EPA-600/2-77-008a
January 1977
Environmental Protection Technology Series
              COMBUSTION ADDITIVES  FOR  POLLUTION
                  CONTROL-  A  State-of-the-Art Review
                                     Industrial Environmental Research Laboratory
                                           Office of Research and Development
                                          U.S. Environmental Protection Agency
                                    Research Triangle Park, North Carolina  27711

-------
                RESEARCH REPORTING SERIES

 Research reports of the Office of Research and Development, U.S. Environmental
 Protection Agency, have been grouped into five  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:

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

 This report has been  assigned  to the ENVIRONMENTAL PROTECTION
 TECHNOLOGY series. This series describes research performed to develop and
 demonstrate instrumentation,  equipment, and methodology to repair or prevent
 environmental degradation from point and non-point sources of pollution. This
 work provides the new  or improved technology required for the control and
 treatment of pollution sources to meet environmental quality standards.


                     EPA REVIEW NOTICE

 This report has been reviewed by  the U.S.  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 through the National Technical Informa-
tion Service. Springfield, Virginia 22161.

-------
                                   EPA-600/2-77-008a
                                   January 1977
       COMBUSTION ADDITIVES

    FOR POLLUTION CONTROL--

   A  STATE-OF-THE-ART REVIEW
                     by

       H.H. Krause, L.J. Hillenbrand,
       A.E. Weller, and D.W. Locklin

       Battelle-Columbus Laboratories
              505 King Avenue
           Columbus, Ohio  43201
           Contract No.  68-02-0262
            ROAPNo. 21ADG-020
         Program Element No. 1AB014


    EPA Project Officer:  W. Steven Lanier

 Industrial Environmental Research Laboratory
   Office of Energy, Minerals, and Industry
      Research Triangle Park, NC 27711


                Prepared for

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

-------
                          ABSTRACT
            Fuel additives have been used, or proposed for use,  to
serve a variety of functions in combustion systems.  This report
covers a state-of-the-art review of combustion-type fuel additives
as  to their potential in reducing air-pollutant emissions for oil
and coal firing.  The report contains two complementary parts:
(1) a review of combustion mechanisms as they relate to additive
action in controlling emissions, and (2) a review of experimental
investigations of combustion-type fuel additives.

            The review of technical literature revealed relatively
limited quantitative data from experimental investigations on
combustion additives in which conditions are well defined.  However,
there is evidence for some measure of control of emissions by fuel
additives.

            The evidence for control by fuel additives of visible
smoke and carbon particulate is relatively strong, and the evidence
for control of polycyclic organic matter (POM)  is somewhat weaker.
The evidence for control of NOX is quite weak.   Significant control
of S02 or total sulfur emissions by fuel additives does not appear
to be possible, although emissions of 503 can be reduced.   Possibili-
ties for the control of emissions of ash, or specific  ash constituents
by fuel additives,  is restricted to enhancing the collectability of
ash particles.   Little experimental evidence is  available for control
of hydrocarbons or  CO emissions by fuel  additives, although theoreti-
cal considerations  suggest  that some degree of  control might  be
possible.   Practical considerations and  other possible limitations to
the use of  additives are also reviewed.
                             iii

-------
                       ACKNOWLEDGEMENT

            The authors wish to acknowledge helpful comments by
W. S. Lanier and G. B. Martin, the EPA project officers assigned
during the course of this study, and by other staff of the EPA
Combustion Research Section — including J. S. Bowen, E. E. Berkau,
D. W. Pershing, and J. H. Wasser.  In addition, acknowledgement is
made to Battelle-Columbus staff members: R. E. Barrett, R. B. Engdahl,
R. D. Giammar, H. R. Hazard, A. Levy, and Consultant W. T. Reid for
contributions to various aspects of the study.
                              iv

-------
                            TABLE OF CONTENTS


                                                                        Page

                                 SUMMARY


INTRODUCTION 	   S-I

      Background of EPA Investigations 	   S-2

      Classification of Pollutants and Fuel Additives  	   S-2
            Pollutant Classes  	   S-2
            Additive Classes 	   S-3

      Scope of Report	   S-5

OVERVIEW AND CONCLUSIONS 	   S-7

      1. Additive Effects on Products of Incomplete Combustion ....   S-7
            Effects on Combustible Particulates  	   S-8
            Effects on Polycyclic Organic Matter (POM) 	   S-9
            Effects on Carbon Monoxide and Hydrocarbons  	  S-10

      2. Additive Effects on Emissions Resulting from Fuel Impurities.  S-10

            Effects on Sulfur Compounds  	  S-10
            Effects on Ash	S-ll

      3. Additive Effects on Nitrogen Oxides 	  S-ll

            Effects on Thermal-Fixation Processes  	  S-12
            Effects on Fuel-Nitrogen Conversion  	  S-12

      4. Additive Effects on Boiler Efficiency 	  S-12

      5. Possible Limitations to the Use of Additives	S-13
            Overall Emission Impact  	  S-13
            Toxic Effects	S-14
            The Effect of Equipment Characteristics
                on Fuel Additive Effectiveness 	  S-14
            Other Practical Considerations 	  S-16
            Alternative Controls 	  S-16

RECOMMENDATIONS FOR FURTHER R&D  	  S-17

-------
                       TABLE OF CONTENTS (Continued)
                                                                        Page
                                 PART I
                 COMBUSTION MECHANISMS AS THEY RELATE TO
              FUEL-ADDITIVE ACTION IN CONTROLLING EMISSIONS
POLLUTANTS CATEGORIZED BY ORIGIN 	   1-1

MECHANISMS OF POLLUTANT FORMATION IN COMBUSTION PROCESSES  	   1-4

      Products of Incomplete Combustion  	   1-5

            Carbon Monoxide  	   1-5
            Hydrocarbons 	   1-5
            Polycyclic Organic Matter  	   1-7
            Soot	   1-8

      Emissions Resulting from Fuel Impurities 	  1-13

            Sulfur Compounds 	  1-13
            Fly Ash	1-15

      Formation of Nitrogen Oxides 	  1-16

            Thermally Fixed Nitrogen 	  1-16
            Chemically Bound Fuel Nitrogen 	  1-18

POSSIBLE OPPORTUNITIES FOR ADDITIVE ACTION IN CONTROLLING
   COMBUSTION-GENERATED EMISSIONS BY FUEL ADDITIVES  	  1-19

      Control of Emission from Incomplete Combustion 	  1-20

            1. Approach of Increasing the Active Oxidant
               Concentration 	  1-21
            2. Approach of Adding an Active Oxidant Species  	  1-22
            3. Approach of Inhibiting Formation of Unreactive
               Combustible Material  	  1-24

      Control of Emissions from Fuel Impurities	1-26

      Control of Nitrogen Oxides 	  1-28

            1. Promotion of Atom Recombination	1-29
            2. Retardation of 0-Atom Production  	  1-29
            3. Promotion of Pyrolysis Reactions  	  1-29
            4. Reduction of NO	1-30
                                    vi

-------
                      TABLE OF CONTENTS  (Continued)
                                                                        Page
                                 PART II
                  REVIEW OF EXPERIMENTAL INVESTIGATIONS
                    OF COMBUSTION-TYPE FUEL ADDITIVES
SCOPE OF REVIEW	   H-l

            Summary Tables  	   II-2
            Organization of Part II Review Discussion 	   II-7

ADDITIVE EFFECTS ON PARTICUATE EMISSIONS AND SMOKE  	   II-7

      Combustion Additives in Fuel-Oil  	   II-7

            Investigations Firing Residual Oil in Boilers 	   II-8
            Investigations Firing Distillate Oil in Furnaces,
               Boilers, and Gas Turbines	   II-9
            Laboratory Scale Investigations Firing Various
               Fuels in Continuous Combustion Apparatus 	  11-11
            Investigations Firing Distillate Oil in I.C. Engines  .  .  II-13
            Additional Investigations 	  11-14

            Survey Conclusions Related to Use of Combustion
               Additives for Reducing Particulate or Smoke
               Emissions from Fuel-Oil Firing 	  11-15

      Combustion Additives in Coal Firing 	  11-15
            Early Investigations on Coal Additives by the
               U.S.  Bureau of Mines	11-16
            Other Investigations on Coal Additives	11-17

            Survey Conclusions Related to Use of Combustion
               Additives for Reducing Particuate Emissions
               From Coal Firing	11-18

ADDITIVE EFFECTS ON POLYCYCLIC ORGANIC MATTER 	  11-19

            Laboratory Investigations in Continuous
               Combustion Apparatus 	  11-20
            Investigations in I.C.  Engines	11-21
            Survey Conclusions Related to Use of Combustion
               Additives for Reducing Emissions of POM	11-23

ADDITIVE EFFECTS ON SULFUR OXIDES 	  11-24

      Combustion Additives in Fuel-Oil Firing 	  11-25
            EPA Investigation of Residual Oil Firing  	  11-25
                                    vii

-------
                      TABLE OF CONTENTS (Continued)
            Other Investigation with Residual Oil
               Firing in Experimental Apparatus  	   11-26
            Full-Scale Investigations with Residual
               Oil Firing in Utility Boilers 	   11-28

      Combustion Additives in Coal Firing  	   11-30

            Survey Conclusions Related to the Use of Combustion
               Additives for Reducing Emission of Sulfur Oxides  .  .   11-32

ADDITIVE EFFECTS ON NITROGEN OXIDES  	   11-32

            Investigations Firing Residual Oil in Boilers  	   11-33
            Investigations Firing Distillate Oil in
               Laboratory Furnaces 	   11-33
            Investigation Directed to Gas Turbines 	   11-34
            Investigation Directed to Diesel Engines 	   11-36
            Survey Conclusions Related to Use of Combustion
               Additives for Reducing NOX	11-36
                                    viii

-------
              COMBUSTION ADDITIVES  FOR POLLUTION  CONTROL
                       A STATE-OF-THE-ART REVIEW
                                   by
                   H.  H.  Krause,  L.  J.  Hillenbrand,
                   A.  E.  Weller,  and D.  W.  Locklin
                                SUMMARY
                              INTRODUCTION

          The use of fuel additives for the purpose of gaining some per-
formance advantage in fossil-fuel fired boilers has been of recurring
interest for many years.   Trials with various fuel additives, usually
used in small amounts, have had different objectives, depending on the
type of fuel involved and the era in which they were conducted.  Earlier
additive trials and laboratory investigations were aimed at reducing the
build-up of deposits on heat-transfer surfaces and minimizing the corro-
sion that resulted from the accumulation of the deposits and from their
interaction with the flue gases.  More recently, some of the emphasis
has shifted to the possible use of fuel additives as a means of control-
ling air-pollutant emissions — mainly visible smoke, particulate, sulfur
compounds, and nitrogen oxides.

          While qualitative observations and commercial claims for the
effectiveness of fuel additives in reducing emissions are numerous,
relatively few quantitative data are available from experimental investi-
gations in which conditions are well defined and in which emissions are
measured.  It is the purpose of this report to review the state-of-the-
art of fuel additives for oil and coal firing, based on published
technical literature, and to assess  (as current understanding permits)
                                  S-l

-------
the theoretical basis for additive action in reducing the formation or
emission of pollutants.

BACKGROUND OF EPA INVESTIGATIONS

          The Environmental Protection Agency (EPA) has completed two of
the more comprehensive experimental programs that have contributed quanti-
tative information on the effects of fuel-oil additives in reducing pollu-
                                (D*
tant emissions.  One EPA program     covered screening tests of numerous
additives in which distillate heating oil was fired in a residential oil
burner.  In general, the additives were found to have minor effect com-
pared to the opportunity for emission control by burner adjustments.
                      (2)
The second EPA program    was an investigation of the effectiveness of
several proprietary fuel-oil additives in reducing sulfur oxide emissions
when firing residual oil in a commercial boiler.  No effect was found for
sulfur oxides or the other gaseous pollutants.

          A third and current program, of which this review is an initial
part, was conducted at Battelle-Columbus Laboratories under contract with
EPA to investigate the effects of combustion-type fuel-oil additives on
                                                    (3)
particulate emissions and on particulate composition   .  Additives for
experimental evaluation under this program were selected on the basis
of this state-of-the-art survey and analysis.
CLASSIFICATION OF POLLUTANTS AND FUEL ADDITIVES

Pollutant Classes

          In considering the effects of additives on pollutant emissions,
it is useful to classify the pollutants as follows:
*  References are listed at the end of this report.
                                  S-2

-------
             •  Products of "incomplete combustion
                (Fuel-derived species capable of further
                reaction with oxygen, such as CO, hydrocarbons,
                polycyclic organics, "carbon" particulates,  or
                soot).*
             •  Nitrogen oxides from thermal fixation of
                nitrogen.
             •  Nitrogen oxides formed from chemically bound
                nitrogen in the fuel.
             •  Elements present in the fuel which are considered
                pollutants regardless of the form in which they
                are emitted.   (Heavy metals, chlorine, sulfur).
             •  Ash components other than those included above
                (Fly ash).
The first three of these pollutant classes are the result of some chemical
reaction (or lack of some reaction) of the fuel, air, or air and fuel.
Thus, one might hope to find an additive which would influence the chemi-
cal reactions in such a way as to minimize the creation of,  or maximize
the destruction of, these pollutants.  Such an influence is  not possible
for the last two pollutant classes which are considered pollutants regard-
less of their chemical form.   At best, one could only hope to retain the
pollutant within the combustion system for later removal or  promote its
removal and retention by some downstream control device.  In assessing the
use of additives to control pollutants, it is important to keep in mind
the importance of not replacing one problem with a new one,  such as unde-
sirable emissions resulting from the additive constituents.

Additive Classes

          Fuel additives have been used as a means of accomplishing a
number of functions.  These functions can be distinguished by whether
they occur before, during, or after the combustion process,  and the addi-
tives classified accordingly.  Table 1 outlines specific functions of
additives in each of these three classes:
*  Emissions of polycyclic organic matter (POM) are of concern because of
   the carcinogenic nature of some compounds in this class(4).

-------
                  Class I.     Fuel-Handling Additives
                  Class II.    Combustion  Additives
                  Class III.   Post-Flame  Treatment Additives.
           Class I  additives may  in some circumstances  contribute  to the

control of  pollutant emission  through  their influence  on the physical

properties  of the  fuel or by stabilizing the  condition of  important fuel-

handling elements; for example,  maintaining spray nozzle cleanliness.

However, such effects are incidental and can  be expected to  be small in

the  presence of proper fuel handling and equipment maintenance.



           Class II additives are specifically intended to  affect  the com-

bustion process and, hence, may  have a direct influence on the pollutants

classed as  products of incomplete combustion.  These additives are often

termed "combustion improvers".



            Class  III additives  that operate  in the  post flame region  include

various scavenging agents and  pollutant property modifiers,  as well  as agents

which modify the  catalytic properties  of exposed heat exchanger  and  breeching

surfaces.


              TABLE 1.  CLASSES OF ADDITIVES BY FUNCTION IN OIL AND COAL COMBUSTION  SYSTEMS
        CLASS I.
   FUEL-HANDLING ADDITIVES

   For Improved Storage
       and Handling
         CLASS II.
     COMBUSTIOH ADDITIVES

  For Improved Combustion and
      Pollutant Reduction
         CLASS III.
 POST-FLAME TREATMENT ADDITIVES


   For Post-Flame Treatment
 Fuel stability additives
   - Sludge and gum inhibitors
   - Detergents
   - Metal deactivalors
   - Color stabilizers

 Flow improvers
   - Pour-point depressants

 Demulsifying agents

 Anti-static additives

 Anti-icing compounds

 Corrosion inhibitors
   (for tank protection)
Combustion improvers

  - To reduce smoke or
   particulates

  - To reduce CO, hydrocarbons,
   or polycyclic organic
   matter

Additives to alter particulate
  size or character

Additives to reduce formation of
  pollutant gases

  - Nitrogen oxides, NO

  - Sulfur trioxide, SO,
Soot removers (from heat-transfer
  surfaces)

Additives to control fireside
  corrosion or slay deposits

Additives to enhance particulate
  collection in electrostatic
  precipitators
S0x scavengers
                                       S-4

-------
SCOPE OF REPORT

          The state-of-the-art study covered in this report was directed
toward additives of Class II, plus those of Class III that are added with
the fuel or are added separately to the combustion chamber and that may
function to control pollutant emmissions from oil-fired and coal-fired
boilers.

          No attempt is made to cover the incidental effects of Class I
additives, as such effects (when they exist) are almost totally dependent
on equipment design and condition.  Also, no attempt is made to review
the literature concerning additions of water in sprays or in emulsified
fuels, or addition of limestone, as in the dry limestone SO- control pro-
cess.  With these exceptions, consideration includes all materials (other
than substitute fuels) which are added to a fuel or to the combustion
chamber with the intended result of reducing pollutant emissions by any
mechanism.

          In addition to information on additives in fuel oil and coal
combustion, data relative to other fuels are included where additive
performance has been recorded.  The main emphasis is on fuel additives
applicable to firing boilers; however, some data on additive effects on
emissions from gas turbines and reciprocating internal-combustion engines
are included where these appear possibly pertinent to combustion in
boilers.

        This state-of-the-art review is based on information available
in published technical literature.*  The report is divided into two
complementary parts:
 *   Results  from  the  current experimental  investigation, conducted as
    part  of  this  contract, are  to be  covered in a separate report.(3)
    Also  included in  the appendix to  that  report is a review of Class  I
    fuel-handling additives.
                                  S-5

-------
Part I.   Analysis of Combustion Mechanisms.

          This part comprises a review of combustion
          mechanisms as they relate to the chemical
          basis for emission control by combustion
          additives.
Part II.  Review of Experimental Additive Investigations.

          This part reviews specific investigations reported
          in the literature with respect to the effect of
          fuel additives on emissions of one or more of the
          following pollutants:

              •  Particulate and smoke

              •  Polycyclic organic matter (POM)

                       of interest because of the
                       carcinogenic nature of some
                       of these materials

              •  Sulfur oxides
                       including the corrosive
                       effects of

                 Nitrogen oxides.
effects of S0_ emissions
          In part II, 43 literature references are cited, in
          which 58 different additive compounds have been
          identified and in which some quantitative data
          have been reported as to their effects on pullutants.
                        S-6

-------
                        OVERVIEW AND CONCLUSIONS

          The conclusions reached by this overall review are discussed
below by the category of pollutants, followed by general comments on  the
use of fuel additives; then, recommendations are presented relating to
further R&D needed to fill gaps in information.  Although a number of
possibilities for advancing the state-of-the-art are suggested, with  few
exceptions, the probability of a successful and practically useful out-
come for emission control is relatively low, compared to opportunities
for improvements by good combustion design and practice.

          The following overview summarizes principal conclusions on
combustion-type fuel additive effects as related to:
             1.  Products of incomplete combustion.
             2.  Emissions resulting from fuel impurities.
             3.  Emissions of nitrogen oxides.
             4.  Boiler efficiency.
Also, for sake of perspective, possible limitations to the use of fuel
additives are discussed in a fifth section.

1.  ADDITIVE EFFECTS ON PRODUCTS OF INCOMPLETE COMBUSTION

          Products of incomplete combustion include CO,  hydrocarbons,  POM,
smoke or soot,  and coke particles.   This class of pollutants offers perhaps
the most fertile field for additive application.   Recorded research is
limited to "particulates" (i.e., smoke or soot and coke  particles,  and
POM), although one might expect additives affecting any  of the products
of incomplete combustion would have detectable effects on the others.
Obviously,  the ash portion of the particulate emission (or fly ash)  can
not be reduced by an additive; however, the combustible  or carbon portion
can be reduced.
                                  S-7

-------
Effects on Combustible Particulates

          The greatest effort in fuel additive studies has been devoted
to particulate and smoke emissions.  For minimizing particulate forma-
tion in fuel-oil combustion processes, experimental evidence shows that
the most effective agents are compounds of some transition metals
(manganese, iron, nickel, and cobalt) and some alkaline-earth metals
(barium and calcium).  Organometallic derivatives of these metals have
given the best results.  The organic portion of the molecule provides
the needed solubility in the oil, while also influencing the stability
and volatility of the additive molecule.

          It is generally proposed that the transition metals catalyze
the oxidation of soot in the hot combustion gasses.  Of the alkaline-
earth metals, barium has been suggested to act as a catalyst for the
decomposition of hydrogen and water, promoting the destruction of soot
by the free radicals thus generated.

          No research clearly pertaining  to coke-particle emissions was
found.  While it may reasonably be assumed that the same type of additives
effective in reducing smoke  or soot  emissions will be effective in reduc-
ing coke particle emissions, the optimum  additive volatility or thermal
stability may differ.  A possibility exists for the application of
selected cracking catalysts  to inhibit  coke formation from  heavy oils.

          Finally,  questions concerning electrical effects  in smoke  forma-
tion  and their  relation  to  the effectiveness  of some  types  of additives
have  not been resolved.  Systematic  research  with  additives selected  or
designed  to  produce specific flame ionization levels  would  be needed  to
explore  this  point.
                                   S-8

-------
Effects on Polycyclic Organic Matter (POM)

          Polycyclic organic matter (POM)  comprises a class of poten-
tially hazardous pollutants generally resulting from incomplete combus-
tion (^.   This class includes (1) polynuclear aromatic hydrocarbons,
commonly identified in the literature as PNA or PAH, and (2) nitrogen-
containing heterocyclic compounds.

          Only a few investigations have reported the effects of additives
on POM emissions, and much of this research has been directed toward
diesel engines rather than boiler applications.  Thus, the reported
effects of such additives as nitropariffins, nitrate esters, organic
peroxides, etc., may have limited meaning in boiler combustion.  Also,
it should be recognized that most of the investigations reporting on POM
were published 5 or 10 years ago; the techniques of collecting and
analyzing POM have advanced considerably during that time and are still
being developed.

          Although the physical state of POM and soot are different, their
probable common ancestry and chemical reactivity suggest that an additive
effective for one would be at least somewhat effective for the other.
This conjecture is generally supported by the experimental evidence, but
with one interesting exception:  an "oil-soluble barium compound",
moderately effective in reducing diesel smoke, was reported as having no
effect on POM.  If, in fact, barium functions as a free-radical promoter,
one would expect it to be particularly effective in reducing POM as com-
pared to the post-flame oxidation catalysts, such as the transition metals
are believed to be.  However, if one accepts the conjecture that an
oxidation catalyst-type additive that is effective in reducing soot would
also be effective in reducing POM, there still is no reason to assume
that the optimum additive stability and volatility would be the same for
both pollutants.
                                  S-9

-------
          While an additive that reduces particulate may affect some
reduction in POM emissions, there is a need for additive investigations
specifically directed to POM control.

Effects on Carbon Monoxide and Hydrocarbons

          CO and hydrocarbons are significant pollutants, but their
emission levels are usually relatively low from continuous combustion
systems like boilers; consequently, few investigations have been reported
on the effects of additives on these pollutants from boilers.  Generally,
the oxidation-catalysts effective in reducing soot or, especially POM,
might be expected to have some effect on CO and hydrocarbon emissions.
In view of the relative unimportance of these emissions from boilers,
investigation of this aspect does not appear to be justified.
2.  ADDITIVE EFFECTS ON EMISSIONS
    RESULTING FROM FUEL IMPURITIES
          The fuel impurities considered here are sulfur and ash.  These
impurities are considered pollutants regardless of the form in which they
are emitted.

Effects on Sulfur Compounds

          Mechanistically, no fuel additive can be expected to have a
direct effect on the emission of sulfur present in the fuel.  Most of the
sulfur in the fuel leaves the system as S02 (assuming that the combustion
is not unusually fuel rich).  Typically, only 1 to 3 percent of the
sulfur leaves the system as 803, depending on combustion conditions.  In
addition, emissions can include particulate sulfates.*
*  It may be noted that some additives may change the chemical or physical
   form of  the sulfur emissions to one that is not detected by the sam-
   pling and analytical techniques commonly used to determine S02 or SOX
   emissions.
                                 S-10

-------
          Various fuel additives, mostly basic in character, have been
used to scavenge SO^ or, used in much larger amounts in downstream addi-
tion, to scavenge SO*.  (This reaction with S0~ is typically incomplete,
and the additive must be used in quantities substantially greater than
the stoichiometric amount to achieve high reductions.)  It is possible that
an additive could be found that would promote the reaction of SOo with
basic materials; the reported ability of sodium chloride to enhance the
absorption of sulfur oxides by the limestone in fluidized-bed coal com-
bustors suggests that such effects exist.  However, the large quantity
of basic material required, even if used stoichiometrically, still may
limit any practical application in ordinary combustion systems.

Effects on Ash

          Additives may change the physical properties of ash, but an addi-
tive cannot be expected to reduce the combustion generation of ash from
constituents contained in the fuel.  However, ash collectability may be
influenced.  863 and other conditioners are sometimes added in the post-
combustion zone to reduce the resistivity of fly ash, and thus promote its
collection in electrostatic precipitators.   There may also be an opportunity
of accomplishing this conditioning through the use of compounds (other than
sulfur) added to the fuel.

3.  ADDITIVE EFFECTS ON NITROGEN OXIDES

          Nitrogen oxide emissions are attributed to two sources:
thermal fixation of nitrogen and the oxidation of organically-bound-
nitrogen compounds in the fuel.  The reported examples of additives
causing significant (though not major) reductions in NOX emissions seem
more puzzling than illuminating.  The mechanisms by which the observed
effects could have been produced are obscure.
                                  S-ll

-------
Effects on Thermal-Fixation Processes

          The importance of 0- and N-atom levels in the mechanism for the
fixation of nitrogen suggests that additives which might reduce, princi-
pally, the concentration of 0-atoms could be effective in reducing NOX.
In practice, additives have not been especially effective in this regard.
One of the problems is that hypothetical additives which might reduce the
0-atom level by promoting atom recombination reactions, or by retarding
the combustion reactions leading to 0-atom formation, could be too detri-
mental to the overall combustion process.

Effects on Fuel-Nitrogen Conversion

          NOX is formed more readily from chemically-bound-nitrogen
species than from the fixation process.  To reduce NOX from fuel-N
species requires that the decomposition of the fuel-N be diverted to N2
or N20, rather than active CN or NH species.  No additives have been found
which accomplish this.  Interestingly, however, there is evidence that NH
species can limit, and perhaps reduce, the conversion efficiency of fuel-
N compounds to NO.

          On the basis of the information available, it is possible that
some minor degree of NO  control may be obtained with fuel additives; the
probability of practical success, however, still appears questionable.

A.  ADDITIVE EFFECTS ON BOILER EFFICIENCY

          Two routes are open to influence boiler efficiency by fuel or
combustion additives:   (1) through cleaning or maintaining the cleanliness
of heat-exchanger surfaces, or  (2) by  controlling pollutants, particularly
visible smoke, thus permitting operation at reduced  excess-air levels.
Additives intended to remove or modify soot and other surface deposits
have  a significant history but are outside  the scope of this report.
However, additives which suppress smoke  (soot) and other carbon
                                 S-12

-------
particulate may both reduce the rate of fouling of the heat-exchange
surface and permit operation at lower excess-air levels.   The possible
efficiency gain depends on the smoke-limited excess-air level of the
specific burner-boiler combination, its normal rate of fouling, and the
frequency of cleaning.  Efficiency gains of a few percent may be possible
                                                             (3)
in small boilers which are reasonably maintained and adjusted   .   In
larger boilers, with well-designed fuel-air mixing, equipped with soot
blowers and possibly heat-recovery equipment, and subject to skilled
maintenance and adjustment, additives offer little promise of efficiency
gains.

5.  POSSIBLE LIMITATIONS TO THE USE OF ADDITIVES

          A perspective of the potential use of fuel additives for control
of air-pollutant emissions from oil-fired and coal-fired boilers should
include these additional considerations:
             •  Combined emission impact, including effects
                of the additive on various pollutants .
             •  Possible toxic effects of trace pollutant
                emissions resulting from materials intro-
                duced with the additives.
             •  Effects of equipment characteristics on
                fuel additive effectiveness.
             •  Practical considerations of additive use,
                other than simply the reduction of one or
                more pollutants.
             •  Comparison of the effectiveness of additives
                with alternative control means.
Each of these possible limitations is discussed in the paragraphs below.

Overall Emission Impact

          The overall or net impact of a fuel additive on a number of
different pollutants must be considered in combination, so that means for
                                 S-13

-------
solving one pollutant problem do not create another problem by bringing
about an increase in another pollutant.  For example, some additives may
affect reductions in one pollutant but increase the emission of another
pollutant.   Fortunately, most additives which have been shown effective
in reducing pollutants that result from incomplete combustion are not
expected to adversely effect other criteria pollutants.

Toxic Effects

          Included in the overall consideration of additive usage is the
possibility that materials present in additives may produce pollutants
which are toxic.  For example, metal-containing additives will result in
emissions of the metal or metals, the possible adverse health effects of
which must be balanced against the reduced emissions of the target pollu-
tant.  In many cases, the potential adverse effects of such emissions are
not adaquately understood.
The Effect of Equipment Characteristics
on Fuel Additive Effectiveness
          In the case of fuel impurities (such as sulfur and heavy metals)
and the nitrogen oxides, the evidence for additive effectiveness is too
uncertain and the possible modes of action too obscure to permit any
rational discussion of how their effectiveness may be influenced by
equipment characteristics.

          In the case of products of incomplete combustion, especially
carbon particulates and soot or smoke, there is both experimental evidence
of additive effectiveness in continuous combustion systems and reasonable
conjecture  as  to how the additives function.  Consequently, for additives
effective in reducing smoke and carbon particulate, some consideration of
the influence of equipment characteristics, or combustion conditions, is
possible.  Such influences may explain the varying degrees of effective-
ness  found for the same or similar additive by different investigators.
Following are some examples of equipment characteristics that may influ-
ence  additive effectiveness.
                                  S-14

-------
          Effects of Fuel-Air Ratio.  It is easy to conceive that fuel-air
ratio may affect the performance of additives intended to control emissions
of products of incomplete combustion, since, in the extreme case of sub-
stoichiometric air, combustion must be incomplete regardless of additive
effectiveness.  The suggestion that barium functions as a free radical
promoter and is thus active in the high-temperature region of active com-
bustion, while the transition metals catalize or accelerate the oxidation
of solid carbon in the post-flame region, has been proposed as an explana-
tion for an observed difference in the effectiveness of these additives
at differing fuel-air ratios.  Thus, the effectiveness of barium in
suppressing smoke was little affected by the overall fuel-air ratio, while
the effectiveness of the transition metal manganese was observed to decline
with excess air.

          Effects of Time-Temperature.  The belief that the transition
metals are active in the post-flame region also suggests that their effec-
tiveness may be influenced by the rate of heat extration from the combus-
tion products.  If the combustion space is small and the combustion products
are cooled rapidly, additives that are intended to catalyze oxidation of the
carbon particles would be relatively ineffective.   Such rapid post-flame
cooling would have a lesser effect on the action of barium or,  assuming a
similar basis of action, the other alkaline earth metals which function in
the combustion zone.   In contrast, low combustion temperatures such as
accompanying two-stage or distributed combustion,  might reduce the effec-
tiveness of the alkaline earth metals without greatly affecting the activity
of the transition metals.

          Effects of Fuel Vaporization and Fuel-Air Mixing.   The possi-
bility exists for the fuel additive to be separated from the region where
its activity is needed.   Thus, an additive retained in a coke or ash
particle would be ineffective in reducing smoke emissions.   Likewise,
smoke formed during the initial burning of the fuel might be little
affected by an additive released from the liquid fuel at a latter stage
of the combustion process.   Separations such as these,  or their reverse,
                                   S-15

-------
will be influenced by the volatility and thermal stability of the addi-
tive, the character of the fuel, and the thermal regime characterizing
the vaporization zone of the equipment.

          At the moment, the above descriptions of possible influences
of equipment types on additive effectiveness must be regarded as provid-
ing a basis for anticipating such influences rather than being definitive
examples.  Experimental investigations under controlled conditions will
be required to identify the parameters of importance for various equip-
ment and additive types.

Other Practical Considerations

          Practical considerations that may limit the feasibility of
additives include possible effects on (1) storage stability* and handling
characteristics of specific fuels, or (2) deposits in burner parts or
flue passes.  Such effects may not be easily predictable without investi-
gation of additives under specific field conditions to be encountered.

Alternative Controls

          Other alternative means of emission control may be more effec-
tive than the use of additives in many situations.  Both practical
effectiveness and cost effectiveness should be considered.  For example,
burner design and adjustment for the desired fuel-air mixing condition
and thermal environment generally can reduce emissions resulting from
incomplete combustion.  However, where conditions are marginal and are
not subject to sufficient improvement, additives may have application for
control of those emissions resulting from incomplete combustion.  A
similar interaction may result from combustion modifications (such as
*  Storage stability includes such aspects as sediment and gum formation
   (especially moisture sensitivity).
                                  S-16

-------
staged combustion) intended to control nitrogen oxides.  If it should
develop that such modifications lead to excessive emissions of products
of incomplete combustion, especially soot, fuel additives may offer a
route to overall emission control.  At present, this possibility has not
been sufficiently investigated.

          In short, the needed perspective must include consideration of
the full spectrum of pollutants and control alternatives applicable to
each type of boiler installation.
                    RECOMMENDATIONS FOR FURTHER R&D

          This review has demonstrated that there is some evidence,
experimental and theoretical, regarding control by fuel additives of most
pollutants emitted by combustion systems.  The evidence for control by
fuel additives of visible smoke and carbon particulate is strong, for
control of POM is somewhat weaker, and for control of NC)  is quite weak
                                                        X
(except by ammonia as a Class III additive).  Control of total sulfur
emissions by additives does not appear to be possible; emissions of SO^
can be controlled, although the equivalent amount of sulfur may still be
emitted as a solid sulfate salt.  Possibilities for the control of
emissions of ash or specific ash constitutents by fuel additives is
restricted to enhancing the collectability of ash particles.  No experi-
mental evidence is available for control of hydrocarbons or CO emissions
by fuel additives, although theoretical considerations suggest that some
degree of control might be possible.

          In assessing the need for further investigation of fuel addi-
tives, the importance of the specific pollutants must be considered, as
well as alternative control techniques.  Hydrocarbons and CO are
not usually important pollutants from continuous combustion sources.  NOX
and SOX are important pollutants, but there is little evidence to suggest
that they can be practically controlled by fuel additives.  But 803
                                  S-17

-------
emission can be controlled by additives, and this control may be signi-
ficant in terms of local air quality.  Additional research appears to be
needed to clairify the role of SO^ emission relative to total sulfur
emission and emission of SO-j as solid sulfate salts.  Carbon particulates
and visible smoke are important pollutants, but may be more easily con-
trolled by proper equipment design and operation.  Ash, appearing as fly
ash, is perhaps a special case in that combustion additives or post-flame
additives may enhance the collectability of the ash and thus aid the con-
trol of this emission by other methods.

          On the basis of these considerations, the most promising area
for further R&D on combustion additives is the control of carbon parti-
culate and related pollutants (such as visible smoke and POM), wherein
modest increases in boiler efficiency can be an added benefit.  Three
specific investigations can be suggested beyond the current EPA/Battelle
laboratory evaluation of combustion additives for fuel oil.  These three
studies involve field and laboratory investigations of the effectiveness
of combustion-type fuel additives in controlling carbon particulate and
visible smoke from:
             1.  Different types and capacities of boilers within
                 the intermediate size range firing residual oil.
             2.  Boilers firing oil or pulverized coal where
                 combustio
                 employed.
combustion modifications for NOX control are
             3.  Coal burning in stoker-fired boilers ranging in
                 size from small-commercial to small-industrial.
          The first of these suggested investigations is intended to
assess the practicality of using residual-oil additives for emission
control for a variety of boilers under field conditions and oil types.
(Additives may offer a practical means of praticulate control for residual
oil-fired boilers in an intermediate range of commercial/industrial sizes,
typically used in applications where burner maintenance is marginal and
which are too small for high-efficiency collectors to be used economi-
cally.)  The second investigation would explore the potential of additives
                                  S-18

-------
in particulate control where combustion modification for NO  control
                                                           X
yields conditions that are conducive  to  the formation of carbon particu-
late.  The third investigation would explore the technical feasibility of
controlling emissions from small commercial coal-burning equipment that
may not be equipped with high-efficiency collectors.

          Depending on the importance attached to S0~ emissions, as con-
trasted to total sulfur emissions, research to determine the mode of
action of known SO., suppresants (i.e., prevention of SO., formation or
scavaging of SO., to form sulfate salts) may be justified.

          Throughout this review, attention is called to possible
research of a fundamental nature directed towards improving the perform-
ance of additives, finding new classes of additives, and explaining the
mode of action of additives.  The possibility that this research would
ultimately result in new or more extensive applications of additives for
emission control is regarded as relatively remote.  Consequently, insofar
as the interest in additives is practically directed, these more funda-
mental research opportunities would be assigned a lower priority than
field and laboratory trials to assess the use of additives over a range
of burner-boiler types, sizes, and operating conditions
          Detailed discussion and other background information that
supports the foregoing Summary are contained in the following parts:
             Part I.   Combustion Mechanisms as They Relate to
                       Fuel-Additive Action in Controlling
                       Emissions
             Part II.  Review of Experimental Investigation of
                       Combustion-Type Fuel Additives
                                  S-19

-------
                                 PART I
                 COMBUSTION MECHANISMS AS THEY RELATE TO
              FUEL-ADDITIVE ACTION IN CONTROLLING EMISSIONS
        The combustion of fuels for the generation of heat is ordinarily
an efficient process, and the available heat of combustion which is
lost in the form of incompletely burned fuel is generally insignificant.
However, it is now recognized that these small quantities of incom-
pletely burned fuel, as well as emissions resulting from certain fuel
impurties and high temperature nitrogen fixation, are extremely signi-
ficant as air pollutants.  This recognition has resulted in considerable
research directed towards gaining an understanding of how and why these
pollutants are formed, with the hope that such understanding will lead
to methods of preventing their formation or of destroying them during
the combustion process.


                     POLLUTANTS CATEGORIZED BY ORIGIN

        In terms which reflect both their source and the possible
methods of control, the pollutants emitted from combustion processes
can be categorized as follows:
           •  Chemical species derived from incomplete com-
              bustion of the fuel and capable of undergoing
              further reaction with oxygen.  These species
              include CO, "Carbonaceous particulates" such
              as soot or coke particles, "hydrocarbons",
              and polycyclic organic matter or POM.
                                     1-1

-------
           •  Nitrogen oxides resulting from the thermal
              fixation of atmospheric nitrogen.
           •  Nitrogen oxides resulting from the oxidation
              of chemically bound nitrogen in the fuel.
           •  Compounds of toxic elements, such as sulfur
              and heavy metals, resulting from the presence
              of these elements in the fuel.
           •  Mineral particles formed from the mineral
              matter or ash present in the fuel.
        It is immediately clear that if the emission of an element is
regarded as undesirable regardless of its form, and that element is
present in the fuel, then nothing can be done in terms of the combus-
tion process that will avoid its release.  The best we could hope for
would be to retain the element in the combustion system for later
disposal rather than emitting it into the atmosphere.

        As an example, sulfur in fuel is ordinarily emitted as SO
together with a small amount of SO,..  Presumably one might (and some-
times does) alter the combustion conditions to favor one or the other
of these oxides, but emissions of either are regarded as undesirable.
By drastic alterations of the combustion process, the sulfur might be
converted to species such as H S, CS , etc., but emissions of these
species are unlikely to be regarded as an improvement.  Finally, one
might add substances to the fuel which will react with the sulfur
oxides, such as sodium carbonate or limestone, forming solid sulfates
and sulfites.  Although less noxious than the sulfur oxides, indis-
criminate emissions of such salts add to the particulate loading of the
atmosphere,  and hence, it is important to collect these solids from
the stack gas and dispose of them in some unobjectionable manner.
                                   1-2

-------
        The same situation pertains to either fuel "impurities" and
ash-forming matter.  The situation is, however, different for the
first three pollutant categories listed above.  All of these are
characterized as being nonequilibrium species in the sense that they
would not be expected to exist in significant quantities in a homo-
geneous fuel-air mixture which passed through the observed temperature
profile while maintaining continuous chemical equilibrium.  Hence,
their existence suggests that either the fuel-air mixture was not
homogeneous, or that continuous chemical equilibrium was not
maintained.

        When viewed in sufficient detail, combustion processes always
prove to have some degree of inhomogeneity:  at the least, some region
containing a composition, temperature, or velocity gradient must exist
to stabilize the flame.  In practical combustion processes, other
inhomogeneities may exist due to the nature of the fuel (e.g., heavy
oils and coal) or may be deliberately introduced to control the nature
of the flame (e.g., luminous or nonluminous) or the heat release
pattern.  However, the design of practical combustion systems attempts
to achieve homogeneity at some point within the combustion system.  If
this condition is not achieved, or is achieved after the gases have
been cooled to too low a temperature, then possible routes to the
emission of various combustible pollutants can easily be visualized.
Thus, if a flow path exists which is never subjected to the conditions
needed for ignition of the fuel, or which never contains enough oxygen
for the complete combustion of the fuel, then species such as hydro-
carbons, CO, soot, etc., may become equilibrium species for that path.
Some mechanism of this type is often suggested as a source of hydro-
carbon and CO emissions from the reciprocating internal-combustion
engine.  Such emissions represent not so much a "failure of chemistry"
as a failure of design.
                                   1-3

-------
        Perhaps more typical of practical combustion processes is the
case where a sufficient degree of homogeneity is achieved (either by
preflame or postflame mixing) but at a temperature too low or, for a
time, too short to permit the desired chemical reactions to proceed to
completion.  The chemical reactions are quenched and equilibrium is
never achieved.

        The following section within Part I of this report considers
the mechanisms of formation for major pollutants.  In the succeeding
section, the mechanisms for the various pollutants are discussed as
they relate to opportunities for fuel additive action in controlling
emissions.
                  MECHANISMS OF POLLUTANT FORMATION
                       IN COMBUSTION PROCESSES

        The oxidation of hydrocarbons in flame is a complex process
accompanied by the formation and subsequent destruction of many inter-
mediate products.  Thus, proposed reaction schemes for the burning of
methane with oxygen may include at least 12 species without considering
minor species containing more than one carbon atom.  For methane, the
reaction scheme is, in outline  '  :
                            CH.    *• CH« ,

resulting from pyrolysis, or attack by 0, H, or OH:
                            CH3 	*-  HCHO ,

                            CH3 	**  HCO  ,

which results from reaction with 0 or 0.:

                            HCHO 	*- CO ,
                            HCO 	*~  CO  ,
                                  1-4

-------
through reaction with OH or thermal decomposition, followed by:
                            CO + OH 	>•  C02 + H  .
The oxidation of more complex hydrocarbons is usually considered to
proceed in a roughly comparable way.
PRODUCTS OF INCOMPLETE COMBUSTION
Carbon Monoxide
        The first important feature of this scheme is that the formation
of CO is an unavoidable step in the oxidation of the carbon to CO-.  The
second important feature is that CO is oxidized or destroyed by reaction
with hydroxyl  radicals.
        The significance of these two features is first, that one cannot
expect to avoid the formation of some CO when burning a hydrocarbon (or
generally, any organic compound), and second, that the destruction of
CO by the normal route (i.e. reaction with hydroxyl) may be a marginal
process.  This second point results from the fact that the concentration
of hydroxyl is strongly temperature dependent:  if the CO-containing
gas is rapidly cooled, the hydroxyl concentration may drop to insigni-
ficant values before the CO is oxidized, regardless of the availability
of oxygen.

Hydrocarbons

        The emission of the materials normally labeled "hydrocarbons"
from combustion processes in which a reasonable degree of homogeneity
is achieved is difficult to explain, if for no other reason than the
fact that with a few exceptions these materials are not well charac-
terized.  Of course, "hydrocarbons" are not usually important com-
bustion-derived pollutants other than from systems well known to
                                   1-5

-------
   4400


   3600


*E 2800


 =^2000
 * 1600
il 1200
    800
400
    0
                                          C2H2/02
                8    12    16   20   24  28   32

                 Height Above Burner Surface,  cm
36
    FIGURE 1.  FLUX  OF  POLYCYCLIC AROMATIC HYDROCARBONS
              VS HEIGHT AND BURNER SURFACE  (from
              Reference 9)
                           1-6

-------
possess important quenching effects or inhomogeneities, such as uhc
reciprocating engines and aircraft gas turbines, or which undergo
repetitive transients, such as the off-on cycle of automatic home
heating equipment    .  If "hydrocarbons" are emitted other than by
the above processes, one might expect that they result from quenching
the combustion process at a point where significant quantities of
"organic" carbon remain, as hydrocarbon radicals, aldehydes, etc.,
with subsequent stabilization by recombination, polymerization, etc.

Polycyclic Organic Matter

        Materials labeled polycyclic organic matter (POM) are emitted
from combustion systems in small quantities.  POM is defined to include
polycyclic aromatic hydrocarbons (PNA or PAH), nitrogen-containing
                                                          (4 )
heterocyclic compounds, and possibly derivatives of these    .  Studies
of the mode of formation and destruction of POM in the combustion
process are difficult to carry out because of the very low concentra-
tions, the lack of simple analytical procedures for identifying them,
and the large number of individual compounds making up this class of
material.

        Studies of rich acetylene and ethylene flames strongly suggest
that POM is formed through the polymerization of a two-carbon radical,
probably acetylenic, and is probably destroyed by some reaction involv-
             (8  9 )
ing hydroxyl        .  The two-carbon radical is not necessarily
derived directly from the fuel, but rather may be built up from single
carbon fragments.

        Figure 1 illustrates the observed flux of POM in rich, low-
                                 (9)
pressure acetylene-oxygen flames     .  Evidently POM is formed rapidly
in the initial part of the flame, and much of this is destroyed in  the
highest temperature region.  However, a secondary process of formation
occurs as the temperature of the burned gases drops into the range  of
                                   1-7

-------
650-900  C.   The fact that POM exists and is even created under such
conditions is indicative of the extreme thermal stability of the poly-
aromatic sturcture.

        It may be noted that the POM flux becomes very small as the
fuel-oxidizer ratios approach stoichiometric, even through the mixture
is still very fuel rich.

        Some fuels contain POM as a constituent.  Specifically, coal
might be regarded as being largely POM, and if combustion conditions
are such that insufficient temperature or oxidant concentration is
encountered by the fuel-POM, this source may dominate the emissions.
In fact, large emissions of POM have typically been associated with
hand-fired coal furnaces, burning coal mine waste, and other poorly
controlled combustion systems    .
Soot
        Soot, or loosely, smoke, was probably the first combustion-
generated air pollutant to be recognized and, consequently, has received
considerable research effort.  Nevertheless, soot remains, perhaps, the
least understood of all the combustion-generated pollutants.  Soot is
formed both by reactions in flames and by pyrolysis of organic carbon
species.  It has long been recognized that the emission of soot
represents incomplete burning of the fuel and that the ample provision
of air and maintenance of high temperatures decreased the quantity of
or eliminated the soot.

        With the introduction of petroleum-based fuels, it was found
that various hydrocarbons had differing soot-forming tendencies.  Thus,
in the smoke lamp, it was found that the various hydrocarbon types
could be rated in terms of smoke-forming tendency as : paraffins < naph-
thenes < olefins < monocyclic aromatics < polycyclic
                                 1-8

-------
        It has also been found that the dilution of the fuel or air or
their mixture with inert gases, steam, carbon dioxide, or cooled flue
                                            /•i 3\
gas tended to suppress the formation of soot    .   Water sprays direct-
ed into the flame and firing water/oil emulsions     have also been
found to suppress the formation of soot.  The mechanisms behind these
phenomena have not been resolved.

        Examination of soot particles with the electron microscope has
revealed that the large, visible particles were agglomerates of small
spherules with diameters in the range of 10 nm.  X-ray diffraction
studies showed the presence of graphite crystallites  '

        In spite of this long history of research, and the discovery of
methods of controlling at least partially the emission of soot from
combustion processes, soot remains an important pollutant for numerous
small combustion system   '

        It is probably incorrect to consider soot  as only carbon;
chemical analysis typically reveals the presence of hydrogen in non-
negligible quantities.  Moreover, it has been found that soot may
contain significant quantities of POM and other hydrocarbons that can
be extracted by solvent or vaporized in a vacuum.   In some soots, the
                                   / Q \
POM may exceed 3 percent by weight    .

        In view of the graphite structure observed in soot, its
hydrogen content, and the simultaneous occurrence  of soot and POM, it
is possible to conceive of soot as POM carried to  an extreme.  Indeed,
various studies have indicated a probably origin of soot in the two-
carbon fragments  found  in flames, and the flux of soot or soot
precursors in rich acetylene flames has been found to be very similar
to the flux of POM(8' 9> 17).
                                 1-9

-------
E

03

i
9)
O
(0
c
X
3
70


50


30


 10
2.8
         <2.0
         o.
            1.2
           0.4
                Flux of PAH
                (a POM  constituent)
                                            Flux of insoluble matter
                                            (soot or soot precursors)
                                            I      i      T     i      I
                                                                           1300
                                                                           1100
                                                                           900
                                                                           700
                                                                           500
                                                                                a
                                                                                o
                                                                                a>
a>
Q-
a>
                          8    12    16    20    24
                             Height Above Burner, cm
                                                       28
                                                  32   36
       FIGURE  2.  COMPOSITE  PLOT OF  FLUX OF  INSOLUBLE MATERIAL  (SOOT), FLUX
                 OF  POLYCYCLIC AROMATIC HYDROCARBONS (PAH) AND OF
                 TEMPERATURE, RESPECTIVELY, VERSUS HEIGHT ABOVE BURNER
                 SURFACE  IN A RICH,  LOW-PRESSURE ACETYLENE-OXYGEN FLAME
                 (from  Reference  9)
                                        1-10

-------
        Figure 2 illustrates this parallelism^  '.   It has been
                                                            /ION
suggested that the oxidation of soot depends primarily on OH    ,
although suggestions have also been made that oxidation by 0 atoms is
important or even dominates  '    .   The strong pro-soot activity of
S0_ in flames has, thus, been explained as resulting from the scaveng-
ing of 0 atoms by the SO  ^6'.
        The problem presented by soot differs from that of POM in at
least one important aspect:  soot is emitted as discrete particles,
whereas the POM is emitted from the flame as a vapor.  Thus, in
addition to the chemical problem of its formation, the formation of
soot must involve the physical processes of nucleation and growth of
a new phase.  Similarly, the oxidation of soot is a heterogeneous
reaction involving the diffusion of species to and away from the
surface.

        The possibility that electric charges and fields may play a
role in either the nucleation, growth, or agglomeration of soot has
frequently been suggested.  Evidence supporting such concepts has been
obtained but this evidence does not exclude mechanisms involving
neutral species ^  '    .

        Emissions of carbon from combustion systems include, in
addition to soot,  relatively large "carbon" particles representing an
unburned residual of the original fuel particle or oil droplet.  These
particles are formed first by devolatization of the fuel particle by
the processes of simple vaporization and pyrolysis, leaving behind a
nonvolatile residue of coke.  If combustion conditions are such that
the hot coke particle co-exists with vaporized fuel, heterogeneous
pyrolysis reactions may lead to growth of the coke particle at the
expense of the fuel vapor.
                                   1-11

-------
       Because the pyrolysis reactions leading to coke formation
occur at appreciable rates only at high temperatures, above perhaps
300 C, the formation of coke particles is usually a problem only
for those fuels having constituents with low vapor pressures at such
temperatures, e.g. heavy oils and coal.  Because survival of the coke
is dependent on its residence time in a high-temperature environment,
coke-particle emissions are typically a greater problem in smaller
equipment, such as commercial-size boilers, rather than in the larger
industrial or utility boilers.

       The amount of coke formed is dependent on the heating rate,
because the pyrolytic polymerization reactions in the fuel particle
proceed at a finite rate.  Thus, studies of coal devolatization
at a heating rate of 1000 C/sec produced larger coke residues than did
devolatilization conducted at heating rates of 10 -10  C/sec    .   The
pyrolysis reactions are usually described as proceeding through
formation of free radicals, chain propagation by hydrogen abstraction,
fission to shorter carbon chains, polymerization to heavier materials
with lower hydrogen content, and various radical destruction pro-
                                       ;ms,
                                       ,(23)
       (22)
cesses    .  However,  molecular mechanisms,  as contrasted to free radi-
cal mechanisms, have also been proposed

       Because of their large size, the burn-out of coke particles
is a relatively slow process, most likely controlled at high
temperatures by the diffusion of oxidant to and products from the
particle surface.  In parallel to the case for soot, the effective
oxidant may be OH or 0  '   '    , with the result that at lower
temperatures the burning rate may be quite small even in the
presence of substantial 0~.
                                  1-12

-------
EMISSIONS RESULTING FROM FUEL IMPURITIES

       Fuel impurities of interest as pollutants currently include
sulfur, inorganic substances which produce flyash, and various toxic
elements.  At the moment, sulfur oxides and "particulates" (regardless
of chemical composition) are recognized as air pollutants in the sense
that governmental restrictions have been applied to their emission.
Although the results of future research are likely to extend this
situation, as more and more gaseous or solid species are recognized
as specific pollutants and subjected to individual regulation, atten-
tion here will be confined to sulfur-containing species and flyash.

       Inasmuch as sulfur and metals occur in the fuel, they will
be released in the combustion process regardless of the detailed
mechanisms involved.  Thus, they can be avoided as pollutants only
by removing them at some point prior to, during, or after combustion.

Sulfur Compounds

       In ordinary combustion processes (as distinct from gasification
processes) sulfur may exist as either SCL or SO .   Although not well
explained, the oxidation of sulfur is believed to proceed through
the initial formation of SO followed by
                        SO + 02 	»  S02 + 0 ,

                        SO + OH 	>• S02 + H .

S09 can then be oxidized to SO  by the reaction
                        S02 + 0 + M 	>• S03 + M ,
and the SO- destroyed by
                        SO. + 0 	*•  S00 + 0,
                          J             L    t.
                                  1-13

-------
                        S03 + H - >•  SO  + OH ,  and possibly
                                                C21
                        so3 + co — >-  so2 + co2k ;

Because S0_ Is unstable at high temperatures, some of the SO  initially
formed by the abundance of 0 atoms in the flame is subsequently decom-
posed.  As the temperature drops, S0_ becomes more stable relative to
SO., but the uncatalyzed oxidation of SO^ is slow.  As a result, the
concentration of SO  in the high-temperature region is typically
observed to be much larger than the equilibrium value, while its
concentration in the low- temperature region is much smaller than the
equilibrium value.
       Although control of the combustion process can do little or
nothing about the total sulfur oxides emitted, some control can be
gained over the SO /SO  ratio.  This is sometimes desirable to avoid
corrosion of low-temperature heat-exchange surfaces, or to avoid the
formation of visible and noxious sulfuric acid mists or acid smuts
(aerosols consisting of sulfuric -acid-moistened flyash and soot).  In
terms of combustion control, this has been accomplished in boilers by
burning at low excess-air levels, such that the oxygen content of the
                                    (24)
flue gas does not exceed 0.2 percent    .  Under these conditions,
the concentration of SO  in the flue gases appears to be substantially
less than 0.1 ppm.  The mechanism by which this result is obtained has
not been explored.  It appears likely that the formation of SO  in the
flame zone would be unaffected, or even increased due to the higher
temperature and, consequently, higher 0 atom concentration.  Thus,
decomposition of SO  at intermediate temperatures and suppression of
the low-temperature oxidation of SO  by limiting the availability of
oxygen are probably responsible for the low SO. levels achieved.
                                  1-14

-------
Fly Ash

       The ash forming constituents of fuels are predominantly metals,
although sulfur, usually as sulfate, may occur in small amounts.  When
present in significant amounts in the fuel, phosphorus may also be
found in the ash.  In coal, the original forms of the ash constituents
are shales, clays, pyrites, and calcite.  In oils, the quantity of ash-
forming material is much less than in coals, i.e., a few hundred to a
few thousand ppm in residual oil and much less in distillate oils, as
contrasted to typically 10-20 percent in coal.  The ash constituents
in residual oil are vanadium, nickel, and possibly iron, occurring as
oil soluble porphyrins; unidentified organic calcium, magnesium and
zinc compounds; various clays, oxides, etc., suspended in the oil; and
water-soluble salts characteristic of oil-field brines or sea water.
The total ash and its composition are highly variable(25),

       The detailed behavior of ash constituents during the combustion
of residual oils and coals has not been investigated.  Probably the
various metals are converted to their oxides, if not originally pre-
sent as such during burn-out of the residual coke particle.  If the
combustion temperature is sufficiently high, some or all of the oxides
may be fused or even volatilized, to solidify and condense as the
temperature falls.

       In the case of coal combustion, some control of the ash pro-
perties and ash emission is possible through equipment design and
operation.  Thus, burning the coal in fixed-bed equipment or cyclone
furnaces causes much of the ash to be retained as a solid.  In
pulverized-coal firing, the furnace can be designed either to fuse
the ash and collect much of it as slag within the furnace, or to
solidity the ash in suspension and thus carry nearly all of the ash
out of the furnace with the gaseous combustion products.
                                  1-15

-------
       In terms of chemistry, the only practically applied control of
ash properties occurs in the low-excess-air burning of residual oils
Although practiced primarily for control of S0_, the reduced avail-
ability of oxygen causes the vanadium to form as V 0  or V 0  rather
                                                  £• j     t, 1
than the V 0  characteristic at normal oxygen levels.  These lower
oxides are less objectionable from the standpoint of boiler corrosion.
FORMATION OF NITROGEN OXIDES

       In the following discussion concerning NO  formation, the roles
                                                X
of thermally fixed nitrogen and of chemically bound fuel nitrogen are
treated spearately.

Thermally Fixed Nitrogen

       The thermal fixation of nitrogen in flames is now  nearly
universally attributed to the Zeldovich mechanism     ,
                         0 + N  -—»  NO + N  ,

                         N + 02 —*-  NO + 0  .

Reactions such as
                         N  + CH 	*•  CN + NH    and

                         CN + 02 	*•  CO + NO

have been suggested as important sources of "prompt NO"  (that NO
                                                           (27)
which is formed early in the combustion process) in flamesv  ' .
These suggestions have been challenged with the recognition that
super-equilibrium concentrations of 0 atoms in flames could lead
to the observed NO formation rates through the Zelovich mechanism.
                                 1-16

-------
        NO  is  not  a  stable  species  at  low  temperatures,  and  the  equili-
 brium  concentration of NO  at  ordinary exhaust  or  stack-gas  temperatures
 is  not significant.  NO persists in the combustion  process  because of
 the extreme temperature dependence (high  activation energy) of  the
 reactions  leading to its formation and destruction.  These  reactions
 are effectively quenched or frozen at temperatures  below about  2500 F.
 Even moderate rates  of heat removal from  the combustion products will
 freeze the NO concentration at a value characteristic of a  high-
 temperature equilibrium.

        Although the  NO concentrations  in  the cooled combustion  products
 far exceed the equilibrium value,  they are typically much lower than
 the NO concentration in equilibrium at the highest  temperature  achiev-
 ed  in  the  flame (possibly  3000 ppm).  This results  from the failure to
 remain at  the highest temperature  for sufficient  time to reach
 equilibrium, in some situations from  decomposition  of the NO before
 the reactions are quenched, and possibly by reactions between the NO
 and other flame species.

        Various methods of  controlling the formation of NO (through
 Zeldovich mechanism) have  been proposed.   These methods are based on
 preventing the formation of appreciable quantities of NO through
 control of the peak  combustion temperature.  Various methods of com-
 bustion control have been  investigated:  dilution by flue gas through
 internal or external recirculation, humidification of combustion air,
 controlled (i.e.,  slow)  heat release combined with heat extraction,
 and  two-stage combustion with intermediate heat extraction.   These
methods have been generally successful in reducing the quantity of
NO  emitted^28).
                                 1-17

-------
Chemically Bound Fuel Nitrogen

       It is well recognized that, besides the Zeldovich mechanism for
explaining the thermal fixation of nitrogen, organic-nitrogen com-
pounds in fuels are also significant sources of NO  in combustion
       (29  30  31  32)                             X
systems   '   '   '   .A number of recent studies have shown that
fuel-bound nitrogen yields high and varying amounts of NO in combus-
tion product gases.  Fuel-bound nitrogen may be in the form of
indoles, carbazoles, pyridines, quinolines, etc.  Since these compounds
only possess limited stability, especially in an oxidizing atmosphere,
in the preflame, high-temperature gradient regions of the flame, they
will break down and oxidize fairly readily.  Mechanistically, the
nature of the active fuel-nitrogen intermediate has not been identi-
fied.  However, as these nitrogen-containing compounds and inter-
mediates break down and oxidize, they yield a variety of NH and CN
species, i.e., N, NH, NH , CN, HCN, CNO, etc., which will be oxidized
to NO.  These species will lead to NO formation by steps such as
                        N + 0  = NO + 0
                        NH + 02 = NO + OH
                        NH2 + 02 = NO + H20
                        CN + 02 = NO + CO
                        CNO + 0  = CO + NO + 0

       If the NH and CN species behaved strictly as indicated here,
one might expect all the fuel-nitrogen species to be converted to NO.
It is noted, however, that the conversion of fuel nitrogen to NO is
                              /QT   *J *J  Q /  1Q  Q£\
not quantitative.  Experiments   »   »   »   »   ' show that the con-
version of fuel nitrogen to NO:
          - decreases as the concentration of fuel nitrogen
            increases
                                 1-18

-------
          - decreases as the amount of excess air is decreased
            (Under fuel-rich conditions, NO can be reduced to
            nitrogen).
To explain the nonequilibrium and "nonstoichiometric" production of
NO, one must invoke reaction steps which remove or prevent NO from
being formed.   Two principal reaction steps generally invoked are
                         N + NO = N  + 0
                         NH + NO = N 0 + H  .

These steps are both thermodynamically favorable and kinetically
rapid.  The first of these reactions, in fact, is faster than N + 0,
                                                                   i
NO + 0 under fuel-rich conditions.
             POSSIBLE OPPORTUNITIES FOR ADDITIVE ACTION
                IN CONTROLLING COMBUSTION-GENERATED
                     EMISSIONS BY FUEL ADDITIVES
       Purely in terms of what may be conceived, in contrast to what
may be possible or practical, the major opportunity for control of
emissions by additives must be confined to those pollutants described
previously as nonequilibrium species, namely, the products of incom-
plete combustion and the nitrogen oxides.  Opportunities are much more
restricted for the application of additives for control of emission of
objectionable elements present as fuel impurities and of fly ash;
generally, at best, one could only expect an additive to alter the
physical or chemical form in such a way as to promote the removal of
these pollutants in a downstream clean-up process.

       In the following sections of Part I, the opportunities for
additive actions are examined successively for control of
          •  Emissions from incomplete combustion
          •  Emissions from fuel impurities
          •  Emissions of nitrogen oxides.
                                1-19

-------
CONTROL OF EMISSION FROM INCOMPLETE COMBUSTION

       Perhaps the most significant feature of the combustion
mechanisms previously discussed is that the global reaction
                      Fuel + 0  	>•   Products
is not an elementary reaction, or reaction mechanism, of the com-
bustion process.  It is, of course, an everyday observation that
fuels do not ordinarily burn simply upon mixing with air or oxygen.
The real oxidants and reductants in flames prove to be radicals and
fragments formed initially by thermal dissociation and oxidation
steps at high temperatures and maintained and increased in concen-
tration by chain-propagating and chain-branching reactions.  The
destruction of the active species by thermal or surface quenching, or
by other chain-terminating events, may leave some of the fuel uncon-
verted to the ultimate products, CO- and H-O.  Emissions of CO and
"hydrocarbons" may result from such a sequence.

       A process which could possibly complicate this situation is
the existence of a fast reaction which converts reactive fuel species
to products which are much less reactive with the oxidants.  Com-
bustibles could thus be protected from rapid oxidation.  With the loss
of heat from the combustion gases and consequent loss of the reactive
oxidants, such materials could appear as stable combustion products.
Soot and POM are representative of pollutants formed in this manner,
and with only a slight alteration of the concept, coke particles as
well.
       Additives might be conceived of as altering the above sequences
by one or more of the following approaches:
          1.  Increasing the concentration of active
              oxidants, providing 0 and OH.  This
              might be accomplished by promoting their
              formation or inhibiting their destruction.
                                1-20

-------
          2.  Introducing an active oxidant species
              not normally present in flames.
          3.  Inhibiting the formation of unreactive-
              combustible materials.
Each of these three approaches to reducing emission from incomplete combus-
tion is discussed below.
1.  Approach of Increasing the Active Oxidant Concentration

        The concentration of 0 and OH in a flame or in hot combustion
gases might be increased by a catalyst promoting the dissociation of,
for example, ^0, by adding a material which itself is dissociated or
decomposed to yield 0 or OH, or by adding a material which would remove
species responsible for the destruction of 0 and OH.  None of these
possibilities are considered likely of success.

        The concept of increasing the concentration of 0 or OH in a
flame (or in the still-hot combustion gases) by some catalyst would be
attractive, except for the observed fact that 0 and OH concentrations,
particularly 0, exceed the thermal equilibrium concentrations not only
in the flame, but for some distance downstream of the visible flame.
Thus, a catalyst would be most likely to promote recombination reaction
and reduce the concentrations of 0 and OH.

        Additives which would decompose in the flame to yield 0 and OH
appear similarly unpromising.  The normal concentrations of 0 and OH
are such that an additive would be needed in amounts comparable to the
fuel if the concentrations are to be significantly increased.  Finally,
0 and OH are capable of reacting with nearly all species present in
flames, and an additive that would scavenge such reacting species would
be needed in amounts comparable to the total mass of fuel and air.
                                    1-21

-------
        The above view is entirely negative for additive control of
pollutants from incomplete combustion.   Nevertheless, it has been
conjectured that barium functions as a smoke suppressant by catalyz-
                                                      (37)
ing the dissociation of hydrogen and water to radicals    .   The
identity of the radicals is not specified, but they could only be H
and OH, with the additional possibility of 0 through the 02~H
reaction if 02 is present.  The evidence for this conjecture is not
expecially convincing, but the reported activity of barium as a
promoter of NOX is intriguing.  If anything like the conjectured
mechanism is correct, barium additives would be expected to have
wider-ranging effects on the combustion process.  Particularly, barium
would be expected to be effective in reducing the emissions of all of
the products of incomplete combustion.   The properties of barium which
are responsible for this conjectured catalytic activity are unknown,
and no rationale can be suggested for finding other such catalysts,
other than investigating additional alkaline-earth metals.  Calcium
and strontium have both been reported as having some smoke suppressing
properties    ' 39' *°'    .  The potential of such additives for reduc-
ing the entire range of products of incomplete combustion is such that
additional investigations to resolve the matter may be warrented.

2.  Approach of Adding an Active Oxidant Species

        The second listed suggestion of introducing a "new" active
oxidant does not appear attractive if it is considered as a homoge-
neous  (gas phase) oxidant.  Any oxygen-based specie would be expected
to rapidly equilibrate to native species such as 0 or OH, and species
based on other elements,  e.g., fluorine, are likely to result in
emissions potentially more objectionable than the products of incom-
plete combustion.  In contrast, a specific class of heterogeneous
(solid phase) oxidants, namely, oxidation catalysts, are known and
applied in many situations.
                                     1-22

-------
        The initial formation of a reactive fuel fragment or oxygen-
 containing radical in a homogeneous system has a large activation
 energy and, hence, occurs at a significant rate only at high
 temperatures.  Oxidation catalysts achieve a similar result at much
 lower temperatures.   This is accomplished by what can be crudely
 regarded as splitting the formation of an active specie into two or
 more reactions, having individually modest activation energies.   Hence,
 while catalyzed combustion reactions at modest temperatures are much
 slower than the high-temperature reactions in flames, they are much
 faster than the homogeneous combustion process at the same modest
 temperature.

        The difficulty in applying heterogeneous oxidation catalysts
 as  additives  to control emissions from combustion processes is that
 the combustible fuel  molecules  or particles  must have physical contact
 with the catalyst.  This can be arranged by  passing  the  gas stream
 through a bed of catalyst,  i.e.,  a catalytic afterburner.   When  the
 catalyst is used as a fuel  additive,  some care is  required  to  insure
 that the additive in  its functional  form is  well dispersed  or, perhaps
 better,  is  concentrated  in  the  region where  its  activity  is  needed.
 Otherwise,  large quantities  would  be required,  and most catalysts are
 relatively expensive  and are  potential pollutants  themselves.  At least
 in  the  case of  fuel additives for  liquid  fuels,  the  initial  dispersion
 can  be  obtained  by using  the  additive in  an oil  soluble form.

        In the literature  reviewed  in Part II of  this report, there  are
 suggestions that  the effectiveness of  smoke-suppressing, metal-based
 additives is related to  the thermal stability of the particular com-
 pound used.  The observed differences might also easily include an
 effect of volatility.   It is not clear that the thermal stability and
volatility of metal-bearing additives has been sufficiently investigat-
ed to assure that optimum effectiveness has been obtained.
                                   1-23

-------
        The possibility  that  oxidation catalysts  might  reduce  emission
 of coke particles  seems not  to  have  been investigated  as  such.   The
 chemical form of  the  additive for  greatest  activity  may well  be
 different than that for smoke suppression.

        Finally, if the  catalytic additive has  been dispersed  as
 extremely fine particles,  say through the vapor-phase  oxidation  or
 pyrolysis of  a metal-containing organic, sufficient  contact with
 gaseous  pollutants may  be  achieved to produce  noticeable  reductions
 in "hydrocarbons", CO,  and POM  emissions.   Thus, a reduction  in  CO
 emissions has  been observed  when using a manganese smoke  suppres-
     (42)
 sant      .  While  this  may not  represent a  practical approach to
 control  of these gaseous pollutants,  additional  investigation might
 be justified.
3.  Approach of Inhibiting Formation
    of Unreactive Combustible Material
       Soot and POM are believed to be formed from some 2-carbon frag-
ment, possibly derived from a precursor 1-carbon fragment.  Thus, one
might look for an additive that would stabilize or destroy these
fragments.  Destruction would be most likely accomplished by an oxidant,
and this has been discussed previously.  Stabilization, by some chain
termination process may have merit, but the choice of an agent is not
obvious.   In view of the many species already available in flames
which might cause termination, it is apparent that the reactive
materials are either resistant to termination or easily reactivated
after termination.  Also, any agent likely to terminate the POM and
early soot-building steps would likely react with other radicals in
the flame, consuming the agent or leading to undesirable changes in
the overall burning process.   Nevertheless, currently available infor-
mation does not preclude the possibility that some smoke-suppressant
additives might function in this manner.
                                  1-24

-------
       In the case of soot, it might be possible to interfere with the
nucleation, growth, or agglomeration steps, even though the production
of "soot molecules" was not prevented.  If this can be done, the
smaller particles or molecules may be destroyed in subsequent parts of
the flame, or they may be emitted in less visible and possibly less
objectionable forms.  It is suspected that gaseous diluents and water
sprays may function in this manner, either by the simple process of
diluting the precursors and thus reducing the particle growth rate, by
some unquantified thermal effect, or perhaps by some as yet unsuspected
mechanism.

       If the suggestion that electric charges and fields play an
important role in the nucleation, growth, and agglomeration of soot is
correct, then a rational basis for investigating potential additives
can be suggested.  If flame ions serve as nuclei for soot particle
growth, then one could try to either reduce the ionization level, thus
eliminating nuclei, or to greatly increase the ionization level, pro-
viding so many nuclei that no individual particle could grow to signi-
ficant size.  Also, if the ionization level in the flame is increased,
electric fields would be reduced by the increased conductivity.  The
easily ionized alkali metals (and to a lesser extent the alkaline earth
metals) are effective in increasing the ionization level in flames,
whereas chlorine (and to a lesser extent hydrogen) are effective in
suppressing ionization.  The effects of alkali metals (introduced as
nitrates and chlorides) on soot emissions have been reported to vary
with the specific metal, the metal concentration, and the flame
temperature.'^)  Both promoting and inhibiting activity was observed.
It should be possible to devise experimental additives that could
either increase or decrease the flame ionization level and quite
possibly control the position in the flame at which the additive became
effective.
                                  1-25

-------
       There would seem to be little that an additive could do to
affect the formation of coke particles in a pulverized coal flame,
but perhaps some effect might be obtainable with residual oil.  If
the pyrolysis reactions leading to coke formation proceed by a free-
radical chain mechanism, then one might envision adding a free-radical
scavenger to the oil in hopes of delaying the pyrolysis reactions and
thus, hopefully, vaporizing a greater portion of the oil.  Free-radical
scavengers are in fact used as stabilizers and oxidation inhibitors in
liquid oils at low temperatures.  However, for coke inhibition, the
agent would need to survive and be active at temperatures above 600 F.
It is not immediately obvious that this is an impossible requirement,
and perhaps such agents might be found.

       An alternative approach might be to incorporate a cracking
catalyst in the oil that would promote the formation of gaseous
species with high carbon-hydrogen ratios, olefins, cyclics, aromatics,
etc., rather than gaseous species with lower carbon-hydrogen ratios
and coke.  The requirements for such a catalyst, both physical and
chemical, may be impossible to meet.
CONTROL OF EMISSIONS FROM FUEL IMPURITIES --
POSSIBLE OPPORTUNITIES FOR ADDITIVE ACTION
       As indicated previously, additives cannot be expected to direct-
ly influence emission of elements that are present as impurities in the
fuel.  However, additives may influence the chemical or physical form
of an impure element, either reducing or eliminating its objection-
ableness or facilitating its removal from the combustion gases.  For
examble, sulfur has been added to low-sulfur coals to improve fly-ash
collection in electrostatic precipitators    ' , and iron scale and lime
were occasionally added to coal to flux the ash and correct operating
problems encountered by some boilers when firing coals with high ash-
fusion temperatures.  More recently, materials such as limestone and
                                   1-26

-------
dolomite have been used in residual-oil-fired boilers as Class III
additives to reduce the SO  content of the flue gas and avoid fireside
corrosion or acid-smut emissions.  MgO has also been so employed and
may function as either a reactant for the SO  or as an inert, non-
catalytic coating on boiler surfaces.
       There is little in the known mechanisms of pollutant formation
from fuel impurities to suggest opportunities for additive action
beyond that which has been demonstrated.  Only one suggestion for a
Class II additive can be made on the basis of chemical mechanism:  To
inhibit the reaction between SO  and 0 atoms and thus reduce the
quantity of S0_ generated.  This could most likely be done by reducing
the 0 atom concentration; such an action would be likely to have many
adverse effects on combustion processes generally.  A more useful
suggestion (constituting a Class III additive approach) might be to
search for an additive that would promote the reaction of SO  with
chemical reactants such as lime, sodium carbonate, etc., and thus
increase the efficiency of these reactants.  As the mechanism by which
these reactions occur is not known, no suggestions of logical candidate
additives can be made.

       For control of fly-ash particulate, there may be opportunity
for an effective additive (other than sulfur) that would decrease the
electrical resistivity of fly ash and thus promote its collection by
                           (44)
electrostatic precipitatorsv   .  Some combustion additives could con-
ceivably have a corollary effect in reducing fly-ash resistivity, and
an investigation of such possibilities may be warranted.
                                  1-27

-------
CONTROL OF NITROGEN OXIDES —
POSSIBLE OPPORTUNITIES FOR ADDITIVE ACTION
        In terms of the objectives of this study, namely the identification
of approaches by which pollutant emissions may be controlled with additives,
the prospects for NO emissions are very poor.  As stated earlier, one can
effect large reductions in NO by combustion modification procedures.

        For example, two-stage combustion operates through a rich initial
zone and then a lean-burning zone.  In effect, the temperature of the com-
bustion process is reduced in the rich zone and unburned hydrocarbon and CO
are oxidized, again at lower temperature, in the second,  lean, zone.   It
appears that this two-stage procedure can be moderately effective in reduc-
ing NO emissions from chemically-bound-nitrogen fuel systems also.   In this
instance, the organic nitrogen compound in the first, fuel-rich, stage of
combustion will break down via a pyrolysis path as well as via an oxidation
path.  Depending on the relative effectiveness of these two paths (i.e., the
quantity of organic nitrogen going to nitrogen RN —>• N£) two-stage combus-
                                    (45)
tion can be effective in reducing NO

        The control of NO with additives requires that the additive operate
on the Zeldovich mechanism to reduce thermal fixation reactions or on the
chemically-bound nitrogen to reduce or divert organic-nitrogen oxidation
processes.  To affect either of these processes requires that an additive
          1.  promote the recombination of 0- and/or N-atoms,
          2.  retard the production of 0- and/or N-atoms,
          3.  promote the pyrolysis of organic-nitrogen
              compounds, or
          4.  promote the reduction of NO.
Each of these approaches to NOX control is discussed below.
                                      1-23

-------
1.  Promotion of Atom Recombination

        A variety of gaseous species can promote the recombination of
0-atoms, i.e., S02, ^0; but generally, these species cannot promote the
recombination at a sufficiently significant level to noticeably reduce NO.
Heterogeneous catalysts may also be effective as third-body surface for
atom recombination, but here too, one would require a large concentration
of additive to affect the reduction, which in turn might act to the
detriment of the total combustion process.

2.  Retardation of 0-Atom Production

        Materials such as halogen compounds and the cyclopentadienyl com-
pounds are known flame retardants.  It is quite possible that,  acting in
this capacity, these inhibitors can also effect a reduction in NO emission
by interferring with the hydrocarbon oxidation process and reducing the
level of 0-atoms.  Although there is evidence that the addition of a number
of organo-metallic compounds has resulted in reduced NO, the additive was
added in most instances for other purposes (to reduce particulate) and the
effect on NO was only of secondary concern.  In most cases, it appears
that these additives were selected by chance more than by design.  A more
concentrated effort might be warranted, therefore, to understand the role
of the additive in these instances.

3.  Promotion of Pyrolysis Reactions

        An additive which might promote the pyrolysis of organic-nitrogen
compounds could be useful in reducing NO in organic-nitrogen fuel systems.
Halogens and oxygen can catalyze the pyrolysis of organic species; whether
or not such species can catalyze organic-nitrogen compounds and whether
they can direct the decomposition toward N2 is not known.  It is also
recognized that different organic-nitrogen compounds pyrolyze by different
reaction paths.  For example, CN-bonded compounds (i.e., cyanogen) yield
much less NO than NH-bonded compounds.  Why these compounds behave as they
do is not known.  One might, therefore, examine this problem in more detail

                                   1-29

-------
to determine whether specific additives might be applied for the purpose
of directing the decomposition process.

4.  Reduction of NO

        There is sound evidence that NO can be reduced in flame processes.
As discussed earlier, NO yields in organic-nitrogen flame systems

          1.  decrease as the concentration of organic nitrogen
              is increased, and
          2.  decrease as the fuel/air ratio is increased.
This reduction appears to be related to reactions of the type

                           X + NO —*- N2, N20

where X may be a nitrogen-containing radical or an organic radical.  It
follows then that ammonia, CO, or hydrocarbons might be employed as
"additives" to reduce NO.  Obviously, if these species are added to the
fuel, the effect will be similar to the combustion of organic nitrogen or
hydrocarbon under richer conditions.  On the other hand, it is also con-
ceivable that these N- or organic species can be added to the postflame
gases to reduce NO   '    .In this capacity, one is seeking additives
that will yield high concentrations of N, NHX, CN, or CHX radicals which
will attack NO by the following types of reactions:

                             N + NO = N2 + 0
                           NHi + NO = N2 + OE±

                            CN + NO = N2 + CO  .

        Exxon's patent     on ammonia addition under controlled tempera-
ture conditions apparently operates via the reaction scheme,

                           m  + NO = N  + on
                                   1-30

-------
K.VB     has also shown ammonia additions to be very effective, reducing
NO by some 90 percent when injected at temperatures of 1300 - 2000 °F to
the combustion gases in a combustion tunnel.  In a sense, these proce-
dures may be referred to under Class III "Additive Control" (See Table 1),
however, the reduction of NO is carried out so close to the region of NO
formation it is warranted to be considered in this discussion.  In a
                                            (47)
similar manner Wendt, Sternling and Matovich     have shown reduction of
803 by additions of CO or methane, and reductions of NO by additions of
ammonia or methane downstream of the flame zone.

        The chemistry and kinetics of these processes have not been explored
in detail yet.  They have to be considered as speculative processes, although
the evidence to date does suggest that such processess are occurring.
Whether and/or to what extent additives can be employed properly to affect
the conversion of NO to N2 is speculative.

        In summary, several approaches are presented here which must be
viewed with some reservation.  It is not inconceivable that additives
might be employed to control NO in combustion processess; however, it is
highly speculative that additives can be employed practically to effect
substantial reductions of NO in full-scale combustion systems.
        Experimental investigations of combustion-type fuel additives are
reviewed in Part II which follows.
                                  1-31

-------
                                PART II
                 REVIEW OF EXPERIMENTAL INVESTIGATIONS
                   OF COMBUSTION-TYPE FUEL ADDITIVES
                            SCOPE OF REVIEW
        A compilation of the additives for petroleum fuels, with a
discussion of their uses and mechanisms of action, was published 12
years ago by Lodwick    .   More recently, Finfer    ,  Agius et al    ,
and Salooja    published reviews of fuel-oil additives that were di-
rected primarily at soot and smoke emissions.   Salooja also considered
corrosion inhibitors in his review.  Pertinent data from these publi-
cations have been combined in this review with more recent information
to determine the state of the art for additives that have been investi-
gated for their effects on one or more air pollutants.  Results of
experiments conducted as part of the current EPA program are to be
                                      ( 3)
covered in detail in a separate report^  ' .
        In this review, emphasis was placed on identification of
technical literature in which quantitative measurements on additive
effectiveness are reported.  Effects of specific chemical compounds
were of interest; proprietary additives have not been considered
unless the compounds that are major components of the additive are
known.  Some data have been included from studies in which only the
metal component of the additive was designated, provided the data were
particularly pertinent to the discussion.
                                  II-l

-------
        Most of the additive studies reported in the literature were
carried out in the laboratory under controlled conditions, but there
are also a significant number of full-scale boiler tests, mostly with
oil firing.  Sulfur trioxide received the greatest attention in past
years because it was chiefly responsible for the corrosion and deposit
problems.  Recently, attention has been directed to particulate,
nitrogen oxides, and polycyclic organic materials.

Summary Tables

        Tables 2, 3, and 4 summarize the quantitative data found in
the technical literature for combustion additives, grouped according
to chemical type of the additive:
           Table 2.  Effects of Organometallic Additives on
                     Combustion Emissions(2 pages)
           Table 3.  Effects of Inorganic Additives on Com-
                     bustion Emissions
           Table 4.  Effects of Organic Additives on Com-
                     bustions Emissions.
For each additive or chemical compound, the following information is
listed under the applicable pollutant:  fuel type, additive concentra-
tion, pollutant reduction, and reference number.   Additive concentra-
tions generally have been converted to volume percent or weight percent
(or are in the units originally reported).   Data on pollutant effects
have been expressed for this review as "percent reduction" for each
pollutant affected; thus, "0" denotes that the additives showed no
effect on the pollutant in question.

        Most of the data in the tables are based on experiments with
boilers or laboratory simulations of boiler or furnace firing.   Obser-
vations applying to gas turbines and 1C engines are noted in the tables.
                                  II-2

-------
                              TABLE  2.   EFFECTS  OF  ORGANOMETALLIC ADDITIVE  COMPOUNDS  ON  COMBUSTION  EMISSIONS

ADDITIVE
Naphthenates
Cobalt

Manganese

Iron
Lead


Zinc


Nickel
Magnesium
Copper
Ferrocene








PARTICULATES
Fuel Additive Reduc- Ref.
Type cone tlon,
percent

No. 2 0.003 w/o metal 70 38
No. 2 0.01 w/o 15 53
No. 2 0.003u/o metal 65 38
No. 2 0.02 u/o 17 53
No. 2 0.01 u/o 50 54
No. 2 0.003u/ometal 30 38
No. 2 .01 u/o 10 53
.05 u/o 22 53
.

-
No. 2 0.02 w/o 25 53
No. 2 0.02 w/o 7 53
No. 2 0.01 u/o 10 53
No. 6 0.003 u/o 37 52
No. 2 0.012 u/o 47 1
No. 2* 0.1 u/o 45 60
0.2 u/o 82 60
No. 2 0.01 u/o 28 53
0.05 u/o 45 53
I'C 0.001 w/j NS 71
I'+A 1.5**" 28 57
2.5«* 30 57
POLYCYCLIC ORGANIC MATTER
(as BaP or Total PAH)
Fuel Additive Reduc- Ref.
Type cone tlon,
percent

-
.
: : : :
-
- -
.
.
- -
.
~ -
.
.
.
.
.
.
.
- -
.
.
-
.
-
SULFUR TRIOXIDE
(unless S02 Indicated)
Type cone tlon,
percent

No. 2 0.5 v/o 7 89
1.0 v/o 10 89
: : : :
No. 2 0.5 v/o 7 89
1.0 v/o 10 89
- -
No. 2 0.5 v/o 0 89
1.0 v/o 0 89
-
- -
-
No. 6 0.2 w/o 100 77
No. 6 0.07 u/o 75 78
0.14 u/o 100 78
-
No. 6 0.15 u/o 100 77
No. 6 0.07 u/o 100 80
No. 6 0.003 u/o 0 52
No. 2 0.012 u/o 0 1
-
-
-
.
-
-

NITROGEN OXIDES
Type cone tlon,
percent

No. 2 0.5 v/o 10 89
1.0 v/o 15 89
J-A* 0.2 u/o metal 25 101
0.5 u/o metal 30 101
No. 2 0.5 v/o 10 89
1.0 v/o 10 89
J-A* 0.2 u/o metal 20 101
0.5 w/o metal 30 101
No. 2 0.5 v/o 6 89
1.0 v/o 9 89
-
-
.
- -
-
-
-
-
J-A* 0.2 w/o metal 16 101
0.5 u/o metal 18 101
No. 6 0.003 u/o 0 52
No. 2 0.012 u/o 0 1
.

-

• •
-

 I
U)
                           Equipment

Boiler, furnace or continuous-combustion laboratory  apparatus, unless
noted by asterisk (* or **).

*  Experiments conducted on gas-turbine combustor.

** Experiments conducted on reciprocating 1C engine (or simulation bomb).
           Fuel

 No. 2 or No. 6 Fuel Oil
 PC  -- Pulverized Coal
 BC  -- Bituminous Coal
 SC  -- Stoker Coal
 CC  -- Coal Gas
J-A  -- Jet A liquid fuel
I'-*-A  -• Pentane + Acetylene
**
-------
                         TABLE  2.    EFFECTS OF  ORGANOMETALLIC ADDITIVE  COMPOUNDS  ON COMBUSTION  EMISSIONS  (Continued)

ADDITIVE

Iron carbonyl

CsHs -Mn-(CO)3

CH3C5HJ -Mn-(CO)3


Tetraethyl lead
Calcium sulfonate
Barium sulfonate
Copper Sulfonate
Ba dlalkyl
phoaphates
Ba + Zn ethyl
hexanoatea
Nickel carbonyl
Chromium carbonyl
Mn veraatate
Co versatate
Zr neodecahoate
Copper chelate
011-aol Ba Compound
PARTICULATES

Fuel Additive Reduc- Ref.
Type cone tlon,
percent
No. 6 0.003 u/o 40 52
P+A 1.5*** 32 57
2.5*** 47 57
No. 2** 0.1 u/o 14 60
0.2 y/o 18 60
PC 0.1 u/o NS 71
P+A 1.5*** 80 57
2.5*** 82 57
No. 2 0.012 w/o 80 38
No. 2 0.01 u/o 44 1
No. 2 0.01 u/o 5 53
0.05 u/o 20 53
No. 2 0.03 u/o 45 38
No. 2 0.2 u/o 66 63
No. 2 0.01 u/o 50 54
No. 2 0.26 v/o 28 39
No. 2»* 0.4 u/o 85 59
No. 2 0.009 u/o 40 38
No. 2 0.013 u/o 60 38
No. 2** 1 u/o 10 64
No. 2** 1 u/o 10 64
.
No. 2 0.0075 u/o metal 50 66
0.001 u/o metal 50 66
No. 2** 0.05 u/o 49 65
0.18 u/o 80 65
POLYCYCLIC ORGANIC MATTER
(as BaP or Total PAH)

Fuel Additive Reduc~ Ref.
Type cone tlon,
percent
P+A 1.5*** 32 57
2.5*** 50 57
.
.
.
P+A 1.5*** 81 57
2.5*** 97 57
.
-
.
-
-
-
-
_
.
-
Gasoline** 1 u/o 66 64
.
.
.
No. 2** 0.1 u/o 0 65
SULFUR TRIOXIDE
(unless S02 Indicated)

Type cone tlon,
percent
No. 6 0.003 u/o 0 52
-
-
.
No. 2 0.01 u/o 0 1
No. 6 NS 45 88
-
.
-
-
-
_
.
-
-
-
.
.
.
-
NITROGEN OXIDES

Type cone tlon,
percent
No. 6 0.003 u/o 0 52
-
.
-
No. 2 0.01 u/o 0 1
-
-
-
-
-
-
* - - -
_
-
-
-
-
J-A* 0.1 u/o metal 11 101
0.2 u/o metal 22 101
-
No. 2** 0.1 u/o 0 65
H
H
                                     Equipment
          Boiler, furnace or continuous-combustion laboratory apparatus, unless
          noted by asterisk (* or **)
          *  Experiments conducted on gas-turbine combuotor.
          ** Experiments conducted on reciprocating 1C engine (or simulation  bomb).
          Fuel

No. 2 or No. 6 Fuel Oil
       Pulverized Coal
       Bituminous Coal
   --  Stoker Coal
       Coal Gas
   --  Jet A liquid fuel
       Pentane + acetylene
       (concentrations are In
        g tnetal/t vapor)
 PC
 BC
 sc
 CG
J-A
Additive Concentration

v/o  --  volume percent
u/o  --  weight percent
 NS  --  not specified
        Pollutant
BaP  --  Benzofajpyrene
PAH  --  Polynuclear aromatic
        hydrocarbons

-------
                                     TABLE  3.    EFFECTS  OF  INORGANIC  ADDITIVE  COMPOUNDS ON  COMBUSTION  EMISSIONS

ADDITIVE

Mg metal

MgO
MgO + AljOj
MgC03
CaO
Dolocalte




C»(N03)2
CaCl2
Zn Metal
ZnO

NaOH
Na2C03
NaHC03
Ba(M>3)2
SrCl2
Mn02

S10
S102
PARTICULATES
Fuel Additive Reduc- Ref.
Type cone tlon,
percent

.
.

No. 6 0.05 v/o 6 51
.
.
.
-
.
-
-
Propane 0.2 w/o 22 40
0.1 w/o 25 41
.
.
.
-
.
: : : :
Propane 0.012 w/o 60 40
Propane 0.074 w/o 54 40
No. 6 0.004 w/o 15 52
Pitch 0.007 w/o 77 68
. ' - - -
.
POLYCYCLIC ORGANIC MATTER
(as BoP or Total PAH)
Fuel Additive Reduc- Ref.
Type cone tlon,
percent
.
.
-

.
-
-
.
- -
.
- -
.
-
-
.
.
.
.
.
-
-
-
.
-
.
-
SULFUR TRIOXIDE
(unless S02 Indicated)
Type cone tlon,
percent
No. 6 NS 60-75 86
No. 6 0.02 w/o 100 87
No. 6 0.35 w/o 100 77

No. 6 0.05 v/o 100 51
No. 6 0.05-0.1 w/o 75 86
PC NS 50 93
No. 6 1 w/o 99 103
No. 6 0.8 w/o 100 77
No. 6 0.1 w/o 42 83
0.2 w/o 50 83
No. 6 0.15-0.2 w/o 80 84
BC 0.375 w/o 50 92
-
- -
No. 6 0.45 w/o 100 77
SC 0.25 u/o 100 94
CG 0.0045 g/i 85 97
PC NS 47(S02) 93
No. 6 3.3 w/o 36(50*) 2
SC 14 w/o 70(S02) 95
3.4 w/o 23(502) 95
-
-
No. 6 0.004 w/o 0 52
* "
CC 2 gr/cu ft 80 97
N'o. 6 O.M w/o 50 77
NITROGEN OXIDES
Type cone t Ion ,
percent
.
-
.

No. 6 0.05 v/o 43 51
-
.
-
-
_
_
- -
-
•
-
.
-
J-A* 0.1 w/o metal 16 101
J-A* 0.1 w/o metal 26 101
.
-
-
No. 6 O.OOi w/o 0 52
.
-
-
 I
Ln
            Boiler,  furnace or continuous-combustion laboratory apparatus, unless
            noted by asterisk (* or **).
            *   Experiments conducted on gas-turbine conbustor.
            **  Experiments conducted on reciprocating 1C engine (or simulation bomb).
 No. 2 or No. 6 Fuel Oil
 PC  --  1'ulverlzed Coal
 BC  --  Bituminous Coal
 SC  --  Stoker Coal
 CC  --  Coal Gas
J-A  --  Jet A liquid fuel
I'+A  --  I'entane •*• Acetylene
***      (concentrations are In
         g metal/.' vapor)
Additive Concentration
v/o --  volume percent
w/o --  weight percent
 NS --  not specified
Bap  --  Benzo|flIpyrene
PAH  --  Polynuclear arora
        hydrocarbons

-------
                                       TABLE  4.    EFFECTS  OF  ORGANIC  ADDITIVE  COMPOUNDS  ON COMBUSTION  EMISSIONS

ADDITIVE

n- Butanol
n- Pentanol
Nitroethane



Nlcropropane



Methanol

Ethylene glycol-
dltnechyl ether
c-Butyl hydroper-
oxlde


Dl t-butyl peroxide
Hydrazlne
1- Amylnltrace

n-Butyl mercapcan
Formaldehyde deriv.
Lube oil

PART1CULATES
Fuel Additive Reduc- Ret.
Type cone tlon,
percent


.
.
.
_
-
No. 2** 2 v/o 30 58
5 v/o 34 58
-
.
_
.
No. 2** 1.87 u/o 49 62

No. 2** 1 v/o 50 58
2 v/o 57 58
-
.
.
No. 2 0.05 w/o 50 61
.

.
No. 2*" 1 w/o 10 58
.
-
POLYCYCLIC ORGANIC MATTER
(as BaP or Total PAH)
Type cone tlon,
percent


.
Propane 8.5 w/o 83 72
17 w/o 88 72
No. 2** 2 v/o 51 58
5 v/o 64 58
Propane 20 w/o 90 58
-
No. 2** 2 v/o 60 58
5 v/o 75 5d
Ethylene 12.7 w/o 15 73
25.4 w/o 18 73
.

I'ropane 20 w/o 92 58
-
No. 2** 1 v/o 85 58
2 v/o 90 58
.
.
-
_
.
-
C.asollne- 3 v/o 40 74
oil mix** 5 v/o 100 74
SULFUR TRIOXIDE
(unless S02 Indicated)
Type cone tlon.
percent
_

-
-
-
.
-
-

.
- - -
.
-
-

-
.
.
-
.
.
-
.
.
-
-
.
NITROGEN OXIDES
Fuel Additive Reduc- Ref.
Type cone tlon,
percent
No. 2 1 v/o 10 100

No. 2 1 v/o 15 100
-
.
-
-
-

-
-
.
™ ~
-

-

.
-
No. 2** 0.5 w/o 9 102
1.5 w/o 14 102
-
No. 2** 0.42 w/o 9 102
1.8 w/o 17 102
J-A* 37 ppm/u 15 101
J-A* 2.6 w/o 11 101
_
-

 I
ON
                                       llqutpnent

           Boiler,  furnace or continuous-combustion laboratory apparatus, unless
           noted by asterisk  (* or *•*) .

           *   Kxpertments conducted on gas-turbine comhustor.
           **  Kxperlmeits conducted on reciprocating  1C engine for simulation bomb).
            Kuel

 No. 2 or No. 6 fuel  Oil
 I'C  --  Pulverized Coal
 BC  --  Bituminous Coal
 SC  --  Stoker Cool
 CC  --  Coal Cas
,1-A  ••  -let  A liquid fuel
l'+A  --  I'eneane + Acetylene
-.vttv:      (concentrat Ions are
         g rneta 1 /• vapor)
Additive Concentration

v/o --  volume percent
u/o --  weight percent
 NS --  not  spec!fled
         Pollutant

Bal1  --  Benzols Ipyrene
I'AH  --  I'olynuclear aromatic
        hydrocarbons

-------
Organization of Part II Review Discussion

         In the report sections which follow, experimental data and
some qualitative comments on additives reported in the literature are
discussed according to observed effects of combustion-type fuel addi-
tives on emissions of the four pollutant classes:
           •  Particulate and Smoke
           •  Polycyclic Organic Matter (POM)*
           •  Sulfur Oxides
           •  Nitrogen Oxides.
Separate sections covering effects on these pollutants follow.
          ADDITIVE EFFECTS ON PARTICULATE EMISSIONS AND SMOKE

         Many of the fuel-additive investigations reported in the litera-
ture have been directed at reduction of particulate emissions.  The
attention has centered primarily on fuel-oil combustion, because it has
been demonstrated that additives can be of value in this area.  For
coal-fired systems, most of the fuel additive work has been designed to
reduce corrosion and deposits, and any effect on particulate emissions
has been observed only as incidental  to  these other objectives.

COMBUSTION ADDITIVES IN FUEL-OIL

         The interest in reducing smoke from residential or commercial
heating units, from gas turbines and from vehicles powered by diesel
engines has provided impetus for much of the study of additives.  Con-
sequently, much of the work has been done with distillate fuels, with
only occasional investigation of heavier oils.
*  POM emissions are frequently reported in the literature as benzo[a]-
   pyrene  (BaP) determinations, or as polynuclear aromatic hydrocarbons
   (PAH) as a class.  These are the  terms used in Tables 2, 3, and 4.
                                   II-7

-------
Investigations Firing Residual Oil in Boilers

        In the course of their study of residual oil additives in low-
pressure heating boilers, Lee and his associates    included measure-
ments of particulate emissions.  In a research boiler fueled with No. 6
oil, they found only a slight reduction in particulates when using a
hydrated magnesium oxide-aluminum oxide mixture as an additive.  This
additive was designed to control SCL and consequent sulfuric acid
corrosion, and it was found to bring about only a 6 percent reduction
in particulates when added to the oil in concentrations of 0.05 to 0.1
volume percent.

        On the other hand, work at the New England Power Company was
aimed at reducing particulate emissions from an oil-fired power
        (52)
station.     These tests were carried out in a 250 megawatt unit at
the Brayton Point Station just after it was put in service following
the annual maintenance shutdown, so that the additive performance was
evaluated in a clean boiler.  A 40-percent reduction in particulate
emissions was obtained with an addition of 8 ppm of iron in the form
of iron carbonyl.  The same amount of iron added as ferrocene resulted
in 37 percent reduction in particulates.  Another additive tested was
an inorganic formulation containing manganese dioxide, but this addi-
tive provided only a 15 percent reduction in particulates although it
was used at a concentration of 25 ppm manganese in the oil.  It also
was noted in this study that the amount of particulate reduction at a
constant additive level was a function of the oil viscosity as well,
and in the case of the iron carbonyl additive, the particulate loading
went through a minimum at a viscosity of 100 SSU and increased sub-
stantially at higher viscosities of the oil.  (This may be due to the
fact that the carbon particulate emission level is influenced by oil
viscosity as fired.)
                                   II-8

-------
Investigations Firing Distillate Oil
in Furnaces, Boilers, and Gas Turbines

        Martin, Pershing, and Berkau     studied the effect of some
200 different materials as additives to No.  2 fuel oil in the EPA
experimental furnace with a conventional high-pressure atomizing, gun-
type burner.  Fewer than 10 percent of the fuel additives produced any
pollutant reduction; in a few cases, proprietary organo-metallic addi-
tives substantially reduced particulate emissions.

        In this work, ferrocene was used at  a concentration of
0.012 weight percent and a particulate reduction of 47 percent was
observed.  The methyl derivative of cyclopentadienyl manganese tri-
carbonyl also was investigated in this study, and a particulate
reduction of 44 percent was observed at an additive concentration of
0.01 weight percent.  It is very significant that Martin et al found
that when additive concentrations were greater than 0.01 weight percent,
the amount of particulate began to increase as a result of the increas-
ing concentration of additive metal.

        The authors raise  the question about possible toxicity of the
resulting metallic emissions.  They conclude  that the use of burner
designs having improved air/fuel mixing can achieve greater carbon
particulate reduction and is recommended over the use of distillate
fuel additives for residential heating applications.

        The greatest reduction in particulates when firing No. 2 oil
                                                                  (38}
in a domestic heater unit was reported by Riggs and his colleagues
The addition of methyl cyclopentadienyl manganese tricarbonyl at a rate
of 0.1 gram of metal per gallon of fuel resulted in an 80 percent
reduction in particulates.  Cobalt and manganese naphtenates were only
slightly less effective, providing reductions of 70 percent and 65
percent respectively at the same concentration in the oil.  Chromium
carbonyl was the next most effective with 60 percent reduction in
                                   II-9

-------
particulates.  The three other compounds studied were less effective,
the percent reduction being:  calcium sulfonate, 45 percent; nickel
carbonyl, 40 percent; and lead  napthenate, 30 percent.
        Weeks and his associates    ' used a conventional gun-type oil
burner firing No. 2 oil in a typical domestic steam boiler for their
additive study.  Ferrocene was the most effective additive that they
found, with particulate reductions ranging from 28 to 45 percent as
the additive concentration increased from 100 to 500 ppm.  The other
additives that they studied did not reduce the particulates nearly as
effectively as did the ferrocene.  The percent reductions in particu-
lates obtained with the various napthenates were as follows:  nickel,
25; iron, 22; manganese, 17; cobalt, 15; copper, 10; and magnesium, 7.
Tetraethyl lead also was used in a concentration of 500 ppm and pro-
vided a particulate reduction of 20 percent.

        Organometallic additives also were investigated by Vaerman^)
who worked with a domestic steam boiler burning No. 2 oil.  At a con-
centration of 0.01 weight percent, both manganese napthenate and copper
sulfonate reduced the particulate emissions by 50 percent.  Vaerman
also noted that the efficiency of the additive was a function of the
C0? content of the combustion gases as well;  he concluded that parti-
culate emissions for domestic units could be kept within acceptable
limits by burner adjustment, and additives would not be necessary.

        Full-scale tests on power-station gas turbines fueled with No. 2
oil were conducted by Plonsker and his colleagues , '  ' using methyl
cyclopentadienyl  managanese tricarbonyl additive.  Turbines at three
utilities were included in this program, and reductions in smoke,
particulates, and carbon emissions ranged from 50 to 90 percent at
manganese concentrations of 20 to 100 ppm.  In general, the greatest
emission reductions were observed at higher turbine load conditions and
higher manganese concentrations.
                                  11-10

-------
        Organometallic compounds of barium, manganese, iron, lead, and
                             (56)
boron were tested by Shayeson     in JP-5 jet fuel.  The Ba, Mn, and Fe
compounds eliminated smoke in laboratory tests at concentrations of 0.5
volume percent, 0.08 volume percent, and 0.05 weight percent, respec-
tively.  Lead and boron compounds were ineffective.  At lower concen-
trations, Mn was most effective.  At 0.04 volume percent, the Mn
compound eliminated smoke in a full-scale engine test.  However, in a
flight test, as much as 0.1 volume percent would not completely prevent
smoke formation.
Laboratory Scale Investigations Firing Various^
Fuels in Continuous Combustion Apparatus
        In a series of experiments with diffusion flames in which
mixtures of pentane and acetylene were burned, Spengler and Haupt
found that methyl cyclopentadienyl manganese tricarbonyl was very
effective in reducing the amount of particulate formed, although the
additive was used at a relatively high concentration.  Eighty percent
reduction in particulates was achieved with an additive concentration
containing 1.5  grams  of metal per liter of combustible vapor.  Ferro-
cene and iron carbonyl were only about half as effective at the same
metal concentration.

        The effects of 40 metals on the amount of soot emitted by a
laboratory-scale propane diffusion flame were investigated by Cotton
et al     .  The metals were added to the flames by atomizing aqueous
solutions of their salts into the fuel with a stream of nitrogen.  The
most effective compound found in their work was barium nitrate, with
which a 60 percent reduction in soot occurred with enough additive to
provide a carbon-to-barium ratio of 3700.  Calcium nitrate and strontium
chloride were found to be slightly less effective.  Some soot reduction
also was obtained with ammonium molybdate and sodium tungstate, but all
the other metals tried were ineffective.  The authors concluded that in
                                   11-11

-------
terms of the metal concentrations required and the equivalence ratio at
which the additive was effective, only the three alkaline earth metals
could be considered as promising additives.

        Cotton et al     hypothesize from this study that metal addi-
tives reduce carbon emission by two basic mechanisms (a) the alkaline
earth metals, and probably molybdenum, catalyze the production of free
radicals via the decomposition of hydrogen or water, and (b) other
metals probably catalyze the oxidation of soot in cooler parts of the
flame after being incorporated with the soot particle.

        Friswell's     later studies in a rig simulating a gas-turbine
combustor with Ba and Mn additives in jet fuel tend to support these
hypotheses.  Support for the radical production mechanism is also found
in the observation that NO production was increased at the same time
that Ba addition reduced soot formation.  Mn, and also Fe, appeared to
operate by catalyzing the burn-out of the soot.  In accord with these
hypotheses, Mn and Fe operated better as the air/fuel ratio approached
the stoichiometric level.

                                                  {I *5 \
        Alkali metals were investigated by Feugier     for their efrec-
tiveness in reducing soot formation in ethylene flames.  Nitrate and
chloride salts of the metals were volatilized and carried into the
flame in a stream of nitrogen.  At the highest flame temperatures
lithium reduced the soot formation while cesium, potassium, and sodium
increased it.  As flame temperatures were lowered, these three metals
gave small reductions in emissions at low metal concentrations, but not
at high concentrations.
                                   11-12

-------
Investigations Firing Distillate Oil in I.C. Engines

        In their work on polycyclic aromatic hydrocarbons, Ray and
    (59)
Long     also measured the effect of additives on the particulate
formation in diesel engines.  They employed fairly large amounts of
the additives, and observed a 57 percent reduction in particulates
when tertiary-butyl hydroperoxide was added at a rate of 0.5 volume
percent.  They found that 1-nitropropane gave only a 34 percent reduc-
tion in particulates when applied at 5 volume percent.   It also was
noted in this series of experiments with additives that the amount of
carbon residue remaining in the engine remained constant even through
the particulate matter emitted from the system was reduced by the
additives.  On this basis, the author suggested that different reac-
tion mechanisms are responsible for the formation of particulates and
of the carbonaceous residue remaining in the engine.

                                                     (59)
        A recent patent by Cities Services Oil Company^   ' claimed a
very significant reduction in particulate emissions from distillate
fuel burned in a diesel engine.  In this case, a solution of barium
and zinc 2-ethyl-hexanoates (10 to 1 barium to zinc by  weight) with
methyl cellosolve in a hydrocarbon reduced the smoke number reading by
an amount equivalent to a reduction of about 85 percent in the particu-
late emission.  (Particulate loading is not a linear function of smoke
number.)

        Experiments conducted in a constant-volume bomb which simulated
the conditions obtained in a diesel engine were reported by Chittawadgi
and Volnov     .  When ferrocene was added to the fuel  in a concentra-
tion of 0.2 weight percent, particulates were reduced about 80 percent.
The smoke reduction dropped to about 45 percent when the ferrocene
concentration was cut in half.  These investigators also tried cyclo-
pentadienyl manganese tricarbonyl as an additive but obtained only 18
percent reduction in particulates at an additive level  of 0.2 weight
percent.
                                  11-13

-------
           Other organic additives are reported to have been effective
in reducing particulate emissions in the combustion of No.  2 oil in a
diesel engine.  Churchill and Mitchell      reported that 0.05 weight
percent of hydrazine gave a 50 percent reduction; a patent granted to
                              (r y\
the Cities Service Oil Company     also claimed 50 percent smoke
reduction by the use of dialkylethers of ethylene glycol when No. 2 oil
was burned in a single cylinder diesel engine.
        Salts of organic acids have been used for smoke reduction.
                //-•IN
Barium sulfonate^  ' , the barium and strontium salts of dialkyl ortho
phosphoric acids v  ' , and the manganese and cobalt salts of versatic
acids     have all been used with some measure of success.
Additional Investigations
        There are some additional reports in the literature of effec-
tive performance by fuel additive compounds which are not completely
characterized.  The oil-soluble barium compound used by Golothan     ,
which was mentioned earlier, reduced the particulate in diesel engine
exhausts by 80 percent when used at a concentration of 0.18 weight
percent.  Amick and Hunt     formed a copper chelate by reacting cupric
acetate with a mixture of paraformaldehyde, dodecylphenol, and ethylene-
diamine; when this chelate was added to distillate oil at a concentration
of only 0.001 weight percent copper, a 50 percent reduction in particu-
late was obtained.  Summaries of various additive field trials by
Kukin'  '   ' have claimed substantial reductions of particulate emission
by using a proprietary formulation that is primarily an organo-manganese
compound, either alone or in conjunction with MgO.
                                   11-14

-------
       Survey Conclusions Related to Use of
       Combustion Additives for Reducing
       Particulate or Smoke Emissions From
       Fuel-Oil Firing
            Experimental evidence thus indicates that additives
       can be effective in lowering the carbonaceous portion of
       particulate emissions.   They are particularly useful in
       oil-fired systems, because carbon constitutes a large per-
       centage of the particulate emission.   Oil-soluble compounds
       of certain transition metals (manganese, iron, nickel, and
       cobalt) or certain alkaline-earth metals (barium and cal-
       cium) have proven to be the most effective additives for
       reducing particulate emissions.


COMBUSTION ADDITIVES IN COAL FIRING

        Very little effort has been devoted  to the use of additives for
the reduction of particulate formation in the combustion of coal.  In
past years, electrostatic precipitators or mechanical collectors have
been used on large coal-fired equipment to reduce particulate emissions.
(Many smaller coal-fired units have had no collectors.)  From an ambient
air-quality standpoint, further reductions in particulate emissions
from coal firing are desirable.

        The relatively large inorganic content of coal results in a
certain irreducible minimum value of the particulate that will be
formed.  It is only the carbonaceous part of the particulate that will
be reduced by the action of combustion fuel  additives.  Thus, it may
not be economical to consider additives for  such a purpose, unless a
marginally-operating boiler that generates particulate emissions slightly
exceeding regulatory limits can be brought into compliance by a small
reduction in particulate.  Also, the reduced carbon loss of the boiler
could be advantageous to overall efficiency.
                                  11-15

-------
        Only a few investigations of coal additives were identified
in the survey, as noted below.
Early Investigations on Coal Additives
by the U.S. Bureau of Mines
        One of the earliest investigations of the effectiveness of
additives was made at the U.S. Bureau of Mines by Nicholls et al     ,
to determine whether there was any merit in intermittent use of soot
removers in coal-fired residential furnaces to keep heat transfer
surfaces and stacks clean.  Soot deposits from bituminous coal were
treated with additives in a laboratory furnace to determine the
                                *
ignition temperature of the soot ; those materials that performed best
were then compared in coal-burning home furnaces with fixed fuel beds.
The results showed that of the many inorganic compounds used, copper
compounds were the most effective in lowering the ignition temperature
of soot and thereby promoting its combustion.  The metal chlorides
tried were ranked in decreasing order of effectiveness:  CuCl , PbCl ,
NiCl2, SnCl2, NH^Cl, KC1, FeCl^ MnCl^ ZnCl2> Ca(OCl)2, CaCl^ and
NaCl.  It was found that the benefits of CuCl  addition also could be
obtained by using a 50-50 mixture of NaCl and CuSO, .  No measurements
were made of stack particulate emission.
        Later, a broad experimental program was carried out over a
2-year period at the Bureau of Mines under the direction of Nicholls
to investigate the effects of inorganic compounds on the combustion of
coal and coke.  Twenty-five compounds mixed with coal were included in
the study; and NaCl, CaCl , and Na CO  received the most attention.
Eleven coals and four cokes were used as fuels.  The fuels were burned
*  It should be noted that such additives that function downstream of
   the  flame are not "combustion-type" fuel additives and are properly
   categorized as Class III additives.
                                   11-16

-------
        Reduction of particulates in pulverized coal firing also was
claimed by Kerley^  ' who added from 0.001 to 5.0 weight percent of
ferrocene or cyclopentadienyl manganese tricarbonyl to the coal during
the grinding process.  The extent to which the particulates were
reduced was not specified.
        There is little likelihood of combustion-type fuel additives
making significant reductions in particulate emissions from large
pulverized coal burning equipment which normally emits only small
quantities of carbon particulate and which in any event is usually
equipped with electrostatic precipitators.  In some circumstances,
additives may be useful in enhancing the performance of the precipi-
tators through decreasing the particulate resistivity, but these are
generally added at the precipitator inlet and are considered Class III
additives.
        Survey Conclusions Related to Use of Combustion
        Additives for Reducing Particulate Emissions
        From Coal Firing
             For large pulverized coal-fired equipment, fuel
        additives appear to hold little promise for directly
        reducing particulate or smoke emissions.

             In smaller coal-fired equipment where the carbon
        content of the particulate emissions is high and me-
        chanical collectors may be used rather than electro-
        static precipitators, the potential of additives for
        reducing carbon particulate emissions may be greater.
        This could be particularly true for stoker firing,
        where much of the ash is retained in the furnace.
                                  11-18

-------
              ADDITIVE EFFECTS ON POLYCYCLIC ORGANIC MATTER

         A class of hazardous pollutants identified as polycyclic
 organic matter (POM) was defined in 1972 by the National Academy of
 Sciences     .   The terminology, POM, now is coming into wider use.
 Broadly, this class of pollutant compounds includes (1)  polynuclear
 aromatic hydrocarbons, commonly referred to in the literature as PNA
 or PAH, and (2) nitrogen-containing heterocyclic compounds.

         Although POM is usually found in the atmosphere  only in trace
 concentrations, it is of concern because of the highly carcinogenic
 nature of some  of its compounds.   The National Academy of  Sciences  has
 rated  many of the individual  compounds according to their  currently
 known  carcinogenicity    .  Benzo[a]pyrene  (BaP),  one of  the highly
 carcinogenic  compounds,  is  sometimes  used as  a measure of  total PAH or
 POM.   Generally,  POM emissions  are  the result  of incomplete  combustion
 and, thus,  depend on combustion conditions.  Hangebrauck and his
 associates      have  characterized POM emissions  from a variety  of
 fossil-fuel combustion  systems  but  did not  investigate fuel  additives.

        Experimental  investigations aimed at the reduction in POM
 emissions from  combustion processes by additives have claimed signifi-
 cant results.  Much of  this work has been on gas flames and  on diesel
 engines burning distillate oil, and it is questionable as to how much of
 the results are applicable to combustion of heavy oils or coal.  No
 investigations were identified where additive effects on  POM emission
were reported for residual oil or coal firing.  There is  the additional
question as to the adequacy of sampling and analytical techniques,
inasmuch as techniques in these areas have advanced considerably in the
past few years.
                                   11-19

-------
Laboratory Investigations in Continuous Combustion Apparatus
        Significant work has been reported by Ray and Long     who
studied the effects of several additives on the production of PAH in
propane diffusion flames and a diesel engine.  Dichloromethane as an
additive increased the formation of soot and carbonaceous residue to a
considerable extent without having any appreciable effect on the
formation of BaP.  However, additions of nitropropane brought about a
90 percent reduction in PAH production in the propane diffusion flames
(and up to 75 percent reduction when burning No. 2 oil in a single-
cylinder diesel engine).  However, 5 volume percent of the nitropropane
had to be added to achieve this amount of reduction.  The same amount
of nitroethane additive resulted in 64 percent reduction in PAH levels.
The most effective agent that these investigators found was tertiary-
butyl hydroperoxide which provided 92 percent reduction in PAH emission
with the propane flame  (and 90 percent reduction when used in a concen-
tration of 2 volume percent in the diesel fuel) .

        A substantial reduction in the formation of BaP also was noted
by Long and Ray'  ' when they added nitroethane  to commercial propane
which was burned as a laminar diffusion flame in air.  The air supply
was reduced to give heavy soot formation in the absence of the addi-
tives, and under these  circumstances it was found that as little as
0.05 moles of nitroethane per mole of fuel brought about an 83 percent
reduction in the BaP formation.  Doubling the concentration of the
nitroethane increased the percent reduction only to 88 percent.  These
authors found that nitroethane reduced the BaP  formation to a greater
extent in the smoky propane flames than it did  in the No. 2 oil
mentioned previously.

        The effect of methanol additions on ethylene diffusion flames
was investigated by Chakraborty and Long'73) on the basis that only
very small amounts of unsaturated compounds had been found in a methanol
                                  11-20

-------
diffusion flame.  It was found that a modest reduction in soot, carbon-
aceous residue, chloroform soluble material, and total polycyclic
aromatic hydrocarbons resulted.  However, the PAH reduction amounted to
only 18 percent when the menthanol concentration was 0.25 moles per mole
of ethylene.

        Significant reduction in polycyclic aromatic hydrocarbon forma-
tion also was observed by Spengler and Haupt ^   ' who studied the  effects
of additives on the combustion of mixtures of pentane and acetylene.
The most effective additive was  raethyl-cyclopentadienyl manganese
tricarbonyl, which reduced the BaP production by 97 percent when added
at a rate of 2.5 grams of metal per liter of combustible vapor.  While
BaP production was greatly inhibited under these conditions, it must
be noted that a relatively high concentration of the additive was used.
When the additive concentration was reduced to 1.5 grams of metal per
liter of combustible, the reduction in BaP was only 81 percent.  Iron
carbonyl was found to be less effective as an additive under their
experimental conditions, providing only 50 percent reduction in BaP
formation when added in a concentration of 2.5 grams of metal per liter
of combustible vapor.

Investigations in I.C. Engines

        A two-stroke gasoline engine was utilized by Kuhn and
Tomingas '   ' to study the effect of oxidation catalysts on the  formation
of BaP.  The BaP in the engine exhaust was reduced by 66 percent by the
addition of 1 weight percent of the manganese salt of versatic acid,
which is a multibranched synthetic fatty acid containing C  to C
carbon chains.  In other experiments, it was found that no BaP was
produced when diisopropyl  ether  was used as a fuel.
                                 11-21

-------
        Other studies have been reported in which the active agent
either was not completely designated or was difficult to identify.
                                             (74)
Thus, for example, Hunigen and his colleagues     reported on the
content of BaP in the exhaust gases of two-stroke engines.  They
found that the lubricating oil in the gasoline-oil mixture which was
burned exercised an important influence on the BaP production.  When
the lubricating oil content of the mixture was 3 percent by volume,
the BaP content of the exhaust gases was 40 percent less than without
the lubricating oil.  When the lubricating oil concentration was
increased to 5  volume percent, the BaP formation was completely elimi-
nated under the combustion conditions that were studied.  The nature of
the oil was not specified, and it is possible that some metal-contain-
ing compound in the oil was the active ingredient.

        Negative results were obtained by Golothan     when an oil-
soluble barium compound was added to diesel-engine fuel.  There was
no reduction in the level of PAH in the exhaust gases when the additive
was used at a concentration of 0.1 weight percent.

        In a general discussion of fuel additives, Van Den Heuvel
claimed that the addition of metal-containing combustion catalysts to a
fuel would reduce the soot quantity produced, and also .would reduce the
concentration of PAH.  In this connection, alkali metal compounds were
mentioned as being most effective because the metal ions produced
prevent the agglomeration of large soot particles from smaller particles
of unburnt carbon.  However, the author did not cite any quantitative
data to support this contention.
                                  11-22

-------
Survey Conclusions Related to Use of
Combustion Additives for Reducing
Emissions of POM
     Although only light oils and gases were involved
in these research programs on POM, significant reduc-
tion in the POM production was reported in a number of
cases.  However, these data from gas flames and
internal combustion engines may not be applicable to
boiler emissions.  Heavier fuels may respond differently
to the few organic and organometallic compounds which
were found effective.  Further investigation of suitable
additives appears to be warranted, using residual oil as
fuel.  (these investigations should use advanced POM
measurement techniques and take into consideration the
conditions encountered in typical combustion equipment.)
The nitrogen-containing compounds previously found effective
for POM are of questionable suitability for additive use,
as they would add to NOX emissions, but the transition metal
compounds may be useful.
                           11-23

-------
                   ADDITIVE EFFECTS ON SULFUR OXIDES

        Much of the experimental work on additives for reduction of
sulfur oxides in oil- and coal-burning systems has been directed at
eliminating the corrosion and deposits which occur throughout the
combustion system as a result of sulfur oxides.  Although occasionally
the effects of the additives on sulfur dioxide was noted in the litera-
ture, most of the effort was devoted to measurements of changes in
sulfur trioxide concentration, particularly in terms of the sulfuric
acid dew point.  However, for the purposes of this review, such data
have been converted to ppm SO  in the gases, using the data of Lisle
and Sensenbaugh^76).  Typically, the S02/S03 ratio is 30 to 100
depending on combustion conditions, especially excess air.

        Because corrosion reduction in various parts of the boiler
was usually the objective of the additive program, the compounds were
introduced where it was thought that they would be most effective,
often downstream of the combustion zone.  For this reason, additives to
reduce sulfur oxide concentrations sometimes were introduced in the
superheater, in the convection passes, at the economizer, or at the air
heater*.  Investigations in which the additives were introduced down-
stream in this fashion, and the many studies of sulfur oxide removal
by scrubbing systems are not covered in this review.  In the investi-
gation reviewed here, the additive was either introduced in the fuel or
into the combustion zone.  Pertinent research has been reported for both
oil and coal firing, from laboratory burner scale to full-scale boiler
trials.
*  Additives that are active downstream of the flame zone are categorized
   as Type III Additives.
                                  11-24

-------
COMBUSTION ADDITIVES IN FUEL-OIL FIRING

EPA Investigation of Residual Oil Firing

        An extensive experimental program to determine the effects of
additives on sulfur oxide emissions from residual oil firing was carried
out by Pershing, Martin, Berkau and Hall^'   as part of an in-house
investigation by the Environmental Protection Agency.  Although the
additives investigated were proprietary mixtures, a complete chemical
and X-ray diffraction analysis identified the compounds involved.

        The test system consisted of an instrumented commercial package
boiler (1.8 million Btu/hr capacity) fired with an air-atomizing-type
gun burner that was capable of modulation.  The fuel contained 0.88-0.90
weight percent sulfur, and produced about 615 ppm of sulfur oxides in
the flue gases.  The additives were injected in powder form into the
highly turbulent region of the combustion air stream just before the
nozzle.  In the course of the program, variations were made in excess
air, boiler load, and additive dose rate.  No reduction in total sulfur
oxides occurred with any of the four additive systems investigated.
These were:
           1.   Clay mixture, consisting essentially of
               sodium or calcium aluminum silicates.
           2.   A mixture consisting chiefly of Nad and
               CaO, with some MgO and Fe 0 .
           3.   A mixture containing mostly NaCl, with
               some CaCO , talc, and iron manganese
               silicate.
           4.   A clay mixture consisting of sodium and
               calcium aluminum silicates.
A comparison test with Na CO  at a feed rate of 1 pound per 30 pounds
of fuel resulted in a 36 percent reduction in sulfur oxides by convert-
ing them to sodium sulfite and sulfate.  This test demonstrated that
                                  11-25

-------
sulfur oxide reduction was possible in the combustion system used, and
the negative results obtained with the other additives were valid.  In
the course of the experiments,  it also was observed that the additives
had no effect on emissions of NO, CO, hydrocarbons under any of the
conditions investigated.
Other Investigations with Residual Oil
Firing in Experimental Apparatus
        Most of the other investigations of sulfur oxide removal by
combustion zone additives have been concerned with the reduction of
the SO  content of the gas stream, and the effects on SO  were not
determined.
        A significant program of this type was conducted by Rendle and
Wilsdon    , who used an experimental furnace which burned 6 to 10
pounds per hour of residual fuel oil.  These investigators examined
materials which could act to reduce sulfur oxides in any of three ways:
(1) physical adsorption, (2) combination with atomic oxygen, or (3) neu-
tralization of acidic gases.  Representatives of each of these classes
of materials were fed into the oil-fired experimental furnace with a
combustion-chamber temperature of 1830 F and 25 percent excess air.  The
results of their experiments can be summarized as follows:

           1.  Materials which physically absorb gases.
               Powdered SiO  gave maximum of 50 percent
               reduction in SO  concentration, and
               carbon blacks did not change SO  concen-
               tration significantly when added to the
               fuel oil at concentrations up to 0.5 per-
               cent by weight.
           2.  Materials which combine with atomic oxygen.
               Tars containing pyridine and other organic
               nitrogen compounds did not change the SO
               levels appreciably when used in amounts up
               to 2.7 percent.
                                  11-26

-------
           3.  Materials which neutralize acidic gases.  Oil
               soluble magnesium and zinc napthenates com-
               pletely eliminated SO  from the gas stream.
               The magnesium compound was slightly more
               effective than the zinc derivative because only
               0.15 weight percent of the magnesium compound
               was needed to eliminate SO  whereas 0.2 weight
               percent of the zinc compound was required.
               With magnesium oxide, 0.35 weight percent of the
               additive was needed to remove all the SO  .  In
               the case of zinc dust, 0.45 weight percent was
               required, and for dolomite, 0.8 weight percent
               was needed.  In each case the amount of additive
               required was equal to or greater than the
               stoichiometric amount required to react with the
               S0_ in the gas stream.  For compounds of the
               same metal, the oil-soluble materials proved to
               more efficient.

In this study no measurement was made of the effects on SO .  However,
some reduction in SO  concentration is inevitable when basic materials
such as dolomite or MgO are used.
        These results have been confirmed by other investigators in the

laboratory and by actual practice in full-scale boilers.  Flint and co-
       (78}
workers     using a small refractory furnace fired with residual oil,
also showed that zinc naphthenates could be used to eliminate SO  from
the gas stream.  In their furnace system 0.07 weight percent of the

zinc naphthenate reduced the S0_ by 75 percent, while twice this amount

of the zinc compound completely eliminated the S0_.

                       (79)
        Peck and Zaczek     observed that metal oxides such as those of

zinc, copper, magnesium, calcium, and aluminum all were capable of re-
ducing the corrosive effects of sulfur oxides, but they did not make

any quantitative measurements of the gas concentrations.
        With a laboratory-scale burner using residual oil, Lewis
observed that SO  could be completely eliminated in his rig by the
addition of 0.07 weight percent magnesium in the form of magnesium
                                  11-27

-------
naphthenate.  He also tried an oil-soluble xinc compound, but found
it to be less effective than the magnesium naphthenate.*

        The mixture of hydrated MgO and A1203 having an Mg to Al
ratio of 9 to 1, developed by G. K. Lee and his associates, mentioned
                                                     (51  81  82)
previously, also has been effective for 803 reduction   '   '
Results of additive trials in both laboratory combustion boilers and
operating boilers demonstrated that SOg can be completely eliminated
by use of the additive in a concentration of 0.1 volume percent in
the oil.

Full-Scale Investigations with Residual
Oil Firing in Utility Boilers

        In full-scale boiler applications of additives, the Florida
Power Corporation found dolomite to be effective.  As reported by Huge and
Plotter^    at  the Inglis Station, the addition of 1/10 weight percent
of dolomite as  a slurry in the fuel oil resulted in about 33 percent
reduction in the 803 concentration.  At the Higgins Station, the dolo-
mite was introduced in the combustion air as a dry powder.  In this
case, 1/10 weight percent of dolomite reduced the 803 by 42 percent.
However, doubling the amount of dolomite used gave only a 50 percent
reduction in 803.

        The effects of additives in the boilers at the Marchwood
                                                         SQ I \
Station in  England were reported by Wilkinson and Clarke     .  In  this
case, residual  oils containing up  to 4.5 percent sulfur constituted  the
fuel.  The  addition of dolomite, which was introduced in the vicinity
of the burners  at a rate of 3  to 4 pounds per ton of fuel,  brought about
an 80 percent reduction in the  803 levels.  Magnesium carbonate  also was
tried, and  a 75 percent reduction  in 80^ was achieved with  the addition
 *   Other  studies  with No.  2  oil by Altwicher and associates     demon-
    strated  that cobalt  and manganese  naphthenate provided only  small
    reduction  (7 to  10 percent) in 803 concentration.   Iron naphthenate
    gave no  reduction.
                                   11-28

-------
of 1 to 2 pounds of additives per ton of fuel burned.  An effort was
made to use magnesium oxide similarly as an additive, but the fouling
and clogging problems which resulted made its use ineffectual.  Dolomite
also has been found effective in Swedish boiler practice.

        A favorable experience with magnesium oxide has been reported
by Exley et al of the Long Island Lighting Company, where four separate
power stations are combining the additive treatment with low-excess-air
         ( Qf.\
operation    .  In addition to reducing sulfuric-acid stack emissions,
boiler efficiency was raised and reliability was improved.  Magnesium
vanadate was recovered from the deposits in the furnace; at the time
of the report (1966), at least 1.5 times the cost of the additives was
recovered through the sale of the vanadium by-product.

        Granular magnesium metal (20 to 50 mesh) injected into the com-
                                     (O~7\
bustion zone was found by Reese et al     to be effective in reducing
503 in an oil-fired boiler.  The unit in which the tests were made was
rated at 185 megawatts and the magnesium was added at the rate of 12
pounds per hour.  A 60 to 75 percent reduction in 863 was found, depend-
ing on what part of the system the measurements were made.

             (QQ\
        Lasch     reported success with injections of granular magne-
sium into the combustion chamber of oil-fired boilers at the Belvedere
and Tidbury Power Stations of the Central Electricity Generating Board
in England.  With oil having a 2 to 4 percent sulfur content, magnesium
injection at a rate of 0.4 pounds per ton of oil eliminated any problems
with S03.

        Several stations of the Consolidated Edison Company that burn
residual fuel oil have used methylcyclopentadienyl maganese tricarbonyl
as an additive.  The SC>3 in the flue gases was reduced by 45 percent,
but the concentration in which the additive was used was not specified
                                   11-29

-------
                                                                  (52)
        Additive trials conducted by the New England Power Company
showed that ferrocene, iron carbonyl, and manganese dioxide all were
ineffective in reducing the 803 content in the flue gases.  [Martin,
Pershing, and Berkau   , firing No. 2 oil in the EPA laboratory-
scale burner, obtined similar negative results with ferrocene and
with methylcyclopentadienyl manganese tricarbonyl.]
COMBUSTION ADDITIVES IN COAL FIRING

        Only a relatively small amount of effort has been devoted to
fuel additives or combustion-zone additives to reduce sulfur oxides in
coal-fired systems.

        In contrast, considerable investigation has been directed to
downstream addition of limestone for sulfur oxide removal.  The lime-
stone must be added in a temperature zone in which it will calcine to
CaO, which then reacts with sulfur oxides to form CaS03 and caSO^.
These systems have encountered serious problems with scaling and foul-
ing.  For example difficulties have been experienced in large-scale
operation of limestone scrubbing at the Meramec Station of the Union
Electric Company     and the Lawrence Station of Kansas City Power and
Light<91>.

        The following discussion is confined to investigation where
the additive is introduced with the fuel or into the combustion zone.

                            (92)
        Michel and Wilcoxson     observed that dolomite used at a rate
of 7.5 pounds per  ton of coal fired, reduced 803 concentration by 50
percent.  Doubling the dolomite addition reduced the amount of 803 by
90 percent.

                                                 (93)
        In a pulverized-coal burner, Borio, et al     observed a 65
percent reduction  in the S02 concentration when NaOH was  added to
coal.  The 803 level remained about the same.  However, anomalous
results were obtianed when dolomite and NaOH were used in combination,

                                  11-30

-------
as the SC>2 concentration increased from 1200 to 2400 ppm.  A similar
effect was noted when CaO was added to the coal.
        Corbett and Flint     added zinc ore concentrate to coal used
in stoker-fired boilers at the Brimsdown Power Station.  The zinc oxide
smoke generated in this fashion eliminated the SOo from the flue gases
when the additive was used at a concentration of 0.25 weight percent.

        Powdered sodium carbonate was aspirated into the combustion
zone of a large spreader-stoker coal-fired boiler in research by
                    (95)
Brancaccio and Flach    .  The sulfur dioxide emissions were reduced
up to 70 percent in this fashion.  Sodium bicarbonate also was promis-
ing, but both of these additives created problems with boiler fouling.

        A test unit consisting of a 750-horsepower boiler fired by a
multiple retort underfeed stoker was used by Land and his colleagues
Additives were either mixed with the coal before it was fired or were
injected with compressed air jets over the fire.  Dolomite chips,
pulverized dolomite, hydrated lime, aragonite, and red mud were used as
additives.  Anomalous results were obtained in the limited tests that
were made, but the results did indicate that sulfur dioxide emissions
from coal burning can be reduced somewhat by the use of additives
introduced in this fashion.

                                           (97)
        A coal related study by Whittingham    , using a coal gas
burner with 1,000 ppm S02 added, showed that silicon-monoxide fume
formed by the reduction of silicates with carbon could be used to
reduce SO^ concentration.  At an additive concentration of 2 grains
per cubic foot, the 803 was reduced by 80 percent.  He also studied
               (98)                               —t±
zinc-oxide fume    .  At a concentration of 2 x 10   grams of ZnO per
liter, this additive gave an 85 percent reduction in the 803 level.
In another study, Whittingham used coal-gas introduction above the flame
                      (99)
to make a carbon smoke    .  The SO-^ content of the gas stream was
reduced by 90 percent when a smoke concentration of 0.24 grains per
                                  11-31

-------
cubic foot was used.  Whittingham suggested that the role of the smoke
particules in reducing sulfur oxide concentration consisted of absorb-
ing the gases by virtue of the large effective surface area of the fume.
No measurement of particulates was made, however, to determine whether
the introduction of fume would cause a significant increase in the
emissions of particulate.

          Survey Conclusions Related to the Use
          of Combustion Additives for Reducing
          Emission of Sulfur Oxides

               Combustion additives containing basic element such
         as magnesium, calcium, and zinc are effective in reduc-
         ing SO  by converting it into metal sulfates which can
         be removed as particulates or, possibly in some cases,
         by inhibiting the catalytic oxidation of SO„ by exposed
         metal surfaces in the low-temperature regions of the
         boiler.  These additives are particularly effective as
         oil-soluble compounds which can be mixed into fuel oils.
         However, these additives are not useful in reducing SO
         in flue gases.  Solids containing calcium and magnesium
         may be capable of reducing SO  from coal combustion.
         The use of additives may thereby reduce the potential
         for sulfuric-acid aerosol formation, although the total
         sulfur emitted may be unchanged.
                  ADDITIVE EFFECTS ON NITROGEN OXIDES
        Most of the effort aimed at reduction of NO  emissions has
                                                   x
involved combustion control techniques.  These methods require modifi-
cation of the operating and design features which affect the various
combustion parameters.  However, some research has been conducted on the
use of additives to reduce NO  emissions.  Some of this effort was
                             x
directed primarily at NO , but in other cases the data on nitrogen
oxide emissions were recorded in the course of a program whose main
                                  11-32

-------
objective was reduction of particulate emissions or sulfur oxides.   All
of these programs were carried out on liquid fuels; no data have been
found for coal-fired systems where additive effects on NOX have been
recorded.

Investigations Firing Residual Oil in Boilers

        Research on additives to residual oil used in low-pressure
heating boilers was conducted by Lee, Friedrich, and Mitchell   '
They found that a mixture of partially dehydrated hydroxides of magne-
sium and aluminum having a 9 to 1 magnesium-aluminum ratio effected a
43 percent reduction in NOX when added to the oil at a concentration of
0.1 volume percent.  However, it was noted that the NOX concentration,
while decreasing with greater additive dosage, was at a higher level
than with untreated oil until the dosage rate reached the 1/10 percent
level.

                            (52)
        Molino and Zobolotny     reported results by the New England
Power Company where additions of ferrocene to residual oil fuel had no
effect on the amount of NOX produced.  Iron carbonyl and manganese
dioxide also were found to be without effect on nitrogen oxides.
Investigations Firing Distillate
Oil in Laboratory Furnaces
        In the very extensive EPA investigation of the effects of fuel
additives on emissions from distillate-oil-fired furnaces, Martin,
Pershing, and Berkau    found that the few additives which reduce
particulate emissions had no effect on the NOX emissions.  This work
was primarily concerned with proprietary additives, but two of the
additives could be specifically identified as the chemical compounds
biscyclopentadienyl iron (ferrocene) and methylcyclopentadienyl magna-
nese tricarbonyl.  Although they had no effect on NOX, these two com-
pounds were shown to have some efficacy in reducing particulate
emissions.
                                  11-33

-------
        One of the few studies directed mainly to nitrogen oxides was
                                 ( 89^
carried out  by Altwicker,  et alv  ' who used No. 2 oil in a small
laboratory burner.  Additions of 0.5 and 1.0 volume percent of cobalt,
manganese, and iron naphthenates reduced NO  emissions only 6 to 15
                                           X
percent.  The cobalt compound was the most effective and the iron
derivative the least effective.  These investigators found the same
small reduction in nitrogen oxides when using proprietary additives
containing manganese.  As might be expected, some commercial amine
additives increased the production of nitrogen oxide.

        Similar studies carried out by Fredette^   ' showed that butanol
and pentanol also would provide 10 to 15 percent reductions in nitrogen
oxides when added to No. 2 oil at a concentration of 1 volume percent.

Investigation Directed to Gas Turbines

        A broad experimental program for the U.S. Air Force reported
by Shaw     , to assess the feasibility of reducing NOX emissions from
aircraft gas turbine engines, included a study of additive treatments.
Jet fuel A was burned in a laboratory combustor.   The selection of
additives was systematized to cover the various mechanisms which could
be hypothesized for reduction of NO  formation.  These mechanisms
included:
           1.  Heterogeneous catalysts for either reduction
               or decomposition of NO
           2.  Oxygen atom scavengers
           3.  Temperature reducers via heat transfer with
               solids or endothermic physical conversion
           4.  Ignition delay agents
           5.  Agents that change spray dynamics
           6.  Agents that react with NO.
Over 70 additives were evaluated in this program.  These included metal
naphthenates and acetylacetonates, metal salts of neoacids, alkali, and

                                  11-34

-------
alkaline-earth metal carbonate suspensions, emulsions of alkali hydrox-
ides, miscellaneous metal compounds, and various organic compounds.
The additives falling in the first category, namely, heterogenous
catalysts, proved to be most effective in reducing NO  emissions.  How-
ever, the greatest reduction obtained was only 30 percent, which was
achieved by adding cobalt or manganese naphthenate at a concentration
of 0.5 weight percent metal.  The iron and copper naphthenates were
somewhat less effective.  The only other effective compound was the
zirconium salt of neodecanoic acid which reduced nitrogen oxides by 22
percent at a concentration of 0.2 weight percent.  Other organometallic
compounds containing calcium, cerium, lead, aluminum, lithium, or
antimony did not reduce NO  significantly.  Four suspensions of metal
                          X
carbonate in a diluent oil were added in amounts of 0.1 percent by
weight metal; sodium carbonate reduced nitrogen oxides by about 16
percent and lithium carbonate, by about 10 percent.  The other carbon-
ates, barium and calcium, were not effective.

        Of the 16 homogeneous additives, which were all organic com-
pounds, only n-butylmercaptan was effective in Shaw's investigation.
This compound reduced NO  by about 11 percent.  Any organic compounds
containing nitrogen were found to increase the emissions.  Various
inorganic salts which were dissolved in water and then emulsified in
the jet fuel, had no effect, but those of alkali hydroxides gave small
reductions.  The best of these was 0.1 weight percent sodium as sodium
hydroxide, which gave 16 percent reduction in nitrogen oxides.  The
only other water emulsion that showed some promise in reducing NO  was
                                                                 X
one that contained 37 ppm by weight hydrazine acetate.  The results
with this emulsion were erratic, but in some replicate experiments,
gave as much as 16 percent NO  reduction.
                             X

        It was concluded that no really effective additive had been
uncovered in this work.
                                  11-35

-------
Investigation Directed to Diesel Engines

        In a diesel engine study, McCreath      found that isoamyl
nitrate addition to the fuel provided about a 17 percent reduction in
NOX, while ditertiary-butyl peroxide gave a 14 percent reduction.  He
also noted that if the additive were mixed with the fuel and left to
age, some reaction occurred which improved slightly the performance of
the additive.  Under these conditions, nitrcinethane reduced nitrogen
oxides about 9 percent, whereas  in  a  fresh  mixture, it actually increased
the production of NOX.  The nature of the chemical changes involved was
not investigated.
          Survey Conclusions Related to Use of
          Combustion Additives for Reducing NOX
          Small reductions in NOX emissions have been
          observed with a variety of additive types.  Oil-
          soluble transition-metal compounds, organic com-
          pounds, and magnesium oxide all provide minor
          reduction effects.  However, it appears that
          there is little evidence of achieving any worth-
          while benefits for NOX control by the use of fuel
          additives.
        The reader is referred to the SUMMARY at the beginning of this
report, especially the Section on "Overview and Conclusions" which com-
bine observations from the analysis of basic combustion mechanisms in
Part I with the review of experimental investigations in Part II.  Also
included  there  are comments on the overall perspective of possible
limitations to the use of additives.
                                 11-36

-------
                                REFERENCES
 1.  Martin, G. B., D. W. Pershing, and E. E. Berkau, "Effects of Fuel
     Additives on Air Pollutant Emissions From Distillate-Oil-Fired
     Furnaces", U.S. EPA, Office of Air Programs, Publication No., AP-87
     pp 86 (June, 1971).

 2.  Pershing, D. W., G. B. Martin, E. E. Berkau, and R. E. Hall,
     "Effectiveness of Selected Fuel Additives in Controlling Pollution
     Emissions From Residual Oil-Fired Boilers", U.S. Environmental
     Protection Agency Report No. EPA-650/2-73-031 (October, 1973).

 3.  Giammar, R. D., A. E. Weller, D. W. Locklin, and H. H. Krause,
     "Experimental Evaluation of Fuel-Oil Additives for Reducing
     Emissions and Increasing Efficiency of Boilers", Final Report by
     Battelle-Colurabus on EPA Contract No. 68-02-0262 to the Environ-
     mental Protection Agency and the Federal Energy Administration.
     EPA Report No. EPA-600/2-77-008b  (January  1977).

 4.  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., pp 375 (1972).

 5.  Mellor, A. M., "Current Kinetic Modeling Techniques for Continuous
     Flow Combustors", Proceedings:  GM Symposium on Emissions From Con-
     tinuous Combustion Sources. Detroit, Michigan, pp 23-53 (September,
     1971), Plenum Press (1972) .

 60  Palmer, H. B., "Equilibria and Chemical Kinetics in Flames",
     Combustion Technology:  Some Modern Developments. Palmer and Beer,
     Ed., Academic Press (1974).

 7.  Barrett, R. E., S. E. Miller, and D. W. Locklin, "Field Inves-
     tigation of Emissions from Combustion Equipment for Space Heating,"
     EPA Report R2-73-084a (API Publication 4180) (June 1973) .

 8.  Long, R., "Studies on Polycyclic Aromatic Hydrocarbons in Flames",
     Report EPA-R3-72-020, Environmental Protection Agency (July, 1972).

 9.  Thompkins, E. E., and R. Long, "The Flux of Polycyclic Aromatic
     Hydrocarbons and of Insoluble Matter in Pre-mixed Acetylene-Oxygen
     Flames", 12th International Symposium on Combustion.  The Combustion
     Institute, pp 625-634 (1969).

10.  Agius, P. J., M. Brod, H. N. Miller, R. C. Price, J.  Ryer, and
     J. M. Tims, "Current Trends in Fuel Additives", Eighth World
     Petroleum Congress,  Products, J5, pp 27 (1971).

11.  Hangebrauck, R. P.,  D. J. Van Lehmden, and J. E. Meeker, "Sources
     of Polynuclear Hydrocarbons in the Atmosphere", PHS Publication
     No. 999-AP-33, pp 44 (1967).
                                  11-37

-------
12.  Schirmer, R. M.,  "Effect of Fuel Composition on Particulate  Emis-
     sions from Gas Turbine Engines", Emissions  From Continuous Combus-
     tion Systems. Cornelius and Agnew,  Editors.  Plenum Press  (1972).

13.  Cooper, P. W., R. Kamo, C. J. Marek,  and C. W.  Solbrig,  "Recir-
     culation and Fuel-Air Mixing as Related to  Oil-Burner  Design",
     API Publication 1723 (May, 1964).

14.  Hall, R. E., G. B. Martin, J. H. Wasser, and J. S.  Bowen,  "Status
     of EPA's Combustion Research Program for Residential Heating
     Equipment - June  1974", presented at  annual meeting of the Air
     Pollution Control Association, Denver,  Colorado (June  14,  1974).

15.  Lodwick, J. R., "Chemical Additives in  Petroleum Fuels:  Some Uses
     and Action Mechanisms", Journal of  the  Institute of Petroleum, 50.
     pp 297-308 (1964).

16.  Locklin, D. W., A. E. Weller, and R.  E. Barrett, "The  Federal R&D
     Plan for Air-Pollution Control by Combustion-Process Modification",
     Final Report, Contract CPA-22-69-147, to Environmental Protection
     Agency (PB 198,066)  (January 11, 1971).

17.  Homann, K. H., and H. G. Wagner, "Some  New  Aspects  of  the  Mechanism
     of Carbon Formation in Premixed Flames", llth International  Com-
     bustion Symposium. The Combustion Institute, pp 371  (1967).

18.  Fenimore, C. P. and G. W. Jones, "Oxidation of Soot by Hydroxyl
     Radicals", J. Phys. Chem., 7±, pp 593 (1967).

19.  Rosner, D. W., and H. D. Allendorf, "Comparative Studies of  the
     Attack of Pyrolytic and Isotropic Graphite  by Atomic and Molecular
     Oxygen at High Temperatures", AIAA Journal, £,  pp 650-654  (1968).

20.  Place, E. R., and F. J. Weinberg, "The  Nucleation of Flame Carbon
     by Ions and the Effect of Electric Fields", llth International Com-
     bustion Symposium, The Combustion Institute, p  245 (1967).

21.  Field, M. A., D.  W. Gill, B. B. Morgan, and P.G.W.  Hawksley,
     Combustion of Pulverized Coal. BCURA, Leatherhead (1967) .

22.  Benson, S. W., The Foundations of Chemical  Kinetics. McGraw  Hill
     (1960).

23.  Homer, J. B. and  G. B. Kistiakowsky,  "Oxidation and Pyrolysis of
     Ethylene in Shock Waves", J. Chem.  Phys., 47, pp 5290-5295 (1967).

24.  Reid, W. T., "Corrosion and Deposits in Combustion Systems",
     Combustion Technology: Some Modern Developments. Academic  Press
     (1974).
                                 11-38

-------
25.  Nelson, H. W., et al., "A Review of Available Information on Cor-
     rosion and Deposits in Coal- and Oil-Fired Boilers and Gas tur-
     bines".  A report by Battelle to the ASME Research Committee on
     Corrosion and Deposits from Combustion Gases, reprinted by The
     American Society of Mechanical Engineers (1959) 234 pages.

26.  Zeldovich, Ya. B., Acta Physiochim, USSR, 2J., pp 577 (1946).

27.  Fenimore, C. P., "Formation of Nitric Oxide in Premixed Hydro-
     carbon Flames", 13th International Combustion Symposium, The Com-
     bustion Institute, p  373 (1971).

28.  LaChapelle, D. G., J. S. Bowen, and R. D. Stern, "Overview of
     Environmental Protection Agency's NOX Control Technology for
     Stationary Combustion Sources," Paper presented to AIChE, 67th
     Annual Meeting, Washington,  D.C. (Dece,ber, 1974).

29.  Martin, G. Blair and E. E. Berkau, "An Investigation of the Con-
     version of Various Fuel Nitrogen Compounds to Nitrogen Oxides in
     Oil Combustion", AIChE National Meeting, Atlantic City, New Jersey
     (August, 1971).

30.  Pershing, D. W., G. B. Martin, and E. E. Berkau, "Influence of
     Design Variables on the Production of Thermal and Fuel NO from
     Residual Oil and Coal Combustion", 66th Annual AIChE Meeting,
     Philadelphia, Pennsylvania (November 11-15, 1973).

31.  Turner, D. W., R. L. Andrews, and C. W. Siegmund, "Influence of
     Combustion Modification and  Fuel Nitrogen Content on Nitrogen
     Oxides Emissions from Fuel Oil Combustion", AIChE paper, December,
     1971, published in Combustion, August, 1972.

32.  Sternling, C. V. and J.O.L.  Wendt, "Kinetic Mechanisms Covering
     the Fate of Chemically Bound Sulfur and Nitrogen in Combustion",
     Final Report to EPA, Contract EHD 71-45, August, 1972.

33.  Wendt, J.O.L. and C. V. Sternling, "Effect of Ammonia in Gaseous
     Fuels on Nitrogen Oxide Emissions", J. Air Poll. Control Assoc.,
     24, (11) PP  1055-1058  (1974).

34.  Fenimore, C. P., "Formation of Nitric Oxide from Fuel Nitrogen in
     Ethylene Flames", Combustion and Flame, .19, PP 289-296  (1972)

35.  Haynes, B. S., D. Iverach, and N. Y. Korov, "The Behavior of
     Nitrogen Species in Fuel Rich Hydrocarbon Flames", 15th Symposium
     on Combustion, Tokyo (August, 1974).

36.  Merryman, E. L., and A. Levy, "Nitrogen Oxide Formation in Flames:
     The Roles of N02 and Fuel Nitrogen", 15th Symposium on Combustion,
     Tokyo, Japan  (August, 1974).
                                 11-39

-------
37.  Friswell, N. J., "Emissions from Gas Turbine-Type Combustors",
     (W. Cornelius and W. G.  Agnew,  eds.), p  161,  Plenum Press,
     New York (1972).

38.  Riggs, R. J., G. J. Wilkinson,  and H. R.  Wolfe,  "Combustion
     Improvers for Fuel Oils", Erdbel Kohle, 18 (4),  pp 282-286  (1965).

39.  Rai, C., S. D. Collins,  and E.  J. Badin,  "Smoke  Suppressant
     Additive for Fuel Mixtures", U.S. Patent  3,577,228,  pp  7
     (May 4, 1971).

40.  Cotton, D. H., H. J. Friswell,  and C. R.  Jenkins, "The  Suppression
     of Soot Emission From Flames by Metal Additives", Combustion and
     Flame, J.7 (1), pp 87-98 (1971).

41.  Doxey, G., G. E. Hodgkinson, and G. Woodhouse, "Some Uses of
     Calcium Chloride as an Additive to Boiler Coals", Journal of the
     Institute of Fuel, 40 (322), pp 521-535 (1967).

42.  Sawyer, R. F., discussion in Emissions from Continuous  Combustion
     Systems, Cornelius and Agnew, Editors, p  181-182, Plenum Press,
     1972.

43.  Feugier, A.,  "Effect of Metal Additives on the Amount of Soot
     Emitted by Premixed Hydrocarbon Flames",  International  Flame
     Research Foundation, Flame Chemistry Panel, Essen
     (October 26,  1973).

44.  Dalraon, J. and D. Tidy, "A Comparison of Chemical Additives  as
     Aids  to the Electrostatic Precipitation of Fly-Ash", Atmospheric
     Environment,  6, pp 721-734  (1972).

45.  Turner, D. W., and C. W. Siegmund, "Staged Combustion and Flue-Gas
     Recycle: Potential for Minimizing NOx from Fuel Oil Combustion"
     Paper presented  at  American Flame  Research Committee Flame Days,
     Chicago,  Illinois  (September 6-7,  1972).

46.  Lyon, R. K.,  "Method for the Reduction of the Concentration of  NO
     in Combustion Effluents Using Ammonia", U.S. Patent 3,900,554
     (August 19,  1975).

47.  Wendt, J.O.L., C. V. Sternling, and M. A. Matovich, "Reduction  of
     Sulfur Trioxide and Nitrogen Oxides by Secondary Fuel Injection",
     Fourteenth Symposium (International) on Combustion, The Combustion
     Institute, Pittsburgh, p 897 (1973).

48.  Muzio, L. J., J. K. Arand, and D. P. Teixeria, "Gas Phase  Decompo-
     sition of Nitric Oxide in Combustion Products", The Proceedings
     of the NOX Control Technology Seminar, EPRI Report SR-39,
     February, 1976, FB 253661.
                                 11-40

-------
49.  Finfer, E. Z., "Fuel Oil Additives for Controlling Air Contaminant
     Emissions", Journal of the Air Pollution Control Association,  17,
     pp 43-45 (1967).

50.  Salooja, K. C., "Burner Fuel Additives", Combustion,  pp 21-27
     (January, 1973).

51.  Lee, G. K., F. D. Friedrich, and E. R. Mitchell, "Control of 803
     in Low-Pressure Heating Boilers by an Additive", Journal of the
     Institute of Fuel, 42 (327), pp 67-74 (1969).

52.  Molino, N., and E. R. Zabolotny, "Operational Method for Reduction
     of Particulate Emissions From an Oil-Fired Utility Plant", New
     England Power Service Company and Stone and Webster Engineering
     Corporation (1973).

53.  Weeks, R. L., W. L. Clinkenbeard, and J. D. Soltis, "Clean Effi-
     cient Combustion and Heating Oil", Proceedings of 5th World Petro-
     leum Congress, New York, Section VI, Paper 25, pp 381-396
     (June, 1959).

54.  Vaerman, J., "Problems of Soot Emission and Low Temperature Cor-
     rosion in Domestic Oil-Fired Boilers", Journal of the Institute of
     Petroleum, 50, pp 155-168 (1964).

55.  Plonsker, L., E. B. Rifkin, M. E. Gluckstein,  and J.  D. Bailie,
     "Reduction of Gas Turbine Smoke and Particulate Emissions by a
     Manganese Fuel Additive", The Combustion Institute, Central States
     Section, Madison, Wisconsin (March 26-27, 1974).

56.  Shayeson, M. W., "Reduction of Jet Engine Exhaust Smoke With Fuel
     Additives", SAE Transactions, _7J>> PP 2687-2694 (1968).

57.  Spengler, Guenther, and Haupt, Gerhard, "Formation of Soot and
     Polycyclic Aromatic Hydrocarbons in Simple Hydrocarbon Flames and
     Its Inhibition by Fuel Additives", Erdbel and Kohle (Hamburg), 22
     (11), pp 679-684 (1969).

58.  Ray, S. E., and R. Long, "Polycyclic Aromatic Hydrocarbons From
     Diffusion Flames and Diesel Engine Combustion", Combustion and
     Flame, 8, pp  139-151 (1964).

59.  "Smoke Suppressant Compositions for Petroleum Fuels", Cities
     Service Oil Company, British  Patent 1,243,264, p  10
     (August  18, 1971).

60.  Chittawadgi,  B. S., and A. N. Volnov, "Mechanism  of Action of
     Ferrocene  on  Smoke Reduction  in Diffusion Flames, "Indian Journal
     of Technology,  3  (M), pp 209-211  (1965).

61.  Churchill, A.  V., and E. Mitchell, "Hydrozine-Containing Fuel Oil
     Composition",  U.S. Patent 2,971,828  (February 4,  1961).
                                 11-41

-------
62.  "Ethers as Smoke Suppressants for Liquid Fuel Compositions",
     Cities Service Oil Company,  British Patent 1,246,853,  p 3
     (September 22, 1971).

63.  Perilstein, W. L., "Overbased Barium Sulfonates for Use in Dis-
     tillate Fuel Oils", U.S. Patent 3,580,707, p 8 (May 25, 1971).

64.  Kuhn, M.,  and R. Tomingas, "Attempts to Prevent the Formation of
     Pollutants in the Exhaust Gases of Two-Stroke Engines  and Diesel
     Engines by Activating Combustion Within the Engine", Staub
     (English Translation), 25. (3), pp 2-17 (1965).

65.  Golothan,  D. W., "Diesel Engine Exhaust Smoke: The Influence  of
     Fuel Properties and the Effects of Using Barium-Containing Fuel
     Additive", Society of Automotive Engineers, Engineering Congress,
     Detroit, Michigan, Paper 670092 (January 9-13, 1967).

66.  Amick, J.W., and R. A. Hunt, Jr., "Metal Chelate Combustion
     Improver for Fuel Oil", U.S. Patent 3,355,270 (November 28, 1967).

67.  Kukin, I., "Chemical Supplements in Air Pollution Control Pro-
     grams", National Fuels and Lubricants Meeting, New York, Paper
     FL-67-65 (September, 1967).

68.  Kukin, I., "Additives Can Clean Up Oil-Fired Furnaces", Environ-
     mental Science and Technology, ]_, pp 606-610 (1973).

69.  Nicholls,  P., and C. W. Staples, "Removel of Soot From Furnaces
     and Flues by the Use of Salts or Compounds", U.S. Bureau of Mines,
     Bulletin 360, p 76 (1932).

70.  Nicholls,  P., W. E. Rice, B. A. Landry, and W. T. Reid, "Burning
     of Coal and Coke Treated With Small Quantities of Chemicals",
     U.S. Bureau of Mines, Bulletin 404, p 158 (1937).

71.  Kerley, R.V., "Catalyzed Combustion of Coal With Restricted Air
     Supply", French Patent 1,547,982, p 4 (November 29, 1968).

72.  Long, R., and S. K. Ray, "Inhibition of the Formation of 3,4-
     Benzpyrene During Hydrocarbon Combustion", Nature (London), 192
     (480), pp 353-354 (1961).

73.  Chakroborty, B. B., and R. Long, "The Formation of Soot and Poly-
     cyclic Aromatic Hydrocarbons in Ethylene Diffusion Flames With
     Methanol as an Additive", Combustion and Flame, 11 (2), pp 168-170
     (1968).

74.  Hunigen, E., N. Jaskulla, and K. Wetting, "The Reduction of Car-
     cinogenic Contaminants in Exhaust Gases of Petrol Engines Through
     Fuel Additives and Choice of Lubricants", Proceedings of the  Inter-
     national Clean Air Congress, London, Part I, Paper VI-12,
     pp  191-193  (1966).
                                11-42

-------
75.  Van Den Heuvel,  H.M.H.,  "Development and Effect  of Fuel Inhib-
     itors", Freiburger Forschungshefte,  A (503),  pp  51-67  (1972).

76.  Lisle, E.  S., and J.  D.  Sensenbaugh, "The Determination of Sulfur
     Trioxide and Acid Dew Point in Flue  Gases",  Combustion, _3£ (7),
     pp 12-15 (1965).

77.  Rendle, L. K. , and R. D. Wilsdon, "The Prevention of Acid Conden-
     sation in Oil-Fired Boilers", Journal of the Institute of Fuel,
     29, p 372 (1956).

78.  Flint, C., A. W. Lindsay, and R. F.  Littlejohn,  "The Effect of
     Metal Oxide Smokes on the S03 Content of the Combustion Gases
     From Fuel Oils", Journal of the Institute of Fuel, 26  (152),
     pp 122-127 (1953).

79.  Peck, W. J., and B. J. Zaczek, "Sulphur in Flue  Gases", Engineer-
     ing, p 697 (November 29, 1957).

80.  Lewis, A., "Deposits From the Continuous Combustion of Oil Fuels",
     Proceedings of 4th World Petroleum Congress, Section VI/D, Rome,
     pp 247-262 (1955).

81.  Lee, G. K. , "Control of Oil Ash Deposits and Pollution Abatement
     by an Additive", Fuel Society Journal, 20, pp 8-17 (1969).

82.  Whaley, H., F.  D. Friedrich, G. K. Lee, and E. R. Mitchell, "Air
     Pollution: Causes and Control", Department of Energy,  Mines and
     Resources, Ottawa, Ontario, Fuels Research Centre Information
     Circular  211, p  25 (November,  1968).

83.  Huge,  E.  C.,  and E.  C.  Piotter, "The Use of Additives  for  the Pre-
     vention of Low-Temperature Corrosion in Oil-Fired Steam Generating
     Units", Transactions  of the ASME, 76_, p 267  (1955).

84.  Wilkinson, G. J.,  and D. G.  Clarke,  "Problems Encountered  With  the
     Use  of High  Sulphur  Content  Fuel Oils at Marchwood Generating
     Station and  Experience  With  Chemical Additives",  Journal  of the
     Institute of  Fuel, 32 (217),  pp  61-72  (1959).

85.  Exley, L. M., A. E.  Temburrino,  and A. J. O'Neal, Jr.,  "Lilco
     Trims Residual  Oil Problems",  Power,  110  (4), pp  69-73  (1966).

86.  Reese, Jr. G.,  J.  Jonakin,  and V. Z. Caracristi,  "Prevention of
     Residual  Oil Combustion Problems  by Use  of  Low  Excess  Air and
     Magnesium Additive", Transaction of the ASME> Journal  of  Engineer-
      ing for Power,  Series A, 87  (2),  pp 229-236  (1965).

 87.   Lasch, L.,  "Magnesium Injection Checks  Corrosion in Oil-Fired
      Boilers,  Engineering Boiler House Review (London), J51  (2),
      pp 42-23  (1966).
                                 11-43

-------
88.  Belyea, A. R.,  "Manganese Additives  Reduces  803",  Power,  110  (11),
     pp 80-81 (1966).

89.  Altwicker, E.  R.,  P.  E.  Fredette,  and T.  Shen,  "Pollutants  From
     Fuel Oil Combustion and  the Effect of Additives",  Air Pollution
     Control Association,  64th Annual Meeting, Atlantic City,  New
     Jersey, Paper  No.  71-14  (June 27-July 2,  1971).

90.  McLaughlin, J.  F., "Sulfur Dioxide Scrubber  Service Record, Union
     Electric Company—Meramec Unit 2", Second International Lime/
     Limestone Wet  Scrubbing  Symposium, New Orleans,  La.
     (November 8-12, 1971).

91.  Devitt, T. W.,  and F. K. Zada, "Status of Flue  Gas Desulfurization
     Systems in the United States", Flue Gas Desulfurization Symposium,
     Atlanta, Georgia (November 4-7, 1974).

92.  Michel, J. R.,  and L. S. Wilsoxson,  "Ash Deposits  on Boiler Sur-
     faces From Burning Central Illinois Coal", ASME Paper No. 55-A-95
     (November, 1955).

93.  Borio, R. W.,  R.  P. Hensel, R. C.  Ulmer,  E.  B.  Wilson, and
     J. W. Leonard,  "Study of Means for Eliminating  Corrosiveness  of
     Coal to High Temperature Surfaces  of Steam Generating Units II",
     ASME Paper No.  67  WA/CD-3, Pittsburgh, Pennsylvania
     (November, 1967).

94.  Corbett, P. F., and D.  Flint, "The Influence of Certain Smokes
     ans Dusts on the 803 Content of the Flue Gases  in Power-station
     Boilers", Journal of the Institute of Fuel,  Z5, 410-417 (1953).

95.  Brancacio, J., and C. V. Flach, "Use of Dry  Fuel Additives to
     Reduce S02 Emissions From a Full Size Industrial Coal-Fired
     Boiler", Air Pollution Control Association,  65th Annual Meeting,
     Miami, Florida, Paper 72-77 (June 18-22, 1972).

96.  Land, G. W., E. W. Linna, and W. G. Earley,  "Controlling Sulfur
     Dioxide Emissions From Coal Burning by the Use of Additives",
     Air Pollution Control Association, 62nd Annual Meeting, New York,
     Paper 69-143 (June,  1969).

97.  Whittingham, G., "The Influence of Silica Smokes on the Dew Point
     of Combustion Gases Containing Sulfur Oxides",  Journal of the
     Society of Chemical Industry, .67, pp 411-414 (1948).

98.  Whittingham, G., Communication on paper by Bowden, Draper, and
     Rowling,  Proceedings of  Institute of Mechanical Engineers, Section
     A,  167. pp  291-300 (1953).
                                 11-44

-------
 99.   Whittingham,  G.,  "The Influence of Carbon Smokes  on the Dew Point
      and Sulfur Trioxide Content of Flame Gases",  Journal of Applied
      Chemistry, I, pp  382-399 (1951).

100.   Fredette,  P.  E.,  "Effect of Additives on Nitrogen Oxides and
      Particulate Emission From Fuel Oil Combustion", Rennsselaer Poly-
      technic Institute,  Troy, New York, Thesis,  University Microfilms,
      Ann Arbor, Michigan, p 164 (June,  1972).

101.   Shaw,  H.,  "Fuel Modification for Abatement  of Aircraft Turbine
      Engine Oxides of  Nitrogen Emissions", Technical Report Af APL-TR-
      72-80, Air Force  Aero Propulsion Laboratory,  WPAFB (October,1972).

102.   McCreath,  C.  G.,  "The Effect of Fuel Additives on the Exhaust
      Emmissions From Diesel Engines", Combustion and Flame, 17.
      pp 359-366 (1971).

103.   Upmalis,  A.,  "Ammonia or Dolomite  Process",  Brennstoff-Warme-
      Kraft, 2,  pp 232  (1957).
                                 11-45

-------
                                TECHNICAL REPORT DATA
                          (Please read Inunctions on the reverse before completing)
1. REPORT NO.
EPA-600/2-77-008a
                           2.
                                                      3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Combustion Additives for Pollution Control—A-
   State-of-the-Art Review
            5. REPORT DATE
             January 1977
            6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
                                                      8. PERFORMING ORGANIZATION REPORT NO.
H.H. Krause, L.J. Hillenbrand, A.E. Weller,
   and D. W.  Locklin
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Battelle-Columbus Laboratories
 505 King Avenue
 Columbus, Ohio 43201
            10. PROGRAM ELEMENT NO.
            1AB014; ROAP 21ADG-020
            11. CONTRACT/GRANT NO.

            68-02-0262
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
            Task Final; 5/72-12/75
            14. SPONSORING AGENCY CODE
             EPA-ORD
is. SUPPLEMENTARY NOTES IERL-RTP project officer for this report is W.S. Lanier,  919/549.
8411 Ext 2432,  Mail Drop 65.
16. ABSTRACT
          The report is a state-of-the-art review of the potential of combustion-type .
fuel additives in reducing air pollutant emissions from oil and coal firing.  It contains
two complementary parts: a review of the relation of combustion mechanisms to add-
itive action in controlling emissions; and a review of experimental investigations of
combustion-type fuel additives.  The technical literature  review revealed relatively
limited quantitative data from experimental investigations on combustion additives in
which conditions are well defined.  However,  there  is  evidence for some measure of
control of emissions by fuel additives.  The evidence  for  control by fuel additives of
visible smoke and carbon particulate is relatively strong; that for control of polycyclic
organic matter  is  somewhat weaker. The evidence  for control of NOx is quite weak.
Significant control of SO2 or total sulfur emissions  by fuel additives does not appear to
be possible, although SOS emissions can be reduced.  Practical  considerations and
other possible limitations to the use of additives are also reviewed.
17.
                             KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
                                          b.IDENTIFIERS/OPEN ENDED TERMS
                         c. COSATI Field/Croup
Air Pollution; Combustion; Fuel Additives *
Organometallic Compounds; Organic Com-
 pounds; Inorganic Compounds;  Residual
 Oils*; Coal; Boilers; Burners; Nitrogen
 Oxides; Carbon Monoxide; Hydrocarbons;
Sulfur Dioxide; Sulfur Trioxide; Polycy-
 clic Compounds: Smoke: Heat Transfer
Air Pollution Control;
Stationary Sources*;
Combustion Additives*;
Distillate Oils*; Parti-
 culate Emission*; Poly-
cyclic Organic Matter;
Cvclic Ooeration
                 13B; 21B; 21D
                 07C
                      07B
                      13A
jpei
FYCL
2QM
18. DISTRIBUTION STATEMENT
 Unlimited
19. SECURITY CLASS (Tliis Rtportj
Unclassified
                  21. NO. OF PAGES

                       96
20. SECURITY CLASS (Tltis page/
Unclassified
                         22. PRICE
EPA Form 2220-1 (9-73)

-------