78-1521!
                               Eiwlrewuwtal Pratwtiw Teclmtoiy Serfes
 EPA
6002/2
  76
152B

-------
               RESEARCH REPORTING SERIES

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

    1.   Environmental Health Effects Research
    2.   Environmental Protection Technology
    3.   Ecological Research            j    ;
    4.   Environmental Monitoring        !
    5.   Socioeconomic Environmental Studies   i

This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment and rpethddology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology rejquired for the control and
treatment of pollution sources to meet environmental quality standards.
                    EPA REVIEW NOTICE
                                     j •    ;       - .    -       •:
This report has been reviewed by the JU. 8. Environmental
Protection Agency, and approved for publication.  Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor  does mention of trade
names or commercial products constitute endorsement or
recommendation for use.          '.
This document is available to the public throjjgh the National Technical Informa-
tion Service, Springfield. Virginia 22161.

-------
                                                EPA-600/2-76-152b

                                                June 1976



                  PROCEEDINGS  OF  THE

   STATIONARY SOURCE  COMBUSTION  SYMPOSIUM

VOLUME H--FUELS AND PROCESS RESEARCH AND DEVELOPMENT
                  JoshuaS. Bowen, Chairman
                 Robert E. Hall, Vice-Chairman


           Industrial Environmental Research Laboratory
             Office of Energy, Minerals, and Industry
               Research Triangle Park, NC 27711
                       ROAPNo. 21BCC
                  Program Element No. 1AB014
                         Prepared for

         U.S. ENVIRONMENTAL PROTECTION AGENCY
               Office of Research and Development
                     Washington, DC 20460
                                    . ...-lESTirN AGENCY

-------

I
L

-------
                               PREFACE
     The Stationary Source Combustion Symposium was held on September




24-26, 1975, at the Fairmont Colony Square Hotel in Atlanta, Georgia.




The symposium was sponsored by the Combustion Research Branch of E.P.A.'s




Industrial Environmental Research Laboratory (IERL).  The Combustion




Research Branch has been involved in developing improved combustion




technology for the reduction of air pollutant emissions from stationary




sources, and improving equipment efficiency.








     Dr. Joshua S. Bowen, Chief, Combustion Research Branch, was Symposium




Chairman; Robert E. Hall, Research Mechanical Engineer, Combustion Research




Branch, was Symposium Vice Chairman and Project Officer.  The Welcome




Address was delivered by Dr. John K. Burchard, Director of the Industrial




Environmental Research Laboratory.  Frank Princiotta, Acting Director of




the Energy Processes Division of E.P.A.'s Office of Energy, Minerals,




and Industry, was the Keynote Speaker.









     The Symposium consisted of four Sessions:




     Session I;  Fundamental Research




     Co-chairmen:   Dr. Joshua A. Bowen




                   W. Steven Lanier, Research Mechanical Engineer, E.P.A.,




                       IERL, Combustion Research Branch
                                 iii

-------
     Session II;  Fuels Research andDevelopment




     Chairman:  G. Blair Martin, Chemical Engineer, E.P.A., IERL,




                   Combustion Research Branch








     Session III;  Process Research and Development




     Chairman:  David G. Lachapelle, Research Chemical Engineer,




                  E.P.A., IERL, Combustion Research Branch








     Session IV:  Field Testing and Surveys




     Co-chairmen:  Robert E. Hall




                   John H. Wasser, Research Chemical Engineer, E.P.A.,




                      Combustion Research Branch








     These Session Chairmen have reviewed the transcriptions of the




question and answer sessions, and, in addition, have worked with authors




to clarify and revise presentations, where appropriate, and to make them




clear and meaningful for these printed proceedings.








     We  are grateful  for  the cooperation  of Marjorie Maws,  Project  Leader;




Anita  Lord, Symposium Administrator;  and  Margaret  Kilburn,  Program  Director,




of Arthur D. Little,  Inc., who coordinated the symposium for  E.P.A.
                                  IV

-------
                                  CONTENTS
Preface ....

Welcome Address

Keynote Address
                                                      ill
J.K. Burchard	1-1

F.T. Princiotta	1-3


    SESSION I - FUNDAMENTAL RESEARCH
Formation of Soot and Polycyclic Aromatic Hydrocarbons (PCAH)
in Combustion Systems 	 1-18

     J.D. Bittner, G.P. Prado, J.B. Howard, R.A. Kites

Questions and Answers 	 1-32


Effects of Fuel Sulfur on Nitrogen Oxide Emissions	1-35

     J.O.L. Wendt, J.M. Ekmann

Questions and Answers	 1-88


Two-Dimensional Combustor Modeling  	 1-91

     R.C. Buggeln, H. McDonald

Questions and Answers 	 (n/a)


Effects of Interaction Between Fluid Dynamics and Chemistry on
Pollutant Formation in Combustion 	 1-109

     C.T. Bowman, L.S. Cohen, L.J. Spadaccini, F.K.  Owen

Questions and Answers	1-119


Fate of Coal Nitrogen During Pyrolysis and Oxidation  	 1-125

     J.H. Pohl, A.F. Sarofim

Questions and Answers	1-147

-------
A Detailed Approach to the Chemistry of Methane/Air Combustion:
Critical Survey of Rates and Applications 	
     V.S. Engleman

Questions and Answers
1-153
1-182
Chemical Reactions in the Conversion of Fuel Nitrogen to NOx	1-185

     A.E. Axworthy, G.R. Schneider, V.H. Dayan

Questions and Answers	1-211


Prediction of Premixed Laminar Flat Flame Kinetics,  Including
the Effects of Diffusion	1-217

     R.M. Kendall, J.T. Kelly, W.S. Lanier

Questions and Answers	1-266


Estimation of Rate Constants	1-267

     S.W. Benson, R. Shaw, R.W. WooIfoik

Questions and Answers	(n/a)


Production of Oxides of Nitrogen in Interacting Flames  	  1-291

     C. England

Questions and Answers	1-316


Concurrent Panel Discussions:

1. Combustion Chemistry and Modeling: An Overview

   A.F. Sarofim - Combustion Chemistry and Modeling 	  1-325

   Questions and Answers  	  1-345


   T.J. Tyson - The Mathematical Modeling of Combustion Devices ....  1-347

   Questions and Answers	1-410
                                    vi

-------
2. Federal, Regional, State, and Local Air Pollution
   Regulations: An Overview

   S. Cuffe - Federal	1-413
   G.T. Helms - Regional	1-429
   R.H. Collum - State	1-443
   H.W. Poston - Local	I- 447
               SESSION II - FUELS RESEARCH AND DEVELOPMENT


Assessment of Combustion and Emission Characteristics of
Methanol and Other Alternate Fuels	II-3

     G.B. Martin

Questions and Answers	11-30


Burner Design Criteria for Control of Pollutant Emissions
from Natural Gas Flames	11-31

     D.F. Shoffstall

Questions and Answers 	 (n/a)


Integrated Low Emission Residential Furnace 	 11-81

     L.P. Combs, W.H. Nurick, A.S. Okuda

Questions and Answers	 . 11-101


The Control of Pollutant Emissions from Oil Fired
Package Boilers 	 11-109

     M.P. Heap, T.J. Tyson, E.  Cichanowicz, R.E.  McMillan, F.D. Zoldak

Questions and Answers	11-160


Pilot Scale Investigation of Catalytic Combustion Concepts for
Industrial and Residential Applications 	 11-163

     J.P. Kesselring, R.M.  Kendall,  C.B. Moyer,  G.B. Martin

Questions and Answers	 . 11-196
                                   vii

-------
The Optimization of Burner Design Parameters to Reduce NOx
Formation in Pulverized Coal and Heavy Oil Flames 	 ...... 11-197

     M.P. Heap, T.J. Tyson, G.P. Carver, G.B. Martin, T.M. Lowes

Questions and Answers	11-239


Pilot Scale Investigation of Combustion Modification Techniques
for NOx Control in Industrial and Utility Boilers 	 11-241

     R.E. Brown, C.B. Moyer, H.B. Mason, D.G. Lachapelle

Questions and Answers  	 11-267



             SESSION III - PROCESS RESEARCH AND DEVELOPMENT


Overfire Air as an NOx Control Technique for Tangential
Coal-Fired Boilers	Ill-3

     A.P. Selker

Questions and Answers  	 Ill-26


Control of NOx Formation in Wall Coal-Fired Boilers 	 111-31

     G.A. Hollinden, J.R. Crooks, N.D. Moore, R.L. Zielke,
     C. Gottschalk

Questions and Answers  	 Ill-77


The Effect of Additives in Reducing Particulate Emissions
from Residual Oil Combustion  	 III-S3

     R.D. Giammar, H.H. Krause, A.E. Weller, D.W. Locklin

Questions and Answers  	 Ill-115


System Design for Power Generation from Low Btu Gas Boilers 	 Ill-119

     M.P. Heap, T.J. Tyson, N.D. Brown

Questions and Answers	(n/a)
                                       viii

-------
                  SESSION IV - FIELD TESTING AND SURVEYS


The Effect of Combustion Modification on Pollutants and
Equipment Performance of Power Generation Equipment 	 IV-3

     A.R. Crawford, E.H. Manny, M.W. Gregory, W. Bartok

Questions and Answers 	 . 	 IV-109


Analysis of Gas-, Oil-, and Coal-Fired Utility Boiler
Test Data	IV-115

     O.W. Dykema, R.E. Hall

Questions and Answers 	 IV-161


Influence of Combustion Modifications on Pollutant Emissions
from Industrial Boilers 	 IV-163

     G.A. Cato, L.J. Muzio, R.E. Hall

Questions and Answers 	 IV-219


Emission Characteristics of Small Gas Turbine Engines 	 IV-227

     J.H. Wasser

Questions and Answers 	 IV-252
Systems Evaluation of the Use of Low-Sulfur Western Coal in
Existing Small- and Intermediate-Sized Boilers  	
     K.L. Haloney

Questions and Answers
IV-255
IV-316
A Survey of Emissions Control and Combustion Equipment Data in
Industrial Process Heating	IV-321

     P.A. Ketels, J.D. Nesbitt, D.R. Shoffstall, M.E. Fejer

Questions and Answers	(n/a)
                                      ix

-------
POM and Particulate Emissions from Small Commercial
Stoker-Fired Boilers  	 IV-411

     R.D. Glammar, R.B. Engdahl, R.E. Barrett

Questions and Answers	IV-439



Concluding Remarks - J.S. Bowen  	 IV-441
Appendix
      List of Speakers	A-l
      List of Participants
              Alphabetically by Name	A-2
              Alphabetically by Organization 	 A-6

-------
          SESSION  II




FUELS RESEARCH AND DEVELOPMENT






     G. B. Martin, Chairman
             II-l

-------

-------
    ASSESSMENT OF COMBUSTION AND EMISSION




       CHARACTERISTICS OF METHANOL AND




            OTHER ALTERNATE FUELS
                     by
               G. Blair Martin




         Combustion Research Branch




   Energy Assessment and Control Division




Industrial Environmental Research Laboratory




    U. S. Environmental Protection Agency









   Research Triangle Park, North Carolina
                  II- 3

-------
                             INTRODUCTION









     The Combustion Research Branch  (CRB) of the EPA's Industrial Envir-




onmental Research Laboratory at Research Triangle Park, N. C., has the




responsibility for carrying out a combustion modification R&D program




directed toward control of nitrogen oxides and other pollutants from




stationary combustion sources.  In addition to pollutant emission con-




trol it appears that the technology can also lead to equal or improved




efficiency of energy utilization.  The majority of the R&D is carried




out under EPA contracts initiated and directed by CRB and performed by




private organizations.  However, CRB also maintains an in-house research




program in a number of areas.  This in-house activity provides direct




project officer expertise in combustion research and provides the cap-




ability for initial evaluation of potential control techniques or




promising novel concepts applicable to combustion systems.  The results




of many of these studies are well documented.









     For several years, one in-house project has been devoted to estab-




lishing the combustion and emission characteristics of alternate fuels




(e.g., alcohol fuels, low Btu gas), an area that is now receiving




increased emphasis in the United States.  The goals of this project




have been threefold:  first, to determine the potential advantages or




problems related to the combustion of these fuels; second, to obtain




some understanding of the important factors governing the combustion
                               II-4

-------
and emission characteristics and of potential control techniques; and,




finally, to define necessary R&D for application of these fuels in




practical combustion systems.  The supply of alternate fuels available




for evaluation has been quite limited to date; however, one class,




the alcohol fuels, is quite readily simulated with chemical grade pure




alcohols.  Therefore the combustion and emission characteristics of




methanol and methanol-isopropanol mixtures have received the primary




emphasis to date.









     The purpose of this paper is to summarize the results of the




alcohol fuel testing and to describe briefly the future direction of




the project.









                             ALCOHOL FUELS








     This section provides a short background on sources of alcohol




fuels, a summary of the available information on combustion and




emission characteristics of alcohol fuels as compared to conventional




fuels, and information on a recent series of tests on emulsions of




methanol and distillate oil.









                              Background









     The bulk of the methanol produced in the United States today is
                                 II-5

-------
 synthesized  from natural  gas  for  end use  as a  chemical.  There is obvi-


 ously little benefit  to be  derived  from using  this type of alcohol as a


 fuel;  however,  other  approaches may be attractive.  It has been shown


 that  it  is feasible to produce methanol from a synthesis gas generated


 by coal  gasification.     In  addition, Duhl    has proposed a Methyl


 Fuel,  consisting of about 90% methanol and the balance higher alcohols,


 as a more economical  and efficient  synthesis process for a variety of


 starting materials.   Finally, biological  synthesis from waste materials


may be possible.  The ultimate place of alcohols in the alternate fuel


picture  will  depend on a number of  factors, including:  fuel cost, over-


all energy efficiency relative to other fuels, and environmental con-


siderations of  production and use.  It would appear that alcohols might


be in competition with other  clean  high Btu fuels (e.g., Synthetic


Natural  Gas)  to  supplement or replace dwindling supplies of natural gas


and light fuel  oils in combustion equipment requiring clean fuels.






     Although there is a large body of information on the use of alcohol


fuels in automobiles,  the combustion characteristics of these fuels in


continuous combustion devices of the type used in stationary sources

                                          (2)
have received very little attention.  Duhl    published very limited


data for a small boiler which showed some promising results;  however,


little detail was given.   This paucity of detailed data led to this


study.
                                II-6

-------
                     Summary of Combustion Results








     The studies were directed toward defining the combustion and emis-




sion characteristics of alcohol fuels and understanding the critical




phenomena involved.  Most of the results have been reported in detail



         (3)
elsewhere    and are summarized in this section.  In addition, the ex-




periments were recently extended to examine the use of alcohols as a




supplement to conventional fuels, similar to the methanol-gasoline




blends being considered for automotive use.  The approach taken was




to study emulsions of distillate oil and methanol.









Exp^er imenta 1 _E_quiipmen t








     The experimental equipment was also described in detail in refer-




ence 3, and the key features are summarized here.  The basic combustion




chamber is a refractory lined cylindrical chamber 0.267 meters in diam-




eter and 1.22 meters long.  The nominal heat input was maintained con-




stant at 88,000 watts for all fuels and the combustion air was 115% of




theoretical for all cases.  The burner is a movable block swirl burner


                    (4)

of well known design    and is capable of controlling the ratio of axial




to tangential momentum of the combustion air over a wide range.   Liquid




fuels were atomized through high pressure nozzles and propane was intro-




duced through a radial hole injector.   Flue gas recirculation was simu-




lated by withdrawing cooled product gas from the stack and r-ixing it with
                                II- 7

-------
 the  combustion  air.   Emulsions were prepared with a Total device on loan




 to our  laboratory.









 Experimental Results









     The initial activity in evaluation of an alternate fuel is to




 compare its combustion and emission performance to conventional fuels.




 The baseline fuels were a distillate fuel oil and propane, represent-




 ing both liquid and gaseous conventional fuels.  The alcohols were




tnethanol, isopropanol and a mixture containing 50% of each, thereby




 allowing an assessment of the effect of the presence of higher alcohols




on the methanol combustion.  Figure 1 shows a comparision of NO emis-




sions as a function of the swirl block position (SBP), where the tangen-




 tial momentum of the combustion air increases as the swirl block posi-




 tion increases.  The salient features of this figure are as follows:









          1)   The NO emission for distillate oil shows a distinct




peak at SBP of 4.









          2)   The NO emissions for propane and the alcohol fuels




decrease continuously as the SBP is increased from 2 to 8.









          3)   The alcohol fuels have lower NO emissions than either




of the conventional fuels.
                               II-8

-------
   300
o
UJ
oc

CO
GO
QC
a
a
x
o
o
oc
200
100
                    ODISTILLATE OIL A ISOPROPANOL  •METHANOL
                    DPROPANE      O50% ISOPROPANOL - 50% METHANOL
                                                   8
                           SWIRL BLOCK POSITION
   FIGURE 1. COMPARISON OF BASELINE NITRIC OXIDE EMISSIONS FOR
   VARIOUS FUELS AS A FUNCTION OF SWIRL PARAMETER.
                                II-9

-------
           4)    The  NO  emissions  of  the alcohol fuels decrease as



 the  percentage  of isopropanol decreases.







 It is  interesting that, although the alcohol fuels are liquids, their



 emission  characteristics are more like those of gaseous propane



 than of the liquid  distillate oil.  This is probably related to vapor-



 ization of the  low  boiling alcohols early in the combustion zone as



 indicated by observation of the  flame.  The spray is only visible



 for  about 2 to  5 cm and the flame has a blue to nonluminous appear-



 ance which is quite similar to propane flames.  The low NO for the



 alcohols and the trend as a function of alcohol type may be postulated



 to result from  the  partially oxidized nature of the alcohol having



 the  effect of diluent addition to the flame zone.  Viewed in this way



methanol can be viewed as a CEL'H-O fuel and compared to a CH. „
                              /   Z                           1. o


 representation  of the distillate  oil.  The methanol therefore requires



nearly twice the fuel mass to supply the same heat input to the



 furnace.  For convenience, the term "bound water" will be used for



 the  preoxidized part of the alcohols.  The bound water is 54, 43 and



 30 mass percent for methanol, the 50/50 mixture and isopropanol,



respectively.   Increasing the diluent mass would reduce flame temper-



ature and therefore thermal NO.







     To examine the validity of the postulated effect of bound water,



a series of tests on water-in-oil emulsions was run and the results
                                 11-10

-------
are plotted in Figure 2 for an SBP of 4. The line was drawn for the




water-in-oil emulsions and the alcohol results plotted for comparison.




For any given alcohol fuel the NO emissions compare quite well to




the emulsion results of the same water content.









     To allow comparision of the alcohol results with another dilution




technique, the effect of the flue gas recirculation was examined and




the results are presented in Figure 3.  This shows that a further




reduction of NO emissions is possible for all of the fuels with about




85% reduction possible at 25% FGR.








     The effect of any diluent on thermal NO is based on a reduction




of peak flame temperatures and, thereby, a reduction in the rates of




formation.  To allow a comparison of all the techniques on a common




basis, the NO emissions are plotted versus theoretical flame tempera-




ture as shown in Figure 4.  Note that the temperature is calculated




based on fully equilibrated products at the overall stoichiometry of




the system for adiabatic conditions.  Since the combustion system




is hot wall (about 1400 C) and designed for minimum heat loss, the




assumption of adiabatic conditions may be reasonably good.  The




assumption that the reaction occurs at the overall stoichiometry is




more open to question for the diffusion flame system and may aid in




explaining the differences seen in the figure.   The most important




factor may be the differences in fuel vaporization and atomization
                              11-11

-------
v»
O
X
O
O
ac
                                      I          I
                                      O DISTIL LATE OIL-WATER EMULSIONS
                                      O 50% ISOPROPANOL • 50% METHANOL
                                      AISOPROPANOL
                                      OMETHANOL
                                      • ISOPROPANOL AND WATER
                           20          30          40

                            FUEL WATER CONTENT, MASS %

          FIGURE 2. COMPARISON OF NITRIC OXIDE EMISSIONS FOR VARIOUS
          FUELS AS A FUNCTION OF WATER CONTENT.
                                   II-12

-------
                                   O DISTILLATE OIL
                                   D PROPANE
                                   A ISOPROPANOL
                                   O METHANOL
            0.05
  0.10         0.15
FRACTION RECIRCULATED.f
0.20
0.25
FIGURE 3. EFFECT OF FLUE GAS RECIRCULATION ON NITRIC OXIDE
EMISSIONS FOR VARIOUS FUELS.
                           11-13

-------
   300
Q
ut

g 200
<
LU
5
CO
<
 *

DC
Q
a.
a.
Q
X
o 100
o
DC
                         • DISTILLATE OIL WITH FGR
                         • DISTILLATE OIL EMULSIONS
                         A ISOPROPANOL
                         O 50% ISOPROPANOL - 50% METHANOL
                         OMETHANOL
                         A ISOPROPANOL - WATER
    1.900
1.800
1.700
1.600
1.500
1.400
              THEORETICAL FLAME TEMPERATURE, 10 °C
 FIGURE 4. COMPARISON OF NITRIC OXIDE EMISSION REDUCTION
 AS A FUNCTION OF THEORETICAL FLAME TEMPERATURE FOR VARIOUS
 DILUENT ADDITION TECHNIQUES.
                            II-14

-------
characteristics coupled with the fuel-air mixing characteristics of




the system.  An explanation may be summarized as follows:









          1)   For the water-in-oil emulsion, the water droplets are




believed to shatter causing a secondary atoraization of the oil.  It




is also possible that at the higher water levels the emulsions may




tend to separate.  In either case the oil may burn in an environment




where the diluent effect of the water is less than that calculated




and the actual flame temperature is higher.








          2)   For the flue gas recirculation the diluent is uniformly




mixed with the combustion air and should have the same effect on the




flame temperature no matter where the fuel burns; however, the fuel




reactions still may occur at conditions closer to stoichiometric than




the overall mixture which would have the effect of increasing peak




temperatures locally.









          3)   For the alcohols, not only is the "diluent" intimately




bound in the fuel molecule, but the early vaporization may allow more




rapid mixing with the combustion air.  This would most closely approxi-




mate the basis on which the flame temperature calculations were made.









          4)   Finally the alcohol fuels have high latent heats of




vaporization which may result in a significant local heat extraction
                              II- IS

-------
and an associated NO reduction.









     The final series of experiments examined the effects of alcohol




and distillate oil emulsions on NO emissions.  The total heat input to




the furnace was kept at a constant value and the percent of heat input




in the form of methanol was increased from zero to 100%.  The average




results of three series of tests are shown in Figure 5.  The first 25%




methanol reduced the NO emissions by about 50% relative to distillate




oil alone.  To allow a comparison on a common basis the "bound water"




in the methanol was calculated as a water content for the total mix-




ture and plotted along with the water in oil emulsion in Figure 6.




For a given water content in the fuel mass the methanol-oil emulsions




give consistently lower NO emission than do the water-oil emulsions.




This may be attributable to a number of factors including:









          1)   The methanol has a lower boiling point than does water




     and may provide a more efficient secondary atomization effect.




     This would promote better mixing and a greater effect of the




     diluent on the combustion temperature.









          2)   Not only does the methanol have a higher heat of




     vaporization per mass than the water, but also there is a




     larger mass to be vaporized for a given "water" content, re-




     sulting in a more significant effect on peak flame temperature.
                              II- 16

-------
I
                                       I           I
                                        O DISTILLATE OIL
                                        O DISTILLATE OIL • METHANOL EMULSIONS
                                        A METHANOL
                                       25          50          75

                                     METHANOL CONTENT, % INPUT BTU
                          FIGURE 5. NITRIC OXIDE EMISSIONS FOR METHANOL
                          DISTILLATE OIL EMULSIONS AS A FUNCTION OF
                          METHANOL HEAT INPUT.


                                               11-17

-------
                                      O DISTILLATE OIL-WATER EMULSIONS
                                      D DISTILLATE OIL -METHANOL EMULSIONS
                                      A METHANOL
                         20         30          40
                          FUEL WATER CONTENT, MASS %
FIGURE 6 COMPARISON OF NITRIC OXIDE EMISSIONS FOR WATER AND METHANOL
WITH DISTILLATE OIL AS A FUNCTION OF WATER CONTENT.
                                   II-18

-------
     It should be stated that, although the explanations offered here




(based on temperature effects on thermal NO) appear to explain most of




the observed phenomena, some anomalies still exist in the methanol




data.  At this time it is not possible to assess the effect of the




methanol on flame chemistry.  It is possible that the methanol reaction




scheme may reduce concentrations of certain radical species and thereby




reduce non-Zeldovitch NO.  This possibility will be explored in the




laboratory by combustion of a mixture of 33% CO and 76% H2 to simulate




a decomposed methanol.
                           H-19

-------
                              Conclusions









     Based on the results presented here and other Information contained




in reference 3, the following conclusions can be drawn:









          1)   Alcohol fuels produce lower emissions of  nitrogen oxides




     than distillate oil or propane.









          2)   The NO emissions of alcohol fuels increase as the per-




     centage of higher alcohols increases.









          3)   The low NO emissions for alcohol fuels appear to function




     from the presence and level of oxygen in the fuel molecule,  which




     can be  viewed as a diluent carried in the fuel.   The operative




     mechanism appears to be related to thermal effects  of the fuel,




     latent  heat of vaporization,  and/or decreased flame temperature;




     however,  chemical effects  cannot  be totally ruled out.








          4)    In the hot wall  experimental  system,  NO levels  similar




     to  those  for methanol (e.g.,  65 ppm)  could be attained  for distil-




     late  oil  and propane with  flue gas recirculation and  could be ap-




     proached  at 50% water for  distillate  oil  emulsions.









          5)    Based on the  calculations,  the  use  of  methanol  and a
                               .11-20

-------
     water-oil  emulsion  appears  to  impose  a  6  to  7%  increase  in  stack




     heat  loss  compared  to  distillate  oil  with flue  gas  recirculation;




     however, it  is  anticipated  that there will be some  compensating




     factors that  can be designed into practical  systems to minimize  the




     losses, based on lower excess  air operation  as  seen in the  package




     boiler simulator.









           6)    The emissions of  CO, UHC, and particulate from alcohol




     fuels were generally the same  as, or  less than, those for the con-




     ventional  clean fuels  tested (propane and distillate oil).  Alde-




     hydes do not  appear to be a significant problem based on limited





     data generated under EPA contract 68-02-1498 with Ultrasystems,




     lnc.<3>









          7)    From a technical standpoint, methanol appears to be a




     satisfactory  fuel for  stationary  combustion systems; however, the




     final commercial feasibility will probably depend more on cost




     and availability than  on the ability  to burn the fuel.









                            FUTURE PROGRAM








     The future work in  this area will be on other alternate fuels




that appear to have some potential for application to stationary com-




bustion sources.  The fuels of interest and current plans of research
                             11-21

-------
direction are summarized below.




                         Background




     In addition to the alcohol fuels there are a number of other fuels


that may be derived from coal, oil shale or refuse at some time in the


future.  These possibilities cover the range of gaseous, liquid and


solid fuels, with a number of candidates for each of these types.




Gaseous Fuels




     The gaseous fuels of interest can be divided into the broad cate-


gories of Synthetic Natural Gas (SNG) and producer gas.  The SNG is a


high Btu fuel to replace or supplement natural gas in pipelines and


the properties of the two types of SNG are compared to natural gas in


Table 1.  Unless some information develops to indicate a large diver-


gence of combustion and/or emission characteristics of SNG from those


of natural gas, the current technology for methane combustion should


be adequate for emission control.  The properties of low Btu gas


vary widely depending on both the feed material and on the process

                                                         3
used; the energy content can range from 4403 to 8806 KJ/M  as shown


in Table 1.  The combustion characteristics of these fuels are


just beginning to be studied; however, several aspects already have


come to light.  First, due to the nature of the fuel, the adiabatic
                              11-22

-------











CO
1-1
e

&
8
%
o
co
s
M
1
CM
0
2
O
CO
1
o

•
.H

iJ
H















•O
ki
0)
C
I-l
iN!
H
^ 	
0
C >i
O X
•H X
*J CJ
•H
01
0
6*
O
O CM
35
<3


o
U

ocNiio ^o vO^ooin
ro ^ co co





^•^ '^^
v2 s5
co to to co 2^
fl tO W cb ^s ^-^ /^\ x*s
OOO O,D Hui.O
cd cd ntc cij bo j>^ cd ^]
M l-< W-' t-i h C 00 CJ
333^3 3 -H >, 1
U 4J 4-1 (1) 4J (-J&3CC3
2KK4JZ W W 0) CO
g"gOHO^ 0(Sc3o
« «o 4-1 n ye 3333

?d ^ 4J QJ 4J tiO
x ai c a c u 3 ^ 3 3

H U C0%4 (AhJ >J i-] hJ »4









•
C
o
J 1
•H
S
• ^
en
-
0
t t
•o
1
14
S-i
<1>
CM
.
S3
X
0
00
(3
•H
M

0)
C
•H
a
iH
cd
a
°s
S
A
o
J^
0
o
s
jg




0}
ON
•
^
I
O
iH
4J
a.
(U
CO
»
>.
CO
o
Q
c
CJ
(U

CA
0
M-l

(U
3
4-1
•H
4J
CO
C
M
^
M
V
ex
CM
1
•H
•0

a.
P^
CO
rt
0
u

B
H
iw
at
rT
C
s
p




^Q
























•a
vO cd
S3
CM O
a «M
T3 *
ed CM
O
•j- o

CJ <4.J
fft
*J 4-1
41 -H
CO CO

-------
flame temperature is lower than conventional fuels and thermal NO prom-




ises to be low.  However, if the high temperature processes for H_S




removal are developed and the fuel gas is presented to the burner at 550




to 850°C, significant thermal NO may result.  Second, the presence of




significant amounts of ammonia appears to exist in the product gas at




the gasifier exit for many processes.  Unless this bound nitrogen is




removed in the gas clean up sequence prior to combustion, significant




fuel NO can result.  Robson    states that 600 ppra of NH^ can be present




in the fuel gas following some low temperature l^S removal processes,




and projects 3800 pptn for some of the hot H2S removal processes under




development.  This then identifies high temperature fuel gas and chem-




ically bound nitrogen as two potential sources of NO that must be con-




sidered in low Btu gas combustion technology development.








Liquid Fuels








     The liquid fuels of interest can be derived from either coal liq-




uefaction or shale processing and some potential products are shown in




Table 2.  It can be noted that, while the sulfur content is relatively




low (0.2 to 0.76%), the fuel nitrogen is quite high (0.68 to 1.77%) as




compared to a residual oil produced from petroleum stock.  While the




hydrotreated COED oil shows that the heteroatom content can be reduced,




this represents another processing step and extra cost for the finished




fuel.  If these fuels can be burned in the crude form and the emissions
                               II- 24

-------























CO

w
5
Q
rH
:=>

M
J

S3
0
M
P£

£>

fc


O



8««
4J
00
•H
Ol tO
13

CO
•H
CO
£>s,
1— 1
cC
C
<£

CU 2
4-1
CD
g
•H
4-1
rH
E>

ar










CJ
m co
O o rH en

o O r^ d




VO
ON ON in rH
O CO O") O
• • • •
O O O O







ro
CM CM rH
O CN rH O
• • » «
0 0 rH O



t— I U"> sf ^O
» » * *
f*"} CM 00 rH
rH I— I rH





CF» 00 O O
» » * *
\jD \^ ro 00
CO 00 00 00

^D -^- O
1 t • 1
rH 1 1 ON 1 O
in




rH ON CO ^Q
CO rH ^
ao
0
rH
O
C
,G
O
0!
H

CO
0
Q-t
o

CU
4-1
3
4-J
•H
4J
CQ
C
M


;
2
O
en
I
O
ffl
           (LI
           3
           C
           •H
              q
           
p«^
II O 1
ii - i
O^i



s~^
.a

J2 T3 H
O X> 4-1 O rH
^ o jr IT)
Q Q VJ 4-1 O
w w T: c u
o o x >^ i
C_J C_) S3 C/3 ffi

|
|



1
1


>J
3
H-l
i— 1
3
V3

U
O
j_J

rH
TO
O
CJ
1
r-»
•
O
00
H
CO

M
r-
en



^— ^
o
x_^
rH

b

rH
CO
O
O
CO
5
vD
*
»^|-
00

ON
^3-
•«
CO

-------
of NO   and  smoke  effectively controlled by combustion technology, sign-
     A


ificant cost  and/or overall energy efficiency savings may be realized.



The  thrust  of the in-house research will be exploring the possible



approaches  to minimizing  the emissions.






                            Planned Program






     The near term emphasis in the in-house research project will be low



Btu gases;  liquid fuels will be included as suitable candidates



become  available.






GasGenerator






     A  catalytic partial oxidation reactor developed by the Jet Pro-



pulsion Laboratory for automotive use is being adapted as a gas genera-



tor to  supply a variety of simulated low Btu gases.  The reactor will



accept  a variety of hydrocarbons and reform these to yield a product



fuel gas closely approximating the predicted equilibrium composition



based on the  input fuel to air stoichiometry.  The reactor can be operated



with about  40  to  70% of theoretical air.  By changing the fuel type



from methane  to benzene and the input stoichiometry, the CO to H? ratio



and the energy content of the product gas can be controlled over wide



ranges.  At design condition the reactor will produce approximately



88,000 watts  of fuel gas at a temperature of 650 to 720°C.  The product
                                 II- 26

-------
gas temperature will be  controlled down  to  120 C by additional heat


exchange.




Research Areas





     The program will address the two areas previously identified as


requiring attention, fuel gas temperature and bound nitrogen, for


several gas compositions.  The gas compositions will strongly depend


on the capability of the gas generator as delivered; however, the varia-


bles of interest are:  1) CO to H,, volume ratio in the range of 0.5 to 2;

                                                3
and 2) energy content between 2980 and 7460 KJ/M .  The composition of the


gas will also be altered by the addition of methane to simulate some


of the processes yielding significant quantities of hydrocarbons.  In


addition the effect of chemically bound nitrogen will be studied by


ammonia addition in the  range of 100 to 4000 ppm.  The gas temperature


will be controlled by a variable heat exchanger and will range from


120 to 720 C in three or four increments.





     In addition to establishing the emission characteristics of the


various gases as a function of composition and temperature, potential


control techniques will be explored for gases that yield high levels


of NO.  The approaches to be considered are the classical techniques


of burner design and/or staged combustion.   The burner design variables


will allow changes in fuel and air mixing rates and should be applicable
                             i r- 27

-------
 to  either  thermal or  fuel NO.  The staged combustion techniques will be
 primarily  directed toward control of fuel NO from ammonia bearing fuels,
 Based on fundamental  and small scale data, there is reason to believe
 that operating with a burner input of 60 to 70% of theoretical air and
 second stage air injection to complete fuel burn out will yield fuel
 NO  levels  below 100 ppm even with high ammonia levels in the input
 fuel.


     The results of this study will complement and reinforce related
 low Btu gas combustion experiments being carried out under contract
 to  EPA by  IGT and Ultrasystems.


                                SUMMARY


     This  in-house research project has examined the combustion of
 alcohol fuels and identified potential advantages for NO  control.
                                                        X
 The next stage is to explore emission and combustion characteristics
 of  low Btu gases and to study applicable NO  control techniques.
                                           A
The alternate liquid fuels have received only limited considerations
because of limited availability;  however, as these fuels are secured
 they will be factored into the effort.
                               11-28

-------
                              REFERENCES








1.   Hoogendoorn, Jan C., "Experience with Fischer-Tropsch Synthesis



     at Sasol."  Symposium Papers - Clean Fuels from Coal, pp 353-336,




     Institute of Gas Technology, Chicago, Illinois (September 1973).








2.   Duhl, R. W. and T. 0. Wentworth, "Methyl Fuel from Remote Gas




     Sources."  Presented at llth Annual Meeting of the Southern




     California Section AIChE, Los Angeles, California, April 1974.








3.   Martin, G. B. and M. P. Heap, "Evaluation of NOV Emission
                                                    X



     Characteristics of Alcohol Fuels for Use in Stationary Combus-




     tion Systems."  Presented at 80th National AIChE Meeting, Boston,




     Mass., September 9, 1975.
4.   Heap, M. P. ,  T. M. Lowes, R. Walmsley, and H. Bartelds, "Burner




     Design Principles for Minimum NO  Emission."  Proceedings,
                                     X


     Coal Combustion Seminar, EPA-650/2-73-021, pp 141-172; NTIS No.




     PB 224-210/AS (September 1973).
5.   Robson, F. L. and A. J. Giramonti, "The Environmental Impact of



     Coal-based Advanced Power Systems."  Symposium Proceedings



     Environmental Aspectsof Fuel Conversion Technology, EPA-650/2-



     74-118, pp 237-257; NTIS No. PB 238-304/AS (October 1974).
                            11-29

-------
    8:30 a.m.
    Assessment of Combustion and
      Emission Characteristics of
      Methanol and Other Alternate
      Fuels
    G. Blair Martin, U.S. E.P.A.
Qs  Can you comment on the quality of the raethanol
    distillate oil emulsions as characterized by dis-
    persed phase particle size, and also, do you have
    any idea of the residence time of the emulsions
    between the point of preparation and the point of
    injection into your combustion?

A:  We did not look at the particle size.  The one
    thing that we did for both the oil and methanol
    emulsions was to spray them at a container and to
    watch the separation characteristics.  At high water
    content, the emulsions tended to separate relatively
    quickly.  The residence time between the emulsifier
    and the burner is fairly short in that the emulsion
    is formed in a device immediately before the fuel
    tube to the burner.  The fuel tube is approximately
    18 inches long from that point to the nozzle and is
    constructed with a  1/8 inch I.D., so the amount of
    fuel that's actually in the fuel tube at any one
    time is relatively small.
                              11-30

-------
     BURNER DESIGN CRITERIA FOR CONTROL OF
POLLUTANT EMISSIONS FROM NATURAL GAS FLAMES
                         by

                   D. R. Shoffstall
                  Paper Presented at

 SYMPOSIUM ON STATIONARY SOURCE COMBUSTION
                     Sponsored by

        COMBUSTION RESEARCH SECTION OF
    U.S. ENVIRONMENTAL PROTECTION AGENCY

               September 24-26, 1975

                   Atlanta, Georgia
                       II-31

-------
               BURNER DESIGN CRITERIA FOR CONTROL
                       OF  POLLUTANT EMISSIONS
                      FROM NATURAL GAS FLAMES
                                   by
                             D, R. Shoffstall
                        Institute of Gas Technology
                         Chicago, Illinois  60616
                               ABSTRACT
   This program has developed a number of simple burner modifications which
resulted in significant reductions in NO emission levels.  These modifications
can be utilized on existing equipment or can be incorporated in the design of
new combustion systems.  Because of the influence of furnace geometry and
multiple flame interactions  on NO emission levels, these modifications may
not result in the size reduction in emission levels observed during the test
program. However, applying the principles of aerodynamic NO control
developed by this program,  coupled with the use of suitable monitoring
equipment,  reductions in NO emission levels up to 60% are possible.
                                    11-32

-------
                BURNER DESIGN CRITERIA FOR CONTROL,
                       OF POLLUTANT EMISSIONS
                      FROM NATURAL GAS FLAMES
                                     by
                              D. R. Shoffstall
                         Institute of Gas Technology
                          Chicago, Illinois  60616
    For 3 years, IGT has conducted a program for the Environmental Pro-
tection Agency on NO  (nitrogen oxides:  NO,  nitric oxide; and NO2, nitrogen
                    X
dioxide) emissions control from natural gas combustion on a pilot-scale test
furnace.  The objective of the program is twofold:  1} to determine whether
combustion aerodynamics can be used  to control NO  emissions from a
                                                  Ji
natural gas flame and 2) if aerodynamic  control is successful, to determine
the lower emission limits that can be achieved by using this technique.  The
resulting  data could be used by industry and/or burner manufacturers to aid in
reducing NO  emissions from existing  combustion processes  and  to help in
            X-
the development of new low-NO  -emission burners.
                              X
    The effect of combustion aerodynamics on NO  emission levels was tested
on a number of  industrial burner types including a kiln,  a flat flame,  a baffle,
a movable-block swirl, and a utility boiler.  This paper deals primarily
with the baffle burner because not only has the investigation resulted in
recommended low-NO -emission operating conditions but also the resulting
data have been used by a  group of gas utility  companies in the development
of a burner with a 60% reduction in NO  emissions.
                                     Jv
RESEARCH INSTALLATION
    Most of the  experimental work described in this paper was conducted on
the IGT rectangular test furnace, which has a 25-sq-ft cross-sectional area
and a length of  15 feet.   This furnace can be  end-  or sidewall-fired at rates
up to 4 million Btu/hr.   The furnace is equipped for in-the-flame sampling
(Figure 1).  The overall furnace system is flexible enough that the following
parameters can be varied independently:
•   Heat input,  up to 4 million Btu/hr
•   Air input, up to  40^  excess
                                 II-33

-------
10/75
                                                                                               o
                                                                                               03
                                                                                               a
                                                                                               f-i
                                                                                               n>
                                                                                               &0

                                                                                               9
                                                                                              -4J
                                                                                               O
                                                                                              0)
                                                                                              IH
                                                                                              0
                                                                                              60
                                                                                              •H

                                                                                              1=4
                                          1 1. - 34

-------
•   Heat losses to the  furnace wall, by changing flow in water-cooling tubes
    cast into the refractory furnace sidewalls
•   Combustion air temperature, up to 1000 F
•   External flue-gas recirculation (FGR),  up to 35% of combustion air
•   Furnace pressure, up to 4-0.05 in H2O.
    The following analytical instrumentation was used for concentration
measurements of chemical species during the program (Figure 2):
•   Beckman 742 Polarographic Oxygen (O2)
•   Beckman Paramagnetic  Oxygen (O2)
•   Beckman NDIR Methane (CH4)
•   Beckman NDJR Carbon Monoxide (CO)
*   Beckman NDIR Carbon Dioxide (COZ)
•   Varion 1200  Flame lonization Chromatograph (Hydrocarbon Cl-C9)
•   Beckman NDIR Nitric Oxide (NO)
•   Beckman NDUV Nitrogen Dioxide (NOE)
*   Hewlett-Packard Thermoconductivity Chromatograph Hydrogen (H2),
    Nitrogen (N2),  Argon  (Ar),  O2,  CO,  COZ, and hydrocarbons C] to Cs
•   Beckman Cherniluminescent NO-NO2.
GENERAL. FLAME CHARACTERISTICS
    A general aerodynamic characterization of a flame can be made by deter-
mining the different types of flow patterns that exist within a combustion
chamber.  A detailed flow analysis of a confined flame reveals that the front
section of a combustion chamber can be divided into four zones;  primary
jet, primary recirculation,  secondary jet,  and  secondary recirculation.  The
primary and secondary jets contain only particles with a forward flow direc-
tion (away from  the burner), whereas  the recirculation zones contain gas
particles moving in the reverse flow direction (back toward the burner).  The
size,  shape, and particle density of the  recirculation zones are determined
by the velocity,  the ratio  of gas to  air, the spin intensity of the secondary jet,
the burner-block angle, and, for the secondary recirculation zone, the size
and shape of the combustion chamber.  Figure 3 shows the three types of flow
patterns that we observed during our pollution control trials-

                                II- 35

-------
10/75
                    Figure 2.  Control room facility
                                 II- 36

-------
                                          TYPE  I


                                       NO SWIRL
                                       MO PRIMARY
                                       RE CIRCULATION
                                     TYPE.  IT


                                 LOW SWIRL  INTENSITY
                                 PRIMARY JET  VELOCITY >
                                 SECONDARY JET VELOCITY
                                HIGH &WIRL  INTENSITY
                                SKONWRY jet VELOCITY
                                PRIMARY JET VELOCITY
                                        A -1.M-21 66
Figure 3.  Flames types tested
                  11-37

-------
    A type I flow pattern is observed when the secondary jet has no tangential
velocity component (no spin) and the primary jet velocity is much larger than
the secondary jet velocity.  Depending on the initial jet exit velocity from the
fuel injector,  the flame can be either attached to or detached from the injec-
tor. A type I flow pattern also can be generated with a secondary jet  with a
tangential velocity component (spin) if the burner-block divergent angle is
equal to or less than the angle of the secondary jet relative to the burner axis.
As a result,  the  burner block restricts the expansion of the secondary jet and
inhibits the formation of a primary re circulation zone.
    A type II flow pattern is generated when the  secondary jet has a tangential
velocity component large enough to cause the particles  to adhere to and pack
tightly against the burner block.  This packing creates  a low or negative
pressure region  in the center of the burner block.  The pressure differential
between the furnace and the central region of the burner block causes gas
molecules to be pulled into this region and back toward the burner, thus creat-
ing the primary recirculation zone.  When the velocity  of the primary jet is
greater than the velocity of the recirculating gases in the primary recircula-
tion zone,  the primary jet penetrates this reverse flow region and a recircu-
lation lobe occurs on each side of the burner axis.   This flow pattern  is
labeled type II.
    If the central region of the burner block has a large enough negative
pressure differential relative to the furnace, such that  the velocity of the
recirculating gases is greater than the velocity  of the primary jet, pene-
tration of the primary recirculation zone is  not  possible and a type III flow
pattern results.  A type III flow pattern also can be  generated mechanically
by introducing the fuel with only a radial velocity relative to the axis of the
burner.
DATA ANALYSIS AND DISCUSSION
Baffle Burner
    Figure 4 illustrates the design of the baffle burner.  The gas nozzle lies
parallel to and along the axis of the burner.  It is inserted into the ceramic
baffle, thus ensuring that the gas enters parallel to  the axis of the baffle
burner.  Combustion air enters perpendicular to the axis and passes through
the six ports in the baffle, which impart a swirl to the air in some designs.
                                      H-38

-------
10/75
      r
    o

    4 O. D.


                                \ Baffle
                                                    I 8"
 A	t\r
  V        'V
                                 U—J?—
                                                   -
                                                   o a.
4r
                                                                      ,  O
                                                                      :  CD

Long Rome
Short Home
infer. Flame
'A'
•fc"
2"
2"
1B'
nV'
14^*
i«i"
Baffle
Straight Thru Holes
25° Spin
15 Spin
             Figure 4.  Assembly drawing of baffle burner
                                H-39

-------
 Two of these baffle ports are shown in Figure 4 with their axes parallel to the
 axis of the burner.  Air exiting from the ports (as illustrated) would have an
 axial velocity component only, resulting in a "long" flame.  To shorten the
 flame length, a tangential-flow component must be added to the combustion air
 velocity. This is accomplished by using a baffle where the combustion-air
 ports are rotated relative to the axis of the burner.  For the  "intermediate"
 flame length baffle (IFLB), the rotation orientation of the ports is 15 deg,
 and for the "short"  flame length (SFLB),  25 deg.
    Standard Conditions
    Figure 5 presents normalized nitrogen oxide (NO) versus excess oxygen
 (O2) test data for burner conditions normal to an industrial situation (stand-
 ard gas nozzle, baffle nozzle position, and 4 deg burner-block angle).  Photo-
 graphs illustrating flame geometry and flame luminosity with these industrial
 conditions are shown in Figures 6 and 7.
    The standard gas nozzle is a 2-inch pipe that rests within a centering tube
 mounted between the baffle and the  burner housing.  Gas nozzle positions are
 illustrated in Figure 8.  The gas nozzle position is denoted as the baffle  posi-
 tion when the nozzle is even with the burner-block side of the baffle.  Other
 nozzle positions investigated were the throat position (gas nozzle midway
 between the baffle and the front wall of the furnace) and the exit position (gas
 nozzle even with the front wall of the furnace).  For the IFLB, the exit gas
 nozzle position produced unstable combustion and therefore is not included
 in the test results.
    At a gas input of 3000 SCF/hr, the relationship between NO and excess O2
 became increasingly nonlinear as a function of temperature; see Figure 5.
 This indicates that in the range of excess O2 we investigated,  the higher the
 preheat, the  more closely the NO-versus-O2 relationship characterizes that
 of a premix flame.
    For the second set of curves in Figure 5, burner conditions were the
 same, but the gas input was reduced to 2000 SCF/hr.  In an attempt to main-
tain the same level of bulk NO formation, the wall temperature was held at
 1420 C,  compared with 1435°C for  a gas input of 3000 SCF/hr.  These curves
 show the same linear relationship between NO and excess O2 that was observed
for the kiln burner.   This linear relationship is characteristic of a diffusion
flame.
                                  11-40

-------
   600-1
              Cos  input   3070  8  2O05  SCFH
              Oos  Nozzle  Bottle  Position
              Won   Temperature  1435* c
              4* Burner Slock  Angle     ,_  	
   500-
  400-
E
a.
a.

O
Z
   300-
IM
"5
E
b
z
  200-
   100-
o
1 6,000

1
t
	 Legend •
A
O
1 I
2 3
Standard
Standard
1
4
Nozzle,
Nozzle.
I
5
Gas
Gas


Input
Input
1
6
3000
1000


CFH
CFH


                           IN  FLUE,  96
           Figure 5.  Normalized NO concentration as a
              function of flue O2 for the IF LB burner
                 with a standard gas nozzle  at gas
                  inputs of 3070 and 2005 SCF/hr
                              11-41

-------
10/75
                       1.  Divergent fuel nozzle
                       2.  Standard fuel nozzle
            Figure 6.   Flame photographs of intermediate
                      flame length baffle burner
                                 11-42

-------
10/75
                       3.  Combination fuel nozzle
                          4.  Radial fuel nozzle
              Figure  7.  Flame photographs of intermediate
                       flame length baffle burner
                                    T T- 43

-------
10/75
       Figure 8,  Nozzle positions tested for the baffle burner
                                11-44

-------
    External FGR
    As a base-line control case,  the effect of external FGR was tested.  The
percentage of FGR is determined by using the relationship —
                               FGR (CF/hr)
       % FGR =
Secondary  Air (CF/hr) -f-
                                         X 100
                                                 (CF/hr)
    By using the data presented in Figure 9, the NO concentration measured
for standard burner operating conditions with 46 2 C  secondary air preheat
can be compared with the concentration at similar burner conditions but with
the addition of 15% and 30%  FGR to the secondary air.  During preheat, both
the flue gas and secondary air were blended and reached the same final pre-
heat temperature, which in this particular test was 460 C.  At 3% excess O2,
a 15%  FGR level reducjed the normalized NO from 575 to 160 ppm, whereas
30% FGR reduced the measured NO concentration to 50 ppm.
    Wall Temperature
    The data for Figure 10 were collected with the  burner operating in the
normal industrial mode; however, the wall  temperature was decreased from
1435  to 965 C by using water-cooling  tubes within the furnace sidewalls.
Although the NO was  depressed by approximately a factor of 2, it is  difficult
to say  what part, if any, was due to a reduction in the bulk NO formation.
Because the secondary-recirculation-zone flue products were  at a much lower
temperature when they  became  entrained by the secondary combustion air
jet, a lower flame temperature  resulted and, consequently, a  lower  concen-
tration of NO.
    Nozzle Type and Position
    Figure II presents  NO concentrations measured under standard burner
conditions but with a  radial gas  injector.  This  injector forces the gas to
enter the burner block radially to the axis of the burner, which causes a
flame flow pattern categorized as type III.  The combination gas nozzle
shown  in Figure 12,  operating with radial injection only, was used.
    The radial nozzle in the baffle position produced more NO  tliar. the stand-
ard nozzle because of the increased rate of heat release, coupled with the
inability of the burner block to be used  as a heat sink.  However,  when the
radial  nozzle was moved into the throat position, there w-rs a dramatic drop
                                    11-45

-------
   600
   50OH
£
d
a.
   3OO
4)
O
   20CT1
            O
          I.7OO
Gas  input   3070  SCFH
Gas  Nozzle   Baffle  Position
Wall  Temperoturt   I39O  C
4   Burner  BlocK
          T2,OOO
                          30
                                                        FOR
I4,OOO
1
1
1
2
1 I
3 4
1
5
1
6
                             N FLUE. 96
        Figure 9.  Normalized NO concentration as
          function of flue O2 for the IF LB burner
              with a standard gaa nozzle and
                    15% and 30% FGR
                          11-46

-------
   300-
   25O-
E
ex
O
Z 200H
4>
N
fj 150-
o
z
   100-
    50-
                Gos  input   3070  SCFH

                Cos  Nozzle  B of fie  Position

                Wall  Temperature   965*  C
                4* Burner Block  Angle
                Air  Preheat AS   Labeled
                                                       232  C
                               O Standard  Nozzle, Baffle Position

                                 Wall  Temperature  965° C
1
\
	 ( —
2
1
3
4
i
e1
                          IN  FLUE,%
        Figure  10.  Normalized NO concentration as a

           function of flue O2 for the IF LB burner

                with a standard gas nozzle at a

                  wall temperature of 965°C
                             H-47

-------
 7OO-1
            Gas inpgt  2970  SCFM,  Radial
           Gas Nozzle  Baffle  &  Throat Position
           Won Temperature   1370° C
           4  Burner  Block  Angle
                                      legeiv;
100-
  Stand-ire! Nozzle Rottle Position
O R,j Jill N- 7.-!C.
O «odiol N--zzJe
                                              Boffl" r<.-siti">n
                                              Throat Position
1
t

1
2
0, IN
1
3
FLUE, 'X
1
4

1
5

1
6

   Figure II.  Normalized NO concentration as a
       function of flue O2 for the IF LB burner
          with a combination gas nozzle and
                    radial injection
                        II-43

-------
10/75
                                                          O
                                                          c

                                                          
-------
 in the measured NO concentration (approximately by a factor of 2-1/2).
 This decrease could be due to the combustion outside the burner block,
 which allows a larger mass exchange to occur among the fuel,  the internal
 recirculation zone,  the secondary air,  and the external recirculation zone,
 thus lowering the flame temperature.
    In an attempt to further slow the rate of heat release,  half the gas was
 introduced radially while the other half was introduced axially (at a high
 velocity of approximately 300 ft/s).  Again, data were collected for the half
 axial-half radial injection in  both the baffle and throat positions.  These
 results are presented in Figure 13.  The axial gas injection had the desired
 effect:  It decreased the NO concentration relative to all-radial injection by
 projecting  the flame further into the furnace and slowing the rate of heat
 release.  Again, moving the  nozzle position into the  throat of the burner
 block further  reduced the NO concentrations.
    Figure 14 is  a composite plot showing all  significant reductions in NO as
 a function of nozzle  type and/or nozzle  position.  All these data wf.'re col-
 lected with a secondary air preheat at 4^0 C.  The base-line conditions were
 those of the standard nozzle in the baffle  position;  this burner configuration
 produced the highest NO concentrations.  Moving the standard nozzle into the
 throat position did not  change the shape of the NO-versus-O2 relationship,
 indicating that the basic mixing patterns were not dramatically altered.  How-
 ever,  it did decrease the NO concentration to approximately 125 ppm.  With
 a divergent gas  nozzle  (Figure 15} similar to  the one used in the kiln  burner,
 an additional decrease  to 100 ppm NO resulted.  The differences in NO con-
 centration  produced with the  divergent nozzle and the standard  nozzle in the
 throat position decreased as  a function  of excess O2, until at 6% excess O2,
the NO concentrations  were equal.  The divergent  nozzle produced a type III
flame.  The decrease  in NO occurred because of the additional entrainment
of combustion products from the  primary ai^i secondary recirculation zones.
Although the data are not plotted, test results with the divergent nozzle in
the throat position paralleled those with the kiln burner; i.e.,  the NO was
 increased by moving the gas  injector toward the combustion chamber. The
difference  in NO concentration as a function of aozzle position, however,  is
very small.
                                 11-50

-------
  50O-
 40O-
•0-3OO-

N
  200-
  100-
           ,700
                     Gas
Gas
Wall
input   3101  SCFH
     1511 Axial. I59O Radial
Nozzle  Baffle  ft  Throat  Position
Temperature   I39O° C
                     4° Burner  Block  Angle
                  Legend 	
             Standard Nozzle, Baffle  Position
           O Hatr Radial Axial Nozzle, Baffle Poeitfon
           D Halt Radial Axial Nozzle, Throat  Position
                                         I
                                        4
                            I
                            S
                          IN FLUE, 96
            Figure 13.  Normalized NO concentration as
               function of flue O2 for the IF  LB burner
                  with a combination gas nozzle and
                       axial and radial  injection
                                   II- 51

-------
   600-,
                  Cos  Input   3070  SCFH
                  Secondory  Air  Preheat   460° C
                  4° Burner  Block  Angle	
   500-
   400-
E
a.
a.
 *

Z 300-|
N
Z 800-
   100-
OStandard  Nozzle. Battle ft«ltIon
V Standard  Nozzte. Throat Position
A Divergent  Nozzle.Baffle Position
D Axial  Nozzle, Baffle Position
OStandord Nozzle, Baffle Position, 15% FOR
I         I         I
234
00 IN  FLUE. %
                                                   I
                                                  5
                          6
             Figure  14.  Normalized NO concentration as a
                function of flue O2 for the IF LB burner
                     with the various gas nozzles in
                           different positions
                                  II-52

-------
10/75
                                                                O
                                                                C
                                                                CO
                                                                a
                                                                bt>
                                                                u
                                                                C
                                                                V
                                                                CD
                                                                f-,
                                                                4)
                                                                3
                                                                00
                                        H-53

-------
    The most dramatic decrease in NO with burner alterations occurred when
the gas velocity was increased by a factor of  16.  This was accomplished
by using all-axial gas injection but decreasing the  injector opening from the
standard E inch to 0. 5 inch.  As  shown in Figure 14, at less than 2% excess
O2, the divergent nozzle produced the lowest NO emissions; however, above
that level of excess O2, the high-velocity axial nozzle produced the lowest
NO emissions.  At 3% excess O2 (11. 4%  excess air, a. typical industrial level),
comparing the normal burner output of 515 ppm with the high-velocity gas-
injection burner output of 250 ppm, we decreased the NO pollutant level by
325 ppm through a low-cost alteration of the gas injector.  Noted that exter-
nal FGR remains the best depressant of NO formation that we tested. At 15%
FGR, the NO level is 150 ppm with the standard baffle-and-nozzle configura-
tion, which represents a total decrease of 425 ppm.
    Burner-Block Angle
    To test the influence of the primary recirculation zones on the level of
NO emissions, the burner-block angle  was increased from 4 to 8 deg.   The
primary recirculation zones contain not only combustion products but also
fuel and air. The entrainment of these gases by the  primary and secondary
jets should help to depress the flame temperature; however,  they are not as
desirable  as the gases in the secondary recirculation zone because they con-
tain both fuel and air.   The  angular difference between the 1FLB (the air ports
are angled at 15 deg relative to the centerline of the  baffle) and the burner
block (air entry being an 8-deg divergent cone) is 7 deg.  Because, the angle
of the ports is greater than that of  the burner block,  the flow will not separ-
ate from the contour of the block.  The IFLB was tested with the  standard,
divergent,  and high-velocity axial gas nozzles with an air preheat tempera-
ture of approximately 460 C.  The results are graphed in Figure 16.
    Table  1  compares the NO levels for integer values of the  percentage of
excess O2 in the flue versus nozzle type between the  4 and 8-deg burner
blocks.  For the standard and divergent gas nozzles  at low levels of excess
air (below 12%),  the emission levels from the two  burner blocks  are compar-
able.  However, as the excess air is increased, the  emission levels  from
the 8-deg block become larger than those from the 4-deg block.   Use of the
                                    II-54

-------
   700
                Cos  input   2298  SCFH
                Gas  Nozzle  Baffle  Position
                Wall  Temperature   1360  C
                   Burner BlocK Angle
8'
   600
   500-
E

N

O
E
   300-
  200-
  IOO'
                 Legend	
              O standard  Nozzle
              A Divergent  Nozzle
              O Axial  Nozzle
	1	
       3
 IN  FLUE, %
                                    T
                                    6
     Figure 16.  Normalized NO concentration as a
        function of flue O2 for the IF LB burner
    with standard, divergent, and axial gas nozzles
                           II-55

-------
   Table  1.  NORMALIZED NO AS A FUNCTION OF NOZZLE TYPE,
      BLOCK ANGLE,  AND EXCESS O2 FOR THE IFLB BURNER
          WITH AN AIR PREHEAT TEMPERATURE OF 46o°C
  Nozzle Type  Block Angle,  deg
  Standard
  Divergent
  Axial
4
8
4
8
4
8
% 02
1
390
408
197
204
227
174
2
487
492
273
272
255
196
3
	 .„. nnm
577
564
313
360
280
238
4
567
612
340
428
300
278
__5 	
542
636-
360
480
313
313
high-velocity (611 ft/s) axial nozzle with the 8-deg block at normal operating
levels of excess air (below 20% ) resulted in a 20% reduction in NO emissions.
Thus the desired mass exchange among the primary jet, secondary jet,  and
the internal recirculation zones occurs  only for the high-velocity axial noz-
zle.  In the case of the standard nozzle  (38 ft/s) gas velocity) and the diver-
gent nozzle (654 ft/s gas velocity), the increased burner volume resulted in
high emissions.  This increase occurred because a) the rate of heat release
resulting from a high mass exchange between the primary and secondary jets
improved more than the mass exchange among  the primary jet, the secondary
jet,  and the internal recirculation zones or b) the volume of the secondary re-
circulation zone was decreased.
SFLB Burner
    Air Velocity
    Initial tests with the SFLB burner were conducted by using a high-velocity
(126 ft/s) baffle and a  low-velocity (80 ft/s) baffle.   (Air injection ports  are
angled at 25 deg relative to the  centerline of the baffle. ) Because of the
extreme pressure drop associated with  the high-velocity baffle (approximately
22 in. H2O  with an air flow rate of 38,000 SCF/hr at 450 °C), the tests were
conducted with a gas load of 2000  CF/hr.  The  8-deg burner-block angle used
in these tests is the manufacturer' s recommended angle for industrial appli-
cations.
    Figure  17 presents the test  results as normalized NO emissions plotted
against the measured percentage of O2 in the flue.  The high-velocity baffle
produced consistently  lower emission levels than the low-velocity baffle as
                                      11-56

-------
   600-1
   500-
a. 400-
o.
Gas  input    2O20  6CFH
Gas  Nozzle  Baffle  Position
Wall Temperature  1295°  C
Air  Preheat  A«  Labeled
6°  Burner  Block Angle
i: 300
o
w.
O
z
   200"
                             Legend
         I3OJXX) o  LOW Velocity (80 fps) Short  Flame  Bafflt
                A  High Velocity (125 fps) Short Flame  Baffle
  8
              34
            IN  FLUE,%
                                               i
                                               5
6
      Figure 17.  Normalized NO concentration as a
         function of flue O2 for the SF LB burner
               with a standard gas nozzle
                           11-57

-------
a result of the larger mass exchange between the secondary recirculation
zone and the secondary jet because of the increased velocity gradient.  Thus,
the higher the velocity of the secondary jet, the lower the NO emission levels.
The magnitude of the reduction in NO depends  not only on the velocity gradient
across the shear layer separating the secondary jet from the secondary
recirculation zone but also on  the volume and temperature of the secondary
recirculation zone.  The cost involved in this method of NO reduction is due
to the additional fan horsepower needed to overcome the  increased pressure
drop in the burner.
    Wall Temperature
    Figure 18 illustrates the effect of wall temperature on the NO emission
levels of the SFLB as a  function of secondary air preheat and excess air.
Conclusions  similar to those for the IFLB can be drawn for the SFLB:  A
wall temperature as low as possible must be maintained  through  increased
wall cooling, lowered gas input,  or furnace design.
    External FGR
    To determine whether combustion aerodynamics can  be used  to recircu-
late combustion products to the base  of the  flame as an emission control tech-
nique, we simulated the idealized case (in which the combustion air  and pro-
ducts are thoroughly mixed before ignition) by mixing flue gases  with the
combustion air outside the burner.  Figure  19 compares the NO levels  as a
function of the percentage of Oz in the flue for  0%, 15%,  and 25% FGR.  Again,
as demonstrated for the kiln burner and the IFLB, external FGR is an ex-
tremely effective method of controlling NO  emissions from the SFLB.
    Nozzle Type
    To determine the type of injector that would reduce NO emissions,  a
series of trials were conducted by using different methods of gas injection;
the results are shown in Figure 20.   The NO-versus-excess O2 relationship
for the standard nozzle (injection method recommended by manufacturer;
2-inch stainless-steel pipe with an inlet velocity of 38 ft/s and a  gas input
rate of 3000 CF/hr) is shown as a reference, so that the  effa:': xf !;Ae gas
injector type on the NO emission levels can be evaluated. Visual observa-
tions wer,e made to determine the flame length for each gas nozzle tested.
This allowed a first-order evaluation of alterations that can be made to obtain
the flame shape desired by the manufacturer.

                                     11-58
\

-------
    800-t
    700-
    6001
CL
o
z
N
£  400H
o
z
    30O-
   2OO-
    100-
               Gas input , 3079  SCFH. Axial
              Gas Nozzle Baffle  Position
              Wall Temperature    1050* a  I45O* C
              Air  Preheat As  Labeled
              8°  Burner  Block  An die
          	Leaend 	
         O 14S06  C Wall  Temperatur«
         A I05O   C Wall  Temperature
              '30,000
            11,000
  9
2

°2
                                      \
                                      4
                           i
                           6
IN FLUE,%
       Figure 18.  Normalized NO concentration as a
          function of flue O2 for the SF LB burner
               with  a standard gas nozzle at
          wall temperatures of 1450° and 1050°C
                              11-59

-------
   800H
  700 H
  600H
O 500H
ISI
-400H
o
  300 H
  8001

Gas input 3070 SCFH, AKiol
Gas Nozzle Baffle Position
Walt Temperature I36O C
81* Burner Block Anflle

0% FOR
        6,000
                                             25% FOR

	 o-o-
i
1
	 o —

I
2
	 o—

i
3
— o

i
4


i
5
O

6
  10
IN FLUE,
      Figure 19.  Normalized NO concentration as a

         function of flue O2 for the SF LB burner

    with a standard gas nozzle and 15% and 25% FGR
                             11-60

-------
O
z
X3
0)
N
b
   IOOO
   900-
    8OO-
    700-
   600-
    500-
   400-
   3OO-
   2OO-
   100-
                Gas input  3O62  8CFH
                Gas Nozzle  Baffle Position
                Wall  Temperature  1420° C
                8"  Burner  Block  Angle
      — Legend —
        V Radial Nozzle (IO43 fps Gas Velocity)
        ^ Radial Nozzle (532 fps G«s Velocity)
        O standard  Nozzle
        O Divergent  Nozzle
        O Axial  Nozzle
        O Standard Nozzl*»   15%
                                                  O	
^o-— — v
6,000
I 1 i
1 2 3
4 i ^
   12
Q,  IN FLUE
            Figure 20.  Normalized NO concentration as a
               function  of flue O2 for the SF LB burner
                     with the various gas nozzles
                                11-61

-------
    The visual flame length observed with the standard gas nozzle with the
SFL.B was  103 cm.  The high-velocity (1043 ft/s) radial injector increased
the NO emissions substantially above those measured for the standard nozzle.
The shape  of the emission curve closely resembled that of a premixed flame;
peak emissions  occurred at 1.5% excess OE. The flame was invisible.  Re-
ducing the  radial velocity to 532  ft/s resulted in a shift of the maximum NO
emission to an excess O2 level of 2. 3%.  Because of the  reduction in the
radial injection  velocity,  the magnitude of the peak NO concentration was
reduced  20%. The visual^flame  length was 72 cm,  still 31 cm shorter than
that desired.  The divergent nozzle, which was the first  altered gas injector
tested with the SFL.B, resulted in lower emission levels  than the standard
nozzle.   The gas velocity from the divergent nozzle was  654 ft/s (at a 3000
CF,/hr gas load); however,  a type III flame resulted because of the wake gen-
erated by the  divergent cone in the nozzle.   Despite the high inlet velocity,
the flame length was 113 cm,  which compares favorably  with the desired
length of 103 cm.
    The injection method that  resulted in the lowest emissions was the high-
velocity  axial nozzle (611 ft/s  and a 3000 CF/hr gas load).  This nozzle pro-
duces a type II flame, even with  the high-swirl SFLB, because the velocity of
the gas is large  enough to split the primary recirculation zone.  Although
the rate  of entrainment per unit area between the primary jet  and the sur-
rounding flow zones increased because of the larger velocity gradient when
compared with that from the standard nozzle, the flame length also increased,
to 286 cm.  The increased flame  length was a result of a lower mass exchange
rate per unit time between the primary and secondary jets.  This reduced
mass exchange  resulted from a decrease in the area of the shear layer caused
by the smaller primary jet.  Thus,  by delaying the mixing between primary
and secondary jets,  we not only slowed the heat release rate of the flame, but
also approached the ideal of the EFGR by allowing more  time for the sec-
ondary jet  and the combustion products in the secondary  recirculation zone
to mix.
    Burner-Block Angle
   »	
    Figure 21 represents NO data from the SFL.B trial with standard, diver-
gent,  and high-velocity axial nozzles and a l6-deg burner block.  Obviously,
the combustion  aerodynamics were  changed drastically as a function of excess
                                     11-62

-------
6OO-,
         Goa  input   2953  SCFH
         Cos  Nozzle  Baffle  Position
         Wall   Temperature  1460°  C
         16  Burner  Block Angle
                         	Legend
                          o standard Nozzl*
                            Divergent Nozzle
                          Q Axial  Nozzle
                     IN   FLUE.
 Figure 21.  Normalized NO concentration as a
    function of flue O2 for the SF LB burner
with standard, divergent, and axial gas nozzles
                     H-63

-------
 air.  There is no one general interaction theory that explains these curves
 in their entirety.  Therefore, we selected a 1%  level of excess O2for analysis.
 This level was chosen because it consistently produced the lowest levels of
 NO, with concentrations below 50 ppm.
    Table 2 lists the nozzle type  as a function of burner-block angle for both
 the SFLB and the IFLB.  The inlet gas velocity is listed in parenthesis  beside

 Table  2.  NORMALIZED NO CONCENTRATION AT 1% EXCESS O2 WITH AN
 AIR PREHEAT TEMPERATURE OF 4&0  C AS A FUNCTION OF BAFFLE
       TYPE,  GAS NOZZLE TYPE, AND BURNER-BLOCK ANGLE
       Nozzle Type
NO Concentrations


4



590
375 -»
175 ->
215 «-
120
IFLB
Block
8


--
__
400 -»
200 ->
175 -»
--

Angle, deg
16 8
nnm ,

950
700
470 490
235 440
200 230
110
SFLB

16


- -
--
*- 240
«- 210
-* 310
—
    Radial (1043 ft/s)
    Radial (532 ft/s)
    Standard (38 ft/s)
    Divergent (654 ft/s)
    Axial (611 ft/s)
    Standard,  15% FGR
the nozzle type.  For 14%  excess air, the velocity is 120 ft/s.  Note that
the air velocity increases with increasing excess air.  The arrows indicate
the direction of increasing NO concentration.  We know from flame-length
observations that the standard and divergent nozzles have  similar rates of
mass exchange between flow zones,  whereas the axial nozzle has a slower
exchange rate. It is not surprising  that the standard and divergent nozzle
show similar trends in NO formation as a function of burner-block angle.
For the IFLB,  the NO concentration increases with burner-block angle; for
the SFLB, it decreases with an increasing angle.
    For the IFLB, we increased the mixing rate  between the primary and
secondary jets more than the size, shape, and mixing rate of the primary
recirculation zone with the primary and secondary jets.  For the SFLB,  we
were  successful in improving the mass exchange between the primary recircu-
lation zone and the primary and secondary jets,  which resulted in approxi-
mately a 55%  reduction in NO emissions.  The high-velocity axial nozzle
showed almost the opposite effect.   For the SFLB, there was a 30% increase
                                   II-64

-------
in the emission level as the block angle was increased.  Thus, for high
swirl intensity and high axial velocity, increasing the burner-block angle
resulted in a larger primary-secondary jet mass exchange.  We investigated
the IFLB with the 16-deg block for flow separation between the secondary jet
and the block and obtained negative results.   Therefore, we assumed that it
also was not present for the SFL.B and the l6-deg  block.  The results for the
IFLB and the high-velocity axial nozzle were somewhat mixed.  There  was
a decrease in the NO level with a block angle increase from 4 to 8 deg fol-
lowed by an increase in the NO level with an angle increase from 8 to 16 deg.
The optimum burner-block angle for producing minimum NO emissions with
the axial nozzle was approximately 10  deg.
Wall Temperature
    Because of the  observed sensitivity of wall temperature to the flue con-
centrations of NO during these tests, we decided to gather data  relating NO
to wall temperature during a warm-up cycle.  The standard nozzle was fired
with 3000 SCF/hr of gas at 3% excess O2, 460°C air preheat, and the SFLB.
The results are graphed in Figure 22.  The dramatic increase in the level of
NO emissions above 1300°C indicates that an accurate control of wall temper-
ature is required to produce a self-consistent set  of experimental data.  It
also confirms that  the furnace should be operated  with as low a wall tempera-
ture as  possible.
Furnace Geometry
    The IFL.B burner was mounted on our cylindrical (tunnel) test furnace,
which has a cross-sectional area of 1.  17 sq m (12. 6 sq ft) and a volume of
6. 4 cu m  (226 cu ft).   The following operating parameters were  investigated:
secondary combustion air preheat, method of gas  injection, and position of
gas injector.
    Figure 23 shows the emissions data collected from the IFLB with the
standard gas injector (2-inch stainless-steel tube), a 2000 SCF/hr  gas  input,
and nonpreheated and preheated secondary combustion air.  To determine
the influence of furnace geometry on  NO emissions.  Figure 23 compares
these data with similar data collected during a test on the larger-diameter
furnace.  The cross-sectional area of this rectangular furnace is 2. 3 sq m
(25 sq ft)  and the volume is 10.6 cu m (375 sq  ft).  With the exception of
                                    11-65

-------
  700^
 6OO-
E
a.
  500-
E
t->
o
  40O-
Gos  Input  3O08  SCFH.AXlfll

Gas  Nozzle  Baffle  Position

3%  Excess   Oxygen

460*  C  Air Preheat

   Burner flipck
           IOOO
        1100
1200
1300   1400    1500
                  Woll  Temperature,  C
       Figure 22.  Normalized NO concentration as a

           function of wall temperature for the

         SF LB burner with a standard gas nozzle
                       11-66

-------
   8001
   700-
   600-
Nl

O

I
   300H
  200-
   100-
Oas  input  1958  SCFH
Gas  Nozzle  Baffle Position
Wall  Temperature   1360*  C
Secondary  Air Preheat As Labtittf
4° Burner ..fljbcfc  Angle	
                            — Legend—
                             O Cylindrical
                             O Cylindrical
                             A Rectangular
                             v Rectangular
                                         454° C
                                                     0-220 c
   14
                                  1        7
                                IN  FLUE , %
        Figure 23,  Normalized NO concentration as a
            function  of flue  O2 for the IF LB burner
                with a low-velocity gas nozzle
                                11-67

-------
 secondary air preheat temperature (350  C for the cylindrical furnace,  com-
 pared with 454 C for the rectangular furnace),  all other burner and furnace
 operating conditions (gas input, gas nozzle position, burner-block angle, and
 wall temperature) were identical.   For both conditions of secondary air pre-
 heat, the cylindrical furnace had the higher levels of emissions. The area
 ratio between the burner-block opening and the burner wall is 35.9 for the
 rectangular furnace and 18, Z for the cylindrical furnace.  These ratios
 reflect the relative sizes of  the secondary recirculation  zones.  The primary
 reason for the difference in  measured  NO concentrations between the two
 combustion chambers is the unequal secondary recirculation zone volume.
    Nozzle Type
    Additional tests were conducted to  investigate the influence of gas  injec-
 tion on total NO emissions.  In these trials, we used a high-velocity (127 m/s,
 407 ft/s) gas nozzle and a divergent gas nozzle (injection velocity 135  m/s,
 436 ft/s).  Both injectors have only an axial velocity component. Figure 24
 compares the normalized NO emissions as a function of  excess  O2 from
 these gas nozzles.  The low-velocity nozzle produces less NO,  by a factor of
 2, than the high-velocity nozzle.  This result contradicts the data collected
 from the rectangular furnace, which are presented in Figure 25.  Although
 the rectangular furnace has  a higher gas input (2998 SCF/hr) and a higher
 secondary air temperature (450 C), the cylindrical furnace had the higher
 emission levels for a given gas nozzle.  This occurs because  NO production
 in gas flames is entirely thermal in origin.  Thus,  any change in the com-
bustion chamber  or fuel-air  mixing that results in lower peak and average
flame temperatures reduces the rate of NO production.   Conditions that
 should reduce NO emissions include decreasing the rate  of heat release,
increasing the mass exchange between the flame zone and secondary recircu-
lation zone, and reducing the wall temperature.
    During previous trials with the rectangular furnace,  the high-velocity gas
nozzle always produced less NO emissions than nozzles with lower injection
velocities.  This result was  explained by the different fuel-air mixing  rates.
An increase in injection velocity was achieved by decreasing the cross-sec-
tional area of the injector, which caused a decrease in the fuel-air mixing
rate and permitted an increase in recirculation and combustion zone mass
exchange.  The combustion zone is blended with combustion products,  which
                                    H-68

-------
   BOOT
  I20O-
   1(00"
  \ooo-
E
ex
 fk
O
Z
o
o
   BOO'
   TOO'
                               Gas  input   1958 SCFH
                               Gas  Nozzle  Baffle  Position
                               WON  Temperature   I36O°  C
                               Secondary  Air  Preheat  350°
                               4*  Burner  Block   Angle
  60O
   5001
                                       	Legend	
                                        O LOW Momentum Nozzle
                                        A High Momentum Nozzle
                                        ^ Divergent Nozzle
                                  3        4
                               IN FLUE, %
                                                              6
        Figure Z4.  Normalized NO concentration as a
           function of flue O2 for the IF  LB burner
               with low-velocity, high-velocity,
                  and divergent gas nozzles
                              II- 69

-------
  6OO
                Gas  Input  3O70  SCFH
                Secondary  Air Preheat
                4  Burner  Block  Angle
               450° C
 100
                                    .Legend
    OStandard  Nozzle, Baffle Position
    A Diver gent Nozzle.  Bottle Position
    Q Axial   Nozzle, Baffle Position
16
    I         T
    3        4
IN FLUE, 96
    Figure 25.  Normalized NO concentrations as a
        function of flue O2 for the  IF LB burner
   •with standard,  divergent,  and axial gas nozzles
                            II- 70

-------
result in lower ilame temperatures.  However,  for the cylindrical fur-
nace, the high-velocity gas nozzle causes higher NO emissions.  This means
that even though the same method of gas injection was used in each furnace,
the thermal history of the two flames was different.  The major reason for
a higher flame temperature in the  cylindrical furnace  would be  a smaller
mass exchange between the secondary recirculation and the combustion zones.
The smaller mass  exchange occurred because of a decrease in  the size of the
secondary recirculation zones, which, in turn,  resulted from a smaller com-
bustor cross-sectional area.
    The divergent nozzle  injects the gas radially at a 45-deg angle to the
burner axis.  Comparison of  Figures 24 and 25  shows that the cylindrical
furnace has the higher levels of emissions for the divergent nozzle.  Argu-
ments similar to those presented for high-velocity nozzle could also be made
for the divergent nozzle.   However,  for both combustion chambers, the
divergent nozzle has lower emissions than the high-velocity nozzle.  The
diverging gas injection causes a primary  recirculation zone,  which allows a
larger dilution of the combustion zone with combustion products than an axial
injector.  This increased dilution leads to lower flame temperatures and
reduced emissions.
Natural-Gas, Low-NO  Burner I
                     Jt
    IGT has been commissioned by three gas utilities (Consumer's Gas Co.,
Consolidated Natural Gas Co. , and Southern California Gas Co. ) to develop
a burner with low NO  emission levels relative to the industrial burner it
                    x
was intended to replace and with a similar flame geometry and  luminosity.
The initial design guidelines for this burner development work were extracted
from the data presented in this paper.  After fabrication, the low-NO  burner
                                                                   Jt
I (LNO-1) was mounted on the cylindrical furnace for NO  emission level
testing.
    Standard Operating Conditions
    The initial set of trials was conducted by using 25% primary t  r.  This
percentage is determined by ratioing the volume of primary air tc  total air
needed for stoichiometric combustion and multiplying it by 100.  !' igure 26
shows the results of these tests as normalized NO plotted against excess O^.
These data were collected by premixing the primary air and fuel, which gives
                                     II- 71

-------
900-
 800-
 7001
200-j
 ioo-
            Gos  Input   1980 SCFH
            Secondory  Air  Preheat  Temperature  AS  Labeled
            !5°  Spin  on  Secondary  Air
            25%  Primary  Air
            ciear_ Flame	
                                O Baffle Burner-Stand. Nozzle
                                A LOW NOX Burn«r-I
                                V LOW NOX Burner-X
 17
                             3       4
                       0  IN FLUE, %
       Figure 26.  Normalized NO concentrations as a
         function of flue O2 with 25% primary air
                            11-72

-------
rise to a clear flame.  For comparison,  Figure 26  also shows the emissions
curve for the IFLB burner with a standard nozzle.  The LNO-I burner was
designed to replace burners in the IFLB burner category.  Comparison of the
emission levels of these burners at  350  C secondary air preheat shows that
the LNO-I burner was extremely effective in reducing NO emissions.  At a 2%
level of excess O2 (approximately 10% excess air),  the LNO-I burner emitted
55% less normalized NO than the IFLB burner.  This decrease in NO emis-
sions dropped to 18% for NO secondary air preheat.
    Flame Luminosity Adjustment
    Adjusting the primary air and fuel so that no mechanically induced pre-
mixing occurred prior to ignition (luminous  flame) did little to alter the
LNO-I burner emissions without secondary air  preheat.  Figure 27 illus-
trates the normalized NO emissions versus  excess  O2 for luminous flame
operating conditions.  With a  350 C secondary  air temperature,  the lumin-
ous flame had an NO level,  at 2% excess O2, 13% lower than that of the clear
flame.  Thus,  the effect of the mixing rate  of primary air and fuel on NO
emissions is directly related to secondary air temperature.  At ambient
combustion air temperatures, the luminous  and clear flames had nearly the
same emission levels, but at the elevated (350  C) combustion air tempera-
ture, the clear flame had the higher emissions.
    Variation in Primary Air
    To determine the influence of primary air volume on NO formation, the
amount of primary air was  reduced to 15%.  Figure 28 shows the normalized
NO-versus-excess O2 relationships determined for  this primary air volume.
The clear flame with 350°C secondary air preheat displayed little change in
NO emissions between 15% and 25% primary air.  The luminous  flame was
somewhat more sensitive to primary air  volume, yielding a 10%  increase in
the NO emission levels at the lower primary air level.  Thus, if a luminous
flame is needed for the desired industrial application, the higher the percent-
age of primary air,  the lower the NO emissions.
    With an ambient secondary air temperature, both flame conditions pro-
duced the lowest levels of NO measured during  the burner trial series, indi-
cating that the  percentage of primary air should be  reduced as the secondary
air preheat temperature is  decreased.  To  completely generalize the
                                    H-73

-------
     400-j
        Cos Inpu     1993  6CFH
        Secondary  Air  Temperature  AS  Labeled
        15*  Spin on Secondary Air
        (5 96 Primary Air
       Clear and  Luminous Flames
     300-
A
O.
 to
O
    200-1
o
6
     too-j
   19
                                                     365° C
                  	Legend	
                  O Clear Flame
                  O Ctsqr Float
                  A Luminous F|0me
                  V Luminous Flome
        I       2       9       4       &
                02  IN  FLUE.K

Figure 27.  Normalized NO concentration as a
   function of flue O2 for the LNO-1  burner
  under luminous-flame operating conditions
                               II- 74

-------
               Gas Input   1985 SCFH
               Secondary Air Temperature  AS
               15° Spin  on  Secondary  Air
               25 %  Primary  Air
               Luminous   Flame	.
                                    Labeled
  30O-
 E
 a.
 a.
 «J2OO-
 E
 o
   IOO
i e
       12345
                 02  IN  FLUE, %

Figure 28.  Normalized NO concentration as a
   function of flue O2 for the LNO-1 burner
           with 15%  primary air
                              II-75

-------
dependence of NO emissions on primary air volume would require testing
above 25% primary air;  however, because of primary fan capacity, these
trials were not possible.
    Although this investigation was very successful in demonstrating that the
LiNO-I burner has low NO emission characteristics,  particularly at elevated
secondary air temperatures,  additional testing is under way to simplify the
burner design and to determine the burner1 s limitation, not only in NO
reduction but also in the types of flames it is capable of producing.
FUTURE WORK
    The final  step of the program has been initiated by the American Gas
Association in its funding of a field test demonstration for the L.NO-I burners.
IGT is currently contacting industries that might be interested in having
L.NO-I burners mounted on their furnaces with IGT furnishing the burners
and monitoring the emissions during process operation.
SUMMARY
    This experimental research program has demonstrated that the NO
emission levels from gas-fired industrial burners can be significantly
reduced by alterations in the combustion aerodynamics.  A synopsis of the
operational variables studied and their test results is presented in Table 3.
The  technique resulting in the  most dramatic reductions of NO emission levels
was  external flue-gas recirculation.  This verified the intimate relationship
between flame temperature and NO emissions; thus,  this program1 s efforts
concentrated on reducing flame temperature either by reducing the rate of
combustion or by diluting the  flame with combustion products recirculated
within the furnace using combustion aerodynamics.
    Although large variations  in NO emissions occur  because of changes in the
amount of excess air, reduction levels can be established in addition to their
relative effectiveness by comparing the emission levels at several fixed levels
of excess air.  The  conclusions reached below are based on  an excess air
level equivalent to 3% oxygen in the flue and a combustion air preheat temper-
ature of 450°C.
                                      II-76

-------
      Table 3.
 Excess Oxygen,
                 SYNOPSIS OF DATA COLLECTED FOR
                     THE BAFFLE BURNER
                                              1
                         Normalized NO,  ppm
 IFLB Burner (Standard operating conditions:  gas input, 3000 SOF/hr;
              450 C,  secondary air preheat;  2-inch-diameter  axial fuel
              injector: baffle fuel injector: baffle position; and a 1400 C
              wall temperature, 4-degree burner block. )
 Standard Operation
 Gas Input, 2000 SCF/hr
 Secondary Air Temperature:
                    1100°C
                             225°C
                              22°C
Wall Temperature,
EFGR 15T
EFGR 30f,
Radial Nozzle-Throat Position
Axial Nozzle-Throat Position
Throat Position
Divergent Nozzle
Axial Nozzle
8-Degree Burner Block (B. B. )
8-Degree B. B. -Axial Nozzle
8-Degree B. B. -Divergent Nozzle
390
250
195
110
225
130
 40
200
110
280
190
210
410
200
220
 580
 310
 235
 125
 260
 180
 60
 260
 190
 390
 300
 250
 560
 230
 350
SFLB Burner (Standard operating conditions:  gas input,  3000 SCF/hr;
              450 C, secondary air temperature; 2-inch-diameter axial
              fuel nozzle; baffle position; 8-degree burner block angle;
              and 1400  C average wall temperature. )
Standard Operation
Gas Input,  2000 SCF/hr
Secondary Air Velocity,  1Z5
Secondary Air Temperature:
                   1000 C
                             225 C
                              22°C
Wall Temperature,
EFGR 15"',
EFGR 30^
Radial Nozzle
Divergent Nozzle
Axial Nozzle
16-Degree Burner Block (B. B. )
16- Degree B. B. -Divergent Nozzle
16-Degree B. B. -Axial Nozzle
450
270
210
165
 95
270
100
 60
660
440
210
210
180
300
660
430
250
270
160
320
150
 80
800
580
390
380
400
390
                       — Cylindrical Furnace —
IFLB Burner (Standard operating conditions:
             350°C,
                                           gas input,  2000 SCF/hr,
                     secondary air temperature; 2-inch diameter axial
             fuel nozzle, baffle position, 4-degree burner block angle
             and 1 100 C average wall temperature. )
Standard Operation
Axial Nozzle
Divergent Nozzle
                                            460
                                            880
                                            825
              610
             1170
             1090
    LNOX-I
                                            180
                                                          280
  Gas input for this test was  2000 SCF/hr
                               11-77

-------
    For the intermediate flame length baffle burner (IFLB), which has a

tangential-to-axial velocity component ratio of 0. 27, we found the following:

a.  Reducing the firing rate from 3000 SCF/hr to 2000 SCF/hr resulted in a
    47% reduction in NO emissions.

b.  Reducing the combustion air preheat from 450 C to 225 °C leads to a 59%
    reduction in NO emissions, and no preheat gives a 78% reduction.

c.  By cooling the walls from  1400°C to 1100°C,  the NO emissions diminished
    by 55%.

d.  External flue-gas recirculation reduced NO emissions  by 69% for 15%
    recirculation and 90%  for 30% recirculation.

e.  Several types of fuel injectors and positions were tested with a maximum
    reduction of 67% measured for the high-momentum axial nozzle in the
    throat position.  The minimum changes in flame geometry and luminosity
    were observed for the divergent nozzle, which showed  a 48% decrease
    in NO emissions.

f.   The burner block angle was increased from 4-degrees  to 8-degrees,
    leading to only a 4% reduction under standard burner operating conditions.
    However, by also changing the method of fuel injection a maximum reduc-
    tion of 60% was  measured  with the high-momentum axial gas nozzle.

For the short flame length baffle burner (SFLB),  which has a tangential-to-

axial velocity component ratio  of 0.47, we  observed the following:

a.  Reducing the firing rate from 3000 SCF/hr to 2000 SCF/hr resulted in a
    35% decrease in NO emissions.

b.  With a 2000 SCF/hr gas input,  the combustion air velocity was  increased
    from 80 ft/s to  125 ft/s,  which diminished the NO emissions by 42%.

c.  Reducing the combustion air temperature from 450°C to 225°C  resulted
    in a 59% NO reduction and, with no preheat, a 76%  reduction.

d.  Decreasing the wall temperature  from 1400°C to 1000°C reduced the
    NO emissions by 52%.

e.  Externally recirculating flue gas  and blending it with the combustion air
    produced a 77%  decrease in emissions for 15% recirculation and 89%
    reduction for 30% recirculation.

f.  Of the  several types of fuel injectors tested, the high-momentum axial
    nozzle produced the minimum emissions with  a 41% reduction.

g.   The burner block angle was increase-:!  from 8-degrees to  l6-degrees.
    This produced a 41% decrease in NO emissions for all  gas nozzles tested.
                                   II-78

-------
    The IFLB burner was tested on our cylindrical test furnace, which has a
volume of 226 cu ft and an area ratio between the burner block opening and
the burner wall of 18. 2 compared with a 35. 9 ratio for the rectangular furnace.
During these trials we have the following observations:
a.  At a 2000 SCF/hr firing ratio, the cylindrical furnace emissions were
    97% greater than the emissions  from the rectangular furnaces.
b.  The high-momentum axial and the divergent nozzles produced increases
    in the NO emissions of 92% and  79%, respectively,  although both produced
    reductions in NO emissions on the rectangular furnace.
    This program has led to the design,  development, and testing of a low-NO-
emissions burner (LNOX-I), which, hopefully because of its versatile design,
could be used in a large number of industrial applications.  The following
conclusion was drawn from the test  data accumulated from LNOX-I:
•   With 2000 SCF/hr firing ratio on the cylindrical furnace,  the LNOX-I
    burner resulted in a 54% reduction in NO emissions when compared
    with the IFLB burner under similar operating conditions.
    Thus, this program has not only demonstrated techniques and variations
in furnace and burner operating conditions, which lead to reduced NO emissions,
but these test results have produced a burner design that has characteristic
low emissions.  The final success of this program will be in the near future
when the LNOX-I burner is field-tested.
                                    11-79

-------

-------
          INTEGRATED LOW-EMISSION




            RESIDENTIAL FURNACE
                   By:




L. P. Combs, W. H. Nurick, and A. S. Okuda




Rockwell International, Rocketdyne Division




          Canoga Park, California
                   11-81

-------
                         ABSTRACT

Analytical and experimental studies are establishing practical
technology for simultaneously minimizing pollutant emissions
from and improving fuel economy of oil furnaces for residential
space heating.  Objectives are to reduce the emission of ni-
trogen oxides from current levels of 2-1/2 to 3 g/kg of fuel
burned to less than 0.5 g/kg, and to raise season-averaged
thermal efficiency by at least 10 percent above that achieved
by typical current conventional furnaces, while maintaining
minimum emissions of smoke, unburned hydrocarbons, and carbon
monoxide.

The approach has involved research on designing conventional
pressure-atomizing oil burners for low emissions, matching
furnace fireboxes to such low-emission burners, and operat-
ing these optimized components under fuel-conservation con-
ditions.  This paper summarizes the research results and
shows how they are being applied in the design, assembly,
and test of a prototype warm-air furnace to verify that the
technology can be used in practical, cost-competitive pro-
duction furnaces.
                      ACKNOWLEDGEMENT

Work upon which this presentation is based, was performed
pursuant to Contract Nos. 68-02-0017 and 68-02-1819 with
the U.S. Environmental Protection Agency.  The EPA Contract
Officer was Mr. G. Blair Martin, Combustion Research Section,
Industrial Environmental Research Laboratory, Research Tri-
angle Park, North Carolina.
                       11-82

-------
                            INTRODUCTION

The Environmental Protection Agency has sponsored studies over the
past few years to document the emission of air pollutants from exist-
ing residential and commercial oil-fired space heating equipment (Ref.
1 and 2).  Concurrently, the EPA has also supported applied research
programs to determine the effects of design and operating parameters
on exhaust gas emission levels and to devise strategies for minimizing
pollutant emissions (e.g., Ref. 3, 4, and 5).  These and related stud-
ies have shown that substantial reductions in total emissions from fur-
naces can be realized by applying combustion control technology, such
as flue gas recirculation, staged combustion, and advanced burner
designs.

As part of those EPA activities, Rocketdyne performed an intensive in-
vestigation of residential and commercial oil burners (Ref. 5) which
led to criteria for optimizing conventional burner designs with re-
spect to emissions.  The optimized burners were shown to be capable of
reducing NOX emissions to below 1 g NO/kg fuel, in contrast to typical
manufactured residential burners that operate at about 1 1/2 to 3 g
NO/kg fuel or higher.  During the investigation of burner optimization,
several variations in the combustion chamber construction (adiabatic
refractory walls vs metal heat-sink or water-cooled walls) and design
(relative orientation of the burner and chamber axes and relative
burner and chamber diameters) were also studied.  The results clearly
demonstrated that the total emissions are sensitive to (1) the design
of various components comprising a residential heating system and (2)
design interactions among the components.  It became obvious, there-
fore, that minimum emissions could be achieved only by systematically
optimizing the burner in conjunction with the furnace combustion cham-
ber and heat exchanger as well as operating mode.  The research de-
scribed in this paper was undertaken to investigate such system opti-
mization and to delineate furnace design requirements for commercial-
izing the optimum furnace technology.

In view of uncertainties in fossil fuel supplies and prices, and of
the national emphasis on achieving energy independence, an optimized
furnace should, in addition to providing low levels of pollutant emis-
sions, also provide some substantial increase in overall thermal ef-
ficiency.  Further, for the developed technology to be adopted commer-
cially, the optimized design must be cost-competitive.  In this regard,
significantly increased fuel economy will be important in allowing con-
sumer recovery, within a reasonable number of heating seasons, of any
increased capital costs that might result from modifications required
for system optimization.  The need to establish technological require-
ments for optimizing residential heating systems is important, there-
fore, because such effort addresses both the improvement of air quality
and a better utilization of the nation's dwindling energy resources.
                               11-83

-------
The principal objective of this research program was to establish the
technology required for engineering optimization of residential heat-
ing furnaces combining minimum pollutant emissions with maximum ther-
mal efficiency.  Primary emphasis was given to systems fueled with No.
2 distillate fuel oil.  General overall goals are:  (1) to reduce nitro-
gen oxide emissions to less than 0.5 g NO/kg fuel burned, while main-
taining minimum emissions of CO, UHC, and smoke, and (2) to increase
overall season-averaged furnace energy efficiencies by 10 percent or
more above those achieved by current conventional systems.

                             BACKGROUND

Typical residential oil furnace emission and efficiency performance data
are summarized in Table 1 for the existing furnace population and for
new furnaces representative of current technology.

Cycle-averaged emissions data, abstracted from Ref. 1 and 4, show that
the products of incomplete combustion (smoke, CO, unburned hydrocarbons,
and particulates) are generally quite low for both the existing popu-
lation and current new furnaces, although on the average those from the
former set tend to be appreciably higher.  The low levels result from
achieving the consistent design goal of accomplishing complete combustion
while the higher levels for the existing population are not surprising,
considering the greater component ages and broad diversity of burner
types, firebox design, conversion units, etc., which make up that
population.

Average emissions of the oxides of nitrogen from current new furnaces are
approximately the same as the average from the existing population.  This
results from the preponderant use of side-fired, refractory-lined fire-
boxes, irrespective of the burner designs.

Wide ranges of thermal efficiency estimates for the existing residential
furnace population may be found -in the technical literature.  Those
listed in Table 1 are representative of the mid-range,  and are generally
consistent with the flue gas C02 concentrations (or excess air levels) re-
ported in Ref. 1 and 4.  Recognizing that gross efficiencies, based on the
higher heating value of the fuel, are about 5 to 7 percent lower than net
efficiencies, it may be estimated that fuel utilization efficiencies
potentially might be increased by as much as 35 to 40 percent.  Achieving
such large gains, however, would entail very significant departures from
current design concepts, and manufacturing, marketing,  and utilization
practices.  To maximize efficiency, considerably larger heat exchangers
would be needed to cool flue gases to near-room temperature and condense
their combustion-generated water vapor.  A number of problems would im-
mediately arise, e.g., inadequate firebox draft, need for corrosion-
resistant furnace and flue construction materials, condensate disposal
                                  11-84

-------
and noncompetitlve Initial costs.  For these reasons, it was deemed
appropriate for the current research to address efficiency gains
achievable by less drastic modifications of existing commercial tech-
nology, i.e., improvements that can be achieved relatively near term
and remain cost-competitive in the marketplace.

IMPROVING THERMAL EFFICIENCY

The greatest source of thermal inefficiency in residential space heat-
ing equipment is the convection of heat up the flue.  When the burner
is being fired, exhausted product gases carry off substantial sensible
heat and the heat of vaporization of combustion-generated water vapor.
Flue gas sensible heat losses may be reduced by lowering either excess
air or flue gas temperatures.  Steady-state net thermal efficiencies
might thus be increased by 5 to 10 percent before furnace operation is
constrained by pollutant formation, low draft, and condensation problems.
During this same time, most installations use heated, humidified living-
space air for burner combustion air and for barometric draft control air.
Although the resultant heat losses are not charged against furnace
thermal efficiency, fuel utilization efficiency can be raised by 10 per-
cent or more by using a sealed air system to bring in outdoor ambient
air for these uses.

When the burner is not being fired, a natural draft flow of air through
the burner, firebox, etc., cools furnace components and continues to
convect heat up the flue.  This loss can reduce net thermal efficiency
by as much as 5 percent and is a major fraction of the transient heat
losses which cause cycle-averaged efficiencies to be lower than steady
state.  The lower the draft air loss is, the less sensitive is cycle-
averaged efficiency to variations of cycle duration and of fractional
burner-on time.  Draft air loss can be eliminated by providing a positive
shutoff device in the combustion air supply.  It can also be reduced or
eliminated by having the burner fire more nearly continuous, e.g., by
using modulated flow or high/low/off burner control.  No real thermal ef-
ficiency advantage can be realized with such control schemes, however,
because the flue gas temperature varies enough to offset the reduction
or elimination of draft air heat losses.

Heat conducted to the exterior cabinet of a furnace is radiated and
convected to the surroundings.  Although treated as a furnace loss,
this heat, in some installations, may contribute directly to heating
the residence and not be a true heat loss.   These furnace setting
losses average about 1-1/2 to 2 percent for warm-air furnaces.   They
are higher (3 to 3-1/2 percent) for hydronic boilers, because compo-
nent temperatures are more nearly constant  during standby than are
                               11-35

-------
 those in warm-air units.  Reduction of these losses nominally depends
 upon a straightforward economic tradeoff of fuel economy vs more ex-
 pensive insulation, although manufacturers must also be concerned
 about first-cost competition.

 The electrical energy expended in operating furnaces amounts to no
 more than 2 to 3 percent of the net heat of combustion of the fuel.
 Nevertheless, the minimizing of electric power consumption is a worth-
while goal because of its greater cost, to the homeowner than fuel oil
 and because of the factor of 2 1/2 to 3 on total energy savings when
 the inefficiency of the electric generating plant is considered.

 To maximize furnace thermal efficiency, without departing substantially
 from current manufacturing and installation practices, it was concluded
 that the following measures should be effected:

     1.  Design the burner/combustion chamber combination to operate
         pollution-free at low excess air levels, e.g., 10 to 15 per-
         cent excess air (13 1/2 to 13 percent C02)

     2.  Lower flue gas temperatures as much as possible.  The minimum
         practical temperature is probably around 200 C (392 F)
     3.  Employ sealed air systems for combustion air and barometric
         control.  Filter the combustion air to prevent long-term
         degradation of performance
     4.  Close the combustion air supply during standby

     5.  Effect savings in electrical power consumption.  Areas to
         consider are proper matching of drive motors to air fans
         and fuel pumps, interrupted-spark rather than continuous-
         spark ignition systems, and solid-state ignition and con-
         trol circuits.

REDUCING POLLUTANT EMISSIONS

When clean fuels, such as natural gas and No. 2 fuel oil, are burned
 in residential space heating equipment, two classes of combustion-
generated air pollutants are emitted, viz., incomplete combustion pro-
ducts (CO, UHC, and smoke) and oxides of nitrogen.  The two classes
are similar in some respects; their concentrations in the flue gases
are usually very low, so that their production does not affect signi-
 ficantly combustor performance or thermal efficiency, and their con-
 centrations usually are not particularly close to those representative
of thermal equilibrium, but are more likely to be determined by poorly
 controlled features of combustion, e.g., regions of poor mixing, spray/
 combustor wall interactions, size and strength of recirculation eddies,
                                 11-86

-------
and flame-zone cooling phenomena•  These characteristics tend to ob-
viate quantitative analytical characterization of the pollutant emis-
sions; as a result, such characterizations are usually studied experi-
mentally.  Conversely, the two classes are dissimilar in some respects,
principally because their usual concentrations lie on opposite sides
of equilibrium.  Thus, for example, NO production is enhanced by higher
temperatures and longer flame-zone residence times, both of which tend
to drive the hydrocarbon combustion processes to completion and reduce
carbonaceous emissions.

It was reasonably well established in the earlier oil burner studies
(Ref. 5) that an optimized conventional burner can operate reliably at
low excess air levels and produce acceptably low CO, UHC, and smoke.
Its NO emission levels, however, were a factor of 2 or more higher
than the target reduction below 0.5 g NO/kg fuel, and it was uncertain
whether that goal could be reached by the optimizing of the furnace
combustion chamber for this burner.

Therefore, the experimental investigation, which used research combus-
tion chambers rather than typical furnaces, consisted of three separ-
ate, but related parts.  One part dealt exclusively with the 1 ml/s
(gph)* conventional burner, optimized to produce low emissions, as de-
scribed in Ref. 5.  This effort sought to optimize the matching of the
fire box (combustion chamber) to that fixed-burner design.   The other
two parts branched out to burners having, in one case, forced flue gas
recirculation (FGR) to the burner air intake and, in the other case,
forced combustion gas recirculation (CGR) from the combustion chamber
to the burner air intake.  These efforts were also focused upon the
optimizing of the burner/firebox combination for low emissions.

The first of these approaches—the optimizing of the combustion cham-
ber matched to the optimum low-emission burner—was found to be the
best approach.  The experimental investigation with the 1 ml/s combus-
tion gas recirculation (CGR) burner was frustrated by an inability to
simultaneously satisfy the NO, CO, and UHC emission goals.   Because of
these problems and the inherent potential problems of moderately high
temperature mixed gases being passed through the combustion air fan,
it was concluded that further work with this burner concept was not
warranted.
*1.00 ml/s = 0.951 gph.  The 5-percent difference is neglected in this
paper when reference is made to burner firing rates.
                                  II- 87

-------
Tests of the 1 ml/s flue gas recirculation (FGR) burner were more suc-
cessful and showed good potential for satisfying the emission goals
under efficient operating conditions.  However, the actual achieving
of all the emission goals would probably be contingent upon the use of
more-complicated burner start sequence to eliminate start-spike emis-
sions of CO, smoke, and UHC which were experienced under conditions
having acceptably low NO.

Thus, it was concluded that the optimized conventional burner in opti-
mized combustion chambers is a preferred choice over both the C6R and
the FGR burners.  With this choice, the emission goals can be met, low-
excess air levels can be employed, and fewer new or peripheral problems
are likely to be encountered in the commercializing of this burner than
with the more complicated recirculation types.  Additionally, burner
simplicity will ensure lower furnace costs with this choice.  The re-
mainder of the paper describes the testing, results, and implementation
of that burner.

                             DISCUSSION

OPTIMUM BURNER/COMBUSTION CHAMBER MATCHING
EXPERIMENTS

Experimental Apparatus

The 1 ml/s (gph) optimum low-emission residential oil burner, whose
development is detailed in Ref. 5, was used during this study (Fig. 1).

The optimum burner was cyclically fired in a set of three cylindrical
research combustors.  In conformity with the testing reported in Ref.
2, the burner was fired for 10 minutes of 30-minute cycles.  The three
combustors constituted a matched set, allowing considerable dimensional,
structural, and coolant variations.  The basic approach is illustrated
in Fig. 2 for one of the chambers.  Each chamber was a 1.27 m (5.00 ft)
long, flanged section of steel pipe with a stubby, flanged, side-arm
section of the same size of pipe attached near one end.  The tunnel-
fired orientation is illustrated in Fig. 2, in which the oil burner
fits into the annular flange depicted at the left end of the chamber.
In this orientation, the side-arm was redundant and so was positioned
at the opposite end from the burner and simply blanked off.  The steel
chamber could be either lined with a refractory fiber insert to form
an adiabatic combustion zone, or left unlined and cooled by some means.
A movable, spiral-wound, finned-tube heat exchanger (shown inserted in
the opposite end of the combustion chamber from the burner) provided
effective variation of the combustion zone length.  To achieve the
side-fired configuration, the combustion chamber was simply turned end-
for-end, with the blank and burner port flanges relocated as appropriate.
                               11-88

-------
Three chamber diameters were selected such that addition of refractory
linings to the larger two would produce lined chambers having inside
diameters comparable with the smaller two unlined chambers, viz.:
Steel Pipe
ID, m (in.)
0.162 (6.36)
0.222 (8.75)
0.222 (lined)
0.279 (11.0)
0.279 (lined)
Nominal Combustion
Chamber ID, m (in.)
0.162 (6.36)
0.222 (8.75)
0.175 (6.89)
0.279 (11.0)
0.22 (8.7)
Nominal Thickness
of Pyroflex, m (in.)
0.024 (0.93)
0.030 (1.18)
In use, the major axes of the chamber were vertical, with the burner
firing vertically upward when tunnel-fired and horizontally when side-
fired.

A water-cooled heat exchanger was used to accomplish rapid cooling of
the combustion gases as they flowed out of the primary "firebox" por-
tion of any of the combustion chambers.  It was intended that the gas
temperature be quenched rather rapidly so that changes in heat ex-
chariger position (i.e., firebox length) could be readily correlated
with variations in pollutant emissions.  The heat exchanger was a
spiral-wound, finned-tube assembly designed to fit inside all three
combustors and to be positioned anywhere along the length of a chamber.
The exchanger was designed to cool the exhaust gases down to 200 to
350 C (392 to 662 F) at a maximum water coolant flowrate of 3.8 x 10~4
ta3/s (6 gal/min).  A 0.0127 m OD by 0.00124 m wall (1/2-in. x 0.049-in.)
stainless steel tubing was helically wound with 0.006 m (0.25 in.) high
by 0.00051 m (0.020-in.) thick carbon steel fins spaced at 394 fins/ra
(10 fins/in.).  The carbon steel fin material was selected over stain-
less steel for its higher thermal conductivity and nickel-chrome clad
plating provided resistance to corrosive exhaust products.  Approxi-
mately 4.3 m (14 feet) of finned tubing was required in the construc-
tion of the heat exchanger, resulting in a total heat transfer surface
area of about 1.4 m2 (15 ft2).

Several semicircular baffle plates were cut from 21-gage stainless
steel sheet, slipped between coils of the heat exchanger, and wired
in place.  Cut so their outside diameters would just fit comfortably
in the inside diameter of a chamber, the baffles ensured that the
gases passed repetitively over the heat exchanger coils and prevented
them from bypassing around the outside of the coils.
                                   11-39

-------
 The  optimum burner was  fired in the various research combustion cham-
 bers at  an  outdoor test facility whose principal components were at-
 tached to a waist-high  steel table.  A Unistrut superstructure at one
 end  of that table supported the vertically mounted combustion chamber
 and  allowed suspension  of the spiral-wound heat exchanger within it.
 The  facility was organized for rapid and easy changing of combustion
 chambers, burner orientation, and heat exchanger position.  Minimum
 protection  from inclement weather was provided by a simple sheet metal
 roof over the test apparatus.

 Experimental data requirements were primarily concerned with flue gas
 pollutant concentrations.  Most pollutant species concentrations were
 measured by conducting  a continuous flue gas sample to a train of anal-
 ysis instruments located within a nearby laboratory.  Flue gas smoke
 content was measured intermittently at the flue with a manual smoke
 meter.  The instruments used, analyses performed, and types of data
 obtained are described  and discussed in Ref. 5.

 Experimental Results

 Data were recorded during 160 tests of the apparatus previously de-
 scribed; in addition to chamber diameter, construction, and burner
 orientation variations, the heat exchanger position and burner ex-
 cess air level, or stoichiometric ratio, were varied systematically.

 Typical measured emission data are illustrated for one subset of these
 tests in Fig. 3.  In this example, flue gas pollutant concentrations
 varied fairly regularly with both stoichiometric ratio and heat ex-
 changer position, although this was not always consistently so.

 The  results of all the parametric variations tested were synthesized
 into a tabular summary  form amenable to extraction of trends and de-
 sign criteria.  That summary is given in Table 2, where the first
 three columns describe  the combustion chamber, and the next three col-
 umns list the chamber lengths and stoichiometric ratios for which the
 pollutant concentrations in the flue gas were acceptable.  Few of the
 NO emissions of the combustor were acceptably low so conditions that
minimized NO production are listed under that pollutant.  Further, the
 right-hand column of Table 2 lists approximate flue gas NO concentra-
 tions at 25-percent excess air (12-percent C02) for 0.50 m (20 in.)
 long chambers.  It was seen that, with a few exceptions, the combus-
 tion chamber designs tested generally failed to achieve the goal of
 NO < 0.5 g/kg fuel.  However, the observed trends in the data support
 confidence  that the goal can indeed be satisfied.
                              I 1-90

-------
Co^bustor Design Criteria

Combustor design criteria that can be derived from Table 2 (and the
larger body of data upon which it is based), are subdivided into cate-
gories of burner orientation, chamber size, and cooling medium in the
following paragraphs.

Burner Orientation.  Combustion chamber design requirements for ensur-
ing acceptable levels of CO, UHC, and smoke emissions were quite simi-
lar for the side- and tunnel-fired burner orientations.  That observa-
tion held for NO emissions from water-cooled combustors, as well.  How-
ever, NO production in refractory-lined and air-cooled chambers was
about 1 1/2 to 2 times as high in the side-fired orientation as in the
tunnel-fired.  Ostensibly, this latter phenomenon resulted from sub-
stantially longer gas residence times at high temperatures in the
stronger and more complicated eddies of a side-fired chamber.

Combustion Chamber Size.  Chambers smaller than about 0.20 m (8 in.)  ID
must be refractory-lined to avoid operating with unacceptable combus-
tion roughness.  Operation of all designs was acceptable in this regard
when their inside chamber diameter was 0.22 m (8 in.) or larger.

A tendency existed for larger-diameter chambers, with or without in-
sulation, to require longer chamber lengths to achieve comparable
levels of carbonaceous pollutants and, concurrently, to produce sub-
stantially lower levels of NO.  Both phenomena were undoubtedly linked
to the ingestion of recirculating combustion gases into the flame zone.
For a given burner, larger and stronger recirculation eddies can be
established in bigger chambers, reducing the combustion intensity some-
what and lowering the rates of burnout of carbonaceous species.  Also,
larger-diameter chambers have greater wall areas for convective and
radiant transmission of heat from the flame zone, which reduces peak
flame temperatures somewhat.  Increased average gas residence times
should produce opposing trends for both NO production and carbon burn-
out, which could help account for some anomalies in the data.  Also,
the steady-state recirculation and radiation effects may be partially
masked by starting transient effects, particularly with respect to
smoke and UHC data.

Short combustion chambers generally were more favorable for low NO but,
if they are too short, the fuel may not be completely burned before
combustion reactions are quenched in the heat exchanger.  A comb ustor
length of 0.5 m (20 in.) is a suitable compromise for most of the cham-
ber designs tested.  Slightly shorter chambers might be appropriate if
refractory linings were to be used.
                                  H-91

-------
Chamber Cooling Medium.  Refractory-lined combustion chambers generally
exhibited less combustion roughness and less sensitivity to starting
conditions but also produced higher NO concentrations than did cooled-
wall combustors.  The air-cooled, side-fired configuration had better
starting characteristics (but higher NO) than did the air-cooled,
tunnel-fired chamber.  Water-cooling appeared to be preferable to air-
cooling, partially because of lower side-fired NO levels, but also be-
cause of smoother starting and more consistent CO and NO emission re-
sults.  The water-cooled chambers in shorter lengths, however, were
prone to produce excessive CO.

It was inferred that the key effect of both air- and water-cooling is
the reduction of flame zone temperatures that results from removing
heat from the combustor walls.  The 0.279 m (11 in.) diameter water-
cooled combustor, which met all emission goals, had approximately 20
percent of the heat of combustion of the fuel extracted from the fire-
box.  The single "forced draft" air-cooled chamber was designed to
duplicate that heat removal rate, whereas considerably less-effective
natural convection prevailed in the other air-cooled configurations.
That forced-draft air-cooled chamber, however, had an intermediate
0.25 m (10 in.) ID.  Its flue gas NO concentrations were approximately
midway between those from the 0.222 m (8 in.) and 0.279 (11 in.) ID
water-cooled combustors, indicating that forced draft air-cooling can
be just as effective as water-cooling in controlling NO emission levels.
Optimized Combustors for the 1 ml/s (gph)
Optimum Burner
To reduce NO emission levels below 0.5 g/kg No. 2 fuel oil burned while
maintaining acceptably low carbonaceous pollutant emissions during
operation at low excess air levels (10 to 15 percent), the 1 ml/s (gph)
optimum oil burner should be fired into a combustor having the follow-
ing design attributes:

     1.  Cooled walls, with either water or forced-draft air coolant,
         which extract approximately 20 percent of the heat of com-
         bustion of the fuel from the flame zone.  It is also impor-
         tant to design for nearly uniform inside wall temperatures
         and for retention of elevated wall temperatures during stand-
         by periods.

     2.  The ID of the combustor should be 0.28 m (11 in.) or greater.
         (It is inferred that NO emission levels will continue to go
         down as diameter is increased above 0.28 m (11 in.), but this
         hypothesis has not been tested.)
                               11-92

-------
     3.  The effective combustion chamber length, from the burner end
         to the location where the furnace heat exchanger begins to
         quench the combustion gas temperature rapidly, should be
         greater than 0.5 m (20 in.) and perhaps as long as 0.75 m
         (30 in.).
     4.  The combustor may be either side-fired or tunnel-fired, which-
         ever is convenient for a particular furnace or boiler design.

CONCEPTUAL PROTOTYPE FURNACE DESIGNS

Preliminary conceptual designs were laid out for a prototype warm-air
furnace and a prototype hydronic boiler, each embodying the preferred
approaches previously discussed for improving efficiency and reducing
emissions.  Each design concept layout was discussed with one or more
representatives of a manufacturing company that makes and markets res-
idential space heating equipment.  Thereafter, the designs, the manu-
facturer's comments, criticisms and suggestions, and appropriate de-
sign modifications were reviewed with the EPA Project Officer prepa-
ratory to selecting one of the designs for fabrication and experimental
evaluation.

The preliminary conceptual designs are illustrated and briefly described,
Greater detail is given in Ref. 7.

Conceptual Design:  Warm-Air Furnace

A conceptual design layout for a prototype low-emission, improved-
efficiency, warm-air furnace is shown in Fig. 4.  It is based upon
modifying an existing, commercially purchased, warm-air furnace* to
accommodate an optimized, low-emission oil burner with sealed air sup-
ply and a large, air-cooled, steel combustion chamber.  This approach
allows key concepts to be tested without the expense and delay of de-
signing and fabricating most furnace components.  Further, it ensures
that commercially practical heat exchangers, air filters, blowers,
cabinets, and controls are imposed upon the concepts to be evaluated.

Referring to Fig. 4, the burner vestibule is a closed volume supplied
with combustion air, through a small filter, from an external, sealed
air system.  Not shown is a positive shutoff device for this air when
the burner is inactive; this device might be a solenoid-actuated but-
terfly valve in the combustion air inlet line.
*A Lennox Model 011-140 furnace was selected.
                               11-93

-------
The combustion chamber is offset slightly toward the back of the fur-
nace to accommodate its large diameter.  The chamber is comprised of a
standard pipe cap welded to an eccentric reducer whose small end is
mated to the furnace heat exchanger.  Welded to its outside are many
longitudinal fins to enhance transfer of heat to the vertically up-
flowing, coolant air.  The fins also act as a heat sink to help keep
the combustion space warm during standby periods.

Conceptual Design;  Hydronic Boiler

The conceptual design layout for a prototype, low-emission, improved-
efficiency hydronic boiler is shown in Fig. 5.  This design is based
on substantially all-new construction, principally because no commer-
cially available unit was identified as being readily adaptable to the
larger-diameter, water-cooled combustion chamber for the optimum
burner.

A horizontal, tunnel-fired, water-cooled combustion chamber was selec-
ted because it offers boiler structural and package volume advantages.
The horizontal firebox is followed by multiple sequential passes
through a water-cooled, fire-tube heat exchanger.  Commercially avail-
able Turboflue internally finned tubing is indicated for this service.
The combustion chamber and heat exchanger tubes are all attached to
the same boiler tube sheets.  Between the combustion chamber and the
first heat exchanger pass, combustion gases are reversed by passage
through an insulated ISO-degree return manifold, whose volume is con-
sidered to be part of the combustion chamber.  Similarly, there are
180-degree return manifolds between successive heat exchanger passes.

The conceptual hydronic boiler design also has the usual plumbing and
control features needed for installation and operation in a residence,
including a tankless coil for heating domestic hot water.  Needed, but
not shown in Fig. 5, is an exterior cabinet that should be well insu-
lated to reduce external heat losses.

FUTURE WORK PLANS

Following comparative assessments of the two preliminary prototype de-
signs, including their potential overall emissions and fuel economy
impact, the warm-air system was selected as the more promising; there-
fore, a prototype warm-air unit will be built for further study in the
continuing work.  The data and experiences gained will be applied to
the final design of a cost-competitive, commercially producible, warm-
air furnace that embodies the derived low-emission, improved-efficiency
technology.
                                I I- 94

-------
                            REFERENCES

1.  Hall, R. E., J. H. Wasser, and E. E. Berkau, "NAPCA Combustion
    Research Programs to Control Pollutant Emissions From Domestic
    and Commercial Heating Systems," New and Improved Oil Burner
    Equipment Workshop, NOFI Tech. Publ. 108 ED, National Oil Fuel
    Institute, Inc., New York, New York, September 1970, pp. 83-93.

2.  Hall, R. E., J. H. Wasser, and E. E. Berkau, "A Study of Air
    Pollutant Emissions From Residential Heating Systems," EPA-650/
    2-74-003, Environmental Protection Agency, Research Triangle
    Park, N.C., January 1974.

3.  Martin, G. B., and E. E. Berkau, "Evaluation of Various Combus-
    tion Modification Techniques for Control of Thermal and Fuel-
    Related Nitrogen Oxide Emissions," presented at the Fourteenth
    Symposium (International) on Combustion, Pennsylvania State
    University, August 1972.

4.  Barrett, R. E., S. E. Miller, and D. W.  Locklin, "Field Inves-
    tigation of Emissions from Combustion Equipment for Space Heat-
    ing," EPA R2-73-084a (API Publ. 4180), Environmental Protection
    Agency, Research Triangle Park, N.C., June 1973.

5.  Dickerson, R. A., and A. S. Okuda, "Design of an Optimum Distil-
    late Oil Burner for Control of Pollutant Emissions," EPA-650/
    2-74-047, Environmental Protection Agency, Research Triangle
    Park, N.C., June 1974.

6.  Peoples, G., "Sealed Oil Furnace Combustion System Reduces Fuel
    Consumption," Addendum to the Proceedings, Conference on Improv-
    ing Efficiency in HVAC Equipment and Components for Residential
    and Small Commercial Buildings, Purdue University, Lafayette,
    Ind., October 1974.

7.  Combs, L. P., and A. S. Okuda, "Residential Oil Furnace Optimi-
    zation," Interim Report for Phase I of Contract 68-02-1819,
    R-9815, Rocketdyne Division, Rockwell International, Canoga Park,
    California, draft submitted to EPA, August 1975.
                            11-95

-------
                  EXTERNAL VIEW
                  OPTIMUM HEAD
Figure 1.  1 ml/s (gph) Optimum Low-Emission Residential
           Oil Burner
                    11-96

-------
     REFRACTORY INSERT
     WPYHOF LEX LINER
     IOOU BLE THICKNESS)
       0.20 m ID    /
      q.JSm  s   -^

HEAT EXCHANGER
      I WATER
                                           \  -
                                                   ,/BLANK
                                               .   // FLAWGE/
                                            SIDE-FIRED      /
                                            iURNER PORT   /
                                        MOVABLE HEAT
                               , -^       EXCHANGER (SPIRAL'
                                        WOUND. FINNED-TUBE)
                                      - 0.2S nvOIAMETER
                                      COMBUSTION CHAMBER
                         TUNNEL-FIRED BURNER PORT
                         END FLANGE
Figure  2.   Experimental  Combustion Chamber Arrangement
X X
o
S
o
s20
X
i 10

Q
2
4
o 0
7
LU
i s
rt
5
1 3
o
CD 1


	 	 — • -| OT 15
0.40 |
0.7S 	 -BACHAHACH SMOKE • 1 •
' ' ' <
'=• fl-SD « : 0 ™
', , '' : |2
osc ow 6
\ '• \ z
• ••'' \ £
x ^\ m
V"^- ^ - r^' ' r ^ n

1.5
1
o > o
0 >S m IMT EXCHANGER POSITIONI .
1 S
1 0 50 X
1 o 40 O
\ 'ODD u 0.5

- xOx 1 ~
X^X. \^ i

-- 	 	 	 ~"\
- - - - aACHARACH SMOKt - 1

-

0.30 m (HT EXCHANGER
040 , POSITION)
075 O.JO
0.90
o.s6""*-»^ — — ^ °75 0*



_

O.n " IHT. EXCHANGER POSITION!
.•-^-.^^.-.-^^s^*"-

^ - ' 030


         1.00  1.10   1.20  1.30  140  1.60    1.00   1.10  I.JO   1.30  1.40  1.50
              STOICHIOMETRIC RATIO               STOICHIOMETRIC RATIO
 Figure 3.   Cycle-Averaged Flue Gas Emissions,  10  ml/s
               Optimum  Oil  Burner in a 0.279  m Diameter,
               Air-Cooled,  Side-Fired Combustor at Various
               Heat Exchanger Positions
                             I L-97

-------
1
U-r-^"^~^...Jl
! • ' v •
pin ....- .ill
°"cl>i
. — f - -----
! ^
E
f?
O
i

v

-------
                                                  u
                                                  •rl
                                                  G
                                                  O
                                                  M
                                                 *C»
                                                  fs
                                                 53

                                                  C
                                                  o
                                                  CO
                                                 •H
                                                  S
                                                  O
                                                 0)
                                                 ex
                                                 o
                                                 M
                                                 
-------
        TABLE 1.  TYPICAL MISSION AND EFFICIENCY PERFORMANCE OF
                        RESIDENTIAL OIL FURNACES

CYCLE-AVERAGED EMISSIONS
SMOKE IBACHAflACK SCALE)
CO
UHC
PARTICUt-ATES
NOX
THERMAL EFFICIENCIES (NET)
STEADY STATE
CYCLE-AVERAGED
AVERAGE OF EXISTING
FURNACE POPULATION
~ NO 2
~ 06G/KG FUEL
~ OOBG/KG FUEL
~ 0.3G/KG FUEL
~ 2.50/KG FUEL
& TO 75%
50 TO 60%
AVERAGE OF CURRENT
NEW UNITS
~NO. t
~ 03G/KGFUEL
--0.06G/KG FUEL
— 0.2G/KG FUEL
~2.SC/KG FUEL
75 TO 80% i
66 TO 70%
TABLE 2.  SUMMARY OF RESULTS FROM 1 ML/S (GPH) OPTIMUM BURNER/COMBUSTION
                      CHAMBER MATCHING EXPERIMENTS
CHAMBER DESIGN ATTRIBUTES
CONFIGURATION
TUNNEL-FIRED
(COAXIAL)



SIDE-FIRED
(PERPENDICULAR
PORT)






COOLING METHOD
AIR-COOLED
(STARTABLE OVER A
VERY NARROW SR
RANGE)
WATER-COOLED
INSULATED

AIR-COOLED

(FORCED DRAFT!
WATER-COOLED

INSULATED

1 D .fli
0.162
0.222
0.279
0.222
O.I 75
0.22
0.162
0.222
0.279
0.250
0.222
0.279
0.175
0.22
REQUIREMENTS FOR ACCEPTABLE (OR
MINIMUM) EMISSIONS
CO & UHC
ROUGH-BURN
Lc > 0.5 in
LC = 0.75m
Lc > 0.5 m
Lc > O.S m
Lc 2 0.5 m
ROUGH-BURN
Lc 2 0.6 m
Lc >0.4 m &
SR > 1.15
Le > 0.5 m
Lc ;> 0.5 m
Lc J 0.5 m
Lc 2 0.3 m
Lc > O.S wi
SMOKE 1 NO. 1)
NG THROUGHOUT OPE
SH > 1.05
SR > 1.1 Oft
LB >0.5m
SR >1.15
Lc >0.7S m
OR SR >1.3
Lc > 0.5 mil
SR > 1.15
NG THROUGHOUT OPE
Lc > 0.5 ni
SR > 1.15
Lc > O.S m ft
SR > 1.08
LC 2 0-5 m, SR > 1.05
Lc >O.S m &
SR > 1.2
Lc >0.5&SR J1.1
Lc > O.S m &
SR > 1.25
Le > 0.5 en &
SR > 1.1S
NO
RATING RANGE
Lc «• 0.4 m
Lc =• 0.4-5.0 m
SHORT LC'S BEST
SHORT LC'S BEST
SHORT LC'S BEST
BATING RANGE
Le - 0.75 m SLIGHT
FAVOR
SHORT LC'S BEST
SHORT Lc'S BEST
SHORT LC'S BEST
SHORT LC'S BEST
SHORT LC'S BEST
DATA MIXED
NO.g/VgFUEL
9 L • 0 .5 m &
1.25 SR
I0.6)
(0.41
0.7
1.3
0.9
0.9
0.6
0.6
0.7
0.4
2.0
1.9
                                      II-100

-------
10:05 a.m.
Integrated Low Emission
L. Paul Combs, Rocketdyne
Am I right that you still need about  20-25% excess
air for a normal operation to obtain  zero smoke number?
That chart was not typical of our current testing.
With the optimized firebox, we have designed  the  proto-
type furnace  to operate at 15% excess  air.
Have you experienced blue flame  phenomenon  in the
course of your work?
With the recirculating  gas burners, yes, we did have
blue flame condition.
Did you think it worthwhile  to consider  that  as an
extension of  your  research?
It has promise for further reducing the  NO  to levels
perhaps half  of our target here, on the  order of  a
quarter of a  gram  per kilogram.  However, it  has
difficulties.  The blue flame burners  are hard  to
start because  the  combustion chamber  is  filled with
colder gases  than  during steady  state.
Do you have to design a combustion chamber  for your
particular burner  or vice versa?  Does  there  have to
be a mating of burner design and combustion chamber
design or can  this be a conversion burner?

The optimization wor.-c. that we have done has all been
done with one gallon per hour firing  levels.  We  have
done a small amount of experimentation at different
firing rates and the indication is that the combustion
chamber can take some latitude in firing rate.  I think
it is best optimized for a particular firing  rate.
We are working on another contract in which we are
assessing the direct commercial application of the
                        £1-101

-------
optimum burner head  for use as a  retrofit device
for existing furnaces.
In the beginning of your  talk, you mentioned  that
one of your goals was  to  make  something  that  would
fit codes, and I question  the  shutoff damper  in  the
exhaust flue line.  I  rather suspect that  that has
got to be a code problem.  Did you look  into  that?
We discussed it with the  enginnering people at Lennox.
The shutoff device that we had originally  sketched
in our design they objected to pretty strenuously,
from that standpoint.  The use of a solenoid  actuated
butterfly valve they thought would have  no problem in
passing code.  It may  introduce other problems that
impact the codes.  The draft air flow through the
burner and firebox helps  to keep components such as
the electrodes, the flame  sensing cadmium  sulfide cell,
the electrical wiring, the control wiring  and the
vestibule cooled.  With that air shut off, we may have
difficulty with too high  temperatures and  need to
put in additional insulation and so on.
Your design burner is  going to operate at  15% excess
air.   Now does that represent  the excess air  level
as measured in the flue also?  Most heat exchangers
like Lennox  (I have a Lennox  in my house)  are not de-
signed to take the flame  temperatures that result with
15% excess air.  And that  is why on a normal  household burner,
there is a large amount of excess air used just because
heat  exchangers are typically not made out of a high alloy
material, and you are talking about gas to gas  heat exchangers.
Now,  you say 15% excess air gives a high thermal efficiency
if you maintain it through the whole thermodynamic boundary
                          11-102

-------
of your system.  Do you do that?
It is our intention to do that.  We are helped  in  that
regard by the  fact that we are going  to remove  20-25%
of the heat from the combustion chamber so  that  the
combustion gases will be at least partially cooled
by the time they hit the heat exchanger.
When you say you hope to, this means  that you haven't
been able to do it in your prototype  furnace?
We haven't built a prototype  furnace.  We have  de-
signed it and  it is being built now.
[Moderator]
Since we shut  off the questions on Paul's presentation,
he has agreed  to endure double jeopardy; and if  Rix Seals
and the other  gentlemen are still here and  would like  to
ask their questions, we would  welcome  them at this  time.
Paul, what was the final  temperature, not after dilution
but after leaving the heat exchanger on the integrated
furnace setup?   I understood it was operating  at  15%
excess air but I couldn't see the chart from the  back
and I wasn't sure what the final temperature was.
The chart didn't tell what the  final temperature  is.
We anticipate  that we will probably have  flue  gas
temperatures on  the  order of 400° F.  We have  run
some computer  analyses of the behavior of the  existing
Lennox compact heat  exchanger that we will  use down-
stream with  the  combuster but our analyses  are a  little
uncertain as to  what is going to happen.  That's  part
of the testing that  will  be performed, to characterize
the behavior of  the  firebox within  the combustion
chamber and  the  heat exchanger  combination.

                          11-103

-------
The second point that I wanted to ask is not intended
controversially or that sort of thing; but I quite
often see and hear comments regarding the advantages of
using an outside air source and I won't attempt to
dispute that there are energy savings.  However, have
you in your consideration or percentage assignments
or evaluations of this taken into account the fact
that the load on the heat exchange or combustor system
is increased when it is looking at colder combustion
air than when it is looking at inside combustion air
(let's say the 70 degree differential that might exist
which in turn might distort the apparent advantages
of using outside air)?  I might re-express that:  you
are raising the same mass or volume of combustion air,
but you are raising it, say, in case  (1) if it were
outside design at zero you might be raising it from zero
to 400 or 500 final exit.   Whereas in case (2) using
inside air one might be raising it from 70 to 400 or 500
outside.   Has that been allowed for in your evaluation?

We ran a substantial number of computer analyses
simulating the furnace system that I didn't talk about
before.   The analytical portion did include examination
of the effect of varying the combustion air temperature and
it does  have a very modest impact on the furnace thermal
efficiency.   I might say that there is a several hundred
page interim report that has been submitted to EPA
in draft form, and if things go according to plan
that will be available toward the beginning of
                        II- 104

-------
    next year,  for those who want more detail on that material.
    I didn't point out at the beginning of my presentation that
    I do have co-authors.  One sent me a note saying I should
    have mentioned that we did run a preliminary test on the
    finned firebox and found that as opposed to the cylindrical
    combustion chamber I illustrated, it was capable of removing
    the heat required and that the emissions were surpressed as
    predicted.
Q:  I would like to offer one comment to clarify the question
    Rix just asked you.  The comment regards the relationship
    between the air that the furnace requires and the contribution
    of that air to the building infiltration.  Now in your analysis,
    if I heard it correctly, you assume that all the air that
    the furnace needs for combustion and for draft control comes
    sooner or later from outside.  Rix wondered what happens if
    you compare that with the other situation where you have to
    increase the heating for the same mass of air through a larger
    differential.  Now, to add to the confusion, we did some
    measurements last heating season, and it turns out that not
    all the air that the furnace needs for combustion and for
    draft control comes directly from outside, or, in other words,
    contributes infiltration.  Typically between 30% and 100%
    of that air contributes infiltration.  Meaning that some of
    the air that goes to the furnace is really coming from a
    reduction in the exfiltrating air through the cracks near
    the attic where you have a higher inside pressure compared to
    the outdoor pressure.  In some houses, depending on how they
    are built, whether one or two stories, and depending on how
                               11-105

-------
     leaky  they  are,  all  the  air  that  Is  required  by  the
     furnace  comes  from  increased  infiltration.  If I  didn't
     know anything  about  a  house,  I would  guess  that  about
     70% of the  furnace  air requirements  contribute to  in-
     filtration.  A year  ago  when  I didn't  know  anything
     about  that,  I  assumed  like you did,  that  100% is  a
     contribution.  The  question  regards  the effect of
     combustion  chamber  size  on smoke  and  NO increase  or
     decrease, and  I  think  your conclusion  was as  you
     increase combustion  chamber  size  your  NO  generally
     goes down.   Is the  same  true  for  smoke or was the
     smoke  going  the  other  way?
     It is  true,  there is a tendency that  as the combustion
     chamber diameter is  increased to  surpress the NO
     formation, a somewhat  longer  chamber  is required  to
     burn out the products  of incomplete  combustion.   CO
     and the unburned hydrocarbons did go  up and so we
     would use a  longer combustion chamber  to keep them
     down.
     I see; so really you have to maintain  the volume,
     so to speak, to maintain the  soot burnout the same;
     but by reducing  the diameter you  can  reduce the NO.
     Is that correct?
    Yes, that is correct.  Concerning your comment, I
    might say that we have not tried  analyzing  the
    residence itself.  Our analysis has been restricted
     to the furnace unit and  the inputs and outputs that
     it experiences.
Q:   When you take the combustion air from outside the
    building, and, of course, assuming that you possibly
                             II- 106

-------
    could have an installation of this type in quite cold
    climates, would not the inlet air temperature have
    some effect on your combustion process, and possibly
    your emission levels?  Also, since the device has a centri-
    fugal fan, the actual air temperature and the consequent
    density is going to determine somewhat the fan output pres-
    sure, which will somewhat affect the combustion air flow
    throughout the system.  Would you have any comment on that?
A:  You are right, there will be effects.  The thermal
    effect is really pretty marginal.  I don't recall, nor do
    I have here, the graphs of the decrease in thermal efficiency
    as a result of the colder air supply.  I don't know about
    the emissions; it is conceivable that there might be a
    slightly higher starting spike on the CO and unburned hydro-
    carbons.  The colder air temperatures, as the airflow is
    established before ignition is established, will tend to
    chill the combustion chamber more, and it could conceivably
    have a starting spike effect.  Yes, I'm sure it will.
                             11-107

-------

-------
  THE CONTROL OF POLLUTANT EMISSIONS
                 FROM
       OIL  FIRED PACKAGE BOILERS
                  By:
M. P. Heap, T. J. Tyson, E. Cichanowicz
          Ultrasysterns, Inc.,
                  and
     R. E. McMillan, F. D. Zoldak
 The Foster Wheeler Energy Corporation
               11-109

-------
                          ACKNOWLEDGEMENTS









     The authors wish to acknowledge the help and guidance given by




Mr. G. B. Martin, EPA Project Monitor for the current projects and




all the members of the EPA-API Steering Committee.  We are grateful




to Mr. W. Axtman of the ABMA for his assistance on various matters




concerning package boiler practice and boiler sales data.  Mr. McComis




was responsible for all the laboratory investigations and the authors




gratefully acknowledge the help of Mr. R. E. Sommerlad of the Foster




Wheeler Energy Corporation.  Dr. L. J. Muzio was responsible for




Phase I and II of the EPA-API program.
                             11-110

-------
1.0    INTRODUCTION




       A package boiler is described in the ABMA Lexicon as:




            "a boiler equipped and shipped complete with fuel




            burning equipment, mechanical draft equipment,




            automatic controls and accessories.  Usually




            shipped in one or more major sections."




Although boilers with a capacity up to 250,000 Ibs of steam per hour




can be shipped as a single unit by rail or truck, larger units




(250,000 to 350,000 Ibs of steam per hour) must be modularized.  Thus




the term packaged encompasses a wide range of equipment, size range,




design type and fuel capability.  This paper is somewhat more limited




in scope since it is mainly concerned with firetube boilers and




practical experiments with equipment in the lower size range (up to




25,000 Ibs of steam per hour).  However, it is contended that the




general principles established from these experiments with regard to




pollutant formation are applicable to the complete range of package




boilers firing oil.




       The genesis for the results presented in this paper was a




recently concluded program jointly supported by the EPA and API.  The




objective of this program was to provide information on the control of




nitrogen oxide emissions from package boilers by the use of flue gas




recirculation or staged combustion techniques, and this work is currently




being continued under EPA support (EPA Contracts 68-02-1498 and 68-02-




1500).  Several workers have established the influence of the fuel/air/




combustion product mixing processes on NO  formation in turbulent
                                         X
                                  II- 111

-------
 diffusion flames £l»2 and  3^j.  Results  of  this  investigation emphasize
 that  due  regard must also  be paid  to  those design features which control
 these mixing processes  in  order  to maximize the effectiveness of both
 staged  combustion and flue gas recirculation as NO  control techniques.
 Thus, in  the ongoing projects emphasis  is  placed upon optimizing the
 total combustion system to achieve NOX  control.
       The EPA-API program consisted of three separate phases:
       Phase T-    -  Construction  of a versatile combustor
       Phase II   -  Experimental  investigations in that versatile
                     combustor to  determine  the optimum method of
                     applying both flue gas  recirculation and staged
                     combustion  to control NO   emissions
                                             x
       Phase III  -  Demonstration of the  applications of these
                     techniques  to operating boilers in the field
The results of the initial phases of this  program have been reported
elsewhere  [4,5] and the third phase has recently been completed [V] •
A decision to proceed with the field tests was made even though the
results of the experimental investigations were not encouraging.  The
results of the laboratory  investigation aimed at optimizing control
techniques have been summarized by Muzio, Wilson and McComis \5~\.
       -    a 20 percent reduction in NO   emission could be achieved
                                        X
            either by reducing the amount  of primary air or by
            increasing the atomizing air pressure;
                              11-112

-------
            maximum reductions of 25 percent could be achieved by
            deliberately staging the heat release before smoke
            emissions became excessive;
            a 30 percent reduction in emission level could be
            obtained by adding cooled flue gases to either the total
            or the primary air streams;
       -    it was possible to achieve a 45 percent reduction in
            NO  emissions by combined application of staged com-
              X
            bustion and flue gas recirculation.
       This paper briefly describes the results of the field test
and recent laboratory investigations which have been conducted both
to explain earlier results and to provide information on the optimum
design of combustion systems for package boilers.  The optimum design
must satisfy the dual criteria of minimum pollutant emissions and
maximum efficiency where efficiency relates to both thermal and
operating efficiency.
2.0    EMISSION CONTROL FROM PACKAGE BOILERS - FIELD TESTS
       The influence of flue gas recirculation on NO  emissions has
                                                    X
been determined in two package boilers of different design types
(watertube and firetube).  The influence of delayed air addition,
in order to stage combustion, has also been established in the same
firetube boiler.  This work was supported in part by the original
EPA-API program and the field tests were carried out by the Foster
Wheeler Energy Corporation.
                                11-113

-------
 2.1     Design of  the  Equipment




        The boilers  tested during  the field demonstration were both




 located in the same boiler house  and represent a compromise between




 that which was considered ideal and that which was practically




 attainable.   In selecting the field units the following criteria




 were considered to  be of particular importance:




            the units to be tested should be typical of modern




            practice.  The value of the investigations would be




            negated if the data were to be obtained on equipment




            of  outmoded design




       -    the units tested should reflect the bulk of the popula-




            tion of industrial boilers both with respect to type




            and size




       -    the units should be capable of burning both natural




            gas and residual oil




            the same oil supply should be burned in both units




       -    it  should be possible to apply both flue gas recircula-




            tion and staged combustion to the units




       -    the owners of the units must be cooperative since the




            tests could not be carried out without some interruption




            of  the normal routine




The aid of the American Boiler Manufacturers Association was




solicited in order to determine the type, size and characteristics




of the "typical" package boiler.   The ABMA were most helpful and
                                 II-ll/j

-------
provided data on the total sales of boilers for a ten year period.


This data  indicated that the bulk of the population was in the range


3,450 to 6,900 Ibs of steam per hour for the firetube design and


21,000 to  50,000 Ibs of steam per hour for the watertube design.


       A decision was made to investigate the influence of boiler


design on  pollutant emissions rather than the effect of boiler size


for the same design.  Both the units chosen for testing were located


in the same boiler house and therefore one important criteria, that


of common  fuels, was satisfied.  A complete description of the boilers


and the modifications made to allow the application of flue gas


recirculation has been given elsewhere [&"]•  The forced draft fire-


tube boiler burned both natural gas and No. 5 fuel oil and was rated

           3
at 12 x 10  Ibs per hour of steam at 15 psig.   The watertube boiler


was also designed to fire the same fuels, however, a steam rather


than an air atomized oil nozzle was used in this unit.   Although


both boilers were built by the same manufacturer,  they were fired by


different burner designs.


       The flue gas recirculation systems designed and installed in


each boiler were capable of recirculating approximately 30 percent


of the combustion products at full load.   The  recirculation was


drawn from the stack through a metering orifice of standard ASME


design and the recirculation was controlled by a damper at the dis-


charge of the recirculation fan.  The wind box of  both boilers was


breached to provide an entry for the cooled recirculation.   Front
                               II- 115

-------
views of both boilers  are presented in Figures  1  and  2  which show the

location of the  flue gas recirculation systems.


               Fon
              Flue Gas.
              Ducting "
                                          Slack
                                                Windbox
              Figure 1.  Front View of Firetube Boiler
                Showing  the  Location of the Flue Gas
                         Recirculation System
                          •Stock
                                              Orifice Plots
                                                    Combustion
                                                    Air Fon
     Flue Gas
     Recirculation
     Fon
      Figure  2.   Front View of Watertube  Boiler Showing
      the Location of the Flue Gas Recirculation System
                              11-116

-------
       Staged combustion investigations were only carried out  in  the

firetube boiler.  Figure 3 shows the general arrangement of  the

staged combustion system.  Air supplied by a separate fan was

injected radially into the firetube through eight fishtail injectors.

The axial location of the injectors (X see Figure 3) could be  varied

to determine the optimum position for staged air injection.  The

staging system was uncooled, and therefore, the staging air was

heated prior to injection.  However, the temperature of the  second

stage air upon injection was not measured.
  Combustion
     Air
                                                          Air
                                                          Supply
                                                          Manifold
                                Staged Air Supply
        Figure 3.  Details of Staged Combustion Equipment
                      in the  Firetube Boiler
                               11-117

-------
       Flue gas samples were withdrawn from the stack through a




stainless steel probe and cooled in an ice bath.  The composition of




the flue gases were monitored continuously and the NO  results
                                                     2t



reported below were obtained by an analyzer based upon the chemi-




luminscent principle.




2.2    Results




       Flue Gas Recirculation




       Figure 4 shows the influence of flue gas recirculation on




NO  emissions from both boilers at fixed excess air and three loads.
  x


Reductions of the order of 30 percent were possible in the firetube




boiler and the limit was imposed by the capacity of the recircula-




tion system.  Only modest reductions could be attained by recycling




cooled combustion products through the windbox of the watertube




boiler.  However, in this case the limit was imposed by flame sta-




bility and not the capacity of the recirculation system.  Figure 4




also shows that for all watertube boiler loads recirculation of




combustion products initially caused an increased emission.  The




addition of flue gases did not significantly affect either smoke or




carbon monoxide emissions in either boiler.
                               11-118

-------
             FIRETUBE
             86%
a
a
i
WATERTUBE

LOAD 65%
                     10             20

                        % FLUE GAS RE CIRCULATION
                                      30
    Figure 4.  The Influence  of Flue Gas Recirculation and Load
      on NOX Emissions from a Watertube and a Firetube Boiler
                      at 20 Percent Excess  Air
                           11-119

-------
        Staged Combustion



        The  insensitivity  of NO   emissions  to  the overall excess air



 during  staging is shown in Figure  5.  However, as expected, NO
                                                              JV


 emissions are strongly dependent upon burner  stoichiometry  (i.e., the



 fraction of  combustion air passing  through the windbox).  The influ-



 ence of the  location of the staged  air  injectors on NO  emissions
                                                      Jt


 can also be  seen in Figure 5.  The  fractional reduction increases as



 the injection point moves downstream.   In  the system tested, there



 appears to be an optimum burner  stoichiometry for minimum NO  emis-
                                                            X


 sions.  However, this minimum optimum level is probably peculiar to



 this system  and is probably associated  due to the presence  of



 unreacted fuel nitrogen compounds at the point of staged air



 injection.   In the laboratory experiments  increased smoke emissions



 were observed as a direct tradeoff  for  decreased NO  emissions,
                                                   X


 whereas in the field tests excessive smoke emissions were only



 observed at  low excess air levels and carbon  monoxide emissions give



 a better indication of deteriorating combustion conditions within



 the firetube.  Figure 6 compares the influence of the location of



 the staging  injectors on both NO  and CO emissions for a fixed
                                X


 burner  stoichiometry.  It can be seen that the initial NO reductions



were not accompanied by a large increase in the CO emission.  How-



 ever, with the injectors located at the furthest distance from the



 burner  the CO  emission increases significantly, suggesting  that the
                                 11-120

-------
    180
    140
    120
5?
O
E
CL
o.
X
O
    180
    140
    120
              l.25
             = l.50
             = 1.75
                                                0%
                  120
100
80
                 % BURNER  STOlCHlOMETRtC AIR

 Figure 5.  The Influence of Overall Excess Air and
   Location of Staged Air Injection During Staged
      Combustion Tests on the Firetube Boiler
                  (60  Percent  Load)
                         It-121

-------
    140

 
M
O
   300
«  200
o
E
    100
       ID
        2D                3D
LOCATION OF AIR  INJECTORS
 Figure 6.  The Influence of the Location of  Staged Air
  Addition on NOX and  CO Formation with 87 Percent of
   the Stoichiometric Air at the Burner and 4 Percent
                 Overall Excess Oxygen
                          II- 122

-------
optimum location  for the addition of staged air  in  this system  is




between 1.75 and  2.5 furnace diameters from the  oil nozzle.




2.3    Implications 'of  the giej-A Test. Results
       Only a selection of the results obtained during the investi-




gations in the two package boilers have been presented in this paper.




Baseline tests were carried out to establish the operating character-




istics of the boilers and the emission levels appear to be typical of




other units of similar size.  The influence of the various NO  con-
                                                             x


trol techniques has been established for both natural gas and fuel




oil.  Of considerably more importance than the specific results are




the implications of the tests with respect to emission control from




this class of equipment.




       Pollutant emission control techniques for package boilers




must be considered from two points of view.  Control of emissions




from those boilers already installed, i.e., retrofit and the control




of emissions from new equipment which either requires some minor




design modification or a radical change in boiler /burner design.




Retrofit would be expected to have the most Immediate effect upon




total pollutant emissions from package boilers.  One of the major




conclusions that can be drawn from the field tests carried out in




this investigation is that many problems can occur if retrofit becomes




mandatory,  which could render NO  control expensive and/or ineffective
                                .X



in certain cases unless further work is carried out to optimize the




total combustion system for low NO  and high efficiency.
                                  X
                                11-12:

-------
       The boilers chosen for testing were done so with a knowledge



of the requirements of the control system to be installed and yet



problems were uncovered that could not have been anticipated.



Changes in the geometry of the firetube boiler windbox to accommodate



the flue gas recirculation inlet resulted in the initiation of



severe pulsations at several boiler loads.  Successive modifications



succeeded in alleviating the problem but not in eliminating it.



Flame instability occurred in the watertube boiler with the addition



of greater than 30 percent recirculation at all boiler loads.  With-



out prior direct experience, neither of these practical problems



could be predicted.  There is no reason to believe that these two



boilers represent special cases and it must be expected that similar



problems would occur with other units.  The age of equipment and



the lack of available space in the vicinity of most package boilers



will tend to extend the problems of retrofit, particularly with



respect to flue gas recirculation even if it were to be shown to be



an effective control technique.


       Based upon the results of the laboratory investigations, flue



gas recirculation was as effective as could be expected In reducing



NO  emissions from the firetube boiler.  However, the recirculation
  x


of flue gases cannot be considered as a cost-effective control



technique for watertube boilers if the results of these tests are



typical.  Comparison with the data of Cato el al Q?^ suggested that



baseline emissions from this boiler are perhaps a little lower than
                             II-124

-------
could perhaps be expected, but that the emission characteristics do




not differ from other boilers of this size.  We believe that the minor




influence of flue gas recirculation on emissions is attributable to




the fact that the total NOX emission at intermediate boiler loads is




produced by the oxidation of fuel nitrogen compounds.  The generous




furnace volume and low combustion intensity minimize thermal NO




formation.  Consequently, reductions in flame temperature will not




significantly influence emissions.   The slight increase in emission at




low recirculation is probably caused by improved air/fuel mixing due




to the increased burner pressure drop.  It may well be that minor




changes in burner design would result in an improvement in the effi-




ciency of flue gas recirculation.




3.0    LABORATORY INVESTIGATIONS




       The overriding objective of the continuing laboratory investi-




gations is to establish the optimum combustion conditions necessary to




ensure maximum thermal efficiency with minimum pollutant emissions




(EPA Contracts 68-02-1498 and 68-02-1500).  Having established these




conditions, a combustion system will be designed, constructed and opera-




ted to demonstrate that they are compatible with the successful opera-




tion of a commercial system.  It is not the intent to write a "text" on




burner design or to manufacture a commercial burner, but to provide




proof of the practical viability of the concepts of low emission, high




efficiency oil burners for package boilers.




       A burner can be described simply as a piece of equipment which




allows safe conversion of chemical energy contained in a fuel into
                                1 I-125

-------
usable thermal energy.  In a package boiler this requires as a



minimum the creation of conditions which ensure flame stability and



complete combustion without accelerating the mechanical deteriora-



tion of any of the boiler/burner components.  These necessary con-



ditions can be created by suitable hardware design to produce the



optimum fuel-air-combustion product mixing patterns within the



combustor.  The practical choices available to the burner designer



are the geometry of the burner, the method of fuel delivery and the



condition of the combustion air at the burner throat.  Prior to the



necessity for NO  control the major problem of ensuring complete
                J\


combustion was solved by efficient fuel air mixing based upon the



combustion engineers dictum that mixed is burned.  However, for fuels



containing bound nitrogen, rapid fuel air mixing maximizes nitric



oxide production.  The objective of the laboratory investigations is



to determine those design options which will provide a controlled



rate of fuel air mixing which will minimize NO production and yet



ensure complete combustion at low overall excess air levels.



       In the investigations in the field, fuel rich combustion was



achieved by dividing the total air supply and injecting some part



of that air at a physical location remote from the burner.  In the



field this exercise was more successful than the original laboratory



investigations.  The major problem in the laboratory was that smoke



emissions increased in almost direct correspondence to the reduction



of NO.  Smoke emission from practical flames is the net result of



excessive carbon formation and poor carbon burnout.  Staging the
                                 11-126

-------
combustion process will undoubtedly cause an Increase in the mass




of soot produced, and therefore, it is essential that optimum con-




ditions for carbon burnout are provided after injection of the second
stage air.  Lee, Thring and Beer
                                     showed that in the temperature
range 1300-1700 K the rate of carbon combustion in gm/cm  of soot
aurface can be represented by





       q  =  1.085 x
4 pn  I"1/2 exp (-39300/RT)
                          0,
                                                                 (1)
where
        0^ is the partial pressure of oxygen (atm)





         T is the temperature (°K)




         R is the universal gas constant (cal/mole K)




Equation (1) illustrates the importance of both oxygen concentra-




tion and temperature on the rate of soot burnout.  Consequently,




even without increased soot formation, soot emissions will increase




with staged combustion if the second stage air is mixed with the




carbon when temperatures are too low or the addition of the staged




air chills the flame products.




       In the laboratory investigations it was found that NO emis-




sions were relatively insensitive to the level of staging when the




second stage air was injected from the combustor axis [_5J.  If the




second stage air was added either from the axis or from the com-




bustor walls too close to the burner NO emissions increased.  Species




concentration maps have been obtained for certain burner conditions
                               H-127

-------
 in an attempt  to explain the  performance  of  the modified commercial




 burner used in the  Phase II laboratory  investigations.  The samples




 were withdrawn from the  flame with  a  cooled  stainless steel probe.




 Figure 7  shows a schematic layout of  the  combustor illustrating the




 relative  position of  the fuel injector, the  air streams and the




 sampled field,  and  Table 1 lists the  burner  conditions for the tests




 which will  be  discussed  in this paper and CL, CO and NO concentra-




 tion contours  are shown  in Figures  8, 9 and  10.  The burner used in




 these investigations  has been described in detail by Muzio ^5^.  It




 was  a commercial  burner which was modified slightly to provide




 independent  control of the air distribution  in the primary and




 secondary streams.  The  original combustion  air damper was retained




 to vary the  swirl intensity of the  secondary stream.




       The  "baseline" conditions for  the laboratory investigations




 correspond  to Test 3, the combustion  air is  split equally between




 the  primary  and secondary flows.  Figure 11  compares 0 , CO and NO




 concentration profiles at the exit  of the burner divergent (x/D =




 0.45)  for the unstaged baseline case  and a staged flame, Test 6.   In




Test  6, 20 percent of the total air flow was added downstream of the




burner through sidewall injectors while maintaining the 50:50 split




between the  primary and secondary air flows at the burner.   The most




remarkable feature of the unstaged profiles is the high axial oxygen




concentration  (almost pure air)  so remote from the burner throat.




This  suggests that the ignition front is situated well downstream of
                                  H-128

-------
S  2
S  u

<75  'Jf

o   -

O  -o
"Z:   
                     o
             -O   "3
                             «w
                              O
                              c
                              o
                              rt
                              o
                              o
                             (J
                                                                         00 O



                                                                         o n

                                                                         en G
                                                                            M

                                                                         O PQ
                                                                         4J
                                                                         tfl U
                                                                         3 X

                                                                         i"
                                                                         o *o

                                                                            Hi
                                                                         O  M
                                                                         4J  O
                                                                         rt  *J
                                •I-)
                                 c
                             M-l  M
                              O  -rt
                              a *a
                              O  0)
                              >s 00
                              tG  rt
                                 §
                              0  -H

                             C/3  CO
                              2
II- 129

-------
3
pa
 o
u

•a
 0)
-H
£
 o
•H
 O
u
 (1)
 .-1

 •a
 H
>.

i^
•H
<
03 >^
01 *^
(1)
O
3
4-1 M
3 J3
O '•••1^
&^
M 4J
«
4J
B)v£)
0) O
P3 ^H
So
cu ^
H ^
X
a
00

o
00
CS CM
. 	 ,
^H *~^
n



r- tx r>.
rH i— 1 f- ^



-*-*-*
• . •
rn ro c^
CO -C vO

J3
a
00

o
o
1— 1



m
00
OO
CN



i


CN
CN

o
m
o
in



r»
>— i



-a-
•
^V"S
* )
00

                             11-130

-------
     
     IU
                                                              O)
CC   Ul
3   I-

I


8

z
UJ
X
o
     UJ
                                                    /^V"
                                                             00
                                                             CO
                                                             CO
                                                                  o:

                                                                  ^

                                                                  UJ
Q

2
3


9
V)
z
UJ
S

Q
                                                                         03
       e
       o
       o

       c
       ID

       S^'
       &*'
       O
                                                                         3
                                                                         Efl
•H

4-1
in
         (U


         •H
         O
                                                                          4-1

                                                                          CO
rt
      00

       O)
       h

       a
      •H
                                    nvixv
                                  II- 131

-------
      fO


      H
      W
      UJ
                                                                 VO.J
                                                                       oo
      *

      I-
      «/>
      111
to
cc

I

o
u
      UJ
      00


      
      Ul
                                                                       O   CC
                                                                            UJ
                                                                            u
                                                Q



                                                <
                                            00
                                                CO
                                                UJ
                                                                            00


                                                                            Ul
                                            O



                                            00
         to
     (VJ


3DNV1SIQ  1VIXV
 o
J2
 l-l
 cd
u

t)
 
-------
                                            ro
                                                                                                          CO
5f

h-

UJ
                                       a:
                                       z
                                       o
                                       u
                                       o
                                       z
                                                                                                 8 —
                                                                                                          o


                                                                                                          CO
                                                                                                               UJ

                                            vt
                                            UJ
                                                              CO
                                                                   en
                                                                   to
                                                                   UJ
                                                                                                               01
                                                                                                               2
                                                                                                               UJ
                                                                                                               2
                                                                                                                       0)
                                                                                                                      •o
                                                                                                                      •H
                                                                                                                       a
                                                                                                                      •H
-H
2

•a
cu
M
3
to
a
a>
                                                                                                                      o
                                                                          at
                                                                          $-1
                                                                                                          00
                                                                     (M
                                                                 3DNV1SIO 1VIXV
                                                                              TI-133
L

-------
                                                            O

                                                            It

                                                            O

                                                            "*

                                                            +J
                                                            a)


                                                            i
                                                            c
                                                            ct)

                                                            O
                                                            co
                                                             CS
                                                            o

                                                            in
                                                            O


                                                            g
                                                            •H
                                                            4-1

                                                            43
                                                            •H
                                                            M
                                                            4J
                                                            CO
                                                            •H
                                                            Q

                                                            iH

                                                            .2
                                                            T3

                                                            j
                                                            oo
                                                            •H
• NOIiDVMJ
               H-134

-------
the fuel nozzle and that the primary air hardly participates in the

combustion process to that point.  Muzio et al [_5~^ reported that

increasing the swirl intensity by changing the position of the

secondary air damper had almost no influence on pollutant emissions.

Species measurements indicated that the mixing process was almost

unaffected by the position of this damper, e.g., the axial oxygen

concentration was independent of "swirl level".  The primary effect

of reducing the combustion air flow through the burner throat in

Test 6 appears to be to slightly increase the carbon monoxide con-

centrations and to reduce the axial oxygen concentration.  Reduced

throat velocities may well allow the ignition front to move upstream

closer to the axis, thus consuming more of the primary air.

       Figures 11 and 12 emphasize the fact, that in the initial

stages of heat release, the major difference between the staged and

unstaged flames is the participation of the primary air stream in

the combustion process.  Figure 12 plots the radial distribution of

three derived properties:  N, the degree of oxidation, M the

stoichiometric mixing factor, and K the mass of NO produced per mass

of carbon.   These properties are defined as:

             	mass of reacted oxygen	
       N
       M  =
mass of reacted oxygen plus mass of oxygen
      necessary to complete reaction

/mass of oxygen\
\ mass  of  fuel /  flame point
/mass of oxygen\
             \  mass of fuel
                              stoichiometric
                                    II-135

-------
K> (0
o
 n
o
                                                  0)
                                                  0)
                                                  
-------
       K  =
                        mass of NO
             mass of fuel reacted  (CO + CO )
A more detailed account of the calculation of these quantities from




flame data has been given by Hemsath [^9j|.  It should be noted that




these derived quantities relate to time average values.  When N = 1




combustion is complete and values less than 1 give an indication of




the completeness of reaction at that point.  In a perfectly mixed




situation (assuming mixing controlled heat release) N would be equal




to 1 whenever M = 1.  M = 1 indicates that the fuel and oxygen are




present in stoichiometric proportions.




       Over a large proprotion of the radial field at x/D - 0.71 the




degree of oxidation is similar for the staged and unstaged flames;




however, the bulk of the unreacted fuel shown by the carbon monoxide




concentration contours presented in Figure 9 lies off the axis and in




this region the degree of fuel air mixing, as shown by M, is almost




the same for the staged and unstaged flame.  This is only true if




the unmixedness of the two conditions is similar, M = 1 could occur




and the unmixedness be such that at some instant in time there could




either be all fuel or no fuel, and the two would never be mixed on




a molecular scale.




       Test 6 with 20 percent staging showed a reduction in flue gas




NO emission of only 18 percent compared to the baseline level.  At




the exit of the divergent and at x/D = 0.71, actual NO concentrations




are higher for the staged case than the unstaged case (see Figures 11
                                  11-137

-------
 and 13)  but the  mass  of  NO per mass  of  fuel  reacted is very similar




 (see Figure 12).




       It  is contended that the method  of  second stage air injection




 should satisfy the  following criteria:




       —    reaction  of  fuel nitrogen intermediates should be




             complete, otherwise the  addition of air will allow




             the  formation  of fuel NO;




       —    rapid mixing of the staged  air and the primary




             products  should take place  to  give the maximum




             opportunity  for carbon burnout.




 Second stage air was  injected at x/D =  2.2 in Test 6 by which time




 reaction was  almost 80 percent complete and  the unreacted fuel was




 concentrated toward the  axis of the  combustor (see Figure 14).  The




 sidewall injectors were  constructed  to  promote swirling second stage




 air  flow £5^.  This construction directed  the air jet away from the




 axis  of  the  combustor.   Thus the air jets  never penetrated to the




 combustor axis where  the bulk of the unburned fuel lay.  Mixing




 between  the  second stage air and the partially combusted products




 of the primary zone was  poor.  Axial oxygen  concentrations shown




 in Figure 15  for two radial  locations graphically illustrates this




 point.   Because combustion was almost 80 percent complete by




x/D =  2.2, it was concluded  that improved mixing of the staged air




would have no influence  upon NO emissions but that smoke emissions




might well be reduced.   Consequently, minor modifications were made




 to the sidewall injectors to test this hypothesis.
                                H-138

-------
        X

    #  I
O  D
                                                       CO
                                                       GO
                                                             QC
                                                       *    5
o

 H

p

 K

4J
 «

Q
                                                                   TJ

                                                                    CB
s
t 9
in
O
0 ^
§
K
^

-------
I-
U)
Ui
                                                                  CO
                                                                       tr


                                                                       8

                                                                       I
                                                                       in
                                                                       o


                                                                       <
                                                                       o
                                                                       <
                                                                       cc

                                                                       in
                                                                       in
o
tn
z
UJ


5
                                                                  CNJ
                                                                               O
                                                                               •H
                                                                               •H  00
                                                                               X  n)
        0) T3
        o
                       N NOIJLVQIXO dO 33U93Q
                                         11-140

-------
   20
    15
 fM

O
ui
s
O
10
Burner Air
% Stoich.
t.17
0.93
Rodiol Position
r/R=0
O
•
r/R=0.7l
D
•
                        I                  2


                  DIMENSIONLESS  AXIAL DISTANCE
  Figure 15.   Axial  Distribution of 0  Concentration at r/R = 0

                 and 0.71 for Test 3  and Test  6
                                H-141

-------
       New sidewall  injectors were constructed to inject the staged




air directly toward  the combustor axis and Table 2 compares emissions




with  the old and new injector designs for various staging levels and




injection locations.  With only 20 percent of the total air added




through the staging  injectors, the design of the injector had no




effect upon the NO emission.  However, for the two cases where the




injection took place close to the burner, smoke emissions were




reduced.  When 40 percent of the air was staged, then both the smoke




and NO emissions were influenced by the injector design.  With the




new injector design NO emissions were higher when 20 of the 40 per-




cent staging air was added at x/D - 2.2.




       After consideration of the results of these laboratory investi-




gations the staging system for the field tests was designed with




radial staged air injection.  The number of injection points was




also increased from four to eight in order to promote mixing.  An




attempt was also made to prevent flame products from impinging upon




the firetube walls, which would cause local chilling.




       Several measures were taken to try to influence the mixing




conditions in the new burner field in order to improve staging per-




formance using the sidewall injectors.  These measures included




varying the atomizing fluid from air to steam, varying the pressure




of the atomizing fluid , air distribution across the throat, nozzle




size and swirl distribution at the burner throat.   The objective




at this time was merely to demonstrate that staging could be more
                              II- 142

-------
 tn
 a
 o
•rl
 (0
 CO
•H

I

-u
 G
 n)
4J
 O
PH

 C
 O
 CJ
 01
 M
3
 60
 cfl
 4-1
 O

 4-1
 HI
S

14-1
 o
 OJ
^=
H
CM

 cu
rH

•s
H
H
CO
CJ
c/1

A
O
CO
)_l
(0
o
CO
rt

•
0

_£!
o
$

T3
0)
V4
3
CO
CO
QJ
a
CO
B
CX
p.
O
!S

(D
T3 03
•O >~i
•* 2
M U
•rl Q)
T-t M
CO
4-* i — 1
O iH
H etJ
IH (U
O *O
•H
QJ C/3
60
4J 00
C 3
QJ O
U »-l
41 H
00
•rl
U)
(U
O

I
"Z.




c
00
•H
41
O


CM CM
^•*- ^"*.
f-l r-l

C1") CO 00 ^4D U">









s^* l/^ OQ QQ f"^
"O ;

00
0)
Q
!n
S5

00
•rH
M
CJ
O
r-t
0
CM
ITI
II
^1 o

VD

CM

II
*/ 1 /-i
?S | f— 1


CM
CM
11
*|Q

CM O O CM OS
i— I O vO 'A 00
CM CM iH i-l i-l





CM i-l i-» iH i-H


1 1001
CM CM






1 O 1 0 0
CM CM CM






0 1 1 1 O
CM CM
 3
 4-1
rt
 3
 Cu
 C
M
                                                                                cfl
                                                                                0)
                                                                                V-l
                                                                               •H
                                                                               
-------
effective than had been demonstrated earlier and to try to approach




the results reported by Siegmund and Turner £lO^.  Figures 16 and 17




illustrate some of the efforts made to minimize both NO emissions




and smoke.  For a given atomizing air pressure the minimum NO emis-




sions appeared to occur between 20 and 40 percent primary air.




Staging was most effective at low primary air flow rates.  The best




results (i.e, low NO, low smoke) were achieved by fitting a 15  vane




swirler to the oil nozzle to rotate the primary air in the opposite




sense to that imposed on the secondary air by the damper.  45 percent




reductions in NO emissions from the "baseline" conditions with a




smoke increase of 1.5 on the Bacharach scale to 5.5.  Even so, this




cannot be considered as satisfactory performance.




       The emission performance of the modified commercial burner




was found to be very sensitive to atomization parameters:




       •    Nozzle size - Figures 18 and 19 illustrate that




            although the trends were similar, smoke and NO
                                                          X



            emissions as a function of primary air percentage




            are dependent upon the capacity of the oil nozzle




            at a given firing rate.




       •    Atomization fluid - Figure 20 illustrates the effect




            of changing from air to steam as the atomizing fluid




            on NO  and smoke.
                 x
                                    11-144

-------
300r
260
•7,  220
o

s«
10


E
o.
Q.

 X
O

2  180
140
                                                           *
           i
Range high to low

secondary swirl
Burner
Modified
Modified
plus
-15°
vanes
% Staged
0
20
0
20
35

•
a
•
o
A
           20
             40



           % PRIMARY AIR
                                              60
80
  Figure 16.  The Influence of Primary Air Percentage and

          Swirl Level on NO Emissions (17 Percent

          Excess Air 3.4 x 104 Btu/hr Heat Input)
                             II-145

-------
   I0r
   8
o

LU
*C

§ 6
10
CD
Burner
Modified
Modified
plus
-15°
vanes
% Staged
0
20
0
20
35

•
D
•
O
A
             20
  40



% PRIMARY AIR
60
80
        Figure  17.  The Influence of Primary Air Percentage and

              Swirl Level on Smoke Emissions  (17  Percent

               Excess Air 3.4 x 10^ Btu/hr Heat Input)
                             11-146

-------
   400
   360
   32O
o1
10

ex
0.
 X
o
   280
   240
   200
              Nozzle Size
                 gph
                 60   A
                 80   D
                too   o
              20
 40             60

7o PRIMARY AIR
80
     Figure 18.  The  Influence of Primary Air Flow and Nozzle Size
                 on NOX Emissions  from Fuel Oil Flames
                (3.4  x  106 Btu/hr 17 Percent Excess Air)
                              11-H7

-------
                                      Nozzle Size
                                        gph
                                        60  A
                                        80  D
                                        100  O
                           40

                          % PRIMARY AIR
60
60
Figure 19.  The Influence of Primary Air Flow and Nozzle Size
           on Smoke Emissions from Fuel Oil Flames
           (3.4 x 106 Btu/hr  17 Percent Excess Air)
                       II- 148

-------
   400r
o

1=

a
x
O
   360
   320
280
240
200
                                                      10
                                                      8
                                                         d


                                                         UJ
                                                         to



                                                      "I
                                                         CO
              10      20     30      40

                         % EXCESS AIR
                                         50
   Figure 20.  NO  and Smoke Emissions as a Function of Excess

               Air Level and Atomization Medium
                         II-149

-------
       •    Pressure of the atomizing fluid - it had previously been




            reported £§] that increasing the air pressure and there-




            for the mass flow of atomizing air decreased emissions.




            The results presented in Figure 21 show that this trend




            holds for larger nozzle sizes and for steam atomization.




The reasons for the observed emission characteristics for steam are




not known.  However, species concentration maps give an indication




why increased atoraization pressure and nozzle size affect the




pollutant emissions.




       Comparing the iso-concentration contours shown in Figures 8, 9,




and 10 for Tests 3, 4, and 8, it is apparent from the oxygen contours




that the fuel air mixing process is strongly dependent upon the oil




nozzle parameters.  Increasing the pressure from 20 to 28 psig produces




more fuel rich conditions in the initial stages (see the CO contours).




Changing the size of the oil tip has a much more radical influence on




the concentration fields close to the burner nozzle (see Figures 21




and 22).




       The investigations with the modified commercial burner served




to emphasize the strong dependence of pollutant formation in practical




flames on those parameters which control fuel air mixing processes.




The burners used in the laboratory investigation and the field test




were of different design and therefore it should not be expected




that optimum staging conditions defined in the laboratory would also




apply to the field boilers.  Consequently, there is an urgent need
                          11-150

-------
 fvi
o

5?
JO


 ex
 Q.
 X
O
    320r
    300
280
260
    240
    220
           O


           D
Air


Steom
                                                           o
                                                           z.

                                                           L'J
                                                           •X.
                                                         40

                                                           to
                                                           CO
                                                         O
        8      12      16      20      24       23      32


                 ATOMIZATION PRESSURE PS!G


   Figure 21.  The Influence of Atomization Fluid  and  Pressure

               Pollutant Emissions from Oil Flames

            (17 Percent Excess Air, 3.4 x  106 Btu/hr)
                           H-151

-------
       20 -
   UJ
   O
                                O  Test  3
                                D  Test  4
                                A  Test  8
Figure 22.  The Influence of Atomization Parameters on
       Radial Oxygen Distribution at x/D = 0.45
                      11-152

-------
 for experiments  to  define optimum mixing patterns before devising



 the most practical  approach  to produce  them  in staged  combustion



 situations.



       As part of the  initial stage in  the production  of this type



 of information a versatile test burner  has been constructed which



 allows the distribution of both axial and tangential velocity across



 the throat of the burner to  be varied continuously.  The air delivery



 system consists  of  four separate plenum chambers which deliver air to



 four annualar channels.  Details of this test burner are shown in



 Figure 23.  Rotation is imparted to each of  the air streams by vary-



 ing the inclination of the air nozzles  and rotation in a clockwise



 or counterclockwise sense is possible.



       It is known  that the  characteristics  of swirling flames are



 dependent upon the  size and  intensity of internal recirculation



 zones £l]].  The  static pressure field generated by the swirling flow



 strongly influences both of  these properties.  The radial distribu-



 tion of static pressure P  is given by
                         s
        dr
             =  P
                 W
(2)
where
       p is density



       W is tangential velocity



       r is radius
                              II- 153

-------
       Operating Handle
       Counterclockwise
          (CCW)Mode
- Clockwise (CW! Mode

     - Elbow with turning and guide vanes
                     Exit Channel
                     (Annulus)
     Test
   Comtxnlor
                                                  Air Inlet (Typ.)
                                                  Channel 4
                                                            Gland for Fuel
                                                             Injector
                                                          Channel t
                                  Channel 2
Figure  23.   Multichannel Continuously-Variable-Swirl Test  Burner
                                   II- 154

-------
Consequently, the distribution of static pressure can be varied by



controlling the radial variation of tangential velocity at the



burner throat.  The multichannel test burner allows the distribution



of tangential velocity to be varied.  With rotation on the inner of



the four channels the tangential velocity distribution approaches



that of a free vortex, i.e., W = C/r.  Conversely, distribution



given by W = c'r, solid body rotation can be produced by applying



rotation to the outer channels.



       Figures 24 and 25 present initial results from this investi-



gation.  NO  emissions are plotted against the dimensionless swirl
           X


parameter S  for two different size steam atomized nozzles.  S  is
*          P                                                  P
given by
       S   =  2 S
        p        n
where S  is the swirl number associated with the flow in each channel
       n
              G   r.

               xl  1
where
       G~ is the axial flux of angular momentum




       G  is the axial flux of linear momentum



        r is the outer radius of that channel



and subscription 1 refers to channel 1.



       The results presented in Figure 24 and 25 were obtained with



a refractory parallel burner exit.  S  is independent of air
                              11-155

-------
r
            700 r
%A1R 01 ST.
1
50
40
2


3,4

30
50
30
ROTATION
1
20
X
X
Z


3

X
40
4
X
X
X
40
X

o
A
O
O
V
a
            200
               0   05    1.0   1.5    ZO   2.5   3.0   3.5   4.0   4.5    5.0   5.5   6.0


                                       SWIRL  PARAMETER



          Figure  24.   The Influence  of  Tangential Velocity Distribution on

           NO Emissions from Oil Flames (3 x 106 Btu/hr 60 gph Nozzle)
                                         II-156

-------

ROTATION
I-'
in
o
K
<
i«
^r
IO
PJ
-

0*
X X
X X

o

•H M
0 43

OJ 4-1
> «
angential
(3 x 106
H tfi
0)
O tw
r- 1
OJ PM
o
C •-!
0) -H
3 O
i-H
^ €
C 0
M (-1
M-l
0)
j: w
H a
o
•H
• (0
IT) Ifl
CM -H
                      3 O
                      00 Z
U-157

-------
distribution between the channels, however, the results suggest that
axial velocity distribution also influences emissions.  In general,
both nozzles exhibit similar trends.  Emissions are maximized by
free vortex type distributions and minimized by solid body type
tangential velocity distributions.
4.0    CONCLUSIONS
       Studies to optimize burner/combustor conditions to allow the
control of nitric oxide formation in fuel oil flames are continuing.
It has been shown to date that
       1.   Oil atomization parameters not only have a strong
            influence upon uncontrolled emissions, they also
            influence the effectiveness of staged combustion as
            a control technique.
       2.   Pollutant formation in confined oil flames is strongly
            dependent upon the distribution of axial and tangential
            velocity at the burner throat.
       3.   Field tests have demonstrated that 50 percent reduc-
            tions in NO  emission from firetube boilers are
                       x
            possible without incurring serious penalties due to
            emissions of other pollutants.
       4.   Thirty percent reduction in NO  emissions from firetube
                                          j£
            boilers firing oil containing nitrogen can be obtained by
            flue gas recirculation.   Reductions in watertube boilers
            will be strongly dependent upon combustion intensity.
                               H-158

-------
                             REFERENCES
 1.   Heap, M.  P., Lowes, T.  M.  and Martin, G.  B, "The Optimization of
     Aerodynamic Design Variables to Control the Formation of Nitric
     Oxide in Fossil Fuel Flames."  Fluid Mechanics of Combustion.
     p 75, The American Society of Mechanical Engineers  (1974)

 2.   Pershing, D. W., Brown, J. W., Martin, G. B.  and Berkau, E.  C.
     "Influence of Design Variables on the Production of Thermal and
     Fuel NOX from Residual  Oil and Coal Combustion", paper presented
     at the Annual A.I.Ch.E., Meeting Philadelphia, 1973.

 3.   Pershing, D. W., Brown, J. W., and Berkau, E.E., "Relationship
     of Burner Design to the Control of NOX Emissions Through
     Combustion Modification."  Proceedings, Coal Combustion Seminar,
     EPA-650/2-73-021, NTIS  PB224-210/AS, pp 87-140  (September 1973)

 4.   Muzio, L. J. and Wilson, R. P. Jr., "Experimental Combustor for
     Development of Package  Boiler Emission Control Techniques,"
     Phase I of III, EPA Report R2-73-292a, 1973.

 5.   Muzio, L. J., Wilson, R. P. Jr., and McComis, C., "Package
     Boiler Flame Modifications for Reducing Nitric Oxide Emissions,"
     Phase II of III, EPA-R2-292b.

 6.   Heap, M.  P., Tyson, T.  J., McComis C., McMillan, R. E.,
     Zoldak, F. D. and Sommerlad, R. E., "Control of Package Boiler
     Emissions by Staged Combustion and Flue Gas Recirculation,"
     Phase III of III, Final Report EPA API Contract 68-02-0222.

 7.   Cato, G.  A. and Robinson,  J. M., "Field Testing Application of
     Combustion Modifications to Control Pollutant Emissions from
     Industrial Boilers," Phase I, EPA Report  650/2-74-078a,
     October 1974,

 8.   Lee, K. B., Thring, M.  W.  and Beer, J. M., 1962 Combust. Flame,
     6, 137-145.

 9.   Hemsath, K., "Mixing factors and Degree of Oxydation:
     Definitions and Formulae for Computation," I.F.R.F.,  Doc.nr.
     G 00/a/l1, 1965

10.   Siegmund, C. W. and Turner, D. W., "NOX Emissions from Industrial
     Boilers:   Potential Control Methods," ASME Paper 73 1PWR10,  1973.
                              H-159

-------
    10:45 a.m.
    The Control of Pollutant Emissions from
      Oil Fired Package Boilers
    Dr. Michael P. Heap, Ultrasysterns
Q:  Could you tell us what  the fuel nitrogen concentration
    was in the oil for the water tube test?
A:  This is another sort of problem at the moment.  Some
    of the analyses were done with a Coleman analyzer
    using the Dumas method, and there is some difficulty
    in reproducibility of results.  In the past I have
    always had problems trying to analyze fuel for nitrogen
    using the Dumas method.  Using the Kjeldahl method  I
    always sort of got reproducible answers.  I believe
    that the 0.1% figure is more or less correct.
Q:
A;
It seems that exhaust gas recirculation would have
some significant thermal efficiency effects.  I was
wondering if you had extended your tests at all to
consider this in relationship to the reductions or
changes in pollutants.
This is being done at the moment but I don't have
any information on it.  It is part of the extension
of the field test that is just being carried out this
week.
It may even pay to be a low polluter in such cases.
Certainly in the case of the fire tube boiler
firing natural gas.  One of the advantages is not only
do you get the NO  down to very low levels (about
                 «V
15 ppmX but it also allows you to fire the boiler
at less than 2% oxygen, which you couldn't do without
the flue gas recirculation.  So you could say it would
                             II- 160

-------
    improve the thermal efficiency that way,
Q:  What kind of furnace heat release rates were you
    talking about in your package burner boiler tests?
    Volumetric heat release?
A:  The laboratory one was designed that way so that
    it would be typical ot a normal fire tube package
    boiler.  The water tube combustion intensities-^off-
    hand I can't say, but we did have difficulty getting
    up to 100% load anyway.  So the combustion intensi-
    ties would be low.   Most of the measurements were only
    done at less than 70% full load.
                                11-161

-------

-------
         PILOT SCALE INVESTIGATION

     OF CATALYTIC COMBUSTION CONCEPTS

FOR INDUSTRIAL AND RESIDENTIAL APPLICATIONS
                    by
             J. P.  Kesselring
                C.  B.  Moyer
               R.  M.  Kendall

       Aerotherm/Acurex Corporation
         Mountain View, California
              J. A.  Cusumano
        Catalytica Associates,  Inc.
           Palo Alto, California
                  II- 163

-------
                       PILOT SCALE INVESTIGATION
                   OF CATALYTIC COMBUSTION CONCEPTS
              FOR RESIDENTIAL AND INDUSTRIAL APPLICATIONS

           J. P. Kesselring,  C. B. Moyer, and R. M. Kendall
                 Aerotherm Division/Acurex  Corporation
                       Mountain View, California

                            J. A. Cusumano
                      Catalytica Associates, Inc.
                         Palo Alto, California
     The use of catalysts in place of conventional burners for promoting

hydrocarbon oxidation reactions appears to have advantages in the con-

trol of emissions.  The operating conditions of these catalytic combus-

tors are limited by the catalyst bed temperature capability, and the

temperature obtained through normal one-stage, low excess air operation

that is necessary for efficiency reasons is outside the current tempera-

ture capability of catalyst systems.  It is therefore necessary to con-

sider other system techniques such as bed cooling, exhaust gas recircu-

lation, or staged combustion to hold the bed temperature down.  Based

on a review of existing application and research programs involving

catalytic combustion, a research and development program to establish

design criteria for the application of catalytic combustion to low emis-

sion, high efficiency stationary combustion systems has been undertaken.

This program consists of experimental catalyst and combustion concept
                                  11-164

-------
screening, experimental and theoretical scale-up criteria, and develop-




ment of prototype systems.






INTRODUCTION





     More than 150 years ago, Sir Humphrey Davy discovered that platinum




wires could promote combustion reactions in flammable mixtures.




The resulting reactions appeared to take place on the surface of the




wires, "without flame" and with the high radiative fluxes associated




with the solid surface emittance rather than the low emittance of the




combustion products (Reference 1).  Subsequent study and investigation




followed a number of paths reviewed by Spalding in Reference 2.  Of this




work, a substantial fraction which might be termed  "fundamental studies




of heterogeneous catalysis" has attempted to identify and quantify speci-




fic heterogeneous reactions and define the mechanisms by which the solid




surface promotes or accelerates oxidation reactions.  A second fraction,




termed "computational", has dealt largely with mathematical theories




and analyses attempting to describe the diffusion of reactants to the




promoting surface, the diffusion of products away from it, and the com-




bined heat and mass transfer.  A final third fraction can be termed




"reduction to practice", which includes attempts to construct and employ




practical devices exploiting heterogeneous combustion.  As is frequently




the case in combustion, work in this area has proceeded with very little




reference to achievements in fundamentals and computation.  Experiments




throughout the 1800*s demonstrated that various materials promoted









                                  11-165


-------
 heterogeneous  combustion reactions to various degrees, and that the num-




 ber  of materials with  the ability to promote reactions increased as sur-




 face temperature increased.  The earliest successful practial devices




 exploited heterogeneous combustion to provide more efficient lamps.




 The  best example is  the mantle lantern or Auer light, still sold in great




 numbers as a "camping  lantern", in which combustion takes place on the




 ash  residue of a fiber bag mantle surrounding the fuel/air orifice.




 This mantle ash contains oxides, including cerium oxide and thorium




 oxide, which in addition to constraining radiation emission to desirable




 lines in the visible spectrum, contribute to the catalytic activity of




 the mantle (Reference 3).





     Concepts close to the original experiments of Davy have been used




as passive ignition devices and as combustible detectors, particularly




in the form of hydrogen leak detectors.   Related applications along this




line include catalytic fume abatement devices and automotive catalytic




converters, both depending on precious metals to oxidize low concentra-




tions of combustibles in gas streams.





     After extensive practical experimentation,  Bone publicized various




high temperature applications of surface combustion, using for the most




part conventional ceramic materials (Reference 4, 5).   He built and




tested surface combustion furnaces, boilers,  and permeable panels.




This line of investigation has not,  however,  proven very fruitful due




to the complexity and expense of the required equipment and the limited
                                  II-166

-------
industrial need for high radiative fluxes, although some commercial



equipment of this type is available (Reference 6).




     The range of commercial devices has recently led to some speciali-



zation of terminology in a field where formerly several terms were used



interchangeably.  The term "radiant burner" refers to burner types em-



ploying a refractory surface situated within or near the flame so as to



provide for radiant heat transfer from the flame area.  Heterogeneous



combustion occurs in these devices to some degree, but generally does



not dominate the process and homogeneous combustion constitutes most



of the reaction mechanism.  "Surface combustors" employ refractory in



a similar manner but use special provisions of air/fuel distribution to



maximize the amount of heterogeneous combustion occurring.  "Catalytic



combustors" employ special additives (usually precious metals) to en-



hance the heterogeneous activity of the surface.




     There are several reasons for the recent upsurge of interest in



catalytic combustion.  Use of the diffusion flame for combustion appears



to place a limit on the lower level of control that can be achieved



economically for clean fuels (approximately 25 to 30 ppm of NO  for small
                                                              A


sources), and catalytic combustion appears to be able to reach lower



levels of NO  control (perhaps <10 ppm).  In addition, catalytic combus-
            X


tion is soot-free, and the potential of high combustion efficiency for



the overall system is present if the excess air can be limited in some



way, such as two-stage combustion, flue gas recirculation, or cooling of



the catalyst bed.






                                  II- 167

-------
 AVAILABLE CATALYST MATERIALS





      The materials needed  for a catalytic combustion system are those




 associated with  the support, the washcoat, and the catalyst itself.




 The importance of  each of  these elements in the system is described




 below.





 Support  Materials





      The catalytic  support serves two 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.





      •   It increases thermal stability of these very small particles,




         thus preventing agglomeration and sintering with consequent




         loss of active surface.





Most common support materials are A£ 0, and Si02.   The application of




these supports in catalytic combustion presents several significant




problems that must be considered:





     •  The large space velocities used in a catalytic combustion unit




        will result in severe pressure drops if the catalyst beds are




        packed with conventional pellets like those used in the petro-




        leum industry.
                                     11-168

-------
     •  The high reaction rates of complete combustion usually present




        severe pore diffusion limitations for the effective operation




        of the system.





     •  The high temperatures and heat fluxes place special constraints




        on the thermal properties of the system.





Over the last decade, a number of new materials and new preparative




techniques have been developed to overcome these problems.  These mate-




rials have emphasized the use of monolithic support structures and sur-




face impregnated pellets.





     Monolithic supports are composed of small parallel channels of a




variety of shapes and diameters.  These structures may be in the form




of honeycombed ceramics extruded in one piece, oxidized aluminum alloys




in rigid cellular configurations, or multilayered ceramic corrugations.




The channels in honeycomb-like structures have tubular diameters of 1




to 3 mm.  The overall diameter of the monolithic support may vary from




2 cm to 60 cm, and is limited, in the case of extrusion processes, by




availability and operation of the metal die.   Materials of fabrication




are usually low surface area ceramics such as mullite (3A£,jO, •  2Si02)




or cordierite (2MgO • 5SiO  • 2A&-0,).  The refractory monolith is pro-




duced with macropores (ly - 10u) and may be coated with thin layers of




catalytic materials, 5-20 wt% coatings being common.





     The two major advantages of monolith supports for catalytic opera-




tions are the high superficial or geometrical surface area and the low
                                H-169

-------
pressure drop during operation.  Table I lists the currently available




monolith materials, and Figure 1 shows typical extruded monolith struc-




tures.




                 TABLE I.  AVAILABLE MONOLITH SUPPORTS
Manufacturer
American Lava
Corporation
Corning
Glass Works
E. I. DuPont
de Nemours
& Company
General
Refractories
Company
Norton
Company
Product
Thermacomb
AASiMag 614
Thermacomb
AASiMag 776
Thermacomb
AZSiMag 795
Celcor 9475(EX-20)
TORVEX
Versagrid
SPECTRAMIC
honeycomb
Description
Dense 96%
Alpha-alumina
Porous 96%
alpha-alumina
Cordierite
Cordierite
Alumina
Mullite
Cordierite
Mullite
Silicon carbide
RX387
Silicon nitride
RX384
Silicon
oxynitride
RX385
Temperature
Limit (°F)
2800
2192
2192
2200
2732
2462
2550
3000
3000
2800
2800
     In addition to the monoliths, highly porous pelleted materials can




also be used as the catalyst support.  Pellets come in a wide variety
                                 11-170

-------
fit ••**«•• •«»*•• •*»•••••••••••••••••••*•
••••••••••••••••••••••••••••••••••••••••
•••••••••••••••••••••••••••••••••••••••a
•••••••••••••••••••••••••••••••••••••••a
•••»••«»•• ••••••*•»••••••••*••••••••••••
•••••••• •••••••••• ••••••••••••••••••••
*•••••• »••••••»••••••• •••••••••••••»•••
!••••»••• •!•••••••••••••••••••••••••
)•••••*•••••••••••••••••••••••••••••
!•••••••••••••••**•••••••••••••••
!**•• •*••«•• •!•*»•• «••••*• •*!*••
::"'9g3$&S
I •»•»»•••••••••»•»*•••»•••*•••
••••»•»***••»•»»«««**•• •••••f
••••••••••••••••••••••••••I*
•*•*••»•«»• *•*»•••«**•» ••••»/
••*•*»•»•»•»•••»»»*• •••••
••»•*•••!• •••••»»••»*• •••
•••••*»•»•••*»•• •••••••••pr
  Figure 1.  Typical extruded monolith structures (courtesy

          of General Refractories Company).
                    II- 171

-------
of materials, shapes, and sizes, but have the disadvantage of a higher




pressure drop through the bed.






Wash Coat Materials





     The low surface area of the monolith structure requires the appli-




cation of a thin coat of oxide such as AA-O,.  This wash coat, which




strongly adheres to the ceramic or refractory support, provides a uni-




form high surface area while still insuring that the catalytic material




which is subsequently impregnated on the wash coat is close to the main




flow of reactants.  The thickness of the wash coat is usually between




10 x 10"6 and 20 x 20~6 m (10-20 microns).





     Many different materials can be used as wash coats.  ZrC^, for




example, has been used in test runs for the catalytic combustion of




hydrocarbons in automotive exhaust.  Variation in the type of wash coat




can have important consequences in the stability and life of the cata-




lyst system.






Catalyst Coatings





     Although the field of catalysis has progressed substantially over




the past decade, its theoretical aspects are not yet at the degree of




sophistication which would enable one,  a priori, to choose or design




an active oxidation catalyst with a given set of catalytic properties.




However, a vast amount of data exists which permits the development of




correlations useful for choosing promising candidates based on excellent
                               11-172

-------
 catalytic  activity and  catalyst stability.  Over the last decade there




 has been a rapid development of the art, science and technology neces-




 sary  to synthesize,  test and manufacture ultrastable oxidation catalysts




 for the abatement of automotive emissions.  These catalysts operate




 under the  most stringent conditions, and sometimes at temperatures close




 to 2000°F.  The expertise developed in this area will be useful in the




 development of a stable catalytic combustion system.





      Two broad classes of catalytic coating materials are available:




 metals and oxides.  The metals of catalytic interest are listed in




 Table 2.   Of these metals, the only ones which have a possibility of






         TABLE 2.  METALS OF INTEREST FOR CATALYTIC COMBUSTION

Fe
Ru
Os
GROUP
Co
Rh
Ir
VIII* GROUP IB
Ni
Pd
Pt

Cu
Ag
Au
                *Enclosed metals are considered noble.






remaining in the metallic state in a high-temperature, oxidizing en-




vironment are the noble metals; the others readily form oxides.  Of




the noble metals, a large volume of data and correlations are avail-




able 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, methanol, and hydrogen.
                                11-173

-------
The high activity of these metals is related to their ability to acti-




vate H_, 0?, and carbon-hydrogen and oxygen-hydrogen bonds.  Palladium




and platinum are readily prepared in a highly dispersed form on a num-




ber of support materials.  Because of the high activity of these metals




per unit area of metal surface (specific activity) and the ability to




attain high dispersions, only small amounts are necessary for catalytic




combustion (0.1 - 0.5 wt%).  Ruthenium metal is another possible candi-




date for catalytic combustion; however, under oxidizing conditions it




forms a volatile oxide (RuO,), which is rapidly removed from conventional




catalyst supports.  One approach to solve this problem has been to anchor




Ru to a support by forming a relatively stable perovskite structure with




certain oxides such as La-O,.  Osmium is even more volatile and poison-




ous than ruthenium in an oxidizing environment.  Also, as with iridium




and rhodium, it is very costly, and available in limited supply.  The




use of the latter two metals, if at all, would be restricted to small




quantities in multimetallic systems.  Silver, while quite active at low




temperatures for the activation of 0., melts at low temperatures and




therefore sinters to a significant degree at the high temperatures of




catalytic combustion.  However, it could find use as an additive in a




multimetallic catalyst.  Gold is very inactive for oxidation, and would




therefore only be considered as a structural or electronic modifier for




multimetallic catalysts.
                               H-174

-------
     Based on  the above  considerations,  those metals which  show  the




 greatest  promise for use in a  single-active-element catalyst  system are




 platinum, palladium, and ruthenium  (stabilized).  Highly active  and




 stable  catalyst systems  can be produced  through the use of multi-




 metallic  systems (such as palladium/platinum), as well.





     The  catalytic properties of metal oxides have been investigated




 by a number of researchers, including research in catalytic oxidation.




 As for  metals, it has been found that the activity for hydrocarbon oxi-




 dation  parallels the ability of the catalyst to catalyze the homomolecu-




 lar exchange of oxygen.   It has also been noted that the oxides of




 transition containing ions with partially-filled d-orbitals have the




 greatest  activity.  Based on these considerations, the most active sim-




 ple oxides are Co^^, MnO,,, NiO, CuO, Co203> ^^2°3' and V2°5*  Multi-




 component metal oxides of interest include stable spinel structures




 such as Co-O,/CuO, Cr 0  /CuO, and La20/Co_0,.






 Temperature Capability of Catalyst/Support System





     The  choice of an optimum catalyst/support combination requires a




 consideration of the temperature capability of a given combination.




 The low temperature limit is dictated by the so-called "light-off"




 temperature.   Solutions to this problem are not complex for most cata-




 lytic combustion systems, since noble metal catalysts exhibit good




 light-off characteristics for most fuels of interest.   The major ther-




mal limitation for catalytic combustion is the high temperature of
                               11-175

-------
 operation.   The phenomena which lead to catalyst deactivation at high




 temperatures are:





     •  Sintering of the catalytic species





     »  Changes in catalyst stoichiometry





     •  Sintering of the wash coat





     •  Thermal degradation of the ceramic support material





     *  Mechanical failures





     Sintering of the catalytic species will lead to a concomitant loss




 in surface area and therefore a loss in catalyst activity per unit mass




 of catalyst.  The temperature at which sintering occurs is a function




 of the catalytic material.  In an oversimplified manner this sintering




 or crystallite growth is a function of the melting point for metals.




A number of  techniques have been developed to minimize or prevent sin-




 tering.  For example, multimetallic systems can be prepared which are




more stable  than the constituent metals.  A number of structural promo-




 ters are also known.





     Stoichiometry changes for oxides can lead to a catalyst with dra-




matically different catalytic properties.   Loss of lattice oxygen can




cause an Increase in the energy required to activate 0» and thereby




decrease combustion activity.   One method for overcoming this problem




 is to dope the lattice with a material which causes a decrease in the




energy necessary to activate oxygen and also stabilizes the stoichiom-




etry of the lattice.







                                 II- 176

-------
     A third mechanism of thermal deactivation is wash coat sintering.




At temperatures above 1650°F high surface area r)~ or y-A£_0_ undergoes




a phase change to a-AJLO,, with concomitant sintering.  The change may




be from a surface area of 300 m2/g to about 5 m2/g.  This sintering




results in pore closure and a "burying" of active catalytic sites in




the AA^Oq-  To minimize this problem, it is possible to work with cata-




lysts for which the catalytic phase has been deposited on a presintered




AJLO,.  Alternatively, other materials which are thermally more resis-




tant may be investigated as wash coats.  These might include ZrO? and
     At very high temperatures, the ceramic support, be it spheres or




monolithic honeycomb structures, begins to degrade.  Typical monolith




fail temperatures have been given in Table 1.  Although catalytic com-




bustion units are not expected to operate continuously at the tempera-




tures listed, the possibility of local hot spots occurring due to uneven




air/fuel mixing and local catalyst bed nonuniformities must be con-




sidered.





     Thermal expansion and contraction may also occur at high tempera-




tures.  Both pellet and monolithic supports have thermal expansion coef-




ficients different than their mounting hardware.  Considerable catalyst




mechanical attrition can occur if the catalyst becomes loose, thus caus-




ing a decrease in catalytic performance.
                                  II- 177

-------
EXISTING APPLICATIONS OF CATALYTIC OXIDATION CONCEPTS





     The existing applications of catalytic oxidation concepts in indus-




try are largely in the low- temperature area at this time.  The primary




applications of catalytic combustion are in nitric acid plant tail gas




cleanup, industrial odor control, small catalytic heaters, and automotive




oxidation catalyst systems.






Nitric Acid Plant Tail Gas Cleanup





     Nitric acid is produced commercially by reacting ammonia with air




to produce nitrogen oxides, which are then absorbed in water to yield




nitric acid.  Large amounts of unabsorbed nitrogen oxides in the tail




gas from the absorption tower pose a pollution problem, which has been




solved in some cases by the catalytic processing of this tail gas over




platinum-group metals to yield useful energy in the form of steam and/or




power.  The basic reactions for the catalytic treatment of nitric acid




tail gas over platinum, palladium, or rhodium for methane reducing fuel




are:
CH, + 4N02 -* 4NO + C02 +
CH, +
                                   20 (decolorization)





                       + 2H20 (combustion)
     •  CH, + 4NO •* C02 + 2H20 + 2N2 (abatement)





The tail gas reactor outlet temperature is usually between 1250°F and




1380°F.  Support materials for the catalyst have included nichrome wire,




alumina pellets, and alumina honeycomb materials.
                                 11-178

-------
Industrial Odor Control





     The elimination of organic fumes is desirable from both air pollu-




tion and fire safety points of view.  The platinum metals have been used




for several years as catalysts for the oxidation of carbon monoxide and




a wide range of organic molecules in the presence of air or oxygen.




Most odors are caused by the emission of low concentrations of organic




molecules to the air, with these organic molecules always containing




carbon and hydrogen and generally sulphur, nitrogen, and oxygen as well.




Since most industrial odor problems are caused by organic compound con-




centrations well below the level required for spontaneous combustion in




air, it is necessary to raise the contaminated air stream temperature to




a level at which combustion can occur.  The use of catalysts lowers the




temperature needed to achieve odor removal, and also lowers the neces-




sary residence time at the combustion temperature.






Low-Temperature Catalytic Heaters





     Low-temperature catalytic heaters are characterized by the small




portable radiant heaters that operate at temperatures below 800°F.




Operating characteristics include heat release rates of approximately




75 Btu/in2 and lifetimes of several thousand hours.   Gaseous fuel is




distributed in a uniform fashion to all parts of a catalytic pad, and




combustion takes place on the surface of the fibers  of the pad.





     The Control Systems Laboratory of the EPA has conducted an extensive




emissions testing program on a number of catalytic heaters, as reported
                                  11-179

-------
in Reference 7.  No correlation between emissions and specific heat



release rates was found, and CO and UHC levels were generally quite



high.  However, NO levels were quite low for nearly all heaters tested.



The high CO and UHC levels were attributed primarily to poor fuel-air




mixing.





Automotive Exhaust Catalysts




     Since platinum metal catalysts were selected for emission control



on conventional automotive internal combustion engines for the 1975



model year, a great deal of development work went into these emission



control systems.  In general, NO  emissions are controlled by a reduc-
                                X


tion mechanism, and CO and UHC emissions are removed by catalytic oxida-



tion with secondary air added to the exhaust stream to ensure complete



oxidation.




     A reactor for catalytic exhaust emission control has three essen-



tial design features:




     •  A ceramic support on which the catalyst is deposited in a manner



        that prevents attrition




     *  Rapid warm-up of the catalyst unit, and hence rapid light-off




     •  Nearly complete conversion efficiency




     The essential features of a satisfactory CO/UHC oxidation catalyst



for automotive emission control are (Reference 8):




     •  A light-off temperature in the region of 250°C
                                  11-180

-------
     *  Conversion efficiencies in excess of 90 percent for both CO and



        UHC at space velocities up to 150,000/hr




     •  High temperature stability of the support and platinic crystal-



        lite system at temperatures to 950°C




     •  Poison resistance to compounds containing lead, phosphorous,



        and sulfur.




Platinum-palladium catalysts have been extensively tested in full-scale



vehicle tests (Reference 9).  The differences between platinum and pal-



ladium catalyst performances were quite small.




     Reference 10 discusses possible substitute catalysts for platinum



in automobile emission control.  In this study, it was concluded that



base metal catalysts generally are not as effective as platinum/palladium



catalysts for automobile emission control, especially for hydrocarbon



and carbon monoxide oxidation.  For NO  reduction, certain base metal
                                      A


catalysts initially are more active and selective than are noble metal



catalysts; however, most suffer from rapid deactivation.





RESEARCH PROGRAM





     Under EPA funding, an 18-month research program has been undertaken



which emphasizes the understanding of the mechanisms of catalyst per-



formance in the catalytic combustion process.   The program includes cata-



lyst screening tests on five different fuels,  where such performance



factors as light-off,  conversion efficiency, extinction, poisoning, and
                                   11-181

-------
temperature limits will be examined for both single- and multi-active




component catalysts; the integration of promising catalyst systems into




a catalytic combustion system for the purpose of evaluating the concepts




of bed heat removal, exhaust gas recirculation, and two-stage combustion;




and the development of an analytical model based on data obtained in the




screening tests which will be capable of evaluating the performance of




the catalyst system.  All of these tasks will be accomplished at both




small-scale (nominally 100,000 Btu/hr heat release rate) and large-




scale (nominally 1,000,000 Btu/hr heat release rate) conditions.  Each




of these program elements will now be described in greater detail.






Screening Experiments





     It is the intent of the screening experiments to provide an effi-




cient means of evaluating catalyst bed performance, in order to permit




the design of a variety of systems incorporating each catalyst bed.




The evaluation of a catalyst for a combustion system differs in at least




one major respect from more conventional catalyst applications (e.g.,




reforming, exhaust gas clean up, partial oxidation of organic compounds,




etc.).  This is the major heat release associated with the process.




Because of the exponential relation between kinetic rates and tempera-




ture, these systems tend to "bootstrap" themselves rather than simply




"cook" at an imposed temperature.





     Any system concept will provide for heat removal in some way to




limit maximum surface temperatures.   To do this uniformly is not
                                 11-182

-------
generally possible for packed bed or monolith systems.  To achieve uni-




formity, varying amounts of inert dilution will be fed with the test




streams into an effectively adiabatic reactor.  The adiabatic flame tem-




perature of this stream will approximate the maximum catalyst surface




temperature within the system.  The relative uniformity of temperature




achieved during the screening experiments will greatly enhance the




ability to obtain quantitative catalyst activity data from these tests.





     The initial screening experiments, all to be run at atmospheric




pressure, will be conducted in the existing Jet Propulsion Laboratory's




catalytic combustion test facility.  A photograph of this facility is




shown in Figure 2.  This facility has previously been used for research




involving the catalytic partial oxidation of kerosene to produce a




hydrogen-rich fuel for automotive applications.  These screening ex-




periments will initially focus on the performance variations of plat-




inum catalysts on several substrates, with varying wash coats, in




order to determine such things as the effect of wash coat deposition




procedure, the effect of thermal deactivation of wash coat and catalyst,




and the effect of chemical stabilization of the wash coat.  Wash-coated




supports will be supplied to Aerothenn by substrate manufacturers, and




Aerotherm will prepare and apply the catalyst to the wash-coated support.




The planned catalyst test matrix for the first 16 tests is shown in




Table 3.  All these tests will be run with natural gas as the fuel.
                                   II-183

-------
                                            o
                                            «
                                           M-l
                                            m
                                            fl)
                                            a
                                            o
                                            a
                                           ,0
                                            u

                                            u
                                            ft
                                            U
                                           cs

                                            01
                                            60
                                            -H
II- 184

-------
X
C/)
CO

§
Cu
       H
       CO
       i-l
       PL,
ro

W
r4
OS

I














0)

•H
4J
U
CU
1-J
XI
o





















u
CO
o
CJ

r*
CO
CO



4J
CO
rH
n)
ij
CO
CJ

o
a

4-1
CO
cu
H

00
e
•H
4-1
CO
OJ


0)
N
•H
e
•H
4J
a,
o

'O
a
CO

4-1
•H
C


,J

>-) CO
01
4-1 V)
3 3
o -o
CU
Q) rj
C 0
•H S-i
hJ a.
01
4-1
, j
CU V4
4J CU
•H -H
rH TJ
rH M
IS

/«v.
CJ
o


CO
O
CM
o?
^
j
^

"^
CJ
O

4-1
PM


in
I
tH



^t
4J
•H
H
•H
,Q
CO
a.
cfl
O

C
0
•H
4-1
CO
S-i
CO
a,
CU
VH
a.

4-1
co
^
rH
CO
4-1
cO
CJ

^
M-l
•H
^
cu



CU
4J
•H
rH
rH
S

/— \
CJ
o


(f)
o
CM
c^

5 co
cu
i.\-j ^|
O
•
4-1 14-1
CJ CJ
cu •*-<
M-l
M-i CU
01 VH
3
01 T3
4-1 01
cd cj
M O
•H (-1
4-1 O.
CO
01 C
> 0
C -H
M 4J


01
4-1
•H
rH
rH
1
fO
XX
g
Vfc*^

co
o
csl
o?
^3
1


o"
OS
S

4-1
PH



CT*



CU
M
CO
a
g
o
u
\-^

C
o
•rt
4-1
Cfl /~*
> TJ
•H l-l
4J CO
O 13
CO C
01 CO
•0 4J
CO
rH
cd cO
£3
H CO
cu  f.
H | ] \
M-l i-l
U 3


x
CJ <^v
O T3
s*-^ CD
l-l
CO CO
0 4J
CM C
<=< iH
<; co
I ^s
*~

o"
M
ci

4_l
PH



O
rH




CU
4J
k|_|
rH
rH
3

y*-^
o ^a
VH/ Qj
^H
CO CU
O 4J
CM C
o? -H
<(J W
1 s>v/
?~

I


4-1
CM



rH
rH



CO
CO
3

*4H
o

a
0
•H
4J
CO
N
•H
rH
•H
XI
CO
4-1
CO

rH
CO
e»
£^
* CJ
•

§
s

•W
PH



CO
rH










4-1
CO
O
U

f.
ca
cO


Q)
rH
^tQ
CO
4-1
CO
CO
M
4-1
tH
3

*4-t
0

4-1
CJ
0)
M-l
MH
H


(V
4J
•H
rH
rH
1





CO
O
CN
o^
^
rH
CO
4-1
Cfl
CJ

M-J
O

4J
U

CM Cfl
O 0)
Vn M
CJ H

o"
Pi


4-1
CM



so
rH


                                                                                                                      a.

                                                                                                                      J
                                                                                                                      to
                                                                                                                      
-------
     Additional screening tests will be conducted with four other fuels:




#2 distillate oil, methanol, hydrogen, and low Btu gas.  The remainder of




the screening experiments at JPL will focus on establishing light-off tem-




perature, extinction limits, and poisoning effects for a variety of cata-




lyst.  Multi-active-component catalyst systems will be developed based on




the most promising single-active-component catalysts tested and will be




tested with the five different fuels mentioned previously.  The evalua-




tion criterion for catalyst performance will be the effect of a test vari-




able on pollutant emissions, e.g., CO, UHC, and NO .





     While the initial screening experiments are being conducted at JPL,




Aerotherm will be constructing a catalytic combustion test facility for




the remainder of the experimental work on this contract.  A schematic




diagram of this facility is shown in Figure 3.  This facility will be




capable of testing at pressures up to 10 atmospheres, and will be cap-




able of doing both small- and large-scale testing.  It will initially




be used in the conduct of the high-pressure screening tests and the




catalyst life tests.  An overview of all testing to be conducted during




this study is given in Table 4.






System Concept Evaluation Experiments





     The focus of the system concept evaluation will be on effectiveness




of various heat extraction techniques and their effect on catalyst per-




formance (kinetic and thermal stability).  All of these experiments




will be conducted in the Aerotherm test facility in the versatile
                                   II- 186

-------
              O
              m
              VI
              ai
                  a_-t-> o o. X
              E » E j| a. Q. o
         >J   OOI3E**— to
         ail      ^ u •
         ^           x   *J    >
              \. ^ ^- o —  tx. •—














1-


( 	
                                                    O
                                                    u
                                                    a
                                                    a
                                                    n
                                                    c
                                                    o
                                                    u

                                                    u
                                                    .
                                                    a
                                                    u
                                                    o
                                                    tn
   ©
                                   X'Lddns

                                       UO

                                  1ujniaa
     -®—
II-  187

-------
TABLE 4.  OVERVIEW OF CATALYST SCREENING AND SYSTEM
                 EVALUATION TESTS
Element
Small-scale
catalyst
systems
Small-scale
system
studies
Large-scale
catalyst
systems
Large-scale
system
studies
Test Series
Single-catalyst
screening (1A)
Multi-component
catalyst screen-
ing (IB)
Selected catalyst
system evaluation
(1C)
Concept evaluation
(ID)
Large-scale
catalyst screen-
ing (2 A)
Concept develop-
ment (2B)
Primary Purpose
Preliminary evaluation
of single-catalyst per-
formance
Preliminary evaluation
of multiple-component
catalyst performance
Perform extensive eval-
uation of selected cat-
alyst systems
Integrate promising
catalyst systems into
a practical combustion
system
Extensive evaluation of
selected catalyst per-
formance
Development of optimum
combustion system con-
figurations
Approximate
Number of
Tests
70
13
28
34
35
34
                     II- 183

-------
catalytic combustion test apparatus shown in Figure 4.  In this figure


many possible concepts are shown connected in series, but only for illu-


strative purposes.  In actuality, only a few of the segments would be

                                     t
assembled to test a particular concept.  The versatile test apparatus


is capable of testing flue gas recirculation, two-stage combustion, and


bed heat removal concepts.  Table 4 gives an overview of the system con-


cept tests to be conducted.



Data. _Ana ly s is


     In order to establish the data reduction procedures associated with


obtaining catalyst activity, Figure 5 must be considered.  In preparing


this idealized representation of wall events it is assumed that heat and


mass transfer coefficients are equal, that diffusion coefficients are


also equal, and that the system is adiabatic, so that enthalpy is con-


stant.  It should be emphasized that these assumptions are for the ideal-


ized representation only, and do not necessarily represent what will act-


ually be used in the data analysis.  The independent variable selected,


K  , represents the wall concentration of the lean reactant.   The depen-

  w                .
dent variables are n^, mR£AC, and TWAU/  mTRAN is the mass flux of

                          *
reactant to the wall, and n^pAp is the mass flux of reactant consumed by
the catalyst, while T  T  is the wall temperature and is proportional to
                     WALLi

the fraction of reactants consumed at the wall.  This proportionality is


assumed to be linear in the sketch.  Thus, when all reactants are con-


sumed away from the wall the adiabatic flame temperature is reached.  A


steady-state solution is reached when the mass flux of reactants to the
                                  II- 189

-------
                      I

                       s.
                       o
                       VI
                       c
                       O)
                      Cfl
                 S-
                 41
 L-
 41
•o


I

•a
           V
           at
       o
      •!-
      *->
O — 00
X <->
    u •«->
•M  4>  V*
I/I (/)  01
                       J.  S_ >,4->  C  10  O •
                          O "— C  O +> ••-
                        .  0_ 10 01 ••-  « +J
                       >     4-» £ 4-> <_>  c
                        •  O> 10 O O     0)
                       z>  c o 01  ai -o **
                       -^  -r-    (/> tfl  0)  V
                   U
                •M ai
                1. (/)
                        i-    O
                        Hi a: 4->
                        oo fi
                        CLl. 3


                        111
                        X'^ (J
                        UJ -M
                           U
             4-> 01  C
             CO I/) -r-
             £
             aiccu
             3  O  OI-P-
             u- <_> i—<•»
             I QC O
L<-> Ol O>    -^        _  „
: .p- C C -O  O>>—  t-     C
I r— •*-'*- 41  C  to  9> •— ••-
)  O i— G*.^ '»- '«-  U 4D  C
  COHSUX-D10C
 co-«-)*^(dP'''~
                                                                           I
                                '•O r— oo O> o •
                                                                           
-------

                                                      o
                                                      *J

                                                      •o
                                                      Ul
                                                      rO
                                                      
                                                        n
                                                      O
                                                »—   i— C
                                                fd    IQ *O
                                                C    C-U
                                                o    ou
                                                f-       

                                                2    £§
                                                Q.    CXU

                                                  O   ^«
                                                  
                                                             u
                                                             o
                                                             M
                                                             O.
                                                                                X
                                                                                o

                                                                                u
                                                                                •H
                                                                                «8
                                                                                4-1
                                                                                
-------
wall is equal to the mass-flux of reactants consumed at the wall, and is


                                       •                    *

represented by the intersection of an J^BAJJ curve with the "^pAp curve

    v


shown.  Both "cold" and "hot" stable solutions can occur, but the "hot"




stable solution is the one of interest.





     Two items of interest in Figure 5 are events concerning self-ignition




and extinction.  Self-ignition of the bed can occur at very low space




velocities if effectively adiabatic conditions occur.  At high values of




space velocity the value of the mass transfer coefficient p u (L. will




also increase, and eventually extinction will occur.  The extinction




space velocity values for a few adiabatic flame temperatures can provide




a characterization of a given catalyst bed/fuel combination.





     In addition to the preceding "thermal" analysis, exit composition




measurements will further assist in the formulation of the surface kine-




tic reaction processes.  In addition, the effectiveness of the catalyst




to yield equilibrium NO  compositions at its surface will also be eval-
                       X



uated.






SUMMARY





     The use of catalysts in place of conventional burners for promoting




hydrocarbon reactions appears to have advantages in the control of emis-




sions.   A research program has been undertaken that addresses the basic




lack of knowledge concerning catalyst characteristics, wash coat, sup-




port,  and bed depth required for an oxidation reaction to go to comple-




tion.   In addition, the program will study system techniques such as
                                II- 192

-------
bed cooling, exhaust gas recirculation,  and staged combustion to hold




the bed temperature down.  The results of this study will be directly




applicable to the design of compact catalytic combustion systems for




residential and industrial applications.
                                   11-193

-------
                              REFERENCES
  1.  Davy, H., "Some New Experiments and Observations on the Combustion
     of Gaseous Mixtures," J. Davy, ed., The Collected Works ofSir
     Humphrey Davy, Vol. VI, Miscellaneous Papers and Researches, Part I,
     On the Safety Lamp for Preventing  Explosions in Mines, Houses
     Lighted by Gas, Spirit Warehouses, or Magazines in Ships, etc.,
     With Some Researches on Flame, Section II, Papers Published in the
     Philosophical Transactions and in  the Journal of Science and the
     Arts, on the Fire Damp, the Safety Lamp, and on Flame, Smith, Elder,
     and Co., Cornhill, London, 1840, pp. 81-88.

  2.  Spalding, D. B., "Heat Transfer from Chemically Reacting Gases,"
     Ibele, W., ed., Modern Developments in Heat Transfer, Academic
     Press, N.Y., 1963, pp. 19-64.

  3.  Liitken, A. and Hoist, H., Opfindelsernes Bog. Vol. IV, Nordisk
     For lag, Kgibenhavn, 1914, pp. 131-139.

  4.  Bone, W. A., "Surface Combustion and Its Industrial Applications,"
     Engineering. Vol. 91, April 14, 1911, pp. 487-489.

  5.  Bone, W. A., "Surface Combustion," Engineering, Vol. 93, May 10,
     1912, pp. 632-634.

  6.  Walker, W. H., Lewis, W. K., McAdams, W. H., and Gilliland, E. R.,
     Principles of Chemical Engineering. 3rd Edition, McGraw-Hill Book
     Company, Inc., New York, 1937, pp. 202-204.

  7.  Thompson, R. E., Pershing, D. W.,  and Berkau, E. E., "Catalytic
     Combustion — A Pollution-Free Means of Energy Conversion," Environ-
     mental Protection Technology Series EPA-650/2-73-018, Aug. 1973.

 8.  Acres, G. J. K., Bird, A.  J., and Davidson, P. J., "Recent Develop-
     ments in Platinum Metal Catalyst Systems," The Chemical Engineer,
     March 1974.

 9.  Barnes, G.  J. and Klimisch, R. L., "Initial Oxidation Activity of
     Noble Metal Automotive Exhaust Catalysts," SAE Paper 730570, pre-
     sented at the Automobile Engineering Meeting, Detroit, Michigan,
     May 14-18,  1973.

10.  National Materials Advisory Board, "A Semi-Delphi  Exercise on Sub-
     stitute Catalysts for Platinum in Automobile Emission Control De-
     vices and Petroleum Refining," NMAB-314,  March 1974.
                                    II- 194

-------
    TABLE OF CONVERSION FACTORS FOR METRIC UNITS
  English Units





atmosphere




Btu




Btu/hour




degrees Fahrenheit




foot




inch
        Metric Units





1.0133 x 105 newtons/meter2




2.52 x 102 (g) cal




2.52 x 102 (g) cal/hour




(tp - 32)/1.8 degrees Celcius




3.048 x 10"1 meters




2.540 x 10~2 meters
                     II-195

-------
 11:40 a.m.
Pilot Scale Investigation of
  Catalytic Combustion Concepts
  for Industrial and Residential
  Applications
Dr. John P. Kesselring, Aerotherm Acurex Corporation

Will you be considering the possibility of using a
catalyst monolithic screen in a location between a

furnace and a heat exchanger as a burnout medium for

soot that is passing through?
We will not be considering that in this program.

In fact,  the avoidance of sooting conditions in cata-

lytic combustion appears important, since active sites

can be buried by soot deposition inside the monolith.
                            11-196

-------
THE OPTIMIZATION OF BURNER DESIGN PARAMETERS

         TO CONTROL NOV FORMATION IN
                      J\
    PULVERIZED COAL AND HEAVY OIL FLAMES
  M. P. Heap, T. J. Tyson and G. P. Carver
             Ultrasysteins, Inc.
                G. B. Martin
      Environmental Protection Agency
                     and
                 T. M. Lowes
  Associated Portland Cement Manufacturers
                11-197


-------
                           ACKNOWLEDGEMENTS
       The work reported in this paper was carried out under EPA
Contracts 68-02-0202 (International Flame Research Foundation) and
68-02-1488 (Ultrasysterns, Inc.).  Two of the authors (M.P. Heap and
T. M. Lowes) wich to express their appreciation to all their former
colleagues at the IFRF Research Station.  Dr. E. E. Berkau and
Mr. D. W. Pershing made significant contributions to this work by
virtue of many stimulating discussions.
                                11-198

-------
1.0    INTRODUCTION



       In recent years considerable effort has been expended in an




attempt to reduce the emission of nitrogen oxides from stationary com-




bustion sources.  Several parallel paths have been followed in an




attempt to achieve this objective, including the modification of




operational procedures and system redesign.  This paper describes a




series of small scale investigations which were carried out to identify




burner parameters which control nitric oxide formation in pulverized




coal and heavy fuel oil flames.  The amount of nitric oxide produced




in turbulent diffusion flames appears to be strongly dependent upon




the rate of fuel/air mixing.  Two burner design options have been




identified which control mixing and, therefore, influence nitric oxide




formation.  These options are:




       •    the method of fuel injection, and




       •    the method of combustion air delivery.




Both options also affect flame length, flame volume and heat absorp-




tion at the boundary of the combustion chamber.  The paramount objec-




tive of the investigations discussed in this paper is the satisfaction




of combustor requirements as well as the minimization of pollutant




emissions.  Plans will also be discussed to investigate the influence




of scale and multiple burner systems on low emission, high efficiency




combustion systems suitable for both industrial and utility boilers.
                              II- 199

-------
 2.0    NITRIC OXIDE FORMATION IN PULVERIZED COAL AND HEAVY OIL FLAMES




       Nitrogen oxides are formed from two sources of nitrogen during




 the combustion of fossil fuels, molecular nitrogen, and nitrogen com-




 pounds which occur naturally in both liquid and solid fuels.  The




 reactions controlling the rate of oxidation of molecular nitrogen,




 producing thermal NO, are strongly temperature dependent and proceed




 at significant rates only above 1600°C.  It was originally  thought




 that the reaction between nitrogen molecules and oxygen atoms was




 mainly responsible for NO production in flames.  However, it is now




 believed that reaction of hydrocarbon radicals with molecular nitrogen




 in a flame zone provides an additional path for thermal NO formation.




 Nitrogeneous material in fossilized animals and vegetation provided the




 source for nitrogen compounds which are present in coal and oil.  During




 the combustion of these fuels some of this nitrogen is oxidized giving




 fuel NO.   The amount of fuel NO formed in simple premixed flames is not




 strongly dependent upon temperature,  but it is most sensitive to oxygen




availability.




       The processes involved in heat release and nitric oxide forma-




 tion in pulverized coal and heavy oil flames are extremely complex.




The fuel and air are supplied separately to the combustion chamber




through a device which ensures that they are mixed and that the fuel




is in a condition to ignite and burn  completely.   In order to discuss




those features of practical flames which influence NO formation the




total process will be considered in three simple stages.
                              11-200

-------
       Fuel Preparation .  Liquid fuels are atomized, mixing takes
       place between the fuel, the combustion air and recirculated
       hot combustion products.  Both liquid and solid fuels evolve
       gaseous components which will ignite under suitable conditions.

       Heat Release.  The nature of the fuels and the turbulent
       mixing process allow heat release to take place in several
       modes.  The volatile fuel fractions can burn:  1) in homo-
       geneous premixed zones, or 2) in diffusion flames around single
       droplets or clouds of droplets.  Soot formed from the volatile
       fuel fractions must be oxidized as must the char which remains
       after the devolatilization of the coal particles.

       Heat Transfer.  The products of combustion leave the heat
       release zone and mix with the bulk gases and lose heat before
       leaving the combustion chamber.

       In cold wall combustion chambers nitric oxide production can

be considered to occur only during the heat release stage.  Extensive

reviews of the kinetic mechanisms associated with thermal NO formation

                                    (1 2)
in hydrocarbon flames are available;  '   but for the purposes of this

discussion it is convenient to refer to the two dominant mechanisms as

Zeldovich, i.e.,

       N   +  0 — > NO  +  N
or Fenimore,  ' i.e.,

       N2  +  R — *• RN  +  N


where R is some hydrocarbon fragment such as CH, and oxidation of the

nitrogen fragments produces NO,  The Zeldovich path will be dominant

in fuel lean zones, and the amount of thermal NO produced will depend

upon:

       —    the stoichiometry of the reaction zone;

       —    the amount of diluent combustion products in the
            reaction zone;
                                11-201

-------
            the temperature of those diluents; and

            the rate at which the freshly formed combustion
            products mix with bulk gases and thus their rate
            of temperature decay.

The Fenimore path will be dominant in fuel rich zones, and the amount

of NO produced will probably depend upon the carbon/hydrogen ratio of

the fuel fragments.
                          (2)
       Sternling and Wendt    discussed the fate of chemically bound

nitrogen during the combustion of solid and liquid fuels.  Lacking

definitive experimental results these workers assembled fragmentary

data and concluded that 80 percent of the nitrogen in coal would appear

in the coal char and that a smaller, but still major, portion in the

solid formed after devolatilization of oil droplets.  Subsequent oxida-

tion of the char would then produce the major portion of the fuel NO.

           (4 5)
Heap et al,  '   postulated that fuel NO accounted for the major portion

of the emission from coal flames and that the fuel NO was mainly formed

from "volatile" fuel nitrogen compounds.  Pershing et al,    burned coal

with both air and oxygen/argon mixtures and showed that approximately

80 percent of the total emission could be attributed to fuel NO.  The

possible fate of chemically bound nitrogen in an oil droplet or coal

particle is traced in Figure 1.

       A simple working hypothesis of fuel NO formation can be con-

structed based upon available information from studies involving doped

fuels and the addition of nitrogen compounds to laboratory premixed
                                II- 202

-------
                          OIL DROPLET
                             OK
                         COAL PARTICLE
          VOLATILE PRACTIONS

        (HYDROCARBONS,  RN etc.)

                UN
Figure  1.   The Possible  Fate of Fuel  Nitrogen Contained
  in Coal  Particles or Oil Droplets During Combustion
                         11-203

-------
and gaseous diffusion flames.  The primary fuel nitrogen compounds

are broken down yielding secondary intermediate compounds RN which are

subject to two competitive reaction paths:

       A  —  reaction with an oxygen containing species producing
             NO, or

       8  —  reaction with a nitrogen containing species producing
             Nr


Thus the formation of fuel NO can be considered simply as
              Fuel
                                         NO + ...
                                    *»
A


B
The production of nitrogen (path B) will be favored under fuel rich

conditions and the conversion of fuel nitrogen to fuel NO in flames

can be limited by ensuring that the majority of fuel nitrogen com-

pounds react under oxygen deficient conditions.

       Once formed, nitric oxide can be reduced in heat release zones,

a process which could allow the net production of NO in flames to be

reduced.  An experimental study by Wendt, Sternllng, and Matovich^ '

demonstrated that "reburning" was effective and that up to 50 percent
                                                             to\
reduction in flue gas NO level was obtained.  Engleman et al,    added

NO to a well-stirred reactor burning methane/air mixture and found that

almost all the NO was retained in lean mixtures, but that the 65 per-

cent of the input NO was reduced when it was added to reactants con-

taining only 65 percent of the stoichiometric air requirements.
                               11-204

-------
               (9)
 Sarofim  et  al,    demonstrated that a "diffusion flame environment"



 facilitated NO destruction.  Myerson     has shown that NO contained



 in simulated combustion products can be reduced by isobutane when



 flowed through a heated ceramic tube at temperatures in excess of



 1000°C.  More recent information on NO reduction by hydrogen has been



 presented by Flower, Hanson and Kruger    .  Thus, the possibility



 exists that NO could be produced in the early stages of heat release



 and reduced before leaving the heat release zone.



       None of the previous discussion has alluded to the possibility



 of N02 production in flames.  Nitrogen dioxide has been observed in



both flat flames and turbulent diffusion flames.  In flat flames the



N02 is produced in the visible flame region and, in most cases, is



 rapidly converted to NO downstream of the visible zone.  High N02



 concentrations have been observed in regions of high shear or on the


                                                (12 13)
periphery of the base of natural gas air flames.   '     The effects



of N02 formation in coal and heavy oil flames are unknown.  In fuel



oil or pulverized coal flames it is unlikely that the rapid formation



and destruction of N02 within the heat release zone would influence



 the net production of NO unless this process competes with those



reactions leading to the formation of N2 from NO.  Consequently, NO



would be converted to N02 only to be reformed at a later stage, rather



than being reduced to N2-



       Two generalized schemes are presented in Figure 2 to illustrate



possible approaches to the control of the net production of NO in
                             H-205

-------
                           HEAT LOSS
                                                  BUT LOSS
            SECOND STACE AIR
TOTAL AIX
// i /

——4-4-
PRIMARY
RICH
ZONE
-x^-,_
PRODUCTS
PARTIAL £ OX
— x-J
OF
-x 	 »
IDAIIOH
SECONDARY'
BURNOUT
ZONE

                                                               TO FLUE
                                          	J
                        STAGED HEAT RELEASE
niMAKY FUEL
1—
i
L_
SECONDARY FUEL
HIGH INTENSIlT
LOU EXCESS AIR
PRIMARY ZONE
	
PRODUCTS OF
COMPLETE"! COMBUSTION*1
1
SECONDARY AIR
REBURNINC
ZONE


. _^J BURNOUT
t\ ZONE

                           REBURNINC
Figure  2.  Generalized Schemes for  the Reduction of
            HO Emission from Turbulent Diffusion in Flames
                       11-206

-------
turbulent diffusion flames.  The more familiar approach is through




staged heat release by reducing the rate at which air is mixed with the




fuel.  Thus a rich primary zone is produced which creates conditions




favorable for path B.  The crucial practical implications of this sys-




tem are that the second stage air must be added in such a way as to




ensure complete combustion within the constraints imposed by practical




systems; i.e., the need to produce steam and all that this implies in




terms of both thermal and operating efficiency.  Although in Figure 2




the primary and secondary zones are shown as being physically separated,




this must not necessarily be the case — some compromise being necessary




to accommodate the heat release zone in the available physical space.




            The alternate scheme for the NOX control during combustion




of nitrogen containing fossil fuels is based upon the concept of




reburning.  High intensity rapid mixing conditions favor both NO forma-




tion and efficient combustion.  In the primary region the major portion




of the fuel is burnt without regard for NO control.  Secondary fuel is




added to the products of combustion and the nitric oxide is reduced




in the fuel rich mixtures.  Subsequent addition of air completes com-




bustion.  This scheme would require rapid mixing of the secondary fuel




with the primary products which should contain the minimum amount of




excess air, probably requiring that the secondary fuel be gaseous.




            This paper is concerned with investigations to determine




the optimum mixing pattern to control nitric oxide formation by the
                                11-207

-------
 former scheme, which  aims  to prevent  excessive formation of nitric



 oxide in the primary  zone  and promote burnout of the partially oxidized



 fuel in the secondary zone.




 3.0    SUBSCALE  EXPERIMENTS TO ESTABLISH OPTIMUM CONDITIONS

        FOR MINIMUM POLLUTANT EMISSIONS



        EPA Contract 68-02-0202 carried  out  by the  Staff of the Inter-




 national Flame Research Foundation,  IJmuiden, Holland, was concerned



 with defining those burner parameters which control  the rate  of NO



 formation in turbulent pulverized coal  and  heavy fuel oil flames.   The



 results of these investigations  will  be described  to illustrate various



 approaches which can  be used to  provide the conditions illustrated  in



 Figure 2 as staged heat release.   Two methods have been identified



 which allow the  control of nitrogen  oxide formation while maintaining



 satisfactory combustion conditions.   These  are:



        •     division of  the total  combustion air supply into separate

             streams, thus  delaying mixing the total air supply with

             all  the fuel,



        •     optimization of the method of fuel injection.



        The  subscale experiments were  carried out in a refractory tunnel



furnace of  2 meters cross  section  and 6 meters long;  complete details


                                                     / 1 0\

of the  experimental system have been given elsewhere.      Figure 3



presents a  schematic showing the arrangement of the versatile test



burner used  in the investigations.  This burner utilized the moving


                                         (4)
block swirl generator designed by Leuckel    and readily allowed the



influence of several parameters on pollutant emissions to  be investi-
                              11-208

-------
                                     N

                                     PM

                                     iH
                                     •H T3

                                     °  §
                                     M
                                     O  eg
                                        0)
                                        M
                                       C  Vi
                                        o FM
                                     IN ••-(
                                     o  w  
-------
gated.  The most significant being 1) the method of fuel injection,




2) division of the total air supply between primary, secondary and




tertiary air streams, 3) the secondary swirl imparted by the moving




blocks, 4) the primary swirl imparted by fixed vanes, 5) the angle of




the burner divergent, 6) the number of injection points for the tertiary




air supply, and 7) the position of the fuel injector relative to the




burner exit.





3.1    Optimization of the Method of Fuel Injection




       Parametric studies very quickly indicated that the method of




fuel injection had a profound effect upon nitric oxide emission from




pulverized coal flames.  Measured emissions ranged from 800 to 200 ppm




for a 1.1 percent nitrogen coal with 300°C preheat and 5 percent excess




air.  Maximum emissions were achieved when the fuel was injected




radially into the air stream.  Heap et al,   '   '   ^  postulated that this




method of fuel injection provided the most oxygen-rich conditions for




the combustion of the volatile coal fractions.  Those injection con-




ditions which tended to minimize the rate of mixing between the fuel




and the total combustion air gave minimum emissions of nitric oxide,




and were provided by a high velocity axial fuel injector utilizing the




minimum amount of primary air.   Figure 4 gives an impression of the




range of emissions obtained for a given primary air percentage, excess




air level and burner geometry.   Although pulverized coal flames could




be produced with low NO emission,  the length of these low NO flames is




unacceptable for use in existing boiler configurations.  Practical




requirements in utility boilers dictate stable, high intensity
                               11-210

-------
   800 r
               A

               D   MEDIUM VELOCITY AXIAL
               O   HIGH VELOCITY AXIAL
                        1                 2

                       SWIRL NUMBER

Figure 4.   Range  of NO Emissions Produced by Change
   of Fuel Injector  (1.1% N2 Coal, 300°C Preheat,
                 5% Excess Air)
                      II- 211

-------
combustion with rapid burnout.  These combustion characteristics are




attained in practice by using a fuel injection system which is similar




to the radial fuel injector used in the subscale experiments.  Thus,




rapid fuel/air mixing provides satisfactory combustion conditions, but




also promotes NO formation.




       Subsequent experiments with a heavy fuel oil produced results




which provide similar conclusions to those obtained with pulverized




coal.  Minimum NO emissions were obtained by maintaining the fuel oil




as a single coherent jet on the flame axis, thus producing a long




thin cylindrical flame similar in appearance to the low NO coal flame.




Further investigations concentrated upon changing the design of the



burner to maintain the required flame pattern, while attempting to




reduce NO emissions to the minimum level possible.  Japanese workers




suggested an attractive technique for achieving these conditions which



involved splitting the flame into several discrete sections     which




gives the flame a flower petal arrangement.  In these investigations



this was achieved by placing a distributor tip over a standard Y-jet.




Details of two types of distributor tips which have been investigated




are presented in Figure 5; their emission characteristics, plotted




as a function of secondary swirl intensity, are shown in Figures 6 and




7.  It can be seen in Figure 6 that at low secondary swirl levels emis-




sions from standard Y-jets are considerably higher than those obtained




with distributor tips.  However, as the swirl level is increased, the




emissions from flames produced with distributor tips increase and, in
                                   11-212

-------

m
O

•
O
55
4
to
0)
o
•H
^W
•H
H
O
-^s
|
QQ
a
o
SB



00 00 CM
i— 1



o m o
CM CM CM
O O O.
o o c4
o o o
o o o
^H CM f)
                                 (fl
                                 I
                                •H
                                O
O
SS
^H CM M

E
pq'
1
*~r
OQ
a
*
o
^H CNJ
CM •*
0 0
in m
O O
m m
CM CM
-» m
Figure 5. Details of Distributor Tips Used
NOTE: All dimensions are in millemeters
11-213

-------
g
0.
i
    150 H
O
O
D
A
$0° Y-JET

70° Y-JET
DISTRIBUTOR TIP 1
DISTRIBUTOR TIP 2
DISTRIBUTOR TIP
                     0.4    0.6   0.8     1.0
                    SECONDARY SWIRL INDEX
Figure  6.   Emission Characteristics of  Fuel
 Oil  Flames (5% Excess  Air, 300°C Preheat,
          Primary Swirl  Vanes 30°)
                  11-214

-------
 260.
DISTRIBUTOR TIP
    NO   CO

 4  O   •
  too
                     .4       .'6       .8

                  SECONDARY SWIRL INDEX
Figure 7.  Emission  Characteristics of Fuel Oil
     Flames  (5% Excess  Air,  300°C Preheat,
           Primary Swirl  Vanes 30°)
                 11-215

-------
 one  instance, approach  the levels obtained with standard Y-jets.




 Distributor  tips 4 and  5, which included an axial port, tended to




 produce considerable amounts of carbon monoxide at low swirl levels.




 As the swirl was increased, nitric oxide emissions increased and




 carbon monoxide emissions decreased.  Further experiments indicated




 that the emission characteristics of fuel oil flames, produced with




 the  distributor tip design shown in Figure 5, were not only dependent




 upon the configuration  of the tip, but also upon other parameters such




 as the angle of the burner exit and the furnace wall temperature.




 From these investigations it can be concluded that this type of fuel




 injection system could  be readily adapted for single flame applica-




 tions.  However, it may well be that the interaction between flames




 in multiple  burner installations will negate any value with respect




 to reducing  nitric oxide emissions.  Consequently, further work is




 necessary to establish  the influence of flame interactions on emis-




 sions before this type  of tip can be recommended for use in multiple




 burner installations.




       The characteristics of flames produced with the distributor tips




 were found to be different from those produced from the standard Y-jet.




 The extent of this difference can be judged by comparing the radial




 profiles of  species concentration and temperature presented in Fig-




 ure 8.  Measured temperature, carbon monoxide, hydrogen,  oxygen and




methane concentrations are shown at an axial distance equal to




 1.6 burner diameters from the face of the firing wall.   It is evident




from these radial profiles that there is considerable difference
                            H-216

-------
               STANDARD Y-JET
Flame

Axis
        * TEMPERATURE

        D CO

        0 H2
                                    •  1400
                                    - 1000
Ul
u
on
\a
a   6
        TIP I
2


|
    <
        TIP 4    /
   IEOO
800
        ta
        U.
        "

        2
        u
                                      1200
                                      800
              Radial Distance
 Figure 8.  Radial Profiles of Temperature

          and Species Concentration
                   11-217

-------
between  the flame produced with the Y-jet and  those produced with




the  distributor  tips.  The low NO flames are considerably longer, as




shown by the carbon monoxide and hydrogen concentrations, and mea-




sured temperatures are much lower.  The distributor tips also produce




wider flames at  this axial location.  Consequently, utilization of




this method of fuel oil injection would require that considerable




attention be paid to the problem of flame impingement.





3.2    Division  of the Total Air Supply into Separate Streams




       The efficiency of staged combustion as a technique for reducing




nitric oxide emissions from utility boilers is well established.  The




methods used in  actual installation normally involve increasing the




fuel flow through several of the burners, and using the remainder of




the burners only to supply air.  Experiments have been conducted at




subscale to establish whether a combustion process can be effectively




staged using the burner as a staging device by delaying mixing of all




the fuel and air.  Once more, the dual criteria of minimum emissions




and acceptable flame characteristics were imposed upon these investi-




gations since this type of control technique must be acceptable in




existing equipment.




       Investigations were carried out with both fuel oil and pulverized




coal and it was found that the burner can be used as an effective staging




device for both these fuels.   Emission characteristics are shown in




Figure 9 for two oil flames burning under identical furnace and burner

-------
Q.
O.
   180
                                 EXIT NO   SOLIDS.
                                 PPM 0% 02  UG/NM*
                                            a
                   .4     .6

                  SWIRL INDEX
Figure 9.  Reductions in NO Emissions From a
  Fuel Oil Flame by Diverting 35 Percent of
    the Total Air Through the Tertiary
       Injectors (5% Overall Excess
            Air, 300°C Preheat)
                   11-219

-------
conditions except that the staged flame is 35 percent of the air delivered




through eight parallel tertiary injectors.  Included on the same figure




are values of solid emissions measured with a 0-Ray absorption device;




it can be seen that the reduced NO emissions achieved by the staged burner




are not accompanied by significant increases in the level of solid emissions.




However, the reduced rate of fuel/air mixing which is believed to be the




cause of the reduced production of NO also causes an increase in flame




length.  This can be appreciated by comparing the axial temperature and




species concentrations distribution for staged and unstaged flames




(see Figure 10).  Temperatures in the unstaged flame are considerably




higher than those measured under staged conditions.  Flame length can



be UMMed either from visual observation, or it can be denoted by a




particular carbon monoxide concentration.  It is not possible from




these measurements to ascertain whether the characteristics of the low




NO produced by the staged oil flame are acceptable for normal boiler




practice.  However, these encouraging results indicate that the con-



cept shows considerable promise, and that further investigations should




be carried out under conditions more representative of large water




walled boilers.



       Similar experiments were carried out with pulverized coal to




establish whether these techniques were applicable to solid fuels.  Fig-




ures 11 and 12 give an indication of the success of this method of



staging for reducing NO emissions from pulverized coal flames.  As the
                                 11-220

-------
                                              CD
                                              H
                                              •H
                                              O

                                              'O
                                              V
                                              00
                                             •o
                                              
-------
                                          FUEL
                                        INJECTOR
SWIRL
INDEX
   600-
g  400
a
a,
i
   200
                                          C.5.

                                          RAD.
 0.6
 0.9
 0
 0.6
O
a
v
A
                                 "3
                                 Mtotal
                                           .4
         .6
  Figure  11.   The Influence of Mass Flow of Tertiary Air (143) on
     NO  Emissions from Pulverized Coal Flames (5% Excess Air,
                    300°C  Preheat, 950°C Wall)
                              11-222

-------
   1000
  FUEL
INJECTOR

  C.S.
                                           RAD.
SWIRL
INDEX

 0.6
 0.9

 0.6

 0.9
                                     0.4
            0.6
                                                              o
                                                              a
Figure 12.  The Influence of Mass Plow of Tertiary Air (Mj)  on
   NO Emissions from Pulverized Coal Falmes (5% Excess Air,
                  300°C Preheat, 1250°C Wall)
                       11-223

-------
percentage of air injected through the tertiary ports increases the

NO emission decreases markedly .  It was found that the efficiency of

staging depends both upon furnace wall temperature and upon the method

of fuel injection.  Figure 11 presents results obtained with an average

wall temperature of approximately 950°C, whereas the furnace cooling

load was reduced before the results presented in Figure 12 were

obtained causing the wall temperature to increase to approximately

1250°C in the region of the burner.  The influence of furnace wall

temperature on emissions from pulverized coal flames had been observed

before} temperature had a strong influence on NO emission from both

natural gas and pulverized coal flames, but only a minor influence on

emissions from oil flames.  The increased wall temperature and subse-

quent increase in the temperature of recirculated gases produce am

increase in NO emission of about 200 ppm for the staged condition.

It is contended that the influence of the method of fuel injection on

the efficiency of staging is associated with the residence time of the

coal particles within the rich primary zone.  The intensity of swirl

in the secondary air appears to have a negligible influence upon

nitric oxide emissions at high staging levels.  Under all circumstances

carbon burnout was virtually complete at the exit of the furnace for

both staged and unstaged flames.  Figure 13 compares the influence of
 The results refer to two methods of fuel injection, radial (rad) and
 a scaled down version of a fuel injector typical of those used in
 practice (c.s).
                               11-224

-------
tertiary air injection on carbon monoxide emissions for the two methods




of fuel injection with the lower furnace wall temperatures.  It can be




seen that staging has no influence upon CO emissions from the radial




fuel injector; however, the six-fold decrease in NO emission with the




coal spreader type injector is accompanied by a three-fold increase in




carbon monoxide emissions at high levels of burner staging.  The experi-




ments made no attempt to optimize conditions for low CO emissions which




could probably be achieved by further minor modifications to the fuel




injector.





3.3    Investigations Involving Petroleum Coke




       Petroleum coke is a high nitrogen, low volatile, low ash solid




which was burned under identical conditions to those for burning




pulverized coal.  NO emissions from petroleum coke were found to be




generally lower than those from coal.  This can be appreciated by




comparing the influence of firing rate on nitric oxide emissions for




the two fuels (see Figure 14).  At the nominal firing rate of approxi-




mately 5 x 10^ BTUs per hour with 300°C preheat and 15 percent excess




air, emissions from pulverized coal flames are approximately twice




those of petroleum coke when both fuels are injected to ensure rapid




mixing between the fuel and combustion air.  Increasing the firing




rate increases the bulk gas temperature within the furnace which




causes an increased NO emission from both fuels.  However, the




influence of firing rate is much stronger with pulverized coal,




particularly at high firing rates, once more indicating that the




temperature environment of pulverized coal flames has a very strong




influence upon nitric oxide production.  The profound influence of
                               II-225

-------
  160
s
M

Q.
Q.
Ul
Q
I
O

i
U
   80-
   40-
                FUEL
              INJECTOR
                C.S.

                RAD.
SWIRL
INDEX
 0.6
 0.9
 0
                            0.6
                                 O
                                 D
                                 V
                      0.2
                                                            0.6
                                total
 Figure 13.  The Influence  of Mass  Flow of Tertiary Air (M )
      on Carbon Monoxide Emissions  (5% Excess Air,
              300°C Preheat, 950°C  Wall)
                            H-226

-------
   100W
g
I  «oo
    HOQ-I

      1.0
                                          COAL
                                  PETROLEUM
                                   COKE
                                               1.6
                   FRACTIONAL LOAD
 Figure  14.   Nitric Oxide Emissions as a Function of

           Load for Coal and Pulverized Coke
                         II-227

-------
 the method of fuel injection upon the emission of nitric oxide from




 pulverized coal flames has been discussed earlier.  It was found that




 emissions from petroleum coke were almost independent of fuel injector




 type.  Emissions from petroleum coke could also be decreased by using




 the burner as a staging device.  Figure 15 indicates that emissions




 could be reduced by almost half if 50 percent of the air was supplied




 through the tertiary injectors.  This reduction was found to be inde-




 pendent of the secondary swirl level.




       Since petroleum coke is a low volatile, high nitrogen solid,




 comparison of the results presented in Figure 14 could be used to pro-




 vide support for the hypothesis that volatile nitrogen compounds are




 responsible for fuel NO formation in pulverized coal flames.  This




 support would be based upon the assumption that the fuel nitrogen com-




 pounds in coal and oil are similar and that their evolution from the




 solid particle would be similar under similar time/temperature histories.




At this time these are rather sweeping assumptions and, although the



 results tend to support the hypothesis, they cannot be considered as




 definitive.





 4.0    SCALE-UP CRITERIA FOR LOW EMISSION, HIGH EFFICIENCY BURNERS




       The results of the subscale experiments discussed in the pre-




 vious section demonstrated that low emission, high efficiency burners




 could be operated with both pulverized coal and heavy fuel oil.




 Burner designs which reduce the rate of coal/air mixing (and thus NO




 emissions) are already being tested in large utility boilers.  The
                                11-223

-------
   400 -
0~
S
CL
a
   ZOO-
                                  PRIMARY S,

                                    0.4  O
                                    0.6  O

                                    0.9  A

                                                        .6

                                M.3_
                                Mtotal
     Figure 15.  The Influence of Burner Staging on
        NO Emissions from Petroleum Coke Flames
                       H-229

-------
relevance of the subscale investigations to practical situations can

be criticized on three points.  Firstly, all the subscale tests were

conducted in a horizontal tunnel furnace where the flames fired

horizontally, and the exit of the furnace was on the flame axis.  The

recirculation patterns were very different from those likely to be

encountered in large industrial and utility boilers.  Secondly, the

results refer to single burner conditions, and the influence of flame

interactions on NO formation is ill understood.  Thirdly, it has not

been established whether the principles discussed in Section 2.0 of this

paper are suitable for scaling to ensure their applicability in practi-

cal boilers.  Investigations to be carried out under EPA Contract 68-02-

1488 with Ultrasystems have been planned to answer all of these questions,

       Emission control techniques which reduce pollutant emissions at

the expense of fuel economy or operating efficiency are not only

financially unacceptable, but they are also socially irresponsible.

The overall objective of this new EPA program is to design, construct,

and test low emission, high efficiency burners in a size suitable for

use in industrial and utility boilers.  This research program will:

       •    Design and construct a test facility which will allow
            large scale burners to be tested under conditions repre-
            sentative of those to be encountered in industrial
            practice.

       •    Establish the influence of burner/burner and burner/
            furnace interactions on pollutant formation in large
            scale fossil fuel flames.
                             H-230

-------
       •    Establish scaling laws to ensure that design methods are
            available to enable low emission, high efficiency burners
            to be constructed for a wide range of firing rates.

       •    Ensure that reduced pollutant emission is not accomplished
            at the expense of reduced thermal efficiency and increased
            operating costs.

       •    Establish the optimum firing method for dual firing  of low
            BTU gas and pulverized coal in both and single and multi-
            ple burner arrays.

       •    Specify burner designs which are suitable for extensive
            testing in industrial and utility boilers.

This program will be carried out with three fuels:  pulverized coal,

low BTU gas, and heavy fuel oil, burned singly and in combination.

The investigations will include both input/output measurements,  and

detailed flame probing to establish the characteristics of the low

emissions flames.

       The goals of the program can only be satisfied if a flexible

experimental facility is available.  The focal point of this facility

must be a combustor which simulates wall temperature and recirculation

patterns typical of modern industrial and utility water wall boilers.

It must also allow rapid interchange of multiple burner arrays to

enable an assessment of the influence of such parameters as burner/

burner spacing and the distance of burners from adjacent walls on

pollutant emissions.  Figure 16 shows a cutaway sketch of a modular

spray-cooled combustor which has been designed to satisfy the require-

ments of this project by providing the representative environment and

the required degree of flexibility.  The major features of this  combustor
                              II-231

-------
SKIN PANELS
EX05KEUTON
(SPRAY MANIFOLD
AND STRUCTURAL
FRAME)
         ASH HOPPER
                                                                           WATER SPRAY
                                                                           MOVEABLE WALL
              Figure 16.  Cutaway Sketch of  a Modular  Moveable
                          Wall Spray  Cooled Combustor
                                     11-232

-------
are its low thermal inertia; up to five burners can be placed anywhere




on the firing wall; one wall of the combustor is tnoveable, thus allow-




ing the influence of burner/furnace interactions to be established;




the modular construction allows probes to be inserted in almost any




location; and Figure 17 presents a schematic showing the arrangement




of a flame probe and the corabustor wall.  The entire facility will




stand alone, will be dedicated to this project, and will have a firing




rate capacity for single burners of 120 x 10  BTU/hr.  Multiple burner




investigations will be able to be carried out with a maximum of five




burners firing at a total rate of 60 x 10  BTU/hr.  Three single burner




sizes of 12, 60 and 120 x 10  BTU/hr will be tested and their charac-




teristics established, thus providing information on scale up criteria




of low emission burners.  Some idea of the extent of this test facility




can be gained from the general layout which is presented in Figure 18,




which gives the approximate location of the heat exchanger, pulverizer,




steam supply and propane reformers which will be used to provide low




BTU gas.  This program is presently in the design phase; it is expected




that construction of the test facility will commence in the near future.




5.0    CONCLUSION




       This paper has presented experimental information which clearly




demonstrates that low emission, high efficiency burners can be opera-




ted in subscale.  At this time it is pertinent to consider the EPA's
                                11-233

-------
 COOLING
 WATER
 HOSES
HYDRAULIC
ACTUATOR
HOSES
                                                 WALKWAY
                                                  SPRAY
                                                  MANIFOLD

                                                  WATER
                                                  COOLED PROBE
                                                  IN COMBUSTION
                                                  CHAMBER

                                                  SPECIAL
                                                  APERTURE
                                                  MODULE
                                                 ROLLERS IN
                                                 OUTER SKIN
                                                 SUPPORT
                                                 CHANNELS
Figure 17.  Schematic  Representation Showing the Location and
            Actuating Mechanism of Flame Probe
                       11-234

-------
                     •?j~Jkizgbi
                    *  «°25  5
                     :u«rtO«w;i
                    osoESi-s
                    _li*iOlu-l<«c
                    momctuSo
i  t i  i i  i i

   m
                                    o
                                    id
                                    PN

                                    4-1
                                    CO
                                    0)
                                    H
                                    4-1
                                    3
                                    O
                                    0)

                                    g
                                   O
                                    0)
                                    M

                                    00
                          O
                          s
                          DC
H-235

-------
projected long and short term goals for utility boilers as stated by
Lachappelle, Bowen and Stern.
               Table 1.  Project Short and Long Term Goals
                 for NOX Control from Utility Boilers
                                      ppm at 3 percent
       Gas
       Residual Oil
       Coal
1980 Goal
   100
   150
   200
1985 Goal
    50
    90
   100
The short term goals can be achieved at present under controlled
experimental conditions.  It remains to be seen whether they are
achievable in practical operating plants.  Section 4.0 of this
paper discussed plans to construct a test facility which will pro-
vide much needed information on the characteristics of low emission
combustion systems for coal and heavy fuel oil.  This information
will give confidence in the design concepts and give a clear indica-
tion of whether the goals presented in Table 1 are achievable.
                              II- 236

-------
                              REFERENCES
 1.    Bowman, C.T., The Fourteenth Symposium (International) on
       Combustion, p. 729, The Combustion Institute, 1973.

 2.    Sternling, C.V. and Wendt, J.O.L., "Kinetic Mechanisms Governing
       the Fate of Chemically Bound Sulfur and Nitrogen in Combustion."
       Shell Development Company, Emeryville, California, EPA-650/2-
       74-017 (NTIS No. PB 230-895/AS), August 1972.

 3.    Fenimore, C.P., Thirteenth Symposium (International) on Com-
       bustion, p. 373, The Combustion Institute, 1971.

 4.    Heap, M.P., Lowes, T.M. and Walmsley, R., "Nitric Oxide
       Formation in Pulverized Coal Flames," Combustion Institute
       European Symposium, p. 493, Academic Press, (1973).

 5.    Heap, M.P., Lowes, T.M., "Burner Design Optimization to Control
       NOX Emissions," Presented at the "Coal Combustion Seminar,"
       Environmental Protection Agency, Research Triangle Park, N.C.,
       June 19-20, 1973, EPA-650/2-73-021, (NTIS No. PB 224-210/AS)
       September 1973.

 6.    Pershing, D.W., Martin, G.B. and Berkau,  E.E., "Influence of
       Design Variables on the Production of Thermal and Fuel NOX
       from Residual Oil and Coal Combustion."  Paper presented at
       the AIChE 66th Annual Meeting,  Philadelphia, 1973.

 7.    Wendt, J.O.L., Sternling, C.V., and Matovich, M.A., Fourteenth
       Symposium (International) on Combustion,  p. 697, The Combustion
       Institute, 1973.

 8.    Bartok, W., Engleman, V.S. and  del Valle, E.G.,  "Laboratory
       Studies and Mathematical Modeling of NOX Formation in Com-
       bustion Processes," Esso Research and Development Co., EPA
       report APTD 1168 (NTIS No. PB 211-480), 1972.

 9.    Sarofim, A.F., Williams, G.C.,  Modell, M. and Slater, S.M.,
       "Conversion of Fuel Nitrogen to Nitric Oxide in Premixed and
       Diffusion Flames."  Paper presented at the AIChE 66th Annual
       Meeting, Philadelphia, 1973.

10.    Myerson, A.L., Fifteenth Symposium (International) on Combustion,
       p.  1085, The Combustion Institute, 1975.
                              H-237

-------
11.  Flower, W. L., Hanson, R. K. and Kruger, C. H., Fifteenth
     Symposium  (International) on Combustion, p. 823, The
     Combustion Institute, 1975.

12.  Cernansky, N. P. and Sawyer, R. F., Fifteenth Symposium
     (International) on Combustion, p. 1039, The Combustion
     Institute, 1975.

13.  Heap, H. P., et al. "Burner Criteria for NO  Control, Volume I-
     Influence of Burner Variables on NO  in Pulverized Coal Flames."
     International Flame Research Foundation, IJmuiden, Holland,
     EPA-600/2-76-061a (NTIS No.-Later), March 1976.

14.  Leuckel, W., Swirl Intensitites, Swirl Types and Energy Losses
     of Different Swirl Generating Devices, Doc. G02/a/16, IFRF,
     IJmuiden, Holland (1968).

15.  Heap, M. P., Lowes, T. M., and Martin, G. B., "Fluid Mechanics
     of Combustion," p. 75, The American Society of Mechanical
     Engineers, April 1974.

16.  Tsuji, S., Tsukada, M. and Asai, M., Ishikawima-Harima
     Engineering Review 13, p. 227 (1973).

17.  Lachapelle, D. G., Bowen, J. S., and Stern, R. D., "Overview/
     Environmental Protection Agency NO  Control Technology for
     Stationary Combustion Sources."  Presented at 67th Annual
     Meeting of the AIChE, December 1974.
                                  11-238

-------
1:30 p.m.
The Optimization of Burner Design Parameters
 To Control NO  Formation in Pulverized
 Coal and Heavy Oil Flames
Dr. Michael P. Heap, Ultrasystems
Mike, I didn't understand when you said that you
added methane.  When you added it the NO went up?
If you were to take an average value of what you
would expect, if you were just to prorate it, you
would find out that you didn't get any advantage
by having the gas and the coal fired together.
And you could see that instead of methane if you
put in blast furnace gas, which did also contain a
small amount of methane, that- the emissions were
lower than what you would have expected from just
the algebraic sum of the two.  I believe it has
something to do with the heating rate of the
particles.
That was what was bothering me.  Because  1 had
just an intuitive feeling that if you nave  a fast
heating rate and get the stuff hot in the absence
of oxygen you can get the nitrogen to combine to
molecular nitrogen N?.
In fact, the burner situation was such that you
mixed the coal very efficiently with the air
anyway.  You are just driving off more fuel nitrogen
where the oxygen was available.  The same kind of effect
appears to happen if you increase the temperatures
considerably.
                         H-2'39

-------

-------
PILOT SCALE INVESTIGATION OF COMBUSTION

MODIFICATION TECHNIQUES FOR NO  CONTROL

   IN INDUSTRIAL AND UTILITY BOILERS
                  by
              R. A. Brown
              C. B. Moyer
              H. B. Mason

     Aerotherm/Acurex Corporation
       Mountain View, California
           D. G. Lachapelle

    Environmental Protection Agency
        Research Triangle Park
            North Carolina
                11-241

-------
               PILOT SCALE INVESTIGATION OF COMBUSTION
               MODIFICATION TECHNIQUES FOR NO  CONTROL
                  IN INDUSTRIAL AND UTILITY BOILERS*

                R. A. Brown, C. B. Moyer, H. B. Mason
                Aerotherm Divlsion/Acurex Corporation
                      Mountain View, California

                          D. G. Lachapelle
                   Environmental Protection Agency
                       Research Triangle Park
                           North Carolina
INTRODUCTION
     As part of a continuing effort to provide better NO  control for

stationary sources, the EPA has conducted supporting laboratory experi-

ments to define NO  formation mechanisms, to identify control tech-
                  X

niques, and to guide development work on commercial equipment.  The

EPA facility described here is an important component of this support

effort.  Unlike most laboratory units, it provides a fairly realistic

modeling of the geometry and aerodynamics of large multi-burner boilers.

It can fire a wide variety of fuels, including pulverized coal, and

can simulate almost all combustion modifications for NO  control:
                                                       x
 This project was funded at least in part with Federal funds from the
 Environmental Protection Agency under contract number 68-02-1503,
 68-02-1318, and 68-02-1861.  The content of this publication does not
 necessarily reflect the views or policies of the U.S. Environmental
 Protection Agency, nor does mention of trade names, commercial pro-
 ducts, or organizations imply endorsement by the U.S. Government.
                                11-242

-------
 load changes,  excess air  control, staging, biasing, flue gas  recircu-




 lation, reduced air preheat, and burner modifications.





     The important goals  for this facility will include:





     •  Further definitions of staging and biasing in wall firing of




        high nitrogen fuels (coal and oil) to provide a more  complete




        mapping than presently available of the effects of first and



        second stage stoichiometries, residence times, and mixing




        rates, so as to




        (a)  Guide experiments of full scale equipment toward the




             most effective retrofit technology




        (b)  Point the way toward second generation or optimized KOX




             control including advanced combined techniques, for new




             designs





     •  A duplicate set of goals for tangentially fired units, but with




        the additional topic of burner modifications, an aspect not




        being fully investigated elsewhere





     •  Exploration of NOX control techniques for mixed fuel firing




        (such as high and low nitrogen fuels) and for waste fuels





     •  A quick look at the emission levels of other pollutants such




        as sulfates and nitrates, as influenced by NOX controls





     This paper describes the physical features and capabilities of




the EPA research test furnace facility and the tentative test plans to



achieve these goals.






                                  II-243

-------
FACILITY DESCRIPTION





     Figure 1 shows the furnace facility.  The multifuel/multiburner




furnace is one of the most versatile pieces of equipment of this size




and type in the United States.  Table 1 lists the components and capa-




bilities relevant to the research program, and the ranges over which




parameters may be varied.





     The furnace may have both front-wall or opposed-wall firing, with




from one to five burners in each wall.  Alternatively, the furnace




can accept tangential burners to simulate the design practice of Com-




bustion Engineering.  The combustion chamber can run with or without




internal cooling tubes.





     The front wall fired burners are in two sizes, 300,000 Btu/hr




(five each) and 1.5 x 106 Btu/hr (two each).  These burners, shown in




Figure 2, are patterned after the IFRF design, with variable swirl and




provision to change many other parameters.  Eight CE type burners,




shown in Figure 3, are also provided.  These burners are capable of




variable tilt (±30°), variable yaw (±10°), have interchangeable air




sleeves and fuel nozzles and are capable of being fired on gas, oils




or pulverized coal.  The relative position of the two banks of corner




fired burners is the same as the Combustion Engineering burners.





     Four identical heat exchangers designed to cool a heat input of




three million Btu/hr to 400°F are stacked atop the main combustion










                             II-244


-------
1.   Combustion chamber (39" cube)
2.   Ignition and flame safeguard
3.   Observation ports
4.   Ashpit
5.   1.5 x 106 Btu/hr I FRF burner
6.   C.E.-type corner fired burners
7.   3200°F refractory
8.   Heat exchange sections
9.   Drawer assemblies
10.   Staged injection ports
                                                                                    0
                     Figure 1.   Experimental multiburner furnace.
                                          11-245

-------
                TABLE  K   PRINCIPAL COMPONENTS AND CAPABILITIES
           Component
               Description
Main Furnace Combustion Chamber
Max. refractory temp:   3200°F
Volume:  47 ft5
View ports:  5, 3" dia., 1  10" x 10"
             bottom with periscope
Ignition ports:  6, 1" dia.
                1,5 hole,  28" dia.
                1, single hole, 28"  dia.
                8, corner,  "x"
                  1-5 horizontal opposed
                  1-5 wall  fired
                  4, 8 tangentially fired
                                       Burner  blocks
                                         and plugs:

                                       Burner  mounting:
Ash Pit
Volume:  8 ft3
Max. temp:  2600°F
Heat Exchangers
Sections:  4 refractory lined
Max. temp:  3000°F
Inside dimensions:  25x25x32" high
Drawers:  24 with 20-5/8" b tubes/DWR -
          removable
Length of Dwr.:  32"
Coolant:  Dowtherm
Access ports:  4/section, 1" dia.
Mixing section:  6" above and below
                 drawers (access ports
                 are located in those
                 sections)
Max. heat abs.:  2.2 x 106 Btu/hr
Wt.:  1500#/section w/o drawers
Burners
2 - l.S x 10s Btu/hr Aerotherm/IFRF
5 - 300,000 Btu/hr Aerotherm/IFRF
    •   Interchangeable fuel tips
    •   Interchangeable quarls
    t   Variable swirl
    •   Air sleeves to change velocity

8 - 375,000 Btu/hr Aerotherm corner fired
    •   Three identical circular air &
        fuel inlets/burner
    •   ±30° tilt - all inlets ganged
    •   ±10° yaw - all outlets ganged
    t   Interchangeable air sleeves for
        each port
    t   Interchangeable fuel nozzles
Air Supply

  Primary
•   800 SCFM 6 8 psig
•   Aftercooler:  70°F dew point @ 8 psig
•   Cold control valve & orifice; separate
    heater
  Secondary
    Hot control valves & orifices
                                  11-246

-------
                            TABLE 1.  (Continued)
           Component
               Description
                                           Individual  control  4 measurement  to
                                           16 lines, 8 on each side  of the fur-
                                           nace.   Allows  flow  control of  second-
                                           ary air to  each I FRF burner and con-
                                           trol of annular and secondary  air
                                           flows  to the corner fired burners.
Staged Air
    Staged air manifold parallel to heat
    exchanger stack.  Mixes hot secondary
    and cold secondary air to achieve any
    temperature up to the secondary air
    temperature.  Presently only total
    staged air is controlled.
Heaters
    Secondary air 200 kw max.
    Temperature at the burner: 800CF
    Primary air heater 12 kw max.
    Temperature at the burner:  250°F
    Continuous control from 70°F to the
    maximum temperature for 10:1 flow
    range
Flue Gas Recirculation
    Take off point downstream of baghouse
    Max. temperature:  400°F
    Max. flow:  120 SCFM
    Max. pressure:  2 psig
    Max. firing rate permissable at these
    conditions: 1.5 x 106 Btu/hr @ 10'*
    excess air
    Present introduction point is in the
    secondary air line downstream of cne
    secondary air heater.  (Simple modi-
    fication could be made to introduce
    the flue gases in the stage air, pri-
    mary air or individual burners.
    No FGR heater at this time
Oil Delivery System
•   Up to 25 gal/hr on #2 or #6 oil
•   Single pumping & supply system for
    both oils
t   Max. temperature 46:  220°F at the
    nozzles
•   Two oil manifolds with 8 taps each on
    either side of the furnace
•   Quick disconnect fittings at the
    manifold and burners
•   Flow control valves to each burner
•   Max. pressure:  250 psig
                                  II-247

-------
                             TABLE 1  (Concluded)
           Component
               Description
Gas System
•   Up to 3000 ft'/hr @ 25 psig
•   Manifold with quick disconnects, 8
    outlets on each side of the furnace
t   Shutoff ball valves for each tap on
    the manifold and needle control
    valves for each burner inlet
Coal System
l   up to 250 Ibs/hr of pulverized coal
•   Ten delivery lines to two manifolds.
    one on each side of the furnace.
    Five flexible lines on each manifolc
    deliver pulverized coal to one to
    five burners.  The small lines must
    be recombined when firing the large-
    burners .
•   The coal and primary air flow rates
    are controlled and measured to each
    of the delivery lines.  Uniform dis-
    tribution is obtained from a fluid-
    ized bed distributor.
•   Bagged pulverized coal is fed into ;
    bagdump from the seconc floor level
    and Into a 50 ft3 hopper.  This rep-
    resents about a 1 day supply of coa".
•   The fuel flow may be stopped in the
    event of a flame-out or unstable co--
    ditlon through a solenoid operated
    air purge system.  This purge syster
    is controlled manually or by the
    flame safeguard systeir.
Dowtherm System
    Two Dowtherm-to-air heat exchangers
    can remove up to 2.5 x 10s  Btu/hr
    from the Dowtherm
    A bypass arrangement around these
    coolers allows control of the heat
    removal rate.
Induced Draft Fan
    An induced draft fan with bypass
    allows control of the hack pressure
    in the combustion charter to ±2"
    H20 over the full range of firing
    rates.
                                   11-243

-------
                                                   o>
                                                   c
                                                  a;
                                                  u.
                                                  
-------



Vj




~— -1
^





y.\v
h^ 	
J
                                      s-
                                      3
                                      J3

                                      i.
                                      0)
                                      c

                                      o
                                      o
11-250

-------
chamber.  Each exchanger consists of six drawer assemblies with twenty




stainless steel u-tubes for easy removal of heat exchange surface.




Refractory plugs have also been provided to fill the drawer windows




when the heat exchange surface is removed.





     Figure 4 shows the relationship between heat release rate per




unit volume and firing rate, and indicates typical industrial design




practice for heat release per unit volume.  The upper boundary is




determined with the full heat exchanger surface installed (or using




the main firebox as the combustion volume).  The lower boundary is de-




termined by the required heat exchange surface to lower the gas tem-




perature to 400°F.  A wider performance map could be obtained if




additional heat exchange surface was added in the crossover duct or if




the baghouse was bypassed.   Water cooling could also be used in the




main heat exchangers, requiring less heat transfer surface and thus




allowing longer residence times.  Table 2 further illustrates the cur-




rent performance of this facility by showing the bulk residence time,




and heat release rates for various firing rates.





     Figure 5 shows the general layout of the facility in Aerotherm's




laboratory.






RESEARCH PROGRAM





     The 28 month operating contract for the EPA furnace is subdivided




into the following phases:




     •  Phase I:    Installation and Shakedown (four months)
                                 11-251

-------

                                                 0-
                                                 
                                                 o
                                                 «d-

                                                 
-------
TABLE 2.   FURNACE PERFORMANCE  PARAMETERS FOR
          VARIOUS CONFIGURATIONS
Configurator
( Furnace
plus n empty
sections)
Furnace
Furnace + 1
Furnace + 2
Furnace + 3
Vol
ft3
46.60
57.27
67.93
78.60
Total Heat Release (Btu/hr
3 x 10'
R.T.
0.67/0.83
X
><

K.R.
64,378
X
kX
x
1.5 x 106
R.T.
1.34/1.68
1.65/2.06
1.96/2.45

H.R.
32,189
26,191
22,068


1.0 x 106
R.T.
2.01/2.51
2.47/3.09
2.93/3.66
3.39/4.24
H.R.
21,459
17,461
14,721
12,723*
KEY:
R.T. = Residence time (sec) at 25 percent excess air/stoichiometric
H.R. = Volumetric heat release rate Btu/hr-ft3
£>
-------
j	.	
             _Q»L_-_.U_f
                                                        V)

                                                        i
                                                        •M
                                                        v>

                                                        5?
                                                        8.
                                                        Q.
                                                        

                                                        O
                                                        8
                                                        o


                                                        I
                                                        s-
3




2
                                                        0)
                                                        s_
                             11-254

-------
     •  Phase II:  Conventional Fossil Fuels Test Program (12 months)





     •  Phase III:  Mixed and Nonconventional Fuels Test Program (12




        months)





The Phase I shakedown tests served to certify the system design and




also yielded the operational and control characteristics needed to




execute the subsequent test program.





     The objective of the Phase II test program is to develop advanced




techniques for the control of combustion generated air pollution for




application to large and intermediate sized steam generators firing




conventional fossil fuels.  For this series the furnace will be run in




the single wall fired mode using five 300,000 Btu/hr burners and in




the corner fired mode using four or eight burner arrays.  Emphasis




will be on fuels with high fuel NO forming potential — coal and re-




sidual oil — with secondary emphasis on distillate oil and natural




gas.  Western Kentucky #9 coal and Chevron bay area residual oil (S =




1%, N « 0.8%) have been selected as the primary fuels.  Two other coals,




Pittsburgh #8 and Big Sky Montana, and two other residual oils will be




fired periodically to generalize the results obtained with the primary




fuels.





     The Phase II test plan is shown schematically on Figure 6.  The




initial furnace characterization studies will define the system oper-




ating conditions and baseline emissions to be used to interpret the
                                11-255

-------





Phase II
Fossil Fuel-
Program


—Furnace ____
Characterization

^_ Pr«>1 iminary . . ,___.
Studies


Baseline
Emissions



Control
Development







1 — Wall Cooling
Wall fired
Mall Cooling
Corner Fired
	 Burner Array
Corner Fired
| — Wall Fired
Corner
Fired
Fuel NOX
Wall Fired
	 Fuel NOX
Corner Fired
1st Stage
Variables -
— Coal
	 2nd Stage var-
iables -on
2nd Stage Var-
iables - Coal
2nd Stage Var-
iables - Oil
	 Combined Con-
trols - Coal
, Combined Con-
trols -011
r" •"»!•»

Burner
— Parameters
	 Flue Gas
Recirculation
	 Combined
Controls
II-A
1I-B
II-C
II-D
II-E
Il-F,
II-F2
II-H,
II-H2
n-J1
II-J2
II-Kj
II-K2
II-L
II-M
Il-N
II-P
II-Q
20
Points
(20)
(4)
(87)
(97)
(15)
(15)
(42)
(42)
(21)
(21)
(65)
(65)
(23)
(50)
(46)
(20)
(25)
Figure 6,   Summary of Phase II test sequences.
              H-256

-------
results of the control development studies.  The preliminary wall
cooling studies, II-A and II-B, will indicate the level of wall
cooling needed to give the best simulation of NOX formation in large
steam generators.  The determining factors will be the level of NO
as well as the trend with excess air, preheat and load for at least
two levels of wall cooling.  Gas and coal will be fired to encompass
the range of thermal/fuel NO  and the range of luminous versus gas
radiation encountered in practical equipment.  The wall cooling scheme
established in II-A and II-B will be carried through all remaining •:
test series.  The third preliminary study, II-C, will determine if
four or eight corner fired burners yield the best simulation of NO
formation in full scale equipment.

     The uncontrolled baseline emission series II-D and II-E will
provide reference emission levels over the range of preheat, excess
air, load and fuel type to be considered in the control development
series.  The baseline series will also give further characterization
data to indicate the correspondence of this facility to full scale
equipment.  An additional baseline series, II-F, will be run to elu-
cidate the thermal NO /fuel NO  split as affected by fuel type, pre-
                     A        X
heat and excess air.  Concurrent tests at the University of Arizona
In which argon is substituted for nitrogen in the combustion air will
indicate the effect of flue gas recirculation on thermal NO  and
                                                           X
thereby show the fractional conversion of fuel nitrogen to NO. These
                                   11-257

-------
results will be translated to the EPA/Aerotherm furnace run with a




duplicate matrix and the same fuel to give semi-quantitative guide-




lines on fuel nitrogen conversion.





     The control development test series will be largely devoted to




determining the best procedures for NO  control through staged com-




bustion.  The wall-fired control development series will involve a




systematic screening of primary first stage variables and second




stage variables followed by combined testing of the best procedures




identified in the screening study.  The first stage screening study




will investigate the following variables:





     •  Primary flame zone mixing due to burner swirl





     •  Primary flame zone stoichiometry





     *  First stage flame heat removal





     •  Bulk residence time to the point of second stage air injection





     •  Burner air preheat





The NO  control results from these tests will be interpreted through
      X



the time-temperature history of the combustion gases up to the point




of second stage air injection as affected by the above variables.




Since the time-temperature histories and combustion intensities in




this facility are similar to full scale equipment, these results will




serve to establish semi-quantitative guidelines for staging in field




application.
                                 II-258

-------
     The second stage screening study,  II-J , will investigate the
following variables:

     •  Residence time from second stage air injection to quench

     •  Stage air preheat

     •  Rate of mixing of stage air with combustion gases

These results will compliment the II-H  results to indicate the best
overall approach to staging which will be tested in series II-K.
Throughout the staging studies, measurement will be made of combustion
gas temperature, local NO  concentration and residence time distribu-
                         X
tion by means of the suction pyrometer, hot sampling probe and helium
tracer measurements respectively.  These measurements will provide
insight into how staging variables affect N0x formation and control
and will also allow some generalization of the results toward full-
scale application guidelines.

     Flue gas recirculation for NO  control with dirty fuels, pri-
marily residual oil, will be studied on a limited basis in series
II-L.  The objectives are (1) determine the effectiveness of FGR as a
function of fuel type, excess air and preheat, and (2) illustrate
qualitatively how fuel NO  conversion is affected by primary system
                         A
variables.  These FGR tests will actually be run in advance of the
staging studies.  If the FGR technique shows promise, it will be In-
cluded in the combined control optimization series, II-K.
                                  11-259

-------
     The corner fired control development series will extend the




staged combustion guidelines established in the wall fired series to




the corner fired configuration.  Additionally, the corner fired series




will study the effect of burner parameters on NO  formation and con-
                                                •X



trol.  Series 11-M provides for study of the following corner fired




burner configurations:





     •  Burner firing orientation





     •  Eccentric fuel/air injection





     •  Biasing of burner stoichiometry between the upper and lower




        nozzles





These tests will identify low NO,, configurations and will indicate
                                Jt



the effects of hardware settings on fuel/air mixing in the jets and




in the central fireball.  The observations on fuel/air mixing will




guide the selection of burner configuration to be used in the sub-




sequent studies on staging, FGR and combined controls.





     The Phase III mixed and nonconventional fuels test program has




the objective of defining practical design guidelines for emission




control with the firing of nonconventional fuels.  Mixed fuel firing




shows promise as a new and powerful NO  control option and so will
                                      X



tentatively occupy eight months of the 12 month test program with
                                11-260


-------
nonconventional fuels covering the remaining four months.  Mixed fuels




approaches considered for Phase III include:





     •  Blends of fuels to achieve low net fuel N





     •  Simultaneous firing of two fuels through two separate sets




        of burners to provide a shielding of the high N fuel and




        hence a delayed exposure of fuel N to combustion air, sus-




        pected to be an effective control technique for fuel NO
                                                               X




     •  Simultaneous firing of two fuels through separate elements of




        the same burner to provide a shielding effect on the scale of




        an individual burner*





     •  Staged fuel addition, or "reburning," to reduce NO in a sec-




        ondary combustion zone





     Table 3 lists the candidate fuels for both individual firing and




dual fuel firing, and briefly indicates the selection rationale in




each case.  The individual tests are limited to important potential




boiler fuels:  oil shale product, solvent refined coal, coal char,




and one high nitrogen waste fuel.  The particular selections are in-




fluenced by the prototype characters of these fuels as being repre-




sentative of general classes of new fuels and as having interesting




NO  control difficulties.  The only conspicuous absence from this
  X



single-fuel-firing list are synthetic fuel gases produced from coal.
*For wall firing, other EPA programs currently are developing this

 technology.  Consequently, the stress here will be on tangential firing.
                               II-261

-------
o
1-^

<


CQ

O
   CD
U. UJ
Q i— i
   LU
   GO
a: a.
o
a. c£.
SO



(U
Li_
5?
»^-
O





S-
ns
.c
<_>
m
o
o
O
Ol
(O
.c
<^
-o
(/>
s
p**

-------
                          ACKNOWLEDGEMENTS






     The contributions of D. W.  Pershing and Professor J.  0.  L.  Wendt,




of the University of Arizona, to the formulation of the test  plan is




gratefully acknowledged.
                              11-265

-------
         TABLE OF CONVERSION FACTORS FOR  SI UNITS
Equivilant SI Units
 0.0254m
 0.3048m
 0.028317m3
 0.45359Kg
 6895 N/m2
 0.003785 m3
 1055.06
 37,259 J/m3
 (Tf-32) /1.8°C
English Units
   inch
   Ft
   Ft3
   1b
   Psi
   Gallons
   Btu
   Btu/Ft3
   °F
                        H-266

-------
2:25 p.m.
Pilot Scale Investigation of
  Combustion Modification Techniques
  for NO  Control in Industrial and
  Utility Boilers
Richard E. Brown, Aerotherm Acurex
In noticing most of the presentations of yesterday
and today, the effect of wall temperatures on NO
                                                X
emissions just seems to be one of the most pronounced
effects.  Now I noticed how with your U-tubes you  can
affect a cooling rate of the flue gases, but they are
not down in the furnace.  By the time the gases are
going up through those cooling tubes there is going
to be no back mixing into the flame zone.  Maybe I
am wrong, but I see that this furnace is compared to
what everyone else seems to see as the most important
parameter, that's wall temperature -- and I don't
see how we can work that into...
We will be putting cooling walls in this furnace,
and vary the spacings to look at various cooling
wall patterns.  That is part of our base line test
series.
Will you be testing two-stage combustion on a single
burner setup?
Perhaps.  Burner configuration will be determined during
our baseline series and we think it will probably be the
five burner array.  Most likely we will not look at
stagedcombustion for the single burner until we go
to some horizontal extensions.  This furnace is also
designed such that we can go to horizontal extensions,
either end.  And in that case we may look at two-staged
combustion with the single burner.
                        II-267

-------
Q:  The word "shielding," could you interpret that please?
A:  The effect of introducing fuel and air in either a
    fuel-rich or fuel-lean mode, and providing another
    fuel to shield the main fuel, say, from the oxygen.
Q:  What method are you going to use to preheat your air?
A:  We are preheating the air with electrical heaters up
    to 800° F.
Q:  And on the flue gas recirculation, are you taking it
    directly from the stack?
A:  We are taking it downstream of the baghouse; we are
    taking the particulate and the fly-ash out.
Q:  And where are you reinjecting it?
A:  We will reinject it at this time into the secondary
    air streams, although we have the ability to inject
    it at any location.
                            II-268

-------
          SESSION III




PROCESS RESEARCH AND DEVELOPMENT







   D. G. Lachapelle, Chairman
             III-l

-------

-------
       OVERFIRE AIR AS A NO  CONTROL
                           x


TECHNIQUE FOR TANGENTIAL COAL-FIRED BOILERS
                    By:



               A.  P.  Selker



       Combustion Engineering,  Inc.



           Windsor,  Connecticut
                   III-3

-------
                       OVERFIRE AIR AS A NO  CONTROL
                                           x
                TECHNIQUE FOR TANGENTIAL COAL-FIRED BOILERS

                               A. P. Selker
                       Combustion Engineering, Inc.
                           Windsor, Connecticut

     INTRODUCTION

     The purpose of this EPA sponsored program was to determine the
     effectiveness of overfire air (OFA) and biased firing* as methods for
     NOX control on tangentially coal fired utility boilers.   In
     addition, the effects of these emission control techniques on overall
     unit performance were evaluated.  The program was divided into two
     contract phases:

     Phase I (EPA Contract 68-02-0264) consisted of selecting a suitable
     utility boiler to be modified for the incorporation of an overfire
     air system and conducting studies in preparation for the actual unit
     modifications and test program.  This phase was completed in 1973.

     Phase II (EPA Contract 68-02-1367)  consisted of modifying and testing
     the utility boiler selected in Phase I to evaluate overfire air and
     biased firing as methods for NOX control.  This phase also included
     the completion of detailed fabrication and erection drawings, instal-
     lation of analytical test equipment, updating of the preliminary test
     program, and analysis and reporting of test results.

     This program was conducted with the cooperation and active participa-
     tion of Alabama Power Company at their Barry Steam Station Unit 2.
     This unit is a natural circulation, balanced draft utility plant
     boiler firing coal through four elevations of tilting tangential fuel
     nozzles.  Unit capacity at maximum continuous rating (MCR) is 408,000
     kg/hr main steam flow with a superheater outlet temperature and pres-
     sure of 538 C and 131.8 kg/cm^.   Superheat and reheat temperatures
     are controlled by fuel nozzle tilt  and spray desuperheating.   A side
     elevation of the unit prior to modification is shown in Fig.  1.

     OVERFIRE AIR SYSTEM DESIGN

     The overfire air system (Fig.  2) incorporated in the unit provided
     for introducing a maximum of 20% of the total combustion air  above
     the fuel admission nozzles at  full  unit loading.   The overfire air
     was introduced into the furnace tangentially through two separate
     *Biased firing consists of removing individual fuel elevations from
     service while maintaining full airflow through these same elevations.
                                 III-4
I

-------

                            ''   ^''
Fig.  1:  Test unit side elevation, Alabama Power Company, Barry 2





                                III-5

-------
           F- FUEL AND AIR
           A-AIR
           0-OVERFIRE AIR
        Fig. 2:  Test unit overfire air system schematic

compartments near each furnace corner located  approximately  2.4 meters
above the fuel admission zone.  While on new unit  designs  the over-
fire air ports (Fig.  3) are designed as vertical extensions  of the
corner windboxes, on the field modified unit,  separate  compartments
were required due to structural considertions.  The  system provided
for air nozzle tilting (-30 deg.  from vertical  plane) and  separate
compartment flow control dampers to permit  studying  the effects of vari-
ous airflow rates,  introduction angles, and compartment airflow dis-
tributions.
                               III-6

-------
      WINDBOX
SECONDARY AIR
       DAMPERS
SECONDARY AIR
 DAMPER DRIVE
           UNIT
                                                   OVERFIRE AIR
                                                   NOZZLES
                                                    SIDE IGNITOR NOZZLE
                                                    SECONDARY AIR NOZZLES
                                                    COAL NOZZLES
HIGH SWIRL OIL  GUN
            Fig. 3:   Corner windbox showing overfire air system

      TEST INSTRUMENTATION

      In order to evaluate the effect of  using overfire air as a combustion
      process modification technique for  emissions control, a grid of stain-
      less steel flue gas sample probes was installed on the unit's econo-

                                    III-7

-------
mizer and air preheater outlets as shown in Fig. 4,
      Fig. 4:
                                            A.  ECONOMIZER OUTLET

                                            B.  AIR HEATER OUTLET

                                               (Oj ONLY)
STAINLESS STEEL  FLUE GAS
 SAMPLE  PROBE  LOCATIONS

Schematic showing location of sample probes
The gas sample from the economizer outlet was blended, transported,
and preconditioned prior to analysis by the following instrumentation
(Fig. 5).

1.  A Scott chemiluminescence NO-NOX analyzer (0-2000 PPM)
2.  An L&N paramagnetic 02 analyzer (0-25% 02)
3.  Beckman non-dispersive infrared CO analyzer (0-1000 PPM)
4.  A Scott flame ionization Total Hydrocarbon (THC) Analyzer (0-100
    PPM)

A schematic of the gas analysis system is shown in Fig. 6.

Unit performance was monitored using the instruments and analytical
procedures shown in Table I.

TEST PROGRAM

The Phase II test program called for the evaluation of overfire air
as an emissions control system.  Also evaluated were existing process
variables, including excess air, unit loading, and air/fuel distri-
bution.

To properly assess the effect of the overfire air modifications on
unit performance and emission levels, the test program was divided
into two distinct phases.

1.  The first phase, run prior to modification, consisted of baseline
                              III-8

-------
                  TABLE I
        INSTRUMENTS AND ANALYTICAL
PROCEDURES FOR MONITORING UNIT PERFORMANCE

PARAMETER
Flow rates
Temperatures
Pressures
Laboratory analysis
Steam and Water
Feedwater flow
Reheat and superheat
desuperheat spray
Reheat flow
Air and Gas
Total Air and Gas Weight
Overfire air
Air heater leakage
Steam and Water Deg. F
Unit absorption rates
Waterwall absorption
Air and Gas Deg. F

Steam and Water PSIG
Unit absorption rates
Unit draft loss
Temperature and Pressure
Fuel and Ash
INSTRUMENT/ANALYTICAL PROCEOURJ
Flow orifice
Heat balance (deg. F & PSIG
around desuperheater
Heat balance around reheat
extraction and estimated
turbine gland seal losses
Calculated
Pitot traverse
Paramagnetic 0. analyzer
Calibrated stainless steel
sheathed CR-C well & butti
TC's
Calibrated stainless steel
sheathed CR-C chordal WW '
CR-C TC's
Water cooled probes
Pt/Pt-IOZ Rh TC's
Pressure gauges and/or
transducers
Water manometers
C-E data logger
capacity: 400 tentperatur
50 pressures
ASTM procedures
                       -, 9

-------
Fig. 5:  Gaseous emissions test system
                 111-10

-------
     JUL
SAMPLE PROBES
GAS DUCTS
                      -TTTTn-fTrTfHii m iTi
                 .1.1. JL
                    SAMPLE PROBES
                                        SAMPLE BLENDER
                                               PRECONDITIONING
                                               CONSOLE




FILTER p=
T~
PUMPQ

^^
1




1


3 r-0-i
COND
*H2O
H — h n
QD
i
3RYER 1
(~U J 1

'
H20_T__


r










1 1 FILTER

THERMAL
coNvl—l _f4
uumv. ij 0_ | 	 t F|LTER
I
*H2° O PUMP
1 n
1 °2 D —

1 NO 	 T
|HUx |_ 	 1
|L
(3 	 VACUUM PUMP
        CONSOLE 1
                 CONSOLE 2
                 Fig.  6:  Gas analysis system

    and biased firing programs as well as 30-day waterwall corrosion
    coupon studies at baseline and optimum biased firing conditions.

2.  The second phase, conducted after unit modification, consisted of
    repeating the baseline evaluation (to determine if the modifica-
    tion had influenced  unit performance) and testing the overfire
    air system.  A 30-day waterwall corrosion coupon evaluation was
    also conducted at the optimum overfire air condition.

The originally planned evaluation of optimum operations with alternate
coal types could not be  performed due to difficulties in obtaining
the necessary coal supplies.  However, this information has been re-
                          Hi- 11

-------
 ported by others as a result  of  tests  on another unit at the same
 station.'-1)

 RESULTS

 The first phase of the test program characterized the baseline emis-
 sion levels and unit thermal  performance including the effects of
 various operating variables including  unit  load,  excess air level,
 and furnace wall deposit variation.  In  addition,  tests were conducted
 to evaluate the effect of biased firing  on  NO   emissions.

 Load and Excess Air Variation

 Figure 7 shows that the NOX emission levels increased with  increased
 excess air but did not change significantly with  changes in unit  loading.
                   BASELINE TESTS BEFORE  MODIFICATION
1,5
; 4
1 3
1SPS
1 2

1 Q
0 9


n fi




















X»








/







4
/
P^

Tl






,
^








/
r
^







yS









^,










*~ —
4









• •









^



























LEGEND
Unit LMjl Furiact Sin
Q HC» OL19«
Q 3/4 «CR O Noderat*
O vz MCR i"«»y


                   no     i?o     no     no     iso

                    THEORETICAL AH TO FUEL FIRING 2OIE, PtRCENT
                                                 160
Fig. 7:  K02 vs theoretical  air  to fuel furing zone — baseline study

An average increase of 0.017  g/10   cal  NCL  was noted for each 1% change
in unit excess air.  Acceptable  CO limits of 0.05 g/10^ cal establishes
minimum excess air levels of  20  to 25%  (Fig.  8).   (Throughout this re-
port, NO  emission levels are expressed as  g/10"  cal NO-.)
        X                                               £*

On these plots, excess air (EA)  is expressed as percent theoretical
air (TA) to the fuel firing  zone which,  for unit  operation without
                               111-12

-------
                 BASELINE TESTS BEFORE flDDIFICATION
        0.6
        0.5
      •» 0.3
      s
                110    IJO    110-     140     150     1*0

                  THEORETICAL All! TO FUtL FIRING ZONE. PERCENT
Fig. 8:  CO vs  theoretical  air to fuel firing zone — baseline study

overfire air, can be  considered to be directly comparable (20% Total
EA - 120% TA).

The excess air  and  load variation tests were conducted at what were
considered to be clean furnace conditions to establish a baseline for
wall deposit variation tests.   Furnace wall deposit variations were
determined by visual  observations by C-E test personnel.

FurnaceWall Deposit  Variation

Changes in furnace  slagging did not exhibit  a discernable  effect  on
NOX or CO emission  levels  (Figs.  7 and 8) with the exception of two
half-load tests when  higher CO levels were obtained at slagged as com-
pared to clean  furnace conditions.   The reasons for this condition are
not completely  clear  and the higher CO levels were not observed under
full load heavy slag  operation.

Maximum slagging conditions were obtained by operating the test unit in
excess of 24-hour periods without operating any wall blowers.  In this
manner, slag deposits up to 4-inches thick could be obtained in and
above the firing zone.  Wall deposit variations were again determined
                              III-13

-------
 by visual observation.

 The slagging conditions indicated should not be considered representa-
 tive of normal operation at the various excess air and unit load
 levels  studied,  but were specifically controlled for test purposes by
 varying soot blower operation.

 Biased  Firing Variation

 Figure  9  shows the results of removing individual fuel elevations from
 service and  operating them on an air-only basis.  Maximum NOX control
was  obtained with  top elevation of fuel nozzles out of service at max-

               B1ASED FIRING TESTS BEFORE  MODIFICATION
1 3
ISPS
1 2
1 ^
IfS

06
07
Of
OC
»








>
— r
0







*/
;

i






/
/


00


f



/



1



A

/




10





s^
O



i



O
^





20




•""





13



,-*•






0



---•






1

I

-"
•





4D










li
                                                   Unit iMd

                                                   O N» Pots.
                                                   (B 3/4 NCR
                                                   0 1/2 NCR
                                                           Fuel Nozzles
                                                           Out af Sent.
Fig. 9:
                 THIORETICAl AIR TO FUEL FIRMS ZONE, PERCENT
NO, vs theoretical air  to  fuel  firing zone — biased firing
study
imum and 75% maximum  loading.

At 50% percent maximum  loading,  the high excess air levels required to
maintain unit steam temperatures appeared to negate any NOX reductions
obtained by biasing the top fuel nozzle elevation; however, the emis-
sions level obtained  was below the current EPA limit (NSPS*) for coal
fired units of 1.26 g/106  cal.

It should be noted that when any of the lower three fuel elevations
were biased (out of service),  the air introduced through the inactive
fuel compartment was  included in the percent TA to the fuel firing
*New Source Performance Standard
                                III-14

-------
zone.

As can be seen  from Fig.  9  biasing any of the bottom three fuel
elevations did  not  have a discernable effect on NOX emission levels,
although the  emission level tended to be higher at reduced  unit  load-
ings for given  percent  TA levels.

Figure 10 indicates that with biased firing, low percent TA levels to
the fuel firing zone were obtained without increasing CO emission
levels.  This condition is due to the ability to maintain acceptable

                 BIASED FIRING TESTS  BEFORE  MODIFICATION
0.3
f-i
0.1
0
90







d







A















I







*















9






























•
•








100 110 \?0 130 140 15
                                                          LEgHP
                                                     Unit Load
                                                             Fuel Nozzles
                                                             Out of Str».
                                                      Mii. Pass.
                                                      3/« NCR
                                                      ^n nc«
                                                     Top Ctr.
                                                     llol. Ctr.
                                                     Sot to*
                     THEORETICAL AIR 10 fUCL FIRING 2011. PERCEKT

Fig. 10:
CO vs theoretical air to  fuel  firing zone — biased firing
study
total unit excess air  levels  as measured at the economizer outlet
during biased firing operation.

The second phase of the  test  program was conducted to determine base-
line operating characteristics of the modified unit and to compare
these to the unmodified  unit  test results.  The overfire air system
optimization program then investigated the effects of varying over-
fire air locations, tilt angles,  and introduction rates with respect
to emission levels and unit  thermal performance.
Load and Excess Air, and  Furnace Wall Deposit Variation (After Modifi-
cation)
The baseline test program conducted prior to unit modification was  re-
peated after installation of the overfire air system with similar re-
                              Ill- 15

-------
suits with respect to NO^-  and CO emission levels being obtained  as
shown in Figs.  11 and 12.   The NO  levels after modification


                   BASELINE  TESTS AFTER  MODIFICATION
1.3
NSPS
1.2
1.1
8 '-0
%
o. 0.9

0.8
0 ?
0.6
0.5










X










X







9



X
B°






**
4

^










•^

<







X
V









X











•^










X










x









.
/










*s




























U6EHO

Unit Lead Furntcc SIM
O NCR O Light
O 1/4 NCR 0) NMtertU
O UZ NCR « HN«y


          100     110     1ZO     130     140     ISO

                       THEORETICAL MB TO FIRING ZONE. PERCENT
Fig.  11:   N02 vs  theoretical  air to firing zone, overfire air  study,
           load and  excess air variation  — modified  boiler
                   BASELINE  TESTS  AFTER MODIFICATION
0.4
,- 0-3
g 0.2
ai
0
10
















k
(1^


O






p
x
& t








^
t> 4









r~~-a










~Q-~






















































J










^

0 110 120 130 140 ISO 16
                                                            LEGtND

                                                        Unit LotJ  Funiic* Sin
O NCR
a 3/4 NCR
O 1/2 NCR
                                                               0 Light
                                                               9 NMtrite
                     THtORCTtCU. AIR TO FUEL FIRING ZONE. PERCENT
Fig.  12:  CO vs  theoretical  air to firing zone, overfire air  study,
           load and excess  air variation — modified boiler
                                III-16

-------
tended to be slightly lower than pre-modification results;  however,
this could have  resulted from updating of the firing system (replace-
ment of fuel/air nozzles,  etc.)  during modification and a  slight  re-
duction in fuel  N2  levels  and higher heating values.  Again,  no dis-
cernable effect  on  NOX emission levels are noted due to furnace wall
cleanliness.

Overfire Air (OFA)  Variation

As shown in Fig. 13,  the effect on NOX levels of varying the  OFA  ad-
mission points was  investigated with the result that NOX levels de-
creased with decreasing percent TA to the fuel firing zone,

                 0 V £ P. f I R E  AIR LOCATION VARIATION
1.3
NSPS
1.2
1.1
3
io.a
0.7
0.6
0.5
0.4
0-3

	






	

	 |







^








, 	
	
	 	

	





	
^

- -






— — -








_ — -—


-








•"tS


w
•4 —
, -. 4>
3^-
r"





^
^-*
rip


—



Of*
1 ^**

t — •

- IOMIR LIMIT OF
11






H- *l

^— —




4
-""
,**
-~~~

ACCEPTABLE


	




vt^


A LEVE

•- •


^

^


LS


-

c
^-*

—





- •>

rt







80 85 90 95 JOO 105 110 115
LECtjID
UK. Pis.
A 0-1
& 0-2
0-3
tote
& No OF*
A 1/2 Man. OF*
A Man. OFA
ABC (§}
BCD (H)
                     THEORETICAL AtR TO FIRING ZOHE, PERCENT
        N0? v^ Theoretical Air to Firing Zone, Overfire Air Location, Rate 4 Velocity Variation

Fig. 13:   N02  vs  theoretical air to firing zone, overfire air location,
           rate and  velocity variation

The  importance of using firing zone percent TA  levels  now becomes
apparent  as  this  procedure provides a mechanism for correlating test
results while  total excess air levels as measured at the  economizer
outlet provide no meaningful base for data correlation.

NOX  levels were found  to be generally higher with the  top 3 elevations
of fuel nozzles in  service than with the bottom three  elevations.  We
think this is  due to the increased distance between the fuel firing
zone and  the admission point for the OFA.
                                111-17

-------
 Operation at TA levels below  95%  did not result in significant reduc-
 tion in NOX emission levels.  As  shown in Fig.  14, CO levels remained
 acceptable for the duration of  the  test series.  Total excess air as
 measured at the economizer outlet was maintained at approximately 27
 percent.
                       OVERFIRE AIR  LOCATION  VARIATION
        0.4
        0.2
                                 LOHE8 LIMIT OF ACCEPTABLE TA LEVELS
  LEGtM)

 Mm. Pts

£ 0-1
t. 0-2
b o-i. 0-2
O 0-3

   tote
& No OF*
A 1/2 MM. OFA
A N>x. OFA

mm in _5tr».

 ABC ®

 KO 151
                      M
                            95      100     105     110

                          THEORETICAL AIR TO FIRING ZONE, PERCENT
                                                     115
                                                           120
Fig.  14:   CO vs theoretical air to firing zone,  overfire air location,
           rate and velocity variation

Overfire Air Tilt Variation

A test series was conducted at full unit loading to  determine the
effect of  varying OFA and fuel nozzle tilts on NOX and CO emission
levels.  These tests were conducted with total excess air and TA
levels to  the firing zone of 24 and 92 percent,  respectively, and mod-
erate waterwall slagging conditions.

Essentially,  varying fuel/air and OFA nozzle tilts moved the firing
zone both  in the furnace and relative to the overfire air.   Fuel
nozzle tilts that are maximum minus combined with OFA tilts of maxi-
mum plus increased the distance between the overfire air and the fir-
ing zone.  As with previous methods of increasing this distance, the
NOX emission levels were decreased.  Figure 15 shows that as the tilts
were moved toward one another (fuel nozzle tilts up,  OFA tilts down),
the OFA/firing zone separation was decreased and the NOX levels were
increased.

When the OFA tilts were maximum minus and the fuel nozzle tilts maxi-
mum plus,  the term overfire air became ambiguous and the actual over-
fire air was  less than the calculated value, as  the  air was being
                               III-18

-------
                        OVERFIRE  AIR  TILT  VARIATION
             0-9
             0.8
             0.6
60
                      SO  40  10  20  10  0  10  20  30  40  SO  SO  70

                       TOWARD EACH OTHER            AMAY fdOM EACH OTHER

                            OFA TILT AND FUEL NOZZLE TUT A . DEGREES
 Fig.  15:   N02 vs OFA tilt and fuel nozzle tilt  differential,  OFA tilt
           variation

 forced down into the raised firing zone.  At this point,  where the
 combined  fuel nozzle and OFA tilt differential  was  52  degrees toward
 each  other, the NOX emission level reached a maximum of 0.846 g/106
 cal.

 CO  emission levels exhibited an increase as the tilt differential  in-
 creased;  however,  enough total excess air was available to maintain
 an  acceptable emission level as shown in Fig. 16.

 Flame stability became a limiting factor as the tilts move substantial-
 ly  away from each  other (40 - 50 degrees).   NOX emission  levels  de-
 creased to  0.596 g/106 cal;  however,  the CO emission levels increased
 and the fire appeared  less stable.   Maintaining the fuel  nozzle  tilts
 and OFA tilts at approximately equal tilt angles resulted in  accept-
 able  flame  stability as well as reduced NO  emission levels.
                                           x
 For all OFA tilt variation tests,  the NO  emission level  obtained was
well  below  the EPA limit of  1.26 g/106 cal.

Load  Variation at  Optimum Conditions

Tests were  conducted to evaluate unit performance and emission levels
at optimum  operating conditions  as  determined during the overfire air
variation tests.   Tests were conducted over  the unit load range at vary-
                               Ill- 19

-------
                       OVERFIRt  AIR  TILT  VARIATION
          _ 0.05 	
          SO. 04
           0.03
           102
             70  60   50  40  30  20  10  0  10  20  30  40  50  SO  70

                     TONARC UCH OTHCfl            AHA* FROH EACH OTHES

                          OF* TILT AND FUEL IWZ2LE TUT A. DEGREES
Fig. 16:  CO vs OFA  tilt  and  fuel nozzle tilt differential, OFA tilt
          variation

ing furnace waterwall  slagging conditions.   The NO  emission level re-
sults of this series of test  versus unit loading, expressed as main
steam flow, are shown  in  Fig.  17.

The wide range of N02  levels  obtained, particularly during the base-
line tests are due to  variations in unit operating parameters such as
excess air level.  During the optimization tests, total excess air at
the unit economizer  outlet was maintained between 20 and 28% at full
and 3/4 load and 45  to 47% at 1/2 load with fuel nozzle tilts varied,
as required, to maintain  acceptable reheat and superheat outlet temper-
atures.  Also, minimum excess air levels were established on the basis
of maintaining acceptable CO  emission levels and flame stability.

Furnace Performance

During the test program,  furnace performance was monitored by use of
chordal thermocouples  installed in the furnace waterwalls.  Tempera-
tures and corresponding absorption rates were found to vary signifi-
cantly with wall slag  conditions making data interpretation difficult.
The method finally arrived at as representing an accurate indication
of furnace performance was as follows:
                              111-20

-------
              BASELINE AND 0 P T I fl U !1 UNIT  OPERATION
                                                      LEGEND
                                                   G Baseline Tests

                                                   A Optimization Tn
                             STI A" FLOW - 10J».r.,'HR

                     PERCENT Of fi'U IOA0 fiflTINO
Fig. 17:  N0? vs main  steam  flow,  ranges for normal and optimum opera-
          tion

The front and right  side wall  centertube profiles were determined and
the average of these plotted as  shown in Fig. 18.  It should be noted
that the maximum and minimum profiles shown do not represent individual
walls in every case, i.e., at  given furnace elevations the maximum
rate shown may switch  from wall  to wall.

For comparison of optimum and  normal unit operation with respect to
furnace performance, three full  load tests with similar furnace slag-
ging conditions, etc., were  selected.   The average centerline profiles
for these tests were determined  and plotted together as shown in Fig.
19.  Furnace performance remained  essentially unchanged when furnace
slagging effects were  taken  into account.

Waterwall Corrosion  Coupon Evaluation

Following completion of the  steady state phases of the baseline,
biased firing, and overfire  air  test programs, 30-day waterwall cor-
rosion coupon evaluations were performed to determine whether any
measurable changes in  coupon weight losses could be obtained for the
various firing modes studied.  The individual probes were exposed at
                              111-21

-------
        &
        UJ
        i
        *
        §

        s
                                                     Date:  6/28/74
                                                     Load:  125 MM
                                                     Furn. Absorp.:  153.17 1
                                                     Total Absorp.:  276.4 1
                                                     TA to Fuel Firing Zone:
                                                     Total Excess Air:  25.9
                                                         KG-CAL/HR
                                                        KG-CAl/HR
                                                        119.9 J
                                                     O -  Front HW Center Tube Profile
                                                        -  Right HW Center Tube Profile
                                                        -  Avg. Center Tube Profile - Both
. FUEL EL£V.  A
<-'-;	t-"-^
"  FUEL ELEV. B
                     FUEL ELEV. D  •
                   	.,_	.	i  	I
                       24     6     6    10    12    14    16   18


                                TUBE CROWN ABSORPTION RATE, KG-CAL/HR-CK2
                                                  20    22
                                                             24
                                                                  26   28
Fig.  18:   Average  centerline absorption  profile — full  load  operation
                                            111-22

-------

g
—


8*
£  18.29

d   (60)
         "2    4    6    8    10   12   14   16   18   20  22   24





                    TUBE CROWN ABSORPTION RATE, KG-CAL/HR-CM2
26   28
 Fig. 19:   Average  centerline absorption profile for  all tests  — full

            load operation with  and without  overfire air
                               111-23

-------
 five  locations on  the furnace wall with coupon temperatures being main-
 tained at the same levels for each 30-day run.

 The individual coupon weights were determined before and after each
 30-day test and the individual coupon and average probe weight losses
 determined.  The weight losses calculated for the biased and overfire
 air portion of the test program were found to be greater than those for
 the baseline test.  The average weight losses for all five probes were:
  Baseline
2.6381 mg/ctn
Biased Firing
            2
4.6429 mg/cm
Overfire Air
4.4419 mg/cm
These values are within the range of losses which would be expected for
oxidation of carbon steel for a 30-day period.  To verify this premise,
control studies were conducted in C-E's Kreisinger Development Labora-
tory using probes exposed during the biased firing study.  These probes
were cleaned and prepared in an identical manner to those used for fur-
nace exposure and placed in a muffled furnace for 30 and 60-day expo-
sures at 400 C with a fresh air exchange.  The test results were as
follows:
             Probe

          M (30 Day)

          Q (30 Day)

          R (60 Day)

          B (60 Day)
          Wt.  Loss mg/cm  - 30 days

                   4.7999

                   4.7741

              5.1571/2 = 2.5785

              8.3493/2 - 4.1746
These results indicate that the test coupons oxidized more rapidly
during the first 30 days exposure with average weight losses decreas-
ing in the second 30 days.  Based on these results, it appears that
the differences in weight losses observed during the test program are
within the ranges to be expected from oxidation alone.

SUMMARY

Under normal unit operating conditions without the use of overfire
air, excess air variation is the most effective means of controlling
NOX emission levels.  The use of overfire air provided a method of
further reducing fuel firing zone stoichiometries while maintaining
acceptable overall unit excess air and CO emission levels.  The use
of overfire air did not affect unit performance and results of the 30-
day corrosion coupon runs indicate that longer term operation could be
                              111-24

-------
achieved without significantly affecting unit operating procedures.
Longer term corrosion studies of a year or longer would be required
before any definite conclusions concerning corrosion effects could be
established.

The results of this study indicate that variables normally used to con-
trol boiler operation should not be considered as NOX controls with
coal firing.  These variables include unit load, nozzle tilts, damper
positions, and total excess air when overfire air is employed.

REFERENCE

1.  CRAWFORD, A. P., MANNY, E. H. and BARTOK, W., "Field Testing:
    Application of Combustion Modifications to Control NOX Emissions
    from Utility Boilers," EPA-650/2-74-066, June 1974.
                          111-25

-------
A:
3:40 p.m.
Overfire Air as an NCbc Control
  Technique for Tangential Coal-
  Fired Boilers
Ambrose P. Selker, Combustion Engineering
When was this last experimental work done?
The last  test work, you mean?   It was completed right
around  the end of  last year.   We do have
a follow-on study  on  Western  coals.  We  got  the  first
baseline corrosion coupon results back,  and  they  fell
in between the baseline and OFA numbers  I got on   this
unit.
Q:  Did you notice in the off-stochiometric combustion
    mode if there was any change, any observed change
    in slagging characteristics between normal firing and
    low NO  firing?
          X
A:  I am going to answer that in two ways.  First, we
    did not see an adverse change in furnace slagging.
    In other words, it didn't change unit operation where
    we had to change retractable soot blower or wallblower
    blowing cycles or something like that.  Secondly, when
    we did try to run the heavily slagged test with the
    overfire system in operation, we would let the unit
    go for 24-48 hours without wall blowing and come in
    and look at it, and it would be dirty everywhere but
    the firing zone.  So we did have trouble getting the
    firing zone to slag with the overfire air system in
    operation.

Q:  Would you care to elaborate on what you consider the
    slagging characteristics of the fuel or the ash of
    the coal that was tested?
                              IJT-26

-------
A:  It wasn't a bad slagging coal;  also, we have a
    unit with a relatively low heat release rate here.
    A pretty good quality coal.  Now, this coal run in
    other units in the same station where you are talking
    about higher heat release units and so forth is still
    not a bad slagging coal.
Q:  Has Riley had any similar or totally different ex-
    perience in their own work?

A:  In some cases we have had increased slagging condi-
    tions and in other cases decreased slagging.  In some
    cases with combined or overfire air conditions, we
    improved the carbon loss conditions, acually decreased
    NO  by a factor of 2 and decreased carbon loss by a
      A
    factor of 2, which was not anticipated.  So we have
    had it go both ways.
Q:  Do you feel that the corrosion rates that you obtained
    on, let's say, a low NO  firing mode, the differences
    over base firing, are really significant?
A:  I don't believe they are significant; I'll tell you why.
    What I showed you were average rates; they were the
    average of 5 probes, 4 coupons apiece.  That is, 20
    coupons.  In the study where we had the low average, we
    had some coupons at very high rates.  Where we had the
    high average, we had some with very low rates.
    There was a lot of overlapping that didn't actually show
    up in this average presentation I have here.  It is in
                               111-27

-------
    the final report though...the actual numbers for each
    coupon.
    Are you aware of the work we did on a prior EPA program
    where we showed no significant difference for corrosion
    rates?  [Reference Exxon field tests.]
Q:  This is on Barry No. 4?
A:  On more than just Barry.
Q:  Right.
A:  Tomorrow I will present some more data.
    I don't know why we saw the difference.  I think it is a  fluke,
    I really do, right now, but I can't prove that.  We are going
    to find out in the Western coal program, hopefully, whether
    or not we would get the difference.
    I don't think anywhere in your paper you give us the
    levels of the fuel nitrogen.  You made some reference
    that they increased at one point; but could you give
    us an idea of how much of the NO  that you were
                                    A
    measuring could be attributed to fuel nitrogen?

    To answer the first part of your question, the fuel
    nitrogen levels ran from approximately 0.9 and 1%
    up to 1.5 or 1.6%, something like this.  How much
    is attributable to it, we haven't really established
    that yet.  I don't know, maybe it is 50%, maybe it is
    100%.

    Some work done by a very distinguished visitor, Mr.
    David Pershing, with argon-oxygen experiments in a
    pilot scale unit» has indicated that for coal firing
    fuel nitrogen conversion can be quite high,
                           III-28

-------
    perhaps 70% or more.  Is that about right, Dave?
A:  Of the total, yes.   Now what the split is, I am not
    sure that we know that yet.
Q:  I'd like to make some comments regarding the Western
    Coal Program that Ambrose has mentioned.  We have a
    follow-on program with Combustion Engineering, that is
    looking at overfire air firing on two units of new
    design that were built with the overfire air as an
    extended wind box design.  One is the Huntington
    Canyon Unit of  Utah Power and Light, and the other
    is the Columbia Unit of Wisconsin Power and Light.
    One will fire a Western coal and one will fire a bi-
    tuminous.  So, we hope to relate the effect of designs
    built as extended windboxes versus retrofit;  we
    hope to be able to relate the effect of the unit size —
    the Barry unit was 125 megawatts, the two new units
    are about 400 megawatts, and we hope to be able to
    get an insight as to the effect of the fuel itself.
    And then, responding to  Ambrose's  conclusion or
    recommendation regarding further work in the corrosion
    area, we are planning, if we get the money from our
    friends in Washington, to start a more detailed corrosion
    program which would be sort of a monitoring type activity
    where we would select one or two units that were equipped
    with overfire air and try to keep sort of a continuous
    monitoring of the corrosion rates on a couple of boilers
    to establish some longer term data than we've been able
    to get thus far.  Technically, our corrosion runs have
    been of 30-day duration.  We'd like to extend that to
    6 months to a year if we can.
                         Ill-29

-------

-------
     CONTROL OF NO  FORMATION IN
                  X


       WALL COAL-FIRED BOILERS
                 By



Gerald A, Hollinden, James R. Crooks



     N, D. Moore and R. L. Zlelke



        Power Research Staff



     Tennessee Valley Authority




                 And



          Chris Gottschalk



      Division Power Production



     Tennessee Valley Authority
              111-31

-------
     This paper has been reviewed by the Environmental Protection




Agency and approved for publication.  The paper summarizes  data




taken from a joint EPA-TVA Study (Interagency agreement EPA-IAG-137D),




entitled "Control of NO  Formation in Wall Coal-Fired Utility




Boilers."  Approval by EPA does not signify that the contents neces-




sarily reflect the views and policies of the agency, nor does mention




of trade names or commercial products constitute endorsement or recom-




mendation for use.  Dr. David G. Lachapelle served as EPA project




officer.
                          111-32

-------
                             ABSTRACT








      It has been  shown  in previous work that biased firing techniques




 effectively reduce  NOX emissions; however, its long-term effects




 on corrosion,  slagging, and general thermal and operational performance




 were  not  evaluated.  The objective of this study was to provide




 quantitative information from which biased firing techniques would




 be evaluated as a method for control of NOX formation with respect




 to its long-term effects on boiler corrosion, slagging, thermal effi-




 ciency, and operational performance.






      In order to accomplish the objective, particulate and gaseous




 data  were taken for a number of boiler load and firing patterns.   In




 addition, two separate furnace corrosion measurement techniques were




 employed during long term runs,






      Analysis of the gaseous emission data showed that certain  combustion




 modifications can reduce NOX emissions from 30 to 50 percent  depending




 on the load and the burner configuration utilized.






     The collected and dried particulate was  analyzed for  unburnt




 carbon.  It was observed that  the difference  between baseline and  the




 "low NOX" condition was significant.   In certain  cases the carbon




losses were more than double for the "low NOX" case when compared  to




normal firing.   This is an important  factor to investigate further.
                              Ill-33

-------
     Of primary importance was the preliminary finding that  for the




"low NO " operating mode, both corrosion measurement techniques
       A



indicated a loss of tube metal significantly higher than that  for




normal boiler operation.  The most apparent tube metal loss  occurred




on the side wall.  More data will be obtained to further quantify




this finding.
                              111 - 34

-------
Introduct_ion_






     In the last few years, TVA has participated in several research




programs aimed at both identifying and reducing NOX emissions from




coal-fired boilers and, to a lesser extent, from oil-and gas-fired




combustion turbines. 1  The overall results provided information  on




control of NOX formation.  However, there remained lack of information




concerning boiler performance and efficiencies.






     The major combustion operating parameters investigated in the




past studies were excess air level and staged or biased firing patterns.




These parameters were investigated mainly at full or slightly reduced




load conditions.






     The effectiveness of combustion modifications will vary with  the




individual boiler.  For example, during the earlier studies it was




shown that at full load conditions NOX emissions could  be reduced




by approximately 30% over uncontrolled operation in certain boilers.




The NOX emissions were reduced by firing the operating  burners in




the lower burner rows with substoichiometric quantities of air and




supplying the remaining air required through the burners in the  upper row.






     The purpose of the present TVA-EPA study has been  to obtain more




detailed information primarily on possible corrosion and efficiency




losses that might occur while utilizing biased firing conditions as  a




control measure for NOX-
                              III- 35

-------
     The approach used in this field testing was first  to  characterize




baseline and "low NOX" firing condition.  Using these results the




operating conditions for NOX emission control were defined.  Second,




the "boiler was operated under the conditions for emission control while




particulate tests were run to evaluate ignition losses  and  overall




"boiler efficiency.  Lastly, two long-term (300 hrs) corrosion tests




were run to determine waste rates.  Corrosion rates were determined




"by two separate methods.  In one method, actual boiler  tube wall




thickness measurements were made before and after the testing program.




The second method involved the installation of air-cooled carbon steel




coupons exposed in the vicinity of the furnace wall tubes.  The coupons




were used during the long-term corrosion tests.  A weight difference




method was used on the coupons to determine the mils/yr loss rate,









Test Program






     The major objective of this study was to determine whether any




significant operating problems occur during long range  HOX  control




testing — such as  reduction in boiler performance efficiency, corrosion,




increased particulate loadings, or higher slagging rates.






     To accomplish the objective, a boiler was chosen with  enough




operational flexibility to permit a thorough evaluation of  many biased




firing combustion modes. Widows Creek Unit 5 was selected.  This Babcock




and Wilcox unit has the following principal features:  single pass,
                                 111-36

-------
divided furnace, water wall, dry bottom, radiant, reheat, 850,000 Ib.
steam per hr., 1825 psig (at superheater outlet), 1003°F (at superheater
outlet), 125 MW, pulverized bituminous coal, wall fired.  The unit went
into service in September, 1955.  Unit 5 has a sister (Unit 6) that was
used as a control for corrosion rate determinations for the long term tests.

     The test program was designed to characterize the pollutant emis-
sions and process variables of both baseline and biased-firing conditions.
The evaluation consisted of 47 separate tests.  Table 1 depicts the
conditions for each run.  Figure 1 shows the numbering of the 16 burners
inside Unit #5.  These individual tests vary the amount of coal and
secondary air flow through the burners.  All necessary information
concerning the boiler operation (steam flow, air flow, etc) was recorded
along with the different emission levels of NO, NOX, CO, C02, 02, and
S02 for all test series.

     During both the baseline and biased firing characterization test
program, the levels of gaseous emissions were continuously monitored and
recorded.  The baseline emission levels of N0_, were used for comparison
                                             A
with both low air and biased firing techniques.  From an evaluation of
the characterization runs (47 tests), both baseline conditions and selected
operating conditions that significantly reduced NOX levels were run again at
each load condition (50,75,100, 125 MW).  The levels of NOX, unburnt
particulate carbon, total particulate and boiler efficiencies were compared
for each of these operating conditions.
                                     Ill-37

-------
 Data Collection and Measurement






     A mobile analytical van capable of monitoring the combustion




 products was fabricated by an outside firm for use in the program.




 The system is by necessity sophisticated, utilizing the accepted




methods in analytically measuring the components of the gas sample.




The gas sample was handled carefully to avoid contamination or loss.




The van is equipped with a nondispersive infrared analyzer to measure




CO and CC>2, a nondispersive ultraviolet analyzer to measure SC^, a




polargraphic 02 analyzer, a chemiluminescent analyzer for measuring




NO and NOX, and a. flame ionization detector for hydrocarbon analysis.




Table 2 depicts the equipment used and the ranges of detection for




 each.  All necessary calibration gases of appropriate concentrations




were stored in cylinders employing dry nitrogen as the carrier gas.




The instruments were calibrated daily before testing and periodically




throughout each shift to assure accurate analysis.






     In running field tests, the samples were drawn before the air




preheater section from six zones in each duct providing 12 sample



regions per boiler.  This sampling procedure reasonably assures a




representative sample.  The sample is removed from the boiler under
                                  111-38

-------
vacuum through stainless steel probes fitted with metal particulate



filters.  The sample passes through teflon lines and its temperature



is maintained above 250°F until it reaches the gas cooler.  The sample



is refrigerated to a 35°F dew point to remove water vapor and then



is sent to the analytical van for analysis.  Figure 2 illustrates the



flue gas sampling and conditioning system.






     Four Research Appliance Company particulate sampling trains



(EPA-method 5) were used.  This setup included four sampling "boxes



equipped with U-inch filters and probes, and two sets of isokinetic



pumping systems.  The measurements were taken after the air preheater



section of the boiler.  Each duct was divided into 12 sample regions



to assure a fairly representive sample.  The data was used to calculate



the amount of particulate emitted (gr/scf) and the uriburnt combustibles



retained in the flyash.  The latter was obtained from an ignition



loss analysis.





     Two different methods were used to obtain corrosion rate data.



The first method relied on actual tube wall measurements prior to



beginning the study and after its completion of both Units 5 and 6.



The tubes were cleaned to bare metal and then measured with a



Krautkamer USK5 Ultrasonic Minature Flaw Detector.
                                 Ill-39

-------
     The instrument was calibrated frequently with a  standard cali-




bration "block.  More than fifty measurements at  different locations




were made on each boiler wall, totally at  least  UOO measurements per




boiler.  Similar measurements were taken after completion of the




testing portion of the program.






     The second approach (for corrosion detection) utilized corrosion




probes exposed in both the control boiler  (normal firing—Unit 6) and




the test boiler (biased firing—Unit 5).  The temperature of the SA192




carbon steel corrosion coupon was set at approximately 650°F for each




boiler.  Two separate corrosion coupon tests were run.  The coupons




were exposed to the test atmosphere with range of times from 315 to




755 hrs.  After exposure, the slag was removed and saved, and the




coupons were then cleaned with a mild acid pickling solution.  The




coupons were re-weighed to determine any weight  loss.  Figures 3 and




h illustrate the details of the corrosion  probes.  The probe design



is similar to Combustion Engineering's which basically consists of




two concentric pipes.  Cooling air from the plant air supply enters




the tube-shaped coupons through a 1/2-inch stainless  steel tube




roughly centered inside the coupons.  The  coupons (outside surface)




are exposed to the furnace atmosphere.  The amount of cooling air is




controlled automatically to maintain the set-point temperature.  Two




thermocouples are used for each probe—one for controlling the temper-




ature and the other for recording the temperature.  The cooling air
                               111-40

-------
travels back through a 2-inch, extention pipe for discharge outside of




the boiler.  Therefore, the cooling air does not interfere with the




test.  Figure 5 gives the location of the probes and figure 6 gives




the operating procedures utilized throughout this portion of the test.
Data Evaluation







     N°x:  The analysis of the short term NOx test  program shows that




the(lbs/MM Btu's) of NOX can be reduced by staged firing and lowering




the excess air from the normal levels at each level of load generation.




It is important to note that the excess air levels  increase as the




load of the unit is reduced and that lowering load  without  either




staged firing or a reduction in excess air from the normal  level for




that load will not in general reduce the NOX (ibs/MM Btu's).
     Table 3 depicts the emission levels  of Q^*  C°2>  CO,  SOo, and




for each short term test.  Figure 7 is  a  graph of NCL (ibs/MM Btu's) vs
                                                   Jt



02$ for normal firing modes at 125, 100,  75,  and 50 MW.   At 50 MW,




there were two distinct coals (in terms of heat  content)  used.  The




higher Btu coal content coal gave the lower levels of NO  and this




higher Btu coal was more representative of the majority of coal burned




during these tests.  Utilizing the data at 50 MW obtained with the




higher Btu content coal and the data at 125,  100, and 75  MW, the following




relationships have been developed for normal  firing conditions:
                             111-41

-------
     loge  63 = 2.8358 -  .0126 MW                                 (1)





     loge  NO = -1.9068 + .02923 MW - 0.000127 MW2                (2)





     d(HO) = d(HO) ^ d(MW)  _ (NO) (.8) (-1.15 + 5.08 log^ NO)1/2 /-x
       y*               /i.                      A                  \ J *

     d(02)   d(MW)   d(02)                     02





where 52 = normal % oxygen content





      NO = normal oxides of nitrogen (Ibs/MM Btu's)





      MW = megawatt load





     Therefore, given the megawatt load,  the expected 02 level  for



normal firing can be determined by equation (l), the NO emission can



be determined by equation (2), and the rate of change of NO with respect



to Op can be determined by  equation (3) and thus the required level



of Op reduction necessary to meet a specified emission level.




     As stated earlier, KOX emissions can be reduced by biased  firing



at each level of load generation.  Figures 8-11 are  graphs of NOX (ibs/MM  Btu's)



vs 02$ for different biased firing patterns and load conditions.  The



approximate range of testing for each biased firing  condition is repre-



sented by the length of the line on each  figure.  Extrapolation is not



recommended.
                               II I-42

-------
     Partieulate:  The particulate emission data are summarized in




Table k.  Total quantities of particulate increased for the biased




firing tests but were not considered significant.  However, carbon




losses increased significantly, particularly the relative differences




at high loads.  Table 5 illustrates the effect of ignition losses  for




both types of firing as a function of increased fuel operating costs.






     Boiler Efficiency;  Boiler efficiencies were calculated using the




ASME STEAM Generating Units, Power TEST Codes, using the Abbreviated




Efficiency Test Heat Loss Method.  The assumptions made were that  no




combustibles remained in the bottom ash and that the unmeasured losses




were 0.5 percent.  Table 6 summarizes the calculated boiler efficiencies




and other pertinent information for normal and biased-firing conditions.




Figure 12 is a graph of mean boiler efficiency versus load (MW).   A




consistent reduction of efficiencies throughout the load range of  the




boiler is apparent under biased firing conditions.






     Corrosjign:  A summary of the boiler tube corrosion measurements




taken on Units 5 {biased firing)  and 6 (normal firing)  is contained




in Table 7.  The following conclusions have been drawn based upon




a statistical analysis of the data:




     (l)  Unit 5,  B furnace, side wall, lost an average of




          7 mils thickness from the time the wall  was first .




          measured ("before" in Table 7) until the wall was




          again measured ("after" in Table 7).  The loss was
                                 111-43

-------
          statistically significant at « = .05 or  only  5 percent




          of time will this difference be due to random variation,




          and not a true loss of wall thickness.




     (2)  Unit 5, B furnace, division wall,  lost an  average  of




          3 mils thickness.  The loss was statistically significant




          at °= = .08.




     (3)  No other walls tested "both before  and after showed a




          statistically significant (a = .10) loss.




     (k)  All walls in Unit 6, B furnace, showed a significant




          gain in wall thickness.




     (5)  Comparison of Unit 6, B furnace, after measurements,




          Unit 5, B furnace, before measurements,  and Unit 5,




          A furnace, before measurements, shows no difference,




          both on an overall average, and a  wall to  wall average.




     (6)  The consistent differences on Unit 6, B  furnace, on a




          wall to wall comparison, in conjunction  with  the




          conclusion above, point to a calibration error in  the




          measurements on Unit 6, before measurements,  of




          approximat ely 5 mils.






     Based on the conclusion in #6  above, the tube wall measurements




for the B furnace on Unit 6 reflect this correction  factor in Table T-






     It is apparent that more data is needed to further quantify the




magnitude of corrosion.  The unit is presently on  line  due to the load
                              III-44

-------
requirements on the power system, but as soon as it can be taken off




line a thorough tube wall measurement will be performed.






     The corrosion probes were used as an indicator of possible  corrosive




atmospheres in the boiler and were not intended to supply  any  signifi-




cant quantitative date.  However, the results shown in Tables  8  and  9




are quite interesting.  For meaningful comparisons, probe  1 can be




compared with probe 3 and probe 2 with probes k and 5-  The comparison




cases are for those probes in similar locations in the boiler.   In




the first series of tests (Table 8), it is apparent that between 2 and




5 times the corrosion rate (mils/yr) occurred on those  elements in the




"low NOX" mode boiler.  In the second series  of tests  (Table 9)» the




data shows significant corrosion increases when comparing  probe U (biased




firing) to probe 2 (normal firing); however,  little difference is




shown for the other location.








Conclusions
     Even though much data has  recently been  published  in the areas




of NO  control, there has been  little  if  any  definitive published
     A.



data on coal-fired units where  such  control has been correlated to




particulate emissions, boiler efficiency  and  corrosion.  However, based




on the results of this test program  the following tentative conclusions




can be made:
                              111-45

-------
Certain combustion modifications can reduce NOX emissions




ffom 30 to 50 percent depending on the load and the "burner




configuration utilized for pulverized coal-fired "boilers.






Carbon losses in the particulate will increase for biased




firing conditions.  This carbon increase will reduce the




conductivity of the ash somewhat and could affect ESP efficiency.




Total particulate increases, however, were not considered




significant.






A consistent reduction of boiler efficiencies throughout




the load range of the boiler is apparent under biased firing




conditions.






Corrosive atmospheres will prevail in the boiler under biased




firing conditions.  Such atmospheres .appear from preliminary




information to cause significant water tube losses in certain




areas of the boiler.
                        111-46

-------
                            REFERENCES






1.  Hollinden, G. A.  et.al., "HO  Control  at TVA Coal-Fired Steam
                                A


    Plants," presented at  ASME Air Pollution Control Division



    National Symposium,  Philadelphia, Pennsylvania, April 2k - 25, 1973-
                              III-47

-------
                           TABLE 1
Test Program for Baseline and Biased Firing Emissions Studies




                   At Widows Creek Unit 5
                        Test Program
                                            Burners

Run;
1
2
3
U
5
6
7
8
9
10
11
12
13
lit
15
16
17
18
19
20
21
22
23
2U
25
26
27
28
29
30
31
32
33
Boiler Load
(MW)
125
125
125
125
125
125
125
125
125
125
100
100
100
100
100
100
100
100
100
100
100
100
100
100
75*
75
75
75
75
75
75
75
75

Excess Air
Normal
Low
Normal
Normal
Normal
Normal
Low
Low
Low
Low
Normal
Low
Normal
Normal
Normal
Normal
Normal
Normal
Low
Low
Low
Low
Low
Low
Normal
Low
Normal
Normal
Normal
Low
Low
Low
Normal
Firing Coal
16
16
lit
lit
lit
lit
lit
lit
11+
lit
16
16
12
12
12
12
12
12
12
12
12
12
12
12
12
12
10
10
10
10
10
10
8
On Air Only
0
0
AI jAlj.
E-i»DU
B2,Bo
C2,C3
Ai,Alj
DI ,DU
B2,B3
c2,c3
0
0
AI ,Al| ,D]_
B2,B3,C2
Aj,A2 ,A3
G! ,c2 ,c3
D-r ,D2 ,D3
A2 ,A3 ,DI
A^jAj. .Dj
B2,B3,C2
An ,A2 ,A3
C-, ,C2 ,C3
D-, ,D2 ,Do
A2,A3,D1
0
0
Bl )Bit
c2,c3
D-, ,Dj|
Bi>Bii
c2,c3
Di»Dit
B-. ,Bp »B-3












'DU
»C3
,Alj
»Cij
,Di|
>DU
,DI^
>C3
,AI^
,c^
,D^
,DU








>Bk
                           111-48

-------
                  Table 1 (cont)
                                             Burners

Run
3k
35
36
37
38
39
ko
1*1
1*2
1*3
hk
1*5
1*6
1*7**
Boiler Load
(MW)
75
75
75
75
75
50A
50A
50B
5QB
50A
50A
50A
50A
50A

Excess Air
Normal
Normal
Low
Low
Low
Normal
Low
Normal
Low
Normal
Normal
Low
Low
Low

Firing Coal On Air Only
8 Ci.ca.cg.c^
8 B^jB^jDj ,D^
8 Bj_ , Bg , B^ , Bjj
8 C-, , Co jCo,^
O Bn , Bj, , D] , Dh
8 0
8 0
8 0
8 0
6 BO ,B^
6 cltck

6 c Ci
6 B-L^Bj^
In all runs at 75 MW, the  A mill  is off.




In these runs, the A and D mills  are  off.




In these runs, the A and C mills  are  off.




 Used in particulate series only
                            111-49

-------
                              TABLE 2
     Instruments
1.  NO, NO  Analyzer
    Thermo Electron
    Model 10A
2.  CO Analyzer
    Becfcman #86H

3.  C02 Analyzer
    Beckman #86U

k.  SOg Analyzer
    DuPont
5.  02 Analyzer
    Beckman (7^2 Model)
       Technique

Chemiluminscent,
negative pressure
nondispersive
infrared absorption

nondispersive
infrared absorption

nondispersive
ultraviolet absorption

Polargraphic sensor
6.  Hydrocarbon Analyzer  flame ionization
    Beckman #UOO
  Measuring Range

     0-10 ppm
     0-25 ppm

    0-100 ppm
    0-250 ppm
   0-1000 ppm
   0-2500 ppm

   0-1000 ppm
   0-5000 ppm

       0-555
      0-20$

   0-1000 ppm
   0-5000 ppm

       0-5?
      0-10J8
      0-2551

   0-1000 ppm
5-100,000 ppm
                                 111-50

-------
                 •o   oi-ooo
                 o^
                 H   ift  I—
                                 fO   f»"»   CO
                                                                                   8?
                                                                               cvi   W   Ji
                                      cviStoo   t-opw  (rico   wo   cj
                                      irvmmjf   Jo3   >i)
51
4J


^
          H
g™        ?
             4»

             s
             «
             •p
                                                                                    -.
                                                                                    1*1

                                                                                   os   o\
                                       111-51

-------
             tfl
         \D    00   ITS    O\   ON   CVI    CM   CO   jgf   CO   JJ3    b---    ON,   rH    U"\   OS   H    Ijr

         OOOOOQOQOOOOOQOaQO
          o ®
          ra
                     IA   _=f
                     d    81
                           UN   CM
                           IT*   VP
                           OX   01
                                                            t**-
                                                                     Q    H    Q    Q    fO
                                                                     C^   V£    ON    C\l    rH
                                                                     CD    Q    lf\    tA    ON
                                                               §
                                                            ir\   b\
                                                 H    CO   CO   ^
                                                                 ft
! "A

i 6-1
               +>

               1
         IfN    ON   Od
         \^    !_|   OD

               O   -5
               4          H



               4>
         i®
                                      8
                                    m   J-   -   OH   CO   VJ3   ^   CO
                      ^t    iTi^fiAro^r    mcy
                     CO    odfnjtcocdojrnvO^vO
                           on   co   co    UN   UN
                                                           vocoj*cMt^pnvoocM    oon
                                                           o    ON   co   on   ON   ON   on   UN   CM    CM   ~t
                     t~-HpnvOtDCOOOON
-------








1
i-3 IA
M
SEH
H

5 o
i|
§B
SS
0
!| ^
*i
*= *
H B
K C
glk.
M Q
§
g«
wo!
cvw
KEH
o|

i^ W
£

u











g*
*,


1°
<\J
O @)

**** o
Q,
Pi-Mi
• — rn
0®

_
S CV
1°
"""en
g*®





1




"S1
i*S
cvia
»i
S3

g
T3
«
43

•"
1
ffl

_p
!

I
Pi

1
<

1
7
ffl

HP
1
J

a
1
»

^
g
T
|
g



*.

of
R
IA CO VD .^ O
-rt vO -3- IA b-
0 0 0 0 O

OO OJ OS Pn r^
C- ^JD J- J1 H
-^ O r- ( CTi O\
rn i-i H

-* ct f: 3 ^
-3- H H t7\ oo
on H r-l
CT\ t1- ON CM rn
O cVJ OO VJD j^j*
IA



t^- CO J" J~ O
\X) OJ CO IA J-
t\j cn

8 f » 8^ rt
m 3^ ro rn tA


IA PO kA CO OS
rn IA m m ,-f


m \o CM on H
^t J>- LA IA CD


vj3 cO vO O^ O
m vO ro on \x)

-if ,=f H OO V£>
^t  CO O CO SjO
_j- ro en IA -nf
o\ t*- t*~ t~ ^-
1 » iSi « S
1A !A
LA LA IA ^- t-
f- t-. t«- 1 1
I I t O O
^O fp L^ H H

jH rH ri rH rH
o t*--moo-4'co ro^fcovo IAO
\jQ CO t*- -3" \O C~- tA j IA ^" ^J" -5T
oooooooooooo

St*- OJ VD W^ OJ CO CO \O VO -^f" (*^
r-t H fcr^C lAj'COOO-S ON \p
^H O O ON -j O\ ON t~- -^ r^ '^O W
rO ^f r-4 Cy OJ rH rH m PO rH F^

f1^ t^ rn ON ON CM p o ON t- ^H w
COPO OJCMHHCOPO«HH
rH CO CO t~- OO ^O r4 ON -^ ON t*- ON
co m^ w-,* iAt-iAm-^^oco



-^ N- CO y? ON \O IA J" IA OO CO PO
CM rO PO ,^ PO ^r CO \O Pn CO O ON

S!$ K^l fc^SI^^?


H^t C*-vD IA t-- H CC t*- <*^ -^ ON
UNPOH LAfO.31 H LA-it *f\ O ON
-^tvO iAcn^ ir\ j* POJ- cnmoj


CMCO ONlAlAjJCOH H-4-ONCn
^t ITX t^ -* CD O QNCOCO\OlfNCO
H

VOVO PO t-- i»t CO rOH f-CD ON^tf-
PO tA \o on t-- co o co t— IA ^f co

ONvj3 tA_4" C*--d'^t' IA ON LA 1A *r>
-* OOCVIIAONOOOCU POO^r^l
rHHH HH*HHHr-IH


o r--ONt\j HCO lAj^- cnofoon
^^^^jjd00 Pn^-^t 
-------
        fi*i
       ^
              *  s
              d  d
i 1
       -as
       o
       o
              s
   a
p p

 wo
cT
          I
ft  I
              
-------
M   CO
 I   X

I   1
O
a m*

•vo
r3 0
i-nl H
oS
•H •— -
-P
•H (0
as
M 1-5
-P
* 0
M* 3
<-( .. 	 ^ f\
G •• N f— i
•ri PL, |
T) u m
0} CO
J en
a
fl -H -p
•H n) o
S3 Q
£5J?
-p WJ
r-J gj
d r,
0 O
h Tl
0) 0>
Pn -P
O
rH 0)
^?
55


rt
^
K




-P
iri
£^

VO
CO
0

OO
vo


ON
O
OJ




•fc
-f
H
O



H
OJ
8
CO

£
H

<
H
1
IA

LA
^—
1
O
H
1
OJ
ON
t-
O

0
ON


t—
vo
(O





VO
00
OJ



LA
O
IA
CO
co
OJ
OJ
oo
H

PQ
f-^
|
VO

LA
^—
1
H
H
1
OJ
ON
LA
O

0
ON


rH
VO
00





_3"
0
CO



|S
t—
ON

t-

^
l/"N
C\J
|
t—

LT\
r-
i
OJ
^
i
OJ
OJ
LA
0

-3"
oo
H

VO
vo
oo





OJ
H
00



OJ
o
OJ
O
LA

<
ON
CO
1
OO

LA
t—
1
OJ
H
1
OJ
H
IA
0

ON
ON

*
_^
t—
H





00

OO



VO
_-,.
H
t-

ffl

OO
1
ON

LA
f-
1
OO
rH
1
OJ
oo
Jt
o

OJ
ON


H

00





OO
o
00



oo
UN
rH
t—
H
t-

o
Cf\

1
o
l-l
LA
t-
1
00
H
1
OJ
00
CO
0

OJ
vo


o
OJ
LA





ON
c-
LA



LA
rH
LA
CO
OJ
H

O
H
1
P-f
H
LA
t—
1

| 	 1
|
OJ
CO
t—
O

ON
tr\


i — i
vo
LA





^3-
t—
-*



t—
IA
O
CO
oo
LA
t~

pq
IA
OJ
|
OJ
H
LA
t—
1
-~f
rH
1
OJ
IA
t—
O

-=r
t-


-{•
t-
_-,.





00
ON
LA



00
CO
ON
00
t—

0
LA
OJ
1
oo
rH
LA
t—
1
CO
H
1
OJ
OO
CO
0

H
j.


ON
oo
t—





oo
H
t—



00
CO
oo
ro
oo
LA
LA
o
H

H
H
1

H
LA
t—
1
ON
H

OJ
VO H
ON -^t
0 0

ON vo
CO O
rH

-3" VO
H t-
C— t-





rH 0

t- -a-



CO LA
H LA
LA CO
00 ON
VO O
LA OO
OJ -*
O -*
H
PQ <
H I-

| I
LA VO
H H
LA LA
t— t—
1 1
rH H
OJ OJ
1 1
OJ OJ
                         111-55

-------
jgg pQ ^
ir\
sin °
IS

•4^ *
Tt 01 CO
as
M K-5

-p
* o -3-
W* P O\
•H EH 1 t—
0} 03
1-3 CO
C
R -H .p
•H aJ o H
£ ^ (S CO
O — - 1 t—
^


 1
-P IA

CM
CM
IA
O

t—
•
CO




\o
OJ
MD




vo
CVi
V^)






|v —
ir\
-^
^o
*
ITS
j-







_jf-
CVI
H





?
CO
rH


LA
T
LA
^

H H H 00 |
rH P
H
g
W
OOOOr— LACOrHLA oj
LA LA ON -^ C— LA VO t«0
CO t— t^ VD -3" OO J" "d
§
M
-p 0)
* to w
t— VO CM O co t— ON »
C! H
•H -d
rt
3 C
T^ O
r< C '
0) -H 60
i — 1 cd W
co t""- ON oo rH ON oo p4 JH
O CM VD -* t— VD LA B-p ON ON t— t— IA LA <;aj+>
PH -p 03
M o
0) En
C HO
•H HO
PQ < pq <• pq << pq Ot—
l^HrHOOVOVD -do
J'CVICVIOOOOOOOO t) O
1 1 1 1 1 1 1 ft (U 43
OvOHCMf^-^lA O-P
HCMCMCVICMCVICVI HOSTS
0> HO
> ^ 4^
11 o o
LALAIALALA "d'HO)
f— t— t— t— t—LALA 4Jf-l
1 1 1 1 1 t- r~- ^ h rH
vot^b— cocol 1 to an o
CVICMCMCVJCM-^--^ (UPUO
1 I i i i 1 I (-5 *
CVICMCMCMCM^fO #*
HI-56

-------
          !H
          O
         K
             M
            •H

            <<


            rH
             fn
            •H

            <
             IH
             O
             K
       4>
       D)
      &&
 h
•H
<
 h
•H
<
 P.
•H
•a;
                  TJ T3
                   s!25^^>
          Q  Q           o Q
     «*>H3l-l  ««t-ljH3
    D  4)        4)  t/>C/)O v
                                       tfl (A


                                       pq pq

CO
O
CO

8
M
      oooooooooooooo
      oooooooooooooo
      O OcOCOONi-lrH O^i O H LfN O\ t— t—
                                                      o o o o o o
         irsJ-OJrHOHO\H

               H
                                                   r-\
                                                            tr\<*">
                                                                  w:
                                                OO
                                                      ON OJ \O
                                                                  O OJ
         VO
                                                               O\VO
                                                                  H
      O O-*
                           OJOOCO PO|>-o\o\VO-3'-3-
         _.     l-UNOJCTkCOVO
         MOJOJCVJOOONONt-
         rH H r-t rH H rH
                                             oo ono i— t—-3- co
                                          t-t-t—IAJ-J-J- m
13
            i~it— co
                        I/NO
                                             CVICM
                                                      pn op J* ^±  rn tn

                                                      o\ovo o\-=«- ir>
                                                         r-lr-l(HCVJCVI
                                                                                       «
                                                                                       w.
                                                                                       >
                                                                                       ^
                                                                 o
                                                                 ON


                                                           £•  3s

                                                           ^g^i^
                                                               w  o
                                                           +3      VO
                                                            rH  -p  «J O
                                                            O  4) -P H
                                                            fVS  «J ->*
                                                            CJ --^«  M
                                                           «  05    43
                                                               IH +»  a
                                                           H  -H  Oj  4)
                                                            05  S  4)  O
                                                            O    W
                                                            CO Oi    CO
                                                           •H CO   • t—
                                                           PL, O\  0)  .
                                                               •  !> O
                                                            CCO < t~
                                                            o

                                                           T^  •   *  •
                                                            4) H OJ CO
                                                            
-------

VO
    Ok   U
    ri   S
    S   5
e
w w
J M
LJ 1-J
O M
CO
CO
Is
CO
w
s
-_
IGNITIO*
%

< **
>-
a c
Xw
o m 6-
^^ p
K^* "
vO
.0 O  rj r*, &} ~3


0) 0) 01
C! C X X C
•H -H O 0 -H
i-H rH S5 S! iH
CO 0) 01
M CO S > CO


m vo oo co co o
m  r- vo



ON i— ! in O ON "^
co in *n ON in r^
iH rH
r-l ON vO r^ O OO
CM oo oo co r^ cr*

vo -* m ON oo m
^\ in in in r*^ r*»
o o o o o o

r-. co r» m m o>
vo m m r-. oo oo
CM ON oo so m «*
^^ ^j^ c^ ^* r^ ^^

pg <• pq •< pg U
rH rH r-l in I/I in
in ci 1-1 i-*. CM co


01 01 <1> 01
C X X C C C
•H O O iH -H iH
01 0) 01 tt)
SU u co w tn
o o at cd co

in
m
00

o
oo



sr
00

CJN

0
m
o

vo
vO
CO
r-.

£
CM



t
3
a

VO
rH
H3

CO
CO
*v'



oo
00
00
vO

0
m
o

0
r*
CO
r-.
•
8
s
CSI



i
J5
3
                             III- 58

-------
3
H U
   Pn

   W
           en
       CO  00
               st
               oo
                   00
                 m
                 00
                            CM
                            CM
                            00
    st
    00
o
       st  00  H
       m  m  en
                        oo
                            00   sD
       I—  OO  OO  00   OO   vO  OO
tn
a
O
CJ
H
O  
o o o o o o o



t^j ,_( cn "*£> rH tn m
00 C^ C7^ O^ C^ f*^ 00




co CM c^ r*^ r"*- ^f oo
in m m *^~ *^ *** ro






^C PQ *^ pQ i^^ ^j pQ
^D ^o ^^ ^^ ^^ r*^ r™^*
CO OO CO CO CO *^ >3"
1 1 1 1 1 1 1
^" in oo o\ o \D CT\
CN] CM iH rH r-|




Q> 
cd
o>
>
ed
0)
,c
4J
M
•H
0)
3
•H

CN|
O

QJ
•E
E-t
*
                        III- 59

-------
s
pgs
pa o. H
ijKX

j s o
o o a
PH H M


t-
%
S3
EH






d,
w
CQ
M
PH
K
0
CQ
g

e
1
M
O
F 1 .n
B s
B EH
a
0
o



1
i
Ix,
«
g
O
W
H
H
CO
pq



ij
^
>
fn
0)
a
h
3
PQ
H

•tH
Q



K
W
§
W
u
<
fe
S
s
H -H
a a> d -H
>H TJ O >
pi -H S -H
W CQ fc P
•P
at
a;
^
o
1
M 4J
1) C
H U
a) 'H
•H
*o 12
a) Si
•P -H
0 CQ
h 0
II
II
CQ
H a
       III- 60

-------
                     P g CO
EH

Q ^ -H JH -H
M pq CO £ P
IS




PM
H
K
•3



,


w
^ S
a o
S fc


\£> O VD t—
oo on on on


<^J



h-1
, "1
1
§
>H -rl
CU -p CO
fl  C -H
fi -d 0 l>
d -H (H .H
ft CO fc P


















-^^>-
SH
0>
CO CO
CO CO S-,
CD 
jjj
cd cd o
•H
-P -P Vl
C S -H
CO CCl JM
O O t£
•H -H -H-
IH SH CO
•H -H

•H -H is;
CQ CO
II
II II
CO
CM on a
                                     III- 61

-------
8
M
CQ
8
B
CO
          CM
           *

          CO
                   OO


                   VO
-4-
  •



CXI
          IA
                              IA
     8
EH
rQ
a
CO O
H
W CO
m §
* 8

p3
13
"
o
s






-------
O
H

O
8
      lr\    J-
       •     *

     jzt    LA
      LT\    CO
                      o\
           CO   CO
           ir\   u\
           fO   no
                      vo
                      LP>
                      m



o

a
9
EH














&H
CQ
a
o
H

K
8
i
W
EH
1
O
a
o
i-i







3 CQ <~-
1-3 VO CQ CQ
Pk OS
S PH

-------
                        FIGURE 1


TVA's 125MW Widows Creek Unit 6 - Burner Configuration
PULVERIZER
                                   ROW
                                2           3
    B
o     o      o      o
o     o      o      o
o     o      o      o
o     o      o      o
                         111-64

-------
111-65

-------
TII-66

-------
111-67

-------
                     FIGURE 5
                  PROBE LOCATIONS
     Unit 6
I   Burner Wall
                                           Unit 5
                                     1   Burner Wall     I /} 3
                                                        V
'C" Burner Elevation
                                       "C" Burner Elevation
V | Burner Wall L
n
n
                                           T           8
                                     "B" Burner Elevation
                   III-6S

-------
                                                                      cu
                                                                      W)

 to
 fl
 O
 o
 o
-p
 
O * !>> O CU O
POH .p 0 ^!
^i <( C X to
0) O H CU
g" CO CO
OJ S-i S d h
CVJ H g Jj 0) 3 C
, — 1 «r rt O rr; o ,~i
a -H
O cfl
.H h a
40 -rt CO n) Qj
• ooo
O'OOO
oooo
• ooo

oooo
oooo
oooo
oooo
l6-Burners in service ill-Burner configura
Reduction of excess air A-^ , Aj^ burners on a
until burners show a only. Reduce exces
rapid change in CO. NOX until burners show
lowest as possible. rapid change in CO.
NOX less than 500 p
                                           111-69

-------
           !H
           •H
           05
               •H
               c8
                0
             -a
             a
        t) >» aJ
         a H
         i S3 H
           o a
        rH
         03 ^l
         O -H
         o ai
           c a 
^a to c8


 fn '  «VO
  fn
                                   cd
                                 "* 6 ""'«•*-• S
                                 ^S  HO I
                        -° 2
                           £H
                           a)
       efl
       3
                 OO
                 oo
          ®ooo
                 oo
                               c
                               o
                               -H

                               Id
              C
              O
              CJ
                                   1
                    a)
                      O
                      §0
                      J-

                    w «
                               0;  rod  3 to
                               C < ni f> to
                               h  -      
-------
                                                                                
-------
                                                                        o\
M   O
Cu   B
       IfN

       (M
                                                                        LTk
                                                                                 (M
                                                                        OJ
                        oo
                         •

                        o
U"\


O
                                      III-72

-------
                                                                  CO
HJ
•H
   
-------
      CO
      si
     o

      0)
O
H
B
E
                                                                           o
                                                                           Os
                                                                           \D
                                                                           if\
                                                  I
  ON
   *
  o
CO

o
                         o          o          o



                                        XOH
                                  111-74

-------
                        !-i h H
                        O -H -H
                                                                          H
                                                                         cr%
                                                                       -oo
H    fn
H "es. CD
     CO
QJ CO
-H O O
P^l S UA
                                                                         cr\
     CO

     d
                             o
ro

O
                                            OH
                                 III- 75

-------
    Figure 12
    Boiler Efficiency versus MW
87
86
              30
75
100
125
                       A  -  Normal Firing
                       B  -  Biased Firing
                       HI-76

-------
    4:20 p.m.
    Control of NO  Formation in
      Wall Coal-fired Boilers
    Dr. Gerald A. Hollinden, Tennessee   Valley Authority
Q:  I tried to glance through your paper quickly  I couldn't
    see it, it might be in there...where did you measure the
    particulate loading?  Was it before  the ESP, after the
    ESP, or where was it?
A:  It was after the air pre-heater section.
Q:  What was your particular collection device for that
    system?
A:  It was a Research Appliance system  mechanical collector,
Q:  You made the statement that the carbon losses were in-
    creasing but there was no overall, there was no increase
    leaving the furnace.  It had to be accounted for
    someplace.
A:  No, I said no significant increase in particulates.
    The difference that we did notice was in the chemical
    content of the fly-ash as opposed to the quantity of
    the fly-ash.
    You are aware that Exxon tested boiler number six
    (Widow's Creek,  No.  6)  a couple of years ago and we
    obtained similar results on increased carbon on
    particulates.  We evaluated that out on the effect
    on performance and there was no significant effect
    on performance or efficiency.   You are also aware
    that we tested your same boiler.
    I am not aware that you did any ignition loss tests
    on the particulate;  I may be wrong.
    We did check the carbon and calculate the  efficiency by the
    same method  you  did.  And we found  even  under  those
                          111-77

-------
A:
conditions, where we had an increase in carbon, that there
was no appreciable difference in efficiency.  Remember, we
reduced the excess air and we had a net gain in efficiency
from that standpoint.  So the two balance each other out.
That was in the prior work.
You know we shouldn't mislead the audience.  We
would have to compare the excess air levels.  I have
already mentioned that we dropped pretty low (excess
air), especially at 125 megawatts, and I don't think
we would do that again.  1 don't really remember
2 years ago what we ran for excess air levels.  I
don't think it was as low as what we did in this
program.
Tomorrow I will present some other data on your
same boiler and in that case we  had a decrease in
carbon loss on low NO  firing on the same boiler.
                     A
Now, I gather from your data you are saying, in
every case, low NO  firing increased the carbon
                  X
loss.   Is that true?
That is true.  I am saying that with  the  four  low NO
                                                    X
conditions that we selected for  50, 75, 100 and 125
megawatts,  compared  to  normal  firing  at  those  four
loads,  we did have a reduction in boiler efficiency.
We did not obtain particulate data for all 47 tests.
We are also going to present some data on corrosion
based on corrosion probes taken  on your boiler #5
and boiler #6.  Our data show that there is no
significant difference in corrosion,  by corrosion
probes.
It is very important, as I stated earlier, where
you put the corrosion probes, since the probes are
                             III-73
                                                                                J

-------
                         just an indicator of corrosion.  In one  case under
                         normal firing conditions we got an equivalent  8 mils
                         loss per year on a particular probe and  under  the
                         same conditions several weeks later got  50 mils loss
                         per year.  Obviously there was not a seven-fold in-
                         crease in corrosion for the normal firing case over
                         that period of time.  So I don't hold much credence
                         from a quantitative standpoint in our corrosion probe
                         information except that it was strange that we always
                         got more corrosion or more loss of the coupons in the
                         stage-fired unit as compared to the normal firing.
                         And that is really my whole point.  The  important
                         issue is that more information is needed on the
                         actual tube-wall measurements.  That will give you
                         your final answer.
                     Q:  What I would like to know is whether or not the 0«
                         levels or excess air levels at the corrosion probes
                         were determined when you saw the increased losses.
                     A:  No, they were not.  We didn't have any sample withdrawn
                         in the furnace area itself.
                     Q:  Can you give us some idea what your overall excess
                         air 'was, and what your burner stoichiometry was at
                         the conditions where you saw the dramatic reduction?
                     A:  At 50 megawatts and 75 megawatts the 0,, level was
                         somewhere between 6 and 9%.  We weren't even near
                         reducing condtions in the burner region.  However,
                         at 125 megawatts, where we were operating under biased
                         firing conditions at 3% 02 we were substoichiometric in
                         the burner region.  I don't know what all of them
                         were, but at 125 megawatts we got down to about
                         .85 to .9.

                                                  111-79
L

-------
Q:  And this is where your corrosion data was taken that
    you showed?
A:  Yes.  Again, the corrosion data for the coupons were
    only done for the four selected tests.  They weren't
    done for all of them.  However, the boiler corrosion
    tube data was done over the entire program.
Q:  Gerry, did you note any difference in furnace per-
    formance under bias-firing conditions?
A:  No, sir.
Q:  No increases in spray-flows and things like that?
A:  No,  every once in a while we saw some soot fly
    across from one region to the other.
    [Comment by Symposium Vice-Chairman:]
    I think it is important to point out, as you have
    alrady said, some of the test data is not in yet.
    There are still some questions remaining on the results
    of this program.  Even assuming that it all pans out the
    way it looks, I think it is important to point out that
    this data was taken on one unit; it is not necessarily
    typical of all boilers, especially different designs,
    and a lot of work in this area still needs to be done
    before conclusions can be drawn.
Q:  On the calculations of efficiency, was it done on a
    heat rate basis, Btu's per kilowatt hour, or did you
    calculate it on another basis?
A:  No.  It was merely a boiler efficiency calculation.
                             ni-30

-------
[Comment by Moderator:]
Sometimes on a heat rate basis, it may not change.  In
other words, it takes the same number of Btu's to generate
the same KWH regardless of your conditions.  And it is my
understanding that some boiler manufacturers (I may get
jumped on for this by some of the people who are manufacturing
wall-fired units), are aware of the potential for corrosion,
especially on the wall, on the side walls, and note that on
this unit we had apparent corrosion on the division wall
and on a side wall.  Some manufacturers are building in a
curtain air concept wherein there is some bleed air that is
brought down the side wall to sort of compensate for this.
                       111-81

-------


-------
                 THE EFFECT OF ADDITIVES IN
               REDUCING PARTICULATE EMISSIONS
                FROM RESIDUAL OIL COMBUSTION
                             By

R. D. Giammar, H. H.  Krause, A. E.  Waller,  and D.  W.  Locklin

                          BATTELLE
                   Columbus Laboratories

                       Columbus, Ohio
                           111-83

-------
                      THE EFFECT OF ADDITIVES IN
                    REDUCING PARTICULATE EMISSIONS
                     FROM RESIDUAL OIL COMBUSTION

                                by

           R. D. Giammar,  H. H. Krause, and A. E. Weller
                               ABSTRACT


          In this study, the effectiveness of proprietary additives

and pure compounds in reducing particulate emissions was evaluated in

a 50-hp (500 kW) oil-fired packaged boiler.  It was observed that

additives containing certain alkaline-earth and transition metals in

concentrations in the range of 20 ppm to 50 ppm in the oil were

effective in reducing carbon particulate by as much as 90 percent.

These additives were also observed to be effective in reducing smoke

and polycyclic organic matter.  Explanations are offered suggesting

possible additive effects on the combustion mechanism that could

account for differences in performances of the various additives in

achieving particulate emission reduction.
                        Ill-84

-------
                      THE EFFECT OF ADDITIVES  IN
                    REDUCING PARTICULATE EMISSIONS
                     FROM RESIDUAL OIL COMUBSTION

                                 by

           R. D. Grammar,  H. H. Krause, and A.  E. Waller
                             INTRODUCTION


          The use of small amounts of fuel additives  as  a  means  of

improving the performance of fossil-fuel boilers  has  been  of recurring

interest for many years.  Trials with various fuel  additives have had

different objectives, depending on the type of fuel used and the era

in which the trials were conducted.  Earlier additive trials and investi-

gations were aimed at reducing the deposition of  soot and  ash on heat-

transfer surfaces and minimizing the corrosion that resulted from the

accumulation of these deposits and from their interactions with  the

flue gases.  Recently, the use of additives has been  emphasized  as  a

possible method of reducing air pollutant emissions—mainly visible

smoke and particulate--although some claims have  been made of additives

reducing emissions of sulfur oxides and nitrogen  oxides.
                              111-85

-------
          Although a substantial level of effort has been devoted by

various organizations to the development of additives to reduce air

pollution emissions, the effectiveness of some of these additives has

been in question because of the lack of consistent results in both

experimental laboratory investigations and in actual commercial usage.

These inconsistencies, in part, could be attributed to differences in

combustion equipment, operation, and maintenance; fuel type and com-

position; and perhaps, most importantly, evaluation methodology.

          Thus, while qualitative observations and commercial claims

for the effectiveness of additives are numerous, relatively few quanti-

tative data are available from experimental investigations in which

boiler operating conditions were well-defined and in which emissions

were accurately measured.  Consequently, there are few scientific data

to support an assessment of how, and under what conditions, additives

work.

          This paper discusses a portion of an ongoing research program

for the U.S. Environmental Protection Agency* to systematically evalu-

ate the effectiveness of additives for reducing combustion-generated

air pollutant emissions, especially particulate, from an oil-fired con-

tinuous combustion system.  This program is unique in that an extensive

and well-documented program was conducted to evaluate a large number

of proprietary additives and pure compounds under controlled conditions
*  EPA Contract No. 68-02-0262, "The Effectiveness of Fuel Additives
   in Reducing Emissions from Oil-Fired Continuous Combustion".
                                Ill- 86

-------
in a single unit—a commercial 50-hp (500 kW) Scotch firetube boiler

firing residual oil.  This work is an extension of the work of G. B.

Martin, et al., who systematically evaluated about 300 additives for

reducing emissions from a distillate oil-fired residential burner   *.

          In general, most proprietary additives have not been found

to be effective in reducing air pollution emissions   .  However, some

additives have been found to be effective in substantially reducing

smoke and particulate   .
            REVIEW OF EXPERIMENTAL ADDITIVE INVESTIGATIONS


          To guide the selection of additives for this program, the

literature pertaining to experimental studies of combustion additives

was reviewed**.  Previous investigators have used a wide variety of

organic, organo-metallie, and inorganic compounds in efforts to reduce

the amounts of pollutants formed during combustion, or to convert the

pollutants into easily removable solids.

          The greatest amount of effort in additive development has

been devoted to particulate emissions.  This emphasis has been the

result of the almost universal recognition of smoke as an undesirable

emission over a period of many years.  For obtaining reduction in

particulate formation in fuel-oil combustion processes, experimental

evidence indicated that compounds of the transition metals, manganese,
 *  Numbers in brackets designate References at the end of the paper

**  This state-of-the-art review was a part of the program on EPA Con-
    tract No. 28-02-0262 and will be submitted as part of the final report,
                              III-87

-------
                                            (2)
iron, nickel, and cobalt, could be effective   .  The alkaline earth



                                                                 (3 4)
metals, barium and calcium, also gave evidence of useful activity  '  .




The best results were obtained with organometallic derivatives of these




metals.  The organic portion of the molecule provides the needed solu-




bility in oil and also influenced the stability and volatility of the




additive molecule.




          Only a few investigations have examined the effect of additives




on potential carcinogenic materials in the particulate emissions, but




this aspect is presently receiving a significant amount of attention.




Such potential carcinogenic agents are found primarily in the polycyclic




organic material (POM) that forms during the combustion of the fuel   *  .




Host of the early research was done with gas flames or light fuels such




as distillate oil.  The additives which were reported as effective in




reducing the POM emissions included organic compounds such as nitro-




paraffins and peroxides   , as well as organo-metallic compounds of


                  /g\

manganese and iron   .  Although the physical state of POM and the




carbonaceous portion of the particulate is different, their probable




common origin and chemical reactivity suggest that an additive effective




for one would have some effect on the other.  This result was generally




observed in the earlier diesel engine studies, and where both types of




emissions were measured, the POM usually was reduced to a greater




extent than was the particulate   .




          The U.S. Environmental Protection Agency has conducted two




comprehensive experimental programs that have contributed quantitative
                                III- 88

-------
information on the effects of fuel-oil additives in reducing pollutant


emissions.  One program involved tests of numerous commercial additives


used with distillate heating oil in a residential oil burner    .  In


general, the additives were found to have a minor effect compared to


the opportunity for emission control by proper adjustment and operation


of the burner.  However, the additives containing manganese, iron, or


cobalt did reduce the particulate emissions.  A few organic formulations


containing no metals gave moderate reductions in particulates, but the


concentrations of the additives required were too high to be practical.


No changes were observed in the amounts of nitrogen oxides or sulfur


oxides in the emissions with any of the additives.


          The second EPA program was an investigation of the effective-


ness of several proprietary fuel-oil additives in reducing sulfur oxide

                                                         (9)
emissions when firing residual oil in a commercial boiler   .  No


effects of the additives were found for sulfur oxide emissions or for


nitrogen oxides.


          The experience of the EPA investigators with the inefficiency


of additives for reducing the nitrogen oxides or sulfur dioxide in the


emissions verified the generally negative results reported by the


earlier studies.  Only in the case of sulfur trioxide, which is present


to the extent of 25 to 30 ppm in the flue gases, has an additive effect-


iveness been demonstrated.   Chemicals which form basic oxides such as


magnesium, calcium, and zinc have reduced sulfur trioxide levels sub-


stantially (10).
                                Ill-89

-------
                         EXPERIMENTAL PROGRAM



          Commercially available additives and other materials selected


as a result of the previous considerations were experimentally evaluated


with fuel oil in a 50-hp (500 kW) firetube, packaged Scotch boiler


capable of firing natural gas, distillate oil, and residual oil at


rates up to 2 million Btu/hr and generating up to 1500 Ib/hr (680 kg/hr)

                             2
of steam at 15 psig (103 kN/m ).  Combustion air was metered with a


thin-plate orifice, distillate oil flow with a rotameter, and residual


fuel oil flow was determined from beam-scale readings.  Fuel was sup-


plied to the burner at a fixed rate by a variable-speed, positive-dis-


placement gear pump.  For residual oil requiring preheating, controllers


were used to provide a constant fuel temperature at the pump inlet and


burner nozzle.


          The combustion gases were vented from the boiler through a


12-in (0.305 m) diameter insulated stack.  A special sampling platform


permitted sampling in a straight section of stack, 15 ft (4.6 m) above
                              111-90

-------
the boiler outlet.  Figure 1 shows the overall experimental facility.



Analy t ica 1 Procedure s


          Particulate and POM sampling and analytical procedures (modi-


fied EPA Method 5 train with an adsorbent column) have been described


by Jones and Giammar    .  Gaseous emissions were determined by: para-


magnetic analysis for oxygen; flame ionization detection for unburned


hydrocarbons; nondispersive infrared for carbon monoxide, carbon dioxide,


and nitrogen oxide; and a dry electrochemical analyzer for sulfur dioxide,


Smoke emissions were determined with a Bacharach smoke tester according

                                                      (12)
to the ASTM filter-paper method for smoke measurements    .



Fuel

                                                                   3
          To provide a uniform reference fuel, 6000 gallons (22.7 m )


of a residual fuel oil were obtained that consisted of:  87.5 percent C,


11.1 percent H, 0.31 percent N, 0.95 percent S, and less than 150 ppm


trace metals.  API gravity was approximately 15 .  The distillate oil


used was a typical No. 2 heating oil, with 86.4 percent C, 13.6 percent

                                                                   3
H, and less than 0.1 percent S.  An additional 6000 gallons (22.7 m )


of a 2 percent sulfur residual oil was also procured and additives are


currently being evaluated with this oil.



Selection of Additives


          Fuel additives have been used as a means of accomplishing a


number of functions that can be distinguished as occurring before,


during, or after the combustion process.  For the purposes of this




                               III-91

-------
investigation, only additives that were intended to improve combustion




and thus reduce emissions, were considered.  However,  it is recognized




that fuel handling and postflame treatment additives can in some circum-




stances improve the overall performance of a combustion system,  especially




in marginal cases, by keeping nozzle or heat transfer surfaces clean.




          To categorize these "combustion improver" additives, the follow-




ing three classes were defined:




          •  Organo-metallic




          •  Organic




          •  Inorganic.




Table 1 categorizes the additives used in this study according to defined




classes and subclasses.  Inorganic additives were not  evaluated  because




of their insolubility in oil.  Selection of these additives was  based




on a state-of-the-art review of additive technology and on theoretical




considerations of the mechanism of pollutant formation and possible




modes of additive action.






Concentrat ions






          Concentrations for the proprietary additives were based upon




the manufacturer's recommendations or the results of an earlier  EPA




study   .  For pure compounds, concentrations were based on approximately




27 ppm of metal in the oil (0.1 g metal/gal of fuel oil).
                                111-92

-------
Procedure






          To ensure a valid and controlled evaluation of each additive,




procedures were developed to minimize operational variations over the




entire period of experimentation (approximately 1.5 hr) and to provide




several measurements for evaluation purposes.




          For each additive evaluation run, the reference fuel oil was




fired at a constant, continuous rate in the boiler facility to establish




a baseline operating performance over a range of air/fuel ratios.  Gaseous




and smoke emission data were obtained over this entire range, while a




filterable particulate emission  was determined at one selected air/fuel




ratio.  The additive containing fuel was fired at the same firing rate




over the same range of air/fuel ratios to generate a boiler performance




curve with the additive.  A particulate sample was also obtained at the




same air/fuel ratio as for the reference oil run.  Thus, the effectiveness




of each additive could be determined by comparing the smoke and gaseous




characteristic curves and the particulate loadings with and without addi-




tive for runs made on the same day.  (Except as noted, all data reported




in this paper were obtained at steady state.)




          Particulate consists of two basic components:  (1) an ash




component originating with the fuel, and (2) a carbon component re-




sulting from incomplete combustion.  For combustion at a constant
   Probe and filter catch only.
                               111-93

-------
air/fuel ratio, the ash loading in the products of combustion should




be constant assuming no accumulation on the boiler surfaces;  the  carbon




loading could vary depending on the completeness of combustion.   To




minimize the variation in ash loading, all particulate measurements




were made at the same air/fuel ratio corresponding to 12.8 percent CO.




(about 20 percent excess air).  A baseline firing rate of 80  Ib/hr




(36 kg/hr) was selected to correspond to a commonly used boiler load  of




80 percent.






                       SUMMARY OF RESULTS






          Although additives were evaluated with both distillate  and




residual fuel oils, this discussion will be limited to the residual




oil runs, because steady-state firing of distillate oil in the 50-hp




(500 kW) boiler did not generate appreciable quantities of carbon




particulate at normal air/fuel ratios (over 8 percent excess  air).




Thus, a valid measure of the effectiveness of additives firing with




distillate oil could not be made.  In fact, for those additives con-




taining metals, particulate loadings were observed to be increased by




the use of additives with distillate oil.




          Table 2 summarizes the results of the additive evaluation runs




firing the 1 percent sulfur residual fuel oil at 12.8 percent CO  .  As




indicated in portions (d) and (e) of Table 2, several runs were con-




ducted at different firing rates and one run was conducted with cyclic




boiler operation so that other modes of boiler operation could be
                                111-94

-------
considered.  The additive concentrations or dosages are either designated




as ppm metal in the fuel oil for those additives containing metal even




in trace quantities, or by volume percent for the pure organic compounds.




As for the naphthenic and 2-ethyl hexoic acids, the volume percentages




were selected to give the same volumes as used for the barium naphthenate




and barium 2-ethyl hexoic runs, while the volume percentage for toluene




was selected to give the same volume percentage of the proprietary hydro-




carbon additives.  The results of the evaluation runs with the 2 percent




sulfur residual oil are incomplete at this time.  However, cursory




analysis of the data indicate approximately the same level of additive




effectiveness with the 2 percent sulfur residual as with the one percent




residual oil.






Stat1stica1 Cons iderat ions






          Table 2 includes the particulate loadings for both the daily




baseline run and the additive run made that same day.  These daily base-




line runs were used as a measure of the variation in particulate emissions.




As discussed earlier, the ash loading for every run should be nearly the




same and, accordingly, and deviation in the baseline particulate loadings




from day-to-day would have to be attributed to the deviations in the




carbon portion of the particulate loading .  The standard deviation of
                                 III-95

-------
 the particulate loadings for the baseline runs of Runs 6 through 31 .

 is 10 mg/sm , less than 10 percent of the arithmetic mean of 104
     2
mg/sm  of these runs.  Consequently, each additive was evaluated by

comparing the particulate loading of the specific additive run with

the arithmetic mean of the particulate loadings of all daily baseline

runs.  This method minimizes the effect of random variations in the

particulate emissions but may bias the results somewhat in that addi-

tives with low (or high) daily baseline particulate loadings (assuming

these observed loadings are "real" rather than instrument or measurement

errors) appear more effective (less effective) than if the particulate

loading of the additive run had been compared with the specific daily

baseline loading.  However, regardless of which method of comparison

is used, the conclusions would be similar.

          Figure 2 graphically shows the effectiveness of the proprietary,

organo-metallic, and organic additives in reducing filterable particulate.

The arithmetic means and the estimated ash component of the particulate

loading are shown.  The data indicate, in general, those additives con-

taining certain alkaline-earth or transition metals in concentrations

of about 20 to 50 ppm in the oil significantly reduced particulate

emissions. The use and effectiveness of these additives at higher con-

centrations is somewhat questionable.  It has been observed that even

though carbon burnout can be further promoted at the higher concentra-

tions, the net reduction in particulate is not substantial because of

the additional particulate from the metallic inert contained in the

additive.
*  The level of 100 mg/m  was determined from the standard error of the
   mean of the baseline runs.

                                ill-96

-------
          Also shown in Figure 2 is a line located at two standard



deviations below the baseline average.  This standard deviation is



based on the individual baseline measurements and is regarded as



equally valid for the particulate measurements with additives.



Although the use of two standard deviations strictly requires the



assumption of normality to correspond to a 95 percent confidence



level, it is used here in a broad sense to indicate a reasonably high,



but unknown, confidence level.  Those particulate measurements for



additives falling below the line are concluded to be statistically



significant (not due to chance) at this reasonably high confidence


                                                             3

level.  Those additive runs falling above this level (84 mg/m )  but


              3*
below 100 mg/m   suggest that these additives are effective,  but more



experimental data are required to establish the level of statistical



significance.



          In addition, the measurements indicate that some additives



are more effective in reducing particulate than others.  However, by

                                             3

assuming that a standard deviation of 10 mg/m  also applies to the



additives runs, a statistical treatment of the data would indicate that



the effective additives are not statistically distinguishable.





Effect of Carbon Particulate





          For the baseline boiler operating condition of 80 Ib/hr (36 kg/hr)
*                       3
   The level of 100 mg/m  was determined from the standard  error of the

   mean of the baseline runs.
                               111-97

-------
at 12.8 percent C0_, the average carbon content  of  the  filterable


participate was observed to be about 50 percent  for a corresponding


filterable particulate loading of 104 mg/sm .  Accordingly, additives


can only reduce particulate loadings by about  50 percent  to about

        3
50 mg/sm  which represents the ash component associated with  the


inerts of the fuel oil, unless the use of additives caused a  larger


portion of the particulate to remain within the  combustion system


(this would probably be undesirable as it would  foul heat transfer


surfaces).


          This observation that particulate loading can be reduced to


about 50 mg/sm  is verified by the results of  Runs  13,  14, 32,  33, 34,


and 35 in which the boiler was fired at 5 boiler loads.  Figure 3 shows


the particulate loading generated by the boiler  operated  at each of  5


loads with and without one additive.  For the  baseline  runs,  particu-


late loadings were higher at the 100 percent load (Run  32) and  the 35


percent load (Run 35) than at the 80 percent load (Runs 13 and  14),


65 percent load (Run 33), and 55 percent load  (Run  34), but with the


use of barium naphthenate, particulate loadings  were reduced  to about

                                        3
the same level of approximately 55 mg/sm , regardless of  load.  This


particulate loading is about 10 percent higher than the anticipated


ash level when firing the referenced fuel alone  and can be attributed


to the increase in metallic inert of the oil by  addition  of the barium


naphthenate.  Analysis of the particulate loadings  of these runs  (except


Run 32) indicated that over 85 percent of the  particulate was ash.
                               111-98

-------
Additive Type and Concentration






          The data obtained here indicate that some additives containing




metals substantially reduced particulate emission.  These additives were




used in concentrations corresponding to the addition of between 20 ppm




and 5fl ppm of metal to the fuel oil.  Proprietary additives considered




organic with some trace metals also were shown to be effective at con-




centrations that corresponded to the addition of at least 15 ppm of




metal to the fuel oil.  As in the case of the additive "Formula LSD",




these concentrations were significantly higher than those recommended




by the manufacturer.




          In general, the organo-metallic compounds listed in Table 2




were effective in reducing particulate except for two additives in the




subclass cyclopentadienes (CI-2 and PD 1654, a ferrocene derivative)




and zinc naphthenate.  The proprietary additives (CI-2, Runs 4 and 20)




and PD 1654 (Run 5), although observed to be effective in reducing




particulate in other types of combustion systems firing distillate




oil   , were found to be ineffective in the 50-hp (500 kW) boiler




firing residual oil; their ineffectiveness could be attributed to the




type of bonding structure of these compounds (for PD 1654, the additive




concentration used was that recommended by the manufacturer rather than




one based upon a 27 ppm metal in the oil).  The data indicate that the




most effective agents are compounds of the alkaline-earth metals (barium




and calcium)  and the transition metals (manganese,  iron,  and cobalt).
                                  111-99

-------
 Zinc,  an element just beyond the above transition metals, was found to


 be an  effective additive.


          Some of the pure organic additives considered appeared effec-


 tive in  reducing particulate emissions.  However, these additives had


 to be  used at impractically high concentrations to be marginally effec-


 tive and, therefore, could not be considered as a viable pollution con-


 trol method.



 Effect on POM



          POM (polycyclic organic matter) measurements were made on a


 limited number of additive runs.  These measurements included both total


 POM and  species quantification for over 15 compounds (the majority of


which are carcinogenic) as listed by the National Academy of Science   .


 In general, additives that were effective in reducing particulate appear


 to be  effective in reducing POM.  Even if more data were available, it


 is improbable that any specific relation between additives and POM

 reduction could be made.  It is believed that subtle changes in the


 combustion process that cannot be detected by the gross measurements


 of stack gas analysis have a significant effect on the formation of POM


 that would tend to mask an additive effect, especially considering that

                                                   3
 POM loadings were observed to be less than 10 p.g/sm  for continuous


 steady-state combustion.  For these relatively small POM loadings, over


 99 percent of the POM found was considered innocuous.
                           111-100

-------
          POM measurements were also made during the cyclic operation



of the boiler in Run 36.  Although this size range of boiler would not



be operated in an on-off cycle, but rather modulated to meet load, it



was felt that some useful data could be obtained by operating the



boiler for a 15-minute on/5-minute off cycle.  As noted in Table 2,



particulate loadings increased by a factor of two for both the baseline



run and the barium naphthenate additive.  However, POM levels of the


                                                                 2

baseline run increased by over two orders of magnitude (920 jjg/sm )



above the steady-state runs.  For this run the additive was signi-



ficantly effective in reducing POM (to 120 p,g/sm ) although the POM



emissions still were much higher than for steady-state runs.  This



observation agrees with that of an early investigation in which it



appeared that POM was reduced to greater extent than particulate by


                  (4)
use of an additivev  .  At the higher POM loadings, several compounds



considered carcinogenic were found in appreciable quantities.






Effect on Smoke
          As a measure of the effectiveness of additives in reducing




particulate over a range of air/fuel ratios, Bacharach smoke measurements




were made at about six excess air levels for every daily baseline run and



every additive run.  Figure 4 shows the effect of cobalt naphthenate




(Run 7) and barium naphthenate (Run 13) on smoke reduction.  In general,




the relation between Bacharach smoke and (XL for the daily baseline runs




as shown by the curve in Figure 4 remained constant from day to day as




would be expected from the small variations in particulate loadings of




these runs.  Accordingly, the effectiveness of the additive could
                              III-101

-------
semiquantitatively be evaluated by comparing the  smoke  characteristic


curves with and without the additive.   The relative differences  of


these curves could be related to the relative differences  in  particulate


loadings at 12.8 percent CO .  From Table 2, for  Run  7,  the ratio of the


particulate loading with and without additive was 0.71  and was 0.46 for


Run 13, at a CO  level of 12.8 percent.




Effect on Efficiency




          Figure 4 is a plot of stack-gas temperature as a function of


time for 3 long-term (60 hours) runs with the 2 percent sulfur residual


oil.  Barium naphthenate was used at two boiler operating conditions;


one condition matched the excess air level of the baseline run while

                                             *
the other condition matched the smoke  density of the baseline run.


          The results of these runs are somewhat  inconclusive because


for the relatively short duration of these runs,  it is  difficult to


predict the stack-gas temperature for  an extended period of time (such


as several months).  It appears that the slope of the baseline curve


is leveling off after 70 hours, and it is possible that by projecting


the curves of all three runs, the curves would intersect.  In all likeli-


hood, however, the slopes of the additive runs should also level off and


not intersect with the slope of the baseline run.


          However, by assuming that the time-rate-of-change of the stack-


gas temperature (the slope of the curves) of all  three  runs is the same


after 60 hours, the effectiveness of barium naphthenate to increase
                           III-102

-------
boiler efficiency can be estimated.  A reduction in either or both the




mass flow-rate and temperature of the stack gas will increase boiler




efficiency.  For the barium naphthenate run at 14.2 percent CO ,  an




increase of boiler efficiency of about 2 percent is realized.  About




half of this increase can be attributed to the lower stack-gas temperature




while the other half can be attributed to the lower excess air level.
                  POSSIBLE EFFECTS OF ADDITIVES

                     ON COMBUSTION MECHANISMS
          To explain the differences in performance by the various metals




which achieved some reduction in the particulate emissions, it is necessary




to hypothesize as to the mechanism by which they act.   The transition




metals--iron, manganese, and cobalt are known to be oxidation catalysts




in a variety of chemical systems.  Consequently, their activity could be




attributed to catalytic enhancement of the carbon burnout in the fly ash.




Differences among these metals could result from the degree of catalysis




displayed by the particular metal.  However, in these  experiments no




statistically significant difference was found in the  performance of these




three metals.




          The alkaline earth metals--barium and calcium—are not noted




for catalytic activity, so the mechanism by which they reduce particulates




is very likely different from that of the transition metals.  It has been



                     (13)
suggested by Friswell     that alkaline earth metals may function by




promoting the dissociation of water to H atoms and OH  radicals,  a process








                              III-103

-------
occurring only at high temperature.  The OH radicals are involved  in




the combustion of carbon:




                         C + OH - CO + H




                        CO + OH -* C02 + H.




The H atoms in turn react with oxygen in the combustion air  to  form




additional OH radicals and 0 atoms, namely, 0 + OH -* H + 0.




          For differences between various compounds of the same metal,




the bond strength between the metal and the remainder of the  molecule




may be the controlling factor.  This aspect was  explored most completely




for barium.  The number of available materials was  limited by the  necessity




that they be soluble in the oil.  In this investigation, the  naphthenate,




hexoate and the sulfonate were used, all of which are barium salts of




organic acids.  Although some differences in performances of  these barium




derivatives can be noted in Table 2, more experiments are needed to make




a statistically valid evaluation.  In the cases  of  iron and manganese,




the situation is similar, although the compounds used cannot  be as




readily compared.  These mechanistic considerations can be used to




explain differences, but the possibility should  not be eliminated  that




other properties such as volatility of the additive and the nature of




the oil may influence the effectiveness of the additive in reducing




particulate emissions.
                           III-104

-------
                   PERSPECTIVE ON FUEL ADDITIVES


          In this study, additives containing certain metals used at

concentrations of approximately 20 ppra to 50 ppro of metal in the fuel

oil were found effective in reducing particulate emissions.   The results

of this study, for the most part, have confirmed the observations of

other researchers and have extended the range of investigation of pro-

prietary additives and pure compounds to commercial size combustion

equipment firing residual fuel oil.  It should be pointed out that the

mechanisms involved are not fully understood and there is evidence to

indicate that additives effective in one system firing one type of fuel

may not be effective in another system or when firing a different fuel.

          Before any general or widespread use of fuel additive can be

recommended, a perspective should be developed for each type of boiler

installation to consider:

          •  Possible toxic effects of pollutant emissions
             resulting from materials introduced with the
             additives

          •  Practical considerations of additives use (such as
             economics, corrosion resulting from use and increase
             in other pollutants) other than simply the reduction
             of one or more pollutants, and

          •  Comparison of the effectiveness of additives with
             alternative control means.

It is anticipated that there should be no serious economic penalty with

the use of additives as some of the pure compounds are relatively in-

expensive.  In addition, it should be noted that the level of iron
                              III-105

-------
 introduced as iron naphthenate was less than the iron present  in the




 oil as ash so that use of iron containing additives for this oil may




 not introduce any new pollutant.  However, the long-term effects of




 additive use are not well-established nor are the nature and potential




 toxicity of these metallic compounds adequately defined.








                         ACKNOWLEDGMENTS






          The research upon which this publication is based was performed




 pursuant to EPA Contract No. 68-02-0262 with the United States Environ-




 mental Protection Agency, Division of Control Systems.  The authors wish




 to thank EPA for granting permission to present this material, and to




 EPA staff members, G. B. Martin and W. S. Lanier who participated in




 planning this program and have provided helpful comments.  The authors




 also thank other members of the Battelle staff who have contributed to




 this study—A.  E. Weller, A. Levy, P. W.  Jones, L.  J. Hillenbrand,




R. E. Barrett,  R. Coleman, T. C. Lyons, and H. G. Leonard--for their




advice and assistance in carrying out the research.
                           Ill- 106

-------
                             REFERENCES
 1.  Martin, G. B., Pershing, D. W. , and Berkau, E. E., "Effects of Fuel
     Additives on Air Pollutant Emissions From Distillate-Oil-Fired
     Furnaces", U.S. Environmental Protection Agency, Office of Air
     Programs, Publication No. AP-87, June, 1971, pp 86.

 2.  Weeks, R. L., Clinkenbeard, W. L. , and Soltis, J. D., "Clean,
     Efficient Combustion of Heating Oil", Proceedings of the 5th World
     Petroleum Congress, New York, Section VI, Paper 25, June, 1959,
     pp 381-396.

 3.  Riggs, R. J. Wilkinson, C. J., and Wolfe, H. R., "Combustion Im-
     provers for Fuel Oils", Erdoel and Kohle, 18 (4) 282-286 (1965).

 4.  Perilstein, W. L.,  "Overbased Barium Sulfonates for Use in Distillate
     Fuel Oils", U.S. Patent 3,580,707, May 25, 1971, 8 pp.

 5.  Hangebrauck, R. P., von Lehmden, D. J.,  and Meeker, J. E., "Sources
     of Polynuclear Hydrocarbons in the Atmosphere", HEW, A.P. No.
     999-AP-33, Durham,  N.C., 1967.

 6.  Committee on Biologic Effects of Atmospheric Pollutants, "Biologic
     Effects of Atmospheric Pollutants; Particulate Polycyclic Organic
     Matter", National Academy of Sciences, National Research Council,
     Washington, D.C.,  1972, pp 375.

 7.  Ray, S. K., and Long, R., "Polycyclic Aromatic Hydrocarbons from
     Diffusion Flames and Diesel Engine Combustion", Combustion and
     Flame, 8, 139-151  (1964).

 8.  Spengler, G. and Haupt, G., "Formation of Soot and Polycyclic Aromatic
     Hydrocarbons in Simple Hydrocarbon Flames and Its Inhibition by Fuel
     Additives", Erdoel  un Kohle, 22  (11)  679-684 (1969).

 9.  Pershing, D. W., Martin, G. B.,  Berkau,  E.  E., and Hail, R.  E.,
     "Effectiveness of Selected Fuel  Additives in Controlling Pollution
     Emissions from Residual Oil-Fired Boilers",  U.S.  Environmental
     Protection Agency Report No. EPA-650/2-73,  October, 1973.

10.  Rendle, L. K. and Wilson, R. D., "The Prevention of Acid Condensa-
     tion in Oil-Fired Boilers", J.  Institute of Fuel, 29, 372 (1956).

11.  Jones, P.  W., Giammar, R.  D., Strup,  P.  E.,  and Stanford, T.  B.,
     "Efficient Collection of Polycyclic Organic  Compounds from Combustion
     Effluents", presented at the 68th Annual Meeting of the Air Pollution
     Control Association,  Paper No.  75-33.3,  Boston, June 15-20,  1975.
                               Ill- 107

-------
12.  Standard Method of Test for Smoke Density in the Flue  Gases
     from Distillate Fuels,  ASTM Designation:  D 2156-65  (Reapproved
     1970).

13.  Friswell, N.  J., Emissions from Continuous Combustion  Systems.
     edited  by W.  Cornelius  and W.  G.  Agnew, Plenum Press,  New York,
     1972, p 161.

                             111-108

-------
                     TABLE 1.  ADDITIVES CLASSES
      ORGANO-MEX

Naphthenates

  Cobalt
  Iron
  Manganese
  Barium
  Iron + barium
  Zinc
  Calcium


Cvclopentadienes

  Manganese - CI-2 (18% Mn)
  Iron - PD 1654 (1% Fe)
Sulfonates

  Barium
Ethyl Hexoate

  Barium

Carbony|s

  Iron
           ORGANIC

Hydrocarbon

  LSD  (C,H0, trace metals)
         / o
  Improsoot  (C_H0, trace metals)
               / o
  Triple E
  Toluene (C?Ha, pure hydrocarbons)

  Watcon 130 (trace metals)


Ether

  Diglyme

Alcohol

  Ethyl
  Hexyl

Acid
  Napthenic
  Ethyl hexoic
  MC-7 soluble (57=, Ba, 2% Mn)
  Rolf it e 404 (27, Mn)
                                  III- 109

-------
                                xuiu a.  suwm or EVALUATION KIMS
torn
                        Additive
                                                           t>OM»«
                                                                             fattlculete toadIng
                                                                             Maltlve   Basel in*
                                                                            •  mrttm*
                                                                                                         rartleulate
                                                                                                          loading
                                                                                                           »atlo(e)
  9
 10
 11
 12

 13
 14
 IS
 16
 1?
 1ft
 19
 20
 90
 31
                      ••  Ittltlil Evaluation Kuns Vith  Proprietary Additives
       HJ-7 soluble  (Apollo)
       ronula tSD (Coisacrelal Chcnleal)
       iBprotoot  (Comercial Chcaical)
       CI-2 (Ethyl Corporation)
       10 1654  (Arapahoe)
                                           35 pp. to. U pp. tt>
                                           3 ppi Co. * Ppm Mn
                                           9 pp. Co, 2 ppmCa, 10 ip.lta
                                           21 pp. Ha
                                           5pp*r«
                         b.  Evaluation Runs Vtth Oreano-Metallic Additives
Zinc naphthenate
Cobalt naphthetut*
Iron B*phth«Mte
Iron a*phthcnac«
Hug*ne*c Mphchenate
Calelua Mphthcnatc
Iron and barium n»phth«n»t«

Bariua n*phthctue*
lariua naphtheiuc*
laxim naphthcMt*
••elm n«phth«i*te
Bcrlua 2-ethyl
Barium tulfocxtt
Iron pent*ttrbonyl
Cl-2 (Ethyl)
21
22
23
2*
25
2*
27
28
2*
Ethyl hexolc *cld
Itopthcnle *cid
Ethyl alcohol
•cxyl alcohol
Digly»e
Tolucn*
t-4000 (Triple E Product!)
. t-«000 (Triple E Products)
ronul* LSD (CcxmwrcUl Ch








cvical)
Vatcon 130 (Induttrial Chemicals)
kolfit* 404 (Andrew Rolfe Chemical Co.)
                 27 pp. ta
                 27 pp. Co
                 27 pp. F.
                 iipp-r.
                 27pp.Mn
                 27pp.Ca
                 27 pp. Fa, 27 pp. ft

                 31 pp. to. 0.051  by
                 51 pp. to
                 26 pp. to
                 10 pp. to
                 40 pp. to, .031 by voi
                 20 pp. to
                 27pp.ro
                 21 pp. to

Evaluation Runs With Organic Additives
0.03X by volume
0.05X by volume
1.5X by voluw
l.Sl by volume
1.5X by voluo*
0.2X by volume

0.031 by volume
0.21 by voluw
0.2X by voluae, 16 pp. Co,
  • pp. MB

0.21 by volume, 16 ppn Fe
0.21 by voluae, 6 pp. Mn
                                    51
                                    77
                                    4*
                                   100
                                    M
                                   106
                                    74
                                    W
                                    71
                                    77
                                    62
                                    35

                                    4$
                                    50
                                    75
                                    •9
                                    55
                                    •3
                                    7i
                                    97
                                                                                    102
                                                                                     •3
                                                                                    1«1
                                                                                     79
                                                                                     90
                                                                                    114

                                                                                    118
                                                                                    101
                                                                                     69
66
86
           •1
           80
           90
           86
           77
          115
          109
          127
          108
          106
          115
          108

          117
           92
          102
          104
           93
           98
          103
          103
                                                                                         102
                                                                                          91
                                                                                         120
                                                                                          86
                                                                                         104
                                                                                         109

                                                                                         115
                                                                                         110
                                                                                         100
 87
102
             0.61
             0.9)
             0.38
             1.20
             1.08
             1.02
             0.71
             0.77
             0.68
             0.74
             0.60
             0.53

             0.46
             0.48
             0.72
             0.86
             0.53
             0.80
             0;76
             0.93
                       0.98
                       0.82
                       0.97
                       0.74
                       0.87
                       1.03

                       1.13
                       0.97
                       0.66
                                                                           0.63
                                                                           0.84
                 4.  Evaluation Run*With Variable Boiler Lead at Constant Air/fuel  Ratio

                                                      Particular Loading
1HH
32
33
34
35
la
36 •
Additive
Barlm naphthenate
ditto
•
AJa-tttve
toxlua naphthenate
Dosage
51 pp.
51 pp.
51pp.
51.PP«
*. EvsltMt
Dosage
51 PP-
tiring Hate
Ib/hr
nc/srnJ
98 47
66 53
54 31
36 55
ion Run Vlth Cvclie-Boller
Partlrul.ite
Additive
134
Loading
Baseline
206
i^.Kaseline
148
115
. 105
147
Operation


 (•)  The ratio of the part leu late  loading of the additive run and the arithmetic .eaa of th«
     paniculate loadings of the appropriate baseline runs.
 (»>  Ib/hr - 0.454 ';c/hr.
                                                    IH-110

-------
FIGURE 1.  EXPERIMENTAL  50-HP BOILER FACILITY





                III-111

-------
[XXXXXXXX 1X\\\X\X\
10
 a
mean
hmet
     t
      is
¥^ft
\\X\\'
            5
        \ \\X\\\\
        k\\\\\
        xx
                  \
          LXXXXXX
                   \
        xxxxxx
^v
                              \\\
              fc
                   XXXXXXXX
                    \\x\\\\\\\\\\\
         \xx\x\\\
§s
             \\
                    •v^Y^
                  \\x\\\\\\\\\
       
                0
                if)
              Q
              K)
                                       O 04
                                       O O
IMENTAL DATA FOR

LOAD AND 12.8 PERC
EXPER
SUMMARY OF THE

RUNS AT 80 PERCE
E 2
FIG
     £uus/6uj '
                   III- 112

-------
    160
    [40
    120
 o»
•o
o
o
 O
O
0_
    100
     80
60
     40
     20



1 (Baseline
-
—






-

-















V///\ Barium Nopthenate
T5
O
51?
O
g
7]
/
/






















X3

~~
ft
b
CO
\
f

^^^













^
o
5
. o
0s-
O
CO
\















•o
0
0
^o

in
ID
\\\X\\\\\\\\\\\\VI

^^H













T^J
o
—
,0
o
m
7
I
I
x


















0
Q
0^
m
rO
!

            32
   14      32
Run Number
                                       34
35
       FIGURE 3.  THE EFFECTIVENESS OF BARIUM NAPHTIIKNATE
                 IN REDUCING PARTICIPATE AT VARIABLE
                 BOILER LOAD FOR CONSTANT AIR/FUEL RATIO
                         III- 113

-------
    490 (254)
    480 (249)
    470 (243)
~  460 (238)
 5  450 (232)
    440(227)
o
o
    430 (221)
    420 (216)
    410 (210)
    400 (204)
                                                   Baseline, 12.8%
                                                            C02
iorium napththenate
(51 ppm Bo)
      12.8% C02 .
                                       Barium napththenate
                                       (SIppmBa)   14.2 %C02
                    10     20     30    40     50
                                     Time, hours
   60     70     80
             FIGURE 4.  STACK-GAS TEMPERATURE AS A  FUNCTION OF TIME
                       FIRING 2-PERCENT SULFUR OIL AT  80 LB/HR
                                 IH-114

-------
5:15 p.m.
The Effect of Additives in
  Reducing Particulate Emissions
  from Residual Oil Combustion
Robert D. Giammar, Battelle Columbus Labs.
Could you elaborate some more by what you meant by
your cyclic operating conditions and why the POM
levels were so high?
The boiler was operated in a 15 minute on, 5 minute
off fashion for 10 cycles.  The particulate measure-
ments were made during this period.  We suspect that
the POM levels were higher during  the cyclic mode
of operation than the continuous mode of operation
because of the characteristic of burner shutdown
and startup.  During these periods there is poor
fuel atomization and some of the oil will get through
the vaporized field partially unburned.  We found
out that our concentrations of POM in raw fuel oil
were orders of magnitude higher than the concentra-
tions that we have seen in the stack gases from con-
tinuous combustion, and so any unburned fuel collected
will give a high POM level.
You had a slide up there which showed stack gas
temperature on the vertical axis vs. time for no
additives and two different additives.   I noticed
for one of the additives, at time zero there was
significantly lower temperature than the other two,
Should all the stack gas temperatures start at the
same levels?
The air/fuel ratios of these curves were different,
In comparison to the two top curves that were for
20% excess air, the lower curve corresponded to
10% excess air.  A lower percentage of excess air
                       111-115

-------
    will  give  lower  stack  gas  temperature.
Q:  That  was at  a  common smoke number.
A:  Yes,it  is  a  common  smoke number.   The  lower  curve
    with  the additive was  a number  4  smoke which was
    the same as  the  baseline.
    You had mentioned  that you  did  a  great  deal  of
    testing not only with pure  compounds  but  also with
    a large number of  proprietary additives.   If you
    were asked to make some suggestion, if  one of these
    manufacturers or users were  to  use a  fuel  additive,
    would you care to  comment on what your  suggestions
    would be in terms  of a particular additive?  Should
    they go to a pure  compound?  Should they buy a
    proprietary additive?  And what are the economics
    involved ?
    In our study we found that  the  alkaline earth metals
    and the transition metals were  effective in reducing
    particulates or carbon particulates.  These were
    procured in pure compounds... the naphanates, the
    hexoates...which were significantly less expensive
    than some of the proprietary additives.
    Is any work being planned on the toxicology of the
    particulates?
    It's not within the scope of this program.
Q:  Would you care to hazard a guess on average residual
    oil and average set of conditions of operation, the
    cost of treatment per thousand gallons or a range
    of possible costs?
                          III-116


-------
A:  I don't have those numbers with me.  I don't have
    the specific costs but I made a calculation based
    upon the following:  if you  could  save a  dollar  in  fuel
    costs by the use of additives with an increase of
    efficiency of a couple percent, then you  would have
    to take 20% or 20 cents of that dollar to purchase
    your additive.  I do want to mention, however , that
    we had to detune this boiler.  This boiler could be
    tuned so that it generated very little carbon particu-
    late and there would have been no  point in doing
    our studies with no carbon particulate.   We tried
    to establish a smoke characteristic curve that was
    typical of that found in the field, and we relied
    upon some EPA field study data for this purpose.
    Sawyer and co-workers out at Berkeley have shown
    that several of these additives can significantly
    affect the particle size distribution by shifting
    them to lower sizes.  Did you try at all to measure
         •
    the size distribution and,based upon your experimental
    techniques,would you expect such a shift to alter
    your conclusions?
    We did not make any particle size measurements at
    this time.
    Did you run a complete metal analysis on your
    fuel to find out if you had any substantial trace
    metals present?
    Yes.  Trace metals were low, less than 150 ppm.
    The question of how these metallic inerts present
    in fuel affect the formation of carbon particulate
    was of concern to us.   Some of these trace metals
    can be present in the  oil at higher concentrations
    than are introduced into the oil with the additive,

                          III-117

-------
    Apparently, these trace metals are not nearly as
    effective as the metallic additives.  We suspect,
    however, that vanadium in some oils might have
    some effect.
Q:  Would you care to rank your alkaline earth metals
    and your transition metals in terms of their
    effectiveness as seen by your test?
Q:  Statistically, they are equivalent.  We just don't
    have enough data.
Q:  Would that be on a molar basis or a weight basis?
A:  On a weight basis.  It appears that the alkaline
    earth metals were more effective than the transition
    metals.  Calcium was our best performer, but on a
    molar basis, calcium and barium gave about the
    same results.
    Several investigators in premixed laboratory flames
    and diffusion flames have noticed the effect of
    additive concentration on either promotion or
    inhibition of soot formation.  Did you notice
    any effect on concentration of your additives?
    Yes.  In our studies we noticed that increasing
    the metal concentration in the oil almost always
    decreased the carbon found in the particulate
    catch.  For example, the metallic cyclopentadiennes
    reduced carbon particulate by only 30% at concentrations
    of  .1 g of metal per gallon of oil but reduced carbon
    particulate by over 90% at concentrations of 1 g of
    metal per gallon of oil.   However, the resulting total
    particulate loading would increase.
                        Ill- 118
                                                                               1

-------
   SYSTEM DESIGN FOR POWER GENERATION
           FROM LOW BTU GAS
                   By:
       M. P. Heap and T. J. Tyson
           Ultrasysteins, Inc.
                 and
             N. D. Brown
       Combustion Engineering

The work reported in this paper was carried
     out under EPA Contract 68-02-1361
              III-119

-------
 1.0    INTRODUCTION




       The conversion of coal to a clean fuel gas is one attractive




method of utilizing the nation's coal reserves in an environmentally




acceptable manner.  Alternate approaches include the use of fluid bed




coal combustors or conventional coal-fired boilers with flue gas




desulfurization.  This paper addresses the question of how existing




technology might best be used to generate electric power from gasifier




off-gas.  The total study has yet to be completed, therefore this




paper represents a partial progress report.  The primary objective of




the study is to identify those areas which are likely to give rise to




emission problems in the design of a 500 MW power plant fired by low




Btu coal-derived fuel gas.  Having identified problem areas and assessed




their magnitude, it is the intent to suggest design modifications to



allow these problems to be overcome.




       The design study is based upon the provision of a 500 MW power




plant with the following characteristics:




       -    low plant heat rate;




            low N0x emissions (0.1 Ibs N02/106 Btu fired);




       -    high turndown ratio (5:1);




       -    low capital cost;




            low operating cost.




The restrictions placed on the study prevented any consideration of




integrated power plant/gasifier systems and it was required that the




economic analysis be conducted on the basis of purchased coal gas, neces-




sitating an assessment of an "over the fence" cost.
                                  III-120

-------
2.0    SYSTEMS STUDY




2.1    Fuel Characteristics




       The systems study was carried out on the basis that two general




classes of fuel gas can be considered as representative of existing




technology:




       -    a low Btu gas typical of that produced by a Lurgi




            air blown gasifier;




            a medium Btu gas typical of that produced by an oxygen




            blown Koppers Totzek (K-T) system.




The compositions of these two fuels are presented in Table 1.
   Table 1.  Fuel Gas Compositions for the Systems Study
                                                        (After 1)

CO
H2
CH4
co2
TT f\
EL*\v
N2
LHV (Btu/scf @
60°F 1 atmos)
Low Btu
(Lurgi)
14.1 %
20.9 %
5.8 %
12.5 %
6.6 %
40.1 %
160
Medium Btu
(K-T)
53
36.4
-
9.25
0.3
1.05
272
       In addition to variations in heating value it  was  also considered



that the fuel gas may be available at high pressure (in excess of




10 atmospheres)  and high temperature (approximately 1500°F).   It has been
                              III-121

-------
shown  that coal gas combined cycle power plant systems only have an



overwhelming economic advantage over conventional plants if advanced



turbine concepts allowing turbine inlet temperatures of 2600 F can be


     (2)
used'  .  High temperature/pressure fuel gas was included in this study



for purposes of comparison, although the assessment of fuel cost pre-



sented a problem.



2.2    Systems Considered



       Over forty conceptual power plant systems were considered which



were capable of using the fuel gases discussed above.  These concepts



can be classified into four groups:



       1.   Conventional boiler systems with low temperature fuel gas or



            systems which include either a fuel gas precooler or high



            temperature burners to utilize high temperature fuel gas.



       2.   Combined cycle systems of the waste heat and auxiliary



            fired steam boiler type.



       3.   Supercharged boiler systems.



       4.   Oxygen-fired boilers - if oxygen is available in bulk



            for the gasification process then it could also be



            utilized in the power plant.



Each of these four general classes includes a wide spectrum of system



concepts.  After all the concepts had been studied, five were chosen



for more detailed consideration.  These systems are listed in Table 2.



System Al was chosen because it represents a baseline case suitable for



retrofit.  Waste heat combined cycle plants were not considered because


                                                                  (2)
they were being studied in detail in another EPA-sponsored project   .






                               Ill- 122

-------
                 i
  3
                     §   2
                                      o— —
                 O  •
                 U


                 §9
                                                   u


                                                   3
<
oc
 at
 4J
 OJ
a
•a
 -
zz ~
23 £
|Si§
1 . sis
CM

a)
                 o
                 m
                     g   2
                        «  I
                        ?  &
                         £
             Ul
y
111
o

u
                                      111
                                          III- 123

-------
Ill-124

-------
                           CROSS SECTION OF
                           PROPOSED BURNER
                                                 IGNITER
Figure 1.  Sketch of High Pressure Boiler
      Showing Detail of the Burner
            III-125

-------
i—
        i
        o:
—VV
	^
   I
  _ J
                   £U

                   V)
                  <5'
                   .8.
A
                           — '
                           ~A-
                                -J
                          -A
                                            tn
                                            M
                                            O
                                            *a
                                            01
                                            
-------
has a maximum continuous rating of 612,000 Ibs/hr main steam flow with




superheat and reheat steam conditions of 1011°F/2685 psig and 1001°F/




501 psig respectively.  Superheat and reheat steam temperature control




is achieved via desuperheat spray.  Flue gas leaves the boilers at 2000°F




and is delivered to three gas turbines; heat losses from the system are




minimized by routing the air from the compressors to the boilers in an




annulus surrounding the flue gas pipe.  Upon leaving the gas turbines,




the flue gas passes through a waste heat recovery chain consisting of an




evaporator-superheaters, economizer and a feedwater heater.  The flue




gas at 300 F then passes to the stack.  The three gas turbine generator




sets deliver a total net output of 191 MW; a single steam turbine genera-




tor delivers 308 MW net.




       Calculated plant performance for each of the five systems is




summarized in Table 3.




2.3    System Economics




       Capital cost estimates have been prepared for each system




assuming all plants were outdoor-type located at an average U.S.  site




using the standard Federal Power Commission account numbering system.




The performance and operating costs for 1974 operation are summarized




in Table 4,  and the terms used in this table are defined in Table 5.




The gross electrical output of each plant is approximately 520 MW and




has been corrected for typical auxiliary power requirements for
                             III- 127

-------

                                                           §
                                                           o
                                       5    m     ui
                                                                                              8
               8
                   ->  tt.

                   z
                                                    §
                                                    in
               O
                    O
                         .
                .^  8
                              8
                              in
                                                    8
                                                    m
                      •    SJ 2
                     
m
in

pi
 CO

H
               ~
             <3
                                                                                            oo

                                                                                       fc    JOS
                 o

                 t-
                        u.  zui
                                                                                              «O
                                               III- 128

-------
 c
 o
            5*:
            Ou
                               OtVMXJ i O
                               B-  :s
                                            ... in   *    *
                                            0 — 0  «o   o
                                            in 01    o>   w
                                            r«- * f\j  ru   W
                                            M •-    <«•   —
                                              ni    cu
                                                    oh-   tntv
                                                    mm   ~f-
                                                    Oin   r»is>
                                                          v ^
                       O
                       CJ
                       S
                          Q o com
                          O   ^«*
                                               CM     N
                                               8
                                                                               ota
                                                                               — o
 CO
 VI
 V
 P.
O

vr
r>-
a\
   OQ
-  22
       O* ^   ftj OO
       v ea   inSS
    llil
    toe
    to.
                                              >O *  f-  O
                                              eo  •  in  nj

                                              so ""  «  2
                                              •-    
                     U
                  U. 5
                  K m 0.00
                  < CK
                  a: D
                  K H u


                  2< t
                  ~ O ^~
                          X
                            ce

                            SK
                              K     v> -J  ?<
                              !»! an:  > 5  X  £
                              z_>.>.  t ^  ri  i~
                                    ? y^gg
                   UJ =1 H
                   z z «
                                       <  l-l-

                                       z
                                                  x^ go 5a
                                              Q. '^1 Zw Z =
                                              < D Z»- Z<
                                              U it. « <2
                                                                    K
                                                                    §x
                                                                    *
                                                              *    §
                                                        ?SS I9S8
                                                             OQ.
                                                             •-O
                                                                             OHH


                                                                             »"*
                                                                             til
                                        III-129

-------
•O
V
n

-------
condenser cooling pumps, boiler circulating pumps, FD, ID and GR fans.




Transformer losses are 0.5 percent of the net output.  The boiler feed




pumps are steam turbine driven and are included in the steam cycle.




The net plant heat rate is the gross (higher heating value plus sensible)




fuel heat input divided by the net output.




       Annual net generations are the net full load output times 5000




and 7500 hours per year.  These are equivalent to load factors of




57 and 86 percent.  A fixed charge rate of 20 percent of capital cost




per year is assumed as typical for an average U.S. site.  Annual fuel




costs for 1974 are based on $1.25/10  Btu as the delivered price for




10 psig Lurgi LBG and $1.40/106 Btu for 147 psig Lurgi LBG.   The




delivered price for Koppers Totzek MBG is $1.50/10  Btu; plant 15 has




an additional cost of $0.55/10  Btu for oxygen for the boiler, resul-




ting in a total of $2.05/10  Btu.  In order to arrive at a realistic




cost of the high pressure Lurgi gas, an evaluation of the relative




gasifier capital costs for low versus high pressure fuel gas yielded an




increased cost of $0.15/10  Btu for the high pressure gas.  This is




associated with the increased auxiliary power requirements for the




gasifier because the product gas        expanded to 10 psig.




2.4    Discussion




       Table 6 summarizes the plant costs and performance of the five




power plant systems considered suitable for the application of existing
                            III-131

-------
 01
 4-1
 in
 o
 o


 00
 c
 cd

 s
 a
o

T3
 C
 rt

 0)
 o
 d
 o
U-4
 M

 (U

PLI
 C
 nJ
vD


(U

t— 1
H





^^
<:




,—1
33





i— 1
O






1—4
OQ





r~f
•<



















3
pQ
§
Z

H
£5
A




O
o
lA











^
s
H

(^
g
O
H
3
CH

S
^
O


£-4
55
O
CM
ON
00



O
— i
CM
^«4

^
*^
CO
CM
^-1

vD
00
f-^
^O
F-H
^


1-^
o
«-4
CM
CM
rH


rH
f^
fVfc
LA
CM
*— i


f^»
st
ON
!A
CM
t— 1





CO
1
O
i— i

X


H
C/l
O
O
9
H

h^
to
w

^J
^J
f— 1
g
CO
CM



CO
^^
CM



^
CO
CM




N^
«*
CM





CM
lA
CM




«M
lA
CM






















J3
S

H
W
^
•co-
co
CO
lA
CM

00
ON
r^b
O

O
i-H
c-
O
z

M
t—l

Z
O
M
to
C/1
M
§

a
H

^2
3
H
W












^^
CO
•
0
4-1




























si



















                                    III-132

-------
  technology to the utilization of gasifier off-gas for power generation.



  The supercharged boiler combined cycle plant  (HI) has the  lowest  capital



  costs  per KW and the lowest net  heat  rate which  results  in the lowest



  fuel costs.   The three  other  Lurgi LBG plants  (Al, Bl  and  Cl) have



  nearly identical unit energy  costs approximately  6 percent higher than



  Plant  HI.  Plant Al  for  the Koppers Totzek fuel has unit energy costs



  approximately  16 percent higher  than Plant HI; the cost for this plant



  exceeds that of  the other Lurgi plants due to the higher fuel cost.



  Koppers Totzek plant 15 energy cost is approximately 36 percent  higher



  than Plant HI; the high energy cost of this plant is  a result of  the



 use of oxygen in the boilers.



        Although the supercharged  boiler combined  cycle system has  both



 the lowest capital and energy  costs,  the advantage is  hardly over-



 whelming.   Supercharged  boilers have been used  throughout Europe,



 but their  use in the States has been restricted to marine boilers.


       (3)
 Robson     had earlier considered  the use of supercharged boilers and



 rejected them for the following reasons:



        -   the development costs associated with  their construction



            was  likely to be too  high;



            as  permissible turbine inlet temperatures increase, the



            efficiency advantage of the supercharged boiler system



            disappears.


                                                           (4)
       In a study conducted for the Office of Coal Research    Com-



bustion Engineering compared the cost  of conventional  steam plants








                              111-133

-------
with an atmospheric gasifier and supercharged boiler combined cycle



plant with pressurized gasifiers.  In this study which considered an



integrated gasification and power generation plant, the supercharged



boiler concept had a 3.0 Mills/KW energy cost advantage.



       Even if a supercharged boiler can be considered as existing



technology, there does not appear to be a clear-out financial incen-



tive to recommend its use.  This is compounded by the environmental



considerations presented in the following section clearly indicating



that further detailed study is required before the supercharged boiler



combined cycle system can be considered to be an acceptable method of



power generation from coal gasifier off-gas.



3.0    ENVIRONMENTAL CONSIDERATIONS IN THE COMBUSTION OF

       GASIFIER OFF-GAS



       The systems study discussed in the previous section was based



upon the delivery of a fuel gas almost free from sulfur (i.e., less



than that which would exceed S0_ emission standards) and particulates.



Since the gasification process is not included in the study, the major



area giving rise to environmental concern is the formation of NO



during the combustion of the low Btu gas.  The kinetic mechanisms con-



trolling the rate of NO  formation in combustion processes are outside
                       JV


the scope of this paper.  However, it is known that the amount of



NO  formed from molecular nitrogen is strongly temperature dependent,
  A.


Thus, thermal NO formation is unlikely to be a problem in any of the



systems discussed earlier (perhaps with the exception of system HI)
                             III- 134

-------
provided the fuel gas is burned with air or oxygen and sufficient
quantities of recycled cooled combustion products.  The emission
levels shown in Table 6 were calculated by the NO  Emission Subroutine
                                         J       X
of the Lower Furnace Program   .  For tangentially fired units there
is sufficient experience to believe that these emission levels will be
within acceptable predictive limits.  The Lurgi fired systems are so
below the target emission level that NO control techniques need not
be considered.  This does not apply to the K-T fired boiler (Al), the
higher heating value and higher flame temperatures necessitate some
type of control to meet the 0.1 Ibs NO /10  Btu target.
       Predictions for the supercharged boiler must be considered as
less acceptable.  It should be recognized that these predictions were
made with a model tuned to a particular furnace design.   The prediction
can be considered to be suspect because
       -    the upfired boiler may give a very different heat flux
            distribution from that expected based upon tangential
            firing experience;
            it may no longer be permissible to treat the furnace as
            a series of horizontal strips of uniform properties;
            the lower furnace model does not include the capability
            of accepting pressures above atmospheric.
       One other real concern is the possibility of trace quantities of
ammonia in the fuel gas.  It is known that ammonia is one of the pro-
ducts of most gasification schemes and that it will be oxidized to NO
                               HI- 135

-------
 during the combustion process.  The emission levels indicated above



 did not include the possible conversion of ammonia contained in the


                                          (2)
 fuel gases.  The best available estimates    indicate that after low



 temperature H_S clean-up the fuel gas may contain 500 ppm of ammonia.



 This could produce as much as 200 ppm of NO  in the combustion pro-



 ducts, depending upon the fuel-air mixing process within the combustor.



 Existing processes for high temperature H S removal do not scrub



 ammonia and the high temperature gas may contain as much as 4000 ppm



 of ammonia which would give NO  levels in excess of the N.S.P.S. (New



 Source Performance Standards) unless appropriate precautions were



 taken.



 4.0    CONCLUSIONS



       A systems study has been conducted to assess the most appropriate



use of existing technology to produce electric power from low Btu gas in



 an environmentally acceptable manner.  Fired gas turbines could ulti-



mately lead to the most economically advantageous use of this fuel pro-



vided turbines can be built to operate satisfactorily for prolonged



periods with inlet temperatures of 2600°F.  The present study indicates



 that a supercharged boiler unfired gas turbine system provides the best



method of utilizing existing technology to convert coal-derived fuel gas



to electric power.  The analysis discussed earlier gives slight economic



 advantage for the supercharged boiler.  However, it should be noted that
                               III-136

-------
there may be considerable economics associated with the design and



construction of an integrated plant.  The net power plant heat rate for



this system is significantly lower than those of the other systems



studied, suggesting that a more detailed study of an integrated plant



might uncover a more significant advantage.  There should be little need



for concern regarding NO  production from low Btu gas provided it is



used cold and free from significant quantities of ammonia.  The study



has highlighted the need for an analytical tool capable of predicting



combustor performance.  Work is in progress to assess the NO  emissions
                                                            A.


from a supercharged boiler fired by coal-derived fuel gas which could



contain up to 500 ppm of ammonia, and to identify those design param-



eters which will allow emissions to be controlled to a minimum without



impairing the thermal efficiency of the system.
                             III-137

-------
                             REFERENCES
1.     Martin, G.B., "Environmental Considerations in the Use of
       Alternate Fuels in Stationary Combustion Processes." (Symposium
       Proceedings:  Environmental Aspects of Fuel Conversion
       Technology, EPA-650/2-74-118, pp 259-275; NTIS No. PB 238-
       304/AS.  (October, 1974).)

2.     Robson, F.L., Blecher, W.A.  and Giramonti, A.J.,  "Combined
       Cycle Power Systems", (Presented at the Symposium on Environ-
       mental Aspects of Fuel Conversion Technology II,  Hollywood,
       Florida, December, 1975.

3.     Robson, F.L., and Giramonti, A.J., "The Environmental Impact
       of Coal-Based Advanced Power Systems."  (Symposium Proceedings:
       Environmental Aspects of Fuel Conversion Technology, EPA-650/
       2-74-118, pp 237-257; NTIS No. PB 238-304/AS.  (October,  1974).)

4.     Review and Evaluation of New Plant Gasifier/Power System
       Alternatives, Task 4, Office of Coal Research, Project
       No. 900129, Combustion Engineering.

5.     Beuters, K.A., Cogoli, J.G.  and Habelt, W.W., Fifteenth
       Sumposium (International) on Combustion, p 1245,  The Com-
       bustion Institute (1075).
                              Ill-133

-------
                                    SPEAKERS LIST
(NOTE:  To facilitate their identification, speakers are listed alphabetically
together with the name of the organization they represent.  The complete address
of each organization represented at the conference appears at the end of the list
of attendees.)
LIST OF SPEAKERS

Name

Axworthy, Dr. Arthur E.
Bittner, James D.
Bowen, Dr. Joshua A.
Bowman, Dr. Craig T.
Brown, Richard A.
Burchard, Dr. John K.
Cato, Glenn A.
Collom, Jr., Robert H.
Combs, L. Paul
Crawford, Allen R.
Cuffe, Stanley T.
Dykema, Owen W.
England, Dr. Christopher
Engleman, Dr. Victor S.
Siammar, Robert P.
Sail, Robert E.
Heap, Dr. Michael P.
'lelms, G. Thomas
:tollinden, Dr. Gerald A.
Kason, T.T.
Kendall, Dr. Robert M
Kesselring, Dr. John P.
Ketels, Peter
Lachapelle, David G.
Lanier, W. Steven
McDonald, Henry
Maloney, Dr. Kenneth L.
Manny, Srwin H.
Martin, G. Blair
Muzio, Dr. Lawrence J.
Pohl, John H.
I'oston, H. Wallace
Frinciotta, Frank
J.osenberg, Dr. Robert B.
Sarofim, Dr. Adel F.
Selker, Ambrose P.
Shaw, Dr. Robert
Shoffstall, Dr. Donald R.
Tyson, Dr. Thomas J.
Nasser, John H.
Wendt, Dr. Jost O.L.
Representing

Rockwell International, Rocketdyne Division
Massachusetts Institute of Technology
EPA, IERL, Combustion Research Branch
United Technology Research Center
Acurex, Aerotherm Division
EPA, IERL
KVB, Inc.
State of Georgia, Department of Natural Resources
Rockwell International, Rocketdyne Division
Exxon Research and Engineering
EPA, Office of Air Quality Planning and Standards
The Aerospace Corporation
Jet Propulsion Laboratory
Exxon Research and Engineering
Battelle-Columbus Laboratories
EPA, IERL, Combustion Research Branch
Ultrasysterns
EPA, Region IV, Air and Hazardous Materials Division
Tennessee Valley Authority
City of Chicago, Department of Environmental Control
Acurex, Aerotherm Division
Acurex, Aerotherm Division
Institute of Gas Technology
EPA, IERL, Combustion Research Branch
EPA, IERL, Combustion Research Branch
United Technology Research Center
KVB, Inc.
Exxon Research and Engineering
EPA, IERL, Combustion Research Branch
KVB, Inc.
Massachusetts Institute of Technology
City of Chicago, Department of Environmental Control
EPA, Energy Processes Division
Institute of Gas Technology
Massachusetts Institute of Technology
Combustion Engineering
Stanford Research Institute
Institute of Gas Technology
Ultrasystems
EPA, IERL, Combustion Research Branch
University of Arizona
                                         A-l

-------
                                  PARTICIPANTS LIST
 (NOTE:  To facilitate their identification, participants are listed alphabetically
 together with the name of the  organization they represent.  The complete address
 of each organization represented  at  the  conference appears at the end of the list
 of attendees.)
 LIST OF PARTICIPANTS

 Name

 Alvey,  Courtney D.
 Anderson, Dr. Larry W.
 Axtman,  William H.
 Bagwell, Fred A.
 Baker,  Burke
 Ban,  Stephen D.
 Barrett, Richard E.
 Barsin,  Joseph
 Bartok,  William
 Batra,  Sushil K.
 Bauman,  Robert D.
 Beals, Rixford A.
 Beatty,  James D.
 Bennett, Dr. Robert
 Blandford, Jr., J.B.
 Blythe, R. Allen
 Buechler, Lester
 Bonne, Ulrich
 Booth, Michael R.
 Bowman, Barry R.
 Bueters, K.A.
 Carpenter, Ronald C.
 Cernansky, Dr. Nicholas P.
 Christiano, John P.
 Chu, Richard R.
 Clark, Norman D.
 Cleverdon, R.F.
 Cotton, Ernest
 Creekraore, Andrew T.
 Daughtridge, Jimmy T.
Degler, Gerald H.
 Demetri, E.P.
DeWerth, D.W.
Dingo, T.T.
Donaldson, Thomas M.
Dowling, Daniel J.
Downey, Thomas A.
Dyer, T. Michael
Dygert, J.C.
Representing

Baltimore Gas & Electric
Acurex, Aerotherm Division
American Boiler Manufacturers Association
South California Edison
Shell Development Company
Battelle-Columbus Laboratories
Battelle-Columbus Laboratories
Babcock & Wilcox
Exxon Research & Engineering
New England Electric Systems
EPA, Office of Air Quality, Planning & Standards
NOFI
Procter & Gamble
Apollo Chemical
Englehard Industries
International Boiler Works
Systems Research Labs
Honeywell
Ontario Hydro
Lawrence Livermore Laboratories
Combustion Engineering
Armstrong Cork
Drexel University
EPA, Office of Air Quality, Planning & Standards
EBASCO
C-E Air Preheater
Chevron Research Company
American Petroleum Institute
EPA, Control Programs Development Division
Pratt & Whitney Aircraft
Systems Research Labs
Northern Research and Engineering Corporation
American Gas Association Labs
General Motors
EPA, Office of Air Quality, Planning & Standards
Union Carbide
Gamlen Chemical Company
Sandia Laboratories
Shell Development Company
                                         A-2

-------
                           LIST OF PARTICIPANTS (CONT'D)
Name

Dzuna, Eugene R.
Erskine, George
Feng, C.L.
Fennelly,  Paul F.
Fletcher, James
Fletcher, Roy J.
Freelain, Kenneth
Frisch, Dr. N.W.
Fuhrman, Jr., Theodore C.
Gibbs, Thomas
Goetz, Gary
Goodley, Allan R.
Graham, David J.
Greene, Jack H.
Grimshaw, Vincent C.
Grossman, Ralph
Hangebrauck, Robert P.
Heck, Ronald
Hensel, Thomas E.
Holden, Edward A.
Honea, Dr. Franklin I.
Howard, Jack B.
Hudson, Jr., James L.
Jackson, Dr. A.W.
Jepson, Dr. A.F.
Karas, Dennis T.
Kemmerer, Jeffrey
Khan, M. Ali
Khoo, Dr. S.W.
Kloecker, J.F.
Kykendal, William
Lahre, Thomas
Lange, Dr. Howard
Lavoie, Raymond C.
Lee, James E.
Lenney, Ronald J.
Levy, Arthur
Lewis, F. Michael
Lii, Dr. C.C.
Lin, Donald J.L.
Locklin, David W.
Lord, Harry C.
Loweth, Carl
Marshall, David
Marshall, John H.
Representing

Gulf Research & Development Company
Mitre Corporation
Selas Corporation of America
GCA/Technology Division
Industrial Combustion
Peabody Engineering Corporation
Federal Energy Administration
Research-Cottrell
Erie City Energy Division
EPA, Region IV
Combustion Engineering
California Air Resources Board
EPA, Office of Research and Development
EPA, Administrative Office
Process Combustion
Ralph Grossman, Ltd.
EPA, Energy Assessment and Control Division
Englehard Industries
Turbo Power and Marine Systems
General Foods Corporation
Midwest Research Institute
Massachusetts Institute of Technology
Tampa Electric Company
Ontario Hydro
Environmental Measurements, Inc.
East Chicago Air Quality Control
Fuller Company
East Chicago Air Quality Control
Canadian Gas Research Institute
Erie City Energy Division
EPA, Process Measurement Branch
EPA, Office of Air Quality, Planning & Standards
Babcock & Wilcox
Rohm & Haas Company
Facilities Engineering Command (U.S. Navy)
Ronald J. Lenney Associates
Battelle-Columbus Laboratories
Stanford Research Institute
EPA, Combustion Research Branch
Forney Engineering Company
Battelle-Columbus Laboratories
Environmental Data Corporation
The Trane Company
Babcock & Wilcox
Combustion Engineering
                                          A-3

-------
      Name

      Marton, Miklos B,
      Mayfield, D. Randell
      Meier, John G.
      Moore, Douglas S.
      Moore, Edward E.
      Morton, William J.
      Moscowitz, Charles
      Hosier, Stanley A.
      Newton, Charles L.
      Nurick, W.H.
      Pantzer, Karl
      Pershing, David W.
      Pertel, Dr. Richard
      Renner, Ted
      Riley, Joseph
      Robert, J.
      Roberts, Dr. George
      Robertson, J.F.
      Roffe, Gerald
      Rosen, Meyer
      Ross,  Marvin
      Rulseh, Roy
      Sadowski, R.S.
      Samples, J.R.
      Scott, Donald R.
      Sheffield, E.W.
      Slack, A.V.
      Smith, Lowell L.
      Spadaccini, L.J.
      Sterman, Sam
      Sullivan, Robert E.
      Swearingen, W.E.
      Takacs, Dr. L.
      Taylor, Barry R.
      Utterback, Paul M.
      Van Grouw, Sam J.
      Vatsky, Joel
      Vershaw, Jim
      Watson, Raymond A.
      Webb,  R.
      Weiland, J.H.
      Weinberger, Dr. Lawrence
      White, David J.
      White, James H.
      White, Phil
LIST OF PARTICIPANTS (CONT'D)

        Representing

        IBM
        EPA, Region IV
        International Harvester, Solar Division
        Chevron Research Company
        Eclipse, Inc.
        E. Keeler Company
        Monsanto Research Corporation
        Pratt & Whitney
        Colt Industries
        Rockwell International, Rocketdyne Division
        Babcock & Wilcox
        University of Arizona
        Institute of Gas Technology
        Fuel Merchants Association of New Jersey
        EPA, Region IV
        Canadian Department of Environment
        Englehard Industries
        Crystal Petroleum Company
        General Applied Science Laboratories
        Union Carbide
        Lawrence Livermore Laboratory
        Cleaver-Brooks
        Riley Stoker Corporation
        Union Carbide
        Columbia Gas System Service Corporation
        TRW
        SAS Corporation
        KVB, Inc.
        United Technology Research Center
        Union Carbide
        General Motors
        Koppers Company
        General Motors
        Massachusetts Institute of Technology
        Babcock & Wilcox
        KVB, Inc.
        Foster-Wheeler Energy Corporation
        The Trane Company
        Florida Power & Light Company
        The Trane Company
        Texaco, Inc.
        Mitre Corporation
        International Harvester, Solar Division
        Coen Company
        Ventura Company
L
                                              A-4

-------
Name

Wiedersum, George C.
Wilhelm, Ronald
Wilson, Jr., R.P.
Winters, Harry K.
Wittig, Dr. Sigmar L.K.
Woolfolk, Dr. Robert
Wright, Richard
Young, Dexter E.
Ziarkowski, Stanley
Zielke, Robert L.
Zirkel, Eric C.
LIST OF PARTICIPANTS (CONT'D)

        Representing

        Philadelphia Electric Company
        Aqua-Chera, Inc.
        Arthur D. Little, Inc.
        Ray Burner Company
        Purdue University
        Stanford Research Institute
        Industrial Combustion
        EPA, Control Programs Development Division
        Gamlen Chemical Company
        Tennessee Valley Authority
        Armstrong Cork
                                        A-5

-------
                          LIST OF ORGANIZATIONS REPRESENTED
Kame (Represented By)

Aeurex Corporation, Aerotherm Division
   (Mr. Anderson)

American Boiler Manufacturers Association
   (Mr. Axtman)

American Gas Association Labs
   (Mr. De Werth)

American Petroleum Institute
   (Mr. Cotton)

Apollo Chemical
   (Dr. Bennett)

Aqua-Chem, Inc.
   (Mr. Wilhelm)

Armstrong Cork Company
   (Mr. Carpenter, Mr.  Zirkel)

Babcock & Wilcox
   (Mr. Barsin,  Mr.  Lange,  Mr.  Moore)

   (Mr. Utterback, Mr.  Pantzer,
     Mr. Marshall)

Baltimore Gas  and  Electric Company
    (Mr. Alvey)

Battelle-Columbus  Labs
    (Mr. Ban, Mr. Barrett,
     Mr. Levy,  Mr.  Locklin)

 C-E  Air Preheater
    (Mr. Clark)

 California Air Resources Board
    (Mr.  Goodley)

 Canadian Department of Environment
    (Mr.  Robert)

 Canadian Gas Research Institute
    (Dr.  Khoo)

 Chevron Research Company
    (Mr. Cleverdon, Mr. D. Moore)
Address^

485 Clyde Avenue
Mountain View, California  94042

1500 Wilson Boulevard, Suite 317
Arlington, Virginia  22209

8501 East Pleasant Valley Road
Cleveland, Ohio  44131

1801 K Street, N.W.
Washington, D.C.  20006

35 South Jefferson Road
Whippany, New Jersey  07981

P.O. Box 421
Milwaukee, Wisconsin  53201

Liberty & Charlotte Streets
Lancaster, Pennsylvania   17604

20 South Van  Buren Avenue
Barberton, Ohio   44203

P.O. Box  2423
North  Canton, Ohio  44720

 2012 Gas  and  Electric Building
 Baltimore, Maryland   21203

 505  King  Avenue
 Columbus,  Ohio  43201
 Andover Road
 Wellsville, New York  14895

 1709 llth Street
 Sacramento, California  95814

 351 St. Joseph Boulevard
 Houll, Quebec, Canada

 55 Scarsdale Road, Den Mills
 Ontario, M3B2R3, Canada

 P.O. Box 1627
 Richmond, California  94802
                                          A-6

-------
                              LIST OF ORGANIZATIONS  (CONT'D)
Name  (Represented  By)

Cleaver-Brooks
    (Mr. Rulseh)

Coen Company
    (Mr. J.H. White)

Colt Industries
    (Mr. Newton)

Columbia Gas System Service Corporation
    (Mr. Scott)

Combustion Engineering
    (Mr. Bueters, Mr. Goetz, Mr. Marshall)

Crystal Petroleum Company
    (Mr. Robertson)

Drexel University
    (Dr. Cemansky)

EBASCO
    (Mr. Chu)

East Chicago Air Quality Control
    (Mr. Karas, Mr. Khan)

Eclipse, Inc.
    (Mr. E.  Moore)

Englehard Industries
    (Mr. Blandford, Dr. Heck, Dr. Roberts)

Environmental Data Corporation
    (Mr. Lord)

Environmental Measurements, Inc.
    (Dr. Jepsen)

Environmental Protection Agency
EPA - Administrative Office
    (Mr. Greene)
 Address

 3707 North Richards Street
 Milwaukee, Wisconsin  53201

 1510 Rollins  Road
 Burlingame, California  94010

 701  Lawton Avenue
 Beloit, Wisconsin   53511

 1600 Dublin Road
 Columbus,  Ohio  43215

 1000 Prospect Hill  Road
 Windsor, Connecticut   06095

 P.O.  Box 4180
 Corpus  Christi, Texas  78408

 Philadelphia
 Pennsylvania  19104

 145  Technology Park
 Norcross,  Georgia  30071

 900  East Chicago Avenue
 East  Chicago, Indiana  45312

 1100  Buchanan
 Rockford,  Illinois  61101

Middlesex Turnpike, Wood Avenue
Edison, New Jersey  08876

 608 Fig Avenue
Monrovia, California  91016

2 Lincoln Court
Annapolis,  Maryland  21401
Research Triangle Park
North Carolina  27711
                                          A-7

-------
                             LIST OF ORGANIZATIONS (CONT'D)
Same (Represented By)

EPA - Combustion Research Branch
   (Dr. Bowen, Mr. Hall, Mr. Lachapelle,
    Mr. Lanier, Dr. Lii, Mr. Martin, Mr. Wasser)

EPA - Control Programs Development Division
   (Mr. Creekmore, Mr. Young)

EPA - Energy Assessment & Control Division
   (Mr. Hangebrauck)

EPA - Office of Air Quality, Planning,
      and Standards
   (Mr. Bauman, Mr. Christiano,
    Mr. Donaldson, Mr. Lahre)

EPA - Office of Research and Development
   (Mr. Graham)

EPA - Process Measurement Branch
   (Mr. Kuykendal)

EPA - Region IV
   (Mr. Biggs, Mr. Mayfield, Mr. Riley)

Erie City Energy Division
   (Mr. Fuhrman, Mr. Kloecher)

Exxon Research & Engineering Company
   (Mr. Bartok)

Federal Energy Administration
   (Mr. Freelain)

Florida Power & Light Company
   (Mr. Watson)

Forney Engineering Company
   (Mr. Lin)

Foster Wheeler Energy Corporation
   (Mr. Vatsky)

Fuel Merchants Association of New Jersey
   (Mr. Renner)
Address

Research Triangle Park
North Carolina  27711
Research Triangle Park
North Carolina  27711

Research Triangle Park
North Carolina  27711
Research Triangle Park
North Carolina  27711
Washington, D.C.  20460
Research Triangle Park
North Carolina  27711

1421 Peachtree Street, N.E.
Atlanta, Georgia  30309

1422 East Avenue
Erie, Pennsylvania  16502

P.O. Box 8
Linden, New Jersey  07036

1200 Pennsylvania  Avenue, N.W.
Washington, D.C.  20461

P.O. Box 013100
Miami, Florida  33101

P.O. Box 189
Addison, Texas  75001

10 South Orange Avenue
Livingston, New Jersey  07039

66 Morris Avenue
Springfield, New Jersey  07081
                                         A-8

-------
                             LIST OF ORGANIZATIONS (CONT'P)
Name (Represented By)

Fuller Company
   (Mr. Kemmerer)

GCA/Technology Division
   (Dr. Fennelly)

Gamlen Chemical Company
   (Mr. Downey, Mr. Ziarkowski)

General Applied Science Laboratories
   (Mr. Roffe)

General Foods - Technical Center
   (Mr. Holden)

General Motors Corporation
   (Mr. Sullivan)

   (Mr. Dingo, Mr. Takacs)
Gulf Research and Development Company
   (Mr. Dzuna)

Honeywell, Inc.
   (Mr. Bonne)

IBM
   (Mr. Marton)

Industrial Combustion
   (Mr. Wright, Mr. Fletcher)

Institute of Gas Technology
   (Dr. Pertel)

International Boiler Works
   (Mr. Blythe)

International Harvester, Solar Division
   (Mr. Meier, Mr. D.G. White)
Address

124 Bridge Street
Catasauqua, Pennsylvania  18032

Burlington Road
Bedford, Massachusetts  01730

299 Market Street
Saddle Brook, New Jersey  07662

Merrick & Stewart Avenues
Westbury, New York  11790

250 North Street
White Plains, New York  10625

5735 West 25th Street
Indianapolis, Indiana  46224

Technical Center
Warren, Michigan  48090

P.O. Box 2038
Pittsburgh, Pennsylvania  15230

Bloomington
Minnesota  55420

1000 Westchester Avenue
White Plains, New York  10604

4465 North Oakland
Milwaukee, Wisconsin  53211

3424 South State Street
Chicago, Illinois  60616

P.O. Box 498
East Stroudsburg, Pennsylvania  18301

2200 Pacific Highway
San Diego, California  92119
                                         A-9

-------
                             LIST  OF ORGANIZATIONS  (CONT'D)
 Name (Represented By)

 KVB Engineering,  Inc.
    (Mr.  L.  Smith,  Mr. Van  Grouw)

 E.  Keeler Company
    (Mr.  Morton)

 Koppers  Company,  Inc.
    (Mr.  Swearingen)

 Ronald J. Lenney  Associates
    (Mr.  Lenney)

 Arthur D. Little,  Inc.
    (Mr.  R.D. Wilson)

 Lawrence Livermore Laboratories
    (Mr.  Bowman, Mr. Ross)

 Massachusetts Institute of Technology
    (Mr.  Howard, Mr. Taylor)

 Midwest  Research Institute
    (Dr.  Honea)

 Mitre Corporation
    (Dr.  Weinberger, Mr. Erskine)
Monsanto Research Corporation
   (Mr. Moscowitz)

NOFI
   (Mr. Seals)

Naval Facilities Engineering Command,
Southern Division
   (Mr. J. Lee)

New England Electric Systems
   (Mr. Batra)

Northern Research & Engineering Corporation
   (Mr. Demetri)
Address

6624 Hornwood Drive
Houston, Texas  77036

Williamsport
Pennsylvania  17701

Koppers Building
Pittsburgh, Pennsylvania  15219

2001 Palmer Avenue
Larchmont, New York  10538

Acorn Park
Cambridge, Massachusetts  02140

P.O. Box 808
Livermore, California  94550

Massachusetts Avenue
Cambridge, Massachusetts  02139

425 Volker Boulevard
Kansas City, Missouri  64110

Westgate Research Park
1820 Dolly Madison Boulevard
McLean, Virginia  22101

Station B. Box 8
Dayton, Ohio  45407

New York, New York
2144 Melbourne Street
P.O. Box 10068
Charleston, South Carolina  29411

20 Turnpike Road
Weston, Massachusetts  01581

219 Vassar Street
Cambridge, Massachusetts  02139
                                           A-10

-------
                             LIST OF ORGANIZATIONS  (COKT'D)
Naroe  (Represented By)

Ontario Hydro Corporation
    (Mr. Booth, Dr. Jackson)

Peabody Engineering Corporation
    (Mr. R. Fletcher)

Philadelphia Electric Company
    (Mr. Wiedersom)

Pratt & Whitney Aircraft
    (Mr. Daughtridge, Mr. Mosier)

Process Combustion Corporation
    (Mr. Grimshaw)

Procter & Gamble Company
    (Mr. Beatty)

Purdue University
    (Mr. Wittig)

Ralph Grossman, Ltd.
    (Mr. Grossman)

Ray Burner Company
    (Mr. Winters)

Research-Cottrell, Inc.
    (Dr. Frisch)

Riley Stoker Corporation
    (Mr. Sadowski)

Rockwell International Corporation,
Rocketdyne Division
    (Mr. Nurick)

Rohm & Haas Company
    (Mr. Lavoie)

SAS Corporation
    (Mr. Slack)
Address

620 Union Avenue
Toronto, Ontario, Canada  M561X6

835 Hope Street
Stamford, Connecticut  06907

2301 Market Street, S10-1
Philadelphia, Pennsylvania  19101

P.O. Box 2691
West Palm Beach, Florida  33402

1675 Washington Road
Pittsburgh, Pennsylvania  15228

610 South Center Hill Road
Cincinnati, Ohio  45224

West Lafayette
Indiana  47907

P.O. Box 70, Town of Mt. Royal
Montreal, Canada  H3P 3B8

1301 San Jose Avenue
San Francisco, California  94112

P.O. Box 750
Boundbrook, New Jersey  08805

9 Neponset Street
Worcester, Massachusetts  01613

6633 Canoga Avenue
Canoga Park, California  91304
P.O. Box 584
Bristol, Pennsylvania  19007

RFD #1
Sheffield, Alabama  35660
                                           A-ll

-------
                             LIST OF ORGANIZATIONS (CONT'D)
Name  (Represented By)

Sandia Laboratories
   (Mr. Dyer)

Selas Corporation of America
   (Mr. Feng)

Shell Development Company
   (Mr. Dygert, Mr. Baker)

South California Edison
   (Mr. Bagwell) '

Stanford Research Institute
   (Dr. Woolfolk, Mr. Lewis)

Systems Research Labs
   (Mr. Buechler, Mr. Degler)

TRW,  Inc.
   (Mr. Sheffield)

Tampa Electric Company
   (Mr. Hudson)

Tennessee Valley Authority
   (Mr. Zielke)

Texaco, Inc.
   (Mr. Weiland)

The Trane Company
   (Mr. Loweth, Mr. Vershaw, Mr. Webb)

Turbo Power and Marine Systems
   (Mr. Hensel)

Union Carbide Corporation
   (Mr. Rosen, Mr. Sterman)

   (Mr. Dowling)
Address

Livermore
California  94550

Dresher
Pennsylvania  19025

P.O. Box 481
Houston, Texas  77001

P.O. Box 800
Rosemead, California  91770

1611 North Kent Street
Arlington, Virginia  22209

2800 Indian Ripple Road
Dayton, Ohio  45440

1 Space Park - R4/2020
Redondo Beach, California  90278

P.O. Box 111
Tampa, Florida  33601

524 Power Building
Chattanooga, Tennessee  37401

P.O. Box 509
Beacon, New York  12508

3600 Pammel Creek Road
La Crosse, Wisconsin  54601

1690 New Britain Avenue
Farmington, Connecticut  06032

Tarrytown Technical Center
Tarrytown, New York  10591

Box 180
Sistersville, West Virginia  26175
                                        A-12

-------
                              LIST OF ORGANIZATIONS  (CONT'D)
Name (Represented  By)

Union Carbide Corporation
   (Mr. Samples)

United Technologies  Research Center
   (Mr. Spadaccini)

University of Arizona
   (Mr. Pershing)

Ventura County A.P.C.D.
   (Mr. P. White)
Address

Box 4361
South Charleston,  West Virginia  25353

400 Main Street
East Hartford, Connecticut  06040

Tucson
Arizona  85721

740 East Main Street
Ventura, California  93001
  *u.-i. coi'rn,NNE?fr HUNTING OFFICE:
                              Region ».
                                          A-13

-------
                                TECHNICAL REPORT DATA
                          (flcast trad liiitructiuns on the rcicrse before completing)
 1. REPORT NO.
 EPA-600/2-76-152b
                           2.
                                                      I. RECIPIENT'S ACCESSIOr+NO.
 Combustion Symposium; Volume n--Fuels and Process
 Research and Development
                                                      5. REPORT DATE
                                                       June 1976
                                                     6. PERFORMING ORGANtZATtON CODE
 7. AUTHOR(S)

 Miscellaneous
                                                      8. PERFORMING ORGANIZATION REPORT NO
 9. PERFORMING OR9ANIZATION NAME AND ADDRESS
 NA
                                                     10. PROGRAM ELEMENT NO.
                                                     1AB014; ROAP 21BCC
                                                     11. CONTRACT/GRANT NO.

                                                     NA (In-house)
 12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC 27711
                                                     13. TYPE OF REPORT AND PERIOD COVERED
                                                     Proceedings; 9/24-26/75
                                                     14. SPONSORING AGENCY CODE
                                                     EPA-ORD
5 SUPPLEMENTARY
 Mail Drop 65, Ext. 2470/2477.
                              Chairman J.S. Bowen, Vice-Chairman R.E. Hall,
6. ABSTRACT
          The proceedings document the 37 presentations made during the Stationary
 Source Combustion Symposium held in Atlanta,  Ga. , September 24-26, 1975.  Spon-
 sored by the Combustion Research Branch of EPA's Industrial Environmental Resea-
 rch Labor atory--RTP, the symposium dealt with subjects related both to developing
 improved combustion technology for the  reduction of air pollutant emissions from
 stationary sources,  and to improving equipment efficiency.   The symposium was
 divided into four parts and the proceedings were issued in three volumes: Volume I--
 Fundamental Research,  Volume  n — Fuels and Process Research and Development,
 and Volume IE—Field Testing and Surveys.  The symposium was  intended to provide
  ontractor, industrial, and Government  representatives with the latest information
 on EPA in-house and contract combustion research projects related to pollution
 control, with emphasis on reducing nitrogen oxides.
                             KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
                                         b.lDENTIFIERS/OPEN ENDED TERMS C.  COSATi Field/Group
Air Pollution, Combustion, Field Tests
Combustion Control, Coal, Oils
Natural Gas, Nitrogen Oxides, Carbon
  arbon Monoxide, Hydrocarbons, Boilers
 Pulverized Fuels, Fossil Fuels, Utilities
Gas Turbines,  Efficiency
                                        Air Pollution Control
                                        Stationary Sources
                                        Combustion Modification
                                        Unburned Hydrocarbons
                                        Fundamental Research
                                        Fuel Nitrogen
                                        Burner Tests
13B
21B  14B
2 ID  11H
07B
07C  13A
13G   14A
 s, DISTRIBUTION STATEMENT

 Unlimited
                                        19. SECURITY CLASS (ThisReport)
                                        Unclassified
21. NO. OF PAGES
   430
                                        20. SECURITY CLASS (Thispage)
                                        Unclassified
                                                                22. PRICE
EPA Fotm 2220-1 (9-73)
                                     A-14

-------
2P 600/2  ^?A
 5-152b   In^ust rial
'nv .
                            S  Lnb.
AUTHOR

    Proceeding^ oF the stationary
TlTLEsource combustion pvmnosium.
    V.2: Fuel"5 :\ uroct-ss research
DATE
OAYLOUD 41

-------

-------
             DATE DUE
i	

-------

-------