United States National Risk Management
Environmental Protection Research Laboratory
Agency Cincinnati, OH 45268
Research and Development EPA/600/SR-96/105 March 1997
oERA Project Summary
Evaluation of Rotary Kiln
Incinerator Operation at Low-to-
Moderate Temperature
Conditions
J, Lee, D, Fournier, Jr., C. King, S. Venkatesh, and C. Goldman
A 12-test program was performed at
the Environmental Protection Agency's
Incineration Research Facility to study
the effectiveness of incineration at low-
to-moderate temperatures in decontami-
nating soils containing organic com-
pounds with different volatilities (boil-
ing points). Test parameters were soil
moisture content, treatment temperature,
treatment time, soil bed depth, and de-
gree of soil agitation. A related objec-
tive was to determine the fate of con-
taminant metals in the contaminated soil
under these conditions.
The data demonstrate that compound
volatility and treatment temperature are
the key parameters that will affect
whether a contaminated soil can be
successfully decontaminated. Low-boil-
ing (volatile) compounds can be rap-
idly (in less than 20 minutes) driven
out of the soil nearly quantitatively to
non-detectable levels at a 316°C (600°F)
kiln exit gas temperature. High-boiling
compounds require higher treatment
temperatures (greater than 482°C
[900°F] kiln exit gas temperature) and
longer treatment times (longer than
30 minutes) in order for more than
99.9% of the compounds to be driven
out.
Increased soil temperature favors
decontamination and is essential for
satisfactory decontamination of high-
boiling compounds. However, while soil
temperature is important, other param-
eters, such as the presence of mois-
ture or the degree of agitation, can af-
fect the decontamination process and
can be beneficial under the right com-
bination of conditions.
The effects of moisture on decon-
tamination effectiveness are manyfold.
Increased moisture reduces soil heat-
up rates and thus tends to slow down
the decontamination process. Test data
suggest that increased moisture, for
some materials, may increase soil agi-
tation at moderate-to-high kiln rotation
speeds. This, in turn, can lead to faster
heat absorption and reduced mass
transfer resistance, but shorter soil resi-
dence time. Faster heat absorption and
reduced mass transfer resistance in-
creases decontamination rate. Shorter
residence time lowers the extent of or-
ganic constituent decontamination, if
the decontamination rate for that con-
stituent is slow. In addition, increased
moisture may enhance decontamina-
tion of the less-volatile organic com-
pounds through steam stripping, pro-
vided that the additional moisture does
not prevent the soil from reaching nec-
essary temperatures.
With the exception of mercury, the
extent of metal volatilization from the
treated soil was not significantly af-
fected by any of the test variables. Mer-
cury was volatile, tending to be equally
distributed between the kiln ash and
the scrubber exit flue gas. At the high-
est kiln exit gas temperature of 649°C
(1,200°F), the extent of mercury volatil-
ization increased with treatment time.
The effects of thermal treatment on
metals leachability in the toxicity char-
acteristic leaching procedure vary from
metal to metal. Among the test variables,
the most influential is treatment tem-
perature. Lead and barium teachabilities
-------
were not affected by any of the test
variables. Leachable fractions of ar-
senic and cadmium decreased when
soil temperature increased. In contrast,
leachable fractions of chromium and
mercury increased when soil tempera-
ture increased.
This Project Summary was developed
by EPA's National Risk Management
Research Laboratory, Cincinnati, OH,
to announce key findings of the re-
search project that is fully documented
in a separate two-volume report of the
same title (see Project Report ordering
information at back).
Introduction
As part of the EPA's efforts to remediate
Superfund sites, several remediation tech-
nologies can be candidates for consider-
ation. One of the more frequently used
technologies to decontaminate soils con-
taminated with organic hazardous constitu-
ents is incineration. High-temperature in-
cineration, while effective in destroying or-
ganic compounds, may not be necessary
for some soils that need treatment, such
as soils contaminated with volatile organic
compounds (VOCs). Also, in soils con-
taminated with toxic trace metals, high-
temperature incineration may increase the
volatilization of some metals into the com-
bustion flue gas. The presence of elevated
levels of volatile trace metals in the flue
gas can pose increased challenges to an
air pollution control system (APCS).
Another thermal treatment technology,
thermal desorption, may be an attractive
alternative to incineration. When success-
ful in decontaminating soils to the neces-
sary degree, thermal desorption treatment
of soils offers the benefits of lower fuel
consumption, avoidance of slag formation,
reduced metals volatilization, and reduced
APCS demands.
Most conventional rotary kiln incinera-
tors can be easily operated at tempera-
tures below those typically employed for
incineration treatment. Thus, the question
arises: how effective is the treatment of
contaminated soils by a rotary kiln incin-
erator operated at the low-to-moderate
temperatures?
To address this question, a series of
tests was conducted in the rotary kiln in-
cineration system (RKS) at EPA's Incin-
eration Research Facility (IRF). In these
tests the kiln of the RKS operated at low-
to-moderate temperatures. The test pro-
gram consisted of 12 tests under 11 dif-
ferent kiln operating conditions; one test
condition was tested in duplicate.
The objective of the test program was
to study the global effects of five param-
eters believed to be of primary impor-
tance in the effectiveness of soil decon-
tamination and in the fate of contaminant
metals. These parameters were soil mois-
ture content, treatment temperature, treat-
ment time, solids bed depth, and degree
of solids agitation.
The results obtained from the test pro-
gram were intended to yield the following
information:
• The relationship between compound
boiling point (vapor pressure) and the
extent of decontamination for each
organic contaminant
• How the solids bed temperature af-
fects decontamination
• How the presence and the amount of
moisture affect organic decontamina-
tion effectiveness
• The relationship between treatment
time, treatment temperature, and or-
ganic constituent decontamination ef-
fectiveness
• The distribution of trace metals in pro-
cess discharges when a metal-con-
taminated soil is treated by thermal
desorption
• Whether thermal desorption treatment
conditions affect a metal's leachabil-
ity from the treated soil
Test Program
Test Facility
All tests were performed in the RKS at
the IRF. A process schematic of the RKS
Quench
Secondary
Burner
Natural
Gas,
Liquid
Feed
Transfer
Duct
Rotary Kiln Incinerator
Packed
/\ Column
V \ Scrubber
Scrubber
Liquid
Recirculation
Primary Air Pollution
Control System
cb
Demister
Carbon Bed Hepa
Adsorber Filter
Redundant Air
Pollution Control
System
Atmosphere
I
Stack
4)
ID Fan
Figure 1. Schematic of the IRF rotary kiln incineration system.
2
-------
is shown in Figure 1. The IRF RKS con-
sists of a primary combustion chamber, a
transition section, and a fired afterburner
chamber. After exiting the afterburner, flue
gas flows through a quench section fol-
lowed by a primary APCS. The primary
APCS for these tests consisted of a ven-
turi scrubber followed by a packed-col-
umn scrubber. Downstream of the primary
APCS, a backup secondary APCS, com-
posed of a demister, an activated-carbon
adsorber, and a high-efficiency particulate
air (HEPA) filter is in place.
Test Contaminated Soil
A synthetic contaminated soil was pre-
pared for testing by mixing a locally ob-
tained topsoil with an attapulgite clay oil
sorbent in a 1:1 weight ratio. This clay
additive was required to allow the test
mixture to be reliably fed continuously to
the kiln of the RKS using a screw feeder.
The local topsoil without the clay additive
readily bridged in the feed hopper, pre-
venting reliable feed.
The test soil/clay mixture was spiked to
contain contaminants reflecting contami-
nation by gasoline, volatile organic sol-
vents, semivolatile organic compounds
associated with coal tar, and trace metals.
Benzene, n-heptane, and n-octane repre-
sented gasoline components; benzene,
toluene, tetrachloroethene, and chloroben-
zene represented volatile organic solvents;
and naphthalene, phenanthrene, and
pyrene represented coal tar constituents.
Spiking levels ranged from 2,000 to
4,800 mg/kg in the final synthetic contami-
nated soil mixture for the volatile organic
compounds (VOCs) added, and from 200
to 600 mg/kg for the semivolatile organic
compounds (SVOCs) added. The test soil/
clay mixture was also spiked to contain
commonly encountered hazardous con-
stituent trace metal contaminants. The
trace metals spiked were arsenic, barium,
cadmium, chromium, lead, and mercury.
Spiking levels ranged from 10 to 200 mg/
kg in the final synthetic contaminated soil
mixture.
The soil/clay mixtures were prepared in
two 3.5-ft3 (100-L) cement mixers via the
addition of weighed quantities of each mix-
ture 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 spiking, the
moisture content of the spiked/soil/clay
mixture was adjusted to one of two test
program targets of 10% or 20% by adding
additional water, it needed. The final soil
mixtures were tumbled to uniform appear-
ance, then transferred to 55-gal (208-L)
drums that were then sealed. Contami-
nated soil mixtures were allowed to age
between 7 and 14 days before use in a
test.
Table 1 summarizes the organic solu-
tion composition used to spike the test
mixtures. The organic contaminant mix-
ture was added to the soil/clay mixtures in
the ratio of 0.02 kg organic liquid per kg
final soil mixture. Resulting contaminated
soil organic constituent concentrations are
also noted in Table 1. The composition of
the concentrated aqueous solution of trace
metals added is summarized in Table 2.
All metal constituents were added as
soluble nitrate salts except arsenic, which
was added as As203 dissolved into the
acid nitrate spike solution. The metals
spike solution was added to the soil/clay
mixtures iri the ratio of 0.05 kg spike solu-
tion per kg of final contaminated soil mix-
ture. Resulting contaminated soil trace
metal concentrations, neglecting native
soil/clay metal concentrations, are also
noted in Table 2.
Test Conditions
As noted above, the test program con-
sisted of 12 tests under 11 different com-
binations of the test variables, with one
test performed in duplicate. The test pa-
rameters were soil moisture content, treat-
ment temperature, treatment time, solids
bed depth, and degree of solids agitation.
Soil moisture content was directly varied,
at 10 and 20%, as noted above. 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 feed
rate, and kiln rotation rate. Three target
kiln exit gas temperatures were tested,
320°, 480°, and 650°C (600°, 900°, and
1,200°F). Two target feed rates were
tested, 70 and 210 kg/hr (150 and 470 lb/
hr). Three target kiln rotation rates were
tested, 0.2, 0.5, and 1.5 rpm.
Kiln exit gas temperature primarily af-
fected 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 soil
residence time in the kiln, or maximum
treatment time. Total kiln soil residence
times corresponding to the above rotation
rates were 60, 40, and 30 minutes, re-
spectively. The combination of feed rate
and kiln rotation rate affected solids bed
depth.
Total treatment times were changed by
varying kiln rotation rates, as noted above.
However, to allow for the evaluation of
treatment effectiveness at partial treatment
times for each test condition, samples of
the soil bed material were taken at four
axial locations along the kiln for each test,
in addition to a soil 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 3. The "center point" of the test ma-
trix is represented by Test 2, with soil feed
rate at 70 kg/hr (150 Ib/hr), kiln exit gas
temperature of 480°C (900°F), kiln rota-
tion rate of 0.2 rpm, and soil moisture
content of 10%. This test condition was
tested in duplicate (Test 12). From this
"center point," kiln temperature was var-
ied (Tests 1 and 3), soil moisture content
was varied (Test 5), kiln rotation rate was
varied (Test 7), and soil feed rate was
varied (Test 10). Additional test combina-
tions were performed for the high mois-
ture soil at the base feed rate and rotation
rate (Tests 4 and 6), at the high feed rate
and base rotation rate (Test 11), and at
the base feed rate and increased rotation
rate (Test 8). The highest rotation rate was
tested at high feed rate with the high mois-
ture soil (Test 9).
For all tests, the afterburner was oper-
ated at 1,090°C (2,000°F) to ensure satis-
factory burnout of all volatilized organic
compounds. The scrubber system was
operated under its nominal design condi-
tions to achieve typical scrubber perfor-
mance. The scrubber was operated at near
total recycle, so there was minimum
blowdown. The synthetic contaminated soil
was fed continuously until all flue gas sam-
pling was completed. Treated soil was
continuously removed from the kiln ash
hopper via an ash auger transfer system,
and deposited in clean 55-gal (208-L)
drums. After the completion of each test
(all flue gas sampling completed) the sys-
tem 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 moni-
tored continuously throughout the test; the
resulting data allowed the calculation of
total kiln soil residence times.
A summary of the actual test conditions
in effect for each test is given in Table 4.
As shown, average kiln exit gas tempera-
ture targets were closely met for all tests.
Soil 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
3
-------
Table 1. Organic Constituents in the Synthetic Contaminated Soil
Organic liquid Concentration in soil
mixture at an organic liquid
Compound
Molecular
weight
Specific
gravity
Melting point,
°C
Boiling point,
°C
composition,
wt%
fraction
mgJkg
Benzene
78.1
0.88
6
80
15
3,000
n-Heptane
100.2
0.68
-91
98
15
3,000
Toluene
92.7
0.87
-95
111
15
3,000
Tetrachloroethene
165.9
1.62
-22
121
24
4,800
n-Octane
114.2
0.70
-57
126
15
3,000
Chlorobenzene
112.6
1.11
50
132
10
2,000
Naphthalene
128.2
1.16
80
218
3
600
Phenathrene
178.2
1.18
100
340
2
400
Pyrene
202.2
1.27
156
404
1
200
Table 2. Trace Metal Constituents in the Synthetic Contaminated Soil
Aqueous spike solution
Metal
Metal
concentration,
ig/L
Compound
Compound
concentration",
gfi-
Resulting soil
feed metal
concentration1',
mg/kg
Arsenic
0.50
Asp3
0.67
25
Barium
4.0
Ba(N03)2
7.61
200
Cadmium
0.20
Cd(N03)2 • 4HsO
0.55
10
Chromium
0.50
Cr(NOJ3 • 9HsO
3.8
25
Lead
0.80
Pb(NOJ!
1.28
40
Mercury
0.20
Hg(N03)2
0.32
10
'Sufficient HNOs added to maintain lead arsenate compounds In solution.
bNegllgible soil metal concentrations and a ratio of 0.05 kg of spike solution per kg of organlc/soil/spike solution mixture assumed.
-------
probe that allowed the immersing of four
thermocouples in the soil bed at the re-
spective axial locations. Soil bed tempera-
tures measured for the tests are also given
in Table 4.
Sampling and Analysis
For all tests, the sampling protocol con-
sisted of
• Obtaining a composite sample of the
contaminated soil feed material mix-
ture
• Obtaining composite samples of the
treated soil in the kiln chamber at
four axial locations corresponding to
the solids bed temperature measure-
ments: 0.6, 1.1, 1.5, and 2.0 m (2.0,
3.5, 5.5, and 6.5 ft) from the kiln feed
face
• Obtaining a composite sample of the
treated soil discharge from the dis-
charge collection drum
• Obtaining composite pretest and
posttest 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 02, CO, and
total unburned hydrocarbon (TUHC)
levels in the kiln exit flue gas
• Continuously monitoring 02 in the af-
terburner exit flue gas
• Continuously monitoring 02 and C02
downstream of the venturi/packed-col-
umn scrubber
• Continuously monitoring O and CO
in the stack downstream of the sec-
ondary APCS (carbon bed/HEPA fil-
ter)
• Sampling the stack gas for particu-
late, and HCI and Cl2 using Method
50
As noted above, 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 were opened and
sampled. Drum contents were then trans-
ferred to the screw feeder hopper for feed-
ing.
Four composite kiln solids bed samples
were also collected for each test. One
sample was collected using a custom-fab-
ricated quartz scoop at each of the four
axial locations where soil bed tempera-
ture was measured. Each of these samples
represented a different treatment time un-
der the set of other test conditions estab-
lished 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 VOC and SVOC 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 poly-
chlorinated dibenzofurans (PCDDs/
PCDFs) by Method 8290.
Toxicity characteristic leaching proce-
dure (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
Table 3. Target Test Conditions
Test
Kiln exit gas
temperature,
°C (°F)
Expected peak
solids bed
temperature,
°C (°F)
Kiln rotation
rate, rpm
Soil feed rate,
kg/hr (Ib/hr)
Soil
moisture
content,
O/
/O
1.
320 (600)
120 (250)
0.2
70 (150)
10
2.
480 (900)
270 (520)
0.2
70 (150)
10
3.
650 (1,200)
430 (800)
0.2
70 (150)
10
4.
320 (600)
120 (250)
0.2
70 (150)
20
5.
480 (900)
270 (520)
0.2
70 (150)
20
6.
650 (1,200)
430 (800)
0.2
70 (150)
20
7.
480 (900)
270 (520)
0.5
70 (150)
10
a.
480 (900)
270 (520)
0.5
70 (150)
20
9.
480 (900)
270 (520)
1.5
70 (150)
20
10.
480 (900)
270 (520)
0.2
210 (470)
10
11.
480 (900)
270 (520)
0.2
210(470)
20
12.
480 (900)
270 (520)
0.2
70 (150)
10
5
-------
Table 4. Actual Test Operating Conditions
Test
Date
1 2
1/29/93 2/2/93
3
2/4/93
4
1/27/93
5
12/4/92
6
12/16/92
7
1/6/93
8
12/9/92
9
2/11/93
10 11 12
1/15/93 1/19/93 1/21/93
Average soil
feed rate,
kg/hr
(Ib/hr)
Kiln exit gas
temperature
Range, °C
CF)
Average, °C
CF)
Kiln rotation rate,
rpm
Total kiln soil
residence time,
min
Average soil bed
temperature at
"x" m ("x" ft) from
feed face,
C° (F°)
0.6(2.0)
1.1 (3.5)
1.5(5.0)
2.0(6.5)
67
(148)
68
(150)
66
(145)
68
(149)
63
(138)
70
(155)
65
(144)
69
(152)
66
(145)
230
(506)
225
(497)
66
(146)
305-331 474-492 641-658 301-330 463-499 633-660 471-494 467-501 471-492 287-508 464-500 471-493
(501-627) (885-918) (1,185-1,216) (524-626) (866-930) (1,171-1,220) (880-921) (873-935) (880-918) (548-946) (867-932) (879-919)
317
(603)
0.2
58
482
(900)
0.2
68
648
(1,199)
0.2
72
316
(601)
0.2
64
482
(900)
0.2
648
(1,199)
0.2
61'
56
482
(900)
0.5
60
480
(896)
0.5
32
482
(900)
1.5
27
481
(897)
0.2
46
482
(900)
0.2
38
482
(900)
0.2
64
86(186) 123(253) 267(512)
109(228) 182(359) 393(740)
115 (239) 228 (443) 434 (814)
123(254) 227(440) 451(844)
69(156) 113(235) 122(251) 144(292) 179(355) 172(341) 130(266) 107(224) 144(291)
88(191) 204(399) 319(606) 204(400) 215(419) 194(382) 161(321) 132(269) 211(412)
98(208) 260(500) 401(753) 249(481) 239(462) 245(473) 175(347) 149(300) 257(495)
123(254) 182(359) 356(672) 241(465) 231(448) 277(531) 253(487) 239(463) 244(471)
'Estimated.
-------
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.
All pretest and posttest scrubber liquor
samples were analyzed for the test trace
metals by the same methods employed
for the TCLP leachate samples. In addi-
tion, one composite pretest scrubber li-
quor and all posttest scrubber liquor
samples were analyzed for the spiked vola-
tile and semivolatile organic soil contami-
nants.
Finally, all multiple metals train samples
were analyzed for the non-mercury test
trace metals, and the Method 101A sam-
pling train samples were analyzed for mer-
cury.
Test Results
The following summarizes major test
program findings. The test program re-
sults and conclusions are summarized ac-
cording to the points noted in the intro-
duction.
Organic Decontamination and
Compound Boiling Point
The test program data clearly demon-
strate that lower-boiling compounds (more
volatile) can be driven out of the test soil
nearly quantitatively (to non-detectable lev-
els) and rapidly (in less than 20 minutes).
In contrast, higher-boiling-point compounds
remain in the soil, and require higher treat-
ment temperatures (greater than 480°C
(900°F) kiln exit gas temperature) and
longer treatment times (longer than 30 min-
utes) for effective soil decontamination.
Treatment at gas temperatures less than
480°C (900°F) can result in very poor
decontamination of the high-boiling-point
organic compounds tested.
Organic Decontamination and
Soil Temperature
The test program data demonstrate that
elevated soil temperature favors decon-
tamination. In particular, elevated tempera-
ture is essential for satisfactory decon-
tamination of high-boiling-point com-
pounds. However, while soil temperature
appears to be correctable to the degree
of organic decontamination, it is not, by
itself, a sufficient predictive indicator of
decontamination level. Other parameters,
such as the presence of moisture or the
degree of soil bed agitation, can enhance
the decontamination process under the
right combination of conditions.
Organic Decontamination and
Soil Moisture Content
The test program data obtained are dif-
ficult to interpret with respect to evaluat-
ing the effects of soil moisture content on
organic compound decontamination effec-
tiveness. Competing phenomena are likely
involved. Three possible explanations for
the test observations are developed as
follows:
Increases in moisture content reduce
solid heat-up rates due to the higher ther-
mal mass (specific heat and latent heat of
vaporization). This tends to reduce the
rate of organic compound decontamina-
tion.
Inferences based on kiln soil holdup
weights and kiln soil residence times sug-
gest that increased moisture in the test
soils within the tested range increased the
soil bed agitation when the kiln was ro-
tated at 0.5 rpm or faster. This increase in
agitation led to faster heat absorption and
reduced mass transfer resistance, but
shorter solids residence times. Faster heat
absorption increases the driving force for
volatilization. Reduced mass transfer re-
sistance increases decontamination rate.
Shorter residence times lower the achiev-
able decontamination extent of an organic
constituent, if the decontamination rate for
that constituent is slow.
Increased moisture content may en-
hance the decontamination rates for the
less volatile organic compounds through
steam stripping, provided that the addi-
tional moisture, with its higher specific heat
and latent heat of vaporization, does not
prevent the soil from reaching tempera-
tures that are necessary for effective strip-
ping of the volatile organic contaminants.
Organic Decontamination and
Treatment Time and
Temperature
The extent of organic compound de-
contamination increases with solids resi-
dence time and also increases with treat-
ment temperature. However, the relation-
ship between these variables is not
straightforward. An attempt to summarize
all the test program data is illustrated in
Figures 2 and 3. The figures include all
test program data collected in the range
of treatment times from 7 to 67 minutes,
with the difference between soil bed tem-
perature and compound boiling points
ranging from -333° to +371 °C (-599° to
+668° F).
Figure 2 includes 432 data points (4
locations per test, 9 compounds per test
and a total of 12 tests for the test pro-
gram) and is a plot of the extent of decon-
tamination achieved versus the difference
between soil temperature and compound
boiling point. The extent of decontamina-
tion achieved is represented by the ex-
pression C/Co, where C is the final con-
taminant concentration in the treated soil
and Co is the feed soil contaminant con-
centration. The difference between soil
temperature and compound boiling point
is represented by (TS01l-Tbp) where Toil is
the soil bed temperature and Tb is the
compound boiling point.
While the data form a definite pattern
showing increased decontamination with
increased soil bed temperature, the scat-
ter in the test data is significant, thereby
indicating some influence by parameters
other than treatment temperature.
Figure 3 is an attempt to show the com-
bined effects of soil bed temperature and
treatment time, where a first-order depen-
dence on each of the two variables is
assumed. In this figure, the horizontal axis
is the product of treatment time with the
difference between treatment temperature
and compound boiling point. This product,
with unit of °C-min can be considered the
cumulative volatilization driving force. As
in Figure 2, the test data show a definite
pattern of increased degree of decontami-
nation with increases in the soil tempera-
ture-time product. Although the data points
appear to be gathered a bit more tightly
than in Figure 2, data scatter remains con-
siderable. This, again, suggests that pa-
rameters other than treatment time and
soil temperature are important, or that the
dependence on treatment time and soil
temperature is not as simple as first order
in both.
Trace Metal Distributions
Overall average metals mass balance
closures achieved in the test program
ranged from 96%—113%. Individual metal
mass balance closures achieved ranged
from 38%—181 % at the temperatures
tested. On average, arsenic, barium, cad-
mium, chromium, and lead were not vola-
tile and remained in the soil. With the
exception of mercury, the extent of metal
volatilization from the treated soil was not
significantly affected by any of the test
variables. No significant reductions in the
fractions recovered in the treated soil oc-
curred with increased treatment time.
Mercury was the volatile metal, as ex-
pected, tending to be equally distributed
between the treated soil and the scrubber
exit flue gas. At the highest soil treatment
temperature tested, the extent of mercury
volatilization increased with increasing
treatment time.
7
-------
-400 -300 -200 -100 0 100
( Tgoil - Tbp),°C
200
300
400
Figure 2. Organic compound decontamination effectiveness versus the differences in soil bed temperature and compound boiling point.
4.0
O
o
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-12,000 -10,000 -8,000 -6,000 -4,000 -2,000
(T*8oiI ~ "^bp)treatment, C-min
~
rfW
2,000
4,000 6,000
Figure 3. Effect of cumulative treatment time and achieved solid temperature (adjusted for compound boiling point) on organic decontamination.
-------
Trace Metal Leachability
The effects of thermal treatment on met-
als leachabilities in the TCLP varied from
metal to metal. Among the test variables,
the most influential one was treatment tem-
perature. The behavior of the metals'
leachabilities are summarized in the fol-
lowing:
Lead and barium leachabilities were not
affected by any of the test variables.
Leachable fractions of arsenic from the
untreated soil ranged from 16%-37%, with
an average of 26% over the 12 tests.
Arsenic leachabilities in the TCLP did not
change significantly at low treatment tem-
peratures (soil temperatures below 123°C
[254°F]). At maximum soil temperatures
ranging from 228° to 277°C (443° to
531 °F), arsenic leachability decreased to
about 15%. At yet higher soil tempera-
tures (400° to 451 °C [753° to 844°F]),
arsenic leachability decreased further, to
about 10%.
While effects of feed rate and rotation
speed were apparent, these effects may
be attributable to the different maximum
solid temperatures that resulted from the
changes in feed rate and rotation speed.
Overall, the test data suggest that a mini-
mum soil treatment temperature of 228°C
(442°F) is required to reduce arsenic leach-
ability from the synthetic soil tested. Soil
moisture content, within the tested range,
had no effect on arsenic leachability.
At the highest soil temperatures
reached, above 300°C (572°F), the soil
contained a smaller leachable cadmium
fraction, at about 2%, compared to about
13%—16% leachable from the feed. It is
possible that cadmium leachability in the
TCLP may be further reduced at higher
soil temperatures.
The chromium in the feed soil was barely
leachable, at about 1% or less. The data
show that soil temperature affects chro-
mium mobility in the treated soil. A dra-
matic increase in the leachable fraction of
chromium, from less than 1% to more
than 15%, was observed when soil treat-
ment temperature was increased from
225°C (437°F) to 275°C (527° F). At higher
treatment temperatures and increased kiln
rotation rates, the onset of the observed
leachability increase occurred at shorter
treatment times.
The mercury leachability data suggest
behavior comparable to that of chromium.
When heated to temperatures of 200° to
275°C (392° to 527°F), the treated soil
gave TCLP leachate mercury concentra-
tions two or three times those from the
feed samples. These corresponded to
leachable fractions of 5%-62%. Unlike
chromium, however, when heated above
300°C (572°F), the treated soil TCLP
leachate mercury concentrations returned
to levels similar to corresponding feed
TCLP concentrations.
Conclusions
In summary, operating an incinerator at
low-to-moderate temperatures can, under
the right conditions, effectively decontami-
nate soils containing organic contaminants,
including high-boiling-point compounds.
The principal parameters that affect de-
contamination effectiveness are compound
boiling point, achieved solids temperature,
and soil moisture content. Agitation can
affect decontamination effectiveness, most
likely by increasing the rate at which heat
is absorbed into the solid, and by de-
creasing the average mass transfer resis-
tance.
Except for mercury, trace metal con-
stituents were not volatile under low-to-
moderate temperature incineration condi-
tions. The leachability of arsenic, cadmium,
and chromium in the TCLP can be af-
fected by treatment temperature.
The full report was submitted in fulfill-
ment of Contract No. 68-C9-0038, Work
Assignment 4-1, by Acurex Environmental
Corporation under the sponsorship of the
U.S. Environmental Protection Agency.
9
-------
J. Lee, D. Fournier, Jr., C. King, S. Benkatesh, and C. Goldman are with Acurex
Environmental Corporation, Jefferson, AR 72079
R. C. Thurnau is the EPA Project Officer (see below).
The complete report consists of two volumes entitled "Evaluation of Rotary Kiln
Incinerator Operation at Low to Moderate Temperature Conditions,"
Volume I. Technical Results (Order No. PB96-210414; Cost: $49.00, subject to
change)
Volume II. Appendices (Order No. PB96-210422; Cost: $49.00, subject to change)
The above reports will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/SR-96/105
10
------- |