United States
Environmental Protection
Agency
National Risk Management
Research Laboratory
Cincinnati, OH 45268
Research and Development
EPA/600/SR-97/061   September 1997

Burner Technology
S. Venkatesh, W. E, Whitworth, Jr., C. Goldman, and L. R, Water-land
  A series of demonstration tests was
performed at the  Environmental Pro-
tection Agency's (EPA's)  Incineration
Research Facility (IRF) under the Su-
perfund Innovative Technology Evalua-
tion (SITE) program. These tests, 12 in
all, evaluated a  pulse  combustion
burner technology  developed by
Sonotech, Inc., of Atlanta, Georgia. The
burner system incorporates  a  pulse
combustor, the pulsation frequency of
which  can be tuned  to  induce  large
amplitude sonic  pulsations inside  a
combustion  process  unit such  as  a
boiler or an incinerator.
  The primary objective of the test pro-
gram was to  develop test data to allow
evaluating whether the Sonotech pulse
combustor applied to the IRF  rotary
kiln system (RKS), when compared to
conventional, non-pulsating combus-
tion, resulted in: decreased flue gas
NOX, CO, and  soot emissions; increased
POHC ORE; decreased combustion air
requirements; decreased  auxiliary fuel
requirements; and increased incinera-
tor capacity. The waste feed for all tests
was a mixture of contaminated materi-
als from two manufactured gas  plant
(MGP) Superfund sites. One component
of the test waste was a combination of
pulverized  coal and contaminated
sludge waste from the Peoples Natural
Gas Company site in Dubuque, Iowa.
The other components of the    waste
consisted  of soil  borings and  a tar
waste from an oil gasification process
obtained from an MGP site in the south-
east U.S. To address the demonstra-
tion objectives 12 tests were performed
under  four different  test conditions.
Each condition consisted of three iden-
tical     to obtain data  in triplicate.
Test data were statistically evaluated
using the rank sum test. When the rank
sum test is  applied to two data sets,
each containing three      points, the
data     are different, at the 95 per-
cent confidence  level, when  corre-
sponding ranges do not overlap.
  The results addressing the primary
    objectives of the     program are
as follows: the average CO emissions
at the afterburner  exit decreased from
20 ppm (range of  8.0 to 40.0 ppm) for
test condition  2 to 14 ppm (range of
12.6 to 16.0 ppm) for test condition 3;
the average NOX decreased from 82 ppm
(range of 78.3 to 85.1 ppm) for test con-
dition 2 to 77 ppm (range of 68.0 to
87.1 ppm) for test  condition  3; the av-
erage soot emissions decreased from
1.9  mg/dscm (range of 0.9 to 2.7 mg/
dscm) to    than 1.0 mg/dscm  (range
of <0.8 to 0.9 mg/dscm). Benzene ORE
for all 12 tests was greater than  99.994
percent. Naphthalene ORE for  all  12
tests was greater than  99.998 percent.
The  combustion  air requirements
showed a decrease from the 38,400 to
40,600 dscfh  range to the  34,800  to
39,900 dscfh  range when the Sonotech
combustor was used; the auxiliary fuel
requirements (natural gas) were  practi-
cally equal  under conventional and
Sonotech conditions;  and  the  waste
feedrate capacity increased by 13 per-
cent with the Sonotech burner  opera-
tional. As the demonstration waste had
a significant heat  content (8,500 Btu/
Ib), the capacity increase may be trans-
lated into a reduction in auxiliary fuel
needed to treat a  unit mass of waste
from 21.1 kBtu/lb for conventional com-
bustion to 18.0 kBtu/lb for the Sonotech
system.
  This Project Summary was developed
by EPA's  National Risk Management
Research Laboratory, Cincinnati, OH,
to announce key findings of the research
project that  is fully documented in a
separate report of the same title (see

-------
Project Report ordering information at
back),

Introduction
  Sonotech, Inc., of Atlanta, Georgia, has
developed a pulse combustion burner tech-
nology that claims to offer benefits when
applied  in  a  variety  of combustion  pro-
cesses. The burner system incorporates a
pulse combustor that can  be tuned to ex-
cite  large-amplitude sonic pulsations in-
side a  combustion  chamber  such as a
boiler or incinerator. Sonotech claims that
these pulsations serve to  increase  the
rates of heat, mixing (momentum),  and
mass transfer in the combustion process;
and  claims that these rate increases  are
sufficient to result in more complete com-
bustion.
  Sonotech has targeted  waste incinera-
tion as a potential application for this tech-
nology.  Accordingly,  to demonstrate  the
claimed benefits  of the technology within
a well-established forum for providing tech-
nically sound and  unbiased evaluations,
Sonotech  proposed a technology evalua-
tion test series under the Superfund Inno-
vative Technology  Evaluation (SITE)
program. The Sonotech proposal was ac-
cepted,  and an evaluation  test program
was performed at  EPA Incineration Re-
search Facility (IRF) in Jefferson, Arkan-
sas.
Description of the Technology
  A pulse combustor typically consists of
an air  inlet,  a combustor section, and a
tailpipe. Periodic variations in fuel oxida-
tion and heat release  produce pulsations
in  the  combustor section pressure, tem-
perature, and gas velocities.  When prop-
erly applied,  a pulse combustor can excite
large-amplitude (150 dB or greater) pulsa-
tions within  a  cavity  downstream of  the
pulse combustor tailpipe. This cavity could
be the  combustion chamber of a boiler or
an incinerator, for example.
  A retrofit  application of the  Sonotech
pulse combustion system was  evaluated
in  this  test program. Specifically,  the  kiln
section of the RKS at the IRF was retrofit-
ted with a pulse combustion burner  ca-
pable  of   delivering  up  to  73  kW
(250,000 Btu/hr) of heat input from natural
gas fuel to the kiln. This corresponds to
15 to 20 percent of the typical heat input
to  the  kiln. The RKS  was configured as
shown  in Figure 1, with the Sonotech com-
bustion system retrofitted into the end plate
at  the ash discharge end of the kiln.

Demonstration Objectives
  Sonotech  claims  that the application of
pulse combustion technology to an incin-
eration system has several significant  ad-
vantages over conventional (non-pulsating)
incineration.  Thus,  the general objective
of the demonstration test program was to
develop the data needed to  allow objec-
tive and  quantitative  evaluation of these
claims. Accordingly, the primary test pro-
gram  objective was to develop test  data
to allow evaluating  whether the Sonotech
pulse  combustion technology applied  to
the  IRF RKS, when compared to  conven-
tional, non-pulsating combustion,  resulted
in:
      Increased incinerator  capacity  or
      productivity
      Increased principal organic hazard-
      ous constituent (POHC) destruction
      and removal efficiency (ORE)
      Decreased flue gas CO emissions
      Decreased flue gas NOX emissions
      Decreased flue gas soot emissions
      Decreased combustion air require-
      ments
      Decreased  auxiliary fuel  require-
      ments
  The secondary test program objective
was to develop test data to allow  evaluat-
ing whether the application of the Sonotech
technology, when compared  to  conven-
tional, non-pulsating combustion:
      Reduced the magnitude of transient
      puffs of CO and total unburned hy-
      drocarbons (TUHC)
      Allowed reduced incineration costs
      Caused significant changes in:

Quench
Y
Secondary
burner „„ ,
H-i Afterburner
Air extension
Natural 1
gas, y | ooiius
liquid Ll Afterburner Bed
feed II — 1 Feeder
Transfer fl 1 I
duct — * \ T-II 	 H-I-I LJ .
1 1 1=0 ^ain
Natural . — __;, I I K^ burner
yd& Ar psh iMJip J

Sonotech 0 '»
Burner ^°tary Natural
system Klm gas
liquid feed
Rotary Kiln Incinerator
^te
tx Pac
/\ Co1
f \ Scr

— "jam '}
Venturi T \/
scrubber ^
Scrubber
Iquor
recirculation
K/1 '
] ID fan

{>< reheater
	 ' \
MM
\ /
Ash hopper | 	 |
Primary Air Pollution Control System

^\7y^
'• ^^ ^' KN
Carbon bed HEPA
adsorber filter
ghouse
Secondary Air Pollution
Control System
Atmosphere
t
n Stack
X
K^O
> (UJ
ID fan
Figure 1.   Schematic of the IRF rotary kiln incineration system.

-------
      —  The distribution of hazardous
          constituent trace metals among
          the  incineration  system dis-
          charge streams  (kiln bottom
          ash, scrubber liquor, baghouse
          flyash, and  baghouse exit flue
          gas)
      —  The leachability of the toxicity
          characteristic leaching  proce-
          dure (TCLP) trace metals from
          kiln bottom ash, scrubber liquor,
          and baghouse flyash
This last secondary objective item does
not relate to any Sonotech claim, but is of
general interest to the overall IRF research
program.

Test Program
  To address the test program objectives,
tests at the following four different incin-
eration system operating  conditions were
performed:
      Test Condition  1: Conventional
      combustion under baseline, typical
      RKS operation (28 kg/hr [61 Ib/hr])
      Test Condition  2: Conventional
      combustion at the maximum RKS
      waste feedrate without  pulsations
      (33 kg/hr [73 Ib/hr])
      Test Condition  3: Sonotech  pulse
      combustion at nominally the  same
      feedrate and  conditions as Test
      Condition 2 (34 kg/hr [74 Ib/hr])
      Test Condition  4: Sonotech  pulse
      combustion at the maximum RKS
      waste feedrate with pulsations  (37
      kg/hr [82 Ib/hr])
  The test waste feed for all tests was a
mixture  of contaminated materials from
two manufactured gas plant (MGP) Su-
perfund  sites. (The  specific  components
of this feed are discussed later.) This waste
feed was  batch  fed to the RKS via  the
system's fiberboard container ram feed
system,  which feeds 1.5-gal (5.7-L) fiber-
board containers  to  the  kiln  at  virtually
any specified feed frequency. When a rela-
tively high heat content material is  being
fed, the maximum allowable waste feedrate
is  established  based  upon  the onset of
puffs of incompletely combusted organic
constituents (CO and TUHC) that survive
the afterburner.
  Given this,  Test Condition 1 was at a
waste feedrate consistent with stable in-
cinerator  operation under  conventional
combustion, with  infrequent spikes of CO
and/or TUHC  at the afterburner exit. Test
Condition  2 was at an  increased waste
feedrate that resulted in routine afterburner
CO spikes. This condition could be termed
borderline acceptable incinerator  opera-
tion under conventional combustion. Test
Condition  3  was  at the  same  waste
feedrate as Test  Condition 2, but with the
Sonotech pulse combustion system in op-
eration. Test Condition 4 was at a further
increase in waste feedrate, with the pulse
combustor in operation, such that routine
afterburner exit flue  gas CO spikes re-
curred. This condition  could  be termed
borderline acceptable  operation under
pulse combustion  operation.  Three  test
runs (triplicate  testing) at each test condi-
tion were completed to allow the precision
of each emission and  discharge  stream
composition measurement to be assessed.
  As indicated above, the test waste  feed
material for the test program was a  mix-
ture  of materials from  two MGP  Super-
fund sites. One component of the material
was a combination  of pulverized coal and
contaminated sludge/soil  waste from the
Peoples Natural Gas Company Superfund
site  in Dubuque, Iowa. This site is an
abandoned coal MGP site, and the sludge
waste at the site contains high concentra-
tions of coal tar constituents. The other
components of the test feed material were
contaminated soil borings and a tar waste
from an oil gasification  process, both ob-
tained from an  MGP site in the southeast-
ern United States.
  The hazardous constituent contaminants
of all three test waste components were
several polynuclear aromatic hydrocarbon
(PAH) compounds, and the  VOCs  ben-
zene, toluene,  ethylbenzene,  and xylenes
(BTEX). Although concentrations of  sev-
eral contaminant compounds were quite
high in at least the tar  component of the
waste mixture,  it was  decided that spiking
the waste feed with  benzene and naph-
thalene would  be necessary to guarantee
meaningful ORE calculations.
  To address the test program objectives,
the composite waste feed, the kiln  ash
discharge, the scrubber system liquor, the
collected baghouse ash,  the  afterburner
exit flue  gas  particulate,  the  afterburner
exit flue gas, and the baghouse exit flue
gas for each test were sampled  and  ana-
lyzed for sample-matrix-specific combina-
tions of PAHs, VOCs, contaminant trace
metals, polychlorinated  dibenzo-p-dioxins
and   polychlorinated  dibenzofurans
(PCDDs/PCDFs),  and TCLP-leachable
trace metals. In addition, the total organic
carbon (TOC)  content of the afterburner
exit flue gas particulate was determined
and used as the measure of soot emis-
sions. Later in  the test  program, measur-
ing the heating value of  each test's  kiln
ash discharge was added as an indication
of waste treatment residue quality in terms
of completeness of waste burnout.
  Test data from the program were evalu-
ated using the rank  sum test. The  rank
sum test says  the two data sets, each
containing three data points (the case for
this program) are statistically different at
the 95 percent confidence level when the
data ranges in each set do not overlap.

Test

Incinerator Operating
Conditions
  Table 1  provides a summary of the av-
erage incineration system operating  con-
ditions for each of the four program test
conditions.  Each  operating  parameter
noted in the table was recorded nominally
every 30 seconds over a 4- to 5-hour flue
gas sampling period for each test by the
RKS data  acquisition system. Test aver-
ages were calculated for each parameter.
The  data  in Table 1  represent the aver-
age value determined for the three tests
performed for each test condition.
  The data in Table 1 show that the kiln
exit gas  temperatures collected for all con-
ditions averaged close to the test program
target of 925°C (1,700°F), and that aver-
age afterburner exit gas temperatures were
right at the test program target of 1,095°C
(2,000°F).  Afterburner exit flue gas O2 lev-
els averaged close to 9 percent for all test
conditions,  although  slightly lower aver-
age  levels  existed for Test Conditions 3
and 4, the two pulse combustion test con-
ditions.
  The baseline (Test Condition 1) waste
feedrate was 27.7 kg/hr (61.0 Ib/hr).  This
feedrate was increased to  33.1  kg/hr
(72.8 Ib/hr) to give the borderline accept-
able operation associated with Test Con-
dition 2.  Test Condition 3, with the  pulse
combustion system in operation, was per-
formed at nominally the same (just slightly
higher)  feedrate  as Test  Condition 2, as
planned. The  Test Condition 3 feedrate
was 21 percent greater than the Test Con-
dition 1  feedrate.  A further  13-percent
feedrate increase  over that for Test Con-
dition 2  was possible before incinerator
operation  entered the borderline  accept-
able regime with the pulse combustion
system in operation. Thus, with respect to
the Sonotech claim that increased  incin-
erator capacity can be realized with pulse
combustion, test data show that a capac-
ity increase in the range of 13 percent
(comparing Test Condition 4 to Test Con-
dition 2) can indeed be realized.
  The data in Table 1 further show that
the total  system heat input needed to main-
tain target incineration temperatures was
relatively constant for all four test condi-
tions at  about 645 kW (2.2 MBtu/hr). Spe-
cifically  comparing the auxiliary fuel use
for Test Condition 3 to that for Test Con-
dition 2  shows that the auxiliary  fuel re-
quirements  were  nominally  the  same.

-------
Table 1.   Test Condition Operating Data
                                                                Test condition average (3 tests)
Parameter
Waste feedrate, kg/hr (Ib/hr)
1
Conventional
baseline
27.7 (61.0)
2
Conventional
max. feed
33.1 (72.8)
3
Pulsations
feed as in 2
33.5 (73.6)
4
Pulsations
max. feed
37.5 (82.4)
Kiln

    Gas temperature, °C (°F)            935  (1,720)

    Solids bed temperature, °C (°F)       965  (1,770)

Afterburner exit gas temperature,         1,095  (2,000)
°C (°F)
                  940  (1,730)

                  990  (1,810)

                 1,095 (2,000)
 925  (1,700)

1,020 (1,870)

1,095 (2,000)
 925  (1,700)

1,030 (1,890)

1,095 (2,000)
Afterburner exit O2, %
Heat input, kW (kBtu/hr)
Waste feed
Kiln auxiliary fuel
Main burner
Sonotech burner
Total kiln
Afterburner auxiliary fuel
Total auxiliary fuel
Total system heat input, kW (kBtu/hr)
Auxiliary fuel consumption per unit of
waste treated, MJ/kg (Btu/lb)
9.3

153 (522)

193 (659)
0 (0)
193 (659)
297 (1,012)
489 (1,668)
642 (2,190)
63.6 (27,400)
9.3

176 (601)

148 (506)
0 (0)
148 (506)
303 (1,035)
454 (1,551)
630 (2,152)
49.2 (21,100)
8.7

184 (628)

82 (282)
59 (200)
141 (482)
321 (1,094)
462 (1,576)
646 (2,204)
49.7 (21,400)
8.5

204 (697)

60 (205)
59 (200)
119 (405)
317 (1,082)
436 (1,487)
640 (2,184)
41.8 (18,000)
Thus, the Sonotech claim that decreased
auxiliary fuel  use would be possible with
the application of pulse combustion is not
supported by the test data. However, be-
cause  the  waste treated  in these tests
had significant heat content, the capacity
increase noted above equates to a corre-
sponding decrease in  the auxiliary fuel
consumed per unit of waste treated. Com-
paring  the auxiliary fuel consumption per
unit of waste treated for Test Condition 4
to that for Test Condition 2 shows that the
feedrate increase allowed by the Sonotech
system yields a  corresponding decrease
in auxiliary  fuel   use  per  unit of waste
treated from 49.2 MJ/kg (21,200 Btu/lb) to
41.9 MJ/kg (18,000 Btu/lb). Visual obser-
vations suggest that the Sonotech system
produced visually improved mixing in the
kiln chamber.
  With respect to combustion air require-
ments, the data in Table 1  show that less
combustion  air was required,  when  the
same feedrate was used,  for the pulse
combustion test condition compared to the
conventional combustion test condition.
POHC DREs and
Quality
  The POHC DREs and emission rates at
the two flue  gas locations sampled  are
summarized in Table 2. As shown  in the
table, naphthalene DREs measured at both
the afterburner exit and the baghouse exit
were uniformly 99.999  percent or greater
for all tests, and were not affected by the
different test conditions or different waste
feedrates. Benzene DREs  measured at
the  two   locations   were   uniformly
99.994 percent or greater (one baghouse
exit  benzene ORE measurement was
99.989 percent),  and again  not affected
by the different test conditions.
  The average naphthalene emission rate
at the afterburner exit was  reduced from
1.2 mg/hr (range of 0.4 to  2.6 mg/hr), for
conventional  combustion  at Test  Condi-
tion 2,  to  1.1  mg/hr (range  of <0.3 to
2.5 mg/hr),  with the Sonotech system at
the Test Condition  3. The  average ben-
zene emission rate at the afterburner exit
was reduced from 7.7 mg/hr (range of 2.1
to 12), for Test Condition 2, to 5.7 mg/hr
(range of 3.4 to 6.9), for Test Condition 3.
However, the significance of the decreases
is also difficult to judge because both fall
 within the precision of the respective flue
 gas concentration measurement methods.
 With  respect to this,  it  bears  noting that
 benzene emissions increased from Test
 Condition 2 to Test Condition 3 when mea-
 sured at the baghouse exit.
   The kiln  ash heating value data shown
 in Table 1  suggest that incineration resi-
 due quality, as measured by residue (kiln
 ash)  heating  value, was improved with
 pulse combustion. A decrease in  kiln ash
 heating value from 3.1 MJ/kg  (1,340 Btu/
 Ib), for Test Condition 2, to  <1.1 MJ/kg
 (<500 Btu/lb),  for Test Condition 3, at the
 same nominal feedrate but with pulse com-
 bustion,  was seen. The solids bed tem-
 perature data  shown  in the table are
 consistent with this decrease.  Solids bed
 temperatures were measured at four axial
 locations in the kiln during the tests. The
 temperature at the  location  recording the
 peak temperature for each test was aver-
 aged over the flue gas sampling period
 for that test. The entries in Table  1 repre-
 sent  the average of these  individual test
 averages for the three test  runs  at each
 test condition. The data show an increase
 in average  peak solids bed temperature
 of from 965°C (1,770°F) (Test Condition 1)

-------
Table 2.  Test Program POHC ORE Summary
                                        Benzene
Naphthalene
Afterburner exit

Condition 1
Testl
Tests
Test 10
Average



Condition 2
Test 2
Test?
Test 11
Average



Condition 3
Test3
Tests
Test9
Average



Condition 4
Test 4
Tests
Test 12
Feed-
rate
g/hr

253.1
253.1
244.1





298.3
307.3
289.2



7

289.2
307.3
307.3



7

343.5
334.4
334.4
Emission
rate,
mg/hr

44
7.6
14.8
8.9
99.99
4
99.99
6

9.0
2.1
12.0
7.7
99.99
5
99.99


6.9
3.4
6.7
5.7
99.99
7
99.99


10.4
11.7
50.9
ORE,
%

99.99
8
99.99
6




99.99
6
99.99
9




99.99
7
99.99
8




99.996
99.996
99.998
Baghouse exit
Emission
rate,
mg/hr

<1.2
2.1
3.4
2.2




31.0
<0.9
0.6
1.5




6.4
1.5
2.9
3.6




2.5
<1.S
1.1
ORE,
%

>99.999
99.999
99.998
99.999




99.989
>99.999
99.999
99.996




99.997
99.999
99.999
99.998




99.999
>99.999
99.999
Feed-
rate
g/hr

378.0
378.0
364.5





445.5
459.0
432.0





432.0
459.0
459.0





513.0
499.5
499.5
Afterburner exit
Emission
rate,
mg/hr

6.2
2.9a
<0.3
3.1




2.6a
0.6a
0.4a
1.2




2.5a
<0.3
0.5a
1.1




0.6b
0.5b
1.3"
ORE,
%

99.998
99.999
>99.999
99.999




99.999
99.999
99.999
99.999




99.999
>99.999
99.999
99.999




99.999
99.999
99.99
Baghouse exit
Emission
rate,
mg/hr

5.9
3.1a
2.5a
3.8




6.0
<0.3
3.5a
3.3




2.4a
0.6a
1.6a
1.5




1.4"
2.2b
0.4b
ORE,
%

99.998
99.999
99.999
99.999




99.999
>99.999
99.999
99.999




99.999
99.999
99.999
99.999




99.999
99.999
99.999
aAverage concentration of three pairs of M0030 VOST tubes.
bAna!yte         below lowest calibrated level.
"<" = Analyte below method detection limit.

-------
to 990°C (1,810°F) (Test Condition 2), for
conventional  combustion,  to  1,020°C
(1,870°F)  (Test  Conditions) to 1,030°C
(1,890°F)  (Test  Condition 4),  for  pulse
combustion. Specifically,  comparing  the
data for Test Condition 3 to those for Test
Condition 2 shows an increase from 990°C
(1,810°F), for conventional combustion, to
1,020°C (1,870°F), for pulse combustion,
at the same waste  feedrate and  other
system  operating conditions. These data
suggest that  the Sonotech  claim  of in-
creased heat transfer rates with pulse com-
bustion, specifically to the solids bed, is
justified. An increase in heat transfer rate
to the solids  bed would cause the in-
creased bed temperatures seen and the
corresponding decrease in  kiln  ash dis-
charge heating value.

CO, NOX,      Soot
  Table 3  summarizes  the continuous
emissions monitor (CEM) and soot emis-
sions data for the test program. As for the
operating conditions data, summarized in
Table 1, CEM  readings were recorded at
nominally 30~second intervals on the RKS
data acquisition  system. These  readings
were averaged over the flue gas sampling
period for each test  The CEM entries in
Table 3  represent the  average of each
test's average for the three  tests at each
test condition.
  The data in Table 3 show that average
kiln exit CO levels substantially increased
with  pulse  combustion,  from 68 ppm at
7 percent O2 for the two conventional com-
bustion  test  conditions  (1  and  2),  to
117 ppm at 7 percent O2 for Test  Condi-
tion 3 and to 153 ppm at 7 percent  O2 for
Test Condition 4. This increase is consis-
            tent with both the kiln solids bed tempera-
            ture and the kiln ash residue quality data
            in Table 1.  As  discussed  above,  pulse
            combustion  caused increased  kiln  solids
            bed temperatures and decreased kiln ash
            heating values, suggesting that pulse com-
            bustion caused a greater degree of waste
            feed organic content devolatilization into
            the kiln combustion gas. The observation
            that kiln exit CO  levels  were  increased
            with pulse combustion  suggests that the
            greater amounts of devolatilized organics
            were not completely destroyed  in the kiln.
              Average afterburner exit CO levels were
            decreased  to  15 ppm  at  7 percent  O2
            (range of 8.7 to 25.6 ppm),  for Test Con-
            dition 1,  and to 20 ppm  at 7 percent  O2
            (range of 8.0 to 40.0 ppm),  for Test Con-
            dition 2.  Compared to conventional  com-
            bustion, pulse combustion  produced
            slightly lower average afterburner exit CO
            levels. Comparing Test Condition 3  (pulse
            combustion) to Test Condition 2 (with con-
            ventional combustion),  both of which had
            the same waste feedrate, shows that pulse
            combustion  resulted in  decreased aver-
            age afterburner exit CO  emissions of
            14 ppm at 7 percent O2 (range of 12.6 to
            16.0 ppm. Even at the increased  waste
            feedrate achieved with pulse combustion
            for  Test  Condition 4, afterburner exit CO
            levels  were  only marginally increased, to
            17 ppm at 7 percent O2 (range of 10.5 to
            26.2 ppm) — higher than the Test Condi-
            tion 3  level,  but still lower than the Test
            Condition 2 level.
              CO  is the final incomplete combustion
            product in the series of reactions that con-
            verts the carbon in organic constituents to
            CO2. Thus, an  explanation  for  why after-
            burner exit CO  levels  under pulse  corn-
Table 3.   Continuous Emissions Monitor Data
                                       Test condition average (3 tests)
    Constituent
                          Conventional
                           baseline
           Conventional
            max. feed
           Pulsations
           feed as in 2
              4
          Pulsations
          max. feed
Kiln exit:

  CO, ppm at 7% O2

Afterburner exit:

  CO, ppm at 7% O2

  NOX, ppm at 7% O2

  Soota, mg/dscm at 7% O2

Baghouse exit:

  NOt, ppm at 7% O7
68



15

90

<1.3b
68



20

82

 1.9



85
117



 14

 77

 <1.



 78
153



 18

 78

  1.3



 72
'Measured as TOC in participate.
bMeasured for only one test.
bustion operation were lower than under
conventional combustion operation, while
kiln exit levels were higher,  may be that
organic constituent combustion in the kiln
was more complete under pulse combus-
tion operation. More complete organic con-
stituent combustion can result in higher
CO (the final incomplete combustion prod-
uct) levels, while other unburned hydro-
carbon levels, including soot,  would be
decreased. In such cases, the burden on
the afterburner to carry the destruction
process to completeness would be less-
ened,  resulting  in  lower afterburner exit
CO levels.
  The afterburner exit soot emissions data
(measured as TOC in the afterburner exit
particulate) show a consistent pattern. The
soot emission levels given in Table 3 again
represent the average of the levels mea-
sured  for each of the three tests at each
respective condition, with the exception of
the level  noted for Test Condition 1. The
afterburner exit  particulate  for only one
Test Condition 1  test  was  analyzed for
TOC,  so  the Test Condition 1  value in
Table  3  reflects  only this one measure-
ment.  Soot emission levels were <1.3 mg/
dscm at 7 percent O2 for Test Condition 1,
the baseline, conventional combustion test
condition. They were increased, to 1.9 mg/
dscm  at  7 percent O2 (range of <0.9 to
2.7), for Test Condition 2. However, with
pulse  combustion  at the same feedrate
for Test Condition  3, soot emissions  de-
creased to <1.0 mg/dscm at 7 percent O2
(range of<0.8 to 0.9). Even at the in-
creased waste feedrate achieved for Test
Condition 4, afterburner  exit soot  emis-
sions, at 1.3 mg/dscm (range of <0.9 to
1.8), were  less than those measured for
Test Condition 2.
  Afterburner and  baghouse exit  NOX
emissions were comparable from test con-
dition  to test condition and were 90 and
88 ppm at 7 percent O2 (ranges of 85.5 to
93.9 and  85.8 to 91.1),  respectively, for
Test Condition 1, and a slightly decreased
82 and 85 ppm at 7 percent O2,  (ranges of
78.3 to 85.1  and 83.6 to 86.3) for Test
Condition  2.  Levels  of  afterburner and
baghouse exit NOX were, respectively, 77
and 78 ppm at 7 percent O2 (ranges of
68.0 to 87.1 and 76.6 to 79.6), for Test
Condition 3 (under pulse combustion), and
78 and 72 ppm at 7 percent O2  (ranges of
73.7 to 81.4 and 68.7 to 75.5), for Test
Condition  4 (also  under pulse combus-
tion).  Although the Sonotech claim that
pulse  combustion would  result  in  de-
creased NOX emissions was confirmed by
the test data, the  reductions achieved were
small,  and from low initial NOX levels.

-------
Composition
  No PAH  or BTEX constituent, with the
exception of benzene in a few cases, was
detected in  any kiln ash or scrubber liquor
sample from any test.  Detection limits for
PAHs were 0.1 to 0.3 mg/kg,  in kiln ash,
and 1 to 3 |ig/L, in scrubber liquor. Detec-
tion limits for BTEX constituents were 1 to
10 mg/kg, in kiln ash, and 1 to 10 (ig/L, in
scrubber liquor.  Benzene was  found in
the kiln  ash samples from one Test Con-
dition 2 test,  from one Test  Condition 3
test, and from all three Test  Condition 4
tests, but at levels only slightly above the
method  detection limit  of 1 mg/kg.
  Incinerator feed, flue gas, and residue
discharge samples were analyzed for an-
timony, barium, beryllium, cadmium, chro-
mium, lead, and mercury. No antimony or
mercury was found in any test program
sample. The data show that the  distribu-
tion of these metals in the kiln ash dis-
charge did not vary from test condition to
test condition.  Concentrations of  barium
and chromium  in the scrubber liquor were
slightly  lower and  in  the  baghouse  exit
flue gas were  higher  for  the  pulse com-
bustion operating conditions.
  The concentrations of the five metals in
TCLP leachates of test feed, kiln ash, and
scrubber liquor samples ranged from not
detected to low.  At  the  concentrations
measured, neither the test waste feed or
the incineration residual discharges would
be  a toxicity characteristic (TC)  hazard-
ous waste for  any  test condition, and no
significant variation in leachate concentra-
tions with test conditions was seen.

Dioxin
  Chlorinated  dioxins and furans were
measured in the baghouse exit flue gas
for  all tests. Test condition average flue
gas  concentrations are  summarized  in
Table 4  in terms of the two summary con-
centration measures commonly reported:
total PCDD/PCDF and 2,3,7,8-TCDD tox-
icity equivalents  (TEQs). The entries in
Table 4  are cited  as ranges.  This arises
from the fact that, in many cases, a con-
Table 4.   Dioxin/Furan Emission Results
         gener group or specific isomer concentra-
         tion is reported as not detected. Thus, for
         concentrations  reported as a  range, the
         low value corresponds to  assuming con-
         stituents not detected were present at zero
         concentration; the high value corresponds
         to  assuming  these  constituents  were
         present at their detection limit.
           The data  in Table 4 show that total
         PCDD/PCDF emissions at the baghouse
         exit were quite low, at  0.15  ng/dscm at
         7 percent O2 or  less, for all test  condi-
         tions, and  were not affected  by the test
         condition differences. On  a TEQ  basis,
         emissions  were 0.005 ng/dscm at  7 per-
         cent O2, or  less, and,  again, were not
         affected  by the test condition differences.

         Conclusions
           The objectives of this test program were
         to develop test data to permit the evalua-
         tion of Sonotech's claims regarding the
         performance of the Sonotech pulse com-
         bustion technology applied to a hazard-
         ous waste  incinerator in comparison with
         the  performance of conventional combus-
         tion technology. In evaluating the claims,
         test data were evaluated  using the rank
         sum test.  The rank sum  test allows an
         assessment of whether observed  differ-
         ences in data sets are statistically signifi-
         cant. When comparing two data sets, each
         containing  three data points, the two data
         sets are different, at the 95 percent confi-
         dence level, when there is no data  over-
         lap. With respect to the  Sonotech claims,
         test data show that the application of the
         Sonotech  pulse  combustion  technology
         resulted  in:
               increased incinerator capacity. Ap-
               plication of the Sonotech pulse com-
               bustion  system allowed a  waste
               feedrate increase  of  13  percent
               compared to corresponding operat-
               ing  conditions  under conventional
               combustion. Because the  waste
               treated had  significant  heat con-
               tent, the capacity  increase equated
               to a corresponding decrease in the
               auxiliary fuel consumption per unit
               of waste treated  from an  average
                                          Condition average concentration,
                                                ng/dscm at 7% O2
Test condition
                               Total PCDD/PCDF
                                                                     TEQ
Condition 1

Condition 2

Condition 3

Condition 4
 0.13to 0.15

0.092 to 0.099

 0.10 to 0.11

 0.13to0.14
0.0004 to 0.0054

0.0004 to 0.0050

0.0003 to 0.0041

0.0006 to 0.0046
of 21.1 kBtu/lb  (range  of  21.0  to
21.3kBtu/lb) to  an average  of
18.0 kBtu/lb (16.6 to 19.0 kBtu/lb).
Increased POHC  ORE.  POHC
DREs measured for all test condi-
tions were uniformly 99.994 percent
or greater, with no observable ben-
efit, or detriment, due to pulse com-
bustion. Benzene emission rates at
the  afterburner  exit  were reduced
from an average of 7.7 mg/hr (range
of 2.1  to 12) to an  average  of
5.7  mg/hr (range of 3.4 to  6.9).
Naphthalene emission rates at the
afterburner exit were reduced from
an average of 1.2 mg/hr (range  of
0.4 to 2.6) to an average of 1.1 mg/
hr (range of <0.3 to 2.5).
Decreased flue gas  CO emissions.
Average  afterburner exit flue gas
CO  levels were reduced,  from
20 ppm at 7 percent O2 (range  of
8.0  to 40.0) in a  maximum waste
feedrate operating condition  under
conventional combustion operation,
to an average of 14 ppm at 7 per-
cent O2 (range of 12.6 to  16.0)  at
the  same feedrate with  pulse com-
bustion. Pulse combustion  allowed
a higher  waste  feedrate to be
achieved, with average afterburner
exit  CO  emissions  of  18 ppm  at
7 percent O2 (range of 10.5 to 26.0),
still  below the conventional  com-
bustion maximum feedrate condi-
tion.
Decreased flue gas NOX emissions.
Both afterburner exit and baghouse
exit NOX emissions were marginally
decreased from 82  (range of 78.3
to 85) to 85 (range of 83.6  to 86.3)
ppm at 7 percent O2, respectively,
in a conventional combustion maxi-
mum waste feedrate operating con-
dition, to  averages of 77 (range  of
68.0 to 87.1) and 78 (range of 76.6
to 79.6) ppm at 7 percent O2  at the
same feedrate with pulse combus-
tion. At the higher waste feedrate
achievable with  pulse  combustion,
average  NOX emissions were 78
(range  of 73.7 to  81.4)  and
72 (range  of 68.7 to 75.5) ppm  at
7 percent O2.
Decreased flue gas soot emissions.
Average  afterburner exit flue gas
soot emissions,  measured  as total
organic carbon or particulate, were
decreased from  1.9mg/dscm  at
7 percent O2 (range of <0.9 to 2.7),
in the  conventional  combustion
maximum waste feedrate operating
condition,  to <1.0 mg/dscm (range
of <0.8 to 0.9), with  pulse combus-
tion at the same feedrate.  Even  in

-------
      the increased-feedrate pulse com-
      bustion  test condition, afterburner
      exit soot emissions were still lower,
      at an average of 1.3mg/dscm at
      7 percent O2 (range of <0.9 to 1.8),
      than in  the conventional combus-
      tion maximum feedrate condition.
      Decreased combustion air require-
      ments. Total system combustion air
      requirements  decreased from  an
      average of 39,500 dscf/hr (range
      of 38,400 to 40,600) under con-
      ventional combustion to an  aver-
      age  of  37,600 dscf/hr (range of
      34,800 to 39,900) with pulse com-
      bustion, based on stoichiometric cal-
      culations.
      Decreased auxiliary  fuel require-
      ments. No  measurable change in
      auxiliary fuel  requirements  to es-
      tablish a given  set of combustion
      conditions was observed.
  A secondary test program objective was
distribution of hazardous constituent trace
metals among the  incineration system dis-
charge streams. Test data show that no
  S.  Venkatesh, W. E. Whitworth, Jr., C.  Goldman, and L. R. Water/and are with
   Acurex Environmental Corporation, Jefferson, AR 72079.
  R.  C. Thurnau is the EPA Project Officer; M. K. Richards is the SITE Project
   Manager, (see below).
  The complete report, entitled "SITE Program Evaluation of the Sonotech  Pulse
   Combustion Burner Technology,"(OrderNo. PB97-189732; Cost: $35.00, sub-
   ject to change)  will be available only from
          National Technical Information Service
          5285 Port Royal Road
          Springfield, VA 22161
          Telephone: 703-487-4650
  The EPA Project Officer and Project Manager can be contacted at
          National Risk Management Research Laboratory
          U.S. Environmental Protection Agency
          Cincinnati, OH 45268
significant differences  in metals distribu-
tions among the test  conditions are ap-
parent, with the exception that barium and
chromium concentrations  were  slightly
lower in  the  scrubber liquor,  and  were
higher  in  the baghouse exit flue gas with
pulse combustion.
  Although not the subject of the initially
defined project objectives,  heating value
determinations performed on kiln ash  resi-
due samples show that pulse combustion
operation improved residue quality by pro-
ducing  a  kiln ash discharge with lower
heating value than that measured  in kiln
ash from  conventional  combustion opera-
tion.  The  observation that pulse combus-
tion operation caused increased  kiln solids
bed temperatures, likely the result of im-
proved heat transfer to the solids bed,  is
consistent with the lower-heating-value kiln
ash observed.
  The  full report was  submitted in fulfill-
ment of Contract  No.  68-C9-0038, Work
Assignments 3-4 and 4-4, by Acurex En-
vironmental Corporation under the spon-
sorship  of the  U.S.  Environmental
Protection Agency.
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
                                                           BULK RATE
                                                     POSTAGE & FEES PAID
                                                              EPA
                                                        PERMIT No. G-35
EPA/600/SR-97/061

-------