The Fate of Trace Metals in a Rotary Kiln Incinerator With a Single-stage Ionizing Wet Scrubber : Volume 1. Technical Results


EPA/600/2-90/043a
September 1990
THE FATE OF TRACE METALS IN A
ROTARY KILN INCINERATOR WITH
A VENTURI/PACKED COLUMN SCRUBBER
Volume I —Technical Results
By
D. J. Fournier, Jr., W. E. Whitwonh, Jr., J. W. Lee, and L. R. Waterland
Acurex Corporation
Environmental Systems Division
Incineration Research Facility
Jefferson, Arkansas 72079
EPA Contract No. 68-03-3267
Work Assignments 2-1 and 3-1
EPA Project Officer: Robert C. Thumau
Technical Task Manager: Gregory J. Carroll
Waste Minimization, Destruction and Disposal Research Division
Risk Reduction Engineering Laboratory
Cincinnati, Ohio 45268
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268

-------
I
DISCLAIMER
The information in this document has been funded by the U.S. Environmental
Protection Agency under Contract No. 68-03-3267. It has been subjected to the Agency's peer
review and administrative review, and it has been approved for publication as an EPA document.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
ii

-------
Today's rapidly developing and changing technologies and industrial products and
practices frequently carry with them the increased generation of materials that, if improperly
dealt with, can threaten both public health and the environment. The U.S. Environmental
Protection Agency is charged by Congress with protecting the Nation's land, air, and water
resources. Under a mandate of national environmental laws, the agency strives to formulate and
implement actions leading to a compatible balance between human activities and the ability of
natural systems to support and nurture life. These laws direct the EPA to perform research to
define our environmental problems, measure the impacts, and search for solutions.
The Risk Reduction Engineering Laboratory is responsible for planning, implementing,
and managing research, development, and demonstration programs to provide an authoritative,
defensible, engineering basis in support of the policies, programs, and regulations of the EPA
with respect to drinking water, wastewater, pesticides, toxic substances, solid and hazardous
wastes, and Superfund-related activities. This publication is one of the products of that research
and provides a vital communication link between the researcher and the user community.
This report describes an extensive series of tests conducted at the EPA's Incineration
Research Facility to evaluate the fate of trace metals fed to a rotary kiln incinerator. For
further information, please contact the Waste Minimization, Destruction and Disposal Research
Division of the Risk Reduction Engineering Laboratory.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory

-------
ABSTRACT
A 5-week series of pilot-scale incineration tests, employing a synthetic waste feed, was
performed at the U.S. Environmental Protection Agency's (EPA) Incineration Research Facility
(1RF) to evaluate the fate of trace metals fed to a rotary kiln incinerator equipped with a venturi
scrubber/packed column scrubber. Completed were three tests to evaluate the valence state of
chromium in emissions and discharges as a function of valence state in the feed and feed
chlorine content, and eight tests to evaluate the fate of five hazardous constituent and four
nonhazardous constituent trace metals as a function of incinerator operating temperatures and
feed chlorine content.
Chromium test results indicated that when no chlorine was present in the feed, 95
percent of the measured chromium was discharged in the kiln ash, 1 to 2 percent in the scrubber
exit flue gas, and 3 percent in the scrubber liquor. With chlorine in the feed, these fractions
were 84, 4, and 11 percent, respectively. Kiln ash contained negligible Cr( + 6) for all tests.
Scrubber exit flue gas Cr( + 6) as a fraction of total chromium was nominally 15 percent with no
feed chlorine, increasing to about 50 percent with chlorine-containing feed. The scrubber liquor
Cr( + 6) fraction was 20 to 30 percent with Cr(+3) feed, increasing to about 60 percent with
Cr( + 6) feed.
Parametric trace metal test results confirm that cadmium, lead, and bismuth are
relatively volatile metals, based on normalized discharge distribution data. Less than 32 percent
of these metals were discharged in the kiln ash. Barium, copper, strontium, chromium, and
magnesium are relatively nonvolatile, with greater than 75 percent of their discharge amounts
present in kiln ash. Surprisingly, arsenic was found to be relatively nonvolatile, Apparent
scrubber collection efficiencies generally correlated with observed volatility. Volatile metals
exhibited collection efficiencies of 36 to 45 percent; nonvolatile metals, with the exception of
copper, exhibited collection efficiencies of 49 to 88 percent. Feed chlorine content had a major
effect on the observed volatility of cadmium, lead, and bismuth, with volatility increasing with
increased feed chlorine content. Afterburner exit flue gas particle size distributions shifted to
smaller size when chlorine was added to the feed.
The average mass balance closures around the kiln ash/scrubber discharge ranged from
48 to 96 percent for individual metals. Overall average closure was 71 percent. From past
experience, trace metals mass balance closure results for combustion sources are in the 30 to
200 percent range.
This report was submitted in fulfillment of Contract Number 68-03-3267 by Acurex
Corporation under the sponsorship of the U.S. Environmental Protection Agency. This report
covers work conducted during August and September, 1988.
iv

-------
CONTENTS
FOREWORD		iii
ABSTRACT		iv
FIGURES 		vii
TABLES 	 x
1	INTRODUCTION 		1
2	FACILITY DESCRIPTION AND TEST CONDITIONS		3
2.1	Rotary Kiln Incinerator System Description 		3
2.1.1	Incinerator Characteristics		3
2.1.2	Air Pollution Control System		6
2.2	Synthetic Test Mixture		6
2.3	Test Conditions		7
3	SAMPLING AND ANALYSIS PROCEDURES 		21
3.1	Chromium Valence Tests		25
3.1.1	Sampling Procedures		25
3.1.2	Analysis Procedures		28
3.2	Parametric Trace Metal Tests	28
3.2.1	Sampling Procedures		28
3.2.2	Analysis Procedures		34
4	TEST RESULTS 		38
4.1	Synthetic Waste Feed Composition		38
4.2	Continuous Emission Monitoring Data		43
4.3	Flue Gas Particulate and HC1		47
4.4	Trace Metal Discharge Data		47
4.4.1	Chromium Valence State Tests 		50
4.4.2	Parametric Trace Metals Tests		56
4.5	POHC Destruction and Removal Efficiencies		114
4.6	Volatile Products of Incomplete Combustion (PICs)		114
5	CONCLUSIONS 		119
v

-------
CONTENTS
6	QUALITY ASSURANCE 		122
6.1	Organic Analysis of Clay/Organic Liquid Feed Samples 		124
6.2	VolatiJe Organic Compounds in the VOST Traps		124
6.3	Metals Analysis 		133
6.3.1	Chromium Valence State Tests		133
6.3.2	Parametric Trace Metal Tests		135
REFERENCES		141
vi

-------
FIGURES
Number	Page
1	Schematic of the IRF RKS	 4
2	Actual versus target operating temperatures for the parametric
trace metal tests	 20
3	Sampling summary for the chromium valence state and parametric
trace metals test series 	 22
4	Generalized CEM sample gas flow schematic	 24
5a Cadmium discharge distributions for the parametric trace metal tests:
effect of kiln temperature	 73
5b Cadmium discharge distributions for the parametric trace metal tests:
effect of afterburner temperature	 74
5c Cadmium discharge distributions for the parametric trace metal tests:
effect of feed chlorine content	 75
6a Lead discharge distributions for the parametric trace metal tests:
effect of kiln temperature	 76
6b Lead discharge distributions for the parametric trace metal tests:
effect of afterburner temperature	 77
6c Lead discharge distributions for the parametric trace metal tests:
effect of feed chlorine content	 78
7a Bismuth discharge distributions for the parametric trace metal tests:
effect of kiln temperature	 79
7b Bismuth discharge distributions for the parametric trace metal tests:
effect of afterburner temperature	 80
7c Bismuth discharge distributions for the parametric trace metal tests:
effect of feed chlorine content	 81
8a Barium discharge distributions for the parametric trace metal tests:
effect of kiln temperature	 82
8b Barium discharge distributions for the parametric trace metal tests:
effect of afterburner temperature	 83
8c Barium discharge distributions for the parametric trace metal tests:
effect of feed chlorine content	 84
vii

-------
FIGURES (continued)
Number	Page
9a Copper discharge distributions for the parametric trace metal tests:
effect of kiln temperature	 85
9b Copper discharge distributions for the parametric trace metal tests:
effect of afterburner temperature	 86
9c Copper discharge distributions for the parametric trace metal tests:
effect of feed chlorine content	 87
10a Strontium discharge distributions for the parametric trace metal tests:
effect of kiln temperature	 88
10b Strontium discharge distributions for the parametric trace metal tests:
effect of afterburner temperature	 89
10c Strontium discharge distributions for the parametric trace metal tests:
effect of feed chlorine content	 90
11a Arsenic discharge distributions for the parametric trace metal tests:
effect of kiln temperature			 91
1 lb Arsenic discharge distributions for the parametric trace metal tests:
effect of afterburner temperature	 92
11c Arsenic discharge distributions for the parametric trace metal tests:
effect of feed chlorine content	 93
12a Chromium discharge distributions for the parametric trace metal tests:
effect of kiln temperature	 94
12b Chromium discharge distributions for the parametric trace metal tests:
effect of afterburner temperature	 95
12c Chromium discharge distributions for the parametric trace metal tests:
effect of feed chlorine content	 96
13a Magnesium discharge distributions for the parametric trace metal tests:
effect of kiln temperature	 97
13b Magnesium discharge distributions for the parametric trace metal tests:
effect of afterburner temperature	 98
13c Magnesium discharge distributions for the parametric trace metal tests:
effect of feed chlorine content	 99
14 Afterburner exit flue gas particulate size distributions for the parametric trace
metal tests	 106
viii

-------
FIGURES (concluded)
Number	Page
15	Hazardous constituent trace metal size distributions in afterburner exit
flue gas particulate for the parametric metal tests 		107
16	Nonhazardous trace metal size distributions in afterburner exit flue gas
particulate for the parametric trace metal tests		107
17	Trace metal size distributions in afterburner exit flue gas for Test 4		108
18	Trace metal size distributions in afterburner exit flue gas for Test 6		109
19	Trace metal size distributions in afterburner exit flue gas for Test 7		110
20	Trace metal size distributions in afterburner exit flue gas for Test 8		Ill
21	Trace metal size distributions in afterburner exit flue gas for Test 9		112
22	Trace metal size distributions in afterburner exit flue gas for Test 11		113
ix

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
TABLES
Page
Design Characteristics of the IRF Rotary Kiln System		5
Target Trace Metal Feed Concentrations for the Parametric Tests		8
Target Test Conditions		9
Kiln Operating Conditions 		10
Afterburner Operating Conditions		12
Air Pollution Control System Operating Conditions		14
Actual Versus Target Operating Conditions for the Trace Metal Test Series . .	19
Continuous Emission Monitors Used for the Tests		23
Sampling and Analysis Matrix for the Chromium Valence State Tests 		26
Multiple Metals Train Impinger System Reagents for the Chromium
Valence State Tests		27
Cr( + 6) Train Impinger System Reagents		27
Summary of Chromium Valence State Test Samples 		29
Sampling and Analysis Matrix for the Parametric Trace Metal Tests 		30
Multiple Metals Train Impinger System Reagents for the Parametric Trace
Metal Tests 		33
Summary of Parametric Trace Metal Test Samples 		35
Volatile Organic Compounds Routinely Analyzed by GC/FID at the IRF ....	37
POHC Concentrations in Clay/Organic Liquid Feed Samples 		39
Aqueous Spike Solution Chromium Concentrations for the Chromium Valence
State Tests		40
Aqueous Spike Solution Metals Concentrations for the Parametric Trace Metals
Test Series		41
Clay Matrix Metals Concentrations		42
Integrated Feed Chromium Concentrations for the Chromium Valence State
Tests 		43
x

-------
TABLES (continued)
Number	Page
22	Integrated Feed Metal Concentrations for the Parametric Trace Metals
Test Series 	 44
23	CEM Data		45
24	Flue Gas Particulate Data		48
25	HC1 Emissions Data		49
26	Synthetic Waste Feedrates and Kiln Ash Discharge Rates 		51
27	Total Chromium Discharge Distributions and Mass Balance for the Chromium
Valence State Test Series 	 52
28	Total Chromium Discharge Distributions for the Chromium Valence State Test
Series	 54
29	Total Chromium Particulate/Vapor-Dissolved Phase Flue Gas Distributions for
the Chromium Valence State Test Series	 55
30	Hexavalent Chromium Fractions for the Chromium Valence State Test
Series	 57
31	Summary of Metal Discharge Distributions in the Kiln Ash and Afterburner
Exit Flue Gas for the Trace Metals Parametric Test Series 	 59
32	Summary of Metal Discharge Distributions in the Kiln Ash, Scrubber Exit Flue
Gas, and Scrubber Liquor for the Trace Metals Parametric Test Series 	 60
33	Summary of Achieved Metal Mass Balance Closure for the Trace Metals
Parametric Test Series 	 62
34	Summary of Flue Gas Metal Particulate/Vapor, Dissolved Phase Distributions
for the Trace Metals Parametric Test Series 	 64
35	Summary of Apparent Scrubber Efficiency Ranges and Averages for the Trace
Metals Parametric Test Series	 65
36	Metal Discharge Distributions and Mass Balance Closure for the Trace Metals
Parametric Test Series 	 67
37	Normalized Metal Discharge Distributions and Apparent Scrubber Collection
Efficiencies for the Trace Metals Parametric Test Series 	 70
38	Phase Distribution of Flue Gas Metals in the Afterburner and Scrubber Exit
Flue Gas for the Trace Metals Parametric Test Series	 103
xi

-------
TABLES (concluded)
Number	Page
39	Flue Gas POHC Concentrations for the Parametric Trace Metal Tests 		115
40	POHC DREs for the Parametric Trace Metal Tests		116
41	Afterburner Exit Flue Gas VolatiJe PIC Concentrations		117
42	Scrubber Exit Flue Gas Volatile PIC Concentrations		118
43	Precision. Accuracy, and Completeness Objectives		123
44	Clay/Organic Liquid Feed Sample POHC Analysis Results 		125
45	Clay/Organic Liquid Feed Replicate VolatiJe Organic Analysis Results 		126
46	Volatile Organic Constituent Recovery from Clay/Organic Liquid Feed Matrix
Spike Samples		126
47	Octane Surrogate Recovery from VOST Traps 		128
48	4-Bromofluorobenzene Surrogate Recovery from VOST Traps		129
49	VolatiJe Organic Constituent Recovery from VOST Matrix Spike Samples ....	130
50	Volatile Organic Constituent Recovery from the VOST Matrix Spike Sample
for Test 10 (9/20/88) by Method 5040 and GC/FID Analysis 		131
51	Volatile Organic Constituent Concentrations for Duplicate VOST Samples
Taken at the Scrubber Exit During Test 10 (9/20/88) and Analyzed by
GC/FID and GC/MS (Method 5040) 		132
52	Summary of Laboratory Matrix Spike Sample Analysis Results for the
Chromium Valence State Tests Series		134
53	Cr( + 6) Matrix Spike Recoveries		135
54	Laboratory Matrix Spike Aqueous Sample Analysis Results for the Parametric
Test Series		136
55	Laboratory Matrix Spike Filter and Ash Sample Analysis Results for the
Parametric Test Series 		137
56	Aqueous Spike Solution Analysis Accuracy and Precision		139
57	Feed Clay Analysis Precision		140
xii

-------
SECTION 1
INTRODUCTION
The RCRA hazardous waste incinerator performance standards promulgated by EPA
in January 1981 established limits on incinerator particulate and HC1 emissions, and on
hazardous organic constituent emissions by requiring that at least 99.99 percent destruction and
removal efficiency (DRE) be achieved for the principal organic hazardous constituents (POHCs)
in wastes fed to an incinerator. Emissions of hazardous constituent trace metals were not
regulated by these standards except in a global sense via the particulate standard. Risk
assessments to date have suggested that, of the total risk to human health and the environment
from properly operated incinerators, hazardous constituent trace metal emissions may represent
the largest component.
Despite its importance, the data base on trace metal emissions from incinerators is
currently very sparse. Data on the effects of waste composition and incinerator operation on
trace metal emissions are particularly lacking. In response to these data needs, an extensive
series of tests was conducted at EPA's Incineration Research Facility (IRF), with funding
support from the Office of Solid Waste (OSW), to investigate the fate of trace metals fed to a
rotary kiln incinerator equipped with a venturi scrubber/packed column scrubber for particulate
and acid gas control.
A primary objective of the tests was to investigate the fate of five hazardous constituent
trace metals fed in a solid waste matrix to a rotary kiln incinerator as a function of incinerator
operating temperatures and feed chlorine content. Of interest were metal partitioning, particle
size distribution, flue gas phase distribution, and scrubber efficiency for each of the metals.
The trace metals investigated were arsenic, barium, cadmium, chromium, and lead.
In another OSW sponsored effort within EPA's Risk Reduction Engineering Laboratory
(RREL) to support trace metal emission regulation development, a numerical model is being
developed by another EPA contractor to aid in predicting the relative distributions of trace
metals in emissions and discharges from incinerators. Thus, a second objective of this test
program was to supply data to evaluate the predictive capabilities of this model, and to perhaps
guide further model refinement. To support this objective, an additional four nonhazardous
constituent trace metals were included in the test feed material. These were bismuth, copper,
magnesium, and strontium.
Finally, there has been a continuing interest in establishing which valence state of
chromium predominates in emissions and discharges from incinerators treating chromium-
containing wastes. Of the two common chromium valence states, [the trivalent (Cr( + 3)) and
the hexavalent (Cr( + 6))], the hexavalent is much more toxic. In the absence of data showing
which form is emitted from an incinerator, risk assessments have generally assumed that the
entire amount emitted is in the hexavalent form. This assumption has resulted in specifying
conservative emission limits in the regulatory development process. Thus, a third objective of
1

-------
this test program was to develop data on the valence state distribution of chromium in emissions
and discharges as a function of feed composition and chromium valence state fed.
The test program to address the above objectives was comprised of both a series of eight
parametric tests in which the test waste feed contained nine trace metals, and a series of three
tests in which the test waste feed contained either Cr( + 3) or Cr(+6).
This report summarizes the results of the test program. Section 2 describes the rotary
kiln incineration system at the IRF in which the tests were performed, outlines the test waste
feed characteristics, and summarizes the incinerator operating conditions for each of the tests.
Section 3 discusses the emissions and discharge stream sampling and analysis protocols.
Section 4 presents test program results. Test conclusions are summarized in Section 5. Quality
assurance and quality control aspects of the test program are discussed in Section 6.
2

-------
SECTION 2
FACILITY DESCRIPTION AND TEST CONDITIONS
The pilot-scale rotary kiln incinerator system (RKS) at the IRF in Jefferson, Arkansas,
was used in this test program. A description of the system is presented in Section 2.1, followed
by a description of the synthetic waste mixture in Section 2.2. The test matrices and operating
conditions are described in Section 2.3.
2.1 ROTARY KILN INCINERATOR SYSTEM DESCRIPTION
As indicated in Figure 1 and Table 1, the RKS consists of a primary combustion
chamber, a transition section, and a fired afterburner chamber. The primary air pollution
control system (APCS) consists of a quench section, a venturi scrubber, and a packed-column
scrubber. Quench section installation was completed immediately prior to the initiation of this
test series to allow the RKS to more closely reflect typical field installations of rotary
kiln/venturi scrubber designs. The quench section modification is discussed further in
Section 2.1.2. Downstream of the primary APCS a backup APCS, consisting of a carbon-bed
adsorber and a high-efficiency particulate (HEPA) filter, is in place. The backup system is
designed to ensure that organic compound emissions and particulate emissions to the
atmosphere are negligible during less than optimal test conditions. The main components of the
RKS and its APCS are discussed in more detail in the following subsections.
2.1.1 Incinerator Characteristics
The rotary kiln combustion chamber has an inside diameter of 0.95 m (37.5 in) and is
2.1 m (7 ft) long. The chamber consists of 13 cm (5 in) of refractory encased in a 6.3 mm
(0.25 in) thick steel shell. The chamber volume, including the transition sections, is 1.74 m3
(61.4 ft3). Four steel rollers support the kiln barrel. A variable-speed DC motor coupled with
a reducing gear transmission tumbles the rotary kiln. Typical rotation speeds range from 0.2 to
1.5 rpm.
The afterburner chamber has a 0.91 m (3 ft) inside diameter and is 2.74 m (9 ft) long.
The afterburner chamber wall consists of a 15 cm (6 in) layer of refractory encased in a 6.3 mm
(0.25 in) thick carbon steel shell. The volume of the afterburner chamber is 1.80 m3 (63.6 ft3).
An American Combustion, Inc. burner system was used in the RKS for these tests. The
system consisted of propane-fired burners in the kiln and afterburner sections, with
computer-based gas metering and process control systems designed to control burner flows of
propane and air. A separate system was used to control scrubber system operation. Both
control systems were interfaced with a personal computer for data acquisition at 20-second
intervals.
3

-------
ATMOSPHERE
PUMP	TANK
Figure I. Schematic of the IRF RKS.

-------
TABLE 1. DESIGN CHARACTERISTICS OF THE IRF ROTARY KILN SYSTEM
Characteristics of the Kiln Main Chamber
Length, outside
Diameter, outside
Length, inside
Diameter, inside
Chamber volume
Construction
Refractory
Rotation
Solids retention
time
Burner
Primary fuel
Feed system
Liquids
Sludges
Solids
Temperature (max)
2.61 m (8 ft - 7 in)
1.22 m (4 ft)
2.13 m (7 ft)
0.95 m (3 ft - 1-1/2 in)
1.74 m3 (61.36 ft*)
0.63 cm (0.25 in) thick cold rolled steel
12.7 cm (5 in) thick high alumina castable refractory, variable depth
to produce a frustroconical effect for moving solids
Clockwise or counterclockwise 0.2 to 1.5 rpm
1 hr (at 0.2 rpm)
American Combustion Burner, rated at 880 kW (3.0 MMBtu/hr) with
dynamic 02 enhancement capability
Propane
Positive displacement pump via water-cooled lance
Moyno pump via front face, water-cooled lance
Metered twin-auger screw feeder or fiber pack ram feeder
900°C (1,650°C)
Characteristics of the Afterburner Chamber
Length, outside
Diameter, outside
Length, inside
Diameter, inside
Chamber volume
Construction
Refractor)'
Gas Residence Time
Burner
Primary fuel
Temperature (max)
3.05 m (10 ft)
1.22 m (4 ft)
2.74 m (9 ft)
0.91 m (3 ft)
1.80 m3 (63.6 ft3)
0.63 cm (0.25 in) thick cold rolled steel
15.24 cm (6 in) thick high alumina castable refractory
1.2 to 2.5 seconds depending on temperature and excess air
American Combustion Burner rated at 440 kW (1.5 MMBtu/hr) with
dynamic O-, enhancement capability
Propane
1,200-C (2,200CF)
Characteristics of the Air Pollution Control System
System capacity
Inlet gas flow	107 m3/min (3,773 acfm) at 1,200CC (2,200°F) and 101 kPa (14.7 psia)
Pressure drop
Venturi scrubber	7.5 kPa (30 in WC)
Packed column	1.0 kPa (4 in WC)
Liquid flow
Venturi scrubber	77.2 L/min (20.4 gpm) at 69 kPa (10 psig)
Packed column	116 L/min (30.6 gpm) at 69 kPa (10 psig)
pH control	Feed back control by NaOH solution addition
5

-------
2.U Air Pollution Control System
As previously mentioned, combustion gas exiting the afterburner flowed through a
primary APCS consisting of a quench section, a venturi scrubber, and a packed column scrubber,
then through a secondary APCS consisting of a carbon bed adsorber followed by a HEPA filter.
As originally instaLied, the afterburner exit flue gas was not quenched prior to passing through
the venturi scrubber; the venturi scrubber itself was used to quench the Que gas. This is atypical
of a con%rentiona! rotary kiln/venturi scrubber installation. In addition, since flue gas quenching
occurred in the venturi scrubber, its particulate removal efficiency was adversely affected. To
rectify this situation, the RKS APCS was modified prior to this test program by installing a
quench section upstream of the venturi scrubber. With the quench section installed, the RKS
more nearly reflects traditional rotary kiln/venturi scrubber installation designs; and particulate
removal characteristics of the venturi scrubber more nearly reflect the capabilities of the
technology.
The quench section reduces the temperature of the combustion gas to approximately
77°C (170°F). The cooled flue gas then enters the venturi scrubber which has an automatically
adjustable throat. The scrubber is designed to operate at 7.5 kPa (30 in WC) differential
pressure, with a maximum liquid flowrate of 77 L/min (20.4 gpm). The scrubber liquor enters
at the top of the scrubber and contacts the gas to remove acid gases and entrained particulate.
Downstream of the venturi scrubber, the flue gas enters the packed-column scrubber
where additional scrubbing occurs. The scrubber column is packed with 5.1 cm (2 in) diameter
polypropylene ballast saddles at a depth of 2.1 m (82 in). It is designed to operate at 1.0 kPa
(4 in WC) differential pressure, with a maximum liquid flowrate of 116 L/min (30.6 gpm).
Ambient air is also drawn in at the packed-column Liquor discharge port. Past experience
indicates that induced ambient air comprises approximately 50 percent of the flue gas
downstream of the scrubbers.
One liquor recirculation system supplies both the venturi and packed-column scrubbers.
The liquor is a dilute NaOH solution and is monitored continuously by a pH sensor. An integral
pH controller automatically meters the amount of NaOH needed to maintain the required pH
for proper HC1 removal.
At the exit of the packed-column scrubber, a demister removes most of the suspended
liquid droplets. In a typical commercial incinerator system, the flue gas would be vented to the
atmosphere at this point. However, a backup APCS is in place at the IRF. The flue gas passes
through a bed of activated carbon designed to adsorb the remaining vapor phase organic
compounds. Typically, the carbon bed operates at 77°C (170°F). Because the flue gas is
saturated with moisture and is cooled as it flows through the ducts, condensate is continuously
formed. The condensate accumulates in the carbon bed and drains through a bottom tap into
the blowdown storage tanks.
A set of HEPA filters designed to remove remaining suspended particulate from the flue
gas is located downstream of the carbon adsorption bed. An induced draft fan draws and vents
the treated effluent gas to the atmosphere.
22 SYNTHETIC TEST MIXTURE
The synthetic waste fired during the test program was composed of a mixture of organic
liquids added to a clay absorbent material. Trace metals were incorporated by spiking aqueous
6

-------
mixtures of the metals of interest onto the organic liquid-containing solid material. This
synthetic waste was fed to the rotary kiln via a screw feeder.
The organic-liquid base (which supplied the heat content and POHC concentrations to
the test mixture) consisted of toluene, with varying amounts of tetrachloroethylene and
chlorobenzene added to give a range of synthetic waste chlorine contents. Synthetic waste
chlorine content was varied from zero (no chlorinated POHCs added) to nominally 8 percent
by weight. The synthetic waste was prepared by mixing the organic liquid with the clay
absorbent material to yield a mixture containing about 30 percent (weight) organic liquid. After
mixing, the test mixture was poured into 55-gallon drums, each holding approximately 140 kg
(300 lbs).
The trace metals of interest were prepared in aqueous solutions. These solutions were
then metered into the clay/organic liquid matrix at the screw feeder introducing the waste into
the kiln. For the parametric series of tests aimed at evaluating the distribution of the five
hazardous constituent and four nonhazardous constituent metals in emissions and discharges,
the aqueous trace meta] feed solution was metered at a rate which would produce the final
nominal synthetic waste feed concentrations noted in Table 2 (assuming no background metal
concentration in the clay matrix). For the chromium valence state tests, the aqueous chromium
feed solution was metered at a rate which would produce a synthetic waste feed concentration
of about 50 ppm Cr( + 3 or +6) (again assuming zero background clay chromium content).
2 J TEST CONDITIONS
As noted in Section 1, the test program consisted of two test series: a three-test
chromium valence state series, and an eight-test parametric trace metals series. For all tests,
an auger was used to feed the clay/organic mixture at a nominal rate of 63 kg/hr (140 lbs/hr).
The metals were premixed in an aqueous solution, then added to the clay/organic liquid feed
just before introducing the mixture into the kiln. A gear pump continuously injected the trace
metal aqueous solution spike into the feed at the auger to produce the desired spike
concentrations. The nominal flowrate through the gear pump was 0.5 to 1 L/hr.
The test variables selected for the three chromium valence state tests were feed chlorine
content and the chromium valence state in the waste feed. Two tests were performed with
chromium fed in the trivalent state. One of these had a waste feed chlorine content of zero; the
other had a waste feed chlorine content of nominally 8 percent. The third test was performed
with the aqueous spike chromium fed in the hexavalent state with no chlorine in the feed. All
three tests were performed under the same nominal incinerator operating conditions: kiln
temperature of 871°C (1,600°F), kiln exit flue gas 02 of 11.5 percent, afterburner temperature
of 1,093°C (2,000°F), and afterburner exit flue gas Oz of 7.5 percent. These conditions are
considered typical of industrial hazardous waste incinerator operation.
The test variables chosen for the parametric test series were feed chlorine content, kiln
temperature, and afterburner temperature. A factorial experimental matrix for three variables
over three levels was chosen, giving eight target test conditions, as shown in Table 3. All tests
were designed for the same nominal excess air levels. These were kiln exit flue ps Oz of
11.5 percent and afterburner exit flue gas 02 of 7.5 percent.
The actual kiln and afterburner operating conditions achieved for each test are
summarized in Tables 4 and 5, respectively. Table 6 provides a similar summary of the air
pollution control system operation for each test. In the tables, the chromium valence state tests
are designated as Tests 1, 2, and 3; the parametric tests are designated as Tests 4 through 11.
7

-------
TABLE 2. TARGET TRACE METAL FEED CONCENTRATIONS
FOR THE PARAMETRIC TESTS
Synthetic solid hazardous
waste concentration
Metal	(ppm)
Hazardous constituent
trace metals
Arsenic
50
Barium
40
Cadmium
10
Chromium
50
Lead
50
Other trace metals
Bismuth
140
Copper
500
Magnesium
80
Strontium
250
8

-------
TABLE 3. TARGET TEST CONDITIONS
Feed CI content	Kiln exit	Afterburner exit
Test	(%)	temperature, °C (°F)	temperature, °C (°F)
Chromium Valence State Tests"
1	0	871 (1600)	1093 (2000)
2	0	871 (1600)	1093 (2000)
3	8	871 (1600)	1093 (2000)
Parametric Trace Metal Tests
4
0
5
4
6
4
7
4
8
4
9
4
I0b
4
11
8
871 (1600)	1093 (2000)
816 (1500)	1093 (2000)
927 (1700)	1093 (2000)
871 (1600)	1093 (2000)
871 (1600)	1204 (2200)
871 (1600)	982 (1800)
871 (1600)	1093 (2000)
871 (1600)	1093 (2000)
aTes: 1 performed with hexavalent chromium; tests 2 and 3 performed with trivalent chromium.
bTest point 10 is a duplicate of test point 7.
9

-------
TAKLE 4. KILN OPERATING CONDITIONS

Test 1
Test 2
Test 3
Test 4
Tcsl 5
Test 6
Parameter
(9-29-88)
(9-28-88)
(9-26-88)
(9-14-88)
(8-25-88)
(9-16-88)
Propane feedrate






range, scm/hr
5.1 to 6.8
5.7 to 7.1
7 4 to 9.1
4.5 to 6.2
1.1 to 6.5
7 4 to 8.8
(scfh)
(180 to 240)
(200 to 250)
(260 to 320)
(160 to 220)
(40 to 230)
(260 to 310)
average, scm/hr
6.2
6.4
7.9
50
5.4
7.8
(scfh)
(217)
(224)
(280)
(176)
(190)
(275)
Combustion air flowrate





range, scm/hr
188 to 235
20! to 251
266 to 321
156 to 217
9 to 190
257 to 304
(scfh)
(6630 to 8310)
(7110 to 8850)
(9400 to 11,330)
(5520 to 7650)
(320 to 6700)
(9090 to 10,720)
average, scm/hr
219
226
283
173
157
274
(scfh)
(7739)
(7994)
(9996)
(6095)
(5542)
(9661)
Exit temperature






range, °C
856 to 893
839 to 893
854 to 886
851 to 892
793 to 843
709 to 946
(°F)
(1572 to 1640)
(1542 to 1639)
(1569 to 1626)
(1564 to 1638)
(1459 to 1550)
(1669 to 1735)
average, °C
873
872
871
874
825
928
(°F)
(1603)
(1602)
(1600)
(1606)
(1517)
(1702)
Pressure






range, Pa
-3 to -25
-5 to -12
-3 to -45
-5 to -16
-5 to -12
0 to -8
(in WC)
(-0.01 to -0.18)
(-0.02 to -0.05)
(-0.01 to -0.18)
(-0.02 to -0.07)
(-0.02 to -0.05)
(0.00 to -0.03)
average, Pa
-10
-6
-7
-9
-8
-6
(in WC)
(-0.03)
(-0.02)
(-0.03)
(-0.04)
(-0.03)
(-0.02)
Exit 02






range, %
10.7 to 11.8
10.9 to 11.8
11.1 to 13.2
11.8 to 12.8
11.0 to 13.6
9.0 to 11.0
average, %
11.2
11.3
11.5
12.2
11.9
10.6
Clay/organic mixture






feedrate






total, kg/hr
62.6
62.0
61.3
59.4
63.5
62.6
(Ib/hr)
138.0
136.7
135.1
131.0
140.0
138.0
organic fraction,






kg/hr
17.5
17.5
18.4
13.8
15.0
17.2
(Ih/hr)
38.5
38.5
40.5
30.4
33.0
38.0
(continued)

-------
TABLE 4. (concluded)
Tcsl 7	Tcsl 8	Tesl 9	Test 10	Test 11
Parameter	(8-30-88)	(9-7-88)	(9-9-88)	(9-20-88)	(9-22-88)
Propane feedrate
range, scm/hr
6.2 to 7.9
7.1 to 10 2
5.4 to 6.8
4.8 to 7.1
7.1 to 9.3
(scfh)
(220 to 280)
(250 to 360)
(190 to 240)
(170 to 250)
(250 to 330)
average, scm/hr
69
92
6.1
6.4
7.5
(scfh)
(244)
(325)
(216)
(225)
(263)
Combustion air flowrate





range, scm/hr
223 to 277
203 to 269
176 to 243
176 to 257
260 to 330
(scfh)
(7880 to 9780)
(7180 to 9500)
(6230 to 8590)
(6230 to 9070)
(9170 to 11,670)
average, scm/hr
246
245
213
233
271
(scfh)
(8685)
(8643)
(7537)
(8226)
(9570)
Exit temperature





range, °C
869 to 886
832 to 916
836 to 891
841 to 894
852 to 885
(°F)
(1597 to 1626)
(1530 to 1680)
(1536 to 1636)
(1546 to 1642)
(1565 to 1625)
average, °C
878
871
875
873
870
(°F)
(1612)
(1599)
(1607)
(1603)
(1599)
Pressure





range. Pa
-5 to -11
-5 to -8
-8 to -20
-3 to -5
0 to -45
(in WC)
(-0.02 to -0.04)
(-0.02 to -0.03)
(-0.03 to -0.08)
(-0.01 to -0.02)
(0.00 to -0.18)
average, Pa
-8
-6
-10
-5
-10
(in WC)
-0.03
(-0.02)
(-0.04)
(-0.02)
(-0.04)
Exit O,





range, %
11.7 to 13.7
12.8 to 14.5
12.1 to 13.0
10.6 to 12.7
10.8 to 11.7
average, %
12.4
13.6
12.7
11.4
11.4
Clay/organic mixture





feedrate





total, kg/hr
63.4
63.4
62.5
62.8
63.5
(Ib/hr)
139.8
139.8
137.8
138.4
140.0
organic fraction,





kg/hr
16.8
15.0
14.1
18.7
18.1
(Ib/hr)
37.0
33.0
31.0
41.3
40.0

-------
TABLE 5. AFTERBURNER OPERATING CONDITIONS

Test 1
Test 2
Test 3
Test 4
Test 5
Test 6
Parameter
(9-29-88)
(9-28-88)
(9-26-88)
(9-14-88)
(8-25-88)
(9-16-88)
Propane feedrate






range, scm/hr
17.0 to 17 8
16 7 to 17.6
17.0 to 17 8
16 4 to 17.3
17.8
16 4 to 17.3
(scfh)
(600 to 630)
(590 to 620)
(600 to 630)
(580 to 610)
(630)
(580 to 610)
average, scm/hr
17.3
170
17.3
16.7
17.8
17.1
(scfh)
(609)
(600)
(611)
(591)
(630)
(604)
Combustion air flowrate





range, scm/hr
534 to 540
510 to 526
491 to 512
540
540
540
(scfh)
(18,860 to 19,060)
(18,020 to 18,570)
(17,330 to 18,070)
(19,060)
(19,060)
(19,060)
average, scm/hr
540
521
497
540
540
540
(scfh)
(19,059)
(18,392)
(17,567)
(19,060)
(19,060)
(19,060)
Exit temperature






range, °C
1090 to 1103
1091 to 1103
1091 to 1102
1083 to 1100
1052 to 1082
1084 to 1112
(°F)
(1994 to 2018)
(1996 to 2017)
(1996 to 2016)
(1982 to 2012)
(1925 to 1979)
(1984 to 2033)
average, °C
1094
1095
1096
1093
1071
1092
(°F)
(2002)
(2003)
(2005)
(1999)
(1959)
(1998)
Pressure






range, Pa
-30 to -95
-30 to -40
-15 to -30
-20 to -30
-20 to -50
-25 to -32
(in WC)
(-0.12 to -0.38)
(-0.12 to -0.16)
(-0.06 to -0.10)
(-0.08 to -0.12)
(-0.08 to -0.20)
(0.10 to -0.13)
average. Pa
-42
-32
-27
-25
-42
-30
(in WC)
(-0.15)
(-0.12)
(-0.11)
(-0.10)
(-0.17)
(-0.12)
Exit O,






range, %
6.6 to 7.9
6.4 to 8.0
6.3 to 7.7
6.1 to 8.0
6.0 to 10.2
5.7 to 7.8
average, %
7.3
7.4
7.2
7.1
8.4
7.1
Exit CO,






range, %
8.9 to 10.1
8.9 to 10.2
9.0 to 10.2
8.6 to 10.2
6.7 to 10.0
8.7 to 10.4
average, %
9.5
9.4
9.5
9.5
8.2
9.4
(continued)

-------
TABLE 5. (concluded)

Test 7
Tcsl 8
Tcsl 9
Test 10
Test 11
I'aramclcr
(8-30-88)
(9-7-88)
(9-9-88)
(9-20-88)
(9-22-88)
Propane feetlrate





range, scm/hr
178
17.6 to 17 8
13.0 to 13.3
17.3 to 17.8
17.3 to 17.8
(scfh)
(620 to 630)
(620 to 630)
(460 to 470)
(610 to 630)
(610 to 630)
average, scm/hr
178
17.8
13 2
17.5
17.5
(scfh)
(630)
(630)
(466)
(619)
(619)
Combustion air flowrate





range, scm/hr
540
445 to 540
539 to 540
540
540
(scfh)
(19,060)
(15,730 to 19,060)
(19,050 to 19,060)
(19,060)
(19,060)
average, scm/hr
540
451
540
540
540
(scfh)
(19,060)
(15,912)
(19,060)
(19,060)
(19,060)
Exit temperature





range, °C
1079 to 1095
1168 to 1213
967 to 991
1081 to 1103
1084 to 1105
(°F)
(1974 to 2003)
(2135 to 2216)
(1791 to 1816)
(1981 to 2014)
(1984 to 2021)
average, °C
1088
1196
983
1094
1092
(°F)
(1991)
(2184)
(1803)
(2000)
(1998)
Pressure





range, Pa
0 to -55
-30 to -37
-15 to -37
-25
-25 to -95
(in WC)
(0 00 to -0.22)
(-0.12 to -0.15)
(-0.06 to -0.15)
(-0.12)
(-0.10 to -0.38)
average, Pa
-52
-30
-22
-25
-35
(in WC)
(-0.21)
(-0.12)
(-009)
(-0.10)
(-0.14)
Exit Oj





range, %
7.1 to 8.7
3.7 to 7.5
7.7 to 9.3
6.5 to 8.5
6.3 to 8.1
average, %
7.9
5.9
8.7
7.6
7.5
Exit CO,





range, %
7.6 to 9.5
8.9 to 11.2
7.7 to 9.0
8.5 to 10.0
4.8 to 6.8
average, %
8.8
10.3
8.3
9.2
6.1

-------

TAIfLK 6.
AIR POLLUTION CONTROL
SYSTEM OPERATING CONDITIONS


Test 1
Test 2
Test 3
Test 4
Test 5
Test 6
Parameter
(9-29-88)
(9-28-88)
(9-26-88)
(9-14-88)
(8-25-88)
(9-16-88)
Quench chamber






scrubber liquor flowrate






range, L/min
68
53 to 68
53 to 68
53 to 72
45 to 49
45 to 68
(epm).
(18.0)
(14.0 to 18.0)
(14 0 to 18 0)
(14.0 to 19 0)
(12.0 to 13.0)
(14.0 to 18.0)
average, L/min
68
66
67
54
45
67
(gpm)
(18.0)
(17.5)
(17.8)
(14.2)
(12.0)
(17.6)
Venturi scrubber






liquor (lowrate






range, L/min
76
76 to 79
45 to 87
64 to 68
64 to 68
64
(BP™) .
(20.0)
(20.0 to 21.0)
(12.0 to 23.0)
(17.0 to 18.0)
(17.0 to 18.0)
(17.0)
average, L/min
76
76
73
66
68
64
(gpm)
(20.0)
(20.0)
(19 2)
(17.5)
(17.9)
(17.0)
Venturi scrubber






pressure drop






range, kPa
6.0 to 6.2
6.0 to 6.7
4.5 to 6.2
5.7 to 7.0
4.7 to 5.7
5.5 to 6.2
(in WC)
(24 to 25)
(24 to 27)
(18 to 25)
(23 to 28)
(19 to 23)
(22 to 25)
average, kPa
6.0
6.1
5.5
6.0
5.3
5.7
(in WC)
(24.1)
(24.6)
(22.0)
(24.2)
(21.2)
(23.0)
Packed column scrubber






liquor flowrate






range, L/min
106
106 to 110
68 to 132
102 to 106
114 to 121
95
(gpm)
(28.0)
(28.0 to 29.0)
(18.0 to 35.0)
(27.0 to 28.0)
(30.0 to 32.0)
(25.0)
average, L/min
106
108
106
104
115
95
(gpm)
(28.0)
(28.5)
(27.9)
(27.6)
(30.4)
(25.0)
Packed column






pressure drop






range, kPa
3.2 to 3.5
2.7 to 3.2
3.0 to 3.7
2.0 to 2.5
0.25 to 1.5
3.0 to 3.7
(in WC)
(13 to 14)
(11 to 13)
(12 to 15)
(8 to 10)
(1 to 6)
(12 to 15)
average, kPa
3.2
3.1
3.6
2.2
0.8
3.4
(in WC)
(13.0)
(12.4)
(14.4)
(8.9)
(3.1)
(13.8)
Scrubber liquor pH






range
6.5 to 7.1
6.5 to 7 2
6.0 to 9.6
7.0 to 7.3
3.3 to 8.8
6.4 to 8.2
average
7.0
7.0
7.6
7.2
6.8
7.1
(continued)

-------
TAIILE 6. (continued)

Test 1
Test 2
Test 3
Test 4
Test 5
Test 6
Parameter
(9-29-88)
(9-28-88)
(9-26-88)
(9-14-88)
(8-25-88)
(9-16-88)
Scrubber blowdown






flowr;ite






range, L/min
0.8 to 3 0
17 to 2 6
1.5 to 6 1
0 0 to 3.0
1.9 to 2 6
0.8 to 3.8
(gpm)
(0 2 to 0 8)
(0.5 to 0.7)
(0 4 to 16)
(0.0 to 0.8)
(0.5 to 0.7)
(0.2 to 10)
average, L/min
1.9
19
2.6
19
2.3
1.9
. (KPm)
(0.5)
(0.5)
(0.7)
(0.5)
(0.6)
(0.5)
Scrubber liquor






temperature






range, (°C)
68 to 76
74 to 75
43 to 76
73 to 75
72 to 75
68 to 75
•p
(155 to 169)
(166 to 167)
(109 to 168)
(164 to 167)
(162 to 167)
(155 to 167)
average, (°C)
74
75
73
74
73
74
"F
(166)
(167)
(163)
(166)
(164)
(166)
ID fan inlet pressure






range, kPa
-6.0 to -9.0
-8.5 to -8 7
-8.5 to -8.7
-7.5 to -8.5
-5.5 to -5.7
-8.7
(in WC)
(-24 to -36)
(-34 to -35)
(-34 to -37)
(-30 to -34)
(-22 to -23)
(-35)
average, kPa
-8.3
-8.6
-8.6
-7.6
-5.5
-8.7
(in WC)
(-33.5)
(-34.4)
(-34.4)
(-30.6)
(-22.1)
(-35)
(continued)

-------
TABLE 6. (continued)

Test 7
Test 8
Test 9
Test III
Test II
Parameter
(8-30-88)
(9-7-88)
(9-9-88)
(9-20-88)
(9-22-88)
Quench ch;i miter





scrubber liquor (lowrate





range, L/min
45
57
49 to 53
53 to 68
68
(KP"i)
(12.0)
(15 0)
(13 0 to 14 0)
(14 0 to 18.0)
(18.0)
average, L/min
45
57
53
55
68
(gPm)
(12 0)
(15.0)
(14.0)
(14.4)
(18 0)
Venturi scrubber





liquor flowrate





range, L/min
53 to 72
68
68 to 72
76
72 to 76
(gpm) .
(14.0 to 19.0)
(18 0)
(18.0 to 19.0)
(20.0)
(19.0 to 20.0)
average, L/min
69
68
69
76
73
(Epm)
(18.1)
(18.0)
(18.3)
(20.0)
(19.3)
Venturi scrubber





pressure drop





range, kPa
2.5 to 7.0
4.7 to 5.2
5.2 to 6.7
5.2 to 6.2
4.0 to 6.2
(in WC)
(10 to 28)
(19 to 21)
(21 to 27)
(21 to 25)
(16 to 25)
average, kPa
5.7
4.9
6.3
5.6
5.1
(in WC)
(22.8)
(19.8)
(25.4)
(22.6)
(20.6)
Packed column scrubber





liquor flowrate





range, L/min
110 to 114
110 to 114
106 to 110
110 to 114
110 to 114
(6Pm) .
(29 to 30)
(29 to 30)
(28 to 29)
(29 to 30)
(29 to 30)
average, L/min
111
112
108
113
111
(gPm)
(29.3)
(29.6)
(28.4)
(29.8)
(29.3)
Packed column





pressure drop





range, kPa
2.7 to 3.7
3.2 to 3.7
1.2 to 2.0
3.0 to 3.7
1.5 to 3.7
(in WC)
(11 to 15)
(13 to 15)
(5 to 8)
(12 to 15)
(6 to 15)
average, kPa
3.1
3.6
1.6
3.2
3.4
(in WC)
(12.3)
(14.4)
(6.4)
(12.9)
(13.6)
Scrubber liquor pH





range
6.0 to 8.9
6.2 to 8.1
5.9 to 8.0
5.1 to 8.2
5.0 to 8.5
average
7.2
7.2
6.9
6.8
7.5
(continued)

-------
TABLE 6. (concluded)
Test 7	Test 8	Tesl 9	Test 10	Test II
Parameter	(8-30-88)	(9-7-88)	(9-9-88)	(9-20-88)	(9-22-88)
Scrubber blowdown
flowrate
range, L/min	1.5 to 5.7	0.8 to 9.5	0.0 to 1.9	1.5 to 5.7	1.5 to 3.0
(gpm)	(0.4 to 1.5)	(0.2 to 2.5)	(0.0 to 0 5)	(0.4 to 1.5)	(0.4 to 0.8)
average, L/min	2.3	1.9	1.5	2.3	2.3
(gpm)	(06)	(0 5)	(0 4)	(Of.)	(0.6)
Scrubber liquor
temperature
range, (°C)	72 to 74	74 to 76	72 to 74	74 to 76	69 to 77
°F	(161 to 166)	(166 to 169)	(162 to 165)	(165 to 168)	(156 to 171)
average, (°C)	73	76	73	75	75
°F	(164)	(168)	(163)	(167)	(167)
ID fan inlet pressure
range, kPa	-5.2 to -8.7	-7.7 to -9.7	-6.5 to -7.5	-8.0 to -8.5	-5.0 to -9.2
(in WC)	(-21 to -35)	(-31 to-39)	(-26 to -30)	(-32 to-34)	(-32 to-37)
average, kPa	-7.9	-8.1	-7.4	-8.2	-8.1
(in WC)	(-31.7)	(-32 6)	(-29 7)	(-32.8)	(-32.4)

-------
Control room records of the operating parameters recorded at 15-minute intervals are given in
Appendix A. Appendix A also contains plots of the operating conditions for the kiln and
afterburner recorded at 20-second intervals on the PC data acquisition system.
The ranges and averages presented in Tables 4 and 5 were developed using the
computer recorded data obtained during periods of flue gas sampling for all tests except Test 6.
During Test 6 a power outage disabled the data acquisition system midway through the test.
The remaining test data were taken from the data log book. The kiln exit 02 monitor also
malfunctioned during this test. However, O, readings were obtained at 30-minute intervals using
an Orsat analyzer; these values were used to develop the values presented in the tables. The
APCS data in Table 6 were taken from the control room log. Only flue gas 02 and C02 are
given in Tables 4 and 5. Flue gas concentrations of other combustion gas species (e.g. CO) are
presented in Section 4.2
Table 7 summarizes the actual incinerator operating condition ranges and averages
achieved for each test (temperatures and flue gas 02 levels) and compares these to the
respective target conditions. Figure 2 presents a graphical summary of the incinerator
temperature data from Table 7.
Figure 2 illustrates that average test temperatures achieved were within 9°C (17°F) of
target conditions for all tests except one. For this one test the afterburner temperature was 22°C
(41CF) lower than the target. The data in Table 7 show that kiln exit flue gas 02 was generally
within 1 percent of the test program target of 11.5 percent, and that afterburner exit flue gas 02
was generally within about 1 percent of the test program target of 7.5 percent for all tests except
one. For this one test, fan capacity limitations resulted in afterburner exit flue gas 02 being
substantially below the target level, while kiln exit flue gas 02 was high.
Brief llnmeouts in the kOn and afterburner did occur during Test 3, generally due to loss
of signal from the flame sensor. Clay/organic liquid mixture and metal feeds were stopped while
the system was restarted, and flue gas sampling had not yet begun. Therefore, this upset is not
expected to affect the data.
18

-------
TABLE 7. ACTUAL VERSUS TARGET OPERATING CONDITIONS FOR THE TRACE METAL TEST SERIES
Kiln exit	Afterburner exit
Temperature, "C (°F)	Temperature, °C (°F)
Actual	Flue	Actual	Flue
K»s			 gas
Test	O,	O,
Test	date	Target Minimum Maximum Average (%) Target	Minimum Maximum	Average (%)
Chromium valence
state tests
1
9/29/88
871
(1600)
856
(1572)
893
(1640)
873
(1603)
11.1
1093 (2000)
1090(1994)
1103 (2018)
1094 (2003)
7.3
2
9/28/88
871
(1600)
839
(1542)
893
(1639)
872
(1602)
11.2
1093 (2000)
1091 (19%)
1103 (2017)
1095 (2003)
7.3
3
9/26/88
871
(1600)
854
(1569)
866
(1626)
871
(1600)
11.5
1093 (2000)
1091 (1996)
1102 (2016)
1096 (2005)
7.2
rametric
tests














4
9/14/88
871
(1600)
851
(1564)
892
(1638)
874
(1606)
12.2
1093 (2000)
1083 (1982)
1100 (2012)
1093 (1999)
7.1
5
8/25/88
816
(1500)
793
(1459)
843
(1550)
825
(1517)
11.9
1093 (2000)
1052(1925)
1082 (1979)
1071 (1959)
8.4
6
9/26/88
927
(1700)
909
(1669)
946
(1735)
928
(1701)
10.6*
1093 (2000)
1084 (1984)
1112 (2033)
1092 (1998)
7.1
7
8/30/88
871
(1600)
869
(1597)
886
(1626)
878
(1612)
12.4
1093 (2000)
1079 (1974)
1095 (2003)
1088 (1991)
7.9
R
9/07/88
871
(1600)
832
(1530)
916
(1680)
871
(1599)
13.6
1204 (2200)
1168 (2135)
1213 (2216)
1196 (2184)
5.8"
9
9/09/88
871
(1600)
836
(1536)
891
(1636)
875
(1607)
12.7
982 (18(H))
967 (1791)
991 (1816)
983 (1803)
8.7
10
9/20/88
871
(1600)
841
(1546)
894
(1642)
873
(1603)
11.4
1093 (2000)
1081 (1981)
1103 (2014)
1094 (2000)
7.6
11
9/22/88
871
(1600)
852
(1565)
885
(1625)
870
(1599)
11.4
1093 (2000)
1084 (1984)
1105 (2021)
1092 (1998)
7.4
'Monitor not in service. Data obtained using Bacharach Fyrile 02 analyzer.
bAfterburner air fcedrate at fan capacity.

-------
1,200
®
o>
eo
ri
5
w
LU
1.150
1,100
©
me
® TARGET CONDITION
*
ACHIEVED CONDITION.
INTERSECTION IS MEAN
CONDITION; BARS DENOTE
RANGE OVER TEST.
1,050
S3

© ©
*4
I® ©
S1
©
1,000
h
J
©
®±	1
800	B50	900
KILN TEMPERATURE fC)
sso
Figure 2. Actual versus target operating temperatures for the parametric trace metal tests.
20

-------
SECTION 3
SAMPLING AND ANALYSIS PROCEDURES
Sampling and analysis for both the chromium valence state- and parametric test series
were designed to meet research objectives and IRF permit-compliance requirements. Sampling
locations were the same for both test series, but specific analysis methods were used for the
different trace metals. Figure 3 identifies the sampling point locations. Stack sampling
(sampling point 7) was primarily for permit compliance. In general, sampling for each test
consisted the following:
•	Obtaining a composite sample of the feed materials (clay/organic liquid mixture,
and aqueous metal spike solution) and of the ash
•	Obtaining several samples of the scrubber blowdown water over time
•	Obtaining samples of the flue gas at the afterburner and scrubber exits for
particulate and vapor phase metal analysis
•	Obtaining samples of the flue gas at the afterburner and scrubber exits for %'olatile
or&anic hazardous constituent analvses using a volatile organic sampling train
(VOST)
•	Obtaining continuous monitor sampling of various combinations of flue gas 02,
CO, CO-,, NOx, and total unburned hydrocarbon (TUHC) at the kiln, afterburner,
and scrubber exits and in the stack
•	Obtaining samples of the flue gas at the afterburner and scrubber exits, and at the
stack for total particulate and HC1
The continuous emission monitors (CEMs) available at the IRF and the locations they
monitored for these tests are summarized in Table 8. This monitoring arrangement was
employed in all tests. Figure 4 illustrates the generalized flue gas conditioning and flow
distribution system at the IRF. Four independent systems, such as those illustrated in Figure 4,
are in place so that appropriately conditioned sample gas from four separate sampling locations
can be routed to any of the available monitors listed in Table 8. The CEM setup described in
Table 8, with appropriate gas conditioning per Figure 4, was employed throughout this test
program.
Details of the extractive sampling and analysis procedures are discussed in the following
subsections.
21

-------
SAMIM.I-: LOCATIONS
Continuous
Feed* and Residual*	Mnniior*	Hoe Gk
VOST Method S
(vnlolilt (parlkulate
one*nic«) and lift)
1	X	X
2	X
1	X
4	XXX	XXX	X'
5	X
A	X	X	X	X	X	X	XX'
7	XXX	X"	X
"Tc«i« •! through 11 tinly
Tcm* 5. H. ami 11 only
Sampling
Point
Clay
Organk
Liquid
Mrtal
Sp4h*
Solution
Kiln
A*h
CO.
CO,
Unhealed
T\IIIC
(leafed
nine
Method S
fTe«l* 13)
with maltipte
metal* Imptngen,
(particulate and Cr)
Simultaneous
Method 5
(Test* I-.')
with no filler and
rausllr impinfer*
(Cr( ~ *) and IICI)
Method 17
ffeM*4-1l)
with maltlpte
melal* Impingtni
(parlkoliie,
metaU and HO)
Method 5
(Test* 4-11)
with multiple
mellH Implnpn
(particulate,
metal*, and IICI)
Figure 3. Sampling summary Tor the chromium valence stale and parametric trace metals
test scries.

-------
TABLE 8. CONTINUOUS EMISSION MONITORS USED FOR THE TESTS
Monitor
Location Constituent Manufacturer Model	Principle
Range
Kiln exit
Afterburner 02
exit
CO
CO,
Scrubber
exit
Stack
Unhealed
TUHC
CO
CO,
NO.
Unhealed
TUHC
Heated
TUHC
CO/CO,
Heated
TUHC
O,
Beckman
Beckman
Horiba
Horiba
Shimadzu
Horiba
Horiba
Thermo
Hiectron
Shimadzu
Shimadzu
Infrared
Industries
Shimadzu
Teledyne
755
755
10AR
Paramagnetic
Paramagnetic
VIA 500	NDIR
PIR 2000	NDIR
OC Mini	FID
VIA 500	NDIR
PIR 2000	NDIR
Chemilumi
nescent
GC Mini	FID
GC Mini	FID
7020	NDIR
GC Mini	FID
326A	Fuel cell
0-10 percent
0-25 percent
0-50 percent
0-100 percent
0-10 percent
0-25 percent
0-50 percent
0-100 percent
0-50 ppm
0-500 ppm
0-20 percent
0-80 percent
0-10 ppm to
0-2,000 ppm in
multiples of 2
0-50 ppm
0-500 ppm
0-20 percent
0-80 percent
0-75 ppm to
0-10,000 ppm in
multiples of 2
0-10 ppm to
0-2,000 ppm in
multiples of 2
0-10 ppm to
0-2,000 ppm in
multiples of 2
0-20 percent C02
0-2,000 ppm CO
0-10 ppm to
0-2,000 ppm in
multiples of 2
0-5 percent
0-10 percent
0-25 percent
23

-------
r
HEATED
TUHC
MONITOR
o2
MONITOR
MONrTOR MONITOR MONITOR TUHC
MONITOR
r
L
HEATED TO
| 150TO 175°C
" Mixed Na and Ca hydroxides for
acid gas removal
Figure 4. Generalized CEM sample gas flow schematic.
24

-------
3.1
CHROMIUM VALENCE TESTS
Table 9 summarizes the sampling and analysis matrix in addition to the CEM
monitoring for the chromium valence state test series.
3.1.1 Sampling Procedures
As indicated in Table 9, the flue gas at two locations, the afterburner exit and the
scrubber exit, was simultaneously sampled for particulate, total chromium (Cr), hexavalent
chromium Cr( + 6), and HC1 (Test 3 only), using variations of a standard Method 5 sampling
train. Two sampling trains were simultaneously operated at each location. One train at each
location sampled for particulate load and total chromium levels. The other train at each location
sampled for Cr( + 6) and, during Test 3, HC1. Particulate load and HC1 were also measured at
the stack for all tests, using a standard Method 5 train.
The impingers in the train used for particulate and total chromium capture were
specified for multiple metals sampling as noted in Table 10. After sampling, the train contents
were collected as specified in Method 5. The probe wash (acetone) and filter were desiccated
to constant weight to provide the particulate load measurement. After final weighing, the probe
wash was resuspended in acetone for later metal analysis. The aqueous condensate from the
first impinger was preserved to a pH of less than 2 with HN03.
The impinger contents of the train used for Cr( + 6) sampling were as noted in
Table 11. This train did not contain a filter since Cr( + 6) in flue gas samples is best preserved
if it is collected in an alkaline solution as quickly as possible. A NaOH solution was used to
wash the sampling probe. This solution was then combined with the solution in impinger 1.
Aliquots of the impinger 1, 2, and 3 solutions were taken for HC1 measurement for Test 3.
All Method 5 trains used glass-lined sampling probes. The trains at the afterburner
and scrubber exits collected at least 2.8 m3 (100 ft3) of flue gas over about a 3-hour sampling
period. This period began no less than 1/2 hour after the start of the test mixture feed. Test
mixture feed continued until sampling was complete. The sampling trains at the stack collected
at least 0.9 m3 (32 ft3).
Grab sampling of chromium spike solution (50 mL) and scrubber blowdown (100
mL) was performed at the beginning of Method 5 sampling. Additional blowdown grab samples
were taken hourly until all Method 5 sampling was complete. The chromium spike solution and
each blowdown sample were split into two samples, one for total chromium and one for Cr( + 6)
analysis. At the conclusion of each test, two composite samples of the kiln ash were collected
from the ash pit, one for total chromium and one for Cr( + 6) analysis. All spike solution and
blowdown samples for total chromium analysis were preserved with HN03 to pH<2. All
blowdown samples for Cr( + 6) analysis were preserved with NaOH solution to pH> 12. Kiln ash
samples for Cr( + 6) analysis were preserved with 3 percent Na2C03, 2 percent NaOH solution
(the solution used for Cr( + 6) digestion in Method 3060, Reference 2).
Composite feed samples from each drum were collected by trier sampling at three
locations in the drum cross-section. Each of these composites was ultimately analyzed for
toluene, tetrachloroethylene, chlorobenzene, and ultimate composition. Samples for volatile
organic analysis were collected from each drum's composite sample when the contents of that
drum were first fed into the kiln. These samples were sealed in 40 mL VOA vials at that time.
An aliquot of each drum's composite sample was also submitted for ultimate analysis and for
total chromium analysis.
25

-------
TABLE 9. SAMPLING AND ANALYSIS MATRIX FOR THE CHROMIUM VALENCE
STATE TESTS
Sample
Location
Sampling
procedure Parameter
Analysis
Method
Frequency
Aqueous
chromium
spike
solution
Feed clav
Kiln iaet
Waste
preparation
Composite Total Cr
Cla\/organic preparation
liquid rruclure storage drum
Flue gas
Composite
Tner
Afterburner and Method 5'
scrubber exits
Stack gas
Method 5'
Stack downstream Method 5'
o.' carbon bed/
HEPA filter
Cr(+6)
Cr(+6)
C,H^ 0,01,,
C,H,Ci'
Ultimate
anaN'sis
(C, H. O, CI)
Total Cr
Digestion by Method 3010	1/tcst
analysis by Method 7191"
Digestion by Method 3060*	1/iest
analysis by Method 7197s
Digestion by Method 3060'	1 composite
analysis by Method 7197°
Purge and trap by	1/test
Method 5030," GC/FID analysis
A003c	1/test
Kiln ash	K_ln ash pi:	Composite
Scrjbbcr blow. Blowdown	Grsb (tap)
down water discharge
Total Cr
Cr( + 6)
Total Cr
Cr( + 6)
Paniculate
load
Total Cr
HCl
Cr( + 6)
Paniculate
load
HCl
Parr bomb ashing'',	1/tcst
Digestion by Method 3050°,
analysis by Method 6010°
Digestion by Method 3050	1/test
analysis bj Method 7191"
Digestion by Method 3060% 4/test
analysis by Method 7197°
Digestion by Method 3010	4/tes:
analysis by Method 7191°
Digestion by Method 3060s,	4/tes:
analysis by 719T
Method 5'	1/tes:
Digestion by Method 3010	1/test
(impingers) or Method 3050
(paniculate), analysis bv
Method 7191*
Analysis of impinger solution for Test 3 orJy
Cl by specific ion electrode
Digestion by Method 3060*, 1/test
analysis by Method 7197"
Method 5*	1/test
Analysis of impinger solution for Test 3 only
Cl by specific ion electrode
"(1)
'(2)
'(3)
V)
'(-)
26

-------
TABLE 10.
MULTIPLE METALS TRAIN IMPINGER SYSTEM REAGENTS
FOR THE CHROMIUM VALENCE STATE TESTS
Impingcr
Number	Reagent	Quantity
1
Empty

2
5 percent HN03 and 10 percent H202
100 mL
3
5 percent HN03 and 10 percent H202
100 mL
4
Silica ge!
750g
TABLE 11. Cr( +6) TRAIN IMPINGER
SYSTEM REAGENTS
Impinger
Number
Reagent
Quantity
1
0.1N NaOH
100 mL
0
A*
0.1N NaOH
100 mL
3
0.1N NaOH
100 mL
4
Silica gel
750g
27

-------
One composite sample of the feed clay material was also taken and preserved for
Cr(-t-6) analysis with 3 percent Na2C03, 2 percent NaOH solution.
3.1.2 Analysis Procedures
Table 12 summarizes the number of samples collected over the three-test chromium
valence state test series. Particulate load determinations from Method 5 train samples were
performed at the IRF in accordance with Method 5 protocol. Chloride determinations for HC1
measurement were performed on aliquots of appropriate impinger solutions via specific ion
electrode analysis at the IRF.
All samples for chromium analyses, preserved as noted in Section 3.1.1, were shipped
to the Acurex analytical laboratory in Mountain View, California. Total chromium analyses were
performed using graphite furnace atomic absorption (AA) spectroscopy via Method 7191 for all
samples except the composite feed samples (clay/organic liquid mixture). Inductively coupled
argon plasma spectroscopy (ICAP) via Method 6010 was used for this sample matrix. Solid
samples (probe wash, particulate, kiln ash) were digested by Method 3050 prior to analysis.
Liquid samples (spike solution, blowdown) were digested by Method 3010 prior to analysis. The
clay/organic liquid mixture samples were Parr bomb ashed, then digested by Method 3050 prior
to analysis.
Samples for Cr( + 6) analysis were digested by Method 3060. The solid samples (kiln
ash and feed clay) had been preserved after collection in the method digestion solution. For
these, digestion volumes were corrected and the method followed through. Digestates were
analyzed by Method 7197 (chelation/extraction). Graphite furnace atomic absorption
spectroscopy (in accordance with Method 7191) was used to analyze extracts from the Method
7197 procedure.
32 PARAMETRIC TRACE METAL TESTS
Table 13 summarizes the sampling and analysis matrix for the parametric trace metal
tests.
3.2.1 Sampling Procedures
As indicated in Table 13, the incinerator flue gas was characterized at three locations:
the afterburner exit, the scrubber exit, and the stack. Characterization at the afterburner and
scrubber exit locations supported test objectives. Stack gas sampling was performed to ensure
compliance with the IRF's operating permit.
The sampling protocols performed in the afterburner exit flue gas were designed to
measure flue gas particulate load and size distribution, HC1, trace metal vapor phase and
particulate emissions by particulate size range, and volatile organic hazardous constituent
emissions.
Volatile organic hazardous constituent emissions were sampled using the VOST
protocol. All other parameters listed above were measured in the afterburner exit flue gas
using a variation of a Method 17 train. The impingers used for the afterburner exit Method 17
sampling train are noted in Table 14. This impinger train is the same as that used for total
chromium in the chromium valence state tests, with the addition of a caustic impinger between
the initially empty impinger and the first HN03/H2O, impinger. After sampling, the contents
of impingers 1 and 2 were combined and aliquots obtained for HCI analysis. The impinger
28

-------
TABLE 12. SUMMARY OF CHROMIUM VALENCE STATE TEST SAMPLES
Sample type
Analyte
Number of
Each test
samples
Total
Feed clay
Cr( + 6)
—
1
Clay/organic liquid mixture
C-jHg, C2C14. C5HjC1
1
3

Ultimate analysis
1
3

Total Cr
1
3
Aqueous chromium spike
Total Cr
1
3
solution
Cr( + 6)
1
3
Kiln ash
Total Cr
1
3

Cr( + 6)
1
3
Scrubber blowdown
Total Cr
4
12

Cr( + 6)
4
12
Afterburner exit



Method 5 Train A



Probe wash
Paniculate, Total Cr
1
3
Filter
Particulate, Total Cr
1
3
1st impinger
Total Cr
1
3
2nd impinger
Total Cr
1
3
3rd impinger
Total Cr
1
3
Method 5 Train B



Is", impinger
Cr( + 6), Cr
1
3
2nd impinger
Cr( + 6), CI'
1
3
3rd impinger
Cr( + 6), Cr
1
3
Scrubber exit:



Method 5 Train A



Probe wash
Particulate, Total Cr
1
3
Filter
Particulate, Total Cr
1
3
1st impinger
Total Cr
1
3
2nd impineer
Total Cr
1
3
3rd impinger
Total Cr
1
3
Method 5 Train B



1st impinger
Cr(+6). CI"
1
3
2nd impinger
Cr( + 6), CI"
1
3
3rd impinger
Cr( + 6), CI'
1
3
Stack gas:



Method 5 Train



Probe wash + filter
Particulate
1
3
1st and 2nd impinger
cr
Test 3 only
1
29

-------
TABLE 13. SAMPLING AND ANALYSIS MATRIX FOR THE PARAMETRIC
TRACE METAL TESTS
Sample
Location
Analysis
Sampling
procedure Parameter
Method
Frequency
Cia>/organic
liquid
rurLre
Preparation
siorage drum
Trier
Aqueous
metal spie
sol j non
Kiin ash
Kilr. inkt
Kiln ash
C7H„ C.CI,
C4H,C1"
Ultimate
analysis
(C, H. O, CI)
Cr and Mg
Purge and trap by
Method 5030", GC/FID
analysis
A003e
Composite As and Pb
Ba. Cd, Cr,
Cu, and Mg
Bi and Sr
Composite As and Pb
Pan bomb ashing,
digestion by Method 3050,
ICAP analysis by Method 6010°
Pb	Parr bomb ashing, digestion by
Method 3050, furnace AA by
Method 7421°
Bi ard Sr Parr bomb ashing, digestion by
Method 3050, AA analysis by
300 series methods'
Digestion by Method 3010,
furnace AA by 7000 series
methods'
Digestion by Method 3010,
ICAP analysis by Method 6010°
Digestion by Method 3010°,
AA analysis by 300 series
methods'
Digestion by Method 30S0,
furnace AA by 7000 series
methods'
Ba, Cd, Cr. Digestion by Method 3050,
Cu, and Mg ICAP analysis by Method 6010°
Bi and Sr Digestion by Method 3050"
AA analysts by 300 series
methods'
1 composite/test
1 composite/test
] composite/test
1 composite/test
1 eompcsiie/tesi
1 composite/test
1 composite/tes:
1 eomposite/tes:
1 /test
1/test
1/test
'O
el3)
(continued)
30

-------
TABLE 13, (continued)
Analysis
Sampling
Samplt	Location	procedure Parameter	Mttbod	Frequency
Scrubber
blowflown
«-aier
Btowdown
discharge
Grab/tap As and Pb
Ba, C4 Cr,
Cu, and Mg
Bi and Sr
F.ue gss
Afterburner
exit
Method 17*
Particulate
load
Particulate
size
distribution
Digestion by Method 3010,	Al least 3/test
furnace AA by 7000 series
methods*
Digestion by Method 3010,	At least 3/itst
1CAP analysts by Method 6010"
Digestion by Method 3010®, At least 3/t«i
AA «»aly»is by 300 series
methods'
Method 17"	1 /test
ASME PTC 2T	1/test
Method
003CT
HCI
As and Pb
Ba, Cd Cr,
Cu, and Mg
B; and Sr
VoSatile
organic
hazardous
constituents
Analysis of impwger	1/test
solution for Cl by specific
ion clecrode
Digestion by Method 3010 or 1 /test
3050, furnace AA by 7000 series
methods* (tmpingers arid
paniculate by size)
Digestion by Method 3010 or 1/test
3050,1CAP analysis by Method
6010" (impmgers and paniculate
by size)
Digestion by Method 3010 or l/iest
3050s, AA analysis by 300 series
methods' (ur.pingers and
paniculate by size)
Thermal desorption, purge	3 trap pairs/
and trap by Method 5040°,	test
GC/FID analysis
*(1)
'Cl
(continued)
'W
31

-------
TABLE 13. (concluded)
Analysis
Sampling
Sample	Location	procedure Parameter	Method	Frequency
FIjc gas Scrubber
(continued) exii
L
Method 5'
Particulate
load
Method 5"
l/test


HC!
Analysis of ur.pir.ger
solution for CI by specific
ion electrode
1/iesi


As and Pb
Digestion by Method 3010 or
3050, furnace AA by 7000 series
methods' (impingers and
particulate)
l/test


Ba, Cd. Cr,
Cu. and Mg
Digestion by Method 3010 or
3050, 1CAP ar.aiysis by Method
6010" (impingers and particulate)
l/test


Bi and Sr
Digestion by Method 3010 or
3050*, AA analysis by 300 series
methods'' (impingers and
paniculate)
l/test

Method 0030°
Volatile
organic
hazardous
constituents
Thermal desorp'.ior., purge and
trip by Method 5040", GC/FID
analysis
Thermal desorption and trap
GC/MS analysis b>
Method 5(U(r
3 trap pairs/tes:
3 Trap pairs,
Test 4 only
Stack gas Stack
dou-nstrcarr,
of carbon
bed/HE Pa
filter
Method 5'
Paniculate
load
HCI
Method 5*
Analysis of impinger solution for
CI by specific ion electrode
l/test

Method
0030s
V'oUtile
organic
hazardous
constituents
Thermal desorption, purge and
trap by Method 5040s, GC/FID
analysis
3 trap pairs/
test, Tests 8,
9, and 1J only
'(J)
'C5;
32

-------
TABLE 14. MULTIPLE METALS TRAIN IMPINGER SYSTEM
REAGENTS FOR THE PARAMETRIC TRACE
METAL TESTS
Impinger
Number
Reagent
Quantity
Empty
2
0.1 N NaOH
100 mL
4
3
5 percent HN03 and 10 percent H202 100 mL
5 percent HN03 and 10 percent H2Oz 100 mL
5
Silica gel
750g
contents were preserved to a pH of less than 2 with HN03 for later metal analysis. The
sampling probe was washed with acetone. The probe wash was desiccated and weighed, then
resuspended in acetone for later metal analysis.
The modified Method 17 train at the afterburner exit also collected particulate in an
oversized alundum thimble placed within the sampling probe. After obtaining the total
particulate weight, the particulate in these samples was divided according to the terminal velocity
in air using a centrifugal classifier. This device is described in the ASME Power Test Code 28
(6). Weights of the resulting eight size cuts were recorded. These samples were later
recombined into four size cuts and subjected to metals analysis to provide data on metal
distribution by particulate size.
The sampling protocol performed in the scrubber exit flue gas was designed to measure
the same parameters measured in the afterburner exit flue gas, with the exception that particle
sizing and metals analysis by particle size were not performed. Accordingly, VOST (Method
0030) sampling of the scrubber exit was performed. However, instead of a modified Method 17
train, a Method 5 train was run in the scrubber exit. The particulate levels in the scrubber exit
were expected to be significantly lower than in the afterburner exit, making the collection of
sufficient particulate difficult. In addition, the high water content of the scrubber exit flue gas
would cause a moist and agglomerated particulate catch. Classification of this catch would likely
not be representative of the actual particulate size distribution in the scrubber exit flue gas.
The impingers used in the scrubber exit Method 5 train were the same as those used
in the afterburner exit Method 17 train. After sampling, the filter and probe wash (acetone)
from the Method 5 train were recovered and desiccated to constant weight per the Method 5
procedure. The probe wash was then resuspended in acetone for later metal analysis. Impinger
collection, aliquoting, combining, and preservation for the scrubber exit Method 5 train were
exactly the same as for the afterburner exit Method 17 train discussed above.
33

-------
A Method 5 train was used to sample the stack gas for each test to ensure compliance
with the IRF operating permit. Particulate load and HC1 levels were measured. A Method 5
train with impinger contents as noted in Table 11 was used at this location. In addition, VOST
(Method 0030) sampling was performed at the stack location for Tests 8, 9, and 11 to supply
continuing data on carbon bed organic component removal efficiency.
As for the chromium valence state tests, all scrubber exit Method 5 train sampling
collected at least 2.8 m3 (100 ft3) of flue gas over a nominal 3-hour sampling period.
Afterburner exit Method 17 train sampling collected at least 19.8 m3 (700 ft3) of flue gas over
the same nominal 3-hour sampling period. This period began no less than 1/2 hour after the
start of test mixture feed. Test mixture feed continued until all sampling was completed.
In addition to the flue gas sampling described above, samples of the clay/organic liquid
feed mixture, the aqueous trace metal spike solution, the kiln ash, and the scrubber blowdown
were collected for analysis. Composite feed samples from each drum were collected by trier
sampling at three locations in the drum cross-section. Each of these composites were ultimately
analyzed for toluene, tetrachloroethylene, chlorobenzene, bismuth, chromium, lead, magnesium,
strontium, and ultimate composition. Samples for volatile organic analysis were collected from
each drum's composite sample when the contents of that drum were first fed into the kiln.
These samples were sealed in 40 mL VOA vials at that time. An aliquot of each drum's
composite sample was also submitted for ultimate analysis and bismuth, chromium, lead,
magnesium, and strontium analysis. Previous data suggested that clay background levels of
chromium and lead might be greater than 10 percent of the final feed material (clay/organic
liquid plus aqueous metal spike solution) concentration. Thus, the contribution of the clay to
final concentrations of chromium and lead was measured. No data on bismuth, magnesium,
and strontium concentrations in the clay existed, so analyses of the clay/organic Liquid for these
were also included in the analysis protocol.
A composite sample of each test's aqueous metal spike solution was also collected for
trace metal analysis. These samples were preserved with HN03 to pH<2.
Blowdown sampling was performed as discussed in Section 3.1.1 for the chromium
valence state tests. To reiterate, a grab (100 mL) sample was taken at the start of Method 17/5
sampling and hourly thereafter until sampling was completed. A final blowdown sample was
taken at the end of the Method 17/5 sampling period. Each individual blowdown sample was
preserved with HN03 to pH<2.
A composite kiln ash sample was collected from the ash pit after the completion of each
test. The ash pit contents were weighed and a clean, empty ash collection drum was placed in
the collection pit for the subsequent test.
2>22 Analysis Procedures
Table 15 summarizes the number of samples collected over the eight-test parametric
trace metal program. Particulate load determinations from Methods 17 and 5 samples were
performed at the IRF, in accordance with respective method procedures, prior to size
classification and combination for trace metal analysis as discussed in Section 3.2.1. Chloride
analyses for determining HC1 emissions were performed on aliquots of appropriate impinger
solution combinations via specific ion electrode analysis at the IRF.
Analysis of VOST (Method 0030) traps was performed at the IRF by thermal
desorption/purge and trap GC/FID analysis. Thermal desorption/purge and trap was in
34

-------
TABLE 15. SUMMARY OF PARAMETRIC TRACE METAL TEST SAMPLES
Number of samples
Sample type
Analyte
Each test
Total
Clay/organic liquid feed
c7h8, C2C14, c6h5ci
1 to 3
26

Ultimate analysis
(C, H, O, CI)
I to 3
23

Bi, Cr, Pb, Mg, Sr
1
8
Aqueous metal spike solution
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg, Sr
1
8
Kiln ash
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg, Sr
1
8
Scrubber blowdown
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg, Sr
3 to 5
39
Afterburner exit flue eas:
VOST (Method 0030)
Sample trap pair
Volatile organic
hazardous constituents
3
24
Field blank trap pair
Volatile organic
hazardous constituents
1
8
Method 17 train
<2 /xm particulate
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg, Sr
I
8
2 to 4 nm particulate
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg, Sr
1
8
4 to 10 nm particulate
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg, Sr
1
8
> 10 fim particulate
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg, Sr
1
8
1st and 2nd impingers
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg, Sr, CI"
1
8
(continued)
35

-------
TABLE 15. (concluded)


Number of samples
Sample type
Analyte
Each test
Total
3rd impinger
As.. Ba, Cd, Cr, Pb,
Bi, Cu, Mg, Sr
1
8
4th impinger
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg, Sr
1
8
Scrubber Exit Flue eas:
VOST (Method 0030)
Sample trap pair
Volatile organic
hazardous constituents
3
24
Field blank trap pair
Volatile organic
hazardous constituents
1
8
Method 5 train:
Probe wash
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg, Sr
]
8
Filter
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg. Sr
1
8
1st and 2nd impinger
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg. Sr, CI'
1
8
3rd impinger
As, Ba, Cd, Cr, Pb,
Bi, Cu, Mg. Sr
1
8
4th impinger
As, Ba, Cu, Cr, Pb,
Bi, Cu, Mg, Sr
1
8
Stack cas:
VOST (Method 0030)
Sample trap pair
Volatile organic
hazardous constituents
3 (Tests 5,
8, 11 only)
9
Field blank trap pairs
Volatile organic
hazardous constituents
1 (Tests 5,
8, 11 only)
3
Method 5 train



1st and 2nd impinger
cr
1
8
3rd impinger
cr
1
8
36

-------
accordance with Method 5040. Analysis was by capillary column GC/FID. The 22 volatile
organic compounds routinely determined via this method at the IRF were analyzed. These
compounds are Listed in Table 16.
The clay/organic liquid feed composite samples were also analyzed at the IRF for
toluene, teirachloroethylene, and chlorobenzene via purge and trap GC/FID. Purge and trap
was in accordance with Method 5030. Composite clay/organic liquid feed samples were sent to
Galbraith Laboratories in Knoxville, Tennessee, for ultimate analysis.
All samples for trace metal analysis were preserved as noted in Section 3.2.1, and
shipped to the Acurex analytical laboratory in Mountain View, California for analysis. Arsenic
and lead analyses were performed via graphite furnace atomic absorption (AA) spectroscopy:
arsenic by Method 7060 and lead by Method 7421. Barium, cadmium, chromium, copper, and
magnesium analyses were performed by ICAPvia Method 6010. Bismuth and strontium analyses
were performed by flame AA: bismuth by Method 303a and strontium by Method 326a (7).
Samples were digested appropriately prior to analyses. Method 3050 was used for solid samples;
Method 3010 was used for aqueous liquid samples. Clay/organic liquid samples were Parr bomb
ashed (4) prior to digestion.
TABLE 16. VOLATILE ORGANIC COMPOUNDS ROUTINELY
ANALYZED BY GC/FID AT THE IRF
Methylene chloride
Benzene
1,1-Dichloroethane
1,1,2-Trichloroethane
t-l,2-Dichloroethylene
Hexane
Chloroform
Bromoform
1,2-Dichioroethane
Tetrachloroethylene + Tetrachloroethane
1,1,1-Trichloroethane
Toluene
Carbon tetrachloride
Chlorobenzene
Bromodichloromethane
Ethyl benzene
1,2-Dichloropropane
1,3-Dichlorobenzene
t-l,3-Dichloropropylene
1,2-Dichlorobenzene
Trichloroethvlene
••
1,4-Dichlorobenzene
37

-------
SECTION 4
TEST RESULTS
Results from the test program are discussed in this section. The discussion is subdivided
as follows: measured synthetic waste feed composition, continuous emission monitoring data,
particulate and HC1 emissions, chromium valence state test results, parametric trace metals test
results, POHC DREs, and PIC emissions.
4.1 SYNTHETIC WASTE FEED COMPOSITION
Section 2 discussed the targeted synthetic waste feed composition planned for each test
so that test program objectives could be attained. Section 3 noted that the various waste feed
components were sampled during each test and analyzed to verify the actual composition of the
feeds prepared. This section discusses results of the actual feed analyses.
Table 17 summarizes the clay/organic liquid mixture POHC concentrations measured
for each test's composite sample and compares these to target values. The data in Table 17
confirm the close agreement between measured and target composition for all tests.
Table 18 summarizes the analyzed aqueous spike solution concentrations for Tests 1
through 3, the chromium valence state tests. Table 19 provides a similar summary for Tests 4
through 11, the parametric trace metal tests. As indicated, with one possible exception, analyzed
concentrations were comparable to prepared concentrations. The exception was Test 5 in which
the analyzed aqueous spike metal concentrations were lower than the prepared concentrations
for all the metals. The reason for this exception is not known.
Available data on the clay matrix composition from a previous analysis indicated that
the clay material contained less than 1 mg/kg of arsenic and cadmium and about 4 mg/kg
copper. These values were negligible compared to the target integrated feed concentrations
noted in Section 2 for these elements. These same data also indicated that the clay matrix
contained about 20 mg/kg barium, 25 mg/kg lead, and 50 mg/kg chromium. These values are
sufficiently high that the background clay matrix contribution to integrated feed concentrations
would not be negligible. No previous data on bismuth, strontium, and magnesium concentrations
were available. As a consequence, the feed clay/organic liquid mixtures for the parametric trace
metals tests were analyzed for bismuth, chromium, lead, magnesium, and strontium, and a
composite of the Test 1 and 2 clay/organic liquid mixture and the Test 3 mixture were analyzed
for chromium to clearly determine the clay contribution to integrated feed metal concentrations.
Results of these analyses are summarized in Table 20.
As indicated in Table 20, the analysis results for 10 separate feed samples were very
comparable. The average values noted in the table were, thus, used to calculate the clay
contribution to the integrated feed concentration for each metal. Bismuth was only analyzed for
in one feed matrix sample. At the 12 ppm level noted for this sample in Table 20, the clay
38

-------
TABLE 17. POIIC CONCENTRATIONS IN CLAY/ORGANIC LIQUID FEE!) SAMPLES
Weight percent in mixture





Chlorine
Test
Test date
Toluene
Tet rach loroet hylcne
Chlornbenzene
content"
Mixture 1





Target





Composition

28.6
0
0
0
Measured composition





1
9/29/88
27.9
0
0
0
2
9/28/88
27.9
0
0
0
4
9/14/88
232
0
0
0
Mixture 2





Target





Composition

21.7
3.4
3.4
4
Measured composition





5
8/25/88
16.7
3.0
3.6
3.7
6
9/16/88
20.5
3.6
3.5
4.2
7
8/30/88
19.7
3.2
3.2
3.8
8
9/07/88
17.1
3.1
3.0
3.6
9
9/09/88
16.5
2.9
2.9
3.4
10
9/20/88
22.5
3.9
3.8
4.6
Mixture 3





Target





Composition

14.9
6.9
6.9
8
Measured composition





3
9/26/88
15.9
7.5
6.7
8.5
11
9/22/88
14.6
7.1
6.9
8.3
"Based on measured tetrachloroethylene and chlorobenzene concenlralions.

-------
TABLE 18. AQUEOUS SPIKE SOLUTION CHROMIUM CONCENTRATIONS FOR THE CHROMIUM
VALENCE STATE TESTS.
Aqueous spike solution concentration (ppm Cr)
Test 1 (9-29-88)	Test 2 (9-28-88)	Test 3 (9-26-88)
Cr valence
state Compound Prepared Analyzed	Prepared Analyzed	Prepared	Analyzed
Cr( + 3) Cr(NO,)j 9H20 — —	3400 —	3400	—
Cr( + 6) CrO, 3400 2700	— 0.93	—	2.6
Total Cr 3400 3000	3400 3100	3400	3200

-------
TABLE 19. AQUEOUS SPIKE SOLUTION METALS CONCENTRATIONS FOR THE
PARAMETRIC TRACE METALS TEST SERIES
Prepared spike Analyzed spike solution concentration (ppm metal)
solution 	
concentration Test 4 TestS Test 6 Test 7 Test R Test 9 Test 10 Test II
Metal	Compound	(ppm as metal) (9/14/88) (8/25/88) (9/16/88) (8/30/88) (9/07/88) (9/09/88) (9/20/88) (9/22/88)
Arsenic
AsjO,
3,230
2,800
2,300
2,900
2,400
2,400
2,300
3,600
3,300
Barium
Ba(NOj)j
2,580
2,300
1,800
2,400
2,400
2,500
2,500
2,500
2,300
Bismuth
Bi(NO,), . 5HjO
9,040
10,000
6,600
11,000
7,400
9,400
11,000
11,000
11,000
Cadmium
Cd(NO,),
650
605
410
540
530
520
540
640
540
Chromium
Cr(NO,)j . 9H,0
3,230
3,100
2,300
2,900
2,900
3,100
3,000
3,300
2,900
Copper
Cu(NOj),. 3HjO
32,300
29,000
23,000
30,000
30,000
32,000
31,000
30,000
30,000
Lead
PMNO,),
3,230
3,400
2,400
3,000
3,100
3,200
3,100
3,500
3,000
Magnesium
Mg(NOj), 6H,0
5,040
4,600
3,400
4,500
4,300
4,400
4,600
4,900
4,300
Strontium
Sr(NO,),
16,100
18,000
11,000
15,000
14,000
15,000
16,000
19,000
17,000

-------
TABLE 20. CLAY MATRIX METALS CONCENTRATIONS
Concentration (nig/kg clay)*
Metal
Test 1/2
composite
(9-28/29-88)
Test 3
(9-26-88)
Test 4
(9-14-88)
Test 5
(8-25-88)
Test 6
(9-16-88)
Test 7
(8-30-88)
Test 8
(9-7-88)
Test 9
(9-9-88)
Test 10
(9-20-88)
Test 11
(9-22-88)
Average
Bismuth
NA*
NA
NA
12
NA
NA
NA
NA
NA
NA
12
Chromium
54
52
55
55
54
53
54
55
50
50
53
Lead
NA
NA
3.2
3.7
3.5
2.9
3.1
23
3.0
2.5
3.0
Magnesium
NA
NA
23,000
23,000
22,000
22,000
22,000
22,000
20,000
22,000
22,000
Strontium
NA
NA
33
37
30
32
33
30
33
35
34
'Clay matri* residue after Parr bomb ashing.
bNA: not analyzed.

-------
matrix would have a minor effect on the integrated (clay/organic liquid plus aqueous spike
solution) feed bismuth concentration. Thus, other clay/organic liquid samples were not analyzed
for bismuth. Barium analyses of the feed samples were inadvertently not performed. The
previous analysis data (20 ppm Ba) was used to calculate the clay contribution to the integrated
feed concentration for barium. A composite clay sample was analyzed for Cr( + 6). The result
was a Cr( + 6) concentration of 2.3 ppm.
Table 21 combines the clay matrix analysis results from Table 20 with the aqueous spike
solution data from Table 18 to give the chromium concentration in the integrated feed for the
chromium valence state tests (Tests 1 through 3). Table 22 provides a similar summary (from
the data in Tables 19 and 20) for the metals concentrations in the integrated feeds for the
parametric trace metal tests (Tests 4 through 11). The clay/organic liquid feedrate data in
Table 4 combined with the aqueous spike solution feedrate for each test were used to ratio
individual component concentrations to give the integrated feed concentrations in Tables 21
and 22.
42 CONTINUOUS EMISSION MONITORING DATA
Table 23 summarizes the continuous emission monitor (CEM) data obtained for each
of the tests performed. The values in the table were developed from the data recorded by the
PC data acquisition system. The data in the table show that throughout the test program CO
levels at the afterburner and scrubber exits remained below the detection limit of the monitors
(5 ppm) CO spikes of up to 200 ppm accompanied a few of the flameouts experienced.
However, as noted above, sampling was suspended during these periods. The CO/C02 monitor
at the stack operated continuously; however, it performed erratically and had excessive drift.
Reliable stack CO and C02 data were, therefore, not obtained. The C0/C02 monitor at the
scrubber exit functioned well, providing reliable data for both CO and C02 emissions. Flue gas
dilution by inleaking air between the afterburner exit and scrubber exit is evidenced by the
reduction in C02 concentration. Finally, average NOx concentrations at the scrubber exit ranged
from 35 to 76 ppm.
TABLE 21. INTEGRATED FEED CHROMIUM CONCENTRATIONS FOR
THE CHROMIUM VALENCE STATE TESTS
Test 1 Test 2 Test 3
(9-29-88) (9-28-88) (9-26-88)
Total Cr
Concentration (ppm)
Cr( + 6)
Concentration (ppm)
Percent of total Cr
79 84 85
41 1.6 1.6
52 2 2
43

-------
TABLE 22. INTEGRATED FEED METAL CONCENTRATIONS FOR THE PARAMETRIC TRACE
METALS TEST SERIES
Metal concentration (ppm)
Test 4	Test 5	Test 6	Test 7 Test 8 Test 9 Test 10 Test 11
Metal	(9/14/88) (8/25/88) (9/16/88) (8/30/88) (9/7/88) (9/9/88) (9/20/88) (9/22/88)
Arsenic
25
23
23
24
19
19
32
33
Barium
36
33
33
39
36
36
36
37
Bismuth
98
75
95
84
85
99
110
120
Cadmium
5
4
4
5
4
4
6
5
Chromium
68
64
61
68
66
66
66
67
Copper
260
230
240
310
260
260
270
310
Lead
33
26
26
34
28
28
33
32
Magnesium
17,700
17,700
16,700
17,000
17,700
17,900
16,100
16,500
Strontium
190
140
140
170
150
160
200
200

-------
TABLE 23. CEM DATA
Parameter
Test 1
(9/29/88)
Test 2
(9/28/88)
Test 3
(9/26/88)
Test 4
(9/14/88)
Test 5
(8/25/88)
Kiln exit
02, range, percent	10.7-11.8	10.9-11.5	11.5-13.2	11.8-12.8	11.0*13.6
average, percent	11.2	11.2	11.5	12.2	11.9
Afterburner exit
O,, range, percent	6.6-7.9	6.4-7.7	6.3-7.7	6.1-8.0	6.0-10.2
average, percent	7.3	7.3	7.2	7.1	8.4
CO,, range, percent	8.9-10.1	9.1-10.2	9.0-10.2	8.6-10.2	6.7-10.0
average, percent	9.5	9.5	9.5	9.5	8.2
CO, ppm	<5	<5	<5	<5	<5
Scrubber exit
C02, range, percent	5.8-6.7	5.8-6.7	4.8-6.7	4.5-6.9	4.5-6.6
average, percent	6.3	6.2	6.3	6.3	5.7
CO, ppm	<5	<5	<5	<5	<5
NOx, range, ppm	33-41	32-37	33-45	33-56	40-57
average, ppm	37	35	41	44	49
Stack
0:, range, percent	11.5-12.3	11.5-12.1	11.4-12.6	11.4-13.1	11.6-13.5
average, percent	11.9	11.8	11.9	11.7	12.1
(continued)
45

-------
TABLE 23. (concluded)
Parameter
Test 6
(9/16/88)
Test 7
(8/30/88)
Test 8
(9/7/88)
Test 9
(9/9/88)
Test 10
(9/20/88)
Test 11
(9/22/88)
Kib exit






02, range, percent
9.0-11.0"
11.7-13.7
12.8-14.5
12.1-13.0
10.6-12.7
10.8-31.7
average, percent
10.6
12.4
13.6
12.6
11.4
11.4
Afterburner exit






Oj, range, percent
5.7-7.8
7.0-8.7
3.7-7.5
6.3-7.7
6.5-8.5
6.3-8.1
average, percent
7.1
7.9
5.9b
7.6
7.6
7.5
CO:. range, percent
8.7-10.4
7.6-9.4
8.9-11.8
7.7-9.0
8.5-10.0
8.5-10.0
average, percent
9.4
8.8
10.2
8.3
9.2
9.1
CO. ppm
<5
<5
<5
<5
<5
<5
Scrubber exi;






CO,, range, percent
4.6-5.9
4.1-6.4
5.1-7.2
4.5-6.1
5.3-7.1
4.8-6.8
average, percent
5.5
5.7
6.7
5.8
6.5
6.1
CO. ppm
<5
<5
<5
<5
<5
<5
NOx. range, ppm
45-63
35-58
59-89
32-48
43-64
39-56
average, ppm
56
50
76
42
55
48
Stack






O,. range, percer.t
11.8-13.0
11.5-14.0
10.6-11.9
12.3-13.4
13.1-14.1
11.1-12.8
average, percent
12.2
12.6
11.1
12.6
13.7
11.9
8CEM not in service. Data obtained using a Bacharach Fryrite 02 analyzer.
•"Afterburner air feedrate at fan capacity-.
46

-------
As mentioned in Section 1, a secondary test objective was to compare heated and
unheated hydrocarbon analyzers. However, hydrocarbon emissions throughout the test program
remained below the instrument detection limit of 1 ppm (except for brief periods just after a
flameout). Thus, useful information for comparing analyzer performance was not obtained.
43 FLUE GAS PARTICULATE AND HC1
Flue gas particulate emissions were measured at the afterburner exit, scrubber exit, and
stack. Particulate data are given in Table 24. HC1 data are given in Table 25.
As shown in Table 24, stack particulate concentrations ranged from 29 to 58 mg/dscm
at 7 percent 02. All levels were below the hazardous waste incinerator performance standard
of 180 mg/dscm at 7 percent 02. Tests 1, 2, and 3 particulate data indicate the levels were
lower at the afterburner exit than at the scrubber exit. Low afterburner values are likely a
sampling artifact related to the difficulty in achieving isokinetic sampling at the afterburner exit
due to the limited distance between flow disruptions. Samples taken at the afterburner exit that
depend on isokinetic sampling must, therefore, be interpreted cautiously. Stack particulate
emissions are generally greater than scrubber exit levels, most likely due to particulate
entrainment as the gas passes through the carbon bed absorber. The greatest confidence is
placed in the scrubber exit particulate concentrations where the requirements for isokinetic
sampling are met.
As shown in Table 25, stack HC1 levels were nondetectable at detection limits up to
6.6 mg/dscm for Tests 1 to 10. Corresponding HC1 emission rates were less than 12.4 g/hr for
these tests. Stack HC1 levels were measurable at 12.0 mg/dscm (24.1 g/hr) for Test 11, one of
the two tests at the highest feed chlorine content. All HC1 stack emissions were substantially
below the hazardous waste incinerator waste incinerator performance limit of 1.8 kg/hr.
Table 25 also includes information on chlorine feedrates and afterburner and scrubber
exit HC1 emission rates. For four of the nine tests in which flue gas HC1 levels were measured,
afterburner exit discharge rates agreed well with chlorine feedrates, as expected. However, for
five tests, agreement was poor. The lack of correlation between chlorine feedrate and
afterburner exit flue gas HC1 emission rate for these five tests may be due to sampling problems
associated with the flow distribution at the afterburner exit. The nonzero value measured in the
afterburner exit for Test 4 (0 percent chlorine feed) may be due to hysteresis emissions of
chlorides deposited during earlier tests. In any event, the afterburner exit HC1 level for Test 4
is significantly lower than the values measured for the other tests.
It is interesting to note that the scrubber exit HC1 levels for the two tests at the highest
feed chlorine concentration (Tests 3 and 11) were above detection limits and comparable.
Scrubber exit HC1 levels for all but one of the other tests at lower feel chlorine concentrations
were nondetectable.
4.4 TRACE METAL DISCHARGE DATA
This section discusses the distributions of the trace metals fed to the RKS in each test
among the discharge streams sampled and analyzed. Distributions discussed are based on the
mass flowrate of a given element in a specified discharge. This mass flowrate is calculated from
the analyzed concentration of each metal in a given stream (e.g., kiln ash, flue gas particulate)
and the total mass flowrate for the given stream. Appendix B includes all the laboratory
analysis reports which serve as the basis for the given stream concentrations. Appendix C
contains the flue gas sampling train data which includes flue gas stream flowrate information.
47

-------
TABLE 24. FLUE GAS PARTICULATE DATA
Particulate concentration
(mg/dscm at 7 percent 02)

Test
Afterburner
Scrubber

Test
date
exit
exit
Stack
1
9/29/88
21
31
58
2
9/28/88
23
32
42
3
9/26/88
36
37
39
4
9/14/88
273
7
44
5
8/25/88
360
7
31
6
9/16/88
162
31
27
7
8/30/88
146
23
29
8
9/07/88
50
40
40
9
9/09/88
152
12
33
10
9/20/88
414
21
47
11
9/22/88
240
51
38
48

-------
1
2
3
4
5
6
7
8
9
10
11
TABLE 25. HCI EMISSIONS DATA
MCI concentration and emission rate
Afterburner exit	Scrubber exit	Stack
Test CI Teed rate
date (kg/hr) (g/dscm) (kg/hr) (mg/dscm) (g/hr) (mg/dscm) (g/hr)
9/29/88
0




<5.8
<11.9
9/28/88
0




<4.7
<10.5
9/26/88
5.23
10.07
11.42
7.6
17.8
<5.6
<12.4
9/14/88
0
0.13
0.12
<3.9
<7.1
<6.6
<8.9
8/25/88
2.35
8.93
8.30
<3.4
<6.4
<5.7
<8.0
9/16/88
2.64
2.37
2.18
<3.9
<7.9
<5.8
<12.2
8/30/88
2.40
2.51
2.35
<4.2
<8.4
<5.8
<8.9
9/07/88
2.28
0.54
0.49
7.6
17.7
<6.6
<9.8
9/09/88
2.12
1.16
1.11
<4.1
<7.7
<6.4
<9.1
9/20/88
2.86
2.72
2.52
<2.3
<5.1
<5.8
<11.8
9/22/88
5.24
2.91
2.64
7.5
15.5
12.0
24.1

-------
Table 26 summarizes the synthetic waste feedrate and kiln ash discharge flowrate data.
It is interesting to note in Table 26 that the fraction of the clay matrix mass fed to the kiln
accounted for by the kiln ash discharge was relatively constant at about 80 percent (range of 73
to 83 percent) for all tests. The remaining 20 percent of the clay fed represents ash entrained
in the kiln exit flue gas and carried out of the kiln into the afterburner in addition to the
moisture content of the clay.
The mass flowrate of a given metal removed by the venturi/packed column scrubber
deserves some discussion. Prior to each test, the RKS was operated for a minimum of 18 hours
fired on propane alone. Following this period, synthetic waste feed was started. Flue gas
sampling was begun only after at least half an hour of waste feeding had elapsed. At the
conclusion of flue gas sampling, synthetic waste feeding was stopped and the RKS was returned
to propane firing. Kiln rotation continued until all ash was discharged from the kiln into the
ash collection pit.
During propane firing, scrubber blowdown continued with fresh water scrubber makeup
introduced to keep the scrubber liquor loop "full". As a consequence, the scrubber loop was
"purged" of a previous test's metal buildup during propane-only firing such that, at the beginning
of a subsequent test, the scrubber liquor was clean; it contained very little to no test trace metal
content. Thus, over the time span of a given test, the recirculating scrubber liquor metal
concentrations increased from near zero to some nonzero value. Further, at the end of a test,
a significant portion of the metal fed to the RKS during the test which was removed by the
scrubber remained in the 760 L (200 gal) scrubber liquor loop.
To account for the nonsteady scrubber loop operation over a test resulting from the
above mode of operation, the scrubber blowdown sampling protocol specifically called for
obtaining a scrubber blowdown sample at the beginning of a test, hourly thereafter, and at the
end of a test. This allowed the buijdup of metal concentrations in the scrubber liquor to be
determined.
The mass flowrate of metal removed by the scrubber was then calculated as the sum of
two terms. The first was that removed in the scrubber blowdown. This term was calculated
from the blowdown rate noted in Table 6 and the average blowdown concentration measured
from the several blowdown samples analyzed. The second term was that remaining in the 760 L
(200 gal) scrubber recirculation loop at the end of a test. This term was based on the metal
concentration in the last blowdown sample taken.
The following subsections summarize trace metal distributions among discharge streams
for the chromium valence state and parametric trace metal tests, respectively.
4.4.1 Chromium Valence State Tests
As noted in Section 1, the objectives of this particular test series were to identify the
relative distributions of chromium in the RKS discharges and to determine the fraction
discharged as hexavalent chromium (Cr( + 6)). The parameters varied were feed chromium
valence state and feed chlorine content.
Table 27 summarizes the total chromium discharge distributions expressed as a fraction
of the amount of chromium present in the feed. Almost half of the feed chromium was present
as background in the feed clay absorbent. Overall mass balance closure around the system for
Tests 1 and 2 determined at the two locations where the flue gas was sampled was
50

-------
Test
1
2
3
4
5
6
7
8
9
10
11
TABLE 26. SYNTHETIC WASTE FEEDRATES AND KIEN ASH DISCHARGE RATES
Synthetic waste feedrate,
kg/hr (Ib/hr)	Kiln Ash
Clay/organic	Clay	Discharge rate, Fraction of clay
Test data	liquid	Traction	kg/hr (Ib/hr) feedrate (%)
9-29-88
62.6 (138.0)
45.1 (99.5)
37.4 (82.6)
83
9-28-88
62.0 (136.7)
44.7 (98.5)
34.9 (76.8)
78
9-26-88
61.3 (135.1)
42.9 (94.5)
31.3 (69.0)
73
9-14-88
59.4 (131.0)
45.6 (100.6)
34.7 (76.5)
76
8-25-88
63.5 (140.0)
48.5 (107.0)
37.8 (83.5)
78
9-16-88
62.6 (138.0)
45.4 (100.0)
33.6 (74.0)
74
8-30-88
63.4 (139.8)
46.7 (103.0)
37.8 (83.4)
81
9-7-88
63.4 (139.8)
48.5 (106.9)
38.3 (84.5)
79
9-9-88
62.5 (137.8)
48.5 (106.9)
39.8 (87.7)
82
9-20-88
62.8 (138.4)
43.9 (96.7)
34.7 (76.4)
79
9-22-88
63.5 (140.0)
45.4 (100.0)
34.5 (76.0)
76

-------
TABLE 27. TOTAL CHROMIUM DISCHARGE DISTRIBUTIONS AND MASS BALANCE FOR THE
CHROMIUM VALENCE STATE TEST SERIES
Total Cr Fraction (percent of feed)
Test 1
Cr( + 6) feed,
no CI
Test 2
Cr( + 3) Teed,
no CI
Test 3
Cr( + 3) feed,
8.5% feed CI
Feed: Clay mixture	48.4	45.4	43.8
Aqueous spike	51.6	54.6	56.2
Total	100.0	100.0	100.0
Kiln ash	74.8	73.6	46.7
Afterburner exit flue gas	0.4	0.5	0.8
Around kiln and afterburber	75.2	74.1	47.5
Kiln ash	74.8	73.6	46.7
Scrubber exit flue gas	1.1	1.3	2.3
Scrubber liquor	2.5	2.3	6.4
Around kiln and scrubber
78.4
77.2
55.4

-------
approximately 75 percent. Closure for Test 3 was poorer, 47 percent around the system with
an afterburner flue gas discharge and approximately 55 percent around the system with the
scrubber discharges.
Most of the chromium fed to the RKS is accounted for by the kiln ash discharge, A
relatively constant, and very small, amount on percentage of feed basis was measured in the
afterburner exit discharge. However, as noted in several instances in the preceding discussion,
the afterburner exit flue gas sampling location is very near a major flow disturbance. Thus, it
is very difficult to obtain a representative flue gas sample at this location. The lower total
chromium fractions measured in the tests are probably influenced by the poor sampling
characteristics of the available sampling location.
The fractions of total chromium feed accounted for in the scrubber exit flue gas and in
the scrubber liquor were comparable for Tests 1 and 2 in which the feed contained no chlorine.
However, these fractions were increased for Test 3 with chlorine-containing feed. This increase
occurs in absolute terms despite a decrease in the overall mass balance around the kiln and
scrubber for this test. This is consistent with expectations if the presence of chlorine gives rise
to more volatile chromium species and more water soluble chromium species.
Table 28 summarizes the normalized total chromium discharge distributions around the
RKS closed around the afterburner exit and closed around the scrubber discharges. Two
discharge streams are considered with closure around the afterburner exit: the kiln ash
discharge and the afterburner exit flue gas. Three discharges are considered with closure
around the scrubber discharges: the kiln ash discharge, the scrubber liquor, and the scrubber
exit flue gas. The distributions have been normalized with the total amount of chromium
measured in the sum of the discharge streams considered for each closure case. This has the
effect of normalizing discharge distributions to what they would have been if mass balance
closure were 100 percent for each closure case.
The data in Table 28 emphasize the point noted above that the kiln ash discharge
accounts for most of the chromium fed. In addition, as also noted above, the fraction of total
chromium in the scrubber exit flue gas and the scrubber liquor were comparable for Tests 1 and
2 in which the feed contained no chlorine. No change with chromium feed valence state is seen.
However, these fractions increase for Test 3 with chlorine-containing feed. As noted above, this
is consistent with expectation if the presence of chlorine in the incinerator gives rise to more
volatile chromium species.
Interestingly, the apparent scrubber chromium collection efficiency (scrubber liquor
fraction/(scrubber liquor fraction + scrubber exit flue gas fraction)) is relatively constant at 63
to 73 percent.
Table 29 shows the distribution of total chromium between the particulate phase and
the vapor/dissolved phase in the flue gas at the two locations sampled. The particulate phase
fraction represents that accounted for in the probe wash and filter catch of the sampling train.
The vapor/dissolved phase fraction represents that accounted for in the impingers of the
sampling train. Presumably, any chromium present in the vapor phase at the sampling train
probe exit would be captured and accounted for in the impingers. In addition, much of the
chromium present as water soluble salts would also be collected in the impinger. The flue gas
at both sampling locations contains significant water vapor. Thus, water soluble salts can be
"washed" through the glass fiber filter in the sampling train and be captured in the impingers.
53

-------
TAIILE 28. TOTAL CHROMIUM DISCHARGE DISTRIHUTIONS FOR THE CHROMIUM VALENCE
VALENCE SI ATE TEST SERIES
Total Cr fraction (percent of measured)
Test I	Test 2	Test 3
Cr( + 6) feed, no CI	Cr( + 3) feed, no CI	Cr( + 3) feed, 8.5% feed CI
Around kiln and afterburner
Kiln ash	99.6	99.3	98.4
Afterburner exit flue gas	0.4	0.7	1.6
Total	100.0	100.0	100.0
Around kiln and scrubber
Kiln ash	95.5	95.2	84.4
Scrubber exit flue gas	1.3	1.8	4.2
Scrubber liquor	3.2	3.0	11.4
Total	100.0	100.0	100.0
Apparent scrubber Cr
removal efficiency	71	63	73

-------
TABLE 29. TOTAL CHROMIUM PARTICULATE/VAPOR-IMSSOLVED PHASE FLUE GAS DISTRIBUTIONS
FOR THE CHROMIUM VALENCE STATE TEST SERIES
Afterburner exit flue gas
Particulate
Vapor/dissolved phase
Total
Scrubber exit flue gas
Particulate
Vapor/dissolved phase
Total
Percent of flue gas total Cr
Test 1	Test 2	Test 3
Cr( + 6) Teed, no CI Cr( + 3) Teed, no CI Cr(+3) feed, 8.5% feed CI
33
_67
100
57
100
68
J2
100
49
J1
100
56
_44
100
79
21
100

-------
The afterburner exit data in Table 29 show that the particulate phase total chromium
was highest for Test 2. the test with Cr( + 3) feed and no feed chlorine; intermediate for Test 3,
with the Cr( + 3) feed and chlorine in the feed; and lowest for Test 1, with Cr( + 6) in the feed
and no feed chlorine. This would be as expected if a significant fraction of the chromium in the
afterburner exit flue gas was present as soluble Cr( + 6) species (e.g., Cr04=, Cr,07=) for the
Cr( + 6) feed case or as soluble chlorides in the feed chlorine case, as opposed to insoluble
Cr;03 in the no chlorine/Cr( + 3) feed case. However, given the difficulty in obtaining a
representative flue gas sample at the afterburner exit, this result should be treated with caution.
If the soluble species are more effectively removed in the scrubber, one would expect
the fraction accounted for in the flue gas vapor/dissolved phase to decrease in the scrubber exit
flue gas for Tests 1 and 3. This is indeed the case.
Table 30 summarizes the fractions of the total chromium analyzed as Cr(+6) in feed
and discharge streams. As shown, the feed was analyzed as 52 percent Cr( + 6) for Test 1, and
as 2 percent Cr( + 6) for Tests 2 and 3. For all tests, the kiln ash chromium was comprised of
negligible amounts of Cr( + 6).
The scrubber exit flue gas Cr( + 6) fraction was the same (12 to 16 percent) regardless
of whether or not Cr( + 6) was present in the feed, in the two cases with no feed chlorine. In
contrast, for the case in which the feed contained chlorine, roughly half the scrubber exit flue
gas chromium was Cr( + 6). This would be as expected if entrained particulate chromium from
the kiln vaporized in the hotter afterburner and reacted with the flue gas chlorine to form
chromyl chloride (Cr02Cl2), a relatively stable compound with chromium as Cr( + 6).
The scrubber liquor Cr(+6) fraction for Test 1 with Cr( + 6) in the feed was significantly
higher than for the other two tests with only Cr( + 3) in the feed. This is as expected if some of
the chromium in the scrubber inlet flue gas were present as soluble Cr( + 6) species (Cr04" and
Cr:07= as noted above).
4.42 Parametric Trace Metals Tests
Experimental and laboratory data were reduced and analyzed to address the
experimental objectives discussed in Section 1. Data obtained during this testing program
provided information on the following:
•	Distribution of metals among the kiln ash, flue gas, and scrubber liquor streams
•	Afterburner exit and scrubber exit flue gas metal distributions between solid and
vapor/dissolved phase
•	Afterburner exit flue gas metal particulate distributions by size
•	Apparent scrubber collection efficiency for each metal
•	Effects of the primary test variables of kiln exit temperature, afterburner exit
temperature, and feed chlorine content on metal distributions
The test results listed above are summarized and reviewed in the following subsections.
To facilitate the discussion, results from all eight parametric tests are first summarized and
discussed independently of the RKS operating conditions, in Section 4.4.2.1. The intent is to
provide a concise summary of the relative distributions of each metal during the test program.
56

-------
TABLE 30. HEXAVALENT CHROMIUM FRACTIONS FOR THE CHROMIUM VALENCE STATE TEST SERIES
Cr( + 6)/lotal Cr (%)
Test I	Test 2	Test 3
Cr( + 6) feed, no CI Cr( + 3) feed, no CI Cr( + 3) feed, 8.5% feed CI
Composite feed
52
2
2
Kiln ash
0.3
0.2
0.1
Afterburner exit flue gas
102
35
4
Scrubber exit flue gas
12
16
48
Scrubber liquor
57
21
28

-------
The relationships between the discharge distributions of the metals and the RKS operating
conditions are then discussed in Section 4.4.2.2. The tables discussed in Section 4.4.2.1,
therefore, contain only ranges and averages of the test results. The individual values obtained
for each test are contained in the tables and figures discussed in Section 4.4.2.2.
For many samples, laboratory analysis results showed that several metals were not
detected at a certain detection limit. In the data evaluation, two calculations were performed:
one with the analysis result assumed to be zero, and one with the analysis result assumed to be
the detection limit. In many cases, distribution conclusions were not affected. In those cases
where distribution conclusions were affected, however, distributions are reported as ranges.
4.4J. 1 Average Trace Metal Discharge Distributions
As previously noted, a major objective of this test series was to identify the relative
distribution of each metal among the several RKS discharge streams. Table 31 summarizes
the relative distributions between the kiln ash and afterburner exit flue gas. Table 32 is a
similar summary among the kiln ash, scrubber exit flue gas, and scrubber liquor. In both tables
the first set of columns represents the fraction of the metal feed accounted for by the noted
discharge (e.g., kiln ash, afterburner exit flue gas). The range (low, high) exhibited over the 8
tests performed and the average for all 8 tests are noted. The second set of columns represents
fractions normalized to the total amount measured in the discharge streams analyzed. These
normalized values represent fractions which would have resulted had mass balance closure been
100 percent. Note that the sum of the normalized values for each element for the discharge
streams in each table is indeed 100.
Since the first set of columns are percent of metal feed, the kiln ash fractions are the
same in both tables. However, normalized kiln ash values change between the two tables
because mass balance closures experienced around the kiln ash/afterburner exit flue gas
discharge differed from those experienced around the kiln ash/scrubber system discharges.
The metals are ordered in the tables by increasing average normalized kiln ash fraction,
an ordering equivalent to decreasing experimentally observed average volatility. Those metals
with low average normalized kiln ash fractions were more volatile (found in kiln ash at lower
percentages) than metals with higher average normalized kiln ash fractions.
Cadmium was not detected in any kiln ash sample. The kiln ash fractions noted
correspond to the kiln ash detection limits. The normalized cadium fractions for the other
discharge streams in the tables assume cadmium was present in kiln ash at the detection limit.
Chromium is not included in Table 31. The afterburner exit flue gas sampling train
contained a stainless steel condenser between the alundum thimble filter, employed to collect
the large quantities of particulate needed for size fractionation, and the impinger train. As a
consequence, all afterburner exit sampling train impinger solutions contained significant
amounts of chromium.
The data in Tables 31 and 32 show that there are two distinct groupings of metals.
Cadmium, lead, and bismuth (Table 32) tend to be relatively volatile and are not found to a
major degree in the kiln ash. Barium, copper, strontium, arsenic, chromium, and magnesium
were found to be more refractory; kiln ash accounted for the major fraction of these metals.
The last column in Tables 31 and 32 contains a value termed volatility temperature (8).
This is the temperature at which the vapor pressure of the most volatile principal species of each
58

-------
TABLE 31. SUMMARY OF METAL DISCHARGE DISTRIBUTIONS IN THE KILN ASH AND
AFTERBURNER EXIT FLUE GAS FOR THE TRACE METALS PARAMETRIC
TEST SERIES
Distribution	Normalized distribution
(% or metal fed) (% or total measured)	Volatility
		temperature*,
Kiln ash	°C (°F)
Metal
Low
High
Average
Low
High
Average

Cadmium
<9.7
<17.1
<12.7
< 13.7
<43.6
<28.9
216(420)
Lead
4.7
12.6
8.3
9.5
61.7
35.5
627(1160)
Bismuth
8.4
24.4
16.2
53.4
89.6
70.0
621(1150)
Copper
34.8
64.2
49.0
63.6
94.1
84.8
1116(2040)
Barium
39.2
127.7
67.0
83.3
96.7
88.6
849(1560)
Arsenic
36.3
72.0
54.8
88.1
97.4
94.6
32(90)
Strontium
25.5
64.5
43.2
99.5
99.8
99.7
1454(2650)
Magnesium
69.2
133.1
91.7
99.7
99.9
99.8
1549(2820)
Afterburner exit flue gas
Metal
Low
High
Average
Low
High
Average

Cadmium
14.2
70.9
29.6
56.4
86.3
71.1

Lead
4.7
12.6
8.3
38.3
90.5
64.5

Bismuth
2.5
19.3
6.2
10.4
46.6
30.0

Copper
2.8
19.5
6.6
5.9
36.4
15.2

Barium
1.4
15.4
7.3
3.3
16.7
11.4

Arsenic
1.6
5.5
2.1
2.6
11.9
5.4

Strontium
0.07
0.3
0.2
0.2
0.5
0.3

Magnesium
0.1
0.3
0.2
0.1
0.3
0.2

Temperature at which the vapor pressure of a principal vapor species of the metal under
oxidizing conditions is 10"6 atm (8).

-------
TABLE 32. SUMMARY OF METAL DISCHARGE DISTRIBUTIONS IN THE KILN ASH
SCRUBBER EXIT FLUE CAS, AND SCRUBBER LIQUOR FOR THE TRACE
METALS PARAMETRIC TEST SERIES
Distribution	Normalized distribution
(% or metal Ted) (% or total measured)	Volility
	 		temperature*,
Kiln ash	°C(°F)
Metal
Low
High
Average
Low
High
Average

Cadmium
<9.7
<17.1
<12.7
<9.3
<29.3
<14.5
216(420)
Lead
4.7
12.6
8.3
5.8
83.7
20.1
627(1160)
Bismuth
8.4
24.4
16.2
20.9
64.8
31.6
621(1150)
Barium
39.2
127.7
67.0
68.8
86.9
76.6
849(1560)
Copper
34.0
64.2
49.0
58.0
97.6
78.6
1116(2040)
Strontium
25.5
64.5
43.2
81.9
94.3
89.3
1454(2650)
Arsenic
36.3
72.0
54.8
84.0
94.4
91.0
32(90)
Chromium
54.2
88.6
68.4
85.9
95.7
92.8
1613(2935)
Magnesium
69.2
133.1
91.7
99.2
99.6
99.4
1549(2820)
Scrubber exit flue gas
Metal
Low
High
Average
Low
High
Average

Cadmium
15.7
74.4
53.2
41.9
67.5
53.5

Lead
0.9
70.6
39.7
11.6
73.6
51.0

Bismuth
6.0
30.4
20.6
15.7
50.7
38.4

Barium
1.3
4.3
2.2
1.6
5.5
2.7

Copper
0.3
19.5
9.9
0.8
33.2
15.6

Strontium
0.5
1.6
1.0
1.1
3.7
2.1

Arsenic
1.4
4.0
2.6
2.2
8.4
4.5

Chromium
0.7
2.7
1.6
1.1
4.2
2.3

Magnesium
0.03
0.3
0.1
0.03
0.2
0.1

(continued)
Temperature at which the vapor pressure of a principal vapor species of the metal under
oxidizing conditions is 10* atm (8).

-------
TABLE 32. (concluded)
Distribution Normalized distribution
(% or metal Ted) (% of total measured)	Volility
	 		temperature*,
Scrubber liquor	°C(°F)
Metal
Low
High
Average
Low
High
Average

Cadmium
10.8
41.1
30.4
23.2
44.9
32.0
216(420)
Bismuth
7.4
27.1
16.2
19.5
36.7
30.0
621(1150)
Lead
0.4
37.2
22.3
4.7
41.1
28.9
627(1160)
Barium
11.2
31.7
17.0
11.5
28.8
20.7
849(1560)
Copper
0.7
6.4
4.1
1.6
9.1
5.8
1116(2040)
Strontium
1.5
6.3
3.9
4.4
16.0
8.6
1454(2650)
Arsenic
1.4
3.8
2.5
2.6
8.2
4.5
32(90)
Chromium
1.9
7.7
3.4
2.3
12.2
4.9
1613(2935)
Magnesium
0.3
0.6
0.5
0.3
0.7
0.5
1549(2820)
Temperature at which the vapor pressure of a principal vapor species of the metal under
oxidizing conditions is 10"6 atm (8).

-------
metal under oxidizing conditions is 10"6 atm. The lower a metal's volatility temperature is, the
more volatile it is expected to be. The data in Tables 31 and 32 show a remarkable
correspondence between the observed volatility and volatility temperature for all the metals
tested except arsenic. That is, except for arsenic, average normalized kiln ash fraction increases
more or less directly with increasing volatility temperature.
Based on volatility temperature, arsenic was expected to have been the most volatile
element, and the one found at lowest percentage in the kiln ash. However, the data clearly show
arsenic to be apparently refractory, remaining largely with the kiln ash. The volatility
temperature for arsenic is based on the vapor pressure of the most volatile arsenic
species under oxidizing combustion conditions. The fact that arsenic is significantly less volatile
than would be expected if Asj03 were the predominant arsenic species suggests that either some
other, more refractor}' arsenic compound is preferred, or that some other chemical interaction,
for example strong adsorption to the clay, occurred.
Table 33 summarizes the range and average of the percentage mass balance closures
achieved for all the tests performed. Again, achieved closures around the kiln ash discharge and
the afterburner exit flue gas, and around the kiln ash and scrubber discharges (flue gas and
scrubber liquor), are tabulated. The data in Table 33 show that average mass balance closure
ranged from 25 percent for bismuth to 88 percent for magnesium around the kiln
ash/afterburner exit flue gas. with an overall average of 55 percent. Individual metal closures
ranged from 11 to 140 percent.
TABLE 33. SUMMARY OF ACHIEVED METAL MASS BALANCE CLOSURE FOR THE
TRACE METALS PARAMETRIC TEST SERIES
Mass balance closure*
{% of metal fed)
Closed around Closed around
kiln ash discharge and kiln ash discharge and
afterburner exit flue gas scrubber discharges
Metal
Low
High
Average
Low
High
Average
Arsenic
42
74
58
39
77
60
Barium
41
140
83
57
147
86
Bismuth
11
44
25
36
74
53
Cadmium
25
84
52
37
120
96
Chromium
—
—

61
94
73
Copper
47
76
57
46
79
63
Lead
12
61
31
8
96
70
Magnesium
69
98
88
70
134
92
Strontium
26
65
47
28
71
48
a(Sum of total metal discharge rate over all discharges)/(metal feedrate).
62

-------
Achieved mass balance closures were better around the kiln ash/scrubber discharge, with
average metal closures ranging from 48 to 96 percent. Overall average closure was 71 percent.
Individual metal closures ranged from 8 to 147 percent. Previous discussion stated that the
afterburner exit flue gas sampling location is a poor one and likely does not allow a
representative flue gas sample to be collected. This is the most likely explanation for the lesser
closures achieved around the kiln ash/afterburner exit flue gas.
Almost all achieved mass balance closures were less than 100 percent. That is, the total
amount of metal which could be accounted for in the sum of discharge streams was almost
always less than the amount fed. The most likely explanation for this is that some quantity of
the metal fed was deposited in slag buildup in the afterburner chamber of the RKS. A gradual
accumulation of slag in the afterburner was observed over the duration of the test program. This
slag was no doubt a source of metal loss.
Despite the above, the mass balance closures achieved (averaging about 65 percent) are
considered good. Typical trace metal mass balance closure results from past experience for
combustion sources are in the 30 to 200 percent range.
Table 34 summarizes the ranges and averages of the relative distribution of each of the
metals between particulate and vapor/dissolved phases in the afterburner and scrubber exit
flue gas. Again, the particulate phase fraction is defined as that analyzed in the probe wash and
filter catches. The vapor/dissolved phase fraction is defined as that analyzed in the sampling
train impingers. The values in Table 34 assume a metal is present in a sample at the detection
limit for sample concentrations reported as less than the detection limit.
The data in Table 34 show that barium was found largely in the vapor/dissolved phase
in both the afterburner exit and the scrubber exit flue gas. This suggests that the specific barium
species predominant in the flue gas is quite water soluble. Bismuth and strontium were found
predominantly in the vapor/dissolved phase in the afterburner exit flue gas, but were found
predominantly in the particulate phase in the scrubber exit flue gas. All other metals, on
average, were found predominantly in the particulate phase at both locations, although individual
test conditions had opposite distributions for a few of these other metals.
Table 35 summarizes the observed range in, and average of, apparent scrubber
efficiencies in removing each metal from the flue gas stream. As defined in Section 4.4.1, the
apparent scrubber efficiency represents the ratio of the normalized metal fraction measured in
the scrubber liquor to the sum of the normalized metal fractions measured in the scrubber liquor
and scrubber exit flue gas.
The data in Table 35 show that average removal efficiencies ranged from 31 to
88 percent; the overall average for all metals was 57 percent. The metals are listed in the table
in order of increasing apparent scrubber efficiency. With two exceptions, the order is generally
the same as the observed metal volatility order in Tables 31 and 32. The volatile metals (those
found to a lesser extent in the kiln ash, such as lead, cadmium, and bismuth) have lower average
apparent scrubber collection efficiencies than do the less volatile metals (those found in the kiln
ash to a greater extent, such as arsenic, chromium, strontium, and magnesium).
This is as expected. The more volatile metals vaporize in the incinerator. When the
flue gas is cooled or quenched, these metals subsequently condense into either condensation
nuclei, resulting in very fine fume, or onto available existing particulate. The result of this
volatilization/condensation occurrence is that volatile metals become enriched in fine particulate.
63

-------
TABLE 34. SUMMARY OF FLUE GAS METAL PARTICULATE/VAPOR-DISSOLVEI) PHASE DISTRIBUTIONS FOR THE
TRACE METALS PARAMETRIC TEST SERIES
Afterburner exit flue gas	.Scrubber exit flue gas
Vapor/dissolved phase	Particulate phase	Vapor/dissolved phase	Particulate phase
fraction (percent)	Traction (percent)	Traction (percent)	Traction (percent)
Low High Average Low High Average Low High Average Low High Average
Arsenic
3
53
19
47
97
81
19
68
32
32
81
68
Barium
90
99
97
1
10
3
45
90
70
10
55
30
Bismuth
63
91
77
9
37
23
7
30
18
70
93
82
Cadmium
8
43
24
57
92
76
5
18
10
82
95
90
Chromium
—
—
—
—
—
—
13
33
21
67
87
79
Copper
20
71
38
29
80
62
I
55
19
45
99
81
Lead
24
84
47
26
76
53
0 1
10
4
90
99.9
96
Magnesium
9
43
25
57
91
75
7
75
32
25
93
68
Strontium
42
75
62
25
58
38
9
68
29
32
99
71

-------
TABLE 35. SUMMARY OF APPARENT SCRUBBER EFFICIENCY
RANGES AND AVERAGES FOR THE TRACE
METALS PARAMETRIC TEST SERIES
Apparent scrubber removal efficiency
(%)
Metal	Low	High	Average
Copper
19
67
31
Lead
22
48
36
Cadmium
26
52
38
Bismuth
34
55
45
Arsenic
36
64
49
Chromium
44
87
65
Strontium
65
93
78
Magnesium
63
92
84
Barium
82
95
88
That is, their concentration on a mass percent basis (^g/g of particulate) is higher in fine
particulate than in coarse particulate, This is so because:
•	Metal condensing into condensation nuclei form fume particles which are very fine
•	Metal condensing onto available existing particulate are mass enriched in the
available fine particulate, since condensation rates per unit particulate surface area
are constant, and since the available surface area to mass is higher for fine
particulate
Since the more volatile metals are expected to become enriched in fine particulate, the
observation that the apparent scrubber efficiencies for the more volatile metals are lower is
understandable, given that a venturi scrubber is significantly more efficient at collecting coarse
particulate than fine particulate.
The two exceptions to the volatility/apparent scrubber efficiency relationship noted in
Table 35 were copper and barium. The average apparent collection efficiency for barium is
higher than might have been expected based on its relative volatility. However, it was noted
above that barium was found predominantly in the vapor/dissolved phase in both the afterburner
exit and scrubber exit flue gas streams. It was suggested that this observation would be explained
if the barium compound predominating in the flue gas was quite water soluble. The observation
of higher-than-expected apparent scrubber collection efficiency would be similarly explained.
An explanation is being sought as to why copper has a lower apparent scrubber
collection efficiency than might be expected based on its relative volatility.
65

-------
4.422 EITecis of Incinerator Operating Conditions on Metal Distributions
A major objective of this test series was to identify the effects of the RKS operating
conditions on the metal distributions among the incinerator discharges. As discussed in
Section 2, the conditions varied in the test program were kiln exit temperature, afterburner exit
temperature, and waste feed chlorine content. Previous discussion in Section 4.4.2.1 considered
the 8 parameter tests performed as one set of tests, without mention of variations in test
conditions. In this subsection, the data from each test are presented in a format that facilitates
interpretation with respect to the primary test variables.
Table 36 summarizes the metal discharge distributions and mass balances, with the
discharge distributions expressed in terms of the fraction of the metal fed. Table 37 summarizes
the normalized distributions in both the kiln ash and afterburner exit flue gas discharges, and
in the kiln ash and scrubber discharges. As before, in cases where some samples were reported
as not containing a given metal, the distributions are reported as ranges. Limits of the ranges
correspond to an assumed sample concentration of zero and the detection limit, respectively.
No afterburner exit data are available for tests 5 and 10, since the alundum thimble particulate
sample was compromised by excessive moisture. Also, as discussed in Section 4.4.2.1,
afterburner exit flue gas chromium data are not reported.
Three groupings of tests appear in Tables 36 and 37., Each grouping corresponds to a
series of tests in which one test variable (kiln exit temperature, afterburner exit temperature, or
feed chlorine content) varied with the other two variables held nominally constant. The center
point, which was tested in duplicate (Tests 7 and 10), is included in all groupings.
The data in Table 36 represent the individual test data comprising the ranges and
averages noted in Tables 31, 32, and 33. The data in Table 37 represent the individual test data
comprising the ranges and averages noted in Tables 31, 32 and 35.
Figures 5 through 13 show the variation in the normalized discharge distributions among
the kiln ash, scrubber exit flue gas, and scrubber liquor for each of the test metals. Three bar
charts are shown in each figure corresponding to the three groupings of tests in Tables 36 and
37. Thus, each bar chart in Figures 5 through 13 corresponds to a series of tests in which one
test variable was varied, with the other two variables held nominally constant. Each figure shows
the normalized discharge distributions for one of the nine test trace metals. The figures are
ordered by decreasing observed metal volatility; this is the ordering of the metals used in
Table 32. Thus, Figure 5 shows cadium distribution (the observed most volatile metal) data;
Figure 13 shows magnesium distribution (the observed most refractory metal) data.
Normalized discharge distributions have been used in Figures 5 through 13, and will be
used in the following discussion, because they remove one significant source for test to test
data variability, and thereby allow clearer data interpretation. Normalized distributions would
be equivalent to distributions expressed as percent of metal fed had mass balance closure been
100 percent in all cases. Use of normalized distributions, therefore, "corrects" for data variability
introduced by less than perfect mass balance closure. Mass balance closure varied significantly
from test to test in this test program, as it has in all past experience in measuring trace metal
discharges from combustion sources. Thus, variations in metal discharge distributions expressed
as percent of metal fed would be affected by both actual variations due to changes in test
variables and by individual test mass balance closure. Variations in normalized distributions
more closely reflect those due to changes in test variables by "forcing" 100 percent mass balance
closure for each test.
66

-------
TABLE 36. METAL DISCHARGE DISTRIBUTIONS ANI) MASS BALANCE CLOSURE FOR THE TRACE
METALS PARAMETRIC TEST SERIES
Test:
5 7 10
6
9 7 10
8
4 7 10
II
Primary variable:
Kiln exit temperature ("(.')

Allcrburner exit temperature (*C)

Teed chlorine content (wt %)

Target:
816 871 871
927
982 1093 1093
1204
0 4 4
8
Test average:
825 878 873
9?*
983 I0HK 1094
11%
00 38 46
83
Meld constant:
Test average:
Test average:
AB exit = I09VC; chlorine - 4%	Kiln c»il =¦ 871'C; chlorine - 4%
1071 1088 I (I'M	1092 875	878	87.1	871
3 7	.18	4.6	4 2	.14	3 8	4 6	3 6
Kiln exit = 87PC. AB exit = I09VC
874	878	873	870
1093 1088 1094 1092
Discharge distribution (% of metal fed)
Arsenic
Kiln ash
719
395
363
408
720
395
.363
666
635
395
.363
476
Afterburner exit flue gas

20
—
5 5
19
20
—
22
16
20
—
38
Maw balance closure

41 5
—
463
73 9
41 5
—
688
65 1
41.5
—
51 4
around kiln/afterburner












Scrubber exit flue gas
1 9 2 3
2.1-27
11-15
3 6-4.0
25-30
2 1-27
11-15
27-32
12-14
2.1-2.7
1.1-1.5
22-25
Scrubber liquor
2 1
38
1 4
3.7
19
38
14
33
27
38
14
14
Mass balance closure
763
460
.392
485
769
460
392
7.3 1
67.6
460
392
51.5
—I	around kiln/scrubber
Bartam
Kiln ash
430
643
550
766
767
64.3
550
127 7
392
643
55 0
524
Afterburner exit flue gas
—
117
—
154
10 0
117
—
12 1
14
117
—
77
Mass balance closure
—
760
—
920
867
760
_
1390
406
760
—
60 1
around kiln/afterburner












Scrubber exit flue gas
2J
17
4 3
18
20
1.7
43
24
IJ
17
4.3
17
Scrubber liquor
12 9
147
194
31 7
112
14 7
194
17 1
165
14 7
19 4
12.7
Mass balance closure
590
807
787
110 1
89.9
807
787
1472
570
807
78.7
668
around kiln/scrubber
Bismuth
Kiln ash
162
165
109
92
84
165
109
207
244
165
109
23.3
Afterburner exit flue gas
—
35
—
80
25
35
—
25
19J
35
—
140
Mass balance closure
—
200
—
17 2
109
200
—
232
437
200
—
37.3
around kiln/afterburner












Scrubber exit flue gas
261
304
127
203
19 1
.304
127
256
60
304
12 7
247
Scrubber liquor
204
27 1
12 7
105
127
27 1
12 7
22.5
74
27 1
127
163
Mass balance closure
62.7
740
363
400
40 2
740
.363
688
378
740
363
643
around kiln/scrubber
(continued)

-------
TABLE 36. (continued)
Test:
5 7 10
6
9 7 10
8
4 7 10
II
Primary variable:
Kiln exit temperature (*C)

Afterburner exit temperature (°C)

Feed chlorine content (wt %)

Target:
816 871 871
927
982 11191 109.1
1204
0 4 4
8
Test average:
825 878 873
928
981 I0K8 1094
11%
00 .18 46
8.1
Held constant:
Test average:
Test average:
An exit = I093"C; chlorine = 4%
1071	1088	1094	1092
3.7	.18	4 6	4 2
Kiln exit = 87I"C, chlorine - 4%
875	878	871	871
.14	.1.8	4 6	.16
Kiln exit = 871-C; At) ciit = I09.TC
874	878	87.1	870
1093 1088 1094 1092
Dlschmjit distribution (% of mrtal fed)
Cadmium
Kiln ash
< 14 9
<112
<9 7
<12 9
< 14 5
<112
<9 7
<171
<109
<112
<9 7
< 10 0
Afterburner exit flue gas
—
20 8
—
70 9
4.11
208
—
24 9
14 2
208
—
62 8
Mass balance closure
—
21-32
—
71-83
4.1-57
21-12
—
25-42
14-25
21-32
	
63-73
around kiln/afterburner












Scrubber exit flue gas
421
61 6
31 1
744
614
61 6
311
647
157
616
31 1
728
Scrubber liquor
41.1
367
33.7
37.7
266
.167
33.7
362
108
.367
337
250
Mass balance closure
83-98
98-110
65-75
107-120
90-105
98-110
65-75
101-118
27-37
98-110
65-75
98-108
around kiln/scrubber
C?\		
00
Chromium
Kiln ash	88 6	68 9	54.9	75.1	79 9	68 9	54 9	54 2	64 2	68 9	54 9	60 9
Afterburner exit due gas	—	—	—	—	—	—	—	—	—	—	—	—
Mass balance closure	—	—	—	—	—	—	—	—	—	—	—	—
around kiln/afterburner
Scrubber exit flue gas	2 7	15	I t	1 7	0 9	15	I I	2 6	0 7	15	1 1	18
Scnibbcr liquor	2 2	2 9	7 9	2 1	2 8	2 9	7 9	4 0	19	2 9	7.9	3 4
Mass balance closure	93 5	73 3	63 9	79 1	83 6	71.1	63 9	60 8	66 8	73 3	63 9	66 1
around kiln/scmbber
Copper
Kiln ash
630
43.0
411
642
527
43 0
41.1
49.3
448
430
41.1
34 0
Afterburner exit flue gas
—
40
—
117
8.5
40
—
59
2.8
40
—
195
Mass balance closure
—
470
—
759
61 2
470
—
552
47.6
470
—
535
around kiln/afterburner












Scrubber exit flue gas
9.6
86
9.5
11.0
10 1
86
9.5
105
0J
86
9.5
195
Scnibbcr liquor
64
5 1
3J
27
36
5 1
3J
5.9
0.7
5 1
3J
5.2
Mass balance closure
790
567
53 9
779
664
567
539
65.7
45.8
56.7
539
58.7
around kiln/scrubber
(continued)

-------
TABLE 36. (concluded)
Test:
Primary variable:
Target:
Test average:
Held constant:
Test average:
Test average:
5 7 10
6
9 7 10
8
4 7 10
II
Kiln exit temperature (°C)

Afterburner exit temperature (°C)

Feed chlorine content (wt %)

816 871 871
927
9H2 109.1 1091
1204
0 4 4
8
825 878 87.1
9W
981 1088 1094
11%
00 .1.8 46
8.1
All exit = I093*C; chlorine = 4%	Kiln exit
1071	I0H8	1094	1092	R7S
.17	3 8	4 6	4 2	14
87I"C, chlorine = 4%	Kiln exit = 87I°C
878	87.1	871	874	878
.18	4 6	1 6	1091 1088
AD exit = I09VC
87.1	870
1094 1092
Discharge distribution (% of metal fed)
Lead
Kiln ash
114
12 5
50
8 1
47
125
50
126
67
125
50
57
Afterburner exit flue gas
—
77
—
.11 4
262
7.7
—
112
57
7.7
—
551
Mass balance closure
—
202
—
19 5
109
202
—
23.8
124
202
—
608
around kiln/afterburner












Scrubber exit flue gas
455
409
184
516
488
409
184
41 1
09
409
18 4
706
Scrubber liquor
.1.11
.101
1.12
17 2
270
.103
132
37.2
04
30 3
132
196
Mass balance closure
902
8.1.7
.166
769
805
8.17
.166
909
80
837
366
95 9
around kiln/scmbber
CTn
V3
Magnesium
Kiln ash
133.1
979
753
904
962
97.9
753
692
891
979
75.3
826
Afterburner exit flue gas
—
0 1
—
03
03
0 1

0 1
0 1
0 1
—
02
Mass balance closure
—
980
—
907
96 5
980

693
892
980
—
828
around kiln/afterburner












Scrubber exit flue gas
03
005
0 1
005
006
005
n i
007
003
005
0 1
007
Scrubber liquor
05
06
04
04
06
06
04
05
0J
06
04
04
Mass balance closure
133.9
986
75.R
909
969
986
758
698
894
98.6
758
83 1
around kiln/scmbber
Strontium
Kiln ash
326
56.7
34 1
64 I
645
567
34 1
348
25.5
567
34 1
31.1
Afterburner exit flue gas
—
03
—
02
0 1
01
—
0 1
007
0J
—
0 1
Mass balance dosure
—
570
—
64.1
646
570
—
349
25.6
570
—
33.2
around kiln/afterburner












Scrubber exit flue gas
0.5
1.0
IJ
1.2
10
10
IJ
16
06
10
IJ
07
Scrubber liquor
6J
3.3
25
59
3 1
3.1
25
6 1
15
3.1
2.5
28
Mass balance closure
394
610
379
71 2
686
61.0
37.9
425
276
610
370
36.6
around kiln/scrubber

-------
TABLE 37. NORMALIZED METAL DISCHARGE DISTRIBUTIONS AND APPARENT SCRUBBER
COLLECTION EFFICIENCY FOR THE TRACE METALS PARAMETRIC TEST SERIES
¦¦ ¦ ¦ " "¦ ¦¦
Tcsl:
5
7
in
6
9
7
10
8
4
7
in
II
Primary variable.
Kiln exit Irmpcralurc (°C)

Afterburner exit temperature (°C)

Krccl chlorine content (wt %)

T argcl:
816
871
871
927
982
ItPM
1093
12(14
0
4
4
8
Tcsl average:
825
878
873
928
983
1(188
1(194
11%
00
38
4 6
8..1
Hekl constant:
AH exit
= I093T; chlorine = 4%

Kiln exit
= 87I"C; chlorine = 4%

Kiln exit
= 871-C; AB exit =
1093"C
Tcsl average
1(171
1(188
1(194
1092
875
R78
873
871
874
878
873
870
Tcsl average:
37
38
4 6
4 2
3 4
38
46
3 6
1093
1088
1(194
1092
Nnrtniilirrd discharge distribution (%)
Anralc












Kiln ash
_
95.1

88 1
973
95 1
_
96.9
974
95.1

92.7
Afterburner exit flue gas
-
4 9
-
119
27
49
-
3 1
26
4 9
—
73
Total
-
100
-
1(10
100
1(10
-
100
100
1(10
-
1(10
Kiln ash
94 4
86 1
923
840
936
86 1
923
91 2
939
861
92J
924
Scrubber exit flue gas
2 1-29
38-58
23-4 1
68 84 2 6-3 8
38-58
2.3-4.1
3.0-4.3
1.7-2 2
38-58
2 3-4 1
4 0 4 8
Scrubber liquor
27
82
36
76
26
82
3.6
46
3.9
8.2
36
27
Total
1(10
1(10
ion
1(10
100
1(10
ion
1(10
100
100
100
100
Apparent scrubber
48-57
59-69
47*2
47-53
40 50
59-69
47-62
52-61
64-70
59-69
47-62
36-41
efficiency












Barium












Kiln ash
_
846
—
833
884
846
_
916
96.7
846
_
872
Afterburner exit flue gas
-
15 4
-
167
116
154
_
84
33
154
-
12.8
Total
-
1(10
-
100
1(10
100
-
100
100
100
-
100
Kiln ash
74 3
796
699
696
852
796
699
869
68 8
796
699
78.6
Scrubber exit flue gas
38
22
55
16
22
22
5 5
16
16-2 4
22
55
24
Scrubber liquor
21 9
182
24 7
288
125
182
24 7
115
288
182
24.7
19 0
Total
1(10
100
100
100
1(10
1(10
100
100
100
100
100
1(10
Apparent scrubber
85
89
82
95
85
89
82
88
92
89
82
89
efficiency












Bismuth












Kiln ash
_
826
_
534
763
826
_
896
560
826
_
62.4
Afterburner exit flue gas
_
17 4
—
466
23.7
17 4
-
10.4
44.0
174
-
376
Total
-
100
-
1(10
100
1(10
-
100
100
100
-
100
Kiln ash
258
22.2
300
229
209
222
300
301
648
222
300
363
Scrubber exit flue gas
41.5
41.1
35.2
507
474
41 1
352
37.1
157
41.1
352
38 4
Scrubber liquor
326
367
34.7
263
31.6
367
347
328
195
36.7
347
254
Total
1(10
1(10
1(10
100
1(10
1(10
100
1(10
100
inn
1(10
100
Apparent scrubber
44
47
50
34
40
47
50
47
55
47
50
40
efficiency












(continued)

-------
TABLE 37. (continued)
Test
5
7
10
6
9
7
10
8
4
7
10
II
Primary variable:
Kiln exit temperature ("<_")

Afterburner exit temperature ("C)

Feed chlorine content (wt %)

Target:
816
871
871
927
982
1091
1091
1204
0
4
4
8
Test average:
825
878
871
928
98.1
1088
1094
11%
00
.18
46
8.1
Held constant:
AH exit
I091X";
chlorine = 4%

Kiln exit
= 871'C; chlorine - 4%

Kiln exit
= 87I°C;
II
f
ffi
<
I09TC
Test average
in/i
1088
1094
1092
875
878
87.1
871
874
878
873
870
Test average:
.1.7
18
4 6
4.2
.14
3 8
4 6
36
1093
1088
1094
1092
Nnrmallred dlscharce distribution (%)
Cadmium












Kiln ash
_
<15 1
_
<15.1
<25 1
<35 1
_
<40 7
<43 6
<35.1
-
<13 7
Afterburner exit flue gas
-
>649
—
>84 7
>74 9
>64 9
—
>59 3
>564
>64 9
-
>863
Total
-
100
-
lilt)
100
100
-
100
100
100
-
100
Kiln ash
<152
< 10.1
<12 9
<10 7
<13 9
<10 3
<129
<14 5
<29 J
<10J
<129
<93
Scrubber exit flue gas
43-49
56-61
42-45
62-69
61-69
56-61
42-45
55-62
42-54
56-61
42-45
68-74
Scrubber liquor
42-51
.14.19
45-55
27-31
25-31
34-39
45-55
31-38
29-46
34-39
45-55
2.1-26
Total
100
100
100
100
100
100
100
1(10
100
100
100
100
Apparent scrubber
51
39
52
31
31
39
52
38
46
39
52
26
efficiency












Chromium


Kiln ash
—
—
Arterbumer exit flue gas
-
-
Total
-
-
Kiln ash
94.7
94 1
Scrubber exit flue gas
30
20
Scrubber liquor
2J
39
Total
100
100
Apparent scrubber
44
66
efficiency
-
-
-
-
-
859
95.3
955
94 1
859
19
2 1
It
20
1.9
12 2
26
34
39
122
100
100
100
100
100
87
56
75
66
87
-
-
-
-
-
893
957
94.1
859
921
42
14
20
19
28
66
28
39
122
5 1
100
100
100
too
100
61
67
66
87
64
Copper
Kiln ash
_
908
—
84.7
860
908
—
894
94 1
908
-
636
Afterburner exit flue gas
-
92
-
I5J
140
92
-
106
5.9
9.2
-
364
Total
-
100
-
100
100
100
-
100
100
100
-
100
Kiln ash
842
758
762
823
792
758
762
75 1
976
758
762
580
Scrubber exit flue gas
12 9
15.1
17 8
14.1
152
15.1
178
160
08
15.1
17 8
332
Scrubber liquor
30
9.1
5.9
36
56
9 1
5.9
8.9
16
9.1
59
88
Total
100
100
100
100
100
100
100
100
100
100
100
100
Apparent scrubber
19
38
25
21
27
38
25
36
67
38
25
21
efficiency
(continued)

-------
TABLK 37. (concluded)
Test
5
7
10
(>
9
7
10
8
4
7
10
II
Primary variable:
Kiln exit temperature ("(.')

Afterburner exit temperature (°C)

feed chlori
ne content (wt %)

Target
816
871
871
927
9H2
1093
1093
1204
0
4
4
8
Test average:
825
878
873
928
983
I0K8
1094
11%
00
3 8
4 6
8 3
Held constant:
AB exit
- I093"C, chlorine =
4%
Kiln exit
= 87l°c:,
chlorine = 4%

Kiln exit -
87|"C; AH exit =
I093-C
Tesl average
1071
I0H8
1094
1092
875
878
871
871
874
878
873
870
TcM average:
3 7
3 8
46
4 2
34
38
4 6
3 6
109.3
I0H8
1094
1092
Nnrmalired dinthaqif distribution (%)
Lead












Kiln ash
.
61 7
_
203
15 2
61.7
_
528
537
61.7
-
95
Afterburner cxil flue gas
-
383
-
79 7
848
383
-
472
463
383
-
905
Total
-
10(1
-
100
100
100
-
100
100
100
-
100
Kiln ash
126
15 0
13 7
104
5 8
150
137
138
83 7
15 0
137
60
Scrubber exit flue gas
504
489
502
67 2
606
489
502
450
116
489
502
736
Scmbbcr liquor
370
36 1
360
224
33 6
36 1
36.0
41 1
4.7
36 I
36 0
203
Total
100
100
100
100
100
100
100
100
100
too
100
100
Apparent scrubber
42
42
42
25
36
42
42
48
29
42
42
22
efficiency












Magnesium












Kiln ash
_
999
_
997
99 7
999
—
999
999
99.9
_
998
Afterburner exit flue gas
-
0 1
-
0 3
03
0 1
—
0 1
001
01
-
02
Total
-
100
-
100
100
100
-
100
100
100
-
100
Kiln ash
994
993
993
995
993
993
993
992
996
993
99 J
994
Scrubber exit flue gas
02
0 1
02
0 1
0 1
0 1
02
0 1
003
0 1
02
0 1
Scrubber liquor
04
06
0 5
04
06
06
05
07
036
06
05
0 5
Total
100
100
100
100
100
100
100
100
100
100
100
100
Apparent scrubber
63
92
75
89
91
92
75
88
91
92
75
85
efficiency
Strnnlium
Kiln ash
_
995
_
99.7
998
995
—
996
997
99.5
—
997
Afterburner exit flue gas
-
0.5
_
03
02
05
-
04
0J
05
-
03
Total
-
100
-
100
100
100
-
100
100
100
-
100
Kiln ash
829
930
898
90 1
94 3
93.0
898
81.9
918
93.0
898
907
Scrubber exit flue gas
It
1.7
3 5
16
13
17
3 5
3 7
2.5
17
35
16
Scrubber liquor
160
53
66
8J
4.4
5.3
66
144
57
5J
66
7.7
To»«
100
100
100
100
100
100
100
100
100
100
100
100
Apparent scrubber
93
76
65
84
77
76
65
80
69
76
65
83
efficiency

-------
Variable: Kiln Exit Temperature
¦o
ft
W
D
v>
n
o
2
«
o
o
c
o
•—
u
u.
Klin Ash
816 C
Scrubber Exit Gas
871 C
Scrubber Liquor
827 C
Figure 5a. Cadmium discharge distributions for the parametric trace metal tests:
effect of kiln temperature.
73

-------
¦D
®
k_
3
v>
n
0>
5
«
o
c
o
o
K
Variable: Afterburner Exit Temperature
Kiln Ash
962 C
Scrubber Exit Gas
1093 C
Scrubber Liquor
1204 C
Figure 5b. Cadmium discharge distributions for the parametric trace metal tests:
effect of afterburner temperature.
74

-------
Figure 5c. Cadmium discharge distributions for the parametric trace metal tests:
effect of feed chlorine content.
75

-------
70
Variable: Kiln Exit Temperature
•o
01
S
O
c
o
u
(O
60
50 -|
40
30
20
10
Scrubber Exit Gas
871 C
Figure 6a. Lead discharge distributions for the parametric trace metal tests:
effect of kiJn temperature.
76

-------
70
60
50
40 -
30
20 -
10
Variable: Afterburner Exit Temperature
Kiln Ash
962 C
Scrubber Exit Gas
1093 C
Scrubber Liquor
1204 C
Figure 6b. Lead discharge distributions for the parametric trace metal tests:
effect of afterburner temperature.
77

-------
90
Variable: Feed Chlorine Content
*o
0)
k_
D
V)
<0
01
2
o
i-
o
c
o
o
(9
Kiln Ash
Egg ov.
Scrubber Exit Gas
4 •/,
Scrubber Liquor
8 %
Figure 6c. Lead discharge distributions for the parametric trace metal tests:
effect of feed chlorine content.
78

-------
Figure 7a. Bismuth discharge distributions for the parametric trace metal tests:
effect of kiln temperature.
79

-------
50
Variable: Afterburner Exit Temperature
40 -
Kiln Ash
982 C
Scrubber Exit Gas
1093 C
Scrubber Liquor
1204 C
Figure 7b. Bismuth discharge distributions for the parametric trace metal tests:
effect of afterburner temperature.
80

-------
Figure 7c. Bismuth discharge distributions for the parametric trace metal tests:
effect of feed chlorine content.
81

-------
XI
6)
L.
3
O
<9
O
2
-------
Figure 8b. Barium discharge distributions for the parametric trace metal tests:
effect of afterburner temperature.
83
-------
Figure 8c. Barium discharge distributions for the parametric trace metal tests:
effect of feed chlorine content.
84
-------
Variable: Klin Exit Temperature
*o
&
u
3
«
8
2
c
o
w
-------
Figure 9b. Copper discharge distributions for the parametric trace metal tests:
effect of afterburner temperature.
86
-------
Variable: Feed Chlorine Content
Kiln Ash
Scrubber Exit Gas
Scrubber Liquor
0 •/.
4 •/.
S V.
Figure 9c. Copper discharge distributions for the parametric trace metal tests:
effect of feed chlorine content.
87
-------
"D
&
3
ifi
G
Q>
5
o
o
c
o
u
K
Variable; Kiln Exit Temperature

Kiln Ash
Scrubber Exit Gas
Scrubber Liquor
816 C
871 C
927 C
Figure 10a. Strontium discharge distributions for the parametric trace metal tests:
effect of kiln temperature.
88
-------
¦D
ID
w
3
V)
re
o>
2
re
o
c
o
o
re
Variable: Afterburner Exit Temperature

Scrubber Exit Gas
1093 C
Scrubber Liquor
1204 C
Figure 10b. Strontium discharge distributions for the parametric trace metal tests:
effect of afterburner temperature.
89
-------
Of
h.
D
4rt
«
6)
2
O
O
c
o
o
Variable: Feed Chlorine Content
E^&zzzzzZZZZz.
Kiln Ash
Scrubber Exit Gas
Serubbtr Liquor
0%
4 '/.
ev.
Figure 10c. Strontium discharge distributions Tor the parametric trace metal tests:
effect of feed chlorine content.
90
-------
&
w
3
iA
«
&
2
to
o
c
D
o
re
Variable: Kiln Exit Temperature

Scrubber Exit Gas
871 C
Scrubber Liquor
927 C
Figure 11a. Arsenic discharge distributions for the parametric trace metal tests:
effect or kiln temperature.
91
-------
100
Variable: Afterburner Exit Temperature
"O
01
w
3
(A
«
0J
5
m
o
o
c
o
o
Kiln Ash
Scrubber Exit Gas
Scrubber Liquor
g 982 C
E2S3 1093 c
1204 C
Figure lib. Arsenic discharge distributions for the parametric trace metal tests:
effect of afterburner temperature.
92
-------
Variable: Feed Chlorine Content

Scrubber Exit Gas
4 %
Scrubber Liquor
8 %
Figure 11c. Arsenic discharge distributions for the parametric trace metal tests:
effect of feed chlorine content.
93
-------
T3
0>
«s
2
c
o
u
re
Variable: Klin Exit Temperature

Klln Ash
Scrubber Exit Gas
Scrubber Liquor
816 C
871 C
927 C
Figure 12a. Chromium discharge distributions for the parametric trace metal tests:
effect of kiln temperature.
94
-------
Variable: Afterburner Exit Temperature
*0
<0
0)
2
o
c
o
o
(0
1093 C
Scrubber Liquor
ES3 1204 C
Figure 12b. Chromium discharge distributions for the parametric trace metal tests:
effect of afterburner temperature.
95
-------
Figure 12c. Chromium discharge distributions for the parametric trace metal tests:
effect of feed chlorine content.
96
-------
•c
c
u
3
V)
K
a
5
re
o
o
c
o
u
re
Variable: Kiln Exit Temperature
Scrubber Exit Gas
871 C
f
Scrubber Liquor
927 C
Figure 13a. Magnesium discharge distributions for the parametric trace metal tests:
effect of kiln temperature.
97
-------
Variable: Afterburner Exit Temperature
"D
0>
D
iA
<9
G)
5
o
c
o
o
1204 C
Figure 13b. Magnesium discharge distributions for the parametric trace metal tests:
effect of afterburner temperature.
98
-------
Variable: Feed Chlorine Content
1
Scrubber Exit Gas
4 y.
Scrubber Liquor
e v,
Figure 13c. Magnesium discharge distributions Tor the parametric trace metal tests:
effect of feed chlorine content.
99
-------
The following discussion focuses on each metal's discharge distribution data in turn.
Cadmium
The data in Figure 5 show that the kiln ash accounted for about 10 percent of the
measured discharge cadmium provided there was chlorine in the feed mixture. The data suggest
that the kiln ash fraction decreases slightly with increasing kiln temperature. A corresponding
increase in the scrubber exit flue gas with a decrease in the scrubber liquor cadmium fraction
seems to have occurred with increasing temperature, although the data show some variability.
This behavior would be consistent with expectations. Increasing kiln temperature would be
conducive to volatilizing more cadmium in the kiln. Any cadmium vaporized in the kiln would
remain in the vapor phase through the afterburner, then condense as the flue gas was quenched.
In accordance with the volatilization/condensation discussion in Section 4.4.2.1, scrubber
efficiency would be expected to decrease. This would give rise to a greater scrubber exit flue
gas fraction and a smaller scrubber liquor fraction.
No clear trend in any discharge stream's cadmium fraction, within the apparent data
variability, is seen with varying afterburner temperatures.
The kiln ash cadmium fraction is clearly greater when no chlorine is in the feed than
when chlorine is present. This would be expected if cadmium chlorides were more volatile than
cadmium oxides.
L«ad
The data in Figure 6 show that, as for cadmium, the kiln ash accounted for slightly
greater than 10 percent of the measured discharge lead, provided there was chlorine in the feed.
Again, there appears to be a slight decrease in kiln ash lead fraction with increasing kiln
temperature, with a corresponding increase in scrubber exit flue gas fraction and a decrease in
scrubber liquor fraction.
Interestingly, there appeared to be a decrease in scrubber exit flue gas lead fraction and
an increase in scrubber liquor fraction with increasing afterburner temperature. This seems
counter to expectations.
As with cadmium, the kiln ash lead fraction was substantially greater for the test with
no feed chlorine than for the tests with 4 percent chlorine in the feed. This is consistent with
expectations The principal vapor phase lead species in a chlorine-free excess oxygen
environment is expected to be lead metal with a volatility temperature of 627°C (1160°F).
However, in a chlorine-containing environment, the principal vapor phase species becomes PbCl4
with a volatility temperature of -15°C (5°F).
The data in Figure 6 show a further decrease in kiln ash lead fraction when feed
chlorine content was further increased from 4 to 8 percent. Scrubber exit flue gas lead fraction
monotonically increases with increased feed chlorine content. Scrubber efficiency
correspondingly decreases.
Bismuth
The data in Figure 7 show bismuth behaves in a manner similar to cadmium and lead
with respect to feed chlorine content. The kiln ash bismuth fraction was in the 20 to 35 percent
range for all tests in which the feed contained chlorine. A marked increase in kiln ash bismuth
100
-------
fraction (a decrease in bismuth volatility) occurred when chJorine was removed from the feed.
A modest increase in scrubber efficiency for bismuth was also observed when feed chlorine was
removed.
The data show no significant variation in any discharge stream bismuth fraction with
changes in kiln or afterburner temperature.
Barium
The data in Figure 8 suggest that no test variable had a significant effect on barium
discharge distributions.
Copper
The data in Figure 9 show that neither kiln nor afterburner temperatures had
measurable effects on copper discharge distributions. However, copper kiln ash fractions clearly
decreased monotonically with increasing feed chlorine content. Scrubber exit flue gas copper
fractions showed a corresponding monotonic increase with increasing feed chlorine content.
Both of these observations would be consistent with expectations if copper chlorides were more
volatile than corresponding oxides.
Interestingly, scrubber liquor copper fractions may have increased slightly with increasing
feed chlorine content. This suggests that the increased solubility of copper chlorides may have
somewhat offset the partitioning of copper to finer particle size with increased copper
volatilization, and thereby offset to some degree the expected decrease in scrubber collection
efficiency.
Strontium
The data in Figure 10 show that neither kiln temperature nor feed chlorine content
affect strontium discharge distributions. There was an apparent steady decrease in kiln ash
strontium fraction with increasing afterburner temperature. This is coincidental, since
afterburner temperature can have no effect on kiln ash fraction. The apparent steady increase
in scrubber liquor strontium fraction is most likely similarly coincidental.
Scrubber exit flue gas strontium fraction appeared to increase slightly with increasing
afterburner temperature. Although this may also have been coincidental, it would be expected
if higher afterburner temperature caused an increase in the amount of strontium in entrained
flyash from the kiln which vaporized in the afterburner.
Arsenic
The data in Figure 11 show no effect of afterburner temperature on arsenic discharge
distributions. A slight decrease in kiln ash fraction, accompanied by a slight increase in scrubber
exit flue gas fraction, with increasing kiln temperature appears to have occurred. A decrease
in scrubber efficiency was observed with increasing feed chlorine content.
Chromium and Magnesium
The data in Figures 12 and 13 show that no test variable measurably affected chromium
or magnesium discharge distributions.
101
-------
Table 38 summarizes the distribution of each metal between the particulate phase and
the vapor/dissolved phase in the afterburner and scrubber exit flue gas for each test. The data
in Table 38 represent the individual test data comprising the ranges and averages noted in
Table 34. Again, distributions affected by samples with nondelegable levels of a metal are
represented by ranges. The range boundaries reflect the assumption that the samples in
question contained zero and the detection limit of the metal, respectively. The data in Table 38
show no consistent relationship between particulate/vapor-dissolved phase ratios and test
variables.
4.423 Afterburner Exit Particle Size Distributions
Recall from the sampling and analysis protocol discussion in Section 3, that the flue gas
at the afterburner exit was sampled with a sampling train designed to collect a large (-lg)
sample of particulate so that this particulate could be size fractionated and trace metal
concentrations as a function of particle size could be determined. Sufficient afterburner exit flue
gas particulate was collected by the sampling train to allow particle size fractionation to be
performed for six of the eight tests performed. Size fractionation was not possible for the
particulate catch from tests 5 and 10.
Figure 14 shows a log-probability plot of the total particulate particle size distribution
data for the six tests for which size fractionation was performed. As indicated, the size
distributions for all tests were roughly log normal. The distributions were roughly comparable
for the five tests in which the test waste feed contained chlorine. The size distribution of total
particulate was shifted to larger particle size for the one test with no feed chlorine. This would
be as expected if the presence of chlorine in the feed served to increase the volatility of feed
inorganic constituents. In such a case the recondensation of volatilized inorganic would tend to
result in a finer particle size distribution for the chlorine containing feed cases.
Particle size fractions were combined to give four size fraction samples for trace metal
analysis. These size fractions were nominally <2, 2 to 4, 4 to 10, and >10 /im respectively.
From the analyses of these size fractions, size distributions of each of the test trace metals were
obtained.
Figure 15 shows the average size distribution for the five hazardous constituent trace
metals included in the test waste mixture (arsenic, barium, cadmium, chromium, and lead) and
compares these to the total particulate size distribution. The size distributions shown in
Figure 15 represent averages over the six tests for which size distribution analyses were
performed. Figure 16 shows analogous size distributions for the four non-hazardous trace
metals included in the test waste mixture (bismuth, copper, magnesium, and strontium).
The data in Figures 15 and 16 show that the average size distributions for chromium,
copper, magnesium, and strontium were coarser (shifted to larger particle sizes) than the overall
particulate size distribution. These are all relatively nonvolatile metals with high volatility
temperatures (See Tables 31 and 32). The average size distributions for barium and bismuth
were finer (shifted to smaller particle sizes) than the overall particulate. These two are more
volatile, having lower volatility temperature than chromium, copper, magnesium, and strontium.
Figures 17 through 22 show individual test metal size distributions for both the
hazardous constituent and non-hazardous constituent trace metals.
102
-------
TABLE 38. PHASE DISTRIBUTION OF FLUE GAS METALS IN THE AFTERBURNER AND SCRUBBER EXIT
FLUE GAS FOR THE TRACE METALS PARAMETRIC TEST SERIES
T CM
5
7
10
6
9
7
10
8
4
7
10
II
Primary variable:
Kiln exit temperature ("C)

Afterburner exit temperature (°( )

Feed chlorine content (wt %)

Target:
816
871
871
027
•W2
1091
1093
1204
0
4
4
8
Test average:
825
878
873
928
983
KIKK
1094
11%
00
38
46
83
Meld constant:
AR exit
= iwrc, chlorine = 4%

Kiln exit -
87I°C, chlorine = 4%

Kiln exit
= 871°C, AB exit =
I09.VC
Test average
1071
1088
1094
1092
875
878
873
871
874
878
873
870
Test average:
3.7
38
4 6
4 2
34
38
46
36
1093
1088
1094
1092
%
Arsenic

Afterburner exit:

Particulate
-
91
_
91
>97
91
-
94
47
91
-
65
Vapor/dissolved
-
9
-
9
<3
9
-
6
53
9
-
35
phase

Scrubber exit:

Particulate
>76
>72
>53
81-95
>74
>72
>53
>74
32
>72
>53
>78
Vapor/dissolved
<24
<28
<47
5-19
<26
<28
<47
<26
68
<28
<47
<22
phase

Barium

Afterburner ait:

Particulate
-
3
-
1
3
3
-
1
6-10
3
-
3
Vapor/dissolved
-
97
-
99
97
97
-
99
90-94
97
-
97
phase

Scrubber exit:

Particulate
55
26-29
10-27
31-33
25-29
26-29
10-27
32-35
25
26-29
10-27
27
Vapor/dissolved
45
71-74
73-90
67-69
71-75
71-74
73-90
65-68
75
71-74
73-90
73
phase

Bismuth

Afterburner exit:

Particulate
_
17
-
31
37
17
-
12-20
23
17
-
9
Vapor /dissolved
-
83
-
69
63
83
-
80-88
77
83
-
91
phase

Scrubber exit:

Particulate
90-94
83
70-76
86-93
73
83
70-76
86
>75
83
70-76
93-99
Vapor/dissolved
6-10
17
24-30
7-14
27
17
24-30
14
<25
17
24-30
1-7
phase

(continued)
-------
TABLE 38. (continued)
Test:
5
7 111
6
9
7
10
8
4
7
10
II
Primary variable:
Kiln exit temperature ("C)

Afterburner cxil
tcmpcralure (°C)

Feed chlorine content (wt %)

Targrt
816
871 871
97.7
982
109.1
1091
1204
0
4
4
8
Test average:
825
878 87.1
9 28
981
1(188
1094
11%
00
38
46
83
Held constant:
AB exit =
I09.VC; chlorine = 4%

Kiln cxil
- 87I°C, chlorine = 4%

Kiln exit
= 87I°C, AB cxil =
I093°C
Test average
1071
1088 1094
1092
875
878
873
871
874
878
873
870
Test average:
3 7
18 4 6
42
3 4
.18
46
3 6
1093
1088
1094
1092
%
Cadmium

Afterburner exit:

Paniculate
-
57
92
73
57
—
82
68
57
-
82
Vapor/dissolved
-
43
8
27
43
-
18
32
43
-
18
phase

Scrubber exit:

Particulate
84-89
>94 >86
87
95-99
>94
>86
>94
>82
>94
>86
>95
Vapor/dissolved
10-16
<6 < 14
11
1-5
<6
<14
<6
<18
<6
<14
<5
phase

Chromium

Afterburner oil:

Particulate
-
- _
-
_
_
-
-
-
_
—
—
Vapor/dissolved
-
_
-
-
-
-
-
-
-
-
-
phase

Scrubber exit:

Particulate
79-84
>82 68-90
81-92
70
>82
68-90
87
77-94
>82
68-90
>85
Vapor/dissolved
16-21
<18 10-32
8-19
30
< 18
10-32
13
6-23
<18
10 32
<15
phase

Copper

Afterburner exit:

Particulate
-
58
80
55
58
-
77
29
58
-
72
Vapor/dissolved
-
42
20
45
42
-
23
71
42
-
28
phase

Scrubber exit:

Particulate
97
97 45
87
80
97
45
89
51-57
97
45
99
Vapor/dissolved
3
3 55
13
20
3
55
11
43-49
3
55
1
phase

(continued)
-------
TABLE 38. (concluded)
Test
5
7
10
6

7
111
8
4
7
10
11
Primary variable:
Kiln exit
temperature (°C)

Afterburner exit temperature (°C)

Feed chlorine content (wt %)

Target:
816
871
871
927
982
11193
1093
1204
0
4
4
8
Test average:
825
878
873
928
983
10X8
1094
11%
00
38
4 6
8 3
Held constant:
AO exit
- I093°C; chlorine =
4%
Kiln exit
= 87I°C, chlorine = 4%

Kiln exit
= 87I°C, AB exit =
low;
Test average
1071
1088
11194
1092
875
878
873
871
874
878
873
870
Test average:
37
38
46
4 2
3 4
3 8
46
36
1093
1088
1094
1092
%
Uad

Afterburner exit:

Particulate
-
28
_
76
54
28
-
66
26
28
-
71
Vapor/dissolved
-
72
-
24
46
72
-
34
74
72
-
29
phase

Scrubber exit:

Particulate
90
>99
93
90
>99
>99
93
>99
>95
>99
93
>99
Vapor/dissolved
10
< 1
7
10
<1
< 1
7
< 1
<5
< 1
7
<1
phase

Magnesium

Afterburner exit:

Particulate
—
69
-
70
91
69
-
84
57
69
-
76
Vapor /dissolved
-
31
-
30
9
31
-
16
43
31
-
24
phase

Scrubber exit:

Particulate
93
73
25
82-88
62
73
25
71
65
73
25
72
Vapor/dissolved
7
27
75
12-18
38
27
75
29
35
27
75
28
phase

Strontium

Afterburner exit:

Particulate
-
41
-
41
58
41
-
19-25
8-34
41
-
25-32
Vapor/dissolved
-
59
-
59
42
59
-
75-81
66-92
59
-
68-75
phase

Scrubber exit:

Particulate
32
79-85
47
89-98
>91
79-85
47
85-89
74-80
79-85
47
76-86
Vapor/dissolved
68
15-21
53
2-11
<9
15-21
53
11-15
20-26
15-21
53
14-24
phase

-------

F««d
Kiln
Aft«r6urr>«r

T»t
CI (%>
T rc;
T (*C)
0
4
0
672
1064
+
e
4.2
627
1092
o
7
3.6
676
1066
A
a
3 6
671
1186
X
«
9 4
676
664
7
11
6.3
671
1092
; i 1 1 1 r-1—i 1 1 1 1 1 1
2 4 10 20 50
Particle Diameter (microns)
Figure 14. Afterburner exit flue gas particulate size distributions for the parametric trace
metal tests.
106
-------
Particle Diameter (microns)
Figure 15. Hazardous constituent trace metal size distributions in afterburner exit flue
gas particulate for the parametric trace metal tests.
Particle Diameter (mierona)
Figure 16. Nonhazardous trace metal size distributions in afterburner exit flue gas
particulate for the parametric trace metal tests.
107
-------
£
« 60 -
>
«
3
E
a
O
C.2
Partiel* DlamoUr (microns)
£
ec
5
c
? 5C
#
£
£
3
E
a
O
7ott> Stmpl*
D)
Cu
M«
—1
10
Partlcl* Dlamotor (microns)
Figure 17. Trace metal size distributions in afterburner exit flue gas for Test 4.
108
-------
E
a
a
a
>
3
E
3
u
Partlel* Dlamatar (micron*)
SO.6
Particfa Dlamaiar (microna)
Figure 18. Trace metal size distributions in afterburner exit flue gas for Test 6.
109
-------
ft.
m
3
E
3
u
Total Sample
Ax
&i
CO
Cf
rt>
Particl* Dlam*t«r (microns)
£
c
c
«
« 6C
«
3
E
Tola' Simpli
Bl
Cj
MC
PartlcU Diamctar (mfcrona)
Figure 19. Trace metal size distributions in afterburner exit flue gas for Test 7.
110
-------
Particle Diameter (microns)
Particle Diameter (mlcrona)
Figure 20. Trace metal size distributions in afterburner exit flue gas for Test 8.
Ill
-------
•
1
E
m
5
2. 64
C
V
«
>
3
E
3
O
60
0.2
Partlcl# Diam»t«r (microns)
E
c
a
<•*
o
3
£
a
o
Particl* DiamtUr (microns)
Figure 21. Trace metal size distributions in afterburner exit flue gas for Test 9.
112
-------
Par tic l« Dlam«t*r (micron*)
Particle Dlam«t*r (mlcrona)
Figure 22. Trace metal size distributions in afterburner exit flue gas Tor Test 11.
113
-------
45 POHC DESTRUCTION AND REMOVAL EFFICIENCIES
The three compounds introduced in the synthetic waste feed as principal organic hazardous
constituents (POHCs) for these tests were toluene, chlorobenzene, and tetrachloroethylene. Flue gas
concentrations of the compounds measured at the three locations sampled (the afterburner exit, the
scrubber system exit, and the stack (three tests)) during the parametric trace metal tests are
summarized in Table 39. The data in the table show that POHC concentrations were generally
decreased in the scrubber exit flue gas compared to the afterburner exit flue gas; although toluene
and chlorobenzene concentrations increased for Tests 5, 6, and 9. Reduced concentrations are
expected due to some combination of flue gas dilution via air inleakage and POHC removal in the
scrubber system. The increased concentrations observed for Tests 5, 6, and 9 are possibly related
to a sampling artifact in either the afterburner or scrubber exit sampling trains.
Table 40 combines the flue gas concentration data from Table 39 with the flue gas flowrate
data in the sampling sheets in Appendix C, the synthetic liquid waste feed composition data from
Table 17, and the liquid waste feedrate data from Table 4, to give the POHC DREs at the three
locations sampled. As shown in the table, POHC DREs in the stack ranged from 99.9975 to
99.9994 percent. With the exception of the afterburner exit for Test 11, POHC DREs were above
99.99 percent at all locations for all tests. For Test 11 with 8.4 percent waste chlorine feed, the
POHC DREs were uniformly lower than the other tests (with lower feed chlorine contents) at all
three sampling locations.
4.6 VOLATILE PRODUCTS OF INCOMPLETE COMBUSTION (PICs)
As noted in Section 3, all VOST traps were analyzed by thermal desorption, purge and
trap, GC/FID for the volatile organic constituents listed in Table 16. This list includes the POHCs
discussed in Section 4.5. However, several other potential volatile PICs are included in this analyte
list as well. Table 41 summarizes the flue gas concentrations of the non-POHC volatile organic
constituents detected in the afterburner exit flue gas for one or more of the tests performed.
Tabulated values represent the average of the three VOST trap pairs operated during each test. For
determining the average emission rate, practical quantitation limits (PQLs) were used for trains
containing concentrations less than the PQL. Table 42 presents an analogous summary for the
scrubber system exit flue gas.
The data in Table 41 show that the major PICs detected in the afterburner exit flue gas
were chloroform, carbon tetrachloride, and benzene. Concentrations ranged from about 1 to
12 /xg/dscm. Note that for Test 4 (0 percent chlorine feed), the chlorinated PICs are present at
significantly reduced concentrations, as expected. Also for Test 4, both tetrachloroethylene and
chlorobenzene were detected, and are reported as PICs since neither was a POHC for this test.
Table 42 shows that chloroform levels in the scrubber exit flue gas were above those at the
afterburner exit for five of the eight tests. The origin of the increased chloroform is likely the
scrubber makeup water. The scrubber water source is a chlorine treated potable supply in which the
trihalomethane (THM) compounds are typically found. These compounds volatile readily; the
scrubber essentially acts as a purger, releasing the THM into the flue gas stream with the entrained
water vapor.
For all tests, carbon tetrachloride levels in the scrubber exit flue gas were less than the
levels in the afterburner exit flue gas. Benzene concentrations increased in the four tests, decreased
in three tests, and remained the same in the one test. No correlations between PIC levels in the flue
gas at either location and corresponding POHC DRE are apparent.
114
-------
TABLE 39. FLUE GAS POHC CONCENTRATIONS FOR THE PARAMETRIC TRACE
METAL TESTS
POHC concentration (gg/dscni)
Afterburner exit Scrubber eult Stack
Test Test date Toluene Chlornheniene Tetrachloroethylene Toluene Chbirnhentene Tetrachloroethylene Toluene Chlornhenzene Tetrachloroethylene
4
9/14/88
85
—
—
48
—
—
—
—
—
5
8/25/88
10
2.1
75
40
3J
56
—
—
—
6
9/16/88
14
18
13
73
74
10
—
—
—
7
8/30/88
140
17
35
21
24
3 1
—
—
—
8
9/07/88
240
29
43
40
4.2
51
99
13
14
9
9/09/88
19
22
5.0
27
3.3
4.2
85
14
79
10
9/20/88
510
86
55
21
1.9
40
—
—
—
II
9/22/88
1300
550
750
62
18
24
120
22
29
-------
TABLE 40. POHC DREs FOR THE PARAMETRIC TRACE METAL TESTS
ORE (%)
Afterburner exit ScniMier fill Stuck
Test Tent date Toluene Chlombenzene Tetrnchlonwlhylene Toluene Ctitonihenzene Tetrarhloroelhylene Toluene Chlornbeniene Tetrachloroethylene
4
9/14/88
99 99943
—
—
99 99936
—
—
—
—
—
5
8/25/88
99.999910
99999913
99.99966
99 99930
99.99973
99 99945
—
—
—
6
9/16/88
99.999903
99999923
99 99948
99.9988
99 99932
99.99907
—
—
—
7
8/.W/88
999990
9999919
99 9984
9999967
99.99976
9999969
—
—
—
8
9/07/88
99.9980
99 9986
99.9980
9999915
99 99949
9999940
999986
99.9990
99 9990
9
9/09/88
99.99982
9999988
99 99974
99 99951
99.99966
99.99957
99 9988
99.9990
99 9994
10
9/20/88
99.9967
999967
99.9979
99.99967
99 99987
9999964
—
—
—
11
9/22/88
99988
99.989
99985
99 9986
99.99918
999989
99.9975
99 9990
999987
-------
TABLE 41. AFTERBURNER EXIT FLUE GAS VOLATILE PIC CONCENTRATIONS
Test
4
5
6
7
8
9
10
11
Test date
(9-14-88)
(8-25-88)
(9-16-88)
(8-30-88)
(9-07-88)
(9-09-88)
(9-20-88)
(9-22-88)
Compound

Afterburner exit due gas concentration (pg/dscm)

Chloroform
2.4
9 1
5.3
12.6
5.4
9.0
64
4.9
Carbon tetrachloride
16
8.4
53
3.1
5.5
5.5
11.9
<1.5
Benzene
7.8
2.7
<16
2.2
9.9
106
7.5
11.7
Trichlorocthylene
<1.7
1.7
<1.7
<1.7
<1.7
1.9
1.9
<1.7
13-D ichlorobenzene
<1.6
2.0
<1.6
<1.6
<1.6
<16
<1.6
<1.6
1,4-D ichlorobenzene
<1.6
2.1
<1.6
1.7
1.6
1.6
<1.6
<1.6
1,2-Dichlorobcnzene
1.9
<1.5
<1.5
<1.5
1.7
<1.5
2.2
<1.5
Tet rachloret hy lene
16.8
V
P
P
P
P
P
P
Chlorobenzene
2.0
P
P
P
P
P
P
P
"P: Compound was a POHC for these tests.
-------
TABLE 42. SCRUBBER EXIT FLUE GAS VOLATILE PIC CONCENTRATIONS
Test 4 5 6 7 8 9 10 II
Test date (9-14-88) (8-25-88) (9-16-88) (8-30-88) (9-07-88) (9-09-88) (9-20-88) (9-22-88)
Compound Scrubber exit flue gas concentration (/ig/dscm)
Chloroform
43
10.0
8.4
3.3
14.7
5.6
17.9
27.7
Carbon tetrachloride
<1.5
2.1
2.1
19
4.7
2.7
3.6
23
Benzene
<1.6
85
48
<1.6
10.0
16.7
15.7
5.8
1,3-Dichlorobenzene
<1.6
1.7
1.6
<1.6
1.9
1.9
<1.6
<1.6
1,4-Dichlorobenzene
<1.6
16
1.5
<1.6
1.8
2.2
2.5
<1.6
1,2-Dichlorobenzene
2.3
<1.5
3.0
<1.5
<1.5
2.4
4.6
1.6
Tetrachlorethane
<1.8
<1.8
<1.8
<1.8
7.8
14.1
<1.8
<1.8
1,2-Dichloroethane
<1.5
<1.5
<1.5
2.1
<1.5
<1.5
23
<1.5
Ethylbenzene
6.4
63
6.0
3.2
<1.5
10.0
7.5
3.1
-------
SECTION 5
CONCLUSIONS
A five-week series of pilot-scale incineration tests was performed to evaluate the fate
of trace metals fed to a rotary kiln incinerator equipped with a venturi scrubber/packed column
scrubber for particulate/acid gas control. Three tests focused on determining the valence state
(trivalent/hexavalent) of chromium in emissions and discharges from the incinerator, and
investigating whether valence state distributions were affected by the chromium valence state in
the waste feed to the incinerator or the feed chlorine content. Eight tests focused on a
parametric evaluation of the fate of five hazardous constituent trace metals (arsenic, barium,
cadmium, chromium, and lead), and four nonhazardous constituent trace metals (bismuth,
copper, magnesium, and strontium) fed to the incinerator. Test variables for the parametric
evaluation were kiln temperature, afterburner temperature, and feed chlorine content. A
factorial experimental matrix was tested in which kiln temperature was varied from 816 to 927°C
(1500 to 1700°F), afterburner temperature was varied from 982 to 1204°C (1800 to 2200°F), and
feed chlorine content was varied from 0 to 8 percent.
For all tests, the waste feed consisted of a synthetic mixture prepared by combining an
organic liquid mixture with a clay absorbent material. The organic liquid mixture consisted of
toluene with varying amounts of tetrachloroethylene and chlorobenzene added to give desired
feed chlorine content. This synthetic solid waste material contained nominally 0.4 kg of organic
liquid to 1 kg of clay absorbent. This mixture was introduced to the kiln via a screw feeder.
Test trace metals were prepared in an aqueous solution. This solution was metered into the
clay/organic liquid mixture in the screw feeder.
For the chromium valence state tests, total chromium and hexavalent chromium
(Cr(+6)) concentrations were measured in the kiln ash and scrubber blowdown discharges as
well as in the incinerator flue gas at two locations: the afterburner exit and downstream of the
scrubber system. For the parametric trace metal tests, trace metal concentrations were
measured in the same emission/discharge streams. In addition, the distribution of flue gas trace
metal between particulate and vapor/dissolved phases was measured, and the particulate size
distribution of trace metals in the afterburner exit flue gas was determined.
Conclusions from the chromium valence state tests include:
• Most of the chromium fed was discharged in the kiln ash. Kiln ash accounted for
95 percent of the discharge amount with no chlorine in the feed, regardless of the
chromium valence state in the feed. The kiln ash fraction decreased to 84 percent
when chlorine was present in the feed.
• Scrubber exit flue gas accounted for 1 to 2 percent of the chromium discharged
with no chlorine in the feed, again regardless of the feed valence state. Scrubber
exit flue gas fraction increased to 4 percent when chlorine was present in the feed.
119
-------
• Scrubber liquor accounted for 3 percent of the discharged chromium with no feed
chlorine. The scrubber liquor fraction increased to 11 percent with chlorine-
containing feed.
• The kiln ash contained negligible Cr(+6) regardless of feed chromium valence
state or chlorine content.
• Nominally 15 percent of the chromium in the scrubber exit flue gas was Cr(+6)
with no feed chlorine. The Cr( + 6) fraction increased to almost 50 percent of the
scrubber exit flue gas chromium with chlorine present in the feed. This result is
consistent with the formation of chromyl chloride frdm entrained chromium
vaporized in the afterburner.
• Scrubber liquor chromium was 20 to 30 percent Cr(+6) with Cr(+3) feed;
scrubber liquor chromium increased to about 60 percent Cr(+6) with Cr(+6) in
the feed.
Conclusions from the parametric trace metal tests include:
• Based on normalized discharge distribution data, cadmium, lead, and bismuth were
relatively volatile metals with an average of less than about 30 percent of the
discharged metal being present in kiln ash. Barium, copper, strontium, arsenic,
chromium, and magnesium were relatively nonvolatile with an average of greater
than 75 percent of the discharged metal being present in the kiln ash.
• Observed metal volatilities from most to least volatile were cadmium, lead,
bismuth, barium, copper, strontium, arsenic, chromium, and magnesium. This
order agrees with the order predicted by metal volatility temperature (temperature
at which the vapor pressure of a principal vapor species is 10"6 atm) with the
notable exception of arsenic. Arsenic has the lowest volatility temperature of the
metals tested, but was observed to be one of the least volatile of the metals. This
suggests that Asj 03 was not a predominant arsenic species in the incinerator, or
that the arsenic is adsorbed by the clay/ash matrix.
• The average apparent scrubber collection efficiency was lower for the volatile
metals (36 to 45 percent) than for the nonvolatile metals (49 to 88 percent) with
the notable exception of copper. Copper's average scrubber removal efficiency,
at about 30 percent, was significantly lower than the efficiencies seen for the other
relatively nonvolatile metals.
• Feed chlorine content had a major effect on the partitioning of the volatile metals
(cadmium, lead, and bismuth), and of copper. The fraction of metal discharged
in the kiln ash for these metals measurably decreased with increasing feed chlorine
content.
• Kiln temperature had a minor effect on the relative volatility of the observed most
volatile metals (cadmium and lead) and of arsenic. The fraction of metal
discharged in the kiln ash decreased slightly with increasing kiln temperature.
• No test variable measurably affected the discharge distributions of barium,
strontium, chromium, and magnesium, among the least volatile metals.
120
-------
• Total afterburner exit flue gas particulate size distributions were roughly log-
normal for all tests in which they were measured. The size distributions for all
tests with chlorine-containing feed were roughly comparable; the size distribution
was shifted to larger particle size with no chlorine in the feed.
• The afterburner exit flue gas average size distributions of the less volatile metals
(chromium, copper, magnesium, and strontium) were coarser than the
corresponding overall particulate average size distribution. The size distribution
for the relatively more volatile metals (barium and bismuth) were finer than the
overall particulate average size distribution.
Test waste POHC destruction and removal efficiencies (DREs) were greater than
99.99 percent as measured in the scrubber discharge flue gas and in the stack downstream of the
system's carbon bed/HEPA filter for all of the parametric trace metal tests. POHC DREs were
similarly greater than 99.99 percent as measured in the afterburner exit flue gas for all tests
except the test with the highest (nominally 8 percent) feed chlorine content. In fact, POHC
DREs at all locations were uniformly lower for this high-chlorine-content feed than for other
tests with lower feed chlorine contents.
Chloroform, carbon tetrachloride, and benzene were the volatile products of incomplete
combustion (PICs) present in flue gas at the highest levels, up to the order of 10 /Jg/dscm.
All test program data quality objectives (DQOs) were met, with the exception of field
matrix spike recovery and the precision of feed clay matrix spike duplicate analyses for Cr(+6).
Precision and accuracy DQOs for laboratory matrix spikes and field spikes for other matrices
were met. The most important Cr(+6) analyses needed to support the test conclusions noted
above were for matrices other than the feed clay. Thus, the failure to achieve the recovery and
precision DQOs for a single analyte in a single field matrix has very little impact on the test
conclusions stated above.
121
-------
SECTION 6
QUALITY ASSURANCE
A number of quality assurance (QA) procedures were followed to assess the data quality
of the laboratory analytical measurements performed in these tests. In addition, QA efforts
performed to ensure that data quality is known for the particulate and CEM measurements
involved adherence to Reference Method procedures and CEM manufacturers' specifications.
No deviations from the quality assurance project plan (QAPP) occurred for these measurements
with the exception that the C0/C02 monitor at the stack did not function properly. The major
laboratory QA efforts focused on the following measurements:
• Volatile organic compounds in the feed mixtures
• VolatiJe organic compounds in the VOST traps
• Metals in the feed and incinerator discharges
The data quality objectives (DQOs) for these measurements are listed in Table 43. The QA
procedures performed included:
• Analyzing replicate feed mixture samples for toluene, chlorobenzene, and
tetrachloroethylene
• Preparing a clay/organic liquid feed mixture matrix spike and analyzing replicate
samples for spike recovery
• Spiking all VOST train resin traps with two method surrogates and measuring
surrogate recovery
• Preparing eight matrix spike VOST traps and analyzing these for spike recovery
(seven analyzed at the IRF using the method described in Appendix D and one
shipped to the Acurex analytical chemistry laboratory in Mountain View,
California and analyzed via EPA Method 5040)
• Obtaining duplicate VOST samples for analysis at both the IRF and the Acurex
Mountain View, California laboratories
• Preparing laboratory matrix spike solutions with the trace metals and analyzing
these for spike recovery
• Preparing four hexavalent chromium field matrix spike samples at the IRF and
analyzing these for spike recovery
122
-------
TABLE 43. PRECISION, ACCURACY, AND COMPLETENESS OBJECTIVES
Measurement
parameter
Measurement/
analytical
method
Reference
Conditions
Precision
(% RSD* or
RPD")
Accuracy
<%)
Completeness
(%)
Volatile
organic
compounds
in clay/
liquid feed
Packed column
GC/FID by
the method
in Appendix D
Appendix D
Direct
injection
50
50-140
80
Volatile
organic
compounds
on Tenax
Thermal
desorption
purge and trap
by Method
5040, GC/FID
analysis by the
method in
Appendix D
SW-846-3**
Appendix D
Thermal
desorption
100
50-1504'
80
Cr( + 6)
Methods 3060s
and 7197°
SW-S46-3",
SW-846-2"
Alkaline
digestion and
AA analysis
25
75-12
80
Metals
Method 6010,
7000 series,"
or 300 series'
methods
SW-846-3",
AWPCF
Acid
digestion
and 1CAP or
AA analysis
25
75-125^
80
c(i)
"Recovery of surrogate spikes of resin samples greater than 20 percent was also an accuracy objective.
'Percent recovery from a spiked Tenax cartridge.
'(2)
£7)
'Recovery of a spiked water solution.
•Relative Standard Deviation.
"Relative Percent Difference.
123
-------
Results of these OA procedures are discussed in the following sections.
6.1 ORGANIC ANALYSIS OF CLAY/ORGANIC LIQUID FEED SAMPLES
As discussed in Section 3, a composite sample of the feed mixture from each drum
prepared was collected and analyzed for the volatile organic hazardous constituents toluene,
chlorobenzene, and tetrachloroethylene. Each drum of a given test mixture was prepared by
mixing together weighed amounts of the individual liquid organic constituents, then mixing
together weighed amounts of the liquid mixture and the clay sorbent. Waste feed for a given
test was taken from one, two, or three drams depending on test conditions and durations. Thus,
comparing analysis results for each drum of a target test mixture preparation gives a measure
of the precision and accuracy of the entire feed preparation, sampling, and analysis procedure.
Table 44 provides such a comparison. For each target test mixture preparation, analysis
results for a sample of each drum prepared are compared to the target composition and to each
other. The accuracy of the feed preparation, sampling, and analysis operation is measured by
the ratio of the analyzed concentration of a mixture component to the target value (termed
recovery in Table 44). The precision of the procedure is measured by the relative standard
deviation (RSD) of all samples analyzed.
The data in Table 44 show that analyzed concentrations ranged from 26 to 135 percent
of target concentrations over 24 samples analyzed. Only 1 measurement out of 63 was out of
the 50 to 140 percent recovery range, the accuracy DQO for the measurement. Thus,
98 percent of the measurements were acceptable. This DQO was met. The precision of the
procedure as measured by the RSD of the analysis of different drum samples of a given target
formulation ranged from 5 to 26 percent. All were within the precision DQO of 50 percent for
this measurement. Thus, this DQO was met.
As a further measure of analytical precision, two clay/organic liquid feed samples were
subjected to duplicate analysis. The relative percent difference (RPD) for these ranged from
1.3 to 10.1 percent for each of the constituents analyzed, as shown in Table 45. The precision
DQO for the volatile organic compounds in the feed mixture was 50 percent RPD. This DQO
was therefore met.
Three samples of a clay/organic liquid feed mixture matrix spike were prepared under
controlled laboratory conditions (as opposed to bulk feed sample preparation) and analyzed to
determine volatile organic constituent matrix spike recovery. The data in Table 46 show that
the recoveries ranged from 75 to 104 percent. All were within the accuracy DQO range for this
measurement of 50 to 140 percent recovery. Thus, this DQO was also achieved. Further, the
RSD of the replicate spike analyses ranged from 8.9 to 11.3 percent, all within the DQO for
precision of 50 percent.
It bears noting that the analytical data in Tables 44 and 45 are different from the feed
mixture compositions used for DRE determinations in Section 4 (i.e., in Table 17). The data
in Table 17 represent analyzed compositions (Tables 44 and 45) corrected for the matrix spike
recoveries experienced (Table 46).
62 VOLATILE ORGANIC COMPOUNDS IN THE VOST TRAPS
To assess measurement accuracy of the VOST samples analyses, each resin trap was
spiked with the surrogates octane and 4-bromofluorobenzene prior to use in sampling. Surrogate
recovery was then measured.
124
-------
TABLE 44. CLAY/ORGANIC LIQUID FEED SAMPLE POHC ANALYSIS RESULTS
Toluene Tetrachloroethylene Chlorobenzene
Measured Measured Measured
Test concentration Recovery* concentration Recovery* concentration Recover)*
Test date Drum (wt %) (%) (wt %) (%) (wt %) (%)
Mixture 1
Target concentration

28.6

1,2 9/2S/8S-
1
26.4
92

9/29/88

2
25.5
89
—
—
—
—

3
24.4
85
—
—
—
—
4 9/14/88
1
22.5
79
—
—
—
—

2
153
53
—
—
—
—

3
26.6
93
—
—
—
—
RSD (9c) 18
Mixture 2
Target concentration

21.7

3.4

5 8/25/88
1
15.1
70
2.7
79
3.2
94

2
15.7
72
23
68
3.0
88
6 9/16/88
1
19.1
88
3.1
91
3.0
88

2
19.0
88
2.9
85
2.9
85

3
18.9
87
3.1
91
3 2
94
7 8/30/88
1
16.0
74
23
68
2.5
74

2
19.2
88
2.9
85
3.0
88

3
19.4
89
2.7
79
3.0
88
8 9/7/88
1
163
IS
2.8
82
2.7
79

2
15.4
71
23
68
2.5
74
9 9/9/88
1
15.2
70
2.4
71
2.5
74
10 9/20/88
I
21.1
97
3 2
94
3.6
106

2
20.1
93
0.9
26
4.6
135

3
21.2
98
3.2
94
3.1
91
RSD (%) 15 26 18
Mixture 3
Target concentration

14.9

6.9

3 9/26/88
1
13.8
93
6.0
87
5.4
78

2
153
103
6.4
93
6.0
87
11 9/22/88
1
12.8
86
5.6
81
5.7
83

2
14.0
94
6.1
88
6.2
90

3
13.6
91
6.0
87
6.0
87
RSD (%) 7 5 5
'Ratio of measured concentration to target concentration.
125
-------
TABLE 45. CLAY/ORGANIC LIQUID FEED REPLICATE VOLATILE ORGANIC
ANALYSIS RESULTS

Toluene

Tetrachloroethylene

Chlorobenzene

Test
Test
date Drum
1st
Analysis
(wt %)
2nd
Analysis
(wt %)
RPD
(%)
1st
Analysis
(wt%)
2nd
Analysis
(wt %)
RPD
<%)
1st
Analysis
(wt%)
2nd
Analysis
(wt %)
RPD
(%)
3
9/26/88 2
153
15.1
13
6.4
63
1.4
6.0
5.9
1.9
1,2
9/28/88- 3
9/29/88
24.4
27.0
10.1

TABLE 46. VOLATILE ORGANIC CONSTITUENT RECOVERY FROM
CLAY/ORGANIC LIQUID FEED MATRIX SPIKE SAMPLES
Percent recovery
Relative
Standard
Replicate Replicate Deviation
Compound Spike spike spike (RSD)
Toluene 86
Chlorobenzene 78
Tetrachloroethylene 75
88 104 8.9
81 100 11.3
79 95 10.4
126
-------
Table 47 summarizes octane recovery from the VOST trap pairs. The data in the table
show that octane recovery ranged from 0 to 217 percent, with 7 of 75 recoveries not within the
DQO range of 50 to 150 percent. Measurement completeness was 91 percent, and met the
DQO for recovery completeness of 80 percent.
Table 48 similarly summarizes 4-bromofluorobenzene recovery from the VOST trap
pairs. The data show that 4-bromofluorobenzene recovery ranged from 0 to 120 percent, with
3 of 75 recoveries not within the DQO range of 50 to 150 percent. Measurement completeness
was 96 percent, and met the DQO for recovery completeness of 80 percent.
As a further measure of VOST analysis accuracy, and as a measure of analysis precision,
seven matrix spike VOST trap pairs were prepared and analyzed at the IRF. Results of these
analyses are summarized in Table 49. The data in the table show that matrix spike recoveries
ranged from 73 to 194 percent, with 2 of 35 recoveries not within the DQO range of 50 to
150 percent. Measurement completeness was 94 percent, and met the DQO for recovery
completeness of 80 percent. Further, the RSD of the matrix spike analyses ranged from 3.6 to
29 percent, all within the DQO for precision of 100 percent.
In addition to the seven traps analyzed at the IRF, an eighth matrix spike VOST trap
pair was prepared and shipped to Acurex's Mountain View, California, laboratory for analysis
by EPA Method 5040. This spike sample was prepared identically to one analyzed at the IRF.
Table 50 summarizes the analysis results for this VOST trap pair and compares results of the
Method 5040 GC/MS analysis with the GC/FID analysis of an identically prepared trap pair.
Matrix spike and surrogate recoveries for the GC/MS-analyzed sample ranged from 46 to
137 percent, with one out of eight compounds not within the DQO recovery range of 50 to
150 percent. Measurement completeness was 87 percent, and met the DQO for recovery
completeness of 80 percent. In addition, the RPD between the GC/MS-analyzed traps and the
GC/FID-analyzed traps ranged from 28 to 80 percent, all within the precision DQO for this
measurement of 100 percent RPD.
To further compare VOST trap analysis results between the GC/FID analysis method
routinely performed at the IRF (see Appendix D) and the GC/MS Method 5040 analysis
method, two sampling trains were operated simultaneously at the scrubber exit during Test 10
to obtain two sets of VOST traps from the same flue gas. One set was analyzed at the IRF by
GC/FID (Appendix D), and one set was analyzed by Method 5040 at the Acurex Mountain View
laboratory. Unfortunately, the GC/MS analysis of one of the trap pairs (the third pair) was
invalidated due to a leak during analysis. Table 51 summarizes the VOST results reported for
each method.
The precision DQO for the measurement of volatile organic compounds in VOST traps
is 100 percent RPD. The data in Table 51 show that reasonably good agreement between the
two analysis methods, within about 100 percent RPD, was achieved for chloroform,
bromodichloromethane, dibromochloromethane, bromoform, toluene, chlorobenzene, and
tetrachJoroethylene (the last three compounds being the POHCs for these tests). However,
agreement between the two analysis methods was not as good for carbon tetrachloride, benzene,
dichlorobenzenes, and 1,2-dichloroethane. Surrogate recoveries were acceptable by both
methods. Given the mixed nature of these results, further method comparisons are
recommended and planned in future tests.
127
-------
TABLE 47. OCTANE SURROGATE RECOVERY FROM VOST TRAPS
Afterburner exit traps
Scrubber system exit traps
Other traps

Octane

recovery

recovery
Sample
m
Sample
(%)
Sample
(%)

Stack Samples

Test 4 (9/14/881

Test 8 (9/07/881

1st trap pair
121
1st trap pair
118
1st trap pair
123
2nd trap pair
101
2nd trap pair
107
2nd trap pair
109
3rd trap pair
76
3rd trap pair
109
3rd trap pair
134
Test 5 (8/25/88^

Test 5 (8/25/881

Test 9 (9/09/881

1st trap pair
137
1st trap pair
217
1st trap pair
102
2nd trap pair
151
2nd trap pair
112
2nd trap pair
34
3rd trap pair
117
3rd trap pair
153
3rd trap pair
119
Test G (9/16/88^

Test 6 (9/16/881

Test 11 (9/22/881

1st trap pair
107
1st trap pair
109
1st trap pair
91
2nd trap pair
114
2nd trap pair
111
2nd trap pair
122
3rd trap pair
115
3rd trap pair
0
3rd trap pair
123
Test 7 (8/30/881

Test 7 (8/30/881

Field blank traps

1st trap pair
114
1st trap pair
111
8/25/88-1

2nd trap pair
110
2nd trap pair
111
36
3rd trap pair
76
3rd trap pair
98
8/25/88-2
94

8/30/88
80
Test 8 (9/07/881

Test 8 (9/07/881

9/07/88
80
1st trap pair
117
1st trap pair
96
9/09/88
120
2nd trap pair
136
2nd trap pair
100
9/14/88
101
3rd trap pair
108
3rd trap pair
114
9/16/88
105

9/20/88
106
Test 9 (9/09/881

Test 9 (9/09/881

9/22/88
100
1st trap pair
79
1st trap pair
133

2nd trap pair
116
2nd trap pair
135
Laboratory method

3rd trap pair
116
3rd trap pair
118
blank traps

8/25/88
72
Test 10 (9/20/881

Test 10 (9/20/881

1st trap pair
122
1st trap pair
81
Matrix spike traps

2nd trap pair
117
2nd trap pair
117
8/25/88
98
3rd trap pair
115
3rd trap pair
178
8/30/88
103

9/07/88
108
Test 11 (9/22/881

Test 1J (9/22/88)

9/09/88
104
1st trap pair
115
1st trap pair
111
9/14/88
107
2nd trap pair
0
2nd trap pair
117
9/16/88
109
3rd trap pair
111
3rd trap pair
112
9/20/88
112

9/22/88
105
128
-------
TABLE 48. 4-BROMOFLUOROBENZENE SURROGATE RECOVERY FROM VOST TRAPS
Afterburner exit traps
Scrubber system exit traps
Other traps

4-Bromo-

4-Bromo

fluoro-

benzene

recovery

recovery
Sample
(%)
Sample
(%)
Sample
<*)

Stack Samples

Test 4 (9/14/881

Test 8 (9/07/881

1st trap pair
107
1st trap pair
100
1st trap pair
96
2nd trap pair
86
2nd trap pair
104
2nd trap pair
98
3rd trap pair
81
3rd trap pair
98
3rd trap pair
111
Test 5 (8/25/88")

Test 5 (8/25/881

Test 9 (9/09/881

1st trap pair
105
1st trap pair
91
1st trap pair
120
2nd trap pair
115
2nd trap pair
105
2nd trap pair
108
3rd trap pair
101
3rd trap pair
112
3rd trap pair
120
Test 6 (9/16/881

Test 6 (9/16/881

Test 11 (9/22/881

1st trap pair
83
1st trap pair
104
1st trap pair
79
2nd trap pair
101
2nd trap pair
106
2nd trap pair
108
3rd trap pair
105
3rd trap pair
0
3rd trap pair
107
Test 7 (8/30/881

Test 7 (8/30/881

Field blank traps

1st trap pair
94
1st trap pair
99

2nd trap pair
99
2nd trap pair
95
8/25/88-1
27
3rd trap pair
80
3rd trap pair
37
8/25/88-2
94

8/30/88
81
Test 8 (9/07/881

Test 8 (9/07/881

9/07/88
88
Is: trap pair
104
1st trap pair
104
9/09/88
96
2nd trap pair
115
2nd trap pair
81
9/14/88
97
3rd trap pair
95
3rd trap pair
103
9/16/88
102

9/20/88
94
Test 9 (9/09/881

Test 9 (9/09/881

9/22/88
88
1st trap pair
89
1st trap pair
102

2nd trap pair
92
2nd trap pair
101
Laboratory method

3rd trap pair
106
3rd trap pair
102
blank traps

8/25/88
92
Test 10 (9/20/881

Test 10 (9/20/881

1st trap pair
107
1st trap pair
101
Matrix spike traps

2nd trap pair
104
2nd trap pair
91
8/25/88
97
3rd trap pair
92
3rd trap pair
110
8/30/88
96

9/07/88
93
Test 11 (9/22/881

Test 11 (9/22/881

9/09/88
110
1st trap pair
96
1st trap pair
106
9/14/88
105
2nd trap pair
110
2nd trap pair
97
9/16/88
106
3rd trap pair
105
3rd trap pair
105
9/20/88
105

9/22/88
98
129
-------
TABLE 49. VOLATILE ORGANIC CONSTITUENT RECOVERY FROM VOST MATRIX SPIKE SAMPLES
Percent recovery
Relative
Compound
Test 4
(9/14/88)
Test 5
(8/25/88)
Test 7
(8/30/88)
Test 8
(9/07/88)
Test 9
(9/09/88)
Test 10
(9/20/88)
Test 11
(9/22/88)
Standard
Deviation
(% RSD)
1,1,1-Trichloroethane
ltl
93
107
113
98
103
126
9.4
Benzene
73
116
130
127
194
92
153
29.0
Toluene
123
124
115
147
124
131
121
10.0
Tetrachloroethylene
104
103
101
100
107
111
100
3.6
Chlorobenzene
102
96
100
100
108
107
97
4.2
-------
TABLE 50.
VOLATILE ORGANIC CONSTITUENT RECOVERY FROM THE
VOST MATRIX SPIKE SAMPLE FOR TEST 10 (9/20/88)
BY METHOD 5040 AND GC/FID ANALYSIS
Percent
recovery
Compound
GC/FID
GC/MS
RPD
<%)
1,1,1-Trichloroethane
103
64
47
Benzene
92
65
34
Toluene
131
71
59
T etrachloroethylene
111
84
28
Chlorobenzene
107
46
80
1,1-Dichloroethane
a
137
—
4-Bromofluorobenzeneb
105
57
59
Toluene-d8b
S
87
—
"Not reported.
bMethod surrogates.
131
-------
TABLE 51. VOLATILE ORGANIC CONSTITUENT CONCENTRATIONS FOR DUPLICATE VOST
SAMPLES TAKEN AT THE SCRUBBER EXIT DURING TEST 10 (9/20/88) AND
ANALYZED BY GC/FID AND GC/MS (METHOD 5040)
Flue gas concentration
Gig/dscm)
1st Trap pair 2nd Trap pair
Relative Relative
Percent Percent
Compound GC/FID GC/MS Difference GC/FID GC/MS Difference
Chloroform
22.0
14.7
40
122
2.7
128
Carbon Tetrachloride
39
0.7
139
1.5
<025
—
Benzene
19.2
0.7
186
8.4
<025
—
Bromodichloromethane
12.7
16.9
2a
3.8
4.8
23
Dibromochloromethane
27.3
11.6
81
77.6
7.0
167
Bromoform
11.6
10.2
13
15.7
15.0
4.6
1,2-DichJoroethane
3.2
<0.2
—
2.2
<025
—
1,4-Dichlorobenzene
44
<0.2
—
<1.5
<0.25
0
1,2-Dichlorobenzene
6.5
<0.2
—
4.6
<0.25
—
Toluene
19 1
13.4
35
15.5
10.0
43
Chlorobenzenc
17
1.1
43
<1.6
0.9
0
Tetrachloroethylene
2.7
2.5
7.7
4.8
1.6
102
Surrogate Recoveries (%)

Octane
117
O
—
81
g
—
4-bromofluorobenzene
91
65
33
101
79
24
Toluene-dS4

"GC/FID surrogate only.
'GC/MS surrogate only.
132
-------
63 METALS ANALYSIS
For metals analysis in the laboratory, matrix spike and matrix spike duplicate samples
were prepared in the laboratory and analyzed to assess analytical precision and accuracy. In
addition, field matrix spike and spike duplicate samples were prepared at the IRF and shipped
to the Mountain View laboratory for analysis for the chromium valence state tests. Analysis
results are discussed in the following subsections.
6J.1 Chromium Valence State Tests
Table 52 summarizes the results of the laboratory matrix spike/matrix spike duplicate
analyses for the chromium valence state test series. Indicated in the table are the group of
samples to which the matrix spike samples apply. Matrix spike recovery for laboratory-prepared
total chromium and Cr( + 6) samples ranged from 89 to 115 percent, all within the DQO range
of 75 to 125 percent. The RPD between the values obtained for the matrix spikes and the
matrix spike duplicates ranged from 0 to 24 percent. All values were within the DQO for
measurement precision of 25 percent.
Table 53 summarizes the results of the Cr(+6) spike analyses of field matrix spike
samples prepared at the IRF. Two sets of spike recovery values are noted in the table, one set
obtained from analyses performed 12 days after sample preparation, and one set obtained
69 days after sample preparation. The QAPP for these tests specified an analytical hold time
for Cr( + 6) samples of 7 days between sample collection and analysis. However, between 11 and
12 days actually elapsed between collection and analysis. As a consequence, it was decided to
reanalyze the matrix spike samples after a second period of hold time to evaluate the possible
effects of exceeding hold time on test conclusions.
The data in Table 53 show that Cr(+6) matrix spike recovery ranged from 32 percent
for the feed clay matrix to 90 percent for the kiln ash matrix for the original analysis. A range
of recoveries is noted for the scrubber blowdown matrix. This is because a blank sample of the
actual matrix spiked was not analyzed. Instead, blank samples with background Cr(+6)
concentrations which were known to bracket the actual matrix spiked were analyzed. Thus,
actual matrix spike recovery would been in the range noted in Table 53.
The accuracy DQO for this measurement was 75 to 125 percent recovery. Kiln ash,
aqueous feed spike solution, and possibly scrubber blowdown matrix spike recovery for the
original analyses met this objective. Cr(+6) recoveries from the feed clay spike recoveries were
low. Furthermore, the precision of the measurement was 48 percent RPD, greater than the
25 percent DQO for the measurement.
Matrix spike recoveries were uniformly increased for the second analyses series,
performed 8 weeks after the original analyses. For these analyses spike recoveries were within
the DQO range for all samples except the scrubber blowdown. The precision of the feed clay
matrix spike duplicate analyses, at 14 percent RSD, was also within the DQO.
The fact that better QA measurement parameter performance was achieved after
samples had aged considerably suggests that exceeding sample hold time did not affect the
quality of the data reported. With respect to the inability to uniformly achieve the DQOs in the
original sample analyses, it must be noted that the sample matrix failing the spike recovery and
precision objective was the feed clay matrix Acceptable spike recovery was achieved for the
matrices of more importance in the tests—the matrices whose analyses were used to formulate
133
-------
TABLE 52. SUMMARY OF LABORATORY MATRIX SPIKE SAMPLE ANALYSIS RESULTS
FOR THE CHROMIUM VALENCE STATE TEST SERIES
Test
Analyte
Corresponding
sample
Spike
recovery
m
Duplicate spike
Relative Percent
Difference (%)
1, 2,3
Total Cr
Ash
115
7.9
1
Total Cr
Water2
105
1.9
2
Total Cr
Water*
101
0.2
3
Total Cr
Water8
98
1.6
1,2,3
Total Cr
Afterburner exit
train filter
89
0.9
1
Total Cr
Scrubber exit
train filter
89
0.9
2
Total Cr
Scrubber exit
train filter
89
0.9
3
Total Cr
Scrubber exit
train filter
89
0.9
1
Cr( + 6)
All
94
24
2
Cr( + 6)
All
94
6.2
3
Cr(+6)
All
102
0
"Applies to feed, impinger, and probe wash samples.
134
-------
TABLE 53. CR( + 6) MATRIX SPIKE RECOVERIES
Cr( + 6) Concentration (ppm)

Blank
level
Spiked
level
Analyzed
Spike recovery
(%)
Matrix
Original*
Repeat*
Original*
Repeat*3
Feed clay
2.3
10
3.9
10.8
32
88
Feed clay duplicate
2.3
10
6.4
12.4
52
101
RPD (percent)

48
14
Scrubber blowdownc
0.10 to
0.28
0.10
0.19
0.49
50 to 95
129 to
245
Kiln ash
0.27
10
9.2
10.8
90
105
Aqueous feed spike
—
3,400
2,700
3,100
79
91
4 Analyzed 12 days after sample preparation.
bAnalyzed 69 days after sample preparation (8 weeks after original analysis).
'Matrix blank samples which bracket the matrix spike sample were analyzed.
test conclusions. These were the kiln ash, aqueous feed spike solution, and possibly the scrubber
blowdown. Given this, it is felt that test conclusions are valid as stated in Section 5.
6.32 Parametric Trace Metal Tests
Table 54 summarizes the trace metal recoveries for the laboratory matrix spikes
corresponding to analysis of water-related samples for the parametric test series. These spike
and duplicate spike samples apply to aqueous feed spike solutions, impinger, blowdown and
probe wash samples obtained during the trace metals parametric test series. Matrix spike
recovery for the trace metal constituents ranged from 81 to 123 percent, all within the DQO for
recovery of 75 to 125 percent. The RPD between the values obtained for the matrix spike and
the matrix spike duplicates ranged from 0 to 37 percent. One out of 72 duplicate spike pairs did
not fall within the DQO limit for precision of 25 percent. Measurement completeness was
98 percent and met the DQO for precision completeness of 80 percent.
Table 55 summarizes the trace metal recoveries for the laboratory matrix spikes
corresponding to analysis of filter and ash samples for the parametric test series. For the kiln
ash samples, matrix spike recovery of the trace metal constituents ranged from 72 to 137 percent.
Five of the 27 spike samples were not within the DQO range for recovery of 75 to 125 percent.
Measurement completeness was 81 percent, and met the DQO for recovery completeness of 80
percent. The RPD between the values obtained for the matrix spike and matrix spike duplicates
ranged from 0.3 to 27 percent. One out of 27 duplicate spikes exceeded the DQO for precision
135
-------
TABLE 54. LABORATORY MATRIX SPIKE AQUEOUS SAMPLE ANALYSIS RESULTS FOR THE PARAMETRIC TEST SERIES
Test 4 Test 5 Test 6 Test 7
(9/14/88) (8/25/88) (9/16/88) (8/30/K8)
Lhipll- Dupli- Dupli- Dupli-
cate cute cate cate
Spike spike Spike ipike Spike spike Spike spike
recovery RPD recovery HPD recovery RPD rrcovery HPD
Metal (%) (%) (%) (%) (%) (%) (%) (%)
Test 8 Test 9 Test 10 Test II
(9/07/88) (9/09/88) (9/20/88) (9/22/88)
Dupli- Dopll- Dupli- Dupli-
cate cate cate cate
Spike spike Spike spike Spike spike Spike spike
muverjr RPD recovery RPD recovery RPD recovery RPD
<*>) (%) (*) (%) (%) (%) (%) (%)
Arsenic
86
92
86
59
88
34
95
3J
81
16
86
42
89
93
94
5J
Barium
97
7.6
97
0.6
105
27
101
3.6
105
30
104
0.9
93
42
96
39
Bismuth
123
37
88
17
98
1.0
93
14
90
15
83
0.8
84
7.4
114
13
Cadmium
103
52
101
0.3
106
25
106
28
105
09
102
17
101
2.9
98
4J
Chromium
97
65
97
02
108
26
102
28
104
22
102
0.8
97
3.9
98
2.9
Copper
92
7.6
92
05
102
2.4
96
36
99
23
84
2.0
83
2.9
95
4.0
Lead
105
11
104
19
116
6.4
104
11
106
67
122
21
109
51
100
34
Magnesium
94
53
91
0.5
103
1.0
95
45
107
2.4
98
1.4
91
2 J8
97
21
Strontium
108
5.1
101
0
106
3.1
105
2.9
109
09
103
33
107
2.6
117
6J
-------
TABLE 55. LABORATORY MATRIX SPIKE FILTER AND ASH SAMPLE ANALYSIS
RESULTS FOR THE PARAMETRIC TEST SERIES

Kiln Ash

Afterburner exit
sampling train
niter

Scrubber nit
sampling train
(liter

Teals
4,5,8,9,10,11

Test
6

Test
7

Tests
Afi,73,9,11
Tests
<.7,9

Tests
6,10,11
Metal
Spike
recovery
(%)
Duplicate
spike
KPD
(*>)
Spike
recovery
<%)
Duplicate
spike
HPD
(%)
Spike
recovery
(%)
Duplicate
spike
HPD
(%)
Duplicate
Spike spike
recovery KPD
(%) (%)
Spike
recovery
(%)
Duplicate
spike
RPD
(*)
Spike
recovery
<%)
Duplicate
spike
RPD
<%>
Arsenic
125
6.5
117
91
124
18
105
56
77
19
70
36
Barium
106
1.3
80
21
111
6.9
100
3.7
104
3.7
90
1.2
Bismuth
136
13
101
27
91
3.6
92
22
64
16
93
1.1
Cadmium
86
1.7
93
0J
98
91
110
39
93
0.4
91
15
Chromium
93
5
90
3.1
115
11
105
6.0
94
1.0
89
0.9
Copper
137
5.7
72
1J
91
18
92
4.1
94
2.2
77
19
Lead
111
18
104
1.6
117
11
108
3.5
100
0
96
1.7
Magnesium
81
5.1
84
7J
100
3.4
81
9.7
122
13
102
4J
Strontium
129
7.6
72
4 J
104
20
97
30
104
23
123
14
-------
of 25 percent. Measurement completeness was 96 percent and met the DQO for precision
completeness of 80 percent.
For the filter samples, matrix spike recovery of the trace metal constituents ranged from
64 to 123 percent. Two of 27 spike samples were not within the DQO range for recovery of 75
to 125 percent. Measurement completeness was 92 percent, and met the DQO for recovery
completeness of 80 percent. The RPD between the values obtained for the matrix spike and
matrix spike duplicates ranged from 0 to 23 percent, all within the DQO for precision of
25 percent.
As a further measure of trace metal analytical accuracy and precision, recall from
Section 4.1 that a composite sample of the aqueous metal spike solution for each test was
analyzed for the 9 test trace metals. These can be considered field aqueous solution matrix spike
samples. Comparing analysis results from test to test will give a measure of the accuracy and
precision of the entire aqueous spike solution preparation, sample collection and preservation,
and laboratory analysis process.
Table 19 summarized the results of all aqueous metal spike solution analyses. Table 56
presents these results in different form. Specifically, Table 56 tabulates aqueous spike solution
matrix spike recoveries where recovery in defined as the analyzed concentration divided by the
target preparation concentration. The data in Table 56 show that analysis recoveries ranged
from 67 to 122 percent. Seven of 8 analyses were within the recovery DQO for this
measurement of 75 to 125 percent for all the metals except arsenic. Thus, measurement
completeness was 88 percent for all metals except arsenic, compared to a completeness DQO
of 80 percent. Only 4 of 8 arsenic spike recoveries were between 75 and 125 percent, for a
completeness of 50 percent. Arsenic measurement accuracy failed its DQO on this basis.
However, the failing recoveries were between 71 and 74 percent. These marginally low
recoveries suggest that failure to achieve the recovery DQO for arsenic from this matrix would
have no impact on test program conclusions regarding arsenic distributions.
The data in Table 56 also show that the precision of individual metal analyses for the
aqueous spike solutions ranged from 9 to 18 percent RSD. All precisions were within the DQO
for this measurement of 25 percent. Thus, the measurement precision DQO was met.
Finally, recall from Section 4.1 that clay matrix samples for all parametric trace metal
tests were analyzed for lead, magnesium, and strontium, and for all tests were analyzed for
chromium. Comparing all these analyses gives a measure of the analytical precision of these
metal analyses in the clay matrix. Table 57 provides this comparison. As indicated in Table 57,
the demonstrated analytical precision for the four metals ranged from 3.6 to 16 percent RSD.
Again, all these results were within the precision DQO of 25 percent RSD.
138
-------
TABLE 56. AQUEOUS SPIKE SOLUTION ANALYSIS ACCURACY AND PRECISION
Analysis recovery (%)'
Prepared
spike solution
concentration Test 4 Test 5 Test 6 Test 7 Test 8 Test 9 Test 10 Test 11 USD
Metal (ppm as metal) (9/14/88) (8/25/88) (9/16/88) (8/30/88) (9/07/88) (9/09/88) (9/20/88) (9/22/88) (%)
Arsenic 32)0 87 71 90 74 74 71 111 102 18
Barium 2580 89 70 93 93 97 97 97 89 10
Bismuth 9040 111 73 122 82 104 122 122 122 18
Cadmium 650 93 63 83 82 80 83 98 83 12
Chromium 3230 96 71 90 90 96 93 102 90 10
Copper 32,300 90 71 93 93 99 96 93 93 9
Lead 3230 105 74 93 % 99 96 108 93 11
^ Magnesium 5040 91 67 89 85 87 91 97 85 10
vo Strontium 16,100 112 68 93 87 93 99 118 106 16
^a—^»¦ ' ¦! ¦» ^^^
'Analyzed amount (see Table 19)/prepared amount.
-------
TABLE 57. FEED CLAY ANALYSIS PRECISION
Clay sample metal concentration (ppm)
Test 1/2
composite Test 3 Test 4 Test 5 Test 6 Test 7 Test 8 Test 9 Test 10 Test 11 RSD
Metal (9/28,29/88) (9/26/88) (9/14/88) (8/25/88) (9/16/88) (8/30/88) (9/07/88) (9/09/88) (9/20/88) (9/22/88) (%)
Chromium
54
52
55
55
54
53
54
55
50
50
3.6
Lead
NA"
NA
3.2
3.7
3.5
2.9
3.1
23
3.0
2.5
16
Magnesium
NA
NA
23,000
23,000
22,000
22,000
22,000
22,000
20,000
22,000
42
Strontium
NA
NA
33
37
36
32
33
30
33
35
6.7
"NA: not analyzed.
-------
REFERENCES
1. Test Methods for Evaluating Solid Waste: Physical/Chemical Methods. EPA SW-846, 3rd
ed., U.S. Environmental Protection Agency, Washington, D.C., November 1986.
2. Test Methods for Evaluating Solid Waste: Physical/Chemical Methods. EPA SW-846, 2nd
ed., U.S. Environmental Protection Agency, Washington, D.C., July 1982.
3. Harris, J. C., et al. Sampling and Analysis Methods for Hazardous Waste Incineration.
EPA-600/8-84-002, U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, February 1984.
4. Lentzen, D. E., et al. IERL-RTP Procedures Manual: Level 1 Environmental Assessment
(Second Edition). EPA-600/7-78-201, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, October 1978.
5. 40 CFR Part 60, Appendix A.
6. Determining the Properties of Fine Particulate Matter. ASME Power Test Code 28.
7. Standard Methods for the Examination of Water and Wastewater. 16th Edition, APHA,
AWWA, WPCF, 1985.
8. Barton, R. G., et al. Development and Validation of a Surrogate Metals Mixture. In:
Proceedings of the Fifteenth Annual Research Symposium: Remedial Action, Treatment
and Disposal of Hazardous Waste. U.S. Environmental Protection Agency, Cincinnati, Ohio,
April 1989.
141
-------
technical report data
(P'.caie 't)J !m:rmr:iont on ific rcient it ~ot< fomn/fnr
R£F>0«* NO. 2.
Ei'A/eOO/.-SO/O-iSa
3 P
. T.TlI AN3 SUBTITLE
Tne Fate of Trace Metals in e Notary Kiln Incinerator
with a Ver.tu-i/Packed Column Scrubber - Volume 1:
Technical Results
- 1990
6 PER'OAm.nG CCaNiZaTion COOE
¦ autmorisj
D.J. Fournie'', Jr., W.E. Whitworth, Jr.,
J.W. lei, arc L.R. KateHand
8.PERFORMING ORGANIZATION REPORT NO.
.. P£R*ORVING ORGANIZATION NAME ANO ADDRESS
Acurex Corporation
4S'j Clyde «verue
Mountain View, CA 54339
to program element no.
11. CON I F ACT.grant NC.
68-03-3267
"|2 SrONSOfiiNG AGENCY NAME ANO ADDRESS
Risk RcCucticn Engineering Latoratcry--Cincinr,it:, Or:
Office cf Hesearcr anc Gevelopnen;
U.S. Environmental Protection Agency-
Cincinnati. CH 4j?58
13 type 0* REPORT ANO PERIOD COvEREO
Final - 1988
l< SPONSORING ASENCy coce
ef;/„co/i4
15. Supplementary notes
Gregory J. Carroll, 7PM F7S Phone !fo. 6c4-794b Co.'.nercial Pr.cne No. 512/565-7948
IS. ABSTRACT
A five wee* series of pilot-scale incineration te.sts, usi-i a synthetic waste
feed, was performed at-EPA's Incineration Research Facility to evaluate the fate of
trace metals fed to c rotary kiln incinerator. Eight tests studied the 'ate cf five
hazardous constituent and four nonhazardous constituent trace metals as a function of
incinerator operating temperatures and feed chlorine content. Three tests evaluated the
valance state of chromium in emissions and discharges as a function of feed valance
state and feed chlorine content.
Parametric tests confirmed that cadmiurr,, lead and tismuth are relatively volatile,
based on normalized discharge distribution data. 3ariun:, copper, strontium, chromium
and magnesium are relatively nonvolatile. Apparent scri.bber efficiencies generally
correlated with observed volatilities; collection efficiency was higher for nonvolatile
metals than for volatile metals. Increased feed chlorine content significantly
increased the volatility of cadmium, lead and bismuth.
Chromium test results indicated that with no fee.-i chlorine, 95 percent of the
measured enromium is discharged in the kiln ash. With chlorine in the feed, this
fraction drepped to 85 percent. Kiln csh contained negligible hexavalent chromium
[Cr(+6)] for all tests. The fraction of scrubber exit flue gas chromium as Cr(+6) was
nominally 15 percent with no feed chlorine, increasing to 50 percent with chlorine-
conts i r,i no feed .
KIT WORM AN5 DOCVMIHT anaiyji:

l. DESCRIPTORS
b.i0ENTi*iEWS'0'£'- sNDE3 TSP.V.S
c. COSati Field/CfOup

0
(
1
li. 0'. IT *: 6 UT; on BTATEmSnT
PE.EASt TO PUSLIC
i
18 JIC jRi'tt /Tnit Krporll
w\r
-------