oERA
United States     Industrial Environmental Research  EPA-600/7-80-043
Environmental Protection  Laboratory          March 1980
Agency        Research Triangle Park NC 27711
Pilot Scale Combustion
Evaluation of Waste and
Alternate Fuels:
Phase III Final Report

Interagency
Energy/Environment
R&D Program Report

-------
                  RESEARCH REPORTING SERIES


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

    1.  Environmental Health Effects Research

    2.  Environmental Protection Technology

    3.  Ecological Research

    4.  Environmental Monitoring

    5.  Socioeconomic Environmental Studies

    6.  Scientific and Technical Assessment Reports (STAR)

    7.  Interagency Energy-Environment Research and Development

    8.  "Special"  Reports

    9.  Miscellaneous Reports

This report has been  assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the  17-agency  Federal Energy/Environment Research  and
Development Program. These studies relate to EPA's  mission to protect the public
health  and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the  rapid development  of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants  and their health and ecological
effects; assessments  of,  and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
                        EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for  publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products  constitute endorsement or recommendation for use.

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

-------
                                       EPA-600/7-80-043

                                               March 1980
Pilot Scale Combustion  Evaluation
    of Waste and  Alternate  Fuels:
          Phase III   Final  Report
                         by

                  R.A. Brown and C.F. Busch

                    Acurex Corporation
                Energy and Environmental Division
                    485 Clyde Avenue
                Mountain View, California 94042
                   Contract No. 68-02-1885
                 Program Element No. EHE624A
              EPA Project Officer: David G. Lachapelle

            Industrial Environmental Research Laboratory
          Office of Environmental Engineering and Technology
                Research Triangle Park, NC 27711
                       Prepared for

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

-------
                                 ABSTRACT

       This report gives results of three studies at EPA's Multifuel
Test Facility.  The first evaluated a distributed-air staging concept
for NOV control in pulverized-coal-fired systems.  The results showed
      A
that minimum NO levels of 140 ppm were achieved at overall residence
times similar to those used during conventional staging tests.  However,
the NO levels achieved with the distributed-air concept were no lower
than those achievable with conventional staging.  The second evaluated
combustion control techniques and NO emissions when firing coal/oil
mixtures.  NO emissions for a given burner and nozzle were generally
proportional to the fuel-nitrogen content of the fuel.  Additionally,
combustion control technology currently used for NO  control from pul-
                                                   A
verized coal was found to be effective with coal/oil mixtures, but to
differing degrees, depending on the coal/oil mixture ratios and compo-
sitions.  The third evaluated emissions and combustion characteristics
of refuse-derived fuel (RDF) co-fired with either natural gas or pul-
verized coal.  Four RDF materials were evaluated for gaseous, particulate,
trace metal, and organic emissions.  In general:  CO and UHC emissions
were low; NOX and SOX emissions decreased with increasing RDF content
when co-fired with coal; particulate levels did not substantially in-
crease with the RDF; and no trace metal emissions correlation was found.
                                    iii

-------
                                 CONTENTS

Abstract	in
Figures	    vi
Tables	vii
Conversion Table  	   xii

     1.   Overview and Summary  	     1
     2.   Distributed Air Tests 	     7
             2.1  Special experimental hardware 	     9
             2.2  Test plan	    15
             2.3  Experimental data	    18
             2.4  Conclusions	    35
     3.   Coal/Oil Mixture Tests  	    37
             3.1  Test plan	    50
             3.2  Test data	    54
             3.3  Baseline tests	    56
             3.4  Control technology tests	    60
             3.5  Fuel nitrogen studies	    69
             3.6  Burner nozzle comparison  	    71
             3.7  Previous testing data	    71
             3.8  Summary and conclusions	    71
             3.9  Recommendation	    75
     4.   Refuse-Derived Fuel Tests	    76
             4.1  Objectives	    77
             4.2  RDF experimental hardware	    78
             4.3  Test theory and plan	 . .   101
             4.4  Analytical procedures 	   106
             4.5  Experimental data	114

References	166
Appendices

     A.   Data summary - distributed air; coal/oil mixture; RDF        169
                         testing; COM/DOE report  	

-------
                                  FIGURES
Number                                                                Page
                          Distributed Air Tests
 2-1   Distributed air concept as per Pershing 	      8
 2-2   Distributed air arrangement in the horizontal  extension ...     10
 2-3   IFRF burner	     12
 2-4   B&W-type coal  spreader	     13
 2-5   Baffle detail	     14
 2-6   Residence time in section Ib	     20
 2-7   Distribured air configurations  	     21
 2-8   Effect of SR]a	     24
 2-9   Effect of SRla	     25
 2-10  Effect of SR]b	     27
 2-11  Effect of la stage residence time	     28
 2-12  Effect of la stage residence time	     29
 2-13  Effect of Ib stage residence time	     31
 2-14  Effect of Ib stage residence time	     32
                          Coal/Oil  Mixture Tests
 3-1    EPA/Acurex multifuel  furnace	     39
 3-2   Facility modifications	     40
 3-3   Delavan swirl-air nozzle	     41
 3-4   Sonic  Corporation Sonicore nozzle 	     42
                                      VI

-------
                            FIGURES (CONTINUED)
Number                                                                Page
 3-5    Coal/oil delivery system 	     49
 3-6    Baseline emissions 	     58
 3-7    Boiling point curves 	     59
 3-8    Nitrogen evolution curves  	     61
 3-9    Staging emissions  	     62
 3-10   Burner air distribution	     64
 3-11   Burner air distribution	     65
 3-12   Air distribution tests	     66
 3-13   Effect of firing rate and residence time	     68
 3-14   Effect of fuel nitrogen	     70
 3-15   Nozzle comparison  	     72
 3-16   Date from earlier work	     73
                         Refuse-Derived Fuel Tests
 4-1    Furnace cross section  	     79
 4-2    Tangential configuration, aerodynamic patterns 	     80
 4-3    Corner-fired burner	     81
 4-4    RDF nozzle	     82
 4-5    Modified corner-burner assembly  	     84
 4-6    Fuel delivery schematic	     85
 4-7    RDF feed system design   	     87
 4-8    Pneumatic transport system 	     88
 4-9    Safety system	     90
 4-10   Sampling system online at experimental multiburner furnace .     95
 4-11   Aerotherm high volume stack sampler   	     96
                                       vii

-------
                            FIGURES (CONTINUED)

Number                                                                Page

 4-12   Source assessment sampling system (SASS)  	    97

 4-13   Ash deposition	    102

 4-14   Test matrix for baseline emissions  characterization   ....    104

 4-15   Test matrix for emissions control  through  theoretical
          air variation	    105

 4-16   Test matrix for baseline emissions  characterization   ....    107

 4-17   Test matrix for emissions control  through  theoretical
          air variation	    108

 4-18   Photographs of fuel  samples	    117

 4-19   NO emissions  during  baseline  testing  (Ames)   	    120

 4-20   NO emissions  during  baseline  testing  (Richmond)   	    121

 4-21    NO emissions  during  baseline  testing  (Americology) 	    122

 4-22   NO emissions  during  baseline  testing  (San  Diego)  	    123

 4-23   Thermal  NO  (previous work)    	    124

 4-24   NO emissions  during  baseline  testing  (all  RDF's)  	    126

 4-25   S02  data  (all  RDF's)	    127

 4-26    NO emissions  during  detailed  testing  (Richmond RDF/
          Pittsburgh  coal	    129

 4-27    Fuel  nitrogen  contribution	    130

 4-28    Stack  gas particle size  vs. cumulative percent less
          than diameter	    137

 4-29    Particulate loading  results   	    138

 4-30    Stack  gas particulate size vs. cumulative  percent
          less than diameter - trace metal Cu	    149

 4-31    Stack  gas particle size  vs. cumulative percent
          less than diameter — trace metal Zn	    150
                                     vm

-------
                            FIGURES (CONCLUDED)

Number                                                                Page

 4-32   Stack gas particle size vs. cumulative percent
          less than diameter - trace metal  Mn	    151

 4-33   Stack gas particle size vs. cumulative percent
          less than diameter - trace metal  Pb	    152

 4-34   Stack gas particle size vs. cumulative percent
          less than diameter — trace metal  Cd	    153

 4-35   Stack gas particle size vs. cumulative percent
          less than diameter - trace metal  Be	    154

 4-36   Stack gas particle size vs. cumulative percent
          less than diameter - trace metal  Ti  	    155

 4-37   Stack gas particle size vs. cumulative percent
          less than diameter - trace metal  Sb	    156

 4-38   Stack gas particle size vs. cumulative percent
          less than diameter - trace metal  Sn	    157

 4-39   Stack gas particle size vs. cumulative percent
          less than diameter - trace metal  Hg	    158

 4-40   Stack gas particle size vs. cumulative percent
          less than diameter - trace metal  As	    159
                                      IX

-------
                                  TABLES
Number                                                                Page
                           Distributed Air Tests
 2-1     Optimum Staging Parameters (Pershing)  	     9
 2-2     Range of Parameters of Interest	    16
 2-3     Revised Distributed Air Studies Matrix  	    17
 2-4     Residence Times in la  Stage	    19
 2-5     Distributed Air Versus Conventional  Staging  	    34
 2-6     Facility Characteristics	    35
                          Coal/Oil  Mixture Tests
 3-1     Emission Monitoring Equipment  	    44
 3-2     Fuel  Oil  Analyses	    45
 3-3     Coal  Analyses,  As-Received Basis   	    46
 3-4     Coal/Oil  Mixture Analyses,  As-Received Basis, 30% Coal
          by  Weight	    47
 3-5     Baseline Matrix   	    51
 3-6     Effect of Load  and Residence Time	    52
 3-7     Distributed Air Burner Tests   	    53
 3-8^   Effect of Fuel  Nitrogen	    55
                        Refuse-Derived Fuel Tests
 4-1     Emission  Monitoring Equipment  	    94
 4-2     Fuel  Analysis	106

-------
                            TABLES (CONCLUDED)
Number                                                                Page
 4-3    Metals Which Were Analyzed .................    109
 4-4    Liquid Chroma tog raphy Elution Sequence  ...........    110
 4-5    Distribution of Compound Classes in Liquid Chroma tographic
          Fractions of Organic Extracts  ..............    Ill
 4-6    Test Matrix .........................    115
 4-7    Fuel Analyses  .......................    116
 4-8    Parti cul ate Analyses:  Effect of RDF Type  .........    131
 4-9    Particulate Analyses:  Effect of Excess Air  ........    133
 4-10   Particulate Analyses:  Effect of Percent RDF   .......    135
 4-11   Particulate Analyses:  Coal vs. 10% RDF + Coal  vs.
              RDF + Gas .......................    136
 4-13   Total Trace Metal Loadings -Coal Cofiring .........   142
 4-14   Trace Metal Concentrations as Vapor - Coal Cofiring  ....   144
 4-15   Total Trace Metal Loadings -Gas Cofiring  .........   145
 4-16   Trace Metal Concentrations as Vapor - Coal Cofiring  ....   146
 4-17   Trace Metal Concentrations - Pilot vs. Full Scale -
          Particulate Only .....................   148
 4-18   Organics Found  .......................   161
 4-19   LC Column Data  .......................   163
 4-20   LC Column Data  .......................   164
 4-21   Possible Compounds in LC Fractions not Analyzed   ......   165
                                       XI

-------
                               CONVERSION  TABLE
                    ENGLISH  TO  SI METRIC CONVERSION  FACTORS
 To  convert  from                    To                      Multiply by
    inch                            m                   2.540 000  E-02
    foot                            m                   3.048 000  E-01
    scfm                          m3/s                  4.719 474  E-04
    gr/ft3                        g/m3                  2.288 352  E+00
    gallon                          m3                   3.785 412  E-03
    gph                           m3/s                  1.051 503  E-06
    Btu/hr                          Wt                   2.930 711  E-01
    Ib/hr                         kg/s                  1.259 979  E-04
    Btu/lb                        kJ/kg                 2.326 000  E+00
    Btu/hr-ft3                    Wt/m3                 1.034 971  E+01
    Ib                              kg                   4.535 924  E-01
    psig                          kPa                   6.894 757  E+03
    ug/Btu                        yg/J                  9.478 170  E-01
    °F                              °C                   t°C = (t°F-32)/1.8
scfm   = standard cubic feet per hour
gr/ft  = grains per cubic foot
gph    = gallons per hour
Btu    = British thermal unit
hr     = hour
Ib     = pound
ft3    = cubic foot
psig   = pounds per square inch (gauge)
yg     = micrograms
m      = metre
 3
m      = cubic meter
s      = second
g/m    = grams per cubic meter
Wfc     = thermal watt
kg     = kilogram
kJ     = kilojoule
kPa    = kilopascal
                                    xii

-------
                                   SECTION 1

                             OVERVIEW AND SUMMARY


       The work summarized in this report was performed during the period

October 1977 to July 1978 as Phase III  of the Pilot Scale Evaluation of

N0₯ Combustion Control Techniques, EPA  Contract 68-02-1885.   This report
  A

discusses

       •   Advanced NOY Control Techniques for Pulverized Coal Through
                      A

           Distributed Air

       •   Emissions and NO  Control Technology Evaluation of Coal/Oil
                           rt

           Mixtures (COMs)

       •   Evaluation of Emissions on Co-firing of Four Refuse-Derived

           Fuels (RDF) with Natural Gas and Pulverized Coal

A brief summary of the scope and results from each of these tasks follows.

Distributed Air Tests

       Tests at the University of Arizona in a bench-scale coal-fired fur-

nace suggested that low NO levels could be achieved in relatively overall

short residence time  by sequencing the air into the burner in three stages.

The primary zones  include the  primary air conveying coal and  some secondary

air.  The  stoichiometric ratio of this first stage would be in the  range

of 0.3 to  0.6 and  a residence  time of 0.3 to 0.75 seconds.  Tertiary air
  It  is  the  EPA  policy  to  use SI metric units; however,  in this report
  English  units  are  occasionally used  for convenience.   See attached con-
  version  table.

-------
was then added through four ports at 90° to the flow of combustion products.
This second stage is held at a stoichiometric ratio of 0.75 to 0.95, and
staging air is then added through four additional ports.
       This series of tests explored a range of stoichiometric ratios and
residence times for each stage.  Unfortunately, the results from these
tests could not duplicate the Arizona tests.  What was found was that even
with the distributed air approach, NO levels increase with decreasing SR
below an SR of about 0.6.  In addition, NO always decayed with increasing
residence time in both the first and second stages.  Minimum NO levels of
about 140 ppm were achieved in overall residence times similar to the con-
ventional staging results.  Thus, no advantage over conventional staging
was achieved.  These results may be partially explained by the fact that a
diffusion burner was used in a relatively low L/D firebox as compared to a
premixed burner in a high L/D firebox in the Arizona tests.
Coal/Oil Mixture Tests
       In order to utilize coal in the near term, it has been proposed to
fire oil or gas boilers with a slurried mixture of coal and oil.  Although
the feasibility has been demonstrated in a number of small and larger scale
demonstrations, there is a need to determine the environmental problems
associated with COMs.  Because of the generally higher fuel-N content of
the mixture as compared to oil, it seems likely that the NO  levels would
                                                           /\
also be higher.  Therefore, the purpose of this study was as follows:
       •   Obtain emission data for coal/oil combustion in an environment
           closely simulating an industrial package boiler
       •   Determine if emissions levels were affected by the fuel  com-
           position
       •   Determine if conventional control technology developed for coal

-------
           is effective in reducing  emissions  levels  produced  by  coal/
           oil combustion
       •   Investigate the effect of burner modification on emission
           levels produced by coal/oil  combustion.
       During this study, two oils and  three coals  were fired  in  a package
boiler simulator.  Baseline emissions of the parent fuels and  the fuel  com-
binations were determined.  Control  technology tests  were run  on  the  COMs
as follows:
       •   Baseline NO emissions from COM were, in  general, proportional
           to the fuel-N for a given burner and nozzle type.
       t   The burner settings and fuel nozzle type have a strong effect
           on NO emissions.
       •   Conventional control technology currently utilized for pulver-
           ized fuel combustion is effective in reducing NO emissions, but
           to different degrees, depending on fuel  composition
       •   NO emissions increase in proportion to the amount of coal  in
           the coal-oil mixture, but fall between the parent fuel oil and
           coal baseline emissions.
If there is to be significant utilization of COM in industry, and if NO
levels are to be controlled, much additional work is needed to understand
the mechanisms which control NOV formation in COMs for different fuel
                               A
combinations.
Refuse-Derived Fuels Testing
       It  is  necessary that investigations regarding the environmental
compatibility of RDF be conducted before this vast, untapped energy source
can be considered a viable supplement  to present energy  resources.  It  has
thus been  suggested that  such  investigations can be carried out  most cost

-------
effectively  in a pilot-scale facility.  To determine the feasibility of
such  pilot-scale testing, the IERL of EPA/Cincinnati funded a study as
part  of the  Phase  III activity.  The goals of this study were as follows:
       •   To design, fabricate, and operate a system for combustion
           testing of RDF in a laboratory scale facility
       0   To characterize RDF emissions of several types of material
           presently available for use as fuel
       •   To evaluate the combustion efficiency of fuels consisting of
           conventional clean and dirty fossil fuels (natural gas and coal)
           mixed with varying percentages of refuse
       •   To evaluate the effects of combustion parameters on emissions
           from RDF/conventional fuel mixtures
       t   To provide direction for future investigations on refuse-1 derived
           fuel to insure solutions to problems associated with its use.
       A feed system was designed to control  and measure from 10 to 60
Ib/hr of a variety of "fluff" refuse-derived materials.   This was accom-
plished using a rotating drum hopper depositing on an internal moving belt
conveyor.  The conveyor, in turn, deposits the material  into a tube where
it is conveyed by air into the top port of a tangential  burner.  The pilot-
scale facility was tangentially-fired at 1.5 x 106 Btu/hr with RDF fed to
two of the four corners.  The test program was designed to determine the
gaseous, particulate, trace metal, and organic emissions of four sources
of RDF.  The four materials were from San Diego, California; Richmond,
California; the Americology Facility in Milwaukee, Wisconsin; and from the
Ames, Iowa Plant.  NO emissions increased with both percent RDF and percent
excess air when fired with natural gas.  NO emissions also varied for the

-------
four fuel types in approximate proportion to the fuel-N content of the RDF.
When co-fired with coal, the NO emissions decreased with increasing percent
RDF, even though the percent N available increased.  This may be attributed
to shielding of the coal by the RDF from the oxygen, flame/flame processing
or locally fuel-rich zones in the coal stream caused by redistribution of
the combustion air during the RDF firing.
       Particulate levels did not substantially increase with the RDF, but
a higher concentration was found in the less than ly size range.  No corre-
lations were found with trace metal emissions, either with respect to
percent RDF or percent excess air.  It is believed that, due to the great
variability in the feed from minute to minute and the problems with holdup
in the heat exchanger sections, a valid trace metal evaluation is not pos-
sible from a single test.  Many tests will be necessary to form a statisti-
cally reliable number.  Lastly, few poly-organic materials (ROMs) were found
and no poly-chlorinated biphenyls (PCBs).
       Combustion efficiency of these pilot-scale  tests was  perhaps better
than  full-scale  tests due  to  higher combustion  temperatures.   Additional
tests are  needed to better determine  the  variability  of gaseous,  trace
metal, and organic  emissions  from a single  source  and from a variety  of
sources.
       In  the  sections  that follow, the  results from  each of these test
programs will  be explained in more  detail.   In  all  cases they were per-
formed in  EPA's  multifuel, multiburner test facility  located at Acurex
 Corporation in Mountain View, California.  The  facility was  either used
 in its normal  utility boiler configuration  (main firebox only) or with
 the horizontal extensions which can simulate a  package boiler configuration

-------
or serve as an asymmetric flow combustion system.  Details of this facility
may be found in the Phase II report (Reference 1).  A description of the
special equipment required for each of the test programs and details of
the configurations for those tests are included within the section on
that program.

-------
                                 SECTION  2
                           DISTRIBUTED AIR  TESTS

       The conventional  staging  studies performed  in  the main  firebox  and
horizontal extensions (Reference 1)  of the  EPA Multiburner Test Facility
revealed that low NO  levels could be achieved with a sufficiently long
                    A
residence time under fuel-rich conditions.   However,  this approach neces-
sitates an exceptionally long residence time under fuel-rich conditions
with potential for corrosion and slagging problems.  Thus, an alternate
approach was sought.
       Tests run on a subscale premixed combustor at the University of
Arizona (Reference 2) indicated that a three-stage approach would be able
to achieve low NOX levels in an overall residence time of less than 1.5
seconds.
       The approach is illustrated for the Arizona facility in Figure 2-1.
In this arrangement premixed coal, primary air, and secondary air enter
the top of the furnace at a stoichiometric ratio SR-|a.  This stoichiometry
is held for a residence time T.J  seconds, whereuopon tertiary air is intro-
duced.  The second part of the first  stage is held at a stoichiometric
ratio of  SR^b for i^ seconds.  The second-stage air is then introduced
for final burnout under excess air conditions for a  residence time of T2
seconds.  Pershing found that for his  facility a unique combination of

-------
  1st
 Stage
 2nd
Stage
                            Coal  +  primary  air
                            + secondary  air
                          la
Tertiary
   air
                                                    •la
                                                    lb
                                          Second
                                           stage 	j-
                                            air
                                                    T2
                                     Flue
Figure 2-1.   Distributed air concept as  per Pershing,

-------
SR, ,  T,  ,  SR,. ,  and T,.  achieved the lowest NOV level.  The optimum condi-
   ia   ia    iD       ID                       x
tions  for the Pershing experiment are tabulated in Table 2-1.  In summary
he found the following:
       •   Stack NO  decreased with decreasing SR,  until an SR-|a = 0.5.
           Below an SR,   = 0.5, stack NO did not decrease further for con-
           stant conditions downstream
       •   An optimum T,  was necessary to achieve minimum NO .  On either
                        i a                                    x
           side of the optimum, NO  would increase.
                                  /\
       •   NOX decreased with decreasing SR,b
       •   NOX increased with decreasing T-,.
       »   T? residence time had no effect on the stack  NO
            L-                                             A
             TABLE 2-1.  OPTIMUM STAGING PARAMETERS  (PERSHING)
SRla
SRlb
Tla
Tlb
0.4 - 0.5
0.85
0.4 sec
1-1.5 sec
(not varied)
This approach was  thus  investigated  using  the  EPA multiburner-horizontal
extension  test  facility.
2.1     SPECIAL  EXPERIMENTAL  HARDWARE
        It  was decided that the  horizontal  extensions  would be the best
equipment  to perform these tests.  The horizontal extensions were set up
as  shown  in Figure 2-2.   Each horizontal  extension is 33-inch inside diam-
eter,  refractory lined,  and  two feet long.  Up to five sections may be

-------
                                    TERTIARY AIR
STAGED AIR
SECONDARY AIR
       PRIMARY
       + FUEL
       PRIMARY
       + FUEL
SECONDARY AIR
                                                                  SRioT - 1.15
           Figure  2-2.  Distributed air arrangement  in the horizontal extension.

-------
joined together to form an overall  length of 10 feet.   A transition  section
connects to the main firebox where  the flue gases  are  then  quenched  by  the
	 .~_r 	   heat exchange sections.   Gaseous emissions  are sam-
pled just downstream of the heat exchange section.  On the  other end of the
horizontal extensions, either a single burner or up to five burners  may be
mounted.  For this test, four of the nominal 300,000 Btu/hr IFRF variable
swirl block burners were chosen.  These burners were fitted with a 2-inch
diameter  sleeve in the air throat to achieve reasonable velocities under
very fuel-rich conditions.  The burner is illustrated in Figure 2-3.  The
burners also used the B&W-type coal spreader illustrated in Figure 2-4,
and the swirl was set at a mid-position of four.  Part of the objective
for utilizing these four burners in this configuration was to achieve a
well-mixed first  stage.  To further enhance the mixing  in each  stage, baf-
fles were used whenever possible at the  end of the  la  stage and Ib  stage,
as was  illustrated  in  Figure  2-3.  These  baffles  also  served to separate
the stages and prevent  backmixing  into the  first  stage.  The baffle  or  choke
was made  from a high  temperature refractory in four sections as shown  in
Figure  2-5.  This arrangement made it relatively  easy  to move the baffle to
any desired  location  within  the tunnel.   At the first  tertiary  air  position,
it was  not possible  to  install  the baffle due  to  the  close  proximity to
the  burners.  The baffle  opening was  16  inches in diameter.  The  tertiary
air  was introduced  just downstream of the first baffle in  four  locations
90  degrees apart.  This air was introduced through 2-inch  diameter  ports
 perpendicular  to  the main flow. The  horizontal extensions have four ports
 90  degrees apart  every foot along  the length  of the furnace.  The first
 four locations,  1 foot apart, were chosen to  vary the tertiary  air residence
                                       11

-------
ro
                                              Figure 2-4.  IFRF burner.

-------
     27.5°,
Figure 2-4.  B&W-type coal spreader.

-------
Figure 2-5.   Baffle detail.

-------
time t,  .   This enabled the la stage residence time to be varied from
      i a
less than  0.5 second to over 2 seconds.   The rlb residence time was then
varied by  positioning the staging air ports from 2 to 7 feet downstream of
the tertiary air position.  No attempt was made to control the second-stage
residence  time, but it was always sufficient to complete combustion.
       Temperature of the secondary and tertiary air was maintained at about
600°F and the stage air at 300°F.  Bare platinum-platinum/rhodium thermo-
couple measurements were made in the la stage and Ib stage.
2.2    TEST PLAN
       The  tests were  structured to explore the following  variables  in  the
distributed air concept:
       •    la  stage residence time            T,
       •    Ib  stage residence time            T-,.
       •    la  stage stoichiometric  ratio      SR-,
       0    Ib  stage stoichiometric  ratio      SR,.
       •    Firing  rate
        •    Temperature
        The  range  of each  parameter  of interest  is given in Table 2-2.
Table 2-3 shows the matrix that was run.   This  matrix is the matrix which
was developed  ;during  the  course of  the  testing  as the results redirected the
effort.   It was found, for instance,  that a fairly dense matrix was needed
to truly  see the  effects  of the various parameters.  In general two to
 three la  stage residence times and two Ib residence times were selected
 for each  tertiary air position.  The la stage stoichiometric ratio was
 varied from 0.3 to 0.7, and two Ib stoichiometries of 0.85 and 0.95 were
 selected.  The firing rate changes not only effect early mixing, but the
                                      15

-------
TABLE 2-2.  RANGE OF PARAMETERS OF INTEREST
    Parameter
         Range
       SR
         la
       SRlb
       Tla
       Tlb
   Firing Rate
   Temperature
       0.3 -0.7
       0.8 -0.95
       0.5 - 3.0
       0.5 - 1.5
0.85 - 1.7 x 106 Btu/hr
    Ambient - 600°F
                      16

-------
                                                TABLE 2-3.  REVISED DISTRIBUTED AIR STUDIES MATRIX3
Distributed
Air Studies
Matrix
r*
r—
S-
C.
~>.
3
4*
CO
PI
u> «
O .—
•a

nj
01
£
d.
U-
o
O
O
*l«S
s
£
Q.
LL.
§
+*«>
3
V
l-
o_
CD
5
in
tn
^j°
ae.
on oo
0
in
en
J3°
ee.
oo oo
d
in
CTl
^o
cc.
00 CO
o
in

en
^i°
cc.
00
00
0
Position #1 - T,
tlb - Short
0.3











209b
(b)
S^
0.45











209a
(b)
a
0.6











209c
(b)
T - Long
0.3









209J
a)

209g
209t
a)
SRla
0.45








2091
(a)
209k
(a)
209h
(a)
209e
(a)
0.6









2091,
(a)

209d
(a)
Position #2 - T]fl
T,. - Short
0.3
21 3x
(f)
21 3u
(f)


21 Si-
te)
21 3 j
(e)




213d
(e)
21 3c
e)
SRla
0.45
21 3v
(f)
213s
(f)


21 3h
(e)
213k
(e)




21 3e
(e)
21 3a
(e)
0.6
21 3w
(f)
21 3t
(f)


213g
(e)
213*
(e)




213f
(e)
21 3b
(e)
tlb - Long
0.3




21 3p
(f)
213o
(f)






SRla
0.45




21 3q
(f)
21 3n
(f)






0.6




213r
(f)
213m
(f)






Position #3 - T,
T,. - Short
ID
0.3
21 2J,
(k)
211b
(k)


21 2g
(h)
211j
(h)


212f
h)
212b
h)


SR1a
0.45
212k
(k)
211a
(k)


21 2h
(h)
211k
(h)


21 2e
(h)
21 2a
(h)


0.6
21 2j
(k)
211c
(k)


212i
(h)
21H
(h)


21 2d
(h)
212c
(h)


Tlb - Long
0.3
21 Ob
(j)
21 Of
(J)


211i
(k)
211f
(k)






SRla
0.45
21 Oa
(j)
210e
(j)


211g
(k)
211d
(k)






0.6
21 Oc
(J)
21 Od
(j)


211h
(k)
211e
(k)






Position *4 - T,
1 
-------
 residence time between any two points in the furnace and the temperature  at
 any point.   A few tests were also run with no preheat and the lowest firing
 rate to further enhance any temperature effect.   When no effect was  seen
 at this lowest load,  further tests of temperature were exchanged for a  more
 complete  matrix in  other regions.
        To aid  in  estimating the effect of  residence  time,  a  plot and  tables
 were prepared  to  determine the residence time for a  given  staging position,
 firing  rate, and  stoichiometric ratio.  Table 2-4 presents this  data  for
 the  la  residence  time  as  a function of tertiary air  position, stoichiomet-
 ric  ratio, and  firing  rate.   Figure 2-6 presents  residence time  plots for
 the  Ib  stage as a function of firing  rate  and SR^, .   The absissa of this
 plot is the distance between  the  tertiary  air position and the stage air
 position.  Also shown  on  this plot are  the various configuration letters
 associated with each test  point.  The matrix  in Table 2-3 gives  the configu-
 ration  letter  for each  test number, and the configurations are depicted in
 Figure  2-7.  Thus,  by  referring to the matrix, the configuration letter can
 be determined  and seen  schematically  in Figure 2-7.   By referring to Figure
 2-6  and knowing the firing  rate and SR,. ,  the  residence time in  the Ib stage
 may  be determined.  This procedure together with  the  table for T,  were used
 to reduce the data to common  residence times.  Thus,  the true effect of the
 stoichiometric  ratios and,  conversely, the effect of  residence time at con-
 sistent stoichiometric ratio could be determined.
2.3    EXPERIMENTAL DATA
       The data have been compiled in this section in its reduced form.
First, all data were reduced to an air-free (0 percent 02) basis.  Then
cross plots were made so that the true effect of any parameter could be
                                      18

-------
TABLE 2-4.  RESIDENCE TIMES IN la STAGE
             (T,  — seconds)



IT
-C
TJ 3
(0 -P
O CO
(A
0
1.7
1.3

0.85
Position 1
SRla
0.3
0.487
0.637

0.974
0.45
0.354
0.464

0.711
0.6
0.280
0.366

0.560
Position 2
SRla
0.3
0.885
1.16

1.77
0.45
0.644
0.84

1.29
0.6
0.509
0.67

1.02
Position 3
SRla
0.3
1.47
1.93

2.94
0.45
1.06
1.39

2.14
0.6
0.84
1.11

1.69
Position 4
SRla
0.6
1.17
--


0.7
1.10
--



-------
ro
O
                2.8

                2.6

                2.4

                2.2

                2.0

                1.8
            o>   1.6
            E
       a,  1.4
       c
       0)  , „
       -a  1.2
       in
       
-------
External configuration
Internal configuration
          "b"
          "d"
          "h"

i

}

* J
1
1
1
i
i
\
i
i
i



i
i
J *
i
i
i
i

i

i

i

i
i


. t
i
't *

*

*


i
Y t *
i
i
















1
1
1
1

                Figure 2-7.  Distributed air  configurations,
                                    21

-------
Configuration
    number
     "n"


















1
1
1
t





1
1
1
t
1
i
1
1
t

1
1
1
t
*
1
1
t
















1 1

1
1. 1

1
I 1

~~*
T



o
o
o
1 i









\ 1
o
o
0

         Figure 2-7.  Distributed air configurations (concluded)
                                     22

-------
determined.   Each of the various parameters  will  be discussed  in  the  sub-
sections that follow.
2.3.1  la Stage Stoichiometric Ratio
       Figure 2-8 shows the effect of the la stage Stoichiometric ratio  at
a constant SR^ of 0.80, la stage residence  time  of 1.00 seconds, and t.,
residence time of 0.97 seconds.  As can be seen on the plot,  the  effect  was
fairly pronounced with NO decreasing with increasing SRla.   Previous  tests
indicated that above an SR of 0.7-0.8, the NO levels would again  increase.
This result is in contrast to the Pershing result which indicated that there
was no effect of SR.  below a level of SR,  0.50.  Actually,  this data is
consistent with the data taken previously in this facility with conventional
staging.  It was suggested in that previous  work that this increase in NO  at
low Stoichiometric ratios was due to the formation of .second-stage NO.
       A similar trend was found at an SR,.  = 0.95 as seen in Figure 2-9.
There is some question as to the validity of the points at 1.7 x 106  Btu/hr,
particularly at the lower SR,  's because the baffle collapsed during these
tests.  Nevertheless, the data still indicate that NO increases with de-
creasing Stoichiometric ratio.  The NO levels are about 100 ppm higher,
however, at an SRlb of 0.95 than at an SRlb of 0.80.  This suggests that
some second-stage NO is being formed and is either dependent on the Stoi-
chiometric ratio in  the Ib stage or on a dependent variable of the Stoi-
chiometric ratio.  Recent tests  (Reference 3), for example, have shown that
second-stage NO is strongly dependent on the initial flame temperature in
the second stage, provided this  temperature is below 2200°F.  However, if
it was an effect of  temperature, in this case we would expect to see an
increase of NO with  firing tate.   In both plots this  is true at  an SR
                                     23

-------
    600
    500
    400
 C\J
0   300
o
o
    200
    100 ~
               0.30
T,  = 1.0 sec
  I a
•»lb ^ 0.97 sec
R   - 0.80
                                Load
                       O  1-7 x 10  Btu/hr
                       A  1 .3 x 10  Btu/hr
                       D  0.85 x 10  Btu/hr
                                                     _L
0.40               0.50
          SRla
                                                                      0.60
                                                               0.70
                                 Figure  2-8.   Effect  of

-------
ro
en
               600
               500
              400
              300
              200
              100
  Effect  of  SR]a
  T.J   =  1.0  sec
  •tlb  = 0.97  sec
      - 0.95  sec
                                                                          O  1 .7  x  10"  Btu/hr
                                                                          A  1 -3  x  10*'  Btu/hr
                                                                          D  0.85 x  10f Btu/hr
                        0.30
0.40
0.50
0.60
0.70
                                                         SR
                                                           la
                                             Figure  2-9.   Effect of SRlg(.

-------
 of 0.3,  but  not  at  0.6.   Previous data  (Reference 1) indicated the higher
 the first-stage  temperature, the greater the decay rate in the first stage.
 Thus,  it is  possible that there are competing effects with an optimum decay
 at the high  load and 0.6 stoichiometric ratio, while at the 0.3 SR-,  greater
 heat release  is experienced in the second stage causing the second-stage NO
 to increase.   In fact, this is a possible explanation for the decrease of
 NO with  increasing  stoichiometric ratio up to an SR,  of 0.7.  (It is known
                                                   i a
 from previous work  that the NO will again increase beyond an SR of 0.7).
 2.3.2  1b Stage Stoichiometric Ratio
       The effect of SRlb is shown in Figure 2-10 for an SRla of 0.3 and 0.6.
 As  can be seen, NO  always increased with an increase in SR,, .  Again, this
 may be either due to the greater degree of oxygen availability or may re-
 flect an increase in flame temperature in the Ib stage.  This data is also
 interpolated data at constant residence times in the la and Ib stages.
 2.3.3  Residence Time, la stage
       The effect of the la stage residence time is shown in Figure 2-11
 and 2-12 for an SRlb of 0.80 and 0.95, respectively, and an SRlg of 0.6
 (the point at which minimum NO occurred).   The data for the three firing
 rates is also included on these plots.  These data are stack emisions at
 an overall excess air level  of 15 percent.   The data indicates a strong
 decay between 0.5 to 1.0 second at all loads and then a decrease in the
decay rate following 1.0 second.   The data  also suggest that at the higher
 load the initial  NO levels are higher, but  the decay rate is also higher
 resulting in lower NO levels after a T,   of 1.0 second.  This is consistent
with previous results which show lower NO levels at  higher load or first-
 stage temperatures.   The main conclusion from this curve is that little
                                     26

-------
ro
            C\J
           o
o
o
                600
                500
                400
                300
                200
                100
                      =  1.0 sec
                      =  0.97 sec
                                                                              SRla = 0.3
                                                                              SRn
                                                                              O  1-7  x  106  Btu/hr
                                                                              A  1.3  x  106  Btu/hr
                                                                              D  0.85 x  10*  Btu/hr
                                   0.80
                                                        0.95
                                                    SR
                                                       Ib
                                             Figure  2-10.   Effect of SR,

-------
PO
oo
               600
               500
               400
            Q.
            Q_
            C\J
           O
               300
               200
               TOO
Western Kentucky coal
4 IFRF burners
Smallest sleeve
B&W spreader, SW = 4
Excess air = 15;
SR]   = 0.60
SRit, = 0.30
     = 0.97 sec (all values converted
Tlb
                                              to this common T,, )
                                                      O  1.7  x  10"  Btu/hr

                                                      Q  1.3  x  106  Btu/hr
                                                      0  0.85 x 106 Btu/hr
                                      1    1
                               1
                                    1
                                                 1    1     1     1
1     1
0.4       0.8       1.2       1.6       2.0

                                Tla  (56C)
                                                                          2.4
                                                      2.8
                                                                3.2
        3.6
                                  Figure 2-11.  Effect of la stage residence time.

-------
               600
               500
               400
ro
vo
            Q.
            CL
            o

            o
               200
               100
                          0.4
        Western  Kentucky  coal
        4  IFRF burners
        Smal lest sleeve
        B&W  spreader, SW  =  4
        Excess air = 15,*
             = 0.60
             - 0.95
             = 0.97 (all
                                                              values  converted
                     to thi s
                                                                  common i ., )
                                                                 I
0.8
1.2
1.6
 2.0

(sec)
0.4
                                               1-7 x 106 Btu/hr

                                               1.3 x 106 Btu/hr

                                               0.85 x 10* Btu/hr
0.8
3.2
3.6
                                  Figure 2-12.  Effect of  la  stage  residence  time.

-------
 additional  benefit  is gained  past a residence time of about 1 second.
 Similar  curves were made for  lower SR-,  's and the general trend is the
 same  as  at  this stochiometry.  These curves were used to determine the
 real  effect of the  la stage stoichiometry at a constant T,  presented
 earlier.
 2.3.4  Residence Time, Ib Stage
       The  slow decay experienced in the la stage past a residence time
 of  1  second  appears to continue in the Ib stage as shown in Figure 2-13
 at  an SR^  of 0.80.  This data is at a constant injection position so that
 T,  will be  varying with SR-,  and firing rate.  However, at the third ter-
 tiary air position, the residence time for most SR^'s and firing rates is
 sufficiently long (>1.0) that the la stage residence time should not seri-
 ously effect the results.  It is thus interesting to note that the data at
 a firing rate of 1.3 and 1.7 x 106 Btu/hr coincides at the same la stage
 stoichiometric rato.  However, at the lower firing rate of 0.85 x 106 Btu/
 hr, the NO levels appear to be a bit lower.   This could possibly be due
 to the longer residence time in the la stage and/or due to a lower tempera-
 ture environment.   In the Ib stage at an SR,.  of 0.95, the decay rate was
 generally less than at an SR,.  = 0.80 as illustrated in Figure 2-14.
       In summary then,  stack NO levels decayed both in the la stage and
the Ib stage.  The decay appears to be fairly rapid in the initial  1  second
 in the la stage,  then drops off to a relatively slow decay in the Ib stage.
2.3.5  Firing Rate
       In the previous sections on the effects of SR,  , SR,b, T,   and T,.,
the effect of firing rate has been discussed.   Since firing rate effects
 both local mixing, temperature and residence time between two given points,
 it is often difficult to determine which of  these effects is is  predominant.
                                    30

-------
    480

    440

    400

    360

    320

    280
 CM
0   240
O

o
200

160

120

 80

 40

  0
               0.5
                Western Kentucky coal
                4 I FRF burners
                B&W-type spreader, SW = 4
                600°F preheat
                15".  excess air
                Third tertiary air position,
                   SRlb = 0.8
                          O  0.85 x  106  Btu/hr
                          O  1 .3 x 106  Btu/hr
                          A  1-7 x 106  Btu/hr
                 I    I     I     I    I     I    I     I     I    I     I
                     0.7
0.9      1.1
1.3
1 .5
                                     T-,,  residence  time  (sec)
                                      I h
                                                                            SRla = 0.3
                                                                            SR1a = 0.45
                                                                            SRla =0.60
1.7
1.9
                   Figure 2-13.  Effect of Ib stage residence time.

-------
CO
ro
                600
                500 -
                400 -
                300 -
                200 -
                100 -
Western Kentucky coal
4 IFRF burners
B&W spreader, SW = 4
_ 600°F preheat
15" excess air A
Third tertiary air position — 	 _
SRlb = 0.95
A_ 	
u + 	 3

O
Fully shaded - SR = 0.3
Half shaded - SR - 0.45
Open - SR = 0.6
1 1 1 I 1 l
) 0.2 0.4 0.6 0.8 1.0 1.2
	 —A
A'

O 0.85 x 106 Btu/hr
O 1 -3 x 106 Btu/hr
A 1.7 x 106 Btu/hr
l 1
1.2 1.4 l.f
                                                             Ib
                                   Figure  2-14.   Effect of Ib stage residence time.

-------
Thus, we see different effects depending on SR,   and SR,. ,  or T,  .   However,
in summary, the only area where the attributes of firing rate are beneficial
is at the optimum SR,  of 0.6 and a  T,   > 1.0 seconds with either an SR,K
                     i a                i a                                i D
of 0.80 or 0.95.  It is believed this is primarily due to a more  rapid decay
rate in the la stage associated with higher temperature.
2.3.6  Effect of Temperature
       Except for the resultant effect of temperature noted in the pre-
vious section due to firing rate, only a few tests were run with no pre-
heat to the secondary air.  These tests run at a firing rate of 0.85 x
106 Btu/hr showed no significant effect.  It is possible that the change
in secondary air temperature was not significant enough to change the com-
bustion temperatures.
2.3.7  Comparison with Conventional Staging
       The question may be asked:  Has anything really been gained by going
to this more complex staging arrangement?  Table 2-5 will aid in answering
this question.  Let's consider the optimum SR,  = ~0.6 seconds in the la
stage with an SR^b of 0.95 at a residence time of 0.97 seconds.  This yields
a stack NO level of  about 400 ppm.  The average stoichiometric ratio  over
this time period is  about 0.82.  Now conventional staging at  an  SR =  0.95
yields an NO level of about 700 ppm.  However, at an  SR close to the  average
for  the total residence  time of 1.57 seconds, 400 ppm is also achieved.  Thus,
it appears that the  distributed air concept has not really improved upon the
conventional staging result unless  it is  better to  be at a very  low SR  for a
brief period followed by a higher  SR for  another  time element as opposed to
being at the average stoichiometric ratio  for the total  time  period.  A simi-
lar  conclusion  is drawn  at an SRla  of 0.6 and SRlb  of 0.80.   In  fact,  it
                                      33

-------
TABLE 2-5.  DISTRIBUTED AIR VERSUS CONVENTIONAL STAGING
Arrangement Stage
Case 1
Distr. air la
Ib
Conventional staging
Conventional staging
Case 2
Distr. air la
Ib
Conventional staging
SR

0.6
0.95
avg. 0.82
0.95
0.85

0.6
0.8
avg. 0.73
0.75
T (sec)

0.6
0.95
Total 1.57
1.57
1.57

0.5
0.97
Total 1.57
1.57
NO (0% 02) ppm


400
700
400


300
250
                         34

-------
appears the conventional staging produces even lower NO levels than the



distributed arrangement for the same time period for this particular com-



bination.



2.4    CONCLUSIONS



       In summary, a number of tests were conceived to explore a distributed



air concept to achieve low NOV emissions in a relatively short overall  resi-
                             J\


dence time.  This concept had proven successful in a premixed, small-scale



facility.  Unfortunately, the results of the current study did not achieve



any improvement in the N0₯ time to staging results achieved with conventional
                         X


staging.  This may be explained partly by the fact that a diffusion flame



was utilized in this experiment as opposed to a premixed flame in the smaller



scale experiment.  Attempts were made to achieve as premix a situation as



possible by utilizing four burners and increasing the burner exit velocity



to effect a high mixing rate near the burner.  However, the diameter of



the firebox (33 inches) results in a relatively low L/D for any given resi-



dence time, especially compared to the Pershing facility (Reference 2).



Table 2-6 compares these two facilities.





                   TABLE 2-6.  FACILITY CHARACTERISTICS
Parameters
Diameters (in.)
Length (in.) for
~1 sec residence
time
L/D
EPA
Hor. Ext.
33
12
0.364
Pershing
6
54
9
                                      35

-------
Thus, at the shorter bulk residence times in the EPA facility, it will  be
difficult to achieve a real bulk residence time.  That is, the real  resi-
dence time, for residence times under 1 second, will be much less than  the
bulk residence time due to a nonuniform velocity profile across the  diam-
eter of the firebox.
       It was found that with the distributed air approach, NO levels  in-
crease with decreasing SR below an SR of about 0.6.   In addition, NO always
decayed with increasing residence time in both the la and Ib stages.
       In conclusion, no advantage was found for the distributed air con-
cept as applied to the diffusion burner arrangement in the EPA Multiburner
Facility.
                                   36

-------
                                 SECTION 3
                          COAL/OIL MIXTURE TESTS
       As part of the Phase III alternate fuels testing,  an emissions
evaluation test program was developed to look at coal/oil  mixtures  (COMs)
fired in a simulated package boiler configuration.   Supplementing  industrial
oil supplies with coal in the form of coal/oil mixtures has been investi-
gated for nearly 100 years (Reference 4).  Over this period, the feasibility
of coal/oil technology has been demonstrated in both small-scale testing
and practical application.
       With new interest in this technology, it is necessary to determine
if technology developed to minimize the environmental effects of coal  com-
bustion is applicable to coal/oil systems or if further work is necessary
to ensure that pollution standards can be met.
OBJECTIVE
       The purpose of this study was:
       1.  To obtain emission data for coal/oil combustion  in an environment
           closely simulating an industrial  package boiler
       2.  To determine if emission  levels were affected  by the fuel com-
           position
       3.  To determine if conventional  control technology  developed for
           coal  is effective in  reducing emission  levels  produced by coal/
           oil combustion
                                       37

-------
        4.  To  investigate the effect of burner modification on emission
           levels produced by coal/oil combustion
 Facility
        This  study was  conducted  in the EPA Multifuel Furnace Facility.  The
 experimental facility, as shown  in Figure 3-1 and described in detail else-
 where  (Reference 1), was designed  to simulate the aerodynamics of either a
 front-wall fired or tangentially-fired boiler.
        In order to simulate the heat release and temperature profiles con-
 sistent with typical industrial package boilers, the additional modifica-
 tions, as shown in Figure 3-2, were made.  This configuration uses horizontal
extension sections that can be attached to the main firebox.  These 33-inch
 inside diameter by 6'  long refractory-lined sections allowed the simulation
of a tunnel-fired unit and staging of combustion air at residence times
typical of what would be available in a package boiler.  Additional  hardware
included water tubes placed in the horizontal  extension sections for addi-
tional heat absorption in the radiant section, and a heat exchanger placed
between the firebox and the horizontal  extensions to achieve a gas tempera-
ture profile consistent with the radiant and convective sections in a typi-
cal industrial  package boiler.
       The burner used for the study was an IFRF 1.5 x 106 Btu/hr wall-
mounted unit.  This burner is a versatile experimental  swirl block burner
patterned after that developed by Beer (Reference 5).  During baseline tests
on the parent coals, a Babcock and Wilcox-type coal  spreader was used to
achieve a turbulent flame condition.  Two commercial fuel oil  atomization
nozzles were tested with the coal/oil mixture.  These were the Delavan Cor-
poration swirl-air nozzle, shown in Figure 3-3, and the Sonic Development
Corporation Sonicore nozzle, shown in Figure 3-4.  The burner was modified
                                     38

-------
    Combustion chamber  (39" cube)
    Ignition and flame  safeguard
    Observation ports
    Ashpit
    1.5 x 10* Btu/hr I FRF burner
    C.E.-type corner fired burners
    3200°F refractory
    Heat exchange sections
    Drawer assemblies
0.   Staged Injection ports
                   Figure  3-1.   Acurex/EPA  multifuel  furnace.
                                            39

-------
O   OJ
                          !'/
                                  ft
                               '/
                              /   //I

                              /''  //i I/
                             /
i '/"/
                               {}  ff
                                                              Existing Heat

                                                              Exchange Sections
R 3 ft ft
"'  *f 'V /V
'  'f /I1 / <
/ll/<',J1/"
'"//I '//I1/'
'['/  '//'»' '
!«"<' '/'
A'  •"  i/'
B  tt  0  b
               I FRF BURNER
                                          V


                                   Radiant Section
                                Convective

                                  Section
                                            V

                                       Main Firebox
                                         Figure  3-2.   Facility modifications,

-------
                                                  MIXING CHAMBER
 AIR INLET
FUEL INLET
                                    AIR INLET TO
                                    MIXING CHAMBER
EROSION
OCCURRED
IN THESE
AREAS

PINTLE  PLATE
                                                             METERING NUT
             Figure 3-3.  Delavan  swirl-air nozzle, 33373-1, 60 gph,
                         70° spray  angle, mild steel  construction.
                                     41

-------
                                Fuel
IN5
                                                   Air
                                                                         Resonator chamber
                                                                                        Erosion  occurred
                                                                                        in  these areas
                                                                     Standing shock wave
                               Figure 3-4.   Sonic Corporation Sonicore nozzle, 281T-B-11  with
                                            stellite resonator chamber.

-------
for a low NOV configuration by placing an  annul us  of tertiary  air  around
            A
the diffusor.  This will  be described later.
Emissions Monitoring Equipment
       Continuous monitoring equipment was utilized to collect and record
data during this study.  Table 3-1  lists the instrumentation used  and the
principle of operation for each unit.
Fuel Preparation
       The coal/oil mixtures examined in this study were prepared from par-
ent fuels which represent a broad range of classifications and fuel compo-
sitions.  The compositional analyses of the parent fuel oils are listed in
Table 3-2 and that of  the parent coals in Table 3-3.  From these parent fuels,
four mixtures of 30 percent by weight coal to oil were prepared.  Table 3-4
lists the mixtures and their  compositional analyses.
        The  coals,  pulverized  to 70  percent through  200 mesh,  were  blended
with  the fuel  oils  in  a  high  turbulence batch mixer supplied  by Littleford
Brothers.   A suspension  additive, supplied by  Carbonoyl  Co.,  was  added to
ensure  a homogeneous mixture.  The  additive was  prepared as  a 5-percent
aqueous solution and  constituted a  3.75-percent  by weight of the  total mix-
ture.   The mixture was prepared on  a batch  basis  and 55-gallon drums were
used to store the fuel.   The mixtures were  stored at ambient temperatures
 (50 to  60°F) for up to 21 days before they  were  fired.
        Approximately  4 hours prior  to use,  each  drum was wrapped  with heating
blankets, and a mixer with a 6-inch propeller  was immersed in the mixture.
The propeller was located approximately 6 inches from the drum bottom.  The
mixture was heated and agitated utilizing a pump recirculation system until
 the mixture temperature  reached 150 to 170°F.   The mixture was then pumped
                                      43

-------
TABLE 3-1.  EMISSION MONITORING EQUIPMENT
Pollutant
NO
so2
CO
coz
°2
Parti cul ate
Loading
Principal of
Operation
Chemiluminescence
Pulsed Fluorescent
Nondispersive
Infrared (NDIR)
Nondispersive
Infrared (NDIR)
Paramagnetic
Cyclone and
Filtration
Manufacturer
Ethyl Intertech
Thermoelectron
Ethyl Intertech
Ethyl Intertech
Ethyl Intertech
Acurex Corp
Models
Air Monitor-
ing
Teco
Model 40
Uras 2T
Uras 2T
Magnos 5A
HVSS
Instrument
Range
0-5 ppm
0-10
0-100
0-250
0-1000
0-5000
0-50 ppm
0-100
0-500
0-1000
0-5000
0-500 ppm
0-2000
«—
0-5%
0-20*
0-5%
0-21%
0-3 pm
Minimum

-------
TABLE 3-2.  FUEL OIL ANALYSES
Specifications^—- — "
^ 	 ' Fuel Oil
API Gravity
Flashpoint, COC°F
Viscosity, SSU at 100°F
Heat of Combustion Btu/lb
Ultimate Analysis (% Wt)
Carbon
Hydrogen
Nitrogen
Oxygen
Sulfur
Ash
Amerada
Hess #6
15.3
204.0
2,500.0
19,867.0

84.71
10.75
0.36
1.93
2.22
0.03
Chevron
#6
12.3
182.0
4,900.0
18,292.0

85.57
10.52
0.81
2.08
0.93
0.09

-------
TABLE 3-3.  COAL ANALYSES, AS-RECEIVED BASIS
Proximate (%Wtl_- 	 "
^^_^— ^"""^ Coal
Moisture
Volatiles
Fixed Carbon
Ash
Rank
Ultimate (% Wt)
Carbon
Hydrogen
Nitrogen
Oxygen
Sulfur
Ash
Heat of Combustion, Btu/lb
Montana

21.23
35.16
34.27
9.34
Sub-bit. C.

53.26
3.35
0.87
11.16
0.78
9.34
8,972
Virginia

0.31
31.9
51.4
16.5
High-Vol. A

71.11
4.46
1.68
4.24
2.02
16.5
14,079
W. Kentucky

5.0
36.55
50.98
7.47
High-Vol. B

69.79
4.79
1.34
8.65
2.95
7.47
12,349

-------
TABLE 3-4.  COAL/OIL MIXTURE ANALYSES, AS-RECEIVED BASIS, 30% COAL BY WEIGHT
^"~"-->^Jixture 30% (Wt;
^~~~"--^^^^ Coal
Analysis ^-^^^
Carbon
Hydrogen
Nitrogen
Oxygen
Sulfur
Ash
Moisture
Heat of Combustion,
Btu/lb
W. Kentucky/
Amerada
80.23
9.00
0.63
3.92
2.44
2.26
1.52
17,600
Montana/
Amerada
75.27
8.57
0.49
4.67
1.79
2.82
6.39
16,600
Montana/
Chevron
75.88
8.37
0.83
4.77
0.89
2.87
6.37
15,500
Virginia/
Chevron
81.23
8.70
1.07
2.73
1.26
4.95
0.09
17,000

-------
 into  the coal/oil mixture delivery system, shown schematically in Figure 3-5.
 After each drum was emptied, the drums were inspected for settling of solids.
 In all cases, little or no deposits were noted.  The delivery system consists
 of a 120-gallon capacity heated storage tank with an agitation system simi-
 lar to the one described above.  The fuel was kept well agitated and at 180
 to 200°F.  The mixture was delivered to the burner, through heat traced
 lines and an electrical heater, via a Viking C-32 rotary pump with a vari-
speed control.  A recirculation loop ensured that a homogeneous mixture was
maintained between periods of mixture firing.
       A summary of the problem areas associated with this flow system
include the following:
       a   Coal  settled out over a period of time in the bottom outlet of
           the fuel  holding tank causing complete plugging.
       •   Plugging of lines in any low section of piping.  Adequate
           velocities must be maintained to keep the material entrained.
       •   Shorting of electrical  heat tape elements on piping and drums.
           It is recommended that steam tracing be used if at all  possible
           in future tests.
       •   Deterioration of pump performance due to wear at  the seals  and
           increased clearances.  Special  seals for handling this  very
           abrasive  mixture  should  be considered when  pumping COM
       •   An immersion heating element was used in the drums for  initial
           heating of the mixture  before transfer to the holding  tank.   If
           the mixture  had settled  these elements would not  heat  the mixture
           uniformly and fires  could  easily develop.   Also the tanks could
                                      48

-------
                                         HEAVY OIL
                                         SOLENOID
                                         VALVE
                                             BURNER
                                             NOZZLE
                                            SAMPLE
                                             LINE
  VIKING
  ROTARY
   PUMP
With Varispeed
  Control
ELECTRIC
 HEATER
   Fuel 3-5.  Coal/oil  delivery system.

-------
            not be  mixed  successfully  until  they were  thoroughly  heated.



            Thus it was found  to  be  quite  difficult  to get  these  drums



            reheated and  well  mixed  after  they  had been setting for several



            days.



        •    The most difficult task  was  pumping the  mixture from  the drums



            to  the  holding  tank.   The  pump performance frequently deteriorated



            to  the  point  where it  would  not  draw from  the drum.  This was



            of  course hampered by  reheating  the drum and getting the mate-



            rial well entrained.



 3.1     TEST PLAN



        The  tests were planned around  a  range of fuel  types, three coals and



 two  oils, to determine if  emission  levels both under  baseline and incor-



 porating NO control technologies are dependent on  fuel types.  Table 3-5
            A                                                         (


 lists the baseline  test  matrix for  the  fuel  combinations of interest.



 Initially a 50  percent mixture of the various  fuels was to be tested but



 because of  budget constraints and the mixing/handling  problems encountered,



 it was decided, with the project officer's concurrence, to limit the test-



 ing to 30 percent concentration of coal.  The  baseline tests were run at 20,



 30, and 40  percent excess air levels at a firing rate  of 1.8 x 106 Btu/hr.



This firing rate gave a  heat  release per unit  volume of about 50,000 Btu/hr-



ft3 which is typical of  package boilers.  The  coal  tests were run with a



B&W-type coal  spreader with 15 percent primary air, and the coal  and oil



tests were run with the Sonicore nozzle.



       Table 3-6 shows the matrix for the effect of load and the effect of



residence time with staging as the NO  control  technique.   All of these tests
                                     A


were run at 20 percent excess air and 30 percent COM.   Table 3-7 lists the

-------
TABLE 3-5.  BASELINE MATRIX

fD
o
O
£
3
Q.
Coal /Oil
Mixture
•r-
O
0)
i.
3
Q.
i
S
s
1.
5
1/1
CO
01
u
X
LLJ


O
^J-
s
o

o
CM
0
1-
O
CO
0
CM
Pure Fuels
Western
Kentucky Coal
22 5d
225c
225b








226g
226f
226e






6
§
>
01
£
O






217c
217a
217b
<5
Pennsylvania






222g
222f
222e
Coal /Oil Mixtures
Coals
Western
Kentucky
Montana
Virginia
Fuel Oils
Chev









PA



21 7d
217e
217f



Chev



2215
221 a
221 c



PA



221 r
221q
221n



Chev



218b
21 8a
21 8c



PA









               51

-------
TABLE 3-6.  EFFECT OF LOAD AND RESIDENCE TIME
O)
01

Residence Time - Short
0)
c
o
_1
1
QJ
*! —
1—
(U
U
c:
O)
-a
•r—
CO
OJ
OL
Tertiary
Distribution
O
CVJ
1
1 —
o:
GO
in
o>
o
1
oT~
oo
m
o>
•
o
1
oc.
oo

rO
oF~
CO
£
fO
E
£
Q_
in
00
0
in
r^«
o
in
UD
•
o
LT>
in
•
o
in
oo
o
in
i-^
o
in
10
o
in
in
o
in
00
o
m
r>-
•
o
in
ID
o
m
*d-
•
o
Coal /Oil Mixture 30%
Mont/Chev WKC/PA VA/Chev
Load x 106 Btu/hr
1.2


224£
224J








1.8
223f
223g
223h
223i
223o
223m
223k
223j




1.2












1.8
217k
1275,
217n
219o
219a
219b
21 9c
21 9d




1.2












1.8
220g
220h
220 i
220j
221 f
220j
220k
220£




                      52

-------
TABLE 3-7.  DISTRIBUTED AIR BURNER TESTS
Ol
CTI
(O
-|J
l/l
Residence Time -Long
Residence Time -Short
oF~
CO
O)
O>
•M
(/)
CO
r—
in

o
in
CO
o
in
r-
o
if)
VD
O
in

-------
 tests  for the  low  NOV  burner  and  combined  low  NO   burner and  staging con-
                     X                           X
 figurations.   The  purpose  of  this matrix was to look at the effect of these
 control  technilogies on  three fuel  combinations.   A range of  the control
 core stoichiometric  ratios were tried with and without staging.  Finally a
 few tests  were run by  doping  the  fuels with pyridine and/or thiophene to
 increase  the nitrogen  and  sulfur  levels appropriately.  These tests are
 shown  in  Table 3-8.  A comparison of stack emissions with these dopants or
 with fuels that naturally had  these levels would then be possible.  On all
 of these  matrices  the  test number has been given so that reference to the
 emission  levels may  be determined from the listing in the appendix.
 3.2    TEST DATA
       The testing was divided into three phases.  In Phase I the combustion
 stability  of each fuel was evaluated and delivery conditions were adjusted
 for optimization of  flame stability and combustion.
       In  Phase II of  the study,  some of the established combustion control
 technologies for pulverized fuel were applied to the coal/oil mixtures.  These
 included staging of combustion air (Reference 6) and burner air distribution
 (Reference 7).   Flue gas recirculation, which has been found to be effective
 in reducing thermal NO (Reference 6), was not applied due to equipment prob-
lems.
       The last phase of the program was to evaluate fuel  nitrogen conver-
sion by addition of dopant to the feed system.   Results from each of these
phases  will be  discussed in the following sections.  In addition, compari-
son will  be made with data taken during a previous  DOE-supported test in
this same facility.  Details of the  DOE tests may be found in the appendix.
                                      54

-------
                        TABLE 3-8.  EFFECT OF FUEL NITROGEN

u.
o
c
(U
O1
O
+J
s-s



\ •
CT» \
i— \
X
\"
0 \
\°
o \
\ e=>
u\
o \
\5
0 \
% Sulfur (DMMF)*
0.9 ^^^
^^^ 1.2


Montana Coal
226a
Virginia/Chev
21 8c
Chevron Oil
217b
Mont/Chevron
221c





1.96


Mont/Amer + N
221o

Mont/Amer + N
221p

Mont/Amerada
221n




2.2 ^^^
^^^ 2.6
Virg. Coal
226e

Va/Chev + S
221k
Va/Chev + S
221m


W. Kent/
Amerada
217f

Amerada Oil
222e

3.4

W. Kent Coal
225b









DMMF:  Dry, mineral matter free
                                      55

-------
 3.3     BASELINE TESTS
        The  baseline tests were designed to determine the optimum burner con-
 ditions  for each fuel and then the NO emissions as a function of excess air
 for each parent fuel and fuel combinations.  Initially the burner was
 adjusted (swirl, axial fuel tube position and nozzle atomization rate) for
 maximum  flame stability for each fuel.  In all cases, the nozzle position
 was virtually the same relative to the IFRF burner, i.e., 2 inches forward
 of the burner throat.  However, atomizing air pressure had to be optimized
 for each of the fuel mixtures.  This was due to carbonaecous deposition or
 "clinkering" on the water-cooled quarl in cases of poor atomization.  In
 general, the atomization pressure ranged from 12.0 to 22.0 psig depending on
 the fuel.   In most cases, this was 2.0 to 4.0 psig greater than the fuel
 delivery pressure.  Throughout the tests, secondary air was preheated to
 300°F to enhance solids ignition in the relatively cold environment of the
 radiant  section.  Burner swirl was optimized on the parent fuels and on
 the coal/oil mixtures.  A swirl of 5 on a scale of 8 was used for the base-
 line coal tests, and a swirl of 0 (of 8) was used for both the oil and the
 coal/oil mixtures throughout the testing.  (A zero setting implies no swirl.)
 Tests conducted earlier at the EPA/Acurex facility for design support of the
 full scale  DOE/Lorillard, Danville, Virginia, COM facility yielded a compari-
 son of the  two commercial nozzles described above.  After approximately
 3 hours of  testing on 30-percent coal/oil mixture, utilizing the 440 hardened
 stainless steel  Delavan nozzle, significant erosion was observed.  Areas of
erosion are shown in Figure 3-3.   Equivalent testing with the stellite
 Sonicore nozzle revealed similar erosion rates in the areas shown on Figure
 3-4.  However, the erosion had less impact on the atomization characteristics,
                                        56

-------
flame stability and emission levels for the Sonicore nozzle.   On  this  basis,
all subsequent coal/oil  mixture tests were conducted with the  Sonicore nozzle.
       Figure 3-6 illustrates the results of the baseline emissions  tests.
While CO, C02, S02, and  NO data were taken during testing, NO  data were
considered primarily and will be discussed here.  Problems with the  S02
analysis unit during testing rendered the data useful only on  a relative
basis.  Generally,  CO  and  C02 levels  which are quantitatively  valid  reflected
good combustion burnout in all cases except during staged combustion tests
at long first-stage residence times (1.5 to 2.0 seconds).  During these
tests, CO levels rose to as high as 800 ppm (0-percent 02).  Detailed emis-
sion levels for each test condition may be found in the appendix.
       Figure 3-6(a) illustrates NO emission levels from the Chevron No. 6
oil base mixtures along with levels from  the parent fuels.  As is expected,
the NO levels for the mixtures fall in an  intermediate range between the
parent fuels.  The same data for the Amerada-based mixtures are  illustrated
in Figure 3-6(b).  In this case, though,  emission levels  for the mixtures
are very near those of the parent  oil.  Little  contribution from  the coal
is evidenced.  Certainly, there are  several possible mechanisms  to which
this may be attributed.   These mechanisms include effects  due  to  sulfur in
the fuel, atomization characteristics  of  the  oil, and  the manner  in which
the particular oil and coal  volatilize.   The  volatilization rate  of the par-
ticular oil surrounding the  coal particles may also  effect the fate of the
fuel  nitrogen coming  out  of  the  coal.   For example,  Figure 3-7 shows  the
boiling  point curve  for the  Chevron  and  Amerada oils.   This  shows  that the
Amerada  has  a much higher boiling  point curve than  the Chevron oil.   The
rate  at which the  fuel  N  comes  off the oil can also influence these results.
                                      57

-------
     14OO  -,
     1200  -
    10OO -
i    800 -
     600 -
     400
    1200  -
    1000  -
 (M
o
I
o"
    800  ~
    600  -
    400 -
                                                Load - 1.8 x 106 Btu/hr
                                                Air preheat - 300° F
                                               Q  Chevron No. 6

                                               Q  Virginia/Chevron 30%

                                               /\  Montana/Chevron 30%

                                                   Montana Coal

                                                   Virginia Coal
                            Chevron  based mixtures
10        20       30

     % EXCESS AIR
                                                 40
                                              Load- 1.8x 106 Btu/hr
                                              Air preheat » 300'F
                                              Q  W. Kentucky Coal

                                              Q  Montana Coal

                                              /\  Montana/Amerada

                                              O  W. Kty/Amerada

                                                  Amerada No. 6
      - Amerada  based mixtures
    200
                    I
                   10
          I
         20

     % EXCESS AIR
I
30
                                               -  40
                            Figure  3-6.   Baseline emissions.
                                              58

-------
    90
OJ
o
s_
O)
Q.
OJ
>
'o
>
OJ
LO
l/l
    80 -
    70
    60
    50
    40
    30
    20
     10
   2
   3
   4
   5
   6
   7
   8
   9
  10
East Coast
Middle East
Indonesian
Venezuelan Desulfurized
Gulf Coast
Venezuelan
Alaskan
Wilmington
Cali fornia
Amerada
Chevron
                          I
                     I
                                          I
      400
          500
500
   600|
     600
700
       8()0
       700
Temperature
900
  | 1000
800°K
,1100°F
             Figure  3-7.   Boiling point curves (Reference 8)
                                      59

-------
 Figure 3-8 shows the percent nitrogen evolved as a function of temperature.
 These curves are shown compared to a variety of other oils  (Reference  8).
 As can be seen from the curve the Amerada oil is one  of the more  "refractory"
 as far as N evolution is concerned.   These mass and nitrogen evolution rates
 could indeed play a role in  performance of the various fuel  combinations  to
 the NO  control  strategies.
        At this point there is insufficient data to ascertain which  of  the
 various  mechanisms  is causing this  effect.   But it is important to  note that
 all  fuel  combinations do not necessarily behave in the same manner.  With
 regard to the  coal  baseline  tests,  the  differences between  the Montana and
 Western  Kentucky coals are quite  minimal  and  are probably within  the error
 band  of  the data.   The Virginia coal, which consistently showed a higher NO
 level  than  the Montana coal,  is probably a  real  effect due  to the higher
 fuel  nitrogen  in the  coal.   The comparison  of the  Montana and Western  Ken-
 tucky  coals  differs from previous data  (Reference  9).   The  previous data
 showed the  Montana  coal  to be  consistently  above the  Western Kentucky  coal
 by about  5  to  10 percent.  The main  difference  between  the  tests  is that in
 this test the  temperatures are 300 to 400°F cooler.   It is  possible that
 the volatilization  rate  of the Montana  coal changes with the temperature,
 and the Montana  coal  may have  had a  higher  fuel  nitrogen evolution rate in
 the previous tests.
 3.4    CONTROL TECHNOLOGY TESTS
       The results of  staged combustion  of  the  coal-oil mixtures are illus-
 trated in Figure  3-9.  As is  illustrated, the  two  Chevron no.  6 oil-based
mixtures responded almost identically to  the  removal  of combustion air in
 the first stage.  However, the Amerada-based mixture  showed  little response
                                       60

-------
    30
           o
           A
           O
           O
           D
           O
           0
 2
 3
 4
 5
 6
 7
 8
 9
10
O)
o
QJ
CL
TD
OJ
OJ
QJ
cn
O
20
    10
                East Coast
                Middle East
                Indonesian
                Venezuelan Desulfurized
                Gulf Coast
                Venezuelan
                Alaskan
                WiImington
                California
                Amerada
                Chevron
                500
                      600      700      800       900
                                 Temperature (°F)
1000
                                                         1100
          Figure 3-8.  Nitrogen evolution curves  (Reference  8).

-------
                                                                                  Stoichiometry = 1.20
                                                                                  Load = 1.8 x 106 Btu/hr
                                                                                  1st Stage Res. time  = 0.75 sec
           800
                                                                                                              30% Montana/Chev.

                                                                                                              30% Virginia/Chev.

                                                                                                              30% W. Kty/Amerada
           600
rv>
        CM
       o
o.
o
           400
           200
                                                                       I
                          0.65
                                  0.75          0.85          0.95          1.05
                                      First  stage  stoichiometric ratio
1.15   1.20
                                                Figure 3-9.   Staging emissions.

-------
(10 percent) in emissions reduction.   It should be noted from Table 2,  that
the Amerada no. 6 oil has a very high sulfur content relative to the Chevron
parent oil.  This compositional  difference could possibly have contributed
to the ineffectiveness of staging on the Amerada-based mixture NO emissions
(References 10 and 11) or it could again be due to the volatilization char-
acter of the fuels.
       Following the staging tests, distribution of burner air was applied
to the mixtures.  Figure 3-10 shows the distribution scheme schematically.
In these tests air which normally makes up a percentage of the secondary air
was injected through an annul us  7 inches radially from the burner throat cen-
terline.  This was done in order to enrich the flame core where fuel-bound
nitrogen is evolved.  The results of this testing are illustrated in Figure
3-11.  Each mixture responded favorably to the air distribution, but it is
evident that the composition of each mixture leads to a unique emissions
curve under these conditions.  In this case the Amerada oil-based mixture
performed about the same as conventional staging, that is showing little
effect to the control technique except when the core was made very fuel-rich.
A moderate effect is seen with the Virginia coal/Chevron oil and is again
very similar to the conventional staging result.  However the Montana/Chevron
mixture shows a much more dramatic effect compared to conventional staging,
reducing the NO levels by a factor of three.  The Montana coal also yielded
the lowest NO levels during staging tests in the main firebox during earlier
testing (Reference 1).
       The curves in Figure 3-12 illustrate the results of applying burner
air distribution plus staging vs. straight burner air distribution.  The
purpose of this comparison was to evaluate the mixture responses to further
                                        63

-------
                                              Stage
                                             I air

                                             if
   Fuel tube
SR,
                                                  EA
                         ~f / > i  i  1  I  T
                                             \
                          Present IFRF burner with staging

Tertiary
cinnu 1 us te "
Fuel tube 	 ^ SRla

///////f
-f
///////


SR,


Stage
r


EA


SR - Stoichiometric ratio
EA - Excess air
                             Modified IFRF burner for low NO
             Figure 3-10.  Burner air distribution.
                                  64

-------
                                                                                                    Stoichtometry = 1.20


                                                                                                           Load = 1.8 x 106Btu/hr
                   800
                   600
               M
               o


               I


               I
               o
               z
400
cn
                  200
                                                                                                                          V 30% Montana/Chev




                                                                                                                          0 30% Virginia/Chev




                                                                                                                          Q 30% W. Kty/Aimrada No. 6
                                   0.55
                                                  0.65
                                                                 0.75
                                                                                0.85
                                                  SR
                                                     1A
                                                Figure 3-11.   Burner  air  distribution.

-------
     800 -
     600  -
I    „-
o
2
     200  -
              Q SRj = 0.95
              O SRi - 1.20
                  i        i        t       \
                0.55     0.65    0.75    0.85

          FLAME CORE STOICHIOMETRIC RATIO 
                                       30% Virginia/Chev.
                                       1st Stige Res. time » 0.75 tec.
                                       15 x 106Btu/hr
                                       SRi ' In Stage Stoichiometry
  CN
 O
 Q
 a
 O
 Z
     800  -i
     600
     400  -
     200 -
SR,
                      0.95
                      1.20
                0.55    0.65     0.75     035
           FLAME CORE STOICHIOMETRY 
-------
enrich the flame zone.  In Figure 12(b), for the Western Kentucky/Amerada
mixture, the further enriching of the flame zone resulted in higher NO
levels.  This result, possibly, further illustrates the response of the
Amerada high-sulfur parent oil for fuel-rich conditions.  The Montana/Chevron
mixture results are illustrated in Figure 12(c).  This comparison validates
the interesting way in which each mixture reacted to the same combustion
conditions.
     Again, we see little effect of the Amerada-based fuels, a moderate ef-
fect on the Virginia/Chevron mixture and a strong effect on the Montana/
Chevron mixture.  It  is interesting to note, however, that there was very
little difference between the air distribution or low NO  burner tests and
                                                        X
the stated low NO  burner tests on the Chevron-based mixtures.  This may
indicate  that with the short  residence time staging there is considerable
backmixing of the stage air into the first stage.  This backmixing would
result  in a higher overall stoichiometric ratio  in the  first stage.  How-
ever,  the fact that  the Amerada-based fuel reacted differently may indicate
there  is  more a dependence on the N and mass evolution  rate and that the
timing  of the distributed air may be unique for  each  fuel in order to  achieve
an optimum low NO  condition.
                 A
     The  effect of firing rate  and residence time on  NO emissions during
staged  combustion is  illustrated  in Figure 3-13.  Figure  3-13(a) represents
staging combustion air with a first-stage residence time  of approximately
0.75  seconds.  A first-stage  residence  time of  approximately 1.50  seconds
is applicable  to Figure 3-13(b).  As expected in both cases, NO emission
levels decreased at  the lower firing rate of 1.20 x  106 Btu/hr.  NO  levels
dropped to  about one-half the baseline  levels by increasing  the residence
                                        67

-------
       800  -
       600  -
    CM
    O
    -   400  -
    I
    a
    O
        200  -
Q ljBx106Btufhr

    1.2 x 106 Btu/hr
                                         30% Monuni/Chev.
                                         1ft Stage He*, time * 1.50 sec
                                         1st Stage stoicriiorrwtry "1.20
                   0.65
                           0.75
                    I
                  0.85

                   SR,
                    B
                                           0.95
1.05
                                                                 I
                                                               1.20
                                              30% Montana/Chevron
                                              Stoichlometry 1.20
                                              1st Stage Res. time 0.75 sec
    is
   O
   O
        •00 -,
        •00 -J
        400 -
        too -
                   O
   1* xlO6 Btu/hr
   1.2x106 Btu/hr
                   0.86
                           0.75
                    I
                  0.85
                    SR,
                                                    1.05
                                                               1.20
Figure 3-13.   Effect  of  firing  rate and  residence  time.
                                        68

-------
time.  This result of residence time is coincident with earlier tests  on
coal only that showed marked decreases in NO with residence time during
staged combustion.  The levels at the longer residence time are similar  to
the low NOX burner tests on the COM.  However, CO levels during the long
residence time tests rose to 800 to 1000 ppm (0-percent 02) at times.
       It should also be noted that the differences in NO at the two loads
decrease with decreasing first-stage stoichiometric ratio.  This is again
consistent with previous coal data.
3.5    FUEL NITROGEN STUDIES
       The last phase of the study was a limited evaluation of fuel nitro-
gen converion in COMs.  This involved addition of a dopant to the feed
system to evaluate the conversion of fuel bound nitrogen.  Pyridine, CgHgN,
was added to the Montana/Amerada mixture upstream of the fuel tube during
two tests, and the percent conversion of nitrogen to NO was calculated based
on the emissions levels obtained.  These data, along with  the emissions
levels of all the mixtures and parent fuels under baseline conditions, are
plotted against fuel nitrogen content in Figure 3-14.  This plot exhibits
a definite trend.  In both cases,  the fuel nitrogen conversion  to NO was
30 to 35 percent which  is typical  of bound-nitrogen conversion  during coal
combustion.  The pyridine dopant points  fall  slightly  below the line, and
we  have no definitive explanation  for  this.   It  is possible that the pyri-
dine volatilizes earlier than  the  oil  or coal/oil mixtures, but this pre-
liminary screening test did  not develop  sufficient data  to draw any defini-
tive conclusions.  No correlation  with fuel  nitrogen  was  seen  with  the coal
data.  This  result  is consistent with  published  data  (Reference 8).   It
appears  that the  COMs behave more  like the  oil with regard to  fuel  nitrogen
than to  coal.
                                       69

-------
   1500
   1300
   1100
CL
CL
 C\J
0   900
o

o
    700
    500
    300
      0.
                                                    O
                                                      1
                                                             1 .8 x 10  Btu/hr
                                                             20; EA
                                                             Sonicore nozzle
                                                                       A
                                                 D

                                           ^  Montana/Amerada
                                           ^  Virginia/Chevron
                                           0  Montana/Chevron
                                           (2i  W. Kentucky/Amerada
                                           O  W. Kentucky coal
                                           A  Virginia coal
                                           O  Montana coal
                                           O  Chevron oil
                                           O  Amerada oil
                                           C^  Montana/Amerac- + dopant
                                           O  Montana/Amerada + dopant
0.50
0.70
0.90      1.10     1.30     1.50
     Fuel nitrogen DMMF  (^, wt)
1.70
1.90
2.10
                          Figure 3-14.  Effect of fuel nitrogen.

-------
3.6    BURNER NOZZLE COMPARISON
       NO emissions data were obtained with both commercial  nozzles  de-
scribed earlier on the parent oils.   Figure 3-15 illustrates the  results
of this comparison.  Emissions data  obtained earlier during  the DOE  tests
(see appendix for report on these tests)  using the Delavan nozzle are
also included.
       In all cases the Sonicore nozzle gave higher NO levels than the
Delavan nozzle.  This result is attributable to the way the  fuel  is  ato-
mized and mixed with the region.  How a nozzle atomizes any  given fuel can
also affect the NO levels.  For example, the Chevron oil showed a more
marked difference between the two nozzles than did the Amerada oil.   This
again shows the difficulty in trying to predict the NO levels for a given
oil and/or nozzle.
3.7    PREVIOUS TESTING DATA
       During previous testing NO emission  levels were obtained as  a  func-
tion of the percentage of coal in the mixture.  These data  are illustrated
in Figure 3-16.  These data were obtained under baseline  conditions with
the Delavan nozzle  (hardened stainless steel) at a firing rate of 1.8 x 106
Btu/hr using  the Virginia coal and Amerada  parent oil.  The  results are not
too surprising for  a given fuel and nozzle.  They indicate  that as  the per-
centage of coal increases, the NO levels increase, although  not quite
linearly.
3.8    SUMMARY AND  CONCLUSIONS
       The following conclusions can  be drawn based  on  this study:
       •   Baseline NO  emissions from COM were,  in general, linearly  re-
           lated  to the  fuel-N  for  a  given  burner and nozzle type
                                        71

-------
                            800
                            600
LOAD: 1.8x 106 Btu/hr
SWIRL:  Sonicore Oof 8
        Delavan 3 of 8
         X Amerada - Delavan DOE
         V Amerada — Sonicore
        O Amerada - Delavan US
        O Chevron — Delavan
        A  Chevron — Sonicore
ro
                       CM
                      o
                      g
                       o
                            400
                            200
                                                     20
                                              30
                                                                                                             40
                                                                % EXCESS AIR
                                                Figure 3-15.   Nozzle comparison.

-------
               Q.
               0.
CO
                  1400 r
                 1200
                 1000
                  800
                  600
                  400
                  200
                                         Virginia/Amerada
                                         Delavan Nozzle
                                         1.8  x  106 Btu/hr
                                                  O  20% EXCESS AIR

                                                  A  30% EXCESS AIR

                                                  D  40% EXCESS AIR
                                                                                    40% EXCESS AIR
                                                                          20% EXCESS AIR
                                  10
                        OIL
                        ONLY
20       30      40       50      60

             PERCENT COAL IN FUEL
                                                                                    70
                                                                                            80
                                                                                                     90
    100

PULVERIZED
COAL ONLY
                                             Figure 3-16.   Data  from earlier work.

-------
 The  burner  settings and  fuel nozzle  type have a strong effect
 on NO  emissions.
 Conventional control technology utilized presently for pulverized
 coal combustion is effective in reducing NO emissions but to
 different degrees dependent on fuel  composition.
 -    It was  difficult to  achieve low  NO  emissions with the Amerada
                                      /\
     based COMs.
 -    Moderate effects were obtained by NO  control technologies
                                        X
     with the Virginia/Chevron mix.
 —    Strong  effects were  obtained by  NO  control technologies with
                                      J\
     the Montana/Chevron.  These were obtained with the low NO
                                                             /^
     burner, low NO  burner plus staging, and staging at long
                  A
     residence times.  However, CO levels are excessively high at
     long residence times.
 NO emissions increase in proportion  to the amount of coal in the
 coal/oil mixture but fall between the parent fuel oil and coal
 baseline emissions.
 Care must be exercised in designing  the pumping systems for coal/
 oil mixtures to avoid regions where  coal may settle out and even-
 tually plug the lines.
 Flow control in small  scale combustion tests of COM are quite
difficult due to the necessity of small orifices which can either
 erode or become plugged.
 Fuel  nozzles which rely on impingement of solid surfaces by high
 velocity fluid jets will be subject  to high erosion rates.  Judi-
 cious selection of materials may help to overcome the problem.
                             74

-------
       t   Pumps must be selected which can handle this highly abrasive
           mixture.
3.9    RECOMMENDATION
       In order to analyze and understand the complex process of coal-oil
mixture combustion, we must obtain a better understanding of the combustion
processes of each parent fuel.  Also, the combustion of coal/oil mixture
needs detailed examination to determine if it is merely a combination of
the two individual processes or if the interaction of these processes re-
sults in a completely different complexity of physical and chemical phenomena.
       Future work should examine the role of fuel-bound nitrogen utilizing
flue gas recirculation and nitrogen evolution studies of the  parent oils.
Also the effect of the physical presence that each fuel exerts  on the other,
such as shielding or physical separation in the droplets, should be examined.
                                         75

-------
                                 SECTION 4
                          REFUSE-DERIVED FUEL TESTS

       During the last decade, it has become apparent that the growing
demand for energy and the resulting scarcity of clean, readily available
sources to meet that demand have increased the need to use our energy re-
sources wisely and efficiently.  The search for new fuels to supplement
present energy sources is taking on a new importance.
       Commercial and municipal refuse has long been recognized as a vast,
untapped  source  of energy.   However, the logistics of efficiently extract-
ing this energy has prevented its serious consideration as a viable source.
Dwindling supplies of clean fossil  fuels have resulted in higher costs
along with restrictions in their use.  While dirty fossil fuels are becoming
attractive alternatives, environmental considerations dictate that a balance
must exist in their use while new technology is developed to reduce their
harmful  environmental effects.   Alternative energy sources such as atomic
and solar are considered valuable but distant energy sources due to tech-
nological  and environmental  considerations.
       In light of these considerations, fuel derived from refuse is a
practical  energy source that is becoming increasingly attractive as tech-
nological  advances overcome the problems inherent with its use.  Fuel  which
is derived from refuse by removal  of noncombustible material has a nominal
                                        76

-------
heating value of 4000 to 7000 Btu/lb.   A significant number of investiga-
tions in which refuse-derived fuel  (RDF) has been used to generate  steam in
full-scale facilities have answered many questions regarding technological
problems associated with the use of this fuel.  However, gaseous, trace
metal, and organic emissions data which can provide answers to environ-
mentally related questions are currently sparse.
       It is necessary that investigations regarding the environmental
compatibility of refuse derived fuel be conducted before this vast, un-
tapped energy source can be considered  a viable supplement to present en-
ergy resources.  These investigations can be carried out most cost effec-
tively in a  laboratory-scale facility.
4.1    OBJECTIVES
       The objectives in this investigation were:
       1.  To design, fabricate and operate a system for combustion testing
           of RDF  in a laboratory-scale facility
       2.  To characterize RDF emissions of several types of material pres-
           ently available for use as fuel
       3.  To evaluate the combustion efficiency  of fuels consisting of
           conventional clean and dirty fossil  fuels  (natural gas and coal)
           mixed with varying percentages of  RDF
       4.  To evaluate the effects of combustion  parameters on emissions
           from RDF/conventional fuel mixtures
       5.  To provide direction for future investigations on RDF to insure
           solutions to problems associated with  its use.
                                      77

-------
 4.2     RDF  EXPERIMENTAL  HARDWARE
        In view  of  the  fact  that the majority of  the U.S. steam-electric
 capacity is produced by  tangentially-fired boilers, all tests were conducted
 in the  tangential  configuration.   The EPA/Acurex multifuel facility, shown
 in Figure 4-1,  is  capable of simulating several  types of industrial and
 utility boilers.   For  this  investigation, the C.E.-type corner mounted burn-
 ers,  shown  in Figure 4-2, were utilized.  The aerodynamic pattern developed
 in this configuration  is shown schematically in  Figure 4-3.  The corner-type
 axial diffusion burner shown in Figure 4-2 allows for gradual mixing of the
 fuel  and oxidant.   In  standard operation, fuel and primary transport air are
 injected through the fuel tube of  the center gun.  A portion of the secondary
 combustion  air  is  injected  through an annulus surrounding the fuel tube.  The
 balance of  the  secondary combustion air is injected through annular exits in
 the upper and lower guns.   Secondary fuel tubes, located in the upper and
 lower guns,  are used for system preheating with natural gas.
 4.2.1   Burner Modifications
        For  this investigation, a fuel gun was designed to inject refuse
 into the combustor and be aerodynamically consistent with the corner-type
 burners.  A modified gun, shown in Figure 4-4, was used for the injections
 of the  refuse material.  The forward end of the nozzles includes an actively
 cooled  section to protect against preignition of the refuse.  A thermocouple
was also installed on  the nozzle outlet surface and connected to an over-
 temperature alarm system.
       The end of the  nozzle was designed for a venturi effect to ensure
material transport velocities.   An air injection port was installed at the
                                    78

-------
1.   Combustion  chamber (39" cube)
2.   Ignition  and flame safeguard
3.   Observation ports
4.   Ashpit
5.   1.5 x  10s Btu/hr IFRF burner
6.   C.E.-type corner fired burners
7.   3200°F refractory
3.   Heat exchange  sections
?.   Drawer assembl ies
0.   Staged injection ports
                                                                                  G
                        Figure 4-1.  Furnace cross  section.
                                             79

-------
       Yaw angle
                              Tangent
                              firing
                              circle
                                                                           Heat  exchange
                                                                           section
                                                                                      RDF + air

                                                                                      Pri. fuel & air

                                                                                      Sec. air
Overhead view
Side view
            Figure  4-2.  Tangential  configuration,  aerodynamic patterns.

-------
           Gas.
     Secondary
          air
     Coal and
   primary a1

   Secondary
        air
 Secondary
      ai

CO
                                                     Figure  4-3.   Corner-fired burner.

-------
             Intct
00
ro
\
 \\  \\  v\  \\  \\
                                  Water Cooling
                                  With Active
                                  Flow Path
                                                           Refuse
                                                            and
                                                          Primary
                                                            Air
                                                             -RJ •*•
                                                                                             Secondary Air
                                                Figure 4-4.  RDF nozzle.

-------
end to supply a portion of the combustion air.   An access port was also in-
stalled to permit manual sweeping of material  clogs with a rod if needed.
       The modified gun assembly is shown in Figure 4-5.  In the tangential
configuration, two modified burner assemblies were installed in opposite
corners, while two of the standard burner assemblies were installed in the
remaining opposite corners.  This configuration is illustrated schemati-
cally in Figure 4-6.
4.2.2  RDF Feed System
       A feed system specifically designed to eliminate problems associated
with the pneumatic transport of refuse material was fabricated for this
investigation.  Handling problems, due to the nominal size of municipal
refuse (1/8" x 0 to 5.0" x 0), were significant at this scale of operation
(1.5 x 10  Btu/hr).  A separate feed system was fabricated for each RDF
burner.  Each unit was specifically designed to meter from 10 to 50 Ibs/hr
of RDF into the refuse nozzles described in the previous  section.  Signifi-
cant design effort was required to overcome many of the  inherent problems
with feed refuse at these  low flowrates.   The system is  shown schematically
in Figure 4-7.  A review of the system follows.
       The refuse feed system is made up of three  main  components which
were each designed to overcome specific  handling problems involved with
refuse transport.  These are:
       •   Fluidized system
       •   Belt transport  system
       •   Pneumatic transport system
Collectively, they result  in a consistent feedrate of refuse  into  the  com-
bustor.
                                     83

-------
       Refuse and

    combustion air
      Annular air


Coal & primary air
             Gas


    Secondary air
  -
14 n LJ u u. a.
V \\ ", ^ v»


AiVWL
                                                                      Cool ing

                                                                      H20 lines
                                                                       I  t
                                                      Side  view
                                 Figure 4-5.  Modified corner-burner assembly.

-------
                             Primary fuel
                                 plus  ai
00
on
                                                                                           RDF plus primary
                                                                                           fuel plus air
                   RDF plus primary
                      fuel plus air
Overhead view
Primary fuel
plus air
                                            Figure 4-6.  Fuel delivery schematic.

-------
 Fluidization  System.  This system  is made up of the refuse
 supply  drum,  support  frame, and a  chain-driven motor system.
 Refuse  is placed  into the chamber  through a scalable door.  An
 electric motor  is used to rotate the drum by way of a chain
 drive/clutch  system.  The drum is  rotated at a rate which ef-
 fectively fills the drum with refuse in suspension, at steady
 state conditions.  This  system results in an even distribution
 of fluffy material which can easily be transported to the
 burners.
 Belt Transport  System.  An electric-motor-driven conveyor belt
 system  is located axially along the centerline of the drum.   As
 fluffed material falls along the moving belt, it is collected
 by 1.0  inch metal stand-offs which are evenly distributed along
 the belt surface area.  The refuse layer thickness is controlled
 by a pointed  knife edge at the drum exit.   The speed of the belt
 is controlled at 0 to 1750 rpm by a varispeed motor.
 Pneumatic Transport System.   As the belt revolves around the
 forward roller, the material  is stripped off the belt using air
 jets.   The angle of these jets is such that they run tangent
 to the roller, thereby effectively sweeping the belt clean of
material.   This system also gives the material momentum into
 the burner tube.  The air used is part of the total combustion
 air required  for the refuse.   The material is conveyed into the
 furnace pneumatically through the RDF nozzle described previously.
 A schematic of  this system is shown in Figure 4-8.
                          86

-------
  RDF feed system and firebox.
          RDF feed system.
Figure 4-7.   RDF feed system design.
                     87

-------
00
oo
                         Nitrogen
                          Supply
Compressed Air
    Supply
                                                                                 Metal
                                                                                 Standoffs
                                       Figure 4-8.   Pneumatic  transport  system.

-------
Safety System
       The flame safeguard system is designed to prevent flash-back or
propagation of refuse flames up the refuse injection tube and onto the belt
system.  This possibility results from the positive pressure in the furnace
and the combustible mixture of air, refuse and dust in suspension in the
downcomer tubes.  The system is shown schematically in Figure 4-9.
       As shown, a Honeywell UV flame detector is positioned with a clear
line of sight into the burner tube.  A flame signal from the detector is
sent to a control panel where three steps are taken automatically:
       1.  Refuse belt is shut off
       2.  Refuse air supply is shut off
       3.  Nitrogen purge is started
These  steps ensure immediate loss of combustion essentials  in  the down-
comer  and burner tubes.  The activated nitrogen purge duration was  set  for
10 to  30 seconds.
4.2.3  Materials Acquisition and Handling
       As stated in  the  Objectives  section,  one of the  purposes  of  this
investigation was to  better characterize  combustion efficiency and  emissions
as a  function of refuse  type.  Therefore, material  had  to  be transported
from  several  locations to  the  Acurex research  facility.   In order  to  assure
that  the  procedures  used in obtaining  this  type of material  complied  with
state and  federal  regulations, coordination with  local, state, and federal
 authorities  was required.   In  particular, coordination  with the  State of
 California  Department of Food  and  Agriculture  was necessary because of
 the  potential  entomological dangers of shipping RDF from various parts of
 the  country into the state.  In  order  to protect  against these dangers,
                                      89

-------
                                                 Nitrogen
                                                 supply
Air
supply
                                                               NO
                                 Mini peeper
                                 UV scanner      ^
Solenoid
   valve
                                                             Plexiglass
                                                             tube
                      Burner tube
                                                                   NO
                                                             Air
                                                            supply
                                 Figure  4-9.   Safety  system.

-------
the State of California Department of Food and Agriculture,  the  Santa
Clara County Department of Agriculture, and Acurex agreed upon the  follow-
ing packing, shipping and inspection procedures.   The RDF material  was:
       •   Fumigated outside of the State of California using the procedure
           recommended in the U.S. Department of Agriculture Plant Protec-
           tion and Quarantine Manual, T403 (e)-(2), section 6,  page 6.
           This treatment is methyl bromide fumigation under a tarpaulin
           using 10 pounds of methyl bromide per 1000 ft  of RDF at atmos-
           pheric pressure for 48 hours at 40°F or above.
       •   Certified in writing by a state or federal agricultural agent
           as to compliance with the fumigation procedure described above.
       •   Sealed and shipped in a sturdy container with a rigid,  insect-
           proof frame and a leak-proof plastic liner.  Wooden  boxes, 44
           inches x 48 inches x 100 inches, were  constructed  and fitted with
           plastic liners for shipping the 2000-pound  lots of RDF  from each
           each of the facilities.
       •   Received and  unopened  until notification was  given to the County
           of Santa Clara Department  of Agriculture  so that  a Santa Clara
           County Agricultural agent  could  inspect the containers and the
           written  fumigation certification  and agree  to unloading.
       After inspection  and  approval,  the RDF was stored on  the Acurex
 premises  in  its shipping containers  on an outdoor concrete  pad  and fully
 covered  by  a tarpaulin.   This protected  the  RDF from any degradation or
 attrition from  rain  and  wind during  storage  and testing.
        Emphasis was  placed  upon safety during all RDF  test  operations.   All
 of the test personnel  in contact  with the RDF received tetanus, typhoid,
 and diphtheria  series  of vaccinations, and personnel  in  direct  contact with
                                     91

-------
 the  RDF were  required  to wear  respiratory  filtermasks.   In addition, daily
 changes of clothing and footwear were required of test personnel in con-
 tact with the RDF so that no contamination was carried around the work-
 place or to the home.  In addition, cleanliness of the test facility was
 maintained.   Floors and equipment were cleaned daily to  insure a safe
 and  hygenic workplace.
       The as-received RDF was not compatible with the test facility feed
 system.  Early calibration tests revealed that some of the large particles
 in the RDF plugged the feeding mechanisms and interrupted testing.  In
 order to uniformly feed the RDF into the combustion test unit, the RDF had
 to be reduced in size.  Early testing with the test feeder system proved
 that RDF particles of 1 inch or less were suitable for controlled combus-
 tion testing.
       Therefore, the RDF was passed through a commercial garbage composter
 and reduced to 1 inch or less with no other alteration in the RDF composi-
 tion.  This size reduction process was conducted at the test site during
 the combustion tests.
       At the  end of testing, any RDF which was unused was hauled to a
 landfill  area  for disposal.   The major consideration was to dispose of
the unused RDF as soon as possible to prevent insect infestation or putre-
faction of the RDF.
 4.2.4  Sampling  Equipment and Procedures
       The sampling required for this project included collection of gaseous
 emissions by continuous monitoring equipment, collection of flue particu-
 late and gases for trace metal, organic and anion analysis, collection of
 residual ash for detailed analysis, and sampling of the  input fuel.  The
                                    92

-------
methods and equipment for each of these sampling tasks are discussed  in  the
following sections.
4.2.4.1  Gaseous Emissions Measurement
       Table 4-1 lists the continuous monitoring equipments utilized  at  the
Acurex Energy Laboratory.  Figure 4-10 shows a schematic of the gaseous
sampling and analysis system.  The system is designed for accurate analysis
of NO, CO, 02, C02, S02, and unburned hydrocarbons.
4.2.4.2  Stack Sampling Equipment
       Two stack sampling systems were used during the course of testing.
The high volume stack sampling system as shown in  Figure  4-11 was used to
determine the particulate grain loadings and size  distribution.  This sys-
tem meets or exceeds the  EPA Method  5 requirements.   The  second system was
the Source Assessment Sampling System (SASS) shown schematically in Figure
4-12.  The SASS is used for both sampling of particulates and organics.   The
sample is drawn through a glass-lined sampling probe  and  routed through  a
series of three cyclones  and a filter which separates the particulate into
four  size fractions.  Both  the probe and particulate  removal  system are  in
a  400°F  oven  to prevent  condensation.  The  gaseous sample then passes
through  an  organic module where  it  is  cooled  and  the  organics are  trapped
on a  polymer  adsorbent.   Condensate from the  module  is also collected for
analysis.   Finally,  the  sample  is  routed through  an  impinger train where
oxidizing solutions  retain  any  remaining sample.   The sample is  then  drawn
through  the control  unit where  pressures,  temperatures, and gas  volume  are
monitored and controlled.  An S-type pitot  is used to measure gas  velocity
 for the  purpose of determining  isokinetic  sampling rates.
                                     93

-------
TABLE 4-1.   EMISSION MONITORING EQUIPMENT
Pollutant
NO
so2
CO
co2
°2
UHC
Participate
Loading
Principal of
Operation
Chemi luminescence
Pulsed Fluorescent
Nondispersive
Infrared (NDIR)
Nondispersive
Infrared (NDIR)
Paramagnetic
Flame lonization
Cyclone and
Filtration
Manufacturer
Ethyl Intertech
Thermoelectron
Ethyl Intertech
Ethyl Intertech
Ethyl Intertech
Ethyl Intertech
Acurex Corp
Models
Air Monitor-
ing
Teco
Model 40
Uras 2T
Uras 21
Magnos 5A
FID
HVSS
Instrument
Range
0-5 ppm
0-10
0-100
0-250
0-1000
0-5000
0-50 ppm
0-100
0-500
0-1000
0-5000
0-500 ppm
0-2000
0-5%
0-20%
0-5*
0-21%
0-100 ppm
0-300 ppm
0-1000 ppm
0-3 vim
Minimum
                   94

-------
                                               Sampling probe heated filter
vo
en
                                                                         filters    Pulsed fluorescent
                                                                         filter
                        Figure 4-10.   Sampling system online at  experimental multiburner furnace.

-------
                                                                                unit
                                                   25-Foot umbilical line
                   Oven with
                   cyclone
                   and filter
Impinger train
and ice bath
       25-Foot sample hose
0 cfm vacuum pump
                                                  Pump-control  unit hose
                   Figure 4-11.  Aerotherm high volume stack sampler.

-------
                                              Fan
                                             Oven
                       P1H«r
     Sttel T.C.
^— Oven               Fine
                                         Sensor
                                      umbilical
                                         AP. T
Gas conditioner I
moisture col lector
                                                                                       Porous  polymer
                                                                                       adsorber
IMtot AP
  Gage
                                               Parallel
                                              Imp/cooler
                                           trace element
                                               col lectors
            Dry Gas Meter
            Orifice Meter
           Centralized Temperature
             and Pressure Rea.1  jt
                                                                   Semiautomatic
                                                                      control
                                                                      module
               Control Module
                                                       10 cfm vacuum  pump


                           Figure 4-12.   Source  assessment sampling system  (SASS).

-------
4.2.5  Problem Area Summary
       As was discussed earlier, one of the objectives of this investiga-
tion was to design, fabricate, operate, and evaluate a laboratory-scale
system for combustion testing of RDF.  Since, to our knowledge, no other
refuse investigation had been performed on this scale, it is our intention
to fully document the areas where problems occurred during this investiga-
tion.
       Since the investigation was conducted in the suspension fired, tan-
gential configuration, as descrbed in Section 4.2.2, two complete refuse
systems were required for opposed refuse input.  The complexity of each
system alone required careful monitoring along with the primary fuel systems.
The RDF feed systems were evaluated as follows.
Drum System
       The mechanical  fluidization system performed very well throughout
testing,  although there were several times the system went down on one side
or the other.   Problems resulting in downtime were:
       1.   The weight of the drum system caused compression of the forward
           cam system and resulted in binding at the teflon bearing-space
           interface.   This problem occurred six to eight times during the
           3-week test period and was quickly corrected each time by adjust-
           ing the cam vertical  position.   This problem may be averted al-
           together by utilizing a noncompressing material  for the cam.
       2.   The nature of the refuse was such that small pieces of paper,
           glass, etc., being fluidized in the rotating drum found their
           way into the space between the nylon bearing seals in the for-
           ward drum.   This resulted in binding at least three times during
                                      98

-------
           testing.   The problem was  resolved  by  removing  the  seal  cover
           and cleaning the entire surface.  This required 1 to  1-1/2  hours
           to complete.
       Fluidization  of the material was excellent throughout testing  as
long as the material  was in a fluffy, dry condition.
Belt System
       As was mentioned earlier, fluidization  of the material  was poor
when it was moist or packed prior to  loading in the drums.  This resulted
in clumping on the belt system which  caused inconsistent input to the
burners.  The drum exit was such that material agglomeration resulted in
build up at the exit.  This periodically stopped the belt or caused the
belt to be stripped of material at that point.
       A more efficient deflector design could solve this problem, but
more important is providing a properly conditioned feed consistently.
       The feed was determined  by calibrating the belt rpm against mass
delivered.  While this  system was somewhat accurate initially, as  testing
on each  fuel  continued,  it became apparent that  the calibrations weren't
holding.  Therefore,  the  RDF input was determined by back calculating from
the flue gas  analysis.   The errors resulted from the differences in  refuse
density  from  layer to layer in  the storage bins.  Visible differences of
the refuse characteristics were evident  from  day to day on the  same  fuel
types.
 Belt  to Burner Transition
        Most  of the  plugging problems incurred in this  area were due  to  in-
 consistency  in the  refuse sizing. When  plugs occurred,  sweeping the tube
 clean was  easy and  quick, if approached  properly.   After initial trial  and
                                      99

-------
 error,  elimination  of  refuse  plugs was a  secondary problem.  However, if
 material  flow  sensors  allowed buildup on  top of  the plug to occur, the
 tube  had  to  be  removed  in order  to sweep  the plug.
        Generally, a  coated or more scratch  resistent tube  is the only im-
 provement that  could have been made.
        Plugging in  the  gun occurred when  compacted material from the down-
 comer was forced into  the gun.   Other plugs resulted from  foreign objects
 such as wire becoming  lodged  within the system.
 Material  Preparation
        Generally, a  great deal of the handling problems would be eliminated
 if a material  preparation system yielded  the same type product each batch.
 The problems resulting  from this are probably nonexistent  on a large scale,
 but become relevant  on  this small scale.  The material is  very absorbent
 in this state  and should be guarded from  heat and humidity.
 Stack Sampling
       The physical  nature of the particulate product of RDF/gas cofiring
 resulted  in extended sampling periods to  collect the required volume of flue
 gas sample.  As  shown  in Figure  4-12, all particles less than 1.0 micron
 sizing were collected  in a fiberglass filter upstream of the gas conditioner.
 In all cases with RDF/gas cofiring, all solids collected were smaller than
 1.0 micron.  Therefore, frequent filter changes were required to complete
 sampling.
       During RDF/coal  cofiring tests,  the small  particulate apparently
adhered to the larger coal  ash particles  and were captured  in the cyclones.
This  eliminated 80 percent of the sample  and were captured  in the cofired
tests.
                                    100

-------
Ash Deposition
       During RDF/gas cofiring, bottom ash deposition was minimal,  although
after approximately 20 hours of testing at concentrations of 30 percent
refuse, on a heat input basis, some ash deposits were removed from the
ashpit.
       The ash deposition during RDF/coal cofiring displayed characteris-
tics unique to that mixture.   Daily ash collection was necessary for all
ratios of coal/RDF.  However,  during approximately 25 hours of testing at
a  refuse concentration of 30 percent, on  a heat input basis, a bridging
of ash occurred across the ashpit entrance.  This occurrence is illustrated
schematically  in  Figure 4-13.  The fused  ash material recovered weighed ap-
proximately 100 pounds.
       This problem  has never  occurred at this  facility  before during  over
3  years of coal testing at somewhat higher ash  input rates.
4.3    TEST THEORY AND PLAN
       In order to achieve the objectives noted in Section 4.1, the experi-
mental program consisted of two basic elements:
       •   Baseline  tests --  Evaluation  of feed systems  and  characteriza-
           tion of emissions  during gas/refuse  cofiring.
       •   Detailed  tests --  Evaluation  of combustion efficiency  and  conven-
           tional emissions  controls  during  coal/refuse  cofiring.
4.3.1  Baseline Tests
       As  stated  above,  the  purpose  of the  baseline tests was  to  evaluate
 the refuse feed systems,  using all  four types of refuse, and characterize
 the combustion performance  and emission levels from each of the materials
 cofired  with gas.  Based on the results of these tests,  one of the refuse
 materials was selected for use as the detail test fuel.   Also, modifications
                                    101

-------
        Firebox  exit
                                Ash bridge
                            Ashpit
                            entrance
Internal firebox, side view
Figure 4-13.   Ash deposition.
              102

-------
and adjustments of the refuse feed system could be made in order to assure
consistent fuel input during the lengthy detailed test points.   The emissions
produced by each of the refuse types were used to illustrate the uniqueness
of each refuse and to obtain background values for the chosen detail test
fuel.
       The test matrix developed for the baseline testing is shown in Fig-
ure 4-14.  As noted on all the matrices, sampling is  divided in  three
levels of detail.  Level 1 consisted of gaseous emissions sampling only.
Level 2 sampling included gaseous emissions and stack particulate loading
tests.  The Level 3 sampling included tests under Levels 1 and 2 plus de-
tailed stack sampling for trace metals and organic compounds in the  stack
flue  gas.
           As  indicated in Figure 4-14, the bulk of the testing was completed
at 20 percent  excess  air conditions at a heat  input rate of  1.5 x  10  Btu/hr.
Tests at other conditions were  necessary for  background levels  to  be used
with  the detailed test  results.
4.3.2 Detailed Tests
       After completion of  the  baseline  tests  and  selection  of  the  detail
test  refuse, the  detailed test  matrix,  shown  in  Figure  4-15, was  addressed.
As noted,  the  purpose of  these  tests  was to evaluate  the  combustion  effi-
ciency of  a  refuse/coal  fuel  mixture  and to evaluate  conventional  emissions
control,  i.e., theoretical  air, on  the  resulting emissions.   Sampling test
nomenclature was  consistent with that used during the baseline  tests.
        As  noted in  Section  4.2.5,  the particle size  of the stack  particulate
 produced during the gas cofired points  was such that sampling  time required
 for  both Level 2 and 3 tests was increased by a factor of 2.  This resulted
                                    103

-------
        4 RDF Types
     1)  Gas coflre
     2)  Theoretical air
     3)  S RDF
     4)  Residence time to convectlve section
     5)  Firing rate
TA - theoretical air;  LRT -  long residence  time; SRT - short residence time;
105X.  110X. IZOt. 130t
  St.   lot,  201
 (short, long)
1.0 x  10* and 1.5 x 10*  Btu/hr
.1.   Gaseous emissions  sampling only
 2.   Gaseous emissions  plus flue gas
     participate loading and size
     distribution
 3.   Detailed emissions sampling

u.
g
X
in
u.
o
Of
t*
o
u.
o
Of.
»•
o
•VI
fc
i/>
£
&
v\
&
_l
&
VI
t
_l
105S TA
1.0 x 10* Btu/hr






1.5 x 10§ Btu/hr

1

1

1
not TA
1.0 x 10* Btu/hr






1.5 x 10« Btu/hr

1

1

1
1201 TA
1.0 x 10* Btu/hr



1


1.5 x 10* Btu/hr

2

3

2
130* TA
1.0 x 10* Btu/hr






1.5 x 10* Btu/hr

1

1

1
                         Figure 4-14.  Test matrix for  baseline  emissions  characterization.

-------
1)  Cofire RDF *1th coal
2)  Firing r«te 1.5 x 10' Btu/hr and 1.0 x 10' Btu/hr
3)  Vary residence tine to corrective section
4)  U.C. RDF, variable X
5)  Vary theoretical air



LL.
O
at
»«
ttt
u.
a
tr
S
Li.
O
a
»*
o
2*
§S5
_l CC »-
4J -D
l«l
VI -
0,2*
c M e
O «J ^
-ICC H-
*> "d
U f- «j
0 m E
^ OJ --
irt a t-
T>
cn-~- u
§vi E
D —
— ICK t-
t2fc
isl
Irt K t-
0,2 .
gjcl
-IQCt-
•
***
J88J1
lAOCI-
1051
1 n MBtu
'•° hr.


1





7A
, c HBtu
KS hr
1

1

3

2

110S
, .. MBtu
1'° hr


1

1

1


, c MBtu
'•5 hr
1

3

2

1

120J
1 0 MEtu
'•° hr
1

2

2

1

TA
, c KBtu
K5 hr
2

3

3

2

130%
, n KEtu
'•° hr


1

3



TA
1 5 M"J
'•5 hr
1

3

1

2

 1.   Gaseous emissions sampling only
 2.   Gaseous emissions sampling plus flue gas
     partlculate loading and size distribution
 3.   Detailed emissions sampling
    Figure  4-15.
Test matrix for emissions control  through
theoretical air variation.
                                  105

-------
 in a loss of sampling points during the detail  testing.   Figures  4-16  and
 4-17 illustrate the completed test matrices.   The detailed testing was
 focused on Level 2 and 3 points as indicated.
 4.4   ANALYTICAL PROCEDURES
       This investigation required detailed chemical analyses of fuel
samples, and gaseous and solid stack product samples.   All  of these analy-
ses were completed at the Acurex Analytical Laboratory with the exception
of analysis of fuel samples.
       The methodology of these analyses is outlined below.
4.4.1  Fuel Sample Analysis
       Representative samples of all fuels tested during  this investigation
were submitted to a certified commercial laboratory for ASTM standard
analyses listed in Table 4-2.
                         TABLE 4-2.  FUEL ANALYSES
                  Proximate Analysis
                  Moisture
                  Ash
                  Fixed Carbon
                  Volatile Matter
Ultimate Analysis
Carbon
Hydrogen
Nitrogen
Oxygen
Sulfur
Chlorine
Heating Value
 4.4.2  Trace Metal  Analysis
        Trace analyses of metals were conducted using atomic absorption
 spectroscopy by standard EPA and ASTM methods.  The metals which were ana-
 lyzed are listed in Table 4-3.   Particulate fractions from the sampling train
                                     106

-------
         4 RDF Types
     1)  Gas coflre
     2)  Theoretical air
     3)  1 RDF
     4)  Residence time to convectlve section
     5)  Firing rate
105X, 110*,  120S, 1301
  5t,  101,  201
 (short,  long)
1.0 x 10' and 1.5 x 10*  Btu/hr
TA — theoretical air; LRT -  long  residence time;  SRT - short residence time;
.1.   Gaseous emissions sampling only
 2.   Gaseous emissions plus flue gas
     partlculate loading and size
     distribution
 3.   Detailed emissions sampling

u.
o
Cf
M
in
i*.
a
Of
M
o
u.
O
ac
•*
o
CM
&
vt
s
_)
f-
ae
i/i
cc
_j
S
i/i
fc
1051 TA
7.0 x 10* Btu/hr






1.5 x 10* Btu/hr

©

©

©
1101 TA
1.0 x 10* Btu/hr






1.5 x 10: Btu/hr

©

©

©
1201 TA
1.0 x 10* Btu/hr



©


1.5 x 70* Btu/hr

©

©

©
130t TA
1.0 x 10* Btu/hr






1.5 x 10* Btu/hr

©

©

©
     O  Completed tests
                       Figure  4-16.   Test matrix  for  baseline  emissions  characterization.

-------
 1)  Coflre RDF *1th coal
 2)  Firing rate  1.5 x 10'  Btu/hr and 1.0 x  10( Btu/hr
 3)  Vary residence time to convectlve section
 4}  U.C. RDF, variable X
 5)  Vary theoretical air

Li.
O
oc
X
U_
0
tt
o


0
o
CM

u.
0
ct
V*
o

•c
(7>^- 0>
C wl £
O OJ —
_J CC t—
1- — Ci
o >^ e
^: c. —
uo ce »-
•o
cm- o>
c m E
.
L_ ^— CJ
O on E
XT Qj ^~
T3
C l/l E
00 —
_4 CC t—
•«-• XI
t-- y
O f» E
to tt; h-
o>— w
C *^l t-
_i ce i—
.
fe— ₯
O kn 6
£ «> •»
to et t—
1051 TA
1 0 HBtu
'•° hr.


1









, MBtu
1-5 hr
©

©



3


2


1101
l n HBtu
'•° hr


1



1


1


, r MBtU
hr
©

©



©


©


120% TA
l n MEtu
'•° hr
1

2



2


1


, ^ H3tu
" hr
2
-
©



©


©


120X TA
K5.tu
hr


1



3




i
, r M3tJ
''* hr
1

3



©


©


1.   Gaseous emissions sampling only
2.   Gaseous emissions sampling plus flue gas
    partlculate loading and size  distribution
3.   Detailed emissions sampling
O  Completed tests
  Figure 4-17.
Test matrix for emissions control  through
theoretical air variation.
                               108

-------
were analyzed after acid or Parr digestion.   For each SASS train,  at least
three samples were analyzed — a proportionally combined representative
participate sample, a sample of the XAD-2 resin, and combined  aqueous
condensate and first impinger solutions after extraction, and the  combined
second and third impinger solutions.  However, only antimony, mercury, and
arsenic were analyzed in the second and third impinger samples.

                   TABLE 4-3.  METALS WHICH WERE ANALYZED
                                 Trace Metals
                                As
                                Be
                                Cd
                                Hg
                                Ti
Sb
Sn
Pb
Cu
Mn
 4.4.3  Organic  Analysis
        Organic  species were  analyzed  by  a  modified  Level  1  analysis  scheme
 (Level  1  Environmental Assessment,  IERL-RTP  Procedures  Manual,  June  1978).
 Basically,  this scheme involves the separation of a sample  extract into
 broad classes  based on liquid chromatography fractionation  and  gravimetric
 analysis.   An  organic extract is placed  on a column of  silica gel  and frac-
 tionated  by elution with increasingly polar  solvents.   Table  4-4 lists the
 solvents  which are used  in the Level  1 scheme.  Each fraction after  solvent
 removal is weighed to yield a rough estimate of material  present.   This
 separation scheme yields seven fractions which will contain the compound
 classes outlined in Table 4-5.
        Selected fractions from the liquid chromatography  separation were
 then scrutinized for specific chemical species.  For this investigation,
                                      109

-------
TABLE 4-4.  LIQUID CHROMATOGRAPHY ELUTION SEQUENCE
Fraction
1
2
3
4
5
6
7
Solvent Composition
Pentane
20 percent methylene chloride in pentane
50 percent methylene chloride in pentane
Methylene chloride
5 percent methanol in methylene chloride
20 percent methanol in methylene chloride
50 percent methanol in methylene chloride
Volume
25 ml
10 ml
10 ml
10 ml
10 ml
10 ml
10 ml
                       110

-------
TABLE 4-5.  DISTRIBUTION OF COMPOUND CLASSES  IN  LIQUID
            CHROMATOGRAPHIC FRACTIONS OF  ORGANIC EXTRACTS
 Fraction                                 Compound  Class

    1                           Aliphatic  hydrocarbons
                               Halogenated aliphatics

    2,3                        Aromatic hydrocarbons
                               Halogenated aromatics (PCB's)

    4,5                        Nonpolar oxygen or nitrogen
                               containing species

    6,7                        Polar compounds - phenols,
                               alcohols, amines, etc.
                               Ill

-------
 the organic  compounds  of  interest are prevalent only in the LC fractions
 2 and  3.  Therefore, samples were collected only from these fractions.
 The sample was then analyzed by gas chromatographic/mass spectrometry
 methods.  During this  analysis the level of investigation was determined
 quantitatively utilizing  the threshold level for nearly all the most toxic
 species as defined by  OSHA, that level being 0.10 mg/m  of sample gas.
 All peaks above this level were analyzed for the following groups or
 species:
       1.  ROM's (polycyclic organic materials)
       2.  PCB's (polychlorinated biphenals)
       3.  Four other  groups or species
       The other groups or species were selected based on the largest
 quantities of materials which did not fall into the two groups specifically
 selected above.
 4.4.4  Quality Assurance and Control
       To assure the quality of the analytical  data, a program used to
 control contamination, calibrate instrument response, and verify qualita-
 tive and quantitative  data is presented below.
 Glassware
       All glassware used in the extraction and analysis of the samples
was cleaned by one of  two methods.   Separatory funnels and volumetric
 glassware were cleaned in a dichromate acid bath, rinsed with deipnized
water, rinsed with acetone, hexane and methylene chloride,  and sealed with
muffled aluminum foil.   All other glassware was washed with soap and water,
 rinsed with deionized  water, rinsed with acetone and muffled at 450°C to
 500°C for approximately 6 hours.   Although not adopted as a standard
                                    112

-------
procedure,  this procedure has  been used by Acurex and  EPA  labs  to  produce
glassware totally free of detectable organic contaminants  for several  years.
Solvents and Standards
       Only Burdick and Jackson "Distilled in Glass" solvents were used
in this program.  Acurex purchsed all solvents in lot quantities to assure
uniform quality throughout the entire study.  A quality check was performed
on each solvent to insure the absence of any interfering  substances prior
to the start of the program.
       All  standards were purchased from commercial supply houses or  from
EPA.   Each  standard was  verified  by GC/MS  prior  to  its  use.
Blanks and  Spikes
       Two  types of blanks were taken:   (1)  a  sampling  train blank  for each
test  and  (2) method blanks  for the solvent extractions.   For each series
of test  runs,  a blank train was set up in the  same  manner as the  actual
operating  train.   The blank train was  capped off at the nozzle and  impinger
 exit  with  aluminum foil.   The train remained assembled  at the test loca-
 tion  for the duration of the  test period.  Sample recovery and analysis
 proceeded  as described for the sampling train.  Method  blanks using the
 same  glassware and solvents as for the actual  samples were taken every 10
 samples and analyzed  as described earlier.
 Metals Analysis
        Trace metal analysis requires a careful adherence to good analytical
 techniques and the measurement of spiked  samples.  To this  end, each  sample
 was  spiked to give an increase in the  initial concentration greater  than
 10 percent but less than 100 percent.  The  recovery was  calculated from
 these data and applied  to the values  found.
                                      113

-------
       Standards were diluted from stock each day and a standard curve
plotted at the beginning and end of each analysis for that element.  The
standard curve was selected in such a way as to bracket all of the sample
concentrations for the run.  After each 10 samples, at least one standard
was rerun at the level that approximated most of the sample concentrations.
Replicates were run at regular intervals to establish precision of the method
and spike, and recovery for the accuracy data.
4.5    EXPERIMENTAL DATA
       In this section, the experimental results for completed tests will
be presented.  This will include data on the fuel samples, gaseous emissions,
particulate emissions, trace metals, and the organic emissions.  Table 4-6
lists the test point designations and their corresponding test conditions
for referral from the test data.
4.5.1  Fuel  Samples
       During the testing phase of this investigation, the fuels were being
continually sampled to better characterize the inputs.  At the completion
of the testing, these gross samples were combined and sent to a commercial
laboratory.   Representative samples were drawn and analyzed as discussed
in the previous section.  The results of those analyses are listed in Table
4-7.   Photos of the fuel samples are shown in Figure 4-18.
4.5.2  Gaseous Emissions
       While the objective of this investigation was primarily to charac-
terize organic and trace metal emissions from conventional fuel/refuse
fuel  mixtures, gaseous emissions were also fully documented.  Discussion
of gaseous emissions will be limited to oxides of nitrogen and sulfur
dioxide primarily, due to their importance in environmental considerations.
Full  gaseous data are documented in the appendix.
                                     114

-------
                TABLE 4-6.  TEST MATRIX
Test
Point
11A
B
C
D
13A
B
C
D
40
15
37
38
32
31
19
35
34
Fuel
Gas/Ames
Gas/Richmond
Gas/Americology
Gas/San Diego
Gas/Ames
Gas/Richmond
Gas/Americology
Gas/San Diego
Pitts Coal
Coal/Richmond












Coal/Richmond
RDF
Cone*
10%


10%
20%


20%
—
5%
10%
10%
20%
20%
30%

30%
Combustion Conditions
1.5xl06Btu/hr 20% EA





























20% EA
10% EA
10% EA
20% EA
10% EA
20% EA
10% EA
20% EA
1.5xl06Btu/hr 30% EA
Heat input basis
                            115

-------
                                                     TABLE 4-7.  FUEL ANALYSES
cr>
Ultimate Analysis*
Carbon %
Hydrogen %
Oxyen %
Nitrogen %
Sulfur %
Ash %
Moisture %
(as received)
Chlori ne %
Heating Value
Btu/lb
Fuel Type
Pittsburg
No. 8 coal
75.23
5.15
8.12
1.49
2.51
7.50
0.93
0.14
13,545
Richmond
refuse
42.60
6.26
37.90
0.83
0.16
12.25
23.8
.46
7696
Ames
refuse
40.49
6.01
30.04
0.73
0.35
22.38
15.2
.43
7R31
Ampri cology
refuse
40.29
5.88
25.20
0.91
0.17
27.55
24.4
.72
7164
San Dieqo
refuse
38.01
5.64
17.40
0.69
0.21
38.05
26.3
.79
7146
                    Dry basis

-------

                   Ames
                Americology




Figure 4-18.   Photographs of fuel  samples
                    117

-------

/
                        Richmond
                                                           ,*
                                                                '
                        San  Diego
                Figure  4-18.   Concluded,
                           118

-------
       It should be noted that during this  and a  previous  investigation,



sulfur dioxide emissions data were inconsistent.   Following this  investi-



gation, the Pulsed Florescent S02 Analyzer  was returned to the manufacturer



for evaluation.   The source of the inconsistent data was determined to be  a



photomultiplier tube which rendered the S02 data during this investigation



invalid on a quantitative basis.  However,  the data is valid on a relative



basis and should be regarded as such.



       As discussed in  the test plan, baseline testing to characterize the



combustion of refuse was conducted first.  This was accomplished by co-



firing each of  the  four refuse  types with natural gas  and by examining



several  variables.  These  variables  included  excess air and concentration



of  refuse on  a  heat input  basis.  All other combustion parameters were



held  constant.



       The  results  of  this baseline  testing are  illustrated  in Figures 4-19



through  4-22  where  NO  is plotted  as  a function of excess  air  percentage for



each  of  the four refuse types.   The  refuse  concentration  effects  are  also



 illustrated.   In each  figure, a baseline point  is  plotted.   This  point,



 taken with  natural  gas as the fuel,  represents  NO  formed  through  thermal



 fixation of the atmospheric bound nitrogen.   Figure 4-23  represents data



 taken during  previous  work on natural  gas.   The baseline  point taken  during



 this investigation is  plotted to demonstrate the validity of the NO level.



 Using this as a baseline illustrates qualitatively the contribution of fuel



 bound nitrogen to the  total NO emission.



         It should be noted in Figure 4-20 that the 20  percent  Richmond curve



 falls below the curve  representing  10 percent Richmond fuel.  This is
                                       119

-------
                             200
                        CM
                       O
                       ^    100
no
o
                                            10           20
                                              Excess air (percent)
30
              O  Gas Only
               0  Ames 5%
              D  Ames 10%
              O  Ames 20%
                   (Heat Input basis)
                   Tangential mode
                   1.50 x 106 BTu/hr
                   300°F Secondary A1r
                            Figure 4-19.  NO emissions during baseline testing (Ames)

-------
r>o
                            200
                         CM
                        O
                        V!
                        O
                        S-  100
                                           10
                                                                                   O    Gas Only
                                                                                    0    Richmond 5%
                                                                                   Q    Richmond 10%
                                                                                         Richmond 20%
                                                                                         Tangential  Mode
                                                                                         1.50  x  106  BTu/hr
                                                                                         300°F Secondary  A1r
                                               Excess  air  (percent)
                         Figure  4-20.   NO emissions during baseline testing (Richmond)

-------
                           200
                       CVI
                      o
                       O.
                       Q.
                           100
IXi
                                           I
I
                                          10          20          30
                                             Excess air (percent)
                   Gas  Only
               Q  Americology
               O  Americology  20%
                   (Heat  Input  basis)

                 Tangential Mode
                 1.50  x  106 BTu/hr
                 300°F Secondary A1r
                       Figure 4-21.  NO emissions during baseline testing  (Americology)

-------
                           200
                        CM
                       O
                       Q.
                       Q.
                           100
ro
oo
                                          10          20

                                             Excess air (percent)
30
                     Gas  Only

                D   San  Diego  10%

                O   San  Diego  20%

                    (heat input basis)
                   Tangential mode

                   1.50 x 106 BTu/hr

                   300°F Secondary Air
                       Figure 4-22.  NO emissions during baseline testing (San Diego)

-------
                   200
                OJ
               o
               Q.
               Q.
                   TOO
ro
                     0
                                         Data  from  previous  work
                      17
18
  19          20          21
Stack Gas Temperature °F x 102
22
                                                      Tangential Mode
                                                      Gas Firing
                                                      1.0 x 106 Btu/hr
                                                      20% Excess Air
                                                      O  Baseline Point
                                       Figure 4-23.  Thermal NO  (previous work).

-------
believed to be the result of lower thermal NO contributions resulting from
cooler flame temperatures.   The 20 percent Richmond/gas flame was extremely
luminous which resulted in higher radiation losses from the flame and an
overall cooler flame.  It is well documented that thermal NO is very sensi-
tive to temperature.
       Note that neither the Americology nor the San Diego fuel curves
contain 5 percent by heat input refuse concentration fuel mixtures.  This
is due to the density of these two fuels.  The feed systems were not capa-
ble of delivering a consistent feed at the required low  flowrates.
       A comparative analysis of  the combustion characteristics  of each of
the individual fuel  types is illustrated  in  Figure 4-24.   Shown  in this
figure are  curves representing a  constant  fuel mixture consisting of na-
tural  gas and each  of  the refuse  types with  all the other  parameters held
constant.   The fuel  nitrogen content of  each mixture  is  listed in the
legend.  The  order  of  the curves  in  Figure 4-24 demonstrates  the fact  that
each  refuse contributed  to  the overall NO level in a  unique manner.  While
the curves  representing  the Ames  and San Diego  source mixtures are  consis-
tent  with  the chemical  relationship, the Richmond source fuel  is clearly
varying  in  fuel  nitrogen concentration.
        As  noted  earlier, all  sulfur dioxide  data  is  valid  only on a  rela-
 tive  basis.  However,  a good comparison  of fuel  types is illustrated in
 Figure 4-25,  where  curves  for  each refuse type  cofired with  gas at  30  per-
 cent  excess air are plotted.   The sulfur analysis of each fuel is also
 listed.   These  curves  demonstrate the  unique characteristics which each
 refuse exhibits in a combustion environment.
                                      125

-------
                                                                                                N
                      zoo
ro
                  CVJ
                 o
                  0.
                  Q.
                      100
                        0
                                      20%  Refuse/natural  gas
                                     I
I
I
                                     10          20           30
                                        Excess air (percent)
               O
               Q
               D
                                Ames
                                Richmond
                                Americolorjy
                                San Dieqo
0.73
0.83
0.91
0.69
                                                                                  Tangential  Mode
                                                                                  1.50 x 106 BTu/hr
                                                                                  300°F Secondary Air
                    Figure 4-24.   NO emissions during baseline testing (all RDF's)

-------
   175 r
    125
 CM
o
    100
CL
Q-

 CO
O
 QJ
Tangential mode
Gas co-fire
1.5 x 10s Btu/hr
30% excess air
Ames

Richmond

Americology

San Diego
S (%)

 0.35

 °-16

 0.17

 0.21
     75
 OJ
ac.
     50
     25
                    5           10          15          20
              Refuse concentration, percent (h.v. basis)


                   Figure 4-25.   S02 data  (all RDF's).
                                   127

-------
       All gaseous emission data for natural gas testing is listed in the
appendix for completeness.  In general, however, carbon monoxide, carbon
dioxide and unburned hydrocarbon measurements were consistently low through-
out the baseline tests.
       The refuse/coal cofired tests were focused on obtaining stack gas
analyses other than gaseous emissions.   However, gaseous emissions were
recorded and are presented in the appendix for completeness.  The NO emis-
sions for coal  cofiring are summarized in Figure 4-26 where the effect of
excess air percentage and refuse concentration in the fuel  mixture are
illustrated.   As is shown, a general downward trend is exhibited as refuse
concentration is increased.  This trend was first believed to be the result
of cooler flame temperatures reducing the thermal NO and a reduction in
the amount of fuel  N available.   However, an examination of the fuel nitro-
gen availability using thermal NO data  taken from Figure 4-23 as a function
of temperature  indicates that a  reduction in fuel nitrogen conversion is
the likely source of lower NO levels.   This data is shown schematically
in Figure 4-27  where fuel nitrogen conversion and fuel  nitrogen availability
is plotted.
4.5.3  Particulate Analyses
       The results of the particulate analyses are presented according to
variation of combustion conditions.
4.5.3.1  Refuse Type
       Table 4-8 lists the results as a function of refuse type for the gas
cofired tests.   Each point is also expressed as percent of total mass to
illustrate where the bulk of the loading lies, according to size.  As noted
                                    128

-------
ro
10
                    CM
                   o
                   E
                   ex
                   Q.
                          600
300
                         200
                         100
                                      Richmond refuse/Pittsurgh coal
                                        I
                                                    D
                           I
I
                                       10          20          30
                                          1 RDF (heat input)
                                                       D   107, Excess Air
                                                       <£>   20Z Excess Air
                                                       Q   30'^ Excess Air

                                                            Tangential Mode
                                                            1.50 x  106 BTu/m
                                                            300"F Secondary Air
                       Figure  4-26.   NO  emissions  during  detailed  testing  (Richmond
                                      RDF/Pittsburg coal).

-------
CO

O
           O>
           c
           O
            (U
            en
            o
            u.
           r-   5.0
            O)

            3
                2.5
Q  Nitrogen Conversion



tD  Nitrogen Available
Tangential mode

1.5 x 106 Btu/hr

300°F Secondary air

10% Excess air
                                                                                                     2.0
                                                                                                          Ot
                                                                                                          cr
                                                                                                          n>
                                                                                                          c
                                                                                                          ro
                                                                     1.0  o
                                                                          to
                                                                                                          cr
                                       I
                               I
                                      10                       20


                                        Refuse content, % (heating value basis)
                                       Figure 4-27.  Fuel nitrogen contribution.

-------
                      TABLE 4-8.  PARTICIPATE ANALYSES:   EFFECT OF RDF TYPE
Type/
% RDF
Ames/
20%
Richmond/ •
20%
Americology/
20%
San Diego/
20%
Ames/
10%
Richmond/
10%
Americology/
10%
San Diego/
10%
Test*
No.
13A
% of Total
13B
% of Total
13C
% of Total
13D
% of Total
11A
% of Total
11B
% of Total
nc
% of Total
11D
% of Total
Excess
Air
(%)
20














Filter
(gr/ft3)
0.03912
68.96
0.03223
90.94
0.04063
85.65
0.06158
79.59
0.03165
81.51
0.02445
94.04
0.06330
87.05
0.06684
88.32
>io „
(gr/ft3)
0.01067
18.81
0.00044
1.24
0.00233
4.91
0.00732
9.46
0.00317
8.16
0.00095
3.65
0.00283
3.89
0.00354
4.68
>3u
(gr/ft3)
0.00343
6.05
0.00042
1.19
0.00232
4.89
0.00245
3.17
0.00213
5.49
0.00036
1.38
0.00288
3.96
0.00137
1.81
>lu ,
(gr/ft3)
0.00351
6.19
0.00235
6.63
0.00217
4.57
0.00601
7.77
0.00169
4.35
0.00023
0.88
0.00370
5.09
0.00391
5.17
Total
(gr/ft3)
0.05673
0.03544
0.04744
0.07737
0.03883
0.02600
0.07272
0.07568
% Ash
22.38
12.25
27.55
38.05
22.38
12.25
27.55
38.05
Fired with natural  gas

-------
in Section 4.2.5, the highest percentage loading consistently occurred
in the less than 10 micron (y) range which was trapped in the filters.  As
can be noted from the table, the grain loading corresponds roughly to the
percent ash in the fuel.  It can also be observed that the Richmond fuel
consistently had the highest percentage of particules in the less than 1  y
size cut.   Similarly, the Ames fuel consistently had the lowest percentage
in this range size cut.
4.5.3.2  Excess Air
       Table 4-9 shows the effect of excess air for the coal cofired tests.
Both 10 percent and 20 percent refuse concentration points are shown.  Note
that in the coal cofired tests, the majority of the loadings were evenly
distributed in the size  ranges larger than 1.0 micron (y).  Several addi-
tional  comments can be made regarding Table 4-9.
       •   The percent material in the less than 1  y size cut increases
           with excess air.   If the particulate is  friable, the increase
           in velocity may cause more of the material to break up into the
           smaller size  fraction.
           The total  grain loadings decreased as the excess air increased,
           but more rapidly than straight dilutions would account for.  This
           could indicate that there is more unburned carbon in the particu-
           late at the lower excess air levels.  Thus it appears that the
           effect of excess air is to lower the overall grain loading while
           concentrating more of the particulate in the respirable size
           fraction.
                                    132

-------
                              TABLE 4-9.   PARTICULATE ANALYSES:  EFFECT OF EXCESS AIR
co
CO
Test
No.
37
% of Total
38
% of Total
34
% of Total
32
% of Total
31
% of Total
Excess
Air
(«)
10

20

30

10

20

y*
h
RDF
10










20






Filter
(gr/ft3)
0.02287
2.90
0.04439
7.60
0.04833
9.25
0.05455
5.27
0.02583
9.13
>10u
(gr/ft3)
0.39145
49.59
0.26900
46.07
0.25437
48.49
0.45018
43.51
0.13386
47.34
>3y
(gr/ft3)
0.29222
37.02
0.22627
38.75
0.20814
39.82
0.39695
38.37
0.10736
37.97
>lu ,
(gr/ft3)
0.08274
10.48
0.04422
7.87
0.03725
7.13
0.13290
12.85
0.01570
5.55
Total
(gr/ft3)
i
0.78930

0.58389

0.52276

1.03459

0.28276

                  Fired with coal

-------
4.5.3.3  RDF Concentration
       The effect of refuse concentration, cofired with coal, is shown
in Table 4-10.  Results are listed for both 10 and 20 percent excess air
levels.  An even distribution again occurred in the larger than 1.0 micron
(y) size ranges.  This table is merely a rearrangement of the previous
table.  The only point that needs to be reemphasized here is that it ap-
pears that the fraction in the less than 1 y size cut increases with in-
creasing percent RDF.  However, these conclusions should be addressed with
a bit of caution because the total grain loadings did not increase with per-
cent RDF in all cases.   The reason for this apparent data scatter is not
clear at this time.   However, it could be caused by holdup in the heat
exchange sections of the furnace, by non-isokinetic sampling in the duct,
or by the wall and probe effects in the small  exhaust duct due to the stand-
ard large EPA method 5 sampling probe.
       A summary curve of the data shown in the table is shown in Figure
4-28 as the cumulative percent less than a given particle size.  This again
shows the trend of a higher percentage in the  less than 1 y size cut as
the percent of RDF increases.
4.5.3.4  Fuel Makeup
       Table 4-11 compares the particulate loadings of the three fuel mix-
tures.  These results are further illustrated in Figure 4-29 where particu-
late loading is plotted along with fuel ash content.  The table illustrates
that the fraction in the less than 1 y size cut is increased when firing
coal  alone.   This indicates that the RDF is contributing to this fraction
and probably not agglomerating to the larger coal particles.  The figure
                                    134

-------
                              TABLE 4-10.   PARTICULATE ANALYSES:   EFFECT OF PERCENT RDF
to
Test
No.
37
% of Total
32
% of Total
38
% of Total
31
% of Total
35
Excess
Air
(%)
10



20

20

20
*
%
RDF
10

20

10

20

30
Filter
(gr/ft3)
0.02287
2.90
0.05455
5.27
0.04439
7.60
0.02583
9.13

>10u
(gr/ft3)
0.39145
49.59
0.45018
43.51
0.26900
46.07
0.13386
47.34

>3y .
(gr/ft3)
0.29222
37.02
0.39695
38.37
0.22627
38.75
0.10736
37.97

>lu ,
(gr/ft3)
0.08274
10.48
0.13290
12.85
0.04422
7.57
0.01570
5.55

Total
(gr/ft )
0.78930

1.03459

0.58389

0.28276


                   *Fired with coal

-------
                   TABLE 4-11.   PARTICULATE  ANALYSES:   COAL  VS.  10% RDF + COAL VS.  10% RDF + GAS
co
cr>
Test
No.
40
% of Total
28
% of Total
11B
% of Total
Excess
Air
(*)
20






% RDF
+ Fuel
Coal

10% RDF +
Coal
10% RDF +
Gas
Filter
(gr/ft3)
0.02052
2.14
0.04439
7.60
0.02445
94.04
>10
(gr/ft3)
0.53895
56.28
0.26900
46.07
0.00095
3.65
>3 T
(gr/ft3)
0.34355
35.87
0.22627
38.75
0.00036
1.38
>1 7
(gr/ft3)
0.05465
5.71
0.04422
7.57
0.00023
0.88
Total _
(gr/ft3)
0.97769

0.58389
0.02600

-------
 100
  80
  60

  40


  20
310
   8
   6
 u
•f-
 o>
 O
 I-
 ID
1.0
 0.8
 0.6

 0.4


 0.2

 0.1
                  20%  EA
             0  Test 38
             ^  Test 31
             &  Test 35
             0  Test 40
                        '
% RDF
  10
  20
  30
   0
                                  ' _ «  I  '   l  ' — L
     0.01   0.1  0.5   2      10    30   50   70     90     98
                      Cumulative percent less than diameter
                                                               99.8  99.99
         Figure 4-28.  Stack gas particle size vs. cumulative percent
                       less than diameter.
                                       137

-------
00
                                                               _ 16.0
                                                               _ 12.0 T
                           10          20           30
                    Refuse concentration (% heat input  basis)
                                                                       in
                                                                      _o
                                                                       c
                                                                       CD
                                                                  8.0  g
                                                                       o
                                                                       c/l
                                                                       fO
                                                               -  4.0
                                                                       0)
                                                                       3
TANGENTIAL MODE
1.50 x 106 BTU/hr
300 F Secondary Air
2Q% Excess Air
Richmond Refuse/
Pittsburgh Coal
                                                                                             Particulate Loading
                                                                                             Ash Content
                                      Figure 4-29.  Particulate loading results.

-------
indicates a rather strange effect,  and  that is  that  the  grain  loadings
decrease with increasing percent RDF with a minimum  at 20 percent RDF.
Possible explanations for this include  a greater hold up in the convective
section, or more material reaching the  ashpit or sticking to the walls  of
the furnace.  It is possible that resultant ash properties or heat trans-
fer conditions are changing such that more material  is deposited either
in the furnace or on the convective tubes.  However, the duration of each
of these tests was not sufficiently long to determine if this hypothesis
is true.   In addition, due to the refractory walls and dissimilar convec-
tive tubes  compared to a full-scale boiler, it is rather speculative to
say a  similar effect would occur in the  full-scale systems.
4.5.3.5  Percent Combustibles in Flyash
       Table 4-12  lists  the results of  the analysis on percent  combustibles
as a function of percent excess air and  percent refuse when cofired with
coal.  As  was the  case  for CO and  unburned hydrocarbons,  these  results
indicate that the  combustion efficiency is quite  good in  the  pilot-scale
facility when cofiring  RDF with coal as long as the  excess  air is above
10 percent.  There is an indication  that even  this  facility  does  not oper-
ate quite  as efficiently with the  refuse as  with  coal alone.   This  certainly
has been the case  in  full-scale units  where  considerable unburned material
has found  its way  to  the ashpit.   The  reference to  the  plugging of  the  ash-
 pit in Section  4.2.5  is another indication of  unique problems with  the  RDF
materials.
        However, it is possible  that the additional  shredding and/or the
 hot  refractory  walls aid in ignition and achieving  complete combustion in
                                     139

-------
TABLE 4-12.  COMBUSTIBLE CONTENT IN FLYASH
l/l
•r-
in
TO
CO
+J
3
Q.
C
»— t
4->
ro
OJ
31
Z
0
1-^
«t
a:
I—
z
UJ
o
•ST.
O
o
U-
D
cc:
*s
o
^s
0
&s
o
00
frS
0
ro
EXCESS AIR
10% 20% 30?;
0.43%
9.56% 1.37%
8.15% 1.35%
1.17% 3.46%
Tangential mode - 1.5 x 106 Btu/hr
Richmond refuse/Pitts #8 coal
300°F sec air
                      140

-------
the pilot-scale facility.   Perhaps  boilers  designed  specifically  to burn
RDF cofired with coal  will  require  a  hotter radiative  section.  Of course,
the resulting ashing problems associated with the coal would  have to  be
taken into consideration.
4.5.4  Trace Metals
       Concentrations of 11  trace metals were determined in the solid  par-
ticulate and condensible vapors collected in the SASS impingers.
       A summary of the total concentrations found on a yg/Btu basis  is
shown in Table 4-13 for all the coal  plus RDF tests  where the SASS train was
used.  A comparison is also made with the gas test using the same Richmond
RDF.  Although few conclusions can be drawn with regard to this  limited
sample, the following commments are in order:
       •   With a few exceptions,  the order of magnitude of each of  the
           trace elements  does not vary greatly  from  test to test.
       •   Exceptions to this comment include the following:
               Cu     Tests  #37 and #40
               Zn     Test #32
                Pb     Test #40
                Sn     Test #llb
               As     Test #32
       t   There appears to  be  no  clear trend on any  of  the  elements with
            regard  either to  percent  RDF or percent  excess air.
       •    There does not  appear to  be  much difference in the  total  trace
            metal concentrations  when  firing gas plus  RDF or  coal plus RDF
            (six approximately the  same, one higher, and four lower).
 This last comment  leaves  the validity of these  measurements  somewhat in
 question  as it would have  been expected that the trace metal concentrations
 when cofiring with natural gas would be considerably lower.
                                       141

-------
                          TABLE 4-13.  TOTAL TRACE METAL LOADINGS (yg/Btu) - COAL COFIRING
                 Fuel
                 % EA
                 % RDF
Coal
 10
 10
Coal
 20
 10
Coal
 10
 20
Coal
 20
 20
Coal
 30
 30
Coal
 20
  0
Gas
20
10
ro
^"v^est No.
Elemeru^^^
Cu
Zn
Mn
Pb
Cd
Be
Ti
Sb
Sn
Hg
As
#37
<1.660
1.978
<0.207
0.697
<0.0081
0.004
<1.125
<0.0047
<0.1101
0.0274
<0.0389
#38


<0.2693




<0.0090
<0.0913
<0.0173
<0.0323
#32
<0.309
5.327
<0.520
1.319
0.0126
<0.0009
<1.2525
<0.0386
0.1060
<0.0293
1 . 3839
#31
<0.184
1.504
<0.3549
<2.591
<0.013
<0.0108
<0.4379
<0.0809
0.1508
<0.0058
<0.0408
#34
0.206
2.025
<0.3512
2.865
0.0037
0.0025
<1.382
<0.010
O.1481
0.0187
<0.0402
#40
1.958
0.686
0.153
17.53
0.009
<0.0176
<1.7540
<0.024
0.130
<0.0015
<0.088
#11 B
0.340
1.821
0.0209
1.500
0.0062
<0.0034
<0.0277
0.0333
<3.503
<0.0009
<0.0184

-------
       However, the nonhomogeneity of the material  must be considered as
well as the influence of the test furnace.  First,  it is possible that
large concentrations of a particular trace metal can be present locally
in the feed and find their way to the stack sampling equipment.  Holdup
of material in the heat exchange sections of the furnace system can also
result in momentary high particulate concentrations if the material breaks
loose from the heat exchange surfaces in  large discrete clumps.  Finally,
metals in the  furnace from the burners (particularly copper, lead, and Zn
from cooling coils, silver solder and brazing compounds) may also  find
their way to the stack.  Due to  these factors,  it will probably  require a
large data base at  any  given test condition to  obtain  a statistically mean-
ingful result.
       Table 4-14 lists the percentages of the  total trace metals  found as
condensible  vapors  collected  in  the  organic module  and impinger  sections
of  the SASS.   Again,  no clear  trends are  present although Hg,  Cu,  Mn,  and
Sn  generally  had high percentages  in the  vapor  phase.   Cd was  usually split
between  the  vapor  and solid and  As  was  almost always  found with  the  par-
ticulate.  The remaining  trace metals had widely varying  concentrations
of  the condensible  material.
       Table 4-15  presents the total trace metal  concentration for the four
 RDF materials when cofired with natural gas.   Again, there appears to be
 wide variations between the different RDF types.   Similarly, Table 4-16
 presents the percent vapor for each of these  materials.   Hg always appears
 in  the vapor and As in the particulate.  8e,  Cu,  Sn and Mn were also usu-
 ally found in the vapor.   Again, the heterogeneous nature of these materials
 must be  considered.
                                       143

-------
TABLE 4-14.  TRACE METAL CONCENTRATIONS (%) - COAL COFIRING
Fuel
% EA
% RDF
Coal
10
10
Coal
20
10
Coal
10
20
Coal
20
20
Coal
30
30
Coal
20
0
Gas
20
10
^\[est No.
Element^v^^
Cu
Zn
Mn
Pb
Cd
Be
Ti
Sb
Sn
Hg
As
#37
99.9
81.1
90.6
57.4
43.2
90.0
1.1
17.0
74.8
31.0
3.9
#38


91.7




1.1
63.9
91.3
10.2
#32
79.1
28.2
89.5
28.1
51.6
11.1
0.9
76.9
68.5
97.6
0.3
#31
45.4
44.9
92.3
3.8
2.3
41.7
30.0
61.8
49.5
100
3.7
#34
73.1
68.2
93.2
84.5
49.9
23.0
2.0
28.7
87.6
81.3
9.3
#40
96.2
77.2
39.4
98
48.4
73.9
25.1
3.3
65
80
18.2
#11B
64.7
24.5
77
40.6
43.5
82
64.1
7.2
99.6
100
9.2

-------
TABLE 4-15.  TOTAL TRACE METAL LOADINGS (yg/Btu)  - GAS COFIRING
Fuel
%EA
% RDF
Gas
20
10
Gas
20
10
Gas
20
10
Gas
20
10
^^^v^est No.
Elementr^v.
Cu
Zn
Mn
Pb
Cd
Be
Ti
Sb
Sn
Hg
As
11A
Ames
0.653
2.972
o.om
4.38
0.0119
0.0008
0.088
0.118
0.090
0.032
0.020
11B
Richmond
0.340
1.821
0.021
1.50
0.006
0.003
0.028
0.033
3.50
0.001
0.018
nc
Americology


0.039




0.222
0.102
0.005
0.0095
11D
San Diego
4.11
13.58
0.238
42.08
0.0613
0.00027
0.281
0.0988
0.722
0.035
0.319
                                 145

-------
TABLE 4-16.   TRACE METAL CONCENTRATIONS  AS  VAPOR  (%) - COAL COFIRING
   Fuel
   % EA
   % RDF
Gas
 20
 10
Gas
 20
 10
Gas
 20
 10
Gas
 20
 10
""^^Jest No.
Elemeiu^v^^
Cu
Zn
Mn
Pb
Cd
Be
Ti
Sb
Sn
Hg
As
11A
Ames
80.2
21.7
78.4
7.5
5.9
94.9
11.3
0.6
68.2
99.9
11.1
11B
Richmond
64.7
24.5
77
40.6
43.5
82
64.1
7.2
99.6
100
9.2
nc
Americology


48




1.2
89
98.8
30.2
nc
San Diego
86.8
8.7
92.3
10.7
98.2
94.8
6.4
39.7
15.7
99.9
55.8
                                  146

-------
       A comparison was also made between  the trace metal  concentrations  in
the participate flyash found in these tests  and data found in  the  literature
for both coal  and coal plus various RDF.   These results  are shown  in  Table
4-17 for each of the test conditions and for three sets of field data.   The
field data is  from the St.  Louis demonstration (Reference  12), Wright Patter-
son Air Force  Base (Reference 13), and the Ames-Iowa facility  (Reference  14).
The data is presented in yg/grain of flyash for the solid  particulate only.
Again, wide variations in both the data developed on this  program as  well
as the field data are seen.  It should also be mentioned that the field
data are average numbers and that there was considerable variation even
from one site.  From the field data, it appears the results generated here
are within the same order of magnitude.  However, trends as a function of
either excess air or percent RDF still cannot be discerned.
       Finally, two sets of particulate data were analyzed for each  of the
trace metals  in each of the cyclone  size cuts.  This was done for the coal
only test  (#40) and for coal plus 20 percent RDF at 20 percent excess air,
Test 31.   Figures 4-30 through 4-40  show the charts of cumulative percent
versus size cut for each trace metal.  As before,  this is  plotted as the
cumulative percent below and including a given  size.  The  first cyclone
catches all material  >10 y, and  the  filter  catches  everything  less than  1 y.
Seven out  of  the  11 elements indicate  that  the  presence of RDF results in
a  higher  percentage in the  smaller  size cuts.   Trace metals which have a
reverse trend include As,  Be,  Mn,  and  Zn where  the coal only  has  high  con-
centrations of these  elements  in the finer  sizes.   However,  in  light of  the
randomness of much  of the  other trace  metal  data,  caution should  be  exercised
 in drawing any definitive  conclusions  from these curves.
                                       147

-------
              TABLE 4-17.   TRACE METAL CONCENTRATIONS  (yg/g OF  FLYASH)  -PILOT  VS.  FULL  SCALE  (PARTICULATE  ONLY)
          Fuel      Coal
          % EA       10
          % RDF      10
Coal
 20
 10
Coal
 10
 20
Coal
 20
 20
Coal
 30
 30
Coal
 20
  0
Coal

  7
Coal

  0
Coal

 34
Coal

 51
Coal

  0
Coal

 20
Coal

 50
Coal

  0
Test
Element
Cu
Zn
Mn
Pb
Cd
Be
Ti
Sb
Sn
Hg
As
#37
1,508
506.8
26
402
6.2
0.6
1,508
5.31
<38
25.5
50.6
#38
--
—
76.5
--
--
--
—
30.3
113
5.1
99
#32
—
4,799
69
1,190
7.6
1.0
1,558
11.1
<42
8.5
55.9
#31
4,611
28,927
619
118,475
515.4
83.8
12,188
1415.7
4,042
1.4
1206.6
#34
126
1,465
54
1,010
4.1
4.3
3,081
16.2
4.2
8.0
82.9
#40
1,304
2,516
600
5,519
42.9
50.0
4,994
29.8
596
1.1
725.7
St. Louis
Ref 12
430
2,534
--
1,681
44
24.3
12,050
17.3
--
13.04
62
St. Louis
Ref 12
236
1,102
--
598
35
8.98
2,584
1.82
—
6.42
189
WPAFB
Ref 13
--
11,433
--
9,880
--
--
--
—
--
—
—
WPAFB
Ref 1"+
--
27,563
—
21,290
--
—
--
—
--
--
- —
WPAFB
Ref 13
--
902
--
493
—
—
--
—
—
--
—
Ames
Ref l1*
472
29,211
414
31 ,684
--
--
3,196
--
--
--
—
Ames
Ref !"«
379
25,211
360
22,815
--
--
235,710
—
--
--
~~
Ames
Ref Vt
153
8,373
628
6,733
--
—
3,625
--
--
--
	
oo

-------
TOO
 80
 60

 40   h
  20
  10
   8
Ol
fM
0)
(_>
   1.0
   0.8
   0.6
   0.4
   0.2
   0.1
- -  20::  RDF  +  Coal
     Coal
                                             TRACE METAL  - Cu
          Ml   II  I	1	1	i  i   I  I  i   i   i
                                       J	L
        I  I
        JL_L
     0.01  0.1  0.5   2
                            10
                      30   50   70
90
98
S9.fi   99.99
                      Cumulative percent less than diameter
          Figure 4-30.   Stack gas particle size vs.  cumulative  percent
                        less than diameter -trace metal  Cu.
                                         US

-------
 £
OJ
tM
to
o>
O
4->
s_
no
Q.
100
 80
 60
 40

 20

 10
  8
  6
  4
1
0.8
0.6

0.4
     0.2  -
                    20% RDF + Coal
                    Coal
                                        TRACE  METAL  - ZN
     o.i
      0.01   0.1  0.5
                                  i   I  i  I  i   l   i
                                                j	i    i  i
                  2     10     30   50   70     90      98
                   Cumulative  percent  less  than  diameter
                                                               99.8   99.99
         Figure  4-31.   Stack  gas  particle  size  vs.  cumulative  percent
                       less than  diameter  -trace metal  Zn.
                                    150

-------
100

 80   -


 60   -



 40   -
   20
   10

    8
c
o
t.
o
o;
ISI
Ol

o
    0.8


    0.6



    0.4
     0.
     0.
20% RDF + Coal


Coal
                                              TRACE METAL -  MN
         -L-LJ—'  i   '    '   '    i   I  I   i  i  I
                                                      '   '    '  '
                                                                    11
      0.01   0.1  0.5   2
                           10
                30   50   70
90
98
99.9  99.99
                        Cumulative  percent  less than diameter
          Figure 4-32.   Stack gas particle  size  vs.  cumulative  percent

                        less than diameter  — trace metal Mn.
                                       151

-------
to
c
o
S-
u
 QJ
 NJ
OJ
O
4-J
s-
o.
100
 80
 60

 40
   20


   10
    8
    6
1.0
0.8
0.6

0.4
    0.2 -
20% RDF
Coal
                              coal
                                          TRACE METAL - PB
    0.1
        1 1  1
                                       1   1  1
                                                         II
      0.01   0.1  0.5   2      10     30   50   70     90     98
                       Cumulative percent less than  diameter
                                                               99.8  99.99
            Figure 4-33.   Stack gas  particle size vs.  cumulative percent
                          less than  diameter — trace metal  Pb.
                                       152

-------
to
c
o
o

-------
  100
   80
   60

   40


   20
^ 10
~  8
    6
to
c
o
s-
u
• ^
E
c
o
re
O-
    1.0
    0.8
    0.6

    0.4
    0.2
    0.1
                 20%  RDF  +  coal
                 Coal
TRACE METAL - BE
     0.01  0.1   0.5  2      10     30   50   70     90      98
                     Cumulative percent less than diameter
                                                                 99.8   99.99
        Figure 4-35.  Stack gas particle size vs.  cumulative  percent
                      less than diameter - trace metal  Be.
                                      154

-------
   100
   80
   60
   40

   20

3: 10
0)
N
I/)
s-
re
    1.0
    0.8
    0.6
    0.4

    0.2
    0.1
- - 20% RDF + coal
	Coal
                                        TRACE  METAL  - Ti
      0.01   0.1  0.5   2     10     30   50   70     90      98
                      Cumulative percent less than diameter
                                                                  99.8   99.99
         Figure  4-36.   Stack gas particle size vs. cumulative percent
                       less than diameter — trace metal Ti.
                                      155

-------
  100
   80
   60

   40


   20
— 10
~  8
 co
 §  6

1  4
 
-------
  100

   80
   60       — Coal


   40
   20
~ 10


    8
 o  c
 j-  b


'i
 0)
 N




 1  2

 



 £  l.C


    0.



    0.6




    0.








    0.








    0.
           - - 20% RDF + coal              TRACE  METAL  - SN
      0.01   0.1  0.5  2      10     30   50  70     90      98      99988  99.99

                     Cumulative percent less than diameter



         Figure 4-38.  Stack gas particle size vs. cumulative percent

                       less than diameter - trace metal Sn.
                                       157

-------
  100
   80
   60

   40


   20
3  10
    8
 o
 N
 (/)
 O)
 u
 J-
 1C
a.
    1.0
    0.8
    0.6
    0.4
   0.2
    0.1
- - 20% RDF + coal
	Coal
TRACE METAL - HG
      0.01   0.1  0.5   2      10    30   50   70     90     98
                       Cumulative percent  less than diameter
                                                      99.8  99.99
        Figure 4-39.  Stack gas particle size vs. cumulative percent less
                      than diameter - trace metal Hg.
                                     158

-------
  100
  80
  60

  40


  20


   10
^  8
CO   r
C   0
O.
rs,
•^   2
QJ
U
to
Q.
1.0
0.8
0.6

0.4
    0.2
    0.1
                   20% RDF + Coal
                   Coal
           iii
                                    TRACE  METAL  - As
     0.01   0.1  0.5   2
                         10
30  50   70
90
                                                           98
                      Cumulative percent less than diameter
99.8   99.99
         Figure 4-40.  Stack gas particle size vs.  cumulative percent
                       less than diameter - trace metal  As.
                                        159

-------
       In summary, it appears that very little can be said about this  data
with regard to either the levels or trends of trace metals when cofiring
RDF with either coal or natural gas.   For future tests, it is recommended
that at least five samples be collected at any given test condition in order
to adequately determine the concentrations.   In addition, background tests
on gas only also need to be taken so that metals coming off the furnace
can be taken into account.
4.5.5  Organics
       As was mentioned in Section 4.4, the organic modules of the SASS
train were analyzed by GC/MS for organic compounds.  Tests 31, 32, 34, 37,
and 40 contained no detectable organic compounds.  Samples from Tests 38,
11A, 11B, and 11C contained polynuclear aromatic hydrocarbons and deriva-
tives in the amounts indicated in Table 4-18.  No PCBs were detected in
any samples.
       Two other compounds were detected in the RDF Test 11B sample.  The
mass spectra of these components were indicative of silicon containing com-
pounds.   They could not, however, be positively identified.  The spectra
of these compounds as well as the total ion current traces for the analyses
are available if needed.
       A final point involves the presence of medium weight polynuclear
aromatic hydrocarbons in these stack samples.  A large volume of literature
indicates that combustion of hydrocarbon fuels gives rise to polynuclear
aromatic hydrocarbons and also to highly polymerized species which are
collectively known as "soot."  The latter species are not readily analyzed,
but the lower homologues are analyzed as the polynuclear aromatics.  In
these samples, the medium weight species such as pyrene, fluoranthene and
                                      160

-------
                TABLE 4-18.   ORGANICS FOUND
 Test Condition
      Organic
     Amount
Gas Cofire
10% RDF
20% EA
Ames Fuel

Gas Cofire
10% RDF
20% EA
Richmond  Fuel
 Gas  Cofire
 10%  RDF
 20%  EA
 Americology  Fuel

 Coal Cofire
 10%  RDF
 20%  EA
fluoranthene
pyrene
phenanthrene
fluoranthene
pyrene
diphenyl ether
biphenyl phenyl ether

phenanthrene
pyrene
 phenanthrene
0.0000102 pg/Btu
0.0003325 ug/Btu
0.0000641 yg/Btu
0.0001601 vg/Btu
0.0005765 yg/Btu
0.003395  yg/Btu
0.001697  yg/Btu

0.0000593   yg/Btu
0.0010369   yg/Btu
 0.0000981   yg/Btu
                                161

-------
 phenanthrene  normally dominate with the higher molecular weight species
 (such as benzo(a) pyrene present also, but at concentrations lower by a
 factor of  10  to 100.  If such were true with the RDF samples, then these
 carcinogenic  compounds would be present, but at concentrations below the
 detection  limit for these analyses.
       In addition, it should be remembered that only two of the LC frac-
tions were analyzed (LC  2,3).  Tables  4-19 and 4-20 show the quantity and
percent of the material  found in all  of the LC fractions for each  of the
tests where a SASS analysis was made.   As  can be seen from this  table, con-
siderable material was found in Fraction LC 1, 6 and 7  in many of  the tests
although these fractions were not analyzed.  Table 4-21 gives a representa-
tive listing of the possible compounds that could make up each of these
fractions and the MEG concentration limit.  Thus, if the material  in these
fractions were made up of any one of these compounds, it could exceed the
MEG criteria.   For this  reason alone,  further analysis on these samples
is warranted.
                                    162

-------
                                             TABLE 4-19.   LC COLUMN DATA
CO
Test
No.
HA
11B
TIC
31
32
34
37
38
40
mg/m3
Ll
0.10728
0.49074
0.16019
0.12811
0.91087
0.49676
0. 38268
1.12582
0.21218
L2
0.01314
0.05708
0.03620
0.00217
0.04315
0
0.00656
0.02124
0.05378
L3
0.08393
0.24537
0.08688
0.03908
0.12758
0
0.02697
0.07379
0.06410
L4
0.09414
0.15677
0.05249
0.03474
0.15572
0.04909
0.07070
0.13416
0.08767

L5
0.08612
0.26326
0.10951
0.03908
0.15197
0.07034
0.09767
0.12522
0.14146
Lfi
0.05692
1.36317
0.05068
0.01954
0.17448
0.02125
0.04155
0.09503
0.90547

L7
1.14943
0.84943
1.11047
0.38144
1.77577
1.20453
1.25519
2.46853
3.79500

-------
                                             TABLE 4-20.  LC COLUMN  DATA
CT>
-p.
Test
No.
11A
11B
11C
31
32
34
37
38
40
LI/LT
%
6.74
14.32
9.97
19.89
27.28
26.97
20.34
27.84
4.03
4/4
%
0.83
1.67
2.25
0.34
1.29
0
0.35
0.53
1.02
L3/LT
%
5.27
7.16
5.41
6.07
3.82
0
1.43
1.82
1.22
VLT
%
5.92
4.58
3.27
5.39
4.66
2.67
3.76
3.32
1.67
L5/LT
%
5.41
7.68
6.82
6.07
4.55
3.82
5.19
3.10
2.69
VLT
%
3.58
39.79
3.15
3.03
5.22
1.15
2.21
2.35
17.22
L?/LT
%
72.25
24.79
69.13
59.21
53.17
65.39
66.72
61.05
72.15

-------
TABLE 4-21.  POSSIBLE COMPOUNDS IN LC FRACTIONS NOT ANALYZED
Test
No.
38







40












ne













Sample
Fraction
LCI
LC7






LCI
LC6




LC7






LCI
LC6






LC7





Concentration
(pg/m3)
1125.82
2468.53






212.18
905.47




3795.00






490.74
1363.17






849.32





Sample Fraction
Tetraethyllead
2 ,4 ,6-Tri ni trophenol
4,6-Dinitro-O-Cresol
4,4'-Methylene-Bis-(2-Ch1oroaniline)
Penthachlorophenol
1 -Ami nonaphtha 1 ene
Dinitro-P-Cresol
Dini trophenol s
Tetraethyllead
2- Ami nonaphtha 1 ene
Dibenz (A,H) Acridine
Dibenz (A,J) Acridine
Anisidines
Perchloromethanethiol
2,4,6-Trinitrophenol
4,6-Dinitro-O-Cresol
4,4'-Methylene-Bis-(2-Chloroaniline)
Penthachl orophenol
1-Aminonaphthalene
Dinitro-P-Cresol
Din1 trophenol s
Tetraethyllead
2-Aminonaphthalene
Dibenz (A,H) Acridine
Dibenz (A,J) Acridine
Anisidines
Perchloromethanethiol
Dibenzo (C,D) Carbazole
Methyl arolne
2,4,6-Trinitrophenol
4,6-Dinitro-O-Cresol
4,4' -Methylene-Bis-(2-Chloroani line)
Penthachl orophenol
1 -Ami nonaph thai ene
Dinltro-P-Cresol
Concentration
Limit
(ug/m3)
100.0
100.0
200.0
220.0
500.0
560.0
680.0
1400.0
100.0
170.0
220.0
250.0
500.00
800.0
100.0
200.0
220.0
500.0
560.0
680.0
1400.0
100.0
170.0
220.0
250.0
500.0
800.0
1000.0
1200.0
100.0
200.0
220.0
500.0
560.0
680.0
                             165

-------
                                 REFERENCES
  1.  Brown R. A., Kelly, J. T., Neubauer, Peter, "Pilot Scale Evaluation  of
     NOX Combustion Control for Pulverized Coal, Phase II Final  Report."
     EPA 600/7-79-132, June 1979.

  2.  Wendt, J. 0. L., Lee, S. W., Pershing D. W., "Pollutant Control  Through
     Staged Combustion of Pulverized Coal.  Phase I -- Comprehensive  Report.
     U.S. Dept. of Energy, Fe-1817-4, February 1978.

  3.  Johnson, S. A., Cioffi, P. L., McElroy, M.  W., "Development of an Ad-
     vanced Combustion System to Minimize NOX Emissions from Coal-Fired
     Boilers."  Presented to 1978 Joint Power Conference, Dallas,  Texas,
     September 11, 1978.

  4.  Demeter, J. J., et al., "Combustion of Coal-Oil  Slurry in a 100-HP
     Firetube Boiler,"  PERC/R1-77/8, Pittsburgh Energy Research Center,
     Pittsburgh, Pennsylvania, May 1977, pp. 3-8.

  5.  BeeV, J. M., Combustion Aerodynamics, John  Wiley and Sons,  New York,
     N.Y., 1972.

  6.  Thompson, R. E., et al., "Effectiveness of  Gas Recirculation  and Staged
     Combustion of Reducing NOX on a 560-MW Coal-Fired Boiler,"  EPRI  FP-257,
     September 1976.

  7.  Heap, M. P., et al., "The Optimization of Burner Design Parameters  to
     Control  NOX Formation in Pulverized Coal and Heavy Oil  Flames,"  Pro-
     ceedings of the Stationary Source Combustion Sumposium, Volume I, EPA-
     600/2-76-1526,  June 1976.

 8.  England, G. C., et al., "The Control  of Pollutant Formation in Fuel Oil
     Flames -- The Influence of Oil  Properties and  Spray Characteristics,"
     Proceedings of  the Third Stationary Source  Combustion Symposium;  Volume
     II.  Advanced Processes and Special  Topics,  EPA-600/7-79-0506, February
     1979, pp. 41-71.

 9.  Brown, R. A.,  "Pilot Scale Investigation of  Combustion  Modification
     Techniques for  NOX  Control  in Industrial and Utility Boilers," EPA-
     600/2-76-1526,  Proceedings of the Stationary Source Combustion Sympo-
     sium, Volume II,  June 1976.

10.  Wendt, J. 0.  L.  and Ekmann, J.  M.,  "Effect  of  Sulfur on NOX -- Emissions
     from Premixed Flames," EPA-600/2-75-075, October 1975.

11.  Wendt, J. 0.  L.,  et al.,  "Interactions Between Sulfur Oxides  and  Nitro-
     gen Oxides in Combustion Processes,"  Proceedings of the Second Stationary
     Source Combustion Symposium,  Vol.  IV, EPA-600/7-77-073d,  July 1977.
                                    166

-------
12.   Gorman,  et al.,  St.  Louis  Demonstration  Project  Final Report:  "Power
     Plant Equipment, Facilities  and  Environmental  Evaluations," EPA Con-
     tract 68-02-1871, Prepared for U.S.  Environmental  Protection Agency,
     Washington, D.C., by Midwest Research  Institute, Kansas City,  Missouri,
     July 1977, pp.  402.

13.   Jackson, J. W.,  "A Bioenvironmental  Study of Emissions from Refuse
     Derived Fuels,"  USAF Environmental  Health Laboratory, McClellan,  Cali-
     fornia, January 1976, pp.  113.

14.   Hall, 0. L., et al., "Evaluation of the  Ames Solid Waste  Resources  --
     An Energy Recovery System, Part  III -- Environmental  Evaluation  of  the
     Stoker-Fired Steam Generators at the City of Ames, Iowa,  Prepared for
     U.S. Environmental Protection Agency,  Cincinnati, Ohio,  and  Energy
     Research and Development Administration by Iowa State University, Mid-
     west Research Institute, and Ames Laboratory, April 1977, pp.  133.
                                      167

-------
            APPENDIX

DATA SUMMARY - DISTRIBUTED AIR
DATA SUMMARY - COAL/OIL MIXTURE
DATA SUMMARY - RDF TESTING
COM/DOE REPORT
                169

-------
TABLE A-l.   DATA SUMMARY - DISTRIBUTED AIR
Test
No.
209a
b
c
d
e
f
g
h
i
j
k
i
21 Oa
b
c
d
e
f
g
h
211a
b
c
d
e
f
g
h
1
j
k
n
212a
b
c
d
e
f
g
h
i

Fuel
1

















































































SR
0.80
0.80
0.80
0.80
0.80
0.80
0.80
0.95
0.95
0.80
0.80
0.80
0.95
0.95
0.95
0.80
0.80
0.80
0.80
0.95
0.80
0.80
0.80
0.80
0.80
0.80
0.95
0.95
0.95
0.80
0.80
0.80
0.80
0.80
0.80
0.95
0.95
0.95
0.95
0.95
0.95

EA
%
15

















































































Load
xlO Btu/hr
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
0.85
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.3
1.3
1.3
1.3
1.3
1.3
1.3
1.3
1.3
0.85
0.85
0.85
0.85
0.85
0.85
1.3
1.3
1.3

Preheat
sec °F
75
78
79
79
79
79
79
80
600
580
580
550
600
600
600
600
600
600
600
600
600
600
600
600
600
600
580
600
550
600
580
600
500
550
550
600
575
575
635
625
650

Stg °F
75
81
82
82
82
82
82
90
-_
227
269
276
350
350
350
296
296
296
322
327
323
371
381
337
336
335
288
242
230
368
422
443
256
379
409
359
334
326
330
312
305

Burners
4 IFRF
j















































































SW/Int
or
Yaw
4














































































i
i

Prim.
Stoich.
I
12















































































Stg Air
Mixing,'
Location
Hor b
Hor b
Hor b
Hor a
Hor a
Hor a
Hor a
Hor a
Hor a
Hor a
Hor a
Hor a
HE-J
HE-J
HE-J
HE-J
HE-J
HE-J
HE-J
HE-J
HE-K
HE-K
HE-K
HE-K
HE-K
HE-K
HE-K
HE-K
HE-K
HE-H
HE-H
HE-H
HE-H
HE-H
HE-H
HE-H
HE-H
HE-H
HE-H
HE-H
HE--H

Temperature
T23°F
1910
2024
2048
2077
2104
2135
2149
2165
2227
2234
2185
2176
2146
2034
2108
--
--
--
—
--
2362
2314
2394
2366
2374
2282
2261
2350
2325
2306
2256
2364
1877
1767
2429
2339
2621
2482
2621
2603
2577

T24"F
1957
2010
2023
2014
2023
2059
2088
2088
2088
2103
2133
2128
2132
2562
2574
2480
2508
2557
2488
2497
2540
2529
2442
2338
2368
2473
2558
2548
2438
2384
2615
2582
1927
1978
1981
1966
2003
2036
2203
2266
2268

NOC
ppm
300
310
377
299
290
257
251
424
480
262
266
250
376
451
284
133
152
163
170
258
215
260
197
185
167
224
305
252
335
289
264
229
190
219
178
222
253
296
346
289
260

Comments
SRla
0.45
0.30
0.60
0.60
0.45
0.30
0.30
0.45
0.45
0.30
0.45
0.60
0.45
0.30
0.60
0.60
0.45
0.30
0.80
0.95
0.45
0.30
0.60
0.45
0.60
0.30
0.45
0.60
0.30
0.30
0.45
0.60
0.45
0.30
0.60
0.60
0.45
0.30
0.30
0.45
0.60
1

-------
TABLE A-l.  DATA SUMMARY -DISTRIBUTED AIR (CONCLUDED)
Test
No.
212j
k
t
21 3a
b
c
d
e
f
9
h
i
j
k
i
m
n
0
P
q
r
s
t
u
V
w
X
21 4a
b
c
d
e
f
9
h
Fuel
1




































































SR
0.95
0.95
0.95
0.80
0.80
0.80
0.95
0.95
0.95
0.95
0.95
0.95
0.80
0.80
0.80
0.80
0.80
0.80
0.95
0.95
0.95
0.80
0.80
0.80
0.95
0.95
0.95
0.95
0.95
0.80
0.80
0.80
0.80
0.95
0.95
EA
%
15


































Load
« I^Btu/hr
1.7
1.7
1.7
0.85
0.85
0.85
0.85
0.85
0.85
1.3
1.3
.3
.3
.3
.3
.3
.3
.3
.3
.3
.3
.7
.7
.7
.7
.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
Preheat
sec °F
675
675
675
100
98
90
90
90
90
575
600
600
600
600
600
620
600
600
600
—
—
—
—
—
600
600
600
625
625
620
580
600
620
620
670
Stg °F
263
269
270
98
—
98
103
104
104
196
250
275
282
276
277
302
300
—
—
—
—
—
—
—
366
382
385
264
244
294
313
100
335
310
314
Burners
4 IFRF




































































SW/Int
or
Yaw
4




































































Prim.
Stoich.
%
12




































































Stg Air
Mixing/
Location
HE-K
HE-K
HE-K
HE-E
HE-E
HE-E
HE-E
HE-E
HE-E
HE-E
HE-E
HE-E
HE-E
HE-E
HE-E
HE-F
HE-F
HE-F
HE-F
HE-F
HE-F
HE-F
HE-F
HE-F
HE-F
HE-F
HE-F
HE-M
HE-M
HE-M
HE-M
HE-N
HE-N
HE-N
HE-N
Temperature
T23 °F
2878
2759
2274
2059
2089
2137
2161
2172
2188
2075
2381
1937
2017
1864
1507
2233
1760
1890
1886
—
—
—
—
—
2507
2886
2572
2220
2691
2855
3861
—
—
—
2532
T °F
>24 >-
2182
2396
2515
1830
1880
1905
1957
1974
1987
2296
2145
2258
2379
2372
2376
2379
2433
2370
2432
2456
2427
2460
2482
2597
2592
2572
2655
1996
2085
2335
2083
2235
2566
2596
2330
NOC
ppm
379
375
480
190
185
215
303
287
284
305
300
466
324
305
281
255
244
285
348
300
312
346
311
397
368
405
381
223
230
151
163
116
122
201
257
Comments
SR1a
0.60
0.45
0.30
0.45
0.60
0.30
0.30
0.45
0.60
0.60
0.45
0.30
0.30
0.45
0.60
0.60
0.45
0.30
0.30
0.45
0.60
0.45
0.60
0.30
0.45
0.60
0.30
0.60
0.70
0.60
0.70
0.70
0.60
0.60
0.70

-------
TABLE A-2.  DATA SUMMARY - COAL/OIL MIXTURE

Test
No.
215a
b
c
d
21 6a
b
c
d
e
f
9
h
i
21 7a
b
c
d
e
f
g
h
i
j
k
i
m
n
0
21 8a
b
c
d
e
f
g
h
i
a
b

Residence
Time
S












































































Firing Rate
(Btu/hr x 106)
1.8













































































Radiant
Heat
Transfer
(Btu/hr x 106)
0.558
0.483
0.571
0.442
0.521
0.429
9.438
0.471
0.438
0.438
0.454
0.492
0.463
0.250
0.238
0.242
0.254
0.254
0.254
0.254
0.254
0.254
0.275
0.304
0.304
0.313
0.313
0.283
0.263
0.263
0.313
0.300
0.296
0.308
0.263
0.321
0.342
0.267
0.258

Excess
Air
20
30
40
20
20
30
20
40
40
30
20
20
40
30
20
40
40
30
20
20
20
20
20
20
20
20
20
20
30
40
20
30
40
20
20
20
20
20
20
i
Coal
Type
(X)
0





























30% W.Kty.





















30% Va.















30% W.Kty.


Oil
Type
Chevron














Penn




Chevron








Penn





















Chevron
















Penn



Preheat
Temp
80
82
77
76
82
83
82
83
82
82
81
83
83
300
















































•


Staged
Air
Preheat
Temp

































277
277
277
290
216
220

Nozzle
Type
DeLavan
























Sonicore



















































-------
                               TABLE  A-2.   DATA SUMMARY - COAL/OIL  MIXTURE  (CONTINUED)
CO
Test
No.
219c
d
220e
f
9
h
i
j
k
i
221a
b
c
d
e
f
g
h
i
j
k
l
m
n
0
P
q
r
222a
b
c
d
e
f
g
223a
b
c
d
Residence
Time
S




























L
S
L
S






















L
L
S






L
S
S
Firing Rate
(Btu/hr x 106)
1.8












1.7
1.7
1.8


























































Radiant
Heat
Transfer
(Btu/hr x 106)
0.258
0.258
0.304
0.304
0.267
0.267
0.271
0.263
0.263
0.263
0.288
0.292
0.292
0.292
0.292
0.292
0.292
0.292
0.321
0.338
0.338
0.267
0.317
0.279
0.275
0.275
0.275
0.267
0.267
0.313
0.304
0.304
0.296
0.300
0.275
0.292
0.292
0.292
0.292
Excess
Air
20
20
20
--
20
20
20
20
20
20
30
40
20
20
20
20
20
20
20
20
20
20
20
20
20
20
30
40
20
20
20
20
20
30
40
20
20
20
20
Coal
Type
(%)
30% W.Kty.





30% Va.










30% Mont.








30% Va.














30% Mont.
















0
0
0
30% Mont






Oil
Type
Penn




__'.
Chevron




































Penn








Chevron






Penn




Chevron






Preheat
Temp
300




_._
300




































































Staged
Air
Preheat
Temp
225
228
251
—
—
—
—
—
246
—
—
—
—
200
238
238 -
261
261
267
274
275
275
—
—
—
—
—
—
—
205
205
205
—
—
—
—
220
220
261
Nozzle
Type
Sonicore













































































-------
TABLE A-2.  DATA SUMMARY - COAL/OIL MIXTURE (CONTINUED)
Test
No.
223e
f
g
h
i
j
k
i
m
n
0
P
q
r
224a
b
c
d
e
f
9
h
i
j
k
«,
225a
b
c
d
226a
b
c
d
e
f
Residence
Time
L
S










L
S
L
S
L
S






L
S






L
S


























Firing Rate
(Btu/hr x 106)
1.8






















1.2


























1.8


















Radiant
Heat
Transfer
(Btu/hr x 106)
0.292
0.292
0.292
0.292
0.292
0.292
0.292
0.292
0.292
0.292
0.288
0.288
0.288
0.288
0.304
0.304
0.308
0.304
0.304
0.304
0.304
0.304
0.308
0.308
0.283
0.288
0.313
0.313
0.313
0.313
0.313
0.317
0.300
0.300
0.288
0.288
Excess
Air
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
15
20
30
40
20
30
40
20
20
30
Coal
Type
(%)
30% Mont.


















































100% W.Kty.






1002 Mont.






100% Va.
1
Oil
Type
Chevron


















































—
—
—
—
—
—
—
—
—
—
Preheat
Temp
300








































276
276
320
320
300
350
330
330
330
320










Staged
Air
Preheat
Temp
261
—
—
—
	
213
223
223
229
229
229
229
—
	
—
224
221
250
250
268
268
282
282
—

—
—
—
—
—
—
	
—
—
—
—
Nozzle
Type
Sonicore







































































-------
TABLE A-2.  DATA SUMMARY - COAL/OIL MIXTURE (CONTINUED)
Test
No.
21 5a
b
c
d
216a
b
c
d
e
f
g
h
i
21 7a
b
c
d
e
f
g
h
i
j
k
J.
m
n
0
218a
b
c
d
e
f
g
h
i
21 9a
b
Burner
Swirl
5
4.5




4
















6.5














































1.0


Atomized
Air
Flow
240







72
81







21









11




2
-------
                                  TABLE A-2.   DATA SUMMARY - COAL/OIL MIXTURE (CONTINUED)
CTl
Test
No.
219c
d
220e
f
9
h
i
j
k
i
221 a
b
c
d
e
f
9
h
i
j
k
i
m
n
0
P
q
r
222a
b
c
d
e
f
g
223a
b
c
d
Burner
Swirl
1.0
1.0
0.5
—
0.5




































































Atomized
Air
Flow
166
166
—
—
150
























143





1!


1





5


0
1
150



1





0


150






Atomized
Air
Pressure
22
22
26
--
22


























17






14






12


22






10




22






Fuel
Pressure
26
26
27
-_
24
27
29
29
28
26
38
38
38
38
38
38
38
38
27
27
27
27
27
27
26
26
25
25
38
38
38
38
20
20
20
38
38
38
38
Fuel
Temp
200
200
210
—
215
210
200
210
210
200
200
200
200
200
200
200
200
200
190
200
200
200
210
220
210
210
210
200
200
200
200
200
180
180
190
200
200
200
200
Stoich.
Ratio
(SR,)
0.95
0.95
—
—
0.85
0.75
0.65
0.55
0.95
—
—
—
—
0.95
0.85
0.85
0.75
0.75
0.95
0.95
0.75
0.95
—
—
—
—
—
—
—
0.95
0.95
0.65
—
—
—
—
0.85
0.85
0.65
Stoich.
Ratio
(SRla)
0.65
0.55
0.55
—
1.20
1.20
1.20
1.20
0.55
0.65
—
—
—
—
—
—
._-
—
0.85
0.75
—
0.65
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Dopant
Type




















Thiophene
—
Thiophene
—
Pyridene
Pyridene
—
—
—
—
—
—
—
—
—
—
—
--.-
~ ~ ""
Total Fuel
Nitro/Sulfur
(*)




















2.151 S
	
2.148 S
—
0.965 N
0.796 N
—
—
—
—
—
—
—
—
—
—
—
—
...

-------
TABLE A-2.  DATA SUMMARY -COAL/OIL MIXTURE (CONTINUED)
Test
No.
Z23e
f
9
h
i
j
k
t,
m
n
0
P
q
r
224a
b
c
d
e
f
9
h
i
j
k
l
225a
b
c
d
226a
b
c
d
e
f
Burner
Swirl
0.5


















































4.0


















Atomized
Air
Flow
150






















110
120
110












120








190


















Atomized
Air
Pressure
22






















12
20
12






16




18








8.4


















Fuel
Pressure
38
38
38
38
30
30
30
30
30
30
12
12
22
22
22
22
22
22
22
22
22
22
22
22
25
24
..
--
--
--
--
--
--
—
--
—
Fuel
Temp
200
200
200
200
200
200
200
200
200
200
200
200
190
190
190
190
190
180
180
180
180
180
180
180
190
180
—
—
—
—
—
—
—
—
—
— —
Stoich.
Ratio
(SRj)
0.65
—
—
—
—
0.95
0.95
0.95
0.95
0.95
0.95
0.95
—
—
—
—
0.95
0.85
—
—
0.75
0.65
—
—
—
—
—
—
—
—
—
—
—
—
—
- --
Stoich.
Ratio

...
0.85
0.75
0.65
0.55
0.55
0.65
0.65
0.75
0.75
0.85
0.85
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
--.
Dopant
Type




































Total Fuel
Nitro/Sulfur
(*)





































-------
                             TABLE A-2.  DATA SUMMARY - COAL/OIL MIXTURE  (CONTINUED)
co
Test
No.
Z15a
b
c
d
216a
b
c
d
e
f
g
h
i
217a
b
c
d
e
f
g
h
i
J
k
1
m
n
0
218a
b
c
d
e
f
g
h
i
21 9a
b

T25
	
2137
2049
1962
2099
1779
1732
—
1730
1706
1682
1790
1819
1823
1805
1816
1831
1730
1774
1899
1864
2042
2059
1834
1800
1775
1816
1865
1900
1959
1857
1856
1824
1928
2025
2061
2209
2085
1882

T26
...
1723
1811
1651
1675
1605
1529
—
1546
1552
1595
1611
1535
1595
1644
1580
1676
1637
1635
1602
1532
1626
1531
1663
1610
1609
1637
1578
1685
1639
1748
1732
1777
1758
1680
1678
1620
1584
1529

T27

788
834
775
766
820
763
—
823
793
786
787
826
823
801
846
872
815
801
776
722
770
706
907
835
839
845
793
880
886
899
930
967
862
822
825
786
784
767

°2
3.6
5.0
6.3
—
—
—
5.0
5.2
5.9
5.1
3.6
3.7
6.1
5.2
3.7
6.4
5.8
4.9
3.8
3.0
3.4
3.2
3.4
4.0
—
3.4
3.9
3.5
5.4
6.6
3.6
4.9
6.1
3.4
3.9
3.6
3.4
4.3
4.1

CO

8.5
14.5
17.5
4.7
6.8
5.8
7.6
—
—
—
41.7
	
12.6
4.8
16.0
45.6
78.0
77.1
85.0
70.0
70.2
38.2
69.5
—
62.5
63.8
91.5
125.7
119.8
123.0
191.0
271.0
139.0
142.0
95.0
91.0
70.7
70.4

co2

	
	
	
	
	
	
	
	
	
	
	
	
13.2
15.9
12.5
12.8
13.3
13.9
14.7
14.3
15.1
15.3
14.9
15.6
15.2
14.7
14.8
11.8
10.8
13.7
13.2
12.2
14.4
14.1
14.5
14.6
12.8
12.9

NO
649
831
890
833
451
510
460
434
358
345
333
409
487
622
508
683
385
427
380
319
328
324
265
372
364
331
332
204
760
812
688
732
763
627
571
510
416
468
422

SO

542
562
574
564
600
605
608
713
677
617
620
607
568
610
661
1756
1917
1931
1991
2017
2088
2098
2063
—
1963
1987
2033
691
843
548
739
786
790
780
786
790
1725
1628
UHC

2.6
0.4
0.0
	
—
—
—
	
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
1

-------
TABLE A-2.  DATA SUMMARY - COAL/OIL MIXTURE (CONTINUED)
Test
No.
21 9c
d
220e
f
9
h
i
j
k
1
221a
b
c
d
e
f
g
h
i
j
k
1
m
n
o
p
q
r
222a
b
c
d
e
f
g
223a
b
c
d
e
f
T25
2014
2025
2021
	
1871
	
1960
1898
2040
	
1799
1850
1835
1918
1996
2001
1917
__._
1887
1929
1977
1995
1883
1806
1834
1847
1786
1680

	
1947
	
1824
1780
1730
1922
1964
Z096
1917
1925
1972
T26
1603
7626
1637
	
1606
	
1560
1553
1691
	
1626
1627
1674
1602
1625
1524
1648
	
1615
1615
1620
1599
1632
1666
1695
1684
1639
1497
	

1588
	
1669
1648
1637
1668
1677
1601
1637
1531
1631
T27
791
803
831
—
831
—
806
817
832
—
885
932
909
836
832
776
835
—
804
790
796
781
834
850
862
862
855
836
	
	
807
	
867
879
915
872
815
806
790
749
823
°2
3.8
3.2
3.3
2.8
4.0
—
3.6
3.6
3.6
—
4.9
6.3
3.8
3.3
3.0
3.5
3.3
4.1
4.7
3.4
4.0
5.9
6.0
3.9
5.4
3.2
5.9
9.3
	
3.3
3.9
	
3.6
5.2
6.2
4.2
3.2
3.3
3.3
3.7
3.9
CO
64.8
60.0
47.0
	
75.0
	
73.0
66.0
83.0
	
113.4
99.9
83.5
82.3
166.5
105.3
132.2
1869.0
63.0
54.0
53.0
91.0
107.0
85.0
106.0
79.0
131.0
268.0
	
39.2
98.5
	
76.9
95.0
160.0
179.6
356.7
110.3
138.6
1147.0
125.7
co2
14.0
14.3
14.0
	
12.2
11.6
12.4
12.4
12.2
	
12.3
12.7
14.6
15.0
15.1
14.6
15.0
14.2
14.0
14.9
14.6
13.7
13.0
14.7
13.9
15.2
13.2
11.0
	
14.3
13.9
	
13.3
12.1
11.5
13.9
14.7
14.7
14.8
14.6
14.6
NO
404
346
389
—
653
623
504
413
606
—
705
782
622
668
580
233
600
228
606
581
541
563
607
397
622
603
471
—
653
533
307
226
370
399
427
630
285
566
418
280
588
SO
1554
1525
1315
	
843
	
921
913
838
	
538
643
701
776
735
841
771
830
596
732
1022
754
1336
1409
1568
1524
1480
1704
	
809
741
	
977
1240
1522
599
774
700
708
689
695
UHC










































-------
                                  TABLE A-2.   DATA SUMMARY - COAL/OIL MIXTURE (CONCLUDED)
CO
CD
Test
No.
223g
h
i
j
k
1
m
n
0
P
q
r
224a
b
c
d
e
f
9
h
i
j
k
1
225a
b
c
d
226a
b
c
d
e
f
T25
2003
2038
2082
2067
1979
1934
1955
1963
1843
1879
1807
	
1923
1842
2134
2194
2138
2087
2196
2115
2129
	
2041
1852
	
	
	
	
2216
2189
2168
2223
2277
2222
T26
1611
1661
15,82
1631
1592
1509
1540
1489
1617
1587
1574
	
1553
1549
1472
1458
1552
1551
1458
1595
1507
	
1400
1274
1664
1669
1664
	
1783
1754
1754
1806
1844
1807
T27
792
819
812
778
805
772
777
759
799
789
712
—
719
682
626
629
681
671
629
669
676
—
670
595
895
924
960
—
995
1014
1040
998
987
1010
°2
3.9
3.3
3.9
3.7
3.2
3.9
3.6
3.8
4.1
3.4
3.5
	
3.3
3.6
2.8
2.9
3.6
3.2
4.1
3.6
3.1
—
3.8
7.6
4.0
3.7
4.5
—
3.7
5.1
6.2
3.5
3.8
4.9
CO
1156.6
122.7
87.1
625.3
161.8
426.1
123.1
116.8
229.0
190.9
68.0
	
84.0
131.0
687.0
	
134.1
96.7
	
	
169.8
	
87.0
119.7
95.2
109.0
136.7
	
138.9
136.5
194.8
64.0
74.0
121.0
C02
14.4
14.9
14.4
14.1
14.7
14.1
14.2
14.0
13.7
13.9
14.5
	
14.6
14.2
14.6
13.9
14.2
14.3
	
13.8
14.3
	
14.8
11.1
15.4
15.6
14.9
	
13.4
12.7
11.9
13.6
13.3
12.6
NO
219
234
256
221
221
611
336
712
524
733
547
	
581
568
246
168
381
480
810
205
427
—
238
—
—
1092
1159
1226
1156
1199
1266
1158
1152
1237
so2
716
726
728
724
717
729
689
700
694
712
683
—
691
764
911
794
792
781
832
726
827
—
783
883
2552
2536
2553
	
1695
1479
1446
1350
1403
1298
UHC



































-------
               TABLE  A-3.  DATA SUMMARY  - RDF TESTING
Test
No.
2Z7a
b
c
d
e
f
g
h
i
228a
b
C
d
e
f
g
h
i
j
229a
b
c
d
e
f
g
230a
b
c
d
e
f
g
231a
b
232a
b
233a
234a
235a
b
236a
237a
238a
239a
240a
241a
b
c
d
242a
243a
244a
245a
b
Load
Btu/hr x TO6)
1.5
















1.0
1.5
















1.0
1.0
- 1.5








1 0
1.5


















































Excess
Air
5
10
30
X5
10
30
30
10
5
20
5
10
30
5
10
30
30
10
5
20
20
5
10
30
30
10
20
30
10
5
30
10
5
5
10
20
20
20
20
20
20
20
20
20
20
20
20
30
10
20
30
10
20
10
20
Primary
Fuel
4at. Gas
























































































Coal














	 _
Refuse*
No.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
4
4
4
4
4
4
4
2
2
2
2
2
1
1
3
4
4
3
1
2
2
2
2
2
2
2
2
2
-
	 r
Refuse
Concen.
5
5
5
10
10
10
20
20
20 .
10
5
5
5
10
10
10
20
20
20
10
10
10
10
10
20
20
10
10
10
10
20
20
20
5
5
10
20
20
20
20
20
20
10
10
10
20
20
20
30
30
30
20
10
10
—
	 T
Preheat
Temp
290
300
310
310
315
315
310
315
300
300
300
300
300
300
300
300
300
303
300
290
310
310
316
313
317
310
304
305
313
313
314
319
319
470
560
300
300
302
302
302
309
300
317
319
319
309
304
300
300
300
300
300
300
317
315
Yaw
+6








































































































Level
of Stack
Testing












\






















3
2
2
2
2
2
2
3
3
3
3
3
-
-
2
3
3
3
3
3
1  = Ames
2  = Richmond
3  = Americology
4  = San Diego
                                   181

-------
TABLE A-3.  DATA SUMMARY - RDF TESTING (CONCLUDED)
Test
No.
227a
b
c
d
e
f
9
h
i
228a
b
c
| d
e
f
9
h
i
j
229a
b
c
d
e
f
9
230a
b
c
d
e
f
g
231a
b
232a
b
233a
234a
235a
b
236a
237a
238a
239a
240a
241a
b
c
d
242a
243a
244a
245a
b
T25
2066
2107
2149
2223
2172
2229
2200
2277
2334
1977
2231
2228
2251
2251
2301
2273
2155
2282
2320
1904
2055
2245
2281
2253
2257
2296
1987
2087
2093
2293
2269
2303
2326
2157
2304
2231
2311
2320
2054
1915
	
2035
2078
1976
1982
2139
2082
	
2017
2052
2066
2289
2196
1996
3452
T26
1802
1843
1885
1927
1892
1955
1958
2011
2049
1721
1956
1956
1969
1969
2011
1999
1920
2007
2047
1663
1795
1971
2006
2002
2022
2041
1750
1850
1858
1989
2000
2015
2034
1904
2014
1919
1966
1977
1751
1653
	
1759
1801
1757
1766
1931
1898
	
1831
1817
1865
1972
1720
1861
1995
T27
2092
2162
2217
2285
2262
2333
2298
2350
2396
2102
2344
2350
2370
2374
2411
2395
2302
2368
2426
2018
2195
2334
2381
2363
2360
2386
2081
2160
2175
2368
2351
2377
2395
2340
2460
2318
2429
2386
2039
1958
	
2050
2110
1995
1928
2117
2059
	
2068
2138
2165
2175
1964
2088
2328
Emission Level
CO
(ppm)
47.8
39
66
147.
31.2
50.8
88.9
	
972.3
29.4
53.7
715
42.4
1068
34.6
45.6
66.3
65.8
350
104
102
107
114.1
154.5
160.1
123.6
0.6
4.5
5.4
6.4
19.1
32.9
10.0
145.1
94.0
24
19
21
11
12
12
5
9
5
6
—
18
18
90.4
56.6
66
90
85
112
16
C02
(ppm)
9.4
11.3
9.6
12.5
11.9
9.7
9.9
	
13.0
10.5
11.0
11.4
8.8
12.2
10.8
9.3
9.8
11.2
12.1
10.2
10.3
11.4
10.8
9.2
9.7
11.7
9.5
8.7
9.9
10.9
8.5
10.5
10.7
17.7
17.8
10.8
11.5
12.9
11.5
9.7
11.0
11.3
11.4
10.6
11.3
	
15.0
14.9
15.9
15.8
13.8
16.0
15.3
16.6
14.4
NO
(ppm)
80
81
116
85
100
143
163
117
112
90
88
104
144
108
116
193
135
110
106
82
99
102
114
144
189
147
93
150
111
98
182
129
115
348
458
123
134
117
130
122
215
161
185
188
129
410
405
456
334
383
427
289
381
348
493
SO
(ppm)


59.7
168.7
169.6
167.2
165.5
—
183.2
3.1
9.8
9.8
14.8
22
24.9
30.7
8.4
34.7
54
UHC
(ppm)










1.0
5.0
8.0
4.0
5.0
4.0
6.0
6.0
6.0
7.4
6.1
13.7
—
—
13.1
15.1
23.0
34.1
10.5
10.7
19.0
19.9
34.9
29.1
23.5
2123.5
1902
6.6
—
18.6
22.1
22.1
	
34.8
35.9
51.4
18.1
1358.6
1321.1
1321.1
1243
1487
1575.3
1508.6
1148.6
1565.9
1692.8
—
—
—
—
4.0
3.8
4.5
4.0
4.0
3.8
—
—
—
—
—
—
—
—
—
0.5
11.2
0.4
1.3
—
—
—
—
—
436
—
—
—
°2,
('/)
3.8
2.0
4.4
1.2
2.2
5.6
5.2
—
i.o-
3.5
1.4
1.8
5.2
1.1
2.1
5.0
5.4
2.1
1.1
3.8
3.7
1.3
2.0
5.1
5.1
2.1
3.9
5.3
2.3
1.1
5.2
1.8
1.1
1.2
1.9
3.5
—
3.9
3.8
3.8
—
—
3.8
3.8
3.8
3.6
3.5
3.5
2.1
3.2
5.2
1.9
4.0
2.0
3.7
                       182

-------
COAL/OIL MIXTURE (COM) SUBSCALE COMBUSTION TEST RESULTS
 CONDUCTED  IN THE EPA/ACUREX  MULTIFUEL  FURNACE  FACILITY
                             183

-------
  INTRODUCTION
        Subscale combustion tests with coal/oil mixtures as fuel were per-
  formed by Acurex in the EPA Multifuel Furnace Facility to provide design
  support for the planned full-scale COM facility at Lorillard Division,
  Loew's Theaters, Inc., Danville, Virginia.  The test objectives were as
  follows:
       •   Determination of emissions for  30 to 50 percent coal in No. 6
           oil using identical fuels as anticipated for use at the Lorillard
           demonstration site
       •   Identification of fouling, piping, and pumping problems resulting
           from fuel handling and combustion
       t   Determination of suitability of the Carbonoyl, Inc., COM additive
           planned for use in the full-scale demonstration program
       The subscale combustion tests consisted of two major activities:
       •   Fuel preparation
       t   Combustion tests
These activities are described in the following sections.

1.     FUELS AND FUEL PREPARATION
       COM fuels for combustion testing were prepared with the coal and oil
identical to those anticipated for use at the Lorillard demonstration site.
No. 6 oil which meets Lorillard specifications and is identical to that
which is presently in use was obtained from Amerada Hess Corporation.  The
high volatile bituminous coal which was determined to have the most desir-
able properties for wet grinding (from subscale wet grinding tests at Colo-
rado School of Mines Research Institute) and which will be used during demon-
stration testing was obtained from Maryland Coal and Coke Company.  Specifi-
cations and chemical analyses of the oil and coal are presented in Tables 1
and 2.
      Although a wet grinding ball mill will be used for demonstration
fuel preparation, dry grinding and subsequent mixing were used for the test
fuels.  This preparation scheme was chosen because a suitable wet grinding
system was not available.  Pulverized coal prepared by C$MRI was blended

                                    184

-------
                   TABLE 1.  NO. 6 OIL ANALYSES
Specifications3
API gravity
Sulfur (X Wt)
Flash point (PMCC °F)
Viscosity (SSF 0 122°F)
Pour point (F)
BS&W (X Vol )
15.3
2.22
204.0
247.0
+50.0
0.4
Ultimate (
Carbon
Hydrogen
Nitrogen
Oxygen
Sulfur
Ash
% Wt)a
84.71
10.75
0.36
1.93
2.22
0.03
  Supplied  by  Amerada  Hess  Corporation
                       TABLE 2.   COAL ANALYSES
              Proximate (% Wt)a
Ultimate (% Wt)b
 Moisture                      4.2
 Volatiles                    33.0
 Fixed carbon                 54.0
 Ash                           8.9
 Ash fusion temp (F)        2700.0
 Hardgrove grindability       68.0
 Btu per pound             13368.0 As Rec'd
                           13954.0 Dry
 Origin:  Clintwood seam, Conoway, Virginia
Carbon
Hydrogen
Nitrogen
Sulfur
Oxygen
79.0
 5.0
 1.5
 0.9
13.4
 Supplied  by  Maryland  Coal  and  Coke  Company
'EPA-650/2-75-046,  May 1975
                                  185

-------
with No. 6 oil and the Carbonoyl additive in a high turbulence batch mixer
supplied by Littleford Brothers.  The grind distribution of the coal was
approximately 80 percent passing 200 mesh and 100 percent passing 48 mesh.
The additive was prepared in a 50-percent aqueous solution and constituted
3.75 percent by weight of the COM independent of coal fraction.
       The blending procedure was as follows:
       1.  Place premeasured No. 6 oil in the Littleford Brothers batch
           mixer (Model FM 13100 20-gallon capacity).  The mixer is main-
           tained in the "on" position.  Mixer is steam jacketed and mix-
           ture is maintained at about 140°F.
       2.  Add premeasured additive solution to oil
       3.  Add premeasured pulverized coal to oil and additive and allow to
           mix for 10 minutes
       4.  Discharge into 55-gallon storage drum
       Fuel mixing occurred between July 11 and July 22, 1977.  Approxi-
mately 1500 total gallons of 50-, 40-, and 30-percent COM were prepared.
No unexpected difficulties arose.  Those problems which did occur were
related to handling of the fuels, particularly the pulverized coal.  None
of the handling problems, however, are related to full-scale operation.
       The COM was stored at ambient temperatures (minimum approximately
50°F) for up to 24 days before use.  About 3 hours prior to use, the
storage drum was wrapped with electrical resistance heating blankets and a
mixer with a 6-inch propeller was immersed in the mixture.  During this
period, the mixture temperature rose to about 140 to 150°F.  A homogeneous
mixture was observed at about 100°F.  The mixture was pumped into tanks
located within the facility.  The empty storage drums were examined for
signs of pulverized coal  which had settled in the mixture during storage
and failed to reentrain during the mixing cycle.  In all cases, no deposits
of pulverized coal  were found.

2.     SUBSCALE COMBUSTION TESTS
       Subscale combustion testing occurred between July 27 and August 11
at the EPA/Acurex facility.  The test facility, shown in Figure 1, is

                                      186

-------
          OBSERVATION
          PORT
00
                                                 PARTICULATE SAMPLING PROBE

                                                 GASEOUS EMISSION SAMPLING PROBE
                                                                                              FURNACE WATERWALLS

                                                                                              REFRACTORY
                                                                                                        SECONDARY AIR
                                                                                                              LINES (5)
        CONVECTIVE
        TUBE BANK
                                                                                                        BURNER
                                                                                                        3 MILLION
                                                                                                        BTU/HR
                                                                                                               AXIAL
                                                                                                               FUEL TUBE
                                                                                                               POSITION
                                                                                                               CONTROL
                              \  /
                                         VARIABLE
                                         SWIRL
                                         CONTROL
                                    V
CONVECTIVE TUBE
OBSERVATION PORT
FLAME
OBSERVATION
WINDOW
                                       ASH PIT
                      Figure  A-1.  EPA/Acurex  Multifuel  Furnace Test Facility - 3 million Btu/hr
                                    capability  — side view.   (See  Figure 2  for Section  A-A)

-------
sponsored by the Environmental Protection Agency to investigate advanced
emission control concepts for utility and industrial boilers.  The configu-
ration additions indicated on the sketch were made to more closely simulate
industrial boiler operating conditions.  One important addition to these
tests was the steam-cooled tube bank across the path of the combustor gases.
The purpose of these tubes was to model the convective section of a boiler
and thereby provide information regarding tube fouling.  Water-cooled tubes
spiralled around the combustion chamber were added to simulate the water-
walled combustion chamber of industrial watertube boilers.

2.1    FACILITY MODIFICATIONS
2.1.1  Convective Tubes (Slagging Probes)
       A.bank of four tubes mounted across the gas flow was designed to
simulate the entrance plane of the convective tube banks of the demonstra-
tion boiler with the primary objective of gaining qualitative information
regarding the fouling tendencies of the Lori Hard fuels.  The tube con-
figuration is shown in Figure 2.   The tube sizing and spacing were selected
to duplicate the velocity through the tubes of the demonstration unit.
Cooling was provided to maintain the tubes below 600°F, the factory esti-
mated temperature of the convective tubes.

2.1.2  Combustion Chamber Waterwalls
       As shown in Figure 1, the combustion chamber preceeding the convec-
tive tube section was lined with several loops of copper tubing for radiant
cooling.   This cooling reduced the bulk gas temperature to below 2300°F
which is the factory estimate of gas temperature entering the convective
section of the demonstration boiler.  The cooling loops were in the three
horizontal extension sections.

2.1.3  Fuel  Supply System
       The fuel supply system is shown schematically in Figure 3.  The item
numbers shown are described in an equipment list in Table 3.
                                    188

-------
 INNER FURNACE
   DIAMETER 30"
   V SQUARE
REFRACTORY
SEE INSERT BELOW


FOUR EQUALLY
SPACED, STAINLESS
STEEL TUBES 1-1/2"
DIAMETER


TUBE SPACING
13/16"
                                                          TWO THERMOCOUPLES PER
                                                          TUBE (ON FLAME SIDE)
                                    INSERT
                                               COOLING AIR AND WATER OUT
                                               COOLING AIR IN

                                               WATER INJECTION
                 Figure A-2.   Convection  tube bank.
                                (Section A-A from  Figure  A-l)
                                     189

-------
                                         MXJ-tXJ-KXMxH
                                                            N-14
                                                           NOZZLE
Figure A-3.   Fuel  supply system schematic.

-------
TABLE 3.  FUEL SUPPLY SYSTEM EQUIPMENT LIST
Item
T-l
T-7
T-9
M-3
M-10
M-ll
HT-2
HT-8
HT-9
TE-101
TE-102
TE-103
TE-106
P-4
P-5
TIC-102
TIC-103
TIC-106
F6A
F6B
TI-104
TI-107
TI-108
H-12
PI-107
PI-108
V-l
V-2
V-3
V-4
N-14
FCV-105
Description
Fuel storage tank
Fuel holding tank
Fuel holding tank
Pneumatic mixer, propeller type
Pneumatic mixer, propeller type
Pneumatic mixer, propeller type
Heating blanket
Strip heaters
Strip heaters
Temperature element
Temperature element
Temperature element
Temperature element
Gear pump
Helical rotor pump
Temp indicating controller
Temp indicating controller
Temp indicating controller
Strainers
Strainers
Temperature indicator
Temperature indicator
Temperature indicator
Circulation heater
Pressure indicator
Pressure indicator
Metering valve
Solenoid valve
Ball valve
Ball valve
Nozzle
Flexible control valve
Comments
55-gal drum
135-gal
135-gal
1.3 hp, variable speed
1.0 hp, variable speed
1.0 hp, variable speed
1200 W
8 per tank, 500 W each
8 per tank, 500 W each

1.0 hp, 450 RPM
0.75 hp, 1200 RPM
70 to 250°F
70 to 250°F
70 to 250°F
1/16" perforations
1/16" perforations
60 to 260°F
60 to 260° F
60 to 260° F
3000 W
0 to 160 psi
0 to 160 psi
Self cleaning
Flame safety
Sampling port
Nozzle flow recirc

Regulates recirculation
                         191

-------
        For  transfer of  fuel  into  the  holding tanks,  (Item Tl, Table 3} a
 COM storage drum is preheated  using electrical  resistance heaters  (Item HT-2)
 and agitated with  a pneumatically driven  shaft  mixer  (Item M-3).   The drum
 is  connected by  flexible hose  to  the  system inlet where the mixture is
 pumped  through either the  gas  pump (Item  P-4) or the  helical rotor pump
 (Item P-5)  (or both pumps) to  holding tank 1 or 2 (Items T-7 and T-9).
        During furnace operation,  COM  is pumped  from either holding tank
 through one of two parallel  strainers (Items F6A or F6B) by either or both
 pumps.  The flow then splits into a recirculation line and a nozzle line.
 The  recirculation  line  returns excess flow to the holding tank.  A flexible
 pinch valve (Item  FCV-105) regulates  the  recirculation flow and as a
 result  acts as a coarse adjustment for flow into the  nozzle.  COM  flow to
 the  nozzle  is directed  through a  circulation heater (Item H-12).   The fine
 adjustment  on flowrate  is  done with a self-cleaning metering valve (Item
 V-l).   Temperatures and pressures are monitored on either side of  this
 valve (Items TI-107 and TI-108, PI-107 and PI-108).    A solenoid valve
 (Item V-2)  is wired to  the furnace flame  safety system.  Flow progresses
 from the system outlet  through flexible hose to the nozzle (Item N-14).
 Fuel samples may be drawn  at any  time (Item V-3).  Prior to light-off,
 nozzle  flow may be redirected into the recirculation  line (Item V-4).
       Several  other flow  options are available.  COM may be transferred
 from either of the holding tanks  to the other holding tank or to the
original storage drum.   Fuel  may  be delivered to the nozzle directly from
 the storage drum.  This operation requires that both  pumps be employed,
however.  The circulation  heater  (Item 38) may be used to augment holding
tank heaters prior to startup.
       The entire system was  electrically heat traced and insulated to
maintain COM temperatures  to at least 140°F.   Temperature was controlled by
five individual  thermostats covering the storage tank to delivery  system
inlet line; the holding tanks,  system piping and the circulation heater.
These on-off type thermostats were capable of controlling fluid temperature
from 60°F to 250°F with a  7°F tolerance.
       A duplex pumping arrangement was chosen such that system shutdown
 would be prevented in case one pump failed and  also to compare operation
                                    192

-------
of the two pumps on COM.   The fuel  flow through each pump was  approximately
8 gpm yielding fluid velocities in  the 3/4-inch lines of approximately
5 ft/second.

2.1.4  Atomization Air System
       Standard shop air at 150 psig at flowrates up to 30 scfm was used
for fuel atomization.  The pressure and flowrate at the nozzle were con-
trolled by appropriate pressure regulators and flowmeters.

2.2    EMISSION MONITORING EQUIPMENT
       Continuous monitors were used to collect emission data.  Table 4
lists the instrumentation used and the principle of operation of each
device.
       Calibration was performed prior to, during, and at the end of  each
test  period.   Correction  in  the emission data  due to calibration shifts,
whenever  present, were taken into account in the calculation of reported
NO  and  CO levels.

2.3    CHECKOUT TESTING
        Prior  to actual testing, several hours  of system  checkout and  equip-
ment evaluation were conducted.
        During operation  on natural  gas,  it  was discovered that the convec-
 tive tube bank was not adequately  cooled  by air alone, and a  water injec-
 tion system was added (see Figure  2).   The  goal  was to maintain the tubes
 at about 600°F, but the coarse control afforded by the water injection
 provided temperatures of about 400°F.   Also, the water in the tubes
 caused differential expansion between the top and bottom of each tube.
 As a result, tube-to-tube and tube-to-wall  spacing changed during the
 test run.
        Two nozzle configurations were evaluated on No. 6 oil and 30-
 percent COM during checkout.  A Delavan Corporation swirl air nozzle and
 a  Sonic  Development Corporation Sonicore nozzle were tested.  The Delavan
 nozzle,  designed for 60  gph maximum flow with a 70-degree spray angle
 performed well with moderate  burner secondary air swirl.  No flame
                                     193

-------
TABLE 4.  EMISSION MONITORING INSTRUMENTATION
Pollutant
NO
so2
CO
co2
°2
Parti cul ate
Loading
Type of Operation
Chemi 1 umi nescence
Pulsed Fluorescent
Nondispersive
Infrared (NDIR)
Nondispersive
Infrared (NDIR)
Paramagnetic
Cyclone and
Filtration
Manufacturer
Ethyl Intertech
Thermoelectron
Ethyl Intertech
Ethyl Intertech
Ethyl Intertech
Acurex Corp
Models
Air Monitor-
ing
Teco
Model 40
Uras 2T
Uras 2T
Magnos 5A
HVSS
Instrument
Range
0-5 ppm
0-10
0-100
0-250
0-1000
0-5000
0-50 ppm
0-100
0-500
0-1000
0-5000
0-500 ppm
0-2000
0-52
0-20%
0-5%
0-21%
0-3 ym
Minimum

-------
impingement on  the  furnace  walls was  observed.   Fuel  pressure  of about  40
psi delivered the desired 10 to 15 gph.   An  atomization  air pressure of 40
psi and a flowrate  of 1000  scfh adequately atomized the  180 to 200°F fuel.
When the fuel temperature was greater than about 200°F,  pulsations in the
flame were observed.   These pulsations,  thought to be due to the vaporiza-
tion of water in the  fuel,  stopped when  the fuel temperature was lowered.
Since atomization was adequate at 180 to 190°F, the remainder of the tests
were run in this temperature range.  Following approximately three hours of
operation on 30-percent COM, extensive erosion was observed on the nozzle
tip.  Figure 4 shows  the nozzle and the areas where erosion occurred.  The
erosion was probably a result of the high fuel-air mixture velocities  (600 to
1000 feet per second) necessary for proper atomization.   Although the
part was supplied as stainless steel, it was discovered later to  be
carbon  steel.  A stainless  steel metering nut with a tungsten carbide  pintle
was  then obtained to minimize  erosion.
        Figure  5  shows a  diagram of the  Sonicore nozzle.  Designed for  flow-
 rates  up  to 60  gph,  it was  operated  at  fuel  pressures of 0 to 5 psig at the
 nozzle  and atomization air pressure  and flowrate of  35  psig and 1500 scfh,
 respectively.   Initial nozzle  operation at  moderate  and low secondary  air
 swirl  resulted in  clinker  formation  at  the  nozzle tip within 15 minutes of
 light-off.   This condition was eliminated at a zero  swirl  setting, but the
 flame was unstable.   Following approximately 3 hours of operation on 30-
 percent COM at various  swirl levels, examination of the nozzle did not
 reveal  any erosion.   A  decision was  made to use the Delavan nozzle.  Even
 though the Sonicore  nozzle was more  erosion resistant,  the Delavan nozzle
 provided superior flame characteristics.
        During the checkout  tests, a thermocouple mounted  adjacent to  the
 furnace wall just prior to  the convective tube  bank was used to  estimate
 combustion  gas temperatures when  the suction pyrometer failed.   Also,  a pro-
 portional  controller was  used to  minimize  fluctuations in fuel  flowrate
 caused by  the  "on-off"  characteristic  of the  circulation  heater thermostat.
 The flow  variation  resulted from the temperature-induced  change in the
 viscosity.
                                       195

-------
                                                   MIXING CHAMBER
 AIR INLET
FUEL INLET
                                    AIR INLET TO
                                    MIXING CHAMBER
EROSION
OCCURRED
IN THESE
AREAS

PINTLE PLATE
                                                             METERING NUT
                  Fiyure A-4.   Dcliiviin swirl-air nozzle
                                  196

-------
                                   RESONATOR CHAMBER
FUEL
                                     STANDING SHOCK WAVE
                    AIR
    Figure A-5.  Sonic Corporation Sonicore nozzle.
                       197

-------
 2.4    TESTING
        The test points completed are shown in the matrix of Table 5

                          TABLE 5.  COM TEST MATRIX
        2 Excess
          Air
                  % Coal
               20
               30
               40
                            1.8 Million Btu/hr
30
40
50
100
                      1.35 Million Btu/hr
30
40
50
                                     100
        For  a  furnace  load of 1.8 million Btu/hr corresponding to the heat
 release  rate  of  the demonstration  boiler at full  load (80,000 pph),  three
 coal-oil  mixtures were  burned  for  a  range of excess  air.   To provide
 reference points, 100-percent  No.  6  oil  and 100-percent  pulverized coal
were fired  at the same  conditions.   Additional  data  was  taken at a reduced
 load of  1.35 million  Btu/hr  corresponding to 60,000  pph  for full-scale.
Test conditions  at 40 percent  of full-scale load  (0.72 million Btu/hr)  were
planned  but eliminated  because of  the  limited range  of burner secondary air
control.

2.4.1  Test Narrative
       Tests were first performed  with 100 percent pulverized coal at a
firing rate of 1.8 million Btu/hr.   Following nearly 3 hours of operation,
approximately 50 percent  of  the  convective passages  were  blocked by  ash
deposits.  The hard,  porous  deposit  was  removed mechanically.
       A  case hardened  (Rc = 58  to 0.03  inches) Delavan  nozzle was used to
obtain test points on No. 6  oil.   This was a higher  capacity nozzle  (100 gph)
than had  been used in the checkout runs,  but no differences were observed
either in flame  shape or  in  emission levels.   After  3 to  5 hours of  testing,
inspection of the nozzle  showed  no signs  of erosion.  Also, no ash deposi-
tion on the convective  tube  bank was noted.
                                     198

-------
       A 30-percent COM was  tested  next  using the  case-hardened  nozzle  used
on the No.  6 oil.   After about 3 hours of operation,  significant deteriora-
tion of the flame  shape suggested nozzle tip erosion  which was confirmed by
inspection.  The erosion pattern was similar to the first eroded nozzle
(see Figure 4); substantial  erosion occurred on the pintle plate with
slight but definite erosion  occurring on the metering nut.  The nozzle  life
was unaffected by the case hardening process.  A new nozzle with tungsten
carbide pintle and stainless steel  metering nut was used to complete the
tests.  This 60 gph nozzle was compared  with the larger nozzle by taking two
duplicate points; no difference was observed in the two nozzles.  After
approximately 2-1/2 to  3 hours, the nozzle was inspected and significant
erosion was observed.   In this case, only the stainless steel metering nut
had eroded  and the tungsten carbide pintle remained unchanged.
       The  convective tubes showed some fouling but this was minimal
compared to the pulverized coal.
       A 50-percent COM was tested next.  A  high capacity  (150  gph)  Delavan
nozzle was  chosen  to minimize  velocities  through the nozzle  tip.  Test  data
was taken  prior to flame deterioration  at the  3-hour point when erosion  was
again observed.  At this point,  approximately  10 percent  of  the total  con-
vective passage was occluded  by  ash  deposition.  Factors  contributing  to this
high  rate  of  deposition were  probably the volumetric heat release  (0.75
 x 106  Btu/hr-ft3)  and  the hot refractory wall  of  the furnace.
        The 40-percent  COM was the  last  fuel  tested.  Equipment  problems
 arose after about  1  hour of testing.  First, the  helical  rotor  pump failed.
 Subsequent inspection  revealed that the rotor was  extensively galled.   The
 gear pump was used for the  remainder of the tests.  Second,  plugging of the
 fuel  metering valve was experienced.  The valve was  removed  and cleaned
 twice without success.
        After flushing  the entire system with No.  6 oil, a second attempt at
 40-percent COM was made.   Similar valve plugging was experienced.   The valve
 was replaced with a conventional needle valve and data points were taken.
 Erosion of the nozzle was observed following these tests, but  1t was less
 than that  leading to eratic flame patterns.  Fouling was  less  than with
                                      199

-------
 50-percent COM but greater than with 30-percent COM.  Again, the high
 volumetric heat release and the hot-wall effect probably contributed to the
 high rate of deposition.

 2.4.2  Emission Tests
       Pollutant emissions for each fuel were measured at excess air levels
 of 20-, 30-, and 40-percent.  For these measurements, the furnace load was
 maintained at 1.8 x 106 Btu/hr.  This load corresponds to a volumetric heat
 release of 0.75 x 106 Btu/hr-ft3, which is approximately the same as the
 demonstration boiler at full load.

 2.4.2.1  Nitric Oxide (NO) Emissions
       Figures 6 through 10 show NO levels as a function of stoichiometric
 ratio for No.  6 oil, 30-, 40-, and 50-percent COM, and pulverized coal.  In
 Figure 8, the data taken in test 201a is believed to be most representative
 as it was taken prior to plugging problems experienced with the fuel supply
 system.  The data recorded in tests 201b, c, and d, are questionable since
 partial fuel supply blockage occurred during these tests.  As expected, NO
 emissions increased with stoichiometric ratio for each fuel.  This was
 probably due to increased oxidation of fuel nitrogen with each increase in
 excess air.   The rate of increase of NO with excess air was greater for coal
 and coal-oil mixtures than for oil alone.  This is attributed to the in-
 creased emissions of fuel NO for the coal-containing fuels.  Fuel NO is
 generally more sensitive to excess air levels than thermal NO which pre-
 dominates with oil combustion.   Table 6 lists general properties of the
 five fuels tested.  Note that fuel nitrogen increases as coal content of
 the fuel (mixture) increases.
       The effect of coal content on NO emissions is shown in Figure 11.
The upper curve represents NO levels recorded at 40 percent excess air,
 and the lower curve represents data taken at 20 percent excess air.  NO
emissions from COM combustion were slightly lower than levels expected
 from a straight proportional weighting of emissions according to weight
 percentage of coal and oil.
                                   200

-------
ro
                            a.
                           O
                                BOO
                                700
                                eoo
                                500
                                400
                                3OO
                                200
                                100
A
                                                                                 FUEL:  NO. 6 OIL
                                                                                 BURNER: 3 MILLION BTU/HR IFRF; OELAVAN NOZZLE
                                                                                 FURNACE CONFIGURATION:  HORIZONTAL EXTENSION
                                                                                 SWIRL:  6 OF 8
                                                                                 LOAD:  1.8 MILLION BTU/HR
                                                                                 TEST ID   SYMBOL    % EXCESS AIR
            198a
            198b
            198c
            198d
            1986
            198f
                                  1.0       1.1       1.2       13       1.4       1.5       1.6       1-7

                                                         STOICHIOMETRIC RATIO
20
20
30
40
10
40
                         Figure A-6.   Nitric oxide (NO)  versus stoichiometric ratio:   No.  6 oil,

-------
ro
o
                       Q.
                       Q.

                       O
                           MO
                           700
                           eoo
                           500
                           400
                           300
                           200
                           100
                                                                         FUEL: 30 PERCENT COM
                                                                         BURNER: 3 MILLION BTU/HH IFRF; DELAVAN NOZZLE
                                                                         FURNACE CONFIGURATION:  HORIZONTAL EXTENSION
                                                                         SWIRL: 5 OF 8
                                                                         LOAD:  1.B  MILLION BTU/HR
                                                                         TEST ID    SYMBOL  % EXCESS AIR
199a
199b
199c
O
Q
O
A
199e
199f
                                                                                                20
                                                                                                30
                                                                                                30
                                                                                                20
                                                                                                30
                                                                                                40
                             "1.0       1.1       12       13       1.4       1.5

                                                     STOICHIOMETRIC RATIO
       1.6
                1.7
                         Figure A-7.   Nitric oxide  (NO versus stoichiometric  ratio:   30-percent COM.

-------
ro
o
oo
                      I
                      Q.
                          800
                          700
                          800
                          500
                         400
300
                         200
                         100
                                             FUEL: 40 PERCENT COM
                                             BURNER: 3 MILLION BTU/HR IFRF: DELAVAN NOZZLE
                                             FURNACE CONFIGURATION:  HORIZONTAL EXTENSION
                                             SWIRL:  5 OF 8
                                             LOAD: 1.8 MILLION BTU/HR
                                                                      TEST ID  SYMBOL
                                                                        201a
                                                                        201 b
                                                                        201C
                                                                        201 d
                                                       o
                                                       GJ
                                                       2
EXCESS AIR

   20
   20
   30
   20
                                                                      FLAGGED SYMBOLS REPRESENT
                                                                      QUESTIONABLE DATA DUE TO
                                                                      FUEL SUPPLY SYSTEM PLUGGING.
                            1.0       1.1       1.2       1.3       1-4       1.5      1.6

                                                  STOICHIOMETRIC RATIO
                                                              1.7
                       Figure A-8.   Nitric oxide (NO)  versus stoichiometric  ratio:   40-percent  COM.

-------
  0.
   EXCESS AIR
     20
     30
     40
         1.0       1.1       1.2       1.3      1.4

                             STOICHIOMETRIC RATIO
                                                    1.5
                                                             1.6
                                                                     1.7
Figure A-9.   Nitric oxide  (NO) versus  stoichiometric ratio:   50-percent COM,

-------
                        1200
                       1000
ro
o
en
                   O
                        800
                        800
                        400
                       200
                           1.0
11         1.2        1.3

        STOICHIOMETRIC RATIO
                                                                      1.4
                                                                            FUEL:  PULVERIZED COAL
                                                                            BURNER: 3 MILLION BTU/HR IFRF; B«W SPREADER
                                                                            FURNACE CONFIGURATION: HORIZONTAL EXTENSION
                                                                            SWIRL:  4 OF 8
                                                                            LOAD:  1.8 MILLION BTU/HR
                                      TEST ID

                                        197«
                                        1976
                                        197c
                                        1970
                                        197e
                                                                                      O
                                                                                      Q
% EXCESS AIR

     20
     30
     40
     10
      5
                                                                                 1.5
PREHEAT TEMP.

    100
    100
    102
    102
    102
                   Figure A-10.   Nitric  oxide (NO)  versus  stoichiometric  ratio:   pulverized  coal

-------
                         TABLE 6.  FUEL  PROPERTIES
Fuel
Type
No. 6 oil
30% COM
40% COM
50% COM
100% Coal
% N
By Weight
0.36
0.70
0.82
0.93
1.5
% S
By Weight
2.22
1.82
1.69
1.56
0.9
% Water
By Weight3
0.0
4.82
5.24
5.66
4.2
% 02
By Weight
0.0
2.67
3.56
4.45
8.9
HHV
Btu/lb
18,800
17,170
16,627
16,084
13,368
Includes 3.56 percent of fuel by weight due to  fuel additive.
                                    206

-------
ro
                   1400
                   1200
                   1000
                    800
                    600
                    400
                    200
                                                                                             O 20% EXCESS AIR

                                                                                             A 30% EXCESS AIR

                                                                                             G 40% EXCESS AIR
                                                                                      40% EXCESS AIR
                                                                            20% EXCESS AIR
                          OIL
                          ONLY
                                    10       20
30      40       50      60

     PERCENT COAL IN FUEL
PULVERIZED
COAL ONLY
                              Figure  A-ll.   Nitric  oxide  (NO) versus percent coal in  fuel

-------
       The  lower NO emissions of the COM fuels could have been caused by
 the water content of  the additive, or  the shielding of coal by the oil
 spray.  Water content in the fuel has  been found to reduce NO emissions.
 During some studies of water emulsions with distillate and residual oils
 the NO levels were reduced an average  of 100 ppm when approximately 5-
 percent water was added to the fuel oil.*  In the COM tests the water con-
 tent of the fuels attributed to the additive were 3.56 percent by weight.
 Accounting  for this water content, on  the basis of the emulsion tests cited
 above, yields NO levels which conform  more closely to proportional levels
 based on No. 6 oil and pulverized coal NO emissions.  The second factor
 which may have affected NO emissions during these tests results from the
 layer of oil surrounding each coal particle.  This oil layer delays oxygen
 diffusion to the coal  and suppresses oxidation of coal fuel nitrogen to NO.

 2.4.2.2  Carbon Monoxide (CO) and Unburned Hydrocarbon (UHC) Emissions
       Carbon monoxide (CO) and unburned hydrocarbon emissions were nearly
 zero for all COM, oil  and coal tests.  CO emissions were insignificant even
 though the excess air levels were reduced to 10 percent on occasion.  UHC
 emissions were undetectable during all tests.

 2.4.2.3  Particulate  Mass Loading
       Particulate stack sampling tests were conducted for the 50-percent
 coal and pulverized coal fuels.   The results indicate that very low frac-
 tions of the ash contained in the fuel went out the stack.  Based on the
 relatively small amounts of ash remaining in the furnace (compared to the
 amount of fuel fired), the stack test  results are in question.  The test
 results are also contradicted by more extensive testing at General Motors
where nearly 100 percent of the ash appeared in the flue gas.
*G.B. Martin, "Evaluation of NOX Emission Characteristics of Alcohol Fuels
 in Stationary Combustion Systems," presented at the Joint Meeting Western
 and Central States Sections, The Combustion Institute, San Antonio, Texas,
 April 21 to 22, 1975.
^Brown, A, "First Report of the General Motors Corporation Powdered Coal-In-
 Oil Mixtures Program," ERDA Contract E(49-18)-2267, December 1976.

                                   208

-------
2.4.3  Discussion
       Pumping of  the COM resulted  in  the  failure  of  the  helical  rotor  pump
after approxiamtely 50 hours of operation.   On  disassembly,  no  damage was
observed to the BUNA-N stator.   However, significant  galling was seen on  the
lobes of the rotor.  Two possible explanations  for this  failure are:  the
pump should have operated at 400 to 500 rpm but instead  operated close-
coupled to a 1200  rpm motor; the pump operated  in  a dry  state for short
periods.  The transfer of COM from the storage  tank to the system feed  tank
was probably responsible for the dry operation  of the pump.   For the transfer
operation, the helical rotor pump was actuated  as soon as the mixture was
thought pumpable.   This  resulted in dry pump operation if the mixture was
not flowing.
       The Viking  gear pump did not fail, but developed  leaks  in the packing.
When disassembled  after  test completion, hardened or congealed  fuel was
detected  between the  shaft  and  the packing  resulting in  leakage.  This prob-
lem could probably be remedied  by  the  use of mechanical  seals.   Inspection
of the pump  components susceptible to  wear  showed no indication of abrasion
after  about  50 hours  of  operation.
        Metering problems resulted  primarily from  the very low  flowrates
associated with test conditions.   The  self-cleaning  micrometering valve  had
a maximum fuel  passage dimension of  about 0.125 inches.   Operation was found
to be  most  effective if  the valve  metering  position  was  between 75 and 100
percent open.   Operation in this range was  possible  by  controlling the
amount of recirculation  by  adjusting the  flexible valve. While the  valve
 performed satisfactorily with  both 30- to 50-percent COM, excessive  plugging
occurred with the 40-percent COM.   For the 40-percent COM,  however,  operation
was impossible even in the  100-percent open position.   The  plugging occurred
 upstream of the metering groove where the self-cleaning feature was ineffec-
 tive and where plugging would not be suspect because of its larger dimensions
 (approximately 0.25-inch).
        No explanation can be given for the failure of the metering valve on
 the 40-percent COM while not on the 50-percent COM.   Subsequent analysis of
                                     209

-------
 fuel  samples  indicated a  +2  percent tolerance  on  the total  solids  content
 of the fuels  indicating that the difficulty was not  in  improperly  prepared
 fuel.
        Although  the  exact cause  of the  failure is  not known,  two definite
 possibilities  exist.
        •    Insufficient additive in the  fuel resulted in  particle  agglomera-
            tion  leading to eventual  plugging
        •    Following  each COM fuel,  the  entire system was flushed  with No. 6
            oil stored in  the  second holding  tank.  Continued  flushings
            resulted  in  a  mixture of No.  6 and  the  coal  in the second holding
            tank  and  in  the system lines.  In standing for any period of
            time, this coal would settle  onto the walls  of the pipes.  Settle-
            ment  occurred  because the dilution  of the  additive rendered it
            ineffective.   Agglomerated coal particles  would then reentrain
            from  the pipe  walls when  flow occurred.
        Of these  two failure modes,  the second  is the  most suspect  since each
 drum of fuel required two and  a  half batches of fuel  preparation.  Carbonoyl
 Company has stated that little difference in fuel  characteristics would be
 observed if the  additive was not included in one of these batches.  Also, it
 is highly improbable  that three  consecutive batches would have been made
 without additive.  The  second  failure mode is  supported by the fact that
 plugging was the result of agglomerated  coal particles.   The packing of
 these particles  was very  similar to that observed  when  unstabilized coal is
 removed from a sample after settling.
       Significant erosion of  the  Delavan nozzle was  experienced.  At the
 time, this occurrence was viewed  as a major problem but further investiga-
 tion indicated that a change in  nozzle design would probably remedy the
 problem.  The Delavan nozzle uses  an impingement type internal-mix atomiza-
 tion scheme which, by design,  is  highly susceptible to erosion.  A recommen-
 dation  for future use of  air atomizing nozzles are the external-mix type
where atomization takes place  after the  fuel and air  have left the nozzle.
 Several types under this  general  design  are available.
                                    210

-------
       Fouling  of the convective  tube  bank was  observed  for  all  coal-con-
taining fuels.   Deposition was  greatest,  as  would be  expected,  for the
pulverized coal and least for the 30-percent COM.   For the pulverized coal,
about half of the convective flow passages were occluded.  Deposition
associated with the 50-percent COM was less  obvious.   Due to the displace-
ment of the convective tubes, some tube-to-tube and tube-to-wall spacings
were reduced.  Increased deposition was noted for these spacings while
generally no fouling occurred in spacings which remained equal  to or wider
than planned.
       Two existing conditions which probably contributed substantially to
tube fouling were the volumetric heat release rate of the test points and
the loss of  radiant cooling in the combustion chamber.  The volumetric heat
rate of the  tests corresponded to that of the demonstration boiler at full
load.  Since the initial  tests were made at this load, fouling  had already
occurred when  the load was  reduced to 1.35 x 106 Btu/hr and any  subsequent
slagging went  unobserved.   The loss of coolant tubes  during checkout prob-
ably contributed to  slagging as well.  The  radiating  refractory  was
sufficiently hot to  melt  the impinging ash  particles.   Slag thickness
averaged  about 0.25  inch  at the completion  of  the  test  program.   These
slagging  results  indicate that tube fouling  is  a possibility at full-scale
operation,  but firm  conclusions are inappropriate  at  this time due to  the
variation in results at  both GM and PERC  (Pittsburgh  Energy Research Center)
                                    211

-------
A     A XNI lOCV                                                     additional routing

.    Corporation


                        INTEROFFICE

                    to:  Craig Derbidge

                  from:  Allen Shimizu

                  date:  September 15, 1977
           The  viscosity tests on the Lorillard fuels have been completed,
     although  their validity is in question.  Doubt arises from the exceedingly
     high viscosities of the 50% slurries.  This indicates either improper
     measurement procedure or extremely high viscosities.  A study of the litera-
     ture for  operation of the Brookfield viscometer did not reveal any error
     in procedure.  In addition, prior to the viscosity measurements, the instru-
     ment was  calibrated with two standard liquids.  This all points to a highly
     viscous mixture at 50%.

           Figure 1 shows viscosities for the following fuels:

                •  GM oil:  472 coal weight COM
                •  Lori Hard oil:  50% COM
                t  Chevron oil & Ptsbg. 18 oil:  30, 35, 40, 50% COM

                t  PERC oil:  20, 40% COM

     Note that the PERC and GM oils are substantially less viscous than the  Lori Hard
     and Chevron oils.  Also note that the slurries made with the GM and PERC oils
     are far less viscous than those made with the Chevron oil.   This leads  one to
     believe that the viscosity of the mixture may be a strong function of the oil
     viscosity.  Figure 2 plots the ratio of the mixture viscosity to the oil vis-
     cosity versus the X coal 1n the mixture.  This figure Indicates that the
     "normalized viscosities" as a function of coal fraction at 150°F are similar
     for the PERC 20 and 40% COM. the GM 47% and the Chevron 30. 35. and 40% COM.
     The two points at 50% seem to be extremely hijjh,  but this may be due to  the
     coal  fraction.  According to Brown at GM, the viscosities of the mixtures In-
     creased significantly as the coal fraction approached and exceeded 50%. His
     comments  are reflected in the Chevron/Ptsbg.  #8 tests  showing the 30,  35
     and 40% viscosities with uniformly Increasing values, while the 50% COM
     exhibits  anomolously high values.

           Also shown on Figure 1 are viscosities of the GM oil and Marathon
     oil  mixtures made with the petrol ite additive.  The GM oil  and -200 mesh
     coal  1s denoted by PGM 30 and PGM 50 for 30 and 50% COM.  The Marathon  oil.
     which 1s  similar In viscosity to the Chevron oil, 1s denoted MAR 30 and
    "MAR 50.   These were also made with -200 mesh coal.  Those marked with FN
     Indicate  -325 mesh coal.  This shows a slight viscosity Increase with finer
     particles.


                                    212

-------
September 15,  1977
Allen Shinrizu  memo
COM Viscosity  (con't)
Conclusions:
       •  Our viscosity measurements were correct
       §  Mixture viscosity 1s a strong function of oil viscosity
       •  Viscosity of the mixture Increases significantly as the
          coal fraction approaches and exceeds 50%
       •  Although decreasing particle size Increases viscosity,
        -  viscosity appears to be only a weak function of particle
          size
       t  GM used a #6 oil that was particularly fluid
       •  The GM 46.6% COM should be questioned
 ABS:mmcL
                                 213

-------
    10*
                50% LORILLARD FUELS
    10s
                                                     MAR 50FN
    104
2
UJ
t-

1U
0
(0
o
    103
        PERCENT SOLIDS
                            PERC OIL
    10'
              80      100      120      140     160      ISO      200



                            TEMPERATURE - *F
           Figure  1.   Viscosity  of various COM  fuels.
                                   214

-------
  iooo r
                                                  LORILLARD
                                                  CHEVRON/W
         TEMPERATURE = 150°F
                            GM + PETROLITE @160*F
                             MAR/CHEVRON @160*F
O  100
(D
6
O
O
    10
                                     O
                                     /
                                          *
                             V
              10
20      30
PERCENT COAL
                                      40
50
      Figure 2.  Relative COM viscosity versus coal
                  mixture ratio  (percent by  weight).
                               215

-------
                               TECHNICAL REPORT DATA
                         (Please read Instructions on the reverse before completing)
1. REPORT NO.
 EPA-600/7-80-043
                          2.
                                                     3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
 Pilot Scale Combustion Evaluation of Waste and
 Alternate Fuels: Phase III Final Report
                                 5. REPORT DATE
                                  March 1980
                                 6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)

R.A. Brown and C. F. Busch
                                                     8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Acurex Corporation
Energy and Environmental Division
485 Clyde Avenue
Mountain View, California  94042
                                 10. PROGRAM ELEMENT NO.
                                 EHE624A
                                 11. CONTRACT/GRANT NO.
                                 68-02-1885
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
                                 Phase III Final; 2-8/78 	
                                 14. SPONSORING AGENCY CODE
                                   EPA/600/13
is. SUPPLEMENTARY NOTES IERL-RTP project officer is David G.  Lachapelle, Mail Drop 65,
919/541-2236. EPA-600/7-79-132 was the Phase II final report; there was no Phase I
final report.              	                                        	
is. ABSTRACTThe report gives results of three studies at EPA's Multifuel Test Facility.
The first evaluated a distributed-air staging concept for NOx control in pulverized-
coal-fired systems. The results showed that minimum NO levels of 140 ppm were
achieved at overall residence times similar to those used during conventional sta-
ging tests.  However, the NO levels  achieved with the distributed-air concept were
no lower than those achievable with  conventional staging.  The second evaluated com-
bustion control techniques and NO emissions when firing coal/oil mixtures. NO emis-
sions for a given burner and nozzle  were generally proportional to the fuel-nitrogen
content of the fuel. Additionally, combustion control technology currently used for
NOx control from pulverized coal was found to be effective with coal/oil mixtures,
but to differing degrees, depending on the coal/oil mixture ratios and compositions.
The third evaluated emissions and combustion characteristics of refuse-derived fuel
(RDF) co-fired with either natural gas or pulverized coal. Four RDF materials were
evaluated for gaseous,  particulate,  trace metal, and organic emissions. In general:
CO and UHC emissions were low; NOx and SOx emissions decreased with increasing
RDF  content when co-fired with coal; particulate levels did not substantially increase
with the RDF; and no trace metal emissions correlation was found.
                            KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
                                         b.IDENTIFIERS/OPEN ENDED TERMS
                                                COSATl Field/Group
Pollution
Nitrogen Oxides
Combustion Control
Refuse
Wastes
Coal
Fuel Oil
Pollution Control
Stationary Sources
Staged Combustion
Refuse-derived Fuel
Coal/Oil Mixtures
Alternate Fuels
13B
07B
21B
                                              21D
13. DISTRIBUTION STATEMEN1

 Release to Public
                     19. SECURITY CLASS (This Report)
                      Unclassified
                         21. NO. OF PAGES
                         227
                     2O. SECURITY CLASS (This page/
                      Unclassified
                                              22. PRICE
IPA Form 2220-1 (9-73)
                                       216

-------