3   Catalyst Screening Tests  ....  	   1-3
                1.1.4   Graded Cell Catalyst Tests	   1-4
          1.2   System Configuration Test Results   	   1-5
                1.2.1   Characterization of Stationary Combustion
                        Systems	  .   1-6
                1.2.2   Combustion System Configuration Tests ....   1-6
                1.2.3   Prototype System Design Concepts	   1-7
          1.3   Conclusions	   1-7
   2      INTRODUCTION	   2-1
          2.1   Related Research Programs in Catalytic Combustion .  .   2-1
                2.1.1   Gas Turbine Applications	2-1
                2.1.2   Residential Furnace Applications  	   2-8
                2.1.3   Domestic Appliance Applications	2-8
                2.1.4   Life Support Systems	   2-9
                2.1.5   Fundamental Programs  	   2-10
          2.2   Program Purpose and Goals	   2-13
          References	   2-16
   3      CHARACTERIZATION OF STATIONARY COMBUSTION SYSTEMS 	   3-1
          3.1   General Considerations  	   3-1
          3.2   Equipment and Operating Characteristics 	   3-2
                3.2.1   Stationary Gas Turbines	3-2
                3.2.2   Supercharged Boilers  	   3-10
          3.3   Air Pollutant Emission Characteristics	   3-18
          3.4   Conclusions	   3-24
          References	   3-25

-------
                       TABLE OF CONTENTS (Continued)

Section                                                                Page
   4      CATALYST MATERIALS REVIEW 	    4-1
          4.1   General Considerations  	    4-1
          4.2   Characteristics and Properties of Catalyst
                Materials	    4-2
                4.2.1   Monolithic and Cylindrical Supports ....    4-2
                4.2.2   Washcoat Substrates 	    4-17
                4.2.3   Catalyst Coatings 	    4-19
          4.3   Conclusions	    4-25
          References	    4-26
   5      CATALYST PREPARATION AND CHARACTERIZATION 	    5-1
          5.1   General Considerations  	    5-1
          5.2   Catalyst Preparation  	    5-2
          5.3   Catalyst Characterization 	    5-2
                5.3.1   Total Surface Area  	    5-4
                5.3.2   Selected Surface Area (Dispersion)	    5-6
                5.3.3   Pore Size and Pore Volume	    5-8
                5.3.4   SEM-EDAX Analysis 	    5-10
          5.4   Catalyst Characterization Laboratory  	    5-11
                5.4.1   Gas Adsorption Apparatus	    5-11
                5.4.2   Test Procedure	    5-16
          List of Symbols	    5-18
          References	    5-19
   6      CATALYTIC COMBUSTOR ANALYSIS	    6-1
          6.1   Fundamentals of Operation   	    6-1
                6.1.1   Graphical Determination of Stable Surface
                        Combustion States 	    6-4
                6.1.2   Conclusions	    6-8
          6.2   The PROF-HET Computer Code	    6-9
                6.2.1   Comparison to Existing Models 	    6-9
                6.2.2   Model Formulation 	    6-11
                6.2.3   Parametric Calculations 	    6-18
          6.3   Conclusions	                6-38
                                     IV

-------
                       TABLE OF CONTENTS (Continued)

Section                                                                Page
   6      6.4   Recommendations	6-39
                6.4.1   Graded Cell Catalyst Optimization Maps.  .  .  .   6-39
                6.4.2   Breakthrough Analysis 	   6-40
                6.4.3   NOX Emission Characteristics of Catalytic
                        Combustors	6-40
                6.4.4   Effect of Transition on Blowout	6-40
          List of Symbols	   6-41
          References	   6-43
   7      CATALYST SCREENING TESTS  	   7-1
          7.1   General Considerations	  .   7-1
          7.2   Catalyst Test Matrix	7-2
          7.3   JPL Test Facility	   7-2
          7.4   Test Data Summary	7-13
          7.5   Conclusions	7-86
   8      GRADED CELL CATALYST TESTS	8-1
          8.1   Introduction	8-1
          8.2   Graded Cell Catalyst Matrix 	   8-1
          8.3   Acurex Test Facilities	8-3
          8.4   Combustion Screening Tests  	   8-10
                8.4.1   Catalyst Comparison Tests 	   8-11
                8.4.2   High Temperature Evaluation 	   8-28
                8.4.3   Catalyst Scaleup  	   8-36
          8.5   Extensive Evaluation Tests  	   8-40
                8.5.1   Fuel Nitrogen Tests	8-40
                8.5.2   High Pressure Tests	8-49
          8.6   Conclusions	8-51
          References	8-53
   9      COMBUSTION SYSTEM CONFIGURATION TESTS 	   9-1
          9.1   General Considerations  	   9-1
          9.2  Two  Stage  Combustor	9-1
                9.2.1    System Design and Fabrication 	   9-2

-------
                      TABLE OF CONTENTS (Concluded)

Section                                                                Page
   9            9.2.2  Test Results	9-5
          9.3   Model Gas Turbine Combustor 	  9-11
                9.3.1  System Design and Fabrication	9-11
                9.3.2  Test Results	9-12
                9.3.3  Advanced Graded Cell Concept Demonstration .  .  9-17
          9.4   Radiative Catalyst/Watertube System 	  9-21
                9.4.1  System Design and Fabrication  	  9-21
                9.4.2  Test Results	9-28
          9.5   Conclusions	9-38
  10      PROTOTYPE SYSTEM DESIGN CONCEPTS  	  10-1
          10.1  Introduction	  10-1
          10.2  Industrial and Commercial  Boilers 	  10-1
                10.2.1  Firetube Boilers  	   10-1
                10.2.2  Watertube Boilers  	  10-6
                10.2.3  Two Stage Catalytic Systems 	   10-6
          10.3  Gas Turbines	   10-8
          10.4  Other Systems	   10-13
          10.5  Conclusions	   10-14
          References	  10-15
  11      CONCLUSIONS AND RECOMMENDATIONS  	   11-1
          11.1  Conclusions   	   11-1
          11.2  Recommendations	11-2
Appendices
    A     SECTION 7 DATA SUPPLEMENT - CATALYST SCREENING TESTS ...  A-l
    B     SECTION 8 DATA SUPPLEMENT - GRADED CELL CATALYST
          TESTS	  B-l
    C     SECTION 9 DATA SUPPLEMENT - COMBUSTION SYSTEM
          CONFIGURATION TESTS 	  C-l

-------
                            LIST OF ILLUSTRATIONS

Figure                                                                 Page
 2-1      NASA multiple conical tube fuel injector                     2-4
 3-1      Simple cycle gas turbine system                              3-4
 3-2      Simple cycle gas turbine combustor inlet temperature
          vs compressor pressure ratio                                 3-11
 3-3      Regenerative cycle gas turbine combustor inlet
          temperature vs turbine inlet temperature-regenerator
          effectiveness 0.90                                           3-12
 3-4      Regenerative cycle gas turbine combustor inlet
          temperature vs turbine inlet temperature-regenerator
          effectiveness 0.70                                           3-13
 3-5      Predicted air-fuel ratio vs turbine inlet temperature
          (natural gas)                                                3-14
 3-6      Marine supercharged boiler                                   3-16
 3-7      The self-sustaining supercharged cycle                       3-17
 3-8      The power supercharged cycle                                 3-18
 3-9      Distribution of stationary anthropogenic NOX emissions
          for the year 1974 (stationary fuel combustion:  con-
          trolled NOX levels)                                          3-20
 4-1      Examples of Thermacomb corrugated ceramics, produced
          by American Lava Corporation                                 4-7
 4-2      Celcor cordierite monoliths produced by Corning Glass
          Works                                                        4-8
 4-3      Corning high temperature graded cell ceramic                 4-9
 4-4      Zirconia spinel monoliths from Corning Glass Works —
          flexible rectangle and square cell geometries                4-10
 4-5      Torvex ceramic honeycomb configurations by DuPont            4-12
 4-6      Versagrid ceramic honeycomb by General Refractories
          Company                                                      4-13
 4-7      Poramic monolith structures by W. R. Grace and Co.           4-14
 4-8      Kanthal metal monolith by Kentucky Metals, Inc.              4-15
 4-9      Spectramic silicon carbide honeycomb by Norton Company       4-16
 4-10     Washcoat structure on monolith - schematic represen-
          tation                                                       4-18
 5-1      Argon BET at 77K                                             5-5
 5-2      Comparison of Type II and Type IV Langmuir isotherms         5-9
 5-3      High vacuum gas adsorption apparatus                         5-12
                                     VII

-------
                      LIST OF ILLUSTRATIONS (Continued)

Figure                                                                 Page
 5-4      Schematic diagram of vacuum control system 	    5-13
 5-5      Detail of gas adsorption system showing dosing volumes .  .    5-15
 6-1      Physical events in a monolith cell	    6~3
 6-2      Wall temperature variation with lean reactant wall
          concentration  	    6-4
 6-3      Mass flux as a function of mass fraction of lean
          reactant at the monolith wall	    6-5
 6-4      Simplified mass balance solution for catalytic combus-
          tion in a monolith bed	    6-7
 6-5      Wall and bulk gas temperature and fuel  concentration
          through bed	    6-19
 6-6      Wall fuel volume fraction distributions for several
          flowrates	    6-20
 6-7      Blowout mass throughput for various channel diameters  .  .    6-21
 6-8      Blowout mass throughput for various gas preheats 	    6-22
 6-9      Blowout mass throughput for various excess air levels.  .  .    6-23
 6-10     Blowout mass throughput for various bed conductivities .  .    6-24
 6-11     Blowout mass throughput for various surface reaction
          activities   	    6-29
 6-12     Bulk gas fuel concentration through bed	    6-31
 6-13     Bulk gas temperature through bed	    6-32
 6-14     Detailed species concentrations through bed  	    6-35
 6-15     Effect of channel diameter on breakthrough 	    6-36
 6-16     Effect of gas preheat temperature on breakthrough	    6-37
 7-1      JPL patio test facility	    7-7
 7-2      Model JPL-001, platinized cordierite with 30 thermo-
          couples placed in monolith	    7-9
 7-3      Quartz reactor with monolithic catalyst bed  	    7-10
 7-4      JPL patio test facility control console  	    7-12
 7-5      Model 004 - mullite/alumina/platinum	    7-14
 7-6      Model 006 - cordierite/alumina/platinum  	    7-15
 7-7a     Screening data, JPL-004X - preheat temperature  (methane/
          air)	    7-17
 7-7b     Screening data, JPL-004X - space velocity (methane/air).  .    7-18
 7-8a     Screening data, JPL-005X — preheat temperature
          (methane/air)  	    7-19

                                       viii

-------
LIST OF ILLUSTRATIONS (Continued)
Figure
7-8b
7-9a

7-9b
7-10a

7-1 Ob
7-lla

7-llb
7-12a

7-12b
7-13a

7-13b
7-14a

7-145

7-15

7-16a

7-165
7-17a

7-175
7-18

7-19

7-20a


Screening data, JPL-005X - space velocity 	
Screening data, JPL-006 — preheat temperature (methane/
air) 	
Screening data, JPL-006 —space velocity (methane/air). .
Screening data, JPL-006X - preheat temperature (methane/
air) 	
Screening data, JPL-006X - space velocity (methane/air) .
Screening data, JPL-007 - preheat temperature (methane/
air) 	
Screening data, JPL-007 -space velocity (methane/air). .
Screening data, JPL-008 — preheat temperature (methane/
air) 	
Screening data, JPL-008 - space velocity (methane/air). .
Screening data, JPL-009 -preheat temperature (methane/
air) 	 	
Screening data, JPL-009 - space velocity(methane/air) . .
Screening data, JPL-010 and -010X —preheat temperature
(methane/air) 	
Screening data, JPL-010 and -01 OX - space velocity
(methane/air) 	
Degradation of test model JPL-010P at lean conditions
(>350 percent T.A.) (methane/air) 	
Screening data, JPL-010P - preheat temperature (methane/
air) 	
Screening data, JPL-010P - space velocity (methane/air) .
Screening data, JPL-010P - preheat temperature (propane/
air) 	
Screening data, JPL-010P - space velocity (propane/air) .
Surface analysis at exit of channel 1, test model
JPL-010P 	
Surface analysis at exit of channel 4, test model
JPL-010P 	
Screening data, JPL-011 -preheat temperature (methane/
air) 	
Page
7-20

7-21
7-22

7-23
7-24

7-26
7-27

7-29
7-30

7-31
7-32

7-34

7-35

7-37

7-38
7-39

7-40
7-41

7-43

7-44

7-45
                IX

-------
                      LIST OF ILLUSTRATIONS (Continued)
Figure
 7-20b
 7-21a

 7-21b
 7-22a

 7-22b
 7-23a

 7-23b
 7-24a

 7-24b
 7-25a

 7-25b
 7-26
 7-27
 7-28

 7-29a

 7-29b
 7-30

 7-31

 7-32

 7-33

 7-34

 7-35
                                                            Page
Screening data, JPL-011 - space velocity (methane/air) .  .   7-46
Screening data, JPL-011 -preheat tempeature (propane/
air)	
                                                            7-47
Screening data, JPL-011 - space velocity (propane/air) .  .   7-48
Screening data, JPL-012 - preheattemperature (methane/
air)	
                                                            7-50
Screening data, JPL-012 — space velocity (methane/air) .  .   7-51
Screening data, JPL-013 - preheat temperature (methane/
air)	
Screening data, JPL-013 - space velocity (methane/air) .
Screening data, JPL-022 -preheat temperature (methane/
air)	
Screening data, JPL-022 - space velocity (methane/air) .
Screening data, JPL-016 - preheat temperature (methane/
air)	
Screening data, JPL-016 - space velocity (methane/air) .
Test specimen preparation, washcoat stabilization tests.
                                                            7-53
                                                            7-54

                                                            7-55
                                                            7-56
                                                            7-58
                                                            7-59
                                                            7-60
Washcoat stabiliziation study results  	   7-61
Fuel flow capability of JPL-021 catalyst at lean
conditions (350% TA, methane/air)  	   7-64
Screening data, JPL-021 - preheat temperature (methane/
air)	
Screening data, JPL-021 - space velocity (methane/air)
                                                            7-65
                                                            7-66
Test model JPL-019, segment from inlet (large cell)
monolith; R = rear, F = front	7-76
Surface analysis at entrance of large cell  segment,
test model JPL-019	7-77
Surface analysis at exit of large cell segment, test
model JPL-019	7-78
Surface analysis at entrance of small cell  segment,
test model JPL-019	7-79
Surface analysis at exit of small cell segment, test
model JPL-019	7-80
Surface appearance at entrance of large cell segment,
test model JPL-019	7-81

-------
                      LIST OF ILLUSTRATIONS (Continued)

Figure                                                                 Page
 7-36      Surface appearance at exit of large cell segment,
           test model JPL-019 	  7-82
 7-37      SEM/EDAX measurements on exit of large cell segment,
           test model JPL-019 	  7-83
 7-38      Surface appearance at exit of small cell segment, test
           model JPL-019	7-84
 7-39      Washcoat comparison —small cell segment, test model
           JPL-019	7-85
 8-1       EPA/Acurex catalytic combustion test facility  	  8-6
 8-2       Air compressor and receiver	8-7
 8-3       Main air preheater	8-8
 8-4       Test section and exhaust system	8-9
 8-5       W. R. Grace graded cell Pt-Ir catalyst segments  	  8-12
 8-6       NOX emissions corrected to 0% Oo vs. maximum bed tempera-
           ture for catalyst A-025	8-14
 8-7       Pre- and post-test appearance - W. R. Grace Pt/Ir
           catalyst	8-15
 8-8       NOX emissions corrected to 0% Oo vs. preheat - UOP
           catalyst  (A-026), natural gas/air	8-18
 8-9       CO emissions corrected to 0% On vs. preheat — UOP
           catalyst  (A-026), natural gas/air   	  8-19
 8-10      Fuel rich emissions corrected to 0% Oo vs. preheat —
           UOP catalyst (A-026), natural gas/air   	  8-20
 8-11      UOP catalyst (A-026) emissions as a function of through-
           put, natural gas/air 	  8-21
 8-12      NO emissions corrected to 0% 02 vs. bed temperature for
           UOP catalyst (A-026), natural gas/air   	  8-22
 8-13      Catalyst A-032 (Matthey Bishop B) bed temperature
           distributions during aging, natural gas/air	8-24
 8-14      Corning high temperature graded cell ceramic 	  8-30
 8-15      Catalyst A-029 (NiO/Pt) NO emissions corrected to 0% Oo,
           natural gas/air	8-31
 8-16      Catalyst A-030 (CooOa/Pt) NO emissions corrected to
           0% 02> natural gas/air	8-33
 8-17      Catalyst A-030 (Co£03/Pt) NO emissions corrected to
           0% 0, natural gas/air    	8-34

-------
                    LIST  OF  ILLUSTRATIONS  (Continued)
:igure
8-18
8-19

8-20
8-21

8-22

8-23

8-24

8-25


9-1
9-2
9-3
9-4
9-5
9-6
9-7
9-8
9-9
9-10

9-11
9-12
9-13
9-14
9-15
9-16
9-17

NOX emissions comparison corrected to 0% excess §2 • • •
Blowout performance - catalyst A-041 , TRFn = 1589 K
(2400°F), natural gas and methane fuels. . 	
Impinger bottle gas sampling system 	
NOX as measured, catalyst A-030, natural gas doped
with ammonia 	
Percentage of NOo converted to NOX, catalyst A-030,
natural gas doped with ammonia 	
NH3 conversion characteristics, catalyst A-036,
natural gas doped with ammonia 	
NH3 conversion characteristics, catalyst A-037,
natural gas doped with ammonia 	
Lean data correlation: conversion of fuel nitrogen
to NOV 	
X
Two stage catalytic combustor concept 	
Two stage heat exchanger design envelope 	
Two stage catalytic arrangement 	
Two stage combustor assembly 	
Two stage combustor details 	
Two stage combustor fuel nitrogen conversion 	
Gas turbine combustor assembly, catalytic combustion . .
Model gas turbine combustor 	
Gas turbine fuel injector assembly 	
Advanced graded cell /model gas turbine combustor
assembly 	
Advanced graded cell fuel nitrogen conversion 	
Radiative catalyst/watertube combustion system concept .
Radiative catalyst/watertube arrangement 	
Catalyst cylinder heat transfer model 	
Catalytic radiative System I assembly 	
Radiative catalyst/watertube test section installation •
Radiative catalyst/watertube system energy release vs.
theoretical air 	
Page
8-35

8-38
8-41

8-43

8-44

8-45

8-47

8-50

9-3
9-4
9-6
9-7
9-8
9-9
9-13
9-14
9-15

9-18
9-19
9-22
9-24
9-25
9-26
9-27
Q-3D
9-18     Radiative catalyst/watertube system emissions  vs.
         percent theoretical  air	    9-31
                                    xn

-------
                       LIST OF ILLUSTRATIONS (Continued)

Figure                                                                Page
 9-19     Radiative catalyst/watertube energy release vs.
          throughput at 100 percent theoretical  air 	  9-32
 9-20     Emissions vs. throughput at 100 percent theoretical  air .  .  9-33
 9-21     Bed temperature profiles at 100 percent theoretical  air .  .  9-35
 9-22     Radiative catalyst/watertube system, ammonia-doped
          natural gas	9-37
 10-1     Four-pass scotch firetube boiler (courtesy of the
          Cleaver Brooks Company) 	  10-2
 10-2     Graded cell  firetube boiler concept 	  10-4
 10-3     Felt pad firetube boiler concept	10-5
 10-4     Radiant catalyst/watertube combustion system  	  10-7
 10-5     Catalytic gas turbine concept 	  10-9
 10-6     Stationary gas turbine graded cell catalytic combustor.  .  .  10-11
 10-7     Low temperature lightoff/preheat section  	  10-12
 B-l      Catalyst A-025 bed temperature distribution, effects
          of aging	.-  B-6
 B-2      Catalyst A-025 bed temperature distribution, effects
          of preheat	B-7
 B-3      Catalyst A-025 bed temperature distribution, effects
          of throughput	B-8
 B-4      Catalyst A-025 bed appearance rear view, varying
          stoichiometry 	  B-9
 B-5      Catalyst A-025 bed appearance rear view, varying bed
          temperature	B-10
 B-6      Catalyst A-025 bed appearance rear view, minimum preheat.  .  B-ll
 B-7      Catalyst A-025 bed appearance rear view, maximum
          throughput	B-l 2
 B-8      Surface appearance and EDAX analysis of small cell seg-
          ment inlet area, W.  R. Grace and Co. catalyst	B-14
 B-9      Surface appearance and EDAX analysis of small cell seg-
          ment outlet area, W. R. Grace and Co.  catalyst	B-15
 B-10     Surface appearance and EDAX analyses of medium cell
          segment, W.  R. Grace and Co. catalyst	B-17
 B-ll     Surface appearance of pretest catalyst surface,  W. R.
          Grace and Co.   	B-18
 B-12     EDAX analysis of untested catalyst surface  	  B-19
                                      xiii

-------
                       LIST OF ILLUSTRATIONS (Continued)
Figure
 B-13     Catalyst A-026 bed temperature distribution - effects
          of preheat, fuel lean	B-27
 B-14     Catalyst A-026 bed temperature distribution - effects
          of preheat, fuel rich 1598 K (2400°F) bed   	B-28
 B-15     Catalyst A-026 bed temperature distribution -effects
          of throughput	B-29
 B-16     Catalyst A-031 Matthey Bishop A bed temperature
          distributions during aging  	 B-36
 B-17     Catalyst A-031 Matthey Bishop A bed temperature
          distributions, effects of preheat 	 B-37
 B-18     Catalyst A-035 bed temperature distributions  	 B-40
 B-19     Catalyst A-035 bed temperature distributions  	 B-41
 B-20     Catalyst A-038 bed temperature distributions  	 B-45
 B-21     Catalyst A-038 bed temperature distributions  	 B-46
 B-22     Catalyst A-038 bed temperature distributions  	 B-47
 B-23     Catalyst A-030 (Acurex Co203/Pt) bed temperature
          distributions 	 B-57
 B-24     Catalyst A-030 (Acurex Co203/Pt) bed temperature
          distributions 	 B-58
                                      xiv

-------
                              LIST OF TABLES

Number                                                                 Page
 3-1      Types of Stationary Gas Turbine Equipment 	   3-3
 3-2      Principal Applications of Stationary Gas Turbines in
          the United States	   3-5
 3-3      Stationary Gas Turbine Power Generation (1973 Data)  ....   3-5
 3-4      Supercharged Boiler Full Load Performance Figures
          (Reference 3-5)	3-21
 3-5      1974 Summary of Air and Solid Pollutant Emission from
          Stationary Fuel Burning Equipment (1000 Mg) 	   3-23
 3-6      NOX Mass Emission Ranking of Stationary Combustion
          Equipment and Criteria Pollutant and Fuel Use Cross
          Ranking	3-24
 4-1      Ceramic Properties (Reference 4-3)	4-4
 4-2      Monolithic Support Data	4-6
 4-3      Ceramic Cylinder Properties 	   4-27
 4-4      Metals of Interest for Catalytic Combustion 	   4-29
 5-1      Catalyst Preparation Procedures 	   5-3
 5-2      Variation of Relative Pressure with Pore Radius 	   5-10
 5-3      Measured Volumes Used in Gas Adsorption Measurements.  .  .  .   5-16
 5-4      Catalyst Characterization Procedures  	   5-20
 6-1      Effect of Parameter Changes on Blowout. . 	   6-28
 6-2      CH4 Combustion Chemical Kinetic Reactions and Rates  ....   6-33
 7-1      JPL Test Model Summary	7-3
 7-2      Summary of Catalyst Characterization Results for
          Screening Catalysts 	   7-5
 7-3      JPL Test Procedure	7-6
 7-4      JPL Emissions Equipment	7-15
 7-5      Washcoat Stabilization Study Test Results 	   7-74
 7-6      Monolith 019 Test Data -JPL Multi-Fuel Tests	7-80
 7-7      Monolith 019 Test Data - Radial Bed Temperature Profiles.  .   7-81
 7-8      Monolith 019 Test Data -Axial Bed Temperature Profiles  .  .   7-82
 7-9      Monolith 019 Test Data -Axial Bed Temperature Profiles  .  .   7-83
 7-10     Monolith 019 Test Data -Axial Bed Temperature Profiles  .  .   7-85
 7-11     Monolith 019 Test Data -Axial Bed Temperature Profiles  .  .   7-86
 8-1      Graded Cell  Catalyst Model  Summary  	   8-2
                                    xv

-------
                        LIST OF TABLES (Continued)





Number                                                                Page
8-2

8-3
8-4
8-5
8-6
9-1
A-l
A- 2
A-3
A-4
A-5
A-6
A-7
A-8
A-9
A-10
A-ll
A-12
A-13
A-14
A-15
A-16

A-17

A-18

A-19

B-l
B-2
B-3
B-4
Summary of Surface Area and Dispersion Measurements on
Graded Cell Catalysts 	
Continuous Gas Analysis Instrumentation 	
W. R. Grace Catalyst Pretest Characterization 	
Oxide Preparation of Catalyst A-038 	
Blowout Data -Catalyst A-041 	
Platinum on Alumina Catalyst Cylinders 	
Screening Test Data Summary 	
Test Data - JPL-007 	
Test Data - JPL-008 	
Test Data - JPL-009 	
Test Data - JPL-010 	
Test Data - JPL-010X 	
Data Summary - JPL-010P 	
Test Data - JPL-011 	
Test Data - JPL-012 	
Test Data - JPL-013 	
Test Data - JPL-022 	
Test Data - JPL-016 	
Test Data - JPL-021 	
Monolith 019 Test Data - JPL Tests 	
Monolith 019 Test Data - Acurex Tests 	
Monolith 019 Test Data -Acurex Tests, Emissions Data
for Simulated Fuel Nitrogen Tests 	
Monolith 019 Test Data -Acurex Tests, High Pressure
Operation 	
Monolith 019 Test Data -Acurex Tests, Emissions Data
for Fuel Nitrogen Simulation Tests at Pressure 	
Monolith 019 Test Data -Natural Gas, High Temperature
Operation 	
Test Data Summary - Catalyst A-025 	
Emissions Data - Catalyst A-025 	
Lightoff Temperature History — Catalyst A-025 	
Results of Semiquantitative Analysis 	

8-4
8-10
8-11
8-27
8-39
9-28
A-2
A-6
A-7
A-8
A-9
A-10
A-12
A-13
A-15
A-16
A-17
A-18
A-19
A-20
A-21

A-22

A-23

A-24

A-25
B-3
B-4
B-5
B-20
                                    XVI

-------
LIST OF TABLES  (Continued)
Table
B-5

B-6
B-7
B-8
B-9
B-10

B-ll
B-12
B-13

B-14

B-15
B-16
B-17
B-18
B-19
B-20
B-21
B-22

B-23
B-24
B-25
B-26
C-l
C-2
C-3
C-4
C-5

UOP Lightoff Temperature Characteristics — Catalyst
A-026 	
Test Data Summary - Catalyst A-026 	
Emissions Data - Catalyst A-026 	
Test Data Summary — Catalyst A-027 	
Emissions Data - Catalyst A-027 	
Matthey Bishop A Lightoff Characteristics - Catalyst
A-031 	
Data Summary - Catalyst A-031 	
Data Summary - Catalyst A-035 	
Matthey Bishop C Lightoff Characteristics —Catalyst
A-035 	
Pfefferle Lightoff Characteristics -Catalyst A-038
(Co203) 	
Data Summary - Catalyst A-038 	
Lightoff Characteristics - Catalyst A-040 	
Data Summary - Catalyst A-040 	
Test Data Summary - Catalyst A-029 (NiO/Pt) 	
Emissions Data - Catalyst A-029 (NiO/Pt) 	
Lightoff Characteristics - Catalyst A-030 (Co203/Pt). . .
Screening Data Summary -Catalyst A-030 (Co203/Pt). . . .
UOP Scaleup Catalyst Lightoff Characteristics -
Catalyst A-041 	
Data Summary - Catalyst A-041 	
Extensive Evaluation Summary -Catalyst A-036 (NiO/Pt). .
Fuel Nitrogen Data - A-037 (Co^/Pt) 	
Fuel Lean Data for Fuel Nitrogen Conversion 	
Data Summary - Two Stage Combustor 	
Emissions Data — Two Stage Combustor 	
Data Summary - Model Gas Turbine 	
Data Summary -Advanced Graded Cell /Model Gas Turbine . .
Radiative Catalyst/Water System Test Matrix 	
Page

B-22
B-23
B-25
B-31
B-32

B-34
B-35
B-38

B-39

B-43
B-44
B-49
B-50
B-53
B-54
B-55
B-56

B-60
B-61
B-66
B-67
B-69
C-2
C-4
C-6
C-7
C-8
            xvn

-------
                        LIST OF TABLES (Concluded)

Table                                                                 Page
 C-6      Test Data Summary - Radiative Catalyst/Watertube
          System	   C-9
 C-7      Emissions Data -Radiative Catalyst/Watertube System  .  .   C-ll
 C-8      Emissions Data - Radiative Catalyst/Watertube Gas
          Chromatography  	   C-12
 C-9      Data Summary - Radiative Catalyst/Watertube System  ...   C-13
 C-10     Fuel Nitrogen Data - Radiative Catalyst/Watertube
          System	   C-14
                                   xv m

-------
                             ACKNOWLEDGEMENTS

       This document presents the results of a research and development
program to determine catalyst and combustion system design criteria for
stationary source catalytic combustors.  During the course of the program,
valuable contributions were made by many individuals.   In particular,
Acurex extends its appreciation to W. A. Boyer and R.  I. Frost of Corning
Glass Works, Dr. J. M. Maselli of W. R. Grace & Co., Dr. W. B. Retailick of
Oxy-Catalyst, Dr. A. S. D'Alessandro of Matthey Bishop, Inc., Dr. G.J.K.
Acres and Dr. B. E. Enga of Johnson Matthey Corp., and Dr. G. R. Lester of
Universal Oil Products for their support in providing test materials.
Special thanks are extended to Dr. G. Voecks of the Jet Propulsion Labora-
tory for his work in the catalyst screening program.  Additional thanks go
to Dr. R. M. Pierce, B. Hinton, and C. Smith of Pratt & Whitney Aircraft
for their support of the gas turbine model combustor test program.  The
significant contributions of Dr. W. C. Pfefferle, private consultant, and
Dr. R. Levy, Dr. J. A. Cusumano, and Dr. R. Dalla Betta of Catalytica
Associates, Inc. in catalyst selection are gratefully acknowledged.
       This program was conducted for the Combustion Research Branch of the
Industrial Environmental Research Laboratory, U.S. Environmental Protection
Agency.  G. B. Martin was the Project Officer.  The Acurex Program Manager
was Dr. J. P- Kesselring.  Valuable contributions to the final report were
made by R. J. Schreiber, A. J. Murphy, and R. E. Maurer.  Additional key
Acurex contributors include D.  Knirck, M. Angwin, M. Greer, W. Nurick,
Dr. J. T. Pogson, and C. D. Hartman.
                                   xix

-------
                                  SECTION 1
                                   SUMMARY

       The use of catalytic combustors in place of conventional burners has
been shown to reduce CO, HC, and NO  emissions in laboratory scale tests
                                   A
with both clean and ammonia-doped fuels.  The operating conditions for these
catalytic combustors are limited by the catalyst bed temperature capability.
Since the adiabatic, one-stage, low excess air operation that is necessary
for high system efficiency produces temperatures that are outside the current
temperature capability of catalyst materials, concepts to operate the cata-
lysts at permissible temperatures are required.  Successful system concepts
tested include direct heat removal from the catalyst through radiant heat
transfer, simulated exhaust gas recirculation, two stage catalytic combustion,
and high excess air (gas turbine) combustion.  A summary of results from the
program and the program conclusions follow.  For simplicity, the program has
been divided into two parts:   (1) catalyst screening tests and (2) system
configuration tests.

1.1    CATALYST SCREENING TEST RESULTS
       The results for catalyst screening tests include information obtained
from the catalyst materials review, catalyst preparation and characterization,
catalytic combustor analysis, and both single cell catalyst and graded cell
catalyst screening tests.  These results are summarized below.

1.1.1  Catalyst Materials Review
       A review of mid-temperature catalyst application literature has shown
that the monolithic honeycomb support is the most technologically advanced
configuration among catalyst carriers.  The configuration minimizes pressure
drop by having a straight through flow channel and large void area and also

-------
minimizes required catalyst volume by providing a large total surface area
per unit of volume.  Monolithic supports are currently available in a variety
of ceramic and metallic materials.
       To obtain maximum combustion catalyst performance at high temperature
(in excess of 1367K), it is desirable to have a high temperature support
material capable of maintaining a uniformly dispersed catalyst.  The catalyst
should exhibit a low lightoff temperature and sustained high activity.  To
obtain these properties, some tradeoffs in performance at varying combustion
conditions may be necessary.
       Previous catalysis applications had shown that the noble metal (plati-
num and palladium) catalysts are the most promising for high activity and
low lightoff temperature.  Simple oxides of the transition metals should have
good catalytic activity but will have higher lightoff temperatures than the
noble metals.  For very high temperature (>1778K) applications, mixed oxides
containing either free nickel oxides or cobalt oxide are the most promising
candidates.

1.1.2  Catalytic Combustor Analysis
       A catalytic combustor code (PROF-HET) has been used to predict the
performance of various catalyst combinations and configurations.  It has
shown that the overall success of the catalyst in reducing HC and CO emis-
sions to low levels in a short bed length is dependent on both surface and
gas phase reactions.  Surface reactions alone are insufficient to achieve
the desired low levels.
       Predictions from the PROF-HET code show the maximum mass throughput
for a catalytic combustor is determined by blowout of the surface and gas
phase reactions.  Blowout is predicted to increase nearly linearly with
increasing cell diameter, thermal conductivity, and catalyst/support surface
activity.  The maximum mass throughput is also predicted to increase expo-
nentially with increasing preheat temperature and fuel/air ratio for lean
operation.
       Small monolith cells allow fuel conversion to occur in a shorter
channel length than for large cells.  An increase in preheat temperature
                                      1-2

-------
also should result in a more rapid conversion of the fuel as it passes through
the channel.
       A catalytic combustor capable of high mass throughput and low emissions
can be constructed by joining two or more bed segments in series.   The first
segment would have large diameter cells to prevent blowout.   The final seg-
ment would have small diameter cells to initiate gas phase reactions and
achieve low emissions of CO, NO. and HC.
                               A
1.1.3  Catalyst Screening Tests
       Thirty-six catalyst material combinations were screened under combus-
tion conditions to investigate the effects of support, washcoat, and catalyst
properties and their interactions on combustion performance.  Mullite and
cordierite substrates performed adequately with platinum catalysts at 1367K
temperatures.  Tests conducted at higher temperatures show that platinum
catalysts on alumina substrates perform well up to 1783K but experience mild
thermal cracking.  Cobalt and nickel oxide catalysts on zirconia spinel  sub-
strates have high use temperatures (up to 1978K) but have severe thermal
shock problems.
       The washcoat provides a great increase in surface area of the support
materials, but exposure of the washcoat to high temperature results in sig-
nificant loss of pore area.  Typical changes in y-alumina surface area for
                                         2                    2
catalysts operated at 1367K were from 8 m /g pre-test to 0.6 m /g post-test.
Presintering tests of washcoats confirmed the loss of surface area but did
not have a negative effect on combustion properties.  This presintering may
reduce burying of subsequently applied active catalyst below the surface
during combustion.  Sintering of both washcoat and catalyst, however,
results in a reduction of the active platinum available to the reactants and
thus, reduced activity.  Therefore, higher catalyst loadings are required on
unsintered supports to provide activity equal to that of a catalyzed pre-
sintered support.
       Precious metal catalysts undergo degradation by oxidation and vaporiza-
tion as they operate at combustor temperatures.  Large decreases in surface
                          2            2
area (from as much as 20 m /g to zero m/g for catalysts operated above 1600K)
                                     1-3

-------
and dispersion (from 100 percent to zero percent) are indicators of this
degradation.  The impact of degradation on combustor performance can be
minimized by using increased catalyst loading and hydrogen sulfide fixation
of platinum to the support.
       A series of tests were conducted with platinum catalysts on varying
cell size supports.  Large cell monoliths exhibited very high mass through-
puts without blowout, but CO and HC emissions were high.  Small cell mono-
liths exhibited very low emissions but were easily blown out.   This result
is a direct confirmation of performance predictions from the catalytic  com-
bustor analysis.
                                                        o             3
       A catalyst with graded cell segments (6.35 x 10~°m, 4.76 x  10  m,
and  3.18 x  10"3m) was tested to verify the concept.  The mass throughput
capabilities were markedly increased over those of small cell catalysts,
such that the  graded cell catalyst could not be blown out at the maximum
flow capacity  of the test facility (0.85 Kg/hr of methane).  Comparable
 small  cell  monolithic catalysts were easily blown out at approximately
 0.20 Kg/hr  of  methane.  No carbon monoxide emissions were measurable, and
 only trace  hydrocarbons (<0.02 Kg/hr) were present in the exhaust.
        The  graded  cell model was further tested with a  variety  of  fuels.
 It was found  that  heavier gaseous hydrocarbon fuels promote lightoff at
 lower ignition temperature.  The lightoff temperature is fairly consistent
 for a given fuel and catalyst  type.  Lightoff between 672K and  788K tempera-
 tures is  typical for methane on platinum catalysts.

 1.1.4  Graded  Cell  Catalyst Tests
        Additional  testing of various catalysts on graded cell supports  pro-
vided further  verification of  the high throughput, low  emissions capability
of these systems.   Sixteen additional graded cell combustors were  fabricated
or procured for these tests.
        Precious metal catalysts degrade rapidly at temperatures above  1589K
(2400°F) due to oxidation and  vaporization of the metals, resulting  in
greatly reduced throughput ability as well as reduced catalyst  lifetime.
Operation of  precious metal catalysts below  1587K temperature  appears  feasible.
                                    1-4

-------
       Simple metal oxide catalysts of NiO and Co90o were tested and found
                                                 £ O
to operate successfully at temperatures to 1978K without noticeably alter-
ing catalyst activity under lean combustion conditions.  Sooting of oxide
catalysts under rich conditions does affect catalyst activity, however.
       Regardless of pretest BET surface area on both precious metal and
metal oxide catalysts, post-test BET areas are always less than 0.50 m2/g.
While this change in surface area does alter lightoff conditions, very
high steady state maximum throughputs have been achieved with catalysts
having almost no BET surface area.  For precious metal catalysts operating
at 1589K  (2400°F), volumetric heat release rates of 2.76 x 106 J/hr-Pa-m3
          /:              o
(7.5 x 10  Btu/hr-atm-ft ) are typical.  For oxide catalysts operating at
1700K (2600°F), volumetric heat release rates in excess of 7.0 x 106 J/hr-
    3
Pa-m  were achieved.
       Catalytic combustors appear to be effective in the control of thermal
NOX emissions by minimizing the gas phase reactions that occur.  Emission's
of less than 5 ppm at zero percent excess oxygen and 1587K temperature are
typical.  Catalysts are also effective in controlling the conversion of a
simple fuel nitrogen compound (NH3) to NH3, HCN, and NOX species at some
fuel/air  ratios, always under fuel-rich conditions.  Conversions of less than
20 percent of the fuel-nitrogen to NOV under rich conditions were measured
                                     A
for the metal oxide catalysts.  Control of fuel NO  under lean conditions
                                                  A
appears to be difficult for single stage combustion for either low or high
pressure  systems.
       Parametric combustion tests indicate that the maximum throughput that
can be obtained with the graded cell catalyst is a linear function of pres-
sure and  an exponential function of preheat temperature.  This implies that,
for a given fuel and preheat, the catalyst is limited by the velocity of
incoming  reactants.

1.2    SYSTEM CONFIGURATION TEST RESULTS
       The results for system configuration testing include information
obtained  from the stationary combustion system characterization study, com-
bustion system configuration testing, and conceptual design of prototype
systems.  These results are summarized here.
                                    1-5

-------
1.2.1  Characterization of Stationary Combustion Systems
       The gas turbine combustor appears well suited to catalytic combustor
redesign/retrofit because it operates with considerable excess air and uses
clean gaseous or light distillate fuels.
       Warm air gas- or oil-fired furnaces are also candidates for catalytic
combustor redesign/retrofit because of their use of clean fuels.  However,
since they are not maintained closely, early application of catalytic com-
bustors to these systems may be difficult.
       Scotch firetube boilers appear to offer advantages for a catalytic
combustor retrofit due to the unique internal, first pass furnace volume.
Watertube boilers, however, appear to require a system redesign incorporating
the catalytic combustor with a compact heat exchanger.  Package boiler
applications also require demonstration with the more common heavy oil fuels.

1.2.2  Combustion System Configuration Tests
       Three combustion system concepts were tested to evaluate heat removal
techniques for catalytic combustors.  The radiative catalyst/watertube con-
cept showed that direct heat removal from a catalyst surface by radiation to
a watertube heat exchanger is a viable concept for system temperature con-
trol.  It is possible that different catalysts (precious metals and metal
oxides) can be used in the system by varying the amount of heat removal from
the catalyst surface and hence the operating temperature.  Stable operation
of the system over a wide range of flowrates and preheat temperatures was
possible with less than 2 ppm of thermal NO  emissions.  Tests with ammonia-
                                           A
doped natural gas indicated a potential for fuel NOY control under rich
                                                   /\
conditions.
       A two-stage catalytic combustor, incorporating a fuel-rich first-stage,
interstage heat exchanger, secondary air injector, and second-stage catalyst
was also tested.  The first-stage catalyst runs at a predetermined stoichi-
ometry (50 to 80 percent theoretical air) to minimize NO  precursor species.
                                                        }\
The second stage completes the combustion with minimum emissions.  Overall
control of fuel-nitrogen species shows conversion rates below 30 percent.
The staged combustor appears applicable to both boiler and gas turbine systems,
                                     1-6

-------
       The third system concept, a model single stage gas turbine combustor,
showed less than 5 ppm thermal NOV emissions with clean gaseous and distillate
                                 A
oil fuels.  The combustor exhaust temperature is adequately controlled by ex-
cess air levels.  Premixed, lean burning fuel injection systems are a major
development area for the catalytic gas turbine combustor.  Single step lean
combustion of ammonia- and pyridine-doped fuels showed conversions of fuel-
nitrogen to NO  of approximately 80 percent.
              A
1.2.3  Prototype System Design Concepts
       Based on the performance of the three system concepts, firetube boiler
and stationary gas turbine applications appear promising for first generation
catalytic combustor retrofit.  Considerable system redesign will probably be
required for applications involving watertube boilers, residential heaters,
and mobile gas turbines.
       The two stage combustor appears promising for all applications involv-
ing nitrogen-containing fuels due to the ability of this combustor to control
the conversion of fuel-bound nitrogen to nitrogen oxides.  Since fuel-nitro-
gen is most prominent in residual and solvent refined oils and coal, staged
combustor demonstration must be extended beyond clean fuels.  Additional
development of fuel and air injection systems, premixing systems, and pre-
vaporizing systems is required for catalytic applications with these fuels.
Ignition system development is also required.

1.3    CONCLUSIONS
       The results of this program have brought catalytic combustion closer  to
concept demonstration.  Before this step can be taken, however, further work
on the integration of the catalytic combustor into the total system must be
undertaken.  At the same time, the search for additional catalyst materials
capable of high temperature operation must be continued.  Therefore, the
following program is recommended:
       •   Screening tests of various oxide and mixed oxide catalyst/support
           combinations at temperatures above 1644K (2500°F) to determine
           combustion performance over a range of fuel/air ratios.
                                      1-7

-------
Further evaluation of high-performance oxide systems with nitrogen-
doped fuels and at pressure, and determination of the operating
limits and fuel-nitrogen performance of the catalysts.
Extended life testing of selected catalysts to demonstrate high
activity over long periods of time.
Additional testing of the radiative  catalyst/watertube system
and the two stage combustor over a wider range of throughputs,
pressures, and fuels.
Testing of subsystems important to the ultimate success of the
catalytic combustor, including lightoff systems, fuel/air mixing
systems, fuel vaporization systems,  catalyst bed temperature con-
trol systems, and heat exchange systems.
Testing of potential staged combustors which require no inter-
stage cooling, with both clean and nitrogen-containing fuels.
Exploratory combustion testing of heavy fuel oils.
Design and development of suitable boiler/furnace systems utiliz-
ing catalytic combu;
actual field tests.
ing catalytic combustors  to  demonstrate  low NOV  performance in
                                              A
                         1-8

-------
                                  SECTION 2
                                 INTRODUCTION

       Interest in materials capable of promoting heterogeneous combustion
reactions has continued in varying degrees ever since Sir Humphrey Davy
discovered over 150 years ago that platinum wires could promote combustion
reactions in flammable mixtures.  An identification of some of the early
work in surface combustion is given in References 2-1 and 2-2.  Since cat-
alytic combustors have excellent potential for low NOX emissions,  a number
of research programs investigating their use in gas turbine, domestic appli-
ance, and home furnace applications are currently underway.  This  section
describes these ongoing research programs in catalytic combustion, and
outlines the program purpose and goals of EPA Contract 68-02-2116.

2.1    RELATED RESEARCH PROGRAMS IN CATALYTIC COMBUSTION
       The main focus in current catalytic combustion research is  the appli-
cation to gas turbine engines.   Additional work is being conducted for
application to residential furnaces, domestic appliances, and life-support
systems.  There is also active  research in the areas of analytical modeling
and fundamental experimentation.  Each of these research areas is  described
below.

2.1.1  Gas Turbine Applications
       Research in catalytic combustion for gas turbines includes  application
to automotive, aircraft, and stationary power generation systems as well  as
laboratory scale tests.

2.1.1.1   NASA Lewis Research Center Program
       NASA has two active programs in catalytic combustion; the Air Breathing
Engines  Division of the Emission Technology Branch is pursuing the application
                                      2-1

-------
of catalytic combustors to aircraft gas turbines, while the Power Generation
and Storage Division of the Combustion Power Section is interested in cat-
alytic combustors for automotive gas turbines.

Automotive Gas Turbine Program
       There are three major research areas in the automotive gas turbine
program:  1) fuel preparation system studies, 2) catalyst evaluation tests,
3) catalyst life tests.  The proposed gas turbine operating conditions are:
       •   Pressure:  1.5 to 4.5 atm
       e   Inlet temperature:  1210 to 970K
       •   Exit temperature:  131 OK
       •   Primary zone temperature:  1350 to 1425K
       t   Reference velocity:  11.4 to 12.9 m/sec
       •   Airflow:  0.1 to 0.5 kg/sec
       The goals of this program are to limit emissions from the combustor
to half of those required by the most stringent emission standards,  and to
keep the total combustor system (fuel preparation plus combustion chamber)
pressure drop under 3 percent.
       For the fuel preparation system, the program goals are:
       0   Spatial fuel distribution within 10 percent of mean
       •   90 percent of fuel vaporized at 800K
       •   Velocity distribution within 10 percent of mean
       •   No autoignition
       •   Less than 1 percent pressure drop
       Four different fuel injectors were tested; (1) air assist sonicore
injector, (2) splash-groove injector, (3) multiple-jet injector, and  (4)
multiple conical tube injector.  Air swirlers were used with the sonicore
and splash-groove injectors to improve spatial fuel/air distributions,.  The
multiple conical tube fuel injector was generally able to meet the program
                                     2-2

-------
goals if sufficient mixing length was allowed.   This system is shown in
Figure 2-1; Reference 2-3 describes the fuel  injection concepts in more
detail.
       In the catalyst evaluation program, the  objectives are the identifi-
cation of a catalytic combustor capable of:
       •   Emissions
           —  1.6 g N02/kg fuel
           —  13.6 g CO/kg fuel
           —  1.64 g HC/kg fuel
       •   Pressure drop
           --  Less than 2 percent
The catalyst evaluation program has two elements;  furnace screening tests
of monoliths and pellets of small size by passing  500 ppm of propane in air
over the catalyst, and combustion tests of monoliths (0.12 m in diameter)
with  800K inlet temperature and premixed propane, diesel, and Jet A fuels
in air at equivalence ratios between 0.1  and  0.3.   Results from the furnace
screening tests are presented as oxidized fuel  fraction versus catalyst
temperature and serve to identify the most suitable catalysts for further
testing.  The furnace tests have indicated that the most effective catalysts
for gas turbine combustor applications will probably be noble metals on
monoliths.  Results of the furnace screening  tests are described in Ref-
erence 2-4.
       The combustion test rig has a long mixing section, and the reactor
can hold from one to six individual catalyst  elements in series.  Each
element is located between thermocouple arrays.  All tests were performed
at an inlet temperature of  800K,  pressure of 3 x 105 Pa (3 atm), and a
range of velocities from 10 to 25 m/sec.   Adiabatic reaction temperatures
from 1100 to 1600 K were obtained by varying  fuel/air ratio.
       Catalyst elements obtained from Engelhard Industries, W. R. Grace
and Co., Johnson Matthey Corporation, and Oxy-Catalyst, Inc., have been
tested.  All elements were 0.12 m in diameter.   The Johnson Matthey elements
                                     2-3

-------
ho
i
                                                         - FUEL TUBE
                                                                                             t

                                                                                            FUEL
                                    Figure 2-1.   NASA multiple conical  tube  fuel  injector.

-------
used a metal substrate; all other elements used a ceramic substrate.  Emis-
sions measurements were made in combustion tests to determine the minimum
exit temperature at which the reactors should be operated to obtain the
steady-state emission objectives.  Effects of reactor length, cell  density,
and gas phase reactions were considered.
       The  feasibility of using a catalytic reactor in an automotive gas
turbine engine and meeting both emissions and pressure drop goals was dem-
onstrated in this program.  Potential problems for such a system include
the loss of catalytic activity with time and transient operation character-
istics.  This work is described in further detail in References 2-5 through
2-7.
       Concurrent with the fuel preparation and catalyst evaluation studies,
Engelhard Industries has conducted durability tests of two proprietary
Engelhard catalysts at one atmosphere pressure.  The test sequence for the
two catalysts included a 24-hour break-in period, a CO activity test, a test
with propane to determine the performance range, another CO activity test,
a 1000-hour life test with No. 2 diesel fuel (with CO activity tests every
250 hours), and a final propane test to determine performance range changes.
       A summary of the test results showed emissions to be the same for
both catalysts during the 1000-hour test with No. 2 diesel fuel.  One cat-
alyst required a higher inlet temperature to maintain low emissions after
600 hours of testing.  The propane parametric studies showed this catalyst
had deactivated completely for high efficiency combustion, and the CO
activity test showed significant deactivation of both catalysts between
24 and 250  hours of aging.
       Emissions with No. 2 diesel fuel after the 1000-hour test were 4 ppm
HC,  50 ppm CO,  and 4 ppm NO .   These emissions are well below the 1977 and
                           /\
1978 automotive standards.  Details of the durability testing are given in
Reference 2-8.   Further work to perform the testing at 5-atmosphere pres-
sure is now underway.
                                     2-5

-------
Aircraft Gas Turbine Program
       In addition to the automotive work, NASA is pursuing aircraft gas
turbine engine emission reduction programs using catalytic combustors.  Two
programs are currently active; the aircraft gas turbine engine Low-Power
Emission Reduction (LOPER) technology program, and the Advanced Low Emis-
sions catalytic combustor program.
       The LOPER experimental program, being conducted by General  Electric
Co., is to evolve jet aircraft engine combustion technology and reduce low-
power CO and HC emissions to extremely low levels.  Three design concepts
will be screened over a limited range of operating conditions.  The three
designs include a hot wall combustor, a recuperative combustor, and a cat-
alytic combustor.  The catalyst system was prepared for GE by Engelhard
Industries.
       The Advanced Low Emissions catalytic combustor program is being con-
ducted by General Electric Co., and Pratt and Whitney Aircraft.  The purpose
of this program is to evaluate the feasibility of employing catalytic com-
bustion technology in aircraft gas turbine engines to achieve the control
of NOX emissions for subsonic, stratospheric cruise operation while retaining
or improving system performance.  This program is jointly sponsored by NASA
and the U.S. Air Force.

2.1.1.2  Air Force Aero Propulsion Laboratory Program
       The AFAPL has been involved in both in-house and contractual programs
in the application of catalytic combustors to aircraft systems.  The in-house
program addresses the use of catalysts in both mainburner and afterburner
applications.  For the afterburner test program, a honeycomb flameholder
consisting of silicon nitride monolith segments coated with Pt/Pd catalyst
has been fabricated.  Test plans are for an inlet temperature of 975K, with
the surface temperature between 1640 and 1920K.
       Contractual work has focused on the development of a hybrid catalytic
combustor and on a fuel preparation system.  Exxon Research and Engineering
Company performed the work on the hybrid system for aircraft turbine appli-
cations, as described in Reference 2-9.  The hybrid catalytic combustor
minimizes pollution problems associated with unburned hydrocarbons and

                                     2-6

-------
carbon monoxide in the idle mode, and NOX and smoke production in the power
mode of aircraft gas turbine operation.  This combustor consists of a fuel
rich thermal precombustor, secondary air quenching zone, and monolithic
catalyst stage which rapidly oxidizes CO and UHC produced in the precombustor.
       Noble metal catalysts on various monolithic support materials and
geometries were found to be the most active materials for CO and UHC oxida-
tion in the temperature range of 700 to 1200K.  The hybrid catalytic com-
bustor combustion efficiency for JP-4 fuel containing 535 ppm sulfur was
found to be 99.8 percent under realistic conditions.   Combustor pressure drop
was less than 6 percent.  For a Johnson Matthey metal-supported Pt catalyst,
average emission indices of CO, UHC, and NOX were 0.95, 0.43, and 1.8 g/kg
of fuel, respectively.  This catalyst was effective in reducing CO by 86
percent and UHC by 94 percent, while increasing NOX by 68 percent relative  to
catalyst inlet values.  It was estimated that the hybrid catalytic combustor
can meet the 1979 new aircraft emission standards, but must be modified
slightly to reduce UHC emissions to meet the 1981 new aircraft emission
standards.
       General Applied Science Laboratories, Inc. conducted the development
work on a fuel preparation system for the catalytic combustor, as described
in Reference 2-10.  Objectives for the fuel system were to provide uniform
velocity and fuel distribution profiles, complete vaporization of the fuel,
and reasonable pressure drop.  Operating conditions for the system include
flow velocities of 15.2 to 38.1 m/sec (50 to 125 ft/sec), pressures of 6.8  to
13.6 atm, temperatures of 645 to 810K, and fuel/air ratios of 0.018 to 0.028.
The approach taken in developing the fuel preparation system was to design
for a limited residence time to prevent autoignition, to provide for adequate
exit blockage to prevent flashback, to produce small  droplets to get good
evaporation, to use a high flow velocity to preclude flame stabilization
prior to entering the combustor, and to obtain the best possible initial dis-
persion of the fuel to enhance mixing.  Three candidate, designs were tested,
consisting of pressure atomization, air blast atomization, and air assist
atomization systems.  Following testing of these systems and some modifica-
tions, an air assist atomization system with an upstream swirl generator was
selected as the final design.
                                      2-7

-------
2.1.1.3  Westinghouse Electric Company-Engelhard Industries Joint Program
       Beginning in 1973, Westinghouse and Engelhard have conducted a joint
program to assess the applicability of catalytic combustors to gas turbines.
As described in Reference 2-11, experimental test results were obtained for
No. 2 distillate oil and low Btu gas.  Pressure, temperature, and mass flow-
rate were varied during the tests.  The catalyst bed exit temperature profile
was very uniform for low Btu gas, but not as uniform for No. 2 oil.  Excep-
tionally low emissions (2-3 ppm NOX, 20-30 ppm CO) were achieved for both
fuels, and unburned hydrocarbons were less than 1 ppm.

2.1.1.4  Johnson Matthey Research Centre Program
       The Johnson Matthey Research Centre in Reading, England, has begun
laboratory scale testing of metal monolith catalyst systems for gas turbine
application.  Results of this testing are not yet available.

2.1.2  Residential Furnace Applications
       A prototype surface combustion residential furnace has been developed
by R. S. Bratko of Slyman Manufacturing Corporation in Parma, Ohio.  The
furnace has been evaluated by the Combustion Research Branch of the Industrial
Environmental Research Laboratory of the U.S. Environmental Protection Agency
as reported in Reference 2-12.  The furnace passes premixed fuel and air
through a refractory matrix, and the premixed gases then burn on the refrac-
tory surface.  Heat is transferred from the surface to the air-cooled firebox
wall by radiation.  Furnace emissions were evaluated over a range of excess
air from 5 to 45 percent with both propane and natural gas fuels.  For a
nominal operating point on natural gas at 10 percent excess air, NO  emissions
                                                                   A
were less than 15 ppm as measured, with correspondingly low CO and HC emis-
sions.  The low NOX emissions are a result of the low surface temperature
(1255K maximum) of the refractory, and the  low excess air capability gives
potential  for high system efficiency.

2.1.3  Domestic Appliance Applications
       The Institute of Gas Technology has been conducting research and  de-
velopment on catalytic combustors for domestic appliance applications  for
several  years.   These programs include:

                                     2-8

-------
       •  Contractor to Southern California Gas Company on conceptual  burners
          and model ventless appliances, for catalytic combustion of hydrogen
          and steam-reformed natural gas (1972-1976),
       •  Contractor on a joint U.S. Environmental  Protection Agency/Southern
          California Gas Company contract to develop a catalytic range-top
          burner (1975),
       •  Contractor since 1974 on a catalytic ignition system for gas
          appliances using hydrogen as an ignition  fuel ignited in air on a
          platinum catalyst.
       The instantaneous ignition system stores hydrogen in the form of a
metal hydride.  When ignition is required, small  quantities of hydrogen are
released by valving.  This catalytic ignition system could become an alter-
nate to standing pilot or electric systems.
       The model catalytic appliances operate at  low combustion temperatures
and produce very low emissions of nitrogen oxides.   Because of the low
emission levels, outside venting of the products  of combustion is not  re-
quired.

2.1.4  Life Support Systems
       Energy Systems Corporation in Nashua, New Hampshire has developed
various thermal protection equipment using catalytic combustors.  Existing
products include belt-mounted Arctic ambulatory heater systems, SCUBA  diver
heaters, hypothermia prevention and treatment systems, casualty evacuation
bag heaters, and downed airman power sources.
       The downed airman power source system supplies warmth to airmen in
life rafts by circulating heated water through turbulated undergarments or
blankets.  These systems use either propane or propylene fuel combusted on
1-percent Pt on alumina pellets manufactured by Matthey Bishop.  Catalyst
bed temperatures are between 922 to 1033K.  Heat  extraction pins or fins
conduct heat to a hot plate and finally to the fluid heat exchanger.
       ESC is also currently developing a catalytic/thermoelectric SCUBA
diver heater for the U.S. Navy.  This system will be capable of delivering
500 thermal  watts to a diver in 275K seawater.
                                     2-9

-------
2.1.5  Fundamental Programs
       Several ongoing research programs are currently developing information
which would have application to a broad class of catalytic combustion systems.
These programs include both analytical  and experimental  projects.

2.1.5.1  Analytical Programs
       Analytical models for application to high temperature catalytic com-
bustors are in development at Exxon Research and Engineering Company and at
Stanford Research Institute.  Each of these programs is  described below.

Exxon Research and Engineering Company Model
       Under joint National Science Foundation/U.S.  Environmental Protection
Agency/Exxon sponsorship, Exxon has developed a practical  model  of catalytic
combustor operating characteristics for analysis of advanced power systems
and test data.  The model is described in detail in Reference 2-13.  The
major physical assumptions of the steady-state model are:
       •  Uniform gas phase properties at a cross section
       t  Uniform catalyst/substrate temperature and fuel  concentration at
          a cross section
       •  Conversion of reactants to products both at the  catalyst surface
          and in the gas phase
       •  Axial variations in velocity are allowed
       •  Axial heat conduction in the substrate is included, but radiation
          and gas phase conduction are neglected
       The model consists of a variable-order finite-difference method to
solve the two-point boundary value problem.  Comparisons of catalytic com-
bustor data and model predictions show excellent agreement.  When the model
is used to examine conversion as a function of gas  inlet temperature for
typical lean operating conditions with a noble metal catalyst, a sharp cata-
lytic lightoff temperature is predicted, as is a 1ightoff/extinction hysteresis
At higher temperatures the onset of gas phase combustion occurs, resulting
in complete conversion.
                                      2-10

-------
       Further work on expanding the model to include internal  heat removal,
CO kinetics, multiple fuel species, and NOY kinetics is planned.
                                          A
Stanford Research Institute Model
       Under Air Force Office of Scientific Research funding,  Stanford
Research Institute has initiated an analytical  study to determine the contri-
bution of catalytic wall  reactions to combustion initiation.   The temperature
distribution on the duct wall is found, taking  into account wall  heat conduc-
tion, convective heat transfer, and heat generation and fuel  consumption.
For Lewis numbers greater than one, the temperature increases  with distance
down the duct, while for smaller Lewis numbers  the temperature passes through
a maximum whose value depends on flow speed.  Ultimately,  this work will  be
used to predict the distribution of catalyst on a substrate required for
startup and shutdown in a practical system.  Work to date  is  described in
Reference 2-14; experimental evaluations will also be conducted in this
program.

2.1.5.2  Experimental Programs
       Laboratory scale experimental programs are currently in progress at
United Technologies Research Center, Princeton  University, Lawrence Berkeley
Laboratory, and the University of California.  These programs  are discussed
below.

United Technologies Research Center Program
       The UTRC program in catalytic combustion seeks to determine the feasi-
bility of utilizing catalyzed surface reactors  in the combustion  of multi-
component fuels.  As described in Reference 2-15, the program is  currently
involved in simple burner experiments using propane fuel and  in an analytical
study which combines heat and mass transfer and homogeneous reactions in  the
analysis.
       The experimental program uses a catalytic burner apparatus to deter-
mine ignition temperature, steady-state operating conditions, and species
concentrations by varying mixture ratio, temperature, bed length, flow
velocity, bed material, and diluent.  The combustible mixture is  usually
                                     2-11

-------
oxygen-propane with argon diluent, but helium and nitrogen diluent have also
been used.  Isothermal experiments are then run to determine ignition tem-
perature.
       The modeling program takes into account both heterogeneous and homo-
geneous reactions, and seeks to fit experimental  data  over a range of tem-
peratures with single heterogeneous and homogeneous rate constants.   Matching
of temperature rise and concentration data has been done for propane combus-
tion with argon diluent.

Princeton University Program
       The Department of Aerospace and Mechanical Sciences at Princeton
University has undertaken a program to clarify the relative influences of
chemical kinetics and transport processes in a catalytic combustion system.
This work is sponsored by the Air Force Office of Scientific Research.  A
steady combustion system has been constructed and some preliminary data ob-
tained.  Measurements to be taken include velocity, temperature,  pressure,
and gas composition within a honeycomb catalyst system and its boundary
layer using both physical and optical techniques.

University of California/Berkeley Program
       The Department of Mechanical Engineering at the University of California/
Berkeley recently completed a study on fuel-nitrogen conversion with catalytic
combustors for the Department of Energy (References 2-16, 2-17).   A platinum
catalyst was operated with propane/oxygen/argon mixtures at equivalence ratios
between 0.7 and 1.6.  Trace amounts of either nitric oxide or ammonia were
added to the gases, and the conversion of these fuel-nitrogen compounds to
nitric oxide was measured as a function of equivalence ratio, adiabatic flame
temperature, and fuel-nitrogen concentration.  It was concluded that surface
reactions dominate the fuel-nitrogen conversion mechanism, with the conversion
found to be strongly dependent on equivalence ratio, weakly dependent on
calculated adiabatic flame temperature, and moderately dependent  on fuel-
nitrogen concentration.
                                     2-12

-------
Lawrence Berkeley Laboratory Program
       Under Department of Energy support, the Lawrence Berkeley Laboratory
has been investigating the effect of a heated catalytic surface on  combustion
in lean hydrogen-air mixtures (References 2-18, 2-19).   The  velocity and
density profiles of the boundary layer have been measured with laser Doppler
velocimetry and Rayleigh scattering, respectively.   Measurements on a plati-
num catalytic surface indicate that, at a surface temperature of 1000K, not
only is there significant surface combustion but that homogeneous combustion
in the boundary layer is induced by active species  generated at the catalytic
surface.  An analytical model has also been developed to aid the investiga-
tion.  This work will help improve the understanding of high temperature
heterogeneous catalysis of combustion reactions and the coupling with homo-
geneous reactions and fluid mechanics.

2.2  PROGRAM PURPOSE AND GOALS
       The objective of the research and development program described in
detail in the remainder of this report is to establish  design criteria for
the application of catalytic combustion to low emission, high efficiency
stationary combustion systems.  This objective was  met  by conducting a two
phase program which included the following tasks:
       t  Phase I — Small Scale Catalyst and Combustion Concept Screening
                  -- Review the available information on stationary combus-
                     tion system design and operating characteristics, in-
                     cluding residential heating systems, commercial and
                     industrial boilers, stationary gas turbines, and super-
                     charged boilers.  Consider system impacts as to the
                     applicability of catalytic combustor retrofit  or
                     redesign.
                  -- Review available catalyst materials, including substrates,
                     washcoats, and catalysts, and  select materials for test-
                     ing at temperatures between 1360 and 1980K»
                                     2-13

-------
            — Determine catalyst preparation and characterization
               techniques.   Catalysts prepared include both  precious
               and base metal  catalysts.   Catalyst characterization
               parameters include total  surface area,  selected  surface
               area (dispersion)  of precious  metal  catalysts, visual
               surface appearance by scanning electron microscopy,
               and surface  composition by energy-dispersive  x-ray
               analysis.
            -- Develop the  basic  understanding of the  catalytic com-
               bustor, including  the effects  of preheat temperature,
               catalyst system material,  and  catalyst  system geometry.
               Use this information to develop a computer  code  capable
               of giving quantitative information on catalyst perform-
               ance as a function of operating parameters.
            ~ Perform catalyst screening tests of at  least  30  cata-
               lyst systems with  up to four different  fuels  (natural
               gas, propane, methanol , distillate oil).  Obtain radial
               and axial temperature profiles within the catalyst bed.
               Test these catalysts at temperatures to 1980K and
               pressures to 1.01  MPa (10  atmospheres).
            -- Perform a more  extensive evaluation of  those  catalysts
               exhibiting good combustion characteristics.   Use
               ammonia dopant  to  simulate fuel-nitrogen compounds and
               measure the  ammonia conversion to nitrogen  oxide.
                                                        o
            -- Build and test  two small  scale (1.05 x  10  joules/hr)
               combustion systems which would have application  to
               practical combustors.  These systems are to include
               FGR, two-stage  combustion, or  direct bed heat removal
               for temperature control.
•  Phase II -- Scale-up of  Catalyst and System Concepts
            — Perform catalyst scale-up tests at a nominal  heat re-
                                      q
               lease rate of 1.05 x 10  joules/hr. Test results were
               compared to  the small scale tests to determine scaling
               parameters

                               2-14

-------
                    -- Perform scale-up testing of one combustion system
                       and compare to small-scale system tests.
                    -- Develop conceptual designs for the application of
                       catalytic combustors to practical combustion systems,
       Sections 3-11 of this report describe the tasks outlined above in
detail.
                                   2-15

-------
                                 REFERENCES


 2-1.    Kesselring,  J.  P.,  et al.,  "Catalytic Oxidation of Fuels for NOX
        Control  from Area Sources," Environmental Protection Technology Series
        Report EPA-600/2-76-037,  February 1976.

 2-2.    Pfefferle, W.  C., "The  Catalytic Combustor — A Look Back, A Look
        Forward,"  Paper 77-31,  presented at the Western States Section/The
        Combustion Institute Meeting, October 17-18, 1977, Stanford, California.

 2-3    Tacina,  R.,  "Experimental Evaluation of Two Premixing-Prevaporizing
        Fuel  Injection  Concepts  for a Gas Turbine Combustor," NASA TM X-73422,
        May 1976.

 2-4    Anderson,  D.  N.,  "Preliminary Results from Screening Tests of Commercial
        Catalysts with  Potential  Use in Gas Turbine Combustors, Part I.  Furnace
        Studies  of Catalyst Activity," NASA TM X-73410, May 1976.

 2-5    Anderson,  D.  N.,  "Preliminary Results from Screening Tests of Commercial
        Catalysts with  Potential  Use in Gas Turbine Combustors, Part II.  Com-
        bustion  Test Rig Evaluation," NASA TM X-73412, May 1976.

 2-6    Anderson, D.  N., et al.,  "Catalytic Combustion for the Automotive Gas
        Turbine  Engine," NASA TM  X-73589, April 1977.

 2-7.    Anderson, D.  N., "Performance and Emissions of a Catalytic Reactor
        with  Propane,  Diesel, and Jet A Fuels," NASA TM-73786, October 1977.

 2-8.    Heck,  R. M.,  et al., "Durability Testing at One Atmosphere of Advanced
        Catalysts and Catalyst Supports for Automotive Gas Turbine Engine Com-
        bustors," NASA  CR-135132, June 1977.

 2-9.    Siminski, V.   J., and Shaw, H., "Development of a Catalytic Combustor
        for Aircraft  Gas Turbine  Engines," Technical Report AFAPL-TR-76-80,
        September 1976.

 2-10.   Roffe, G., "Development of a Catalytic Combustor Fuel/Air Carburetion
        System," Technical  Report AFAPL-TR-77-19, March 1977.

 2-11.   De Corso, S.  M., et al.,  "Catalysts for Gas Turbine Combustors-
        Experimental  Test Results," ASME Paper 76-GT-4, March 1976.

2-12.   Martin, G. B.,  "Evaluation of a Prototype Surface Combustion Furnace,"
        published in  "Proceedings of the Second Stationary Source Combustion
        Symposium, Volume III," EPA-600/7-77-073c, July 1977.

2-13.    Cerkanowicz,  A. E., et al., "Catalytic Combustion Modeling:   Comparisons
       with Experimental Data,"  ASME Paper 77-GT-85, March 1977.
                                      2-16

-------
2-14.  Ablow, C. M. and Wise, H., "Contribution of Catalytic Wall  Reaction
       to Combustion Initiation," Paper 77-39 presented at the Western States
       Section/The Combustion Institute Meeting, Stanford, California, October
       17-18, 1977.

2-15.  Marteney, P. J., and Kesten, A.  S., "Studies of the Rate of Oxidation
       of Propane on a Catalytic Surface," Paper 77-36 presented at the
       Western States Section/The Combustion Institute Meeting, Stanford,
       California, October 17-18, 1977.

2-16.  Matthews, R. D., "The Nature and Formation of Nitrogenous Air Pollu-
       tant Emissions from Combustion Systems," University of California/
       Berkeley, Energy and Environment Division Report LBL-6850,  October
       1977.

2-17.  Matthews, R. D., and Sawyer, R.  F., "Fuel Nitrogen Conversion and
       Catalytic Combustion," University of California/Berkeley, Energy and
       Environment Division Report LBL-6396, October, 1977.

2-18.  Robben, F., et al., "Catalyzed Combustion in a Flat Plate Boundary
       Layer.  I.  Experimental Measurements and Comparison with Numerical
       Calculations," Lawrence'-Berkeley Laboratory, Energy and Environment
       Division Report LBL-6841, September 1977.

2-19.  Schefer, R. and Robben, F., "Catalyzed Combustion in a Flat Plate
       Boundary Layer.  II.  Numerical  Calculations," Lawrence Berkeley
       Laboratory, Energy and Environment Division Report LBL-6842, September
       1977.
                                     2-17

-------
                                  SECTION 3
               CHARACTERIZATION OF STATIONARY COMBUSTION SYSTEMS

       Background  information on stationary combustion systems which is re-
quired for future  catalytic combustor retrofit or redesign is presented in
this section.  Discussion includes:
       t   General considerations in which the nature of area source sectors
           is  described
       0   A description of the design and operating characteristics of gas
           turbines and supercharged boilers
       •   Air pollutant emission characteristics
       •   Conclusions

3.1    GENERAL CONSIDERATIONS
       Emissions of nitrogen oxides from stationary source combustion account
for 94 percent of  the nation's stationary source total of 1.14x10   g/year
(12.6 million  tons/yr.).  Developing technology to ultimately produce a re-
duction in these NOV emissions is the objective of the present program.
                   A
       The options available for controlling these NOX emissions include fuel
modification,  flue gas treatment, modification of combustor operating condi-
tions, or use of alternate processes.  The first three methods have, in the
recent past, been vigorously investigated, and one of these control tech-
niques, combustion modification, has proven successful for the more prolific
NO  emitters.  In  order to have the technology available to further reduce
  A
NO  emissions from sources as well  as to bring those sources which do not
  X
respond to the simpler reduction options into compliance with future emis-
sion regulations, alternate processes must be explored.  One of these, cata-
lytic combustion, is the focus of this report.

                                     3-1

-------
       Catalytic combustion offers significant NO ,  CO,  and unburned
                                                 A
hydrocarbon emission reduction potential  due to a low operating temperature
and a concurrent promotion of oxidation reactions.  An early assessment of
the applicability of catalytic concepts to gas turbines  and utility boilers
was performed by the Aerospace Corporation in 1973 (Reference 3-1).  This
report concluded that catalytic oxidation might be applicable to gas turbines,
but that application to a utility boiler  would require a new system design.
The report also indicated that only gaseous fuels and light, sulfur-free
hydrocarbons could be used in catalytic systems due  to system requirements
and catalyst poisoning difficulties.
       Acurex extended the applicability  assessment of catalytic com-
bustors to gas- and oil-fired home heaters and commercial  and industrial
boilers in 1976 (Reference 3-2).  These area-source  combusters use gaseous
or light distillate oil fuels, and may be amenable to redesign.  It was
concluded that there are no theoretical barriers to  the  application of
catalytic combustors in any of these area sources on either a retrofit or
redesign basis.

3.2    EQUIPMENT AND OPERATING CHARACTERISTICS
       A detailed discussion of residential heating  equipment and commercial
and industrial boilers is given in Reference 3-2.  This  section will describe
equipment and operating characteristics for gas turbines and supercharged
boilers.

3.2.1  Stationary Gas Turbines
3.2.1.1  Introduction
       The gas turbine is a rotary internal combustion engine based on the
Brayton cycle.  All gas turbines use these same equipment components:
       t   Compressor — pressurizes and  forces primary air into the combus-
           tion chamber,
       •   Combustor -- combusts the injected fuel and primary air and allows
           for injection of secondary air to complete the combustion reaction,
                                    3-2

-------
       •   Turbine -- transforms the enthalpy of the gas into mechanical
           energy by expanding the gas through the turbine blades to drive
           the compressor and do useful shaft work.
       Table 3-1 provides a breakdown of the many types of stationary gas
turbine equipment. | The first column shows the initial  design either as a
derivative of an existing aircraft unit or exclusively for stationary use.
The remaining columns break down the major equipment components as listed
previously.  Each of these equipment subcategories is described below.
                    TABLE 3-1.  TYPES OF STATIONARY GAS
                                TURBINE EQUIPMENT
Design
Stationary
design
Aircraft
derivative

Compressor
Axial flow
Axial and
Centrifugal
flow
Centrifugal
flow
Combustor
Annular
Can
Canannular

Turbine
Axial flow
Radial inflow

       In addition to the available component combinations, several  cycle
options exist.  The most important are:
       t   Simple cycle -- Air and fuel are burned in the combustor, and the
           hot combustion products are expanded through the power turbine
           and exhausted to the atmosphere.  Figure 3-1 is a schematic of
           this cycle.
       •   Regenerative cycle — The combustion products are directed through
           a heat exchanger to preheat the primary combustion air.
       t   Combined cycle — The hot exhaust gases of the simple cycle turbine
           pass through a waste heat boiler to increase the thermal  efficiency
           of the unit.
       Several subclassifications of these cycles exist.  Intercooling is a
modification of the simple cycle which uses two compressors separated by a
                                     3-3

-------
                              Comburtor
               jtPDtRl
                                   Shaft
           ComprtMor
Turbine
                                                      Exhaust
                                                        t
                  Load
                 Coupling
                Figure 3-1.   Simple cycle gas  turbine system.

heat exchanger which lowers  the second stage compressor inlet  temperature,
decreasing its energy requirements and thus increasing thermal  efficiency.
Reheating is another modification of the simple cycle wherein  two combustors
and turbines are used to raise the average temperature of heat addition,
also increasing thermal efficiency.
       The stationary gas turbine serves in a  variety of applications,  the
most common of which are electricity generation and  pumping.   Table  3-2 gives
the principal applications of turbines in the  United States.   Also shown  are
the typical locations of these installations,  which  is an important  factor
when emission regulations are being considered.  Table 3-3 shows  the relative
installed horsepower and associated fuel consumption for the three primary
users.  Although the electric power generation community appears  to  be  the
largest user of gas turbines, the use factor for these installations is far
lower than for the other sectors.  The majority of electric generating  units
are used for peaking power only.
       A major disadvantage  of the gas turbine is its requirement for clean
fuels such as natural gas and distillate oil.   Newer units, however, are
being designed to use low sulfur residual oil, and the number  of units fir-
ing residual oil is expected to increase.
                                    3-4

-------
TABLE 3-2.  PRINCIPAL APPLICATIONS, Of STATIONARY
            GAS TURBINES IN THE UNITED STATES
Industry
Electric utility
Natural gas
utility
Petroleum
Chemical
General
industrial
Commercial &
municipal
Applications
Base load
Intermediate load
Peak load
Standby power
Total energy
Compressor drives
Pumping
Compressor drives
Electric power
Compressor drives
Electric power
Electric power
Mechanical drive
Electric power
Total energy
Pumping
Typical Locations
Near populated
areas
Remote
Remote
Near populated
areas
Near populated
areas
Near populated
areas
    TABLE 3-3.  STATIONARY GAS TURBINE POWER
                GENERATION (1973 data)
Industry
Electric
power
Oil & gas
pipelines
Natural gas
processing
Rated
Capacity
Mw (103 Bhp)
22.83 (30,440)
2.64 (3,520)
1.15 (1,530)
Power
Generation
M >oule (106 Bhp-hr)
89,750 (33,240)
57,402 (21,260)
40,527 (15,010)
Fuel
Consumption
1015 J (1012 Btu)
Natural No. 2
gas oil
134 (127) 295 (280)
846 (802) 43 (41)
457 (433) -- --
                      3-5

-------
3.2.1.2  Equipment Characteristics
       The wide variation in gas turbine equipment design  is  determined  by
the various combinations of compressor,  combustor, and  turbine.   The  first
stationary units were modifications of existing  aircraft designs.   At present,
units specifically designed for stationary use are common.  Stationary de-
signers need not conform to the aircraft engine's  geometry  and weight con-
straints, although in many instances uniform compressor and turbine designs
are used in both applications.   The remainder of this subsection  will  describe
the various types of compressors, combustors, and  turbines.

Compressor Types
       The compressor is one of the most important gas  turbine components
since the energy released in the combustor is directly  proportional to the
mass of air consumed.  Basically, the compressor provides high pressure  air
to the combustion chamber.  The most efficient compressor operation provides
maximum compression with minimum temperature rise.  Two basic compressor
designs exist:
       •   Centrifugal flow
       •   Axial flow
       A centrifugal compressor consists of an air impeller,  a diffuser  mani-
fold, and a compressor manifold to direct the compressed air  to the combustor.
The primary advantages of the centrifugal compressor are its  simplicity,
ruggedness, and low manufacturing cost.   They are  capable  of  compression
ratios of about 4 or 5 to 1, with an efficiency  of about 80 percent.   Because
compression ratios above 5 to 1 reduce specific  fuel consumption, these  com-
pressors are generally ruled out for use in very large  units. However,  5  to  1
compression ratio regenerative turbines  are competitive in  thermal efficiency
with high compression ratio simple cycle turbines.
       The axial flow compressor consists of a  series of rotating airfoils
(rotor blades) and a stationary set of airfoils  (stator vanes).   Though  capable
of producing very high compression ratios, the  design has  its disadvantages.
Manufacturing is very costly, and it is  very susceptible to particulate  damage.
Axial flow designs find greatest use where requirements for high  efficiency
and output predominate considerations of cost and  flexibility.
                                     3-6

-------
Combustor Types
       The combustgr converts chemical  energy to thermal  energy through
combustion of a fuel/air mixture.  Of all  gas turbine components,  the  com-
bustor presents the most difficult design  problems since  it must withstand
the high temperatures of combustion in  addition to diluting and cooling  the
combustion products prior to their entry into the turbine.
       The following are important goals in the design of a combustor;
       t   High combustion efficiency.
       t   Stable operation, free from blowout over large operating  ranges.
       t   Low pressure loss, ensuring  that pressure drop will  be  taken
           across turbine.
       •   Uniform temperature distribution.
       •   Low carbon deposition.
       •   Long material lifetimes.
       Only about one-fourth of the compressor's air is used in the  primary
combustion process.  The remainder is used as a film coolant for the combus-
tor and as secondary dilution air.
       Three basic combustor designs have  evolved throughout the developmental
stages of the aircraft gas turbine:
       •   can
       •   annular
       •   canannular
       In an aircraft engine utilizing the can design, several  combustors are
mounted around the engine axis.  Each consists of a cylindrical shroud with
an inner liner, a fuel nozzle, and an igniter.  The ease  of maintenance, due
to the modular design, is this engine's major advantage.   Its main disad-
vantage, the poor use of available space,  is generally not important in
stationary applications.
       The annular type combustor is a single combustor composed of two  large
concentric cylinders.  Structural problems are usually a concern due to the

                                    3-7

-------
large diameter, thin wall construction that is typical  of this design.   Such
problems are magnified with increasing engine size and  diameter.   Other in-
herent difficulties include maintenance and parts replacement.  The major
advantage of this type of combustor, especially in aircraft applications, is
its effective use of available space.
       The canannular type of combustor is a combination of the can and the
annular.  It consists of inner and outer combustion casings mounted coaxially
about the engine axis, in addition to  several cylindrical  combustion cans
mounted within the annular housing. This design reduces the structural prob-
lems found in the annular design and in general experiences a lower pressure
drop than the can design.
       The stationary application has  spurred several new designs in com-
bustors, primarily because of the absence of the geometry constraints in
aircraft turbines.  Chief among these  developments has  been the vertical
combustor.  These are typically very large, modified versions of the can
design which are mounted vertically.
       In the vertical combustor, fuel is introduced by a fuel nozzle which
creates a highly atomized, accurately  shaped pattern to facilitate rapid
mixing and high combustion efficiency.  The two basic nozzle designs are
known as the simplex and duplex.  The  simplex nozzle is generally unable to
provide satisfactory spray patterns under varying conditions but is entirely
adequate for continuous loads, while the duplex nozzle  is capable of opti-
mizing spray patterns to suit the operating conditions  and is preferred for
varying load conditions.
       Burner performance is determined by the interaction of several operat-
ing variables and various design factors.  The operating variables are:
       •   pressure
       •   inlet air temperature
       t   fuel/air ratio
       •   flow velocity
                                    3-8

-------
Important design factors are:
       t   methods of air distribution
       •   physical dimensions of the combustor
       •   fuel/air operating ranges
       •   fuel nozzle design
       The interrelationships between these operating and design  variables
must be exploited in order to optimize the performance of the combustor.

Jurbine Types
       Following their formation in the combustor, the exhaust products  enter
and expand through the turbine.  The extracted energy is  used to  drive the
compressor and other turbine accessories, including generators and  pumps.
       Two turbine designs exist:
       t   axial flow
       •   radial inflow
The axial flow turbine is the most widely used of the two designs and con-
sists of one or more turbine rotors and an accompanying set of stationary
vanes.  These vanes, sometimes called the turbine nozzle, are configured so
that the gas is discharged directly onto the turbine blades.
       The radial inflow turbine is less common than the previous type,  but
smaller units are gaining in popularity.  In this design, the gas flows
through peripheral nozzles and enters the wheel passages in an inward, radial
direction.  The primary advantages of this type of turbine are its  rugged-
ness and ease of manufacture relative to the axial flow design.

3.2.1.3 .Operating Characteristics
       In simple-cycle gas turbines, the optimum pressure ratio increases
substantially with increasing turbine inlet temperature.   For instance,  for
a turbine inlet temperature of 1256K, which is close to current practice,
the optimum pressure ratio is about 18:1 and the cycle efficiency is approxi-
mately 30 percent.  At 1922K, the optimum simple cycle pressure ratio
                                   3-9

-------
increases to over 30:1.  The corresponding cycle efficiency is of the order
of 40 percent, which is comparable to the efficiency of current steam power
plants.
       Substantially higher thermal efficiencies can be achieved at much
lower cycle pressure ratios by means of regeneration.   The thermal  efficiency
of current steam power plants can be matched with regenerative gas  turbines
operating at a turbine inlet temperature of about 1367K and a regenerator
effectiveness of 90 percent.  Under these conditions,  the optimum pressure
ratio is only about 4:1.
       Predicted combustor inlet temperatures for simple-cycle gas  turbines
are presented in Figure 3-2 as a function of compressor pressure ratio.   The
data are based on a compressor efficiency of 86 percent, a turbine  efficiency
of 88 percent, and an ambient temperature of 289K.   For the region  of
interest, the combustor inlet temperature of simple-cycle gas turbines  varies
between about 672K and 81 IK.
       The combustor inlet temperatures  of regenerative-cycle gas turbines
are depicted in Figures 3-3 and 3-4 for regenerator effectiveness values of
90 percent and 70 percent, respectively.  Because of the high degree of
regeneration, the temperatures are substantially higher than those  of the
simple-cycle gas turbines.  The curves of Figures 3-3 and 3-4 are based
on the same turbomachinery efficiencies used in the simple-cycle calcula-
tions and a combined regenerator and combustor pressure loss of 4 percent
of the compressor discharge pressure.
       Predicted air/fuel  ratios for natural  gas are presented in Figure 3-5
for both simple-cycle and  regenerative cycle gas turbines.  As expected, the
required air/fuel ratio decreases with increasing turbine inlet temperature
and decreasing degree of regeneration.

3.2.2  Supercharged Boilers
       A supercharged boiler cycle refers to a system where a steam boiler is
operated in conjunction with a gas turbine in such a manner that combustion
air for the boiler comes  from a compressor driven by a gas turbine  which, in
turn, is driven by the expansion of the combustion gases leaving the boiler.
                                    3-10

-------
CO
I
                 900
                 800
                 700
               LU
               CC
               3



               £ 600

               a.
               w 500
               CC

               8
               
-------
             2000
   1300
              1800
   1200
   1100
k  looo
           u- 1600
          o
           of
           cc
             1400
           ui
    90O -
    80O -
             1200
           ffl
           o
           u
             1000
    700
    600
800
600
                                                       12.0
   NOTE: Compressor Inlet Temperature 289K (60°F)
         Compressor Efficiency 0.86
         Turbine Efficiency 0.88
         Combustor Pressure Loss Factor 0.04
         rc = Compressor Pressure Ratio
                                                               I
                2000       2200       2400        2600        2800       3000
                                    TURBINE INLET TEMPERATURE,°F
                     I	I	I	I	1	I
                   1400
               1500
1600
                                              K
1700
1800
1900
            Figure  3-3.  Regenerative-cycle  gas turbine combustor
                          inlet temperature vs  turbine inlet  temperature-
                          regenerator  effectiveness  0.90.
                                         3-12

-------
           2000
1300
           1800
1200
           1600
1100
1000
   ui
   
-------
   100
   80
2   60
oc
_l
Ui
u.
K
<
                      REGENERATOR EFFECTIVENESS 0.90
                           REGENERATOR EFFECTIVENESS 0.70
                                             NONREGENERATED
    20
          NOTE:  Compressor Efficiency 0.86
                 Turbine Efficiency 0.88
                 Combustor/Regenerator Pressure Drop Factor 0.04
                 Lower Heating Value 47.7 MJ/Kg (20.500 Btu/lb)

            I	|	|	I            I
2000      2200        2400        2600       2800        3000
             TURBINE IN LET TEMPERATURE, °F
_|	I	I	I	I	|_
       1400
            1500
1600
1700.
1800
1900
    Figure 3-5.   Predicted air/fuel ratio vs turbine inlet
                  temperature  (natural gas).
                              3-14

-------
There are two types of supercharged cycles, differing by the operation of
the gas turbine.  The first is termed the self-sustaining cycle and is similar
to the Velox system where the gas turbine drives only the compressor.   In
the second type, called the power cycle, the gas turbine generates electrical
power as well as driving the compressor.
       The self-sustaining cycle is of special interest to the marine  industry
due to its compact size.  Because of its improved cycle efficiency, the power
cycle is of particular interest in large operating stations.
       In general, the supercharged power cycle has several advantages over
conventional steam boilers.  Primary among them are:
       t   A 5- to 9-percent improvement in plant heat rate over that  of a
           conventional cycle of equivalent rating and steam conditions
       •   Greatly reduced boiler size and weight
       •   Improved load response and reduced starting time
       •   Savings in piping due to closer equipment grouping (Reference 3-3)
       The Foster Wheeler Corporation has been the leading figure in the
development and manufacture of the supercharged boiler.   Both Foster Wheeler
and the Sulzer Company in Europe have seriously investigated the adaptation
of the supercharged cycle to electrical power generation, with the latter
company having built several European pilot plants (Reference 3-4).
       As discussed previously, marine applications have been of primary
interest for the self-sustaining cycle.  Figure 3-6 shows the Foster Wheeler
unit as presently installed in several destroyer escort (DE) class naval
vessels, and Figure 3-7 shows the cycle schematically.
       A schematic of the power cycle for supercharged boilers is given in
Figure 3-8.  The plant utilizes a gas turbine which drives an electrical
generator in addition to the combustion air compressor, while the steam
turbine drives a 400 Mw generator.  Fuels for both cycles are limited  to
natural gas, fuel oil, and other clean fuels.
       The design of the supercharged boiler differs from the conventional
boiler in that it must contain the relatively high pressure combustion gases.
                                    3-15

-------
Figure 3-6.   Marine supercharged boiler.




                   3-16

-------
         Exhaust
Economizer
Compressed
Combustion
Air
  Turbine
                          Steam 0 8.27 MPa, 783K
                          (1200 piig, 950° F)
                         Compression Gases @ .276-.680 MPa, 700K
                         (40-100 piig, 800°F)
            Figure  3-7.   The self-sustaining supercharged cycle.


This  constraint generally results in a multi-walled,  cylindrical  shell
design.   Also,  supercharged boilers are smaller and lighter  than  conventional
ones, due mainly to the higher heat transfer rates.   An  increased average
furnace  temperature as well as the increased emissivity  of the combustion
products due  to higher pressure cause the higher transfer rate.   Heat trans-
fer surface areas are reduced to about one-third of conventional  boilers.
       While  little general data exist on the operating  characteristics of
supercharged  boilers, some data on specific units  is  available.   Full load
performance figures for a gas-fired self-sustaining supercharged  cycle, in
addition to selected physical data, is presented in Table 3-4.
                                    3-17

-------
         Exhaust
           i,
Economizer
Compressed
Combustion
Air
  Turbine
                                Steam 0 8.27 MPa, 783K
                                (1200p«ig,950°F)
                                 Compression Gases @ .276-.690 MPa, 700K
                                 (40-100psig.800°F)
                         Figure 3-8.  The power  supercharged  cycle.
              At the outset, it appears that the compact  heat  transfer configuration
        of this type of boiler predisposes it to the application  of catalytic com-
        bustion concepts.  No requirement seems to exist  for the  large catalytic
        surface area that earlier investigators identified  as  a necessity for utility
        boiler applications.  The feasibility of high  pressure catalytic combustion
        has been demonstrated by other programs in catalytic combustion (References
        3-6, 3-7).

        3.3   AIR POLLUTANT EMISSION CHARACTERISTICS
              Emissions data for stationary fuel burning  equipment has been the sub-
        ject of several recent studies.  The most recent  year  for which complete fuel
        consumption  data is available is 1974.  Figure  3-9  gives  the distribution of
        anthropogenic NO  emissions for 1974, showing  that  gas turbine NO  emissions
                        A                                                 X
                                          3-18

-------
          TABLE 3-4.  SUPERCHARGED BOILER FULL LOAD PERFORMANCE
                      FIGURES (Reference 3-5)
Steam Flow
Superheater Outlet Pressure
Superheater Outlet Temperature
Total Air Flow
Boiler Air Flow
Afterburner Air Flow
Air Pressure
Air Temperature
Excess Air
Total Fuel Flow
Boiler Fuel Flow
Afterburner Fuel Flow
Gas Leaving Afterburner
Gas Flow to Turbine

Net Heat to Furnace (LHV + Air)
Net Heat Release Per Cu. Ft. Furnace
  Vol ume
Net Heat Release Per Sq. Ft. Projected
  Furnace Envelope
Furnace Radiant Heat Absorption Per Sq.
  Ft. Projected Surface
126,100. Kg/hr (278,000 Ib/hr)
8.72 M Pa (1250 psig)
783 K (950 °F)
157,800 Kg/hr (348,000 Ib/hr)
154,400 Kg/hr (340,480 Ib/hr)
3410 Kg/hr (7520 Ib/hr)
0.59 M Pa (86.0 psia)
535 K (503 °F)
10%
9741 Kg/hr (21,476 Ib/hr)
9530 Kg/hr (21,014 Ib/hr)
210 Kg/hr (462 Ib/hr)
1061 K (1450 °F)
167,557 Kg/hr (369,476 Ib/hr)

4.49xlOn J/hr (425.2 x 106 Btu/hr)
7.23xl08 J/hr (685,000 Btu/hr)

l.lBxlO9 J/hr (1.09 x 106 Btu/hr)

2.24xl08 J/hr (211,900 Btu/hr)
account for 2.9 percent of the total.  Table 3-5 shows emission inventory
results for other pollutants as a function of equipment type.  The relatively
low contribution of SOX emissions from gas turbines is indicative of the fuel
types used.
      Table 3-6 ranks equipment/fuel combinations by annual, nationwide NOX
emissions and lists corresponding rankings for these combinations by fuel con-
sumption and emissions of criteria pollutants.  Although over 70 equipment/
fuel combinations were inventoried (Reference 3-8), the 30 most significant
                                   3-19

-------
Industrial Process Combustion 2.9%
         IMoncombustion 1.7%
    Warm Air Furnaces 2.8%
     Gas Turbines 2.9%
     Fugitive 4.3%
Incineration 0.3%
                  Reciprocating
                  1C Engines
                  16.2%
            1974 Stationary Combustion Source NO  Emissions
Utility Boilers
Packaged Boilers
Warm Air Furnaces
Gas Turbines
Reciprocating 1C Engines
Industrial Process Combustion
IMoncombustion
Incineration
Fugitive
TOTAL
1,OOOMg
5,553
2,345
321
338
1,857
333
193
40
498
11,478
1,000 Tons
6,116
2,583
353
372
2,045
367
212
44
548
12,640
Percent
Total
48.4
20.4
2.8
2.9
16.2
2.9
1.7
0.3
4.3
100
 Figure  3-9.   Distribution of stationary anthropogenic
               NO   emissions for  the year 1974  (stationary
               fuel  combustion:   controlled NOX levels).
                               3-20

-------
     TABLE 3-5.   1974 SUMMARY OF AIR AND SOLID  POLLUTANT EMISSION FROM  STATIONARY
                   FUEL BURNING EQUIPMENT  (1000 Mg)



oo
ro


Utility Boilers
Packaged Boilers
Warm Air Furnaces
& Misc. Comb.
Gas Turbines
Recip. 1C Engines
Process Heating
TOTAL
N0xb
5,553
2,345
321
338
1,857
333
10,747
sox
16,768
6,420
232
10.5
19.6
622
24,122
HC
29.5
72.1
29.7
13.7
578
166
889
CO
270
175
132.6
73.4
1,824
9,079
11,554
Part Sulfates POM Ash ^val
5,965 231 0.01 - 1.2 6.18
5,221 146 0.2 - 67.8 4.41
39.3 6.4 0.06
17.3
21.5 a d
4,861 3 *
16,125 382 69
Sluiced
Ash Removal
24.78
1.07
—
—
—
--
~
 No emission  factor available

Controlled N0x

°Based on 80  percent hopper and flyash removal by sluicing methods; 20 percent dry solid removal

-------
               TABLE 3-6.  NOX MASS EMISSION  RANKING OF STATIONARY  COMBUSTION  EQUIPMENT AND  CRITERIA POLLUTANT
                           AND FUEL USE CROSS  RANKING
Sector
1 Utility Boilers
2 Reciprocating 1C
Engines
3 Utility Boilers
4 Utility Boilers
5 Utility Boilers
6 Utility Boilers
7 Utility Boilers
8 Reciprocating 1C
Engines
9 Packaged Boilers
10 Packaged Boilers
11 Utility Boilers
12 Packaged Boilers
13 Utility Boilers
14 Packaged Boilers
15 Packaged Boilers
16 Utility Boilers
17 Packaged Boilers
18 Industrial
Process Comb.
19 Utility Boilers
20 Packaged Boilers
Equipment Type
Tangential
>270 MJ/hr/cyl
Wall Firing
Cyclone Furnace
Wall Firing
Wall Firing
Horizontally Opposed
270 MJ/hr - 270 MJ/hr/cyl
Watertube >105 GJ/hr
Watertube Stoker >105 GJ/hr
Horizontally Opposed
Watertube >105 GJ/hr
Tangential
Firetube Scotch
Watertube <105 GJ/hr
Horizontally Opposed
Watertube <105 GJ/hr
Forced & Natural Draft
Refinery Heaters
Tangential
Firetube Firebox
Fuel
Coal
Gas
Coal
Coal
Gas
Oil
Gas
Oil
Gas
Coal
Coal
Oil
Oil
Oil
Gas
Oil
Coal
Oil
Gas
Oil
Annual NOX
Emissions
(Mg)
1,410,000
1,262,000
946,000
863,500
738,300
481,000
378,700
325,000
318,500
278,170
270,800
232,480
208,000
203,990
180,000
177,900
164,220
147,350
146,000
139,260
	
Cumulative
(Mg)
1,410,000
2,672,000
3,618,000
4,481,500
5,219,800
5,700,800
6,079,500
6,404,500
6,723,000
7,001,170
7,271,970
7,504,450
7,712,450
7,916,440
8,096,440
8,274,340
8,438,560
8,585,910
8,731,910
8,371,170
Cumulative
(Percent)
13.1
24.8
33.5
41.5
48.4
52.8
56.3
59.4
62.3
64.9
67.4
69.5
71.5
73.4
75.0
76.7
78.2
79.6
80.9
82.2
Fuel
Rank
1
21
3
6
4
8
14
>30
16
7
23
26
12
11
5
>30
>30
>30
13
17
SOX
Rank
1
>30
2
3
>30
9
>30
>30
>30
4
5
16
10
11
>30
17
8
29
>30
13
CO
Rank
7
4
6
12
13
17
24
3
29
11
>30
>30
27
>30
>30
>30
>30
>30
>30
>30
HC
Rank
16
1
23
9
28
27
>30
3
19
4
>30
26
>30
>30
22
>30
>30
18
>30
>30
Part
Rank
2
>30
5
13
>30
18
>30
26
>30
1
7
22
19
16
>30
27
9
21
>30
20
CO
I
ro

-------
                                                  TABLE  3-6.   Concluded
Sector
21 Packaged Boilers
22 Gas Turbines
23 Packaged Boilers
24 Warm Air Furnaces
25 Packaged Boilers
26 Packaged Boilers
27 Gas Turbines
28 Reciprocating 1C
Engines
29 Industrial
Process Comb.
30 Utility Boilers
Equipment Type
Watertube Stoker
13.5 - 54 GJ/hr
Watertube <105 GJ/hr
Central
Firetube Stoker <105 GJ/hr
Firetube Scotch
>54 GJ/hr
>270 MJ/hr/cyl
Forced & Natural Draft
Refinery Heaters
Vertical and Stoker
Fuel
Coal
Oil
Oil
Gas
Coal
Gas
Oil
Oil
Gas
Coal
Annual NOX
Emissions
(Mg)
125,350
118,500
116,430
106,300
102,040
98,010
97,400
94,000
92,608
90,900
Cumulative
(Mg)
8,996,520
9,115,020
9,231,450
9,337,750
9,439,790
9,537,800
9,635,200
9,729,200
9,821,808
9,912,708
Cumulative
(Percent)
83.4
84.5
85.6
86.5
87.5
88.4
89.3
90.2
91.0
91.9
Fuel
Rank
>30
30
27
2
29
19
>30
>30
15
>30
SOX
Rank
7
>30
15
>30
6
>30
>30
>30
>30
12
CO
Rank
28
15
>30
10
>30
>30
>30
22
>30
>20
HC
Rank
29
14
>30
8
10
>30
30
13
7
>30
Part
Rank
8
>30
23
25
6
>30
>30
>30
30
>10
no
co

-------
combinations account for over 90 percent of the NOX emissions.   The ranking
of a specific equipment/fuel  type depends both on  total  installed  capacity
and emission factors.   A high ranking,  therefore,  does  not  necessarily imply
that a given source is a high emitter;  large installed  capacity may offset a
low emission factor to give the high ranking.   In  general,  coal-fired  sources
rank high in SOX and particulate emissions, while  1C engines  rank  high in
emissions of CO and hydrocarbons.
       Because of the extremely low population of  supercharged  boilers, no
emissions data for this equipment class is  available.

3.4    CONCLUSIONS
       The characterization of residential  furnaces and  commercial  and
industrial boilers was reported in Reference 3-2.   This  characterization was
extended to gas turbine and supercharged boilers in Section 3.2.   Based upon
this characterization  of area sources,  the  principal  conclusions regarding
catalytic combustion redesign or retrofit are:
       •   The gas turbine combustor is well  suited to  a catalytic  combustor
           redesign/retrofit  because it operates with considerable  excess  air
           and uses clean gaseous or light  distillate fuels.
       •   No theoretical  barriers exist for a catalytic combustor  redesign/
           retrofit on warm air residential  gas- or oil-fired furnaces.
           These systems are  ordinarily not maintained  closely, but they
           may represent a good system  for  early application  of catalytic
           combustors.
       t   Scotch firetube industrial boilers  appear to  offer some  advantages
           to a catalytic  combustor retrofit due to their unique internal,
           first pass  furnace volume.
       •   Watertube boilers  appear to  offer very  little hope for  a catalytic
           combustor retrofit,  but would be adaptable to a  redesigned  system
           using compact heat exchangers.
       •    The  limitation  of  catalytic  combustors  to systems  burning only  clean,
           sulfur-free  gases  or oils may not be necessary.  Further work to
           establish the need for fuel  prevaporization  is required.
                                   3-24

-------
                                REFERENCES


3-1.  Roessler, W. U., et al.,  "Investigation of Surface  Combustion  Concepts
      for NOX Control  in Utility Boilers  and  Stationary Gas Turbines,"
      Environmental Protection  Technology Series Report EPA-650/2-73-014,
      August 1973.

3-2.  Kesselring, J. P., et al., "Catalytic Oxidation  of  Fuels  for NOX
      Control From Area Sources," Environmental  Protection Technology Series
      Report EPA-600/2-76-037,  February 1976.

3-3.  Daman, E. L., and Zoschak, R.  J., " Supercharged Boiler Design,
      Development, and Application," presented at the  Eighteenth  Annual
      Meeting of the American  Power Conference,  March  1956.

3-4.  Aquet, E., "Technical and Economic  Advantages  of Combined Gas  Turbine
      and Steam Power Stations," Sulzer Bros. Technical Review  No. 3/0315/
      Aq. BG-MS, 1971.

3-5.  Zoschak, R. J.,  and Gorzegno,  W.  P., "The  Supercharged Unit -- A
      Projection of Experience," presented at the Pacific Coast Electrical
      Association Engineering  and Operating Conference, Los Angeles, March
      1965.

3-6.  Blazowski, W. S., and Bresowar,  G.  E.,  "Preliminary Study of the
      Catalytic Combustor Concept as Applied  to  Aircraft  Gas Turbines,"
      Technical Report AFAPL-TR-74-32,  Air Force Aero  Propulsion  Laboratory,
      May 1974.

3-7.  De Corso, S. M., et al.,  "Catalysts for Gas Turbine Combustors --
      Experimental Test Results," J. Eng'g for Power,  vol. 99A, pp.  159-167
      (1977).

3-8.  "Environmental Assessment of Stationary Source NOX  Control  Technologies
      -- First Annual  Report,"  Aerotherm  Report  TR-77-58, July  1977.
                                    3-25

-------
                                 SECTION 4
                         CATALYST MATERIALS REVIEW

       This section provides a review of the characteristics  and properties
of currently available catalyst materials.  In particular,  properties  of
monolithic supports, washcoat substrates, and catalyst coatings  are reviewed.
This discussion provides the background to assess the current state of the
art in catalyst materials for combustion applications.

4.1    GENERAL CONSIDERATIONS
       An earlier review of catalyst materials which are suitable for  the
oxidation of hydrocarbon and other fuels was reported in Reference 4-1.
This review included information on
       t   Support types, including monolithic ceramics, pellets, and
           ceramic fiber pads
       •   Substrate materials used as washcoats on monolithic ceramics
       t   Catalyst coatings which were known to be active  in lower-
           temperature oxidation processes
       •   Temperature capability of the catalyst/substrate/support system
       t   Poisoning effects which were known to be deleterious  to catalyst
           materials at lower temperatures.
       To obtain maximum performance from a catalytic combustion system,
the system materials should have the following properties:
       •   A catalyst coating capable of igniting fuel/air  mixtures at,the
           lowest possible temperature (low lightoff temperature),
       t   A catalyst coating of sufficient activity to maintain complete
           combustion conditions in the bed at the lowest possible values
           of combustion air preheat,
                                   4-1

-------
       •   A catalyst coating of sufficient activity to maintain complete
           combustion conditions in the bed at the highest possible values
           of mass throughput,
       •   A washcoat substrate capable of maintaining high surface areas
           (-10 m2/g of monolith plus washcoat)  under high temperature
           (>1370K) combustion conditions, and
       0   A catalyst/washcoat/support system which maintains the catalyst
           in a highly dispersed form under high temperature conditions,
           which is capable of operation at temperatures in excess of
           1755K without thermal degradation or  complexing of the mate-
           rials, and which does not exhibit thermal  shock.
       The identification of a single catalyst system that is capable of
the performance indicated above is probably not  possible.   The identifica-
tion of desirable system properties which are not necessary for good
performance is therefore of high priority.
       For catalyst systems operating in the combustion mode (high bed
temperatures and mass throughputs), the monolithic honeycomb support is
superior to the pellet support in terms of both  catalyst volume required
and pressure drop.  The fiber pad support, while used successfully for
clean, prevaporized fuels as described in Section 2,  is not suitable for
partially vaporized fuels or pressure-drop limited systems.   Therefore, the
primary support material reviewed in this study  is the monolithic honeycomb.

4.2    CHARACTERISTICS AND PROPERTIES OF CATALYST MATERIALS
       The performance of any catalytic system depends on  many factors in
addition to the intrinsic catalytic properties of the active substance.  In
addition to the catalytic properties, the important factors include the
thermal, structural, and chemical properties of  catalyst,  substrate, and
support materials.

4.2.1   Monolithic and Cylindrical Supports
       The support serves three important functions in a catalyst system:
       •   It increases surface area of the active metal or metal  oxide by
           providing a matrix that stabilizes the formation of very small
           particles.
                                   4-2

-------
       t   It increases thermal stability of these very small particles,
           thus preventing agglomeration and sintering with consequent loss
           of active surface.
       •   In some cases, it provides catalytic activity due to special
           properties of the support.
       The monolithic honeycomb is the most technologically advanced support
for purposes of catalytic combustion.  Monolithic supports are composed of
small parallel channels, available in a variety of shapes and hydraulic
diameters.  These structures may be in the form of honeycombed ceramics ex-
truded in one piece, oxidized aluminum alloys in rigid cellular configura-
tions, or multilayered ceramic or metal corrugations.  The channels in
honeycomb-like structures have hydraulic diameters of 1 to 7 mm.   Materials
of fabrication are usually low surface area ceramics, although metal mono-
liths are now being made to overcome the thermal shock and material stability
problems sometimes encountered in the use of ceramics (Reference 4-2).  Table
4-1 lists some of the properties of high temperature ceramics.
       As shown in Table 4-1, the most common high temperature ceramic family
is alumina.  For strengthening purposes, the alumina is alloyed with silica
and/or chromium.  Aluminas are relatively inexpensive, reasonably resistant
to thermal shock, and can operate to high (>1756K) temperatures.   Beryllia
ceramics are about as strong as aluminas and have excellent thermal shock
resistance but are highly toxic.  This latter property makes them undesir-
able for many applications.  Zirconia ceramics can be used at temperatures
up to 2480K, the highest use temperature of all ceramics.  Zirconia is
extremely inert to most metals, even at high temperatures, making it attrac-
tive as a support for metal oxide catalysts.
       The so-called "glass ceramics" are composed of large proportions of
several metal oxides that form complex microstructures.  The three common
glass ceramics are:
       1)  Lithium-aluminum-silicate, or beta spodumene
       2)  Magnesium-aluminum-silicate, or cordierite
       3)  Aluminum-silicate, or mullite (3Al203«2Si02)
                                    4-3

-------
TABLE 4-1.  CERAMIC PROPERTIES (REFERENCE 4-3)
Ceramic
Family
Alumina
Beryl lia
Zirconia
Lithium-
Aluminum-
Silicate
(Beta
Spodumene)
Magnesium-
Aluminum-
Silicate
(Cordierite)
Aluminum-
Silicate
(Aluminous
Keatite or
Mullite)
Silicon
Carbide
Silicon
Nitride
Boron
Carbide
Material
Cost
Low
High
Moderate-
High
Moderate
Moderate
Moderate
Low-High
Low-High
High
Thermal
Shock
Resistance
Fair
Excellent
Fair- Good
Good
Good
Fair- Good
Excellent
Excellent
Fair
Therma 1
Strength
Good
Good
Good
Good
Good
Good
Excellent
Excellent
Poor
Thermal
Conduc-
tivity
Low
High
Low
Low
Low
Low
Low
Low
Low
Other
Most common
high-temperature
ceramic
Highly toxic
Can be used at
temperatures
above 2478 K
Not resistant to
sulfur, sodium
More corrosion
resistant than
LAS
Good corrosion
resistance
Does not self-
bond easily
Does not self-
bond easily
High hardness,
low density
                     4-4

-------
These three ceramics are nearly as low in cost as alumina, but have very
low coefficients of thermal expansion.  Lithium-aluminum-silicate (LAS) is
attacked by sodium and sulfur, however.  Magnesium-aluminum-silicate is more
corrosion resistant and also stronger.  Aluminum-silicate, prepared by
leaching lithium out of LAS particles prior to forming, has both good corro-
sion resistance and high strength.
       Silicon carbide and silicon nitride are both capable of stable high
temperature operation.  However, self-bonding of particles of these two
materials is difficult to achieve.  They are usually prepared by hot press-
ing, giving a dense material that is resistant to thermal shock.  Boron
carbide, which has high hardness and low density, has low strength at high
temperatures.
       Manufacturers of monolith supports include American Lava Corporation,
Corning Glass Works, E. I. DuPont de Nemours & Company, General Refractories
Company, W. R. Grace & Company, Johnson Matthey Corporation, Norton Company,
and Kentucky Metals, Inc.  A variety of materials and configurations are
available, and Table 4-2 presents some of the significant characteristics
of the materials.
       American Lava's Thermacomb corrugated ceramics are available in six
different ceramic compositions, as well as in two structure types (honeycomb
and split-cell).  These ceramics are prepared in corrugated layers and are
very rugged.  Figure 4-1 shows examples of Thermacomb corrugated ceramics.
       Corning Glass Works produces Celcor, a porous cordierite ceramic, in
a honeycomb structure with square or triangular cells.  Examples of Celcor
monoliths are shown in Figure 4-2.  Celcor has been extensively used as a
catalyst support for controlling automotive emissions.
       In addition to cordierite, Corning has prepared higher temperature
ceramics in monolith configurations.   Figure 4-3 shows Coming's zircom'a
spinel  ceramic in varying cell  geometries.  To help suppress thermal shock,
Corning has developed a flexible rectangle geometric configuration which
allows  the cell walls to bend rather than crack.  The flexible rectangle
geometry is compared to conventional  square cell geometry in Figure 4-4 for
a zirconia monolith.
                                    4-5

-------
TABLE 4-2.  MONOLITHIC SUPPORT DATA
Manufacturer
American Lava
Corporat i on
(3M Company)




Corning Glass Works


E. I. DuPont de
Nemours & Company

General Refractories

W. R. Grace & Company
Johnson Matthey
' Corporation
Norton Company

Kentucky Metals, Inc.
Product
Thermacomb 843
Thermacomb LIE
Thermacomb 795
Thermacomb 784
Thermacomb MD-3
Thermacomb 614
Celcor 9475


Torvex

Versagrid

Porami c
Fecralloy
Steels
Spectramic
Honeycomb
RX 387
Spectramic
Honeycomb
RX 384
Kanthal (Metal)
Ceramic
Lithia-Alumina-Silica
Cordierite
Cordierite
Zircon-Mull ite
Mullite
Dense 96% Alumina
Cordierite
Mullite-Alumina-Titanate
Zirconia-Spinel
Alumina
Mullite
Cordierite
Mullite
Cordierite
--
Silicon Carbide
Silicon Nitride
--
Temperature
Limit, K (°F)
1367 (2000)
1478 (2200)
1478 (2200)
1756 (2700)
1700 (2600)
1811 (2800)
1478 (2200)
1923 (3000)
1978 (3100)
1773 (2732)
1623 (2462)
1672 (2550)
1922 (3000)
1478 (2200)
1573 (2371)
1922 (3000)
1811 (2800)
1678 (2560)
Thermal Shock
Resistance
Excellent
Excel lent
Excellent
Good
Good
Fair
Excellent
Good
Fair
Fair
Fair
Good
Fair
Excellent
Excellent
Good
Good
Excellent
                4-6

-------
                    a.  8x enlargment

  b.  Rolled structures
c.  Rolled and stacked structures
Figure 4-1.   Examples  of Thermacomb  corrugated  ceramics,
             produced  by American  Lava  Corporation.
                           4-7

-------
co
                        Figure 4-2.   Celcor cordierite monoliths produced by Corning Glass Works,

-------
10
                          jf&K- - •* m^V PP^^ ^^^" ^^^ —
                        *••••••
                        PP" pm* PMW> nv  — —
                       *••••••

                Figure 4-3. Corning high temperature graded cell ceramic.

-------
I
o



    Figure 4-4.  Zirconia spinel monoliths from Corning Glass Works — flexible rectangle and square cell
                 geometries.

-------
       DuPont produces Torvex ceramic honeycomb in two compositions, alumina
and mullite.  Three geometric configurations (straight honeycomb, slant cell
honeycomb, and cross-flow honeycomb) are available, and examples of these
configurations are shown in Figure 4-5.
       Versagrid ceramic honeycomb is produced by the General Refractories
Company.  Versagrid is available in four cell shapes (round, square, tri-
angular, and rectangular) and two compositions (cordierite and mullite).
Samples of Versagrid honeycomb are shown in Figure 4-6.
       Poramic monolith structures by W. R. Grace & Company have also seen
use in automotive catalytic converter systems.  The ceramic components are
fluxed with polyethylene and passed through rollers to form ribs on sheets
of material, which are then rolled and fired as described in Reference 4-4.
Examples of Poramic materials are shown in Figure 4-7.
       Johnson Matthey Corporation has developed a new metal monolith for
use with platinum catalysts, as described in Reference 4-2.  These metal
monoliths are composed of Fecralloy steels, consisting of up to 20 percent
chromium, 0.5 to 12 percent aluminum, 0.1  to 3 percent yttrium, and the
balance iron.  They are reported to have greater resistance to thermal
shock and mechanical failure than conventional ceramic monoliths.
       Kentucky Metals, Inc. is also fabricating metal monolith materials,
with catalyst application performed by Oxy-Catalyst, Inc.  These monoliths
are made from Kanthal A-l, consisting of 5.5 percent aluminum, 22 percent
chromium, 0.5 percent cobalt, and the balance iron.  A photograph of a cat-
alyzed graded cell metal monolith appears in Figure 4-8.
       Norton Company produces Spectramic honeycomb products in two composi-
tions (silicon carbide and silicon nitride) and in either circular or rec-
tangular cell shapes.  The products can be supplied with the cell axis at
any specified bias angle, and have high maximum use temperatures.  Figure
4-9 shows Spectramic silicon carbide honeycomb.
       Cylindrical supports can also be adapted to certain applications for
catalytic combustion.  Manufacturers of ceramic cylinders include the Coors
Porcelain Company and Coming's Zircoa Products Department, as shown in
Table 4-3.   Coors produces mullite and alumina cylinders In sizes from
                                   4-11

-------



Figure 4-5.   Torvex  ceramic  honeycomb configurations  by  DuPont.
                            4-12

-------
   ft»**a•a»»»»»*»**•»••••*•••••••••••••*••
   •••••••••••••••••••••••••••••••••••••a**
   •••••••••••••••••••••••••••••••••••••••a
   «l««l»««tt»«l»«••«»••••«•»•••»••«
   ••••••••••••••••••••••••••••••••a
   ••••••••••••••••••••••••••••••••I
Figure 4-6.
Versagrid ceramic  honeycomb by General
Refractories  Company.
                         4-13

-------
-pi



-p=>
                           Figure 4-7.   Poramic monolith  structures by W. R. Grace  & Co.

-------
-p.
I
—'
en
•I
     Figure 4-8. Kanthal metal monolith by Kentucky Metals, Inc.

-------
cr>
                          Figure  4-9.   Spectramic silicon carbide honeycomb by Norton Company

-------
                  TABLE 4-3.  CERAMIC CYLINDER PROPERTIES
Manufacturer
Coors Porcelain Company
Zircoa Products Department,
Corning Glass Works
Ceramic
Composition
Mullite
Alumina
Zirconia
Maximum Use
Temperature,
K (°F)
1973 (3100)
2223 (3540)
2478 (4000)
6.35 x 10"3 m (0.25 in) to 0.178 m (7 in) outside diameter while Zircoa
manufactures zirconia cylinders in a variety of sizes.   These ceramics  have
very low porosity, but can be adapted to system configurations.   The maximum
use temperature of the cylinders listed in Table 4-3 are higher  than for mono-
liths of similar materials because of trace element variations.   Section 9.2
describes one suitable system using ceramic cylinders.

4.2.2  washcoat Substrates
       As mentioned earlier, the low surface area of the monolith structure
can be increased by the application of a thin coat of metal oxide material,
such as A^Og.  This washcoat strongly adheres to the ceramic support and
provides a high surface area.  At the same time, because the washcoat is
thin (between 10 x 10~  and 20 x 10~  m), the catalytic material which  is
subsequently impregnated on it is close to the main flow of reactants.
Figure 4-10 shows a schematic representation of the washcoat structure  on
a monolith cell wall.
       The most common washcoat material is y-Al203.   At temperatures  above
1172K (1650°F) the high surface area y-AO  undergoes a phase change  to
relatively low surface area a-
                                    with concomitant sintering.   This  phase
change gives a resultant change in surface area from 300 m/g of washcoat
      o
to 5 m/g of washcoat.  This sintering thus results in pore closure and a
burying of active catalytic sites in the alumina washcoat.   The use of
presintered ^2^3 washcoats, AlgOg washcoats stabilized with Ce02 or CS20
or more thermally resistant washcoats such as Zr02 and Th02 can also be
                                   4-17

-------
              280y
                                Y////A
                              a.  View of washcoat and monolith cell  wall
00
             7777
              Washcoat
                                                                  Macropores ~ 1000A
                                                                  Micropores ~  100A
                                 b.  View of washcoat pore structure
                  Figure 4-10.  Washcoat structure on monolith  - schematic representation.

-------
considered.  Finally, the necessity of maintaining a high surface area,
which is required for low temperature catalysis, may not be required for
high temperature catalytic combustion systems.  Thus, systems using no
washcoats should also be considered.

4-2.3  Catalyst Coatings
          I
       Two broad classes of catalyst coating materials are available:
metals and oxides.  Much of. the work on metal catalysts has been performed
under the automotive emission abatement program where the catalyst operating
temperature is usually around 1000K (1340°F).  Metal and oxide catalysts
are discussed separately in the following sections.

4.2.3.1  Metal Catalysts
       The metals of catalytic interest are listed in Table 4-4.  Of these
metals, the only ones which have a possibility of remaining in the metallic
state in a high-temperature oxidizing environment are the noble metals.
The others readily form oxides and are discussed in  Section 4.2.3.2.  Of
the noble metals, a large volume of data exists for  platinum and palladium
because of their use as automotive oxidation catalysts.  They are among the
most active catalysts for the oxidation of a number  of fuels, including
methane (Reference 4-5), methanol (Reference 4-6), and hydrogen (Reference
4-7).  The high activity of these metals is related  to their ability to
activate H2, 02, C-H, and 0-H bonds.
       Problems in the operation of platinum and palladium catalysts do
exist, however.  Even at the low temperature (<825K) of catalytic reforming,

         TABLE 4-4.  METALS OF INTEREST FOR CATALYTIC COMBUSTION*
                       Group VIII
                     Fe    Co    Ni
                     Group IB
                        Cu
                     Ru
                     Os
Rh
Ir
Pd
Pt
         *Enclosed metals are considered noble.
                                   4-19

-------
a  loss in platinum surface area occurs.  This platinum crystallite growth
can be explained by the volatility of platinum under oxidizing conditions.
On the other hand, palladium tends to be converted to an inactive oxide at
temperatures of 973-1073K in an oxidizing atmosphere and rapidly loses the
ability to burn methane.
       In spite of these problems, the development of a stable metal catalyst
capable of long-term operation at temperatures of 1525 K (2285°F) appears
very realizable.  At 1525K a high platinum dispersion cannot be maintained,
but this loss of platinum surface area may not affect low temperature light-
off characteristics to a great extent.  Thus, a high activity catalyst may
not be required for catalytic combustion applications.
       In addition, although catalyst sintering is severe at 1525K, catalyst
poisoning is much less of a problem.  Sulfur compounds  decompose at catalytic
combustor operating temperatures.  Even lead is not a serious problem, since
compounds such as lead oxide have an appreciable vapor pressure at combustor
temperatures.
       Finally, the use of traces of noble metals may greatly enhance the
performance of certain base metal catalysts.  One example of such behavior
involves the use of platinum with mixed oxides of perovskite structure.
Traces of platinum have been shown to render certain base metal perovskite
catalysts insensitive to sulfur poisoning (Reference 4-8).   Such catalysts
can be expected to operate at a much higher temperature without loss of
platinum than would be the case for a straight platinum catalyst because
the platinum is chemically bound in the perovskite structure and the vapor
pressure of platinum oxide is accordingly reduced.  The perovskite struc-
ture is described in detail in Reference 4-S.
       The use of the other noble metals listed in Table 4-4 appears limited
for catalytic combustion applications.  Ruthenium is known to form a vola-
tile oxide (Ru04) under oxidizing conditions, and this  oxide is rapidly
removed from conventional catalyst supports.  It may be possible to anchor
ruthenium to a support by forming a relatively stable perovskite structure
with certain oxides such as La£03 (Reference 4-10).  Osmium is even more
volatile than ruthenium, and the oxide is poisonous.  It is also very costly
and available only in a limited supply as are iridium and rhodium.  Thus
                                   4-20

-------
the use of these metals would be restricted to small quantities in multi-
metallic systems.  Silver melts at low temperatures, and gold is very
inactive for oxidation.
       Based on these considerations, platinum and palladium show the great-
est promise for use in a catalytic combustion system.  They also have appli-
cation to mixed metal/metal oxide systems.

4.2.3.2  Metal Oxide Catalysts
       The catalytic properties of metal oxides have been studied extensively
by a number of research groups for low temperature applications (Reference
4-11).  Some of the simple oxides, such as €0364, have oxidation activities
comparable to the very active noble metals.  The primary difference between
these oxides and the metals is the "lightoff" temperature, which relates
the ability of a catalyst to reach a significant conversion level at low
temperatures and in short periods of time.  Oxides typically have lightoff
temperatures significantly higher than noble metals with the same fuel.
Since it is important for an oxidation catalyst to reach operating tempera-
ture quickly, doping a high-lightoff-temperature oxide with small amounts
of noble metals to initiate lightoff is feasible, as is the use of a multi-
bed catalyst.
       Oxides of the transition metals have been shown to be the most active
simple oxide catalysts.  These oxides are:
       Co304, Mn02, NiO, CuO, Co203, Fe203, V205.
       Inasmuch as solid state reactions can be very rapid at the high tem-
peratures of catalytic combustor operation, the oxide catalyst and the
monolithic support structure must be compatible.  This means there must
either be mutual insolubility (or at least limited solubility) or that the
melting points of the materials used must be greater than the maximum oper-
ating temperature by a factor of at least 1.5.  Recrystallization becomes
appreciable at about one-half the melting point of a material.  Thus alumina,
with a melting point of about 2305K (3689°F) would be expected to react
readily with most catalytic materials at temperatures above about 1673-1773K
(2551-2731°F).  Fortunately, the reaction product itself may be a suitable
catalyst.   Thus cobalt oxide reacts with alumina to form the less active,
                                   4-21

-------
 though still  catalytic,  cobalt aluminate  (Reference 4-12).  It should be
 noted that the  interaction  of the catalyst and support can alter the
 strength  and  thermal  shock  capabilities of the support.
        With the proper choice of materials, highly active stable catalysts
 of mixed  oxides are  believed possible.  Also, it may be possible to fabri-
 cate honeycomb  structures directly from catalytic oxide compositions.
 Perovskites,  for example, have been proposed for catalyst supports (Refer-
 ence 4-13).   The high sintering temperatures required for development of
 structural  strength  will of course result in a very low surface area.
 However,  any  catalyst intended for use at combustor conditions will rapidly
 sinter in use,  resulting in loss of the high catalyst surface area.  It is
 far better to start  with a  low surface area since this should result in
 better retention of  strength and thermal shock properties as well as
 catalytic activity.
        There  is essentially no literature specifically on catalysts (either
 noble metal or  oxide) suitable for use under catalytic combustor conditions.
 Consequently, a literature  review was conducted on catalyst compositions --
 and  materials likely to  possess catalytic properties -- which could reason-
 ably be expected to  be candidates for use at temperatures as high as 1773K
 (2731°F).   The  emphasis  was placed on compounds of the perovskite type
 since  many  of these  compounds are known to be relatively refractory and have
 attracted much  interest  as  catalytic agents.  Also covered were spinels,
 scheelites, etc.  In particular, melting point and thermal decomposition
 data was  sought.
       While materials of potential high temperature utility have been
studied in the  past, no  studies have been reported for temperatures above
 1273K  (1831 °F).    Based on their promise for use as combustion catalysts
at high temperatures, the most promising metal oxide compositions are:

                1-a)   Ni0_5Mg00i5 (A1Q>5 CrQ_3 ^^^ ^
                1-b)   NT   (A1Q_3 Cr0<5 FeQ>2)  04 + 1XN10
                     Co (A10>3 CrQ-5 FeQ_2)2 0, + 5%CoO
                                  4-22

-------
These materials were developed for use in high temperature thermistors and
were prepared by solid state sintering at 1893K (2947°F)  (References  4-14
and 4-15).  Materials 1-b) and 1-c) should be especially  interesting  since
they contain free nickel oxide and cobalt oxide, respectively.
       The following materials were studied as catalysts  for use at con-
ventional low temperature conditions but based on their composition can be
expected to be useful at high temperatures.
            2-a) LaMn03
       Compound (2-a) is the base compound for a variety  of analogs
       (Reference 4-16).
            2-b)  La0>5 Sr0>5 Mn03
       Compound (2-b) is sulfur sensitive at low temperatures.   Addition of
       traces of platinum is said to render this composition insensitive
       to sulfur poisoning (Reference 4-17).
            2-c)  La0.8 K0>2 RhQ.l Mn0.9 °3
       Compound (2-c) is an example of the almost infinite variations
       possible (Reference 4-18).
       No catalytic data was found for the following materials,  but many
variations should exist which combine both catalytic activity and a reason-
ably high melting point.
            3-a)  ZrXW03
       This class of zirconia-based compounds (3-a)  contains candidates for
       high temperature operation.  Stability of these materials has  been
       studied (Reference 4-19).  Compounds of this  type  warrant extensive
       investigation because many species should exhibit  both catalytic
       activity and a very high melting point.  Relatively high  surface area
       catalysts should be possible with compounds of this type.
            3-b)  LaCoFe03
       The catalytic and thermodynamic properties have not been  determined
for material  (3-b) (Reference 4-20).
                                   4-23

-------
            3-c)   LaMNiOa   ,   M  =   Ca,  Sr,  or  Ba
       The calcium and strontium compounds  should  be  very  interesting.  The
       barium compound would  likely  not  be  stable  because  of  the volatility
       of barium oxide (Reference 4-21).
       The last group is of special  interest because  it  teaches the  use of
perovskite oxides as carriers or structural  supports  for transition  metal
oxide catalysts.   Examples  are:
            ( Sr0>2 La0.8  ) Co03
       or
4)          ( Sro.4 La0.6  )(  Coo.g ptQ.l  )  °3   as catalysts
       on
            LaAl203
       or
            ( Sr0-4 La0i6  )(  Co0.5 V0>2  ) 03   as  catalyst support
(Reference 4-13).
       It should be noted  that all compounds containing  metals with  volatile
oxides are unsuitable for  combustor  service. Therefore, the  very many
compounds containing barium,  lead, rhenium,  ruthenium, sodium, etc.  have
been excluded from consideration. Similarly, compounds  containing halogens
are not considered useful  for combustor  service.   In  addition, all compounds
with known melting points  below about 1873K (2911°F)  have  been excluded
Melting points of at least 2273K (3631°F) are sought.
       In summary, it appears that there are many  mixed  oxide catalysts
likely to be operable at 1773-1873K  (2731-2911°F).
                  Ni1+x (  A1Q>3 Cr0_5 Fe0_2  )g  04+x and
                  c°l+x (  AlQ.3 Cr0.5 Fe0.2  ) 04+x
represent compounds which  should be  usable  up to 1773K.  However,  it is
almost certain that superior  compositions exist, and  a systematic study of
the high temperature capabilities of the many types of mixed  oxide compounds
possible should be made.
                                    4-24

-------
4.3    CONCLUSIONS
       Based upon a review of the available catalyst materials  and  their
properties, the following conclusions have been made:
       •   To obtain the maximum performance from a  catalytic combustion
           system, it is desirable to have a high temperature support
           material, a washcoat capable of maintaining  a  high surface  area
           under high temperature operation, and a highly dispersed catalyst
           having a low lightoff temperature and high activity.   Such  a
           system may not exist, making performance  trade-offs  necessary.
       •   The monolithic honeycomb support is  the most technologically
           advanced geometric configuration for catalyst  carriers,  minimiz-
           ing both pressure drop and volume of catalyst  required.
       •   Newly developed metal monolith materials  may have advantages over
           ceramics in terms of thermal shock characteristics.
       •   It may not be necessary to maintain  a high surface area  in  a
           catalytic combustor system, and hence, systems using  no  wash-
           coats should be considered potentially viable.
       0   The most promising noble metal catalysts  are platinum and
           palladium because of their high activity  and relatively  low
           lightoff temperature.
       •   Simple oxides of the transition metals should  have good  catalytic
           activity but will have higher lightoff temperatures  than the
           noble metals.
       •   The most promising high temperature  (>1773K) catalysts are
           mixed oxides containing either free  nickel oxide or  cobalt  oxide.
                                   4-25

-------
                                REFERENCES
4-1.   Kesselring, J.  P., et al.,  "Catalytic  Oxidation  of Fuels  for  NOX
       Control  from Area Sources,"  Environmental  Protection  Technology
       Series Report EPA-600/2-76-037,  February 1976.

4-2.   Pratt, A.  S., and Cairns, J.  A., "Noble  Metal  Catalysts on  Metallic
       Substrates," Platinum Metals  Review, Vol.  21,  No.  3,  July 1977.

4-3.   Bittence,  J. C., "Sorting Out the Ceramics," Machine  Design.  Vol.  49,
       No. 28,  December 8, 1977.

4-4.   Lundsager, C. B., et al.,  "Thermoplastics  Process  for Catalyst
       Substrates," SAE Paper 760311.

4-5.   Dixon, J.  K., and Longfield,  F.  E.,  in Catalysis,  edited  by P.  H.
       Emmett,  Reinhold, New York,  1960, Vol. 7,  p. 183.

4-6    Bond, G. C., Catalysis by Metals. Academic Press,  New York, 1962,
       p. 464.

4-7.   Kowaka,  M. J.,  Japan Inst.  Metals, Vol.  23,  1959,  p.  659.

4-8.   Johnson, D. W., et al.,  "The  Nature  and  Effects  of Platinum in
       Perovskite Catalysts," J. Catalysis, Vol.  48,  1977, pp. 87-97.

4-9.   Evans, R.  C., An Introduction to Crystal  Chemistry, Cambridge
       University Press, London, 1952.

4-10.  Shelef,  M., and Gandhi,  H.  S.,  "Stabilization  of Ru-Containing  Nitric
       Oxide Reduction Catalysts,"  publication  preprint from Scientific
       Research Staff of Ford Motor  Company,  Dearborn,  Michigan.

4-11.  Margolis,  L. Y., Advances  in  Catalysis,  Vol. 14, p. 429,  Academic
       Press, New York.

4-12.  Schachner, H.,  "Cobalt Oxides as Catalysts," Cobalt,  December 1960,
       pp. 1-10.

4-13.  McCann,  E. L.,  "Perovskite  Oxides as Carriers  for  Transition  Metal
       Oxides," German Pat. Offen  2,615,352.

4-14.  Matsuo,  Y. and Hayakawa,  S.,  Japan Kokai,  76-22,093.

4-15.  Matsuo,  Y. and Hayakawa,  S.,  Japan Kokai,  76-22,094.

4-16.  Voorhoeve, R.J.H., Remeika,  J.  P. and  Tremble, L.  E., "Defect
       Chemistry  and Catalysis  in  Oxidation and Reduction over  Perovskite-
       Type Oxides," Ann. N.Y.  Acad. Sci.,  Vol.  272,  1976, pp.  3-21.
                                    4-26

-------
4-17.  Johnson, D. W. Jr., Gallagher, P.  K.,  Wertheim,  G.  K.,  and Vogel,
       E. M., "The Nature and Effects of Platinum in Perovskite Catalysts,"
       J. Catalysis, Vol. 48, 1977, pp.  87-97.

4-18.  Voorhoeve, R.J.H., Remeika, J. P.  and  Trimble, L.  E.,  "NO and
       Perovskite-Type Catalysts," Cat.  Chem.  Nitrogen  Oxides,  Ed.  by
       Klimish, R. L., and Larson, J. G., Plenum.

4-19.  Ekstrom, T. and Tilley, R.J.D., "Stability of Perovskite-Type  ZrXW03,"
       J. Solid State Chem. 1_9, 1976, pp. 227-33.

4-20.  Rao, C.N.R., Parkash, 0. and Ganguly,  P.,  "Electronic  and Magnetic
       Properties of Perovskites," J. Solid State Chem. Vol.  15, 1975,
       pp. 186-92.

4-21.  Obayashi, H. and Kudo, T., "Crystallographic, Electric  and Thermo-
       chemical Properties of Lanthanum M Nickelate," J.  Appl.  Phys., Vol.
       14, 1975, pp. 330-5.
                                   4-27

-------
                                 SECTION 5
                 CATALYST PREPARATION AND CHARACTERIZATION

       An essential feature of design criteria for catalytic combustion sys-
tems is information concerning the catalyst/substrate/support system in terms
of catalyst type, catalyst loading, washcoat type, support type,  pre- and
post-test total surface area, pre- and post-test dispersion (for  precious
metal catalysts), and microscopic surface characteristics.  Due to the pro-
prietary nature of much of the current work in catalyst systems,  it was nec-
essary to prepare most catalysts in-house and to supplement these with
conmercially available systems.  All systems were then subjected  to pre-
and post-test characterization analyses.  This section provides information
on the catalyst preparation and characterization techniques (Sections 5.2
and 5.3) used in this study.  Also included is a description of the EPA/Acurex
characterization laboratory.

5.1    GENERAL CONSIDERATIONS
       A review of the available materials for use in catalytic combustion
systems is presented in Section 4.  While an identification of these mate-
rials is important in characterizing the catalyst system,  additional infor-
mation is required to fully define the system.  This additional information
includes:
       •   Catalyst loading
       •   Catalyst pre-treatment techniques
       t   Total (BET) surface area, pre- and post-test
       •   Precious metal dispersion, pre- and post-test
       •   Pore volume and pore size distribution
       t   SEM-EDAX analysis of catalyst surface

                                    5-1

-------
       Measurement of these physical  properties before and after combustion
reaction can be correlated with catalyst performance to understand the basis
for specific catalyst behavior.  Combining the measurement of these physical
properties with the laboratory preparation of catalyst systems provides the
maximum amount of information to be gained from combustion testing.

5.2    CATALYST PREPARATION
       Since a very limited number of commercially available catalysts suit-
able for combustion applications could be identified,  most catalyst systems
were prepared at Acurex.   Preparation of these systems is performed with
standard laboratory equipment.  All catalyst systems include a monolithic
honeycomb support.  For those systems employing a washcoat, the washcoat is
usually placed on the monolith by an  outside vendor.  The washcoat is  either
used as supplied, or pre-sintered by  placing the washcoat/monolith in  a muf-
fle furnace for 8 hours at approximately 1360K.  The washcoat/monolith is
then dried in a muffle furnace, cooled in a desiccator, and weighed.   For
platinum catalysts, an aqueous solution of chloroplatinic acid (H-PtClg) is
then prepared to give the appropriate platinum loading.  This aqueous  solu-
tion is applied to large cell monoliths by brushing and to small  cell  mono-
liths by dipping.
       For metal oxide catalyst systems using cobalt or nickel oxides, the
monolith is dipped into a melt or an  aqueous solution  of the metal nitrate.
Compressed air is used to gently blow out any clogged  cells.   The system is
then dried, calcined, and weighed to  determine the catalyst weight percent.
       Catalyst preparation procedures are given in tabular form in Table 5-1.

5.3    CATALYST CHARACTERIZATION
       Once the catalyst system has been prepared, it is further character-
ized in terms of total and selected surface areas.  Additional characteri-
zation of the surface can be done by  measurement of pore volume and pore
size distribution as well as by surface microscopy.
       The monolithic support is ordinarily a ceramic of low surface area
(-0.01 nr/g).  The washcoat substrate, also a ceramic, is applied to increase
the surface area.  This increased surface area allows higher precious  metal
                                    5-2

-------
                TABLE 5-1.  CATALYST PREPARATION PROCEDURES
   Operation
             Procedure
   Materials
   Required
Monolith
Wet Point
Determination
(for dipped
catalysts)
Washcoat
Application
Monolith
Impregnation
with Noble
Metal by
Dipping
Monolith
Impregnation
with Noble
Metal by
Brushing
Monolith
Impregnation
with Base Metal
1. Dry monolith at 673K for 3 hours.
2. Weigh monolith.
3. Cover monolith with 500 ml de-
   ionized water.
4. Remove monolith from water,
   measure water uptake.

1. Determine monolith wet point.
2. Dip monolith in Alon 3-D (30%
   aqueous dispersion of alumina) or
   50% Zr(N03)4 in water.
3. For Al^Og washcoat, fix Alon with
   NHs; for Zr02 washcoat, calcine at
   773 K for two hours.
4. Repeat step 3 until desired weight
   percent is achieved.

1. Determine monolith wet point.
2. Prepare solution of metal  salt
   with proper concentration  to give
   desired weight percent of  metal.
3. Dip monolith in solution until all
   solution is taken up.
4. For Pt catalysts, fix metal by
   spray while catalyst is wet.
5. Calcine at 673K for 3 hours.

1. Calculate metal salt needed to
   give desired loading.
2. Prepare solution of metal  salt.
3. Using small brush or cotton swab,
   paint monolith with solution.
4. Fix Pt by H2S spray.
5. Dry, calcine, and weigh.
6. Repeat step 3 until all solution
   is used.

1. Dry monolith at 673K for 3 hours.
2. Dip monolith in melt of appropri-
   ate nitrate, or in solution of
   appropriate nitrate.
3. Calcine at 673K for 3 hours.
4. Repeat step 2 until desired load-
   ing is obtained.
De-Ionized Water
Beakers
Zr(N03)4
                                                         De- Ionized Water
                                                         Beakers
                                                         NH3 (g)
Metal Salt
De-Ionized Water
Beakers
H2S (g)
Metal Salt
De-Ionized Water
Beaker
Watch Glass
H2S (g)
Paint Brush
Base Metal Nitrate
Beakers
De-Ionized Water
                                   5-3

-------
dispersions to be achieved by reducing agglomeration of precious metal
crystallites.  Thus a high total surface area measurement indicates a wash-
coat material of high surface area, and a high selected surface area, or
dispersion, measurement indicates a precious metal that has been placed on
                                    o
the surface in extremely small (<50 A) crystallite sizes.
       A description of total and selected surface area, pore size and pore
volume, and SEM-EDAX analyses follows.

5.3.1  Total Surface Area
       The basis for surface area measurements is the Langmuir gas adsorp-
tion theory in which the surface of a solid is regarded as an array of active
adsorption sites, each site with the capacity to adsorb one molecule.  It is
assumed that when a gas molecule strikes an active 'site it will remain for a
period of time and then re-evaporate.  Brunauer, Emmett and Teller (Refer-
ence 5-1) extended the Langmuir theory to apply to the second and higher
layers of adsorbed molecules.  Based on this work, the total surface area
is given by
                                                                       (5-1)
where W = catalyst weight in grams
                                                          o
      A = area occupied per molecule (for argon, A = 14.6 A2)
      and Vm is found from the BET equation
                             - P)   V C    V C  P
                                '    m      m    o
       with C being a constant at a given temperature.
                HI - Hi
       C = exp
where H-| = heat of adsorption of first layer
      HI = heat of liquefaction
      R  = universal gas constant
      T  = temperature
     Experimentally, it is found that a plot of P/V(P  - P) versus P/P  is
linear in the range of relative pressures from 0.05 to 0.35 (see Figure 5-1)
The slope is (C - l)/VmC and the intercept is l/VmC.  Thus Vm = (slope +
                                    5-4

-------
RuCu Alloy  Powder
         P(cm)
         1.0625
         2.0408
         4.5073
       P/Pp
       .0506
       .0972
       .2146
V.(cc/.g)
 4.2545
 5.1681
 6.6749
P/V(PQ-P)
.012526
.020828
.04094
Q-
 I
 O
Q.
     .0500
     .0400
     .0300
     .0200
     .0100
Slope = 0.171
Intercept =  .0039
                                             vm MS  +  i)
                                                         -1
                                                        ,-1
                              = (.1749)"1 = 5.73 cc/g

                           SA = 5.73 cc/g x 3.927 m2/cc
                              =22.5 m2/g
                        .100
                     .200
                      P/P,
             .300
                        Figure 5-1.   Argon BET at 77K.
                                     5-5

-------
intercept)"1.   An example of a total surface area calculation using the BET
equation is given in Figure 5-1.
     One should note that the number of molecules adsorbed on the monolayer
is related to the volume adsorbed by the expression
                                       23
where  N = Avogadro's number (6.03 x 10   molecules/mole)
  and mv = gram molecular volume (22,400 cc/mole)
       The gas adsorbate may be nitrogen, krypton, or argon.  Measurement
of  total surface area by physical adsorption of argon is carried out at the
liquid nitrogen temperature (77K).

5.3.2  Selected Surface Area (Dispersion)
       For a noble metal supported on an inactive refractory oxide, the
catalytic activity should be a function of the metal surface area.  The
metal surface area can be determined in a number of ways, including x-ray
diffraction line broadening and electron microscopy.  The most convenient
and accurate technique, however, is by the chemical adsorption of a gas.
       The stoichiometry of selective adsorption is established on pure
metal powders or on supported metal samples characterized by the techniques
described above.  The usual gas adsorbate is hydrogen or carbon monoxide.
The two adsorption techniques described below are used in the determination
of  catalyst dispersion.

Hydrogen Chemisorption
       After a platinum catalyst has been reduced at high temperature (673K)
and the sample cell evacuated, the metal surface is measured by hydrogen
chemisorption and expressed as percent dispersion.  The stoichiometry is:

                    Pt(surface) + | H2— -PtH(surface)

On  very small metal crystallites the stoichiometry may change such that two
atoms are adsorbed for each atom of surface Pt:

                    Pt(surface) + H2 — -PtH2(surface)

                                      5-6

-------
Calculated dispersions of over 1QQ percent indicate well  dispersed catalysts
and should be rounded to 100 percent.

Hydrogen Titration
     When, after reduction with hydrogen and removal of the gas, the plati-
num catalyst is exposed to an atmosphere of air or oxygen, a surface layer
of platinum oxide forms.  The stoichiometry of the reaction of hydrogen with
PtO, or titration, is:

                PtO(surface) + 3/2 Ho—^PtH(surface) + H20

In the case of platinum on an alumina washcoat, the water formed is adsorbed
by the alumina.  The consumption of  1.5 Ho molecules for each atom of sur-
face Pt results in an increased sensitivity in the surface area measurement
compared with chemisorption.
       Dispersion is expressed as

                             nPt active sites
                                "total  Pt

Wlth                                    2/ymoleH^
                                       _  _
                       Pt active sites   3
                                           (u
from the stoichiometry of the H£ titration, and

                        -    g Pt      106      p.  _ y mole Pt
              "total Pt " g catalyst 195.09 g-^-r-      g
                          3      j          » mole      3
Uptake of H~ for each expansion is
             .n _  D" lu-Uit  pi + pi-UPC-1   ,\   piPC-1 ymole
             An ~    W T     K1   V2  lPC-2   7   P2 PC-2    g
                                                 Pi PC-ll
                                              "  Yl PC-2J
where  VQ = dosing volume (cc)
       W  = weight of catalyst (g)
       T  = temperature (K)
       P  = initial pressure (torr)
                                     5-7

-------
       P? = pressure after expansion (torr)
       i  = expansion number
     Pf 1
     pp~' = average ratio of initial and final pressures from dead space
            expansions

Extrapolation of the ZAn vs. P2 plot gives total uptake of H2 at P0; the
dispersion of Pt is then calculated.  A complete description of this tech-
nique is given in Reference 5-2.

5.3.3  Pore Size and Pore Volume (Reference 5-3)
       A Type II Langmuir isotherm (see Figure 5-2) describes the condensa-
tion of gas molecules onto a nonporous surface.  When the surface has micro-
porosity, the relation between pressure and physical adsorption is shown by
a Type IV isotherm (also shown in Figure 5-2).  The differences between the
two adsorption isotherms are due to the finite size of the internal  pores
and allow an assessment of the internal pore volume of the surface.
       Type II and Type IV isotherms are nearly identical at P/P0 <0.3 such
that the BET equation is applicable to both nonporous and porous materials.
The differences arise at moderate and high relative pressures.
       A Type II isotherm approaches the P/P0 = 1  limit asymptotically,
whereas the Type IV isotherm breaks at some finite P/P0 value less than
unity and becomes nearly horizontal at the limit.   The break occurs  at point
                                     o
G, when the pores smaller than -1000 A diameter (the major contributors to
total surface area) are filled.  Beyond this point adsorption is slight be-
cause the available surface area is relatively low.
       Another feature of a Type IV isotherm is the hysteresis which occurs
when the relative pressure is reduced and desorption takes place.  Due to
the relatively small  size of the pores (assumed to be cylinders with diameter
d) and the concave meniscus formed by the adsorbate liquid in the pore, the
effective vapor pressure of the liquid in the pores is lower and desorption
occurs at a lower pressure.  The vapor pressure is a function of the meniscus
curvature as given by the modified Kelvin equation

                               £nZ.= _2iy_                             (5_4)
                                  P0    rRT                             (b 4'

                                    5-8

-------
c
o

•Jj
Q.
s_
o
l/l
-o

-------
where  P  = saturation vapor pressure at temperature T (K) of the system
       r  = radius of pore
       y  = surface tension of the adsorbate in liquid form
       v =  molar volume of adsorbate in liquid form
       R =  gas constant per mole
Thus, at any point H on the evaporation leg of a Type IV isotherm, the radius
(r) of the pores (which contain saturated liquid at the relative pressure
P/P0) can be calculated.  The variation of P/P0 with cylindrical pore radius
varies, as shown in Table 5-2.
                TABLE 5-2.  VARIATION OF RELATIVE PRESSURE
                            WITH PORE RADIUS
               radius (A)
                    50
                   500
                  5000
•readily measured
0.66
0.96
0.996 [difficult to measure
A practical upper limit for evaluation of pore volume radii  is about 1000 A.
       Calculation of internal pore volume from a Type IV isotherm is done
in the following manner.  For idealized cylindrical  pores of diameter d which
are many diameters deep, total surface area SA and volume V  are related by
                                  __
                                  SA
       d
       4
                                  (5-5)
Extension of the isotherm to the high relative pressure break point G (Figure
5-2) allows the calculation of the radius r from Equation (5-4).  Total  sur-
face area is determined as described previously; V can then be calculated.

5.3.4  SEM-EDAX Analysis
       Use of the scanning electron microscope and energy dispersive x-ray
analysis is helpful  in determining changes in the catalyst surface structure
and precious metal  crystallite size.  Using SEM-EDAX analysis, the surface
                                    5-10

-------
composition can be identified by element, and the location and size of plati-
num crystallites can also be determined.  Examples of results from SEM-EDAX
analysis are given in Sections 7 and 8.

5.4    CATALYST CHARACTERIZATION LABORATORY
5.4.1  Gas Adsorption Apparatus
       The gas adsorption procedures described in Sections 5.2 and 5.3 re-
quire a finely calibrated high vacuum control volume for pretreatment and
measurement of total and selected surface areas of catalyst systems.  In
addition, gases are delivered to and stored within the apparatus.  The gas
adsorption system used in this study is shown in Figure 5-3.  The major
components are:  (1) the pumping and vacuum control system, (2) the gas
delivery, storage, and cleaning system, and (3) the working volume and sam-
ple cell.  These parts are described in the following sections.

5.4.1.1  Vacuum Control System
       Figure 5-4 is a schematic diagram of the pumping and vacuum control
system.  A mechanical vacuum pump is used first to rough the unit to approx-
imately 0.1 torr pressure, which is read on the thermal-conductivity gauge.
The oil diffusion pump then is used for evacuation to 10   torr as read on
the ionization gauge.  A liquid nitrogen cold trap between the pumps and
the working system condenses volatiles released into the unit by the pumps,
mainly oil vapors.

5.4.1.2  Gas Delivery, Storage, and Cleaning System
       Three gases are used routinely in adsorption measurements:  helium
for dead space detejminations, argon for total surface area, and hydrogen
for platinum dispersion.  The helium and argon cylinders are connected to
the adsorption apparatus by a gas line equipped with a safety pressure valve
rated at 15 pounds per square inch (gauge).  The inert gases also pass
through a dry ice trap which condenses any water present.  A second inlet
port is used exclusively for hydrogen.  From the cylinder, hydrogen passes
first through a De-Oxo unit which removes oxygen by conversion to water,
and then through a zeolite trap which dries the gas.  A mercury manometer
serves as a relief valve.   Gases thus introduced into the apparatus are
                                    5-11

-------
                                                              GAS STORAGE BULBS
on
 i
                        GAS MANIFOLD
                       GAS INLET LINES
                         DIFFERENTIAL
                      PRESSURE GAUGE
                            THERMAL-
                         CONDUCTIVITY
                              GAUGE
                         TEMPERATURE
                         CONTROL UNIT
                                                                                                         IONIZATION
                                                                                                         GAUGE

                                                                                                         CALIBRATED
                                                                                                         MERCURY
                                                                                                         MANOMETER

                                                                                                         VACUUM MANIFOLD
                                                                                                         TO DIFFUSION
                                                                                                            PUMP
COLD TRAP
SAFETY
MERCURY
MANOMETER
                                                                                                         LARGE DIAMETER
                                                                                                         RESISTANCE FURNACE
                                                                                                         MECHANICAL
                                                                                                         PUMP

                                                                                                         SMALL DIAMETER
                                                                                                         RESISTANCE FURNACE
                                                                                                         (STORAGE LOCATION)
                                       Figure  5-3.    High  vacuum  gas  adsorption  apparatus,

-------
              lonization
              Gauge
                                 To working system
                                    Liquid N£
                                    Cold Trap
01
/Thermal-Conductivity
 Gauge
                    Mechanical
                    Vacuum
                    Pump
           Diffusion
           Pump
                               Figure  5-4.   Schematic  diagram  of  vacuum  control  system.

-------
stored  in 5-liter bulbs above the main manifolds.  Note that a gas is  intro-
duced into a section only after the storage bulk has been evacuated or
thoroughly purged with the gas.

5.4.1.3 Working Volume and Sample Cell
        Figure  5-5 shows the locations of the dosing volumes (V-, and V^),
working cross  (V2), cell connect, and pumping connection on the adsorption
apparatus.  The Wallace and Tiernan differential pressure gauge (shown in
Figure  5-3) contains volume V.,.
        The volumes  used in gas adsorption measurements have been calibrated
with  the small dosing bulb used as a reference.  Before the small bulb was
sealed  into place,  it was weighed, filled with mercury and reweighed.   The
volume  V] was  calculated from the density of mercury at ambient temperature.
On  the  basis of four measurements, the accuracy of this determination  is
estimated to be ±2.5 percent.
        System  calibrated volumes are listed in Table 5-3.
                  TABLE  5-3.  MEASURED VOLUMES USED IN GAS
                             ADSORPTION MEASUREMENTS

                                 =  57.12 cc
                                     3.16 cc
                                 =  19.24 cc
                                   493.40 cc
The volume extending into the Wallace and Tiernan gauge, V3, varies slightly
with pressure as the diaphragm inside the gauge shifts.  The value of V3 was
established with several calibrations.
       The sample cell is positioned at the pumping connection for outgassing
and reduction of the catalyst system.  This minimizes the time required for
pretreatment since the working cross, which is made of capillary tubing, is
bypassed.
                                   5-14

-------
  TO W&T
PRESSURE
  GAUGE
WORKING
CROSS
                                           CELL   HIGH VOLUME
                                       CONNECT   PUMPING CONNECTION
          Figure 5-5.  Detail  of gas adsorption  system showing  dosing volumes,
                                            5-15

-------
5.4.2  Test Procedure
       For the determination of BET and selected surface area by the volu-
metric calculation of an adsorption isotherm, successive charges of adsorb-
ate are admitted from the dosing volume.   After each admission, readings
of pressure are taken until there is no further detectable change.  Equi-
librium is assumed at this state.  The amount adsorbed is calculated for
each equilibrium pressure as the difference between the total quantity
of gas which has been admitted and the quantity remaining in the dead space.
The catalyst characterization procedures  are given in Table 5-4.
       Total surface area and dispersion  measurements have been made for
many of the catalysts in this study.  Catalyst type and application have,
as a result, been shown to directly influence combustion performance in a
predictable way.  Catalyst measurements are discussed with combustion test
data of Sections 7 through 9.
                                    5-16

-------
           TABLE 5-4.  CATALYST CHARACTERIZATION PROCEDURES
  Operation
                Procedure
  Ma ten' al s
   Required
Total Surface
Area (BET)
Measurement
Selected
Surface Area
(Dispersion)
Measurement
1. Evacuate dosing volume, pressure gauge,
   and working cross; load monolith in
   cell.
2. Fill dosing volume with helium and
   measure pressure PC-1, the initial
   pressure in the dosing volume, and
   room temperature.
3. Expand helium into test cell; read
   pressure PC-2, the final pressure in
   the total volume after expansion into
   the test cell, at 5-minute intervals.
   Stop when equilibrium is reached.
4. Pump cell to vacuum and repeat steps
   2-3 at different values of PC-1.
5. Steps 2-4 determine dead volume.
6. Fill dosing volume with argon and
   measure PC-1.
7- Expand argon into test cell; read pres-
   sure PC-2 and room temperature.
8. Repeat steps 6-7 for a minimum of five
   expansion points.
9. Calculate BET surface area as described
   in Section 5.3.1.

1. Reduce precious metal catalyst with h^.
2. Evacuate dosing volume, working cross,
   and catalyst system.
3. Fill dosing volume with hydrogen and
   measure pressure PC-1 and room temper-
   ature.
4. Slowly expand hydrogen into cell and
   read pressure PC-2.  Repeat until pres-
   sure has stabilized.
5. Close test cell stopcock and repeat
   steps 3-4 at least three times.
6. Measure dead volume as in BET proce-
   dure.
7. Calculate dispersion as described in
   Section 5.3.2.
Nitrogen
Helium (g)
Argon (g)
Hydrogen (g)
                                  5-17

-------
                              LIST OF SYMBOLS

English
   A   -   area
   C   -   constant
   d   -   diameter
   i   -   expansion number
   mv  -   gram molecular volume
   n   -   number
   N   -   Avogadro's number
   P   -   pressure
   R   -   gas constant per mole
   r   -   radius
   SA  -   total surface area
   T   -   temperature
   V   -   volume
   W   -   weight

Greek
   y   -   surface tension

Subscripts
   o   -   initial or saturation condition
   m   -   molecules
   D   -   dosing
   1   -   initial condition
   2   -   expanded condition

Superscript
   i   -   expansion number
                                   5-18

-------
                                REFERENCES
5-1.   Brunauer, S., Emmett,  P.  H.,  and Teller,  E.,  "Adsorption of Gases in
      Multimolecular Layers," Journal  of the American  Chemical Society,
      Vol.  60, p.  309, 1938.

5-2.   Benson, J.  E., and Boudart,  M.,  "Hydrogen-Oxygen Titration Method for
      the Measurement of Supported Platinum Surface Areas,"  Journal of
      Catalysis,  Vol. 4, p.  704, 1965.

5-3.   Gregg, S. J., and Sing, K.S.W.,  Adsorption,  Surface  Area, and Porosity,
      Academic Press, New York, 1967.
                                   5-19

-------
                                 SECTION 6
                       CATALYTIC COMBUSTOR ANALYSIS

       Catalytic combustion in a honeycomb monolith is a complex process
which involves the interaction of several physical and chemical phenomena.
Of primary importance are (1) radial heat and mass transport between the gas
and wall, (2) axial heat and mass transport in the gas, (3) axial radiative
and conductive wall heat transfer, (4) heterogeneous surface and bulk gas
phase chemical kinetic reactions.  The interaction of these phenomena de-
termines the maximum mass throughput and fuel conversion efficiency of the
catalytic bed.
       In this section, the analysis of a catalytic combustor is discussed
with regard to:
       •   Fundamentals of operation, where a simplified version of the
           catalytic combustion process is described and the important sys-
           tem implications are introduced,
       •   The PROF-HET computer code, which models all of the important
           physical phenomena occurring within monolithic catalytic com-
           bustors and verifies some of the conclusions of the simplified
           model,
       •   Conclusions and recommendations regarding major system impacts
           and further analytical requirements.
The following discussion provides the information needed to understand the
operation of the  catalytic combustion system.

6.1    FUNDAMENTALS OF OPERATION
       Catalytic  combustion in a monolith bed includes the interaction of
chemical reactions (surface and gas phase), diffusive heat and mass transport

                                    6-1

-------
(laminar or turbulent), convection, bed conduction, and radiation.  These
phenomena are depicted schematically in Figure 6-1.  During steady operation,
the catalytic combustion process can be described as follows:
       •   Premixed fuel and air are introduced into the combustor.
       t   These gases diffuse to the catalyst-coated surface of the com-
           bustor and react on the active sites at and within the surface.
           Near the cell entrance, where most of the gas is at low tempera-
           ture, gas phase chemical reactions are unimportant.
       •   At the entrance, heat release is controlled by catalytic wall
           chemical reactions.  This heat is transferred by conduction,
           radiation, and convection.  Further down the channel, where the
           gas has been preheated to a high temperature, gas phase reactions
           become active.  In this region fuel is rapidly consumed by a
           "flame-type" phenomenon which controls the amount of unburned
           hydrocarbon emissions that escape the system.
       •   Surface reaction products diffuse back to the main flow of gases
           and are carried downstream.
       Under normal operating conditions, wall and gas phase reactions are
active and very little unburned hydrocarbon escapes the bed for lean and
stoichiometric initial mixture ratios.  However, it has been experimentally
observed that above a certain mass flow limit, small increases in flowrate
cause an abrupt rise in unburned hydrocarbon emissions.  The abruptness of
the increase indicates that a "flame type" phenomenon has been extinguished.
This condition, called breakthrough, represents an upper mass throughput
for low unburned hydrocarbon emissions.
       Increasing the mass throughput in a catalytic bed to levels much
above the breakthrough point can cause the front of the bed to become cool.
It has been experimentally found that small increases in mass throughput,
once the front end of the bed has become cool, can cause the cool region to
spread downstream.  At this point, all wall reactions are extinguished and
the entire bed becomes cold.  This condition, called blowout, represents
the maximum mass throughput for hot bed operation.  It  is very important to
know when this blowout condition occurs for a given catalyst system.
                                    6-2

-------
                                                 Surface Reaction
                                                   Controlled
                                         Gat Phaie Reaction
                                            Controlled
CO
               m
                                                                    ^radiation
Kiw Tw   Wa" rMCtk>n
                                                                                                      =—  Gas phase
                                                                                                               reaction
                                                   q conduction
                                          Figure  6-1.   Physical events  in  a monolith  cell.

-------
6.1.1  Graphical Determination of Stable Surface Combustion States
       For the purpose of understanding system characteristics, considerable
simplification of the catalytic combustion process can be made if it is as-
sumed that:
       •   No conductive or radiative heat transfer occurs
       •   The Lewis number is unity for all species
       •   The combustion reaction is a single global reaction described by
           an Arrhenius law equation
       Based on these assumptions, the temperature at the monolith wall,
TW,  can be related to the residual concentration of lean reactant at the
wall, K,  by
                           TW = TA -
                   0, the surface reac-
tions approach equilibrium conditions, and the wall temperature is equal to
the adiabatic flame temperature.  This relationship is shown in Figure 6-2.
         Figure 6-2.
Wall temperature variation with lean reactant
wall concentration.
                                      6-4

-------
        Once the relation  between  Tu  and  K.,  is  established,  it  is  possible
                                   IN       Vfl
 to  perform a simple mass  balance  on  the  lean reactant  at  the wall of  the
 monolith  bed;  that is,  the  mass of lean  reactant  transported to the wall
 1s  equal  to the mass of lean  reactant  consumed at the  wall.  In equation
 form  this can  be written  as

                      m  =  Nu ^ (KE - Kw)  =
                                             V
 where  m  = mass of lean reactant  transported to and consumed at the wall,
            per unit area

      Nu  = Musselt number for  mass transfer (Sherwood  number)
        p  = gas density
        V  - diffusion  coefficient
        D  = diameter of  one  channel of monolith bed
        A  = preexponential factor
      AE  =  activation energy
        R  =  universal  gas  constant
        Each of  these  expressions for the mass flux can be shown graphically
by plotting the mass  flux vs.  the mass fraction of lean reactant at the
monolith wall  (i.e.,  plot m vs. fy).   This is shown in Figure 6-3.

                        m-AKwrAE'RTW
      m
                                       m
                                                         Nu    (K  - K)
            Figure 6-3.
                                                       (b)
Mass flux as a function of mass fraction of
lean reactant at the monolith wall.
            6-5

-------
       By combining the two curves on a single graph, solutions to the
simplified equation may be identified as intersections of the two curves.
As shown in Figure 6-4, there are generally three solutions that can exist:
a hot stable solution, an unstable solution, and a cold stable solution.
Only the hot stable solution is of interest here.
       For fixed preheat and inlet composition the parameters defining the
rate of reactant consumption at the wall are fixed.  However, the parameters
defining the ordinate intercept of the reactant transport equation are  func-
tions of configuration and flowrate conditions.  To  study effects of these
conditions on the ability to achieve hot stable solutions, it is  reasonable
to consider the curve of Figure 6-3(a) as fixed, and to then evaluate the
cause and effect of changing the ordinate intercept  of Figure 6-3(b).   The
solid line of Figure 6-4 represents the general case of three solutions,
while the upper dashed line represents the transition to a single cold  stable
solution and  is the blowout condition.  The lower dashed line represents the
transition to a single hot solution.  This represents the maximum ordinate
 intercept to  achieve self-ignition of the catalyst without an auxiliary
 lightoff system.
        To assist  self-ignition and to avoid blowout, the ordinate intercept
Nu(pP/D)K£ should  be minimized.  For fixed entry temperature and  concentra-
tion, this is achieved by
        •   using  large diameter cells in the bed
        •   operating at  a small value of Musselt number
The  use of large  diameter cells is simply achieved,  but operation at a  small
value of Nusselt  number  is more subtle.  The Nusselt number at  the  inlet of
the  catalyst  is of greatest importance if hot  stable operation  is to be
achieved.  In this zone, the Nusselt number increases with increasing ap-
proach  velocity and decreases with the effective thickness of the cell  web.
Thus, blowout of  the catalyst is a consequence of the increase  in Nusselt
number  when the flowrate is increased.  The benefit  of using  large  cells
should  be enhanced by  increasing cell web thickness  at the same time.
                                     6-6

-------
M
                                                                                   Cold stable
                                                                                   solution
      Figure 6-4.   Simplified mass balance solution  for catalytic combustion
                    in  a  monolith bed.

-------
       Changes in the inlet concentration and temperature result in more
complex effects on Figure 6-4.   Of primary significance is the effect on
adiabatic flame temperature, because of its exponential impact.   Graphi-
cally, this causes an increase in the maximum value of the curve shown in
Figure 6-3(a).  To increase the adiabatic flame temperature, either preheat
temperature can be raised or a concentration nearer to the stoichiometric
value can be used for a given fuel/oxidizer combination.
       In a practical combustion system, values of preheat temperature,
composition, and flowrate are usually imposed.   Thus, the most effective
way to achieve stable operation is to use large diameter cells with rela-
tively thick webs at the bed entrance.  However, using large cells through-
out the bed would result in poor surface conversion of combustibles to pro-
ducts.  The amount of surface conversion is directly related to the number
of transfer units in the bed, where the length of each transfer unit is
equal to
                                Pr -Re   n
                                 4 Nu  ' u
where  Pr = Prandtl number
       Re = Reynolds number
       Nu = Nusselt number
       D  = cell diameter
Thus, to get complete conversion on the catalyst surface, it is necessary
to have many transfer units available by minimizing the length of each
transfer unit.  This suggests use of small diameter cells.  The small cells
will also accelerate gas phase reactions, which are helpful for full con-
version.
6.1.2  Conclusions
       As a consequence of these operational fundamentals, it appears that
a catalytic monolith bed used for the purpose of combustion should use large
diameter cells at the front of the bed to prevent blowout, and small diameter
cells at the back of the bed to maximize the number of transfer units in a
given length of bed.  Therefore, for a given catalyst, it was postulated that
superior performance can be obtained by using the catalyst in a graded cell
                                     6-8

-------
configuration, with large cells at the front end, small cells at the back
end, and perhaps one or more intermediate sized cells between.  A complete
discussion of the model used to support these conclusions is given in Sec-
tion 6.2.

6.2    THE PROF-HET COMPUTER CODE
       The PROF-HET computer code models the important physical phenomena
occurring within monolithic catalytic combustors.  An efficient numerical
technique which includes axial and radial heat and mass transport, axial
radiative and conductive wall heat transfer, and heterogeneous surface and
bulk gas phase chemical kinetic reactions has been developed to establish
how performance varies with bed operating and design parameters.  In this
technique, matrix procedures are used to solve the finite difference form of
the governing differential equations.  The axial distribution of both wall
and bulk gas properties, as well as wall temperature, are output by the code,
The solution procedure is reliable and stable for the range of input param-
eters used to date.

6.2.1  Comparison to Existing Models
       The PROF-HET model differs from previous models in that it can handle
the high temperature effects of catalytic combustion, where bed radiative
heat transfer and "flame type" phenomena are important.  Most of the models
constructed to date have focused on catalytic cleanup devices (e.g., automo-
tive catalytic mufflers, and industrial process exhaust cleanup devices for
sludge drying, PVC processing, foodstuff processing, etc.) where the temper-
ature rise due to catalytic oxidation is small compared to that which occurs
during combustion.  For example, Votruba et al. (Reference 6-1) published an
analytical study on the heat and mass transfer in monolithic catalysts.  In
their model, the heat and mass transfer normal to the channel walls was
treated by a transfer coefficient approach where the detailed distribution
of properties across the channels need not be known.  This reduced an essen-
tially two- or three-dimensional (for noncircular channels) problem to one
dimension.  Equations for the variation of wall temperature and gas condi-
tions as a function of distance along the channel were developed.  Calcula-
tions made using this model  are more economical than higher dimensional

                                     6-9

-------
models.  Illustrative prediction'! for catalytic monoliths where the wall
temperature rise was both low and high were given.  These predictions showed
that monolithic structures give more stable operation and less pressure drop
than packed beds.  However, since radiative heat transfer and gas phase re-
actions were ignored and only fully developed heat transfer parameters uti-
lized, the model cannot be accurately applied to catalytic combustors.
       Cerkanowicz, et al. (Reference 6-2) presented results generated using
a catalytic combustor model similar to Votruba's model, except for the in-
clusion of a gas phase chemical reaction.  As with Votruba, only constant
fully  developed heat transfer coefficient values were utilized in the model.
Also,  the gas phase chemical reaction was assumed to be a one-step process.
This approach is useful for illustrative predictions to show global effects
due to gas phase reactions.  However, the method is not sufficiently funda-
mental for exploring the types of "flame" phenomena and pollutant formation
processes which occur in catalytic combustors.
       Young and Finlayson (Reference 6-3) developed one-, two-, and three-
dimensional models for monolithic catalytic converters.  Their two- and
three-dimensional models treated heat and mass transfer perpendicular to
the wall by an orthogonal collocation method which determines the detailed
distribution of properties across the tube.  As in Votruba's model, bed
radiative transfer and gas phase reactions were not included, making appli-
cations to catalytic combustors limited.  Young and Finlayson's predictions
showed that three-dimensional effects associated with peripheral temperature
and concentration variations are not important to overall bed operation.
They also showed that holding transfer coefficients constant over the bed
length is not an adequate approach.
       Heck (Reference 6-4), using a finite difference procedure, compared
one- and two-dimensional catalytic converter model predictions.  Like
Finlayson, Heck concluded that one-dimensional constant heat transfer solu-
tions  were not adequate.  However, he showed that if the transfer coefficients
are allowed to be functions of distance and wall temperature, characteristic
of developing boundary layers, predictions comparable to the two-dimensional
results could be achieved at a fraction of the cost.
                                    6-10

-------
       Following Heck, this study employs a one-dimensional model with trans-
fer coefficients given as functions of distance, wall temperature, initial
gas temperature, and inlet flow conditions.  In addition to including wall
reactions and bed heat conduction, the present model also includes bed radi-
ative heat transfer and gas phase chemical reactions and axial diffusion.

6.2.2  Model Formulation
       The important bed operation characteristic of blowout is controlled
by events occurring near the combustor inlet.  At this location, the bulk
gases are relatively cool and heat release due to gas phase chemical reac-
tions is small.  Heat release in this zone is solely due to wall surface
chemical reactions.  The heat produced by these reactions is carried away by
radial transport of heat to the bulk gases, heat conduction towards the front
of the bed, and radiative transfer towards the front of the bed and into the
upstream reservoir.  Because the thermal boundary layer is developing in this
region, the heat transfer coefficient is large.  Above a certain limiting
mass flow, the heat transfer coefficient for the developing boundary layer
becomes so large that the radial convective heat loss plus the radiative heat
loss exceeds the heat produced by wall reactions.  The wall reactions are
then extinguished and the bed becomes cold.  To model this blowout state,
the following phenomena must be treated:
       t   Heterogeneous surface chemical reactions
       •   Radial heat and mass transport
       •   Axial convection
       t   Axial bed conductive and radiative heat transfer
       Unlike blowout, breakthrough and certain emissions phenomena are con-
trolled by processes which occur away from the front of the bed where the
bulk gases are hot.  Breakthrough occurs when there is insufficient preheat
of the bulk gases by the wall reactions to "light off" gas phase "flame-type"
phenomena.   Since breakthrough is believed to be flamelike in nature, in ad-
dition to the above phenomena, the following phenomena must be observed:
       •   Homogeneous gas phase chemical reactions
       •   Axial gas phase heat and mass transport
                                    6-11

-------
       Since the controlling regions for blowout and breakthrough are spa-
tially separated, two separate but compatible models, optimized for their
respective regions, have been developed to treat blowout and breakthrough
The MET numerical model is employed to determine blowout and the PROF code
is used to predict breakthrough and emissions.  Both codes employ compa-
tible thermal, transport, and chemical  reaction data and formulations.  They
differ primarily in the amount of detail included in their respective regions
of application.

6.2.2.1  HET Blowout Model
       The HET code governing equations are developed by integrating the
steady two-dimensional  gas phase species,  mass, and energy equations across
a plane perpendicular to the axis of a  channel.  This results in a set of
quasi-one-dimensional equations which can  be written in terms of bulk gas
properties and wall heat and mass fluxes.   As previously indicated,  wall
fluxes are treated by a transfer coefficient approach where the fluxes are
directly proportional to bulk and wall  gas states.   The proportionality
factor varies over the bed length and is a function of channel  diameter,
distance down the channel, inlet gas temperature, wall  temperature distri-
bution, and flow Reynolds and Prandtl numbers.  The use of transfer  coeffi-
cients permits the application of efficient one-dimensional  solution proce-
dures to a two-dimensional problem.  Heat  transport in the bed  is determined
by heat conduction in the solid and radiative transfer within the channel
and outside into the upstream and downstream reservoirs.  The radiative
transfer is modeled through a view factor  approach, in which all sections of
the channel are able to radiatively communicate with each other and  with the
upstream and downstream reservoirs.
       The quasi-one-dimensional  governing equations are:
       Species balance in the gas phase
       Species balance at the wall
                                     - \                              <6-2>
                                    6-12

-------
       Energy balance in the gas phase
       Overall  energy  balance
                      m
                                                                        (6-3)
                                                                        (6-4)
Global continuity  has  been  incorporated  into these equations, and the mo-
mentum equation  has  been  replaced by an  assignment of fixed pressure.  This
assignment  is  usually  reasonable for the flow rates and channel dimensions
of  interest.
becomes
       Applying  the  transfer  coefficient approach, wall species flux, Jw^,
                                                                        (6-5)
Similarly, wall  heat  flux, q  ,  including a term due to chemical reaction be-
tween  the bulk and wall  gas,  becomes


                                             m
                                                   w
                                                                        (6-6)
The heat transfer  coefficient  formulation, taken from Kays  (Reference 6-5),
is applicable to circular channels with variable surface temperature, and is
given by:
          "~                   '  "                         2W+
CH ~ RePr
                                                       -X.
                          n  \An
                                    n
16b£ -T -
     n    n
          -Xn2X+
        le  n   +
n   \xn
                                                     n  \ n
where
                                     dT.
                     Y+ = 2x/D   b = __w       T    _
                          RePr  ' D   dx+  '      w   'c
                                                            (6-7)
                                     6-13

-------
and G  and x  are constants and eigenvalues whose magnitudes (given in Ref-
erence 6-5) depend on whether the flow is laminar or turbulent.  This expres-
sion is valid for flows with fully developed velocity profiles and developing
thermal profiles and is assumed to be adequate for the problem of interest.
Entrance effects due to channel web thickness and developing velocity pro-
files are not considered in the code, except to define a limiting heat
transfer coefficient at the channel entrance.  The limiting value is based
on stagnating flow on the channel web.
       The mass transfer coefficient Cmi  is developed from the heat transfer
coefficient by:
                                                                     (6-8)
       Wall reactions are given by
                            w

                                                                     (6-9)
where k
               -E.../RT
                 w
       w  = Awe
oxidizer concentrations at the wall.
               and a and m are arbitrary exponents  on fuel  and
               s at the wall.
Bulk gas phase reactions are given by
                                                                     (6-10)
                                  m
where y. are the stoichiometric coefficients of the reaction m, and Rm
is given by
                 Rm = k
                       fm
                              R
                                                      m
                                     - e
and kf  has the Arrhenius form kf  = aT e  '
      m                         Tm
       The wall radiative heat flux q  at station j is given by:
          qr. = ea
                                       -  K.
                                    k
                                    6-14
- K   T1*
  Kjr2'r2
                                                              (6-11)
                                                                     (6-12)

-------
where  K  is  the  channel  segment  view  factor,  k  denotes  all  other  stations
except j  and  rl,  r2  denote  upstream  and downstream  reservoirs.

       Completing the definition of  the problem, the boundary conditions
f°r  Equations (6-1)  through  (6-4) are:
s = 0        YI = Y0,, h = h0, -"
                                                     = 0,T
                                                  o
                                                          rl
              s = L
                                   dT,
                                     w
                                    ds
                                                       (6-13)
                              '  T
                                 r2
       Applying a straightforward linear finite differencing technique, the
differential Equations 6-1 through 6-4 are reduced to algebraic form.  The
resulting algebraic equations are solved by a Newton-Raphson matrix pro-
cedure which includes a predictor-linearized corrector step.  Very briefly:

       1.   Initial values are guessed for TW at all grid points
       2.   By applying known upstream conditions and the initial guessed
            T 's, grid point values for Y., T, h, Yw , T . h,, etc. are
            w                          i         "i   w   w
            found through Newton-Raphson solution procedures
       3.   Using the derivatives obtained from the solutions at each
            grid point, the rate of change of wall temperatures with
            respect to initially guessed wall temperatures at each
            grid point is constructed
       4.   Assuming the system is linear, corrections to all T  's are
            made by applying the derivatives from step 3
       5.   Using the corrected T 's as new guesses, steps 2 through
            4 are repeated until guessed TW'S equal corrected TW'S

6.2.2.2  PROF Breakthrough and Emissions Model
       The  PRemixed One-dimensional Flame (PROF) code has been described
in detail elsewhere (Reference 6-6).  Very briefly, the two-dimensional
governing equations are integrated across the channel, producing a set of
                                     6-15

-------
quasi-one-dimensional  equations similar to the  HET  model  formulation.  These
equations are:

       Species
                   m
                     _
                      ds
                                    AJ-h-
                                      11
U I  \   p
rlc-  I    WHW
                                     (6-14)

                                     (6-15)
They differ from the HET equations  by the addition  of axial  gas  phase  dif-
fusion, J., and heat conduction £ J.h.  + k (dT/ds)  terms,  where  J.  is  given
by                              ""
^Ii+W*l   ,  M"
ds     M  I ds    i  ds y
                                                                      (6-16)
where the binary diffusion coefficient,  D-j .,  is  replaced  by  the  bifurcation
                                          J
approximation (Reference 6-6)
                                n   =
                                D   -
and
                                      F,F
                                      (6-17)
The development of Equation (6-16)  is  given in  Reference 6-6.
       The inclusion of the axial  diffusion terms  makes  the  PROF model  a
multivariable (Y^T) boundary value problem.  A difference between  the  PROF
and HET models is the assignment of wall  state  (YWi,  Tw) in  the PROF model,
whereas in the HET model the wall  state is calculated as part  of the solution,
For PROF calculations, the wall  state  is  found  by  running the  HET code  for
                                     6-16

-------
the same inlet flow conditions.  The predicted wall state is then input as
a boundary condition into the PROF code.  Additional boundary conditions for
PROF calculations are the initial gas composition, pressure, and temperature;
and the condition of no heat and mass diffusion at the downstream boundary.
       The PROF code can handle many chemical  species, including those which
model detailed combustion and pollutant formation mechanisms.  As discussed
in Reference 6-6, solving the species equations, including chemical  production
terms, R^, is difficult.  The PROF code has been optimized so that chemical
reaction rates -- from nearly equilibrated to inactive -- can be handled
reliably, accurately, and efficiently.  This has been demonstrated by compari-
son of detailed free flame species predictions and data in Reference 6-9.
Also, a favorable comparison in Reference 6-9 of PROF predictions of flame
quench in small diameter tubes and data demonstrates the accuracy of the
code when applied to confined flame problems such as catalytic combustors.
       To reduce the PROF differential equations to algebraic form,  straight-
forward linear finite differencing is used.  The resulting system of simul-
taneous equations are solved by a predictor-linearized corrector solution
procedure which consists of the following steps:
       1.  Initial values are selected for Y., T, h at all  grid points.
           These may be output from a prior run or may be generated  by a
           linear interpolation between initial and guessed final values.
       2.  By applying known upstream conditions and the initial  guessed
           downstream values, grid point values for Y^, h, T, etc.,  are
           found through matrix solution of the equation set
       3.  When the downstream boundary is reached, the no-diffusion bound-
           ary condition is applied.
       4.  Using the derivatives of Y. obtained from chemistry solutions at
           all grid points, the rates of change of all YI-'s with respect to
           initially guessed Y.'s at each grid point are constructed
       5.  Assuming the system is linear, corrections to all YVs are made
           by applying the derivatives from Step 4
                                   6-17

-------
       6.  Using the corrected Y.'s as new guesses, Steps 2 through 5 are
           repeated until the guessed Y.. equals the corrected Y^

6.2.3  Parametric Calculations
       MET code predictions illustrating the effect of bed channel diameter,
preheat gas temperature, mixture ratio, conductivity (and/or void fraction)
and surface activity on the important bed operating characteristics of
blowout are presented.  Also, PROF code predictions which show the effect
of channel diameter and preheat gas temperature on the important bed operat-
ing characteristic of breakthrough are given.  These results have signifi-
cant system design implications for achieving catalytic combustors with high
heat release (high blowout limit) and low emissions (high breakthrough limit),
       Numerical results presented in Figures 6-5 through 6-10 use a surface
reaction rate based on an assumed activation energy and blowout condition
found experimentally at Acurex.  Methane fuel and air are assigned as ini-
tial reactants in all of the calculations.  For methane fuel on a platinum
catalyst, Anderson (Reference 6-7) experimentally found an activation energy
of 96 kj/mole.  This value of activation energy was used in all of the cal-
culations.  The pre-exponential factor was obtained by matching predicted
blowout mass flowrates to experimentally found values for a bed operating
on methane fuel with 0.00635 m diameter channels, initial gas preheat of
672K and 193 percent excess air.  The pre-exponential factor, A , found
                               f       o                       W
using this approach is 6.5 x TO  mole/m /sec for £ and m in Equation (6-9)
assumed to be one and two, respectively.  This pre-exponential rate factor
incorporates effects of surface area, catalyst dispersion, catalyst activity,
etc., and is a global rate for the experimental support/catalyst system.
All calculations presented below are for a pressure of 101.3 KPa.

6.2.3.1  Blowout
       Figure 6-5 gives the HET code predicted distribution of wall and
bulk gas fuel concentration and temperature through a monolith bed.  Initial
gas conditions and bed geometrical  and material properties are listed on the
figure.  These results are typical  of catalytic combustor operation at high
mass throughput and graphically illustrate how reactants are consumed.
                                     6-18

-------
CTl
I
1500
1400

1300



1200


1100
?
2 1000
2
75
o> 900
a
E
0)
*" 800

700
een

	
^.^"" Wall Temperature
•



•
§^^
5
T-
X
I4

Is
•M
- 0)
E
a 2
o

ll
UL
0
^-
. 0.003175 m Channel Diameter
' 0.8 Void Fraction
, ' 0.8661 W/m/K Conductivity
/ 150 Percent Excess Air
/ 0.0929 m Diameter
0.8 Wall Emlssivlty
/ 0.04 kgm/s Flowrate
= 	 --?£- 	 	 	 Bulk Fuel Concentration
^. s
'' \
\
\
\
x x Bulk Temperature
^^ 	 	 — 	
	 	 — ' — . 	 __Wall_Fuel Concentration
, _ . , . ,
                                             Distance (m x 102)
                  Figure 6-5.  Wall and bulk gas temperature and fuel concentration through  bed.

-------
ro
o
                 x   4


                     3
o

                       0.003175 m Channel Diameter
                       0.8 Void Fraction
                       0.8661 W/m/K Conductivity
                       160 Percent Excess Air
                       0.0929 m Diameter
                       0.8 Wall Emlssivity
                                                      0.075 kgm/s Mass Flow
                           0.1    0.2   0.3    0.4    0.5    0.6    0.7
                                                 Distance (m x 102)
                                                      0.8    0.9  1.0
                    Figure 6-6.  Wall fuel volume fraction distributions for several flowrates.

-------
ro
                   o>
0.14


0.12
                   o>  0.10
                   en
                   CO
                   (A
0.08


0.06


0.04


0.02
    0.8 Void Fraction
\   0.8661 W/m/K
 \  193 Percent Excess Air
  \  0.0929 m Diameter
    0.8 Wall Emissivity

    \
     \
      \
        \
          \
                                                                    672K
                       Re = 2000
                                                                   550K Tlnlet
                                 0.2      0.4      0.6      0.8      1.0
                                        Channel Diameter (m x 102)
                                     1.2
                     Figure 6-7.   Blowout mass throughput for various channel  diameters,

-------
ro
ro
                0.14
                0.12
              * 0.10
0)

E
a
.*
             3
             Jc  0.08
             en
             3
             O
I-
(0

-------
   0.3
550K Tln,et
 o
 o
 CO

 1,
 3
 0.
 £
 O)
 3
 O
0.2
    0.1
 to
 (Q
   0.003175 m Diameter Channel
   0.8 Void Fraction
   0.8661 W/m/K Conductivity
   0.929 m Diameter
   0.8 Wall Emissivity
   672K Tln,et
            100    200    300
             Percent Excess Air
Figure 6-9.  Blowout mass throughput for various excess
           air levels.
                       6-23

-------
o
o
o
   14
   12
x
o 10
0)
0)
    Q
    8
1  6
O)
o
a  «
(A
W  ~
0.003175 m Diameter Channel
0.8 Void Fraction
193 Percent Excess Air
0.0929 m Diameter Bed
0.8 Wall Emissivlty
550K Inlet Temperature
           0.2    0.4    0.6   0.8    1.0
                  Conductivity  (W/m/K)
                           1.2   1.4
      Figure 6-10. Blowout mass throughput for various bed
                 conductivities.
                           '6-24

-------
       Close to the channel entrance radial heat transport and radiative
heat losses are large, causing the surface temperature to be much lower
than the adiabatic flame temperature.  At blowout, due to the low wall
temperature, surface reactions become much slower than the radial trans-
port of fuel and oxidizer to the wall, and fuel concentration at the wall
is a substantial fraction of the bulk gas value.  The wall reactions are
controlling heat release in this case and the front of the bed is said to
be kinetically controlled.  Further down the channel, heat losses decrease
and the wall temperature rises.  This drives the wall reaction rate to much
higher levels than the radial mass transport which is now controlling heat
release.  This region of the bed is said to be mass transfer controlled
and any fuel reaching the wall is rapidly consumed giving low values of
fuel concentration at the wall.  As the mass flow through the channel is
increased, the heat transfer coefficient and radial heat transport away
from the wall is increased and a greater portion of the front of the bed
becomes kinetically controlled.  This is illustrated in Figure 6-6 where
wall fuel concentration distributions are given for several bed mass through-
puts.  These results show that the kinetically controlled region spreads
downstream as mass throughput increases.  The wall then becomes cooler and
the surface chemical reactions are extinguished.
        In this study, blowout is defined as the condition where the kineti-
cally controlled region sweeps down the bed and the wall reactions are ex-
tinguished.  It should be noted that the movement of the kinetically
controlled region to locations downstream where the channel flow is more
hydrodynamically developed does not ease the problem of the extinguishing
of wall chemical reactions.  This is due to the thermal boundary layer
initiation point moving concurrently with the kinetically controlled region,
resulting in locally high values of heat transfer coefficient which can
extinguish wall reactions.
       Predictions of blowout mass throughput as a function of channel
diameter are presented in Figure 6-7.  The upper curve in Figure 6-7 is for
a preheat temperature of 672K and the lower curve is for a preheat tempera-
ture of 550K.  The parameters held constant for these calculations are
                                    6-25

-------
listed on the figure.   These include mixture ratio,  flow area,  wall  conduc-
tivity (or product of conductivity and wall  solid cross sectional  area)  and
surface emissivity.  The dashed curve at the top of the figure  is  a  constant
Reynolds number of 2,000 line.   For fully developed pipe flow,  this  curve
represents an approximate upper limit for purely laminar flow.   Between
Reynolds numbers of roughly 2,000 and 10,000 is a transitional  flow  regime
where part of the tube flow is  laminar and part is turbulent.   If the length-
to-diameter ratio of the channel is less than 50, entrance effects and the
transition of laminar to turbulent flow in the developing boundary layer
must be considered.
       The point of transition  to turbulent flow within the channel  depends
on the entering freestream turbulence level, disturbances due  to entrance
geometry, rate of wall heating  and roughness of tube wall.  If the channel
length is short and disturbances due to entrance effects and roughness are
not severe, laminar flow can be maintained within the entire channel for
values of Reynolds numbers based on channel  diameter much above the  fully
developed 2,000 value.  For the geometries and flowrates of interest, fully
developed flow is never achieved in the channel  and Reynolds numbers do  not
exceed the 2,000 limit by a large amount.  Therefore, laminar  flow should
prevail for most of our cases of interest.
       The mass throughput curves in Figure 6-7 show that blowout increases
almost linearly as tube diameter increases for both the 550K and 672K pre-
heat cases.  This is primarily the result of the heat transfer coefficient,
for fixed mass throughput, decreasing with increases in diameter of the cells
and thickness of the web.  Increasing the channel diameter permits more mass
to pass through the channel before blowout occurs.  However, increasing
diameter also decreases the mass transfer coefficient which reduces  fuel con-
version efficiency.  A longer bed is then required to convert all  of the fuel
to combustion products.  Comparison of the two curves shows that preheat has
a very strong influence on blowout.  This is even more dramatically demon-
strated in Figure 6-8 where the channel diameter is held constant at 0.003175
meters along with all the other parameters and gas preheat temperature is
varied.  These predictions show that preheat has a very strong influence on
                                     6-26

-------
blowout with higher preheat producing more than a proportionate increase in
blowout mass throughput.  This is due to the "activation" nature of the wall
reaction rate, which is an exponential function of wall temperature.  These
results indicate that, for maximum heat release, beds should be operated at
as high a temperature as is compatible with the degradation of the catalyst
or is acceptable from an NO emissions point of view.
       Figure 6-9 shows the effect of mixture ratio on blowout.  The upper
curve is for a preheat temperature of 672K and the lower a preheat of 550K.
Numerical values of the other parameters held fixed during the calculations
are listed on the figure.  Both curves show a rapid rise in blowout mass
flowrate as the amount of excess air is decreased.  In Figure 6-9, the param-
eter controlling the high blowout mass throughput rates is surface tempera-
ture.  For low excess air levels and no wall cooling, surface temperatures
(~2200K at 0 percent excess air) are very high, exceeding present bed ma-
terial operating limits.  These high temperatures drive the surface chemi-
cal reaction rates to very high levels which far exceed radial (and radiative)
heat losses under laminar flow conditions.  Large mass flowrates are needed
to produce blowout at these conditions.  Once again, the strong influence of
wall temperature on blowout is evident.
       For the 50 percent excess air, 550K preheat case, the blowout
Reynolds number based on tube diameter is in the transitional flow regime.
To investigate the effect of fully turbulent flow on blowout, the MET tur-
bulent developing boundary layer heat transfer coefficient model was acti-
vated for the 50 percent excess air case.  These predictions show that the
high turbulent heat transfer coefficient forces the blowout mass throughput
down to very low values.  At these low flowrate conditions, laminar flow
would prevail and the fully turbulent flow model is not applicable.  From
these results, it may be conjectured that as mass throughput approaches a
value such that the developing channel boundary layer transists from laminar
to turbulent flow within the channel, the surface reactions in the down-
stream turbulent flow portion of the tube could be extinguished.
       Figure 6-10 gives the effect of bed material conductivity on blowout.
Since the solid cross sectional area enters the governing equations coupled
                                     6-27

-------
with conductivity, the variation of blowout with conductivity can also be
interpreted as blowout variation with solid cross sectional  area for fixed
conductivity.  The blowout trends in Figure 6-10 show that the blowout mass
throughput varies almost linearly with conductivity.  However, at low values
of conductivity, the blowout limit levels off and reaches a constant value
for no wall conductivity.  It should be noted that radiative heat transfer
is included in these calculations and, therefore, the zero conductivity
calculation does not represent adiabatic conditions.
       Figure 6-11 gives the effect of wall activity on blowout.  Variations
in wall activity model the effect of increasing surface area, catalyst load-
ing and dispersion on blowout.  Since the exponential factor in the catalyst
rate expression was held fixed during these calculations, the results repre-
sent a single catalyst whose amount and distribution on a monolith bed has
been varied.  Results in Figure 6-11 show that the effect of surface activity
on blowout is nearly linear.
       The results of the parametric blowout calculations are summarized in
Table 6-1.  These results indicate that for maximum mass throughput, surface

             TABLE 6-1.  EFFECT OF PARAMETER CHANGES ON BLOWOUT
       Parameter
Effect of Increase on Blowout
     Comments
   Channel diameter
   Gas inlet
   temperature
   Initial fuel/air
   mixture ratio
   Conductivity
   Surface activity
    Linear increase
    Exponential increase

    Exponential increase
    in lean systems

    Linear increase
     Linear increase
Same type of
behavior as temper-
ature
No variation in  blow-
out as conductivity
bcomes smaller than
0.2 W/m/K.
temperature should be as high as is compatible with the support/catalyst ma-
terial combination and the channel diameter should be large.  However, large
                                    6-28

-------
                     12
                     10
o
o
o
X
o



O)
                      8
ro
                   a
                   JZ
                   o>
                   !   4

0.003175 m Diameter Channel
0.8 Void Fraction
193 Percent Excess Air
0.0929 m Diameter Bed
0.8 Wall Emissivity
550K  Inlet  Temperature
                              200    400    600    800   1000   1200   1400
                                     Kf (mole/mVs x 104)
                Figure 6-11.  Blowout mass throughput for various surface reaction activities.

-------
diameter channels also have poor fuel conversion performance and long beds
are needed to convert all the fuel to combustion products by wall chemical
reactions.  This dilemma of high mass throughput but poor fuel conversion
for large diameter channels can be solved by adding additional beds behind
the first bed to efficiently convert the remaining fuel.  Blowout should
not be a severe problem for these additional beds because the entering gases
are highly preheated.  The next section addresses the fuel conversion or
breakthrough problem.

6.2.3.2  Breakthrough
       As indicated previously, large diameter channels increase blowout
mass throughput but decrease fuel conversion by wall chemical reactions.
Therefore, small diameter channels should be used to minimize unburned fuel
emissions.  However, blowout, channel mechanical forming, and pressure drop
considerations limit the minimum channel diameter that can be applied.  If
only wall reactions are assumed to occur, complete fuel conversion in chan-
nels of practical size requires long beds.  To minimize bed length, homo-
geneous chemical reactions must be activated to rapidly consume any fuel
remaining in the bed.
       As discussed previously, homogeneous gas phase reduction of fuel is
postulated to be "flame-like" in nature.  The PROF predictions in Figures
6-12 and 6-13 support this postulate by illustrating the importance of up-
stream diffusion of reactive chemical species and heat on homogeneous reac-
tions.  Boundary conditions for these calculations are listed on the figures,
and Table 6-2 gives the elementary chemical kinetic reactions and associated
rates applied in the calculations.  The prediction represented by the dashed
curve in Figure 6-12 includes both diffusion and chemical reactions.  This
prediction shows a rapid decay of fuel concentration, indicative of "flame-
type" phenomena.  In Figure 6-12, predictions represented by the solid line
and circular symbols include only chemical reactions or diffusion, respec-
tively.  These predictions show the less rapid fuel decay characteristic
of wall reactions only.  The closeness of these results shows that axial
diffusion does not significantly impact fuel concentration if homogeneous
                                    6-30

-------
o
CO
0)
     10
       -1
     10
       ,-2
     10
       -3
0)

I   10
       -4
     10
     10'
       -5
Gas Phase Reaction and Axial Diffusion
No Axial Diffusion-Gas Phase Reaction
No Gas Phase Reaction-Axial Diffusion
     0.00325 kg/sec Mass Flow
     0.003175 m Channel Diameter
     550K  Preheat Temperature
     100 Percent Excess Air
         0123456789  10
                    Distance (mx 100)
     Figure 6-12.  Bulk  gas fuel  concentration through bed.
                            6-31

-------
    1600
    1500
    1400
    1300

^  1200
*  1100
 S  1000
1-°*   900
     800
     700
     600
     500
             Gas Phase Reaction and Diffusion
           o No Gas Phase Reaction-Diffusion
             No Diffusion-Gas Phase Reaction
               0.00325 kg/sec Mass Flow
               0.003I75 m Channel Diameter
               550K Preheat Temperature
               100  Percent Excess Air
    1234567
            Distance (mx 100)


Figure 6-13.   Bulk gas temperature through bed.
                  6-32

-------
                              TABLE 6-2.   CH4  COMBUSTION CHEMICAL KINETIC REACTIONS AND RATES
co
co
               KINETIC REACTION DATA

                           OF REACTIONS*  ?fl

                           HEACTION
1
?
3
I*
5
6
7
6
9
10
11
1?
13
It*
:5
16
17
IB
19
20
?1
??
?3
2>t
25
?(,
27
2*
CH«* +OH
CHt* +H
CHI* *0
CHS +0
CH3 +02
CH20+
CH20+OH
CH20+0
CH20+H
CHO +02
CHO +OH
CHO +0
CHO +
CO +OH
CO +0
MO? +0
H02 +OH
HO? +H
H02 +H
H +02
H +02
0 +H2
OH +H2
OH +OH
H +OH
0 +H
H +H
0 +0
+ ••>
+ -->
+ ••>
+ -->
+ -->
+ M •->
* -->
* -->
* -•>
* •->
* ••>
* •->
*« -->
* -->
*H -->
+ -->
+ -->
* -->
+ ••>
+i -->
+ -->
+ -->
+ •->
+ -->
+ M -->
*M -->
*i» -->
** -->
CH3 *H20
CHS +H2
CH3 +OH
CH20+H
CH20+OH
CO +H2
CHO +H20
CHO +OH
CHO *M?
CO *H02
CO +H20
CO +OH
CO *H
C02 +H
C02 +
C? *OH
08 +H20
OH *OH
02 *H2
MO? +
OH +o
OH *H
H?0 +H
H20 +n
H20 +
OH +
H2 +
P2 *
                                                  PRE FXP FACTOR

                                                      .iooo+m
                                                      .POOO+15
                                                      .2000 + 11*
                                                      .35oo+i<»
                                                      .1000*13
                                                      .2000*17
                                                      .2500*1<»
.1700 + 11*
,3000*1<*
.1000*15
.P«»oo*i2
.2000*13
,fi500*12
.3600*19
,?5on*iH
.2500+m
.2500*15
,?500*1<»
,?onn*i6
.?200*l*i
.1700*14
.??no*m
.(.000*13
.7000*20
.U000*19
.?000*20
. 1000*19
               TEMP EXP  ACTIVATION ENERGY
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .500
  .500
  .000
•1.000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
-1.000
•1.000
-1.000
-1.000
 6.0000
11.9000
 6.9000
 3.3000
15.0000
35.0000
 1.0000
  .0000
 3.0000
  .0000
  .0000
  .0000
28.8000
 1.0800
 2.5000
  .0000
  .0000
 2.0000
  .0000
  .8700
16.6000
 9.1600
 5.2000
  ,7flOO
  .0000
  .0000
  .0000
  .0000
                              INDIVIDUAL THIRD BOOT EFFIC.
                                                                                                    H90
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
20.000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
  .000
 .000
 .000
 .000
 ,noo
 .000
 .000
 .000
 .000
 .000
 .000
 .000
'.000
 .030
 .noo
 .000
 .000
 .000
 .000
 .000
 .noo
 .000
 .000
 .000
 .000
 .000
 .000
 .000
 .000

-------
reactions are inactive.  The effect of including both diffusion and chemi-
cal reactions can also be seen in Figure 6-13.  This figure shows that
predictions which include both diffusion and chemical reaction have a rapid
rise in temperature indicative of "flame-type" phenomena whereas other cal-
culations show a much slower rise to the final temperature, indicative of
wall reactions.  The "flame-type" nature of the homogeneous fuel concentra-
tion reduction is further shown in Figure 6-14, which presents detailed
species concentrations through the channel.  From this figure it can be seen
that the rapid decay of CH/j is accompanied by an increase in free radicals
(0 atoms for example) and production of CO and H2.  The CO and H2 are sub-
sequently oxidized to C02 and H20.  This sequence of events is very similar
to those which occur in free methane/air flames, and demonstrates the "flame-
type" nature of the processes occurring within the catalytic combustor.  It
may be concluded that, to accurately predict breakthrough, the analytical
model must include the effects of axial heat and mass diffusion, as well as
homogeneous chemical kinetic reactions.
       PROF code predictions demonstrating the effect of channel diameter
on breakthrough are given in Figure 6-15.  Axial heat and mass diffusion,
as well as chemical kinetic reactions, are included in these and all sub-
sequent calculations.  The boundary conditions for this case are listed on
the figure and chemical kinetic reactions and rates are given in Table 6-2.
Figure 6-15 shows that, as channel diameter is decreased, the rapid fuel
decay region associated with "flame-type" phenomena moves towards the front
of the bed.  This is due to the acceleration of bulk gas heating through
the increase in heat transfer coefficient which accompanies reductions in
channel diameter.  Examining detailed computer printout shows that for this
100-percent excess air case, the  "flame-type" phenomena is initiated at a
channel bulk gas temperature of approximately HOOK.  These results show
that wall reactions play an important role in preheating the gases to tem-
peratures sufficiently high to light off the "flame-type" phenomenon.  The
importance of  gas preheat is shown in Figure 6-16, where predictions of fuel
concentrations for a fixed channel diameter and several inlet bulk gas mix-
ture temperatures are presented.  As can be seen, preheating the inlet gas
to higher temperatures causes the "flame-type" rapid fuel decay region to
                                     6-34

-------
0.00001
                     CO,
                   0.00325 kg/sec Mass Flow
                   0.003175 m Channel Diameter
                   550K Preheat Temperature
                   100 Percent Excess Air
                2345
              Distance (mx 100)
   Figure 6-14.  Detailed species concentrations
               through bed.
                     6-35

-------
   10
     -1
   10
     -2
o
      •>
     "
£  10
0)
O

I  ID"4
0)
3
   10
      -5
   10
      -6
                  0.0064 m
           0.00325  kg/s  Mass Flow
           550K Preheat Temperature
           100 Percent Excess Air
                       0.0046 m
0.0032 m
       01234567

           Distance (mx 100)
  Figure 6-15.  Effect of channel diameter on
              breakthrough.
                   6-36

-------
   10'2
o
75
Z  10'3
o
o
0)
3
U.
   10'6
                   0.00325 kg/s Mass Flow
                   100 Percent Excess Air
                   0.0046 m Channel Diameter
                    Tjn,et - 530K
           12345
             Distance (mx 100)
Figure 6-16.
            Effect of gas preheat temperature on
            breakthrough.
                     6-37

-------
approach the front of the bed.   For these cases the "flame-type" phenomenon
is initiated at a bulk gas temperature of approximately HOOK.
       In summary, the parametric breakthrough calculations show that initia-
tion of "flame-type" phenomena  in catalytic combustors requires high channel
bulk gas temperatures.  These temperatures can be achieved by a combination
of wall reaction heating, which is a function of channel  diameter, and/or
inlet gas preheat.  Once the "flame-type" phenomena are active, rapid decay
of fuel and fuel fragments, characteristic of free flame behavior, is achieved,
If "flame-type" phenomena are active, only short bed lengths are needed to
reduce unburned fuel and fuel  fragments to extremely low concentrations.

6.3    CONCLUSIONS
       The PROF and MET codes have been used to characterize catalytic com-
bustor performance.  The calculations are in qualitative agreement with
experimental results, but a detailed comparison at matching conditions has
not been attempted.  HET code predictions indicate that blowout mass through-
put increases roughly linearly with increases in channel  diameter, conduc-
tivity, and catalyst/support surface activity.  Also, blowout increases
roughly exponentially with increases in inlet mixture preheat temperature
and fuel/air ratio for lean operation.  Therefore, for maximum catalytic
combustor mass throughput, surface temperature should be as high as is com-
patible with the support/catalyst material combination and channel diameter
should be large.  Maximum channel diameter, however, is limited by fuel con-
version requirements.
       PROF code predictions show that homogeneous chemical kinetic phe-
nomena in catalytic combustors are "flame-like" in nature and proper model-
ing of this effect requires treatment of axial heat and mass diffusion as
well as homogeneous chemical kinetic reactions.  Predictions which include
axial diffusion indicate that the rapid decay of fuel, associated with
"flame-type" phenomena, moves towards the front of the bed as channel diam-
eter is decreased and initial preheat temperature is increased.  This occurs
with the wall at an equilibrium condition and the surface at the adiabatic
flame temperature, as long as the flow conditions are not near blowout.
                                      6-38

-------
These results show that high gas temperatures, produced by either wall
reaction heating or high preheat, are needed to "light off" the "flame-
type" phenomena in catalytic combustors.

       Predictions suggest that a catalytic combustor system which has high
mass throughput (high blowout limit) and low emissions (no breakthrough)
could be constructed by joining two or more bed segments in series.  The
first segment would have channels large enough to prevent blowout and yet
small enough to convert sufficient fuel to meet the preheat/blowout re-
quirement of the second bed segment.  The second segment would have smaller
diameter channels to convert more of the fuel  to products and further heat
the gases.  The last segment would have very small  diameter channels and the
entering gas preheat would be sufficient to "light  off"  homogeneous "flame-
type" phenomena.  Any fuel remaining in this segment would be rapidly con-
sumed by homogeneous reactions.  This system design, called the graded cell
concept, is described in detail in Section 8.   As described in Reference
6-8, tests have shown this system to have very high mass throughputs and
heat release while maintaining very low unburned hydrocarbon, CO, and NOX
emissions.

6.4    RECOMMENDATIONS
       The PROF-HET catalytic combustion code has been used to characterize
blowout and breakthrough phenomena.  Further use of the code to aid in
catalytic combustor system design would be beneficial.  Four recommended
tasks are described below.

6.4.1  Graded Cell Catalyst Optimization Maps
       The length and stability of graded cell catalytic combustors can be
designed by matching the fuel conversion and preheat requirements for all
combustor segments.  The optimization procedure can be developed by prepar-
ing maps of fuel conversion, pressure, and bulk gas temperature as a func-
tion of combustor length for a variety of channel diameters and preheat
temperatures.  Blowout mass flowrate limits for these parameters should
also be noted.  Performing this procedure would result in a combustor design
of two or more segments having optimal stability and minimal pressure drop.
                                    6-39

-------
6.4.2  Breakthrough Analysis
       Calculations have shown that the initiation of gas phase reactions
for a given mixture ratio and fuel occurs at a relatively constant bulk gas
temperature, independent of channel diameter and initial preheat temperature.
It is possible that the homogeneous phenomena behave as a free flame.  Exist-
ing calculations should be further examined and additional calculations
made to establish the nature of the initiation of gas phase reactions in a
catalytic combustor.  Results would be compared to free flame data, and
determine if gas phase reaction initiation is similar to free flame behavior.
If this is the case, it may be possible to apply free flame results to
catalytic combustor gas phase phenomena, and to describe the initiation of
gas phase reactions in terms of fuel  type, initial  mixture ratio, and bulk
gas temperature alone.

6.4.3  NO  Emission Characteristics of Catalytic Combustors
         X
       Experimental results reported in Section 8 show that the presence
of a surface markedly reduces NO  emissions.  It is currently unknown how
                                X
much NO  is formed within the combustor and how much is formed downstream
       /\
of the combustor in gas phase reactions.  To gain insight into the potential
NO  formation mechanisms within the combustor, a NO  formation model
  X                                                X
(Zeldovich) which has been successfully applied to free flames should be
applied to catalytic combustors.  Wall conditions would be varied from heat
loss only to full equilibrium conditions.  These calculations would then
be compared qualitatively to experimental results for catalytic combustors
and free flames to determine if NOX formation processes within free flames
are similar to those within catalytic combustors, and to determine how much
NOX formation takes place within the combustor.

6.4.4  Effect of Transition on Blowout
       Catalytic combustors for gas turbine (and some boiler) applications
will  operate near the laminar/turbulent transition flowrate.  The transition
to turbulent flow can significantly increase the potential for blowout.
Existing information on transitional  flow heat and mass transfer should be
included in the code to enhance its use in practical system design work.
                                      6-40

-------
                             LIST OF SYMBOLS




A          cross sectional area



C^         Mussel t number divided by Reynolds and Prandtl numbers



Cm         dimensionless mass transfer coefficient defined by Equation (6-8)



C          specific heat



GW         circumference of bounding tube



D . -        binary diffusion coefficient
  ' 0


D          diffusion constant defined by Equation (17)



E          activation energy for kinetic reaction



F.         diffusion factor of species i



h          enthalpy



J          species flux



k          thermal conductivity



K          channel segment radiative heat transfer view factors



K          equilibrium constant



Le         Lewis number



m          mass rate of gas



M          molecular weight



p          pressure



Pr         Prandtl number



q          heat flux



R          gas constant



Re         Reynolds number



s          distance along flame axis



T          temperature



W          chemical production rate
                                    6-41

-------
                        LIST OF SYMBOLS (Concluded)
X          mole fraction
Y          mass fraction
e          wall emissivity
p          density
a          Stefan-Boltzmann constant
Superscripts
P          reaction products
R          reaction reactants
Subscripts
i          denotes species
m          denotes reaction
o          inlet conditions
r          radiation
rl         upstream reservoir
r2         downstream reservoir
s          solid bed material
w          wall
                                    6-42

-------
                                REFERENCES


6-1.   Votruba, J., Sinkule, J., Hlavacek, V-  and Skrivanek, J., "Heat and
       Mass Transfer in Monolithic Honeycomb Catalysts-I.",  Chemical  Engi-
       neering Science, 1975, Vol. 30, pp. 117-123, Pergamon Press,
       Great Britain.

6-2.   Cerkanowicz, A. E., Cole, R. B. and Stevens, J.  G.  "Catalytic  Com-
       bustion Modeling; Comparisons with Experimental  Data," ASME paper
       77-GT-85, presented at the ASME Gas Turbine Conference, Philadelphia,
       Pennsylvania, March 27-31, 1977.

6-3.   Young, L. C., and Finlayson, B. A., "Mathematical  Models of the
       Monolith Catalytic Converter; Part II.   Application to Automobile
       Exhaust," AIChE Journal, Vol. 22, No. 2, pp. 343-353, March 1976.

6-4.   Heck, R. H., Wei, J., and Katzer, J. R., "Mathematical Modeling of
       Monolithic Catalysts," AIChE Journal, Vol. 22, No.  3, pp. 477-484.

6-5.   Kays, W. M., Convective Heat and Mass Transfer,  McGraw Hill, New
       York, 1966.

6-6.   Kelly, J. T., and Kendall, R. M., "Premixed One-Dimensional  Flame
       (PROF) Code Development and Application," Proceedings of the 2nd EPA
       Stationary Source Combustion Symposium, Volume IV,  EPA-600/7-77-073d,
       July 1977.

6-7.   Anderson, R. B., Stein, K. C., Feenan,  J. J. and Hofer, L.J.E.,
       "Catalytic Oxidation of Methane," Industrial and Engineering Chemistry,
       Vol. 53, No. 10, pp. 809-812, October 1961.

6-8.   Kesselring, J. P., Krill, W. V., and Kendall, R. M.,  "Design Criteria
       for Stationary Source Catalytic Combustors," Hestern  States Section/
       The Combustion Institute Fall Meeting on Catalytic  and Fluidized Bed
       Combustion, Paper Number 77-32, Stanford, California, 17-18 October
       1977.

6-9.   Kendall, R. M. and Kelly, J. T., "Premixed One-Dimensional  Flame Code
       (PROF) - Its Formulation, Manipulation, and Evaluation," Aerotherm
       Report TR-75-158, July 1975.
                                      6-43.

-------
                                 SECTION 7
                           CATALYST SCREENING TESTS

7.1    GENERAL CONSIDERATIONS
       A series of catalyst combustion tests were performed at the Jet
Propulsion Laboratory (JPL) in Pasadena, California, to identify those
catalysts which are most suitable for stationary combustion system develop-
ment.  A suitable combustion catalyst has the following characteristics:
       •    Low ignition temperature (both initially and at restart
            conditions)
       •    Low preheat requirements for sustained combustion
       •    Combustion uniformity throughout the catalyst bed for a
            variety of test conditions
       t    High heat release capability
       •    High combustion efficiency
       •    Low pollutant emissions
       t    High temperature operation capabilities (material-limited)
       t    Operational  with a variety of fuels, both gaseous and
            liquid
       •    Long life
       Catalysts were selected based on the review reported in Section 4.
The JPL test series identified catalyst properties which are important for
each of the above combustion characteristics.   The following subsections
discuss the program approach to combustion testing, the actual test data
that was obtained, and conclusions and recommendations for further system
                                     7-1

-------
development.   Pre- and post-test catalyst measurements (as described in
Section 5) were used to support combustion findings.

7.2    CATALYST TEST MATRIX
       Combustion screening was preceded by the development of a test
matrix of catalyst models.   Each model  is made up of  the support, washcoat,
and catalyst materials which became the main variables in the matrix.  More
specifically, the matrix was developed  to:
       •    Identify appropriate ceramic support materials
       •    Compare washcoat materials  and application techniques
       •    Investigate catalyst types  and the effects of loading
       t    Verify Acurex catalyst coating techniques
       •    Investigate the effects of  catalyst bed geometry
The test matrix, therefore, provided a  systematic investigation of each of
these combustion-related variables.
       The matrix of models tested at JPL is shown in  Table 7-1.  A summary
of pre- and post-test surface area and  dispersion measurements is presented
in Table 7-2.  Other tests originally existed in the  matrix but were
eliminated as test data was accumulated.  A total of  22 models were tested
in the screening program.  The catalyst materials and  the purpose of each
test model are also listed in Table 7-1.
       Test procedures were based on a  review of catalyst materials and
their expected performance under combustion conditions.  The result was a
10-point test procedure (shown in Table 7-3) which would give primary data
on ignition temperature, maximum heat release, uniformity, efficiency, and
emissions as well as secondary data on  temperature, fuel, and lifetime
capabilities.

7.3    JPL TEST FACILITY
       The JPL patio test stand (shown  in Figure 7-1)  includes air supply
and fuel feed systems and a vertical quartz test chamber.  The test stand
is supported by a control console, full instrumentation and emission
                                     7-2

-------
                                                        TABLE 7-1.   JPL TEST MODEL SUMMARY
Sample
No.
JPL-001



JPL-002

JPL -003

JPL-004


JPL-004X

JPL-005

JPL-005X

JPL-006


JPL-006X

JPL-007
T.C.
Instr.
30



9

0

30


10

30

21

21


21

21
Substrate
Manuf.
G-R



G-R

G-R

G-R


G-R

Corn

Corn

Corn


Corn

G-R
Type
Cord



Mull

Mull

Mull


Mull

Cord

Cord

Cord


Cord

Mull
Washcoat
Manuf.
0-C



0-C

0-C

0-C


0-C

0-C

0-C

0-C


0-C

0-C
Type
10 Wt %
G-Alumina


10 Wt %
G-Alumina
10 Wt %
G-Alumina
10 Wt %
G-Alumina

10 Wt %
G-Alumina
10 Wt %
G-Alumina
10 Wt %
G-Alumina
10 Wt %
G-Alumina

10 Wt %
G-Alumina
10 Wt %
Catalyst
Manuf.
0-C



0-C

0-C

0-C


0-C

0-C

0-C

Acur


Acur

Acur
Type
PT



PT

PT

PT


PT

PT

PT

PT


PT

PT
Weight
%
.28



.22

.22

.22


.22

.30

.30

.31


.31

.22
Grams
.879



.957.

.957

.957


.957

.975

.975

.975


.974

.957
Dates
Tested
11/06 -
11/14/75


10/20 -
10/29/75
11/03 -
11/06/75
12/04 -
12/12/75

01/22 -
01/28/76
12/15 -
01/08/76
01/29 -
01/30/76
01/15 -
01/21/76

02/06 -
02/09/76
02/11 -
Pres
Atm.
1.0



1.0

1.0

1.0


1.0

1.0

1.0

1.0


1.0

1.0

Fuel
M



M

M

M


M

M

M

M


M

M
Mono
No.
015



007

036

038


037

054

055

049


052

018

Purpose of Test
GR(001)/Corning (Baseline)
Comparison (Check with Base-
line, 10% Difference in PT
Loading)
Facility Checkout

Facility Checkout

Mullite/Cordierite Comparison
(Check with 001 and Baseline,
Note PT Loading Difference)
Rerun of JPL-004

Baseline

Rerun of JPL-005

Verify Acurex Coating
Technique (Cordierite),
Compare to Baseline
Rerun of JPL-006

Verify Acurex Coating
JPL-008      21       G-R    Mull     M-B    10 Wt %      Acur    PT      .23    .957   02/23 -     1.0     M
                                           G-Alumina                                02/27/76


JPL-009      21       DPnt   Alum    0-C    10 Wt %      0-C      PT      .30    .975   03/18-     1.0     M
                                           G-Alumina                                03/19/76


JPLrOlO      21       G-R    Cord    0-C    10 Wt %      Acur    PT      .73  2.320   02/02 -     1.0     M
                                           G- Alumina                               02/04/76
      Technique (Mullite),
      Compare to JPL-004

045   Compare Washcoating Techniques
      (Compare 008-M-B, 007-0-C)
      Compare to JPL-004

072   Effect of Increased Cell Size
      (1/4 in Cell), Compare to
      Baseline

012   Sensitivity to Increased PT
      Loading, Compare to 001
      Compare to Baseline

-------
                                                  TABLE 7-1.  JPL TEST MODEL SUMMARY (Concluded)
Sample
No.
JPL-010X
JPL-010P
JPL-011
JPL-012
JPL-013
JPL-016
JPL-019
JPL-021
T.C.
Instr.
12
9
12
12
12
9
9
8
Substrate
Manuf.
G-R
Corn
Corn
Corn
Corn
Corn
DPnt
DPnt
Type
Cord
Cord
Cord
Cord
Cord
Cord
Alum
Alum
Washcoat
Manuf.
0-C
0-C
0-C
0-C
0-C
0-C
DPnt
0-C
Type
10 Wt %
G-Alumina
10 Wt %
G-Alumina
10 Wt %
G-Alumina
10 Wt ^
G-Alumina
10 Wt %
G-Alumina
10 Wt %
G-Alumina
CE-Stable
A-Alumina
7 Wt %
Catalyst
Manuf.
Acur
Acur
Acur
Acur
Acur
Acur
Acur
Acur
Type
PT
PT
PT
PT
PT
PT/PD
2:1
PT
PT
Weight
%
.75
.72
.29
.30
.30
1.00
vars.
5.27
Grams
2.426
2.397
.975
.975
.975
—
vars.
5.003
Dates
Tested
02/28 -
03/10/76
04/09 -
04/23/76
03/16 -
03/25/76
03/11/76
03/15/76
04/13 -
04/15/76
09/06 -
10/21/76
05/05 -
Pres
Atm.
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
Fuel
M
M,P
M,P
M
M
M
IJMe
M
Mono
No.
009
103
051
050
048
107
vars.
125
Purpose of Test
Rerun of JPL-010
To Establish Propane Operating
Procedure
Effect of H2S Platinizing
Technique, Compare to 005
Effect of Presintered Washcoat
(1000 C), Compare to 006
Effect of Sintering Washcoat-
Platinum, Compare to 006
Multimetallics (PT/PD) Cerium
Stabilized Catalyst
Graded cell catalyst
One Inch Long Torvex
JPL-022
12
                                           G-Alumina
                                                                       05/13/76
                                                                                               1/4" Cells
                                                                                               (5/13 Tests with 2  - 1"
                                                                                                       /Cordierite Segments
DPnt   Alum    DPnt   A-Alumina    DPnt    PT     .00   .000   03/12  -      1.0
                                         (Stab)                 03/15/76
076   Single Metal (DuPont
      Stabilized PT)

-------
TABLE 7-2.  SUMMARY OF CATALYST CHARACTERIZATION
            RESULTS FOR SCREENING CATALYSTS
Model
Number
001
002
003
004
004X
005
005X
006
006X
007
008
009
010
01 OX
01 OP
on
012
013
016
019
021
022
Dispersion (%)
Pre-Test Post-Test
--
--
--
--
—
—
40.69 0
60.33
—
--
—
31.1
--
94.1
--
70.48
49.0
0.7
--
1.1-4.4
2.6
56.1
0.6
—
0
0.5
1.6
0.8
.7-1.3
—
0.6
0
--
--
0.4
0.07
—
0.75
2.4
--
—
—
—
6.7
Surface
Pre-Test
—
—
--
--
—
--
10.59
8.40
—
9.58
13.14
6.69
7.8
7.92
10.87
6.87
1.33
0.93
6.39
1.10-3.20
4.68
10.4
Area (m2/q)
Post-Test
0.03
—
0.70
0.64
0.73
0.625
0.81
0.61
0.65
0.69
1.54
--
0.61
0.41
1.05
0.925
0.56
0.77
2.08
—
0.59
0.62
                      7-5

-------
                       TABLE 7-3.  JPL TEST PROCEDURE
       a.   Start and record light-off temperature of the catalyst at 35
            percent T.A.
       b.   Stabilize reaction at 50 percent T.A. and record conditions.
       c.   Stabilize reaction at stoichiometric and record conditions.
       d.   Stabilize reaction at 150 percent T.A. and record conditions.
       e.   Stabilize reaction at 200 percent T.A., if possible, and
            record conditions.
       f.   Stabilize reaction at 250 percent T.A., if possible, and
            record conditions.
       g.   Stabilize reaction at 300 percent T.A., if possible, and
            record conditions.
       h.   Return to 200 percent T.A., determine maximum throughput
            conditions and record, if possible.
       i.   Return to 50 percent T.A., and determine maximum throughput
            conditions and record.
       j.   Determine change in light-off temperature after testing.
Throughout testing, the following conditions were set:
       •    Try to maintain the reaction on the front face of the catalyst
            or within 0.0635m thereof.
       t    Try to maintain a maximum reaction temperature in the catalyst
            within 14K (25°F) of 1367K (2000°F).
       •    Conduct tests at a minimum of 21.1 MJ/hr (20,000 Btu/hr)  and
            maximum of 105.6 MJ/hr (100,000 Btu/hr)  heat release.
                                     7-6

-------
I
—I
                                    VERTICAL
                                     QUARTZ
                                     REACTOR
VERTICAL
CATALYTIC
 REACTOR
HORIZONTAL
 CATALYTIC
  REACTOR
                                                                                                       AIR
                                                                                                   COMPRESSOR
                                        Figure 7-1.  JPL patio test facility.

-------
measurement equipment, and data handling and monitoring systems located in-
doors.
       The JPL facility operates at near-atmospheric pressure only.  The air
compressor is capable of delivering 4.96 m3/min (175 SCFM) of air,and is
metered (by rotameter), valved, and passed to the system preheaters.
The inlet air can be heated either electrically (to 81 IK) or by tandem gas-
fired heaters (to 922K).
       Both liquid and gaseous fuels can be used in the JPL patio test
facility.  The gaseous fuels are supplied either as pipe line natural gas
(boosted in pressure by an in-line gas compressor) or as manifolded bottles
for methane or propane.  Liquid fuels (distillate oil and methanol) are
supplied in 55-gallon drums.  The fuel is extracted from the drum by an
aircraft fuel pump fitted with a return line for bypassing the excess flow-
rate.  Both gaseous and liquid fuels are metered, throttled, and injected
directly into the heated air stream without further handling.
       All catalyst beds tested at JPL were instrumented with in-depth
thermocouples at Acurex.  Normally, three thermocouples were placed at
varying pre-determined locations within a single channel, and the channel
was then sealed at both ends.  The thermocouples then read the ceramic wall
temperature.  As indicated in Table 7-1, between 8 and  30 thermocouples were
placed in each monolith.  Figure 7-2 shows Test Model 001 with 30 in-depth
thermocouples.  The test models were placed in the vertical quartz reactor
at JPL.
       The quartz reactor test section with monolithic  catalyst bed in
place is shown in Figure 7-3.  The fuel/air mixture enters vertically at the
bottom of the quartz tube, passing through several screens to promote mixing.
The outer surface of the reactor is insulated (to maintain adiabatic com-
bustion), and the exit end is capped to channel the exhaust gases away from
the reactor.  In-bed thermocouple leads exit at the downstream end of the
reactor.  An additional probe for exhaust gas sampling is inserted from the
downstream opening.
       The JPL test stand is instrumented for temperatures, pressures, and
flowrates.  Orifice meters are connected directly into the data acquisition
                                    7-8

-------
Figure 7-2.   Model  JPL-001,  platinized cordierite with 30 thermocouples
             placed in monolith.

                                7-9

-------
Figure 7-3.   Quartz reactor with monolithic catalyst bed.



                              7-10

-------
system to record the fuel and air flow rates automatically.  Rotameters are
used to permit local observation and flowrate control.  Temperature measure-
ments of the gas stream are made both upstream and downstream of the catalyst
bed in addition to the in-bed measurements mentioned above.
       Exhaust gas analysis consists of continuous analyzers for CO, C02,
C^j H£, UHC, and NO    Heated sample lines convey all gases to the respec-
tive analyzers.  Table 7-4 gives further information on the JPL emissions
equipment.  The combined control console and instrumentation readout
equipment is shown in Figure 7-4.  Data can be recorded by strip chart or
magnetic tape for computer analysis.
                    TABLE 7-4.  JPL EMISSIONS EQUIPMENT
Component
Oxygen
Carbon dioxide
Carbon monoxide
Unburned hydro-
carbon
Nitric oxide
Hydrogen
Analyzer
Paramagnetic
Nondispersive
infrared
Gas chromatograph
Nondispersive
infrared
Gas chromatograph
Flame ionization
Gas chromatograph
Chemiluminescent
Gas chromatograph
Range(s)
0-25%
0-10%
0-20%
0-10%
0-250 ppm
0-1000 ppm
0-30%
0-10 ppm
0-5000 ppm
0-10%
0-1 ppm
0-10,000 ppm
0-50%
Fuel
Applicability
Rich Lean
yes yes
yes yes
yes yes
no yes
yes no
no yes
yes no
yes yes
yes no
                                   7-11

-------
                  •
XJ


rv
                                      r—Sr-t
                                            (    ^T= ^= 	
                                                           I
                                                 '"MM, *

                                              "•""",,,/
                                    -  -J  . ,       I. I I n I
                                     - •   '    '••",,.,,
                           Figure 7-4.  JPL patio  test  facility control console.

-------
7.4    TEST DATA SUMMARY
       The test data presented in this section has been prepared to facilitate
comparison between test models.  A table of test data points (Appendix A)
and two plots are shown  for each model.  Data for models JPL-001,  -002,
-003, -004, and -005 are not included since they were used almost solely
for facility checkout and test procedure verification.  The two data plots
show preheat requirements and space velocity (indication of maximum heat
release) capabilities as a function of stoichiometry (percent theoretical
air).  In some instances, additional data showing bed temperature distri-
butions, emissions, and bed degradation with time are also presented.

Test Models JPL-Q04X. 005X. 006. and 006X
       Models 004X, 005X, 006, and 006X were tested as baseline cases for
future test models and to verify Acurex catalyst application techniques.
Photographs of models 005 and 006 are shown in Figures 7-5 and 7-6.  Model
006X was a rerun of 006 to verify past results.   Monolith 004X used a mullite
support while 005X, 006, and 006X were cordierite and had a somewhat higher
platinum loading than 004X.  Table A-l summarizes the test points achieved.
The tests determined two catalyst characteristics.   One is the maximum
throughput for a given air/fuel ratio (50 percent to 200 percent theoretical
air), and the other is the minimum preheat needed to maintain a uniform bed
reaction for various air/fuel ratios (50 percent to 300 percent T.A.).
       All  test models were tested with methane  fuel, and test model  006 was
the first_of the four to be tested.   Lightoff at 35 percent T.A.  was  accomp-
lished at 642K (695°F).  The lightoff temperature increased to 672K to 683K
(750°F to 770°F) when a second lightoff was performed following some  testing,
illustrating the initial degradation of a virgin catalyst.
       Light-off temperatures for JPL-004X and -005X were 650K (710°F) and
661K (730°F), respectively.  These correspondingly degraded to lightoff
temperatures of 714K (825°F) and 706K (810°F) after finishing the first run
on each.  While lightoff on virgin catalysts takes place at the front face,
subsequent starting may initially take place further into the catalyst.
The reaction zone then normally moves back to the front face.  This was
observed in JPL-004X where the mid-bed temperature rose at lightoff on the
                                   7-13

-------
I
4^
                             Washcoated
                            10 Wt % Alo
                                                    UO3
                                                     Platinized
                                                    0.22 Wt % Pt
Figure 7-5.  Model 004 - mull ite/al umina/pl at i
                                                                               num.

-------
en
                                           Catalyst coating 0 0.2 Wt % Pt
                                             circular cell geometry
                                  cell  diameter (washed and coated) = 0.058 inch
                               Figure  7-6.   Model  006 — cordierite/alumina/plati
num.

-------
second run.   When the reaction left the front face of JPL-004X, two-thirds
of the face darkened first, leaving a single 2 cm circular spot showing
activity within that area, and the remaining one-third of the face was also
still active.  Breakthrough occurred off-center near one side the first
time but near an adjacent side at approximately 90° when breakthrough
occurred in the second run.  If the temperatures are maintained near 1367K
(2000°F) and breakthrough is controlled, however, the section of the bed where
breakthrough first occurs will also normally be the area in which subsequent
breakthrough occurs.  The areas where breakthrough occurs will be in the
areas where the front face loses the reaction first and have varied from
near center!ine to near the outside edge for this catalyst.
       Minimum preheat temperatures were found throughout a range of theo-
rectical air from 50 to 500 percent, rising from 478K (400°F) to over 811K
(1000°F) as conditions went from fuel-rich to fuel-lean.  Nitrogen diluent
was used for operation near 100 percent theoretical air.  Maximum space
velocity was determined in a separate run.  Figure 7-7b shows that it was
initiated at a heat release rate of 21.1 MJ/hr (20,000 Btu/hr).  (See
Figures 7-7 through 7-10).  Data points were then obtained at 27.9, 34.8,
46.5 and 58.1 MJ/hr.  These were all at 50 percent theoretical air.  Further
data was taken at a heat release of 32.5 MJ/hr and breakthrough was found
to occur at a space velocity of about 48,000 hr'1.
       Returning to Figure 7-7a reveals another interesting phenomenon.
The data points of the second run show that a preheat temperature of 724K
(844°F) was required for lightoff.  This is an increase of more than 56K
beyond the preheat temperature required for the first run.  This indicates
that a degradation of the catalyst occurred during the first run.  The
third run was then performed which required an additional 39K preheat
(to 763K) for lightoff.
                                     7-16

-------
900
800-
700-
600-
500 -
400-
300-
        01
        t-
i.
o>
        ID

        01
           1200
           1000
    800
            600
            400
            200



£
21.1 MJ/hr ' ' 1
(20,000 Btu/hr) y ~ 450« TA |
A ^"^ / u •
vT X
X^6* Breakthrough 1
/» XjO _^_ \
* -*^"^
/ '/>

-vr- TT
N
1
- 500X TA

$ A/V
!• 	 	 	 / /•* .,.,._ /7\ 1-Lt run
...... • . / * ^j iat run . .
/ ^
II / />
OX A / /
1 1 /
. 	 j i
IA x
^ /
igj /
6 y
° j

Q 2nd run
^ 3rd run
1 transient














•














                                          100                        200


                                              Theoretical air, percent
                                                                                      300
              Figure  7-7a.  Screening data, JPL-004X-preheat temperature  (methane/air).

-------
co
                  GOOOO
                  50000  -
                  40000
                  30000
               01
               u
               «J
               a.
                  20000
                  10000
46.5


34.8

27.9'
*Q


.—or
                                                      .  21.1 MJ/hr
                                                     '  (20,000 Btu/hr)
M
71
                             21.1
O  1st run

Q  2nd run

A  3rd run

 T  tra slent
                                                   100                        200
                                                      Theoretical  air, percent
                                                                           300
                                        Figure 7-7b.   Screening data,  JPL-004X  -  space  velocity
                                                        (methane/air).

-------
UD
           900-
           800-
           700-
           600-
           500-
           400 -
           300 -
                     1200
                     1000
   800
3

fe  600
                      400
                       200

jJQ 	
r
\
a '
|!
1 !
M


21.1 MJ/hr
(20 000 Btu/hr)
--•%---— o 	

O 1st
EJ 2nd
T tra




-^Brea k th rough

run
run
nslent



                                                   100                        200
                                                        Theoretical air, percent
                                                                                   300
                     Figure  7-8a.  Screening data, JPL-005X-preheat  temperature  (methane/air).

-------
ro
o
60000 -
50000
dnnnn .
.e
J?
u
o
•— innnn -
>
HI
(j
m
a.
'
34.8-^ /
21.1-^
10


^^____ 	 Rreakthrough
	 XD 	 -^
^--o"""'
x^r
/
$



i 2(





0 30
                                                            Theoretical air, percent
                                            Figure 7-8b.   Screening data, JPL-005X — space velocity

                                                           (methane/air).

-------
 i
ro
           900 -
          800 -
          700-
600 -
          500-
          400-
          300-
                   s
        ID
        01
        j=
        01
        L.
        Q.
                      1200
                       1000
             800
             600
                        400
                        200


rjt
f ^
\ Tl
: /v
K
0 d

21.1 MJ/hr
(20,000 Btu/hr)
1^ Incipient 1
P^ breakthrough ^
mJjP 	 Q 	 -Q
iSt/f7i
S O 1st
Q 2nd
T. tra





— ___
run
run
nslent
^
.

                                                    100                        200

                                                         Theoretical air,  percent
                                                                                               300
                      Figure  7-9a.   Screening  data, JPL-006-preheat temperature  (methane/air).

-------
-4
 I
ro
ro
01
u

-------
i
ro
CO
            900-
            800-
            700-
            600-
            500-
            400-
            300-
                        1200-
                        1000-
                         800
pera


cr>
o
o
                     4->
                     ro

                         400
                         200-
                                                                  JEL
                                            88.2 MJ/hr
Q 46.3 MJ/hr   £>

                21
                (20,000 Btu/hr)
                                    a
                                                   /
                                                     / 21.1 MJ/hr
                                                     '
        /-
                                                                                10.6 MJ/hr
O 1st run, preheat

£ 1st run, bed lightoff

0 2nd run, preheat

(I 2nd run, bed lightoff
                                                       100                         200

                                                           Theoretical air, percent
                                                                       \
                                                                                                                     I TA = 370.
                                                                      300
                       Figure 7-10a.   Screening  data, JPL-006X-preheat temperature  (methane/air).

-------
                                                    JPL-006X
•-J

ro
o
(O
0.
t/1




50,000 ~



10,000


•}

.30,000 '
>
>
i
L
)
20,000

10,000 -


Heat release
MJ/hr
88.2 g
1 -
1
1
1
1
1

1.
JL /
69.6 [pJ /
1 '
: /
58.0 £) /
J /
1 .
1 /
46.5 ^ /
' / 0/
/
2i,y
m-



£>
/
/21.1 MJ/hr
.x (20,000 Btu/hr)
/
/
/





	 s~\
<*^"^^—' *^^.
/ x\.n




















-^
10.5 MJ/hr
O 1st run
E) 2nd run

















-QBre
\_



                                100                       200

                                    Theoretical air, percent
                                                                                       Breakthrough


                                                                                            TA = 370.
                                                                                                          300
               Figure  7-10b.   Screening data,  JPL-006X - space velocity
                                (methane/air).

-------
       The results of testing monoliths 004X,  005X,  006,  and 006X can be
summarized as follows:
       1.    644K to 764K (700°F to 915°F) lightoff temperatures are
            common for these catalysts under fuel  rich conditions with
            methane fuel.
       2.    Catalyst degradation occurs rapidly with time as noted by
            increased lightoff temperature.
       3.    Both mullite and cordierite appear to be acceptable support
            materials at 1367K (2000°F) bed temperatures.
       4.    Preheat requirements increase substantially (478K to over
            811K) from fuel-rich to fuel-lean stoichiometry.
       5.    Breakthrough occurred typically at a space velocity between
            40,000 and 60,000 per hour.
       6.    Successive breakthoughs occur at the same bed location
            each time, probably at the area of minimum catalyst activity.
       7.    Acurex platinum application techniques were adequate.
       8.    The somewhat higher platinum loading of model  006 helped to
            promote initial lightoff at a lower temperature.
       The next two catalysts to be tested, monoliths JPL-007 and -008, were
to further demonstrate the effects of coating and washcoat application
(see Table 7-1).  The results are given below.

Test Models JPL-007 and -008
       Monolith JPL-007 was tested to verify mullite coating techniques.
JPL-008 compares washcoat preparation of two different manufacturers.  In
general, test results varied somewhat from baseline tests.
       Data for JPL-007 is shown in Table A-2 and Figure 7-11.  During the
first run, preheats were surprisingly low in going from rich to lean
conditions.  Data was taken at five points between 51.8 percent theoretical
air and 211 percent theoretical air, all with preheat temperatures between
317K and 497K (110°F to 435°F) (most previous runs required ">811K preheat
for operation above 100 percent T.A.).  Operation beyond 211 percent T.A.

                                    7-25

-------
                          1200
—i
 i
ro
01
              900  -
              800  -
              700 -
              600 -
              503 -
              400 ~
              300 -
                          1000
                         0)
                         i.
                         3
t
a.
                          400
        Olst run

        Q2nd run




        • Bed lightoff
               O
Note:

  Catalyst

  degradatiojn
                                100                        200

                                    Theoretical air, percent
                                                                                                               300
                      Figure  7-1 la.   Screening data, JPL-007-preheat  temperature  (methane/air).

-------
ro
                             60,000
                              50,000
                        o
                        o
01
(J
ro
0.
oo
                             40,000
                             30,000
                             20,000
                             10,000
                                      Heat release

                                          HJ/hr
                    92.9



                    81.2
                                     :   21.1  CK
 /O

//I      21.1  MJ(hr)
' '  ' —— (20,000 Btu/hr)
                                      100
                                                                                              [laxinium " TA attainable
                                                                                          200
                                                         300
                                                                  Theoretical air, percent
                               Figure 7-11b.   Screening data, JPL-007  - space velocity (methane/air).

-------
was not possible, however.  The second run indicates considerable catalyst
degradation.  The operating point at 100 percent theoretical air indicates
that a preheat of 563K (553°F) was required to sustain the reaction, where
only 401K (262°F) was required for the previous run.  A short maximum
throughput test at 50 percent theoretical  air indicated that 81.2 MJ/hr
(76,900 Btu/hr) (S.V. = 37,603 hr"1) could be produced at a 395K preheat.
       JPL-008 (Table A-3 and Figure 7-12) performed roughly comparable to
JPL-004X and -005X in that extremely lean  operation was attained;but at
high values of preheat temperature (>680K).  Maximum throughput at 50 percent
theoretical air was similar to JPL-007 in  that 81.2 MJ/hr (S.V. = 39,254 hr"1)
was attainable at a low preheat temperature (384K).  Lightoff temperatures
were again comparable to all other catalysts tested at 640K, 682K, and 706K
for the first, second, and third runs, respectively.  Another interesting
point was the high pretest surface area of 13.14 m2/g achieved by the
Matthey Bishop washcoat (see Table 7-2).
       It was concluded from these tests that again, platinum coating tech-
niques on the mullite support were adequate.  Other than the low preheat
temperatures noted for the first run on JPL-007, no significant differences
in combustion characteristics were noted in testing the different manufac-
turer's washcoats.

Test Model JPL-009
       Model JPL-009 was a 6.35 x 10"3m (0.25 inch) cell size monolith
(larger than other cell sizes tested) with the baseline 0.30 weight percent
platinum applied.  Thus, the effects of cell size on combustion were
investigated.  Results appear in Table A-4 and Figure 7-13.
       Testing of this catalyst revealed interesting operational charac-
teristics.  Lightoff was attained at 653K (716°F, similar to others), and
data were recorded for 38.7, 51.1, and 105.5 percent theoretical air as
operation on the lean side was approached.  Lower preheats than all the
catalysts previously discussed were needed at these points (339K - 367K).
At 105.5 percent theoretical air, high levels of UHC's were recorded
(1200 ppm) and further testing at lean conditions was not attempted.  A
maximum throughput test was  also tried and, a 69,6  MJ/hr  (S.V.  = 22,869  hr"1)
heat release was reached when high  UHC values were  evident  (5360  ppm).  This
                                    7-28

-------
ro
           900-
           800-
           100-
           600-
           500-
           400-
           300-
                      1200
                      1000-
                    2
                    2.
                    £
                    a.
                       800'
                       600-
                       400
                       200-
O 1st run
Q 2nd run
& 3rd run


• Bed lightoff
                                     81.2 MJ/hr
                                                                /
                                                                 /
                                               21.1 MJ/hr    O"
                                               (20.000 Btu/hr)
                                                                                     Incipient
                                                                                     Breakthrough
                        100                         200
                             Theoretical air, percent
\
                                                                                     % TA
                                                                                     401.
                                                                                                           300
                       Figure  7-12a.   Screening data, JPL-008-preheat  temperature  (methane/air).

-------
co
o
              60,000
              50,000-
              40,000
°  30,000"
01


01
o

-------
CO
            900-
            aoo-
            700-
            600 -
            500-
            400-
            300 -
\
-------
 i
OJ
ro
    60000
    50000  -
    40000 - -
,2   30000  '
o>
EL
OO
    20000
    10000 -
                          O  1st run
                0)

                TO
                01
                                   69.6   Q
                                   MJ/hr  T
            Q
             i
            /
                  46.5
$
                          /
                                               /21.1  MJ/hr
                                                (20,000 Btu/hr)
                                                             UHC  high
                                                        -no lean testing
                                   100                        200
                                       Theoretical air, percent
                                                                                                           300
          Figure  7-13b.   Screening data, JPL-009 -space velocity (methane/air)

-------
catalyst showed promise for use with others in a series arrangement.  Using
this catalyst upstream of others (others possibly being non-noble metal and
needing high preheat to operate) it appeared that it could be used to sustain
a reaction at low preheat conditions but with poor conversion.  The down-
stream catalysts could then be used for cleanup of the UHC.

Test Models JPL-010. -010X. and -010P
       This series of three catalysts was tested to evaluate the effects
of high platinum loading and the properties of propane fuel.  JPL-010 was
tested for lightoff and one steady operating point, at which time the front
face was over-temperatured.  The data for both 010 and 01 OX are shown in
Tables A-5 and A-6 and Figure 7-14.
       JPL-010X was a duplicate of JPL-010 as the lightoff temperatures
indicate (608K as compared to 596K).  This was slightly lower than the
other catalysts discussed previously.  The most important feature of this
catalyst was indicated by the preheat temperatures required to go from 27.5
percent theoretical air to 280 percent theoretical air.  They were all less
than 350K  (170°F), something none of the other catalysts could approach.
This indicates that the higher platinum loading had a marked effect in
increasing catalyst performance.  This was the first significant improve-
ment in catalyst performance to date.
       During the first run, a breakthrough test was performed at 219 percent
theoretical air.  A heat release of 31.7 MJ/hr (30,000 Btu/hr) (S.V. =
70,846 hr  ) was achieved at a preheat of 638K.  This had been the highest
space velocity recorded to date in the JPL test series.  The third run on
this catalyst illustrates catalyst degradation.  The operating point at
150 percent theoretical air was repeated after about 8 hours of testing.
The preheat needed was 685K (773°F) as compared to 343K (157°F) for the
first run.  The fourth run attempted to measure the maximum throughput at
50 percent theoretical air.  A heat release of 139.5 MJ/hr  (132,000 Btu/hr)
(S.V. = 53,819 hr~ ) was attained without breakthrough occurring.  This
represents the maximum fuel flow capability for the JPL test facility.   In
summary, this was found to be the best catalyst tested to date, although
degradation was still evident.
                                   7-33

-------
900-



300'

700-
/ uu

•^ * 600-
co
-p.



500-




400"
"3 An
oUU
\euv



1000
u.
0

J 80U
s_
3
•t-)
10
i-
01
Q.
^ 60Q
4->

-C
O)
0.

4on




200

O 1st Run
Q 2nd
^ 3rd
^ 4th
0 Bed light off
A*



_ JPL-010
<


^- JPL-010
/^
/
GT



^-139.5 MJ/h»
O^^""""' 	 ^
»U/hr^







.
^^
/
/
1
, i
/ f
/ 1
/Note:
/ Catalyst degradatic
/ after -3 hours of
/ testing
' /
"** \
''^5 — — :=«21.1 MJ/hr








/
pO 31.7 MJ/hr
T (30,000 Btu/hr)






n



^* ^^^^^ ^*— _-* _^^^
\ Jpr n ^^ ^^ ^^^^
^~*"















r>
r
/
/
/
/
*

                           100                       200
                              Theoretical air, percent
300
Figure 7-14a.   Screening data,  JPL-010 and -OlOX-preheat temperature  (methane/air),

-------
co
tn
                    80000
                    70000  - -
                    60000
                     50000 • -
                     40000
                 01
                 u
                 IO
                 O-
                     30000 • -
                     20000
                     10000 - -
                                               100                   200

                                                 Theoretical air, percent
                           Figure  7-14b.
Screening data,  JPL-010 and -010X -space velocity
(methane/air).                                        J

-------
       JPL-010P was tested with both methane and propane to compare combustion
characteristics of the two fuels.  The primary test objective was to demon-
strate catalyst degradation with time.  Therefore, minimum preheat data was
not obtained.  Rather, a constant preheat of nearly 811K (1000°F) was used,
mainly under fuel-lean conditions.  Data is shown in Table A-7.
       Figure 7-15 shows that the catalyst lit off at a methane flow rate of
0.41 Kg/hr (0.91 Ibm/hr) and attained steady-state conditions after 20
minutes.  The flowrate was then dropped to 0.23 Kg/hr (0.50 Ibm/hr) as the
catalyst was allowed to age (from an elapsed time of 39 to 95 minutes).  Fuel
flow was then increased to determine the maximum throughput capability.  This
procedure was repeated until the maximum was found at 0.41 Kg/hr at an
elapsed time of 208 minutes.  After aging the catalyst at 0.23 Kg/hr for
approximately 10 hours, another maximum was found at a total elapsed time of
818 minutes  (13.6 hours).  Here the maximum throughput capability was found
to be 0.23 Kg/hr, a considerable drop from 0.41 Kg/hr.
       Further  testing was done at rich conditions (~50 percent T.A.) where
the catalyst was able to combust as much as 2.0 Kg/hr (4.5 Ibm/hr) of methane
at a preheat temperature of only 301K (82°F).
       Finally, the catalyst was tested with propane.  Initial startup was
attained at  a fuel flow of 0.32  Kg/hr (0.7 Ibm/hr).  During the  propane  test-
ing, the catalyst experienced melt down in various locations and was found  to
be extremely unreactive.
       Figures  7-16 and 7-17 illustrate the preheat temperatures and space
velocities for  each test point.  Most points cannot be used as a comparison
to previous  data since minimum preheat conditions were not sought.  The
exception is the fourth methane run performed at rich conditions.  Lightoff
occurred near 700K (800°F, see Figure 7-16a) and minimum preheats were found
to fall below 375K (215°F).  In comparison to previously tested catalysts,
JPL-010P appeared to be very active at rich conditions after more than 13
hours of aging  at lean conditions.
       A number of significant conclusions were made from the tests on these
three catalysts.  They are:
       1.    High catalyst loading facilitates both low lightoff  and low
             preheat conditions.
                                    7-36

-------
0.5  -
0.4  -
0.3  -
0.2  -
0.1  -
1.0-


s-
~j= 0.8-
•~
f\
cu
ro
- s-
5 0.6.
o
«4-
 — Mdx
i/^
E D V




\
\
\
\
Q El fc^





imum through







\
^^
>«


Q EJ D PUP D








•^

puL












CD Q




































y — Ma
/
Q ^"Qi — EJ^ 	








ximum throug







iput






0        200
400        600        800

         Time,  minutes
                                                           1000       1200       1400
   Figure 7-15.
      Degradation of test model  JPL-010P at lean conditions
      (>350 percent T.A.) (methane/air)

-------
900-
800-
700-
600-
500-
400-
300-



1000-

u_ 800-
O
e
4->
ra
i-
01
i" 600-
0)
4-J
4-)
QJ
OJ
400
200.
0-
O 1st Run
Q 2nd Run



A





A
^^k
^\
>3rd Run
^ 4th Run -












a













a o
%









PSr
ir
























                         100        200        300        400

                                    Theoretical air, percent
500
600
           Figure 7-16a.   Screening data, JPL-OlOP-preheat temperature
                           (methane/air).
                                  7-38

-------
I
CO
                                  60,000
                                  50,000
                                  40,000
                                  30,000
                                a)
                                u
                                10
                                CL
                                  20,000
                                  10,000
A



A
     O 1st Run

     QJ ?nd Run

     ^ 3rd Run

     A 4th Run
                                            .A
                                                                                   iff
                                                  100        200       300        400


                                                              Theoretical  air, percent
                                              500
600
                                         Figure  7-16b.   Screening data, JPL-010P  - space velocity
                                                          (methane/air).

-------
•^1



o
                         900-
                         800-
                         700-
                         600-
                         500-
                         400-
                         300-
"i nnn.

800 •
u_
o
QJ
S-
3
-4-1
(O
S-
0 finn-
OJ
-i->
+j
n3
0)
£
?nn .


c
x-'
/
Q




Q

6

1C
1
3 1st Run










10 2
<;— — 	










30 3(
	 (§1










)0 4'











^0 5C











)0 6
                                                            Theoretical air,  percent
                           Figure 7-17a.   Screening  data, JPL-OlOP-preheat temperature (propane/air)

-------
  60,000
  50,000
  40,000
u 30,000
o


s

a>
o
a
a.
if)

  20,000
  10,000
                   O 1st Run
                                                  9
a
                  100       200        300        400


                             Theoretical air, percent
           500
600
          Figure 7-17b.   Screening data, JPL-010P -space velocity

                           (propane/air).

-------
       2.    Catalyst  degradation  with  time is  still  significant  but
            lifetime  is  increased with increased  catalyst  loading  at
            1367K operating  temperatures.
       3.    Increased catalyst loading allowed higher heat release and
            space velocities to be achieved.
       4.    Maximum throughput indicates  catalyst degradation occurs
            in a similar manner to lightoff temperature.
       Segments of test  model  JPL-010P were prepared and  analyzed  at  JPL  by
Scanning Electron Microscopy (SEM) and energy-dispersive  analysis  by  x-ray
(EDAX).  The results  for the exit ends of two  channels are shown in  Figures
7-18 and 7-19.  The micrographs show visible platinum globules in  both
locations, verified by the EDAX scans. Aluminum  and silicon of the catalyst
support and washcoat were also detected.
       The detection of  platinum  at the catalyst  surface  by the JPL  analyses
shows that active catalyst remains at  the surface (exposed to the  gas flow)
despite operating temperatures in excess  of 1367K (2000°F).  Larger  platinum
globules indicate that some agglomeration (loss of dispersion) has occurred.
Both catalyst and washcoat structure appear similar for the two areas analyzed.

Test Models JPL-011,  -012, -013,  and -022
       JPL-011, -012, and -013 were tested to  learn more  about washcoat and
platinum behavior.  JPL-011  was treated with H?S  gas to improve the axial Pt
distribution through the monolith.  JPL-012 and -013 were presintered to
study the effects of the alumina  washcoat phase change from gamma  to  alpha.
Keep in rnind that unlike JPL-010X these catalysts had the low platinum load-
ing (0.22 - 0.30 weight percent).
       JPL-011 was tested both with methane and propane.   The results are
shown in Table A-8 and Figures 7-20 and 7-21.   The methane tests on this
catalyst resulted in performance  similar  to JPL-006X and JPL-007.   Lightoff
was 637K (686°F) and 676K (756°F) for the first and second runs, respectively.
Unlike the others a maximum throughput test was performed at 50 percent
theoretical air first.  Performance was excellent, requiring a low preheat
(< 340K) to reach a heat release of 139.5  MJ/hr (132,000 Btu/hr, S.V-  =
50,835 hr" ) without breaking through.  A subsequent walkthrough to lean

                                    7-42

-------
         r . ;ip
                                         -Channel  1
                                         -Channel 2
^-J

OJ
               1.0 mm
 a.
20 x magnification of
channels 1  and 2
                                                                                  0.2 mm
100 x magnification of
channel  1  bottom
c.   1000 x magnification
    showing Pt crystallites
                                                                Al  Si  Pt
                                                                                 Pt  Pt
                                                      d.  EDAX scan  showing  presence
                                                          of aluminum, silicon,  and platinum
           Figure  7-18.   Surface  analysis  at  exit  of  channel  1,  test model  JPL-010P.

-------
r
 i   . &,$   ,
•o-

                     1.0 mm
Channel  3
                                               Channel 4
       a.   20x magnification  of
          channels  3  and  4
      c.  lOOOx magnification
         showing Pt crystallites
                           "*   _,,  ;-
                                                                                           IP
                                                                                           I
•1
                                         ...  _ *

                                        \ 0.2 mm
                        b.  lOOx magnification of
                            channel 4 bottom
                                                                   Al  Si  Pt
                                         Pt  Pt
                d.  EDAX scan showing presence
                    of aluminum, silicon, and platinum
                Figure  7-19.  Surface analysis at exit of channel  4,  test model  JPL-010P.

-------
en
               900 -
               800-
               700 -
               600 -
               500 -
               400 -
               300 -


1000 •

u. 80°
o
n

/i(20.0C
^v^-58.0 MJ/hr
-\
J


U/hr
)0 Btu/hr)
Maximum % TA
attainable for
baseline heat
release










Maximum % TA
attainable with
low heat release


-




-


-
                                                       100                       200
                                                           Theoretical  air, percent
                                                                                                         300
                         Figure 7-20a.   Screening data,  JPL-Oll-preheat temperature (methane/air).

-------
    60000
~-J


CTi
    50000
    40000
o   30000
0)
u
 /
' /
1 /
92.7 (•) /
1 /
j /


Note: catalyst degratatio
from 1st to 2nd run

P
/
/
/ 21.1 MJ/hr
/ (20,000 Btu/hr)
/
/
69.5 Q/ J


1








Tf P ^-Q
EKBreakthrough/ ^ -- ~"
1 /
46.4 rS /
T /
1 /
1 /
1 /
1 /
1 ,'
--Xj

^•^^
^ *~
	 ^^ 11.6 MJ/hr
or


K1
Max " Ta
(21.1 MJ/hr)






r1
Max % Ta
(11.6 MJ/hr)


-















-


                                  100                       200

                                     Theoretical air, percent
                                                                                                300
                         Figure 7-20b.   Screening data, JPL-011 — space  velocity
                                          (methane/air).

-------
900-
800-
700-
600-
500-
400-
300-

1000 -
^800 .
o
£
+j
A
S-
§600 '
4->
CO
O)
2!
O.
400 -
200 •
& 3rd run
A Bed lightoff
•





A
y
X
/v x-X 21.1 N
"^-^sr*^ (20, oc









J/hr
0 Btu/hr)










                                       100                       200
                                           Theoretical air, percent
300
         Figure 7-21a.   Screening  data,  JPL-Oll-preheat temperature (propane/air).

-------
— I
I
CO
                    O)
                    u
                    CL

                    CO
50000-
40000-

30000"
i nnno-

^ 3rd run


A
/ 21.1 M
(?n nnn R





]/hr
tn /hl-1







                                                  TOO                     200

                                                      Theoretical air, percent
300
                              Figure 7-21b.   Screening  data, JPL-011 - space velocity
                                              (propane/air).

-------
conditions was not possible beyond 147 percent theoretical air.   Decreasing
the heat release allowed a maximum of 211 percent theoretical  air to be
reached with a very high preheat (782K).  To confirm that the catalyst had
degraded substantially, an attempt was made to repeat the maximum throughput
test at 50 percent theoretical air.  Breakthrough was found to occur at a
space velocity of 22,848 hr~^.  In the first run, no breakthrough occurred
up to the maximum capability of the test rig at 50 percent theoretical air.
A direct comparison of this run with JPL-006X to deduce the effect of the
H2S platinizing technique is difficult, since the tests were not carried out
in the same order.  JPL-006X tried a rich-to-lean walkthrough followed by a
maximum throughput test at 50 percent theoretical air.  JPL-011  was exposed
to the reverse of that procedure.  Since degradation plays such an important
role, the two catalysts could not be totally compared.  However, from the
data obtained, it was noted that JPL-011 required a lower preheat temperature
at the same values of theoretical air and heat release than JPL-006X, in
spite of the early maximum throughput test.  Thus, H^S treatment of platinum
appears beneficial in terms of extending platinum activity with time.
       Propane testing consisted of a lightoff and three test points.  The
bed lightoff temperature  (469K) was much lower than all catalysts tested
with methane.  Preheat temperatures for operation near 100 percent theoreti-
cal air were also surprisingly low (<391K) considering the long testing
that the catalyst had undergone with methane.
       JPL-012 was prepared by sintering the washcoat at 1273K for 48 hours
before applying the platinum.  Data for this test model is given in Table A-9
                                                     2
and Figure 7-22.  The pretest surface area was 1.33 m /g, considerably below
the unsintered values of the previous catalysts.  The dispersion of platinum
was measured as 49 percent.  In general, this catalyst performed similar to
other catalysts after they had aged for a few hours, in that operation at
conditions leaner that 111 percent theoretical air was not possible.  Also,
maximum heat release at 50 percent theoretical air was 81.2 MJ/hr (76,900
Btu/hr), S.V.  = 42,373 hr   at a relatively high preheat temperature  (704K).
Bed lightoff temperatures were similar to the previous catalysts reported;
614K (645°F) and 674K (754°F) for the first and second runs, respectively.
In general, the performance of this catalyst was similar to JPL-006X which
did not have a presintered washcoat.
                                    7-49

-------
                       1200
i
en
O
             900
             800-
             700-
             600-
500'
            400'
            300-
                       1000
                       800
        10
        s-
        01
        §•  600
        
-------
I
tn
                      60000
                      50000
                      40000
                   _o 30000

                    (U
                    
-------
       JPL-013 was prepared by sintering at 1273K for 48 hours, after the
washcoat and platinum had been applied.  An extremely low pretest surface
              p
area of 0.93 m /g and a dispersion of only 0.7 percent were measured.  These
are comparable to the previous post-test measurements of the other catalysts.
As expected, the performance was very poor.  Lightoff was high, at 734K
(861°F).  Breakthrough occurred at 51.1 percent theoretical air at only the
baseline heat release.  UHC was measured at 6300 ppm at this condition.
Further testing was not attempted.  Data for JPL-013 is shown  in Table A-10
and Figure 7-23.
       The fourth catalyst in this series, JPL-022, was prepared with a
DuPont stabilized washcoat.
       Performance with this catalyst was comparable to JPL-011, the Acurex
prepared catalyst with H?S treating.  An initial maximum throughput test at
50  percent theoretical air showed the performance to be good.  Maximum heat
release of the  test facility was reached (139.2 MJ/hr) at a very low preheat
(306K).  Upon attempting to "walk" the catalyst through to lean conditions
it  was  found that it  could not exceed 103.5 percent theoretical air.  This
is  roughly what was found to occur for JPL-011 at 147 percent  theoretical
air.  Note that like  JPL-008 (Matthey Bishop washcoat) the pretest surface
area was slightly higher than Oxy-Catalyst prepared monoliths.  (See Table
7-2).  Data is  shown  in Table A-ll and Figure 7-24.
       Several  important conclusions resulted from these washcoat evaluation
tests.
       1.   Acurex-prepared catalysts performed comparably to  commercial
            catalysts prepared by DuPont, Oxy-Catalyst, and Matthey
            Bishop.
       2.   \\2$ treatment of the platinum appears effective  in increasing
            combustion performance
       3.   Propane is more active with a combustion catalyst  than methane.
       4.   Presintering the washcoat does not appear to have  a deleterious
            effect on the performance of the baseline catalyst, implying
            that relatively little platinum is buried during sintering of
            the washcoat and that much of the initially available surface
            area is unused.
                                     7-52

-------
i
in
co
900-
r
800-
7QQ-

600 -
500 '
400 •
300 -

ICUU
1000 -
u. 800 -
o
»
OJ
!_
3
-(->
n>
s-
a)
§• 600 -
•p
+->

-------
    60000
    50000
    10000
o   30000
O)
o

S.
oo


    20000
    10000
                   O 1st
                               21.1 MJ/hr

                               (20,000 Btu/hr)
                        Terminated test  due to

                        high bed lightoff temp and high UHC
                                  100                        200

                                      Theoretical air, percent
300
             Figure  7-23b.   Screening  data,  JPL-013 - space  velocity
                               (methane/air).

-------
en
tn
              900-
              800-
              700-
              600-
              500-
              400-
              300-
                        1000
                         800
                       e
ID


S.600
                       0)


                         400
                         200
                                                      100                        200

                                                          Theoretical air, percent
                     Figure  7-24a.   Screening  data, JPL-022-preheat temperature (methane/air).

-------
60000 '
50000 •
40000 -
S-
jr
~~
,0 30000 -

0>
-J re
1 0.
on ">
W 20000 -
10000 -
n -

139.2 9
• MJ/hr i
1
127.6 6

104.3 (j>
I ^
81.1 6 /
1 /--
1 /
1 /
1 /
57.9 (S /
1 '/
!/
21.1 GP^
O 1st run
D 2nd run
	 Incipient
J
21.1 MJ/hr
(20,000 Btu/hr)


Maximum % TA attainable

-


-

-
1 	
                TOO                      200

                    Theoretical  air, percent
300
Figure 7-24b.   Screening  data, JPL-022 - space velocity
                (methane/air).

-------
       5.   Presintering the platinized washcoat at 1273K for 48 hours has
           a marked effect on catalyst performance, reducing the catalyst
           activity to a very low level.

Test Model JPL-016
       Stabilization techniques are available for catalyst materials as well
as for washcoats.  JPL-016 consisted of a Corning cordierite support with
an Oxy-Catalyst 10 weight percent gamma-alumina washcoat.   The washcoat was
stabilized by the addition of cerium oxide and impregnated with a 2:1 molar
ratio of platinum stabilized with palladium.
       Table A-12 summarizes the operating points attained.  These are
illustrated in Figure 7-25.  Initially, lightoff occurred  at 608K (635°F) at
37.3 percent theoretical air.  Figure 7-25a shows that preheats of less than
324K (123°F) were able to maintain activity up to 106.8 percent theoretical
air.  Beyond this point activity was difficult to maintain as indicated by
several aborted attempts to reach theoretical airs beyond  113.4 percent.
The furthest lean condition attainable is shown to be 113.4 percent theoretical
air with a preheat of 357K (182°F),  The catalyst was restarted and operated
at rich conditions, hitting almost identical operating points as in the first
run.  In an attempt to reach leaner conditions, the fuel flow was then
reduced to 0.23 Kg/hr (0.5 Ibm/hr).  A preheat of 764K (915°F) was needed
at 109.7 percent T.A. to maintain activity.  Maximum throughput at
50 percent T.A. was then determined by operation at 1.59 Kg/hr (3.5 Ibm/hr)
and then at 1.81 Kg/hr (4.0 Ibm/hr).  Note the low preheats required for
these two conditions.  The maximum throughput was even higher, attaining
nearly 2.72 Kg/hr (6.0 Ibm/hr).  A third run was attempted which required
approximately 22K more bed preheat for lightoff than the previous two runs.
Operation at 90.4 percent theoretical air required 763K (914°F) of preheat,
showing a large amount of degradation over the last two runs.
       In summary, this catalyst appeared to be comparable to JPL-022 (pre-
pared by DuPont).  It also was unable to operate at lean conditions but
performed well at rich conditions with high throughputs.
                                                                         o
       Since the post-test surface area of JPL-016 was quite high (2.08 m /g),
as  shown  in Table 7-2, a further investigation into the area of washcoat

                                    7-57

-------
I
un
00
                900-
                800-
                700 -
               600 -
                500 -
               400 -
               300 -
                           1200
                           1000 •
                        ^ 800
                        S. 600
                        OJ
                           400
                           200 -
O1st run

CJ 2nd run
A 3rd run
9 Bed lightoff

                                                                    21.1 MJ/hr
                                                                    (20,000 Btu/hr)
                                                    MJ/hr
                                                                    •Maximum % TA attainable
                                                          100                         200

                                                              Theoretical air, percent
                                                                           300
                              Figure  7-25a.   Screening  data,  JPL-016-preheat temperature (methane/air).

-------
^J

en
                          60,000-
                          50,000
                          40,000-
                        o  30,000-
                        01
                        u
                        10
                        a.
                          20,000-
                          10,000-
O 1st run
0 2nd run
A 3rd run

MJ/hr
92.8 -*-Q &f
81.2 -*H3



/— 21.1 MJ/hr
(20,000 Btu/hr)


P
$
/
/
•s /
/•
O1^





x-11,6 MJ/hr



^ 	 Maximum % TA attaina












ble

                                                         100                       200

                                                            Theoretical  air, percent
300
                                     Figure 7-25b.  Screening  data, JPL-016 - space  velocity
                                                      (methane/air).

-------
stabilization was conducted.   The objective was to determine the change in
surface areas of Y-alumina as a function of the amount of ceria (Ce02)  or
cesium oxide (Cs20) deposited on the y-alumina washcoat, when these wash-
coats are held at high temperatures.
       To perform the study,  a Corning cordierite monolith was cut in half,
and then each half was cut into 12 pie-shaped wedges as shown in Figure 7-26.
  Figure 7-26.  Test specimen preparation, washcoat stabilization tests.

Duplicate samples were dipped in cerium or cesium-doped solutions of alumina
to give loadings of 0 percent, 0.3 percent, 1 percent, and 5 percent Ce02
or CS20.  The wedges were dried at 423K.  Samples were then placed in a cold
oven, heated to 1273K, and held at that temperature for 16 hours.  After
cooling down for 8 hours, the samples were reheated to 1423K (2100°F) for
18 hours.  Each sample was ground up, weighed, loaded in the BET cell, and
the  cell evacuated for 30 to 60 minutes at 523K.  Surface area measurements
were then made.  Data for the analysis is presented in Table 7-5 and the
results are plotted in Figure 7-27.
       Based on the results of the surface area measurements, the ceria
treated samples show no stabilization in surface area with up to 5 percent
ceria applied.  Rather a decrease in surface area to a value about 10 percent
below the unstabilized samples is noted for ceria loadings of 1 percent and
5 percent.  Samples treated with 0.3 weight percent cesium oxide also show
no effect on surface area stabilization.  However, the samples treated with
1 percent and 5 percent Cs20 show nearly a factor of 2 increase in surface
                                    7-60

-------
   3.0-
   2.5 H
   2.0-
OJ
o
       i
    1.0-
    0.5'
                                 Washcoat stabilizabion samples
                                 A    Untreated
                                 Q
                                           treated
                                 Q    Cs20 treated
                                 Samples held at  1423 K for 18 hours
                                                            -O
                             2          3^
                             Stabilizer (percent)
   Figure 7-27.   Washcoat stabilization  study results.
                             7-61

-------
area over presintered values for untreated washcoats.   Based on these results,
it appears desirable to stabilize y-A^*^ washcoats with Cs20.

           TABLE 7-5.  WASHCOAT STABILIZATION STUDY TEST RESULTS
All samples listed were heated at 1273K for 16 hours,
cooled for 8 hours, and reheated to 1423K for 18 hours.
Sample

3
5
7
10
12
14
15
16
17
18
Ce02 loading

5%
1%
0.3%
0
0
0
0
0
0
0
Cs20 loading

0
0
0
5%
1%
0.3%
0.3%
0
0
0
2
Surface Area (m /g)

0.96
0.96
1.23
2.14
2.06
1.06
1.07
1.12, l.lla
1.16
1.20, 1.15a
 Two BET runs were made on these samples to check data.   Results show
 excellent agreement.

Test Model JPL-021
       The catalyst used for this test consisted partly of a 1-inch thick
DuPont alumina support with 6.35 x 10  m (0.25 inch) cells and was washcoated
with approximately 7 weight percent alumina.  Acurex impregnated the bed
with over 5 weight percent of platinum which exceeds by more than a factor
of six that used in earlier tests.  In previous testing JPL-009 (which also
             _o
had 6.35 x 10  m cells) had shown much promise because of its capability of
maintaining a reaction without experiencing breakthrough.  The only drawback
was the high amount of UHC allowed to pass through the bed due to the large
cell size.  This led to the system concept having a large cell catalyst
upstream to maintain a reaction followed by a small cell catalyst downstream
to clean up the UHC.  The purpose of this test was to do more extensive
testing of the large cell monolith.  The latter part of testing also
included downstream small cell segments to confirm system concepts.
                                   7-62

-------
       The test procedure was similar to that performed on JPL-010P.   It was
intended to show the degradation of the 1-inch segment by testing at lean
conditions and finding the decrease in maximum throughput with time.   Table
A-13 summarizes the data points taken.  Note that the first 1187 minutes
were all at very lean conditions with the 0.0254m (1-inch) large cell  segment
only.  From a total elapsed time of 1187 minutes to 1370 minutes the system
concept was tested at lean conditions.  For the last four operating points
the system concept was tested at rich conditions (50 percent theoretical air)
to roughly compare it to previous tests.  Figure 7-28 illustrates the fuel
flow capability of the catalyst for over 20 hours of testing.   Unlike the
degradation tests of JPL-010, this test did not show a decreasing throughput
capability of the catalyst.  This is because the improved catalyst performance
allowed operation at the  maximum flow capability of the test facility  (at ~400
percent theoretical air) throughout the test.  Notice that this catalyst sus-
tained 0.68 to 0.86 Kg/hr (1.5 to 1.9 Ibm/hr) of methane for over 20 hours
while JPL-010 showed a decrease in capability from 0.41 to 0.23 Kg/hr (0.91)
to 0.5 Ibm/hr) of methane over a 13-hour period.  This demonstrated an
excellent capability to maintain a reaction since it remained at or near
the front face throughout the test.  Like JPL-009, poor conversion was seen,
as expected.  Typical amounts of UHC ranged from 0.077 to 0.213 Kg/hr methane.
       To reduce the amount of UHC, two segments of JPL-015 were included
                                                              _-3
downstream after 1187 minutes of testing.  These had 1.59 x 10" m (0.0625 in.)
cells and were impregnated with 0.7 weight percent Pt, identical to the
JPL-010 series of catalysts which were tested previously.  Testing of this
concept was successful in that UHC decreased substantially to less than
0.0045 kg/hr (0.01 Ibm/hr) of methane.  In the 3 hours of testing, the
cleanup capability was seen to decrease slightly as the downstream small
cell segments degraded.
       Figure 7-29 illustrates the preheat temperatures and space velocities
attained.  Like JPL-010P, these cannot be used for a basis of comparison
with other catalysts since the operating points are not at a minimum pre-
heat condition.  Exceptions are the last three data points taken at rich
conditions.  Figure 7-29a shows that after 25 hours of testing this catalyst
still maintained 1.8 kg/hr (4.0 Ibm/hr) of methane at 51.3 percent theo-
retical air with a very low amount of preheat (<322K).  Also notice the

                                   7-63

-------
 .90-
 .80-
 .70-
.60-
.50-
1.8.
-C
Il.6-
O)
s <
o
M-
(Ul,4 -
ra
_c
4->


& A A A

System
concept

hrs
200 140
Time, minutes
   Figure 7-28.  Fuel  flow  capability of JPL-021  catalyst at lean conditions
                 (350% TA,  methane/air).

-------
^J



on
   900-
   800-
   700^
  600 -
  500 -
  400 ~
  300 -
              1200
               800-
                                    £
                                    01
                                    0.
             O)
             s_
             o.
               400'
                                            O
                                            Q

                                            O
 1st run

 2nd run

 3rd run

 4th run

 5th run

V
                                                                 o
                                       200                  400


                                        Theoretical air, percent
                                                                                                        600
Figure 7-29a.   Screening data, JPL-021-preheat temperature (methane/air).

-------
   300,000
   250,000
   200,000
^  150,000
u
o
O)
>
 0)
 o
 g. 100,000
    50,000
                    100
1

1st run \
Q 2nd run > ] lar
O 3rd run J
A 4th run \
V 5th run ^ ] 1ar
4

J
1
Q
ge cell segment
ge cell and 2 small cell
3>


O Q
/
segments

+

200        300        400

 Theoretical  air, percent
                                                              500
600
       Figure 7-29b.   Screening data,  JPL-021  - space velocity
                        (methane/air).
                                   7-66

-------
unusually high space velocities attained in Figure 7-29b.  These are due pri-
marily to the factor of three decrease in catalyst volume when testing with
a 0.0254m (1-inch) thick segment rather than a 0.0762m (3-inch) thick one.
       The large cell/small cell catalyst with high platinum loading showed
two very promising characteristics:  substantially higher throughput capa-
bilities and increased operational lifetime.  The graded cell  concept was
to be further tested on the last of the JPL test monoliths, JPL-019.

Test Model JPL-019
       The JPL-019 monolith was termed the "graded cell catalyst."  Three
cell size supports (6.35, 4.76, and 3.18 cm cells) were bonded together to
reduce cell size as flow passed through the catalyst bed.  The catalyst was
tested with four fuels at JPL (methane, propane, indolene, and methanol) in
demonstrating its performance.
       Test point summary, emissions, and bed temperature data are shown in
Tables A-14 and 7-6 through 7-8.  As can be seen from the tables, the bed
was also run at temperatures up to 1672K (2550°F) with no substantial increase
in pollutant emissions.  The higher bed temperatures generally resulted in
more uniform temperature distributions.  The JPL test data shows somewhat less
uniform temperatures when operating with the liquid fuels (methanol and
indolene) than with the gaseous fuels.  This may have resulted from incomplete
fuel vaporization.
       Based on the success of Test Model JPL-019, it was brought to Acurex
for additional testing, including operation at pressure and at higher temper-
atures (to 1756K) and  investigation of conversion of fuel-nitrogen species.
       Atmospheric pressure tests for conversion of fuel-nitrogen were con-
ducted first.  Natural gas and propane were used as the test fuels and doped
with ammonia.  Tables  A-15 and A-16 give the test conditions and resulting
emissions.  Under fuel-rich conditions, nearly all injected ammonia came
through the combustor  as ammonia for both test fuels.  The measured NH3 in
the combustion products was higher than that measured for the  incoming gases
in two cases, indicating some measurement error.  For fuel-lean conditions,
the NH3 was totally broken down by the catalyst.  Conversion  of  NH3  to  NOX
was measured as approximately 20  to 26 percent for lean  conditions.   Bed tem-
perature measurements  for  selected test points are shown  in Table  7-9.
                                     7-67

-------
                             TABLE  7-6.   MONOLITH  019 TEST  DATA - JPL  MULTI-FUEL  TESTS
cr>
CO
Typical Emissions Data Kg/hr (Ibm/hr)

1367K
(2000°F)
Bed
Temp

1589K
(2400°F)
Bed Temp
Fuel
Methane
Propane
Indolene
Methanol
Methane
Propane
TA, %
320
350
290
326
211
248
CO
0
0
0
0
0
0
H2
0
0
.0045(0.01)
.0045(0.01)
.0045(0.01)
.0090(0.02)
UHC
.0227(0.05)
0
.0045(0.01)
0
0
0
Run -Scan No.
A41A-6
A41C-5
A41 D-5
A41E-7
A41 F-4
A41F-15
                 NO   measurements  were not made on this series of tests;  based on previous test
                   A

                 results,  no  NO  emissions were expected.
                               /\

-------
                        TABLE 7-7.  MONOLITH 019 TEST DATA - RADIAL  BED  TEMPERATURE PROFILES
Methane A41A-6
TC
1

2

3

4

K (°F)
1369
(2004)
1371
(2008)
1362
(1992)
1340
(1952)
Propane A41C-5
TC K (°F)
1

2

3

4

1366
(1998)
1373
(2011)
1344
(1959)
1367
(2001)
Indolene A41D-5
TC
1

2

3

4

K (°F)
1295
(1871)
1289
(1861)
1412
(2082)
1381
(2025)
Methanol A41E-7
TC
1

2

3

4

K (°F)
1318
(1913)
1349
(1969)
1467
(2181)
1372
(2009)
Methane A41 F-4
TC
1

2

3

4

K (°F)
1636
(2485)
1621
(2457)
1627
(2470)
1521
(2277)
Propane A41F-15
TC
1

2

3

4

K (°F)
1636
(2484)
1615
(2447)
1622
(2460)
1473
(2191)
•~g
cr>
                                                           Section taken  .0635  m from front face  -
                                                           four thermocouples in  that  plane.
                                                           Gaseous fuels  show more uniform  temperature  profiles
                                                           than liquid fuels.

-------
                       TABLE 7-8.  MONOLITH 019 TEST DATA -AXIAL BED TEMPERATURE PROFILES
Methane A41A-6
TC
5
6
7
8
9
K (°F)
1344
(1960)
1321
(1917)
1352
(1973)
1369
(2004)
1317
(1910)
Propane A41C-5
TC
5
6
7
8
9
K (°F)
1359
(1986)
1208
(1714)
1261
(1810)
1366
(1998)
1351
(1971)
Indolene A41D-5
TC
5
6
7
8
9
K (°F)
1348
(1966)
1133
(1580)
1239
(1771)
1295
(1871)
1384
(2031)
Methanol A41E-7
TC
5
6
7
8
9
K (°F)
1379
(2022)
1324
(1923)
1402
(2063)
1318
(1913)
1412
(2082)
Methane A41F-4
TC
5
6
7
8
9
K (°F)
1596
(2412)
1523
(2282)
1566
(2358)
1636
(2485)
1547
(2325)
Propane A41F-15
TC
5
6
7
8
9
K (°F)
1594
(2409)
1513
(2263)
1614
(2446)
1636
(2484)
1541
(2313)
•-J
o
                                i
                                         t
Thermocouples 5, 6, 8, and 9 lie in same plane.
Axial temperature for gaseous fuels at 1589K (2400°F)
operating temperature are in excellent agreement.
Significant variations in axial temperature exist
for indolene.

-------
TABLE 7-9.  MONOLITH 019 TEST DATA - AXIAL BED TEMPERATURE PROFILES



c
I

I


I
I
7




Typical test points
Thermocouples 5, 7, 8, 9 lie in same plane
FUEL RICH FUEL LEAN
Natural
Gas
Run 1122-2 Run 1124-2 Run 1201-1 Run 1201-4
1C
5
6
7
8
9
K_ £°fl TC JC1 (°F) JC K J^F)_ . J_C JC (°F)
1273 (1831) 5 1518 (2273) 5 1144 (1600) 5 1457 (2162)
1303 (1886) 6 1501 (2242) 6 1432 (2118) 6 1591 (2404)
1342 (1956) 7 1647 (2504) 7 1486 (2214) 7 1620 (2456)
1359 (1986) 8 — — 8 1434 (2121) 8
1333 (1940) 9 1597 (2414) 9 1441 (2134) 9 1561 (2349)
Propane Run 1123-2

TC
5
6
7
8
9
L (°F)

1359 (1986)
1367 (2001)
1309 (1896)
1284 (1852)
1293 (1867)

-------
       The fuel-nitrogen conversion tests at atmospheric pressure indicated
the potential  of the platinum catalyst to control NO  emissions due to fuel-
                                                    X
bound nitrogen species under lean conditions.  Further, tests on the rich
side should be conducted between 40 and 100 percent theoretical air.  One
additional NH3 conversion point at 0.303 MPa (3 atm) was included with sub-
sequent pressure tests.
       Data for pressure tests with methane between 0.101 and 0.606 MPa (1 and
6 atm) are given in Tables A-17 and A-18.  In general, NOX, CO, and unburned
hydrocarbon levels did not vary with pressure under lean conditions.  Run
number 1206-6 with NFL dopant at 0.303 MPa showed an increase in conversion
to NO  (68 percent) over the baseline (1 atm) condition.  This result suggested
     A
that catalytic control of fuel nitrogen conversion to NO  is more effective
                                                        J\
in lower pressure combustion systems.
       Table 7-10 shows bed temperature distribution with increasing pressure.
No changes in relative uniformity were found.
       Following fuel-nitrogen and high pressure testing, the catalyst was
removed from the test fixture for surface area and platinum dispersion
                                      2
measurements.  The results were 0.02 m /g and zero, respectively.  Since the
catalyst still retained good activity, it was re-instrumented with thermo-
couples and returned to the facility for testing at high temperature.
       Tests were conducted with natural gas fuel at 0.101 MPa (1 atm)
pressure and approximately 589K (600°F) preheat temperature.  Bed temperatures
of 1644 to 1700K (2,500°F to 2,600°F) were maintained for approximately one
hour, with the maximum bed temperature of 1761K (2,710°F) maintained for
approximately 15 minutes.  Following a brief cooldown period, the catalyst
was  successfully relit under fuel-rich conditions at approximately 744K
(880°F), virtually the same lightoff temperature as was found previously.
It was then operated briefly under fuel-lean conditions to demonstrate
continued successful performance and removed and sectioned for SEM and EDAX
analyses.
       High temperature test conditions are summarized  in Table A-19 and
temperature measurements are given in Table 7-11.  No emissions data was
obtained.  Much greater bed temperature uniformity was  noted as test temper-
ature increased to the 1761K maximum obtained.
                                   7-72

-------
                     TABLE 7-10.   MONOLITH  019  TEST  DATA -AXIAL  BED TEMPERATURE PROFILES
CO






5
•

1
6 1
' ! .!
1 7
I '
1




                                                  Pressure data, methane, lean conditions
                                                  Thermocouples 5, 7, 8, and 9 lie in same plane
1 ATM
Run
TC
5
6
7
8
9
1206-3
K (°F)
1576
(2376)
1637
(2487)
1636
(2484)
( " )
1593
(2408)
3 ATM
Run
TC
5
6
7
8
9
1206-5
K (°F)
1561
(2350)
1601
(2422)
1604
(2428)
( - )
1563
(2354)
6 ATM
Run
TC
5
6
7
8
9
1206-7
K (°F)
( " )
1478
(2200)
1400
(2060)
1417
(2091)
1527
(2288)

-------
TABLE 7-11.  MONOLITH 019 TEST DATA -AXIAL BED TEMPERATURE PROFILES

~J
* Fuel
Rich
Fuel
Lean
5.
n
6.

1
9.
•7
Run 1228-1
K
5
6
.7
9
Fuel : natural gas
i nerinocoup i es o, /, y are in tne s
Run 1228-11
ame plane
958 (1264) 5 1244 (1780)
1365 (1997) 6 1478 (2201)
1533 (2299) 7 1703 (2605)
1209 (1716) 9 1319 (1914)
Run 1228-3 Run 1228-6 Run 1228-8
TC
5
6
7
9
K (°F) TC K (°F) TC K (°F)
1369 (2005) 5 1386 (2034) 5 1724 (2643)
1624 (2464) 6 1656 (2521) 6 1737 (2666)
1636 (2484) 7 1707 (2613) 7 1760 (2708)
1296 (1872) 9 1567 (2360) 9 1592 (2406)
Run 1228-12
TC K (°F)
5 1404 (2067)
6 1646 (2503)
7 1691 (2583)
9 1519 (2274)

-------
       A total test time of 74 hours was accumulated on test model  JPL-019,
demonstrating that the graded cell catalyst system had long life character-
istics even under severe test conditions.  Thus, an important program objec-
tive was met by development of a catalyst with adequate lifetime for system
application.
       Test model JPL-019 was sectioned for SEM/EDAX analyses at JPL.
Sections from both the large and small cell segments were cut as in Figure
7-30 for mounting.  Initial analyses of these segments showed little or no
detectable platinum at either the inlet or exit region of either of the cell
sections (Figures 7-31 to 7-34).  A small platinum response on the EDAX scan
of Figure 7-31 was the only evidence of catalyst at the surface.  Subsequent
searching of the outlet regions of both large and small cell channels
revealed additional platinum but with very low dispersion and great non-
uniformity from location to location.  Additional platinum can be seen in
the results of Figures 7-35 to 7-38.  It was clearly evident that the higher
operating temperatures of test model JPL-019 resulted in substantial catalyst
removal from the surface.
       Additional micrographs were taken to investigate washcoat structure
in the small cell segment.  Figure 7-39 shows that variations in surface
texture were apparent which could affect catalyst adherence at the surface.
                                    7-75

-------
                                 R
                                               Flow
                                               direction
Figure 7-30.   Segment  from  inlet  (large  cell)
              monolith,  R =  rear,  F  =  front,
              test model JPL-019.
                    7-76

-------

a.
        *"j 1.0  mm
       16x magnification at
      entrance of segment
 c.   1600x magnification

b.  400x magnification
                                                            Al   Si  Pt
                                                             d.   EDAX scan showing presence of
                                                                 aluminum, silicon, and platinum
Figure 7-31.   Surface analysis  at entrance  of large  cell  segment,  test model JPL-019.

-------
^J

co
                 H 1.0  mm
  a.   16x magnification at
      rear of segment
                       50y
        b.   400x magnification
  c.  1600x magnification


Figure 7-32.   Surface analysis at exit of large cell  segment,  test  model  JPL-019.
d.   EDAX scan showing presence of
    aluminum and silicon

-------
     18x magnification  at
     entrance  of segment
                                                               b.   450x  magnification
c.
1800x magnification                                       d.   EDAX scan  showing  presence
                                                              of aluminum  and  silicon
Figure 7-33.  Surface analysis at entrance of small cell  segment, test model JPL-019.

-------
co
                           . 0 mm
              a.   18x magnification  at
                  exit of segment
      b.   450x magnification
              c.   1800x  magnification
                                                                                   Si
d.  EDAX scan showing presence
    of aluminum and silicon
                Figure 7-34.  Surface analysis at exit of small  cell  segment,  test model  JPL-019.

-------
CO
                                  1.0 mm
                     14x  magnification  at  entrance
                     of segment
        200x magnification - larger
        platinum globules visible
                  c.   lOOOx  magnification  — fine
                      globules  of platinum visible
                                                                                                          F4
d.   BOOOx magnification — small  globules
    of platinum clearly visible
                 Figure 7-35.   Surface appearance at entrance of large cell  segment,  test model  JPL-019.

-------
I
CO
ro
            a.  14x magnification at exit of
                segment
       75x magnification — dispersed
       platinum globules visible
                                                                                                     Rl
                                                                                                     R3
          c.   375x  magnification — large  and
              small  platinum  globules  visible
d.   3750x - platinum globules visible
    in center of photograph
             Figure 7-36.  Surface appearance at exit of  large cell segment, test model JPL-019.

-------
I
co
co
           Al-1  LPt
           Figure 7-37.  SEM/EDAX measurements  on  exit  of  large  cell  segment,  test model  JPL-019.

-------
CO
                Extent of Pt beads
               ••a*
              |Monol ithgp
                 exit
                     a.  16x magnification of
                         small cell  monolith
                 c.  750x magnification —
                     platinum globules distinct
      b.
150x magnification -
platinum globules apparent
                                      10
d.   1600x magnification
    showing large and small  platinum globules
                   Figure 7-38.  Surface appearance at exit of small cell, test model JPL-019.

-------
CO
en
                                                                                                           FM3
                   a.   160x magnification  at  entrance
                       of segment
b.   800x magnification at entrance
    of segment
                                                                                                           RM3
                    c.  T60x magnification at exit of
                        segment
d.  800x magnification at exit of
    segment
                        Figure  7-39.   Washcoat  comparison - small  cell segment,  test model JPL-019.

-------
7.5    CONCLUSIONS
       The large body of data Obtained during combustion screening tests  at
the Jet Propulsion Laboratory and at Acurex has led to the  development of
catalysts that have far better c'ombustion  characteristics than  those  ini-
tially tested.   Further, an extensive fundamental  understanding of the
catalytic combustion phenomenon was  developed.   The tests demonstrated that
catalytic combustion is a high efficiency,  very low emission  process.   More
specifically, there is good promise  of being able  to develop  catalysts with
all the properties enumerated in, the introduction  to this section,  including
low ignition and preheat temperatures, high heat release, and long lifetime.
       The series of catalyst screening tests showed that catalyst perform-
ance could be improved in a number of ways:
       1.  Increased catalyst loading resulted in  lower initial  lightoff
           temperatures, higher mass throughputs,  and increased lifetime  at
           1367K (2000°F)
       2.  Increased cell size resulted in  higher  possible  mass throughputs
           at the expense of increased hydrocarbon emissions.
       3.  Fuel-rich operation allowed lightoff, at lower temperatures  and
           increased mass throughput at a  given preheat temperature than
           fuel -lean operation.
       4.  Heavier hydrocarbon fuels promote lightoff at lower  ignition
           temperatures.
           Hydrogen sulfide ^S)  fixation  of platinum catalysts  promotes
           retention of platinum surface  area.
           Presintering of catalyst washcoats may  reduce  burying  of active
           catalyst below the surface during combustion.
       7.  Stabilization of y-Al203 washcoats  with cesium oxide (Cs20)  up to
           5 weight percent increased surface  area.   Stabilization of alumina
           washcoats with ceria up to 5 weight percent had a negative effect
           on surface area.
       8.  Decreased cell size significantly reduced unburned hydrocarbon
           and carbon monoxide emissions.
                                   7-86

-------
       9.   Catalyst beds of combined .large cell  and small  cell  monoliths
           significantly increased throughput (at a given  preheat tempera-
           ture) and overall catalyst life with  low emissions.
      10.   Bed temperature uniformity was increased by operation at higher
           temperatures.
       In addition to catalyst preparation and operating conditions which
were found to enhance combustion characteristics, the following information
was obtained which impacts ongoing catalyst development.
       1.   Catalyst lightoff temperatures are fairly consistent for a given
           fuel and catalyst type.  672K to 783K is typical  for methane
           on platinum catalysts.
       2.   Higher preheat conditions are required under fuel-lean than fuel-
           rich conditions for the monolithic catalysts.
       3.   Mullite and cordierite perform adequately at 1367K use tempera-
           tures.  Alumina performs well at its  maximum use temperature
           (1783K) but experiences mild thermal  cracking.
       4.   Active platinum migrates and agglomerates at the surface at
           1367K.
       5.   Washcoat presintering reduces surface area but  does  not have a
           negative effect on combustion properties.
       6.   Presintering of washcoat and catalyst results in reduction of
           active platinum available to the reactive stream and hence
           reduces both initial and ultimate activity.
       7.   Catalyst degradation results in large reductions in  both
           active surface area and dispersion.
       The graded cell platinum catalyst represents the best concept developed
by the test series.  It has been shown that washcoat preparation techniques
are of lesser importance to combustion operation than bed geometry and cata-
lyst loading.  The graded cell catalyst also shows promise of increased
activity and lifetime at high operating temperatures (1489K).
                                    7-87

-------
                                 SECTION 8
                        GRADED CELL CATALYST TESTS

8.1    INTRODUCTION
       The experimental  data obtained from catalyst screening tests  (Section
7) verified the high throughput, high efficiency, and low emission character-
istics of the graded cell concept.   The combination of large cells at the
bed inlet to prevent blowout and small cells at the bed outlet for high fuel
conversion provided a significant improvement in catalyst performance.
       Additional catalyst screening tests were conducted at Acurex  on  the
graded cell configuration.  The objective of these tests was the identifica-
tion of the best catalysts that would be appropriate for system development
and testing.  A wide cross-section  of available combustion catalyst  types
was obtained by enlisting the support of catalyst manufacturers.  An initial
goal of 1756K (2700°F) catalyst operation and 1978K (3100°F) support use-
temperature was offered as a guideline.  Discussion of the catalyst  matrix
which follows identifies the manufacturers and catalyst types that were
evaluated.

8.2    GRADED CELL CATALYST MATRIX
       A summary of the sixteen graded cell catalysts prepared is given in
Table 8-1.  These catalyst models were used for the following test purposes:
       1.  Identification of a superior catalyst at small scale by screen-
           ing of manufacturer-prepared catalysts
       2.  Identification of high temperature (to 1978K) capability cata-
           lysts and comparison to  low temperature data
       3.  Scaleup of the best identified screening catalyst to verify
           scaleup criteria
                                    8-1

-------
                                              TABLE 8-1.  GRADED CELL CATALYST MODELS
Sample No.
AERO- 025



AERO-026
AERO-027


AERO-028


AERO-029

«RO-030


AERO-031

AERO-032

AERO-033


AERO-034


AERO-035

AERO-036



AERO-037

AERO-038
.ifPO-W

AERO-141
No. Of
TC
9



6
6


-


6

6


6

6

7


7


7

7



4

5
5

4
Substrate
Hanu. Type
DuPont



OuPont
DuPont


OuPont


Corning

Corning


DuPont

DuPont

Corning


Corning


DuPont

Corning



Corning

DuPont
DuPont

DuPont
Alumina



Alumina
Thoria im-
pregnated
alumina
Zirconia im-
pregnated
alumina
Zirconia
spinel
Zirconia
spinel

Alumina

Alumina

Zirconia
spinel

Zirconia
spinel

Alumina

Zirconia
spinel


Zirconia
spinel
Alumina
Alumina

Alumina
Hashcoat
Manu. Type
Grace



UOP
None


None


None

None


Hatthey
Bishop
Hatthey
Bishop
None


Oxy-
Catalyst

Hatthey
Bishop
Aero



None

None
Johnson
Hatthey
UOP
Rare earth
stabilized
alumina
10-18 Ut. I
Proprietarv
_


_
Catalyst
Manu. Type
Grace



UOP
H. Pfefferle


y. Pfefferle


H. Pfefferle




Proprietary
"

Proprietary




Hatthey
Bishop A
Hatthey
Bishop B
Aero

1
Y-Alumina
4 Wt. %
Aero


Proprietary

Zi rconia
Matthey
Bishop C
Aero
Magnesia
.8-1.0 Wt.%

- Aero

- !
Proprietary

Proprietary

W. Pfefferle
Johnson
Hatthey
UOP
Pt/Ir



Proprietary
Pt/Ir/Os


Pt/Ir/Os


NiO/Pt


°2 3

Stab. Pt

Stab. Pt

NiO/Pt-Pd


C0203/PI


Stab. Pt

NIO/Pt



Co,0,/Pt
C J
C°2°3
roprietary

Proprietary
Catalyst Loading
»t I of GUIS per
Segment Segment
0.6-1.0



-
0.29


0.29


0.29-0 »t
.Z9/.467
2.9 ,11 0
2 4 0 Pt
5.4-7.8
Co203
.33-. 76

.86-1.09

.67-0 Pt
2.27.94,'
2.0 NiO
2.2-0 Pt
9.3/5.67
4.3 C0203
1.9671.27
0.9
1.8/2.1,
2.1 NIO
1.2/0.9;
0 Pt
2.7/2.8;
3.4 Co,0,
4.0/0/6 Pt
15.9/4.1,I5.I
-


.47/.7G/.80



-
.36/.24/.17


.52/.08/.08


0.5/.31/0
.5/. 8/5.0
4. 1/0/0


-

-

15/0.7/0 Pt
5. 0/2. I/
4.5 NiO
3.0/0/0 Pt
7.7/9.97
20.8 C0203
_

3.0/3.2/
4.0 NiO
2.0/1.4/
0 Pt
4.5/4.S/
10.1 Co,0,
6.7/0/O^Pt
10.3/2.9/8.8
-


Dates
Tested
4/25 - 4/29/77



6/24 - 6/29/77
6/30 - 7/1/77


_


7/10 - 7/11/77

7/M 7/21/77


8/2 - 8/10/77

7/26 - 7/29/77

9714 - 9/20/77


9/22 - 9/27/77


10/1 - 10/6/77

10/10 - 10/12/77



12/5 - 12/23/77

12/28 - 12/29/77
1/25 - 2/6/78

12/30/77- 1/3/78
Fuel
Nat. Gas



Nat. Gas
Nat. Gas


-


Nat. Gas

Nat. Gas


Nat. Gas

Nat. Gas

Nat. Gas


Nat. Gas


Nat. Gas

Nat. Gas



Nat. Gas
I
Methane
Nat. Gas
Nat. Gas

Nat. Gas
Test Purpose
Screen Grace Pt/Ir catalyst



Screen UOP catalyst
Determine effects of high melting point precious metals and
catalyst without washcoat

Not to be tested based on results of A-027.


Investigate metal oxide catalyst capabilities (no
washcoat). Perform high temp, operation (3100°F).



Compare to A-032

Screen Matthey Bishop catalyst
1
Fuel nitrogen and pressure testing 1 Test difficulties
\ precluded data
1 results
Comparison to A-033 )


Screening comparison

Fuel nitrogen testing



Fuel nitrogen and pressure testing

Investigate catalyst-support interactions
Screen Johnson Hatthey catalyst

Catalyst scaleup, 6.06-inch diameter
00
I

-------
       4.  Extensive evaluation to investigate fuel nitrogen conversion to
           nitrogen oxides and operational characteristics at high pressure
The tests performed are discussed separately in Sections 8.4 and 8.5.
       Screening catalysts were obtained from six sources, including W. R.
Grace and Company, Universal Oil Products Company, William Pfefferle (a
private consultant), Matthey Bishop, Inc., Johnson Matthey, and Acurex.
Support materials were either DuPont alumina or Corning zirconia spinel.
Washcoats varied from proprietary preparations with high pre-test surface
area to no washcoat and a subsequent low pre-test surface area.  The catalyst
loadings, test dates, and fuels used are listed in Table 8-1.
       In support of combustion test results, catalyst physical measurements
were made both pre- and post-test.  These measurements included catalyst
surface area and dispersion performed in the catalyst characterization lab-
oratory described in Section 5.4.  Additional scanning electron microscopy
(SEM) and energy dispersive analysis by x-ray (EDAX) tests were performed
at the Jet Propulsion Laboratory in Pasadena, California, as required.
Table 8-2 is presented here for reference, summarizing all surface area and
dispersion measurements.  Specific results will be discussed with each cata-
lyst in Sections 8.4 and 8.5.

8.3    ACUREX TEST FACILITIES
       In order to conduct catalyst screening and system development tests,
Acurex designed and constructed a catalytic combustion test facility.   The
facility provides a preheated, premixed fuel/air mixture for combustion
by the catalyst and instrumentation for the monitoring of catalyst perform-
ance.
       Basic facility capabilities include:
       •   Pressure:  1.01 to 10.1x10  Pa (1-10 atmospheres)
       •   Air capacity:  0.042 to 4.49 m3/min (5-540 SCFM)
       t   Preheat:  to 81 IK (1000°F)
       •   Fuel type:  gaseous and liquid
       •   Heat release:  1055 MJ/hr (106 Btu/hr) maximum
                                    8-3

-------
              TABLE 8-2.  SUMMARY OF SURFACE AREA AND DISPERSION  MEASUREMENTS ON  GRADED CELL CATALYSTS
CO
I
-pi
Sample No.
AERO-025
AERO-026
AERO-027
AERO-028
AERO-029
AERO-030
AERO-031
AERO- 03 2
AERO-033
AERO-034
AERO-035
AERO-036
AERO-037
AERO-038
AERO- 040
AERO-041
Surface Area m2/g
Pretest Post-test
1.55-2.49
5.94
0.44
0.06
0.60
0.15
24.38
11.99
-
-
5.17
-
-
0
4.00
6.37
0
0
-
-
0
.01
0
0
-
-
0.09
—
-
-
0
0.50
Dispersion %
Pretest Post-test
1.5-4.9
20.64
8.33
-
-
36.16
9.11
-
-
4.09
-
-
-
-

0
-
-
—
-
0
0.20
-
-
0
—
-
-
—

Catalyst
Pt/Ir
Proprietary
Pt/Ir/Os
Pt/Ir/Os
NiO/Pt
Co203/Pt
Stab. Pt
Stab. Pt
NiO/Pt-Pd
Co203/Pt
Stab. Pt
NiO/Pt
Co203/Pt
C°2°3
Proprietary
Proprietary
SEM/EDAX Results and Comments
No Pt or Ir found on back two
segments by SEM/EDAX

Not combustion tested




Invalid test data
Invalid test data


Zero surface area on each segment




-------
Additional capabilities allow a variety of system testing configurations,
including inert gas dilution, staged combustion, inter- and intra-bed cool-
ing, and fuel doping with nitrogen compounds.
       A schematic of the combustion facility is shown in Figure 8-1, and
photographs of the system components appear in Figures 8-2 to 8-4.   A
1.01 x 10  Pa (150 psia) air compressor delivers the required air flow to a
receiver tank which damps the pressure oscillations at the compressor out-
let.  The air is regulated in pressure and throttled at the tank outlet
prior to passing through a mass flowmeter.  Just prior to entering  the air
heater, inert gas (nitrogen) can be introduced to dilute the air supply.
The electric air heater has an 81 IK maximum temperature capability.
       Fuel and fuel dopants (when used) are injected into the preheated
air stream downstream of the heater.  A sufficient length of pipe (dependent
on fuel injector type) allows thorough mixing before entrance to the com-
bustor.  The combustor is modular, refractory-lined, and water-cooled (as
shown in Figure 8-4) to allow flexibility in test configuration. A quartz
viewport allows visual observation of the catalyst bed during operation.
The exhaust gases are cooled (by water spray) before passing through the back-
pressure valve to the stack.  The backpressure valve regulates system pres-
sure to the maximum 1.01 x 10  Pa capability.
       Facility instrumentation includes pressure, temperature, and flowrate
measurements.  Catalyst beds are routinely instrumented with in-depth thermo-
couples such that bed temperature profiles and histories can be determined.
Optical pyrometry is also used.  Bed temperature acts as the main variable
for system control.  In addition, the catalyst exhaust gases are sampled
and analyzed continuously for COp, CO, Op, NOX, and unburned hydrocarbons.
Table 8-3 lists the on-line analyzer types and normal operating ranges.
Gas chromatography (Carle 8500) is also utilized for analysis of C02, CO,
Op, Np, Hp, CH*, and other hydrocarbons.
       The facility is operated from a remote control room with the power
and flow controls mounted in a single console.  All instrumentation is fully
electronic which allows pressure, flow, and temperature measurements to be
monitored from the control room.  A computerized data acquisition system is
employed to record test data and continuously display real-time conditions.
                                    8-5

-------
oo
 i
01
                                    -^-i-^-HH F1ow t-r-
                                                      Dilutlon
                                                      nitrogen
                                                                    Gaseous and
                                                                    liquid  fuels
                                                                                   Fuel
                                                                                   Dopants
                                                                                Mixing
                                                                                section
K
Combustor
Ua tfl r -M-



A
Stack
-i View
— 1 port
, E
* s
Spray
chamber
                                                                                                                             Emissions
                                                                                                                             sampling
                                                                                                               Back pressure
                                                                                                                  valve
                                    Figure  8-1.   EPA/Acurex catalytic  combustion test facility.

-------
CO
I
                                        Figure 8-2.  Air compressor and  receiver.

-------
CO
co
                                                 MAIN AIR
                                                 PREHEATER
                                               Figure  8-3.   Main  air  preheater.

-------
oo
i
UD
                             Figure 8-4.  EPA/Acurex  test  section  and  exhaust systems,

-------
            TABLE 8-3.   CONTINUOUS GAS ANALYSIS INSTRUMENTATION
       Species
               Instrument Manufacturer
Measurement Ranges
    Carbon
    Dioxide - C02
    Carbon
    Monoxide - CO
    Oxygen - 02
    Nitrogen
    Oxides - NOV
               Intertech 2T

               Intertech 2T
               Intertech 5T

               Air Monitoring,  Inc.  32C
0-5, 0-20 percent

0-500, 0-2000 ppmv
0-5, 0-21 percent

0-10,000 ppmv
The instrumentation is linked to a central  computer system which provides
analog-to-digital  conversion, calculations  and conversions to engineering
units, display capabilities, and permanent  data storage.   The facility oper-
ator keys in commands at a remote CRT terminal to display test variables and
store operating data.
8.4
COMBUSTION SCREENING TESTS
       The manufacturer-supplied graded cell  catalysts were tested under a
uniform test procedure.   This test procedure  included:
       •   Investigation of lightoff requirements with catalyst life,
           various fuels, and preheat.
       •   A 10-hour initial  operation  period to age the catalyst to a near
           steady state level of activity.  Aging was performed at nominally
           1589K (2400°F) and fuel-lean conditions at a heat release rate
           of 105.5 MJ/hr (105 Btu/hr).
       t   Investigation of minimum preheat requirements for both fuel-rich
           and lean operation.
       0   Maximum throughput capabilities  at 1589K, fuel-lean conditions.
       •   Additional  tests based on the initial catalyst performance, includ-
           ing variations of preheat temperature and throughput at higher
           temperatures.
Catalyst operating temperature was normally varied by varying fuel/air ratio.
For operation near 100 percent theoretical  air, diluent nitrogen was used.
                                    8-10

-------
Each of the tests performed is described in the following sections.  The
primary test fuel for combustion testing was natural gas.  References are
made to Appendix B of this report for additional test data.

8.4.1  Catalyst Comparison Tests
       Catalysts were received from several manufacturers and screened by
combustion testing.  The best catalysts were then to be used for system
development phases of the program.

8.4.1.1  W. R. Grace and Company (A-025)
       As shown in Table 8-1, W. R. Grace applied a precious metal platinum/
iridium catalyst on DuPont alumina support with stabilized washcoat.  The
three segments of the graded cell configuration are shown in Figure 8-5.
The metal loadings and surface area and dispersion measurements performed
on each segment at Acurex are shown in Table 8-4.  The dispersion was
quite low, ranging from 1.5 to 4.9 percent for the three segments.  The
catalyst segments were instrumented with a total of nine in-depth thermo-
couples and joined together prior to screening.  Screening tests were per-
formed at atmospheric pressure with natural gas.
       The catalyst performed very well through the first 10 hours of aging,
with some bed nonunifortuities noted.  Following this period, minimum pre-
heat and maximum throughput tests were conducted.  The nominal  test condi-
tion of 1589K bed temperature, 672K (750°F) preheat temperature, and
105.5 MJ/hr heat release rate was repeated periodically to check catalyst
degradation with time.  Bed temperatures in excess of 1756K for extended
periods and space velocities of 380,000 per hour were achieved.  The signif-
icant test points are summarized in Table B-l  of Appendix B.
         TABLE 8-4.  W. R. GRACE CATALYST PRETEST CHARACTERIZATION
Segment
Large cell
Medium cell
Small cell
Noble Metal
(mg)
474
756
797
Noble Metal
(wt %)
0.60
0.99
0.76
BET Surface Area
(m2/g)
1.55
2.49
1.85
Dispersion
(*)
1.5
2.4
4.9
                                   8-11

-------
co
i
(VI
wm
                 Figure 8-5.  W. R. Grace graded cell Pt-Ir catalyst segments.

-------
       Measured emission levels are summarized in Table B-2 and Figure 8-6.
No increase in either CO or NOX emissions was noted over the entire 20-hour
test time at the nominal 1589K test condition.  Slight increases in NOV
                                                                      A
production with increased bed temperatures were noted from the data.   In-
creased throughput did not affect CO and NOX emissions.
       Catalyst lightoff temperature data (Table B-3) indicated degradation
of catalyst activity with time.  The catalyst initially lit off under lean
conditions at 761K.  Following only 2.5 hours of operation, lean lightoff
was no longer possible.  Successive rich lightoff temperatures remained
essentially constant until 1756K bed temperature operation.  The following
lightoff attempt (fifth) resulted in a substantial increase in preheat --
772K under rich conditions.
       In general, the bed operated very nonuniformly even under steady-
state conditions.   Bed temperature distributions and visual observations
reported in Figures B-l to B-7 indicate the existence of relatively inactive
combustion sites under various test conditions.  The nonuniformity of the
bed also impacted minimum preheat and maximum throughput values that  could
be achieved.  Conclusions drawn from the data were:
       •   Bed temperature distributions did not change significantly during
           10 hours of aging
       •   Operation could be maintained at reduced preheat (down to  533K),
           but bed nonuniformities became more pronounced
       •   Throughput could be increased to at least 274.3 MJ/hr (260,000
           Btu/hr) at the expense of bed temperature uniformity
       •   Bed temperatures were more uniform at 1756K than at lower  tem-
           peratures under all conditions
       Although significant nonuniformities were observed throughout  the
test period, no substantial  increases in nitrogen oxide or carbon monoxide
emissions were measured.  This was to be true for several additional  cata-
lysts tested at later dates.  A picture of the post-test catalyst compared
to an untested Grace catalyst is shown in Figure 8-7.  A whitening of the
catalyst is apparent.  Post-test surface area and dispersion measurements
both resulted in zero values, showing washcoat and precious metal sintering
had occurred.
                                    8-13

-------
>
p;
Q.
Q.
ft
CVJ
O
O
o
+J
-o
0)
o
OJ
i-
s-
o
o
in
0° 1
*" 1
i
o
•z.




10
14


12


10


8




6


4

2
o
'. °


—
..
(SfC
^^^
^^^
0 ^"
jQ ^^ ©
®V? 0 x*

^^^
^^
	 . ^* © a&QS)
^^. ^^^^ v^*^
©
-


i i i i i i
2200 2300 2400 2500 2600 2700 2800
°F
llil
     1500
  1600                 1700

Maximum bed temperature, K
1800
Figure 8-6.   NOX emissions corrected to Q% 02 vs.  maximum bed temperature
             for catalyst A-025.

-------
CO




en
                       Figure 8-7.   Pre- and Post-test appearance - W. R. Grace Pt/Ir catalyst.

-------
       Since significant nonunifertilities in appearance were observed,  seg-
ments of the catalyst from relatively active and inactive areas (observed
at 1589K) were sent to the Jet Propulsion Laboratory for SEM/EDAX and
chemical composition analyses.  A pretest segment was also included for
comparison.  The samples of the post-test catalyst were taken from the
middle and aft end segments of the catalyst bed.
       The results of those analyses are presented in Appendix B of this
report.  Included are electron micrographs of the catalyst surface and X-ray
diffraction spectra.  The results indicated that:
       1.  Very little active catalyst (platinum or iridium) existed at the
           surface in either the active or inactive area following testing,
       2.  Platinum existed at 0.19 weight percent in the pretest catalyst
           with only trace amounts of iridium, and
       3.  The remaining catalyst material was well-embedded within the
           washcoat structure.
From these results and the results of the combustion tests, it appears that
the relatively low precious metal loading and subsequent sintering resulted
in zones of low catalyst activity.

8.4.1.2  Universal Oil Products  Company (A-026)
       Test model A-026 was obtained from Universal  Oil  Products (UOP).   A
proprietary catalyst was applied to a washcoated DuPont alumina support.
                                           2
Pretest surface area was measured at 5.94 m/g.   Six thermocouples were in-
stalled for bed temperature measurements.  Lightoff characteristics of the
catalyst were very good (Table B-5).  Initial lightoff was accomplished
fuel-lean at 741K (875°F).  Lightoff after 10 hours operation at 1589K,
however, could only be done fuel-rich.  After 23 hours of testing, rich
lightoff was still possible at only 622K (660°F).
       A nominal test condition  of 1589K bed temperature, 644K preheat
temperature, and 105.5 MJ/hr heat release rate was repeated to check cata-
lyst degradation with time.  Bed temperatures in excess of 1756K and space
velocities of 420,000 per hour were achieved (see Table B-6).
                                   8-16

-------
       Measured emission levels are summarized in Table B-7 and Figures 8-8
to 8-12.  No increase in either CO or NOX emissions was noted over the entire
23-hour test time at the nominal  1589K  test  condition.  Variations in emis-
sions with preheat temperature, stoichiometry, throughput, and bed tempera-
ture were noted as shown.
       The UOP catalyst bed experienced some nonuniformities similar to the
Grace catalyst.  These nonuniformities occurred much later in the test period
(at ~12 hours), however.  Temperature nonuniformities became quite severe by
22 hours of testing, essentially ending testing at the 1756K temperature.
Emissions were not affected significantly:  Resulting bed temperature dis-
tributions are shown in Figures B-13 to B-15.
       The following conclusions were drawn from the data:
       t   The UOP-prepared catalyst had exceptional lightoff characteristics
           over the entire test time,
       •   Operation at low preheats was possible (as low as 394K) without
           significant CO or NOX emissions,
       t   Throughput could be increased significantly at 1589K (to
           420,000 hr   space velocity).  The high throughput at 1589K ac-
           celerated catalyst degradation, however, so that little testing
           could be done at 1756K, and
       t   Bed nonuniformities did not cause significant increases in emis-
           sions.
                                                                      2
       Post-test surface area for the UOP catalyst was found to be 0 m /g.
This result, in consideration of the fact that the catalyst was still able
to light off and operate at the end of the test period, suggested that a
high catalyst surface area was not a significant parameter at lightoff.  It
further suggested that combustion catalysts could be operated under steady
state conditions without high surface area.

8.4.1.3  W. Pfefferle (A-027 and A-028)
       The results of the first two screening tests indicated that catalysts
with low surface areas may operate well under combustion use.  The services
of Dr.  William Pfefferle, a private consultant, were obtained to manufacture
two precious metal catalysts without washcoat.  These platinum/iridium/osmium
                                   8-17

-------
         26 r-
00
oo
     CL
     Q.
     e\j
    o
    o

     o
-a
01
4->
O

-------
                     500 r-
oo


10
                 >  400

                 CL
                 a.
                  evj
                 O
                 O

                 o  300
                 0)
                 -p
                 O
                 O)

                 i-
                 O
                 O
c
O
                 o>
                     200
                     100
                       0

                       200
             105.5  MJ/hr (100,000 Btu/hr)






            01589 K  (2400°F)  lean,  TA =  275%



            Q1756 K  (2700°F)  lean,  TA =  230%
                  300
                                   400
                                                     I
700
800
                                                         I
                              400
                                                                              700
                                        500                   600

                                      Preheat temperature,  K

Figure 8-9.  CO emissions corrected to 0% 02 vs. preheat — UOP catalyst  (A-026),  natural  gas/air

-------
                   c
                   O)
                   


 QJ

 U

 s_
 i.
 o
 O

CM

 o
                   o
    OJ
    o

    o>
    Q.
                   CM
                  O
                      O
                      +J
                      (U
                      +J
                      O
                      d)
                      S-
                      s-
                      o
                      o
    t/>
    •r-


    0>
                   E O
                   CVJ O
                   : o
                         16
14
                          12
                         10
0

200
      105.5 MJ/hr  (100,000  Btu/hr)
      1589 K (2400°F),  fuel-rich
                                       300
                                  400
                                      500

                                       °F
600
700
800
                                  400
                                       500                   600

                                     Preheat temperature,  K
                                                                           700
                           Figure 8-10.  Fuel-rich  emissions corrected to 0%  02  vs.  preheat -- UOP
                                         catalyst  (A-026), natural gas/air.

-------
00
I
ro
   Q.
   Q.
o

o


O)

o

s_

o
u
   o
   O)

   o
                                                                0

                                                                ED
16



14



12



10



 8



 6



 4



 2
        0

       100,000
                                                                 - NO at 1589 K (2400°F)

                                                                 - NO at 1756 K (2700°F)


                                                                 - CO at 1589 K (2400°F)


                                                                 - CO at 1756 K (2700°F)
                            200,000
                                               300,000

                                         Space velocity,  1/hr
400,000
                                                                                                          -I 4000
                                                                                                            3000
                                                                                                            2000
                                                                                                            1000
     0

500,000
                                                                                                               Q.
                                                                                                               Q.
                                                                                                                  CM
                                                                                                                 o
                                   o

                                   o
                                   4->


                                   O)

                                   O
                                   O)


                                   O
                                   O

                                   I/I
                                   c
                                   o
                                                                                                                 o
                                                                                                                 o
                Figure 8-11.  UOP catalyst (A-026) emissions  as  a  function of throughput, natural gas/air.

-------
00
I
ro
ro
                    ex
                    CL


                    oo 6
                   o
                   o

                   o
                   -o
                   O)
                   +J
                   o
                   
-------
                                                     2
catalysts had very low surface areas (0.44 and 0.06 m /g respectively) and
dispersions of 20.64 and 8.33 percent.
       Catalyst A-027 proved to be difficult to light off (778K, fuel-rich
on the virgin catalyst) due to the low loading of precious metals.  Subse-
quent lightoff attempts after 4.5 hours of testing were unsuccessful  at
762K preheat.
       The test points that were achieved are shown in Tables B-8 and B-9.
Initial operation at the nominal 105.5 MJ/hr heat release rate produced
temperature and visual nonuniformities.  This situation was alleviated by
reduction of mass throughput to a heat release of 79.1 MJ/hr.
       It was concluded that the low metal/no washcoat preparation was not
sufficient for complete hydrocarbon conversion.  As a result, testing of
catalyst A-028 was not performed.  The data had not totally indicated the
combustion characteristics of a no-washcoat catalyst since precious metal
loadings were also significantly lower than for previous test models.  Dem-
onstration of catalysts without washcoats was left to later testing.

8.4.1.4  Matthey Bishop A, B, and C Catalysts (A-031, A-032, and A-035)
       Three precious metal catalysts were obtained from Matthey Bishop for
combustion screening.  Test results of the three were similar and are re-
ported below.
       The Matthey Bishop B catalyst (A-032) was the first to be tested.
                                        2
It had a high initial surface area (12 m /g) with a dispersion of 9 percent.
The catalyst was generally unstable during operation, however, and the de-
cay in activity was quite rapid.  After an initial 6 hours of constant lean
operation at 105.5 MJ/hr and 1589 K bed temperature with natural gas, the
bed lost activity and blew out.   Attempts to relight the bed were unsuccess-
ful.   Although some areas of the bed showed activity, uniform combustion
could not be achieved and blowout occurred rapidly at low space velocities
(150,000 1/hr).  The decay in bed activity is shown in the temperature
profiles of Figure 8-13.  A several hundred degree drop in surface tempera-
ture at the front of the bed was experienced within just 2 hours of opera-
tion.  Post-test surface area was reduced to zero with 0.20 percent
dispersion.
                                   8-23

-------
00
I
f\3
                                                                                  2
                                                                                  •
                                                                                           4,5
                                                                                          -•—
t = 3 hours
0726-4
TC
1
2
3
4
5
K (°F)
1208 (1715)
1395 (2051)
1587 (2396)
1607 (2432)
1569 (2365)
t = 5 hours
0726-06
TC
1
2
3
4
5
K (°F)
1079 (1482)
1346 (1963)
1587 (2397)
1607 (2432)
1561 (2350)
                    Figure 8-13.
Catalyst A-032 (Matthey Bishop B) bed temperature distributions
during aging, natural gas/air.

-------
       The Matthey Bishop A catalyst (A-031) showed significantly higher
activity than that of the B catalyst.  Surface area and dispersion were very
              2
high at 24.4 m /g and 36.2 percent respectively.
       The catalyst bed had average lightoff characteristics for a precious
metal catalyst, as shown in Table B-10.  After 16 hours of operation at
1589K, however, lightoff with natural gas was no longer possible.  Although
lightoff was difficult at this point in the test period, the bed still  showed
good activity in terms of throughput.  The data summary in Table B-ll  shows
a high throughput of 6.9 Kg/hr was obtained with natural gas under lean
operation.
       Early in the test period (at 3.5 hours), the catalyst showed the
same unstable operation at 2.1 Kg/hr of natural gas as the Matthey Bishop B
catalyst had shown.  Under steady operating conditions, the catalyst appeared
to blow out and had to be restarted.  Temperature profiles were essentially
constant, however, as shown in Figure B-16.  Activity remained good follow-
ing that incident until 23 hours had elapsed at which time the catalyst
could not be restarted.
       Preheat effects on combustion uniformity were also investigated  for
catalyst A-031.  The results are shown in Appendix B, Figure B-17 and are
typical of other precious metal catalysts tested to date.  Again, the post-
test surface area was reduced to zero for this catalyst.
       Performance of the Matthey Bishop C catalyst was similar to the  A
catalyst in that relatively good lightoff and steady state operating charac-
teristics existed until approximately 20 hours of test time had elapsed.
The catalyst activity had degraded significantly following 20 hours of
testing.
       A summary of the data points taken for catalyst aging, minimum pre-
heat, and pressure operation is given in Table B-12.  Maximum throughput
test points normally included in the test matrix were not obtained due  to
catalyst degradation.
       The generally good lightoff characteristics of the catalyst are  shown
in Table B-13.   These results were similar to those of the Matthey Bishop A
catalyst reported above with lightoff occurring consistently between 700K
                                   8-25

-------
and 767K.  Figures B-18 and B-19 give information on bed temperature dis-
tributions.
       No significant trends in emissions were found with varying preheat,
pressure, or operating time.  Both CO and NO emissions generally remained
below 10 ppm throughout the 21-hour test period.   Aging was found to promote
greater bed  uniformity in temperature, as did lowering the preheat tempera-
ture (at constant fuel rate) on the rich side.   Lower preheat on the lean
side resulted in the expected decrease in bed uniformity.
       The results of the Matthey Bishop catalyst screening tests can be
summarized as follows:
       •   The C catalyst generally performed better than the A and B
           catalysts
       •   Catalyst life was limited to 20 hours
       •   Lightoff temperatures, emissions, and  preheat requirements were
           similar to other precious metal catalysts tested
       •   Catalyst instabilities existed for the A and B models.

8.4.1.5  W.  Pfefferle Co203 Catalyst (A-038)
       Dr. W. Pfefferle prepared a cobalt oxide catalyst on DuPont alumina
support to investigate complexing interactions between the Co^O., catalyst
and support.  Cobalt oxide catalyst interaction with the alumina present in
support materials had been suspected from an earlier Co?0~ catalyst test
(see Section 8.4.2.2).  Four segments were prepared without washcoat, two
of which included calcium oxide to act as a buffer between catalyst and sup-
port.  The cobalt and calcium oxide additions are as shown in Table 8-5.
All four segments were tested as a single bed,  stacked in order as shown in
the table.
       Initially, the catalyst could not be lit off under normal preheat
conditions (to 756K).   A small  amount of platinum was applied to the front
end segment  to facilitate lightoff.  A total test time of 13.5 hours pro-
duced the lightoff history shown in Table B-14.  The added platinum resulted
in lightoff  temperatures very similar to other oxide/platinum combinations
tested.
                                   8-26

-------
              TABLE 8-5.  OXIDE PREPARATION OF CATALYST A-038

Channel Size

Calcium oxide addition, g
Cobalt oxide addition, g
Percent cobalt oxide
Percent calcium oxide
Large


LA
None
10.3
15.9
-

LB
3.7
4.6
6.7
5.4
Medium


M
5.9
2.9
4.1
8.4
Small


S
None
8.8
8.5
-
       Table B-15 summarizes the test data points.  Emissions and bed oper-
ating characteristics were like those of other oxide catalyst tests during
the 10-hour aging period.  Bed temperature distributions at the beginning
and end of aging are shown in Figure B-20.
       At preheats below 533K under fuel-lean operation (1589K bed temper-
ature), the front segments began to break through, producing the bed temper-
ature distributions shown in Figure B-21.  Homogeneous bed reactions could
not be supported under fuel-rich combustion regardless of preheat.
       Since many test points appeared to provide marginal operation at
1589 K, the bed temperature was increased to 1644K for the final maximum
throughput test.  An exceptionally high throughput was achieved (443,100
per hour space velocity) without experiencing blowout.  This throughput
represents a total heat release rate of over 464.2 MJ/hr (440,000 Btu/hr)
and a volumetric heat release rate of 5.10xlQ6 J/Pa-hr-m3 (13,850,000 Btu/
         o
atm-hr-ft ).  The bed temperature distributions at the low and high through-
put test conditions are shown in Figure B-22.
       At maximum throughput, the catalyst came apart and was blown from
the test section by the high velocity gases.  Cracking had occurred as a
result of catalyst/support complexing.  The cobalt aluminate compound formed
by the complexing of the cobalt catalyst with the alumina support at high
temperatures has severe effects 'on support strength.  No apparent differences
were observable on segments where calcium was used as a buffer.  The results
of these tests show that further research on catalyst/support interactions
is warranted.
                                   8-27

-------
8.4.1.6  Johnson Matthey Catalyst (A-040)
       The Johnson Matthey test model (A-040) was a proprietary catalyst
                                     2
with a pretest surface area of 4.00 m /g (see Table 8-2).  The catalyst
operated exceptionally well under fuel-lean conditions but unstably under
rich conditions.  A history of catalyst lightoff is given in Table B-16 and
shows continuous possibility of lean lightoffs at relatively moderate pre-
heat temperatures.  Lightoff under rich conditions generally resulted in
nonuniform combustion and unstable operation.
       Table B-17 summarizes the test points obtained.  Lean operation at
low preheats was quite good.  Total  mass flowrates to 367 Kg/hr were ob-
tained (497.6 MJ/hr heat release rate) without experiencing blowout.
       Emissions data showed somewhat higher NOX values than are typical
for lean-operating catalytic combustors.  In some instances, readings were
not repeatable at similar test conditions at later times.  NOX emissions
were highest during the middle of the test period and lower during early
and late testing stages.

8.4.2  High Temperature Evaluation
       Two graded cell catalyst models were developed for high temperature
operation (to 1978K).  These high temperature catalysts required advanced
support materials such as Corning zirconia spinel with use temperatures
above the 1756K limit typical  of alumina.  In addition, catalysts which
are stable above 1756K are required, with metal  oxides selected as candi-
date catalysts.  The results of the two high temperature catalyst tests
(A-029 and A-030) are summarized below.

8.4.2.1  W.  Pfefferle NiO/Pt (A-029)
       The services of Dr. W.  Pfefferle were used to develop a catalyst for
high temperature applications (catalyst bed temperatures to 1978K).  A
base metal (nickel oxide) was selected for the catalyst since catalyst A-025
had demonstrated that very little precious metal remains after 1756K oper-
ation.
                                   8-28

-------
       The nickel oxide catalyst was prepared on Corning zirconia spinel
graded cell supports, without washcoat, as shown in Table 8-1.  The zirconia
spinel material, shown in Figure 8-14, has a 1978K use temperature.  Plat-
inum was added to the front two segments to enhance lightoff characteristics,
with palladium and iridium added to the large cell segment as well.  The pre-
                                        p
test surface area was measured as 0.60 m /g as shown in Table 8-2.  This low
value was consistent with the no-washcoat application technique.  The rear
segment of the bed (small cell) had fractured during catalyst preparation
but was restored for testing.
       The NiO catalyst proved difficult to light off even with the added
platinum.  Therefore, all catalyst lightoffs were performed with propane at
a bed temperature of approximately 672K.  At approximately 922K, the fuel
was then switched to natural gas to complete the lightoff.  The first suc-
cessful lightoff showed that the rear segment of the bed had refractured
and fallen away from the front two segments.  The test points summarized in
supplement Tables B-18 and B-19 were all taken with the bed in this config-
uration with combustion occurring on the front two segments and downstream
of the catalyst in the gas phase.
       This catalyst was the first to be tested to the maximum use tempera-
ture of the Corning support.  Significant NO emissions (>100 ppm) were
measured above 1880K, as shown in Figure 8-15, and were somewhat higher
than for the precious metal catalysts tested in the range of 1589K to
1756K.  No CO emissions were measurable throughout the test period.   On
subsequent blowout testing, the bed further separated between segments and
became disoriented to the flow direction.  Testing was terminated at that
time.
       In general, it appeared from this test that the metal oxide performed
well at the higher bed temperatures (>1700K) and could be restarted even
after high temperature operation (to 1978K).  The Corning zirconia spinel
support did experience some thermal cracking.  Since only large and medium
cell segments were operational, the catalyst acted as a flameholder for
downstream thermal  reactions.  With this geometry, fairly high levels of
NOX were observed at high temperatures.  The success of high temperature
catalyst operation led to the construction of a second test catalyst.

                                   8-29

-------
co
i
CO
o


                        Figure 8-14.  Corning square-celled extruded monolith structures.

-------
00
I
CO
             400
          Q.
          Q_
          o

          o
•M
u

-------
8.4.2.2  Acurex Co203 (A-030)
       The second high temperature catalyst was cobalt oxide applied to
Coming's zirconia spinel  support at Acurex.  Platinum was also applied to
the front segment to facilitate lightoff.  No washcoat was used in the
                                           o
preparation.  Very low surface area (0.15 m/g) and dispersion (0.01 percent)
were obtained by this technique.
       The cobalt oxide catalyst was screened for a total  of 24 test hours
with natural gas at atmospheric pressure.  Two lightoffs (one on the virgin
catalyst and a second after 7 hours of testing) were accomplished fuel-rich
with natural gas.  Subsequent lightoffs could only be achieved with propane
fuel.  At ~1756K temperature operation, however, the catalyst maintained
uniform conditions over the entire test time.  Lightoff characteristics are
summarized in Table B-20.
       A summary of all test data is given in Table B-21.   Bed temperatures
up to 1978 K and space velocities to 490,000 per hour were achieved.  Bed
temperature profiles varied similar to the lower temperature precious metal
catalysts under varying preheat and throughputs (see Figures B-23 and B-24).
The effects of space velocity and preheat temperature on NO  emissions
                                                           A
under fuel-lean combustion conditions are shown graphically in Figures 8-16
and 8-17.  These curves appear similar to those of precious metal catalysts
tested at 1589K to 1756K.
       Under high temperature operation, NO  emissions of the (^Og catalyst
were substantially lower than those of the NiO catalyst tested previously.
Figure 8-18 shows this distinction in the metal oxide catalysts for refer-
ence.  Since the three-segment Co,,03 catalyst had a much greater surface
area than the two-segment NiO catalyst, it appears that maximizing the
surface reactions (and thereby minimizing gas-phase reactions) minimizes
NO  formation.
  A
       The following conclusions were drawn from the test data:
       •   Low surface area, non-washcoated base metal catalysts have good
           combustion properties
       •   The metal oxide catalysts had good relight characteristics even
           after 1756K operation
                                   8-32

-------
00
I
co
CO
                           40
                           30
                        CU
                        (O
                        ex
                        Q.
                         CM
                       O
O

O
                        O
                        O)
                        S-
                        J-
                        O
                        O
I  10
to
l/l



-------
    35 ,-
    30
-o
cu
i.
3
to
to
CD
00
ra
 a.
 CL
 CM
O
o

o
 QJ
 +J
 O
 o;

 i.
 o
 oo
 c
 o

 CO
 CO

 i
 (1)
25
     20
15
10
       300
                               1811  K (2800°F) lean



                                                0
                           1700 K (2600°F) lean
                         1756  K  (2600°F)  rich
               400
                                               _L
             450        500        550        600

                       Preheat  temperature, K
                                                    650
     Figure 8-17.  Catalyst A-030 (Cc^C^/Pt) NO emissions  corrected
                   to 0% $>  natural  gas/air.
                                8-34

-------
   350
                 Test Model
   300
 CM
O
   250  -
   200
X
CO
o

o
cu
+->
O


t
O
O
S- 150
O
(/)
•i—

§ 100


o
    50   -
  A-025


OA-026


B A-027



  JPL-019
      2000
                                                 3200
          1400      1500     1600      1700     1800      1900


                      Maximum bed  temperature,  K
                                              2000
  Figure 8-18.
NOX emissions comparison corrected  to 0 percent

excess 00.
                                 8-35

-------
       t   Metal oxides show excellent potential for low NOX emissions at
           high temperatures (to 1756K)
       •   High throughputs are possible with metal oxide catalysts in the
           graded cell configuration without experiencing blowout
       t   The increased surface area of the small cell catalyst segment
           had significant effect in reducing thermal NOX emissions
       •   Significant amounts of the metal oxide remain on the support
           material after high temperature operation
       The last conclusion resulted from a deep blue appearance of the post-
test A-030 catalyst, indicating that cobalt compounds were still evenly
distributed over the entire catalyst surface.  It was speculated that the
cobalt had complexed with the support material  to form cobalt aluminate.
This catalyst/support complexing may also impact the thermal shock character-
istics of the ceramic support.
       Post-test surface area and dispersion measurements were not made on
A-030 due to the low pretest values obtained.  Since catalyst activity re-
mained good following screening, the catalyst was later used in fuel  nitrogen
extensive evaluation tests described in Section 8.5.

8.4.3  Catalyst Scaleup
       Based on the results of all  screening catalysts tested, one catalyst
was selected for evaluation of scaling parameters to larger size systems.
It was assumed that combustion throughput would scale proportionately to bed
frontal area.  Therefore, bed diameter was increased to provide a scaleup
factor of 2.7 increase in frontal area over that of small scale screening
catalysts, while bed length remained at 0.0762 m (3.0 inches) of graded
cells.
       The scaleup catalyst was prepared by Universal Oil Products Company
on 0.154 m diameter DuPont alumina supports.  The initial surface area was
                            p
measured at Acurex at 6.37 m /g.  This area was only slightly greater
than the small scale UOP catalyst (A-026) at 5.94 m /g.  Some variations in
preparation technique were reported by UOP based on test results of model
A-026.
                                    8-36

-------
       The  scaleup catalyst was tested for a total period of 27 hours to
determine the catalyst scaling properties.  Test sequences included:
       t    10 hours aging at 1561K (2350°F) with natural gas
       t    fuel-lean minimum preheat operation
       •    fuel-rich minimum preheat operation
       •    blowout at 608K preheat, 1589K bed temperature, 0.101 MPa (1  atm)
       •    blowout at 478K preheat, 1589K bed, 0.101 MPa
       t    blowout at 394K preheat, 1589K bed, 0.101 MPa
       •    blowout at 608K preheat, 1589K bed, 0.303 MPa
       •    blowout at 672K preheat, 1589K bed, 0.101 MPa
       •    fuel-lean minimum preheat at 1055 MJ/hr (10  Btu/hr) heat release
            rate
       The  catalyst lightoff history shown in Table B-22 was very similar to
that of test model A-026, with initial lightoff performed under fuel-lean
conditions  at 722K.  Subsequent lightoffs could only be accomplished fuel-
rich but at preheats as low as 622K.
       A summary of all data points is given in Table B-23.   Screening  test
results were similar to those of test model A-026.  Maximum throughput  re-
ported for  the small scale catalyst was 258.5 MJ/hr (245,000 Btu/hr) and
3.42xl06 J/hr-Pa-m3 (9.3xl06 Btu/hr-atm-ft3) volumetric heat release  rate.
                                                      fi          "3
This compares to a volumetric heat release of 4.38x10  J/hr-Pa-m  (11.9 x
106 Btu/hr-atm-ft3) at 926.3 MJ/hr (878,000 Btu/hr) for the scaleup catalyst
at 672K preheat.  Emission characteristics were also similar.
       A series of blowout tests were conducted to determine the operational
mass throughput limit of the catalyst for varying preheat and pressure  con-
ditions.  The blowout points used are shown in Table 8-6.
       Data points 1  and 4 were compared to determine the effect of pressure
on -blowout, and a maximum fuel  flowrate at 0.101 MPa was calculated for all
points based on a linear variation of maximum fuel flowrate with pressure.
Additional  blowout curves were then calculated using the same linear vari-
ation for operation at 0.136 MPa,  0.202 MPa,  and 0.303 MPa pressures.   Figure

                                   8-37

-------
  40-
  30
  20
S-
-C
S1  10

oJ  9
•M  O
1C  °
                   0.134 MPa

                   0.140 MPa

                   0.195 MPa

                   0.301 MPa

                   Data corrected to indicated pressure
                                   ©
                                                                   0.303 MPa
                                                                   (3 atm)
                                                                   0.202 MPa
                                                                   '(2 atm)
                                                                   0.136 MPa
                                                                    1.35 atm)
                                                                    p = 0.101  MPa
                                                                       (1 atm)
              200
                         300
400
500
600
700
                                                                              800
                                         I
           350       400       450      500        550
                                 preheat temperature,  K
                                                           600
                                 650
                                700
 Figure 8-19.
                 Blowout  performance  -- catalyst  A-041,  TgED  = 1589 K
                 (2400°F),  natural gas and  methane fuels.
                                       8-38

-------
                TABLE 8-6.  BLOWOUT DATA -- CATALYST A-041
Data
Point
1
2
3
4
Bed Temp,
K (°F)
1588 (2400)
1588 (2400)
1588 (2400)
1588 (2400)
Preheat Temp,
K (°F)
608 (635)
478 (400)
389 (240)
603 (625)
Max. Fuel Flowrate
Kg/hr (Ibm/hr)
15.0 (33.0)
10.5 (23.1)
8.4 (18.5)
22.2 (49.0)
Pressure
MPa (atm)
.195 (1.93)
.140 (1.39)
.134 (1.33)
.301 (2.98)
8-19 shows the resultant blowout curves, and also the experimental data for
each pressure.  Two things are shown on Figure 8-19:
       •   Blowout scales linearly with pressure ("ifuei  = Patm x if)fue-|  )
                                                   max
max
 1 atm
       •   Blowout for catalyst A-041 is approximately exponential in pre-
           heat temperature, although a relatively weak exponential factor
           is shown.
       Operation of the catalyst was, of course, possible at any combination
of preheat and fuel flowrate below the blowout curve.  Figure 8-19 can be
employed as a set of design curves for operation of catalyst A-041 under
varying conditions.
       The final minimum preheat test conducted at 1055 MJ/hr (10  Btu/hr)
showed that it was also possible to operate outside of the blowout limit by
reducing preheat temperature at a fixed heat release rate.  (Previous blow-
out tests were conducted by increasing throughput at a fixed level of pre-
heat.)  The observed hysteresis may be of only nominal interest, however,
in system applications.
       The catalyst scaleup testing demonstrated that scaleup is a direct
function of the bed frontal area for the graded cell configuration.  In-
creased mass throughput capability (proportional to frontal area) and sim-
ilar emission levels at all test conditions demonstrated this property.
                                   8-39

-------
8.5    EXTENSIVE EVALUATION TESTS
       Duplicates of selected screening catalysts were constructed for ex-
tensive evaluation testing.  These tests included investigation of the con-
version of fuel-bound nitrogen to nitrogen oxide and the effects of high
pressure operation on combustion characteristics.  Both have implications
in system development where fuels will contain certain quantities of bound
nitrogen and higher pressure operation is required (as in turbine applica-
tions).  The objectives of extensive evaluation were:
       1.  Demonstration of fuel nitrogen conversion characteristics for
           catalytic combustion systems
       2.  Development of operating constraints for catalytic combustors
           in pressurized applications
As shown in Table 8-1, test models A-036 and A-037 were constructed for
additional extensive evaluation.  Catalyst A-030 had previously provided
some fuel-lean nitrogen conversion data, which is discussed below.

8.5.1  Fuel Nitrogen Tests
       Fuel nitrogen tests were conducted with natural  gas.  Ammonia (NHg)
was added in known quantities to simulate fuels of varying nitrogen content.
Exhaust gas analyses for nitrogen oxides (NOX) by chemiluminescent analyzer
and for ammonia (N^) and cyanide (HCN) by specific ion electrode were per-
formed routinely.  Ionic solutions were obtained by bubbling gas samples
through the impinger train shown in Figure 8-20.

8.5.1.1  Acurex Co203 (A-030)
       Following high temperature screening, fuel nitrogen extensive evalua-
tion of the 60203 catalyst was conducted.  The test points achieved were all
under fuel-lean operation, with and without ammonia (NHg) as the fuel  dopant.
The catalyst nitrogen conversion characteristics were  investigated at the
following dopant rates:
       •   5000 ppm NH3 (ppm of fuel)  at nominal space velocities of 40,000,
           150,000, and 250,000 per hour and 1700K, 1811K, and 1922K bed
           temperatures
                                   8-40

-------
00
-p.
                                  Figure 8-20.  Impinger bottle gas sampling  system.

-------
       •   2500 ppm NH3 at 250,000 hr"1 space velocity and 1700K, 1811K,
           and 1922K bed temperatures
       The fuel-lean test results are shown in Figures 8-21 and 8-22 as
measured NOX and percent of NH3 converted to NOX-  The percentage of NH3
converted to NO  in the combustion process is shown to increase significantly
               n
with throughput (space velocity), with dopant concentration, and slightly
with bed temperature.  Restated, low NO  emissions under fuel-lean combus-
                                       X
tion were favored by low throughput rates.

8.5.1.2  Acurex NiO/Pt (A-036)
       A nickel oxide/platinum catalyst was prepared at Acurex and tested
over a range of stoichiometries from 55 to 200 percent theoretical air.
Space velocity was held constant at 100,000 per hour and bed temperature
was nominally 1589K.  Fuel dopant concentration ranged from 2500 to 10,000
ppmv NH3 in the fuel.
       A summary of test model A-036 data is provided in Table B-24.  Sev-
eral points in the data should be clarified.
       1.  Early thermocouple failures required optical pyrometer readings
           of bed temperature.  The bed emissivity was assumed to be 0.6.
       2.  Variations in NO readings with time were noted at the baseline
           5000 ppm NH3 dopant rates in some circumstances.  The variations
           were traced to low flowrate measurement problems.  Larger dopant
           rates were then used to obtain steady readings.
       3.  Test points 1012-11 and 1012-12 were obtained when the bed was
           not fully active (only the front two segments were active).  The
           NH3 conversion data varies significantly from earlier established
           trends.
       The data of Table B-24 is plotted  in Figure 8-23 as the percentage
of the incoming NH3 converted to NH3, HCN, and NO.  The NH3 conversion to
NO increased from zero under very fuel-rich conditions to better than 90 per-
cent on the fuel-lean side.  NH3 conversion to HCN showed the opposite
trend -- high under rich conditions and decreasing to zero on the lean side.
                                     8-42

-------
00


-p»
oo
                      300
                 •o
                 
    200
                      100
                  - Thermal


                  - Total, 5000 ppm



                  - Total, 2500 ppm
                                                   150,000 SV
                                                                 40,000 SV
                               250,000  SV
                        1500
                            2000
2500
3000
                         1100    1200
                       1300     1400    1500    1600    1700


                             Average bed temperature, K
            1800    1900
                    Figure 8-21.
                NO  as measured, catalyst A-030, natural gas doped with ammonia.
                  A

-------
                100
              c
              HI
              o
              
-------
CO

#=>
en
                                                                                                         NO
                       60
80
                           Figure 8-23.
      100         120         140


           Theoretical  air, percent


NH3 conversion characteristics, catalyst A-036, natural gas

doped with ammonia.

-------
Unconverted ammonia was highest below 70 percent theoretical air and de-
creased to low levels under lean combustion.
       The total of these three curves (dashed line and cross symbols) is
considered to represent all NOV precursor species for the NiO/Pt catalytic
                              A
combustor.  A distinct minimum occurs between 70 and 80 percent theoretical
air, where only 20 percent of the fuel nitrogen is converted to NO  pre-
                                                                  A
cursors.
       The partially active bed condition mentioned earlier is also of
interest.  The minimum amount of surface reactions occurring with the small
cell segment not fully reactive showed a conversion of NHo to NO of 100 per-
cent.  It is therefore suggested that the catalytic surface reactions play
a dominant role in minimizing NO  formation under fuel-rich conditions.  It
                                A
should be noted that significant amounts of h^ were measured by gas chroma-
tography for all rich combustion conditions.
       The low fuel nitrogen conversion measured at 70 to 80 percent
theoretical air has important system implications.   Combustors which could
operate fuel-rich, possibly in a two-stage arrangement with secondary air
injection, have potential for very low fuel nitrogen conversions to NOX.

8.5.1.3  Acurex Co203/Pt (A-037)
       A cobalt oxide/platinum extensive evaluation catalyst was tested for
fuel nitrogen conversion and pressure operation, as shown in Table B-25.
No surface area measurements were performed.  Natural  gas was used as the
fuel and doped with 1 to 2 percent of ammonia.
       The fuel nitrogen conversion data is summarized in Table B-25 and
Figure 8-24.  The ammonia conversion to nitric oxide provided the same
curve as the previous nickel oxide/platinum catalyst (A-036).  Differences
in the HCN and NH3 species measured, however, resulted in lower total NOX
precursor (NO + NHg + HCN) levels under fuel-rich conditions.  The minimum
occurred at a lower value of theoretical air (60 percent) than that of the
previous nickel oxide catalyst (75 percent), and the conversion remained low
over a much broader range of theoretical air.  The cobalt oxide catalyst
could thus be operated fuel-rich without dilution to achieve low conversion
of fuel-bound nitrogen to nitrogen oxides.
                                     8-46

-------
         100
CO
          80
      O)
      o

      

      o
      o

       CO
40
          20
                NH3 + HNC + NO
                   NH
                                                                                  NO
                                                                - NO + NH3 + HCN
                                                                                I
                                                                                   I
                   40
                Figure  8-24.
                     60
80
140
160
                                                                                                    - 2 atm
                                                                                          - 3 atm
                                                             J
180
                     100          120

               Theoretical air, percent


conversion characteristics, catalyst A-037, natural gas doped with ammonia.

-------
       The testing included one partially active bed condition at 53 percent
TA when the rear segment had blown out,  As shown in previous test data for
catalyst A-036, the loss of catalytic reactions during the blowout condition
of the rear segment resulted in 100 percent conversion of fuel nitrogen to
NO  precursors.
  A
       Fuel nitrogen conversion at 0.202 MPa and 0.303 MPa (2 and 3 atmos-
pheres) pressures is also shown in Figure 8-24 at 175 percent theoretical air.
A trend to decrease nitrogen conversion with increasing pressure was found.
       Attempts were made to perform blowout tests on the catalyst at both
0.101 MPa and 0.303 MPa pressures.  Blowout was not reached at either pres-
sure at the fuel flow limit of the test facility in the screening configura-
tion.  At 0,101 MPa, a heat release of 480 MJ/hr (455,000 Btu/hr) was achieved
                                                    fi          o
for a volumetric heat release rate of over 7.54 x 10  J/hr-Pa-nr (20.5 million
Btu/hr-atm-ft3).  At 0.303 MPa, the results were 434.7 MJ/hr (412,000 Btu/hr)
and 3.17 x 106 J/hr-Pa-m3 (8.6 million Btu/hr-atm-ft3).
       Based on the results of the fuel nitrogen studies, a data correlation
for fuel nitrogen conversion to NOX under fuel-lean conditions was developed.
The data included both Acurex test results and similar results reported in
the literature.  These included:
       •   Acurex data on a platinum catalyst (A-019).  The fuel used was
           natural gas and methane doped with ammonia.
       •   Acurex data on a nickel oxide/platinum catalyst (A-036).   The
           fuel was natural gas doped with ammonia.
       •   Acurex data on two cobalt oxide/platinum catalysts (A-030 and
           A-037).  The fuel was natural gas doped with ammonia.
       •   NASA Lewis Research Center data (Reference 8-2), consisting of
           two segments of Johnson Matthey metal monolith, the first using
           platinum and the second palladium.  The fuel was No. 2 diesel
           with 135 ppm by weight of nitrogen.
       •   Engelhard Industries data (Reference 8-3).  No information was
           given on the catalyst type; the fuels used were ammonia-doped
           propane and No. 2 oil with 0.94 weight percent nitrogen as
           pyridine.

                                     8-48

-------
       The data was correlated assuming that only space velocity, bed tem-
perature, nitrogen concentration in the fuel, and pressure are significant
variables.  The correlation equation assumed was of the form:

                   % conv = A (SV)B (TBED)C (KQ)D (p)E .

Table B-26 lists the lean data used in the correlation procedure.  Figure
8-25 shows the results of the analysis for A = 1.152 x 10~5, B = 0.7,
C = 1.5, D = 0.17, and E = 0.8.   Considerable scatter from the correlation
is noted.  Neither catalyst type (precious or base metal) or stoichiometry
have been included in the correlation and may be important parameters in
improving the correlation.
       On the basis of this analysis, it appears that high conversion of
fuel nitrogen to NOX is to be expected in lean combustion unless space veloc-
ity, bed temperature, and pressure are all minimized.

8.5.2  High Pressure Tests
       The results of high pressure testing of graded cell catalysts have
been discussed in preceding sections.  They can be summarized as:
       1.  Catalyst throughput capabilities scale linearly with pressure.
           Design criteria for mass throughput for variable pressure and
           preheat were developed for catalyst A-041.
       2.  Limited data on catalyst A-037 showed a slight decrease in fuel
           nitrogen conversion to NOX with pressure increases to 0.303 x 10
           Pa (3 atm).
       The latter result is not consistent with the results of the lean-
combustion, fuel nitrogen conversion correlation of Figure 8-25 where in-
creased pressure was found to increase conversion.  Both  additional  experi-
mental data and refinement of the data correlation are required to resolve
this inconsistency.  Additional  high pressure fuel nitrogen conversion is
reported in Section 9 under system testing.
                                   8-49

-------
00
en
o
        OJ
        o

        

d)
>
c
o
o
        (U
        3
            100
             90
     80
     70
             60
             50
             40
             30
             20
             10
      DATA:



© NASA

Q ACUREXPt


   ACUREXNiO/Pt


   ACUREXCo203/Pt

   ENGELHARD


   A037Co203/Pt
                     10    20    30    40    50    60    70    80    90    100


                Percent  conversion  =  1.152  x  10"?(SV)°"7 (Tbed)1-5 (KN)0'17 (p)°-e
                      Figure 8-25.   Lean data correlation:   conversion  of  fuel  nitrogen  to  NOX.

-------
8.6    CONCLUSIONS
       Graded cell catalyst screening tests have identified many important
performance parameters for different catalyst types in the graded cell con-
figuration.  Specifically, these parameters include mass throughput and
heat release capabilities, emissions under varying operating conditions,
lightoff requirements, and limited lifetime capabilities.  Based on this
data, it is possible to estimate catalyst size, length, temperature, and
operational constraints for specific applications.  Further work is required,
however, to develop optimum catalyst formulation for a given application, as
well as to develop long-life catalyst systems.
       The maximum combustion throughput characteristics of graded cell
catalyst A-040 have been shown to be directly proportional to bed frontal
area and directly proportional to pressure at a given level of preheat.
Further, the maximum throughput was found to increase exponentially with
preheat.  These results allow sizing of the catalyst bed for specific
system applications.  It is certain that other catalysts of different types
will have different performance levels, but it is expected that the scaling
of mass throughput with frontal area, pressure, and preheat will follow
the same relationships.  The maximum throughput levels of many catalyst
types have been identified in the graded cell tests and variations with
pressure, size, and preheat can be estimated.
       The catalyst size  for  an application with given preheat, pressure,
and throughput constraints can be generated.  The final design is also im-
pacted by the catalyst emissions characteristics for the given operating
conditions, operational transients, and fuel properties.  The graded cell
tests discussed in this section have systematically identified catalyst
emissions (primarily nitric oxides, carbon monoxide, and hydrocarbons) over
wide ranges of throughput, preheat, pressure, bed temperature, and catalyst
type.  Emissions with varying quantities of fuel-bound nitrogen have also
been evaluated.  Clearly, catalyst specification for high efficiency, low
emission systems must (as a minimum) include both catalyst sizing and
emission level considerations.
       The catalyst lightoff temperature has been shown to increase rapidly
during early use, and then to remain fairly constant under subsequent
                                    8-51

-------
startups.   Since catalyst lightoff temperature varies with catalyst type,
catalyst surface area and dispersion, and fuel type, as well as with op-
erational  time, it remains difficult to predict.   The use of non-catalytic
lightoff aids in practical combustion systems should be pursued, since low
temperature catalyst lightoff is not required for good high temperature
steady state performance.
       Optimum catalyst formulation and increased catalyst life are two
additional factors requiring further work.   Based on the results of the
graded cell testing, it appears that noble  metal  catalysts on alumina sup-
ports are capable of operation at bed temperatures to 1589K (2400°F), but
degrade rapidly at temperatures above 1589K.   Metal  oxide catalysts placed
on metal oxide supports exhibit material  incompatibilities, causing exces-
sive thermal shock and subsequent support failure.  No apparent degradation
in metal oxide catalyst performance occurs, however.   Thus, it appears that
monolithic systems which include the active metal  oxide in the support formu-
lation would improve both catalyst formulation and catalyst life.
       The following section (Section 9)  describes system application studies
for the graded cell and other catalyst configurations where graded cell data
were used to predict system performance.
                                      8-52

-------
                                REFERENCES
8-1.  Kelly, J. T. et al., "Development and Application of the PROF-HET
      Catalytic Combustor Code," presented at the 1977 Fall  Meeting of the
      Western States Section of the Combustion Institute,  Paper No. 77-33,
      October 1977.

8-2.  Anderson, D. N., "Performance and Emissions of a Catalytic Reactor
      with Propane, Diesel, and Jet A Fuels," NASA TM-73786, October 1977.

8-3.  Carrubba, R. V. et al., "The Proceedings of the NOX  Control  Technology
      Seminar," EPRI SR-39, February 1976.
                                   8-53

-------
                                  SECTION 9
                    COMBUSTION SYSTEM CONFIGURATION TESTS

9.1    GENERAL CONSIDERATIONS
       The design criteria generated for graded cell catalyst configurations
were used in the specification of small scale systems incorporating heat
extraction techniques.  Three system concepts were tested using a variety of
test fuels and catalyst types.  Additional system design criteria were then
generated for advanced system development.
       Two of the small scale system configurations utilized the graded cell
catalyst directly.   The third was based on a cylindrical catalyst geometry
derived from catalyst preparation information obtained from the graded cell
system.  The three systems are:
       1.  The two-stage combustor, utilizing rich first-stage catalytic
           combustion with secondary air injection and interstage cooling
       2.  The gas turbine system, utilizing high excess air levels to main-
           tain low combustor operating temperatures
       3.  The radiative catalyst/watertube system, utilizing intrabed heat
           extraction by watertubes (cylindrical catalyst support)
The design details and experimental results for the three systems are pre-
sented in the following sections.  Appendix C provides supplementary data.

9.2    TWO-STAGE COMBUSTOR
       The two-stage catalytic combustor appears attractive for two reasons.
First, it allows control of bed temperatures to those compatible with the
support material without large excess air requirements.  Second, the first
stage can be operated fuel-rich, which has been shown (in extensive evaluation
                                     9-1

-------
testing) to be advantageous for reduced conversion of fuel nitrogen to
nitrogen oxides.   A two-stage combustor was designed and constructed to
demonstrate these concepts.
       The two-stage combustor is shown schematically in Figure 9-1.  A fuel-
rich mixture is introduced into the primary stage which contains a graded
cell catalyst bed.  The fuel is partially combusted, and the energy released
is removed by an interstage heat exchanger.  Sufficient secondary air is
then injected into the combustion products to complete combustion of the re-
maining fuel in the second stage.  The full system combustor would also in-
clude a second heat exchanger to remove the combustion energy released in
the second stage.

9.2.1  System Design and Fabrication
       The combustor design had to meet several constraints in order to
interface with the Acurex test facility.  A design was sought that would
utilize available test section hardware.  Combustor operating conditions
were selected as:
       •   Overall heat release rate -- 211 MJ/hr (200,000 Btu/hr) and
       •   Bed temperature limit -- 181 IK (2800°F).
       In order to define the temperature control requirements of the inter-
bed heat exchanger, the design curve of Figure 9-2 was constructed.  This
curve identifies the resulting inlet temperature to the second stage cata-
lyst for a given heat exchanger outlet temperature and secondary air injec-
tion temperature.  Two regimes are identified:
       1.  Lightoff, where 811K to 922K (1000°F to 1200°F) temperatures
           are required, and
       2.  Steady state operation where 589K to 811K (600°F to 1000°F)
           preheat is desired to limit second stage flame temperatures.
       In order to control the second stage inlet temperature between the
two regimes, a variable heat exchanger concept was developed.  The heat
exchanger was constructed with two coils.  The primary coil provided
                                     9-2

-------
                                                                          Secondary air
                Fuel
                rich
                mixture
10
co
                                  Primary
                                  stage
Water heat
exchanger
Secondary air
injection
Second
stage
                                      Figure 9-1.   Two  stage  catalytic combustor concept.

-------
1500r
         2000
         1500
1000
      s.
      
-------
sufficient cooling to control lightoff while a secondary coil could be
turned on to provide additional cooling for steady state operation.  The
heat exchanger would thus provide a 1033K to 1256K (1400°F to 1800°F)
outlet temperature.  The coils were constructed of AISI 310 stainless steel
to resist corrosion in the high temperature reducing environment.
       Secondary air injection involved an injector design that would pro-
vide an even distribution across the duct, promote rapid mixing, and avoid
flow separation regions that could result in flame-holding.  Seven conical
nozzles were designed to inject air axially with the first-stage products
at ambient temperature.  Stainless steel (310) was again selected for the
high temperature environment.
       A configuration drawing of the final design is shown in Figure 9-3.
Three available test section spools were utilized, resulting in an overall
combustor length of 1.17 m (46 inches).  All sections were refractory lined
to prevent heat losses.  Cooling water and secondary air were supplied from
available sources.  Photographs of constructed hardware are shown in Figures
9-4 and 9-5.

9.2.2  Test Results
       The two-stage combustor was tested with natural gas at 0.101 MPa and
0.202 MPa pressures (1 and 2 atmospheres).  Lightoff and steady state opera-
tion presented no  unusual control problems.  The combustor was tested at an
overall stoichiometry varying from 70 to 150 percent theoretical air at a
nominal fuel flow  rate equivalent to 211 MJ/hr (200,000 Btu/hr) heat release
rate.  The first-stage stoichiometry was varied from 40 to 70 percent theo-
retical air.  Ammonia was added to the natural gas fuel at a rate of 0.2 to
0.4 percent.  The  test data  is summarized in Tables C-l and C-2.
       Bed temperatures ranged from 1256K to 1660K depending on theoretical
air for a relatively constant preheat of 617K (650°F).  The energy extracted
in the interstage  heat exchanger represents 50 to 60 percent of the combus-
tion energy generated in the first stage.
       The results of the fuel nitrogen conversion data are shown in Figure
9-6 as a function  of overall combustor stoichiometry.  When operating above
                                      9-5

-------
                                                                                                                                                o*n | ffffmtp
 I
cn
               FUOW  —**••>

              DiKsCTiO" —^
                                                                         & J-2 PL.CS
                                                                                    4. SeCONPfcKY AIB INJECTION POKTS

                                                                                      (a FOtZTS <£• CZO" APftET^

                                                                                      fMANSet-t C^_£q END M&UB CONK1. 3 PLCS\
rf
95

§
                                                                                                              SEE S£PAEHTE PACTS LIST PL-12S7— O2O
                                                                                                                 'BISS

                                                                                                                     ]g^^>/
                                                                                                                     51
                                                                                                                       K-f.
                                                                                                                                M
                                                                                                                                    STASE CATALYTIC
                                                                                                                                   50726
                                                                                                                                        T28T-02O
                                                                                                                                    >ii   I*"
                                                                                                                                              _li
                                                  Figure  9-3.   Two stage catalytic arrangement.

-------
            I
Figure 9-4.   Two stage combustor assembly,

-------
I
CO
                                     Figure  9-5.   Two  stage  combustor  details.

-------
   100 r-
S-
OJ

c
o
o

CO
    80
 x
o
&  60
£
O
40
    20
     0
                                       1  Stage
                                                          Q One-stage combustion


                                                          A 50% First Stage  Theoretical Air

                                                          O 60% First Stage  Theoretical Air

                                                          Q 70% First Stage  Theoretical Air

                                                       A O D  1 atm

                                                          A    2 atm
                                        \
                                               .202 MPa
                                                          QJ- 70%
                                               60%
                                                                    J
       50                             100                           150


                      Overall  theoretical  air,  percent


           Figure 9-6.  Two stage combustor fuel nitrogen conversion.
First stage
stoichiometry
50% T.A.

-------
100 percent theoretical air, only nitrogen oxides are normally present.
Under overall fuel-rich conditions, fractions of ammonia and cyanide were
also present.  These results are consistent with fuel nitrogen data ob-
tained on the single cobalt oxide catalyst (Model A-037, Figure 8-24).
The data of Figure 9-6 show a nominal 30 percent conversion rate of fuel
nitrogen to NOX precursors with a value of approximately 27 percent near
overall stoichiometric conditions.  A slight decrease in conversion was
noted at 0.202 MPa pressures.
       It appears as though the primary stage catalyst had slowly degraded
with time, evidenced by the variations in bed temperatures and decreases
in interstage heat extraction of Table C-2 at nearly constant stoichiometry
and mass flowrate.  Degradation may have resulted from either loss of cata-
lyst activity during high temperature exposure or by the deposition of soot
on the surface under fuel-rich conditions.  Table C-2 also indicates de-
creasing emission levels of thermal NOV with time.  The decreased thermal
                                      J\
NO  emissions are generally accompanied by increased CO emissions, indicat-
  s\
ing that a reduction of NOV in the presence of CO is occurring.  There are
                          X
three initially high values of thermal NOV (102, 55, and 28 ppm) during
                                         A
operation of the system under overall rich conditions with very high CO
levels which require further resolution.
       The data shown in Figure 9-6 at approximately 10 percent conversion
levels varied in test conditions from the other data in two respects:
       1.  The first stage was operated at higher values of theoretical air
           (60 and 70 percent) compared with 50 percent for the initial data
       2.  The first-stage catalyst had experienced some sooting by later
           test times when the data were taken, causing the catalyst to
           operate at lower temperatures with less complete combustion
           under the fuel-rich conditions.
The first-stage sooting of the cobalt catalyst proved to be a limiting factor
in the test  life of the system.  The incomplete combustion occurring at the
final test times is evident from the increasing measured carbon monoxide
levels in Table C-2.
                                    9-10

-------
       The demonstration of the two stage catalytic combustor showed a number
of important results.
       1.  The two stage combustor is effective in controlling conversion
           of fuel nitrogen to nitrogen oxides under stoichiometric and fuel-
           lean conditions.
       2.  A slight decrease in nitrogen conversion was found at 0.202 MPa
           (2 atmospheres) pressure.
       3.  The variation of first stage stoichiometry may impact overall
           fuel nitrogen conversion.
       4.  First stage sooting of the cobalt oxide catalyst was a limiting
           factor in combustor operating life and decreased fuel conversion
           rate capabilities.
The identification of a catalyst suitable for first stage operation in the
system is required to obtain additional data with varying first stage stoi-
chiometry and at higher pressures.

9.3    MODEL GAS TURBINE COMBUSTOR
       The gas turbine combustion system was selected for scaleup to a 1056
MJ/hr (10  Btu/hr) heat release rate.  The graded cell  catalyst was demon-
strated to have the low preheat, high heat release, and pressure capabilities
required for this application.  A model combustor can,  fuel  injection system,
and catalyst were prepared.  Subsequent testing was performed at Acurex and
Pratt and Whitney Aircraft (West Palm Beach, Florida) facilities.

9.3.1  System Design and Fabrication
       The model gas turbine combustor was designed to  interface with both
Acurex and Pratt and Whitney test sections whose diameters are 0.22 m and
0.33m(8 and 12 inches) respectively.  The combustor can was selected as a
nominal  0.14 m (5-inch) internal diameter.  A fuel injection system was also
required to provide both gaseous (natural  gas and propane) and liquid (diesel
and No.  2) fuels to the combustor.  The fuel injection  system had to provide
a premixed, fully vaporized fuel/air mixture to the catalyst.  A uniform mix-
ture across the duct area with a minimum mixing length  was also required.
                                   9-11

-------
       Catalyst operating temperature was selected as 1367K (2000°F)  with
air preheat as high as 811K (1000°F), setting the material  thermal  require-
ments for the injector and can.   The maximum heat release rate was  to be
1056 MJ/hr at pressures up to 1.01  MPa (10 atmospheres).   The combustor it-
self did not require pressure design since no differential  would exist
across its surfaces.
       An assembly of the final  design is shown in Figure 9-7 as interfaced
with test sections for the two facilities.  Both combustor  and injector were
constructed of stainless steel.   The fabricated parts are shown in  Figure
9-8 with a detail of the upstream face of the injector in Figure 9-9.
       The fuel injector was a multiple conical tube type similar to a con-
cept developed at NASA Lewis Research Center for automotive gas turbine
applications.  The  incoming air enters the apex end of the cones at high
velocity where the  fuel is injected.  The fuel mixes with the air stream as
it expands through  the cones and is injected into the combustor.  Additional
mixing occurs downstream of the injector exit plane.  Large gaseous fuel
tubes inject axially into the center of the cone inlet.  Smaller fuel tubes
inject the diesel fuel normal to the incoming air for vaporization  and mix-
ing.  Multiple fuel ports were provided to satisfy facility interface re-
quirements.
       The graded cell catalyst used in the system is shown in Figure 9-8.
Six  segments of 0.0254m(1.0 inch) length DuPont AA washcoated alumina Torvex
were coated with platinum catalyst at Acurex.  The six segments were two
pieces of each of the standard 6.84, 5.13, and 3.42 cm cell sizes.   The cat-
alyst was instrumented with six type K thermocouples (chromel-alumel) and
bonded together prior to testing.  A backup catalyst was similarly constructed
by Universal Oil Products Company.

9.3.2  Test Results
       The model gas turbine combustor was first tested at Acurex with pro-
pane between 0.117 MPa and 0.345 MPa (1.16 to 3.42 atm) pressures.  Additional
testing was conducted at Pratt and Whitney Aircraft in West Palm Beach,
Florida at pressures ranging from 0.299 MPa (2.96 atm) to 1.014 MPa (10.04
atm) with propane,  No. 2 oil, and No. 2 oil doped with 0.5 weight percent
nitrogen as pyridine (Cs^N).
                                   9-12

-------
                                                                           GAS TURBINE COMBUSTOR
                                                                                ASSEMBLY
                                                                            CATALYTIC COMBUSTION
Figure 9-7.   Gas turbine combustor  assembly, catalytic combustion.

-------
                           |T':i Will
                          I W .
Figure 9-8.   Model  gas  turbine combustor.
                    9-14

-------
                              ^(C-*—-—
Figure 9-9.  Gas turbine fuel injector assembly.
                    9-15

-------
       Lightoff with propane under fuel-lean conditions was found to occur
repeatedly at a preheat of approximately 644K (700°F).  The Acurex test
points with propane are shown in Table C-3 for varying pressures.  Heat re-
lease rates to 263.8 MJ/hr (250,000 Btu/hr) at approximately 1478K (2200°F)
bed temperature were run as the nominal  test conditions.   No significant
emissions of either carbon monoxide or oxides of nitrogen were obtained.
The fuel injection system and catalyst performed well  at the listed condi-
tions.
       The test data obtained at Pratt and Whitney Aircraft with propane,
No.  2 oil, and No,  2 oil with 0.5 weight percent nitrogen are shown in Table
C-3.  Early propane test points were obtained with an  Acurex catalyst.  Later
propane and all oil tests were conducted with the UOP  catalyst.   Heat release
rates to 844 MJ/hr (800,000 Btu/hr) were achieved with low NOV emissions for
                                                             A
both propane and No. 2 oil.  Some difficulty was encountered with flashback
and flameholding on the fuel  nozzles when running No.  2 oil.  High CO and
unburned hydrocarbon emissions resulted from operating at low bed tempera-
tures (near the breakthrough limit) required to avoid  flashback.
       Tests run with pyridine-doped No. 2 fuel  oil increased the NOV emis-
                                                                    A
sion levels to the values shown in Table C-3.  These emission levels repre-
sent percentage conversions of fuel nitrogen to NO  of 100, 61,  and 55
                                                  A
percent for test pressures of 0.303, 0.505, and 0.707  MPa, respectively.
       The results of the model gas turbine testing exhibited catalyst tem-
perature and pressure operating conditions similar to  those of current tur-
bine combustors at steady state operation.  High mass  rates were achieved
in a relatively small volume combustor.   Overall pressure drop for the com-
bustor and fuel injector were measured at less than one percent at 0.303 MPa
(3 atmospheres) test pressure.
       Close examination of the fuel injector hardware following testing
indicated fabrication errors, resulting in nonuniform introduction of the
diesel fuel at low inlet velocities relative to the air stream.   This low
liquid fuel velocity was responsible for the flashback and flameholding
that occurred.  The velocity mismatch was rectified prior to the next test
series, although total uniformity in fuel distribution among the seven  cones
could not be achieved.

                                    9-16

-------
9.3.3  Advanced Graded Cell  Concept Demonstration
       A modification in the geometry of the graded cell catalyst was con-
ceptualized, based on obtaining an equal number of transfer units in each
segment of the catalyst bed.  The result was a three-segment bed with 6.35x
10"3, 4.73 x 10~3, and 1.80 x 10"3 m (0.250, 0.188, and 0.071 inch) cell
sizes in 0.76, 0.038, and 0.019 m (3.0, 1.5 and 0.75 inch) lengths, respec-
tively.  A proprietary catalyst was applied to Corning zirconia spinel sup-
port by Universal Oil Products Company.
       The advanced graded cell concept was developed for gas turbine appli-
cations (see Section 10,3).   As shown in Figure 10-6, the concept included
an initial large cell catalytic lightoff segment isolated from other ele-
ments by radiation shields.   The concept as tested included only the graded
cell catalyst and the upstream metal radiation shield.  The catalyst was
instrumented with six thermocouples and mounted in the model gas turbine can
for testing.  Figure 9-10 shows the catalyst, radiation screen, fuel in-
jector, and combustor can of the turbine assembly.
       Based on initial test results of the model turbine at Acurex and
Pratt and Whitney Aircraft, design of the fuel injector was modified to
increase local air and fuel velocities.  Both gaseous and liquid fuel
injection capabilities were retained.
       The advanced graded cell catalyst was operated with natural gas at
pressures between 0.117 and 0.824 MPa (1.16 to 8.16 atm) and with diesel
fuel at 0.145 to 0.545 MPa (1,44 to 4.95 atm).  All test points were taken
under fuel-lean conditions.   Ammonia was added to the natural gas at
selected test points to determine fuel nitrogen conversion to NO  charac-
                                                                A
teristies with varying pressure.
       A summary of the test data is shown in Table C-4.  The first series
of test points (0412-02 to -19) was conducted at a nominal 1422K  (2100°F)
bed temperature with ammonia added to the natural gas fuel in various con-
centrations.  A decrease in ammonia converted to NO  with pressure was ob-
                                                   A
served.  The results are shown graphically in Figure 9-11.  These results
are consistent with those obtained at Pratt and Whitney with pyridine-doped
No. 2 oil between 0.101 and 0.707 MPa pressure.  An increase in conversion
                                    9-17

-------
I



CO
                     Figure  9-10.  Advanced  graded  cell/model  gas  turbine  combustor  assembly,

-------
VD
 I
               c
               
-------
at 0.824 MPa pressure was obtained.   Additional  data at these elevated pres-
sures is required to explain the trend noted.  Changes in fuel.nitrogen
concentration produced only small changes in nitrogen conversion at elevated
pressures.
       The  second series of test points (0413-02 to -10) was conducted to
investigate maximum catalyst throughput at 0.303 MPa pressure, 1422K bed
temperature, and 561K (550°F) preheat temperature.   At space velocities
near 200,000 per hour, the catalyst began to break through with increasing
CO and unburned hydrocarbon emissions.   Nitrogen oxide emissions remained
at near zero levels throughout the test.   Full blowout was not achieved as
control of catalyst temperature during breakthrough produced difficulties
in system control.  The maximum heat release obtained was 615 MJ/hr
(583,000 Btu/hr).
       Final tests were conducted with diesel fuel  for comparison of emis-
sions to those of natural gas.  An increased bed temperature was maintained
for the oil tests in order to maintain uniform bed conditions and suppres-
sion of soot formation.  The NO  levels reported in Table C-4 are somewhat
                               A
higher than for natural gas (15 ppm compared to 3 ppm) due primarily to the
small amount of nitrogen in diesel fuel.
       It appears that the shorter small  cell segment of the bed (0.019 m
length rather than 0.051 m) decreased catalytic reactions and allowed a
significantly higher amount of gas phase reactions to occur.  Essentially
complete burnout of CO and hydrocarbons was observed, however.  Increased
combustion  pressure appeared to aid in CO burnout but had no noticeable
effect on NOV emissions.
            A
       The  advanced graded cell catalyst performed similarly to other geome-
tries tested with only minor variations in lightoff, preheat, and through-
put characteristics.  Measured emissions indicate that the 0.019 m length
small-cell  segment may be marginal in providing sufficient catalytic activity
to minimize NO   CO, and hydrocarbon emissions.   The test results did not
              X
conclusively indicate that an equal  number of transfer units in each bed
segment is  advantageous, and additional catalyst geometry testing is required
to determine optimum catalyst configurations.  Further development of fuel
injection for distillate oils is also needed.

                                 9-20

-------
9.4    RADIATIVE CATAL.YST/WATERTUBE SYSTEM
       The objective of the radiative catalyst/watertube system was the
demonstration of stoichiometric catalytic combustion with controlled cata-
lyst temperatures.  High combustion efficiency and low pollutant emissions
have been demonstrated for off-stoichiometric catalytic combustion and were
incorporated into the system design.  The system was also designed to demon-
strate the applicability of catalytic combustion to existing systems, i.e.,
watertube boilers.
       The concept is shown schematically in Figure 9-12.  A stoichiometric
fuel/air mixture is fed to the radiative section which contains a close-
packed array of catalyst elements and watertubes.  The mixture is partially
combusted by the catalyst which is kept at a low surface temperature by
radiation heat loss to the watertubes.  The combustion products and remain-
ing unburned fuel and air are then passed to a dowstream catalytic adiabatic
combustor to complete combustion reactions.  A final convective section is
utilized to extract energy from the fully combusted gases.  Both catalyst
sections operate well below the maximum use temperature of the catalyst sup-
ports -- the radiative section by radiative cooling and the adiabatic section
by dilution of the fuel/air mixture with exhaust products from the radiative
section.

9.4.1  System Design and Fabrication
       Design calculations for the radiative section were performed for one
specific operating condition.  The section was then built and tested to
define its performance for comparison to analytical predictions.  Integra-
tion of the radiative section with the adiabatic combustor and downstream
heat exchanger was not performed.
       System requirements are:
       •   Overall heat release rate of 211 MJ/hr (2 x 105 Btu/hr) — con-
           sistent with volumetric heat release rates established by cata-
           lyst screening tests
       •   Operating pressure of 0.101 MPa (1 atmosphere)
                                   9-21

-------
IN3
        Stoichiometric

        Fuel/air
 8°W
°oom°
    Og?
                                                              Monolith
                          Radiative section
Transition
                                  Adiabatic
                                  comb us tor
                                                 To stack
                                                                                  o   o
Convective
  section
                        0- Catalytic cylinders

                        O - Watertubes
                     Figure  9-12.   Radiative catalyst/watertube combustion  system  concept.

-------
       t   Interface with the existing test facility — appropriate sectional
           area and physical interfaces
       •   Ease of fabrication and low cost
Additional requirements for the radiative section include:
       t   Catalyst elements that can be removed when required
       •   Refractory-lined test sections to minimize energy losses.
       The radiative stage geometric configuration was selected based on its
calculated ability to achieve 50 percent combustion efficiency.  The arrange-
ment is shown schematically in Figure 9-13.  A square-packed array was se-
lected to minimize the number of watertubes required per catalytic cylinder.
Large circular watertubes were selected for ease in fabrication.  A cylinder
length of 0.133 m (5.25 inches) was exposed to the incoming flow.
       Heat transfer performance predictions of the configuration in Figure
9-13 were used to finalize the design.  Critical design parameters included:
       •   Steady state catalyst cylinder surface temperature of 1367K
           at a stoichiometric fuel/air ratio and 211 MJ/hr heat release
           rate
       t   Watertube heat removal rates -- both radiative and convective
       •   Test section refractory thickness to maintain exterior steel
           surfaces at safe temperatures.
       The catalyst cylinder steady state wall temperature can be determined
by equating the convective energy gain to the radiative losses (Figure 9-14).
The convective gain, Q , is given by the difference between the freestream
                      \f
fuel/air mixture adiabatic flame temperature and the actual surface temper-
ature of the cylinder multiplied by the convective transfer coefficient of
the cylinder in crossflow.  The radiative transfer, QR, is dictated by the
cylinder wall temperature, surrounding watertube wall temperatures, and
respective emissivities and absorptivities, since the view factor is essen-
tially unity.  This analysis yielded a 1317K (1910°F) catalytic wall  tem-
perature at stoichiometric mixture ratios and 211 MJ/hr heat release condi-
tions.  This result verified that radiation transfer was an acceptable bed
heat removal technique.  It was then necessary only to define the heat load
to the cooling tubes (combined radiation and convection) and to size the
                                     9-23

-------
                                Watertube
Catalyst cylinder
                 ^-&^
                 -.^0>s\>\J^.!"^^^^^
Flow
                      _ y




                      "/
                         ^;"-'-^!^^)^^
                       .L;v.T^;.^-^:vMrli'-'^^/^X'-^j.7^:r^^'>-M'>-.vr--i-v'^-i^^^c^4ij^'
                  Figure 9-13. Radiative catalyst/watertube arrangement.

-------
test section refractory to complete the design calculations.  Watertube heat
transfer was calculated to be half radiative and half convective with a
total value of 693 MJ/hr-m2 (61,000 Btu/hr-ft2) of tube surface.  The refrac-
tory was sized at 5.08 x 10"2 m thick sidewalls and 2.54 x 10"2 m thick top
and bottom walls for the firebrick material selected.
            Figure 9-14.  Catalyst cylinder heat transfer model.

       The final radiative section configuration is shown in Figure 9-15.
The catalyst cylinders  (1.27 x 10"^ m OD) are supported by the top and
bottom refractory and can be removed by screw access in the top plate.  The
                                           i)
stainless steel watertubes (also 1.27 x 10"  m OD) are fitted at both ends
with smaller tubes to minimize the hole size penetrating through the re-
fractory.  All tube manifolding is flexible hose on the exterior of the
section.  The tubes were manifolded in series to provide two complete flow
paths.  During operation, the water flowrate and inlet and outlet tempera-
tures were measured in  order to determine experimental heat flux to the
tubes.  A picture of the radiative section installed in the test facility
is shown in Figure 9-16.  Thermocouple wires from the catalyst cylinder
surfaces can be seen at the side of the section.
       Thirteen cylinders of Coors alumina were coated with an alumina
washcoat by Oxy-Catalyst, Inc., and with platinum catalyst by Acurex.  A
summary of the catalyst loading for each cylinder is given in Table 9-1.
A nominal loading of five weight percent was achieved.  Cylinder number 237
was used for pretest surface area and dispersion measurements, both of which
were essentially zero.
                                    9-25

-------
                                                                    SEE SEPARATE PARTS LIST -7287-O25
                                                                                E [50726] 7287-OES  {_
Figure 9-15.   Catalytic  radiative System I  assembly.

-------
I
ro
                            Figure 9-16.  Radiative catalyst/watertube test section installation.

-------
           TABLE 9-1.   PLATINUM ON ALUMINA CATALYST CYLINDERS
Cylinder
Number
226
227
228
229
230
231
232
233
234
235
236
2371
238
Original Weight
(KgxlO3)
32.05978
33.23824
33.46360
30.52612
32.21470
33.68275
31.78670
30.82062
31.97873
31.99540
32.65464
33.20850
32.80017
Final Weight
(Kg xlO3)
33.60413
34.87402
34.96856
32.17465
33.74632
35.41054
33.41083
32.34902
33.47561
33.53665
34.17952
34.60557
34.54772
Platinum Loading
(wt %)
4.82
4.92
4.50
5.40
4.75
5.13
5.11
4.96
4.68
4.82
4.67
4.21
5.32
   1
    Used for pretest BET surface area and dispersion measurements.
9.4.2  Test Results
       The objectives of the radiative system tests were to:
       •   Identify the feasible stoichiometric operating range
       •   Identify mass throughput and combustion efficiency
           characteristics
       •   Determine lightoff and preheat requirements
       •   Evaluate the heat extraction technique
The test matrix of Table C-5 was formulated to satisfy these objectives
by varying stoichiometry, fuel flowrate, and preheat.
       Lightoff temperatures for the radiative system were typical of other
platinum catalysts tested.  After 20 hours of testing, the lightoff tempera-
ture under fuel-rich conditions (40 to 50 percent theoretical air) remained
between 700 and 728K (800-850°F).
                                   9-28

-------
       Significant test data is summarized in Tables C-6 and C-7.  A range
of stoichiometries from 50 to 219 percent theoretical air was run by varia-
tion of combustion air at a constant fuel flowrate.  Fuel mass flowrate
was later varied from 2.13 to 6.70 Kg/hr (4.7 to 14.8 Ibm/hr) of natural
gas to investigate the effects of mass throughput at stoichiometric condi-
tions.  Finally, operation at reduced values of preheat was investigated.
Figure 9-17 shows the energy extracted by the cooling tubes out of the
total available energy at the bed inlet as a function of stoichiometric
ratio.  The total available energy includes the fuel heating value (22,000
Btu/lbm) and the sensible preheat energy.  Thermal input to the catalyst
cylinders is primarily controlled by the adiabatic flame temperature of
the fuel/air mixture.  This temperature peaks near unit stoichiometry,
and as a consequence the tube temperatures have a corresponding maximum
As theoretical air percentage increases above 100 percent, catalyst sur-
face temperature begins to decrease, decreasing the radiant exchange to
the watertubes.  The higher total mass throughput, however, also increases
convective heating of the watertubes such that at fixed fuel flowrate the
energy exchange does not fall off rapidly.
       Measured CO and CH^ emissions versus percent theoretical air are
shown in Figure 9-18.  CO levels increase rapidly at approximately 100 per-
cent theoretical air going towards fuel-rich combustion.  Measured methane
(CH^) increased only slightly on the fuel-rich side.  No oxides of nitrogen
were measured for any of the test conditions of Table C-6.
       The methane measurements of Figure 9-18 were made by gas chromatog-
raphy (GC).  Data was taken at several test points to supplement routine
continuous gas analysis for CO, C02, 02, and NOX.  The species normally
detected by the gas chromatograph include CO, C02, 02, H2, N2, CH^, and
other trace hydrocarbons.  A summary of the GC data is given in Table C-8.
       The effects of mass throughput on the radiative system are shown  in
Figures 9-19 and 9-20 for stoichiometric conditions.  The energy absorbed
by the cooling tubes increases with mass throughput due to both increased
radiation (increased bed temperature) and convection.  At the 4.1 Kg/hr
                                    9-29

-------
                 Available energy of fuel and air
1 bU



100



£ 75
-~^
2

i
OJ
0 50


25

i-i
140
120

100

I
O
x 80

.c
3
OQ
. 60
O)
i.
Q>
c=
LU
40 '

t
20
(
0
-
Assuming total
oxygen -^ 	
consumption
r
y
fS
/
X
/
/
/
f^
UJ
/
/
/

\ ^^ *
3^^^ .^^^^^J
'+^6
i i i

^ Assuming total -_
fuel conversion __i— 	 	 "
^_^ 	 — Q 	 ' | Air preheat
,,^-Q1"""" | energy





Fuel mass flowrate = 2.1 Kg/hr






Total energy release to cooling
al_energy extracted J^coohng tubes in radiative sect^n
o

111 111
                           Theoretical air, percent
Figure  9-17.   Radiative catalyst/watertube  system energy  release vs,
               theoretical  air.

-------
oo
           7000
         CVJ
        o
        o
        •u
        Ol
        
a
0}
o
c
o
(J

o
o
           3000
           2000
           1000
                                                              0  CO measurement


                                                              0  CH. measurement
                        60.
                          80
                                                                                          Inlet fuel
                                                                                         concentration
                     25
                                                                                                20  £
                                                                                                    o

                                                                                                    O)
                                                                                                    Q.
                                                                                                       15
                                                                                                       10
180
200
220
                             100       120        140      160

                                 Theoretical air, percent


Figure 9-18.  Radiative catalyst/watertube system emissions vs. percent theoretical  air.
                                                                                                    03
                                                                                                    S-
                                                                                                           0)
                                                                                                           o
                         o
                         u

-------
I
GO
INi
                        400
              400
                        300
              300
              200
                     x
                     s-
                     co
0)

LU
                        200
              100
                        100
                0*-
                             2.0
                      Figure 9-19.
                     3.0
                                                                              Total
                                                                              available
                                                                              energy at
                                                                              100% TA
                                                                                     s
                                                          V
                                                  /

                                        64.4 MJ/hr (61,000 Btu/hr)
                                      / prediction
                                                                                            f^    Cooling
                                                                                           •vr~~ tubes
                                                           A  31.8 MJ/hr  (30,100 Btu/hr) to test  sections
                                                           A  3.69 MJ/hr  (3500 Btu/hr) to stack
8          10           12
Fuel  mass flowrate, Ibm/hr
                                                                                       14
                                                                              16
    4.0
5.0
6.0
7.0
                                                              Kg/hr
               Radiative catalyst/watertube energy release vs. throughput
               at 100 percent theoretical air.

-------
OJ
CO
              CM
             o
             o

              o
             T3
              CD
             4->
              O
              O)
              S-
              S-
              o
              o


              a.
              o.
              5-
             4-3
              sr
              cu
              o
              o
             o
                  12,000
                  10,000
8,000
                   6,000
4,000
                   2,000 -
                                                          0	
                     0
                                                     Maximum Measurement
                                                     Capabi1ity
                                                         Inlet Fuel
                                                         Concentration
                                                0   CO  Measurement

                                                CD   CH»  Measurement
                                                                            0
                                                                         Exhausted Fuel
                                                                                          TIT
                                                             10

                                                            Kg/hr
                                                      12
14
                        Figure 9-20.  Emissions  vs.  throughput at 100 percent  theoretical  air.
                                                                                  12
                                                                                  10
                8
                  CD
                  O
                  i.
                  CU
                  Q.
                                                                                     O
                                                                                     cu
                                                                                     (J
                  o
                  u
               4S
16
                                                                                                     0

-------
design point, the actual cooling of 57.0 MJ/hr (54,000 Btu/hr) extracted
by the watertubes is slightly below the predicted value of 64.4 MJ/hr
(61,000 Btu/hr).  At the 4.3 Kg/hr test point, an additional 31.8 MJ/hr
was extracted by a separate cooling system for the surrounding test sec-
tions, and 3.69 MJ/hr was exhausted at the stack.  Therefore, the combus-
tion efficiency at the design point can be calculated as approximately
37 percent.  The emission levels of Figure 9-20 show a fairly even level
of unburned hydrocarbons with carbon monoxide increasing significantly
above 2.7 Kg/hr of natural gas.  This trend is consistent with the reduced
efficiency at higher mass throughputs shown in Figure 9-19.
       Typical bed temperature profiles are shown in Figure 9-21.  Note the
increase in catalyst surface temperature with throughput due to increased
convection to the catalyst cylinders.  The temperature observed near the
front of the bed at the nominal 4.3 Kg/hr design point matches the predicted
temperature of 1317K (1910°F) quite well.  Preheat temperatures as low as
394K  (250°F) were achieved with no effect on combustion stability.
       A second test series was conducted with the radiative catalyst/water-
tube system with a base metal oxide (cobalt oxide) catalyst.  Cobalt oxide
was applied to Coors alumina cylinders at Acurex.  Three cylinders at the
entrance region of the bed had platinum added to facilitate system lightoff.
The configuration was tested at atmospheric pressure with natural gas and
propane.
       Attempts to light off the system with natural  gas at 694K (790°F)
were unsuccessful.  Lightoff with propane at 672K (750°F) showed initial
catalyst activity as the bed temperature on the upstream cylinders increased
to 950K (1250°F).  Full lightoff could not be achieved, however, even at
stoichiometric conditions.
       The results of these tests and test data obtained for cobalt oxide
catalysts in the graded cell configuration confirm that near-adiabatic sur-
face conditions or higher ignition temperatures are required for lightoff
of base metal oxide catalysts.   The required heat retention at the surface
was not available in the radiant system, making lightoff very difficult to
achieve.  It had also been observed that operation of cobalt oxide graded
                                    9-34

-------
CO
on
                   2500
           1500
                   2000
                   1500
           1000
                   1000
                         Preheat
            500
                    500
                            -2
                -0.1
                                                                            O 2.1  Kg/hr nat. gas

                                                                            Q 4.3  Kg/hr nat. gas

                                                                            & 6.7  Kg/hr nat. gas
2468


   Distance from bed inlet, inches
                                                                                           10
                                                           0.1
                              0.2
                                                              Meters
12
                                                                                                                14
                                                                                                    0.3
16
                  0.4
                         Figure  9-21.   Bed temperature  profiles at  100  percent theoretical  air.

-------
cell catalysts (model A-038) was less stable at the low surface temperatures
of the radiant system (1367K) than at higher temperatures.  This low activ-
ity of the cobalt catalyst at low temperatures explains the lightoff diffi-
culties experienced in the non-adiabatic watertube system.
       A third test sequence was conducted with the original platinum on
alumina cylinders used for initial system testing.  Test objectives were
to evaluate fuel nitrogen conversion characteristics of the system over a
range of stoichiometries for comparison to the graded cell configuration.
Stoichiometry was varied from 52 to 120 percent theoretical air.  Natural
gas and propane were used as test fuels.
       The nitrogen conversion test data is summarized in Tables C-9 and
C-10 and in Figure 9-22.  The test points were run in the order shown in
Table C-9.  Following natural gas operation at 75 percent theoretical air,
catalyst degradation was apparent.  Operation below 75 percent or above
120 percent theoretical  air with natural gas was no longer possible.   Op-
eration with propane fuel allowed additional test points below 75 percent
theoretical air, but some cylinders were observed to be inactive.   Higher
thermal NOV values (Table C-10)  over those of the original test series re-
          A
suited due to the loss in surface activity and increased nitrogen oxidation
in the gas phase.  The poorer fuel conversion obtained (Table C-9 shows a
maximum conversion of 22.7 percent compared to an initial  37 percent) also
indicates the loss in catalyst activity with time.
       Figure 9-22 shows the fuel nitrogen conversion characteristics of
the radiative system for natural gas doped with 2000 ppm of ammonia.   Low
NO  and high NH., values above 100 percent theoretical air are consistent
  X            «j
with the incomplete combustion characteristics of the radiative system.
The low point in the total NO  precursor curve (Nl-U + HCN + NOV) at 60
                             X                    j           X
percent theoretical air is similar to those obtained for metal oxide graded
cell catalysts (see Section 8.5.1).  Although operating under fuel-rich
conditions, excess oxygen is still present in the exhaust from the radia-
tive system due to incomplete conversion of the fuel/air mixture.  The
low conversion of the fuel nitrogen to NO  precursors even in the presence
                                   9-36

-------
I
co
          100
           80
      4J
      C
      <3J
      (J
      S_
      O)
      X

       o
c
O

t/1
5-
(LI
       o
       O
           60
           40
       ro
           20

                                    '*

                                     0
                                                  0
                                                              0

                                                                       0
                                                  ©
                                                                  Legend

                                                            O  NH3

                                                            Q  HCN


                                                            A  NO

                                                            O  NH3 + HCN + NO
                                                                                     110
                                                                                           120
                                   80           90           100

                                    Theoretical  air,  percent

Figure 9-22.  Radiative catalyst/watertube system,  ammonia-doped  natural  gas,  fuel  nitrogen
              conversion.
130

-------
of oxygen suggests the action of a NOV reduction mechanism by the CO and
                                     A
unburned hydrocarbons in the exhaust stream.   The data scatter in the 80
to 100 percent theoretical  air range should be resolved with additional
testing.
       Since the actual measured heat release of the radiative catalyst/
watertube system was not as high as that predicted at the nominal 4.3 Kg/hr
(9.5 Ibm/hr) of methane design condition (see Figure 9-19), the radiant sec-
tion as tested is not fully suited for complete system development.  The
addition of the downstream adiabatic catalytic combustor would result in
too high a temperature in that region at stoichiometric conditions since
combustion efficiency in the first stage is not as high as expected.  The
excellent performance of the radiative catalyst/watertube section at stoi-
chiometric conditions with very low levels of NOX makes it attractive for
further optimization to increase efficiency and compatibility with the
adiabatic section.  The system also appears attractive for fuels with high
nitrogen content.  The extremely stable operation of this first radiative
system under all test conditions makes it an ideal candidate for life test
considerations.

9.5    CONCLUSIONS
       The combustion system data obtained from tests of the radiative
catalyst/watertube, two-stage, and gas turbine systems established three
potential applications of the catalytic combustor.  The control of nitrogen
oxide emissions from both thermal fixation and conversion of fuel-bound
nitrogen was also shown.  The emission characteristics of nitrogen oxides,
carbon monoxide, and unburned hydrocarbons were evaluated over a range
of operating conditions.  The results thus provide design criteria for
system combustors and their operation.
       The radiative catalyst/watertube system exhibited stoichiometric
catalyst operation by direct bed cooling with potential for both low thermal
and fuel NO  emissions.  It is apparent that the thermal NOV emissions
           A                                               X
obtained are sensitive to the activity of the catalyst surface.  Thermal
NOX increased significantly during later test times when deactivation of
                                   9-38

-------
the surface was observed -- allowing greater oxidation of atmospheric
nitrogen by gas phase reaction.  The conversion rate of fuel nitrogen to ..
NOV appears to be influenced by the presence of CO and UHC species which
  A
reduces the NO,to N9.  Therefore, the partial combustion characteristic
              A    L-
of the radiative section may be very attractive for the control of fuel
NOV.  Coupled with the downstream adiabatic combustor, full fuel conver-
  /\
sion can be achieved.  Advanced designs are required to achieve greater
first stage efficiency and operation of the complete system concept.  Sys-
tem stability during operation makes it suitable for constant operation
steam raising applications.
      The two stage combustor combined the advantages of low fuel nitrogen
conversion under fuel-rich conditions with high overall system efficiency
achieved at stoichiometric conditions.  Seventy percent control of fuel NO
conversion was achieved by the two stage combustor tested.  It appears
that even higher levels of control can be achieved by optimization of first
stage stoichiometry and by operation at higher pressures.  Production of
thermal NO  by the system appears to be dependent on both catalyst activity
          X
and emission levels of CO.  Reduction of NOV formed during combustion by
                                           A
CO appears probable under overall lean combustion but ineffective for over-
all fuel-rich operation.  These combined properties make two stage combus-
tion attractive to a number of applications (see Section 10).  Advancements
in first stage catalyst application are currently required to achieve long
life in that stage without the effects of sooting.  Consideration of mixed
catalytic and conventional burners in two stage arrangements should also
be investigated.
      The model gas turbine system shows the direct application of the
graded cell concept to turbine systems.  Exhaust temperature control by
high excess air levels was demonstrated at high volumetric heat release
rates for both gaseous and liquid fuels.  Low thermal NOV emissions at up
                                                        A
to 1.01 MPa (10 atmospheres) and decreasing conversion of fuel nitrogen
with increasing pressure were observed.  Advancements in fuel injection
system design, graded cell design, catalyst support, and flashback con-
trol contributed to the success of the model turbine system.
                                   9-39

-------
       Advanced design and testing of each of the three systems is required
to obtain complete design data for prototype development.  Advanced develop-
ment data would lead directly to laboratory installation of prototype sys-
tems for long-term demonstration and finally to field process applications.
Projected field applications for catalytic combustors based on the results
of the system tests of this study are presented in Section 10.
                                     9-40

-------
                                 SECTION  10
                     PROTOTYPE  SYSTEM DESIGN CONCEPTS

10.1    INTRODUCTION
       The characterization of  stationary combustion systems to which
catalytic combustor  retrofit and/or redesign may be applicable (Section 3)
and the catalyst performance test data generated under this program (Sections
7 to 9) have provided information for conceptualization of catalytic combus-
tion systems.   Those systems which appear most promising for future applica-
tion include industrial  and commercial firetube and watertube boilers, and
gas turbines.   Other applications, including utility boilers and mobile gas
turbines, as well  as rangetop burners and home furnaces (which are less fre-
quently maintained), also appear feasible for catalytic combustors but would
require more extensive development.  Prototype catalytic concepts for com-
mercial and industrial boilers  are discussed in Section 10.2, for stationary
gas turbines in Section 10.3, and for other systems in Section 10.4.

10.2   INDUSTRIAL AND COMMERCIAL BOILERS
       Both firetube and watertube boilers appear to be attractive applica-
tions for catalytic  combustion.  The concepts that are developed below apply
equally well to both industrial and commercial sized units.

10.2.1   Firetube Boilers
       The firetube  boiler utilizes radiative and convective heating by
combustion products  on the inside of tubes immersed in water.  A thorough
discussion of firetube boiler types is given in Reference 10-1.  A typical
scotch  firetube boiler is shown in Figure 10-1.  A relatively large combus-
tion chamber exists  at the center of the  unit and is fed directly by the
                                    10-1

-------
                            Flue
                                Firetubes, showing flue gas
                                directions 	
Blower
      Figure 10-1.
Four-pass scotch firetube boiler (courtesy of the
Cleaver Brooks Company).
                                  10-2

-------
burner.  The conventional  burner could be simply replaced by a redesigned
catalytic burner with no additional system modification necessary.
       Two firetube boiler conceptual design concepts were completed.  The
first utilizes a graded cell catalyst which has been shown by screening
tests to possess appropriate throughput, preheat, and emissions character-
istics.  The boiler burner would be replaced with the graded cell burner as
shown in Figure 10-2.  The graded cell burner has a preheat section in which
air and/or fuel is passed through the inside of a set of cylindrical spines
to provide the necessary preheat.  This preheated air and fuel is then mixed
and passed over the catalytically treated exterior cylindrical surfaces
where combustion is initiated, thus providing the preheat energy.  Following
this initial section is a region in which heat transfer occurs between the
partially combusted products and the furnace chamber.  A graded cell cata-
lyst then accepts the cooled, partially burned mixture and completes combus-
tion.  The products of combustion are then passed down the remaining central
furnace chamber.  The heat transfer rate to this water-cooled furnace is con-
trolled by the center ceramic flow restrictor.  Proper sizing of this cylin-
drical support (flow restrictor) would control the exhaust product velocity
and, hence, the heat transfer rate.
       This design is attractive in that it utilizes the extensive test ex-
perience and property knowledge of the graded cell configuration and would
be a simple retrofit.  The actual geometry of the catalyst would require
further optimization for operation at the appropriate stoichiometry and for
full fuel conversion in a minimum volume.
       The second firetube boiler concept utilizes a felt-like matrix mate-
rial as the catalyst element.  This material has been shown to operate
effectively in residential furnace application (Reference 10-2) and should
be directly applicable to boiler systems.  In this concept (Figure 10-3),
a gaseous fuel/air mixture is passed down a metering manifold located
in the center of the cylindrical felt-like matrix shell.  The fuel/air
mixture passes radially outward and diffuses through the matrix material.
Combustion occurs on the outer matrix surface.  This surface then radiates
its energy to the water-cooled furnace wall.  The concept again utilizes a
proven catalyst type (felt-matrix) and would be a simple burner retrofit.
                                    10-3

-------
                                 GRADED CELL
                                 COMBUSTOR
          RADIATIVE CATALYTIC
          FUEL/AIR COMBUSTOR AND
          PREHEAT SECTION
STOICHIOMETRIC
PREVAPORIZED
FUEL/AIR
MIXTURE
                                                                                   CENTER CERAMIC
                                                                                   SUPPORT AND
                                                                                   FLOW RESTRICTOR
                                                                                   STRUCTURE
FIRETUBE
COMBUSTION
CHAMBER
                                          FUEL/AIR MANIFOLD
                                          SECTION MOUNTED
                                          IN CENTER CAVITY
                              Figure 10-2,   Graded  cell  firetube  boiler  concept.

-------
o
en
                            FELT-LIKE MATRIX
                            RADIATIVE
                            BURNER
                  FUEL AIR
                  MIXTURE
                                                                                               FIRETUBE
                                                                                               COMBUSTION
                                                                                               CHAMBER
                                        PRE-VAPORIZED FUEL/AIR MANIFOLD
                                        Figure 10-3.   Felt  pad firetube boiler concept.

-------
10.2.2  Watertube Boilers
       In the watertube boiler, hot combustion products contact the boiler
outer tube surfaces and water is heated within the tubes.  The radiative/
watertube catalytic combustor concept discussed in Section 9 utilizes radia-
tive and convective heat transfer within the bed to heat water-carrying tubes
       As described in Section 9, the radiative/watertube system achieves
partial combustion of the fuel at stoichiometric conditions.  Some geometric
optimization is possible to increase the amount of combustion occurring
within the radiative/watertube system, but it is expected that, for a short
system length, a secondary graded cell stage will  be necessary to combust
remaining hydrocarbons following first stage heat release.
       Two concepts are presented here for the watertube boiler.  The first,
Figure 10-4, assumes that efficient combustion can be achieved in a single
radiative combustor, A.  Heat transfer to watertubes occurs both within the
bed (to keep bed temperatures at acceptable levels during stoichiometric
combustion) and in a downstream convective section, C, consisting of addi-
tional watertubes to extract the remaining combustion energy.  Considerable
extension of existing bed heat transfer experience would be required to fully
implement the concept and achieve efficient single stage combustion.  Simple
retrofit of existing horizontal straight or bent tube boiler units would not
be possible, and new designs would be required.
       The second watertube boiler concept would add a graded cell catalyst
stage B to the system, Figure 10-4.  This second stage would be capable of
converting all unburned hydrocarbons passed by the radiative/watertube stage.
The final convective watertube section would be downstream of the second
stage to extract remaining heat from the combustion products.  The concept
has the advantage of less stringent requirements on the radiative/watertube
stage performance and hence less development time for that stage.  Again,
as with the single-stage concept, full boiler redesign would be required.

10.2.3  Two Stage Catalytic Systems
       Graded cell catalyst extensive evaluation tests (Section 8) and sub-
sequent two stage combustor system tests (Section 9) demonstrated that two
                                    10-6

-------
                                         .Steam
                                          drum
             -Refractory lining
o
 i
                                             Catalyst
                                             coated cylinder
                 Hud
                drum
                                                                      Radiative heat
                                                                      transfer section
                                                                                               . Monolith bed
                                                                                               - Adlabatlc
                                                                                                Combustor
                                                                                                                   Convectlve heat exchanger
                                             Hater-tube
                                  Figure  10-4.   Radiant  catalyst/watertube  combustion  system.

-------
stage combustors operating fuel-rich in the primary stage show promise for
reduced conversion of fuel-bound nitrogen to nitrogen oxides.  Two stage
catalytic combustors can be conceptualized that would be applicable to both
firetube and watertube boilers.   The only difference for the two applications
is the geometrical constraint involving energy extraction following the first
and second stage beds.
       The two stage combustor with graded cell catalyst tested in this
program (see Figure 9-3) is applicable to the watertube boiler since inter-
stage heat removal occurs by watertubes.   A second convective heat recovery
exchanger downstream of the second stage would also be required for added
energy extraction and increased boiler efficiency.  Secondary air injection
complicates application of the concept to the boiler system.   A two-stage
combustor could also be built for watertube boilers by use of the radiative
catalyst/watertube system where the radiative zone operates under fuel-rich
conditions and secondary air is added prior to the adiabatic combustor.
Special fuel delivery and injection systems would need to be developed for
high nitrogen fuel oils for this application.
       The two stage combustor could also be built within the long combustion
chamber of the firetube boiler with the hot gases contained within the boiler
tubes.  Again, secondary air injection poses a difficult design problem which
would probably rule out burner retrofit of existing units.

10.3   GAS TURBINES
       The stationary gas turbine has long been considered the natural appli-
cation for the catalytic combustor since large excess air levels can be
utilized to maintain appropriate catalyst bed temperatures and control tur-
bine inlet temperature levels.  In general, the graded cell catalyst could
be applied directly to the gas turbine combustor provided sufficient air
preheat is available.  The graded cell system has been demonstrated to be
operable over a wide range of temperature, mass throughput, and varying pre-
heat conditions.  Demonstrated high volumetric heat release rates also show
that a relatively compact combustor can be achieved.
       A concept for a canannular gas turbine combustor where the conventional
burner is replaced by a catalytic one is shown in Figure 10-5.  Stationary
gas turbines are predominantly of the canannular type.  Only the  internal
                                    10-8

-------
                                          GRADED CELL
                                            CATALYST
COMPRESSED AIR
                                TRANSITION PIECE
      TO
   SECOND
    STAGE
FIRST STAGE
   ROTOR


FIRST STAGE
  NOZZLE
             Figure 10-5.  Catalytic gas turbine concept.
                              10-9

-------
details and the fuel  injection for the combustion can are modified.  A
special fuel injector to provide even fuel/air distribution without swirl
would be required.  The fuel would be premixed with all combustion air
rather than adding additional secondary air as in present turbine systems.
The manifold and turbine nozzles would not require modification.
       The above concept appears feasible based upon experimental data
obtained on the model turbine combustor tested under this program (Section
9).  A graded cell monolith in a single can configuration was tested from
0.101 MPa to 1.01 MPa pressure with propane and from 0.303 MPa to 0.707 MPa
with diesel fuel.  Heat release rates as high as 844.8 MJ/hr (800,000 Btu/
hr) were achieved.  Simple modifications to the combustor and fuel injection
system could be made for stationary gas turbine applications.
       The graded cell catalyst design used in the model turbine could also
be improved to achieve maximum conversion efficiency in a small combustor
volume with a reasonable factor of safety.  One possible concept for a more
optimum gas turbine graded cell catalyst is shown in Figure 10-6.  This sys-
tem includes a large cell section for system lightoff and preheat.  Forward
of this preheat section a high emittance radiation shield is located which
will absorb the radiant energy for this section and provide both preheat
to the reactants and flameholding of possible flashbacks.  Aft of the large
cell preheater is a high reflectance metallic radiation shield which reduces
radiant energy transfer from the primary bed to the preheat section, thereby
minimizing the initiation of gas phase reactions in this region.  The pri-
mary combustion section consists of a three-segment graded cell catalyst bed.
All dimensions would be analyzed to provide the most efficient combustor.
        Concepts for the larger cell preheat section are shown in Figure 10-7.
  Concept A  is a large cell  (0.0127 m) opening with 0.00318 m blunt edges to
  provide for lightoff with  low transfer coefficients.  Concept B would utilize
  smaller cells with larger diameter openings.  The smaller cells would not
  support surface reactions, whereas the larger diameter openings would supply
  the preheat combustion.  The third concept is a staggered tube arrangement,
  similar to the radiatively-cooled combustor described in Section 10.2.2.
        Further development of the graded cell system  for gas  turbine applica-
  tion could be aided by PROF-HET code predictions of catalyst  performance  for
                                   10-10

-------
                                    HIGH EM1TTANCE
                                    RADIATION SHIELD
                                                                                 GRADED CELL
                                                                                 CATALYST BED
o
i
PREMIXED
FUEL/AIR
  INPUT
                                                                                                          COMBUSTION
                                                                                                          PRODUCTS
                       0.013m OPEN LARGE
                       BLUNT EDGED LOW
                       TEMP. LIGHTOFF AND
                       PREHEAT SECTION,
                       0.0095m THICK
                                                          HIGH REFLECTANCE
                                                          METAL RADIATION
                                                          SHIELD
                                                                     0.089m ALUMINA
                                                                     SHROUD TUBE
                       Figure 10-6.  Stationary gas turbine graded cell  catalytic combustor.

-------
o
 I
                       Concept A
                    Large cell (0.013m)
                      blunt edge bed
    Concept B
Combined small cell
 large opening bed
  Concept C
Staggered tube
    bed
                                   Figure 10-7.   Low  temperature lightoff/preheat section.

-------
various geometries.  Experimental development would further prove the validity
of various design parameters.  System lightoff and temperature control tech-
niques also require further definition.
       The two stage combustor could also be applied to the gas turbine.
A catalytic fuel-rich first stage would be utilized for control of thermal
nitrogen fixation and fuel nitrogen conversion, followed by a second stage
which could be either catalytic or conventional combustion.  No interstage
heat removal would be required since the second stage would be operated
lean (to control turbine inlet temperature) by addition of large amounts of
excess air.  Current advanced stationary turbine technology involves the
investigation of two stage thermal combustion for NO  control.  A catalytic
                                                    A
combustor in one or both stages may be a natural advancement.

10.4   OTHER SYSTEMS
       A number of other combustion systems have been considered for catalytic
applications.  These include residential and industrial furnaces, rangetop
burners, mobile turbines, and utility boilers.  Although they are considered
to be less probable applications, they are briefly mentioned here.
       The residential furnace is the one system application that is cur-
rently marketed.  The Bratko furnace has been described in Section 2 and in
Reference 10-2.  Additional schemes for both retrofit and redesigned resi-
dential furnaces are presented in Reference 10-1.  Industrial combustion
furnaces pose additional application possibilities but have not been inves-
tigated by this study.
       Home rangetop burners pose applications problems that are similar to
residential furnaces.  Relatively small heat release rates and simple control
systems are required.  Only limited interest in the application currently
exists due to their relatively small impact on the NO  emissions inventory.
                                                     X
       Mobile gas turbines, either automotive or aircraft, may be a natural
extension of stationary turbine combustor development.  Catalyst size and
weight are not expected to be critical elements in the development of these
systems.  A high degree of system reliability is required, however, and poses
the most serious current development problem.
                                    10-13

-------
       Catalytic combustion application to utility boilers poses a large
development problem due to extremely large heat release rate requirements.
Demonstrated high combustion efficiency and simultaneous low emissions may
warrant projected development costs.  The use of lean catalytic combustors
in overfire air ports may be the first generation of utility boiler applica-
tions.

10.5   CONCLUSIONS
       Firetube boiler and stationary gas turbine applications appear promis-
ing for first generation catalytic combustor retrofit.  The concepts of
catalytic combustion have also been demonstrated for watertube boilers, home
heaters, and mobile turbines, but radical system redesigns will probably be
required for these applications.
       The success of the two stage combustor in this program in controlling
conversion of fuel-bound nitrogen to nitrogen oxides makes it appear promis-
ing in all applications.  System redesign would necessarily occur in all
two-stage applications since secondary air injection and/or interstage cool-
ing systems are required.  Additional work is also required to determine the
applicability of mixed catalytic and conventional burner systems in two-stage
combustors for the control of NO .
                                X
       Further, all catalytic applications require additional development of
fuel and air injection, premixing, and pre-vaporizing systems.  Combustor
control by bed temperature or stoichiometry and ignition systems also require
further consideration.
       The demonstration of catalytic combustor concepts in this program by
the radiative/watertube, two stage, and gas turbine systems has shown that
catalytic combustion is a viable technique and only awaits further develop-
ment to accomplish these promising applications.
                                   10-14

-------
                                 REFERENCES
10-1.   Kesselring,  J.  P.  et.  al.,  "Catalytic Oxidation of Fuels for NOX
       Control  from Area  Sources,"  EPA-600/2-76-037,  February 1976.

10-2.   Martin,  G.  B.,  "Evaluation  of a  Prototype Surface Combustion Furnace,"
       EPA-600-7-77-073C, July 1977.
                                     10-15

-------
                                SECTION 11
                      CONCLUSIONS AND RECOMMENDATIONS

       As a result of the extensive research and development program described
in this report, significant progress has been made toward the development of
a practical catalytic combustion system.  Before the step to demonstration
can be taken, however, additional work relating to the integration of the
catalytic combustor into the total combustion system must be performed.   This
section briefly presents the conclusions reached under this program and  makes
recommendations for further work.

11.1   CONCLUSIONS
       Based upon the analysis and test results of this program, the design,
fabrication, and operation of catalytic combustors with high volumetric  heat
release rates and low emissions have been shown.  Both precious metal
and oxide catalysts have been tested over a wide operating temperature range.
The precious metal catalysts should be limited to temperatures below 1589K
(2400°F) for catalyst life considerations while oxide catalysts can be oper-
ated for long periods at temperatures above 1644K (2500°F).  Catalyst per-
formance has been greatly enhanced through the use of graded cell monoliths
and higher catalyst loadings.
       Catalytic combustors have been shown to be effective in controlling
both thermal and fuel NOX emissions.  The thermal NOX control appears to
result from maximizing surface reactions in the combustor, while fuel NOX
can be minimized by operating at a rich fuel/air ratio which minimizes the
formation of NHa, HCN, and NO, with complete combustion of CO and HC at a
later time.  High pressure operation appears to give higher conversions  of
fuel nitrogen to NOX if space velocity, bed temperature, and nitrogen con-
centration in the fuel are held constant.  This implies that gas turbine
                                   11-1

-------
systems will have higher NOX emissions if only one stage lean combustors
are used with nitrogen containing fuels.
       The maximum throughput of a catalytic combustor is a linear function
of pressure and an exponential function of preheat.  Thus, for a given pre-
heat, the catalyst is face velocity limited in throughput ability.  Hyster-
esis is also exhibited by the combustor in terms of preheat required, with
less preheat required when the combustor has been operating than during the
early combustor startup period.
       Small scale catalytic combustion system configurations have been
tested and indicate the feasibility of direct radiative removal of bed
heat for temperature control, two stage catalytic combustion for temperature
and fuel NOX control, simulated exhaust gas recirculation through the use of
nitrogen diluent for temperature control, and high excess air operation.
The combustion system concepts that have been operated show that it is pos-
sible to operate stoichiometric conditions with less than 10 ppm NOX and CO
in a natural gas-fired catalytic combustor.  While this program has provided
much information on system applicability, further work with catalytic com-
bustors in actual systems is required.

11.2   RECOMMENDATIONS
       A number of areas in catalytic combustion need to be addressed to
capitalize on the progress to date.  Further oxide catalyst development work
is required to minimize catalyst/support interactions and subsequent loss
of thermostructural ability.  Additional  testing of simple and mixed oxide
catalysts for combustion and fuel nitrogen conversion abilities is needed,
along with life testing of selected catalysts to 1000 hours at various pres-
sures.
       Exploratory work with heavy fuel oils (#4, 5, 6) and pulverized coal
should be conducted to determine system feasibility and fuel preparation
problems.  The potential of catalytic combustion in controlling NOX emis-
sions from the combustion of these fuels  is great and needs early experi-
mental  verification.
       Development of auxiliary systems required to interface with the cat-
alytic  combustor is also needed.  This includes lightoff systems, temperature

                                   11-2

-------
control systems, and fuel  and air introduction systems.   In addition,  further
testing of the radiative catalyst/watertube, two stage combustor,  and  gas
turbine combustor systems  is needed to more thoroughly define operating ranges
with a variety of fuels.
       Finally, the design, fabrication, and operation of a demonstration
unit should be undertaken  when the above work is completed.  The demonstra-
tion unit would be operated as a laboratory device for several  months  prior
to the initiation of field demonstration tests.
                                   11-3

-------
             APPENDIX A

SECTION 7 DATA SUPPLEMENT -- CATALYST
           SCREENING TESTS
                 A-l

-------
TABLE A-l.   SCREENING TEST DATA SUMMARY
            a.  JPL-004X
Fuel
(Ib/hr)
.91



















^ t

.91
.91
.91
.91
.97
1.2
1.2
1.5
2.0
2.5
2.5
1.4
1.4
1.4
1.4
1.4
1.4

.91

















4


Air
(Ib/hr)
6.3
6.2
6.3
6.2
8.3
14.2
14.9
15.9
15.8
18.4
20.8
20.4
22.3
21.0
21.8
22.6
40.8
61.1
76.4
76.7
73.4

6.9
6.9
6.9
6.8
9.0
14.4
11.7
14.1
18.8
22.7
21.2
21.5
21.7
22.8
24.3
26.2
23.2

6.4
6.4
6.8
6.9
6.8
8.8
11.7
14.0
14.6
15.0
17.9
20.8

20.5

18.9
24.7
52.9
69.9
66.8

N2
(Ib/hr)

—
—
—
—
16.1
19.4
22.8
25.3
28.1
—
33.4
35.3
35.7
37.2
38.4
—
—
—
—
—

—
—
—
—
—
14.7
14.7
14.7
14.7
20.2
27.8
30.0
32.0
33.3
34.6
36.6
37.0

—
—
—
—
—
—
14.8
19.0
26.0
30.8
36.0
38.0

38.0

38.0
38.0
18.0
—
-

Preheat
Temp.
(°F)
670
581
504
400
373
731
781
826
865
910
939
990
1028
1047
1059
1065
1040
1042
1041
1046
1039

844
645
592
476
377
617
527
482
498
777
893
948
983
1005
1018
1034
1041

914
775
694
650
534
665
837
772
890
1045
1075
1128

1133

1185
1204
1192
1198
1073

% TA
41.



54.
93.
98.
104.
104.
121.
136.
134.
146.
138.
143.
148.
267.
400.
500.
500.
481.

45.
45.
45.
45.
55.
72.
58.
56.
56.
54.
51.
92.
92.
104.
104.
112.
99.

42.
42.
45.
45.
45.
58.
77.
92.
96.
98.
117.
136.

134.

124.
162.
347.
458.
438.

Space
Velocity
(1/hr)
5655.
5655.
5655.
5655.
7084.
23495.
26495.
29484.
31606.
35585.
—
41029.
43826.
43200.
49908.
46388.
30314.
44823.
55759.
55973.
53615.

6083.
6083.
6083.
6012.
7660.
22945.
21015.
23110.
27102.
34688.
39372.
39860.
41518.
43289.
45346.
48218.
46449.

5726.
5726.
6012.
6083.
6012.
7442.
20723.
25548.
31279.
35200.
41211.
44799.

44584.

43441 .
47586.
52595.
51113.
48897.

Approximate Bed
Tfront ave

710.
1110.
1180.
1500.












2040.
2040.
1980.
1050.


930.
1450.
1720.












1300




720.
1760.






1150.

1170.




1700.
1300.

730.




1990.
2000.
2000.
2000.
2010.
2000.
2020.
2020.
2020.
2010.
2030.
2060.





830.



1870.
1850.
1970.
1900.
1990.
1950.
1920.
1920.
1980.
1990.
1980.
1990.


880.
860.
850.


1980.
1990.
1970.
1970.
1590.
2020.




1960.
2110.
2080.



Temperature (°F)
Tback

790.
900.
1150.
1470.












2080.
2100.
2070.
2070.


1000.
1100.
1550.












1900:




1300.
1650.






1780.

2040.




2180.
2330.

Comments
Start
Transient
ii

















Break-
through
Restart
Transient
"







i





Break-
through
Restart
Transient
n
H







Break-
through
Break-
through




Break-
through
                    A-2

-------
TABLE A-l
Continued
b.  JPL-005X
Fuel Air
(Ib/hr) (Ib/hr)
.91 6.









1
6.
6.
8.
10.
18.
19.
24.
33.
40.
r
.91 8.
1.5 13.
2.0 18.
2.5 23.
2.75 24.
2.75 24.
3.01 26.
.91 6.
.91 6.
.91 6.
.91 8.
2.0 18.
2.8 25.
2.8 24.
4
4
4
5
7
5
1
9
9
7

6
1
0
0
0
8
8
1
2
9
9
6
0
7
3.0 27.
3.5 31.
3.
3 33.
3.8 33.
1.5 22.
1.5 25.
1.

5 25.

7
1
1
5
8
5

(Ib/hr)

—
—
—
16.
40.
40.
38.
34.
28.

—
—
—
6.7
12.
19.
21.
—
—
—
—
13.5
20.
20.
26.
34.
35.
36.5
40.
45.3
46.5

Preheat
Temp.
793
648
504
590
1023
1043
1018
1055
1083
1082

443
551
320
522
741
778
754
837
748
507
460
801
845
790
847
925
967
973
1087
1092
1067

% TA
42.
42.
42.
56.
70.
121.
125.
163.
222.
267.

56.
52.
53.
54.
52.
53.
53.
40.
41.
45.
58.
55.
53.
52.
53.
54.
51.
51.
89.
102.
101.














7
9
0
8
1




5
2
6
7
0
9
9
4
6
4

Space
Velocity
(1/hr)
5726.
5726.
5726.
7227.
20918.
44670.
45099.
47729.
51133.
51449.

6870.
11262.
15397.
24678.
29723.
35596.
38870.
5512.
5583.
6083.
7513.
26050.
36560.
36346.
42787.
52838.
54976.
56112.
48275.
54648.
55343.

Approximate Bed Temperature (°F)
Tfront Tave. Tback Comments
720. Start
860. Transient
1250.
1930.
1930.
1960.
1960.
2020.
2010.
1200. 2000. Break-
through
1840.
1870.
1930.
1910.
1910.
1910.
1890. End
810. Restart
1150. 840. Transient
1720. 1500.
1840.
1790. I
1890.
1790.
1880.
1900.
1820.
1920.
1960.
2000.
1150. 1900. Break-
through
          A-3

-------
TABLE A-l .


Fuel Air
(Ib/hr) (Ib/hr)
.91 6.2
6.4




i













i
|

I
i
1
i

i

!
1






1
6.3
8.4
9.7
10.6
12.1
12.6
13.0
13.7
14.1
14.9
14.7
15.5
15.7
16.5
18.6
15.8
6.1
6.8
8.4
10.5
12.4
12.5
13.9
14.0
15.9
14.3
15.9
16.8
18.0
16.3
18.5
19.4
24.5
30.1
15.8
' 17.8

M
N2
(Ib/hr)

—
—
—
4.5
8.5
12.8
17.0
19.5
22.5
23.5
25.0
26.5
27.5
27.5
28.0
28.0
28.0
_
—
—
—
—
22.4
24.
25.
26.7
27.5
29.2
29.2
30.3
31.8
33.0
34.4
30.1
26.1
25.6
29.1

Preheat
Temp.
(°F)
707
597
381
357
506
588
672
742
784
814
836
871
885
888
902
903
715
899
772
477
447
644
747
784
819
853
880
893
920
925
939
956
969
979
981
981
886
895


% TA
41.
42.
41.
55.
64.
69.
79
83.
85.
90.
92.
98.
96.
102.
103.
108.
122.
104.
40.
45.
55.
69.
81.
82.
91.
92.
104.
94.
104.
110.
118.
107.
121.
127.
161.
197.
104.
117.
Continued
c. JPL-006
Space
Velocity
(1/hr)
5583.
5726.
5655.
7156.
11493.
15166.
19495.
23033.
25212.
27985.
29028.
30736.
31729.
33058.
33201.
34152.
35653
33651.
5512.
6012.
7156.
8657.
10015.
27051.
29264.
30093.
32738.
32201.
34632.
35275.
36966.
36887.
39368.
41072.
41460.
42433.
31834.
35914.






Approximate Bed Temperature (°F)
Tfront
Tave. Tback Comments
710. Start
770. Transient
















1210.
1890.
1920.
1930.
1960.
1960.
1960.
1970.
1970.
1970.
1970.
1970.
1970.
1980.
1970.
1980 End
















800. Restart
1630.








1680.
1690.
1210.
1200.
1180.
1100.
1050.
1000.

1530.
1370.
1870.
1860.
1960.
1950.
1980.
1990.
2000.
1990.









2010. Incipient
2030. break-
2030. through
1990.
2000.
2010.
2060.
2000. i
1970.




r

1940. Incipient
break-
through
A-4

-------
                                                     TABLE A-l.
Concluded

d.  JPL-006X
Fuel
(Ib/hr)
.91










1











.50








.91
1
2.0
2.0
2.5
3.0
3.0
3.8
Air
(Ib/hr)
4



1
7
15
22
24
14
14
23
7
12
15
31
31
3
4
4
15
16
21
26
25
33
4




5
4





5




9
0
5
6
6
1
3
8
8
N2
(Ib/hr)
—
-
-
-
-
-
35.
40.
45.
37.
48.
54.
14.
22.
12.
-
-
-
9.
11.
15.
19.
19.
30.
Preheat
Temp
783.
633.
571.
511.
477.
95.
653.
607.
800.
852.
901.
922.
876.
901.
869.
847.
841.
879.
764.
370.
637.
684.
690.
690.
698.
698.
% TA
28.8



1 '
49.1
101.
144.
157.
92.
92.
151.
89.4
143.
179.
370.
370.
25.6
26.2
29.5
46.5
49.5
50.3
52.3
51.3
53.0
Space
Velocity
(1/hr)
4,295
4,295
4,295
4,295
4,295
6,509
38,642
47,141
52,353
39,156
47,481
58,451
16,586
25,855
20,430
22,776
22,776
3,937
4,009
4,366
20,486
22,714
29,589
36,963
36,606
51,658
Approximate Bed
Temperature (°F)
Tfront
683
661
850
1243
1264
1959
2003
1264
-
1934
2007
1974
2000
1966
1766
1522
1475
791
813
1717
1955
1988
2004
1976
2005
2002
^ave
—
-
-
-
-
-
-
-
2008
-
-
-
-
-
-
-
-
-
-
-
-
-
-
—
Tback
662
689
935
1210
1224
1792
1955
1980
-
1891
1960
1980
1911
1910
1930
1975
1970
748
763
1492
1848
1887
1905
1836
1875
1861
Emissions (ppm)
CO












11,500.
<1
<1


13,800.
15,200.
10.
9,500.
<1


i







D.










UHC





20.
<5.









i











28.
15.
<5.


1



NO





<1.





















r

<1.




^





Comments
Start
Transient





Unstable Point

Not Minimum
Not Minimum *
Not Minimum



Incipient Breakthrough
Incipient Breakthrough
Restart
Transient





End (Breakthrough Not
Reached)
i
en.
         *Trying to get to 200 percent TA.

-------
                                                 TABLE A-2.  TEST DATA - JPL-007
Fuel
(Ib/hr)
.91










1











3.5
4.0
Air
(Ib/hr)
4.9
5.0
5.3
5.1
7.9
15.1
23.6
23.5
32.2
5.1
4.8
7.9
15.5
8.0
30.0
33.9
N2
(Ib/hr)
—
-
-
-
-
20.
26.
27.
22.
-
-
-
26.
-
10.
11.
Preheat
Temp
(°F)
836.
733.
521.
512.
110.
262.
307.
302.
435.
871.
708.
117.
553.
180.
282.
251.
% TA
32.1
32.8
34.7
33.4
51.8
99.0
155.
154.
211.
33.4
31.5
51.8
102.
52.4
51.1
50.5
Space
Velocity
(1/hr)
4,652
4,723
4,937
4,795
6,795
27,075
37,688
38,373
40,803
4,795
4,580
6,795
31 ,902
6,866
33,427
37,603
Approximate Bed
Temperature (°F)
Tfront
697
746
1346
1291
1949
1979
1975
1967
843
1119
1940
1907
1802
1562
1565
ave
—
-
-
-
-
-
1984
-
-
-
-
-
-
-
Tback
670
683
1304
1254
1785
1909
1929
1927
830
861
1765
1897
1667
1788
1774
Emissions (ppm)
CO





16,300.
<10.
1



19,400.



UHC




17.
2.
2.
2.
5.


3.
44.
6.
582.
383.
NO




<1.






<1.



i




Comments
Start
Transient
Transient



End (No Breakthrough)
Restart
Transient



Incipient Breakthrough
Incipient Breakthrough
I
01

-------
                                              TABLE A-3.  TEST DATA - JPL-008
Fuel
(Ib/hr)
.91










1











.91
3.5
.91
Air
(Ib/hr)
4.5
4.4
4.7
8.3
14.5
61.2
15.6
15.8
22.5
23.7
34.8
42.5
4.2
4.2
7.4
31.8
3.8
N2
(Ib/hr)
—
-
-
-
16.5
16.
23.
40.
43.
31.
14.
-
-
-
16.
-
Preheat
Temp
(°F)
707.
572.
243.
112.
117.
765.
107.
460.
880.
981.
941.
883.
804.
669.
277.
232.
828.
% TA
29.5
28.8
30.8
54.4
95.0
401.
102.
104.
147.
155.
228.
279.
27.5
27.5
48.5
54.2
24.9
Space
Velocity
0/hr)
4,366
4,295
4,509
7,080
23,997
44,867
24,405
29,845
47,498
50,625
49,472
42,106
4,152
4,152
6,438
39,254
3,866
Approximate Bed
Temperature (°F)
Tfront
_
846
-
1882
1980
1432
1965
1959
1717
1953
1856
1859
-
1170
1835
1983
-
ave
692
-
1201
-
-
-
-
-
-
-
785
-
-
-
811
Tback
—
750
-
1740
1928
1891
1898
1913
2072
2035
2081
2075
-
865
1639
1818
-
Emissions (ppm)
CO




19,400.
18,200.
13,900.
<10.
1
1
t





UHC






37.
<
i
8.
23.
<


1








1600.
2800.


NO



<1.





i








<1.
<1.

Comments
Start
Transient


Incipient Breakthrough
Not Minimum



Incipient Breakthrough
Restart
Transient

Incipient Breakthrough
Restart
-•J

-------
                                                 TABLE A-4.  TEST DATA - JPL-009
Fuel
(LB/HR)
.91



^




2.0
3.0
.91
Air
(LB/HR)
5.5
5.9
7.8
16.1
8.0
17.9
26.7
15.1
N2
(LB/HR)



16.



15.
Preheat
Temp
(°F)
689.
197.
164.
161.
177.
168.
159.
161.
% TA
36.0
38.7
51.1
105.5
52.4
53.4
53.1
99.0
Space
Velocity
(1/HR)
5,080
5,366
6,723
24,762
6,866
15,317
22,869
26,670
Approximate
Bed Temp (°F)
T Front








ave
716
1,366
1,765
1,849
1,774
1,916
1,890
1,867
TBack








Emissions
PPM
CO







9630
UHC



1200.
84.4
1350.
5360.
1170.
NO


< 1.










Comment
Start

High
High


High
End
I
oo

-------
TABLE A-5.  TEST DATA - JPL-010
Fuel
(LB/HR)
.91


















Air
(LB/HR)
5.0
5.0
5.0
5.0
7.5
13.5
17.2
22.8
33.8


N2
(LB/HR)
—
—
—
—
—
36.0
40.0
40.0
31.0


Preheat
Temp.
(°F)
684.
574.
518.
405.
638.
795.
735.
822.
836.


% TA
32.8
32.8
32.8
32.8
49.1
88.5
112.7
149.4
221.5


Space
Velocity
(1/HR)
4,723
4,723
4,723
4,723
6,509
38,042
43,712
47,712
48,758


Approximate
Bed Temp (°F)
TFront


1234








ave
635
670

1198

BAD
DATA




TBack


704








Emissions
PPM
CO





14,000.
3,440.
.2
644.


UHC




47.4
0.3
0.3
2.8
2010.


NO




0.










Comments
Start

Transient





End





Catalyst
Damaged
at Front
Face

-------
                                   TABLE A-6.  TEST DATA - JPL-010X
Fuel
(LB/HR)
0.91









1.









5
1.5
0.91
!
Air
(LB/HR
5.25
5.1
4.2
7.9
14.2
23.0
33.0
42.7
50.0
33.3
56.1
55.2
4.5
8.2
15.9
N2
(LB/HR)
—
—
—
—
16.5
21.0
13.0
—
—
20.0
38.0
39.0
—
13.0
	
Preheat
Temp.
(°F)
643.
462.
165.
117.
96.
157.
118.
107.
407.
516.
698.
688.
738.
192.
154.
% TA
34.4
33.4
27.5
51.8
93.1
151.
216.
280.
328.
218.
223.
219.
29.5
53.7
104.2
Space
Velocity
(1/HR)
3,865
4,795
4,152
6,795
23,783
33,475
34,563
31,653
36,867
40,075
70,732
70,846
4,366
7,009
22,348
Approximate
Bed Temp (°F)
TFront

494.









781.


ave*
619.

1197.
1727.
1930.
1957.
1946.
2017.
1993.
1954.
2105.

726.
1792.
2060.
TBack

840.









1811.


Emissions
PPM
CO




3370.
< 10.





>







59,300
16,700
UMC



40.
< 5.






>








44.0
17.6
NO



< 1.

















< 1.
1
Comment
Start
Transient







Not Minimum

Breakthrough
Restart
End
^Thermocouples 0.1" back were not included  in  average.

-------
                                           TABLE A-6.  Concluded

                                                  JPL-010X
Fuel
(LB/HR)
0.91














4.0
6.0
Air
(LB/HR)
4.5
4.5
7.6
15.7
15.7
23.3
5.9
7.8
35.
52.
N2
(LB/HR)
—
—
—
18
27
39
—
—
—
12
Preheat
Temp.
(°F)
786.
708.
195.
214.
644.
773.
784.
125.
62.
243.
% TA
29.5
29.5
49.8
103
103
153
38.7
51.1
52.2
51.7
Space
Velocity
(1/HR)
4,366
4,366
6,580
25,990
32,801
47,312
5,366
6,723
30,063
53,819
Approximate Emissions
Bed Temp (°F) (PPM)
TFront

1632.








ave*
839.

1786.
1944.
1978.
1970.
858.
1785.
1889.
1931.
Tback

847.








CO



22,200
2,600
< 10.


59,700
53,600
UMC


30
< 5

4-

52.2
4.8
227.
NO


< 1.


-




< 1.




Comments
Restart
Transient


Not Minimum
End**
Restart


End
 thermocouples 0.1" back were not included in average.
**Approximately 8 hours of testing has occurred by this point.

-------
TABLE A-7.  DATA SUMMARY - JPL-010P
Fuel
(Ib/hr)
0.91



<
0.5
0.5
0.5
0.7
0.91
0.5
0.7
0.91
0.5
0.7
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.91


3.0
3.8
4.5
4.5
0.7*
0.6*
0.5*
1.0*
1.0*
1.0*
Air
(Ib/hr)
54.9
55.5
54.1
54.8
58.6
56.8
22.7
28.8
35.3
49.3
65.0
35.8
49.2
67.1
36.1
50.4
36.1
37.2
37.2
36.8
36.6
36.3
35.7
36.3
36.2
5.4
5.7
8.2
25.7
31.9
40.5
40.6
48.0
39.0
32.4
4.55
4.5
4.7
N2
(Ib/hr)










































































Preheat
Temperature
(°F)
908
880
880
882
894
829
971
902
1059
1067
1050
1090
1105
1091
1121
1112
1141
1220
1120
1119
1097
1126
1121
1120
1118
825
215
107
82
78
124
82
1119
1126
1121
827
521
339
Percent
TA
359.7
363.7
354.5
359.1
384.0
372.2
270.7
343.5
421.0
420.0
426.0
427.0
419.1
439.7
430.6
429.4
430.6
443.7
443.7
438.9
436.5
432.9
425.8
432.9
431.7
35.9
37.4
53.7
51.1
50.1
53.7
53.8
408.9
387.6
386.4
27.1
26.8
28.0
Space
Velocity
(1/hr)
40367
40796
39796
40296
43010
41724
16848
21205
25848
36101
47582
26205
36030
49082
26419
36887
26419
27205
27205
26919
26777
26705
26134
26705
26491
5009
5223
7009
22154
27595
34624
34696
35173
28617
23776
4516
4480
4623
Comments
Methane start
Transient

1
Stable
Restart








End
Restart







End
Restart





End
Propane restart

End
Restart
Restart
End
Time
(min)
0



20
39
39
39
95
100
105
185
198
208
283
293
295
308
362
428
444
512
639
741
818
830
853
873
890
899
926
1003
1063
1113
1128
1138
1138
1141
*
Propane
                 A-12

-------
                               TABLE A-8.  TEST DATA - JPL-011
Fuel
(LB/HR)
.91

<


2.0
3.0
4.0
5.0
6.0
• .


11


0.5
1
Air
(LB/HR)
5.7
5.9
7.7
17.4
26.4
35.3
42.9
51.0
15.9
22.4
5.7
6.1
9.2
17.7
N2
(LB/HR)



5.5
4.5
4.5
5.5
9.0
17.
36-


12
15
Preheat
Temp.
(°F)
642.
183-
158.
143.
149.
151.
152.
146.
141.
655.
804.
299.
312.
947.
% TA
37.4
38.7
50.5
51.9
52.5
52.6
51.2
50.7
104.2
146.8
37.4
40.0
109.7
211.1
Space
Velocity
(1/HR)
5,223
5,366
6,652
19,123
26,060
33,683
41,134
50,835
25,376
44,399
5,223
5,509
16,287
24,629
Approximate
Bed Temp (°F)
T Front













Tave*
686
1,358
1,792
1,662
1,766
1,927
1,934
1,951
1,896
2,005
756
1,515
1,916
1,952
TBack













Emissions
PPM
CO








14,800.
< 10.


< 10.
I
UHC


26.2
6.9
< 5.



-






< 5.
1
NO


< 1.














< 1.
1
Comments
Start







End - (Could
not get to
200% TA)
Restart



Thermocouples 0.1" back were not included in average.

-------
                                               TABLE A-8.   Concluded
                                                           JPL-011
Fuel
(LB/HR)
2.5

Air
(LB/HR)
21.2

N2
(LB/HR)
6

Preheat
Temp.
(°F)
118.

% TA
50.6

Space
Velocity
(1/HR)
22,848

Approximate
Bed Temp (°F)
TFront


TAve*
1,894

TBack


Emissions
PPM
CO


UHC
1,000.

(PROPANE)
1.0



5.9
5.1
7.7
14.8
""" " '
—
—
21
400.
115.
95.
244.
37.7
32.6
49.2
94.6
4,694
4,122
5,980
26,945




384
1,725
1,769
1,911







5120.



< 5.
NO
< 1.

Comments
Break-
through
(High
UHC)




< 1.
Restart


End
I
-p>
       Thermocouples 0.1" back were not  included  in  average.

-------
                                          TABLE  A-9.  TEST DATA - JPL-012
Fuel
(LB/HR)
0.91






,







3.5
Air
(LB/HR)
6.2
6.6
8.0
15.5
17.0
6.6
6.2
7.7
29.1
N2
(LB/HR)



19.
38.



27.
Preheat
Temp.
(°F)
567.
188.
186.
290.
751.
707.
127.
138.
808.
% TA
40.6
43.3
52.4
102.
111.
43.3
40.6
50.4
49.6
Space
Velocity
(1/HR)
5,580
5,866
6,866
26,604
42,055
5,866
5,580
6,652
42,373
Approximate
Bed Temp (°F)
TFront




967




ave*
645
1,523
1,832
1,904

754
1,518
1,757
1,979
TBack




1,974




Emissions
PPM
CO



10,100.
6,750.




UHC

< 5.
35.5
< 5.
153.


40.6
417.
NO

<


«


<
\

1.





1.
i
Comments
Start



END**
Restart


END***
en
      **
     ***
 Thermocouples 0.1"  back not included  in ave.
 Could not go to leaner conditions  (150% TA)
'Maximum throughput  at  50%  TA

-------
                                         TABLE A-10.  TEST DATA - JPL-013

Fuel
(LB/HR)
.91











Air
(LB/HR)
5.8
5.9
7.9
7.8




N2
(LB/HR)



6.




Preheat
Temp.
(°F)
757.
666.
141.
201.




% TA
38.0
38.7
51.8
51.1




Space
Velocity
(1/HR)
5,295
5,366
6,795
11,264



Approximate
Bed Temp (°F)
1 Front

1,824
2,079




ave*
861 *


1,718 *



TBack

892
1,803




Emissions
PPM
CO







UHC


600.
6,300.



NO


< 1.
1




Comments
Start
Transient

Break-
through
(High
UHC)
I
__l

CTl
         thermocouples 0.1"  back were not included  in  average.

-------
TABLE A-ll.  TEST DATA - JPL-022
Fuel
(LB/HR)
.91

, ,
2.5
3.'5
4.5
5.5
6.0

.91


• -

Air
(LB/HR)
6.6
5.8
8.1
21.0
30.8
38.8
46.2
50.6

6.5
7.8
15.1
15.8

N2
(LB/HR)




6.0
6.0
8.0
11.0



30.
31.

Preheat
Temp.
(°F)
594.
114.
109.
86.
87.
97.
93.
90.

542.
148.
705.
679.

% TA
43.2
38.0
53.1
50.1
52.5
51.4
50.1
50.3

42.6
51.1
99.0
103.5

Space
Velocity
(1/HR)
5,866
5,295
6,938
15,317
Approximate
Bed Temp (°F)
! Front




30,971
37,951
46,016
52,063
i
5,795
6,723
34,643
35,900


1,285

ave*
606
1,320
1,782
1,821
1,864
1,844
1,824
1,845

597
1,860
1,959


TBack












1,977

Emissions
PPM
CO











12,700.
10,800.

UHC




23.7
5.7
<


•


5.


-


15.4
<
,

5.
\

NO


< 1.












< 1.

•



Comments
Start






END

Restart


Incipient
Break-
through

-------
                                           TABLE A-12.  TEST DATA - JPL-016
3=




CO
Fuel
(Ib/hr)
0.91










0.5
3.5
4.0
0.91
0.91
Air
(Ib/hr)
5.7


8.1
16.3
17.3
16.6
17.1
6.1
7.9
16.0
9.2
31.6
36.0
5.6
13.8
(Ib/hr)
0
0
0
0
18.4
20.5
35.9
28.8
0
0
17.9
22.7
7.5
5.0
0
31.0
Preheat
Temp
650
498
123
93
85
182
798
589
625
66
68
915
112
91
665
914
% TA
37


53
106
113
108
112
40
51
104
109
53
53
36
90
.3


.1
.8
.4
.8
.1
.0
.8
.8
.7
.8
.7
.7
.4
Space
Velocity
(1/hr)
5223


6938
26720
29025
40180
35164
5509
6795
26128
24385
32678
34562
5152
34472
Approximate
Bed Temp (°F)
ave
687
895
1403
1859
1931
1971
1950
1958
752
1859
1931
1935
1820
1934
1139
1954
Comments
Start
Transient




Not minimum
Not minimum
Restart





Restart


-------
                   TABLE  A-13.   TEST  DATA - JPL-021
Fuel
(Ib/hr)
1.5
1.5
1.7
1.6
1.5








1.7








1.9
1.9
1.5












2.0
2.0
3.0
4.0
Air
(Ib/hr)
105.6
109.7
118.2
115.4
109.8
112.4
114.8
118.1
107.7
111.4
107.5
110.3
58.4
108.8
109.0
111.2
106.2
113.1
108.1
109.7
108.0
108.0
107.5
16.8
17.1
26.9
34.4
N2
(Ib/hr)
0








•»











































Preheat
Temp
(°F)

1029
1060
1059
1054
917
1114
1069
1053
1046
1046
1051
1072
1052
1047
1043
1059
1029
1044
1055
1053
1050
1047
967
117
101
98
% TA
419.8
436.1
414.6
430.1
436.5
446.9
456.4
469.5
428.2
390.8
377.1
386.9
204.9
381.6
342.1
349.0
422.2
449.6
429.8
436.1
429.4
429.4
427.4
50.1
51.0
53.5
51.3
Space
Velocity
(1/hr)
231,990
240.774
259,749
253,368
240,990
246,561
251 ,703
258,774
236,490
245,178
236,820
242,820
131,601
239,604
240,792
245,508
77,758
82,687
79,115
80,258
79,044
79,044
78,687
14,531
14,746
23,012
29,634
Comments
Start





Restart





Restart



Restart*






Restart*



Time
(min)
0
23
65
122
183
190
190
192
223
239
434
599
606
617
908
1187
1187
1187
1192
1201
1253
1298
1370
1370
1446
1487
1503
*
 Two 1-inch segments of JPL-015 included downstream.

-------
                                  TABLE A-14.   MONOLITH 019 TEST DATA - JPL TESTS
1 atm Pressure
Fuel
Methane
Methane
Propane
Indolene
Methanol
Methanol
Methane
Methane
Propane
Propane
TA, %
320
320
350
290
326
241
211
200
248
261
SV, hr-1
87,800
88,200
78,800
80,800
64,400
72,500
66,300
74,500
72,700
61 ,000
ifif uel ^ 1 bm/hr
2.2
2.2
1.9
2.5
4.0
6.0
2.5
3.0
2.5
2.0
ma-jrj 1 bm/hr
118.1
118.7
106.1
108.2
84.2
93.4
88.4
99.1
97.1
• 81.6
T °F
'PH, F
852
957
864
855
689
293
854
843
862
752
TBED, °F
1980
2040
1980
1950
2000
1930
2400
2550
2400
2340
Run-Scan No.
A41A-6
A41A-17
A41C-5
A41D-5
A41E-7
A41E-15
A41 F-4
A41 F-8
A41F-15
A41F-19
I
ro
o

-------
ro
                                   TABLE A-15.   MONOLITH 019 TEST DATA - ACUREX TESTS



                                                     1  ATM PRESSURE
FUEL
Natural Gas
Natural Gas
Natural Gas
Natural Gas
Natural Gas
Propane
Propane
Natural Gas
Natural Gas
Natural Gas
Natural Gas
Natural Gas
TA, %
28
28
42
42
42
38
37
40
227
217
205
196
SV, hr'1
21,100
21 ,400
29,300
29,200
28,400
40,000
39,100
37,800
105,000
97,600
92,700
92,500
mfuel, Ibm/hr
4.4
4.4
4.4
4.4
4.4
5.3
5.0
5.8
3.5
3.4
3.4
3.6
mair, Ibm/hr
20.9
21.5
32.1
32.1
31.1
50,6
49.5
39.8
135.9
127.4
120.3
120.0
V °F
660
460
520
524
527
417
421
634
721
738
740
730
Tbed, °F
1,950
1,940
2,240
2,250
2,230
2,010
1,970
2,400
2,120
2,350
2,340
2,360
RUN #
1122-2
1122-3
1122-4
1122-5
1122-7
1123-1
1123-2
1124-2
1201-1
1201-2
1201-3
1201-4

-------
TABLE A-16.   MONOLITH 019 TEST DATA - ACUREX TESTS, EMISSIONS DATA FOR SIMULATED FUEL NITROGEN TESTS



3=
I
l\o
PO

FUEL
Natural Gas
Natural Gas
Natural Gas
Natural Gas
Natural Gas
Propane
Propane
Natural Gas
Natural Gas
Natural Gas
Natural Gas
Natural Gas
TA, %
27
28
42
42
42
38
37
40
227
217
205
196
NH, fuel ppm NH., comb ppm Nf) NH
'bed, F J cone, J gases, CO, ppm UHC, ppm NUx ppm HCN, ppm NH3' ppm
i OCA __, , -•- ? 0
1,940 1,000 270
2,240 	 	
2,250 1,000 200
2,230 6,000 1,212
2m n
1 ,970 1 ,000 62
^ flnn n

, ItU 	
, JOU
2,340 6,000 292
00 > 3C


15 N(
17 1
13
Wo
4 2,3 173
•3 	 	
4 35 223
25 NG 964
T 1
NG NG 77
<1 NG 0
7
m?
2 78 5 3.4
2,360 1,000 51 16 1 10 6 0

-------
                         TABLE A-17.   MONOLITH  019 TEST DATA - ACUREX TESTS, HIGH PRESSURE OPERATION
i
ro
cx>
FUEL
Methane
t




P, Atm
1
1
2
3
3
6
TA,%
33
212
197
187
210
232
SV, hr"1
21,700
98,200
88,000
84,000
89,300
123,300
"fuel' 1bm/hr
4.0
3.6
3.4
3.3
3.3
4.1
mair,lbm/hr
22.8
130.5
114.9
106.6
120.3
161.5
T -°F
V F
520
688
697
702
710
762
T °F
'bed' h
1980
2430
2350
2400
2400
2170
RUN #
1206-2
1206-3
1206-4
1206-5
1206-6
1206-7
                                                   Total bed operating time:  >74 hours

-------
                           TABLE  A-18.   MONOLITH 019 TEST DATA - ACUREX TESTS, EMISSIONS DATA
                                        FOR FUEL NITROGEN SIMULATION TESTS AT PRESSURE
3s

-Pa
FUEL
Methane
Methane
Methane
Methane
Methane
Methane

TA, %
32
212
197
187
210
232

P,
atm
1
1
2
3
3
6

Tbed, *
1930
2430
2350
2400
2400
2170

NH, ppm
Jfuel ,
_._
-..
---

6,000
—

NH3 ppm
comb,
gas
___
_..
___
—
286
.__
	 . 	 ___
CO, ppm
>2000
30
23
24
22
--

UHC.ppm
>300
2
4
2
1
--

NO , ppm
A
1
2
6
5
195
_-.

HCN.ppm
---
...
__.
—
0
—

NH3,ppm
.__
.-_
— -
.__
0
._.


-------
                  TABLE  A-19.   MONOLITH 019 TEST DATA - NATURAL GAS, HIGH TEMPERATURE OPERATION
TEST PT.
1228-1
1228-3
1228-4
1228-6
1228-7
1228-8

1228-11
1228-12
TA,
(%)
42
238
250
225
211
195

48
236
SV,a
(hr-1)
30,300
141,200
142,200
126,400
120,200
111,900

32,000
136,800
•
m/- T
fuel,
(Ibm/hr)
4,6
4.5
4.3
4.3
4.3
4.3
— Shutdown and
4.4
4.4
""air,
(Ibm/hr)
34.0
184.1
185.9
164.4
155.9
144.5
Relight —
35.6
178.4
Tph,
(°F)
532
678
625
594
581
582

642
720
Tbed,
(°F)
1820
2210
2070
2382
2500
2610

2110
2360
T °F
max,
bed
2300
2485
2530
2615
' 2660
2710

2600
2580
r
ro
01
           *Space  velocity based on standard conditions

-------
               SECTION B

SECTION 8 DATA SUPPLEMENT — GRADED CELL
             CATALYST TESTS
                   B-l

-------
     SECTION  B-l

W. R. GRACE CATALYST
  TEST MODEL A-025
         B-2

-------
                                      TABLE B-l.   TEST DATA SUMMARY - CATALYST A-025
DO
I
co
Run #
0425-01
0426-10
0427-03
0427-05
0427-07
0427-08
0428-04
0428-07
0428-10
0428-12
0428-14
0428-18
0428-20
0429-01
0429-02
0429-05
0429-07
0429-09
0429-11
TA%
269
233
250
264
253
272
306
308
302
321
291
297
272
251
232
234
223
252
287
SV, hr"1
146,700
135,700
143,000
150,100
146,600
157,500
186,500
248,700
327,400
193,300
175,900
176,900
163,500
156,400
143,100
204,900
273,800
383,300
182,300
""fuel' lbm/hr
4.2
4.4
4.4
4.4
4.4
4.4
4.7
6.2
8.3
4.6
4.6
4.6
4.6
4.8
4.7
6.7
9.3
11.6
4.9
m. , lbm/hr
air
193.3
177.8
187.8
197.6
192.7
207.6
246.8
329.2
433.2
256.4
232.5
234.0
215.6
205.6
187.5
268.5
358.0
503.8
240.8
Tph,°F
761
676
697
602
500
755
758
778
797
785
778
600
501
759
752
769
785
800
788
T °F
'BED' r
2027
2251
2263
2092
1967
2265
2129
1780
1806
2144
2455
2365
2406
2024
2437
2203
2096
2145
2092
T °F
BEDmax,
2231
2431
2432
2461
2414
2500
2551
2466
2511
2523
2711
2721
2709
2542
2717
2705
2699
2735
2511
Total test time = 20 hrs
at Q = 100,000 Btu/hr nominal heat release rate

-------
    TABLE  B-2.   EMISSIONS DATA - CATALYST A-025
Run
No
11W •
0425-06
0426-10
0427-03
0427-08
0428-04
0428-07
0428-10
0428-12
0429-01
0429-11
0429-02
0429-07
0429-09
Test Time,
hrs
2.0
10.0
11.7
12.5
15.0
15.5
16.0
16.5
17.7
20.0
18.2
18.7
19.6
TA
%
269
233
250
272
306
308
302
321
251
287
232
223
252
mf uel ,
Ibm/hr
4.2
4.4
4.4
4.4
4.7a
6.2a
8.3a
4.6
4.8
4.9
4.7
9.3
11.6
CO,
ppm
0
0
0
0
0
0
48b
437b
0
0
0
0
0
NO,
A
ppm
0
3
3
3
2
2
2
2
2
2
4
3
4
                                                          Emissions time history
                                                          and throughput effects
                                                          at the nominal condition
                                                               « 2400°F
                                                          Tph  «  750°F
                                                          Emissions vs throughput

                                                               Tbed  t
 Increased mass throughput to demonstrate the effects
 on emissions.

 May not be steady operating values.  Wide variations
 (0 to 1200 ppm) in CO emissions were noted as test
 conditions were changed.

cNo emissions changes with reduced preheat were noted.

 All NOV measured was present as NO.
       X

-------
                                TABLE  B-3.   LIGHTOFF  TEMPERATURE  HISTORY - CATALYST A-025
CO
I
en
Cumulative
Test Time
(hrs)
0
2.5
10.0
13.0
17.5
Lightoff
Stoichiometry
Fuel lean
Fuel rich
Fuel rich
Fuel rich
Fuel rich
Lightoff
Temperature
(°F)
910
900
830
830
930
Comments
Uneven bed temperature distribution
immediately apparent
Unsuccessful lean lightoff at 990°F.
May not have been minimum rich
lightoff temperature. Very sooty
combustion.
Soot not apparent
Combustion on one side of bed only


-------
CD
I
                                                            Flow
                                                                         1
                                                                         0
                                                                         2
                                                                         O
      3
     0
                                                                                          8
                              BEGINNING  OF AGING

                                   t  = 1  hr

                                  Run 0425-01
10.
1
2
3
4
7
8
9
IE
2032
1947
2205
1457
2222
2093
2231
END OF AGING

 t = 10 hrs

 Run 0426-10
TC
1
2
3
4
7
8
9
IE
2306
2314
2375
1657
2431
2358
2314
                      Figure B-l.  Catalyst A-025 bed temperature distribution, effects of aging.

-------
00
                                                                 A
                                2400°F BED
TPH = 697°F
Run 0427-03
TC
1
2
3
4
7
8
9
IE
2277
2110
2432
1964
2390
2341
2324
TPH = 500 °F
Run 0427-07
TC
1
2
3
4
7
8
9
IE
1754
2015
2109
969
2414
2334
2173
                                                          Flow









1
o



2
0


3
0


4





9
/%


7
O
8
O
2700°F BED
TPH = 778°F
Run 0428-14
TC_
1
2
3
4
7
8
9
IE
2461
2203
2577
2415
2464
2391
2711
TPH = 501 °F
Run 0428-20
1C
1
2
3
4
7
8
9
IE
2342
2551
2588
1454
2649
2546
2709
                    Figure B-2.  Catalyst A-025 bed temperature distribution,  effects  of  preheat.

-------
CO
CO
                                                        Flow
                              2400°F BED
                      = 4.7 Ibm/hr  mf = 8.3 Ibm/hr
                     Run 0428-04      Run 0428-10
TC
1
2
3
4
7
8
9
IE
1899
1725
2218
2270
2246
1997
2551
1C
1
2
3
4
7
8
9
IE
1365
1625
1641
1742
2297
1458
2511








t
1 3
0 0


4
2
O



9
f\
\j

1
O
8
O
2700°F BED
mf = 4.7 Ibm/hr
Run 0429-02
1C
1
2
3
4
7
8
9
IE
2428
2245
2587
2407
2496
2419
2717
•
mf = 11 .6 Ibm/hr
Run 0429-09
1C.
1
2
3
4
7
8
9
IE
2287
1716
1835
2403
2489
1551
2735
                  Figure B-3.  Catalyst A-025 bed temperature distribution, effects of throughput.

-------
DO
                             D
                    Nominal  test  condition
                          TA =  200  %
                                                  Q  =  100,000  Btu/hr
                                                  TPH  *  750°F
                                                  TBED = 2400°F
TA = 50%
                     Figure B-4.  Catalyst A-025 bed appearance rear view,  varying stoichiometry.

-------
                                               Varying bed temperature*
CO

o
                       TBED = 2200°F
                                                 Q = 100,000 Btu/hr

                                                 TpH = 750°F

                                                 LEAN
TBED = 24oo°F
       At Tg£D * 2700°F, combustion appeared uniform from the rear face until approximately 15 hours of test
       time was accumulated.  The appearance then began to approach that of TBED * 2400°F.

                      Figure B-5.  Catalyst A-025  bed appearance rear view, varying bed temperature.

-------
00
I
           A
                       TBED = 2400°F
                                                 TpH = 500°F

                                                 Q = 100,000 Btu/hr

                                                 TA « 200%
                                                                                 BED = 2700°F
                        Figure B-6.  Catalyst A-025 bed appearance rear view,  minimum preheat.

-------
CO
i
-  ,
fuel
                        = 8.3 Ibm/hr


                       » 2400°F
                                TA « 200%


                                TPH  « 750°F
mf  ,-  = 11.6 Ibm/hr


TBED = 27°°°F
                     Figure B-7.   Catalyst A-025 bed appearance rear view, maximum throughput.

-------
                    SEM/EDAX  Analyses  of Catalyst A-025

Small  Cell  Segment (Aft-End)  Results
       An extensive visual  and X-ray evaluation  of  sections  from  the  down-
Stream end of the catalyst  bed showed  no variations in either  appearance  or
chemical  composition of the observed active  and  inactive  catalyst sites.
Figure B-8 shows a representative series of  photomicrographs from the flow
inlet region of the small  cell segment and a high resolution EDAX count
result for the area depicted.   No platinum or iridium was detectable  at the
specimen  surface.  An EDAX  result of  Figure  B-8c (Figure  B-8d)  shows  the
aluminum  (line at 1.5), silicon (line  at 1.7), and  cerium (line at 4.8)
composition of the catalyst washcoat.
       A  significant difference in surface appearance was noted,  however,
between the flow entrance  and exit regions of the small cell segment. This
difference can be seen by  comparing the micrographs of  Figure  B-9 to  those
of Figure B-8.  In general, the surface at the outlet site (Figure B-9)
appears to have a much larger surface  area than  that of the  inlet region.
The characteristic surface  cracking seen at  low  magnifications is predomi-
nant in all areas analyzed.  Again, no platinum  or  iridium was detectable
for the outlet area sample, indicating at least  that active  catalyst  sites
are not available at the surface within the  detectable  limits  of  the  EDAX
equipment (1 percent).  The aluminum/silicon/cerium washcoat composition  is
again apparent in Figure B-9d (analysis of the area in  B-9c),  although in
varying ratios as is typical  over the  catalyst surface.   Additional trace
quantities of iron were also  detected  in some areas.

Medium Cell Segment (Mid-Bed) Results)
       The appearance of the  medium cell segment is similar  to that of the
rear, small cell segment.   Micrographs of two different areas  and their
                                    B-13

-------
i
**


                     •'--•/*••
                            -xrt'
                              \	'«.*	
                                      ', -  ?»*
   ^   : V^       V> T
             cV*
""'^*   ,1/'-
-  ' ^v *~                   s(|
    '  !   ,',^'    .

      '   •   i' •   -      is
                             2 mm
           a.  35  x magnification at segment inlet

             ,4-
            A
                             10 jum
                c.  2000 x magnification
               /^\
*,        jOv
                                                        -


                                                                                          -
                                                                                         »
                                                           40 /im

                                              b.  500 x magnification
                                           d.  EDAX  analysis of area c

-------
CD


cn
                                                    s*
                                                      X
                                                                      y        ?m.
                                                                    •-•*•    LJtVSSs
a.   200 x magnification at segment outlet

  »*g   /  *
                                                                b.  260 x magnification of a similar area
                                                                                I
                                                                   Al  Si      Ce    Fe

                         2000 x magnification                          d.  EDAX analysis  of area  c

                         Figure B-9.   Surface appearance and EDAX  analysis of small cell  segment
                                      outlet area, W. R. Grace  and Co.  catalyst.

-------
respective EDAX results are shown in Figure B-10.  Variations in appearance
along the cell length were not significant.
       Extensive EDAX data was taken at specific surface sites -- with
broad area scans, at specific grain sites and into cracked areas.  Again,
no active precious metal was detectable.   Wide variations in the washcoat
composition are apparent in Figures B-lOb and B-lOd.
       It was concluded from the above results that for the catalyst to
remain active at the end of the 20-hour test period,  the precious metal must
be embedded well below the surface or finely dispersed such that local  con-
centrations were below the 1-percent detectability of the EDAX analyzer.  A
series of data was therefore generated on an untested catalyst segment pre-
pared at the same time as the tested segments.

Untested Catalyst Segment Results
       Photomicrographs of the untested segment are shown in Figure B-ll.
Note the change in washcoat surface structure following testing (Figures B-8
and B-9).  Figure B-lla shows the washcoat edge with  a relatively smooth
surface (foreground) covering the more granular structure below the surface.
Figures B-llb through B-lld show these edge details at greater magnification.
The EDAX analysis for Figures B-llc and B-lld are shown in Figures B-12a and
B-12b respectively.  Extensive searching of the surface composition revealed
platinum in some areas (B-12b) but not in others (B-12a).  No iridium was
detected at any location.

Chemical Analysis of a Post-test Segment
       Since very little catalyst material (platinum  or iridium) could be
detected by the EDAX analyses on the post-test catalyst, a chemical  analysis
was performed.  This technique would verify the relative quantities of
catalyst within the complex structure that might not  be detectable by the
surface measurements of the EDAX.
       The results of the semiquantitative analysis are listed below in
Table B-4.
                                    B-16

-------
CO
              a.   285 x magnification at segment outlet
                                                     Alt          fee

                                                       b.  EDAX analysis of area a,
             c.
d.  EDAX analysis of area c.
265 x magnification of a similar area

          Figure B-10.  Surface appearance and EDAX analyses of medium cell
                        segment, W. R. Grace and Co. catalyst.

-------
CO

CO
                      i
               •
                                                       •

                       s~



                                   10



                                                                         *

                                                                                      1


                                 "
                               »  4 V

                                                                                                       I


                                                                                                           1
                            	-j 20 /urn
              a.  700 x magnification of washcoat edge     b.
               h	H 10 jum
1500 x magnification of  typical  washcoat  surface
   d.  3500 x magnification of washcoat edge
             c.   1400 x magnification  of  unusual  washcoat
                         crystalline  growth
            Figure B-ll.  Surface appearance of pretest catalyst  surface,  W.  R.  Grace  and Co. catalyst.

-------
CD
I
                      f  I
                      AT   S       Ce

                      a.  EDAX analysis of  area  c.
b.   EDAX analysis of area d.
                              Figure  B-12.   EDAX  analysis  of untested catalyst surface.

-------
             TABLE B-4.  RESULTS OF SEMIQUANTITATIVE ANALYSIS
Al-
Si-
Mg-
Mn-
Fe-
La-
Ce-
Ca-
Pt-
Ni-
Ag-
Ti-
Na-
Cu-
Cr-
Pb-
Ga-
Sm-
Ir-
Other elements
44.%
4.8
0.52
0.015
0.081
0.89
0.79
0.026
0.19
0.0013
0.0012
0.0081
0.26
0.00022
0.066
TR < 0.01
0.0080
TR < 0.03
ND < 0.07
nil
       The measurements indicate that platinum was  present  in  the  sample at
0.19 percent by mass, and iridium was not detectable  at  the 0.07 percent
sensitivity limit.
       The presence of platinum (not detected  by EDAX) indicates that  plati-
num is not present at the surface at the minimum detectable limit  of
1 percent.  Because EDAX measurements were made extensively at varying
magnifications, it does not appear that the catalyst  materials have accumu-
lated at the surface of the tested catalyst.   Similarly,  the relative  dif-
ficulty of detecting platinum on the untested  segment indicates that the
metal is probably dispersed within the washcoat during preparation rather
than applied at the surface.   Finally, the relative activity of the catalyst
after 20 hours of testing indicates that although the catalyst may not exist
primarily at the washcoat surface, adequate contact between it and the
reactants is provided to sustain combustion.
                                    B-20

-------
          SECTION  B-2

UNIVERSAL OIL PRODUCTS CATALYST
       TEST MODEL A-026
              B-21

-------
                        TABLE B-5.   UOP LI6HTOFF TEMPERATURE CHARACTERISTICS -CATALYST A-026
Cumulative
Test Time
(Hrs)
0
10
14
17
21
23
Lightoff
Condition
Fuel Lean
Rich
Rich
Rich
Rich
Rich
Lightoff
Temperature
(°F)
875
900
780
765 j
750 >
660 \
Comments
May not be minimum lightoff temp.
nonuniform temp, profiles
Not a minimum. Unsuccessful lean
lightoff at 940°F more uniform
temp, profiles

Not minimum lightoff temps.

Subsequent lightoff at 635°F was
unsuccessful
ro
IND

-------
                                   TABLE  B-6.   TEST DATA SUMMARY - CATALYST A-026
Test Pt.
0624-02
0624-05
0624-09
0624-13
0624-17
0624-21
0627-02
0627-03
0627-04
0627-05
0627-06
0627-07
0627-08
0627-09
0628-16
0628-07
0628-08
0628-09
0628-10
0628-11
0628-14
TA%
292
270
267
274
288
265
50
50
55
56
58
269
268
289
253
253
247
224
228
234
227
SV,h
173,
161,
160,
162,
165,
153,
33,
33,
36,
37,
37,
153,
151,
165,
147,
144,
146,
135,
139,
142,
134,
r~
400
100
300
500
400
100
100
400
200
300
700
200
500
500
000
100
900
800
700
500
000
""fuel
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
, Ibm/hr
.5
.6
.6
.5
.4
.4
.4
.4
.4
.4
.4
.4
.3
.4
.4
.4
.5
.6
.7
.6
.5
""air'
229
212
211
214
218
201
37
37
41
43
43
201
199
218
193
189
193
177
182
187
175
Ibm/hr
.1
.3
.1
.3
.6
.6
.5
.8
.6
.1
.8
.9
.6
.7
.3
.5
.0
.6
.8
.6
.4
T °F
V F
667
651
647
646
645
718
570
501
350
300
251
633
602
549
388
597
550
496
448
401
326
TBEIW
2350
2409
2421
2438
2432
2376
241 0^
2419 1
2409
2435 |
2431 ;
2405 "j
2460 j
2430 j
2535 J
2702
2702s]
2713 j
2698
2707 I
2680J
Test
°F Objective


10-hr
agi ng



Minimum preheat
fuel -rich
2400°F


Minimum
preheat
fuel-lean :
2400°F '


Minimum preheat
> fuel -lean
2700°F
1
1
DO

r\i
CO

-------
                                                  TABLE B-6.   Concluded
Test Pt.
0627-11
0627-16
0627-12
0627-15
0627013
0627-14
0629-03
0629-04
0629-05

0629-06
0629-07
TA%
256
305
321
253
275
290
213
206
219

197
186
SV, hr"1
145,600
216,900
262,300
292,300
343,900
420,700
126,100
144,700
165,800

183.500
217,500
"fuel* 1bm/hr
4.4
5.4
6.5
8.8
9.6
11.1
4.5
5.3
5.7

7.0
8.8
mair> lbm/hr
191.5
287.1
347.5
384.3
453.5
556.0
164.5
188.5
216.6

238.5
281.8
T °F
V F
603
732
719
740
725
751
642
587
601

606
611
Test
TRFn , °F Objective
BEDmax
2425
2504
2419
2285
2406
2492
2680
2713
2669

2716
-2650 f


Maximum thruput
fuel lean
2400°F


Maximum thruput
fuel lean
2700°F


D3
I
ro

-------
                                     TABLE  B-7.   EMISSIONS DATA - CATALYST A-026
Test Pt.
0624-02
0624-05
0624-09
0624-13
0624-17
0624-21
0627-02
0627-03
0627-04
0627-05
0627-06
0627-07
0627-08
0627-09
0628-16
0628-09
0628-10
0628-11
0628-14
Test
Time, hrs
0.7
2.7
4.2
6.2
8.2
10.0
10.3
11.6
12.1
12.3
13.5
14.0
14.3
14.5
20.6
19.0
19.3
19.7
20.2
TA%
292
270
267
274
288
265
50
50
55
56
58
269
268
289
253
224
228
234
227
'"fuel' lbm/hr COs PP"1 N0» PPm UHC» °l° H2' °/0
4.5 0 000
4.6 ' 0 ' ' '
4.6 0
4.5 1
4.4 , 1
4.4 0100
4.4 1
4.4 1 0.72 0.04
4.4 0 0.32 0.04
4.4 0
4.4 0 0.22 0.03
4.4 105 1
4.3 76 1 o
4.4 13 2
4.4 0 1
4.6 5
4.7 , 7
4.6 0 6
4.5 15 3
03

ro
en

-------
                                                  TABLE B-7.  Concluded
Test Pt.
0627-07
0627-08
0627-09
0628-16
0628-07
0628-08
0628-09
0628-10
0628-11
0628-14
0627-11
0627-16
0627-12
0627-15
0627-13
0627-14
0629-03
0629-04
0629-05
0629-06
0629-07
Test
Time, hrs
14.0
14.3
14.5
20.6
17.3
17.6
19.0
19.3
19.7
20.2
15.2
16.5
15.4
16.2
15.5
16.1
21.2
21.8
22.2
22.5
22.6
TA%
269
268
289
253
253
247
224
228
234
227
256
305
312
275
253
290
213
206
219
197
186
^fuel »
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
5.
6.
9.
8.
11.
4.
5.
5.
7.
8.
Ibm/hr CO, ppm NO, ppm UHC, %
4 105 1
3 76 1 0
4 63 1
4 13 2
4 0 1
5
6
7 1
2
5
7
6 06
5 15 3
4 52 1
4 75 2
5 48 1
6 1393 0
8 565 0
1 771 1
5 0 1
3
7
0
3
3
4
3 04
no
en

-------
CO
i
ro
                                                                  1.2
                                                                  —•-
                                                                             4
                                                                             •
                                                          5,6
                                                          —•-
                            2400°F Bed
                                                2700°F Bed
                   0627-07
                   TC
                   1
                   2
                   3
                   4
                   6
              0628-16
              TC

              1
              2
              3
              4
              6
2305
2299
1571
2535
2352
                       0628-07
TC
1
2
3
4
6
2017
2248
1951
2702
2532
                   0628-14
1
2
3
4
6
2500
2680
2501
2519
2677
Figure B-13.
                                        Catalyst A-026 bed temperature  distribution  --
                                        effects of preheat, fuel  lean.

-------
                                                                     1,2
                                                                    —•-
                                                     5,6
I
ro
CO
                              Ph

                               0627-02
                               TC
                               1
                               2
                               3
                               4
                               6
2202
2214
2411
2313
2411
                                          Tph ' 251°F

                                            0627-06
TC

1
2
3
4
6
2212
2175
2407
2253
2426
                          Figure B-14.  Catalyst A-026 bed temperature distribution --
                                        effects of preheat, fuel rich 1598K (2400°F) bed,

-------
CO

ro
                           2400°F Bed
4.4 Ibm/hr
0627-11
TC IF
1 2287
2 2307
3 2214
4 2425
6 2380
mf * 11.1
0627-14
TC JT
1 2212
2 2005
3 1753
4 2492
6 1351
                                                                   1,2
                                            5,6
                                2700°F Bed*
                                                                ny = 4.5
                                                                0629-03
                                                8.8
IP.
1
2
3
4
6
mf
0629-07
TC   °F
                                                                    2211
                                                                     678
                                                                    1868
                                                                    2680
                                                                     670
                                           1
                                           2
                                           3
                                           4
                                           6
     699
     635
     707
    2650
     642
                                                                *  Bed  showed  significant  degredation
                                                                  at 21  hours of  testing
                           Figure B-15.
Catalyst A-026 bed temperature distribution  ~
effects of throughput.

-------
             SECTION B-3

W. PFEFFERLE PRECIOUS METAL CATALYST
          TEST MODEL A-027
                 B-30

-------
TABLE B-8.  TEST DATA SUMMARY - CATALYST A-027


03
CO



Test Pt.
0630-02
0630-04
0630-05
0630-07
0630-11
0630-14
Test
Time, Hrs.
0.2
0.4
0.8
1.0
3.0
4.5
TA% SV, hr"1
35 25,500
296 172.300
297 155,800
299 127,500 ;
302 125,400
280 118.500
mfuel» lbm/hr
4.5
4.5
4.0
3.3
3.2
3.2
'"air* lbm/nr
26.8
227.8
206.1
168.7
165.9
156.4
T °F
V F
677
724
707
675
688
504
T °F
BEDmax'
2358
2467
2444
2408
2450
2432
Initial Lightoff: Fuel rich at 935°F
Subsequent Lightoff: Unsuccessful at 965°F

-------
TABLE B-9.  EMISSIONS DATA - CATALYST A-027
- Test Pt.
0630-02
0630-04
0630-05
0630-07
0630-11
0630-14
TA%
35
296
297
299
302
280
""fuel* lbm/hr
4.5
4.5
4.0
3.3
3.2
3.2
CO, ppm
—
240
231
70
0
21
NO, ppm
—
0
0
0
0
0
                     B-32

-------
       SECTION B-4

 MATTHEY BISHOP CATALYSTS
TEST MODELS A-031 AND A-035
            B-33

-------
                        TABLE B-10.  MATTHEY BISHOP A  LIGHTOFF  CHARACTERISTICS -CATALYST A-031
Cumulative
Test Time
(Hrs)
0
10
14
16
23
Lightoff
Condition
Fuel Lean
Lean
Rich
Rich (Propane)
Rich
Lightoff
Temperature
(°F)
790
840
950
735
870
f
Comments

Very slow lightoff (6 to 8 min); unstable com-
bustion until 2200° F temperature was reached
Unsuccessful lean lightoff at 930°F
Unsuccessful rich lightoff with natural gas
Would lightoff with natural gas but would not
sustain combustion —testing terminated
03
I
CO

-------
                                             TABLE  B-ll.   DATA SUMMARY - CATALYST A-031
Test
Pt.
0802-03
0802-13
0804-02
0804-03
0804-04
0804-05
0804-07
0804-12
0808-01
0808-14
0808-12
0808-11
0808-10
0808-02
0808-03
0808-05
0808-09
0808-16
0808-17
0808-18
0808-20
0808-22
0808-25
0808-24
0808-23
TA
(%)
232
258
245
231
250
241
245
242
31
34
34
34
36
37
39
39
40
210
207
175

173



SV
(1/hr)
140,600
146,400
146,100
137,800
148,800
142,100
143,000
143.800
24,400
27,000
27,100
27,100
28,300
28,100
29,500
29,700
30,700
136,900
143,600
154,600

203,000



""fuel
(Ibm/hr)
4.6
4.3
4.6
4.5
4.6
4.5
4.5
4.5
4.7
4.8
4.8
4.8
4.8
4.8
4.8
4.8
4.9
4.9
5.2
6.6
7.3
8.8
12.2
13.1
15.1
*
mair
(Ibm/hr)
184.1
192.6
191.8
180.5
195.5
186.5
187.6
188.7
24.9
28.3
28.4
28.4
30.1
29.9
31.8
32.0
33.2
78.5
187.2
199.6
250. Oa
262.0
419.0a
450.03
519.0a
TPH
(°F)
578
635
596
508
443
352
296
255
632
493
452
409
358
324
276
198 .
173
626
591
592
602
609
628
622
615
TBed
(°F)
2154
2118
2073
2182
1754
1679
1738
1509
2050
2067
2062
2027
2035
2012
2031
1996
2021








Bed max
(°F)
2406
2406
2489
2620
2473
2440
2434
2300
2394
2444
2431
2384
2442
2394
2440
2411
2445
2446
2431
2455
2467
2450
2545
2489
2442
CO
(ppm)
	
0
0
0
0
3
1
1
	
1141
1139
1141
1137
>2000
>2000
1134
1129
	
—
0
0
0
0
0
30
NO
(ppm)
_
0
	
1
1
1
1
—
	
0
0
0
0
1
1
0
0
13
10
6
12
10
19
20
_
Test
Type
Aging

Minimum
preheat
fuel lean
2400°F


Minimum
preheat
fuel rich
2400° F





Maximum
Throughput






CO
CO
en
             aAirflow values estimated from temperature data.

-------
DO
I
co
cr>
                                                                      1,2
                                                 5,6
                                                 -•—
                                        t = 0.5 hours
                                           0802-03
TC
1
4
5
6
7
°F
1990
2026
2335
2406
2013
                       t =  5.5  hours
                         0802-13
TC
1
4
5
6
7
°F
1970
2060
2077
2406
2076
                          Figure  B-16.
Catalyst A-031 Matthey Bishop A bed temperature
distributions during aging.

-------
CO
 I
OJ
                                                                            1,2
                                                                        5,6
                           Fuel Rich
                                                        Fuel Lean
                pH
                 0808-01

                TC     °
                 1
                 4
                 5
                 6
                 7
2273
2394
2084
1923
1575
TpH = 173'F
0808-09
TC
1
4
5
6
7
°F
2198
2445
2018
2292
1151
TpH = 596°F
0804-02
TC
1
4
5
6
7
°F
2111
2024
2325
2489
1417
'PH "°J '
0804-12
TC
1
4
5
6
7
°F
1070
2160
1057
2299
959
                           Figure B-17.
                   Catalyst A-031 Matthey Bishop A bed temperature
                   distributions, effects of preheat.

-------
                                        TABLE B-12.  DATA SUMMARY - CATALYST A-035
1
Test
Pt.
1001-03
1003-02
1003-10
1004-02
1004-08
1004-09
1004-10
1004-12
1004-14
1004-16
1004-17
1004-20
1004-23
1004-26
1005-02
1005-03
1005-06
2
TA
(«)
255
196
175
188
199
196
186
188
179
184
177
181
178
172
41
41
226
3
SV
(1/hr)
155,700
120,600
104,700
109,700
120,000
116,900
112,100
114,000
108,700
113,100
109,200
111,100
108,100
101,600
28,500
30,100
125,100
4
m,- ,
fuel
(Ibm/hr)
4.7
4.6
4.5
4.4
4.5
4.5
4.5
4.6
4.6
4.6
4.6
4.6
4.5
4.4
4.4
4.6
4.2
5
"•air
(Ibm/hr)
204.8
156.7
135.2
142.2
156.0
151.9
145.3
147.8
140.6
146.5
141.1
143.8
139.8
131.1
31.0
32.9
163.7
6
TPH
(°F)
560
510
535
790
600
600
550
500
470
435
400
365
310
260
580
350
610
7
TBED
(°F)
2031
229C
2368
2046
2307
2301
2288
2270
2209
2178
2106
2080
2075
2095
2207
2290
1213
8
TBED,MAX
(°F)
2417
2429
2489
2486
2432
2425
2407
2426
2440
2432
2454
2427
2379
2374
2429
2432
2514
9















10
CO
(ppm)
—
0
5
49
0
0
0
0
0
0
0
5
9
0
>2000
>2000
4
11
NO
(ppm)
_
6
11
2
4
4
5
5
4
5
4
3
4
5
1
1
3
12
.














13















14
Test
Type

Aging



Minimum
preheat,
fuel lean
2400°F






Minimum
preheat,
rich
2400°F
Pressure
comparison
2 Atm
DO
I
CO
CO
                                                    TOTAL TEST TIME = 21 HOURS

-------
                         TABLE B-13.  MATTHEY  BISHOP  C  LIGHTOFF  CHARACTERISTICS -CATALYST A-035
Cumulative
Test Time
(Hrs)
0
1
7
14
18.5
Lightoff*
Condition
Fuel Rich
Rich
Rich
Rich
Rich
Lightoff
Temperature
920
820
850
890
920
Comments
Lean lightoff at 950°F
was unsuccessful
Rich lightoff at 750°F
was unsuccessful


Probably not a minimum
lightoff temperature
co
ID
          ""All  lightoffs  performed with natural  gas.

-------
CD
 I
                                                               1,2
                                                               —•-
5,6
                                                                                             DURING AGING
t = 1.0 hr
1001-03
TC
1
2
3
4
°F
1676
2182
1847
2417
t = 10.0 hrs
1004-08
10.
1
2
3
4
°F
2212
2432
2199
2385
                             Figure  B-18.   Catalyst A-035 bed temperature distributions,

-------
03
                                                                     1,2
                           5,6
                          —•-
                                                                                                  EFFECTS OF PREHEAT
                FUEL LEAN, 2400° F

    1004-09. TPH = 600          1004-20, TPH = 365
1C
1
2
3
4
°F
2212
2425
2193
2375
1C.
1
2
3
4
°F
2202
2108
1582
2427
             FUEL RICH, 2400°F

1005-02, TPH = 580          1005-03, TPH = 350
1C.
1
2
3
4
°F
2122
2056
2220
2429
1C.
1
2
3
4
°F
2220
2140
2403
2432
                              Figure B-19.   Catalyst A-035  bed  temperature  distributions.

-------
           SECTION B-5

W. PFEFFERLE METAL OXIDE CATALYST
        TEST MODEL A-038
               B-42

-------
TABLE B-14.  PFEFFERLE LI6HTOFF CHARACTERISTICS -CATALYST A-038 (Co203)
CUMULATIVE
TEST TIME
(MRS)
0
10.8
11.0
11.5
LIGHT OFF
CONDITION
FUEL-RICH
RICH
RICH
RICH
LIGHT OFF*
TEMPERATURE
(°F)
890
890
850
850
COMMENTS




PLATINUM ADDED TO
CATALYST TO PROMOTE
LIGHT OFF WITH
NATURAL GAS
\
                                    B-43

-------
                                         TABLE  B-15.   DATA SUMMARY  - CATALYST A-038
































1
Test
Point
1228-03
05
07
10
11

1228-12
13
14
15
16



1229-05
06
07
08
09
10
11
12
13
14
15
16



2
TA
('»)
172
157
162
160
167

159
164
157
149
150



174
168
170
177
189
177
184
190
192
188
216
221



3
SV
0/hr)
80,400
82,100
78,300
73,700
77,000

74,500
77,100
76,600
76,800
73,200



135,100
130,600
151,700
183,100
214,400
223,200
257,200
284,600
316,400
328,200
425,900
443,100



4
"1 .r T
fuel
(Ibm/hr)
4.4
4.9
4.5
4.3
4.3

4.4
4.4
4.6
4.8
4.6



7.5
7.5
8.6
10.1
11.2
12.4
13.8
14.9
16.4
17.3
19.8
20.1



5
"'air
(Ibm/hr)
130.9
132.7
126.6
118.5
124.1

120.0
124.6
123.5
123.5
117.4



224.8
216.5
253.2
308.0
363.0
377.0
436.5
484.6
539.7
559.7
733.5
764.3



6
Tph
(°F)
732
605
614
607
615

613
574
548
524
498



696
692
668
656
645
642
637
632
628
625
622
628



7
TBED
(°F)
2392
2474
2219
2348
2369

2376
2298
2209
2261
2201



2481
2476
2464
2476
2474
2510
2517
2515
2521
2534
2347
2426



8
TBED, MAX
(°F)
2423
2492
2459
2474
2474

2465
2455
2449
2467
2469



2517
2512
2503
2498
2503
2530
2539
2530
2542
2550
2445
2479



9






























10
CO
(ppm)
9
0
0
0
0

0
0
0
0
0



0
0
0
0
0
0
0
0
0
0
0
3



11
NO
(ppm)
5
17
9
10
8

9
9
11
15
14



7
7
6
5
5
5
5
4
4
4
7
4



12






























13

•v



_,>






}



)



•^










J















14
TEST
TYPE

ging at
400°F




uel lean
inimum
)reheat
400°F







naximum
through-
put
at
2500°F





CO
I

-------
DO




on
                                                                                             DURING AGING
                               1228-03
1228-11
TC
1
2
3
4
5
°F
2423
2342
2386
2415
2374
TC
1
2
3
4
5
°F
2216
2331
2474
2456
2442
                            Figure B-20.  Catalyst A-038 bed temperature  distributions.

-------
DO
I
                                                                                •
                                                                                4
                                                                                          •
                                                                                          5
                                                                                                   EFFECTS  OF
                                                                                                   PREHEAT
                                                FUEL LEAN, 2400°F

                                  1228-12, Tp(_, = 613 °F           1228-16,  TpH = 498 °F
TC
1
2
3
4
5
°F
2251
2328
2459
2465
2462
TC
1
2
3
4
5
°F
1760
2109
2469
2467
2467
                             Figure B-21 .   Catalyst A-038 bed temperature  distributions.

-------
DO
 I
                                                                       •
                                                                       3
                                                                                        EFFECTS OF THROUGHPUT
                                                   FUEL LEAN, 2400° F
                                 1229-05. TPH= 696 °F
1229-15,  'PH  = 622 °F
TC
1
2
3
4
5
°F
2467
2517
2494
2447
2561
TC
1
2
3
4
5
°F
2445
2195
2389
2357
2463
                            Figure B-22.  Catalyst A-038 bed temperature distributions.

-------
      SECTION B-6

JOHNSON MATTHEY CATALYST
    TEST MODEL A-040
          B-48

-------
                                    TABLE B-16.  LIGHTOFF CHARACTERISTICS -CATALYST A-040
CO

-p>
10
Cumulative
Test Time
(Mrs.)
0
0.2
7.8
14.5
17.3
18.3
Lightoff
Condition
Fuel Lean
Lean
Lean
Lean
Rich
Rich
Lightoff
Temperature
(°F)
760
790
800
850
800
850
Comments



Rich lightoffs were
unsuccessful
Apparent poor conversion
Nonuniform combustion

-------
TABLE B-17. DATA SUMMARY -CATALYST A-040
Test
Pt.
0126-02
0126-03
0126-04
0126-06
0126-07
0126-08
0126-09
0127-02
0127-03
0127-04
0127-05
T3 0127-07
o 0127-09
0127-10
0127-12
0127-14
0127-15
0130-01
0130-02
0130-04
0130-06
0130-07
0130-09
0130-12
TA.
240
244
240
240
240
246
250
234
232
236
246
230
236
216
206
203
201
258
250
230
236
231
230
214
SV
(1/hr)
149,700
144,100
143,100
130,600
131,300
133,100
132,700
120,000
121,900
116,800
137,500
124,200
161,600
165,000
135,500
142,800
138,900
127,200
158,900
233,600
287,300
383,000
438,900
602,500
m fuela
(Ibm/hr)
4.77
4.52
4.56
4.16
4.19
4.15
4.07
3.94
4.01
3.78
4.28
4.13
5.24
5.82
5.00
5.34
5.25
3.78
4.87
7.76
9.31
12.68
14.58
21.44
m air
(Ibm/hr)
196.9
189.7
188.3
171.8
172.7
175.3
174.9
158.6
160.1
153.5
181.1
163.1
212.4
216.1
177.1
186.5
181.3
167.8
209.4
306.8
377.7
503.6
576.4
788.7
Tph
793
700
692
714
701
698
699
665
665
678
688
622
528
499
327
304
277
852
707
689
591
643
637
568
fbed
2303
2301
2306
2308
2316
2316
2302
2341
2341
2359
2283
2320
2333
2324
2341
2342
2324
2341
2343
2358
2349
2369
2369
2340
Tbed max
2484
2353
2371
2377
2373
2366
2353
2389
2393
2414
2327
2386
2377
2371
2388
2385
2360
2401
2406
2452
2403
2403
2407
2402
CO
(ppm)
22
13
5
1
0
26
1
-
-
5
42
0
0
0
7
27
28
19
17
15
18
15
14
12
NOX Test
(ppm) Type
6
8
9
11
9
6
11
-
-



Aging



-
26
20
19
19
45
39
0


Lean
Minimum
Preheat


-
18
8
5
3
2
1

Maximum
throughput




-------
Test
Pt.
0130-13
0130-14
0130-16
0130-18
0130-20
0130-22
0130-24
0130-25
01 30-26
01 30-27
TA
224
212
218
200
198
194
192
189
56
57
SV
(1/hr)
585 ,800
552,500
527,200
502,200
480,000
479,300
478,600
482,900
34,900
35,700
m fuel3
(Ibm/hr)
19.96
19.84
18.43
19.06
18.39
17.86
18.88
19.34
4.18
4.25
TABLE
m air
(Ibm/hr)
768.4
722.9
690.7
655.3
625.9
595.6
623.1
628.3
40.4
41.4
B-17. Concluded
Tph Tbed
523
466
396
347
299
250
199
175
633
720
2310
2280
2303
2367
2360
2356
2362
2352
^
-
Tbed max
2376
2346
2352
2408
2401
2403
2397
2397
_
-
CO
(ppm)
12
12
14
8
7
5
5
5
_
-
NQV Test
X
(ppm) Type
1
1
3
5
5
5
5
6


High through-
put
, Lean Minimum
preheat



10 Rich
.„ Operation
03
 I
cn
         Fuel flow  rates  and  stoichiometry  are approximated based on

         bed temperature  due  to  fuel  flow meter calibration errors.

-------
        SECTION B-7

 HIGH TEMPERATURE CATALYSTS
TEST MODELS A-029 AND A-030
             B-52

-------
TABLE B-18.  TEST DATA SUMMARY -CATALYST A-029 (NiO/Pt)
Test Pt.
0710-03
0710-04
0710-05
0710-06
0710-08
0710-09
0710-11
0710-13
0710-15
0710-16
0710-18
TA%
231
219
210
207
189
189
186
177
171
171
161
SV, hr'1
93,400
148,200
141,200
142,400
129,600
129,300
126,600
121,900
119,400
120,200
113,400
•
mfuel ,
Ibm/hr
2.6
4.4
4.3
4.4
4.4
4.4
4.4
4.4
4.5
4.5
4.5
•
mair,
Ibm/hr
104.0
164.6
156.6
157.8
143.0
142.6
139.6
133.9
130.9
131.8
124.0
TpH, °F
638
644
645
643
642
641
643
644
644
644
642
Tbed
Avg, °F
2423
2454
2557
2669
2776
2803
2845
2905
2947
3004
3056
Tbed
max, °F
2437
2543
2648
2745
2805
2856
2902
2952
3005
3049
3100
                          B-53

-------
TABLE B-19.  EMISSIONS DATA -CATALYST A-029 (NiO/Pt)
Test Pt.
0710-03
0710-04
0710-05
0710-06
0710-08
0710-09
0710-11
0710-13
0710-15
0710-16
0710-18
TA%
231
219
210
207
189
189
186
177
171
171
161
A
""fuel, Ibm/hr CO, ppm NO, ppm
2.6 0 6
4.4
4.3
4.4
4.4
4.4
4.4
4.4
4.5
4.5
9
11
20
33
43
58
86
119
166
4.5 0 213
                       B-54

-------
TABLE B-20.  LIGHTOFF CHARACTERISTICS -CATALYST A-030 (Co203/Pt)
Cumulative
Test Time
(Hrs)
0
7
14
21.5
Lightoff
Condition
Fuel Rich
Rich
Rich
Rich
Lightoff
Temperature
(°F)
960
950
800
750
Comments
Lean lightoff at
860°F unsuccessful
Natural gas
Lean lightoff at
950°F unsuccessful
Natural gas
Propane lightoff,
switch to natural gas
Propane lightoff
                                B-55

-------
                                TABLE B-21.   SCREENING DATA  SUMMARY - CATALYST A-030  (Co203/Pt)
tn
Test Pt.
0714-04
0714-10
0715-04
0715-07
0715-08
0715-11
0715-05
0715-06
0715-09
0715-10
0715-13
0715-14
0715-15
0715-16
0715-17
0715-18
0715-19
0715-20
0715-21
0715-22
0715-23
0715-24
0715-25
0719-03
0719-04
0719-06
0719-08
0719-10
0719-12
0719-14
0719-21
0719-22


TAX
249
253
252
250
243
243
256
232
224
221
274
265
255
256
279
260
239
232
226
229
241
251
237
-200
-250*
TA




-50*
TA


SV, hr-1
170400
174000
177100
173200
169500
169800
160500
164000
156900
154400
185000
225700
309900
342700
455500
175900
163500
215000
260700
337400
427900
489700
163800











""fuel, Ibm/hr
4.4
4.5
4.6
4.5
4.5
4.5
4.6
4.6
4.5
4.5
4.4
5.6
7.9
8.7
10.7
4.4
4.4
6.0
7.5
9.5
11.5
12.5
4.5



-4.4a
Ibm/hr






"'air, Ibm/hr
190.4
194.5
197.9
193.5
189.1
189.5
178.6
182.7
174.5
171.6
207.4
252.8
346.5
383.2
511.1
196.9
182.4
239.4
290.1
375.6
477.4
547.7
182.6
186.7
176.3
188.4
194.2
216.6
202.8
184.6
39.4
37.5


Tph, °F
652
597
675
'576
502
394
649
591
471
409
786
747
683
676
683
662
652
653
658
661
670
676
649
652
648
644
644
643
640
633
559
439


TBED, °F
2084
2054
2181
2020
1970
2214
2561
2490
2459
2432
2235
2091
2211
1869
1764
1999
2670
2233
2233
2052
1965
1935
2747
1914
1968
2151
2365
2741
2862
2950
2734
2679


TBED, Max Test CO, ppm NO, ppm
2597
2756
2710
2665
2647
2657
2856
2839
2835
2845
2683
2647
2666
2618
2657
2664
2863
2872
2842
2869
2875
2846
2876
2553
2598
2719
2830
2913
2997
3111
2773
2741


Aging 0 5

Min.
preheat
2600°F
lean
Min.
preheat
2800°F
8
13
6
6
6
14
13
17
lean 0 15
Max. 18 5
thruput 0 4
2600°F
lean


Max.
thruput
28000F
lean



3100°F
emissions
lean



5
4
4
6
13
12
10
9
8
8
15
3
4
6
10
14
25
0 46
Min. >2000 1
preheat >2000 0
2700°F
rich
             Inaccuracies in fuel flow measurements make exact values unknown.

-------
DO
 I
Ol
                                                    flow
                                                                                5,6
                                                                                         Bed  temperature distributions
                                                                                         effects of preheat
                                                                     2800°F Bed
                                                           TPH =650

                                                           ...0715-05

                                                           TC      °F
                                 TPH = 410

                                  0715-10

                                TC     °F
                                                                 2665
                                                                 2542
                                                                 2719
                                                                 2856
                                                                 2141
                                      2664
                                      2605
                                      2845
                                      2791
                                      2563
                               Figure  B-23.
Catalyst A-030 (Acurex COpO./Pt)  bed  temperature
distributions.

-------
co
 i
en
co
                                                              1,2
                                                             -•—
                                                                       4
                                                                       •
                                                       5,6
                                                      —•—
Bed temperature distribution
effects of throughput
                           2600°F Bed
                mfuel  = 4-4

                  0715-13
                TC
                1
                2
                4
                5
                6
2564
1925
2341
2683
1661
mfuel =10.7
0715-17
TC
1
2
4
5
6
IF
1023
1140
1807
2657
1126
                                             2800°F Bed
""fuel
= 4.5
0715-19
It
1
2
4
5
6
IE.
2672
2629
2723
2854
2461
mfue]
= 12.5
0715-24
TC
1
2
4
5
6
If.
1851
1143
2604
2831
1176
                             Figure B-24.   Catalyst A-030 (Acurex Co203/Pt)  bed temperature
                                             distributions.

-------
   SECTION  B-7

CATALYST SCALEUP
TEST MODEL A-041
    B-59

-------
TABLE B-22.  UOP SCALEUP CATALYST LIGHTOFF CHARACTERISTICS - CATALYST A-041
Cumulative
Test Time
(Hrs)
0
6
12.5
18.0
19.5
22.0
24.0
Lightoff
Condition
FUEL LEAN
RICH
RICH
RICH
RICH
RICH
RICH
Lightoff
Temperature
(°F)
840
750
770
660
700
860
860
Comments
NATURAL GAS





•






                                    B-60

-------
                                         TABLE B-23.  DATA  SUMMARY — CATALYST A-041


































1
Test
Point
1230-02
05
07
10
0102-02
05

0102-06
07
09
11
13
15
16
17
18
19

0103-02
03
04
05
06
07
08
09
10




2
TA
(*)
248
259
244
255
230
220

231
221
212
210
215
200
188
189
186
176 '

55
54
55
54
56
56
54
57
58




3
SV
(Vhr)
127,800
133,200
124,800
129,600
120,100
107,500

115,800
112,200
108,400
107,700
109,600
106,200
102,800
104,700
'108,000
106,300

30,500
30,700
31 ,200
31,300
31 ,900
32,500
33,400
34,100
35,200




4
"fuel
(Ibm/hr)
10.4
10.4
10.3
10.3
10.7
10.0

10.3
10.4
10.4
10.4
10.4
10.3
11.1
11.3
11.8
12.2

10.0
10.2
10.2
10.3
10.2
10.4
11.1
10.8
11.0




5
"air
(Ibm/hr)
442.4
462.2
.431.7
449.8
422.8
377.4

407.6
394.0
379.9
377.3
384.5
371.3
358.1
364.8
376.0
369.1

94.1
94.4
96.2
96.4
98.7
100.7
102.8
105.8
109.6




6
TPH
(°F)
662
636
639
607
630
642

640
599
549
500
451
349
299
250
201
177

501
449
448
400
349
299
250
225
200




7
TBED
(°F)
2349
2354
2344
2332
2354
2353

2401
2400
2389
2389
2380
2398
2395
2397
2408
2393

2373
2352
2352
2364
2379
2375
2383
2376
2384




8
TBED, MAX
(°F)
2351
2354
2347
2334
2356
2355

2402
2405
2396
2402
2394
2402
2397
2401
2411
2396

2406
2376
2375
2388
2407
2400
2406
2397
2404




9
































10
CO
(ppm)
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0

>2000
>2000
>2000
>2000
>2000
>2000
>2000
>2000
>2000




11
NO
(ppm)
3
4
3
3
3
3

5
5
5
5.
5
5
* 7
8
10
17

1
1
1
1
1
1
1
1
2




12
































13
































14
TEST
TYPE

Aging







lean fuel
minimum
preheat
2350°F








rich fuel
minimum
preheat
2400°F





CO
I
CT>

-------
                                                TABLE B-23.  Continued
































1
Test
Point
0103-13
14
15
16
17
18
19
20
21
22

0104-03
04
05
06
08

0104-09
10
11
12
13
14

0105-03
04
05
06

2
TA
(*)
221

240
235
240
256
258
252
266
274

202
189
198
230
229

174
180
200
203
215
232

175
221
221
234

3
SV
(1/hr)
137,800

173,600
199,500
233,900
281,200
310,100
338,600
386,850
433,700

108,000
117,000
137,600
209,700
253,700

97,900
105,700
125,700
143,800
174,700
207,100

96,600
151,400
213,200
297,300

4
"fuel
(Ibm/hr)
12.8

14.9
17.4
20.0
22.6
24.7
27.6
30.0
32.6

10.9
12.5
14.1
18.6
22.6

11.3
11.8
12.8
14.4
16.6
18.3

11.1
14.0
19.7
26.0

5
"air
(Ibm/hr)
483.9

612.0
702.8
824.7
994.0
1096.3
1195.9
1369.4
1537.2

377.9
408.9
480.8
738.9
892.7

339.6
367.6
439.3
503.1
613.0
728.9

335.3
531.8
749.0
1047.1

6
TPH
(°F)
616

644
637
634
_ 636
634
634
635
637

399
379
382
395
407

251
248
241
238
242
239

615
604
613
628

7
TBED
(°F)
2339

2348
2372
2372
2351
2384
2373
2334
981

2371
2363
2359
2355
1413

2367
2370
2342
2331
2326
2015

2386
2381
2389
2374

8
TBED, MAX
(°F)
2354

2363
2388
2391
2371
2394
2397
2349
1249

2377
2378
2371
2368
1868

2385
2379
2375
2354
2375
2183

2390
2386
2398
2377

9






























10
CO
(PPm)
44

20
21
31
0
2
42
149
52

15
11
6
7
0

1
0
0
0
19
10

1
0
0
14

n
NO
(ppm)
3

2
2
2
1
1
1
1
. 1.

6
9
7
7
3

10
10
7
6
12
6

5
2
2
. 2

12






























13






























14
TEST
TYPE
maximum
through-
out
2400°F







blowout
@400°F Pl-
2400°Fbed



blowout
@2500°F PI
2400°F be<




blowout
@635°F PH
2400° Fbed

DO
I
cn

-------
                                                  TABLE B-23.  Continued


































1
Test
Point
0105-08
09
10
11
12
13
14


0105-17


20
21
22
23
24
25
26
27
28
29
30








2
TA
(%)
248
242
240
238
262
248
250


197


245
249
255
263
259
257
261
252
246
237
236








3
SV
(1/hr)
385,900
430,000
512,700
504,400
601 ,800
589,500
588,600


104,700


265,300
307,400
314,600
345,300
410,100
406,100
429,300
432,800
433,800
443,000
453,300








4
"fuel
(Ibm/hr)
32.0
36.4
43.9
43.5
47.2
48.7
48.3


10.8


22.2
25.4
25.3
27.0
32.6
32.5
33.9
35.2
36.2
38.3
39.3








5
""air
(Ibm/hr)
1362.1
1516.5
1807.2
1777.6
2129.1
2081.0
2078.4


365.8


936.2
1085.3
1112.0
1221.8
1450.1
1435.7
1518.5
1528.7
1530.9
1561.0
1597.0








6
TPH
(°F)
626
624
626
625
626
625
625


776


761
765
768
765
761
761
758
756
756
753
754








7
TBED
(°F)
2347
2405
2408
2431
2324
2268
856


2361


2405
2381
2388
2372
2369
2370
2390
2355
2344
2371
2361








8
TBED, MAX
(°F)
2349
2408
2418
2437
2346
2313
973


2370


2411
2387
2397
2379
2374
2379
2398
2367
2355
2380
2373








9
































10
CO
(ppm)
15
15
17
21
20
22
21





38
31
33
0
0
0
0
0
0
0
0








11
NO
(ppm)
2
1
1
1
1
1
2


5


1
1
1
1
1
1
1
1
T
1
2








12
































13
































14
TEST
TYPE









blowout^
50°F PH
400°F be<


















O1
CO

-------
                                                  TABLE  B-23.   Concluded


































1
Test
Point
0105-31
32
33
34
35
36
37
38
39
40
41




















2
TA
(X)
235
230
229
225
210
192
191
183
169
164
159




















3
SV
0/hr)
457,600
454,600
450,700
448,300
429,800
396,500
389,200
395,500
383,200
369,900
367,300




















4
""fuel
(Ibm/hr)
39.9
40.4
40.3
40.7
41.7
41.9
41.4
43.7
45.5
45.4
46.3




















5
"air
(Ibm/hr)
1611.7
1600.0
1585.6
1575.8
1506.4
1383.4
1357.3
1376.5
1327.7
1279.1
1267.9




















6
TPH
(°F)
704
652
650
595
506
386
296
248
200
150
125




















7
TBED
(°F)
2329
2360
2355
2367
2367
2365
2379
2367
2382
2346
2362




















8
TBED,MAX
(°F)
2344
2372
2367
2379
2379
2380
2397
2391
2397
2379
2385




















9
































10

0
0
0
0
0
0
0
0
0
0
0




















11
CO
(ppm)
1
2
1
2
2
2
3
4
11
18
17




















12
NO
(ppm)































13
































14
TEST
TYPE
ean
inimum
preheat?
O'Btu/hr
400°F
1 Atm
























DO
I

-------
        SECTION B-8

    EXTENSIVE  EVALUATION
TEST MODELS A-03'6 AND A-037
            B-65

-------
                              TABLE B-24.   EXTENSIVE EVALUATION  SUMMARY  - CATALYST A-036 (NiO/Pt)
1
Test
Point
1010-03
1010-04
1010-05
1010-06
1010-07
1010-08
1010-09
1010-10
1010-11
1010-12

1012-03
1012-04
1012-05
1012-06

1012-07
1012-08
1012-09
1012-11
1012-12
2
TA
(%)
57
57
65
65
75
75
93
93
100
100

190
190
145
145

101
101
101
80
80
3
SV
Hr"1
113,800
115,700
103,100
103,600
119,800
119,600
115,700
115,300
106,200
106,900

103,600
104,000
101,100
98,400
98,400
98,400
108,100
108,200
109,500
111,200
107,500
4
ppm
NH3 Fuel
0
5000
0
5000
0
5000
0
5000
0
5000

0
5000
0
5000
2500
10,000
0
5000
10,000
0
5000
5
Vfuel
SCFM
4.44
4.45
3.46
3.47
2.91
2.91
1.90
1.90
1.83
1.84

1.36
1.36
1.37
1.36
1.36
1.36
1.48
1.49
1.48
2.03
2.03
6
Vair
SCFM
23.60
24.12
20.84
20.94
19.65
19.60
16.13
16.11
16.58
16.82

24.00
24.09
19.56
18.85
18.85
18.85
13.55
13.55
13.90
16.02
15.12
7
ppm
NH3 Gas
0
779
0
711
0
646
0
527
0
493

0
267
0
336
168
673
0
495
991
0
592
8
Bed max.
Temp max.
2650
2630
2685
2680
25801
2580
24801
2480
25001
2500

2340
2340
2375
2375

2400
2400
2400
2440
2440
9
ppm
NO
Thermal
0

0

0

45

47

9

22

30

26

10
ppm
NO
Total

0

0

15

298
140-2
360

187
80-460
60-255
625
100-360
600

605
11
ppm
Fuel NO

0

0

15

253
93-
313

178
58-438
38-233
603
70-330
570

579
12



















13
% NH3
Converted

0

0

2.32

48.0
41. 23


66.7
73.73
80. 83
89.6
40. 43
57.5

97.8
14



















CO
I
cr>
cr>
       1.   Pyrometer estimates (e = 0.6)

       2.   Continuously varying data

       3.   Averaged values

-------
                                     TABLE B-25.  FUEL NITROGEN DATA - A-037 (Co^/Pt)






























1
Test
Point

1222-04
1222-05

1222-07
1222-08

1222-09
1222-10

1222-11
1222-12

1222-13
1222-14

1222-15
1222-16

1222-17
1222-18

1222-19
1222-20
1222-21


2
Theore-
tical
Air
(%)

53
53

48
48

63
63

-v-71
72

82
78

89
88

97
99

105
103
104


3
Space
Velocity
(Hr'1)

39,000
40,200

37,000
36,800

45,100
45,100

47,600
57,000

60,700
60,500

56,300
56,800

58,500
59,900

66,400
65,400
65,800


4
Dopant
Concen-
tration
(ppm)

0
20,770

0
20,770

0
20,770

0
15,980

0
15,980

0
22,550

0
22,550

0
21,760
29,260


5
Fuel
Rate
(SCFM)

1.7
1.7

1.7
1.7

1.7
1.7

1.7
1.9

1.8
1.9

1.4
1.5

1.5
1.5

1.6
1.6
1.6


6
Air
Rate
(SCFM)

8.4
8.7

7.8
7.8

10.0
10.0


12.9

13.9
13.8

13.1
13.1

1.3.7
14.0

15.6
15.4
15.5


7
Max. Bed
Temp.
(°F)

2483
2438

2542
2538

2564
2549

2552
2528

2573
2554

2545
2537

2534
2536

2559
2571
2579


8


Rear Segt.
blowout

Raise bed
temp.





















9
ppm NH-,
in gas


3395


3717


3018


2051


1934


2317


2182


2048
2747


10
Thermal
NOX
(ppm)

1


5


3


4


9


24


25


25




11
Total
N0x
(ppm)


95


22


30


66


440


775


1090


1350
1635


12
Fuel
NOX
(ppm)


94


17


27


62


431


751


1065


1325
1610


13




























14
Conver-
sion
NOX
(%f


2.8


0.46


0.39


3.0


22.3


32.4


48.8


64.7
58.6

CO

-------
                                                        TABLE B-25.  Concluded






























1
Test
Point
1222-22
1222-23
1222-24


1223-04
1223-05

1223-08
1223-09

















2
Theore-
tical
Air
(*)
131
131
131


178
178

176
176

















3
Space
Velocity
(Hr'1)
80,700
81,100
81 ,000


195,600
197,000

205,200
207,900

















4
Dopant
Concen-
tration
(ppm)
0
21,630
29,360


0
17,330

0
20,860

















5
Fuel
Rate
(SCFM)
1.6
1.6
1.6


2.7
2.7

2.9
2.9

















6
Air
Rate
(SCFM)
19.3
19.4
19.4


46.0
46.2

48.2
48.3

















7
Max. Bed
Temp.
(°F)
2555
2568
2569


2532
2503

2545
2522

















8






2 Atm


3 Atm


















9
ppm NHj
in gas

1648
2237



985


1170

















10
Thermal
NOX
(ppm)
55




54


88


















n
Total
NOX
(ppm)

1490
1780



865


789

















12
Fuel
NOX
(ppm)

1435
1725



831


701

















13




























14
Conver-
sion
NOX
(*)

87.1
77.1



34.4


59.9
















co
i
cr>
CO

-------
                                   TABLE  B-26.   FUEL LEAN DATA FOR  FUEL  NITROGEN CONVERSION
CO
 I
CTi
ID
Type
1 NASA
2
3
4
5
6
7
8
9
10
11
12
13
14
15 AERO Co203/Pt
16
17
18
19
20
21
22
23
24
25 AERO Pt
26
27
28 AERO NiO/Pt
29
30
31 ENGELHARD*
32
33 AERO Co203/Pt
34
35
36
37
38

SV, 1/hr
. 2842+06
. 2842+06
.2842+06
. 2842+06
.2842+06
.2842+06
.2842+06
.2842+06
. 2842+06
.2842+06
. 2842+06
. 2842+06
. 2842+06
. 2842+06
.1583+06
. 1 489+06
.1406+06
.2739+06
.2685+06
.2525+06
.2503+06
.2407+06
.2407+06
. 3960+05
.9270+05
. 9250+05
.8930+05
.1040+06
. 9840+05
. 1 095+06
.1900+06
. 1 300+06
.2176+05
.2936+05
. 81 1 0+05
.8100+05
.1970+06
. 2079+06

TBED, °F
1826.
1880.
1898.
2006.
1952.
1988.
1970.
1979.
2015.
2060.
2060.
1916.
1925.
1961.
2658.
2866.
2985.
2685.
2663.
2846.
2842.
3070.
3050.
2681.
2340.
2400.
2400.
2340.
2375.
2400.
2340.
2340.
2573.
2579.
2568.
2569.
2504.
2522.

Ko, % N in fuel
.1350-01
.1350-01
.1350-01
.1350-01
.1350-01
.1350-01
.1350-01
.1350-01
.1350-01
.1350-01
.1350-01
.1350-01
.1350-01
.1350-01
.4370
.4370
.4370
.4370
.2180
.4370
.2180
.4370
.2180
.4370
.5240
.8730-01
.5240
.4370
.8730
.8730
.1700
.9400
.1901+01
.2566+01
.1891+01
.2566+01
.1558+01
.1823+01

P, atm
3.000
3.000
3.000
3.000
3.000
3.000
3.000
3.000
3.000
3.000
3.000
3.000
3.000
3.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
3.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
2.000
3.000

Conversion
Measured, %
85.30
77.20
75.00
82.80
75.60
74.00
72.30
71.10
71.60
72.50
69.70
66.70
66.40
63.90
68.60
79.20
78.80
84.00
64.00
89.50
64.90
96.10
79.30
23.40
26.70
19.60
68.20
66.70
89.60
57.50
81.50
81.50
64.7
58.6
87.1
77.1
84.4
59.9

Conversion
Calculated, %
68.40
71.46
72.48
78.76
75.60
77.70
76.65
77.17
79.29
81.96
81.96
73.52
74.04
76.12
59.81
64.15
65.51
89.13
77.14
91.88
80.96
99.55
87.58
22.97
35.03
26.12
85.37
36.82
40.73
44.59
47.81
49.02
18.2
23.8
45.6
48.0
79.1
85.3

              *Engelhard .bed  temperature was assumed to be 2340°F based on  reported  stoichiometry.

-------
                  APPENDIX C

SECTION 9 DATA SUPPLEMENT — COMBUSTION SYSTEM
              CONFIGURATION TESTS
                      C-l

-------
                                 TABLE C-l.  DATA SUMMARY - TWO STAGE COMBUSTOR
Test
Pt.
0302-03
0302-04
0302-05
0302-06
0302-07
0302-08
0302-09
0302-10
0302-11
0302-12
0302-13
0302-14
0302-15
0302-16
0302-17
0302-18
0302-19
0302-20
0302-21
0302-22
0302-23
0302-24
Fuel
Rate
(Ibm/hr)
8.76
8.73
8.50
8.48
8.80
8.74
8.83
8.84
8.74
8.65
8.64
8.61
8.97
9.00
9.07
9.08
9.40
9.44
9.49
9.42
9.02
9.02
Primary
Air
(Ibm/hr)
60.9
61.0
74.7
75.0
82.1
81.5
76.6
76.7
76.5
75.5
76.0
75.6
77.1
77.3
78.0
77.7
78.6
78.9
78.7
77.8
77.8
77.5
Primary
TBed
(°F)
2250
2228
2230
2165
2232
2120
2314
2183
2319
2235
2507
2527
2393
2386
2466
2341
2197
2227
2275
2256
2554
2430
Primary
SV
0/hr)
66,100
66,200
77,600
77,800
84,400
83,800
79,700
79,900
79,500
78,500
78,900
78,600
80,400
80,600
81 ,300
81 ,000
82,400
82,700
82,500
81 ,600
81 ,000
80,800
Tph
(°F)
665
649
635
635
642
642
641
643
644
644
645
646
645
645
646
647
649
650
650
651
656
657
Primary
TA
(%)
40
40
51
51
54
54
50
50
51
51
51
51
50
50
50
50
49
49
48
48
50
50
Interstage
Energy
Extracted
(Btu/hr)
43,900
46,800
53,300
54,600
65,600
68,200
60,500
66,900
64,900
66,500
67,400
66,900
66,000
65,500
65,300
65,500
63,300
62,700
60,900
62,300
55,700
55,300
Secondary
Air
(Ibm/hr)
41.9
41.9
40.0
40.0
40.0
40.0
66.4
66.4
76.5
76.5
89.6
89.6
30.8
30.8
85.5
85.5
115.2
115.2
155.7
155.7
92.7
92.7
Secondary
TBed
(°F)
2204
2138
2269
2251
2309
2289
2343
2318
2206
2177
2256
2241
2257
2253
2271
2278
2333
2326
2231
2225
2338
2341
Overall
SV
(1/hr)
50,400
50,400
55,300
55,400
58,700
58,400
67,700
67,700
71,900
71 ,400
77,300
77,100
74,200
74,300
76,700
76,600
89,900
90,100
107,500
107,000
79,700
79,500
Overall
TA
(*)
68
69
79
79
81
81
94
94
102
102
112
112
102
102
105
105
120
120
144
144
110
110
Pressure
(Atm)
1



































i
1
2
2
o
I
ro

-------
                                                    TABLE C-l.   Concluded
Test
Pt.
0308-02
0308-03
0308-04
0308-05
0308-07
0308-08
0308-09
0308-10
0308-11
0308-12
0309-02
0309-03
Fuel
Rate
(Ibm/hr)
4.45
4.45
4.45
4.45
4.42
4.42
4.47
4.47
4.56
4.56
8.03
7.99
Primary
Air
(Ibm/hr)
37.6
37.4
38.1
38.6
52.4
52.6
41.1
42.0
42.4
42.2
86.9
86.6
Primary
TBed
(°F)
2350
2341
2345
2355
2406
2476
2197
2206
2269
2271
2096
2092
Primary
SV
0/hr)
46,900
46,600
46,300
47,000
67,800
67,300
53,000
53,900
50,500
50,200
87,300
88,700
V
(°F)
642
642
637
651
690
693
659
662
654
654
662
656
Primary
TA
(%)
49
49
50
50
69*
69*
54*
54*
54*
54*
63
63
Interstage
Energy
Extraction
(Btu/hr)
25,400
25,600
25,500
26,000
38,300
39,100
30,900
30,900
30,000
29,600
37,800
37 ,800
Secondary
Air
(Ibm/hr)
49.1
49.1
41.6
41.6
49.1
49.1
45.0
45.0
45.0
45.0
72.9
72.9
Secondary
TBed
(°F)
2195
2212
2104
2098
1807
1805
1932
1772
1952
1952
2020
1952
Overall
SV
(1/hr)
40,700
40,600
37,700
37,900
•47,100
47,100
40,500
40,900
41,100
41 ,000
74,900
74,700
Overal 1
TA
(%)
113
113
104
104
134
134
113
113
111
111
116
116
Pressure
(Atm)
1







\








1
1
1
o
U)
        Nith nitrogen dilution

-------
                                    TABLE  C-2.   EMISSIONS DATA - TWO STAGE COMBUSTOR
Test
Pt.
0302-03
0302-04
0302-05
0302-06
0302-07
0302-08
0302-09
0302-10
0302-11
0302-12
0302-13
0302-14
0302-15
0302-16
0302-17
0302-18
0302-19
0302-20
0302-21
0302-22
0302-23
0302-24
CO
(ppmv)
> 2000
> 2000
> 2000
> 2000
> 2000
> 2000
90
74
95
65
120
80
33
29
52
35
40
47
310
165
22
13
NH3 Dopant
Cone.
(ppm fuel )
0
2080
0
2080
0
2080
0
2080
0
2100
0
2110
0
2040
0
2030
0
2140
0
2140
0
2040
ppm NH3
Overall

274

242

237

207

194

180
,
188

183

171

144

176
Thermal
NOX
(ppm)
102

55

28

28

24

19

23

23

24

14

13

Total
NOX
(ppm)

125

90

88

105

72

65

70

76

75

69

55
Fuel
NOX
(ppm)

23

35

60

77

48

46

47

53

51

55

42
Percent
Conversion
NH3

0


















1


















'
0
HCN

18.6

31.8

8.9

0












\












r
0
NOX

8.4

14.5

25.3

37.2

24.7

25.6

25.0

29.0

29.8

38.2

23.9
%
NH3 + HCN
+ NOX

27.0

46.3

34.1

37.2

24.7

25.6

25.0

29.0

29.8

38.2

23.9
o
I

-------
                                                    TABLE  C-2.   Concluded
o
en
Test
Pt.
0308-02
0308-03
0308-04
0308-05
0308-07
0308-08
0308-09
0308-10
0308-11
0308-12
0309-02
0309-03
CO
(ppmv)
230
170
> 2000
> 2000
> 2000
> 2000
1666
700
> 2000
1650
> 2000
> 2000
NH3 Dopant
Cone.
(ppm fuel )
0
4080
0
4020
0
4100
0
4000
0
3920
0
2020
ppm NH3
Overall

343

362

271

322

334
276
Thermal
NOX
(ppm)
8

11

3

4

5

7
Total
NOX
(ppm)

39

39

19

87

40
7
Fuel
NOX
(ppm)

31

28

16

83

35
0
Percent
Conversion
NH3

0

0

7.0

5.6

0
10.8
HCN

0

0

0

0

0
0
NOX

9.0

7.7

5.9

25.8

10.5
0
%
NH3 + HCN
+ NOX

9.0

7.7

12.9

31.4

10.5
10.8

-------
                                            TABLE C-3.   DATA SUMMARY - MODEL  GAS TURBINE
Test
Pt.
TA.
SV
(1/hr)
rti fuel
(lbrn/hr)
rfi air
(Ibm/hr)
Tph
fbed
P
(atm)
CO
(ppm)
NO
(ppm)
UHC Fuel
(ppm)
Acurex Tests
0112-05
0112-06
0112-03
0112-09
Pratt and
1976
1977
1978
1981
1982
1983
I98b
1986
1987
1988
1989
1991
1992
1993
~250


Whitney Tests
312
350
283
397
504
326
' 860
752
829
819
1277
541
583
573
91,500
96,600
87,300
92,400

162,900
165,800
167,100
185,000
428,400
141,300
288,400
291,900
644,000
901,100
276,600
291,900
564,000
699,500
11.3
12.5
11.3
11.9

16.5
15.1
18.6
14.87
27.24
13.75
11.8
13.7
27.5
38.5
7.7
19
34
43
460.3
426.7
441.7
465.3

806.4
823.7
824.4
922.0
2147.0
700.2
I47b
1493
3294
4609
1415
1493
2885
3578
713
724
801
806

703
722
711
723
829
679
894
901
819
681
881
778
754
746
2200
2200
2200
2200

-
-
~
-
-
1/00*
1850*
1850*
1600*
1600*
-
-
1.16
2.06
3.13
3.42

3.06
4.97
4.97
6.77
10.04
3.10
2.99
3.03
5.21
7.01
2.96
3.06
5.07
6.77
0
0
0
0

10
9
10
no
23
9
Il9b
710
1592
2202
1860
82
1285
1362
2
1
3
2

2
1
2
1
1
0
3
5
3
3
44
145
74
68
-
—
-

Propane


0.6
0.3
0
0.3
0
0.6

Propane


/9.b \
34'7 No. 2
23.9 j Oil
222.6 )
high
80.1
24.8
15.6

No. 2 oil
+ pyridine

o
I
             Bed temperature estimates due to bed  nonuniformities.

-------
TABLE C-4.  DATA SUMMARY -ADVANCED GRADED CELL/MODEL GAS TURBINE
Test ~TA SV
Pt. (X) (1/hr)
0412-02 270 42,500
03
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
41,900
42,000
43,400
35,300
34,200
34,600
33,700
41,600
40,700
42,300
41,900
59,600
57,500
62,100
89,000
91,600
0413-02 230 44,900
03
04
05
06
07
08
09
10
46,000
64,000
82,800
98,400
133,100
192,900
200,200
221,700
0427-02 300 47,900
04 220 47,100
05 127 46,300
0428-01 300 27,000
02 250 85,000
03 250 180,800
05 245 162,500
07 140 161,000
m fuel
(Ibm/hr)
4.22
4.15
4.20
4.35
3.48
3.36
3.41
3.31
4.12
4.02
4.20
4.15
5.95
5.71
6.14
8.96
9.20
5.32
5.38
7.54
9.78
11.55
15.8
23.0
23.9
26.5
5.4
7.2
12.3
2.3
9.6
20.6
25.3
39.2
m air
(Ibm/hr)
195.9
192.6
195.0
202.1
161.4
155.9
158.2
153.6
191.1
186.6
194.9
192.7
276.0
265.0
285.0
416.0
427.2
210.3
212.9
298.0
386.8
456.5
623.1
910.1
946.6
1,046.6
242.5
238.3
234.4
116.9
410.7
885.7
827.9
819.9
TPH
730
733
861
864
775
767
765
766
779
779
781
780
790
778
782
787
787
825
680
582
576
573
562
542
542
537
834
839
839
877
745
857
911
923
TBED
2,112
2,115
2,111
1,953
2,105
2,093
2,093
2,080
2,098
2,089
2,079
2,076
2,041
2,030
2,145
2,101
2,084
2,109
2,109
2,089
2,082
2,141
2,130
2,127
2,106
2,106
1,864
2,400
2,500
2,100
2,350
2,350
2,175
2,700
P
(Atm)
1.16
1.16
1.21
1.21
1.22
3.16
3.16
3.14
5.03
5.03
5.02
5.02
6.97
6.96
6.94
8.16
8.12
1.99
3.00
2.97
2.88
2.98
2.82
2.84
2.79
2.79
1.44
1.45
1.46
1.07
3.15
4.64
4.67
4.95
NH,
Addgd
ppm fuel

5,610



5,290
10,420


5,040
11,290
18,960

5,160
18,650

3,820

















NO
1
320
1
—
1
120
228
6
1
60
125
165
5
44
125
1
58
3
0
0
0
0
0
0
0
0
10
15
200
	
—
6
66
263
CO
(ppm)
20
8
556
556
610
62
32
536
383
252
162
151
339
170
16
249
18
560
509
639
714
399
870
1,524
459
79
0
1,800
292
17
18
13
0
0
UHC
(ppm)
6
—
—
—
—
—
—
—
150
8
0
3
69
40
0
16
0
50
2
40
75
17
160
820
730
870
__
—
—
	
—
0
0
0
% NH,
Converted Fuel
Natural
65.0 Gas



61.2
59.0


32.1
29.9
23.5

23.0
18.1

41.0
Natural
Gas







I Diesel
Fuel
1
NG
NG
NG
1 Diesel
1 Fuel

-------
                              TABLE C-5.  RADIATIVE  CATALYST/WATERTUBE  SYSTEM  TEST MATRIX
Test Pt.*
1
2


3
4
5
6
7-10

11
12
13
14
Stoichiometry
(% TA)
40
80


100
200
150
100
100

100

200
100
Surface Temp.
Heat Removal Rate

Determine T
• w




\





Vary water velocity
to optimize Tw
Optimum cooling


\


r
Fuel Rate
kg/hr (Ibm/hr)
2.06 (4.55)








i








r 1








r
4.1 (9.1)
8.2 (18.2)
4.1 (9.1)
2.06 (4.55)
Preheat
temp.
K (OF)
478-533 (400-500)























r '











r
Identify minimum
Lightoff
Characteristics



To be determined -
specific test points
not required








o
CO
       Full emission measurements to be taken at all test points.

-------
                         TABLE  C-6.  TEST  DATA  SUMMARY - RADIATIVE CATALYST/WATERTUBE SYSTEM
Run #
0610-02
0610-03
0610-04
0610-05
0610-06
0610-07
0610-08
0610-09
0610-10
0610-11
0610-12
0610-13
0610-14
0610-15
0610-16
0610-17
T«
40
60
78
88
99
110
120
146
219
100
100;
100
97
99
100
100
SV, hr"1
8900
12600
15800
17500
19200
21400
23300
27800
40700
19500
27500
39800
50100
60900
26500
26300
rf'fuel , lbm/hr
4.7
4.7
4,7
4.7
4.7
4.7
4.7
4.7
4.7
4.6
6.6
9.5
12.4
14.8
6.3
6.3
ITI .jrs lbm/hr
31.3
47.8
62.4
70.6
77.9
88.4
96.9
117.6
173.6
79.6
114.1
166.3
205.9
246.9
109.3
108.1
T °F
V p
557
617
690
722
740
742
741
721
714
742
761
756
816
756
754
675
Xax' °F
1513
1662
1842
1816
1788
1760
1738
1669
1540
1804
1853
1898
1960
2005
1853
1815
o

-------
                                               TABLE C-6.  Concluded
Run #
0613-03
0613-05

0613-07
0613-09
0613-11
TA%
102
101
\
100
99
100
SV, hr"1
33700
32700

35100
35000
33800
"VueT lbm/hr
7.8
7.7

8.4
8.4
8.3
m . , lbm/hr
a 1 r
138.7
135.2

144.8
143.8
139.7
V °F
610
500

400
300
230
T °F
BEDmax'
1824
1742

1689
1700
1667
o
1

-------
                         TABLE  C-7.   EMISSIONS  DATAd - RADIATIVE  CATALYST/WATERTUBE  SYSTEM
o
Run #
0610-02
0610-03
0610-04
0610-05
0610-06
0610-07
0610-08
0610-09
0610-10
0610-11
0610-12
0610-13
0610-14
0610-15
0610-16
0610-17
TA% m, ,, Ibm/hr
fuel
40 4.7
60
78
88
99
110
120
146






f
219 4.7
100 4.6
100 6.6
100 9.5
97 12.4
99 14.8
100 6.3
100 6.3
CO, ppm
>2000
>2000
>2000
>2000
577
34
23
9
0
292
792
>2000
>2000
>2000
850
1004
CH4, Vol %
4.0
-
-
-
2.6
-
5.6
1.2
1.4
3.6
1.5
1.5
3.4
2.9
-
-
                      aNo measurable NO  emissions (>1  ppm)  at any test condition
                                       A

-------
                           TABLE C-8.   EMISSIONS  DATA - RADIATIVE  CATALYST/WATERTUBE GAS

                                        CHROMATOGRAPHY
o
i
INO
Run No.
0610-02
0610-06
0610-08
0610-10
0610-11
0610-12
0610-14
0610-15
Concentrations, Volume
Percent, Dry Basis
H2
4.6
0
0
0
0
0
0
0
°2
15.6
16.5
14.7
19.0
16.9
16.5
15.0
15.7
N2
75.4
81.0
79.7
79.6
79.5
82.0
81.6
81.5
CH4
4.0
2.6
5.6
1.4
3.6
1.5
3.4
2.9

-------
                                 TABLE  C-9.  DATA SUMMARY  - RADIATIVE CATALYST/WATERTUBE SYSTEM
Test
Point
0221-02
0221-04
0223-02
0223-03
0223-04
0223-05
0223-06
0223-07
0223-08
0223-09
0223-10
0223-11
0223-12
0223-13
0224-03
0224-04
0224-05
0224-06
0224-07
0224-08
TA
(*)
59
59
97
97
90
89
35
86
80
80
75
75
120
120
70
70
65
65
52
52
SV
(1/hr)
39,500
29,400
39,400
39,300
38,200
38,100
37,300
37,800
38,300
38,200
38,600
38,600
37,700
37,600
41 ,300
41,100
27,300
27,100
28,500
27,900
'"fuel
(Ibm/hr)
14.9
14.9
9.6
9.6
10.1
10.0
10.3
10.3
11.1
11.0
11.9
11.9
7.6
7.6
13.1
13.4
9.3
9.6
11.5
11.3
'"air
(Ibm/hr)
151.7
151.5
161.0
160.7
154.9
154.3
150.4
152.7
153.3
153.0
153.3
153.4
156.9
156.5
155.8
154.2
101.6
100.1
101.6
99.4
TPH
(°F)
726
646
767
760
751
752
749
748
745
746
744
742
757
758
622
626
624
616
666
660
Fuel3
1
2
1
2
1
2
1
2
1
2
1
2
1
2
3
4
3
4
3
4
Heat Balance (Btu/hr)
Inlet
354,500
351,200
240,300
240,000
249,700
247,500
253,300
253,700
271 ,400
269,200
289 ,000
289,000
194,900
194,900
284,800
290,800
201,100
206,700
245,500
242,000
Tubes
58,810
57,310
37,910
37,410
38,820
38,690
39,680
39,560
44,170
45,050
46,840
49,300
31 ,660
28,530
39,290
39,290
39,480
38,560
46,600
51,260
Test
Sec.
24,610
19,950
18,500
25,420
12,680
12,610
10,810
13,190
16,880
17,160
13,350
14,780
8,410
7,664
10,470
11,920
16,470
18,230
17,250
14,760
Stack
4840
4660
5910
5900
5890
5870
5680
5820
6040
5970
6240
6250
5010
4770
6270
6196
3770
3720
2880
3720
Fuel
Conv.
(«)

17.5

17.7

17.6

17 .5

18.6

18.8

17.1

14.7

20.1

22.7
CO
(ppmv)
>2000
>2000
31
35
33
39
37
42
51
53
100
106
7
9
>2000
>2000
>2000
>2000
>2000
>2000
NOX
(ppmv)
0
9
7
24
19
44
35
50
1.8
22
41
54
0
22
57
62
11
8
8
17
o
I
co
          JFuel Code:     1  -Natural gas
                        2  — Natural gas + ammonia
                        3  - Propane
                        4  — Propane + ammonia

-------
                           TABLE  C-10.   FUEL  NITROGEN  DATA - RADIATIVE  CATALYST/WATERTUBE SYSTEM
I
-ta
Test
Point
0221-02
0221-04
0223-02
0223-03
0223-04
0223-05
0223-06
0223-07
0223-08
0223-09
0223-10
0223-11
0223-12
0223-13
0224-03
0224-04
0224-05
0224-06
0224-07
0224-08
Dopant
Conv.
(ppmv)
0
2000
0
1900
0
2000
0
2000
0
2000
0
2000
0
2000
0
7200
0
6000
0
7000
ppm NH3
in gas

302

191

212

216

229

243

160

401

347

487
Thermal
NOX
(ppm)
0

7

19

35

18

41

0

57

11

8

Total
NOX
(ppm)

9

24

44

50

22

54

22

62

8

17
Fuel
NOX
(ppm)

9

17

25

15

4

13

22

5

0

9
Percent Conversion
NH3

15.3

78.0

93.4

40.3

95.2

61.3

63.1

81.6

83.6

42.1
HCN

5.20

5.24

4.72

4.63

4.37

4.12

6.25

3.49

2.88

2.05
N0x

2.98
--
8.90

11.79

6.94

1.75

5.35

13.75

1.25

0

1.85
%
NH3 + HCN
+ NOX

23.5

92.1

109.9

51.9

101.3

70.8

83.1

86.3

86.5

46.0

-------
                                TECHNICAL REPORT DATA
                         (Please read Instructions on the reverse before completing)
 . REPORT NO.
 EPA-600/7-79-181
        2.
                                   3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
 Design Criteria for Stationary Source Catalytic
   Combustion Systems
                                   5. REPORT DATE
                                    August 1979
                                   6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
J.P.Kesselring, W.V.Krill, H.L.Atkins,
   R. M. Kendall, and J. T. Kelly
                                   8. PERFORMING ORGANIZATION REPORT NO.
                                     78-278
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Acurex/Energy and Environmental Division
485 Clyde Avenue
Mountain View, California 94042
                                   10. PROGRAM ELEMENT NO.
                                   E HE 62 4 A
                                   11. CONTRACT/GRANT NO.
                                   68-02-2116
 12. SPONSORING AGENCY NAME AND ADDRESS
                                                      13. TYPE OF REPORT AND PERIOD COVERED
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC  27711
                                   13. TYPE OF REPORT AND PE
                                   Final; 6/75 - 8/78
                                   14. SPONSORING AGENCY CODE
                                     EPA/600/13
 15.SUPPLEMENTARY NOTES jERL-RTP project officer is G. Blair Martin, Mail Drop 65,
 919/541-2235.
 is. ABSTRACT The repOr|- gives results of an investigation of the applicability of catalytic
 combustion to stationary gas turbine, boiler, and furnace systems, identifying sys-
 tem operating characteristics and potential for NOx emissions reduction. An experi-
 mental program was conducted to develop catalyst materials and combustor concepts
 with useful heat extraction.  Catalyst development included: screening of over  30
 single-cell reactor material combinations, development of a graded-cell reactor con-
 cept, and extensive testing of graded-cell catalysts to determine catalyst perfor-
 mance  as a function of pressure, temperature, heat release rate, and fuel nitrogen
 content. Catalyst development was supported by complete surface  characterization
 analyses and a computer model of catalytic combustion in honeycomb structures.
 Small scale concept testing included a lean-burning model gas turbine,  a radiative
 catalyst/watertube configuration for watertube boiler application,  and a two-stage
 combustor for fuel-NOx control. Tests of these cqncepts focused on techniques for
 energy extraction, obtaining high combustion efficiency,  and minimizing emissions
 of both thermal and fuel NOx. Test results from both catalyst and system concept
 development were used to conceptualize prototype design for gas turbine, boiler,
 and other systems.
17.
                             KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
                                          b.IDENTIFIERS/OPEN ENDED TERMS
                                                c.  COSATI Field/Group
 Pollution
 Combustion
 Catalysis
 Gas Turbines
 Boilers
 Furnaces
Nitrogen Oxides
Pollution Control
Stationary Sources
Catalytic Combustion
13 B
2 IB
07D
13 G
ISA
07B
 . DISTRIBUTION STATEMENT
 Release to Public
                       19. SECURITY CLASS (ThisReport)
                       Unclassified
                         21. NO. OF PAGES
                             478
                       20. SECURITY CLASS (Thispage)
                       Unclassified
                                                22. PRICE
EPA Form 2220-1 (9-73)
                                         F-l

-------