United States
Environmental Protection
Agency
Hazardous Waste Engineering
Research Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/S2-87/054 Sept. 1987
SERA Project Summary
Distribution of Trace Element
Emissions from the Liquid
Injection Incinerator Combustion
Research Facility
Johannes W. Lee, Robert W. Ross, II, Ralph H. Vocque, Jerry W. Lewis, and
Larry R. Waterland
The EPA is currently developing
regulations on trace element emissions
from hazardous waste incineration.
However, the data base on trace ele-
ment emissions from incinerators which
can be used to support regulations is
very sparse. Data on the effects of
waste composition and incinerator
operation on trace element emissions
are particularly lacking. In response to
these data needs, a series of tests were
conducted at EPA's Combustion Re-
search Facility (CRF) to investigate the
fate of volatile trace elements in liquid
injection hazardous waste incineration.
In these tests, arsenic in the form of
arsenic trioxide (As2O3) and antimony
in the form of antimony trichloride
(SbCI3) were added to a methanol base
containing varying amounts of chloro-
benzene and carbon tetrachloride, and
fired in the liquid injection incinerator
at the CRF. Test variables included
incinerator temperature and excess air
level, and feed chlorine content. Test
results show a relatively even distribu-
tion of both elements between scrubber
exit flue gas and scrubber blowdown.
Both elements are found in the vapor
phase at high temperatures, although
most condenses to particulate at
scrubber exit temperatures. Designated
POHC destruction and removal effici-
ency (ORE) ranged from 99.99 to
99.999 percent at the afterburner exit
to 99.999 to 99.9999 percent in the
scrubber exit flue gas. Typical levels of
common products of incomplete com-
bustion were measured.
This Project Summary was developed
by EPA'* Hazardous Watte Engineering
Research Laboratory, Cincinnati, OH, to
announce key findings of the research
project that Is fully documented In a
separate report of the same title (see
Project Report ordering Information at
back).
Introduction
In support of the EPA Office of Solid
Waste (OSW) hazardous waste incinerator
regulation development, a series of in-
cineration tests to determine the fate of
two Appendix VIII metals, antimony and
arsenic, were conducted at the EPA
Combustion Research Facility (CRF) in
Jefferson, Arkansas. These metal emis-
sion tests are important, because risk
assessments show that metal emissions
from otherwise properly operated incin-
erators can be the largest component of
risk to human health and the environment.
These tests incinerated mixtures of
methanol, chlorobenzene, and carbon
tetrachloride, spiked with SbCI3 and
As203. The objectives of the test program
were:
• To track emissions of arsenic and
antimony through the incinerator
system
• To study the distribution of metals
between flue gas and scrubber blow-
down water
• To explore the relationship between
metal emissions and incinerator
operating conditions and feed
chlorine content
• To evaluate whether the metals
-------
affect waste component destruction
• To extend the data base on volatile
products of incomplete combustion
(PIC) emissions from the incineration
of chlorinated hydrocarbons
Approach
The tests were conducted in the liquid
injection incineration system (LIS) at the
CRF. The LIS can incinerate pumpable
and atomizable liquid wastes.
Figure 1 shows a simplified schematic
of the incinerator system which consists
of two refractory-lined combustion
chambers. The system cleans the com-
bustion product gases with a packed-
column scrubber, and an ionizing wet
scrubber, followed by an activated carbon
bed adsorber and a high efficiency par-
ticulate (HEPA) filter. An induced draft
fan downstream of the HEPA filter draws
a slight vacuum (draft) throughout the
incinerator/scrubber system and vents
the combustion products via a stack.
The research program studied the ef-
fects of three operating parameters: feed
chlorine content, incinerator temperature,
and combustion excess air. A Box-Wilson
factorial experimental design specified
18 test conditions which included 5 levels
of variation for each parameter.
Blends of methanol, carbon tetrachlo-
ride, and chlorobenzene in the feed
produced the various inlet chlorine con-
centrations (0 to 33.8 percent). Arsenic
and antimony were added to these blends
in proportions designed to give constant
feed concentrations of 12 and 40 ppm
respectively for all tests. These concentra-
tions in the feed material ensure that the
worst-case concentrations in the stack
gas will never exceed the threshold limit
values (TLV), of 0.2 mg As/m3 and 0.5
mg Sb/m3. Furthermore, if all the trace
metals leave the incinerator via the
scrubber blowdown water, their concen-
trations will not exceed the EP toxicity
limits (5 mg As/L, no limit established for
Sb).
The liquid feed entered the main com-
bustion chamber via a steam-atomized
nozzle at about 45 kg/hr (100 Ib/hr).
Auxiliary propane maintained the incin-
erator at the specified temperature which
ranged from 1,137° to 1,450°C (2,079°
to 2,642°F). Combustion air entered via
swirl registers in the burner to produce
from 5.9 to 11.7 percent excess oxygen
at the incinerator exit.
Figure 2 illustrates the sampling pro-
tocol. In addition to the continuous
monitors, which measured 02, CO, and
C02, volatile organic sampling trains
(VOST) collected samples at the after-
burner and the scrubber exits. Thermal
desorption purge and trap GC/FID
analyzed for 22 volatile organic com-
pounds.
EPA reference Method 5 was used to
collect samples for particulate load and
arsenic and antimony analyses. The
Method 5 train was modified to ensure
collection of any antimony and arsenic
which passed through the particulate
filter. It consisted of a probe and a glass
cyclone, followed by a filter and five
impingers. The first impinger contained
0.1 N NaOH. The second, third and fourth
impingers contained 0.2 M (NH4)2S208 +
0.02 M AgN03. The last impinger con-
tained silica gel. Following collection and
digestion, furnace atomic adsorption (AA)
methods were used to analyze for arsenic
and antimony.
Results
Of the 18 planned tests, 12 were com-
pleted. The remaining 6 tests were gen-
erally unattainable due to flame stability
problems at low flame temperatures and
low excess air conditions.
Trace Metal Discharges
Good mass balance closure for anti-
mony and arsenic could not be established
during these tests. Table 1 lists the frac-
tions of inlet antimony found at the after-
burner exit, the scrubber exit and the
scrubber blowdown water. As shown in
the table, for the three tests where the
feed contained nominally 26 percent
chlorine, no more than about 1 percent of
the inlet antimony, was measured in the
afterburner exit flue gas. For one test
with the nominally 19 percent chlorine
feed, less than 0.1 percent of the inlet
antimony was detected in the afterburner
exit flue gas. For the remaining tests with
nominally 19 percent chlorine feed and
the tests with nominally 8 percent and 0
percent chlorine feed, between 25 and
Carbon Bed
Filter
Building Wall
Afterburner Chamber
(Unfired)
Sampling Port
Quench Chamber
Main Chamber
Aux Propane
Liquid Feed
Atomizing Steam
Combustion Air
Venturi Scrubber /
Scrubber Liquid Tank
Figure 1. Simplified schematic of the EPA liquid injection incinerator system
2
-------
Primary
Combustor
Secondary \
Combustor J~
H Packed Tower
Scrubber
Ionizing ^
Wet Scrubber I F
1 Carbon
Bed
Sampling
Point
1
2
3
4
5
Parameter
Waste
Feedrate
X
Waste
Feed
X
Auxiliary
Fuel
Feedrate
X
Air
Feedrate
X
CM's
Scrubber I (Oz, CO,
Slowdown CO2, NO,)
<- - \ - j
X
X
X
Volume
Flow
X
X
X
X
M5
(Paniculate, •
Metals)
X
X
X"
VOST
(Volatile
Organics)
X
X
Temperature
X
X
X
Pressure
X
X
X
Relative
Humidity
X
X
X
"Paniculate only.
Figure 2. Sampling protocol
101 percent of the input antimony was
measured in the afterburner exit flue
gas.
Between 15 and 84 percent of the inlet
antimony was measured in the scrubber
discharge streams (flue gas plus blow-
down). Scrubber blowdown water ac-
counted for 5 to 75 percent of the inlet
antimony. Scrubber exit flue gas ac-
counted for between 3 and 21 percent of
the inlet antimony. Comparable fractions
of antimony were generally measured in
these two streams for a given test.
Mass balance closure in the afterburner
Table 1. Antimony Discharge Distributions
exit flue gas was quite poor for four tests.
However, for 4 of 5 tests at the nominally
19 percent chlorine feed, two at nominally
8 percent chlorine feed, the one with no,
and the one with about 34 percent chlo-
rine feed, mass balance closure was
within a factor of 3. This is in the range of
acceptable mass balance closure based
on past experience. Closure results are
generally better at the scrubber discharge.
Still, closure within a factor of 3 was
obtained for only 7 tests.
The inability to attain better mass
balance closure in the tests was likely
Antimony discharge distribution3
(percent of feedrate)
Test
No.
14
4
5
7
13
18
16
2
3
6
9
15
Feed Cl
content,
percent
33.8
26.4
27.1
26.4
18.8
18.8
18.8
18.8
18.8
7.6
7.9
0.
Primary
combustor
temperature,
°C(°F)
1.293 (2.359)
1,307 (2,384)
1.137(2,079)
1.399(2,551)
1.203 (2, 198)
1,176(2,148)
1.450 (2.642)
1,336 (2,437)
1.228 (2.243)
1,290 (2.354)
1.425(2.597)
1.265(2.310)
Afterburner Afterburner
exit 02, exit
percent flue gas
9.2
10.5
10.8
6.1
11.8
10.7
7.4
9.2
8.4
10.1
5.9
8.1
34 (52)
0.1 S/3
0.3 (0.24)
1. 1 (1.8)
0 (0. 1 1)
32 (53)
83 (146)
79 (104)
89 (99)
56 (47)
101 (83)
25 (38)
Scrubber
discharge streams
Flue gas
8(12)
14"
17(13)
18 (30)
3(6.6)
9(14)
21 (36)
15 (20)
10(11)
20(17)
17(14)
9.4(14)
Blowdown
7(11)
22*
21 (16)
8(12)
40 (87)
75(124)
5(8.3)
9(12)
13 (15)
10(18)
18(15)
7(11)
Total
15(23)
36"
38 (29)
26 (42)
43 (94)
84 (138)
26 (44)
24 (32)
23 (26)
30 (35)
35 (29)
16(25)
" Numbers in parenthesis are based on a feed composition calculated from feed constituent
blending proportions
No feed sample analysis, value based on a feed composition calculated from feed constituent
blending proportions.
affected by the fact that accumulation of
element concentrations in recirculating
scrubber liquor was not taken into
account.
Table 2 shows the distribution of anti-
mony between the particulate and vapor
phases in the flue gas at both the after-
burner exit and the scrubber exit. The
table shows that, at the afterburner exit
where temperatures were above 760°C
(1,400°F), the vapor phase (impinger
samples) contained an average of 55
percent (range 5 to 90) of the collected
antimony. At the scrubber exit where the
temperature was generally at 74°C
(165°F), the vapor phase accounted for
an average of 6 percent (range 0 to 38) of
the collected antimony.
Tables 3 and 4 list the arsenic mass
balance and phase distribution data. In
the afterburner exit flue gas, detectable
levels of arsenic were measured for 9
tests. The measured fractions of feed
arsenic ranged from 1.4 to 245 percent.
Eight tests showed better than 10 percent
arsenic recovery. Seven of these cor-
responded to mass balance closure within
a factor of 3.
The combination of the scrubber exit
gas with the scrubber discharge blow-
down accounted for 3.8 to 299 percent of
the input arsenic. Mass balance closure
around the scrubber was within a factor
of 3 for eight tests. More often than not, a
greater proportion of arsenic was found
in the flue gas than in the scrubber
blowdown.
At the afterburner exit, arsenic is dis-
tributed between the vapor and particulate
-------
Table 2. Antimony Flue Gas Distributions
Antimony distribution between
paniculate and vapor phase in the
flue gas (percent)
T.«o*
lest
No.
14
4
5
7
13
18
16
2
3
6
9
15
Table 3.
Feed Cl
content.
percent
33.8
26.4
27.1
26.4
18.8
18.8
18.8
18.8
18.8
7.6
7.9
0.
Arsenic
Primary
combustor
temperature.
°C(°F)
1,293 (2,359)
1,307 (2.384)
1, 137 (2,079)
1,399 (2,551)
1,203 (2. 198)
1.176(2.148)
1,450(2.642)
1,336 (2,437)
1,228(2.243)
1,290 (2,354)
1. 425 (2,597)
1,265(2.310)
Afterburner Afterburner exit
exit 02,
percent
9.2
10.5
10.8
6.1
11.8
10.7
7.4
9.2
8.4
10.1
5.9
8.1
Paniculate
32
62
49
53
42
21
76
10
22
41
34
95
Vapor
68
38
51
47
58
79
24
90
78
59
66
5
Scrubber discharge
Paniculate
99
99
62
99
99
100
99.9
82
97
100
98
99
Vapor
1
1
38
1
1
0
0.1
18
3
0
2
1
Discharge Distributions
Arsenic discharge distribution11
(percent of feedrate)
Test
No.
14
4
5
7
13
18
16
2
3
6
9
15
Feed Cl
content,
percent
33.8
26.4
27.1
26.4
18.8
18.8
18.8
18.8
18.8
7.6
7.9
0.
Primary
combustor
temperature,
°C (°F)
1,293 (2.359)
1.307(2.384)
1.137(2,079)
1.399 (2.551)
1.203(2.198)
1.176(2,148)
1.450(2,642)
1.336 (2,437)
1,228 (2,243)
1,290 (2.354)
1.425 (2.597)
1.265 (2.310)
Afterburner Afterburnei
exit O2, exit
percent flue gas
9.2
10.5
10.8
6.1
11.8
10.7
7.4
9.2
8.4
10.1
5.9
8.1
126(26)
NO
NO
1.4(1.7)
ND
25(4.0)
245 (77)
45 (10 2)
77(20)
125(32)
153(177)
103(21)
Scrubber
discharge streams
Flue gas
30 (6. 1)
3.06
15 (5. 1)
49 (58)
ND
13 (2.2)
30 (9.4)
19 (4.3)
289 (73)
60 (15)
26 (20)
5(1.1)
Slowdown
2.5(0.5)
3.8"
6.9 (2.3)
5.4 (6.4)
3.8 (2.8)
19 (3. 1)
6.0(1.9)
2.3 (0.51)
9.8 (2.5)
21 (5.3)
20(15)
32 (6.6)
Total
33 (6 6)
J2"
22(7.4)
54 (64)
3.8 (2.8)
32 (5.3)
36(11)
21 (4.8)
299 (76)
81 (20)
46 (35)
37(7.7)
ND — Not detected.
" Numbers in parenthesis are based on a feed composition calculated from feed constituent
blending proportions.
b Feed sample not analyzed; value based on feed composition calculated from feed constituent
blending proportion.
phases. More often than not, a greater
proportion was found in the participate.
At the scrubber discharge, usually all of
the arsenic was measured in the
particulate.
Destruction and Removal
Efficiencies (ORE)
The tests demonstrated acceptable
DREs for both carbon tetrachloride and
chlorobenzene. Chlorobenzene DREs
were higher than those for carbon tetra-
chloride. Carbon tetrachloride DREs were
greater than 99.99 percent at the after-
burner exit. These increased to greater
than 99.999 percent at the scrubber exit.
Similarly, chlorobenzene DREs ranged
from 99.999 percent at the afterburner
exit to generally better than 99.999 per-
cent at the scrubber exit. The available
data do not suggest any discernible effect
of temperature, feed composition, excess
air, or the presence of arsenic or antimony
on DRE. This is in general agreement
with previous CRF data which has con-
sistently shown high in DREs for con-
centrated feed materials.
Products of Incomplete
Combustion
Incineration produces low levels of
PICs. The CRF routinely analyzes VOST
samples for 22 organic compounds with
GC/FID. For these tests at the afterburner
exit, of these 22 compounds, carbon
tetrachloride (POHC) was present at the
highest levels (>100 /ug/dscm). Other
chlorinated alkanes and alkenes were
present in the 10 to 100 /ug/dscm range.
Chlorobenzene (POHC) concentrations
were similar to those of the common
chlorinated PICs.
At the scrubber exit, PIC concentrations
are about one-tenth of those at the
scrubber inlet, i.e., afterburner exit.
Carbon tetrachloride (a POHC) levels
ranged from 4.4 to 31 ug/dscm. The
highest concentrations occurred at low
excess air conditions. Except for one test,
chlorobenzene (the other POHC), levels
ranged from 1.4 to 14 /ug/dscm and did
not appear to correlate with temperature
or excess air. Other PICs include methy-
lene chloride which was present in all
tests at relatively high levels (11 to 139
ug/dscm). Other chlorinated alkanes and
alkenes, hexane, benzene, and toluene
were present at levels generally below
10/ig/dscm.
All the data reported above for methy-
lene chloride, 1,1 -dichloroethylene, 1,1,1-
trichloroethane, carbon tetrachloride, tri-
chloroethylene, benzene, hexane, and
toluene are somewhat compromised due
to laboratory solvent contamination.
However, since the reported results are
quite consistent with past experience,
they are reported here.
Conclusions
Major conclusions from the test results
addressing each of the test objectives
noted in the Introduction are:
• Antimony was relatively evenly dis-
tributed between the particulate and
vapor phases in flue gas samples
taken at the afterburner exit. The
same was generally true for arsenic,
although a greater fraction was found
in the particulate phase for the
greater number of tests. Most, if not
all of the antimony and arsenic was
found in the particulate catch of the
sampling trains run at the scrubber
exit.
-------
The success in achieving mass
balance closure in the flue gas at
the afterburner exit, and in the
scrubber exit flue gas and scrubber
blowdown discharge was in keeping
with past experience. Mass balance
closure was within a factor of 3 for 6
to 8 of the 12 tests performed for
each element/location (afterburner
exit or scrubber discharge) combina-
tion. Four of the 12 tests had very
low (less than 2 percent) both anti-
mony and arsenic recovery in the
afterburner exit flue gas. The lowest
recovery in scrubber discharges (flue
gas plus scrubber blowdown) were
better; 15 percent for antimony and
4 percent for arsenic. However, no
account was taken for accumulation
of the elements in the recirculating
scrubber liquor. This may have af-
fected the ability to achieve better
mass balance closure.
Comparable fractions of the input
antimony were accounted for in the
scrubber exit flue gas and the
scrubber blowdown. The same was
largely true for arsenic, although a
greater fraction of arsenic was found
in the scrubber exit flue gas in a
greater number of tests.
No clear dependence of the distribu-
tion of arsenic and antimony dis-
charges with the primary test vari-
ables was apparent. However, a
statistical evaluation of the data
obtained is planned in future efforts.
This analysis will establish whether
statistically significant relationships
exist.
Carbon tetrachloride ORE was
greater than 99.99 percent as mea-
sured at the afterburner exit; this
increased to greater than 99.999
percent as measured at the scrubber
exit. Chlorobenzene ORE was
greater than 99.999 at the after-
burner exit, increasing to generally
greater than 99.9999 percent at the
scrubber exit No clear dependence
of DRE on incinerator operation was
apparent. This DRE performance is
consistent with past experience at
the CRF, so the presence of the
elements fed had no apparent effect
onPOHCDRE.
Several commonly measured chlori-
nated C, and C2 hydrocarbons,
benzene, and toluene were present
in the afterburner exit flue gas at
levels in the 10 to serveral hundred
//g/dscm range. These same com-
pounds were generally present in
Tab/tf 4. Arsenic Flue Gas Distributions
Arsenic distribution between
paniculate and vapor phase in the
flue gas (percent)
Feed C1
Test content.
No. percent
14
4
5
7
13
18
16
2
3
6
9
15
ND —
33.8
26.4
27.1
26.4
18.8
18.8
18.8
18.8
18.8
7.6
7.9
0
Not detected.
Primary
combustor
temperature,
°C (°F)
1,293 (2,359)
1.307 (2,384)
1,137(2.079)
1.399 12,551)
1,203(2,198)
1,176(2,148)
1,450(2,642)
1,336 (2.437)
1.228 (2.243)
1.290 (2.354)
1.425 (2.597)
1.265(2.310)
Afterburne
exit 02,
percent
9.2
10.5
10.8
6.1
11.8
10.7
7.4
9.2
8.4
10.1
5.9
8.1
r Afterburner exit
Paniculate
70
ND
ND
43
ND
32
76
83
41
71
69
98
Vapor
30
ND
ND
57
ND
68
24
17
59
29
31
2
Scrubber discharge
Paniculate
85
100
83
100
ND
100
10O
94
100
100
100
100
Vapor
15
0
17
0
ND
0
0
6
0
0
0
0
the scrubber exit flue gas at levels
about a factor of 10 lower.
Data quality assurance objectives for
measurement precision and accuracy
were not fully met for many measurement
parameters. However, failure to attain
these had little impact on the above-
stated conclusions.
Johannes W. Lee, Robert I/I/. Ross, II, Ralph H. Vocque, Jerry W. Lewis, and
Larry R. Water/and are with A cur ex Corporation, Jefferson, AR 72079.
Robert E. Mournighan is the EPA Project Officer (see below).
The complete report, entitled "Distribution of Trace Element Emissions from
the Liquid Injection Incinerator Combustion Research Facility." (Order No.
PB 87-224 689/AS; Cost: $13.95, subject to change) will be available only
from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Officer can be contacted at:
Hazardous Waste Engineering 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 P
EPA
PERMIT No G-3E
Official Business
Penalty for Private Use $300
EPA/600/S2-87/054
0001961
LIBRARy REGION V
r_
CHICAGO
ST
60604
------- |