&EPA
United States
Environmental Protection
Agency
Office of Solid Waste and
Emergency Remedial Response
Washington, DC 20460
Office of
Research and Development
Cincinnati, OH 45268
Superfund
EPA/540/S-92/014
September 1992
Superfund
Engineering Issue
Considerations for Evaluating the Impact
of Metals Partitioning During the Incineration
of Contaminated Soils From Superfund Sites
Index
Introduction
Overview
Metals Behavior
Metals Partitioning
Fundamental Parameters
Matrix
Operational
Combustion Chamber Design
Air Pollution Control Devices
Material Balance Closure Issues
Conclusion
Technology Contacts
Acknowledgements
References
Introduction
The shortage of landfill space, EPA's land ban restrictions,
and the demonstrated effectiveness of incineration are making
incineration a preferred treatment for large volumes of con-
taminated soils. Because of the limited capacity in suitable
RCRA incinerators, the advantage of using transportable incin-
erators on a "campaign basis," and its effectiveness, on-site
thermal treatment of Superfund soils is being chosen frequently
as the method of remediation in the Superfund Record of
Decisions (RODs).
When Superfund soils containing metals are incinerated,
the metals vaporize, react to form other metal species, or
remain with the soil residuals. The vaporized metals can un-
dergo reactions to form other species or condense to form fine
particulates and/or fumes. Some metals oxidize to form new
species with lower oxidation states or react with other elements
such as chlorine and sulfur. The new species will either volatilize
or fall out as ash. The majority of metals will simply remain as
ash or be entrained in the form of fly ash.
Metals in ash, scrubber sludge, and/or stack emissions, if
improperly managed, can result in potential exposures and the
resultant adverse health effects. Metals that remain with the
ash have the potential to leach when disposed of in improper
landfills [6]. Scrubber sludge must be stabilized before being
landfilled if it exhibits leachability characteristics. Metal par-
ticles, vapors, and fumes can become environmental and health
concerns when released from a stack. Table 1 shows metals of
interest in stack emissions. The metals under the first column
are metals for which EPA guidance on emissions has been issued
[10]. The second column includes additional metal constitu-
ents that are listed in Appendix VIII of 40 CFR Part 261 [14].
Metals can also react with other elements in the feed stream,
such as chlorine or sulfur, forming more volatile and toxic
compounds than the original species [6].
Conducting treatability studies is an important step in the
selection of treatment alternatives for the remediation of
Superfund sites as well as other hazardous waste sites. By
gaining valuable information on the feasibility and cost of treat-
ment options, such as off-site or on-site incineration, the risk of
failure at full-scale implementation due to the selection of an
inappropriate technology can be minimized. Bench- and pilot-
scale treatability testing, as well as limited full-scale testing,
can provide valuable information in evaluating remedial
alternatives [11 ].
fes) Printed on Recycled Paper
Superfund Technical Support Center
for Engineering and Treatment
Risk Reduction Engineering
Laboratory
Technology Innovation Office
Office of Solid Waste And Emergency
Response, U.S. EPA, Washington, DC
WalterW. Kovalick, Jr., Ph.D.
Director
-------
Table 1. Metals of Interest [4]
RCRA Guidance
Other RCRA Constituents
Antimony (Sb)
Arsenic (As)
Barium (Ba)
Beryllium (Be)
Cadmium (Cd)
Chromium (Cr)
Lead (Pb)
Mercury (Hg)
Silver (Ag)
Thallium (Tl)
Osmium (Os)
Nickel (Ni)
Vanadium (V)
Selenium (Se)
This paper provides guidance for Remedial Project Manag-
ers (RPMs) and On-Scene Coordinators (OCSs) considering
incineration of soils containing metals. Important consider-
ations impacting metals partitioning are addressed. If a
treatability study is necessary, certain steps must be taken to
ensure the success of the test and to define the fate of the
metals in discharge streams; these steps are also addressed.
This paper examines the available metal partitioning data
for incineration of soils from treatability studies of three Superfund
sites (Baird and McGuire, Florida Steel, and McColl) and from
synthetic soil matrix studies conducted at the U.S. Environmen-
tal Protection Agency's Incineration Research Facility (IRF) in
Jefferson, Arkansas. (This facility was formerly known as the
Combustion Research Facility, or CRF.) The rotary kiln incinera-
tor at the IRF was used for the Baird and McGuire treatability
study. A pilot-scale treatability study was conducted at the
Ogden Environmental Services research facility using a circulat-
ing bed combustorto incinerate soils from the McColl site. Soil
from Florida Steel was incinerated as part of a Toxic Substances
Control Act PCB destruction demonstration at the site using an
infrared incineration unit.
Overview
Incineration has been shown to be effective in treating
soils, sediments, sludges, and liquids containing primarily or-
ganic contaminants. Organic contaminants are destroyed by
subjecting them to temperatures typically greater than 800° C
in the presence of oxygen, which causes the volatilization and
combustion of these compounds. For any soils incineration
system, the efficiency of the combustion process depends upon
the following: the temperature the contaminated soil is sub-
jected to, the time the soil is subjected to that temperature, and
the degree to which the contaminants are exposed to oxygen
during combustion through the mixing process.
Incineration systems are composed of several integrated
unit operations. A generic flow diagram showing these opera-
tions is presented in Figure 1. A typical incineration system
consists of a primary combustion chamber (PCC) such as a kiln
Figure 1. Incineration System Concept Flow Diagram [9]
or furnace, a secondary combustion chamber (SCC) or after-
burner, and air pollution control devices (APCDs). The APCDs
often consist of a venturi scrubber and/or packed tower scrub-
ber for particulate and acid gas removal. The scrubbers may be
followed by an electrostatic precipitator or fabric filter for addi-
tional particulate collection. Many other APCDs or APCD trains
are also available.
The general combustion chamber operating conditions
favorable to the destruction of contaminated soils are well
established and are presented in Table 2. Oxygen (in the form
of air) is supplied to the combustion chamber in excess of what
is stoichiometrically required to compensate for incomplete
mixing. There are several different types of incineration systems
based on the design of the primary combustion chamber. The
most common of these is the rotary kiln incinerator. Other
types of incineration systems include fluidized bed, infrared,
and liquid injection. Liquid injection incinerators will not be
reviewed herein, since Superfund soils are not amenable to
treatment by such systems.
Table 2. General Combustion Chamber Conditions Favorable to
the Destruction of Contaminated Soils [9]
Combustion chamber temperature level: 800°C - 1500°C
(1472°F-2700°F)
Residence time:
Excess combustion air:
Several minutes to 1/2 hour
for solids/sludges
60 -130% of stoichiometric
requirements
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
-------
A rotary kiln incinerator is a long, inclined cylinder that
rotates on its longitudinal axis. This rotating cylinder provides
agitation to ensure that all of the soil is exposed to the oxygen
present and to improve heat transfer. The contaminated soil
and auxiliary fuels are usually introduced to the kiln at the high
end and are transferred to the low end by gravity. Ash residue
is collected at the low end of the kiln. Exhaust gases typically
pass to a SCC or afterburner for further oxidation. Rotary kilns
are used primarily for the treatment of solids;
but liquids, sludges, and gases can be co-incinerated with
solids [9].
Fluidized bed incinerators use high air velocity to suspend
a bed of inert granular material. The bed is preheated by an
auxiliary burner, and fuel and contaminated soil are introduced
into the bed. Air passing through the bed causes turbulence
which facilitates intimate mixing of the air, heated bed mate-
rial, fuel, and contaminated soil. A modification of this technol-
ogy, called circulating fluidized bed incineration, operates at
higher velocities and with finer bed particles than fluidized bed
combustors. The bed materials, including the contaminated
soil, are passed through a combustion loop consisting of the
combustion chamber and a second chamber (cyclone) where
the flue gas is separated from the solid particles. A major
portion of the solid particles separated in the second chamber is
reinjected into the first chamber. Circulating fluidized beds do
not require a SCC or afterburner. The technology is applicable
for liquids and sludges and may be used for solids with small
particle sizes [9].
Infrared radiation may also be used as a heat source for
thermal destruction of contaminated soils. Contaminated soils
are fed into the combustion chamber on a woven wire con-
veyor belt. Electrical resistance heating elements are used to
generate thermal radiation within the chamber. Solids are
carried through the chamber on the belt until reaching the
discharge end of the chamber where ash drops off r'nto a
hopper. Exhaust gases pass through a secondary combustion
chamber. This technology is intended for treatment of solids,
stabilized sludges, and contaminated soils. As with the fluidized
bed incinerators, feed preparation must include specific sizing
of the material to maintain a consistent layer on the belt [9].
Metals Behavior
A significant amount of research has been conducted on
the incineration of contaminated materials. Although histori-
cally the primary focus has been on the destruction of organics,
increased emphasis is being placed in the fate of metals. Unlike
the organic portion, the metal fraction may change form but is
not destroyed. Metals can remain with the ash when dis-
charged from the primary combustion chamber [6]. They can
also volatilize or oxidize to form fumes or fine partjculates and
pass through the incineration system [6][13]. Metal vapors,
fumes, or particles can be collected by the APCDs, exit out the
stack, or be deposited along the walls of the combustion cham-
ber and remain in the incineration system [6][13]. Potential
pathways for metals are illustrated in Figure 2. This distribution
of metals into the various incinerator system components is
called "partitioning." A number of factors affect how various
metals will partition, or behave, during incineration.
Metals Partitioning
Partitioning is highly dependent on the volatility of the
metal [2][6][13]. The temperature of volatilization of a metal
can be predicted using basic laws of physical chemistry and the
concentration of the metal in the waste stream. Volatility
temperature is defined as the temperature at which the effec-
tive vapor pressure of the metal is 1 x 10~6 atm. At this vapor
HOMOGENEOUS
CONDENSATION
'VAPOR
FUME NUCLEI
HOMOGENEOUS"
CONDENSATION
CHLORIDES
SULFIDES
OXIDES
FLY ASH
ENTRAINMENT
OLID
FUME
RESIDUAL ASH
Figure 2. Metals Behavior in Combustion Devices [4]
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
-------
Table 3. Predicted Metals Volatility Temperatures [4]
With 0% Chlorine
With 10% Chlorine
Metal
Chromium
Nickel
Beryllium
Silver
Barium
Thallium
Antimony
Lead
Selenium
Cadmium
Osmium
Arsenic
Mercury
Volatility
Temperature (°C)
1613
1210
1054
904
849
721
660
627
318
214
41
32
14
Principal
Species
CrO/CrOj
Ni(OH)2
Be(OH)2
Ag
Ba(OH)2
TI203
Sb203
Pb02
Se02
Cd
Os04
As203
Hg
Volatility
Temperature (°C)
1610
693
1054
627
904
138
660
-15
318
214
41
32
14
Principal
Species
Cr(yCr03
NiCI2
Be(OH)2
AgCI
BaCI2
TIOH
Sb203
PbCI4
Se02
Cd
Os04
As203
Hg
pressure, a measurable amount of vaporization is likely to occur
[1]. Although total metal concentrations are typically deter-
mined, the specific form of the metal in the waste stream is
generally not measured. Determination of the "volatility" tem-
perature of the metal compounds within a given soil would
require significant analytical costs; therefore, it is generally not
done. Table 3 lists predicted volatility temperatures for the
metals (except vanadium) from Table 1 and their principal
species with and without the presence of chlorine [4]. Table 3
data are based on a model rather than actual field measure-
ments. Volatility temperatures range from 14°C to 1613°C
without chlorine and -15°C to 1610°C with chlorine. Those
metals exhibiting volatilization temperatures below about 900°C
are commonly referred to as volatile metals, while those having
higher volatilization temperatures are referred to as refractory
metals. The volatile metals are more likely to pass through the
combustion chambers as vapor, fume, or fine particulars. The
refractory metals are more likely to remain in the PCC ash,
although they can vaporize or be entrained with the flyash
under certain operating conditions (e.g., high temperature,
high flue gas flowrate, high turbulence).
Fundamental Parameters
Table 4 lists fundamental parameters that influence metals
behavior. These include those associated with the matrix to be
treated, the operation of the incinerator, the design of the
combustion chamber, and the type of APCD used [4].
Matrix
The concentration and species of metal in the contami-
nated soil help determine the degree to which the metal will
volatilize. The soil matrix, organic constituents, and halogen
content also impact the volatilization of certain metals. As the
metal feedrate is increased, more metal will vaporize until a
saturation point is reached, at which additional metal is parti-
tioned to the ash rather than vaporized. The particle size
distribution and propensity to fragment affect the amount of
metals that can become entrained and carried from the PCC as
fly ash [3].
Table 4. Fundamental Parameters That Influence Metals Behavior [4]
Matrix Parameters
Type and Concentration of Metals
Particle Size Distribution of Metals
Propensity to Fragment
Presence or Concentration of Organometals
Halogen Content
Operational Parameters
PCC Temperatures
SCC Temperatures
Stoichiometric Ratio of Oxygen to Contaminant
in Combustion Zone
Combustion Chamber Design Parameters
Degree of Mixing
Combustion Zone Velocity
Air Pollution Control Device Parameters
Parameters that Control Fine Particle Capture - Specific to
Type of Device
Temperature at APC
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
-------
One metal compound in the feed may behave differently
than another compound of the same metal. Volatilization
temperatures for each compound may be significantly
different.
One of the most important parameters affecting metals
behavior is halogen (e.g., chlorine) content. Since chlorinated
metal compounds are generally more volatile, the greater the
chlorine content the greater the degree of vaporization for
certain metal species [2][6].
Increasing chlorine concentration gives the metal more
opportunity to react and form chlorinated metal compounds.
Tests at Florida Steel and at the IRF spanned significant varia-
tions in chlorine content of the waste stream. The tests with a
high chlorine content showed an increase in metal volatility
and a corresponding decrease in metal partitioning to the PCC
ash. In addition to having a significant effect on the
partitioning of metals, halogen content can also affect the
efficiency of APCD's. These studies are discussed below.
Florida Steel—
In 1982, it was discovered that the soil of Florida Steel
Corporation's metal-recycling plant at Indiantown, Florida, was
contaminated with PCBs and metals, including cadmium and
lead. In 1987, a Toxic Substance Control Act (TSCA) PCB-
destruction demonstration was performed by OH Materials
Corporation (OHM) using its 100 ton per hour transportable
infrared unit. Table 5 summarizes the soil characteristics of the
site. One of the objectives of the trial burn was to determine
the impacts, if any, that chlorine has on the amounts of cad-
mium and lead that are volatilized. The test conditions for the
demonstration are shown in Table 6. The chlorine content
ranged from below 0.19 to 0.79 percent. PCC exhaust tem-
perature was varied slightly while feedrate and retention time
were kept relatively constant [7].
Metals emissions data for the furnace ash, scrubber water,
and stack are given in Table 7. The normalized data for both
metals are shown in Table 8 (in order to compare relative
distributions, mass fractions were normalized to the total amount
of metal measured in the output. Normalization of mass
fractions results in mass balance closure of 100%, thereby
eliminating significant test-to-test data variability). Based upon
the results obtained in Table 8, cadmium and lead became
slightly more volatile (mass fraction in PCC ash decreased) as
the chlorine content rose. Cadmium seemed to be affected
more by chlorine content than lead. (Note that volatility
temperature predictions suggested a strong impact of chlorine
on lead behavior and no impact on cadmium. A plausible
explanation for this is that the rnetal species associated with the
tests differed from those in Table 3.) The proportion of mass
fraction of scrubber solids to mass fraction of stack particulate
also decreased for both metals as the chlorine content in-
creased [7].
Synthetic Materials Tests at IRF—
In 1989, a research project investigating the fate of trace
metals in rotary kiln incineration with venturi- and packed
tower-scrubber particulate control was conducted at the EPA's
IRF in Jefferson, Arkansas. This testing was conducted in part to
support the development of a metal partitioning model and to
Table 5. Florida Steel Soil Characteristics [7]
Test Run No.
1
Test Date
Ash, wt %a
Chlorine, wt %
Metals (ppm):
Cadmium
Lead
9/28/87
87
<0.19
8.3
375
9/29/87
87
<0.26
9.9
459
10/3/87
87
0.79
7.4
332
10/4/87
87
0.68
11.3
434
10/5/87
87
0.71
7.9
352
Based on vendor information during phone conversation of October 1,1990.
Table 6. Florida Steel Operating Conditions [7]
Test Run No.
1
Test Date
Soil Feedrate (kg/hr)
PCC Temperature (° C)a
PCC Solids Residence Time (Min.)
SCC Temperature (" C)
SCC Excess 02 (%)
SCC Gas Residence Time (Sec.)
9/28/87
5203
731
22
1148
6.9
5.58
9/29/87
5289
802
25
1111
7.0
5.77
10/3/87
5351
843
23
1073
7.0
5.18
10/4/87
5360
714
23
1057
6.9
5.45
10/5/87
5508
783
23
1132
6.0
5.35
Primary exhaust temperature.
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
-------
Table 7. Florida Steel-Effluent Metal Concentrations [7]
Test Run No.
Furnace Ash, ppm '
Cadmium
Lead
Scrubber Water. mg/L
Cadmium
Lead
Stack mg/m3
Cadmium
Lead
1
10.9
772
1.57
7.96
0.517
13.5
2
9.0
489
2.12
7.73
2.38
1.07
3
6.0
494
0.97
9.73
0.89
10.7
4
7.2
447
1.18
11.9
5.74
1.34
5
7.0
510
0.892
9.92
1.34
11.7
Increase in concentration overfeed material partially accounted for by weight loss during incineration.
Table 8. Normalized Mass Fraction of Metals for Florida Steel [7]
Test Run No.
Test Date
Cadmium
PCCAsh1
Scrubber Solids1
Stack Particulate
Totalb
Lead
PCCAsha
Scrubber Solids4
Stack Particulate
Total"
1
9/28/87
83
7
10
100
99
0
1
100
2
9/29/87
75
9
16
100
97
1
2
100
3
10/3/87
61
4
35
100
93
1
6
100
4
10/4/87
69
6
25
100
94
1
5
100
5
10/5/87
65
4
31
100
94
1
5
100
a Used vendor estimate of 87% for ash content and 10 gpm for makeup water flow. Accumulation within system assumed to be neglible.
b Normalized to 100% of emissions as basis.
evaluate the predictive capability of the model. The feed
material consisted of a synthetic soil matrix made by adsorbing
aqueous mixtures of trace metals onto a clay material [2].
Feed concentrations of various metals during the tests are
summarized in Table 9. All tests were conducted at the same
nominal exit flue gas oxygen content (12 percent), afterburner
exit flue gas oxygen content (7.5 percent), synthetic soil feedrate
(63 kg/hr), and kiln rotational speed. The normalized metal
discharge distributions (percent of total measured) for the syn-
thetic material tests are presented in Table 10 [2].
As chlorine content increased from 0 percent to 8.3 per-
cent, the mass fraction of metals in the PCC ash (kiln ash)
decreased for the lead and bismuth. Cadmium was not de-
tected in the kiln ash from any of the tests. These three metals
were considered to be the most volatile metals. One of the
refractory metals, copper, showed an increase in volatility as the
chlorine content increased. The other refractory metals showed
Table 9. Nominal Feed Metal Concentrations for IRF Tests3 [2]
Metal
Concentration (ppm)
Arsenic
• Barium
Bismuth (Bi)
Cadmium
Chromium
Copper(Cu)
Lead
Magnesium (Mg)
Strontium (Sr)
50
50
180
10
90
500
50
17,000
300
a Based on average clay matrix metals concentrations of Bi (12 ppm);
Cr (53 ppm); Mg (2.2) percent); Sr (34 ppm).
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
-------
no significant changes in volatility with increasing levels of
chforine in the feed. Apparent scrubber efficiencies for lead,
cadmium, bismuth, and copper decreased as the chlorine
content increased [2].
Operational
The operating conditions of any incinerator affect metals
behavior. An important variable affecting metal volatilization is
PCC temperature. The lower the PCC temperature, the less
likely it is that the metal will volatilize. Although increasing
combustion air in the PCC increases vaporization, the impact
of this variable is small compared to bed temperature and
chlorine content of the feed stream [6]. Other potential im-
pacts on partitioning and scrubber efficiency include SCC tem-
perature and the stoichiometric ratio of air to feed in the
combustion chamber. The feed rate and percent excess oxy-
gen determine the stoichiometry in the combustion chamber.
While no test data were reviewed that specifically studied the
effect of changes in the feed rate, it can be expected that
excessive feed rates would result in higher mass flowrates enter-
ing the APCD. Assuming APCD efficiency remains relatively
constant, increases in metal feedrates can be expected to cause
increases in stack emissions.
The temperature in the PCC has an effect on the partition-
ing of some metals. Generally, as the PCC temperature in-
creases, the amount of metals that partition into PCC ash will
decrease. Treatability studies in which PCC temperature effects
Test Number:
Table 10. Normalized Metal Discharge Distributions for IRF Tests [2]
(% of Total Measured)
8
Test Average:
Primary Variable: Feed Chlorine Content (wt %)
0 3.8 4.6
Target Temperatures: PCC = 871°C: SCC = 1093°C
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
8.3
Arsenic
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Barium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Bismuth
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Cadmium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Chromium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Copper
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Lead
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Magnesium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Strontium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
93.9
1.7-2.2
3.9
68.8
2.0
28.8
64.8
15.7
19.5
<29.3
42-54
29-46
95.7
1.4
2.8
97.6
0.8
1.6
83.7
11.6
4.7
99.6
0.03
0.36
91.8
2.5
5.7
86.1
3.8-5.8
8.2
79.6
2.2
18.2
22.2
41.1
36.7
<10.3
56-61
34-39
94.1
2.0
3.9
75.8
15.1
9.1
15.0
48.9
36.1
99.3
0.1
0.6
93.0
1.7
5.3
92.3
2.3-4.1
3.6
69.9
5.5
24.7
30.0
35.2
34.7
<12.9
42-45
45-55
85.9
1.9
12.2
76.2
17.8
5.9
13.7
50.2
36.0
99.3
0.2
0.5
89.8
3.5
6.6
92.4
4.0-4.8
2.7
78.6
2.4
19.0
36.3
38.4
25.4
<9.3
68-74
68-74
92.1
2.8
5.1
58.0
33.2
8.8
6.0
73.6
20.3
99.4 .
0.1
0.5
90.7
1.6
7.7
.
-------
were evaluated include studies of soil from the Baird and McGuire
Superfund site and studies of a synthetic soil material.
Baird & McGuire—
Once a chemical manufacturing plant, the Baird and
McGuire site was placed on the National Priorities List (NPL) by
EPA Region I as a Superfund site. Pilot-scale incineration tests
were conducted during late September through early October
1989 in the rotary kiln incineration system at EPA's IRF [5].
The primary contaminants of concern in the soil and sedi-
ments at the site were volatile organic compounds, polynuclear
aromatic hydrocarbons, dioxins, pesticides, lead, and arsenic.
For the IRF testing, soil was excavated and mixed at the site,
then packaged In 55-gallon drums. One drum was used In each
of the first four runs. The fifth run was a composite of soil from
the four drums, The characterization of the feed soil for each
day of testing Is listed In Table 11. Table 11 shows arsenic levels
In the feed soil ranged from 81 ppm to 93 ppm while lead levels
were from 16 ppm to 27 ppm. One objective of this series
of tests was to evaluate the effects of incinerator operating
conditions on lead and arsenic distributions in the waste streams.
Table 12 lists the operating conditions for the five tests
conducted [5].
The normalized mass fractions of metals at Baird and
McGuire are shown in Table 13. For the high PCC temperature
tests (runs 3 through 5) the mass fractions of lead and arsenic in
the PCC ash were significantly reduced [5].
During runs 3 and 4 at Baird and McGuire the percent
oxygen was varied. The change in PCC percent oxygen from
10.4 to 7.5 percent has no significant effect on lead, but
arsenic mass fractions in the kiln ash increased from 36 to
56 percent [5].
Synthetic Materials Tests at IRF—
During the same research project at the IRF mentioned
earlier, tests runs were conducted to investigate the effect of
PCC (kiln) temperature on the fate of metals. Data on these
tests are presented in Tables 9 and 14 [2],
Minor effects due to variations in PCC temperature were
observed for arsenic and lead,1 For these two metals, the kiln
ash distributions decreased slightly as the PCC temperature
increased. PCC temperature effects on the other metals were
not significant [2].
Table 11. Baird & McGuire Soil Characteristics [5]
Test Run No.
Test Date
Ash,wt%
Chlorine
1
9/26/89
83
<0.19
2
9/29/89
83
<0.28
3
9/27/89
83
<0.23
4
9/28/89
84
<0.18
5
10/5/89
a
a
Metals (ppm):
Arsenic
Lead
82
21
83
16
93
27
a Soil feed for Test 5 consisted of soils from the other 4 tests.
Table 12. Baird & McGuire Operating Conditions [5]
8 Average exit temperature.
b Average exit 02, dry basis.
c Approximate.
'Cadmium not detected in ash; difficult to determine effect of PCC or SCC temperature.
81
17
84
20
Test Run No.
Test Date
Soil Feedrate (kg/hr)
PCC Temperature (°G)a
PCC Excess 02(%)b
PCC Solids Residence Time (Min.)
SCC Temperature ("C)8
SCC Excess Os,(%)b
SCC Gas Residence Time (Sec.)
1
9/26/89
50
832
11.3
30°
1,094
7.9
1.8
2
9/29/89
50
844
6.8
30°
1,089
6.3
2.7
3
9/27/89
56
994
10.4
30°
1,105
7.4
2.0
4
9/28/89
54
994
7.5
30°
1,099
7.3
1.9
5
10/5/89
60
839
11.2
30C
1,083
8.1
2.0
8
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
-------
Table 13. Normalized Mass Fractions of Metals for Baird & McGuire [5]
Test Run No.
Test Date
Arsenic
PCCAsh
Scrubber Solids
Stack3
Total
Lead
PCCAsh
Scrubber Solids
Stack3
Total
1
9/26/89
72
23
5
100
89
4
7
100
2
9/29/89
76
22
2
100
93
3
4
100
3
9/27/89
36
55
9
100
69
12
19
100
4
9/28/89
56
38
6
100
69
13
18
100
5
10/5/89
66
29
5
100
91
3
6
100
Method measured particulate and vapor-phase metals at stack.
The SCC (afterburner) temperature was also varied during
the IRF tests. An effect of SCC temperature on partitioning was
not measurable [2].
Combustion Chamber Design
The design of the combustion chamber can affect the
partitioning and APCD efficiency. While such parameters as
PCC or SCC temperature can be controlled with any type of
incinerator, different types of combustion chambers have in-
herently different mixing capabilities and gas flowrates. Turbu-
lent mixing and high air velocities can lead to entrainment of
particles to the APCD rather than the kiln ash.
Mixing is generally greatest in fluidized beds or rotary kilns
and least in infrared incinerators. Air velocities are greatest in
fluidized beds and least in rotary kilns and infrared incinerators.
An example of how the incinerator's design possibly affected
the partitioning of metals is the McColl Superfund Site treatability
study.
McCoff—
The EPA selected contaminated materials from the McColl
Superfund Site in Fullerton, California, as feed for a treatability
study of circulating bed combustor (CBC) technology. The
McColl soil treated had high sulfur content and elevated levels
of barium, chromium, and nickel. EPA selected several drums
that were representative of the soil to be used for the treatabil-
ity tests. Soil from McColl was sent to the Ogden research
facility in San Diego, California. The incineration system con-
sists of a CBC that uses high-velocity air to entrain solids in a
turbulent combustion zone. Soil characteristics, operating con-
ditions, and the normalized mass fractions for the McColl
treatability study are given in Tables 15 through 17. It can be
seen from Table 17 that most of the metals partitioned to the
flyash. Since these metals are predicted to be refractory, it can
be speculated that the partitioning was not due to volatil-
ization. The design of the CBC allows the metals to pass
through the chamber as discrete particles rather than metal
fumes. Because of their fine size, the metal particulates are not
captured in the cyclone, but are transferred to the APCD [8].
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
-------
Table 14. Normalized Metal Discharge Distributions for IRF Tests [2]
(% of Total Measured)
Test Number:
Primary Variable:
Test Average:
Targets:
Arsenic
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Barium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Bismuth
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Cadmium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Chromium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Copper
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Lead
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Magnesium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
Strontium
Kiln Ash
Scrub. Ex. Gas
Scrub. Water
2
4
8
3
6
PCC Exit Temperature (°C)
825
875
876
927
984
SCC = 1093°C;CI = 4%
94.4
2.1-2.9
2.7
74.3
3.8
21.9
25.8
41.5
32.6
<15.2
43-49
42-51
94.7
3.0
2.3
84.2
12.9
3.0
12.6
50.4
37.0
99.4
0.2
0.4
82.9
1.1
16.0
86.1
3.8-5.8
8.2
79.6
2.2
18.2
22.2
41.1
36.7
<10.3
56-61
34-39
94.1
2.0
3.9
75;8
15.1
9.1
15.0
48.9
36.1
99.3
0.1
0.6
93.0
1.7
5.3
92.3
2.3-4.1
3.6
69.9
5.5
24.7
30.0
35.2
34.7
<12.9
42-45
45-55
82.9
1.9
12.2
76.2
17.8
5.9
13.7
50.2
36.0
99.3
0.2
0.5
89.3
3.5
6.6
84.0
6.8-8.4
7.6
69.9
1.6
28.8
22.9
50.7
26.3
<10.7
62-69
27-31
95.3
2.1
2.6
82.3
14.1
3.6
10.4
67.2
22.4
99.5
0.1
0.4
90.1
1.6
8.3
93.6
2.6-3.8
2.6
85.2
2.2
12.5
20.9
47.4
31.6
<13.9
61-69
25-31
95.5
1.1
3.4
79.2
15.2
5.6
5.8
60.6
33.6
99.3
0.1
0.6
94.3
1.3
4.4
4
8
5
SCO Temperature (°C)
1088
PCC = 871
86.1
3.8-5.8
8.2
79.6
2.2
18.2
22.2
41.4
36.7
<10.3
56-61
34-39
94.1
2.0
3.9
75.8
15.1
9.1
15.0
48.9
36.1
99.3
0.1
0.6
93.0
1.7
5.3
1093
°C;CI-4%
92.3
2.3-4.1
3.6
69.9
5.5
24.7
30.0
35.2
34.7
<12.9
42-45
45-55
85.9
1.9
12.2
76.2
17.8
5.9
13.7
50.2
36.0
99.3
0.2
0.5
89.8
3.5
6.6
1196
91.2
3.0-4.3
4.6
86.9
1.6
11.5
30.1
37.1
32.8
<14.5
55-62
31-38
89.3
4.2
6.6
75.1
16.0
8.9
13.8
45.0
41.1
99.2
0.1
0.7
81.9
3.7
14.4
Considerations for Evaluating the Impact of Metals Partitioning
10 During the Incineration of Contaminated Soils from Superfund Sites
-------
Table 15. McColl Treatability Study Soil Characteristics [8]
Test Run No.
1
Test Date
Ash, wt %
Chlorine, wt %
Heating Value, BTU/lb
Metals (ppm)
Barium
Beryllium
Chromium
Nickel
Silver
3/29/89
75.8
ND
986
65
0.21
45
15
ND
3/30/89
71.1
0.02
1430
126
ND
61
17
1.1
3/30/89
70.9
0.08
1344
136
0.7
65
19
ND
ND - Not detected above the quantitation limit.
Air Pollution Control Devices
The type of APCD, or train of devices, that must be em-
ployed for incineration of a contaminated soil will depend on
the-type of incinerator and the characteristics of the soil inciner-
ated or produced during incineration. Incinerators designed to
treat soils with a high ash or toxic metals content will generally
be equipped with an APC train consisting of two to four APCDs.
A number of typical APC trains are as follows:
• Quench/wet scrubber
• Quench/spray dryer/cyclone/electrostatic precipitator
• Quench/spray dryer/cyclone/fabric filter
• Quench/wet scrubber/ionizing wet scrubber/mist
eliminator
• Quench/wet electrostatic precipitator/venturi scrubber/
packed scrubbers
• Fabric filter/wet scrubber.
Table 16. McColl Treatability Study Operating Conditions [8]
Test Run No.
1
Test Date
Soil Feedrate (kg/hr)
PCC Temperature (° C)a
PCC Excess 02 (%)b
PCC Solids Residence Time of Soil (Win.)
Chlorides in feed (ppm of feed)
3/29/89
147.9"
938
11.0
30°
3/30/89
77.5
941
9.9
30°
170
3/30/89
91.1
932
11.8
30°
780
a Mid PCC temperature.
b Exit 02, dry basis, measured downstream of the flue gas cooler.
0 Approximate values.
d Includes 74 kg/hr of sand.
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
77
-------
Table 17. Normalized Mass Fraction of Metals for McColl Treatability Study [8]
Test Run No.
Test Date
Barium
PCCAsh
Fly Ash
Stack Particulate
Total1
Chromium
PCCAsh
Fly Ash
Stack Particulate
Total*
Nickel
PCCAsh
Fly Ash
Stack Particulate
Total1
1
3/29/89
16
84
0
100
15
85
0
100
28
72
0
100
2
3/30/89
6
92
3
100
3
95
2
100
5
90
5
100
3
3/30/89
3
97
0
100
4
95
1
100
ND
96
4
100
ND - Not detected at levels greater than quantitation limit.
a Normalized to 100% of emissions accounted for by analysis.
Operational and design variables that can impact the per-
formance of APCDs include temperature, pressure drop, liquor
flow rate, and number of ionizing stages. Most toxic metals, or
their compounds, condense as solids if incinerator combustion
gases are cooled. Because of this, a quench chamber is com-
monly used to cool incineration flue gas by the evaporation of
water injected into the hot gas stream. The efficiency of venturi
scrubbers, especially on submicron particles, increases as the
gas stream pressure drop across the unit increases [10]. Lower-
ing the liquid mass flow rate will negatively impact scrubber
efficiency due to a corresponding decrease in the diffusional
driving force. Multi-stage ionizing units are usually necessary
for high efficiency removal of small particles. Up to four stages
in series have been used in industry [10].
In addition, APCDs will have different efficiencies depend-
ing on the specific metals being treated and on the amount of
chlorine present. In Table 18, various APCDs are given conser-
vatively estimated efficiencies for particulates and toxic metals.
It should be stressed that these estimates are intentionally
biased low and that higher efficiencies than those shown in the
table are achievable [10].
Material Balance Closure Issues
Complete mass balance closure for metals in combustion
systems is rarely achieved. The limitations to closure include
metals accumulation within the incineration system, difficulty
in obtaining representative samples of heterogeneous streams,
and potential limitations of analytical methods.
Steps can be taken to maximize the closure. The sampling
methods used to determine metals concentrations from a par-
ticular inlet or outlet stream should be appropriate. For deter-
mination of multiple metals emissions in exhaust gases, it is
recommended that EPA draft Method 0012 be used [12]. All
metal-containing streams entering or exiting the incinerator
should be analyzed for the metals of concern, using totals
expressed as elemental metal. Care should be taken to obtain
representative samples. Implementation of operational and/or
design changes to minimize accumulation of feed or waste
residual within the incineration system would also be beneficial.
12
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
-------
Table 18. Air Pollution Control Devices (APCDs) and Their Conservatively Estimated Efficiencies for Controlling Toxic Metals J10J
APCD
ws*
VS-20*
VS-60*
ESP-1
ESP-2
ESP-4
WESP*
FF*
PS*
SD/FF;C/FF
DS/FF
FF/WS*
ESP-1/WS; ESP-1/PS
ESP-4/WS; ESP-4/PS
VS-20/WS*
WS/IWS**
WESP/VS-20/IWS*
C/DS/ESP/FF;C/DS/ESP/FF
SD/C/ESP-A
Pollutant
Ba, Be
50
90
98
95
97
99
97
95
95
99
98
95
96
99
97
95
99
99
99
Ag
50
90
98
95
97
99
97
95
95
99
98
95
96
99
97
95
99
99
99
Cr
50
90
98
95
97
99
96
95
95
99
98
95
96
99
97
95
98
99
98
As, Sb, Cd,
Pb.TI
40
20
40
80
85
90
95
90
95
95
98
90
90
95
96
95
97
99
95
Hg
30
20
40
0
0
0
60
50
80
90
50
50
80
85
80
85
90
98
85
* It is assumed that flue gases have been pre-cooled in a quench. If gases are not cooled adequately, mercury recoveries will diminish, as will cadmium
and arsenic to a lesser extent.
** An IWS is nearly always used with an upstream quench and packed horizontal scrubber.
C = Cyclone
WS = Wet Scrubber including: Sieve Tray Tower
Packed Tower
Bubble Cap Tower
PS = Proprietary Wet Scrubber Design
(A number of proprietary wet scrubbers have come on the market in recent years that are highly efficient on both
particulates and corrosive gases. Two such units are offered by Calvert Environmental Equipment Co. and by Hydro-Sonic
Systems, Inc.)
VS-20 = Venturi Scrubber, ca. 20-30 in W.G. Ap
VS-60 = Venturi Scrubber, ca. > 60 in. W.G. Ap
ESP-1 = Electrostatic Precipitator; 1 stage
ESP-2 = Electrostatic Precipitator; 2 stages
ESP-4 = Electrostatic Precipitator; 4 stages
IWS = Ionizing Wet Scrubber
DS = Dry Scrubber
FF = Fabric Filter (Baghouse)
SD = Spray Dryer (Wet/Dry Scrubber)
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites 73
-------
Conclusion
Technology Contacts
Contamination of Superfund site soils is rarely limited to
one particular compound. Oftentimes, selection of remedial
options must take into consideration a variety of both organic
and inorganic contaminants. The effectiveness of incineration
for the treatment of hazardous organic compounds has been
well demonstrated; consistently high destruction and removal
efficiencies for such compounds can be expected. In contrast,
the fate of metals subjected to incineration is less certain.
In recent years, a considerable amount of research has
been conducted in an attempt to better understand the behav-
ior of metals during incineration. Theoretical predictions have
been complemented by laboratory and pilot-scale studies of
both real-world and synthetic contaminated soils. Current
information suggests that estimation of metal behavior should
not be over-simplified. It is dependent on a number of factors
that should be considered on a case-by-case basis, among
them: matrix parameters (e.g., type, concentration of metals);
operational parameters (e.g., temperature, oxygen); design
parameters (e.g., mixing, combustion zone velocity); and air
pollution control device parameters (e.g., type, temperature).
Since incineration residuals (e.g., ash, scrubber water) are
collected, they may be further treated following incineration in
order to minimize adverse impacts of remaining metals. In
contrast, stack emissions of metals, which are not collected,
represent a potential risk element and should therefore be
evaluated prior to undertaking remedial action. Risk assess-
ments for metal emissions may be performed by examining
estimated metal emission rates, site-specific dispersion factors,
and health effects data.
As a result of soil-specific/site-specific factors, similar metal
feedrates for two sites may yield very different metal emission
rates. While conservative theoretical assumptions may be made,
treatability studies are suggested as a means by which to
develop a more accurate expectation of metal partitioning
associated with a specific contaminated soil. It should also be
noted that similar metal emission rates for two sites may yield
very different risk estimates as a result of differences in terrain
complexity, stack height, meteorology, and other factors; site-
specific dispersion modeling is suggested in cases where the
potential for significant metal emissions exists.
In summary, incineration of metal-bearing wastes should
be approached with caution from two perspectives. On the
one hand, care should be taken to avoid overlooking potential
risks associated with stack emissions. On the other hand, one
should avoid rejecting incineration from consideration based
on generalized concepts (e.g., "arsenic is volatile") that may
not always prove true.
The following individuals can be contacted with technical
questions concerning treatability studies to evaluate the impact
of metals partitioning during incineration:
Greg Carroll
(513)569-7948
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 W. Martin Luther King Drive
Cincinnati, Ohio 45268
Paul Leonard
(215)597-8485
U.S. Environmental Protection Agency
Region III
841 Chestnut Building
Philadelphia, Pennsylvania 19107
The Incineration Research Facility in Jefferson, Arkansas,
has been involved in three major metal partitioning studies as
well as a number of Superfund treatability studies for the EPA
Regions. For information on the availability of the IRF for both
bench- and pilot-scale treatability studies, the following indi-
vidual can be contacted:
Robert Thurnau
(513)569-7692
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 W. Martin Luther King Drive
Cincinnati, Ohio 45268
Acknowledgements
This issue paper was prepared for the Superfund Engineer-
ing Forum under the direction of Joan Colson, Overall Project
Manager. Considerable help and constructive suggestions were
provided by Greg Carroll (RREL), Paul Leonard (Region III), and
the Remedial Project Manager (RPM) for each of the case
studies. This report was prepared by Science Applications
International Corporation (SAIC) under EPA Contract No. 68-
C8-0062, Work Assignment No. 3-43. SAIC's Work Assignment
Manager for this project was Gary Baker. This paper was
authored by George Wahl of SAIC.
14
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
-------
References
1. Barton, R.G., et al. Fate of Metals in Waste Combustion
Systems. Presented at: First Congress on Toxic By-
Products of Combustion, Los Angeles, CA, August, 1989.
2. Carroll, G. J., et al. The Partitioning of Metals in Rotary
Kiln Incineration. EPA/600/D-89/208. U.S. Environmental
Protection Agency, Cincinnati, Ohio, 1989, p.15.
3. Energy and Environmental Research Corporation.
Engineering Analysis of Field Data. Presented at ASME/
EPA Metals Workshop, Cincinnati, Ohio, November 7,
1991.
4. Energy and Environmental Research Corporation.
Overview of Metals Behavior in Combustion Systems.
Presented at ASME/EPA Metals Workshop, Cincinnati,
Ohio, November 7,1991.
5. King, C. and LR. Waterland. Pilot-Scale, Incineration of
Arsenic-Contaminated Soil From Baird and McGuire
Superfund Site. EPA Contract No. 68-C9-0038, Work
Assignment 0-5. U.S. Environmental Protection Agency,
Cincinnati, Ohio 1990.
6. Lee, C.C. A Model Analysis of Partitioning in a Hazardous
Waste Incineration System. APCA, 38(7): 941 -945,1988.
7. Rosenthal, S. Shirco Infrared Incineration System:
Applications Analysis Report. EPA/540/A5-89/010. U.S.
Environmental Protection Agency, Cincinnati, Ohio, 1989,
pp. 65-82.
8. Sudell, G.W. Treatability Study Report: Results of
Treating McColl Superfund Waste in Ogden's Circulating
Bed Combustor Research Facility Draft. EPA-600/X-89-
342. U.S. Environmental Protection Agency, Cincinnati,
Ohio, 1989.
9. U.S. EPA. Experience in Incineration Applicable to
Superfund Site Remediation. EPA/625/9-88/008. U.S.
Environmental Protection Agency, Cincinnati, Ohio,
1988.
10. U.S. EPA. Guidance on Metals and Hydrogen Chloride
Controls for Hazardous Waste Incinerations: Volume IV of
the Hazardous Waste Incineration Guidance Series (Draft
Final Report). Office of Solid Waste, Waste Treatment
Branch, Washington, D.C., 1989.
11. U.S. EPA. High Temperature Thermal Treatment for
CERCLA Waste: Evaluation and Selection of Onsite and
Offsite Systems. EPA/450/X-88/006. U.S. Environmental
Protection Agency, Office of Solid Waste and Emergency
Response, Washington, D.C., 1988.
12. U.S. EPA." Proposed Methods for Stack Emissions
Measurement of CO, O2, THC, HCI, and Metals at
Hazardous Waste Incinerators: Volume VI of the Hazard-
ous Waste Incinerator Guidance Series. Office of Solid
Waste and Emergency Response, Washington, D.C.,
1989.
13. Waterland, L.R. Operations and Research at the U.S. EPA
Incineration Research Facility: Annual Report for FY89.
EPA/600/9-90/012. U.S. Environmental Protection
Agency, Cincinnati, Ohio, 1990.
14. 40 CFR, 261.30 Subpart D - List of Hazardous Wastes.
Considerations for Evaluating the Impact of Metals Partitioning
During the Incineration of Contaminated Soils from Superfund Sites
15
'U.S. Govarnmant Printing Offlea: 199S—643-OSO/SO13B
-------
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
Please make all necessary changes on the below label,
detach or copy, and return to the address In the upper
left-hand comer.
If you do not wish to receive these reports CHECK HERE D;
detach, or copy this cover, and return to the address in the
upper left-hand comer.
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/540/S-92/014
------- |