EPA/600/R-94/091
June 1994
OPERATIONS AND RESEARCH AT THE U.S. EPA
INCINERATION RESEARCH FACILITY:
ANNUAL REPORT FOR FY93
By
L. R. Waterland
Acurex Environmental Corporation
Incineration Research Facility
Jefferson, Arkansas 72079
EPA Contract 68-C9-0038
EPA Project Officer: Robert C. Thurnau
Waste Minimization, Destruction, and Disposal Research Division
Risk Reduction Engineering laboratory
Cincinnati, Ohio 45268
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
Printed on Recycled Paper
-------
NOTICE
The information in this document has been funded wholly or in part by the United
States Environmental Protection Agency under Contract 68-C9-0038 to Acurex Environmental
Corporation. It has been subjected to the Agency's peer and administrative review, and it has
been approved for publication as an EPA document. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
11
-------
FOREWORD
Today's rapidly developing and changing technologies and industrial products and
practices frequently carry with them the increased generation of materials that, if improperly
dealt with, can threaten both public health and the environment. The U.S. Environmental
Protection Agency is charged by Congress with protecting the Nation's land, air, and water
resources. Under a mandate of national environmental laws, the Agency strives to formulate and
implement actions leading to a compatible balance between human activities and the ability of
natural systems to support and nurture life. These laws direct the EPA to perform research to
define our environmental problems, measure the impacts, and search for solutions.
The Risk Reduction Engineering Laboratory is responsible for planning, implementing,
and managing research, development, and demonstration programs to provide an authoritative,
defensible engineering basis in support of the policies, programs, and regulations of the EPA
with respect to drinking water, wastewater, pesticides, toxic substances, solid and hazardous
wastes, and Superfund-related activities. This publication is one of the products of that research
and provides a vital communication link between the researcher and the user community.
This document reviews the accomplishments at the Incineration Research Facility (IRF)
in Jefferson, Arkansas, during Fiscal Year 1993. In the 12-month period, two major test
programs were completed at the facility. The major EPA program supported through test
activities was the Superfund site remediation program within the Office of Emergency and
Remedial Response (OERR) as administered by OERR and EPA Region 3. In addition, a third
major test program was completed in support of the Department of Energy's efforts to install
and permit a mixed-waste incineration system at the Savannah River Plant. The report outlines
all efforts completed or ongoing at the facility during FY93.
-------
ABSTRACT
The U.S. Environmental Protection Agency's Incineration Research Facility (IRF) in
Jefferson, Arkansas, is an experimental facility that houses a pilot-scale rotary kiln incineration
system (RKS) and the associated waste handling, emission control, process control, and safety
equipment; as well as onsite laboratory facilities.
During fiscal year 1993, two major test programs were completed at the IRF: an
evaluation of rotary kiln incinerator operation at low to moderate temperatures, and a series of
tests in which simulated mixed wastes were incinerated to support the Westinghouse Savannah
River Company.
Results of a pilot-scale test program previously completed, a parametric evaluation of
the fate of trace metals fed to a rotary kiln incinerator equipped with a Calvert Flux-Force/
Condensation scrubber system, were reported during FY93.
Detailed plans were developed for four test programs to be completed in FY94: an
evaluation of the incinerability of waste and contaminated soil from the M. W. Manufacturing
Superfund site; an evaluation of a pulse combustion burner technology under the Superfund
Innovative Technology Evaluation (SITE) program; a program demonstrating that U.S. and
Russian Federation environmental regulations can be complied with during the incineration of
liquid ballistic missile propellant components; and an evaluation of candidate additives as trace
metal sorbents for incineration applications.
Finally, a fabric filter air pollution control system, including flue gas reheat, was
incorporated into the RKS.
This report summarizes all efforts completed or ongoing at the IRF during FY93.
IV
-------
CONTENTS
Section
NOTICE if
FOREWORD Hi
ABSTRACT iv
FIGURES viii
TABLES . x
1 INTRODUCTION . 1
2 FATE OF TRACE METALS IN THE ROTARY KILN INCINERATION
SYSTEM WITH A CALVERT FLUX-FORCE/CONDENSATION
SCRUBBER 3
2.1 TEST PROGRAM . r ........ 4
2.1.1 Synthetic Waste Mixture 4
2.1.2 Test Conditions 4
2.2 TEST RESULTS , 4
2.2.1 Average Trace Metal Discharge Distributions 4
2.2.2 Effects of Incinerator Operating Conditions on Metal
Distributions 10
2.2.3 Scrubber Collection Efficiencies 15
2.3 CONCLUSIONS 25
3 EVALUATION OF ROTARY KILN INCINERATOR OPERATION AT
LOW TO MODERATE TEMPERATURES 26
3.1 TEST PROGRAM 27
3.1.1 Test Contaminated Soil 27
3.1.2 Test Conditions 28
3.1.3 Sampling and Analysis 32
3.2 CURRENT STATUS 34
-------
CONTENTS (continued)
Section
4
INCINERATION OF SIMULATED MIXED WASTE .................... 35
4.1 TEST PROGRAM .................................... ....... 35
4.1.1 Test Waste Description .................................. 36
4.1.2 Test Conditions ............................ . ........ ---- 37
4.1.3 Sampling and Analysis ................................... 42
4.2 TEST RESULTS ............................... ....... ............ 45
4.3 CONCLUSIONS ............. . ..... . .......... .............. 52
TESTING OF FLUFF WASTE AND CONTAMINATED SOIL FROM
THE M. W. MANUFACTURING SUPERFUND SITE .............. ...... 54
5.1 BACKGROUND ...... . ...................... ........ ...... 54
5.2 TEST PROGRAM .......................................... 56
5.2.1 Waste Description ................ .......... . . ....... - 56
5.2.2 Test Conditions ........................... ............. 59
5.2.3 Sampling and Analysis Procedures .......... ................ 61
5.3 CURRENT STATUS ........................................ 63
EVALUATION OF THE SONOTECH FREQUENCY-TUNABLE PULSE
COMBUSTION TECHNOLOGY ........................ ............. 67
6.1 DESCRIPTION OF THE TECHNOLOGY ....................... 67
6.2 TEST PROGRAM ..................................... ...... 69
6.2.1 Test Facility ............ , .............................. 71
6.2.2 Test Feed Material ................... ................. 72
6.2.3 Test Conditions ...... ______ .............................. 72
6.2.4 Sampling and Analysis Procedures .......................... 75
6.3 CURRENT STATUS ............... ----- ..................... - 77
VI
-------
CONTENTS (concluded)
Section
7 TEST INCINERATION OF BALLISTIC MISSILE PROPELLANT
COMPONENTS 80
7.1 TEST PROGRAM . 81
7.1.1 Environmental Regulations < 81
7.1.2 Test Conditions . ., 85
7.1.3 Sampling and Analysis Procedures 86
7.2 CURRENT STATUS 89
v
8 EVALUATING THE EFFECTIVENESS OF ADDITIVES AS SORBENTS FOR
METAL CAPTURE 92
8.1 TEST FACILITY DESCRIPTION 93
8.2 TEST PROGRAM 95
8.3 CURRENT STATUS 96
9 INSTALLATION OF A FABRIC FILTER AIR POLLUTION CONTROL
SYSTEM WITHIN THE IRF RKS . 100
10 EXTERNAL COMMUNICATIONS 103
11 PLANNED EFFORTS FOR FY94 107
REFERENCES 108
Vll
-------
FIGURES
Number Page
1 Schematic of the IRF rotary kiln incineration system 5
2 Schematic of the Calvert flux-force/condensation scrubber system 6
3 Distributions of metals in the discharge streams expressed as a fraction of the
metal fed 8
4 Normalized distributions of metals in the discharge streams 9
5 Cadmium distributions expressed as a fraction of cadmium fed, showing
variations with kiln exit temperature at constant waste feed chlorine
content 11
6 Normalized cadmium distributions, showing variations with kiln exit
temperature at constant waste feed chlorine content 12
7 Cadmium distributions expressed as a fraction of the cadmium fed,
showing variations with waste feed chlorine content at constant kiln exit
temperature 13
8 Normalized cadmium distributions, showing variations with waste feed
chlorine content at constant kiln exit temperature 14
9 Copper distributions expressed as a fraction of the copper fed, showing
variations with kiln exit temperature at constant waste feed chlorine
content 16
10 Normalized copper distributions, showing variations with kiln exit
temperature at constant waste feed chlorine content 17
11 Copper distributions expressed as a fraction of the copper fed, showing
variations with waste feed chlorine content at constant kiln exit temperature ... 18
12 Normalized copper distributions, showing variations with waste feed chlorine
content at constant kiln exit temperature 19
13 Lead distributions expressed as a fraction of the lead fed, showing variations
with kiln exit temperature at constant waste feed chlorine content 20
Vlll
-------
FIGURES (concluded)
Number
14
Page
Normalized lead distributions, showing variations with kiln exit temperature
at constant waste feed chlorine content 21
15 Lead distributions expressed as a fraction of the lead fed, showing variations
with waste feed chlorine content at constant kiln exit temperature 22
16 Normalized lead distributions, showing variations with waste feed chlorine
content at constant kiln exit temperature 23
17 Apparent Calvert scrubber collection efficiencies for metals 24
18 Schematic of the IRF rotary kiln incineration system 60
19 Test sampling locations -. , 62
20 Test sampling locations j 76
21 Test sampling locations 88
22 The IRF TTU 94
23 Baghouse APCS schematic 101
IX
-------
TABLES
Number
1 Average organic concentrations in the waste feed 6
2 Average integrated feed metal concentrations 7
3 Average achieved values for the test variables 7
4 Organic constituents in the synthetic contaminated soil 29
5 Trace metal constituents in the synthetic contaminated soil 30
6 Target test conditions 31
7 Actual test operating conditions 33
8 Hazardous constituent metals spiked in waste E 36
9 Target test conditions 38
10 APCS operating conditions 39
11 Actual test conditions 40
12 Solids bed temperatures 43
13 Collected ash weights 44
14 Afterburner exit particulate load 46
15 POHC DREs 47
16 Flue gas PCDD/PCDF analysis results 49
17 Trace metal concentrations in feed and discharge samples 50
18 Trace metal distributions 51
-------
TABLES (concluded)
Number page
19 M. W. Manufacturing site waste contaminants from the ROD 57
20 M. W. Manufacturing site characterization sample analysis results 58
21 M. W. Manufacturing site characterization sample hazardous waste
characteristics analysis results 59
22 Test matrix gj
23 Analysis protocol 54
24 Analysis results for the test material feed preparation 73
25 . Analysis protocol , ; 78
26 Russian Federation environmental regulations for UDMH incineration 83
27 European hazardous waste incinerator emission limits 83
28 Expected approximate test conditions 87
29 Analysis protocol 90
30 Trace metals to be determined in selected test program samples 91
31 Approximate sorbent mineral composition 95
32 Test matrix 97
33 Metal spike solution concentrations 99
34 Test program metal dispersion concentrations 99
35 Design characteristics of the RKS baghouse collector 102
36 IRF program reports and presentations in FY93 . . . 104
37 Visitors to the IRF 105
xi
-------
-------
SECTION 1
INTRODUCTION
The U.S. Environmental Protection Agency (EPA) Incineration Research Facility (IRF)
in Jefferson, Arkansas, is an experimental facility that currently houses a pilot-scale rotary kiln
incineration system (RKS) and its associated waste handling, emission control, process control,
and safety equipment, and a bench-scale thermal treatment unit (TTU) for performing thermal
treatability studies on a smaller scale. The IRF also has onsite laboratory facilities for waste
characterization and analysis of process performance samples.
The objective of research projects conducted at the IRF have been and continue to be
as follows:
To develop technical information on the performance capabilities of the hazardous
waste incineration process to assist EPA Regional Offices and state environmental
agencies in the review, assessment, and issuance of reasonable and responsible
permits for regulated hazardous waste incineration facilities, and to assist waste
generators and incinerator operators in the preparation of permit applications
To develop incinerator system performance data for regulated hazardous wastes
to support current Resource Conservation and Recovery Act (RCRA) incinerator
regulations and performance standards, and to provide a sound technical basis for
any necessary future standards
To promote an understanding of the hazardous waste incineration process and '
develop methods to predict the performance of incinerators of varying scale and
design for the major classes of incinerable hazardous wastes as a function of key
process operating variables
To develop methods of improving reliability and control of the incineration
process, including the use of destruction and removal efficiency (DRE) surrogates
To provide a means of conducting specialized test burns (particularly for high-
hazard and special waste materials such as Superfund site wastes) in support of
specific Regional Office permitting or enforcement actions and Regional Office
or private party Superfund site remediation efforts
To test the performance of new and advanced incinerator components and
subsystems, and emission control devices
1
-------
Fiscal year 1993 (FY93, October 1, 1992 through September 30? 1993) saw the
continuation of routine incineration testing efforts at the IRF. During the year, two major
programs were completed and plans for an additional four test programs developed.
The major EPA program supported through test activities in FY93 was the Superfund
site remediation program with the Office of Emergency and Remedial Response (OERR). A
major test program to evaluate rotary kiln incinerator operation at low to moderate temperatures
in decontaminating soils contaminated with organic constituents was completed.
FY93 also saw the initiation, completion, and reporting of the first major test program
funded by a source other than EPA. This program, a series of tests feeding simulated mixed
waste, was performed for the Westinghouse Savannah River Company (WSRC), the operating
contractor for the Department of Energy (DOE) Savannah River Plant and Savannah River
Laboratory.
In addition, the results of a major series of tests to evaluate the fate of trace metals fed
to a. rotary kiln incinerator equipped with a Calvert Flux-Force/Condensation scrubber system,
completed in FY91, but with significant analytical effort extending through FY92, were assembled
and reported in FY93. The test planning documents (test plan outline, test plan, and/or Quality
Assurance Project Plan) for three major series of incineration tests and one series of tests in the
thermal treatment unit (TTU), all to be performed in FY94, were completed.
Finally, a major enhancement to facility capabilities was completed in FY93.
Specifically, a flue gas reheat and fabric filter air pollution control system (APCS) was installed
for use on the RKS.
Activities completed during FY93 are discussed in more detail in the following sections.
Section 2 discusses the results of the Calvert scrubber trace metal tests. Section 3 discusses the
low to moderate temperature kiln operation tests. Section 4 discusses the results of the WSRC
simulated mixed waste tests. Section 5 discusses the detailed test plan for the RKS test program
to evaluate the incinerability of fluff waste and contaminated soil from the M. W. Manufacturing
Superfund site in Danville, Pennsylvania. Section 6 covers the upcoming Superfund Innovative
Technology Evaluation (SITE) of a pulse combustion burner technology. Section 7 discusses the
upcoming program to demonstrate that U.S. and Russian Federation environmental regulations
can be complied with during the incineration of liquid ballistic missile propellant components.
Section 8 outlines the detailed plans for a series of tests in the TTU at the facility to evaluate
the effectiveness of candidate additives as trace metal sorbents. Section 9 discusses the fabric
filter APCS installed on the RKS. Section 10 discusses external communication activities
associated with the facility and its operation. Finally, Section 11 presents an outline of plans for
activities to be completed in FY94 other than those discussed in Sections 5 through 8.
-------
SECTION 2
FATE OF TRACE METALS IN THE ROTARY KILN INCINERATION SYSTEM
WITH A CALVERT FLUX-FORCE/CONDENSATION SCRUBBER
During FY88, a research program to develop data on the partitioning of trace metals
among the discharges from hazardous waste incinerators was initiated at the IRF. Over the
succeeding 4 years, three series of parametric tests were performed to quantify the distribution
of trace metals among the discharge streams of the IRF RKS, and to assess the effects of system
operation and feed composition on these distributions. Although certain details of the test
program design differed from series to series, the major differences between the test series were
the primary APCSs used for particulate and acid gas control. In the first series, completed in
FY88, a venturi scrubber/packed-column scrubber was used as the primary APCS. In the
second series, completed in FY89, a single stage ionizing wet scrubber was used.2
The third series of parametric trace metal partitioning tests investigated the use of a
Calvert Flux-Force/Condensation scrubber system for air pollution control. This series of tests
was completed during FY91. However, the trace metal analyses of the extensive number of test
program samples collected were not completely reported until mid-FY92. In addition, initial
evaluation of the original trace metal analysis data raised several questions regarding reported
concentrations. Consequently, several test program samples were reanalyzed. Reanalysis data
were received near the close of FY92. Thus, detailed data evaluation and reporting became
FY93 efforts; the results of these evaluations are summarized in this section.
The objectives of the third series of parametric trace metal tests were to identify:
The partitioning of metals among the kiln ash, the scrubber liquor, and the flue
gas discharges
The effects of kiln exit gas temperature and waste feed chlorine content on metal
partitioning
The efficiency of the Calvert scrubber in collecting flue gas metals
The effects of scrubber pressure drop on metal collection efficiencies
The test series to address these objectives consisted of 11 parametric tests. Two
additional baseline tests were also performed with no metals fed to determine the extent of
hysteresis in the RKS caused by test-to-test carryover of metals. Results of the test program are
discussed in the subsections that follow.
-------
2.1
TEST PROGRAM
All tests were conducted in the IRF RKS. The RKS as it was configured for these tests
is illustrated in Figure 1. Figure 2 is a schematic of the Calvert scrubber pilot plant installed on
the RKS for this test program.
2.1.1 Synthetic Waste Mixture
The synthetic waste contained a mixture of organic liquids added to an attapulgite clay
absorbent material. This attapulgite clay material was comprised of hydrated magnesium
aluminum silicate, free silica, dolomite, and calcite. Trace metals were incorporated by spiking
an aqueous mixture of the metals of interest onto the clay/organic-liquid material as it was fed
to the rotary kiln via a screw feeder.
The organic-liquid base consisted of toluene, with varying amounts of tetrachloroethene
and chlorobenzene added to provide a range of chlorine contents. Synthetic waste chlorine
content was varied from 0 to nominally 3.4 percent. The analyzed organic fractions for the three
waste feed mixtures are given in Table 1. .Table 2 summarizes the average metal concentrations
in the combined waste feed over the 11 parametric tests.
2.1.2 Test Conditions
The test variables were kiln exit gas temperature, chlorine content of the synthetic waste
feed, and scrubber pressure drop. Eleven specific combinations of these variables were selected
as test points. In addition, two baseline tests were conducted in which the clay/organic mixture
was fed, but the metals spike solution was not. Average achieved values for the three variables
for all 13 tests are summarized in Table 3. One of the baseline tests was conducted several days
before the test series began to establish baseline conditions for metals present in the incineration
system. The second baseline test was performed 2 days after the test series was completed. The
target conditions for the two baseline tests were the same as test condition 8. .
All tests were performed at the same nominal afterburner exit flue gas O2 (9 percent),
afterburner exit gas temperature (1,093°C [2,000°F ]), and synthetic waste feedrate (63.5 kg/hr
[140 Ib/hr]), of which 14 kg/hr (30 Ib/hr) was the organic-liquid mixture. For all tests, the solids
residence time in the kiln was about 1 hour. Average kiln exit flue gas O2 concentrations ranged
from 10.6 to 17.7 percent.
22.
TEST RESULTS
2.2.1 Average Trace Metal Discharge Distributions
Figure 3 shows the amount of each test metal, except mercury, found in each discharge
stream, as a fraction of the amount of metal fed. Figure 4 is a corresponding illustration of the
amount of test metal found in each discharge stream as a fraction of the total measured in the
three discharge streams: kiln ash, scrubber exit flue gas, and scrubber liquor. In the figures, the
tick marks at each end of the vertical line for each metal represent the range in the fraction
accounted for by each discharge stream over the 11 parametric tests. The average fraction over
all 11 tests is noted by the midrange tick mark. Metal discharge distribution data in the figures
-------
UJ O «
OC < L
IU 1- t
x w -
M-* 1 fxl 1
O rN I (
P ^
< ;
.
3
.\
) 1
y
-------
FROM IRF
HEAT *-+
EXCHANGER
FROM IRF
QUENCH __(
UNIT 1
TOIRF ^ .
HE A 1" -"1
EXCHANGER
TO
QUENCH -*-
SUMP
TO
SLOWDOWN
fr~~
.j
~L
/ s. COLLISION
1 1 SCRUBBER
Sp(7
_b
r-,1
PUMp CONDENSOR
ABSORBER
ENTRANMENT
SEPARATOR
DRAM
STORAGE ~*~^
TANK
H
'
H
J
s
TT
IIM
INI
mi
1 1 1 1
1 1 1 1
1 1 1 1
1 1 1 1
MM
\ /
65 KVDC
6 TO 7 SEC
mA 1
('
A ^
""ENTRANMENT
SEPARATOR (VAHlAt
, "*" v.^r
,
WET
ELECTROSTATC
PRECPITATOR
ARCS
Figure 2. Schematic of the Calvert flux-force/condensation scrubber system.
TABLE 1. AVERAGE ORGANIC CONCENTRATIONS IN THE
WASTE FEED
Weight % in mixture
Test
1 through 3
4 through 6
7 through 11
Toluene
20.2
18.3
16.2
Tetrachloroethene
0
0.6
2.7
Chlorobenzene
0
0.6
2.7
-------
TABLE 2. AVERAGE INTEGRATED FEED
METAL CONCENTRATIONS
Metal
Arsenic
Barium
Bismuth
Cadmium
Chromium
Copper
Lead
Magnesium
Strontium
Mercury
Concentration,
mg/kg
34
465
371
20
283
347
74
34,500
388
4
TABLE 3. AVERAGE ACHIEVED VALUES FOR THE TEST VARIABLES
Test
Baseline
1
2
3
4
5
6
7
8
9
10
11
Repeat
baseline
Kiln exit gas Feed mixture
Date temperature, °C (°F) chlorine content, %
5/29/91
6/5/91
6/6/91
6/13/91
6/18/91
6/19/91
6/21/91
6/25/91
6/28/91
7/9/91
7/11/91
7/16/91
7/18/91
831 (1,528)
541 (1,006)
819 (1,507)
909 (1,669)
555 (1,031)
842 (1,547)
919 (1,686)
543 (1,010)
817 (1,502)
944 (1,731)
829 (1,524)
827 (1,521)
834 (1,534)
2.8
0
0
0
0.6
0.6
0.8
3.6
3.4
3.1
2.3
3.4
2.3
Calvert scrubber pressure
drop, kPa (in WC)
12.4(50)
12.9 (52)
12.4 (50)
12.4 (50)
12.4 (50)
12.4 (50)
12.4 (50)
12.4 (50)
12.4 (50)
12.2 (49)
8.2 (33)
16.9 (68)
12.4 (50)
-------
KILN ASH
«iUU
o5 160
SEui 120
u.
O 3? 80
£ 40
0
0
-
Pb
_j- Cd -J '
--
200 400 600
--Cr
"As Mg
~t~Ba _i_sr~t~
+Cu _J_ ~ "
AEC 030-94
800 1000 1200 1400 1600 1800
VOLATILITY TEMPERATURE (°C)
5
OC
UJ
ffl *
g O a
||£ ^
O t ^* o
O liJ
E 1
0
SCRUBBER-EXIT FLUE GAS
-
"
Cd
- + Pb
I" 1 BiU-
200 400 600
-4- Cu ^
T Ba _L Sr Cr
^1 m * I ' M_|_II 1 ^nQ "T"
I I 1 I 1 | 1 | 1 ^ - II
800 1000 1200 1400 1600 1800
VOLATILITY TEMPERATURE (°C)
DC 50
§ 40
DC
Ul
SS 30
ra u.
o11-
2? 20
cc
LL. Q
SCRUBBER LIQUOR
Pb
Cd
Bi-4.
TT" 1 H&
200 400 600
Cu
I3 Ba
T "I" -- Sr Mg 9r
+ 1 i 1 -- 1 1 ± d=±
800 1000 1200 1400 1600 1800
VOLATILITY TEMPERATURE (°C)
Figure 3. Distributions of metals in the discharge streams expressed as a fraction
of the metal fed.
8
-------
KILN ASH
CO
< Q
_J DC
^c io
z <
UJ
Ou-
PO
Ovv$
cc
UL
IUU
90
80
70
60
Bi-
_Cd
:± JBa :
4- As
Sr M9 Cr
-Cu
_Pb
'
-
1 1 1 1 1 1 1 1
200 400 600 800 1000 1200
VOLATILITY TEMPERATURE (°C)
SCRUBBER-EXIT FLUE GAS
1400 1600 1800
cc
UJ
Igjjj
CO UJ CO
Z 3 UJ
z " 2
|ujj
cc
U_
3
4
3
2
1
f\
-
-
-
-
-Cd pb -
B''1
f-
-
0 200 400 600
-As
Sr
JBa _j_Cu V Cr
, T , , I M9 ±
III 1 T^ i TT"^
800 1000
VOLATILITY TEMPERATURE
1200 1400 1600 1800
(°C)
SCRUBBER LIQUOR
cc
O
O
_i
LU W
CQ £
m 7Z
§1
CO ^
Z ^
O
g o
o
1
u.
50
40
30
20
10
/\
~
M
_
T
-Cd
Bi-
i r
-Pb
. As
.T -j-Ba
--I- ,4= . :
u
0 200 400 600
800 1000
-Cu
Or
MQ ^*
I i i ± -4- i|rcr
1200 1400 1600 1800
VOLATILITY TEMPERATURE (°C)
Figure 4. Normalized distributions of metals in the discharge streams.
9
-------
are plotted versus the volatility temperature for each metal, based on the metal's elemental and
oxide forms. The volatility temperature3 is the temperature at which the effective vapor pressure
of the metal is 10"6 atm. The effective vapor pressure is the combined equilibrium vapor
pressures of all species containing the metal, reflecting the quantity of metal that would vaporize
under a given set of conditions. A vapor pressure of 10~6 atm was selected because it represents
a measurable amount of vaporization. The lower the volatility temperature, the more volatile
the metal is expected to be.
Figures 3 and 4 represent two ways of presenting metal partitioning data. The first way,
illustrated in Figure 3, presents metal partitioning to discharge streams as fractions of the
amount of metal fed to the incinerator. The second way, illustrated in Figure 4, presents
discharge distributions normalized to the total amount of metal measured in the sum of the three
incinerator discharges the kiln ash, the scrubber liquor, and the scrubber exit flue gas.
Both methods of presenting metal discharge distributions are useful in the process of
data interpretation. Expressing distributions as a percent of the metal,fed provides a more
accurate representation of individual discharge stream fractions if there are metal losses in the
incineration system that are not quantified. However, this method introduces a source of test-to-
test data variability caused by less than perfect mass balance closure.
The use of normalized distributions attempts to correct for this source of data variability.
Mass balance closure varied from test to test in this program, as it has in all past experience in
measuring trace metal discharges from combustion sources. Thus, variations in metal discharge
distributions expressed as a percent of the metal fed are affected by both actual variations due
to changes in test variables, and by individual test mass balance closure. Variations in
normalized distributions more closely reflect those due to changes in test variables because
variable mass balance closure is removed by the normalization. In discussing metal partitioning,
it is important to present metal distributions as both a fraction of the metal fed and normalized
to the total amount measured. The greatest credence is placed on conclusions supported by both
methods of presentation.
In all tests, mercury concentrations in the kiln ash were below a detection limit of
0.1 mg/kg. The volatile behavior of mercury was expected based on mercury's high vapor
pressure and low volatility temperature of 14 °C (57 °F). The normalized data in Figure 4 show
that the other test metals were relatively refractory. On average, kiln ash fractions accounted
for greater than 86 percent of each of the metals except for mercury; the lowest measured
normalized kiln ash fraction was 72 percent, for cadmium. In addition, Figure 4 shows a gradual
increase in kiln ash fraction with increasing volatility temperature, an expected observation. The
less volatile a metal is, as reflected in its higher volatility temperature, the less likely it is to
volatilize in the kiln and be carried out of the kiln in the vapor phase in the combustion flue gas.
Thus, it will more likely be retained in the kiln ash discharge.
22.2 Effects of Incinerator Operating Conditions on Metal Distributions
The effect of incinerator operating conditions and feed chlorine content can be seen via
plots such as those shown for cadmium in Figures 5 through,8. Figures 5 and 6 show the
variation in cadmium distribution with varying kiln temperature at each of the three feed
chlorine contents tested. Figures 7 and 8 show the variation in cadmium distributions with
10
-------
VARIABLE: KILN EXIT TEMPERATURE
CADMIUM, NO CHLORINE
1UU
£
ffi 80
1-
g
< 60
0
2
0
H- .»
on
15
*V
v*>
*JCJ
>
-------
VARIABLE: KILN EXIT TEMPERATURE
CADMIUM, NO CHLORINE
ON OF TOTAL MEASURED (%) FRACTION OF TOTAL MEASURED (%)
_L <
co to o ^j co to c
0 0 00 0 0 C
FRACTION OF TOTAL MEASURED (%) FRACTI
«g oo
-------
100
o
ffi 80
I-
60
40
20
VARIABLE: WASTE FEED CHLORINE CONTENT
CADMIUM, 538°C KILN EXIT TEMPERATURE
IUU
Q
a so
fe
FRACTION OF AMOUf
ro .b. en
o o o
15
|
1
10
5
rxTi iy/1 R??
^
CADMIUM, 816-C KILN EXIT TEMPERATURE
20
15
10
CADMIUM, 927°C KILN EXIT TEMPERATURE
0%
0.7%
3.4%
CJ
o
O
LLJ
FRACTION OF AMOUNT FED (%)
to *. 05 co c
o o o o c
I
I
15
10
5
KILN ASH SCRUBBER EXIT
FLUE GAS
1
P
//
SCRUBBER
LIQUOR
Figure 7. Cadmium distributions expressed as a fraction of the
cadmium fed, showing variations with waste feed
chlorine content at constant kiln exit temperature.
13
-------
VARIABLE: WASTE FEED CHLORINE CONTENT
CADMIUM, 538°C KILN EXIT TEMPERATURE
FRACTION OF TOTAL MEASURED (%)
vi oo to c
0 0 0 C
j
^
20
10
_^ R5 ^
CADMIUM, 816°C KILN EXIT TEMPERATURE
0%
| | 0.7%
3.4%
%
o
FRACTION OF TOTAL MEASURED (%)
3 vl 00 (O C
30 0 0 C
FRACTION OF TOTAL MEASURED (%)
VI 00 (O C
o o o c
I
s
/
/
/.
>
/
/
/
f
20
10
<
PV1
1
i
CADMIUM, 927°C KILN EXIT TEMPERATURE
I
\
20
10
x>
VSi
1
I
KILN ASH SCRUBBER EXIT SCRUBBER
FLUE GAS LIQUOR
Figure 8. Normalized cadmium distributions, showing variations with
waste feed chlorine content at constant kiln exit temperature.
14
-------
varying feed chlorine content at each of the three kiln exit gas temperatures tested. Figure 5
and 7 use distribution fractions as a percent of cadmium fed. Figures 6 and 8 use normalized
distribution fractions as discussed in Section 2.2.1.
All four figures show an indication of a range in kiln ash fraction for the 3.4-percent
chlorine feed tested at the 816°C (1,500 °F) kiln exit gas temperature. Three tests with this feed
at this nominal kiln exit gas temperature were performed while the Calvert scrubber system
pressure drop was varied. However, because changes in the scrubber system pressure drop
should not affect metals partitioning to the kiln ash, these three tests represent replicates for
evaluating data variability for metals partitioning to the kiln ash. Kiln ash partitioning values for
these three tests have been averaged and plotted as a single bar. The observed range for
partitioning to the kiln ash for these three tests is shown as an indicator of data variability.
In many cases, cadmium was not detected in the scrubber exit flue gas. The "<"
indications above the bars for the corresponding scrubber exit flue gas fractions are used to
highlight this occurrence. The height of the bar in these cases is the distribution fraction
corresponding to the detection limit flue gas concentration. Also note that two scales are shown
in each distribution plot; one (usually 70 to 100 percent) applies to the kiln ash fractions, and
the other (usually 0 to 30 percent) applies to the scrubber exit flue gas and scrubber liquor
fractions.
The normalized data in Figure 6 show a weak increase in cadmium volatility with
increased kiln exit gas temperature for all three feed chlorine contents, whereas the percent-of-
fed data in Figure 5 confirm this only for the chlorine-free feed. The data in Figures 7 and 8
suggest that feed chlorine content did not affect cadmium volatility within the range of data
variability experienced. The normalized data in Figures 10 and 14 show that the observed
volatilities of copper and lead were not affected by kiln exit gas temperature at 0 and 0.7 percent
chlorine content in the feed. For a 3.4-percent feed chlorine content, these figures show that
both metals exhibited increased volatility with increasing kiln temperature. However, the
percent-of-fed data in Figures 9 and 13 do not confirm this observation. The normalized data
in Figures 12 and 16 show that the volatilities of copper and lead also apparently increased with
increasing feed chlorine content. The percent-of-fed data in Figure 11 suggest the same for
copper at the two lower kiln exit gas temperatures. However, the percent-of-fed data in
Figure 15 do not confirm the observation for lead.
In all cases, each of these metals remained relatively refractory and were found primarily
in the kiln ash for all test conditions. In addition, there was no clear relationship between the
partitioning of arsenic, barium, bismuth, chromium, magnesium, mercury, or strontium and the
test variables over the ranges tested.
2.2.3 Scrubber Collection Efficiencies
The apparent scrubber collection efficiency for flue gas metals was determined for each
test. The apparent scrubber efficiency represents the ratio of the metal fraction measured in the
scrubber liquor to the sum of the metal fractions measured in the scrubber liquor and scrubber
exit flue gas. Figure 17 summarizes the efficiency data. The bar for each metal represents the
range of apparent scrubber collection efficiencies over the 11 tests, with the overall average for
the 11 tests noted by the midrange tick mark. Mercury recovery from scrubber liquor samples
15
-------
VARIABLE: KILN EXIT TEMPERATURE
COPPER, NO CHLORINE
100
Q
UJ
U.
I
O
60
40
20
100
ffi 80
O
Q
&
Q
111
U.
f-
O
O
60 _
40
20
100
80
60
40
20
1
1
15
10
5
COPPER, 0.7% CHLORINE
n/\
1
1
15
10
5
COPPER, 3.4% CHLORINE
I
T
J_
\
15
10
5
' I
v^
^, %>
KILN ASH SCRUBBER EXIT SCRUBBER
FLUE GAS LIQUOR
Tf
0>
o>
C4
O
O
UJ
<
538° C | |816°C
927° C
Figure 9. Copper distributions expressed as a fraction of the copper
fed, showing variations with kiln exit temperature at
constant waste feed chlorine content.
16
-------
VARIABLE: KILN EXIT TEMPERATURE
COPPER, NO CHLORINE
FRACTION OF TOTAL MEASURED (%)
SCO CD CD C
Ol O UI C
I
!
15
10
5
R53 V/\
COPPER, 0.7% CHLORINE
Q
LU
cc
3 95
CO
s
_l
!* 90
LL
O
O 85
l-
"- 80
«
~~
PyS
KTV
^IfS
y'VH
iX,>
VN
Vi
KX
L/XX
>sX
8$
?9
<§
s8
Vi
*xs
KX
^X
£x
vs
^
jfj
Ys
%
V,
^
/ x
15
10
5
^
R^i
UI O Ol
80
1
vxO
T
1
15
10
5
/yi
- J
r^ ^
P^^^J i Xx* Xx
KILN ASH SCRUBBER EXIT SCRUBBER
FLUE GAS LIQUOR
538° C | |816°C
927° C
co
01
O
LU
Figure 10. Normalized copper distributions, showing variations with kiln
exit temperature at constant waste feed chlorine content.
17
-------
VARIABLE: WASTE FEED CHLORINE CONTENT
COPPER, 538°C KILN EXIT TEMPERATURE
FRACTION OF AMOUNT FED (%) FRACTION OF AMOUNT FED (%) FRACTION OF AMOUNT FED (%)
tvj jx CD co o ro 4x a> co o r& u ococ
oo ooooo ooooo ooc
:
I
//
I
15
10
5'
h^y j
^
COPPER, 816°C KILN EXIT TEMPERATURE
I
T
I
I
15
10
5
ss
COPPER. 927°C KILN EXIT TEMPERATURE
I
%
6
15
10
5
KTTj
I
KILN ASH SCRUBBER EXIT SCRUBBER
FLUE GAS LIQUOR
AEC 028-94
0%
0.7%
3.4%
Figure 11. Copper distributions expressed as a fraction of the copper
fed, showing variations with waste feed chlorine content at
constant kiln exit temperature.
18
-------
VARIABLE: WASTE FEED CHLORINE CONTENT
COPPER, 538°C KILN EXIT TEMPERATURE
, luu
% '
Q
LU
FRACTION OF TOTAL MEASUR
co oo (o
*sH
KX
$
KX
KX
KX
KX
KX
KX
KX
KX
i
S
j" i«
J3y
I
10
5
^
100
s?
o
UJ
IT
1
I
LL
O
o
s
<
£E
ID
cn
-------
VARIABLE: KILN EXIT TEMPERATURE
LEAD, NO CHLORINE
0
LU
LL.
I
<
LL
O
g
<
a:
LU
100
00
o
o
*.
o
20
1
\
15
10
5
< < < E£T~^/
Tf
cn
(O
CO
o
O
LU
LEAD, 0.7% CHLORINE
FRACTION OF AMOUNT FED (%)
§*. CD 00 C
O O O C
I
^
15
10
5
s /
< < < 5£> '//
LEAD, 3.4% CHLORINE
10UNTFED(%)
§ 1
< 60
O
0 40
EC
LL,
20
1
K
$$
'&
*Os
g
T
1
1
%
%
//(
'/,
//
~^u
15
10
5
.
//
< < < ^ //
KILN ASH SCRUBBER EXIT SCRUBBER
FLUE GAS LIQUOR
^^ 538° C | |816°C [/yd 927° C
Figure 13. Lead distributions expressed as a fraction of the lead fed, showing variations
with kiln exit temperature at constant waste feed chlorine content.
20
-------
VARIABLE: KILN EXIT TEMPERATURE
LEAD, NO CHLORINE
IUU
g.
Q
LU
CC
3
CO
2 90
s
_l
<
1-=
o
i-
LL
O 80
z
O
1-
o
2
LU
70
!X>
wxO^
*SfSi
xjrjc^
^JC5
*ScS
y>
jxfj
*XXX
//
s J
fj
jT
SS
ss
f
r f
jX>
J* y
f
fjf
f /
fjf
//
f J
fjfj
//
jSj
* f .
20
10
i*^xT)4 Tx'xxi
L^'V^f l^^r'l KXl r jr A
LEAD, 0.7% CHLORINE
inn °"
1 W
g.
O
LU
FRACTION OF TOTAL MEASUR
3 § 8
* \s
1 00
Q
FRACTION OF TOTAL MEASURE
j co //
vS /s
KILN ASH SCRUBBER EXIT SCRUBBER
FLUE GAS LIQUOR
538" C
816°C
927° C
S
O
LU
Figure 14. Normalized lead distributions, showing variations with kiln
exit temperature at constant waste feed chlorine content.
21
-------
VARIABLE: WASTE FEED CHLORINE CONTENT
LEAD, 538°C KILN EXIT TEMPERATURE
100
g.
O
m
LL
K-
z
3
80
0%
| | 0.7%
3.4%
in
CM
o
O
FRACTION OF AMOUNT FED (%) FRACTION OF AMOUNT FED {%) FRACTION OF A
8 £ g g g 8 S ggggS i
I
\
10
^
LEAD, 816°C KILN EXIT TEMPERATURE
1
~r.
SY
I
30
20
10
rvd
^
LEAD, 927°C KILN EXIT TEMPERATURE
1
|
30
20
10
^
!
KILN ASH SCRUBBER EXIT SCRUBBER
FLUE GAS LIQUOR
Figure 15. Lead distributions expressed as a fraction of the lead fed, showing variations
with waste feed chlorine content at constant kiln exit temperature.
22
-------
u.
O
VARIABLE: WASTE FEED CHLORINE CONTENT
LEAD, 538°C KILN EXIT TEMPERATURE
TOTAL MEASURED (%) FRACTION OF TOTAL MEASURED (%)
o?
~r
20
I
8
8
o
LU
<
80
70
LEAD, 927°C KILN EXIT TEMPERATURE
1UU
g
Q
UJ
CC
OT
2i 90
5
_i
<
1-
O
h-
u_
O 80
z
0
g
2
LL.
yrt
C7W1
Vs
'SfS
Kx
s/V
*vs
KX
KX
KX
CK>
KX
SOf
5O
5o
KX
kV"V^
KX
KX
vs
KX
KX
±S\S
*<>
*//
s/
KILN ASH
20
10
< < < F77
i rx^
^
f^
/[/
ff
fjf
fJ
jf j
fjf
f^
jfj
/J
jfj
jfj
%
//
SCRUBBER EXIT SCRUBBER
FLUE GAS LIQUOR
0%
| |
0.7%
3.4%
Figure 16. Normalized lead distributions, showing variations with waste
feed chlorine content at constant kiln exit temperature.
23
-------
100
o
o
u.
u.
UJ
O
LU
O
cc
o
UJ
80
60
Cd
Mg
Cu
As
Ba
Hg
Sr
Cr
cc
Q_
O.
20
i
i
i
i
200 400 600 800 1,000 1,200
VOLATILITY TEMPERATURE (°C)
I
1,400
I
1,600
1,800
Figure 17. Apparent Calvert scrubber collection efficiencies for metals.
was poor for the tests without chlorine. For this reason, the mercury scrubber collection
efficiencies shown in Figure 17 are based on flue gas measurements, which, for mercury alone,
are thought to be more representative of the actual system performance.
Average metal collection efficiencies shown in Figure 17 ranged from 78 to 96 percent;
the overall average for all metals was 90 percent. Individual scrubber collection efficiencies for
each test ranged from 52 to greater than 99.7 percent. Figure 17 shows that there was no clear
relationship between apparent scrubber collection efficiency and metal volatility temperature.
There was also no clear relationship between the test variables and scrubber collection efficiency.
The observed scrubber collection efficiencies are somewhat lower than expected based
on past experience with this scrubber system. However, the average concentration of the metals
in the scrubber inlet flue gas ranged from 23 to 580 /ig/dscm. For any APCS, there will likely
be a contribution to the scrubber exit metals concentrations due to metals penetration that is
relatively insensitive to inlet metal concentrations. Therefore, the removal efficiency of an APCS
will likely be lower when inlet concentrations are lower than the efficiency will be when inlet
concentrations are higher. Scrubber inlet metals concentrations were relatively low for these
tests. Higher removal efficiencies might have been measured had inlet loadings been higher.
24
-------
A relationship between scrubber pressure drop and removal efficiency would also be more likely
at higher inlet loadings.
23 CONCLUSIONS
Test program conclusions include the following:
All of the metals were relatively nonvolatile except mercury. On average, kiln ash
fractions accounted for greater than 86 percent of each of the metals except for
mercury; the lowest measured normalized kiln ash fraction was for cadmium, at
72 percent. As expected, mercury was very volatile and was not found in the kiln
ash sample of any test.
On average, the relative volatilities of the metals observed agreed with the order
predicted by metal volatility temperatures
There was no clear relationship between the partitioning of arsenic, barium,
bismuth, chromium, magnesium, mercury, or strontium and the test variables over
the ranges used
Cadmium volatility may have showed a weak increase with increasing kiln exit gas
temperature; feed chlorine content did not affect cadmium volatility within the
range of data variability experienced
The observed volatilities of copper and lead were not affected by kiln exit gas
temperature at 0- and 0.7-percent chlorine content in the feed. For a 3.4-percent
feed chlorine content, both metals exhibited increased volatility with increasing
kiln exit gas temperature based on normalized data. Percent-of-fed data do not
confirm this observation, however. The volatilities of copper and lead also
increased with increasing feed chlorine content based on normalized data. This
trend is not apparent in the percent-of-fed data for lead, however.
There was no clear evidence of hysteresis in the RKS caused by test-to-test
carryover of any of the metals
The average scrubber collection efficiencies for metals ranged from 78 to
96 percent. The overall average collection efficiency for all metals was 90 percent.
That scrubber efficiencies were lower than expected is likely due to the low metal
concentrations at the scrubber inlet. No clear variation in scrubber collection
efficiency with variations in the test variables was observed. The lack of a
relationship between scrubber pressure drop and scrubber collection efficiency is
most likely due, in part, to the low metal concentrations at the scrubber inlet.
Test results outlined above were reported in the following report:
Fournier, D. J., Jr., and L. R. Waterland, "The Fate of Trace Metals in a Rotary
Kiln Incinerator with a Calvert Flux-Force/Condensation Scrubber System," draft,
January 1993.
25
-------
SECTION 3
EVALUATION OF ROTARY KILN INCINERATOR OPERATION
AT LOW TO MODERATE TEMPERATURES
As part of the EPA's efforts to remediate Superfund sites, several remediation
technologies can be candidates for consideration. One of the more frequently used technologies
to decontaminate soils contaminated with organic hazardous constituents is incineration. High-
temperature incineration, while usually effective in destroying organic compounds, may not be
necessary for some soils that need treatment, such as soils contaminated with volatile organic
compounds (VOCs). Also, in soils contaminated with toxic trace metals, high-temperature
incineration may increase the volatilization of some metals into the combustion flue gas. The
presence of elevated levels of volatile trace metals in the flue gas can pose increased challenges
to an APCS.
Another thermal treatment technology, thermal desorption, may be an attractive
alternative to incineration. When successful in decontaminating soils to the necessary degree,
thermal desorption treatment of soils offers the benefits of lower fuel consumption; avoidance
of slag formation; and less tendency to volatilize certain toxic metals such as lead and cadmium
into the effluent gas discharge, thereby decreasing the demands on an APCS to control metal
emissions.
Most conventional rotary kiln incinerators can be easily operated at temperatures below
those typically employed for incineration treatment. Thus, the following question arises: how
effective is the treatment of contaminated soils by a rotary kiln incinerator operated at the low
to moderate temperatures associated with the thermal desorption processes?
To address this question, a series of tests was conducted in the IRF RKS with the kiln
of the RKS operated at low to moderate temperatures The test program consisted of 12 tests
under 11 different kiln operating conditions; one test condition was tested in duplicate.
The objective of the test program was to study the global effects of five parameters
believed to be of primary importance in the effectiveness of soil decontamination and the fate
of contaminant metals in the treatment of contaminated soil in an incinerator operated at the
low to moderate temperatures associated with thermal desorption processes. These parameters
were soil moisture content, treatment temperature, treatment time, solids bed depth, and degree
of solids agitation.
The results obtained from the test program were intended to yield the following
information:
26
-------
The relationship between compound vapor pressure characteristics (boiling point)
and the extent of decontamination for each organic contaminant
How peak solids bed temperature affects the extent of decontamination
The differences in organic constituent decontamination effectiveness for solids
beds of different depths, when the same peak solids bed temperatures are reached
How the presence and the amount of moisture affect organic constituent
decontamination effectiveness
The relationship between treatment time and temperature and organic constituent
decontamination effectiveness
The distribution of trace metals in process discharges when a metal-contaminated
soil is treated by thermal desorption
Whether thermal desorption treatment conditions affect a metal's teachability
from the treated soil
The test program to address these issues was performed during late December 1992, and
January and early February 1993. The RKS was configured as shown in Figure 1, with the
exception that the Calvert scrubber system indicated in the figure was not in place. The venturi/
packed-column scrubber system shown in the figure served as the primary APCS for these tests.
3.1 TEST PROGRAM
3.1.1 Test Contaminated SoU
Original plans for the test program were to prepare a contaminated soil mixture by
combining a mixture of organic contaminants and an aqueous trace metal solution with a local
topsoil obtained from a garden supply vendor. After preparation of this synthetic contaminated
soil mixture, requisite quantities of water were to be added to achieve one of the two target
moisture contents selected for this test program variable. Prepared contaminated soils were to
be continuously fed to the RKS kiln over a test's duration using either a screw feeder for free
flowing (lower moisture content) soil or a progressive cavity pump for sludge-like (higher
moisture content) soil. However, initial attempts to feed unsupplemented local topsoils with
target moisture contents ran into problems.
Attempts to feed several local soil types adjusted to the low target moisture content of
10 percent were unsuccessful due to problems with soil bridging in the feed hopper supplying the
screw feeder. Even the use of an oversized screw and the placement of bridge breakers in the
feed hopper did not prevent bridge formation problems. Mixing soil with up to 30 percent sand
was similarly unsuccessful.
A different set of problems was experienced in the feeding of soil with the initial high
target moisture content of 30 percent with a progressive cavity pump. Pumping this high
moisture content material into the kiln was not a problem. However, once heated in the kiln,
27
-------
the dried soil caked and adhered to the kiln refractory wall and would not flow along the kiln
axis to the kiln discharge end.
After many trials, soil mixtures containing equal weights of local topsoil and attapulgite
clay absorbent material, with mixture moisture contents of 10 and 20 percent, were found to be
feedable when the screw feeder with bridge breakers and the oversized screw was used. Thus,
these mixtures became the two moisture content bases for the subsequent tests.
Synthetic contaminated soil mixtures for the actual tests were prepared as follows. The
local topsoil was delivered to the IRF in quantity by a dump truck. Delivered soil was placed
on a polyethylene-liner-covered area outside of the IRF building. This soil pile was mixed for
1 hour or longer with a tractor bucket that was dragged in an alternate perpendicular (crisscross)
pattern. After mixing, the soil was transferred into the IRF building and spread onto
polyethylene-liner-covered floor space to air dry for 36 hours or more.
The soil/clay absorbent mixtures were prepared in two 3.5-ft3 (100-L) cement mixers
via the addition of weighed quantities of each mixture component into a mixer. The organic
contaminants were added to the soil/clay,mixtures as a combined organic solution. Trace metal
contaminants were added in an aqueous solution. After the organic and trace metal solutions
were added, additional water (if needed) was added to adjust the contaminated soil mixtures to
the desired moisture content. The final soil mixtures were tumbled to uniform appearance, then
transferred to 55-gal (208-L) drums that were then sealed. Contaminated soil mixtures were
allowed to age between 7 and 14 days before use in a test.
The organic compounds included in the test feed materials were n-hexane, octane,
benzene, toluene, tetrachloroethene, chlorobenzene, naphthalene, anthracene, and pyrene. This
set of compounds includes common volatile organic solvent contaminants (benzene, toluene,
tetrachloroethene, and chlorobenzene), common gasoline constituent contaminants (n-hexane,
octane, and benzene), and common semivolatile organic contaminants associated with coal tar
or manufactured gas plant materials (naphthalene, anthracene, and pyrene). Compound boiling
points ranged from 69° to 156°C. Table 4 summarizes the organic mixture composition used to
spike the test mixtures. The organic contaminant mixture was added to the soil/clay mixtures
in the ratio of 0.02 kg organic liquid to 1.00 kg final soil mixture. Resulting contaminated soil
organic constituent concentrations are also noted in Table 4.
Contaminant trace metals were added to the soil/clay absorbent feed materials as a
concentrated aqueous solution. The trace metals added and their concentrations are summarized
in Table 5. All metal constituents were added as soluble nitrate salts except arsenic, which was
added as As2O3 dissolved into the acid nitrate spike solution. The metals spike solution was
added to the soil/clay mixtures in the ratio of 0.05 kg spike solution per kg of final contaminated
soil mixture. Resulting contaminated soil trace metal concentrations, neglecting native soil/clay
metal concentrations, are also noted in Table 5.
3.1.2 Test Conditions
As noted above, the test program consisted of 12 tests under 11 different combinations
of the test variables, with one test performed in duplicate. The test parameters were soil
moisture content, treatment temperature, treatment time, solids bed depth, and degree of solids
28
-------
ri
o
co
a
w
g
|
H
8
U
£
H
S3
|r^
Z
Cfl
W
H
Z
^H
^
1
£
OT
O
O
o
o
o
_ Ji
£j
tt
£;
32
M 3 ~
apyt fc5
2 *«*
^ rq
C ^} sw OD
S c c^^
^ L* *"^ 3
e o tJ
g c g
c **
a«
"«
'3 e
o- >
O IO TJ-
T-H T-H T-H CM
T-H T-H
CTs C3 T-H fM
VO OO T-H T-H
? « ^ ^
VO OO C^- CM
vo oo oq \q
O O O T-H
o\
«o
VO OO CM vo
- S
C
o o
T-H T-H CO
vo CM OO
CM CO T-H
T-H T 1 CM
£ 0 0
i >o oo
O T-H VO
C~- T-H T-H
O T-H T-H
<* CM oo
T-H r-H CM
C
M ^
flj « rd
C ^5 S
rt O -5
s ° -t
0 3 &
a U Z
o
o
T-H
rM
^
-------
TABLE 5. TRACE METAL CONSTITUENTS IN THE SYNTHETIC CONTAMINATED
SOIL
Metal Compound
concentration, concentration3,
Metal g/L Compound g/L
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
0.50
4.0
0.20
0.50
0.80
0.20
AS203
Ba(NO3)2
Cd(N03)2 4H2O
Cr(NO3)3 - 9H20
Pb(N03)2
Hg(N03)2
0.67
7.61
0.55
3.8
1.28
U.32
jveauiiuig auii
feed metal
concentration13,
rag/kg
25
200
10
25
40
10
Sufficient HNO3 added to maintain lead arsenate compounds in solution.
bNegh"gible soil metal concentrations and a ratio of 0.05 kg of.spike solution per kg of
organic/soil/spike solution mixture assumed.
agitation. Soil moisture content was directly varied, at 10 and 20 percent, respectively. Changes
in the other test parameters were caused by changing the RKS operating conditions. The
operating conditions varied from test condition to test condition were kiln exit gas temperature,
contaminated soil feedrate, and kiln rotation rate. Three kiln exit gas temperatures were tested,
320°, 480°, and 650°C (600°, 900°, and 1,200°F). Two feedrates were tested, 70 and 210 kg/hr
(150 and 470 Ib/hr). Three rotation rates were tested, 0.2, 0.5, and 1.5 rpfn.
Kiln exit gas temperature primarily affected peak solids bed temperature. Peak solids
bed temperatures corresponding to the above kiln exit gas temperatures were about 120°, 260°,
and 430 °C (250 °, 500°, and 800 °F), respectively. Kiln rotation rate affected both degree of
agitation and solids residence time in the kiln, or maximum treatment time. Total kiln solids
residence times corresponding to the above rotation rates were 60, 40, and 30 minutes,
respectively. The combination of feedrate and kiln rotation rate affected solids bed depth.
Total treatment times were varied by varying kiln rotation rates, as noted above.
However, to allow the evaluation of treatment effectiveness at multiple, partial treatment times
for each test condition, samples of the solids bed material were taken at four axial locations
along the kiln for each test in addition to a solids discharge sample. These additional samples
corresponded to four different treatment times at each test condition.
A summary of the target test operating conditions and soil moisture contents for each
of the specified 12 tests is given in Table 6. The "center point" of the test matrix is represented
by Test 2, with soil feedrate at 70 kg/hr (150 Ib/hr), kiln exit gas temperature of 480°C (900°F),
30
-------
TABLE 6. TARGET TEST CONDITIONS
Test
1
2
3
4
5
6
7
8
9
10
11
12
Kiln exit gas
temperature,
°C (°F)
320 (600)
480 (900)
650 (1,200)
320 (600)
480 (900)
650 (1,200)
480 (900)
480 (900)
480 (900)
480 (900)
480 (900)
480 (900)
Expected peak
solids bed
temperature,
°C (°F)
120 (250)
270 (520)
430 (800)
120 (250)
270 (520)
430 (800)
270 (520)
270 (520)
270 (520)
270 (520)
270 (520)
270 (520)
Kiln rotation
rate, rpm
0.2
0.2
0.2
0.2
0.2
0.2
0.5
0.5
1.5
0.2
0.2
0.2
Soil feedrate,
kg/hr (Ib/hr)
70 (150)
70 (150)
70 (150)
70 (150)
70 (150)
70 (150)
70 (150)
70 (150)
70 (150)
210 (470)
210 (470)
70 (150)
Soil
moisture
content,
%
10
10
10
20
20
20
10
20
20
10
20
10
kiln rotation rate of 0.2 rpm, and soil moisture content of 10 percent. This test condition was
tested in duplicate (Test 12). From this "center point," kiln temperature was varied (Tests i
and 3), soil moisture content was varied (Test 5), kiln rotation rate was varied (Test 7), and soil
feedrate was varied (Test 10). Additional test combinations were performed for the high
moisture soil at the base feedrate and rotation rate (Tests 4 and 6), at the high feedrate and
base rotation rate (Test 11), and at the base feedrate and increased rotation rate (Test 8). The
highest rotation rate was tested at high feedrate with the high moisture soil (Test 9).
For all tests, the afterburner was operated at 1,090°C (2,000°F) to ensure satisfactory
burnout of all volatilized organic compounds. The scrubber system was operated under its
nominal design conditions to achieve typical scrubber performance. The scrubber was operated
at near total recycle, so there was minimum blowdown. The synthetic contaminated soil was fed
continuously until all flue gas sampling was completed. Treated soil was continuously removed
from the kiln ash hopper via an ash auger transfer system, and deposited in initially clean 55-gal
(208-L) drums. After the completion of each test (all flue gas sampling completed) the system
continued to operate at the specified test conditions, without soil feed, until all treated soil was
cleared from the kiln. The weight of treated soil collected was monitored continuously
throughout the test; the resulting data allowed the calculation of total kiln solids residence times.
31
-------
A summary of the actual test conditions in effect for each test is given in Table 7. As
shown, average kiln exit gas temperature targets were closely met for all tests. Solids bed
temperatures were measured at four locations along the kiln axis 0.6, 1.1, 1.5, and 2.0 m (2.0,
3.5,5.0, and 6.5 ft) from the kiln feed face. Measurements were made with a specially-fabricated
probe that allowed the immersing of four thermocouples in the solids bed at the respective axial
locations. Solids bed temperatures measured for the tests are also given in Table 7.
3.13 Sampling and Analysis
For all tests, the sampling protocol consisted of:
Obtaining a composite sample of the contaminated soil feed material mixture
Obtaining composite samples of the treated soil in the kiln chamber at four axial
locations corresponding to the solids bed temperature measurements 0.6, 1.1,
1.5, and 2.0 m (2.0, 3.5, 5.5, and 8.6 ft) from the kiln feed face
Obtaining a composite sample of the treated soil discharge from the discharge
collection drum
Obtaining composite pre-test and post-test scrubber liquor samples
Sampling flue gas for trace metals using an EPA multiple metals train at the
venturi/packed-column scrubber exit
Sampling flue gas for mercury using a Method 101A train at the venturi/packed-
column scrubber exit
Continuously monitoring O2, CO, and total unburned hydrocarbon (TUHC) levels
in the kiln exit flue gas
Continuously monitoring O2 in the afterburner exit flue gas
Continuously monitoring O2 and CO2 downstream of the venturi/packed-column
scrubber
Continuously monitoring O2 and CO in the stack downstream of the secondary
APCS (carbon bed/HEPA filter)
Sampling the stack gas for particulate, and HC1, and C12 using Method 50
As noted in Section 3.1.1, contaminated soil feed was prepared, placed into 55-gal
(208-L) drums, and allowed to age between 7 and 14 days prior to use in a test. Just prior to
a test, the drums of soil to support the test were opened and sampled. Drum contents were then
transferred to the screw feed hopper for feeding.
Four composite kiln solids bed samples were also collected for each test. One sample
was collected using a custom-fabricated quartz scoop at each of the four axial locations where
32
-------
I
Q
z
o
O
S
04
O
%
Si
t>
w
s
x-v
s
M
^f \o
VO 00
S3
A P
T li o\
S
T-H OO C*l
^SS 3
oo
s
a
*O o
E; *
rj -1 00 «H
3 C- 3 C/
g» S
3a
11
u b
O O
8
O
1
"5
S
s
VO
?=>
t^
CM
3!
f 1 -
3 S -S
*O >n.
IIS
^ ^f "*s*
C> CL/ o
r-t r-- «t
s a
?, £
s
ON I/I
C- *-!
O 00''
S s
e, &
a §,
S S3
,- & E
JJ} ^
21 £2
co ^r
33
-------
solids bed temperature was measured. Each of these samples represented a different treatment
time under the set of other test conditions established for each test. A sample of the final
treated soil discharge was also collected from the discharge collection drum after the completion
of each test.
Test program samples were analyzed as follows. Unspiked soil/clay absorbent mixture,
each test's feed mixture, and all treated soil samples were analyzed for the spiked volatile and
semivolatile organic contaminants and the spiked trace metals. A composite of the Test 1,2, and
3 soil feed, and the final treated soil discharge samples for each of these tests, were analyzed for
polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans (PCDDs/PCDFs) by
Method 8290.
Toxicity characteristic leaching procedure (TCLP) leachates of two composite soil feed
samples and of all treated soil samples were prepared and analyzed for the test trace metals.
Specifically, leachates were digested by EPA Method 3010 and analyzed for barium, cadmium,
chromium, and lead by inductively coupled argon plasma (ICAP) spectroscopy by Method 6010;
leachates were digested and analyzed for arsenic by Method 7060; and leachates were digested
and analyzed for mercury by Method 7470.
Allpre-test and post-test scrubber liquor samples were analyzed for the test trace metals
by the same methods employed for the TCLP leachate samples. In addition, one composite pre-
test scrubber liquor and all post-test scrubber liquor samples were analyzed for the spiked
volatile and semivolatile organic soil contaminants.
Finally, all multiple metals train samples were analyzed for the non-mercury test trace
metals, and the Method 101A sampling train samples were analyzed for mercury.
32 CURRENT STATUS
As noted in Section 3.1.2, the testing portion of this program was completed in February
1993. However, due to the large number of samples collected, the organic analysis results were
not completely reported until May. In addition, the trace metal analysis results were not
completely reported until mid-August. Evaluation of the very large quantity of test data collected
in the program was underway at the close of FY93. The test report will be completed in FY94.
34
-------
SECTION 4
INCINERATION OF SIMULATED MIXED WASTE
The Savannah River Laboratory (SRL) is providing technical support for the design and
eventual operation of the Consolidated Incineration Facility (GIF), a rotary kiln incineration
system to be installed at the Savannah River Plant to treat hazardous and combustible low-level
mixed waste. As a key part of this technical support, SRL (D. Burns, Coordinator) requested
that a series of pilot-scale incineration tests be performed at the IRF to supply data to support
final equipment specifications for GIF subsystems, to guide trial burn planning efforts, and to
supply kiln ash samples for solidification/stabilization experiments. These tests were performed
under contract to SRL within the third-party use provisions of the IRF operations contract.
In this series of tests, simulated waste feeds representing typical Savannah River Plant
low-level mixed waste were incinerated in the IRF RKS. Three simulated waste types were
tested in a 22-test program in which waste type, waste density, waste feedrate, solids residence
time in the kiln, and incineration temperatures were varied. The objectives of the test program
were to:
Measure the incineration system flue gas particulate load and size distribution
upstream of any APCS, as a function of the test variables
Collect sufficient kiln bottom ash and flue gas particulate upstream of the APCS
for further characterization by SRL, as a function of the test variables
Measure waste volume and mass reduction as a function of the test variables
In a limited set of three tests with simulated waste spiked with organic hazardous
constituents and hazardous constituent trace metals:
Measure organic constituent DREs as a function of the test variables
Evaluate the fate of the trace metals with respect to their distribution among
incinerator discharge streams as a function of the test variables
Results of the test program are discussed in the subsections that follow.
4.1 TEST PROGRAM
As noted above, all tests were performed in the RKS. For the tests, the RKS was
configured as shown in Figure 1 with the exception that the Calvert scrubber system was not in
35
-------
place. Sections 4.1.1 through 4.1.3 discuss the simulated mixed wastes tested, the matrix of
incinerator operating conditions tested, and the sampling and analysis matrix completed.
4.1.1 Test Waste Description
Three simulated waste feed mixtures were incinerated in different tests during the
program. The three waste types were:
Design waste mix A, which contained, by weight, 30 percent white copier paper,
15 percent newsprint type yellow paper, 10 percent polyvinylchloride (PVC)
scraps, 25 percent polyethylene (PE) bags, and 20 percent latex gloves
Spiked design waste mix E, which consisted of waste type A spiked with hazardous
constituent trace metals and organic constituents. The organic constituents,
considered as the principal organic hazardous constituents (POHCs) requiring
demonstration that incineration achieved 99.99 percent DRE, were
hexachloroethane and hexachlorobenzene. Each POHC was added at a
concentration of 0.6 percent by weight of the waste A base. In addition,
hazardous constituent trace metals were spiked into this waste mkture. The
metals spiking compounds and spiking levels are shown in Table 8.
« Design waste mix B, a high-ash waste that consisted of 100 percent copier paper
TABLE 8. HAZARDOUS CONSTITUENT METALS SPIKED IN WASTE E
Metal
Spiking compound
Metal concentration
in simulated waste,
mg metal/kg feed
Antimony
Barium
Antimony potassium tartrate
[K(SbO)C4H406 y^ H20]
Barium acetate
pa(CH3COO)2]
Chromium Chromium nitrate, 9-hydrate
[Cr(N03)3 9 H20]
Lead Lead nitrate
Mercury
Thallium
Mercuric nitrate,
monohydrate
[Hg(N03)2 H2OJ
Thallium acetate
[TI(CH3COO)]
730
10,000
550
9,000
590
730
3ft
-------
The bulk raw waste materials (the two types of paper, PVC, PE, and gloves) were
delivered to the IRF by SRL. Waste blending, spiking (for waste E feed), and packaging for
feeding to the RKS were performed at the IRF. The waste was packaged into 1.5-gal (5.7-L)
cubical cardboard boxes with length, width, and height of 7 in. (17.8 cm). These boxes were fed
into the RKS via the ram feed system. Three target densities of waste type A were tested: 3,
6, and 9 lb/ft3 (48, 96, and 144 g/L). An industrial shredder was used to shred the paper and
PVC. In packing the waste A boxes, each box was first lined with a PE bag, then packed
sequentially with each item, and then sealed with paper tape backed with water-soluble glue.
The PE bag liner was not used for the waste B boxes, and only one waste B density, 9 lb/ft3
(144 g/L), was tested.
Waste E boxes required the addition of the POHCs and spike metals. These boxes were
packaged with the required density of waste A, which was then spiked with the POHCs and
metals. Each POHC was preweighed into plastic test tubes that were emptied into each box.
The metals were added as aqueous solutions. The PE bag liners were then sealed with a plastic
tie and the boxes sealed with paper tape.
4.12 Test Conditions
The test program variables were waste type (A, B, and E), waste feed density, waste
feedrate, kiln and afterburner temperatures, and kiln solids residence time. Table 9 summarizes
the matrix of target test conditions. The primary waste feedrate settings desired by SRL for the
test program variable were in terms of the volumetric heat release rate from the waste (kW/m3
[Btu/hr/ft3]). Given that the volume of the RKS is 1.9 m3 (67.2 ft3), required waste heat input
rates (kW [Btu/hr]) can be specified. Knowledge of the heating values of the waste components
allowed the specification of corresponding waste weight feedrates (kg/hr [lb/hr]) as shown in
Table 9.
The waste feeding procedure involved placing the entire quantity of feed for a test on
a weigh scale and manually loading each box onto the ram feeder. The ram feeder's operation
was remotely controlled to deliver each box into the kiln at the specified feedrate.
For all tests, the scrubber system was operated at its nominal design operating
conditions, listed in Table 10. The scrubber system was operated as close to complete
recirculation (minimum blowdown) as possible for all tests in order to maximize the total solids
content for each waste type. The same scrubber liquor was recirculated in all tests for a given
waste type.
A total of 765 gal (2,900 L) of scrubber liquor with a solids content of 1 percent was
collected from 17 waste A tests. From the waste B tests, a total scrubber liquor volume of
265 gal (1,000 L) with a 1 percent solids content was collected. The total scrubber liquor volume
collected from the waste A tests and the total from the waste B tests were shipped separately
to SRL for further characterization. Scrubber liquor from waste E tests was disposed of.
Ash from each test was initially collected in 20-gal (76-L) drums in the kiln ash pit and,
after cooling, transferred to 5-gal (19-L) plastic drums. In addition, some amount of flyash was
deposited in the afterburner, and in the entry and observation hatches to the kiln. Ash from the
37
-------
CO
i
s
Q
2
O
O
en
13
o
H
9
tn
I
g a
-S -
a*
«n "x^
^>
5 *e
Sf
e
OJ
E
'5b
2
O)
gf
C3
J3
O
2
25,
O
B
I J3
!=.
,
o o o o o
T* VO CO vS CO
£-* oq*
oq
oq* oq* £~* £~* £5* oq* ^T ^ ^ oq* f--' f--* o*
O\ r-H iIO\O\r-lt-la\»-lT-IO\O\iIT-HIl»-liI
?a
oqoooqoqoo
"
R i^. ^ R R S"
oqoooqoqcsocsc^csooqoqoqoq
R ^" & S. R S. ?. £ s?" 5' .^. S; R 5. S
a
a a
a
a
\qvovovooqoqoqoqtscstSNoqoqoqoqvqvqvooqoqoq
C5C3C5C3r-
-------
TABLE 10. APCS OPERATING CONDITIONS
Venturi liquor flowrate 76 L/min (20 gpm)
Venturi pressure drop 6.2 kPa (25 in WC)
Packed tower liquor flowrate 115 L/min (30 gpm)
Scrubber liquor temperature 49°C (120°F)
Scrubber blowdown rate 0 L/min (0 gpm) or minimum operable
afterburner was recovered whenever the kiln and the afterburner were shut down and sufficiently
cool between tests.
On any given test day the RKS was brought to steady operation at the desired test
condition, firing natural gas only. Waste feed was then initiated. Flue gas sampling was started
only after at least 1 hour of steady operation in which waste feeding had occurred.
Table 11 summarizes the actual RKS operating conditions for each test, along with the
respective test targets. As indicated in the table, actual test waste feedrates were within 6
percent of target for all except three tests. Even for these three, actual waste feedrates were still
within 9 percent of target. Target kiln solids residence times were always achieved.
Average kiln exit gas temperatures achieved were within 14 °C (25 °F) of target
temperatures for all but three tests. For Test 6, the average kiln exit gas temperature, at 825 °C
(1,517°F), was 65°C (117°F) higher than the 760°C (1,400°F) target. As noted in Table 11, the
target kiln exit gas O2 level was 10 percent for all tests. The attempt to maintain a near target
flue gas O2 level at the kiln exit for the high waste A feedrate specified for Test 6 resulted in
higher than target kiln exit gas temperatures. At the specified heat input rate for Test 6, gas
temperatures could not be held as low as 760°C (1,400°F) without the addition of cooling air in
excess of that associated with an exit gas O2 of 11 percent. For subsequent tests, the SRL
Coordinator agreed that maintaining target kiln gas temperatures was more important than
achieving target flue gas O2 levels. Thus, for subsequent tests, the low kiln exit gas temperature
target, 760°F (1,400°F), was more nearly achieved for low waste feedrate (13.1 kg/hr[28.8 lb/hr])
conditions at the expense of needing excess air resulting in 12.5 to 13 percent O2 in the exit flue
gas. The next low kiln temperature, high waste A feedrate (24.6 kg/hr [54.0 lb/hr]) test, Test 2,
still had an average kiln exit gas temperature of 54°C (97°F) higher than the target, even though
excess air corresponding to an exit gas O2 of 14 percent was allowed. The two tests at low kiln
temperature for the intermediate waste A feedrate (19.6 kg/hr [43.2 lb/hr]) were associated with
the next two largest differences between average test kiln exit gas temperature and test target.
For these two tests, Tests 16 and 17, actual average kiln exit gas temperatures were 23° and
13 °C (41° and 24°F) higher, respectively, than the kiln exit gas temperature target. This
occurred despite allowing excess air levels corresponding to kiln exit gas O2 concentrations of
12.1 percent to be introduced.
39
-------
CO
z
I
Z
O
u
3
H 0
c
a
^ '
°- 0°>
rH "^ O\ "1
jo S e^ |o.
o
Re
"~! CQ **:
CO oS CO
a
o
STT
^
§2-
ll
S co 3
2 S S o
.
TH ^*^ ON
si?
d ».
.
r-< N / O\
§^* trt <* £
** OO SO """^
*O T-; rX ^ vrl in
1 «
? £f
& « 2 I
£4 3 V ca
« jj >; 05
C>
o o
o °i
rH OS
rM~ O R
s S
a- *. 4 "o.
SS 8
2 ^
5 a
s ^ 3
VO
G
o
o
i
-
1 in
so *f so Qrt U-)
o ad; id;
«- s g> | &
y g -g ta 5 >
< § S H .< <
40
o
Tf
O\ rH
O O
CO CO
00
VO «3
oo
CO CO
O
so
O
\O
O O
CO m
-------
S
S ***
1>
.S
c
ea
E-i
2
2
O^L
VOr-T
0
ON ^-^ O\
K> «-" KJ "-r
^
rH* i
o _ sp i ^r
85> o o o\
9^ o S o^
-CM °^ of *°" T-T
s
VO VQ
d) S3
§iS
S "°- !S lrt-
oo vi oo ^
8S
Si
s
C §S
? & P v'^
rf-S-l is si
|S
THS5|
S S- g °-
TH-CS TH-G,
I
"O
OV 00
^^ OV TH
m
00
s s
4> g « 0
-_ac^ S a
JilljI i1 fl J
8
g
1
3
&'H
S fla
41
-------
Average actual afterburner exit gas temperatures were uniformly within 5°C (8°F) of
the target for all tests. However, for many tests, additional excess cooling air was needed so that
the exit gas O2 levels were often as high as 14 percent compared to the test program target of
9 percent.
Table 12 summarizes the solids bed temperatures measured at four locations along the
kiln axis. These locations were 0.6, 1.1, 1.5, and 2.0 m (2.0, 3.5, 5.0, and 6.5 ft) from the feed
face of the kiln. The solids bed temperature measurement system discussed in Section 3.1.2 was
used to make the bed temperature measurements. The data in Table 12 show that the highest
solids bed temperature always occurred at either the 0.6 m (2.0 ft) or 1.1 m (3.5 ft) distance from
the feed face. These locations were where visual observation showed that the most vigorous
waste combustion occurred after waste charge ignition.
As noted above, kiln ash was collected in a 20-gal (76-L) drum placed in the kiln ash
discharge pit. However, some of the flyash entrained in kiln exit combustion gas dropped out
in the horizontal afterburner of the RKS or in hatches between the kiln and afterburner. The
afterburner and hatches were not cleaned after each test. Instead, they were usually cleaned at
the beginning of a test week while the incinerator was cool, before the start of heating up the
incinerator for the week's tests. Table 13 summarizes the weight of kiln ash collected in the
collection drum during each test, and collected from the afterburner and hatches after groups
of tests.
4.13 Sampling and Analysis
The sampling protocol followed to satisfy the test plan objectives entailed:
Collecting and shipping all kiln bottom ash from each test to SRL
Collecting and shipping all composite scrubber liquor from the test series with
waste A and waste B to SRL
Continuously measuring O2 levels at the kiln exit; O2, CO, CO2, NOX, and TUHC
levels at the afterburner exit; O2, CO2, and NOX levels at the scrubber exit; and
O2 and CO at the stack
Sampling the flue gas in triplicate for each test at the afterburner exit for
participate loading using Method 5, and for particle size distribution using an
Andersen cascade impactor train
Sampling the flue gas in triplicate for each test at the afterburner exit to obtain
gram-sized particulate samples, using a modified Method 17 train
Sampling the flue gas in triplicate for Tests 7, 10, and 14 at the afterburner exit
for trace metals, using the EPA multiple metals train
Sampling the flue gas for Tests 7,10,14, and 18 at the afterburner exit for the test
POHCs, using Method 0010
42
-------
TABLE 12. SOLIDS BED TEMPERATURES
Temperature, °C (°F)
Test
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Kiln exit gas
766 (1,410)
814 (1,497)
998 (1,829)
1,003 (1,837)
773 (1,423)
825 (1,517)
1,000 (1,832)
995 (1,823)
1,007 (1,844)
767 (1,413)
769 (1,417)
1,006 (1,843)
1,003 (1,838)
874 (1,606)
878 (1,612)
783 (1,441)
747 (1,376)
880 (1,616)
1,004 (1,840)
759 (1,398)
997 (1,827)
997 (1,827)
Solids bed
0.6 m (2.0 ft)
885 (1,625)
887 (1,628)
986 (1,807)
1,091 (1,996)
851 (1,564)
832 (1,530)
842 (1,547)
968 (1,775)
1,098 (2,009)
679 (1,255)
864 (1,588)
1,053 (1,927)
989 (1,813)
726 (1,339)
794 (1,462)
875 (1,607)
916 (1,681)
853 (1,569)
1,072 (1,961)
856 (1,573)
881 (1,617)
857 (1,574)
at a distance of V m
1.1 m (3.5 ft)
779 (1,434)
867 (1,596)
1,094 (2,001)
1,086 (1,987)
801 (1,474)
919 (1,687)
1,133 (2,072)
1,003 (1,838)
1,131 (2,067)
879 (1,614)
765 (1,409)
1,047 (1,916)
1,087 (1,988)
1,037 (1,898)
994 (1,822)
829 (1,525)
829 (1,524)
1,030 (1,886)
1,048 (1,919)
779 (1,434)
980 (1,796)
1,007 (1,844)
("x" ft) from the
1.5 m (5.0 ft)
717 (1,323)
777 (1,431)
1,003 (1,837)
1,030 (1,886)
742 (1,368)
808 (1,487)
1,000 (1,832)
967 (1,773)
1,046 (1,915)
717 (1,322)
714 (1,317)
1,009 (1,849)
997 (1,826)
863 (1,585)
856 (1,573)
753 (1,387)
729 (1,345)
890 (1,634)
993 (1,819)
704 (1,300)
922 (1,691)
936 (1,716)
kiln feed face
2.0 m (6.5 ft)/
691 (1,275)
746 (1,375)
975 (1,787)
937 (1,718)
701(1,293)
760 (1,400)
977 (1,791)
947 (1,736)
971 (1,780)
696 (1,285)
667 (1,232)
928 (1,703)
972 (1,782)
828 (1,522)
813 (1,496)
717 (1,322)
698 (1,288)
866(1,590)
971 (1,779)
694 (1,281)
920 (1,688)
941 (1,726)
43
-------
TABLE 13. COLLECTED ASH WEIGHTS
Test Date
21 3/24/93
22 3/25/93
After Test 22
Total Waste B
6 3/30/93
5 3/31/93
8 4/1/93
After Test 8
20 4/6/93
17 4/7/93
After Test 17
3 4/13/93
2 4/27/93
13 4/28/93
19 4/29/93
After Test 19
16 5/5/93
1 5/6/93
After Test 1
11 5/11/93
9 5/13/93
After Test 9
12 5/18/93
4 5/19/93
After Test 4
18 5/25/93
15 5/27/93
After Test 15
Total Waste A
10 6/3/93
After Test 10
7 6/8/93
14 6/10/93
After Test 14
Total Waste E
Waste
type
B
B
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
E
E
E
Collection drum
5.0 (11.0)
5.9 (13.0)
3.0 (6.5)
1.1 (2.5)
1.4 (3.0)
1.4 (3.0)
3.2 (7.0)
2.5 (5.5)
2.5 (5.5)
1.1 (2.5)
1.4 (3.0)
1.8 (4.0)
1.6 (3.5)
0;9 (2.0)
2.3 (5.0)
0.9 (2.0)
1.1 (2.5)
2.7 (6.0) .
1.4(3.0)
5.0 (11.0)
1.8 (4.0)
2.0 (4.5)
Ash collected, kg (Ib)
Afterburner
and hatches
11.4 (25.0)
5.5 (12.0)
2.5 (5.5)
4.8 (10.5)
3.4 (7.5)
2.7 (6.0)
2.7 (6.0)
3.2 (7.0)
1.6 (3.5)
8.9 (19.5)
Total per
test group
22.3 (49.0)
22,3 (49.0)
10.9 (24,0)
74 (15.5)
12.3 (27.0)
6.8 (15.0)
5.9 (13.0)
4.7 (10.5)
7.3 (16.0)
55.0 (121.0)
6.6 (14.5)
12.7 (28.0)
19.3 (42.5)
44
-------
Sampling the flue gas for Tests 7, 10, 14, and 18 at the scrubber system exit duct
for PCDD/PCDFs, using Method 23
Sampling the flue gas for Tests 7, 10, and 14 at the stack downstream of the
secondary APCS for particulate and HC1, using Method 5
A kiln ash sample from each test was analyzed for total carbon using ASTM Method
D-3178. Kiln ash samples from Tests 7, 10, 14, and 18 were analyzed for total chlorine content.
The Method 0010 train samples were Soxhlet extracted (EPA Method 3540) and analyzed for
the test POHCs by Method 8270.
The front half (filter plus probe wash) and back half (impinger contents) of the multiple
metals train were analyzed separately for the spiked trace metals. Microwave multi-acid (HF
plus HNO3) digestion was used to prepare the front half, and conventional digestion to prepare
the back half. Digestates were analyzed by ICAP spectroscopy in accordance with Method 6010
for all test trace metals except mercury, which was determined by cold vapor atomic absorption
spectroscopy (CVAAS) according to Method 7470. Kiln ash samples for Tests 7,10, and 14 were
also analyzed for the spiked metals by microwave multi-acid digestion followed by ICAP, and for
mercury using CVAAS by Method 7471.
The Method 23 train samples from the scrubber exit were extracted and analyzed by
Method 8290 for total tetra- through octa-chlorinated PCDDs and PCDFs, with specific
quantitation for all 2,3,7,8-chlorine-substituted tetra- through octa-chlorinated isomers.
Stack gas particulate load was determined by desiccating the Method 5 train filter and
probe wash. The stack gas HC1 levels were determined by analyzing the combined impinger
solutions from the train according to Method 9057.
The composite scrubber liquor sample from the waste E tests was subjected to TCLP
extraction and analyzed for trace metals and the test POHCs before disposal.
4.2
TEST RESULTS
Table 14 summarizes the particulate levels measured in the afterburner exit flue gas for
all tests. Average levels (over three Method 5 runs for a test) ranged from 175 to 743 mg/dscm
corrected to 7 percent O2. The general, though not universal, trend is that the higher
afterburner exit particulate levels are associated with the higher waste feedrates for waste A.
Two POHCs, hexachlorobenzene and hexachloroethane, were spiked into the waste feed
for three tests, Tests 7, 10, and 14, in order that the achieved DREs could be measured. The
spiked waste became waste E. Afterburner, exit flue gas concentrations and emission rates of
these POHCs were measured for the three tests. In addition, afterburner exit flue gas
concentrations of the two POHCs were also measured for Test 18 to establish a background flue
gas concentration for a test in which no spiking was performed. Measured flue gas
concentrations and corresponding DREs are summarized in Table 15. As shown, neither POHC
was detected in the afterburner exit flue gas of any test at detection limits of about 0.6 /ig/dscm
for hexachlorobenzene and 1.3 to 1.4 jig/dscm for hexachloroethane. Corresponding DREs
45
-------
bfi
t o\ co o\ f- >o
"
vo
vo t-~ T-I
oo vo og
T
»/S VQ
O4 04
co
o o\ uv
* »0 ?-l
co co co
vo T-I
10
oo
OX
CO O O\
t< * TH
04
CO
\0
9
O4 O\
04
VO
f- t-~
0\
s a
>0 CO
o >o
i
s
CO \0
VO
rf 04
O CO
s a s s
oo
T-T T-T r-f T-T T-T
CO
OO
*
O>
oo
.
ON T-l
OO
s
8. fe
cq c> o oq oq C3 oq o oq oq oJ
ol OJ o) oq oq
jo jo
oq oq
a
fcD
CO CO. CO CO
o\ o\
CO CO
vo. os. p\, p\, co. co, co. ps. p\.
oooo
T? Tf T-l
rHo4cO'*>OVOOOOST-(
a -as
04
46
-------
1
o
u
ffl
o
OH
*
!H
fVt
r_3
i
in
i
1
o>
1
3
*
Q
S
u
OT
1
a.
c
,g
1
c
1
-_
fa
O
O
0
1
«s
o3
&
i
0>
2
1
c
f
p
o
1
tS
i
c
c^
A
2
o
1
ffi
0)
1
1
1
E
e
N
e
exachlorobe
a
|i
U
a* ^j
§
e
a w
S So
95
r j
.
^t" ON ^~
OO OO OO
ON ON ON
ON ON ON
ON ON ON
A A A |
v^ oo ID
OJ ^" C^J
ON ON ON
ON ON ON
ON ON ON
ON ON ON
ON ON ON «
A A A |
OO VO >O CM
V V V V
TJ- ON CO OJ
VO *O NO VO
C3 C3 C3 O
V V V V
^^ >O oo^ G^
^ 1 O ^ O
o^ vo^ c^ c^
CO OO f-
O ^-. ON ON
r-^ ji O^ C3^
r-l OO r-H r-l
OJ CO *O VO
CO rH O r-l
T T r-T rH~ rH~
O
§ t- ^ o
*-^ vo t^- oo
r-< C"> OO OO
0 ^J 00
§
CL,
a
1
Q
o
c
o"
-------
achieved for the three POHC-spiked tests were greater than 99.99925 percent for
hexachlorobenzene and greater than 99.9984 for hexachloroethane.
Eevels of PCDDs/PCDFs were measured in the scrubber exit flue gas for the same four
tests for which afterburner exit flue gas POHC concentrations were measured. PCDD/PCDF
results are summarized in Table 16. As shown in the table, the highest PCDD/PCDF emission
levels were measured in Test 18, the one test of the four for which no hexachlorobenzene or
hexachloroethane was spiked. All total PCDD/PCDF concentrations were less than the recent
EPA guidance level of 30 ng/dscm, and all 2,3,7,8-TCDD toxicity equivalent levels were less than
to slightly higher than the European Community (EC) guidance level of 0.1 ng/m .
Table 17 summarizes the trace metal data from the three tests for which simulated waste
was spiked with metals. The data in the table show that the kiln ash antimony, barium, and
chromium concentrations were greater than the corresponding waste feed concentrations. This
would be expected for metals not volatilized to a significant extent into the kiln combustion gas
because a given weight of waste feed produces a smaller weight of kiln ash. In contrast, kiln ash
lead concentrations were lower, and mercury and thallium concentrations much lower, than the
corresponding waste feed concentrations. This suggests that some, perhaps most, of the feed
quantity of these three metals was volatilized under test conditions and was carried out of the
kUn in the combustion flue gas.
This point is further illustrated by Table 18. This table summarizes the fractional
distributions of the amount of metal fed for each test among the two incineration system
discharge streams sampled, the kiln ash and the afterburner exit flue gas. Afterburner exit flue
gas distributions are based on the average concentrations for each test given in Table 17. It
bears noting that the three metals sampling train concentration measurements for each test gave
quite comparable results for each metal.
Two sets of metal distributions are given in Table 18. The first set, comprising the
upper half of the table, represents the fraction of the metal fed in the spiked waste accounted
for by the respective discharge. This set of distributions also notes the total fraction of metal
fed measured in the two discharges. This total fraction represents the degree of mass balance
closure achieved for each metal. The second set of distributions, comprising the bottom half of
the table, represents fractions normalized to the total amount of each metal measured in the two
discharge streams analyzed. These normalized values represent fractions that would have
resulted had mass balance closure been 100 percent. Note that the sum of the normalized
fractions is, indeed, 100 percent.
Focusing on the percent-of-fed distributions, the data in the upper half of Table 18 show
that metal mass balance closures ranged from 18 to 130 percent, with an average closure of
59 percent and median of 39 to 56 percent. That most closures achieved are less than
100 percent is expected because of flyash settling and accumulation in the horizontal afterburner.
As noted above, the accumulated ash in the afterburner was removed after Test 10 and after the
combination of Tests 7 and 14. This ash was assumed to have the same metal concentrations
as bottom ash, and a proportional contribution to the bottom ash factions was included in the
distributions in Table is. However, it is possible that not all of the ash that accumulated in the
afterburner was recovered. This would likely have been the case if some ash fusion occurred
leading to an accumulation of slag in the afterburner. Because slag is difficult to remove, it is
48
-------
TABLE 16. FLUE GAS PCDD/PCDF ANALYSIS RESULTS
Afterburner exit flue gas concentration,
ng/dscm
Analyte
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,6,7,8-HpCDD
OCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
3,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDF
Total TCDD
Total PeCDD
Total HxCDD
Total HpCDD
Total TCDF
Total PeCDF
Total HxCDF
Total HpCDF
Total PCDD/PCDF
2,3,7,8-TCDD toxicity equivalents
Test 7
< 0.004
< 0.025
< 0.019
< 0.013
0.013
0.047
0.138
0.119
<0.019
0.075
0.103
0.038
0.081
< 0.015
0.084
0.028
0.144
0.030
0.081
0.041
0.059
0.210
0.312
0.344
0.097
1.46
0.097
Test 10
< 0.003
< 0.009
< 0.006
0.003
0.006
0.031
0.097
0.034
0.014
0.023
0.031
0.011
0.028
0.003
0.037
0.020
0.094
0.006
0.020
0.020
0.031*
0.123
0.088
0.120
0.094
0.69
0.033
Test 14
< 0.006
<0.013
< 0.009
0.006
< 0.006
0.051
0.102
0.074
0.013
0.058
0.074
0.023
0.070
< 0.009
0.080
0.023
0.141
0.013
0.022
0.040
0.096
0.289
0.282
0.273
0.180
1.44
0.064
Test 18
< 0.001
<0.015
0.012
0.009
0.015
0.078
0.330
0.126
0.033
0.099
0.121
0.048
0.108
< 0.009
0.014
0.048
0.332
0.015
0.069
0.045
0.145
0.393
0.785
0.513
0.332
2.96
0.116
49
-------
TABLE 17. TRACE METAL CONCENTRATIONS IN FEED AND DISCHARGE SAMPLES
Concentration
Simulated waste feed,
mg/kg
Test?
Kiln ash, mg/kg
Afterburner exit flue gas,
mg/dscm
Train 1
Train 2
Train 3
Average
Test 10
Kiln ash, mg/kg
Afterburner exit flue gas,
mg/dscm
Train 1
Train 2
TrainS
Average
Test 14
Kiln ash, mg/kg
Afterburner exit flue gas,
mg/dscm
Train 1
Train 2
TrainS
Average
Antimony
730
4,160
2.86
4.29
3.93
3.69
1,540
2.71
2.35
2.21
2.42
.6,990
4.31
4.63
3.62
4.19
Barium
10,000
61,100
8.5
14.2
7.9
10.2
46,200
2.45
3.00
2.62
2.69
29,900
5.75
5.48
7.29
6.17
Chromium
550
2,200
0.57
0.29
1.41
0.76
3,050
0.29
0.51
0.48
0.43
1,690
0.32
0.32
0.68
0.44
Lead
9,000
67.1
44.8
49.1
54.7
49.4
14.6
94.7
58.6
53.8
69.0
1,620
75.9
72.9
65.5
71.4
Mercury
590
0.03
5.10
8.61
7.82
7.18
0.04
10.6
10.3
10.3
10.4
0.1
8.54
8.05
9.43
8.67
Thallium
730
0.1
93
10.5
11.3
10.4
*
0.1
13.1
11.8
11.2
12.0
0.1
14.7
14.4
11.7
13.6
50
-------
TABLE 18. TRACE METAL DISTRIBUTIONS
Discharge stream
Test 7, kiln temperature:
Kiln ash
Afterburner exit flue gas
Total
Test 10, kiln temperature:
Kiln ash
Afterburner exit flue gas
Total
Test 14, kiln temperature:
Kiln ash
Afterburner exit flue gas
Total
Antimony
1,000°C (1,832°F)
29
36
65
767°C (1,413°F)
10
16
26
874°C (1,60S°F)
45
40
85
Barium Chromium
31
7
38
21
2
23
14
4
18
Distribution, °/
20
10
30
25
5
30
14
6
20
Lead
'c of metal
<0.04
39
39
<0.01
32
32
0.8
55.2
56
Mercury
fed
<0.01
85
85
<0.01
94
34
<0.01
102
102
Thallium
<0.01
100
100
<0.01
82
82
<0.01
130
130
Distribution, % of metal measured
Test 7, kiln temperature:
Kiln ash
Afterburner exit flue gas
Test 10, kiln temperature:
Kiln ash
Afterburner exit flue gas
Test 14, kiln temperature:
Kiln ash
Afterburner exit flue gas
1,000°C (1,832°F)
45
55
767°C (1,413°F)
37
63
874°C (1,605°F)
53
47
81
19
93
7
77
23
68
32
84
16
72
28
0.1
99.9
<0.01
100
1.5
98.5
<0.01
100
<0.01
100
<0.01
100
<0.01
100
<0.01
100
-
<0.01
100
51
-------
possible that not all the metal accumulated in the afterburner was recovered. In addition, the
assumption that the afterburner accumulation has the same composition as bottom ash may not
be true. Metal volatilization, and subsequent condensation on available flyash, may have led to
metal enrichment of the flyash.
Mass balance closures achieved for mercury and thallium were quite good, at 85 to
130 percent. Essentially all of the mercury and thallium fed was accounted for in the afterburner
exit flue gas. This confirms the expectation that these very volatile metals are vaporized in the
kiln and carried out in the vapor phase through the higher temperature afterburner. These
metals would not be expected to be present in any unrecovered afterburner accumulation. Thus,
this route of possible loss would not have affected mass balance closure for these metals. Mass
balance closures for barium, chromium, and lead were uniformly in the 20 to 40 percent range.
Closures for antimony were better, at 26 to 85 percent.
Focusing on the normalized distribution data in the bottom half of Table 18 gives a
clearer picture of metal partitioning in the tests. This clearer picture occurs because variable
mass balance closure is removed as a source of test-to-test data variability. Highly variable and
less than perfect mass balance closure has been the universal experience in attempts to measure
metal partitioning in combustion sources. Because this variable and less than perfect mass
balance closure is invariably experienced, the use of normalized distributions represents a best
attempt to quantify metal partitioning phenomena.
The normalized distributions in Table 18 show that, in addition to mercury and thallium,
lead exhibited quite volatile behavior in all three conditions tested. Essentially all, greater than
98.5 percent, of the lead discharged from the RKS was accounted for in the afterburner exit flue
gas. At most only 1.5 percent was accounted for in the kiln ash at the intermediate kiln
temperature tested, 874°C (1,605°F). Barium and chromium were relatively nonvolatile under
the conditions tested; 77 to 93 percent of the barium and 68 to 84 percent of the chromium
discharged remained with the kiln ash. The remaining fractions accounted for in the afterburner
exit flue gas most likely represent the amount of metal contained in entrained flyash. Antimony
exhibited intermediate volatility behavior, and was relatively evenly distributed between the two
discharges. The variations in kiln (and afterburner) temperatures tested appear not to have
affected any metal's partitioning tendencies.
43 CONCLUSIONS
Test program conclusions include the following:
Afterburner exit flue gas particulate levels ranged from 175 to 743 mg/dscm
corrected to 7 percent O2. In general, higher particulate levels were associated
with higher waste feedrates.
Greater than 99.9984 percent POHC (hexachlorobenzene and hexachloroethane)
DREs were achieved as measured in the afterburner exit flue gas under the three
operating conditions in which POHC-spiked waste was tested. Neither POHC was
found above method detection limits for any test in which flue gas was sampled
for POHCs.
52
-------
Scrubber exit flue gas total PCDD/PCDF levels ranged from 0.69 to 2.96 ng/dscm,
well below a recent EPA guidance level of 30 ng/dscm. 2,3,7,8-TCDD toxicity
equivalent levels were 0.033 to 0.116 ng/dscm, just at or below the EC directive
level of 0.1 ng/m3. The highest PCDD/PCDF levels occurred in a test in which
no POHCs were spiked into the feed. Lower levels were measured in the tests
with POHC-spiked feed.
Lead, mercury, and thallium exhibited quite volatile behavior over the range of
kiln temperatures tested exit gas at 767°, 874°, and 1,000°C (1,413°, 1,605°, and
1,832°F). Essentially all of the discharged amount of each metal was found in the
afterburner exit flue gas, and essentially none in the kiln discharge.
Barium and chromium exhibited relatively nonvolatile behavior, with 77 to
93 percent of the barium and 68 to 84 percent of the chromium discharged in the
kiln ash. The 7 to 19 percent of the barium and 16 to 32 percent of the chromium
accounted for in the afterburner exit flue gas most likely represents respective
amounts in entrained flyash.
Antimony exhibited intermediate volatility behavior, being relatively evenly
distributed between the two discharge streams analyzed
53
-------
SECTION 5
TESTING OF FLUFF WASTE AND CONTAMINATED SOIL
FROM THE M. W. MANUFACTURING SUPERFUND SITE
One of the IRF's primary missions is to support Regional Offices in evaluations of the
potential of incineration as a treatment option for wastes and other contaminated materials at
Superfund sites. One priority site is the M. W. Manufacturing site in Danville, Pennsylvania.
EPA Region HI (B. Khona, Remedial Project Manager) and the U.S. Army Corps of Engineers
(USAGE) (H. Santiago, Project Manager) requested that a pilot-scale test program be conducted
at the IRF to support evaluations of the suitability of incineration as a treatment technology for
wastes and contaminated soil at the site. FY93 efforts in support of this test program are
discussed in this section.
5.1
BACKGROUND
The M. W. Manufacturing site began operation in 1966. M. W. Manufacturing
Corporation reclaimed copper from scrap wire using both mechanical and chemical processes.
Reclamation activities began in 1969 and continued until 1972 when M. W. Manufacturing filed
for bankruptcy. Warehouse 81, Inc., acquired the site in 1976 and began mechanical recovery
operations from the existing waste piles onsite. The mechanical recovery operations-generated
large volumes of waste material, termed fluff. The chemical recovery processes used by M. W.
Manufacturing also led to site contamination.
The fluff waste produced by the mechanical stripping process consists of fibrous
insulation material mixed with plastic. Phthalate esters, copper, and lead are the major
contaminants in this material. The chemical recovery process used by M. W. Manufacturing was
a two-step process. The first step involved the use of a hot oil bath to melt the plastic insulation
away from the metal in the scrap wire. Residual oils were removed from the separated copper
in the second step through the use of chlorinated solvents, including trichloroethene and
tetrachloroethene. Thus, these solvents are waste and soil contaminants at the site.
The June 1990 record of decision (ROD) document for the site identified five wastes
and contaminated materials for remedial treatment:
Fluff waste piles
Organic- and trace-metals-contaminated surface soils
Organic- and trace-metals-contaminated subsurface soils
54
-------
Lagoon water
Contents of drums and tanks
Onsite incineration was identified as the selected treatment for the fluff and the soil. The
remedy includes possible stabilization of the incineration ash prior to landfill disposal.
In subsequent efforts, a series of treatability studies was completed to aid in the
selection of incineration system design and operating variables. Initially a set of muffle furnace
tests was performed using site wastes. The test results suggested that effective organic
decontamination and destruction could be achieved at incineration temperatures of 871° and
982°C (1,600° and 1,800°F) and perhaps at as low as 760°C (1,400°F). The residual ash from
the muffle furnace tests of both fluff waste and a blend of fluff and site soil exhibited the toxicity
characteristic for lead.
The findings of the muffle furnace tests were used to guide a series of small pilot-scale
tests. These small pilot-scale batch-mode tests were performed to further aid in the selection
of the best set of incinerator operating parameters. The small pilot-scale tests confirmed that
871 °C (1,600°F) appeared to be the optimum incineration temperature. However, these tests
raised several issues, including the following:
PCDDs/PCDFs were apparently formed during the incineration of both fluff
waste and a fluff/soil mixture. However, this observation may have been an
artifact of the system's design and operating mode, which used only air quenching
of combustion gas to the 360° to 475 °C (680° to 887 °F) range. Rapid quenching
of flue gas to temperatures below 90 °C (200 °F) using water may prevent
PCDD/PCDF formation.
Trace metal mass balance closures were poor for all tests
The fabric filter (baghouse) collection efficiencies for both particulate and metals
were 90 to 94 percent, below the expected 98 percent
To extend the data base on the incineration characteristics of the fluff waste and
contaminated site soil, Region III requested that pilot-scale testing on a larger scale, more in
keeping with the size of incinerator envisioned for use at the site, be performed. Testing at this
larger scale would provide flue gas emission and ash residue characteristics data that will better
reflect expected actual conditions during the site remediation. Thus, a series of large pilot-scale
incineration- tests at the IRF was designed to supply these data. The specific data quality
objectives (DQOs) of the IRF tests were defined as follows:
Verify that the fluff waste and the contaminated soil at the site can be incinerated
in compliance with the hazardous waste incinerator performance standards and
permit requirements of:
99.99 percent principal organic hazardous constituent (POHC) DRE
55
-------
5.2
HC1 emissions less than 1 percent of the APCS inlet flowrate or 1.8 kg/hr,
whichever is greater
CO emissions of less than 100 ppm at 7 percent O0, 1-hour rolling average
and the recently announced performance guidance of:
Particulate emissions of less than 34 mg/dscm (0.015 gr/dscf) corrected to
7 percent O2
Total tetra- through octa-PCDD/PCDF emissions of less than 30 ng/dscm
corrected to 7 percent O2
Measure the effectiveness of incineration treatment in decontaminating fluff and
soil of their organic contaminants and evaluate whether incineration temperature
affects the effectiveness of fluff decontamination
Measure the distribution of the contaminant trace metals in the fluff and the
contaminated soil among the incineration system discharge streams
Determine whether the bottom ash residue and the APCS discharges from the
incineration of fluff and contaminated soil will be toxicity characteristic (TC)
hazardous wastes
Determine whether the bottom ash residue from the incineration of contaminated
soil meets the cleanup levels for soil given in the ROD
TEST PROGRAM
5.2.1 Waste Description
Data on the contaminant concentrations in the fluff waste and the surface and
subsurface soil, taken from the ROD, are summarized in Table 19. Only contaminants present
at an average concentration of 1 mg/kg or greater in one or more contaminated site matrix are
listed in the table. The data in Table 19 show that the major site contaminants are the two
phthalate esters, bis (2-ethylhexyl) phthalate (BEHP) and di-n-octyl phthalate (DNOP). Thus,
these compounds would be considered the POHCs in the site wastes. In addition, Region III
was interested in establishing that tetrachloroethene is effectively destroyed by incineration, so
tetrachloroethene was also defined to be a POHC. Site wastes were also highly contaminated
with copper and lead, and with lesser, though still significant amounts of antimony, barium,
chromium, nickel, and zinc. PCB-1254 was also reported in the ROD to be present at an
average of 9.4 mg/kg in the fluff. Site soil PCB levels were significantly lower.
Samples of the fluff waste and surface and subsurface soil were sent to the IRF for
characterization analyses. Results of the analyses are summarized in Table 20. Presuming that
the characterization samples analyzed typify the test materials to be received for testing in this
test program, the soil material for testing will have contaminant concentrations in the range
reported in the ROD for site soils. Contaminant concentrations in the fluff material for testing
56
-------
TABLE 19. M. W. MANUFACTURING SITE WASTE CONTAMINANTS FROM THE ROD
Concentration, nig/kg
Fluff
Contaminant
Volatile Organic Constituents
2-Butanone
Tetrachloroethene
Trichloroethene
1,1,2-TrichIoroethane
1,2-Dichloroethene
Methylene chloride
Semivolatile Organic Constituents
Bis(2-ethylhexyl)phthalate
Di-n-octyl phthalate
Di-n-butyl phthalate
PCB-1254
Trace Metals
Antinomy
Barium
Cadmium
Chromium
Copper
Lead
Nickel
Silver
Zinc
Range over
17 samples
2.8-6.4
0.72-18.0
up to 7.7
72,000-230,000
1,800-13,000
-
0.90-18.1
80-143
20-232
0.65^t.4
24-59
5,910-130,000
1,600-3,600
4.1-15
1.6-5.7
135-2,580
Average
1.6
4.4
0.45
149,000
4,400
9.4
65
93
2.4
40
50,000
2,400
5.6
1.8
620
Surface
Range over
21 samples
a
0.023-67
0.002-21
0.003-2.8
0.002-10
up to 0.83
3.9-3,000
0.2-140
0.48
0.061-3.7
62-118
22-107
1.2-12
7.1-59
742-171,000
32-9,770
8.5-40
8.6
55-787
soil
Average
10
1.0
0.28
0.49
0.04
836
37
0.02
0.21
16
74
2.0
27
21,600
1,450
22
0.4
240
Maximum average
subsurface soil
Range
up to 3.9
0.001-l,600b
0.002-2.6
up to 5.4
0.004-0.58
0.30-30,000b
0.038-150
0.036-130
0.077-1.0
47-218
1-13
14-70
24-38,900
7-741
42-50
56-319
Average
0.78
56
2.7
1.1
0.04
1,480
7,850
3.9
0.043
107
1
20
1-.850
160
46
144
Depth,
ft
16-18
4-6
12-14
16-18
8-10
12-14
0-2
0-2
0-2
0-2
0-2
0-14
12-14
16-18
6-8
6-8
a = not reported.
Maximum value in range represents an estimated value above minimum detection limit but below lowest calibration standard ("J"
flag). , .
57
-------
TABLE 20. M. W. MANUFACTURING SITE CHARACTERIZATION SAMPLE
ANALYSIS RESULTS
Sample
Parameter
Fluff Surface soil Subsurface soil
Characterization
Moisture, % 7.7 18 9.8
Ash, %
at550°C 41 77 89
at900°C 14 76 90
Heating value, MJ/kg 6.50 0.07 Will not burn
(Btu/lb) (2,800) (30)
Volatile organic constituents, mg/kg
Tetrachloroethene 146 69 18
1,1,2-Trichloroethane 4.8 1.5 NDa
Semivolatile organic constituents, mg/kg
BEHP 124,000 47.6 4.62
DNOP 17,800 1.95 ND
Trace metals, mg/kg
Antimony
Barium
Cadmium
Chromium
Copper
Lead
Nickel
Silver
Zinc
230
64
3.5
57
31,000
2,700
6.1
4.0
890
51
60
<0.2
30
8,300
1,800
15
<0.4
76
<5
78
0.93
21
160
180
31
<0.4
62
aND = not detected.
are in the range reported in the ROD for most contaminants. However, the fluff
characterization sample contained substantially more tetrachloroethene, 1,1,2-trichloroethane,
DNOP, and antimony than did fluff samples reported in the ROD.
Characterization samples received were also analyzed for hazardous waste
characteristics. Results are summarized in Table 21.
The two semivolatile POHCs in site materials, BEHP and DNOP, are poor candidates
for testing the incineration process with regard to destroying other site waste organic
contaminants because they are ranked as relatively easy to thermally destroy compounds in the
58
-------
TABLE 21. M. W. MANUFACTURING SITE CHARACTERIZATION SAMPLE
HAZARDOUS WASTE CHARACTERISTICS ANALYSIS RESULTS
Characteristic
Reactivity -S, mg/kg
Reactivity -CN, mg/kg
Corrosivity, pH
Ignitability, °F
TCLP leachate, mg/L
Arsenic
Barium
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Silver
Zinc
Pesticides and other organics
Fluff waste
<0.01
<0.01
6.85
>200
<0.10
0.10
0.11
0.11
199
3.1
< 0.002
<0.01
<0.10
<0.01
5.1
NDb
Surface soil
<0.01
<0.01
8.15
>200
<0.10
0.37
0.07
0.09
158
3.2
< 0.002
<0.01
<0.10
<0.01
0.49
ND
Subsurface soil
<0.01
<0.01
6.37
>200
<0.10
0.21
0.07
0.09
1.88
0.20
< 0.002
<0.01
<0.10
<0.01
0.12
ND
Regulatory
level
Contains
and reacts
<2, >12
<140
5.0
100
1.0
5.0
a
5.0
0.2
«
1.0
5.0
a = no regulatory level.
bND = not detected at detection limits ranging from O.Q04 to 0.01 mg/L.
thermal stability based incinerability ranking.4 This ranking groups the 333 compounds ranked
into seven stability classes from most stable, or most difficult to destroy (Class i), to least stable
or easiest to destroy (Class 7). Both BEHP and DNOP are ranked in Class 6, or relatively easy
to destroy. To present a challenge to the incineration process and develop data that suggest
incineration is capable of achieving sufficient DREs for other site organic contaminants, it was
decided to spike the test waste with naphthalene, a Class 1 (most difficult to destroy) POHC.
The spiking level was defined to be 2 percent by weight. In addition, it was decided to spike the
volatile POHC of greatest interest to Region HI, tetrachloroethene, into test materials at a level
of 4,000 mg/kg by weight. Tetrachloroethene is a Class 2 POHC. Spiking is needed because
site material concentrations of tetrachloroethene, as confirmed by the characterization samples,
are too low to allow establishing 99.99 percent DRE at achievable flue gas concentration
quantitation limits.
5.2.2 Test Conditions
The test program defined will consist of six tests. All tests are to be conducted in the
RKS at the IRF. The configuration of the RKS shown in Figure 18 will be used. This
configuration includes the fabric filter APCS, installed as discussed in Section 9.
59
-------
66-890 03V
oo
TH
2
60
-------
In the planned test program, two sets of duplicate tests feeding fluff waste alone and one
set of duplicate tests feeding contaminated soil alone will be performed. The two sets of fluff
. feed tests will be conducted at different kiln temperatures. Soil and fluff will be tested separately
because the eventual site remediation may treat each material separately for logistical reasons.
In addition, Region III would like data to determine whether the ash from incinerated soil alone
meets the cleanup levels given in the ROD. The test matrix will be as shown in Table 22.
For all tests the kiln exit flue gas O2 will be nominally 10 percent, afterburner exit gas
temperature 1,090°C (2,000°F), and afterburner exit flue gas O2 nominally 8 percent. The
venturi/packed-column scrubber and baghouse APCS units will be operated at their normal
design settings. Kiln rotation rate will be set to give a 30-minute kiln solids residence time.
Test material feedrate will be 54.5 kg/hr (120 Ib/hr) for all tests. Test materials will be
fed to the RKS via the fiberboard container ram feed system. This system sequentially batch
feeds 1.5-gal (5.7-L) fiberboard containers (cubical cardboard boxes for these tests) to the kiln.
It is.planned that, for all tests, the scrubber system will be operated at its design settings,
listed in Table 10, and at as close to total recirculation (zero to minimum blowdown) as possible.
For the fluff waste tests, kiln ash will be continuously deposited in initially clean 20-gal (76-L)
drums placed in the RKS ash pit. For the soil tests, kiln ash will be continuously removed from
the kiln ash hopper via the ash auger transfer system on the kiln and deposited into initially
clean 55-gal (208-L) drums.
5.2.3 Sampling and Analysis Procedures
The RKS sampling locations and the scope of the planned sampling effort are shown in
the process schematic given in Figure 19. For all tests, the planned sampling matrix will entail:
Obtaining a composite sample of the test material feed
Obtaining a composite sample of the kiln ash discharge
Obtaining a composite sample of the pre-test and post-test scrubber system liquor
TABLE 22. TEST MATRIX
Test
Feed
Kiln exit gas temperature,
°C (°F)
1
2
3
4
5
6
Fluff
Duplicate of Test 1
Soil
Duplicate of Test 3
Fluff
Duplicate of Test 5
870 (1,600)
870 (1,600)
750 (1,400)
61
-------
16-itt QS3
X X
X
if
1
X
62
X
X
191)
X
X
X
X
o>
X X
X X
X X
X
X
II
X
i.
,tii w
cd * S>
i i
11
f
-------
Obtaining a composite sample of the baghouse ash
Continuously measuring O2 concentrations in the kiln exit flue gas; O2, CO2, and
NOX in the afterburner exit flue gas; CO, CO2, NOX, and TUHC concentrations
in the baghouse exit flue gas; and O2 and CO concentrations in the stack gas
Sampling flue gas at the baghouse exit for trace metals using the EPA multiple
metals train
v..
~ Sampling flue gas at the baghouse exit for the waste and spiked POHCs using a
Method 0010 train
Sampling flue gas at the baghouse exit for the waste and spiked volatile organic
contaminants using Method 0030 (VOST)
Sampling the flue gas at the baghouse exit for PCDDs/PCDFs using Method 23
Sampling the baghouse exit and the stack for particulate and HC1 using Method 5;
the stack sample is needed to comply with the IRF's permit requirements
Test program samples will be analyzed for matrix-specific combinations of semivolatile
POHCs, volatile organic contaminants, semivolatile target compound list (TCL) organic
constituents, volatile TCL organic constituents, PCDDs/PCDFs, contaminant trace metals, and
chloride. The analysis matrix is summarized in Table 23.
53
CURRENT STATUS
All planning and preparatory efforts toward initiating the planned test program were
completed during FY93. Specifically:
A preliminary test plan outline was prepared and distributed in November 1992.
A revised test plan outline with revisions to address Region HI comments was
distributed in December 1992.
A complete test plan was prepared and distributed in April 1993
A stand-alone quality assurance project plan (QAPP) was prepared and
distributed in September 1993
Installation of the baghouse system required for the tests by Region HI and
discussed in Section 9 was completed in September 1993
At the close of FY93, all preparations needed to allow scoping tests to be initiated in mid-
October, with planned tests following in November, were nearly complete.
63
-------
TABLE 23. ANALYSIS PROTOCOL
Sample
Test feed
material
Test feed TCLP
Icachate
Feed packaging
material
Feed packaging
material ash
*SW-846.5
Parameter
Proximate analysis
(moisture, volatile matter,
fixed carbon, ash)
Elemental analysis
C, H, O, N, S
Cl
Heating value
Test semivolatile POHCs
Test volatile organic
contaminants
PCDDs/PCDFs
Trace metalsb
TCLP extraction
Trace metalsb
Proximate analysis
(moisture, volatile matter,
fixed carbon, ash)
Elemental analysis
C, H, O, N, S
Cl
Heating value
Test semivolatile POHCs
Test volatile organic
contaminants
PCDDs/PCDFs
Trace metalsb
Ash preparation
Test semivolatile POHCs
Test volatile organic
contaminants
Trace metalsb
TCLP extraction
Analysis method
ASTM D-5142
ASTM D-3176
ASTME-442
ASTM D-3286
Soxhlet extraction by Method 3540A, GC/MS
analysis by Method 8270Aa
Purge and trap GC/FID of methanol extract by
Method 8015A8
GC/MS by Method 8290a
Digestion by the multiple metals filter method0,
ICAP analysis by Method 6010A8
Method 1311a
Digestion by Method 3010A, ICAP analysis by
Method 6010A"
ASTM D-5142
ASTM D-3176
ASTM E-442
ASTM D-3286
Soxhlet extraction by Method 3540A, GC/MS
analysis by Method 8270A8
Purge and trap GC/FID of methanol extract by
Method 8015Aa
GC/MS by Method 8290°
Digestion by the multiple metals filter method0,
ICAP analysis by Method 6010A3
Per ASTM D-5142
Soxhlet extraction by Method 3540A, GC/MS
analysis by Method 8270Aa
Purge and trap GC/FID of methanol extract by
Method 8015A8
Digestion by the multiple metals filter method0,
ICAP analysis by Method 6010A"
Method 1311"
Frequency
1 composite for
each test material
1 composite for
each test material
I/fluff test,
1 composite soil
I/fluff test,
1 composite soil
1 composite for
each test material
I/fluff test,
1 composite soil
I/fluff test,
1 composite soil
I/fluff test,
1 composite soil
1 composite
1 composite
1 composite
1 composite
I/fluff test,
1 composite soil
1 composite for
each test material
I/fluff test,
1 composite soil
1 composite
.1 composite
1 composite
1 composite
1 composite
(continued)
bAs, Sb, Ba, Cd, Cr, Cu, Pb, Ni, Ag, and Zn.
C40 CFR 266, App. IX.6
64
i
-------
TABLE 23. (continued)
Sample
Parameter
Analysis method
Frequency
Feed packaging . Trace metalsb
material ash
TCLP leachate
Digestion by Method 3010A, ICAP analysis by
Method 6010Aa
Kiln ash
Kiln ash TCLP
leachate
Pre-test
scrubber liquor
Post-test
scrubber liquor
Scrubber liquor
TCLP leachate
Baghouse ash
Baghouse ash
TCLP leachate
Test semivolatile POHCs
Test volatile organic
contaminants
PCDDs/PCDFs
Trace metalsb
TCLP extraction
Trace metalsb
Test semivolatile POHCs
Test volatile organic
contaminants
Trace metalsb
PCDDs/PCDFs
Test semivolatile POHCs
Test volatile organic
contaminants
PCDDs/PCDFs
Trace metalsb
TCLP extraction
Trace metalsb "
Test semivolatile POHCs
Test volatile organic
contaminants
PCDDs/PCDFs
Trace metalsb
TCLP extraction
Trace metalsb
Soxhlet extraction by Method 3540A, GC/MS
analysis by Method 8270A8
Purge and trap GC/FID of methanol extract by
Method 8015A8
GC/MS by Method 8290"
Digestion by the multiple metals filter method0,
ICAP analysis by Method 6010Aa
Method 1311a
Digestion by Method 3010A, ICAP analysis by
Method 6010A8
Extraction by Method 3520A, GC/MS analysis by
Method 8270A8
Purge and trap by Method 5030A, GC/FID by
Method 8015A8
Digestion by Method 3010A, ICAP analysis by
Method 6010A8
GC/MS by Method 82908
Extraction by Method 3520A, GC/MS analysis by
Method 8270A8
Purge and trap by Method 5030A, GC/FID
analysis by Method 8015A8
GC/MS by Method 82908
Digestion by Method 3010A, ICAP analysis by
Method 6010A8
Method 13118
Digestion by Method 3010A, ICAP analysis by
Method 6010A8
Extraction by Method 3520A, GC/MS analysis by
Method 8270A8
Purge and trap GC/FID of methanol extract by
- Method 8015A8
GC/MS by Method 82908
Digestion by the multiple metals filter method0,
ICAP analysis by Method 6010A8
Method 13118
Digestion by Method 3010A, ICAP analysis by
Method 6010A8
I/test
I/test
I/test
I/test
I/test
I/test
I/test
I/test
I/test
1 sample before
the first test
I/test
I/test
I/test
I/test
I/test
. I/test
I/test
I/test
I/test
I/test
I/test
I/test
8SW-846.5
bAs, Sb, Ba, Cd, Cr, Cu, Pb, Ni, Ag, and Zn.
C40 CFR 266, App. IX*
(continued)
65
-------
TABLE 23. (continued)
Sample
Baghouse exit
flue gas
Stack gas
Parameter
Scmivolatile TCL organics
Volatile TCL organics
PCDDs/PCDFs
Trace metalsb
Particulate
HC1
Particulate,
HC1
Analysis method
Soxhlet extraction of Method 0010 samples by
Method 3540A, GC/MS analysis by
Method 8270A*
Analysis of Method 0030 samples by Method 5040"
GC/MS of Method 23 samples by Method 8290s
Digestion of multiple metals train samples by
multiple metals procedure0, ICAP analysis by
Method 6010Aa
Method 3d
ICAP analysis of combined impinger solution by
Method 9057°
Method 5d
1C analysis of combined impinger solution by
Method 9057°
Frequency
I/test
3 trap pairs/test
I/test
I/test
I/test
I/test
I/test
I/test
"SW-846.5
bAs, Sb, Ba, Cd, Cr, Cu, Pb, Ni, Ag, and Zn.
C40 CFR 266, App. IX°
d40 CFR 60, App. A.7
66
-------
SECTION 6
EVALUATION OF THE SONOTECH FREQUENCY-TUNABLE PULSE
COMBUSTION TECHNOLOGY
Sonotech, Inc., of Atlanta, Georgia, has developed a pulse combustion burner technology
which claims to offer benefits when applied in a variety of combustion processes. The burner
system incorporates a pulse combustor which can be tuned to excite large amplitude sonic
pulsations inside a combustion process such as a boiler or incinerator. These pulsations serve
to increase the rates of heat, momentum (mixing), and mass transfer in the combustion process.
Sonotech claims that these heat, momentum, and mass transfer rate -increases are sufficiently
significant that faster and more complete combustion is accomplished.
Sonotech has targeted waste incineration as a potential application for the technology.
As an initial demonstration that its pulse combustion system offered claimed benefits, Sonotech
participated in an EPA Phase I Small Business Innovative Research (SBIR) program. In this
recently completed program, Sonotech retrofitted a pulse combustion burner to the EPA bench-
scale rotary kiln incinerator at the Air and Energy Engineering Research Laboratory (AEERL)
in Research Triangle Park, North Carolina. Tests which measured the effect of pulsations on
incinerator emissions of soot, CO, and TUHC were completed.
Based on the initial experience in the SBIR program, Sonotech proposed a followup
demonstration under the SITE program. Specifically, Sonotech proposed that its pulse
combustion technology be evaluated on a larger scale incineration system, the IRF RKS. Efforts
completed during FY93 to support this test program are discussed in this section.
6.1
DESCRIPTION OF THE TECHNOLOGY
The phenomenon of pulse combustion has been studied since 1900. Active study of the
phenomenon accompanied the advent and development of jet engines and rocket motors. The
focus of most research has been on avoiding pulse combustion because of its deleterious effect
on jet engine and rocket motor performance (e.g., structural failure). Perhaps the best known
commercial application which takes advantage of the combustion efficiency improvement aspects
of pulse combustion is the Lennox pulse furnace.
A pulse combustor typically consists of an air inlet, a combustor section, and a tailpipe.
In pulse combustion, fuel oxidation and heat release rates vary periodically with time. These
variations produce periodic variations (pulsations) in combustor section pressure, temperature,
and gas velocities. The frequency of pulsations is generally close to the resonant frequency of
the fundamental longitudinal acoustic mode of a duct consisting of the combustor section and
tailpipe. Thus, by changing combustor and tailpipe geometry, for example by varying the length
67
-------
of the tailpipe, the frequency of pulsations can be changed, or tuned. Further, if properly
applied, a pulse combustor can excite large amplitude (150 dB or higher) resonant pulsations
within a cavity downstream of the pulse combustor tailpipe. This cavity could be the combustion
chamber of a boiler or an incinerator, for example.
The periodic pulsations of pressure, gas velocities, and temperature give rise to
increased rates of mass, heat, and momentum (mixing) transfer in pulse combustion compared
to non-pulsating combustion. Thus, improved combustion efficiency and more complete organic
compound oxidation (destruction) might be expected. Such are the claims of this SITE
program's process developer. Sonotech claims that "large amplitude resonant pulsations excited
by a tunable pulse combustor [will] significantly improve the incinerator's performance by
increasing the rates of mass, momentum (i.e., mixing), and heat transfer within the incinerator.
These, in turn, increase the incineration rate, reduce the amount of air required for incineration,
reduce the severity of puffs, and reduce pollutant emissions. These improvements are expected
to reduce capital investment and operating costs of a wide variety of incineration systems and
improve their performance."
With the development of frequency-tunable pulse combustors to excite large amplitude
pulsations in combustion chambers downstream of the pulse combustor, it became possible to
apply pulse combustion to a variety of combustion processes such as boilers, dryers, calciners,
and incinerators. In such applications, the pulse combustor could be used as the combustion
process burner, supplying all of the heat input to the process. Alternatively, the pulse combustor
could be used only as the driver to excite pulsations in the combustion process. In such
applications, the pulse combustor needs deliver only a fraction, as little as 2 percent, of the
combustion process heat input, while still exciting resonant pulsations in the process combustor.
The remaining heat input would be suppk'ed via the normal process means, e.g., the process
conventional burner.
To excite large amplitude pulsations inside an incinerator, for example, the pulse
combustor must operate at a frequency that equals one of the natural acoustic mode frequencies
of the incinerator. When this condition is satisfied, the pulsations inside the pulse combustor
and the incinerator are in resonance. Resonant driving of large amplitude pulsations is achieved
by retrofitting a tunable pulse combustor to a wall of the incinerator, and varying its frequency
until one of the natural acoustic modes of the incinerator is excited. The desired resonant
operating condition is established in practice by using one or more pressure transducers to
monitor the changes in the amplitude of pulsations inside the incinerator in response to changing
the pulse combustor frequency. The desired operating condition is reached when these
transducers indicate that the ampk'tude of pulsations inside the incinerator has been maximized.
The first trial of the Sonotech technology was in the SBIR program noted above. In this
application, the pulse combustor was the sole heat input source to the bench-scale incinerator
at AEERL. In the study, the effect of exciting large amplitude pulsations in the incinerator on
the formation of transient "puffs" was studied. Transient puffs of incompletely burned organic
constituents arise in the kiln exit flue gas of incinerators when a batch charge of organic waste
demands more oxygen for complete destruction than available at the set kiln air feedrates. This
puff consists of CO, TUHC, and soot.
68
-------
Sonotech's interpretation of the findings of the SBIR program were that the excitation
of pulsations within an incinerator with a frequency-tunable pulse combustor improved its
performance as a result of increased mixing rates between the fuel and air and between reactive
gas pockets and ignition sources, and due to increased rates of heat and mass transfer to and
from the burning waste. These effects should reduce the amount of excess air required to
completely burn a waste, increase the POHC DREs, minimize the formation of products of
incomplete combustion (PICs), and eliminate or minimize the formation of detrimental puffs.
Sonotech expects to confirm the advantages of its frequency-tunable pulse combustion
technology in the SITE demonstration at the IRF. This demonstration will entail retrofitting the
kiln section of the RKS at the IRF with a pulse combustor "driver" which will deliver a heat input
of 73 kW (250,000 Btu/hr), or roughly 15 to 20 percent of the typical heat input to the kiln of
the RKS. The advantages Sonotech claims the application of the frequency-tunable pulse
combustion technology to an incineration system has over conventional (nonpulsating)
incineration are:
1. Lower CO, soot, and NOX emissions. The high mixing, mass, and heat transfer
rates within the incinerator produce complete combustion, which practically
eliminates soot and CO production. Furthermore, the periodic mixing process
and improved heat transfer minimize NOX production.
2. Lower combustion air requirements. The improved mixing within the incinerator
reduces the amount of air required for complete combustion.
3. Lower energy requirements. The improved heat transfer within the process
increases the rate of drying and heating the waste, which may increase the burn
rate and reduce the required fuel input into the system.
4. Higher incinerator capacity. The increase in mixing, mass, and heat transfer rates
produced by the pulsations within the incinerator accelerate the transfer of heat
and oxygen to the burning waste and CO away from the burning material, which
shortens the waste burning time. Consequently, the waste can be moved faster
through the incinerator, resulting in higher incinerator capacity.
5. Reduced severity of transient puffs. The improved mixing, mass, and heat transfer
within the incinerator eliminate or reduce temperature and concentration
nonuniformities within the process, resulting in better utilization of available
oxygen and more uniform incineration through the incinerator volume. This
lowers the likelihood of emissions of transient puffs due to the occurrence of
"rich" combustion within the incinerator.
6. Reduced incineration system capital and. operating costs due to 2, 3, and 4 above.
62 TEST PROGRAM
The objective of the demonstration test program is to develop the data needed to allow
objective and quantitative evaluation of the above Sonotech claims for their frequency-tunable
pulse combustion technology. The focus of the program will be on the evaluation of the claims
69
-------
that lower combustion pollutant emissions result from the application of the Sonotech
technology, and that higher incinerator capacity results. However, test program data will also
be taken to evaluate whether the Sonotech technology affects trace metal partitioning in the
incinerator, the leachability of trace metals from incinerator discharges, or the severity of
transient puffs.
To address the test program objectives, tests at four different incineration system
operating conditions will be performed. The four planned conditions are:
Conventional combustion:
1. Baseline, typical operation
* 2. Maximum waste feedrate
Sonotech pulse combustion
3. Same feedrate and conditions as 2
4. Maximum waste feedrate
The waste to be fed for all tests will be a relatively high heat content contaminated
material from a manufactured gas plant (MGP) Superfund site, specifically, the Peoples Natural
Gas Company site in Dubuque, Iowa. The test material will be batch fed to the RKS via the
fiberboard container ram feed system, which will feed 1.5-gal (5.7-L) cubical fiberboard
containers to the kiln. When a relatively high heat content material is being fed, the maximum
allowable waste feedrate is established based on the onset of puffs of incompletely combusted
organic constituents (CO and TUHC) which survive the afterburner. Thus, a feed regimen that
results in routine system exit flue gas CO spikes of over 300 ppm, lasting 30 to 60 seconds, which
in turn cause 1-hour rolling average flue gas CO levels to be about 50 ppm, is the maximum
feedrate that can be tolerated and still be characterized as barely acceptable operation.
Given this, Test 2 will be at a feedrate and other operating conditions which result in
frequent (every charge to every third charge) CO spikes, however with 1-hour rolling average
system exit CO levels of nominally 50 ppm or less. The 50 ppm rolling average maximum is
selected to ensure that the usual permit limit for hazardous waste incinerators (the IRF RKS
included) of 100 ppm, 1-hour rolling average, is not exceeded.
Test 1 will be at a decreased feedrate which results in only infrequent (less than one
every tenth charge) system exit flue gas CO spikes. This feedrate will be no greater than
80 percent of the Test 2 feedrate. Test 1 represents baseline, typical, well-controlled incinerator
operation.
Test 3 will be at the Test 2 feedrate and incineration temperatures; however, the
Sonotech pulse combustion system will be in operation. Test 4 will be at whatever increased
feedrate can be sustained with 1-hour rolling average system exit flue gas CO levels of 50 ppm
or less.
70
-------
This test matrix will allow the development of data to evaluate the Sonotech claims as
follows. Test 1 will provide emissions, POHC DRE, metals partitioning, and metals leachability
data corresponding to baseline, typical, well-controlled incinerator operation while feeding the
test contaminated material. Past IRF experience is that few CO spikes will occur, and those that
do will peak at less than 100 ppm. The 1-hour rolling average CO emissions should be less than
10 ppm. NOX emissions are expected to be in the 30 to 70 ppm range.
Test 2 will provide emissions, POHC DRE, metals partitioning, and metals leachability
data corresponding to an operating condition at the maximum contaminated material feedrate
possible under conventional incinerator operation. This operating condition will correspond to
borderline incinerator failure and noncomph'ance with typical permit limits. Frequent system exit
flue gas CO spikes of 300 to over 1,000 ppm are expected, and 1-hour rolling average CO
emissions will likely be about 50 ppm.
Test 3 will provide emissions, POHC DRE, metals partitioning, and metals leachability
data for the Test 2 operating condition with pulse combustion. Comparing the NOX, CO, and
TUHC emission data and POHC DREs for Test 3 to those for Test 2 will allow the evaluation
of claims that lower emissions result from the application of pulse combustion.
In addition, Sonotech expects that, with the improved combustion efficiency resulting
from the application of pulse combustion, acceptable incinerator operation (infrequent system
exit flue gas CO spikes) will be possible at a decreased combustion air requirement. With a
decreased combustion air supply, decreased auxiliary fuel will be needed to maintain incineration
temperatures. If acceptable operation with decreased burner air and auxiliary fuel feedrates can
be achieved for Test 3 at the Test 2 contaminated-material feedrate and incineration
temperatures, evaluation of these claims will be possible.
Test 4 will provide emissions, POHC DRE, metals partitioning, and metals leachability
data at the maximum incinerator feedrate possible under pulse combustion operation.
Comparing Test 4 emissions and POHC DRE data to those from Tests 2 and 3 will allow
evaluation of the claim that increased capacity at acceptable emissions is possible.
Each test will be performed three times (triplicate testing) to allow assessment of the
precision of each emission and discharge stream composition measurement.
6.2.1 Test Facility
As noted above, the test program will be conducted in the RKS at the IRF. A process
schematic of the RKS as it will be configured for these tests is shown in Figure 18. The RKS
will be retrofitted with a Sonotech frequency-tunable pulse combustor system with firing rate
capacity of 73 kW (250,000 Btu/hr) for the test program. The Sonotech burner will be mounted
into the stationary end wall at the ash pit end of the kiln section of the RKS. The retrofit system
will consist of a frequency-tunable pulse combustor, associated fuel and air flow controls, a
process control system, and an appropriate structural support. Fuel (natural gas) and air supply
will be taken from existing IRF supply lines. Appropriate safety interlocks between the Sonotech
control system and the RKS system will be defined and installed.
71
-------
62.2 Test Feed Material
The test material for the test program will be a mixture of pulverized coal and
contaminated sludge waste from the Peoples Natural Gas Company Superfund site in Dubuque,
Iowa. This site is an abandoned MGP site, and the sludge waste at the site contains high
concentrations of coal tar constituents.
The test feed mixture was prepared at the Peoples site by mixing and grinding a
combination of 30 to 35 percent sludge with 65 to 70 percent coal. After being screened through
a 2.5-in (6.4-cm) screen, the resulting mixture was a relatively free-flowing solid. After
preparation, the material was transferred to 20 55-gal (208-L) drums and shipped to the IRF in
late September 1993.
Samples of the test feed material taken from one of the drums received at the IRF were
analyzed for ash content, moisture content, heating value, semivolatile and volatile organic
constituents, and trace metals, and the TCLP leachate of the feed material was analyzed for
volatile organic constituents and trace metals. Results of these analyses are summarized in
Table 24. As shown, the test feed material contains several PAH compounds at concentrations
ranging from 100 to over 1,400 mg/kg. In addition, it contains several volatile aromatic
constituents at levels ranging from 3.7 to 55 mg/kg.
None of the feed material organic contaminants, however, is present in the feed material
mixture at concentrations high enough to allow a clear determination of whether 99.99 percent
DRE can be achieved. Thus, for the test program each feed-containing fiberboard container will
be spiked with naphthalene and benzene. The spiking level will be at 10,000 mg/kg. Both
compounds are feed-material contaminants and are ranked as very difficult to thermally destroy
in the thermal-stability-based incinerability ranking.4 Both compounds are ranked as class 1, the
most difficult to thermally destroy of seven classes, with class 7 being easiest to destroy.
A solution of 50 percent (weight) naphthalene dissolved in benzene (equal weights of
both) will be used to spike the feed material. Weighed quantities of the solution will be placed
in appropriately sized high-density polyethylene (HOPE) bottles with polypropylene screw
closures, and a bottle of the solution will be added to each feed container during feed
repackaging.
The data in Table 24 show that the test feed material contains antimony, barium,
beryllium, cadmium, chromium, lead, and mercury. Thus, metals partitioning will be measured
for these metals.
6.23 Test Conditions
Current plans are to perform all tests at the same overall incinerator operating
conditions, with a kiln exit gas temperature of 870°C (1,600°F) and an afterburner exit gas
temperature of 1,090°C (2,000°F). In Test 1, the baseline test under conventional combustion
conditions, and Test 4> the maximum feedrate test under pulse combustion operation, average
flue gas O2 levels are expected to be about 10 percent at the kiln exit and 8 percent at the
afterburner exit. Lower average kiln exit O2 levels will be experienced in Test 2, the maximum
feedrate test under conventional operation. Sonotech claims that lower kiln exit O2 levels
72
-------
TABLE 24. ANALYSIS RESULTS FOR THE TEST MATERIAL FEED PREPARATION
Analyte
Proximate analysis
Moisture, %
Ash, %
Higher heating value, MJ/kg
(Btu/lb)
Trace metals, mg/kg
Sb
As
Ba
Be
Cd
Cr
Pb
Hg
Se
Ag
Tl
Polynuclear aromatic hydrocarbons, mg/kg
Acenaphthylene
Anthracene
Benz(a)anthracene
Benzo(k)fluoranthene
Benzo(ghi)perylene
Benzo(a)pyrene
Chrysene
Dibenzofuran
Fluoranthene
Fluorene
Indeno(l,2,3;-cd)pyrene
2-Methylnaphthalene
Naphthalene
Phenanthrene
Pyrene
Volatile organic constituents, mg/kg
Benzene
Ethylbenzene
Styrene
, Toluene
Total Xylenes
Sample
concentration
13.8
27.2
17.7
(7,620)
17
<5
110
0.54
" 3.8
10
130
3.4
<0.7
<4
440
390
310
190
110
230
290
330
1,080
490
100
450
1,430
150
730
7.8,
3.7
10.1
21.7
55.2
TCLP
Concentration
,mg/L
a
<0.05
0.34
< 0.004
< 0.007
< 0.045
< 0.001
<0.075
< 0.007
'
. , mg/L
0.66
leachate
Regulatory level
5.0
100
1.0
5.0
5.0
0.2
1.0
5.0
0.5
m = Not measured.
73
-------
associated with decreased combustion air requirements will also be the case for Test 3, pulse
combustion operation at the Test 2 feedrate and temperature conditions. For all tests, kiln
rotation rate will be that required to provide a feed material solids residence time of 1 hour, and
the APCS components will be operated under their nominal design conditions.
The two test program variables will be test material feed regimen (feedrate and feed
charge frequency) and whether conventional or pulse combustion is in operation. The specific
feed regimens to be used for each test condition will be defined based on a series of scoping
tests. The focus of the scoping tests will be to identify the feed regimen which gives the
maximum waste feedrate for a given mode of operation (conventional or pulse combustion).
As noted above, test material will be batch-charged to the kiln in 1.5-gal (5.7-L)
fiberboard containers via the RKS's ram feed system. Past IRF testing experience has shown
that, depending the test material, puffs of uncombusted organic material which survive the
afterburner can be achieved by feeding batch charges of between 11 and 53 MJ (10,000 and
50,000 Btu) at a rate of 12 to 20 per hour, with kiln and afterburner gas temperatures at the
above-noted plan values.
Given the past experience, scoping tests will be initiated with containers charged with
2.27 kg (5.0 Ib) of test material (40.2 MJ [38,100 Btu] per container based on the sludge/coal
mixture heating value noted in Table 24). Charge feed frequency will start at 12 charges per
hour. Variations in charge frequency, and test material weight per charge will be tried until the
maximum test material feedrate, as defined in introductory paragraphs to Section 6.2, is
discovered for both conventional combustion and pulse combustion. For conventional
combustion operation, the Sonotech system will not be operated. For pulse combustion
operation, the Sonotech system will fire natural gas at its design rate, and pulse frequency will
be tuned to excite resonant pulses with amplitude greater than 160 dB in the RKS kiln.
It is possible that maximum test material feedrate for the Peoples Natural gas site
material may be constrained by kiln temperature at minimum kiln burner fuel feedrate, instead
of by high 1-hour rolling average CO emissions, as desired. The minimum total kiln burner
auxiliary fuel feedrate consistent with safe operation is about 150 kW (500,000 Btu/hr). This
minimum auxiliary fuel feed would be needed in the kiln main burner for the conventional
combustion tests, or divided between the main burner and the Sonotech burner for the pulse
combustion tests. The maximum total kiln heat input consistent with the target kiln operating
temperature of 870°C (1,600°F) and exit gas O2 of about 10 percent is approximately 530 kW
(1.8 MMBtu/hr). Thus, the maximum test material feedrate will be at a heat input rate of
380 kW (1.3 MMBtu/hr). If the test material mix has a heating value of 23 MJ/kg
(10,000 Btu/lb), then the maximum test material feedrate will be about 59 kg/hr (130 Ib/hr).
It is possible that a feed regimen that produces high 1-hour rolling average CO emissions with
test material feedrate so constrained cannot be found. If this is the case, the option of replacing
kiln main burner natural gas heat input by firing, or cofiring, a soot-producing solvent such as
toluene (a sludge contaminant) to stimulate CO puffs will be recommended. Past experience has
shown that puff production is enhanced when a soot-producing liquid waste or fuel that burns
with a radiant flame is fired, with or without auxiliary natural gas, in the kiln burners.
After appropriate RKS operating conditions have been defined for the four planned
tests, the evaluation testing will be initiated. As noted above, test material was prepared in one
74
-------
batch at the Peoples site and packaged into 20 55-gal (208-L) drums for shipment to the IRF.
Test material for the evaluation tests will be repackaged at the IRF into the 1.5-gal (5.7-L)
fiberboard containers, with container contents as specified from scoping test findings.
For a given evaluation test, the RKS will be brought to steady operation at the desired
conditions beginning on the prior day by firing natural gas only. Test material feed will then be
initiated and steady RKS operation reestablished with waste feed. Kiln and afterburner auxiliary
fuel-fired (natural gas) burner fuel and air flows, along with secondary combustion air flows, will
be controlled to give the desired temperature and excess air conditions. Flue gas sampling will
be started no sooner than 0.5 hour after the initiation of test material feed. Feed will continue
until all flue gas sampling has been completed.
For all tests, the scrubber system will be operated at its design settings, listed in
Table 10, and at as close to total recirculation (zero to minimum blowdown) as possible. Kiln
ash will be continuously removed from the kiln ash hopper via the ash auger transfer system on
the kiln .and deposited into initially clean 55-gal (208-L) drums.
6.2.4 Sampling and Analysis Procedures
The RKS sampling locations and the scope of the sampling effort are shown in the
.process schematic, Figure 20. For all tests, the sampling matrix will entail:
Obtaining a composite sample of the test feed material
Obtaining a composite sample of the kiln ash discharge
Obtaining a composite sample of the scrubber system liquor
Obtaining a composite sample of the baghouse flyash
Continuously measuring O2 concentrations in the kiln exit flue gas; O2, CO, CO2,
NOX, and TUHC concentrations in the afterburner exit flue gas; O2, CO2, and
NOX in the baghouse exit flue gas; and O2 and CO concentrations in the stack gas
Collecting a gram-sized sample of the afterburner exit particulate using a
high-volume Method 17 sampling train
Sampling flue gas at the baghouse exit for trace metals using the EPA multiple
metals sampling train, Method 29
« Sampling the flue gas at the baghouse exit for mercury using Method 101A
Sampling the flue gas at the baghouse exit for the semivolatile POHCs and other
PAH constituents using a Method 0010 train
Sampling the flue gas at the baghouse exit for PCDDs/PCDFs using Method 23
75
-------
'iGZ QS3
w-£ _,
1
X
X
X
X
X X
X X
X
X
X
X
X
X
X X
X
X
s,
I.
en
I
76
-------
Sampling the flue gas at the baghouse exit for volatile organic constituents using
the VOST, Method 0030
Sampling the flue gas at the baghouse exit and the stack for particulate and HC1
using Method 5; the stack sample is needed to comply with the IRF's permit
requirements
Test program samples will be analyzed for matrix-specific combinations of semivolatile
POHCs and PAH constituents, volatile organic constituents, PCDDs/PCDFs, contaminant trace
metals, total carbon, and chloride. The matrix of planned analyses is summarized in Table 25.
63 CURRENT STATUS
Much of the planning for this test program was completed during FY93. Specifically:
A test planning initiation meeting among RREL, PRC Environmental
Management (the RREL SITE program contractor), Sonotech, Industrial Gas
Technology Commercialization Center (the funding source for the development
of the Sonotech system), and IRF representatives took place at the IRF in
November 1992
A data quality objectives development meeting among RREL, RREL QA,
S-Cubed (an RREL QA support contractor), PRC, and IRF representatives took
place at RREL in February 1993
A test plan outline was prepared and distributed in June
A meeting among RREL, PRC, EPA Region VII, Midwest Natural Gas (the
company funding the remediation of the Peoples Natural Gas site), the Midwest
Natural Gas remediation contractor, and IRF representatives was held in
Dubuque, Iowa, in July to discuss test material preparation procedures and
shipment schedules
Test material was prepared, shipped, and received at the IRF in September
A draft of the stand-alone QAPP was prepared and distributed for review in
September
Further efforts culminating in installing the Sonotech combustion system and initiating the test
program in March 1994 will proceed during FY94.
77
-------
TABLE 25. ANALYSIS PROTOCOL
Sample
Test feed
material
Test feed TCLP
leachate
Kiln ash
Kiln ash TCLP
leachate
Pro-test
scrubber liquor
Parameter
Proximate analysis
(moisture, volatile matter,
fixed carbon, ash)
Elemental analysis
C, H, 0, N, S
Cl
Heating value
Semivolatile organic
constituents
Volatile organic
constituents
Trace metalsb
Mercury
TCLP extraction
Trace metalsb
Mercury
Semivolatile organic
constituents
Volatile organic
constituents
Trace metalsb
Mercury
TCLP extraction
Trace metalsb
Mercury
Semivolatile organic
constituents
Volatile organic
Analysis method
ASTM D-5142
ASIM D-3176
ASTME-442
ASTM D-3286
Soxhlet extraction by Method 3540A, GC/MS
analysis by Method 8270Aa
Purge and trap GC/MS by Method 8240A8
Digestion by the multiple metals filter method0,
GFAAS analysis by Method 7000Aa
Digestion and CVAAS analysis by Method 74718
Method 1311*
Digestion by Method 3015, GFAAS analysis by
Method 7000Aa
Digestion and CVAAS analysis by Method 7471"
Soxhlet extraction by Method 3540A, GC/MS
analysis by Method 8270A8
Purge and trap GC/MS by Method 8240A8
Digestion by the multiple metals filter method0,
GFAAS analysis by Method 7000Aa
Digestion and CVAAS analysis by Method 7471a
Method 1311"
Digestion by Method 3015, GFAAS analysis by
Method 7000A8
Digestion and CVAAS analysis by Method 7470
Extraction by Method 3520A, GC/MS analysis by
Method 8270Aa
Purge and trap GC/MS by Method 8240A8
Frequency
1 composite
1 composite
1 composite
2/test condition
2/test condition
2/test condition
2/test condition
2/test condition
2/test condition
2/test condition
I/test run
I/test run
I/test run
I/test -run
I/test run
I/test run
I/test run
I/test run
I/test run
constituents
Trace metalsb
Mercury
Digestion by Method 3015, GFAAS analysis by
Method 7000Aa
I/test run
Digestion and CVAAS analysis by Method 7470 I/test run
SW-846.5
bAs, Ba, Be, Cd, Cr, Pb, and TL
c40CFR266,App.K.
-------
TABLE 25. (continued)
Sample
Post-test
scrubber liquor
Scrubber liquor
TCLP leachate
Baghouse ash
Baghouse ash
TCLP leachate
Afterburner exit
particulate
Baghouse exit
flue gas
Stack gas
Parameter
Semivolatile organic
constituents
Volatile organic
constituents
Trace metalsb
Mercury
TCLP extraction
Trace metalsb
Mercury
Semivolatile organic
constituents
Volatile organic
constituents
Trace metalsb
Mercury
TCLP extraction
Trace metals"
Mercury
Semivolatile organic
constituents
Total carbon
Semivolatile organic
constituents
Volatile organic
constituents
PCDDs/PCDFs
Trace metalsb
Mercury
Particulate
HC1
Particulate
HC1
Analysis method
Extraction by Method 3520A, GC/MS analysis by
Method 8270A"
Purge and trap GC/MS by Method 8240Aa
Digestion by Method 3015, GFAAS analysis by
Method 7000Aa
Digestion and CVAAS analysis by Method 7470
Method 1311a
Digestion by Method 3015, GFAAS analysis by
Method 7000Aa
Digestion and CVAAS analysis by Method 7470
Soxhlet extraction by Method 3540A, GC/MS
analysis by Method 8270Aa
Purge and trap GC/MS by Method 8240A*
Digestion by the multiple metals filter method0,
GFAAS analysis by Method 7000Aa
Digestion and CVAAS analysis by Method 7471"
Method 1311a
Digestion by Method 3015, GFAAS analysis by
Method 7000A8
Digestion and CVAAS analysis by Method 7470
Soxhlet extraction by Method 3540A, GC/MS
analysis by Method 8270A
ASTM D-3178
Soxhlet extraction of Method 0010 samples by
Method 3540A, GC/MS analysis by
Method 8270Aa
Purge and trap GC/MS analysis of Method 0030
samples by Method 5040
GC/MS analysis of Method 23 samples by
Method 8290a
Digestion of multiple metals train samples by
multiple metals procedure0, GFAAS analysis by
Method 7000Aa
Sample preparation by Method 101Ad, CVAAS
analysis by Method 7470a
Method 5e
1C analysis of combined impinger solution by
Method 9057"
Methods'
1C analysis of combined impinger solution by
Method 9057*
Frequency
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
3 trap pairs/test
run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
I/test run
"SW-846.5
bAs, Ba, Be, Cd, Cr, Pb, and Ti.
°40 CFR 266, App. IX.6
d40 CFR 61, App. B.8
e40 CFR 60, App. A.7
79
-------
SECTION 7
TEST INCINERATION OF BALLISTIC MISSILE PROPELLANT COMPONENTS
Presidents Clinton and Yeltsin agreed in principle to a draft agreement on April 4,1993,
between the U.S. Department of Defense (DOD) and the Committee for Defense Industry of
the Russian Federation concerning cooperation in the elimination of Russian Strategic Offensive
Arms. This agreement was signed and became effective on August 26, 1993. The agreement
obligates the DOD to provide the Russian Federation with a list of equipment for the
Federation's possible use in eliminating strategic offensive arms in accordance with schedules
negotiated in the Strategic Arms Reduction Treaty (START). Included in this equipment list
are eight mobile incinerators each capable of destroying at least 750 metric tons per year of
ballistic missile liquid propellant comprised of unsymmetrical dimethyl hydrazine (UDMH) liquid
fuel and nitrogen tetroxide (N2O4) oxidizer.
The Defense Nuclear Agency (DNA) is responsible for providing the incinerators. One
requirement for the incinerators to be provided is that they meet both U.S. and Russian
Federation environmental regulatory requirements. Thus, to minimize the possibility of Russian
rejection of the incinerators, and the associated adverse effects on the schedule for the
destruction of the ballistic missile propellant, the DNA has sponsored a series of tests at the IRF
to demonstrate that Russian ballistic missile propellant can be incinerated in compliance with
U.S. and Russian environmental laws and regulations. These tests will also supply the design
information required to commercially procure the eight incinerators.
The general objectives of the initial tests are to:
Demonstrate the U.S. and Russian environmental certifiability of the incineration
of former Soviet Union ballistic missile fuel UDMH
Demonstrate the U.S. and Russian environmental certifiability of the incineration
of former Soviet Union ballistic missile oxidizer N2O4
Environmental certifiability will be established by showing that both UDMH and N2O4 can be
separately destroyed in an incinerator to levels which meet both U.S. and Russian environmental
regulations, while resulting in emissions of incineration by products considered acceptable under
those regulations. Efforts completed in FY93 to support this test program are discussed in this
section.
80
-------
7.1
TEST PROGRAM
The planned test program will be conducted in the IRF RKS. A process schematic of
the RKS as it will be configured for these tests is shown in Figure 18. However, because very
little flue gas paniculate is expected from the incineration of either component of the ballistic
missile liquid propellant, the baghouse system shown in Figure 18 will be bypassed.
7.1.1 Environmental Regulations
As noted above, the objective of the initial tests are to establish that UDMH and N2O4
can be destroyed in an incineration system in a manner that meets U.S. and Russian
environmental regulations. The applicable U.S. environmental regulations are the hazardous
waste incinerator performance standards established under RCRA. These standards require that
the incinerator achieve:
At least 99.99 percent DRE of the POHCs in,the waste feed to the incinerator
HC1 emissions of less than 1 percent of the HC1 entering the incinerator's APCS
or 1.8 kg/hr, whichever is greater
The promulgated regulations require that particulate emissions be no greater than 180 mg/dscm
(0.08 gr/dscf) corrected to 7 percent O2. However, recent EPA guidance, planned for
incorporation into hazardous waste incinerator operating permits as they are issued or renewed,
states that particulate emissions be limited to 34 mg/dscm (0.015 gr/dscf) corrected to 7 percent
O2.
In addition, hazardous waste incinerator permits currently being enforced in the U.S.
require that CO emissions be no greater than a 1-hour rolling average of 100 ppm, corrected to
7 percent O2, and limit hazardous constituent trace metal feedrates to levels designed to prevent
exceeding risk-based ambient levels. The hazardous constituent trace metals are antimony,
arsenic, barium, beryllium, cadmium, chromium, lead, mercury, silver, and thallium. Finally,
current guidance states that total tetra- through octa-chlorinated PCDDs/PCDFs be limited to
30 ng/dscm corrected to 7 percent O2.
Discarded or off-specification UDMH to be destroyed or disposed of would be the listed
hazardous waste U098. The POHC for this waste for which an incinerator would need to
achieve 99.99 percent DRE would obviously be UDMH. Discarded or off-specification N2O4
would be listed waste P078. P078 is listed as nitrogen dioxide (NO2). However, N2O4 is the
term used to refer to the equilibrium mixture of N2O4 and NO2 expressed as
2NO
(7-1)
Because neither N2O4 nor NO2 is an organic constituent, no DRE requirement would apply.
81
-------
The Russian environmental regulations limit the emissions of UDMH and several
potential UDMH PICs from the incineration of UDMH. These limits are summarized in
Table 26. The limits noted in the table are occupational exposure limits in terms of maximum
permissible concentrations in workplace air. Corresponding ambient air standards for general
population exposure are a factor of 100 lower. Region-specific regulations may further constrain
a sources' duration of operation to ensure maintenance of the ambient standards For example,
in a highly industrialized region with many CO or NOX sources, a UDMH incinerator may be
constrained to operate only a set number of hours per day or days per week to ensure that
ambient CO or NO2 levels from resulting from the collection of sources in the region are not
exceeded. The candidate locations for operating the transportable incinerators to be supplied
to the Russian Federation are sufficiently remote from highly industrialized urban areas that no
additional region-specific constraints will apply.
European hazardous waste incinerator regulations might also be considered in
addressing the environmental certifiability of missile propellant incineration. A summary of
select European incinerator regulations is given in Table 27. Of the European regulations noted
in Table 27, the German regulations are currently the most stringent. Other European countries
either do not specifically regulate incinerator emissions or have less stringent emission limits,
such as the French limits noted in Table 27. Recognizing this, and coupled with the desire to
have a common set of regulations within Europe, the European Community (EC) proposed a
new directive in 1992, which is also given in. Table 27. The EC directive is expected to be
adopted by EC members in 1994.
Comparing Russian and the EC directive limits in Table 27 to the U.S. incinerator
standards noted above shows that the EC particulate emission standard, which equates to
6.5 mg/dscm at 7 percent O2, is significantly more stringent than even the U.S. guidance level
of 34 mg/dscm at 7 percent O2. The EC CO standard equates to 56 ppm daily average at
7 percent O2. This is comparable to the U.S. standard of 100 ppm, 1-hour rolling average at
7 percent O2. The Russian regulations cannot be directly compared to either the EC or U.S.
standards because the Russian regulations are in terms of ambient concentrations. The EC
metal emission limits also cannot be directly compared to the U.S. limits because the U.S. limits
are feedrate based. However, the EC limits for antimony, lead, mercury, and thallium are likely
to be significantly more stringent than the U.S. requirements. In addition, the EC directive
extends to cobalt, manganese, nickel, tin, and vanadium, but does not address barium or silver.
With respect to UDMH and N2O4 incineration, the primary requirements that need to
be demonstrated are that 99.99 percent UDMH DRE can be achieved with acceptable CO and
NOX emissions are possible for both UDMH and N2O4 incineration. The U.S. incinerator
standard of 100 ppm CO, 1-hour rolling average at 7 percent O2, might represent an appropriate
target. This equates to 183 mg/dscm at 7 percent O2. Thus, only a 10-fold dilution of stack
emissions into ambient air would be needed to meet the Russian workplace standard of
20 mg/m3. Typical stack to maximum ambient concentration dilution factors are much larger,
generally 100 to several thousand.
The German standard of 200 mg/Nm3 as NO2 at 11 percent O2 might be considered
an appropriate target. This level equates to an emission limit of 136 ppm NOX at 7 percent O2.
The U.S. hazardous waste incinerator standards do not address NOX emissions. However, new
source performance standards (NSPS) for similar combustion processes established under the
82
-------
TABLE 26. RUSSIAN FEDERATION ENVIRONMENTAL REGULATIONS
FOR UDMH INCINERATION
Compound
Maximum permissible
concentration in workplace air,
mg/m3
UDMH
Dimethylamine
N-Nitrosodimethylamine
Hydrogen cyanide (HCN)
1, l,4,4-Tetramethyl-2-tetrazene
Formaldehyde
CO
NO7
0.1
1.0
0.01
0.3
3.0
0.5
20
2.0
TABLE 27. EUROPEAN HAZARDOUS WASTE INCINERATOR EMISSION LIMITS
Pollutant
Participate
HC1
HF
S02
NOX
CO
Total organic carbon
Germany,
17th BImSch V,
mg/Nm3,
11% O2, dry,
daily average
10
10
1
50
200b
50
10
France
Regulation
mg/Nm3,
7% CO2, wet
150
100
a
_
Recent permit
requirements,
mg/Nm3,
7% CO2, wet
30
50
_
EC directive,
mg/Nm3, 11%
O2, dry,
daily
average
5
5
1
25
50
5
Heavy metals
Cd + Tl
Hg
Others0
Dioxins and furans,
TEQd
0.05 1 ' 1
0.05 [ 5 total [ 5 total
0.5 J j
0.1 ng/Nm3
0.05
0.05
0.5
0.1 ng/Nm3
a = No standard.
bAs NO2.
°Sb, As, Cr, Co, Pb, Mn, Ni, Sn, and V.
dTEQ = 2,3,7,8 tetrachloro dibenzo-p-dioxin (TCDD) toxicity equivalents.
83
-------
Federal Clean Air Act are comparable to the German incinerator standard. For example, the
NSPS for large municipal waste incinerators (greater than 250 tons/day [227 Mg/day] capacity)
is 180 ppm NOX at 7 percent O2; and the NSPS for residual oil fired utility, industrial,
commercial, and institutional boilers with heat input rates greater than 100 MMBtu per hour
(29 MW) is 0.3 Ib NOX (as NO2) per MMBtu heat input (129 ng/J), which equates to about
180 ppm NOX at 7 percent O2. A 100-fold dilution of stack emissions of 200 mg/Nm3 would
satisfy the Russian workplace standard for NO2. This is at the lower bound of typical dilution
factors, as noted above.
Test data will be compared to the NOX and CO emission standards noted above.
Dispersion modeling will also be performed to allow comparing test emission level data to the
Russian ambient workplace standards.
Emissions of particulate, HC1, HF, SO2, heavy metals, and dioxins and furans, regulated
in Germany and proposed for the EC, are expected to be negligible. Total organic carbon
emissions should also be negligible if the CO emission limit is met. Emission rates of several
of these pollutants will be measured in the planned initial tests, however. In addition, emission
rates of UDMH and the UDMH PICs having Russian occupational exposure requirements,
noted in Table 26, will also be measured.
In summary, the specific test program objectives are:
To develop the data to evaluate whether UDMH and N2O4 can be incinerated in
compliance with the U.S. hazardous waste incinerator performance standards and
recent permitting guidance of:
99.99 percent POHC DRE
HC1 emissions less than the greater of 1 percent of the APCS inlet flowrate
of 1.8 kg/hr .
CO emissions of less than 100 ppm hourly rolling average at 7 percent O2
Particulate emissions of less than 34 mg/dscm (0.015 gr/dscf) corrected to
7 percent O2
Total tetra- through octa-PCDD/PCDF emissions of less than 30 mg/dscm
corrected to 7 percent O2
To develop trace metal emission rate from the incineration of UDMH and N2O4
data for comparison to both the EC directive limits and the U.S. incinerator
performance Tier n limits
To develop particulate, HC1, total organic carbon, and PCDD/PCDF emission
rate data from the incineration of UDMH and N2O4 for comparison to the EC
directive limits
84
-------
To develop CO and NOX (NO plus NO2) emission rate data from the incineration
of UDMH and N2O4 for comparison to German incinerator standards and the
emission rate limits corresponding to the Russian occupational exposure limits
To develop UDMH PIC emission rate data from the incineration of UDMH for
comparison to the emission rate limits corresponding to the Russian occupational
exposure limits
7.1.2 Test Conditions
The planned initial test program will consist of nine incineration tests. Three tests will
be performed under the same incineration system operating conditions feeding each component
of the missile propellant. Triplicate testing is a requirement for U.S. hazardous waste incinerator
trial burns. In addition, triplicate testing establishes a measure of confidence in test results by
allowing the precision of the test program sampling and analysis procedures to be evaluated.
Two sets of triplicate tests feeding UDMH (six total) will be required to complete all the flue
gas sampling procedures planned for the UDMH feed tests, as noted in Section 7.1.3. Thus, nine
tests in total, six feeding UDMH and three feeding N2O4, will be performed.
As noted in Section 7.1.1, the initial tests will focus on demonstrating that 99.99 percent
UDMH DRE can be achieved under at least one set of incineration conditions, with acceptable
CO, NOX, particulate, HC1, total organic carbon, trace metal, PCDD/PCDF, and UDMH PIC
emissions. Thus, the incineration conditions selected for the tests are those expected to achieve
the most effective propellant component destruction. The initial tests will, therefore, serve as
a baseline to show that environmentally acceptable incineration can be achieved under at least
one set of incineration conditions, although perhaps not the optimum set of conditions.
The six UDMH destruction tests will be performed at kiln exit gas temperature of 980 °C
(1,800 °F). Only UDMH will be fed to the kiln along with the required combustion air. The
RKS auxiliary fuel, natural gas, will not be fed during actual testing, although natural gas will be
used for incinerator heat up, and to maintain incinerator temperatures overnight between tests.
UDMH will be fed to the kiln via the liquid waste/fuel nozzle of the kiln's dual fuel burner. The
UDMH will be directly pumped and metered from its nitrogen-blanketed storage container to
the burner nozzle.
The three N2O4 destruction tests will also be performed at kiln exit gas temperature of
980°C (1,800°F). Diesel fuel will serve as the material to be oxidized by N2O4 for its
destruction. The diesel fuel will be fed to the kiln via the liquid nozzle of the kiln's dual fuel
burner. The N2O4 oxidant will be added to the burner primary air supply.
This mode of operation will require an N2O4 evaporator between the N2O4 storage
container and the kiln burner, which will deliver N2O4 vapor to the burner at the correct
pressure. The evaporator will be fed from the N2O4 storage container by pressurization of the
storage container nitrogen blanket. The N2O4 supply line to the kiln burner will be heat-traced
to ensure that no N2O4 recondensation occurs. The feed line will tee into the burner primary
air supply line at the kUn burner.
85
-------
For both test feeds (all initial tests) the RKS afterburner will be fired with natural gas
to maintain an afterburner exit gas temperature of 1,090 °C (2,000 °F). The IRF hazardous waste
management permit specifies a minimum .afterburner temperature of 1,017°C (1,863°F)
whenever a hazardous waste is being fed to the system. The 1,090 °C (2,000 °F) set temperature
is a typical afterburner operating temperature and also allows a margin of flexibility above the
permit-mandated temperature.
The feedrates of both UDMH and N2O4/diesel fuel will result in kiln chamber heat
input rates of between 290 and 440 kW (1 and 1.5 MMBtu/hr), typical RKS operation. Actual
test weight feedrates will be established based on preliminary scoping tests to be performed
before the evaluation tests are initiated. These scoping tests wii confirm the ability to incinerate
both propellant components, while maintaining planned incineration conditions, with acceptable
CO, total organic carbon, and NOX emissions as discussed above. In particular, it is expected
that several N2O4/diesel fuel ratios will need to be tried during scoping tests to identify ratios
that give acceptable NOX emission rates.
Current expectations are that the propellant component feedrates summarized in
Table 28 will yield the target test conditions. For the N2O4 tests, the Table 28 entries assume
that a 100 percent excess of diesel fuel will give acceptable NOX emissions.
For all tests, the RKS scrubber system will operate at its nominal design conditions,
listed in Table 10, and at as close to total recirculation (zero to minimum blowdown) as possible.
The feed materials will be supplied by the San Antonio Air Logistics Center's
Directorate of Aerospace Fuels (SA-ALC/SF) at Kelly Air Force Base, Texas. Thirteen 55-gal
(208-L) drums of UDMH containing a total of 1,930 kg (4,240 Ib) will be shipped to the IRF.
Two cylinders of MON-1 grade N2O4, each containing 910 kg (2,000 Ib) (total of 1,820 kg
[4,000 lbj)will be shipped.
7.13 Sampling and Analysis Procedures
The RKS sampling locations and the scope of. the sampling effort are shown in the
process schematic given in Figure 21. For all tests, the sampling matrix defined to meet the test
program objectives listed in Section 7.1.1 will entail:
Continuously measuring O2, CO, NOX, and TUHC concentrations in the kiln exit
flue gas; O2, CO2, and NOX concentrations in the afterburner exit flue gas; O2,
CO2, and NOX concentrations in the scrubber exit flue gas; and O2 and CO
concentrations in the stack gas
Sampling flue gas at the scrubber exit and the stack for particulate and HC1 using
Method 5; the stack gas sample is needed to comply with the IRF's permit
requirements
Additional sampling procedures will be performed for the UDMH incineration tests.
These are:
Sampling flue gas at the kiln exit, afterburner exit, and scrubber exit for:
86
-------
TABLE 28. EXPECTED APPROXIMATE TEST CONDITIONS
Parameter
Kiln
Propellant component feedrate, kg/hr (Ib/hr)
Diesel fuel feedrate, kg/hr (Ib/hr)
Combustion air feedratea, scm/hr (scfh)
Exit gas:
Temperature, °C (°F)
Moisture, %
O2, % dry
Flowrate, dscm/hr (dscfh)
Afterburner
Natural gas feedrate, scm/hr (scfh)
Combustion air feedratea, scm/hr (scfh)
Exit gas:
Temperature, °C (°F)
Moisture, %
O2, % dry
Flowrate, dscm/hr (dscfh)
UDMH tests
45 (100)
740 (26,000)
980 (1,800)
9.2
11.4
710 (25,000)
31 (1,090)
350 (12,500)
1,090 (2,000)
11.5
8.6
1,030 (36,300)
N2O4 tests
91 (200)
36 (80)
800 (28,400)
980 (1,800)
6.9
13.8
830 (29,400)
36 (1,280)
380 (13,400)
1,090 (2,000)
10.3
9.7
1,220 (43,200)
Includes routinely experienced inleakage.
UDMH and dimethylamine using a variation of the National Institute for
Occupational Safety and Health (NIOSH) Method S143
N-nitrosodimethylamine and l,l,4,4-tetramethyl-2-tetrazene
(tetramethyltetrazene) using Method 0010
HCN using the California Air Resources Board (CARB) Method 426
Formaldehyde using Method 0011
Sampling flue gas at the scrubber exit for PCDDs/PCDFs for one test using
Method 23
87
-------
16-2.6Z QS3
W)
ec
d>
E
a
o
i
|
1
u 5 W
V
S
o" g
I'll
9 S *6
o 2 S H
Ilfl
1 S3 .
f S ^
2^1
M *v T3
|is
i!§
5 *£ "§
*^ ea «
2 0,
||
Is
til
«« a<
S -
|ll
O rj
w i*!t
^ O
S p
t
f
U
8
o"
hi
11
u v««
g
a
1
*c
I
SB P
S S S
P P P
3 x
z"
P
z"
z
X
XX X
X X
X X
XX XX
X
i)
S bfl
a S s I
1 5 f 1
- E ^ s s
e o "§ "§ "t
.-s a g g c
^ <^ C/3 C/5 V
rJ f^
-------
Additional sampling procedures will also be performed for the N2O4 tests. These
are:
Sampling flue gas at the scrubber exit for trace metals using the EPA multiple
metals train
Sampling flue gas at the scrubber exit for PCDDs/PCDFs using Method 23
Measurements of NOX, UDMH, and UDMH PICs are specified at the three locations
noted, specifically to supply data to allow evaluating the need for a secondary combustion
chamber (afterburner) and/or a wet scrubber APCS. Additional sampling and analysis is also
planned for all tests to allow a wider scope of environmental acceptability to be evaluated. This
additional sampling will measure particulate, HC1, trace metal, and PCDD/PCDF emission rates.
This additional sampling will be performed in the scrubber exit gas. Pre- and post-test scrubber
liquor samples will also be taken and analyzed for UDMH (for UDMH incineration tests),
chloride, nitrate, nitrite, and trace metals.
The number of sampling procedures specified for the UDMH tests cannot be performed
simultaneously at the IRF due to the unavailability of sampling ports in all the locations
specified. Thus, the UDMH sampling matrix will be completed over two sets of tests. The
procedures denoted Ul in Figure 21 will be simultaneously completed over one set of three test
days; the procedures denoted U2 in the figure will be completed during a second set of three test
days.
The analysis procedures to be applied to test program samples are summarized in
Table 29.
7.2 CURRENT STATUS
The planning for this test program was initiated in June 1993 during a visit to the IRF
by two DNA representatives. Following the discussions at this meeting, a test plan outline was
prepared and distributed in July. With DNA approval of the test plan outline in September, a
complete test plan was prepared for distribution in October.
Current plans are to complete all test planning documentation in early FY94 and initiate
the testing in December 1993. Sampling tests will follow in January and February, 1994. All
FY94 efforts will be supported by DNA under interagency cost reimbursement order (IACRO)
No. 93-691, which will become EPA interagency agreement (TAG) No. RW97936997.
89
-------
TABLE 29. ANALYSIS PROTOCOL
Sample
Analyte
Analysis method
Frequency
NIOSH Method
S143 combined
impingcr solutions
UDMH
Dimethylamine
Phosphomolybdic acid addition,
colorimetric analysis by NIOSH
Method S143"
Neutralization then GC/FID by
NIOSH Method 2010"
3 per test (kiln,
afterburner, and
scrubber exit), UDMH
tests only
3 per test (kiln,
afterburner, and
scrubber exit), UDMH
tests only
Method 0010 train N-nitrosodimethylamine,
tetramethyltetrazene
Extraction by Method 3540A,
GC/MS analysis by Method 8270Ab
3 per test (kiln,
afterburner, and
scrubber exit), UDMH
tests only
CARB Method 426 Cyanide
train
Distillation and colorimetric analysis
by CARB Method 426°
3 per test (kiln,
afterburner, and
scrubber exit), UDMH
tests only
Method 0011 train Formaldehyde
Extraction with methylene chloride
analysis, by HPLC with UV/vis
detection by Method 8315b
3 per test (kiln,
afterburner, and
scrubber exit), UDMH
tests only
Method 23 train PCDD/PCDF
Extraction and analysis by
Method 8290b
1 per test for N2O4 and
set 2 of UDMH tests
Multiple metals train Trace metals"
Digestion by multiple metals
procedure', ICAP analysis by Method
6010Ab
Gravimetry by Method 5f
1 per test for N2O4 and
set 2 of UDMH tests
Method 5 train Particulate
cr
IC by Method 9057b
1 per test at stack,
1 per test at scrubber
exit for N2O4 and set 2
of UDMH tests
1 per test at stack,
1 per test at scrubber
exit for N2O4 and set 2
of UDMH tests
Pretest scrubber
liquor
Trace metalsd
CT, NO3', NO2"
Digestion by Method 3010A, ICAP 1 per test
analysis by Method 6010Ab
1C by Method 300.0s 1 per test
Post-test scrubber
liquor
Trace met
or, N03-, N02-
UDMH
Digestion by Method 3010A, ICAP
analysis by Method 6010Ab
1C by Method 300.0s
Phosphomolybdic acid addition,
colorimetric analysis by NIOSH
Method S1438
1 per test
1 per test
1 per test for UDMH
tests
Reference 9, NIOSH Methods.
Reference 5, SW-846.
Reference 10, CARB Methods.
dSee Table 30.
Reference 6, 40 CFR 266, App. IX.
Reference 7, 40 CFR 60, App. A.
^Reference 11, Water and Wastes.
90
-------
TABLE 30. TRACE METALS TO BE
DETERMINED IN SELECTED
TEST PROGRAM SAMPLES
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Nickel
Silver
Thallium
Tin
Vanadium
91
-------
SECTION 8
EVALUATING THE EFFECTIVENESS OF ADDITIVES
AS SORBENTS FOR METAL CAPTURE
There is currently considerable interest in the potential use of mineral-based sorbents
for capturing and retaining hazardous constituent trace metals in incineration processes. A
number of fundamental, bench-scale research programs are currently underway at several
universities and research laboratories. These studies are investigating the application of sorbents
both in the combustion flue gas and in the solids bed.
Most of the research completed to date has focused on quantifying the effectiveness of
various proposed sorbents for capturing vaporized metals from the flue gas. In such applications,
it is theorized that vaporized metals will react with the sorbent particles at the elevated
incinerator temperatures or heterogeneously condense onto the sorbents as the flue gas cools.
In the absence of available condensation sites, vaporized metals will primarily undergo
homogeneous condensation, forming a fine fume. Thus, the goal of this approach is to make
particles available in the flue gas with which the metals can react or upon which they can
condense. Metals bound to larger sorbent particles will be more effectively collected by APCSs
than metals presented as a fine fume. In addition, studies completed to date suggest that
chemical reaction between the metal and the sorbent dominates over physical adsorption,
offering the additional advantage of reduced potential for metal leaching from collected
particulate.
Other researchers have studied the incorporation of sorbents into the solid feed. This
approach seeks to capture and bind the metals in the incinerator ash, thereby preventing them
from exiting with the combustion gases. For this approach to be effective, research completed
to date suggests that the metal should become volatile in the incinerator environment and
chemically react with the sorbent material.
The test program described in this section is designed to further investigate this second
approach by screening several minerals for their suitability as sorbent materials for capturing
metals in the solid bed and preventing their release to the flue gas. In addition to capturing the
metals, an ideal sorbent would retain them in the ash when disposed, so that a TCLP leachate
of the ash would contain metals concentrations below respective regulatory levels. Thus, the
objective of this screening test program planned for completion at the IRF is to evaluate several
candidate sorbents with respect to:
The degree to which they facilitate retention of trace metals in the solid bed that
would be the bottom ash discharge from a rotary kiln incinerator
92
-------
The degree to which they retain trace metals in the solid bed when subjected to
TCLP extraction
If screening test results show that one or more of the candidate sorbents test
significantly alters test metal behavior in terms of fraction of metal feed retained in the solids
bed, or decreased fraction of solids bed metal leachable in the TCLP test, then a confirmation
test program in the IRF RKS may be planned.
The screening tests in this program will be conducted in the IRF TTU. A brief
description of this unit, and the outline of the planned test program, are discussed in the
following subsections.
8.1
TEST FACILITY DESCRIPTION
The TTU at the IRF consists of a small commercial pathological incinerator that has
been modified to allow for continuous test material feed and treated-material (e.g., ash) removal,
for variable and controlled thermal treatment temperatures, and for expanded process operation
monitoring. The TTU is illustrated in Figure 22.
The combustor portion of the TTU consists of three chambers: the charge chamber, the
retention chamber and the breeching chamber. The charge chamber is designed to accept the
TTU's solid material feed stream. It corresponds to the primary combustion chamber, or kiln
portion, of a waste incinerator. Its inner cross section is 0.66 m (2 ft 2 in) square, its height
1.9 m (6 ft 2 in), and its chamber volume 0.82 m3 (29 ft3). The retention chamber, which directly
follows the charge chamber, is designed to effect complete organic constituent destruction. It
corresponds to the secondary combustion chamber, or afterburner portion, of a waste incinerator.
Its inner cross section is also 0.66 m (2 ft 2 in) square, its height 1.5 m (5 ft), and its chamber
volume 0.67 m3 (23.5 ft3). The breeching chamber serves as a second-stage afterburner. Its
inner diameter is 0.4,1 m (1 ft 4 in), its total height 0.76 m (2 ft 6 in), and its chamber volume
0.10 m3 (3.5 ft3). All chambers are lined with a 13-cm (5-in) thickness of refractory.
As received from the incinerator vendor, all three chambers were designed to be fired
with natural-gas-fueled burners. The burners installed in the charge and retention chambers are
natural-gas-fired, with 350 kW (1.2 million Btu/hr) capacities and 5-to-l turndowns. Modulating
burner controls allow variable firing rates to control temperatures in each chamber at preset
levels between 260° and 1090 °C (500° and 2,000 °F) with variable air-to-fuel ratio. The
breeching chamber has a manually adjustable 220 kW (750,000 Btu/hr) burner.
Test material is fed to the charge chamber via a feed system that transports quartz trays
containing the test material through the chamber via a variable-speed chain-drive mechanism.
Each quartz tray is 23 cm (9 in) long by 13 cm (5 in) wide by 5 cm (2 in) deep, and holds up to
2.3 kg (5 Ib) of test material. The variable-speed chain drive allows trays containing test material
to have charge chamber residence times of between 20 minutes and 1 hour. Multiple trays can
be fed in sequence to simulate continuous feed to a thermal treatment system.
Combustion gas temperatures are recorded using type K or R thermocouples at the
following locations in the system:
93
-------
TEMP INDICATING
CONTROLLER
TEMP INDICATING
CONTROLLER
FEED
CONVEYOR
STACK
. SAMPLING
~zJ PORTS
BREECHING
CHAMBER
RETENTION
CHAMBER
CHARGE
CHAMBER
- , _. 1
^^ ^^
TEMP INDICATOR
TEI-2
3
Q
(fl
111
&
BURNER 3
BURNER 2
CHART
RECORDER
BURNER 1
ASH
TEMP INDICATOR t t .
jir_riririri^ririririri_"L \
Figure 22. The ERF TTU.
94
-------
Inside feed door
Inside discharge door
Bottom of charge chamber center
Charge chamber exit gas
Retention chamber exit gas
Breeching chamber exit gas
Stack gas
82 TEST PROGRAM
The test program to be performed will consist of 50 tests. The test variables will be
sorbent material, solid bed temperature, feed chlorine content, and metal form in the feed. Six
sorbents will be evaluated in this study. Four of the six, diatomaceous earth, kaolinite, bauxite,
and alumina, were selected based on the most promising results from other researchers. The
attapulgite clay used in past IRF trace metal studies will be tested as the fifth sorbent to serve
as a link to the past IRF studies. The sixth material, quartz, was selected to be the neutral
material to serve as a system blank or background material The trays used to feed the test
materials to the TTU are quartz.
The approximate mineral content of diatomaceous earth, kaolinite, and bauxite is given
in Table 31. Quartz is presumed to be pure silica (SiO2); alumina is presumed to be pure A12O3.
This attapulgite clay is a hydrated magnesium aluminum silicate [(Mg,Al)5Si8O22(OH)4-4H2O]
containing some dolomite [Ca,Mg (CC^, calcite [CaCO3], and silica.
Three solids bed temperatures will be tested: 540°, 700°, and 870°C (1,000°, 1,300°, and
1,600 °F). Two feed chlorine contents will be tested, 0 and 4 percent by weight. Polyvinyl
chloride (PVC) powder will be added to chlorine-containing feed mixtures to provide the desired
chlorine content.
Sorbent behavior in retaining five trace metals will be evaluated in the test program.
The five metals are arsenic, cadmium, chromium, lead, and nickel. Two forms of incorporating
TABLE 31. APPROXIMATE SORBENT MINERAL COMPOSITION
Diatomaceous
earth Kaolinite Bauxite
SiO2
A1203
FejOa
TiO2
CaO
MgO
Other oxides
90.4
6.5
2.3
0.2
0.3
0.3
52
45
1
2
_~
-
11
84
5
95
-------
the metals into metals/sorbent mixtures for testing will be investigated. Past trace metal-tests
at the IRF1'2 have used aqueous metal spike solutions containing soluble nitrate salts of the
metals, with the exception of As which has been added as As2O3. This form will be one of the
two used in these tests. The second form of metal spiking will use metal compound dispersions
or suspensions. Such dispersions consist of metal compound powders suspended in a liquid
carrier analogous to pigments dispersed in paint or ink.
Table 32 summarizes the matrix of test variable combinations planned for testing. The
table notes 48 combinations to be tested. Test condition 36 in Table 32 will be repeated twice
(performed in triplicate) over the course of the test program, giving a total of 50 tests. For each
test, weighed amounts of appropriate mixtures of sorbent, PVC (for tests with chlorine-
containing feed), and metal spiking preparation will be charged to one of the quartz trays used
to introduce feed to the TTU. Charge depth will be nominally 2 cm.
The composition of the metal spiking solution will be as given in Table 33. The
appropriate quantity of the spiking solution will be combined with the solid sorbent and PVC (if
added) to result in solid charge metal concentrations approximating those rioted in the rightmost
column of Table 33 for sorbents that dp not contain the spiked metals.
The metal dispersion to be used as the second form of metal spiking will consist of finely
ground (particle size typically between 0.1 and 5 /tm) metal compounds in a matrix of fuel oil
and vegetable oil. The metal compounds to be used for these tests will be metal oxides. Stable
dispersions can be made with the compounds chosen.12 Table 34 gives the metal and metal
oxide concentrations in the dispersion to be prepared for these tests. The appropriate quantity
of metal dispersion will be combined with the solid charge to result in a mixture with total metal
concentrations approximating those noted in the rightmost column of Table 34, again for
sorbents with negligible test metal concentrations.
For each of the planned tests, the TTU will be allowed to reach steady state at the
desired temperature condition before feeding the test tray for each test. The feed tray will be
fitted with thermocouples so that the bed temperatures can be monitored. The feed tray will be
fed to the TTU to start the test, then removed when the feed material has reached the target
temperature.
Four sample matrices will be collected or prepared for the analytical measurements:
unspiked sorbent, TTU feed, TTU residual discharge, and TCLP leachates of,TTU residual
discharges. These samples will be analyzed for the test trace metals.
83 CURRENT STATUS
All planning efforts for this test program were completed during FY93. After a
substantial literature review and extended telephone discussions with several researchers in the
field, a test plan outline describing the recommended test program was prepared and distributed
in May. The complete test program QAPP, incorporating test program modifications in response
to technical review of .the test plan outline, was prepared and distributed in September. Current
plans are to initiate testing in November, in parallel with the M. W. Manufacturing Superfund
site treatability tests discussed in Section 5.
96
-------
TABLE 32. TEST MATRIX
Test Sorbent Metal form
1 Quartz beads (silica) Aqueous nitrates
2
3
4
5
6
7 Metal dispersions
8
9
10
11
12
13 Attapugite clay Aqueous nitrates
14
15
16
17
18
19 Metal dispersions
20
21
22
23
24
Feed chlorine,
%
0
0
0
4
4
4
0
0
0
4
4
4
0
0
0
4
4
4
0
0
0
4
4
4
Solids bed temperature,
°C
540
700
870
540
700
870
540
700
870
540
700
870
540
700
870
540
700
870
540
700
870
540
700
870
(continued)
97
-------
TABLE 32. (continued)
Test Sorbent
25 Diatomaceous earth
26
27
28
29
30
31 Kaolinite
32
33
34
35
36
37 Alumina
38
39
40
41
42
43 Bauxite
44
45
46
47
48
Feed chlorine,
Metal form %
Aqueous nitrates 0
0
0
4
4
4
Aqueous nitrates 0
0
0
4
4
4
Aqueous nitrates 0
0
0
4
4
4
Aqueous nitrates 0
0
0
4
4
4
Solids bed temperature,
°C
540
700
870
540
700
870
540
700
870
540
700
870
540
700
870
540
700
870
540
700
870
540
700
870
98
-------
TABLE 33. METAL SPIKE SOLUTION CONCENTRATIONS
Metal
Arsenic
Cadmium
Chromium
Lead
Nickel
Metal
concentration,
g/L
7.75
1.55
4.65
7.75
4.65
Spike solution
Compound
As2O3
Cd(N03)2-4H20
Cr(N03)3-9H20
Pb(N03)2
Ni(NO3)2-6H2O
Compound
concentration3,
g/L
10.2
4.25
35.8
12.4
23.0
Approximate
resulting feed
metal concentration,
mg/kg
250
50
150
250
150
Sufficient HNO3 will be added to maintain lead arsenate compounds in solution.
TABLE 34. TEST PROGRAM METAL DISPERSION CONCENTRATIONS
Metal
Arsenic
Cadmium
Chromium
Lead
Nickel
Total
Metal
concentration,
g/kg
7.73
1.49
9.31
7.80
4.45
Dispersion
Compound
As2S3
CdO
Cr2O3
PbO
NiCO3
Compound
concentration,
g/fcg
12.7
1.7
13.6
8.4
9.0
45.4
Approximate
resulting feed metal
concentration,
mg/kg
250
50
300
250
150
99
-------
SECTION 9
INSTALLATION OF A FABRIC FILTER AIR POLLUTION
CONTROL SYSTEM WITHIN THE IRF RKS
The RKS primary APCS, used in most testing efforts completed over the past 6 years,
represents an APCS typical of that employed in industrial hazardous waste incineration systems
built in the 1970s and 1980s. However, more recent installations have incorporated more state-
of-the-art APCSs employing a fabric filter for particulate removal. Thus, to upgrade the IRF-
RKS capabilities to allow testing with a state-of-the-art APCS, it was decided to install a fabric
filter system during FY93.
Conceptual design efforts leading to a system specification were initiated in April. As
acid gas (e.g., HC1) control would continue to be a critical requirement of the RKS APCS, it was
decided that the baghouse system be located downstream of the existing wet scrubber. Thus,
because flue gas exiting the existing wet scrubber is water-saturated, flue gas must be reheated
to comfortably above the dew point before entering the baghouse. Although several approaches
can be employed to effect flue gas reheat, the availability of a 100-kW airheater resulted in this
approach being selected.
Figure 23 is a schematic of the system designed. The location of the baghouse system
within the RKS process flow schematic is illustrated in Figure 18. As shown in Figure 23,
baghouse system inlet flue gas can be taken from the induced draft (ID) fan at the discharge
from the packed column. Flue gas at this location is saturated with water vapor at typically 65°
to 75°C (150° to 170°F). Typical flue gas flowrates are 20 to 35 dscm/min (800 to 1,200 dscfm).
The water-cooled heat exchanger shown in Figure 23 cools the flue gas by about 5°C (10 °F) to
lower its dew point slightly. The 100-kW reheater then heats the flue gas to nominally 120 °C
(250 °F), well above its dew point, so that no water condensation can occur in the baghouse.
After the baghouse, flue gas is routed through the carbon bed/HEPA filter secondary APCS as
required by the IRF's hazardous waste management permit when hazardous waste is being
burned. Either the baghouse system or the secondary APCS, or both, can be bypassed, as shown
in Figure 23. Table 35 summarizes the design specifications of the baghouse collector.
The system design specified placing the baghouse system physically between the
incinerator and the horizontal length of ductwork between the scrubber system and the secondary
APCS. The location formerly housed the single stage ionizing wet scrubber system occasionally
used in RKS tests. The ionizing wet scrubber system was removed and stored to make room for
the baghouse.
Most of the facility modification work required to install the baghouse system was
performed in July and August. The baghouse was delivered to the IRF by the vendor in early
100
-------
Z6-ZZZ 03V
1
s
VI
»a
tj
a>
VI
§
a
03
M
101
-------
TABLE 35. DESIGN CHARACTERISTICS OF THE RKS BAGHOUSE COLLECTOR
System capacity, inlet gas flow 70 m3/min (2,500 acfm) at 120 °C (250 °F)
Operating temperature, max 200 °C (400 °F)
±12.4kPa(±50inWC)
1.8 m (6 ft)
4.2 m (13 ft, 8.375 in)
Operating pressure
Diameter
Overall height
Filter elements (bags)
Material
Length
Number
Total filter area
Materials of construction
Collector internals
Airlock
Venturi nozzles
Insulation
16 oz. Nomex
1.8 m (6 ft)
69
45 m2 (488 ft2)
304 SS
316 SS
Aluminum
Heat loss less than 8.8 kW
(30,000 Btu/hr) at 200°C (400°F)
September. Because much of the preparatory work had been completed during the previous
2 months, actual baghouse installation and final connection proceeded rapidly such that system
startup and shakedown was possible at the close of FY93.
102
-------
SECTION 10
EXTERNAL COMMUNICATIONS
During FY93,7 research reports were prepared or finalized and 6 technical papers were
presented or published. These are listed in Table 36. This level of external communication and
technology transfer testifies to the high level of research being supported at the IRF.
Table 37 lists some of the visitors to the IRF during FY93. The list attests to the
visibility of the work being performed at the IRF to the incineration research community.
103
-------
TABLE 36. IRF PROGRAM REPORTS AND PRESENTATIONS IN FY93
Reports
Waterland, L. R., "Operations and Research at the U.S. EPA Incineration Research Facility,
Annual Report for FY92," draft October 1992, revised December 1992, published as
EPA/6QO/R-93/087, June 1993
Whitworth, W. E., Jr., and L. R. Waterland, "Evaluation of the Impact of Incinerator Waste Feed
Cutoffs," draft November 1992
ICing, C., and L. R. Waterland, "Pilot-Scale Incineration of Contaminated Sludges from the Bofors-
Nobel Superfund Site," revised October 1992, published as EPA/600/R-92/240, March 1993
Siag, A., D. J. Fournier, Jr., and L. R Waterland, "Pilot-Scale Incineration of Contaminated Soil
from the Chemical Insecticide Corporation Superfund Site," revised September 1992, published as
EPA/600/R-93/036, April 1993
King, C., J. W. Lee, and L. R. Waterland, "Pilot-Scale Incineration of Contaminated Soils from the
Drake Chemical Superfund Site," published as EPA/600/R-93/047, May 1993
Fournier, D. J., Jr., and L. R. Waterland, "The Fate of Trace Metals in a Rotary Kiln Incinerator
with a Calvert Flux-Force/Condensation Scrubber System," draft January 1993
Siag, A., and L. R. Waterland, "Pilot-Scale Incineration of Contaminated Soils from the Scientific
Chemical Processing Superfund Site," revised May 1993
Papers and Presentations
Waterland, L. R., and W. E. Whitworth, Jr., "Evaluation of the Impacts of Incinerator Waste Feed
Cutoffs," presented at the Air & Waste Management Association International Conference on
Waste Combustion in Boilers and Industrial Furnaces, Clearwater, Florida, March 1993; published
in Waste Combustion in Boilers and Industrial Furnaces. Proceedings of an International
Conference. A&WMA SP-86, Air & Waste Management Association, Pittsburgh, Pennsylvania,
1993
Richards, M. K., W. E. Whitworth, Jr., and L. R. Waterland, "Hazardous Waste Incinerator
Emissions Resulting from Waste Feed Cutoffs," presented at the Nineteenth Annual RREL
Hazardous Waste Research Symposium, Cincinnati, Ohio, April 1993
Waterland, L. R., M. K. Richards, and H. O. Wall, "Comparison of Thermal Treatment PCB Data
from Two Superfund Sites," presented at the Nineteenth Annual RREL Hazardous Waste Research
Symposium, Cincinnati, Ohio, April 1993
Lee, J. W., S. Venkatesh, and R. C. Thurnau, "The U.S. EPA Incineration Research Facility,"
presented at the Nineteenth Annual RREL Hazardous Waste Research Symposium, Cincinnati,
Ohio, April 1993
Waterland, L. R., and D. J Fournier, Jr., "Potential Surrogate Metals for Incinerator Trial Burns,"
presented at the 1993 Incineration Conference, Knoxville, Tennessee, May 1993; published in
Proceedings of the 1993 Incineration Conference. University of California, Irvine, California, 1993
Lee, J. W., L. R. Waterland, W. E. Whitworth, Jr., and M. K. Richards, "Evaluation of the
Emissions Impact of Repeated Incinerator Waste Feed Cutoffs," presented at the Third
International Congress on Toxic Combustion Byproducts, Cambridge, Massachusetts, June 1993
104
-------
TABLE 37. VISITORS TO THE IRF
Person
R. Wade
S. Makhijani
A. Senthil
K. Partymiller
R. Argus
Z. Plavnik
D. Brooks
M. Richards
W. Schofield
G. Patterson
A. George
B. Blackburn
S. Schulberg
M. Richards
J. Chancy
J. Truholt
A. Doribati
E. Rothschild
G. Stryker
M. Subbaruman
K. Knudsen
T. Holt
D. Burns
H. Burns
D. Burns
H. Pope
M. McDerment
D. Burns
M. Looper
C. McVay
R. Thurnau
Affiliation
USAGE, Waterways
Experiment Station
CPCB, New Dehli, India
PRC
PRC
Sonotech
Industrial Gas Technology
EPA/RREL
DRE Southwest
Marsten Bentley
Marsten Bentley
S-Cubed
S-Cubed
EPA/RREL
USAGE, PBA
ADPCE
ADPCE
Geraghty & Miller
Four Nines
Rockwell, ETEC
Rockwell, ETEC
Rockwell, ETEC
WSRC
WSRC
WSRC
DOE/SRP
DOE/SRP
WSRC
WSRC
WSRC
EPA/RREL
Date
10/1/92
11/2/92
11/17,18/92
11/19/92
12/7-10/92
1/8/93
2/18/93
3/2/93
3/15/93
3/17/93
3/18/93
3/24/93
3/30/93
Purpose of visit
To witness New York Harbor
TTU testing
Facility tour
Preliminary Sonotech SITE test
program planning
To discuss possible metal
dispersion testing
QA technical systems review
Facility tour
Hazardous waste inspection
Facility tour
Facility tour
Preliminary molten salt
oxidation process test planning
To witness initial simulated
mixed-waste testing
To witness simulated mixed-
waste testing
To discuss additional WSRC
test needs and initiate
DOE/EPA interagency
agreement planning
(continued)
105
-------
TABLE 37. (continued)
Person
C. Cooley
G. Knight
W. Quapp
L. Apel
A. Wollerman
J. Petersen
R. Thurnau
D. Burns
P. Lowe
A. Baker
W. Fortner
G. Thomasson
L. Seaton
M. Garvey
K. Shuster
D.Bell
D. Wallace
D. Burns
R. Schultz
H. Thompson
J. Anthony
R. Thurnau
O. Vincent
R. Darner
G. Carroll
G. Lewis
D. Donton
E. A. Burns
M. Richards
G. Carroll
P. Lin
J. Fuhrmann
A. Doribati
M. Adams
G. Carroll
Affiliation
DOE/EM-50
WPI
INEL
INEL
SAIC
CaUidus
EPA/RREL
WSRC
WSRC
EPA
USACE/PBA
USACE/PBA
EPA/Region VI
EPA/OE
EPA/OSW
DOD
DOD
WSRC
DNA
DNA
SAIC
EPA/RREL
EPA/OSORD
EPA/RREL
EPA/RREL
Sparkman HS
SRA
EPA/RREL
EPA/RREL
EPA/RREL
EPA/RREL
ADPCE
ADPCE
EPA/RREL
Date
4/1/93
4/19,20/93
4/19-21/93
5/27/93
6/8/93
6/10/93
6/27,28/93
9/7/93
9/14/93
9/27-29/93
9/30/93
Purpose of visit
To discuss DOE mixed-waste
thermal destruction
research/technology
development needs
DOE QA surveillance for
WSRC tests
EPA properly audit
Facility tour
To witness WSRC tests
Preliminary ballistic missile
propellant test planning
Health and safety technical
assistance visit
Facility tour
Facility tour, discussion of
SBIR test program
Project review and planning
Permit renewal, RFI
discussions
106
-------
SECTION 11
PLANNED EFFORTS FOR FY94
In addition to the test programs planned for completion in FY94, as discussed in
Sections 5 through 8, several other candidate programs are being considered for inclusion in the
IRF FY94 schedule beginning in May 1994. The current most likely candidate is a program to
evaluate the treatability, via low-temperature thermal desorption, of three contaminated
materials from the Rocky Mountain Arsenal Superfund site. Initial test planning discussions
have begun with Argonne National Laboratory staff. Argonne is supporting the Department of
the Army in its efforts to remediate the site.
Other candidate test programs for FY94 include:
Testing to support the DOE's mixed-waste thermal destruction technology
development program
Evaluation of other innovative incineration systems
Extended liquid ballistic missile ppopellant testing
Additional trace metal fate testing to support the installation and permitting of the
GIF at DOE's Savannah River Plant
Testing to evaluate whether the major constituents comprising the TUHC
emissions from an incinerator can be identified
Detailed plans for testing during the second half of FY94 will be firmed up as the year
progresses.
107
-------
REFERENCES
1. Fournier, Jr., D. J., W. E. Whitworth, Jr., J. W. Lee, and L. R. Waterland, 'The Fate of
Trace Metals in a Rotary Kiln Incinerator with a Venturi/Packed Column Scrubber,"
EPA/600/2-90/043, February 1991.
2. Fournier, Jr., D. J., and L. R. Waterland, "The Fate of Trace Metals in a Rotary Kiln
Incinerator with a Single-Stage Ionizing Wet Scrubber," EPA/600/2-91/032, September 1991.
3. Barton, R. G., W. D. Clark, and W. R. Seeker, "Fate of Metals in Combustion Systems,"
Combustion Science and Technology. Vol. 74, pp. 327-342, 1990.
4. "Guidance on Setting Permit Conditions and Reporting Trial Burn Results, Volume II of
the Hazardous Waste Incineration Guidance Series," EPA/625/6-89/019, January 1989.
5. 'Test Methods for Evaluating Solid Waste: Physical/Chemical Methods," EPA SW-846,
Third edition, Revision 1, July 1992.
6. 40 CFR Part 266, Appendix IX.
7. 40 CFR Part 60, Appendix A.
8. 40 CFR Part 61, Appendix B.
9. Eller, P. M., ed., "NIOSH Manual of Analytical Methods," Third edition, February 1984, with
supplements 1, 2, 3, and 4 (1985, 1987, 1989, 1990).
10. "Stationary Source Test Methods, Volume III, Methods for Determining Emissions of Toxic
Air Contaminants from Stationary Sources," State of California, Air Resources Board,
Sacramento, California, March 1988.
11. "Methods for Chemical Analysis of Water and Wastes," EPA-600/4-84-017, March 1984.
12. W. Schofield, L., Weitzman, and A. George, "Metal Dispersion Spiking Systems for RCRA
Trial Burns," in Waste Combustion in Boilers and Industrial Furnaces. Proceedings of an
International Specialty Conference. Air & Waste Management Association, Pittsburgh,
Pennsylvania, March 1993.
108
tlU.S. GOVERNMENT PRINTING OFFICE: 1991 - 550-001/80409
------- |