United States
Environmental Protection
Agency
Office of
Research and Development
Washington, DC 20460
EPA/625/R-92/002
May 1992
Technology Transfer
&EPA Handbook
Vitrification Technologies
for Treatment of Hazardous
and Radioactive Waste
-------
-------
EPA/625/R-92/002
May 1992
Handbook
Vitrification Technologies for
Treatment of Hazardous and
Radioactive Waste
U.S. Environmental Protection Agency
Office of Research and Development
Center for Environmental Research Information
Cincinnati, OH 45268
Printed on Recycled Paper
-------
Disclaimer
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
-------
TABLE OF CONTENTS
DISCLAIMER ii
LIST OF TABLES iv
LIST OF FIGURES v
ACRONYMS vi
ACKNOWLEDGEMENTS viii
1 INTRODUCTION 1-1
1.1 Purpose 1-1
1.2 Overview 1-1
1.3 Handbook Organization 1-2
2 GLASS STRUCTURE AND ITS RELATIONSHIP TO VITRIFICATION 2-1
2.1 Glass Structure 2-1
2.2 Stabilizing Mechanisms 2-4
2.3 Chemical Attack Mechanisms 2-5
3 TYPES OF VITRIFICATION PROCESSES 3-1
3.1 Electric Process Heating 3-1
3.2 Thermal Process Heating 3-8
4 APPLICABLE WASTE TYPES AND CONTAMINANTS 4-1
4.1 Applicable Waste Types 4-1
4.2 Applicable Contaminants 4-3
5 PRODUCT CHARACTERISTICS 5-1
5.1 Product Durability 5-1
5.2 Product Volume Reductions and Densities 5-5
5.3 Product Use 5-5
6 OFF-GAS TREATMENT 6-1
6.1 Off-Gas Components 6-1
6.2 Constituents of Concern 6-1
6.3 Means of Off-Gas Control 6-2
7 CAPABILITIES AND LIMITATIONS 7-1
7.1 Capabilities 7-1
7.2 Limitations 7-2
8 PHYSICAL AND CHEMICAL TESTING 8-1
8.1 Physical Tests 8-1
8.2 Chemical Tests 8-3
9 PROCESS EVALUATION 9-1
9.1 Selection of Vitrification Processes 9-1
9.2 Initial Testing and Scaling-Up 9-1
9.3 Cost 9-5
REFERENCES R-1
APPENDIX A A-1
-------
LIST OF TABLES
Page
2-1 Sample Compositions of Soda Lime Glass, Borosilicate Glass, and ISV Glass 2-3
2-2 Inorganic Colorants for Glass j. 2-4
2-3 Effects of Waste-Glass Components on Processing and Product Performance 2-7
3-1 Classification of Vitrification Processes j. 3-1
4-1 Approximate Solubility of Elements in Silicate Glasses 4-4
4-2 Metals Retention Efficiency Test Results for ISV 4-5
4-3 ISV Organic Destruction and Removal Efficiencies 4-10
4-4 Demonstrated Organic Destruction Efficiencies for Vitrification Systems 4-10
5-1 TCLP Leach Data for selected Processes and Selected Metals 5-2
5-2 Strength Comparisons of Waste Glasses Produced by ISV and a JHCM 5-3
6-1 Off-Gas Systems for Selected Processes..! 6-4
6-2 Radionuclide Distribution in the Off-Gas System During an ISV Pilot-Scale Test 6-5
6-3 Hypothetical Distribution of 137Cs Activity in JSV Off-Gas System After Vitrifying 10,000 Ci 6-5
7-1 Comparison of Soil Composition (wt%) from Selected Sites: 7-3
9-1 Determination of Preferred Melter System for Beta-Gamma, Low Level Mixed, Inorganics (Heavy
Metals), Asbestos, Organics, and Soils Waste 9-2
9-2 Criteria Raw Scores and Weighted Overall Scores for the INEL Thermal Process Evaluation Study 9-3
9-3 Testing Units for Developing ISV Technology 9-4
9-4 Major Components of ISV Costs .' 9-6
9-5 Time Requirements for Each ISV Setting...: 9-7
9-6 ISV Electrode Setting and Vitrification Settings 9-7
9-7 Manpower Requirements for ISV Process Preparation 9-7
9-8 Labor Estimate for ISV Processing Operations at a Radioactive Site 9-7
9-9 Power Requirements for ISV Rate as a Function of Moisture Content 9-8
9-10 Sample ISV Cost Estimates 9-8
9-11 ISV Equipment Costs 9-9
9-12 ISV Site Operating Costs 9-10
9-13 Equipment Required for JHCM Processing 9-11
9-14 Capital Cost Summary for JHCM 9-12
9-15 Comparison of Capital Costs and Operating Costs for a JHCM 9-12
9-16 Throughput Rate for Selected Vitrification Processes 9-12
IV
-------
LIST OF FIGURES
Page
2-1 Silicon-Oxygen Tetrahedron 2-1
2-2 Example Silicate Glass Network Structure 2-2
3-1 Relationship Between Resistivity and Temperature for Selected Glasses 3-2
3-2 TypicalJHCM Process Flowsheet 3-3
3-3 Generalized JHCM Showing Components of Melter and Molten Materials 3-5
3-4 Schematic of ISV 3-5
3-5 Pilot-Scale Process ISV 3-6
3-6 Comparison of a Transferred Arc and a Non-Transferred Arc 3-7
3-7 Schematic of the Demonstration PCR Showing the Bottom-Pour Configuration for Exit Gas
and Molten Glass 3-8
3-8 Schematic of a Full-Scale PCR 3-9
3-9 Microwave Melter 3-9
3-10 Flow Diagram of the IRI Process 3-12
3-11 Simplified System Schematic of MSP's Process 3-13
4-1 Element Retention Versus Burial Depth During Pilot-Scale ISV Tests 4-6
5-1 Leach Resistance of Selected Materials 5-2
7-1 Schematic of a Pilot-Scale ISV Hood Assembly 7-5
7-2 Predicted Versus Achieved Large-Scale ISV Melt Shape 7-7
7-3 The Effect of a Molten Metal "Passive" Electrode on Electrical Current Distribution in an ISV Melt 7-7
-------
ACRONYMS
AC Alternating Current
AEDC Arnold Engineering Development Center
ANS American Nuclear Society :
ASTM American Society for Testing Materials
AVM Atelier de Vitrification Marcoule !
BOAT Best Demonstrated Available Technology
CERCLA Comprehensive Environmental Reponse, Compensation, and Liability Law
CMS Cyclone Melting System
DC Direct Current
DE Destruction Efficiency
DNT Dinitrotoluene
DOD Department of Defense
DOE Department of Energy
ORE Destruction and Removal Efficiency
DWPF Defense Waste Processing Facility
EDTA Ethylenediaminetetraacetate
EMF Electromotive Force
EFS Electrode Feed System
EP Tox Extraction Procedure Toxicity
EPA Environmental Protection Agency
GRI Gas Research Institute
HEPA High-Efficiency Paniculate Air
HLLW High-Level Liquid Wastes
HLW High-Level Waste
IAEC International Atomic Energy Commission
INEL Idaho National Engineering Laboratory
IRI Inorganic Recycling, Inc.
1SV In Situ Vitrification
JHCM Joule Heated Ceramic Melter i
LFCM Liquid Fed Ceramic Meiter
LX Leachability Index
MCC Materials Characterization Centej-
MIIT Materials Interface Interactions Tests
MSP Marine Shale Processes '
MSW Municipal Solid Waste
NBS National Bureau of Standards
NEPA National Environmental Policy Act
ORNL Oak Ridge National Laboratory |
OSHA Occupational Safety and Health Administration
OSWER Office of Solid Waste and Emergency Response
PCB Polychlorinated Biphenyl '
VI
-------
ACRONYMS (continued)
PCR Plasma Centrifuge Reactor
PCT Product Consistency Test
PEC Plasma Energy Corporation
PNC Power Reactor and Nuclear Fuel Development Corporation
PNL Pacific Northwest Laboratory
RCRA Resources Conservation and Recovery Act
ROW Recomp of Washington
RWMC Radioactive Waste Management Complex
SITE Superfund Innovative Technology Evaluation
SRL Savannah River Laboratory
SRS Savannah River Site
TCLP Toxicity Characteristic Leaching Procedure
TNT Trinitrotoluene
TRU Transuranic
USATHMA United States Army Toxic and Hazardous Materials Agency
USEPA United States Environmental Protection Agency
UST Underground Storage Tanks
V/S Volume of leaching solution/Surface area of solids
VOC Volatile Organic Compound
WIPP Waste Isolation Pilot Plant
ZHE Zero Headspace Extractor
VII
-------
ACKNOWLEDGEMENTS
This document is a compilation of material applicable to the vitrification of hazardous wastes. The information was
compiled and organized by Timothy Voskuil of EQUITY ASSOCIATES, Inc., Knoxville, Tennessee under the technical
direction of Ed Barth of the Center for Environmental Research Information, Cincinnati, Ohio. Many individuals
contributed to the preparation and review of this document; a partial listing of reviewers appears below.
Technical Reviewers:
Dennis F. Bickford, Savannah River Site, Aiken, South Carolina
Tony Eicher, Focus Environmental, Knoxville, Tennessee
Jeff Means, Battelle, Columbus, Ohio j
J.M. Perez, Pacific Northwest Laboratory, Richland, Washington
M J. Plodinec, Savannah River Site, Aiken, South parolina
J.W. Shade, Westinghouse Hanford Company, Richland, Washington
Brian Spalding, Oak Ridge National Laboratory, Oak Ridge, Tennessee
Laurel Staley, EPA Risk Reduction Engineering Laboratory, Cincinnati, Ohio
L.E. Thompson, Pacific Northwest Laboratory, Richland, Washington
Peer Reviewers:
Bill Bonner, Pacific Northwest Laboratory, Richland, Washington
Jim Cudahy, Focus Environmental, Knoxville, Tennessee
George Wicks, Savannah River Site, Aiken, South| Carolina
VIII
-------
CHAPTER ONE
INTRODUCTION
1.1 Purpose
Vitrification, the process of converting materials into a
glass or glass-like substance, is increasingly being con-
sidered for treating various wastes. Vitrification is con-
ceptually attractive because of the potential durability of
the product and the flexibility of the process in treating a
wide variety of waste streams and contaminants. These
characteristics make vitrification the focal point of treat-
ment systems for high-level radioactive waste (HLW)
around the world.
This handbook presents the theory behind the vitrification
process and overviews the applications and limitations of
vitrification for waste treatment. Accordingly, it classifies
the types of vitrification processes which have been
applied to waste treatment, explains why vitrification may
be considered as a treatment process, identifies waste
streams and contaminants to which vitrification may be
applied, and discusses other issues important to the
application of vitrification to waste treatment.
In overviewing vitrification in this way, this handbook
relies primarily on publicly available information and re-
ports. Rather than attempting to evaluate the quality of
such information, this document simply presents the
conclusions as stated in the reports. Wherever possible,
it avoids unpublished vendor information and personal
communications; however, it includes some vendor infor-
mation in order to provide information and direction not
publicly available. Even so, it does not use vendor
information as the sole source to make new, unique, or
unsupported claims about vitrification processes.
Vitrification may proceed in situ (in situ vitrification or ISV)
or ex situ. ISV and ex situ vitrification have opposing
strengths and weaknesses. An advantage of ISV is that
it proceeds in situ without requiring that the material be
removed prior to treatment. With very hazardous con-
taminants, such as radionuclides or dispersible volatile
compounds, this may be a significant advantage. Fur-
thermore, the product remains buried under ground and
on-site, thus limiting liability by keeping the waste product
on-site. Ex situ vitrification, on the other hand, does
require excavation, or at least increased materials han-
dling prior to vitrification, but is not limited to waste in the
immediate area of electrode placement.
The advantages of ex situ over in situ lie primarily in the
increased amount of control that can be exerted during
processing. This control extends to feed composition and
melt conditions and this in turn allows for greater control
of product characteristics. For processes aimed at a
specific waste stream, control is increased because of
the relative homogeneity of the waste stream. Secondly,
ex situ vitrification allows greatercontrol overthe combus-
tion of non-pyrolyzed organics escaping from the melt.
For these processes, the environments in the molten
glass melt and in the secondary combustion area can
be more easily regulated to facilitate efficient organic
destruction. However, ISV does operate at higher tem-
peratures (typically 1600-2000°C) than ex situ processes
(typically 1000-1600°C) and thus may sometimes avoid
the need for fluxants.
1.2 Overview
Glass is a rigid, noncrystalline material of relatively low
porosity. It is often composed of constituents such as
oxides of silicon, boron, aluminum, and alkali and alkaline
earth elements. While phosphate, sulfide, and oxynitride
glasses are also important glass types, most glasses
used in waste immobilization are borosilicate, sodium
silicate, or aluminosilicate glasses and this handbook
limits its review to these glasses.
Vitrification is the process of converting materials into a
glass or glassy substance, typically through a thermal
process. Although heat is not necessarily required for
vitrification (for example, vapor deposition, solution hy-
drolysis, and gel formation can also form glassy materi-
als), this document considers only vitrification processes
which use heat.
When accomplished through a thermal process, vitrifica-
tion may destroy organic contaminants via pyrolysis or
1 -1
-------
combustion. As a stabilization process, vitrification may
immobilize inorganics by incorporating them into the
glass structure or by encapsulating them in the product
glass. Glass's ability to incorporate metals has a long
history: selenium isfound in clear glass bottles; chromium
gives green glass its color; and lead oxide, at levels up to
30%, is found in lead crystal glass (McLellan and Shand,
1984). I
Many contaminated materials contain adequate quanti-
ties of the raw ingredients needed for forming [glass.
When such materials are heated, the ingredients melt
together and actually form the glass in which the contami-
nants are immobilized. Because not all contaminated
materials do contain proper ratios of the materials for the
formation of a glass, additives may be required for some
materials to address these deficiencies. Additives may
also be needed to create the special characteristics of
some glasses. ;
Vitrification has four major advantages over other 'meth-
ods of waste management. The primary advantage is the
durable waste glass it produces. In most instances, this
waste glass performs exceptionally well in leach tests.
Because of its chemical and physical durability, the
vitrification product has been considered for recycling as
aggregate and other products. The second major ad-
vantage of vitrification is the flexibility of the waste, glass
in incorporating a wide variety of contaminants and
accompanying feed material in its structure without a
significant decrease in quality. The third advantage js that
vitrification processes can accommodate both organic and
inorganic contaminants of various amounts. Lastly, vitri-
fication may reduce the volume of waste material.!
Vitrification's major limitation is that it is energy intensive
and, thus, may be more expensive compared to other
remedial technologies. A second major limitation [is the
potential for some contaminants, both organic and inor-
ganic, to volatilize. This limitation applies to both ex situ
processes and ISV. For ISV, there is some concern that
certain contaminants may migrate into the surrounding
soil. These limitations may be amenable to modification
of process parameters given site characteristics and
management goals.
Given these advantages and limitations, vitrification's
niches may include waste with great potential hazard,
waste which is highly concentrated, waste with a complex
mix of contaminants, specialized industrial waste streams,
and wastes where a high quality product is required. For
example, vitrification is the treatment of choice for!high-
level waste (HLW) and is well worth the associated 'costs
intreatingthiswaste. Infact, EPA has declared vitrification
to be a Best Demonstrated Available Technology (B^DAT)
for HLW (40 CFR 268.42, Table 3). Site conditions that
may make any vitrification process attractive are low unit
costs for electricity. ISV may be attractive at sites where
concerns about company liability from off-site disposal
drive treatment objectives.
1.3 Handbook Organization
Chapter One. "Introduction," (the present chapter) identi-
fies the purpose of this handbook, defines vitrification as
it is used in this handbook, overviews vitrification applica-
tions, and summarizes handbook organization.
Chapter Two. "Glass Structure and Its Relationship to
Vitrification," describes the structure of glass and relates
this structure to the vitrified product. Because of its
amorphous, non-crystalline structure, glass can immobi-
lize a wide variety of inorganic contaminants, either by
chemical incorporation into the actual glass matrix or
physical encapsulation. Leaching of immobilized
inorganics occurs via matrix dissolution and
interdiffusion.
Chapter Three. "Types of Vitrification Processes," classi-
fies vitrification into two majorcategories: electric process
heating and thermal process heating. Electric process
heating is the more important vitrification category and
includes joule heating processes, both ex situ and in situ,
plasma heating processes, microwave heating processes,
as well as several miscellaneous heating processes.
Thermal process heating is dominated by processes
using rotary kiln incinerators, but otherthermal processes
may also be applicable to waste vitrification.
Chapter Four. "Applicable Waste Types and Contami-
nants," discusses waste streams and contaminants to
which vitrification may be applicable. Waste streams
discussed include radioactive wastes and sludges, con-
taminated soils, contaminated sediments, incinerator
ashes, industrial wastes and sludges, medical wastes,
underground storage tanks (USTs), drummed wastes,
shipboard wastes, and asbestos wastes. Contaminants
discussed include inorganics (metals, radioactive wastes,
asbestos, and others) and organics. This chapter is
meant to give the readeran understanding of vitrification's
potential; it is not necessarily comprehensive or limiting in
its scope.
Chapter Five. "Product Characteristics," addresses vari-
ous components of product quality, volume reductions
achieved with vitrification, and potential uses of the prod-
uct glass. Generally, the vitrified waste is a high quality
product. Waste glasses have performed well in a variety
of leach tests, thus indicating high chemical durability.
They have also shown high physical integrity and gener-
1 -2
-------
ally perform well even when devitrified, or re-crystallized.
Furthermore, estimates indicate that waste glasses may
exhibit these properties over geologic time spans.
Chapter Six. "Off-Gas Treatment," discusses off-gas con-
cerns by describing typical off-gas components, off-gas
constituents of concern, and potential means of off-gas
control. Off-gas control may be approached in two ways:
reducing emissions and treating evolved off-gases. Re-
ducing emissions is accomplished through control of
various process parameters. Methodsfortreating evolved
off-gases are similar to those for other waste treatment
processes.
Chapter Seven. "Capabilities and Limitations," summa-
rizes the capabilities and limitations of vitrification.
Chapter Eight. "Physical and ChemicalTesting,"describes
the physical and chemical tests used to determine the
properties of targeted waste streams and of waste glasses.
Described tests include, but are not limited to, the leach
tests so important in determining waste form quality.
Chapter Nine. "Process Evaluation," addresses various
issues which may be important in selecting a vitrification
technology. Presented in this chapter are examples of
technology screening studies, initial testing and scaling-
up concerns, and a discussion of cost components. The
discussion of cost components emphasizes cost catego-
ries and their relative importance to total clean-up costs.
1 -3
-------
-------
CHAPTER TWO
GLASS STRUCTURE AND ITS RELATIONSHIP TO VITRIFICATION
Vitrification is attractive as a waste treatment process
primarily because of the properties of glass. These
properties give vitrification its high-quality product and
flexibility in inorganic incorporation. Because an under-
standing of the properties of glass is fundamental to
understanding the advantages of vitrification as a waste
treatment process, this chapter offers a brief overview of
glass structure and discusses how this structure relates to
thedurability of vitrified glass containing hazardous waste.
This chapter is summarized mainly from McLellan and
Shand (1984). Sections 2.3.3 and 2.3.4 are summarized
from Wicks (1985,1986). Because this chapter deals with
the incorporation of inorganic contaminants into the glass
structure, organic contaminants will not be addressed
here.
2.1 Glass Structure
Glass is a rigid, noncrystalline material of relatively low
porosity, often composed primarily of silica, alumina, and
oxides of alkali and alkaline earth elements. While
phosphate, sulfide, and oxynitride glasses are also im-
portant glass types, most glasses used in waste immo-
bilization are silicate glasses. Therefore, this handbook
limits its review to silicate glasses.
Thermally-formed glasses are produced by fusing or
melting crystalline materials and/or amorphous materials
(e.g., previously formed glasses) at elevated tempera-
tures to produce liquids. These liquids are subsequently
cooled to a rigid condition without crystallization. Most
thermally-formed waste glasses, however, also have a
crystalline phase. For example, while the ISV product is
substantially glassy, it is actually a mixture of glass and
microcrystalline phases. Glass composition is largely
inorganic, with silica (SiO2) being the most common con-
stituent. From an engineering standpoint, what distin-
guishes glass from crystalline substances is the lack of a
definite melting point temperature. When glass is heated,
it will gradually deform and, at high enough temperatures,
form a viscous liquid.
Silicate glasses are not composed of discrete molecules,
but are three-dimensional networks. The basic structural
unit of the silicate network is the silicon-oxygen tetrahe-
dron in which a silicon atom is bonded to four oxygen
atoms (Figure 2-1). The silica tetrahedra are linked at the
corners, where each shares one oxygen atom with an-
other tetrahedron (Figure 2-2). Some, or all four, of the
oxygen atoms from the tetrahedron can be shared with
other tetrahedra to form a three-dimensional network.
What prevents these tetrahedra from forming a crystalline
network is that the extended 3-dimensional network is
irregular and the Si-O-Si bonds random (McLellan and
Shand, 1984).
The shared oxygen atoms are called bridging oxygens. In
pure silica glass, the ratio of silicon to oxygen is ideally 1:2
and all oxygen atoms are bridging. Some atoms, such as
sodium, are ionically bonded to oxygen when present in
Figure 2-1. Silicon-Oxygen Tetrahedron (McLellan
and Shand, 1984)
2-1
-------
Figure 2-2. Example Silicate Glass Network
Structure (McLellan and Shand, 1984)
glass and thus interrupt tetrahedra linking and the conti-
nuity of the network. An oxygen atom ionically bonded to
another atom is called nonbridging. '.
Appreciable amounts of most inorganic oxides can be
incorporated into silicate glasses. Elements that can
replace silicon are called network formers. By replacing
silicon in the glass network, some inorganic species (such
as some metals found in the middle portion of the periodic
table) can be incorporated into a glass. Most mondvalent
and divalent cations (such as sodium, calcium, and some
other metals and metalloids grouped near either side of
the periodic table) do notenterthe network, but form ionic
bonds with nonbridging oxygen atoms, and are termed
network modifiers. The effect of variation in the network
integrity and the constituents of the glass are manifested
in changes in glass properties such as softening point
temperature and chemical durability (i.e., teachability and
solubility) (McLellan and Shand, 1984).
The role of elements in the glass may vary with conditions.
For example, aluminum may be a network former or a
modifier depending on the ratio of aluminum to alkali and
alkaline earth ions and is thus called an intermediate.
The role of iron depends on redox state or oxygen
availability in the molten material. For example, Fe(lll) is
a network former (McLellan and Shand, 1984). |
i
Because of the network structure of glasses, it doks not
help to express their composition as chemical formulae.
The most common way of describing glass is to list
relative amounts of oxides derived from the raw materials
used in a glass formulation, even though these oxides do
not exist, per se, in the glass network.
Many types of glass can be formed depending on the raw
materials used. The glass industry prepares special
formulations to obtain glasses with properties desirable
for various uses. Important considerations for the treat-
ment of hazardous wastes include processing character-
istics, such as melt viscosity and redox conditions, and
product characteristics, such as durability.
Vitreous silica, the simplest glass, can be prepared by
heating silica (SiO2) above its melting point and quickly
cooling to the solid state. In order to decrease the
viscosity of molten glass from that of pure silica and allow
it to melt at a lower temperature, it is necessary to add a
flux, or network modifier, that will soften the glass by
generating nonbridging oxygen atoms. Alkali metals,
such as sodium, make excellent fluxes in their oxide
forms.
Alkalis can be incorporated into the glass as carbonates
or other salts that react, at elevated temperatures, with
silica to form a siliceous liquid. The reaction of fluxes is
complex, but aside from lowering the viscosity of the
glass, they also have the effect of lowering the melting
point of the raw material mix. This helps decrease the
energy requirements of the melting process.
Unfortunately, adding alkali to the glass generally de-
creases its chemical resistance from that of silica glass.
At high alkali concentrations, the glass will even become
water soluble (the basis for the soluble silicate industry).
To decrease the aqueous solubility of alkali glasses, but
to maintain the lower melting points, alkaline earth fluxes
may also be used. Oxides of calcium and magnesium are
the most common alkaline earth or stabilizing fluxes.
However, adding too much calcium can cause calcium
silicates and aluminatestoform and these may crystallize
(devitrify) on cooling.
Soda ash (sodiumcarbonate)iscommonly used in industry
to supply alkali fluxes, while lime (calcium oxide) is
commonly added to supply alkaline earth fluxes. Thus,
glass made from silica and alkali and alkaline earth fluxes
is commonly called soda-lime glass. Soda-lime glass is
the most common type of glass, and is used in most
container glass and window glass applications. The
typical composition of soda-lime glass is compared with
the composition of two waste glasses in Table 2-1.
Typical raw materials for industrial glass making consist
of various formulations of the following main ingredients:
Sand - SiOa
Feldspar -
2-2
-------
Table 2-1. Sample Compositions of Soda-Lime Glass, Borosilicate Glass, and ISV Glass
Oxide
SiO2
AI203
Na20
K20
MgO
CaO
BO
Fe203
FeO
La203
Li20
MnO
NiO
Ti02
Zr02
SrO
BaO
1From McLellan
2From Goldston
3This glass was
Typical
Soda-Lime Glass1
(wt %)
65-75
1-2
12-16
0.1-3
0.1-5
6-12
-
-
-
-
-
.
-
-
-
-
-
and Shand, 1984.
and Plodinec, 1 991 .
produced by ISV of INEL soils. From
SRS Borosilicate
Benchmark Glass2
(wt %)
48.95
3.67
16.71
0.04
1.66
1.13
11.12
8.08
0.89
0.41
4.28
1.34
0.61
0.71
0.41
-
-
Farnsworth, Oma, and Reimus, 1990.
Sample
ISV Glass3
(wt %)
71.20
13.50
1.55
2.47
1.87
3.58
-
4.63
-
-
-
0.11
0.12
0.76
0.07
0.02
0.10
Dolomite - CaMg(COs)2
Limestone - CaCOs
Soda ash -
These are mixed with a variety of other constituents to
produce glasses with whatever physical and chemical
properties manufacturers may desire, such as heat resis-
tance, chemical inertness, various optical properties,
various colors, etc. The selection of materials from which
to make a waste glass, on the other hand, generally
involves compromises based on the product and process-
ing characteristics desired.
While soda-lime glass may serve as a waste glass, many
waste glasses are borosilicate glasses and contain 6203.
Waste glasses also generally contain less silica and more
aluminum and iron than soda-lime glasses. Most soils
and the ISV glass derived from their melting also have
more aluminum and iron and less silica and sodium than
typical soda-lime glass. The "aluminum-bearing glasses"
are generally more typical of glass compositions pro-
duced in waste vitrification.
It is interesting to note that many metals of environmental
concern are readily incorporated into a glass matrix and
are commonly used as colorants in glassmaking. This
suggests that vitrification processes may be particularly
attractive for immobilizing metals found in certain waste
streams. Table 2-2 presents a list of metal compounds
commonly used as glass colorants.
2.2 Stabilizing Mechanisms
Hazardous constituents can be immobilized in vitrification
processes by two main interactions with the glass matrix:
Chemical bonding
Encapsulation
Certain inorganic species can be immobilized by chemi-
cal bonding with the glass-forming materials, particularly
2-3
-------
silica, present in the wastes to be vitrified. The most
notable chemical bonding within a vitrified material oc-
curs when certain metals or other inorganics bond cova-
lently with the oxygen atoms in a silica network and thus
become part of the network. Inorganics that interact in
this way are network formers since they essentially re-
place silicon in the glass network structure. \
\
Other inorganic species can bond ionically with oxygen or
other elements in the glass network. This ionic bonding
incorporates the material into the glass but disrupts the
network's continuity, thereby modifying the vitrified
material's physical and chemical properties. As men-
tioned earlier, materials that interact in this fashion are
called network modifiers. [
Hazardous constituents may also be immobilized without
direct chemical interaction with the glass network. [Since
vitrification constitutes a molten phase during some portion
of the process, materials that do not interact chemically or
have not completely entered solution can be surrounded
by a layer of vitrified material and encapsulated, as the
melt cools. This layer of vitrified material protects the
encapsulated constituents from chemical attack and in-
hibits their ability to escape from the vitrified product
(McLellan and Shand, 1984).
2.3 Chemical Attack Mechanisms
Vitreous materials are often thought of as being "inert,"
which is somewhat justified since these materials exhibit
high corrosion resistance compared with many other
materials. It is important to note, however, that all vitrified
products are chemically reactive to some degree. This
section discusses the nature of chemical attack on vitre-
ous silicate materials and the factors that affect the rate
and degree of attack.
There are two major forms of chemical attack on vitrified
materials:
Matrix dissolution
Interdiffusion
2.3.1 Matrix Dissolution
Matrix dissolution is characterized by alkali attack. It
begins by hydration of the silica network and may proceed
todissolution of the vitreous material. In pure silicaglass,
the matrix dissolution process can be described by the
following equation:
2 NaOH + SiO2 -> Na2SiO3 + H2O
The alkali silicate (Na2SiO3 in the example shown) is
water soluble, so as the silica network is attacked and
dissolved congruently, the other constituents in the vitri-
fied material are released. The rate of alkali attack is
generally linear with time; however, the rate can change
if soluble materials accumulate in solution, or if insoluble
reaction products adhere to the material's surface, blocking
the reaction.
Table 2-2. Inorganic Colorants for Glass (Tooley, 1984)
Color Produced
Material
Under Oxidation
Under Reduction
Cadmium Sulfide
Cadmium Sulfide, Selenium
Cobalt Oxide
Copper Oxide
Cuprous Oxide
Cerium Oxide
Chromic Oxide
Gold
Iron Oxide
Manganese dioxide
Noodymium oxide
Nickel oxide
Nickel oxide
Selenium
Sulfur
Uranium
None
None
Blue-violet
Greenish blue
Gjreenish blue
T|tania Yellow
Ypllowish green
Ruby
Yellowish green
Amethyst to purple
Violet
Violet in K2O glass
Brown in Na2O glass
Fugitive
Npne
Yellow with green fluorescence
Yellow
Ruby
Blue-violet
Greenish blue
Ruby
Yellow
Emerald green
Bluish green
None
Violet
Violet in K2O glass
Brown in Na2O glass
Pink
Yellow to amber
Green with fluorescence
2-4
-------
Alkali attack is highly pH dependent. The rate of attack
generally increases by a factor of 2 to 3 for each pH unit
increase. The influence of temperature on the rate of
alkali attack follows an Arrhenius relationship with the rate
of attack increasing by a factor of 2 to 2.5 for each 10° C
temperature rise.
2.3.2 Interdiffusion
Interdiffusion is typified by acid attack on vitrified materi-
als. While alkali attack (matrix dissolution) leads to
surface dissolution of the vitreous material, interdiffusion
is an ion exchange process which preferentially extracts
elements present as network modifiers, leaving the silica
structure almost intact. Generally, interdiffusion involves
the exchange of hydronium ions in solution for ionically
bonded elements in the vitreous network (McLellan and
Shand, 1984).
Interdiffusion has sometimes been called leaching, but
interdiffusion is the more precise term. "Leaching" is
commonly used to denote loss of constituents from a
material without specifying a mechanism. As used here,
interdiffusion is a mechanism; thus, to call it "leaching" is
confusing.
The reaction rate in interdiffusion is influenced by tem-
perature in a relationship similar to that for alkali attack;
however, the interdiffusion reaction rate increases only by
a factor of 1.5 to 2 for each 10°C temperature rise.
Depending on the composition of the vitrified material,
especially its silica content, the pH of the leaching solution
influences the rate of acid attack. Generally, that influ-
ence is not as strong as the influence on the rate of alkali
attack.
The rate of acid attack on glass is generally proportional
to the square root of time. Since the process is controlled
predominantly by diffusion, the rate of leaching decreases
as the thickness of the leached layer near the glass
surface increases. However, this effect can be limited if
the layer dissolves or sloughs off.
The teachability of trace constituents is difficult to predict,
but it is reasonable to assume that in addition to the alkali
and alkaline earth elements (sodium, potassium, cal-
cium) there may be preferential extraction of other network
modifiers of potential environmental concern, such as:
barium (Ba), beryliium (Be), cobalt (Co), copper (Cu), lead
(Pb), magnesium (Mg), manganese (Mn), nickel (Ni),
silver (Ag), strontium (Sr), and zinc (Zn).
Water attacks vitrified materials to some degree, although
the attack is much less aggressive than that of alkali and
is generally less vigorous than acid attack. Water can
exhibit both acid and alkali attack mechanisms since it can
produce both hydronium and hydroxyl ions. However, in
a static environment water attack quickly becomes alkali
attack as the alkali present in the glass is extracted into the
water and then takes part in the reaction.
Attack by salt solutions is thought to correspond to the
attack by water, but the mechanism has not been thor-
oughly defined. However, typical attack rates at room
temperature are still very low. Many chelating compounds
attack glasses at a rate comparable to that of strong alkali.
Citrate, gluconate, oxalate, tartrate, EDTA, and malate all
attack glass in alkaline solution. Alkaline phosphate and
acetate also attack glass readily. Hydrofluoric acid has a
unique ability to dissolve silicate glasses, forming a solu-
tion of alkali fluorides and silicon fluorides.
2.3.3 Three -Stage Model of Waste Glass
Corrosion
While dissolution and interdiffusion describe leaching
under many conditions, the leaching of many waste
glasses appears to be modified by the formation of
surface gel layers (Wicks, 1985). Layer formation is
favored in static or near-static conditions and where silica
is present, as in many groundwaters. As matrix dissolu-
tion occurs, the surface layers, composed of insoluble
glass components, arise. The formation of these layers
proceeds in a three-stage process.
Stage one is dominated by interdiffusion as network
modifiers, such as sodium, diffuse out of the glass and
into solution, and water diffuses in. The result is a
modifier- deficient surface layer. During this stage the pH
of the leachant increases (becomes more basic), because
alkali hydroxides form in solution.
Stage two is dominated by matrix dissolution. As de-
scribed earlier, matrix dissolution is an alkaline attack;
thus, its rate is primarily governed by the pH of the
leachate, glass composition, and temperature.
Stage three is characterized by the formation of surface
layers. These surface layers are formed from the precipi-
tation and adsorption of insoluble compounds onto the
surface of the glass. These compounds are the more
insoluble waste glass constituents that are "left behind" as
more soluble constituents dissolve and move into solutions.
For example, these surface layers may contain substantial
iron and manganese hydroxides. Where a surface layer
forms, it can exert astrong limiting effect on leaching of the
waste glass underneath. Under static or near-static
conditions, leaching may be reduced further as silica
concentrations build up in the leachate and approach
2-5
-------
saturation, thus reducing even more the tendency of silica
in the glass to move into solution. !
i
2.3.4 Factors Impacting Waste Glass Leaching
The use of vitrification to treat HLW has produced a
wealth of knowledge about waste glasses and their
production, particularly in terms of chemical composition,
waste loading, temperature, time and pH.
Chemical Composition. Chemical composition plays an
important role in product durability (Wicks, 1985). In
general, as the ratio of oxygen to network formers (such
as silicon) decreases, more bridging oxygens are; pro-
duced, resulting in a more durable product. Netj/vork
modifiers such as alkalis and alkali earth oxides tend to
decrease glass durability. This occurs because these
oxides increase the oxygen-to-network former ratio and
produce more singly-bonded oxygen, thus breaking up
the glass network. However, these elements do lower
melt viscosity and lower processing temperatures and
therefore have potential as fluxing agents. In general,
oxides with valences greater than 1 may increase glass
durability.
Composition of the incoming feed can have enornSous
effects on product durability and processing parameters.
Table 2-3 displays some of the effects of various inorganic
oxides on processing and glass durability. Modification of
the waste stream through additives and/or material re-
moval can have dramatic impacts on processing and
product characteristics. However, as Table 2-3 shows,
most additives have both desired and undesired effects.
Therefore, modification of the feed will often involve
compromises based on treatment goals, processing
limitations, and waste character.
Waste Loading. Increased waste loading does notjnec-
essarily increase product teachability (Wicks, 1985;
Mendel, 1973). Research on borosilicate glass for the
immobilization of nuclear waste has indicated that glass
teachability is reduced as the waste loading increases
from 0 wt% to 35 wt%, with only small changes in
teachability as the waste loading increases from 35jwt%
to 50 wt% (Rankin and Wicks, 1983). Thus, the amount
of waste immobilized by borosilicate glass may not be
limited by product durability, but by processing consid-
erations. The reason forthe beneficial effects of increased
waste loading on durability is due to the formation of
surface layers that form during leaching and thai are
made up of the major constituents found in the waste
composition.
Temperature. Leachability of waste glass increases with
temperature (Wicks, 1985). The mechanism of corrosion
varies With temperature: at temperatures near ambient
conditions, diffusion effects can dominate glass corro-
sion, but at temperatures near 100°C or higher, network
dissolution can dominate. The exact temperature forthe
shift in mechanism varies with test conditions and glass
composition.
Time. At a given temperature, the largest leach rates
occur during the early stages of leaching (Wicks, 1985).
Therefore, leach rates usually decrease over time. Two
mechanisms appear to be involved in this leach rate
decrease. First, under static or near static conditions,
such as groundwater in proposed repositories, the solu-
tion becomes saturated as elements are extracted from
the glass and enter solution. Increased saturation
reduces the solution's solubility and its ability to corrode
the glass. Secondly, with time, a layer forms on the
glass's surface, thereby further inhibiting leaching
(Jantzen, 1988).
pH. In solutions of about pH 3 to 9, glass leaching may be
substantially or minimally affected by solution pH, de-
pending on the chemical composition of the glass (Wicks,
1985). At pH values above 9 (basic conditions), two
mechanisms function to increase leaching: silica solubility
increases and matrix dissolution dominates. The effect of
acidic conditions on glasses varies more than the effect of
basic conditions. Most silicate glasses are dominated by
interdiffusion at low pH values. For these glasses, leach
rates are proportional to the square root of time and the
effect of low pH values is small. However, borosilicate
glasses are dominated by matrix dissolution at low pH
values. Their leach rate increases linearly with time and
the effect of acid attack (below pH 5) may be quite
dramatic. Thus, the expected pH of the disposal site or
use location may be important in determining the desired
composition of the waste glass.
2-6
-------
Table 2-3. Effects of Waste-Glass Components on Processing and
Product Performance (adapted from Plodinec, Wicks, and Bibler,1982).
Frit Components
Processing
Product Performance
Si02
8203
Na2O
Li20
K20
CaO
MgO
Ti02
Increases viscosity greatly; reduces waste solubility
Reduces viscosity; increases waste solubility
Reduces viscosity and resistivity;
increases waste solubility
Same as Na2O, but greater effect;
increases tendency to devitrify
Same as Na2O; decreases tendency
to devitrify
Increases then reduces viscosity
and waste solubility
Is same as CaO; reduces tendency to vitrify
Reduces viscosity slightly;
increases then reduces waste solubility;
increases tendency to devitrify
ZrO2, La2O3 Reduces waste solubility
Increases durability
Increases durability in low amounts,
reduces in large amounts
Reduces durability
Reduces durability, but less than
Na2O
Reduces durability more than Na2O
Increases then reduces durability
Is same as CaO, but more likely
to decrease durability
Increases durability
Increases durability greatly
Waste Components
Processing
Product Performance
AI2O3 Increases viscosity and has tendency to devitrify
Fe2Os Reduces viscosity; is hard to dissolve
UsOs Reduces tendency to devitrify
NiO Is hard to dissolve; increases tendency
to devitrify
MnO Is hard to dissolve
Zeolite Is slow to dissolve; produces foam
Sulfate Is an antifoam, melting aid; increases
corrosion of processing equipment
Increases durability
Increases durability
Reduces durability
Reduces durability
Increases durability
Increases durability
Too much causes foam or
formation of soluble second phase
2-7
-------
-------
CHAPTER THREE
TYPES OF VITRIFICATION PROCESSES
This report divides vitrification technologies into two cat-
egories: electric process heating and thermal process
heating using fossil fuels. Electric processing can be
subdivided into 3 primary groups: (1) joule heating, (2)
plasma heating, and (3) microwave heating. Both joule
heating and plasma heating are based on well-developed
electric-furnace technologies for metal melting, metal
smelting, and glass melting. Joule heating includes ex
situ furnaces and ISV. In addition, several alternative
electric heating processes of varying applicability to vit-
rification are described under Section 3.1.4, "Miscella-
neous Electric Heating." AH of these categories are de-
scribed in greater detail in Maurice Orfeuil's Electric
Process Heating. Much of the information in this chapter
is summarized from Orfeuil's book (see also Pincus and
Diken, 1976; Trier, 1976). The discussion of vitrification
technologies in this chapter follows the outline presented
in Table 3-1. Identified studies of the applicability of
these processes to various wastes are presented in
Appendix A.
During research forthis document, several high-tempera-
ture, non-vitrification methods were discovered that im-
mobilized waste in a crystalline rather than glass form.
These included Ceramic Bonding, Inc. (Melzer, 1990)
Table 3 -1. Classification of Vitrification Processes
1. Electric Process Heating
A. Joule Heating
(1) ex situ
(2) in situ
B. Plasma Heating
C. Microwave Heating
D. Miscellaneous Electric
Heating
2. Thermal Process Heating
Examples
Ceramic Melter
In Situ Vitrication
Plasma Furnace
Microwave Melter
Resistance Heating,
Induction Heating,
Electric Arc Heating
Rotary Kiln Incinerator
(operated in slagging
mode)
and Molten Metals Technology, Inc. (Roy, 1991). While
these technologies may be applicable to many of the
same wastes as vitrification, it was felt that inclusion of
these types of processes was beyond the scope of this
document. Therefore, this document is limited to pro-
cesses that use high-temperatures to produce a waste
form that is all or largely glass.
3.1 Electric Process Heating
Many types of electric heating processes are potentially
applicable to vitrification. Joule, plasma, microwave,
induction, and electric arc heating are the electric pro-
cesses currently being applied to vitrification.
3.1.1 Joule Heating
In joule heating, an electric current flows through the
material. As the material internally resists the current, the
current loses power and transfers heat energy to the
material. The dissipated power is predicted by Joule's
Law:
P = I2R
where P = dissipated power (watts, W),
I = current through the material (amperes, A), and
R = resistance of the material (ohms, Q.).
Thus, with increased electrical resistance, if current can
be maintained, additional power is dissipated and the
material heats more rapidly. However, unless the voltage
is increased, an increase in resistance will also decrease
current. This is predicted by Ohm's Law:
R = V/lorV=IR
where V = voltage (volts, V),
I = current, and
R = resistance.
Ohm's Law explains why materials with low resistivity are
often heated at low voltages (5 to 48v) in non-vitrification
3-1
-------
processes. However, soils and other materials heated by
joule heating are frequently quite resistant and require
higher voltages.
I
Several properties of glass impact the joule heating pro-
cess. Among these properties is glass's poor electrical
conductivity (high resistivity) as a solid. Conversely, at
high temperatures, especially in the liquid state, glass is
a more efficient electrical conductor and can be heated
directly by electric current. Glass resistivity decreases by
a factor of 1013 to 1014 as temperatures increase from
ambient temperature to 1300-1400°C. This is explained
by the structure of glass: current flow takes place due to
the mobility of ions in the silica framework. As increased
heating input breaks apart the framework, ions are
increasingly able to carry the electric charge (qrfeuil,
1987). Rgure 3-1 illustrates the decreased resistiyity of
selected glasses as temperature increases.
i
i
The resistivity and melting pointtemperature of a particular
glass are also influenced by the chemical composition of
that glass. Alkali content is particularly important in
carrying charge. For equal alkali content, electrical con-
ductivity is inversely proportional to the size of the ions.
Therefore, sodium glasses have a higher electrical con-
ductivity than do potassium glasses. However, conduc-
tivity is not related by a simple equation to the concentration
or size of ions, and in general, only measurements can
provide electrical resistivity values (Orfeuil, 1987):
Melt viscosity is the most important processing property;
it controls processing rate, product homogeneity, and
heat transfer within the molten glass. It exerts this control
primarily by impacting convection currents (Orfeuil, 1987).
Viscosity is modified by changing feed composition or
process temperature.
Mechanical Constraints. Characteristics of the molten
glass place mechanical constraints on the design of a
joule heating system. Forexample, since the conductivity
of molten glass is ionic, an alternating current (AC) must
be used to avoid the risk of electrolysis, annodization of
electrodes, and the depletion of charge carriers (Qrfeuil,
1987).
Electrodes must withstand corrosion from the molten
glass bath, offer adequate mechanical strength at high
temperatures, and have low resistivity. The commercial
glass Industry generally uses graphite and molybdenum
for electrodes.
The position of the electrodes in the furnace controls the
buildup of convection currents in the melt and, subse-
quently, homogeneity in the melt. Theirarrangementwith
respect to each other, and with respect to the top melt
level, controls the energy given off and enables the best
possible glass melting conditions to be obtained. The
concentration of energy around the electrodes causes
local heating, resulting in an upward movement of the
glass and convection currents in the bath.
Joule heating vitrification can be carried out both ex situ
and in situ.
3.1.1.1 Ex Situ Joule Heating
Joule process heating furnaces for the treatment of
hazardous wastes evolved directly from glass melters in
the glass industry. The electric furnace/melter category
includes processes that use aceramic-lined, steel-shelled
melter to contain the molten glass and waste materials to
be melted.
Some melters are much like electric glass furnaces used
to manufacture glass products (e.g., bottles, plates).
Such melters receive waste materials and glass batch
chemicals directly on the surface of a molten glass bath.
Most melting occurs at the waste/molten glass interface
as heat is transferred from the molten glass. As waste is
heated, volatiles may be released and organics are either
pyrolyzed (in an oxygen-poor environment), or oxidized
(in an oxygen-rich environment). Off-gas treatment is
required to minimize air emissions. Figure 3-2 shows a
process flow-sheet for a typical joule-heated ceramic
melter (JHCM).
60
Legend:
1,2 Ordinary glass
3 Glass lamp
4 Neutral glass
5 Borosilicate
6,7 Pyrex
8 Lead crystal
1000°
1100° 1200
1 300° 1 400°
Temperature
Figure 3-1. Relationship Between Resistivity and
Temperature for Selected Glasses ( Orfeuil, 1987)
3-2
-------
Quench
Scrubber
Joule-Heated
Ceramic Melter
Glass
Quencher
Roughing
Filter Heat
Exchanger HEPA
Filter
Blower
*->,
o
Stack
Off-Gas Stream
Melter Feed
Scrub Solution
Glass Stream
Landfill Disposal
Figure 3-2. Typical JHCM Process Flowsheet (adapted from Koegler et al., 1989)
The molten glass melt has several distinctive character-
istics which influence processing conditions and, ulti-
mately, contaminant destruction and product formation.
The more important of these will be briefly described here.
The melt is initiated by some form of pre-heating. Once
the glass is fluid and conductive, heating continues by
joule heating, as described earlier. Melt temperatures
generally range from about 1000°-1600°C (Chapman,
1984). Maximum temperatures are limited to prevent
corrosion of electrodes or refractory material and volatil-
ization of constituents.
For many glass melters, an important part of the vitrifica-
tion process is the formation of a cold cap, or crust on the
top of the melt (Figure 3-3). The cold cap forms from the
feed as it is introduced from the top of the melter and
functions as the interface between the incoming material
and the molten glass. Water evaporates from the top of
the cap and enters the off-gas system. The cap's bottom
contactsthe glass and isthe interface where feed material
melts and forms the waste glass matrix. The cold cap
performs the important function of filtering and holding
volatilized wastes for possible re-incorporation into the
melt.
Incoming feed off-gases
1__L- .plenum
cold cap
refractory wall
molten glass
undlssolved
metal precipitate
molten metal tap
Figure 3-3. Generalized JHCM Showing
Components of Melter and Molten Material
3-3
-------
In addition to the cold cap, other zones of non-glassy
material may form in the melter. If the melt is strongly
reducing, metals in their elemental form may sink (and
form a layer on the bottom of the melt. This electrically
conductive layer may short the system and shut down the
melter. Solutions for this difficulty include melter design,
electrode placement, feed modification, and an additional
tap on the bottom to remove metal slag separately from
the glass. Metals which commonly form a slag include
palladium, ruthenium, rhodium, silver, iron, and other
heavy metals. This metal layermay potentially be recycled.
A salt layer may also form. Salt layers float on top oi the
melt and could cause shorting and corrosion (Eiseristatt
and Chapman, 1986).
Electric furnace melting may result in several types of
processing problems. Among these are:
1. Foaming (possibly caused by oxidizing condi-
tions) may lead to unstable operations and pres-
sure surges. Prolonged glass foaming may,also
lead to corrosion of refractory walls (Holton et al.,
1988). i
2. Cold-cap bridging (occurring when liquid flows
underthe cold cap) creates a high pressure zone
which may result in uncontrolled glass discharge
(Holton etal., 1988). |
3. High electrical conductivity in the melt may cause
the current required to heat the glass to exceed
the recommended maximum current density for
the melter electrodes. i
4. Low electrical conductivity in the melt may result
in a high voltage potential, causing conduction
within the refractory material. Low conductivity
also requires large electric power systems
(Koegler etal., 1989). j
5. High viscosity may slow the processing rate
because the interaction rate between feed( and
glass is slowed (Koegler et al. 1989).
6. Low viscosity (<100 poise) may result in in-
creased melter corrosion.
However, these problems are amenable to feed modifi-
cation and other types of processing adjustments.
Other melters involve feeding mechanisms that introduce
waste materials below the molten glass surface. Such
methods of introduction result in the pyrolysis of organic
contaminants within the molten glass, followed by eyolu-
tion of pyrolyzed off-gases to the plenum (the space
above the glass surface) where they may combust.
Undestroyed organics and organic by-products then move
to the off-gas treatment system for removal. Both types
of melters result in the incorporation of low-vapor-pres-
sure inorganics into the molten glass.
Electric melters must periodically be tapped (drained) to
remove the accumulated glass product. The molten
glass may be cast directly into containers or sand. An-
other alternative uses a water bath (quench bath) to
produce a granular residual product (aggregate).
Following is a description of several innovative melters
that fall into the category of joule heated melters.
Stir-melters. Stir-melters are joule heated melters in
which the molten material is agitated by a stirrer (Richards
and Lacksonen, 1991; Bickford etal.,1991). Because this
increases efficiency in heat distribution, stir-melters have
a high throughput rate for their size. Throughput rates
with the stirrer operating have been eight times greater
than those without the stirrer operating. The greater
efficiency in heat distribution also permits operation of the
stir-melter at lowertemperatures, thus allowing increased
flexibility in selection of materials for melter components
and increased contaminant incorporation into the waste
glass. The increased throughput rate means the stir-
melter can be constructed small enough to be used in
gloveboxes for the treatment of radioactive materials.
The smaller size and lower operating temperatures also
reduce costs by reducing heat losses.
Liquid-fed Ceramic Melters (LFCM). The LFCM is cur-
rently the state-of-the-art melterfor HLW. The advantage
of the LFCM is that it is capable of converting high-level
liquid wastes (HLLW) directly into glass without
pre-calcination. Because it avoids calcination, the entire
process is simplified and costs are substantially reduced.
Seven projects are formally committed to the LFCM:
Savannah River's Defense Waste Processing Facility
(DWPF), USA; West Valley Demonstration Project, USA;
Hanford Waste Vitrification Project, USA; Germany's
PAMELA plant at Mol, Belgium; Wackersdorf, Germany;
and Japan's Vitrification Facility (Chapman and McElroy,
1989).
3.1.1.2 In Situ Joule Heating
In situ joule heating is represented by ISV. ISV evolved
from joule-heated glass melters developed to immobilize
radioactive wastes. It was developed by Battelle at
Pacific Northwest Laboratory (PNL) for the U.S. Depart-
ment of Energy (DOE). The ISV process has been
developed and demonstrated through large-scale test-
ing. Wastes treated include a variety of hazardous
chemical, radioactive, and mixed (hazardous chemical
and radioactive) wastes.
3-4
-------
ISV converts contaminated soil and other substrates into
a stable glass and crystalline product. Figure 3-4 depicts
the process. The Electrode Feed System (EFS) inserts
a square array of four graphite electrodes into the con-
taminated site. This mechanism allows the electrodes to
sink to increasingly greater depths as the molten glass
increases in volume. Processing continues until the de-
sired treatment level is reached, or until a process-
limiting depth is reached. If processing difficulties are
encountered, then EFS can "grasp" the electrodes and
thus prevent their downward movement until the difficulty
is addressed. Previously, ISV required insertion of the
electrodes into boreholes prior to vitrification.
Because soil is not electrically conductive when moisture
has been driven off, a conductive mixture of flaked
graphite and glass frit is placed between the pairs of
electrodes as a starter path. An electrical potential is
applied to the electrodes to establish an electrical current
in the starter path. The resultant power heats the starter
path and surrounding soil to 2000°C, well above initial
soil-melting temperatures of 1100°C to 1400°C. The
graphite starter path is eve ntually consumed by oxidation
and the current is transferred to the molten soil, which is
electrically conductive when molten. As the molten or
vitrified zone grows, it incorporates radionuclides and
nonvolatile hazardous elements, such as heavy metals,
into the melt and pyrolyzes organic components. The
pyrolized by-products migrate to the surface of the vitri-
Electrodes
\
Floating Layer
(Rocks, Ceramics)
Surface
Combustion
fied zone where they combust in the presence of oxygen.
A hood placed over the vitrified area directs the gaseous
effluents to an off-gas treatment system (Buelt,
Timmerman, and Westsik, 1989).
Attempts to reduce costs by utilizing a fabric hood were
not successful. Fabric hoods have caughtfire twice in ISV
tests, once during a PNL test and once during a Geosafe
test. Both fires started when molten material splashed on
the hoods. The hoods used at the time of the fires were
fabric hoods coated with heat-resistant sealants. Since
these fires, both PNL and Geosafe have reverted to
previous steel hood designs. This change from fabric
hoods to steel hoods has delayed the application of ISV
to several sites.
As the melt grows downward and outward, power is
maintained at sufficient levels to overcome heat losses
from the surface and to the surrounding soil. Generally,
the melt grows outward beyond the electrodes to a
distance equal to about half of the spacing of the elec-
trodes. For example, if the electrode spacing is 5.5 m, a
melt width of about 8.5 m would normally be observed.
The molten zone is roughly circular and somewhat flat-
tened. The tendency to flatten increases as melt size
increases (Buelt, Timmerman, and Westsik, 1989).
In order to control the amperage during ISV processing,
operators use a power transformer with multiple voltage
. Off-Gases
To Treatment
(Some Cases) ^< A
--'T''-'-"^ •T"'-" ••• ^T-~
Volatiles
(Dissociation,
Destruction)
Nonvolatiles
(Distribution,
Incorporation)
Maximum Extent
of Melt
(Mixture of Soil and
Melt at Surface;
Size Depends on
Electrode Spacing,)
Denser Layer
(Ceramics, Pure Metals)
Figure 3-4. Schematic of ISV (adapted from USEPA, I989b)
3-5
-------
taps. At start-up, the ISV process requires high voltage
(up to 4,000 V) to overcome the resistance of the soil.
Current is relatively low (400 A) at this time. As the melt
progresses and resistance decreases, voltage is de-
creased (down to 400 V by the end of processing) to
compensate for the decreased resistance of the molten
glass and the resulting increase in current (up to 4,pOO A
by the end of processing). Processing continues until
heat loss from the melt approaches energy delivered to
the soil via the electrodes, oruntil powerto the electrodes
is shut off (Jacobs et al., 1988). |
Rve major subsystems comprise the process equipment
to perform ISV: (1) electrical power supply, (2) off-gas
hood, (3) off-gas treatment, (4) off-gas support, and (5)
process control (Buelt, Timmerman, and Westsik, 1989).
These five major subsystems and their set-up at a typical
site are depicted in Figure 3-5. Except for the off-gas
hood, all components are contained in three transport-
able trailers. The off-gas hood and off-gas line, which are
installed on the site for collecting gaseous effluentb, are
dismantled and placed on a flatbed trailer for transport
between the sites to be treated.
The normal processing rate for the large-scale system is
3 to 5 tons/hour (t/h). The maximum depth demonstrated
thus far has been 5 m (17 ft) by PNL and 5.8 m (19ft) by
Geosafe. The average processing operation lasts about
150 to 200 hrs, depending upon the depth and electrode
spacing (Buelt, Timmerman, and Westsik, 1989). !
ISV processing is termed 'In situ" when the soils are
processed where they presently exist. Placing soil in a
trench or container for treatment is termed "staged"
processing. For example, a staged application may in-
volve consolidating contaminated soil by removing the soil
and placing it in a trench. The filled trench could then be
vitrified. Typically, staged application would be most
effective where the contaminants are widely distributed in
the top few feet of the site. Because ISV is a batch
process, it may not be cost effective to move the hood
from setting to setting to vitrify the top few feet of the
contaminated material.
A predictive model of the ISV process has been developed
at PNLto assist engineers and researchers in the applica-
tion of ISV to different sites. The model, configured on a
Macintosh personal computer, predicts vitrification time,
melt depth and width, and electrical consumption. Pre-
dictions are based on datainputs of electrode configuration,
soil parameters, and molten-glass characteristics. The
model's predictions are useful for operations planning,
cost estimates, and melt locations. The depth and width
predictions, for example, can be used to locate the melts
to help ensure that the entire contaminated region is
treated and that adjacent structures are not damaged by
ISV treatment. Using the model to predict the shape of a
large-scale ISV melt indicated close agreement between
model prediction and actual monolith shape. Further
validationtesting is needed, however(Koeglerand Kindle,
1991).
3.1.2 Plasma Heating
Plasma heating is an electrical heating process which
relies on the conversion of a gas into a plasma through the
application of energy by an electric arc. Plasma heating
offers high operating temperatures and high powerdensi-
Emeraency
Diesel
Generator
Electrical
Power
Supply
Process
Control
Trailer —
Q
-D-
Off-Gas Hood
Off-Gas Line V ' V
Support t
r Outdoor
Lighting (lyp)
3S,
Off-Gas Support and Treatment
500 KVA [ 3 CT
Supply Cables
In PVC Conduit
600 Volt 1
600 AMP
Fused Safety
switch ;
~~600 KVA Substation
Figure 3-5. Pilot-Scale Process ISV (adapted from Callow, Weidner, and Thompson, 1991)
3-6
-------
ties. Unlike joule heating vitrification, which grew out of
the glass-making industry, plasma heating vitrification
has grown out of the specialty metals industry.
A plasma is an ionized gas. At high enough temperatures
(e.g., 20,000°K for argon), electrons are stripped from
their nuclei and the matter exists as a mixture of negative
electrons, positive nuclei, and atoms. The ionized par-
ticles make plasma an excellent electrical conductor
(Jacob 1991; Orfeuil, 1987).
There are two types of plasmas: plasmas in which the
degree of ionization is close to unity and plasmas which
are only partially ionized (Jacob, 1991). The first type of
plasma occurs in thermonuclear fusion, in which tem-
peratures reach several millions of degrees. This is the
type of plasma found in the sun and which has no
industrial application at present. In partially ionized plas-
mas, the degree of ionization varies from 2 to 50%. The
temperatures of partially ionized plasmas vary between
2,000°K and 5,000°K. It is these plasmas that have
industrial application (Jacob, 1991; Orfeuii, 1987).
Plasma heating equipment must perform two basic func-
tions: creating the plasma and effectively heating the
product.
Plasma is commonly created by passing a gas through an
electrical arc. The arc can be generated by direct current
(DC) or alternating current (AC). With a DC arc the
cathode generally consists of tungsten and the anode
generally consists of copper. The anode also typically
functions as a nozzle directing the plasma. In contrast, in
a single phase AC arc plasma generator, the electrodes
act as the cathode and anode alternately, and must
therefore be made from the same material.
Gases used in generating a plasma arc include nitrogen,
oxygen, noble gases, air, and mixtures of these gases.
Electrode life is a major concern and is influenced by
electrode material, the gas used, and electrical current
levels. Electrode structure, gas injection method, and
nozzle design help shape the plasma and determine
heating efficiencies.
The product is heated in one of two ways: by a
non-transferred arc or by a transferred arc (see Figure
3-6). A non-transferred arc uses two internal electrodes.
A small column of injected gas is heated by the electric
arc, creating a plasma flow that extends beyond the tip.
Non-transferred arcs heat only via conduction and pro-
duce a dispersed heat that is needed for tasks such as air
and gas heating and drying. Non-transferred arcs have
been applied to hospital wastes.
A transferred arc uses the working material as one of the
electrodes. Therefore, in a transferred arc application,
heating occurs via convection, radiation, and electrical
resistance. It is the transferred arc that is the heat source
in hazardous and radioactive vitrification applications. In
these applications, the plasma arc melts the material to
form a molten bath from which glass is periodically re-
moved to form the immobilized waste product.
The application of plasma heating to hazardous material
is international in scope. Kupp, a German firm that was
recently purchased by Mammesman Demag, has devel-
oped an AC transferred arc torch with a tungsten tip that
has application to hazardous materials. Aerospatiale, a
French company, has a non-transferred arc torch with
application to medical wastes. Tetronics Research and
Development Company in Faringdon, England, has re-
searched treatment of contaminated soil and incinerator
ash. Davy McKee's Research and Development Group in
Stockton-on-Tees, England, is working on a plasma fur-
nace for treating arc furnace dusts by recovering the
metals and leaving a material suitable for landfill (Jacob,
1991).
In the United States, Plasma Energy Corporation (PEC)
has a transferred arc plasma torch that has been used in
industrial applications in the past and is now being applied
to the vitrification of ash from the incineration of municipal
solid waste (MSW) in Japan. In one effort, Ebara and
Rear
Klcctrocle
Injectors
Front
Kkctroile
Transferred Arc
Non-Transferred Arc
Figure 3-6. Comparison of a Transferred Arc
and a Non-Transferred Arc (Source: Plasma
Energy Corporation)
3-7
-------
Infilco are applying the PEC torch at a pilot-level plant. In
the United States, PEAT, Inc. is researching the applica-
tion of the PEC torch to medical wastes and incinerator
ash. I
Retech, Inc. of Ukiah, California has developed a plasma
heating furnace called the plasma centrifugal reactor
(PCR). In the PCR, prepared waste materials are fee- into
a rotating reactor in which a transferred-arc plasma torch
is operating. The rotating reactor also serves as one
electrode forthe transferred arc. The plasma torch, which
is capable of temperatures exceeding 10,000°C, heats
the waste material beyond the point of melting to about
1,600°C. Centrifugal force created by the rotating reactor
prevents waste and molten material from flowing out of
the reactorthrough the bottom. The rotation of the reactor
also helps to transfer heat and electrical energy evenly
throughout the molten phase. Periodically, the melted
material is allowed to fall into a slag chamber where it is
collected in waste containers (Eschenbach, Hill, and
Sears, 1989). Rgure 3-7 is a schematic of a demonstra-
tion PCR; it shows the location of the electrodes ancl the
way in which the molten glass pools due to centrifugal
forces.
Organics and other volatiles emitted during the plasma
heating pass from the reactor chamber to a secondary
combustion chamber into which an oxidizing gas is added,
thus allowing for further destruction of any organics re-
maining in the gas phase. Resulting off-gases are then
transferred to an off-gas treatment system to ensure safe
air emissions.
Figure 3-8 illustrates the components of a full-scale PCR,
including the feed system, reactor, secondary combus-
tion chamber, slag chamber, and off-gas system.
3.1.3 Microwave Heating
In microwave heating, a form of dielectric heating, the
body to be heated absorbs electromagnetic radiation.
More specifically, a dielectric is a material which is an
electrical insulator. A dielectric becomes polarized when
it is placed in an electric field. If the electric field is
alternating, successive distortion of the molecules causes
heating (Orfeuil, 1987). Ceramic-like wastes such as
incinerator ash, thermal insulators, concrete, soil, and
sand are mostly composed of dielectric material and can
WATER-COOLED COPPER ELECTRODE
PLASMA GAS INJECTION
ARC
TERMINATION
SPINNING
REACTOR
WELL
EXIT GAS AND
SLAG REMOVAL
o
Figure 3-7. Schematic of the Demonstration PCR Showing the Bottom-Pour Configuration
for Exit Gas and Molten Glass (Eschenbach, Hill, and Sears, 1989)
3-8
-------
ROTATING REACTOR WELL
Figure 3-8. Schematic of a Full-Scale PCR (Eschenbach, Hill, and Sears, 1989)
be directly melted by microwave radiation (Komatsu et al.,
1990).
Dielectric heating is usually classified into two
sub-categories on the basis of frequency ranges used:
radio frequency heating using frequencies between 10
and 300 MHz, and microwave heating using frequencies
between 3,000 and 30,000 MHz (Orfeuil, 1987). Of these
two forms of dielectric heating, only microwave heating
has been used to vitrify hazardous wastes.
A microwave installation consists of a microwave genera-
tor, a waveguide, an applicator, and ancillary monitoring,
handling, and safety devices (Fig. 3-9). The microwave
generator produces the microwaves that dielectrically
heat the load material. The waveguide directs the micro-
waves from the generatortothe load material by reflecting
the microwaves from its metal walls; it also keeps
radiowavesfrom propagating in all directions. Applicators
define the way in which the microwaves are applied to the
load material. There are many types of microwave appli-
cators. These applicators vary depending on the type of
process, continuous or batch, and the nature and shape
of the load material (Orfeuil, 1987). Ancillary monitoring,
handling, and safety devices work much as those used in
other types of treatment processes.
The main advantage of microwave heating is that the heat
is produced directly and solely in the mass of the material
to be heated. Another advantage is high power density.
The main disadvantage is relatively high energy con-
sumption and corresponding costs (Orfeuil, 1987). Arcing
resulting from induced currents in metallic components of
waste may damage the microwave generator unless
special provisions are made.
Kobe Steel, Ltd. has developed an incinerator/microwave
melter treatment process for plutonium contaminated
solid waste at the Plutonium Waste Treatment Facility
(PWTF) in the Tokai Works of the Power Reactor and
Nuclear Fuel Development Corporation (PNC) (Miyata et
al., 1989, Ohuchi etal., 1989). Inthis process, plutonium
contaminated solid waste is incinerated and the ash is
passed to the microwave melter. The microwave system
consists of a melter, ash feeding system, microwave
Waveguide
Conveyor
oven
2,450 MHz
generator
Figure 3-9. Microwave Melter (Orfeuil, 1987)
3-9
-------
feeding system, and the waveguide are all contained
within a glove box for safety in handling the radioactive
material. Only the microwave generator is outside the
glove box. Generated microwaves are introduced into the
glove box via the waveguide. The microwave rrielter
operates in batch feed. In this process, the material to be
treated is placed in crucibles before vitrification. These
crucibles serve as melt containers and, ultimately, jstor-
age containers for the waste glass.
The microwave melter has also been demonstrated on
slurry produced from a nuclear reactor. Melting and
immobilization occurred in crucibles that were ;later
capped and then welded by a remote-controlled plasma
arc welder (Komatsu et al., 1990). '
Kobe Steel is now seeking to apply microwave heating to
a wide variety of non-combustible wastes, including liquid
and sludge wastes, inorganic insulators (such as asbes-
tos and rock wool), residues of acid digestion and direct
liquid wastes, concrete, contaminated soil and sand, and
radioactive contaminated wastes. !
In America, the DOE is researching the application of
microwave vitrification to radioactive wastes. Bench-- and
pilot-scale tests have been conducted using actual trans-
uranic (TRU) waste from Rocky Flats Plant. Results were
similar to those from cold bench-scale tests and encour-
aged further research (Petersen, 1990). Methodology
similar to the Japanese microwave methodology is also
being developed at Oak Ridge National Laboratory
(ORNL). I
3.1.4 Miscellaneous Electrical Processes
Several electrical heating methods have been used in
vitrification, but not extensively. These include resistance
heating, induction heating, and electric arc heating. Of
these, electric arc heating probably represents the greatest
potential for broad application to the treatment of
hazardous waste, but it is still in the early stages of such
development.
3.1.4.1 Resistance Heating
Initial large-scale testing of vitrification for HLW was clone
in crucibles heated by external resistance heaters. Jheir
design represented a direct increase in scale from glass
development crucible tests. Crucible heating was dis-
carded as a treatment optionfor HLW because of low melt
rates caused by slow heat transfer and lack of agitation
and because temperature non-uniformities made itjdiffi-
culty to homogenize the glass (Bickford, Hrma, and Bowan,
1990).
3.1.4.2 Induction Heating
Currently, induction heating application to hazardous and
radioactive wastes is represented by the French AVM
process (Atelier de Vitrification Marcoule) and its de-
scendents. However, because induction heating is also
used in commercial glass manufacturing, it is potentially
applicable to hazardous and radioactive wastes and will
be briefly described here.
Induction heating is accomplished by inducing currents in
the material to be heated. For example, a solenoid can be
used to create a variable magnetic field inside the coil and
around it. If an electrically conductive body is placed
inside the magnetic field, the variation in the magnetic
field causes a variation in the magnetic flux passing
through the material and induces an electromotive force
(EMF) current. The EMF current causes eddy currents,
and these are converted into heat due to the Joule effect.
Induction heating can also be created using highly varied
induction configurations (flat inductors, linear inductors,
tunnel inductors, etc.) and a wide range of relative part/
inductors (Orfeuil, 1987).
The French have developed an induction-heating vitrifi-
cation process preceded by calcination for their process-
ing of HLW (Jouan, Ladirat, and Moncouyoux, 1986;
Bonniaud et al., 1986; Baehr, 1989). This system, the
AVM, has been operating since 1978 and is located at
Marcoule, France. As of October, 1988 the AVM had
vitrified 1,225 m3 of concentrated fission product solu-
tions. These operations generated 540 tons of glass
packaged in 1,547 metallic canisters (Baehr, 1989).
The AVM facility treats HLW in two primary steps: calci-
nation; and glass formation. The calcination process
occurs first and drives off water, converts hydroxides to
oxides, and sinters the material, thereby reducing surface
area. The resulting calcine is mixed with appropriate
glass-forming materials and melted in the induction-heated
glass furnace.
Vitrification processes in several other locations are
modeled on the AVM facility. In France, two new, sister
vitrification plants are being built at La Hague. The
English are employing a similar system to vitrify English
HLW at Sellafield (Nuclear Engineering International,
1990).
3.1.4.3 Electric Arc Furnaces
Electric arc furnaces also are being applied to vitrification;
they heat by creating current flow between two electrodes
in an ionized gas environment. They differ from plasma
furnaces in that a plasma is not created and therefore not
3-10
-------
part of the heat transfer mechanism. The electric arc
furnace was first developed in the metal industry.
A group from Electro-Pyrolysis, Inc. is working with a
group from Massachusetts Institute of Technology to
develop an innovative vitrification process. In this pro-
cess, a DC electric arc is used in connection with a plasma
heating arc to pyrolyze solid hazardous materials. The
electric arc provides the primary energy for the heating
and melting of the target material. This occurs in a sealed
unit, thus reducing overall the amount of gases produced
during pyrolysis and allowing the gas to be removed from
the system in a non-oxidizing atmosphere. Furthermore,
because the chamber is sealed, generated gases are
forced to exit upward through the hollow arc-generating
electrode and must pass through the electric arc. In
addition, a plasma-heated zone created by electron-
beam ionization and microwave heating is located at the
tip of the electrode; gases must also pass through this.
Thus, the plasma functions as a scrubber for off-gases
generated by the electric arc. The electric arc provides
target material heating and also off-gas treatment
(Bromberg et al., 1991).
An electric arc is also being used in the vitrification tests
in Albany, Oregon of MSW bottom ash and fly ash and the
ash from sludge incineration. These tests are in the
shakedown stage in preparation for round-the-clock
testing. The Bureau of Mines and the American Society
of Mechanical Engineers are the primary sponsors of
these tests. The Japanese are also working on electric-
arc vitrification.
3.2 Thermal Process Heating
Thermal process heating differs from electric process
heating in that the heat for melting is produced by the
burning of the waste and/or fuel. The melting most com-
monly occurs in a rotary kiln operated in a slagging mode
to produce a glass product, but other incinerators are also
used to vitrify wastes. Fossil-fuel-fired glass furnaces
have been used in the glass industry and may also be
applicable to waste vitrification. This section describes
several rotary kiln processes and one other thermal
process used to vitrify wastes.
Rotary Kiln Incineration. A rotary kiln is a cylindrical,
refractory-lined shell mounted at an incline from a hori-
zontal plane. This cylinder is rotated to facilitatejjiixing of
wastes under incineration with combustion air, as well as
to promote transfer of wastes through the reactor. Con-
stant rotation of the kiln also provides continuous expo-
sure of fresh surfaces to oxidation to promote destruction.
A rotary kiln system includes the waste feed system,
rotary kiln incinerator, auxiliary fuel feed system, after-
burner, and air pollution control systems.
Wastes and auxiliary fuel are injected into the high end of
the kiln and pass through the combustion zone as the kiln
slowly rotates. Retention time can vary from several
minutes to an hour or more. Wastes are substantially
oxidized to gases and inert ash within this zone. Ash is
removed at the lower end of the kiln, while flue gases pass
through a secondary combustion chamber and then
through air pollution control units for paniculate and acid
gas removal. Residual streams generated during rotary
kiln incineration include bottom ash, fly ash, and scrubber
wastewater(Johnson and Cosmos, 1989; USEPA, 1988).
Rotary kiln incinerators operated in the slagging mode
may produce a vitrified product. At high enough tempera-
tures, the material in the kiln will deform, producing an
amorphous state in that material. This molten slag can
then be tapped and may harden into a glass or glass-like
product upon cooling, based on material composition
(Brunner, 1984). Leachability tests were conducted on
the hardened slag produced in a 50,000 metric tons/year
rotary kiln operating at Rijnmond, Holland. Results indi-
cated that the slag, as produced, would pass the EPA
Toxicity Characteristics Leaching Procedure (TCLP) tests
(Schlegel, 1989).
Kiln incineration may be used as a vitrification process by
itself or prior to a vitrification step in a treatment train.
Inorganic Recycling, Inc. (IRI) has developed a vitrifica-
tion process using only incineration, while Marine Shale
Processors (MSP) hasdeveloped a vitrification process in
which only a portion of the incineration products are
vitrified. These processes are described below.
IRI's kiln-driven process uses F006 waste (wastewater
treatment sludges from electroplating) as feedstock to
produce ceramic products. Metals in the waste feed
increase the hardness of the glass-like products and also
affect their color (The Hazardous Waste Consultant,
1990a). Figure 3-10 shows a flow diagram of IRI's
recycling process. The process involves two primary
operations: mixing and vitrification.
The mixing system operates in a batch mode. Before
being mixed, each batch of F006 feedstock is tested to
determine the amounts of other raw materials that must
be added to the batch. In the mixing vessel, water and
various chemicals are added to the waste and a series of
oxidation-reduction reactions take place. After the reac-
tions are complete, silicates, such as sand and clay, are
blended with the feed. The mixture is then pumped into
an agitated holding tank.
3-11
-------
FEED SAMPLING
AIR POLLUTION
CONTROL SYSTEM
WATER TANK
Figure 3-10. Flow Diagram of the IRI Process (adapted from The Hazardous Waste Consultant, 1990a)
The vitrification system operates continuously. Material
is pumped from the holding tank into the kiln at a controlled
rate. Kilntemperature is varied based on the composition
of the feed. A pool of molten material forms in the kiln and
rises to an overflow level. When it reaches the overflow
level, the molten material flows out of the kiln and injto the
exit system.
According to IRI, the material produced in the recycling
process has numerous potential uses. These include:
architectural products, such as wall and floor tiles,jpatio
stones, mosaics, sinks, tubs, and countertops; abrasive
products, such as sandpaper, shot blast, and grinding
media; and refractory products, such as high-tempera-
ture bricks and otherinsulating materials (The Hazardous
Waste Consultant, 1990a).
In the MSP incineration/vitrification process, the hazard-
ous materials may form the raw ingredients for an aggre-
gate material. The primary elements of the processing
system are a 275-foot, counter-current rotary kiln where
incineration occurs, a puddling furnace where vitrification
occurs, and an off-gas treatment system where off-gases
are treated (see Figure 3-11).
Sludges and solids are prepared for processing by blend-
ing. Included in this blend are the shredded containers in
which the waste was stored or transported. The produc-
tion of feed material by blending is controlled to produce
a feed with a heat content between 18,600 and 25,570
joules/gram (8,000-11,000 BTU/lb). Raw ingredients are
fed into the elevated end of the kiln and move toward the
lower end with a residence time of 120 to 150 minutes.
The lower end of the kiln isfired with natural gas and liquid
fuels. Oxygen and air are also introduced at the lower end
to support oxidation and maintain temperatures at ap-
proximately 1200°C. Solids exiting the lower end of the
kiln are separated by size. Fine materials are sent to the
puddling furnace to be vitrified, while large materials, such
as gravels and ferrous materials, are stored for testing.
3-12
-------
Gases travel up the kiln and enter the off-gas treatment
system where the remaining organic materials may be
destroyed thermally (temperatures range from 870 to
1260°C) in a series of oxidizers (Harlow et al., 1989).
The primary source of energy for melting process residue
in the puddling furnace is the gases from the incineration
process. This is augmented with a natural gas/oxygen
lance thatfiresuponthe molten surface. Puddling furnace
by-products are fed to the smelting section while the lava
migrates to the lower section of the smelter and enters a
pooling pot (Harlow et al., 1989).
Multi-fuel Glass Melter. Vortec has developed a multi-fuel
glass melter with application to hazardous wastes (Hnat
et al., 1990b). The Cyclone Melting System (CMS) is
composed of three primary components: a multi-fuel-
capable batch preheater, a cyclone melter, and a glass
melter reservoir. Preheated combustion air, pulverised
coal, and glass-forming ingredients enter the preheater
from the top. The batch rapidly preheats in suspension by
radiative and convective heat transfer. The preheater is
designed to burn pulverized coal or a variety of gaseous,
liquid, and coal-slurry fuels. The preheated batch ingre-
dients are separated against the walls of the cyclone
melter by centrifugal forces. The liquid phase reactions
occur along the walls, and the melted glass and combus-
tion gases exit the melter to the melt reservoir. The melt
reservoir gives material more time to form a glass, and is
designed to hold an adequate supply of glass for level
control or temperature conditioning. The melted glass
may then be delivered to a glass forming process, orother
glass conditioning device, for integration with a glass
manufacturing process. The combustion gases exit the
melt reservoir to a high-temperature recuperator where
waste heat is recovered and recycled to the preheater.
Off-gas contaminants may also be recycled to the preheater
to increase process destruction efficiencies (DE's).
TO WASTE FEED VIA CONVEYOR
VITRIFICATION
OF FINES
MIXING/BLENDING TANKS
Figure 3-11. Simplified System Schematic of MSP's Process (adapted from Harlow et al., 1989)
3-13
-------
-------
CHAPTER FOUR
APPLICABLE WASTE TYPES AND CONTAMINANTS
This chapter discusses waste and contaminant types to
which vitrification applies. The wastes and contaminants
identified here are not inclusive. Vitrification potentially
appliesto a very wide variety of wastes and contaminants.
As Chapter One made clear, inclusion of a waste and a
contaminant in this chapter does not mean that vitrifica-
tion is the preferred technology for this waste at all sites.
Furthermore, inclusion is not meant to suggest that all the
technical problems for application have been solved, or
even that they will be solved. The tests and studies
included in this chapter serve to indicate the potential of
vitrification for consideration in early screening studies.
4.1 Applicable Waste Types
Some vitrification processes may be able to handle a wide
variety of waste types, while others are applicable only to
a very specific waste stream. Wastes to which vitrification
potentially applies include:
Radioactive wastes and sludges
Contaminated soils
Contaminated sediments
Incinerator ashes
Industrial wastes and sludges
Medical wastes
Underground storage tanks (USTs)
Drummed wastes
Shipboard wastes
Asbestos wastes
Radioactive Wastes and Sludges. The global implemen-
tation of vitrification to treat nuclear wastes demonstrates
vitrification's potential for this waste type. Mixed wastes,
in which radioactive contaminants are combined with
Resource Conservation and Recovery Act (RCRA) haz-
ardous inorganic and/or organic contaminants, also pose
a challenge to remediation and/or disposal. Because
vitrification may destroy organics and incorporate
inorganics, it may be applicable at sites with difficult
wastes such as mixed wastes.
Radioactive inorganic contaminants are not destroyed
during vitrification, but, as most are metals, are generally
incorporated in the glass during vitrification. Thus, vitri-
fication puts them in a waste form which is more man-
ageable and decreases the probability of their escape into
the environment. The discussion of metal inorganics
(later in this chapter) will address the fate of radioactive
contaminants more fully.
Radioactive sludges (or slurries) will be the incoming feed
for the LFCM vitrification at West Valley, New York, and
the Savannah RiverSite (SRS), South Carolina (Bjorklund,
Mellinger, and Pope, 1984; Wicks and Bickford, 1989). At
the SRS, the sludge and the supernatant salt solution
from HLLW storage tanks will first be separated. Each of
these waste streams will then be treated to concentrate
the radioactive contaminants found in each. Prior to
melting, these streams will be mixed to form the slurry that
will feed the LFCM (Wicks and Bickford, 1989).
Contaminated Soils. Treatment of contaminated soils
has been proposed for all types of vitrification processes.
Generally, soils are amenable to vitrification since they
often contain high percentages of silica, alumina, and
other glass-forming raw materials. Soil composition will
impact product characteristics such as density and
chemical durability. Soil composition will also impact
processing parameters by helping to define thermal con-
ductivity, fusion temperature, specific heat, electrical
conductivity, and melt viscosity (Buelt et al., 1987).
PNL evaluated soils across the United States and felt that
most were amenable to ISV (Shelley, 1990). Theprimary
soil characteristics limiting ISV application are high
quartz content and low alkali flux content without flux
addition. Other vitrification processes should also be
applicable to a variety of soils. In fact, the various ex situ
processes may be better able to vitrify a variety of soils
because of the greater ease with which incoming soil can
be modified through feed additives.
Site characteristics and treatment objectives will play an
important role in determining which type of vitrification
4-1
-------
process, if any, is applicable. ChapterSeven discusses in
more detail the limitations placed on vitrification by site
characteristics. !
One important issue in the application of ISV to contami-
nated soils is whether volatile organic contaminants mi-
grate away from the melt, or are captured by the melt and
destroyed. This issue will be addressed later in this
chapter. i
i
Contaminated Sediments. As most sediments are Com-
posed of run-off soils and other components amenable to
vitrification, vitrification should be applicable at many
sites with contaminated sediments. However, sedirnents
generally have high moisture contents. Any material with
high watercontent increases processing time and energy
demands by first requiring that the water be driven off.
Thus, vitrification maybe limited economically in its ability
to treat sediments. If vitrification is to be used to treat
sediments, the demands on time and energy need^o be
addressed. This is generally accomplished by dewater-
ing or drying prior to vitrification. •
ISV engineering-scale tests have been performed on
polychlorinated biphenyl (PCB) contaminated sedirnents
from New Bedford Harbor. Results indicated destruction
and removal efficiencies (DRE's) of greater than
99.99999%. TCLP testing resulted in leach extract that
contained metal concentrations below the regulatory
limits (Reimus, 1988).
Incinerator Ashes. While incineration significantly re-
duces volume of waste materials, the resulting ash may
concentrate undesirable inorganics and heavy metals.
Vitrificationfurtherreduces ash volume, destroys residual
organics, and immobilizes heavy metals. In addition, the
vitrified ash may become a useful construction material
and thus avoid the need to landfill the ash (Chapman,
1991). Treatment of incinerator ash is one of the growing
areas of interest in vitrification. In Japan, the aggregate
produced in the vitrification of incinerator ash is used in
road construction (GRI, 1989). Vitrification is potentially
applicable to the ash from MSW incinerators, hazardous
waste incinerators, and other incinerators. Both bottom
ash and fly ash may be amenable to vitrification.
Firms in Europe are increasingly looking to vitrification to
deal with the ash from hazardous waste incinerators.
Tougher pollution control legislation, the expense and
regulations involved in landfill disposal, and the closing of
international borders to the importation of hazardous
wastes are all helping to make vitrification cost effective
as the tail end of atreatmenttrain focusing on incineration
as the main agent of toxin destruction (Gilges, 1991). In
the United States, Recomp of Washington (ROW) opened
a facility in 1991 to vitrify MSW incinerator ash. Feasibility
studies in preparation for this facility have indicated an
80% volume reduction of the ash (Chapman, 1991).
Used in the context described above, vitrification no
longer functions as a stand alone technology, but rather
as part of a treatment train. Darnell (1990) proposed that
vitrification be used in a treatment train preceded by
incineration and followed by solidification.
Industrial Wastes and Sludges. Because of vitrification's
ability to immobilize inorganics, it is considered for the
treatment of industrial waste streams containing con-
taminant metals. The United States Army Toxic and
Hazardous Materials Agency (USATHMA), for example,
has conducted bench-scale studies on the vitrification of
paint sludgewastesandfoundvitrificationto be applicable
to these (Balasco et al., 1987). The vitrification system of
IRI handles liquid, solid, orsludge typeinorganicfeedstock
input such as metal-bearing sulfates, metal-bearing car-
bonates, and metal-bearing phosphates. The IRI system
has been tested and has produced a potentially usable
product for the following EPA listed waste streams: F006
(electroplating wastes); K061 (electric arc furnace ducts);
and D004 through D006 (inorganic transition metals,
arsenic (As), barium (Ba), cadmium (Cd), chromium (Cr),
and lead(Pb)); (The Hazardous Waste Consultant, 1990a).
As with sediments, industrial sludges are vitrifiable, but
water content may increase processing costs. Tests of
ISV with zirconia-lime sludges showed that the material
was vitrifiable and that the level of radon emanation was
reduced by a factor of 104 to 10^ after processing. Mea-
sured radon emanation rates were in the femtocurie
range (Buelt, Timmerman, and Westsik, 1989).
Medical Wastes. The vitrification of medical wastes may
destroy potential pathogens, provide very large volume
reductions, and immobilize any metals. A plasma heating
system using a non-transferred arc has been tested on
medical wastes by Aerospatiale. Results indicated a very
low (0.25%) concentration of unburned waste in the solid
residue and no biological activity in the ashes.
Underground Storage Tanks. USTs containing sludges
and salt cakes of radioactive and/or hazardous chemicals
are present at many DOE sites. Recent enactment of
stringent environmental regulations requires timely
remediation of certain inactive tanks. Studies evaluating
tank remediation alternatives show that many of the tanks
may potentially be treated in place using ISV, although
significant technical issues need to be resolved. Tanks
containing material that cannot be economically removed
and tanks with outlying soil contamination are likely can-
didates for in-place treatment by ISV (Campbell,
Timmerman, and Bonner, 1990).
4-2
-------
The ISV of USTs proceeds by filling the tanks with clean
or contaminated soil and then melting the tank, the tank
contents, and the surrounding soil. Processing results in
the vitrification of tank wastes and the surrounding con-
taminated soil. The tanks themselves form a metal ingot
at the bottom of the melt upon cooling (Campbell,
Timmerman, and Bonner, 1990).
So far, researchers have tested the applicability of ISV to
USTs at three levels: engineering-scale, pilot-scale, and
large-scale.
The engineering-scale test was conducted on a 30 cm
steel tank coated with concrete. The tank contained a
sludge made of the contents fromORNL USTs. Contami-
nants included uranium (U), technetium (Tc), lead (Pb),
mercury (Hg), chromium (Cr), cesium (Cs), and strontium
(Sr). The tank, contents, and surrounding soil were all
vitrified.
The pilot-scale test was conducted on a 1 m steel tank
coated with concrete. The tank contained a 35-cm-deep
sludge layer containing hazardous wastes typical of a
"worst case" scenario for ORNL inactive USTs. Both the
glass and metal products produced in this test passed the
TCLP (Campbell, Timmerman, and Bonner, 1990).
PNL completed a large-scale test of a 6000-gallon, 10-
foot-diameter, steel and concrete UST in July of 1991.
The tank did not contain hazardous or regulated materials
but did contain a layer of water saturated soil, with the
balance of the tank filled with pumice. The test was
terminated earlier than planned when a sudden release of
water vapor caused the containment hood to pressurize.
Since gas and vapor generated below the melt inside the
confines of the tank can only escape by venting through
the melt, it was discovered that under these conditions,
unpredictable, sudden releases of steam or vapor can
cause the containment hood to pressurize. Consequently,
a key understanding of this behavior and identification of
potential methods to deal with it is necessary to mitigate
the consequences of gaseous releases under these
conditions.
Drummed Wastes. Vitrification is attractive for treating
drummed waste because the drums, as metals, may be
incorporated into the waste glass. At least three processes
have developed methodologies for treating drummed
wastes: MSP, Retech and ISV. In the MSP process,
liquids are removed from the drummed materials and
used as supplementary fuels. The sludges and solids in
the containers and the shredded containers themselves
are blended to create a waste feed to a rotary kiln (Harlow
et al., 1989). The Retech process is similar in that liquids
are first removed and fed separately to the furnace. The
partially emptied drums are shredded in the furnace
above the melt using a copper electrode that creates an
arc with the drum. This arc melts and cuts the drum at the
arc contact point. Contents remaining in the drum fall into
the melt as the drum is shredded. Eventually, all the
pieces of the drum itself fall into the melt chamber and are
incorporated into the melt (Schlienger and Eschenbach,
1991). While not ready for wholesale remediation of
drummedwaste, ISV has been used to process drummed
waste in tests. It may also be possible to add intact, filled
drums along with othertrash to a properly designed ex situ
vitrification system.
Shipboard Wastes. Concerns governing the disposal of
wastes at sea are driving the re-evaluation of waste
disposal options for shipboard wastes. Vitrification offers
a volume reduction of wastes and a chemically durable
product that may potentially be dumped overboard. Be-
cause of these attractive benefits, the U.S. Navy is ex-
amining this option. Furthermore, PenberthyElectromelt,
Inc. is marketing a version of their glass melterfor ship-
board vitrification. However, no vitrification units are
known to be presently operating on board a ship.
Asbestos Wastes. Asbestos frequently contaminates a
wide variety of materials. Asbestos-contaminated materi-
als are amenable to vitrification because it thermally
destroys asbestos. Vitrification of asbestos is described
in greater detail later in this chapter.
4.2 Applicable Contaminants
Vitrification is potentially applicable to a wide range of
organics and inorganics, including both radioactive con-
taminants and asbestos. Because vitrification may immo-
bilize inorganics and destroy organics, it is also applicable
to wastes with organic and inorganic compounds.
Vitrification has four possible effects on contaminants:
1. Destruction through pyrolysis or combustion
2. Removal in off-gas treatment
3. Chemical and/or physical immobilization in the
glass product or metal slag
4. Escape into the environment
Pyrolysis in the intense heat within the molten bath
generally destroys organic wastes. Organic wastes not
destroyed by pyrolysis are generally destroyed by com-
bustion in a region separate from the melt. Most often,
combustion occurs in the plenum, or the area above the
melt surface, but within the furnace. For ISV, the plenum
is defined as the space above ground level, but within the
hood. In other processes, combustion may occur in a
4-3
-------
secondary combustion chamber. For example, in Retech's
plasma heat process, combustion takes place primarily
in a secondary combustion chamber adjacent to jthe
centrifugal furnace (Eschenbach, Hill, and Sears, 1989).
Organic contaminants that fail to pyrolyze or combust
must be removed by the off-gas treatment system.
Pyrolysfs destroys not only organics, but asbestos as well.
In the melt, asbestos breaks down into its constituent
groups (atoms or molecules), and these constituents are
incorporated into the waste glass or removed by 'the
off-gas system.
Metals (including radioactive metals) are not destroyed
but are immobilized in the solidified glass or metal slag or
are vaporized. Immobilization may occur when the con-
taminant is incorporated into the glass network or encap-
sulated (or surrounded) by the glass. These two immobi-
lization processes also prevent some radioactive decay
products from escaping into the environment. The off-igas
treatment system should be designed to capture vapor-
ized inorganics.
During ISV treatment, contaminants may migrate aljsng
three different pathways. The first pathway occurs when
vitrification fails to either destroy or immobilize the c'on-
taminant and the contaminant subsequently pasjses
through the off-gas system without being removed. The
second pathway is the movement of contaminants jnto
uncontaminated, adjacent soil during ISV. The existence
and importance of this pathway are debated at present.
Finally, contaminants may also migrate during excava-
tion, transportation, pre-treatment, and other steps de-
manding handling of the contaminated material. Con-
taminant migration during material handling is a common
concern for all ex situ treatments, vitrification as well as
non-vitrification treatments, and so will not be addressed
in this document.
Following is a discussion of the applicability of vitrification
to metal and radioactive inorganics, non-metal inorganics,
and organics.
4.2.1 Metal and Radioactive Inorganic
Contaminants
Metals are not destroyed during vitrification; therefore,
there are only three possible pathways for metals during
treatment: (1) removal in the off-gas treatment, j(2)
chemical and/or physical immobilization in the glass prod-
uct or metal precipitate, (3) escape into the environment.
Depending on treatment goals, chemical and/or physical
immobilization is generally preferred to off-gas treatment.
But when vitrification fails to incorporate metals into !the
melt, they must be removed by the off-gas system and
receive additional treatment as secondary wastes. How-
ever, it is sometimes desirable not to chemically or
physically immobilize metals in the vitrification process.
For example, mercury is removed during pre-treatment'
prior to HLW vitrification at the SRS DWPF. Or, if
recovery of the metals is a concern, the metals may be
recovered from the off-gas system and thus reused. In
this scenario, non-incorporation in the melt and removal
by the off-gas system would be preferred. Recovery of
mercury in this way is being explored by the Department
of Defense (DOD) for remediation of the M-1 holding
ponds at the Rocky Mountain Arsenal, Colorado.
Retention efficiencies vary with type of metal; different
metal oxides will have different solubility limits in glass.
The solubility limits of most metal oxides and salts in glass
can be found in the Handbook of Glass Manufacture
(Tooley, 1984) and other documents on glass production.
Oxides for which extensive solubility information is
available are: alumina, antimony oxide, arsenic oxides,
barium oxide, cadmium oxide, chromium oxide, copper
oxides, cobalt oxides, iron oxides, lead oxides, manga-
nese oxides, nickel oxides, selenium oxides, tin oxides,
and zinc oxides (USEPA, 1990a). Waste glass will retain
metals with varying efficiency depending on the type of
vitrification process used and its operating parameters.
These limits will also be influenced by other metals in the
waste and the chemical composition of the glass. Table
4-1 presents measured solubilities of elements in silicate
waste glass. These values should be read very generally
due to the multitude of processing variations which can
affect element solubility.
Data for retention efficiencies of selected metals by ISV
is presented in Table 4-2. ISV is not as amenable as other
vitrification types to manipulation of operating param-
eters and so its retention factors give a rough estimate of
difficult metals. Forthat reason, the data presented in this
Table 4-1. Approximate Solubility of Elements
in Silicate Glasses (adapted from Volf, 1984)
less than 0.1 wt%: Ag, Ar, Au, Br, H, He, Hg, I,
Kr, N, Ne, Pd, Pt, Rh, Rn, Ru,
Xe
between 1 and 3 wt%: As, C, CI, Cr, S, Sb, Se, Sn,
Tc, Te
between 3 and 5 wt%: Bi, Co, Cu, Mn, Mo, Ni, Ti
between 5 and 15 wt%: Ce, F, Gd, La, Nd, Pr, Th, B,
Ge
between 15 and 25 wt%: Al, B, Ba, Ca, Cs, Fe, Fr, K, Li,
Mg, Na, Ra, Rb, Sr, U, Zn
greater than 25 wt%: P, Pb, Si
4-4
-------
Table 4-2. Metals Retention Efficiency Test Results for ISV (Hansen, 1991)
Class Metal
Volatile Mercury
Semi-Volatile Arsenic
Cadmium
Cesium
Lead
Ruthenium
Antimony
Tellerium
Non-Volatile Americium
Barium
Cerium
Cobalt
Copper
Chromium
Lanthanum
Molybdenum
Neodymium
Nickel
Plutonium
Radium
Strontium
Thorium
Uranium
Zinc
(Hg)
(As)
(Cd)
(Cs)
(Pb)
(Ru)
(Sb)
(Te)
(Am)
(Ba)
(Ce)
(Co)
(Cu)
(Cr)
(La)
(Mo)
(Nd)
(Ni)
(Pu)
(Ra)
(Sr)
(Th)
(Th)
(Zn)
(a) Percentage of original amount remaining in the
(b) Engineering-scale tests involve a melt depth of
Pilot-scale tests involve a melt depth of 3-7 ft.
Retention Efficiency, % (a)
0
70-85
67-75
99-99.9
90-99
99.8
96.7-99.9
50-99
99.99
99.9
98.9-99.9
98.7-99.8
90-99
99.9
98.9-99.98
99.9-99.999
99-99.98
99.9
99.99
99.9
99.9-99.998
99.99
99.99
90-99
melt.
1-2 ft.
Scale'b>
Engineering
Engineering
Pilot
Pilot
Pilot
Pilot
Pilot
Pilot
Pilot
Engineering
Pilot
Pilot
Engineering
Engineering
Pilot
Pilot
Pilot
Engineering
Pilot
Engineering
Pilot
Engineering
Engineering
Engineering
table should not be regarded as precise measurements of
expected retention efficiencies. Table 4-2 also shows how
metals can be divided based on tendency to volatilize.
4.2.1.1 Increasing the Retention of Metals
Retention of metals, if that is the treatment goal, may be
increased by a number of mechanisms. These include:
Reduction of generated gas
Presence of a cold cap
Recycling volatilized metals
Decreasing melt temperature
Modification of melt composition through
additives
Reduction of Generated Gases. Gases evolved during
vitrification can help carry metal particles and vapors to
the surface. Greater gas evolution results in a more rapid
movement to the surface, decreased exposure of the
metals to the melt, and thus, decreased probability of the
metals dissolving in the melt. Because the burning of
combustibles during vitrification produces increased
quantities of gas, gas-assisted movement of contami-
nants to the melt surface is one reason that combustibles
are of concern during vitrification (see Chapter Seven).
4-5
-------
Presence of a Cold Cap. Presence of a cold cap increases
the contact-time between metals and the melt and thus
increases the probability of metals dissolving in the rrjelt.
Incold-topglass melters, metal vaporization has tradition-
ally been solved by the creation of a "cold-top" layel- or
crust on the surface of the melt. This layer is formed by
the incoming feed as it floats on the liquid melt, warmis to
melt temperature, and eventually dissolves into the melt.
Because it is cooler than the melt, metals migrating to the
melt surface may be trapped in the cold-top and sink bjack
into the melt to be possibly incorporated into the glajss.
ISV can be modified to increase metal retentions! by
implementing a similar mechanism. For example! in
certain ISV applications, soil may be added above [the
melt to increase the amount of glass that has formed
before contacting the contaminated layer. This increases
the length of contact between the metals and the glass
and increases the probability that the metals will be
incorporated into the glass. Figure 4-1 illustrates the
effect of melt depth on retention efficiencies for several
metals. !
Recycling Volatilized Metals. Metals that escape
treatment zone and enter the off-gas may be removed
from the gas stream by the off-gas treatment system
(typically removed by the scrubbersolution). At this point,
the contaminants may be considered for recovery,! for
recycling back to the molten glass to attain increased
retention in the waste glass, or for separate disposal.
Contaminants to be recycled may be removed from [the
off-gas system component, for example, by passing the
scrubber solution through a filter aid and activated car-
99,9
3
f
90
La/Nd i
0
l_
0
0,2
0,4
0,6
0.8
1.0
1.2
1.4 en
~ ft
Burial Depth
Figure 4-1. Element Retention Versus Burial Depth
During Pilot-Scale ISV Tests (Buelt et al., 1987)
bon. Contaminants can be returned to the melt to in-
crease the overall immobilization efficiency of the vitrifi-
cation treatment (Hansen, 1991). Contaminants to be
recycled may also be recycled with the off-gas system
component in which they were collected (e.g., by placing
a contaminated high-efficiency paniculate air (HEPA)
filter directly into the melt) (USEPA, 1987).
Decreasing Melt Temperature. An important processing
variable which affects metal incorporation is melt tem-
perature. Generally, the lower the temperature at which
the melt proceeds, the lower the quantity of volatilized
metals. This is apparently because the metals are
incorporated into the molten glass before they volatilize
(Hansen, 1991).
A primary factor controlling melt temperature is composi-
tion of the incoming feed. Feeds with high quantities of
fluxing agents will melt at lower temperatures. This must
be balanced against possible loss of product durability
and the potential impact on other processing variables
(Hansen, 1991; USEPA, 1989c).
For ISV, the development of the EFS may permit greater
control over the rate of melt advance than previously
indicated.
Modification of Melt Composition through Additives. The
solubility of metals may be affected by changing the
chemical composition of the melt. For example, reducing
agents such as carbon and ferrous salts may reduce
arsenates and selenates to lower valence compounds
that are more volatile, thus reducing incorporation effi-
ciencies of these metals (Schreiber et al., 1988).
4.2.1.2 Experience with Selected Metals
Following are selected treatment data for several metals.
Arsenic. Arsenic (As) is a semi-volatile metal which can
be difficult to incorporate into waste glass. Vitrification
has been evaluated as a BOAT by EPA for the following
arsenic and selenium (chemically similarto arsenic) waste
streams: D-004, D-010, K-031, K-084, K-101, K-102, P-
010, P-011, P-012, P-036, P-038, P-103, P-114, P-204,
P-205, and P-336 (Federal Register, 1991).
In general, waste glass containing arsenic exhibits re-
duced concentrations in the leachate for both Extraction
Procedure Toxicity (EP Tox) and TCLP tests. Twidwell
and Mehta (1985) found that glass made from slag
containing arsenic in concentrations of 0.3 - 23.5 wt%
showed 0.007-1.791 mg/L of Asinthe leachate (EPTox).
Chapman (USEPA 1990a) reported that waste glass
composed of 17 mg/kg As showed <0.005 mg/L As in the
4-6
-------
leachate (EP Tox). Rhone-Poulenc (USEPA, 1990a)
found that glass made from sludge containing arsenic
sulfide in concentrations of 2.0 - 2.5 wt% showed 0.5 ppm
(EP Tox) and <0.5 - 2.5 ppm (TCLP).
Arsenic is more volatile in some forms than in others. For
example, arsenicoxide may be more volatile than calcium
or iron arsenates. Certain waste feeds may require
chemical or thermal pre-treatment to convert arsenic
oxide to less volatile forms before vitrification (USEPA,
1990a). Twidwell and Mehta (1985) converted arsenic
oxide (AsaOa) in flue dust to a mixed calcium oxide
(Cas(AsO4)2)via slow roasting. The calcined mixture
was dissolved in a molten iron silicate slag at tempera-
tures up to 1290°C. These results indicated that arsenic
oxide, although volatile, may be successfully vitrified.
Cesium. Cesium (Cs) is also a semi-volatile metal. It is
common at DOE waste sites in its radioactive isotope,
137Cs. Due to its volatility and radioactivity, 137Cs pre-
sents a difficult remediation challenge. Researchers have
studied its behavior during vitrification in tests for a
number of vitrification processes. Several of these tests
are summarized below.
The volatilization of cesium and other semi-volatile radio-
active metals has been a concern at several potential ISV
sites, including ORNL(Spalding and Jacobs, 1989). This
concern arises because volatilized cesium must be re-
moved by the off-gas system, which increases the quan-
tity of secondary contamination that must be handled. At
ORNL, the treatment of a typical trench would require ten
ISV settings, each of which would produce a quantity of
waste. Therefore, despite retention efficiencies of 99.88%
in pilot-scale tests, the total waste generated over the
course of a trench clean-up was considered too high for
remediation goals.
The initial method to minimize 137Cs volatilization fo-
cused on adding sodium oxide (Na2O) or sodium car-
bonate (NaaCOs) to the soil priorto melting (Spalding and
Jacobs, 1989). This reduces the soil-melting temperature
and Cs is captured in the melt before it can volatilize.
However, bench-scale testing indicated that use of so-
dium as a flux increased gas evolution and caused
additional amounts of 137Cs to be carried to the surface.
These two mechanisms balanced and the net result was
no difference of 137Cs incorporation in the melt, whether
or not a flux was added. As a result, recent attempts to
control cesium volatilization focus on use of the EPS to
control processing characteristics and on recycling off-
gases by placing a HEPA filter prior to off-gas entry into
the off-gas system. The results of a second pilot-scale
test indicated that these methods successfully curtailed
generation of secondary off-gas system wastes.
Composition of the feed may also impact retention effi-
ciencies for cesium. Early tests that prepared for use of
vitrification at West Valley studied the ability of a LFCM to
incorporate cesium. Those tests found that the predomi-
nate variable controlling cesium incorporation was the
halogen concentration in the feed. Increased chlorine
content in the feed decreased the incorporation of cesium
in the glass. Air inleakage into the melter, plenum
temperature, feed rate, and waste loading were the pro-
cessing variables examined and found not to be important
in cesium retention (Goles and Anderson, 1986).
Tests in Japan using a microwave vitrification process
also found that feed composition influenced cesium re-
tention. This research reported that the amount of vola-
tilized cesium could be reduced by half by adding a flux of
20 wt% 8263 to the feed. The 6203 also reduced the
teachability of the waste glass (Komatsu et al., 1990).
At the SRS, cesium volatilized during HLW vitrification is
scrubbed from the off-gas with a 99.99999+% efficiency
and recycled to the melter feed (Wicks and Bickford,
1989).
4.2.2 Non-Metallic Inorganic Contaminants
Non-metallic inorganics found in waste include, but are
not limitedto, cyanides, ammonia, various acids, asbestos,
radon (a radioactive gas), halogens, and oxides of nitro-
gen, sulfur, and phosphorous. These inorganics are a
concern because they may adversely impact processing.
These non-metallic inorganics react to vitrification in a
variety of ways. This is because some are compounds,
such as asbestos and cyanides, and some are elements,
such as the halogens. Compounds such as cyanide and
asbestos decompose to their constituent molecules and
atoms and then follow the path typical of inorganics or
organics, as identified in this or other sections of this
chapter. Elements, such as halogens, form compounds
of different types depending on melt conditions, feed
composition, and other factors. In these various forms
they may either incorporate into the glass, evolve as off-
gases, or follow other applicable pathways.
The following sections address asbestos, radon, halo-
gens, and important inorganic oxides. Other inorganics
(such as cyanides and acids) will not be addressed since
they are compounds that primarily decompose into con-
stituents and description of their behaviorwill therefore be
redundant. Asbestos will be described in greater detail
than the other substances because it is a contaminant
widely targeted for treatment by vitrification. Halogens
and the addressed inorganic oxides are generally not the
contaminants targeted for treatment by vitrification, but
4-7
-------
evolve as part of the treatment process and thus require
attention. ;
4.2.2.1 Asbestos
Asbestos is a fibrous material composed of silicates,
metals, and either water of hydration or hydroxides. For
example, thechemicalformulaforcrocidolite (blue asbes-
tos) is NaFe(SiO3)-FeSiO3-H2O. When subjected to
temperatures of 400-900°C, water is driven off and the
rest of the asbestos fibers are broken down to their
constituent molecules or atoms. The constituents dis-
solve in the melt and are ultimately incorporated into the
glass framework as the melt cools. Asbestos wastes are
particularly amenable to treatment through vitrification
because their high silicate content helps form part of the
glass network (Roberts, 1989).
In order to successfully vitrify asbestos, the vitrification
process must be capable of handling all the materials
likely to have asbestos in them or to be mixed in with
asbestos waste. These materials include paper, plastic,
wood, concrete, brick, steel lath, copper, aluminum, rock
wool, glassfiber, gypsum, plaster, clay, quartz, refractory
material, and other materials (Roberts, 1989). j
Destruction efficiencies of asbestos are primarily con-
trolled by temperature and residence time in the furnace.
In the Vitrifix process the furnace operates at 1300°C.| At
these temperatures destruction of asbestos takes 2-3
minutes. However, because asbestos is an insulator and
a poor conductor of heat, residence time in the melte^ is
12 hours. Therefore, the probability of asbestos contact
with the molten glass and its resulting destruction is
greatly increased. Furthermore, the melter has a sub-
merged throat which is not heated in any way. JAt
temperatures less than 1100°C glass flow through the
throat stops due to the increased viscosity of the glass.
Glass exiting the furnace is thus ensured of exposure to
temperatures in excess of those required for asbestos
destruction (900°C or less).
The product of asbestos vitrification is a dark green1 to
black silicate glass. Its physical properties are similar to
those of container glass, but asbestos glass is more
chemically resistant. Because this glass is produced from
a waste stream with highly variable characteristics', it
does not have the precisely controlled properties found in
industrial glasses. However, it may possibly be used as
a hard core, or in place of the ordinary glass found; in
sandpaper (Roberts, 1989).
Asbestos vitrification has been tested in both England
and in the U.S. and the product reported free of asbestos
fibers. In the Faslane site clean-up, a 5 ton/day furnace
was used to remediate "hot-spot" soils contaminated with
asbestos (Denner, Langridge, and Affleck, 1988). In 1987
at the Dalzeil Glassworks in New Martinsville, West Vir-
ginia, a proprietary asbestos process was demonstrated
for the EPA. Results of this test showed safe handling of
materials as well as an absence of asbestos in the waste
glass. Air monitoring indicated that the process met
Occupational Safety and Health Administration (OSHA)
air quality standards for asbestos fibers, both inside and
outside the materials handling area (Roberts, 1989).
Asbestos has also been incorporated in waste glass in
bench-scale ISV melts (Farnsworth, Oma, and Bigelow,
1990).
4.2.2.2 Radon
Radon exists in rocks and soils and is spontaneously
produced by radioactive decay. Because radon is a gas,
it represents a difficult challenge to waste treatment pro-
cesses. Vitrification is a potential option because the
glass matrix severely limits the diffusion of gases (ap-
proaching no diffusion) with atomic radii greater then
krypton (1.03 A) or xenon (1.24 A). Radon's radius is
1.343 A. Thus, the release of radon from the vitrified
residues should be limited to that from externally exposed
surfaces (Sing and Swallow, 1960). Pre-existing radon will
be released during vitrification.
Results of bench-scale tests conducted on the Fernald K-
65 residue indicated the potential of vitrification to effec-
tively immobilize fission-product radon produced after
vitrification. The non-vitrified K-65 residue tested "hazard-
ous" by the EP Tox, and the radon emanation rate of
52,400 pCi/m2/s was over 2500 times the EPA limit of 20
pCi/m2/s. After vitrification, the K-65 residue tested
"nonhazardous" by the TCLP, and the radon emanation
ratewasl .56pCi/m2/s (Janke, Chapman, and Vogel, 1991).
4.2.2.3 Halogens
Halogens of primary concern are chlorine and fluorine.
They are a concern because of their tendency to form
compounds, such as dioxins, and because of the impor-
tant ways in which they may affect vitrification processing.
Halogens exhibit low solubility in silicate glass and may be
difficult to incorporate in the glass. However, this will vary
with halogen and glass composition. Chloride exhibits a
solubility of less than one percent, while fluoride has been
incorporated into glass up to 9 wt%. Tests on wastes from
the Weldon Spring site, Missouri, indicated that the glass
produced from soils at that site would hold no more than
5 wt% fluoride. As the waste feed held more than 5 wt%
fluoride (about 10 wt% fluoride), additives were requiredto
4-8
-------
dilute the fluoride and bring it within the solubility of the
glass (Koegler, Oma, and Perez, 1988). Other reports
have indicated volatilization of about 50% of the fluorides
in the feed (Loewenstein, 1983).
Halogens may also enter the off-gas system. If chlorine
enters the off-gas system as hydrochloric acid (HCI) gas,
it can be removed by the spray chambers and transformed
into salts such as NaCI, CaCI2, or some other innocuous
chemical. Fluorine in the off-gases may corrode the
melter (Bonniaud et al., 1986).
4.2.2.4 Inorganic Oxides of Concern
Primary inorganic oxides of concern include nitrogen
oxides (NOX), sulfur oxides (SOx^ and phosphorous
compounds (such as P2®5 °r PO42').
NOX and SOX will exit the melt to the off-gas system and
are regulated compounds. In general, nitrogen and sulfur
do not exhibit high solubility in silicate glass (< 2%) and
thus may necessitate treatment by the off-gas system. As
NOx and SOX are common process emissions, off-gas
systems designedtotreatthem are available. ChapterSix
discusses off-gas treatment in greater depth.
Nitrogen, sulfur, and phosphorous oxides are also a
concern because of the way in which they may influence
processing. For example, PaOs increases glass corro-
siveness and glass viscosity. Both tendencies may in-
crease the cost of vitrification, although by different
mechanisms. Sulfur, on the other hand, may increase the
tendency for the glass to foam, increase metal corrosive-
ness in the off-gas system, and form molten salts in the
melt (Chapman, 1984). The decomposition of NOa may
cause oxidizing conditions in the melter plenum and
thereby may contribute to the volatilization of iodine and
ruthenium (Smith, Nyman, and Anderson, 1990).
4.2.3 Organic Contaminants
The treatment of organics is, in one sense, a by-product
of vitrification. This does not mean that vitrification is
ineffective in treating organics. Organics, both contami-
nants and non-contaminants, are primarily destroyed
thermally during vitrification. Tables 4-3 and 4-4 present
organic DE's and DRE's for ISV and glass melters.
Potential fates of organics include the following: (1) de-
struction via pyrolysis or combustion, (2) removal in the
off-gas system, (3) migration to adjacent soil during the
vitrification process.
Destruction of organics occurs primarily via pyrolysis in
the melt and combustion in the plenum orin the secondary
combustion chamber. Ideally, organics will degrade com-
pletely to form products such as carbon dioxide
water, and HCI. Conditions in the plenum or combustion
chamber may often be controlled to maximize the com-
bustion of escaping organic products and the production
of the desired off-gases. Those organics and organic
by-products that still exist as contaminants are removed
in the off-gas system.
As described previously, organics may be released dur-
ing vitrification to the off-gas system or by migrating into
the surrounding soil. One criticism of ISV is the possibility
that organics may migrate into the surrounding soil. If this
criticism is correct, ISV may potentially transform a small
area of manageable contamination into a much larger
area of contamination, even if a high percentage of
organics are destroyed.
4.2.3.7 IncreasingDestruction Efficiencies ofOrganics
Methods to increase DE's of organics include:
Cold cap
Secondary combustion chambers
Recycling to melt
Cold Gap. The creation of a cold cap increases the length
of time organics are exposed to the melt and thus also
increases DE's via pyrolysis. As with metals, one poten-
tial method to increase organic DE's of ISV is to cover the
site with a layer of clean soil. Engineering-scale ISV tests
have shown that DE's for organics may increase from
97% (when the contaminated soil is not covered with a
layer of clean soil) to greater than 99.99% when an
uncontaminated layer of soil is placed over the contami-
nated site (Buelt, Timmerman, and Westsik, 1989).
Secondary Combustion Chambers. Control of conditions
in a secondary combustion chamber or in the plenum may
increase organic DE's. Forexample, in early tests, Retech's
PCR produced unacceptable levels of carbon monoxide
(CO). By increasing the quantity of oxygen (02) in the
secondary combustion chamber, organics in the off-
gases burned more completely and COa was produced
instead of CO (Eschenbach, Hill, and Sears, 1989).
Recycling to Melt. Finally, as with metals, organics that
are not destroyed but that are captured in the off-gas
system may be recycled to the melt to increase DE's.
They may first be removed from the off-gas component
which captured them orthe organic contaminants may be
recycled intact with the component.
4-9
-------
I
Table 4-3. ISV Organic Destruction and Removal Efficiencies(77;e Hazardous Waste Consultant, 199Gb)
Contaminant
Aldrin
Chlordane
DDD.DDE.DDT
Dieldrin
Dioxins
Fuel Oils
Furans
Glycol
Heptachlor
MEK
PCBs
Pentachlorophenol
Toluene
Trichloroethane
Xylenes
Initial
Concentration
(ppb) :
113
535,000
21-240,000
24,000
>47,000
230-11,000
>9,400
NA
61
NA
19,400,000
>4,000,000
203,000
106,000
3,533,000
Percent
Destruction
>97
99.95
99.9-99.99
98-99.9
99.9-99.99
>99
99.9-99.99
>90
98.7
>99
99.9-99.99
99.995
99.996
99.995
99.998
Total DRE
(including off-gas
removal)
>99.99
>99.999
>99.999
>99.99
>99.9999
>99.999
>99.9999
>99.99
>99.99
>99.999
>99.9999
>99.99999
>99.99999
>99.99999
>99.99999
Table 4-4. Demonstrated Organic Destruction Efficiencies for Vitrification Systems1
Compound
Hydrocyanic Acid
Chlorobenzene
Formic Acid
Phosgene
Methylene Chloride
Phenol
Acetone
Isodrin
Ethanol
Mustard Gas
Nitrogen Mustard
Carbon Tetrachoride
Aldrin
Dieldrin
Sulfoxide
Endrin
Dithlane
Sulfone
Xylenes
DIMP
DMMP
ACN
AN
'Data collected from Armstrong and Klingler,
°C for 99%
Destruction in
2 Seconds
482-866
482-866
31 8-368
^27-479
427-479
374-421
374-421
374-421
374-421
318-368
318-368
318-368
318-368
318-368
218-316
!38-160
182-213
> NA
i NA
NA
NA
NA
NA
1985; USATHMA, 1,988; Klingler and Abellera,
Measured DE (%)
NA
99.99986
NA
NA
>99.9995
99.99992
>99.9995
>99.9998
>99.9995
NA
NA
99.99988
99.99994
>99.9995
>99.99
>99.998
>99.96
>99.995
99.99817
>99.8
>99.8
99.99996
99.9994
1989.
4-10
-------
4.2.3.2 Contaminant Migration During ISV Treatment
As mentioned, one important question concerning ISV is
whether contaminants migrate into the adjacent soil dur-
ing treatment. This question is examined in general
fashion below viathe presentation of key empirical studies
and a theoretical model. This question may also need to
be addressed for a specific site via bench- and pilot-scale
tests priorto implementation of ISV at that site. While this
question applies primarily to organics, some volatile and
semi-volatile inorganics, such as mercury, may require
consideration.
Empirical Data. Numerous tests have failed to demon-
strate significant contaminant migration during ISV
(Campbell and Buelt, 1990; Campbell, Timmerman, and
Buelt, 1990; Farnsworth, Oma, and Bigelow, 1990; Landau
Associates, 1991; Timmerman and Peterson, 1990). For
example, an initial engineering-scale test was conducted
to address the question of organic migration. The soils in
this test were contaminated with 500 ppm of PCBs.
Results indicated that process destruction of PCBs was
slightly greater than 99.9%. After off-gas treatment, the
system ORE was >99.9999%. Analysis of the vitrified
block showed no residual PCBs; considering the pro-
cessing temperature, the data are reasonable. The soil
adjacent to the vitrified area was also examined for PCBs
but only limited quantities were detected (a maximum of
0.7 ppm of PCBs). These data were interpreted to
indicate that the soil vitrifies faster than the PCBs diffuse
and that for this reason they are unlikely to migrate from
thevitrificationzone during processing (Buelt, Timmerman,
andWestsik, 1989).
The movement of organic chemicals has also been ex-
amined through mass balance studies. A water mass-
balance study indicated a net migration of water into the
vitrification and off-gastreatmentzone (Buelt, Timmerman,
and Westsik, 1989). In this study, researchers estimated
that 33,800 liters of water were in the soil. Negative
pressure under the hood withdrew another 9,000 liters of
water from outside air. The amount of water leaving
through the stack was estimated to be 47,500 liters. From
this latter figure, 5,100 liters was subtracted to account for
the water lost from the scrub tank in the off-gas system.
The final balance was equivalent to 42,800 liters of water
prior to vitrification and 42,400 liters of water exiting the
off-gas system. The indicated net migration of water into
the vitrification and off-gas treatment zone suggested at
least partial movement of organics toward the melt, rather
than migration from the melt. Another ISV test was
performed on arsenic and mercury contaminated sludges
from the M-1 holding ponds at Rocky Flats, Colorado
(JacobsonandMears, 1991). A mass balance conducted
for this test failed to account for 37.5% of the arsenic and
53% of the mercury originally in the test chamber.
Theoretical Model. In addition to the empirical data
presented above, geochemical and soil chemistry
mechanisms have been presented to identify possible
contaminant behavior in soil adjacent to the melt. Pre-
sented below is a brief summary of a theoretical model of
contaminant migration (Dragun, 1991).
Seven soil mechanisms impact contaminant migration
during ISV:
• Concentration diffusion
Gaseous convection
Thermal diffusion
Chemical reaction acceleration
Pressure diffusion
Capillary water migration
Adsorption of water and chemicals onto soil
particles surfaces.
Dragun argued that these mechanisms function differently
in the five soil zones surrounding an ISV melt. Dragun
hypothesized that, during ISV, the following soil zones
form and remain in quasi-equilibrium: a melt zone (of
molten glass), a pyrolysis zone (where organic destruc-
tion is initiated), a heat affected zone (where soil moisture
is vaporized), a transition zone (where soil is heated from
ambient temperatures to 100°C), and an ambient soil
zone (of normal soil conditions).
After examining the mechanisms and the way in which
they functioned in the five soil zones, Dragun argued that
the net direction of organic contaminants would be toward
the melt and ultimate destruction and not away from the
melt to spread contamination.
4-11
-------
-------
CHAPTER FIVE
PRODUCT CHARACTERISTICS
The characteristics of the waste-glass that vitrification
produces are among the primary reasonsthatthis process
is considered as away of treating hazardous waste. This
product is typically a dark-colored, glassy substance, but
crystalline or non-vitrified material may also be present in
the product.
This chapter describes the various components of waste
glass durability, the volume reductions accompanying
vitrification, and potential uses of the product glass.
5.1 Product Durability
Product durability is associated with important chemical
and physical properties of waste glasses. These proper-
ties are closely interrelated and result from the structural
characteristics of glass as described in ChapterTwo. This
section discusses these properties under three sub-sec-
tions: chemical immobilization, physical durability, and
devitrification (i.e., the formation of a crystalline phase in
a glass). In a fourth sub-section, estimations of waste
glass's long-term stability are presented. Such estima-
tions are important because the long-term durability of
waste glass over geological time spans has not been
directly measured.
5.1.1 Chemical Immobilization
The single most important characteristic of waste glass is
chemical immobilization, orthe ability of the waste glass
to resist leaching of the immobilized contaminants when
contacted by water or other liquids. Without this charac-
teristic, the cost of vitrification is most likely not justified;
similar destruction efficiencies may be achieved using
other processes. Furthermore, it is because of the chemi-
cal durability of waste glass that it may potentially be used
and re-used in a variety of applications.
Although everyday experience suggests that glasses are
durable, all glasses do leach to some degree. For
example, recent experiments with lead crystal have shown
elevated lead concentrations of up to 21,530 u,g/l in wine
and brandy that was stored in lead crystal decanters from
six months to five years (Graziano and Blum, 1991). This
value is well above the maximum value of 50 u,g/l allowed
by EPA in drinking water. Additionally, elevated lead
levels (mean of 68 u.g/1) were identified within one hour in
wine poured into lead crystal glasses. However, this does
not address directly the leachability of waste glass: lead
crystal contains 24-32 wt% lead oxide (PbO), significantly
higher than values commonly found in waste glass.
The chemical durability of waste glasses has been
evaluated for a wide variety of glasses, wastes, process-
ing types, and storage conditions. In addition, tests have
indicated that non-glassy by-products of vitrification (i.e.
metal wastes and devitrified or crystalline wastes) may
also demonstrate high contaminant immobilization.
Leach Tests of Waste Glasses. Waste glasses have been
subjected to numerous leach tests. These tests include
the 24-hr Soxhlet Leach Test, the 28-Day Materials Char-
acterization Center Test (MCC-1), the EP Tox, the TCLP,
and the method used by the International Atomic Energy
Commission (IAEC) (Buelt et al., 1987; Komatsu et al.,
1990). Results generally indicate leach rates below the
levels set as acceptable by the EPA. Sample leach rates
for selected metals are presented in Table 5-1. Based on
the results of EP Tox and TCLP tests conducted thus far,
it is likely that waste glass may be below regulatory levels
under the provisions of these tests.
One cautionary note concerning leach rates: low leach
values of the product do not necessarily indicate chemical
immobilization; they may indicate failure to incorporate
the metal of concern into the waste glass. For example,
mercury may volatilize during vitrification and fail to be
incorporated in the melt. If this happens, low leach values
for mercury may be more indicative of the particular
process's difficulty in incorporating mercury into glass
than of a leach resistant glass. Therefore, for volatile
contaminants, a mass-balance may need to be performed
to fully address the chemical immobilization afforded by a
particular glass.
The leach rate of ISV waste glass was compared with the
5-1
-------
Table 5-1. TCLP Leach Data for Selected Processes and Selected Metals*
Glass Kiln/Vitrification ISV
Metal Melter1-" Process2-" Glass2-0
i
Arsenic <0.02 <0.01 ; <5
Barium <0.05 <0.175 0.05
Cadmium 0.007 0.015 <1
Chromium 0.03 0.825 <1
Lead <0.05 0.15 <1
Mercury <0.0002 0.00035 <0.03
Silver <0.01 0.01 <0.1
1in ppm
2in mg/l
"Penberthy Electromelt International, Inc., vendor information
bHartow et al., 1989 !
eFamsworth, Oma, and Bigelow, 1990.
* As original contaminant concentrations and prpcess DRE's were not
directly comparable. This data is presented to [show that, in general,
i
ISV TCLP
Metal2-0 Limits2
<5 5.0
<1 100.0
<1 1.0
2.7 5.0
<1 5.0
<0.03 0.2
<0.1 5.0
always supplied, this leach data is not
vitrification products pass TCLP limits.
leach rate of other durable glasses using data from [the
Soxhlet Leach Test. Results indicated that the leach rate
of the ISV waste glass is significantly less than thajt of
marble or bottle glass and is comparable to Pyrex glass
and granite (Buelt et al., 1987). Figure 5-1 presents ^his
comparison.
Leach Tests of Non-glass Waste Forms. In addition toithe
leaching of waste glass, the leaching of non-glass waste
forms has been studied. Non-glass waste forms produbed
during vitrification include crystalline material mixed in'the
product and the metal that may settle to the bottom of [the
melteror molten region.
Researchers at ORNL have compared the leaching of!the
crystalline-phase ISV product to the glass-phase ISV
product. Results from a field demonstration indicated that
every element tested, with the exception of Cs, was more
extractable into 0.1 normal HCI from the crystalline phase
than from the glass phase, but generally by a factor of ibss
than 10, Although the absolute magnitude of the concen-
trations of the elements released by acidic extraction is
not directly comparable with the release rates under
environmental conditions, it does support the conclusion
that most elements will be more susceptible to leaching
from the crystalline phase than from the glass phase in
these systems (Spalding and Jacobs, 1989). This conclu-
sion is supported by research for both commercial and
defense nuclear waste glass. Under most conditions,
teachability can increase after samples are devitrified, put
this increase is generally less than a factor of 10 (Wicks,
1985).
The metal product formed from the pooling of metal at the
bottom of an ISV glass monolith has also been leach
tested. The metal product from bench-scale tests on
simulated wastes in Idaho National Engineering Labora-
tory (INEL) soils indicated that it would pass the TCLP.
These tests included arsenic, barium, cadmium, chro-
mium, and silver in the melt (Farnsworth, Oma, and
Bigelow> 1990). The metal ingotformedfrom an engineer-
ing-scale test of ISV of a UST passed the EP Tox for the
8 metals tested: arsenic, barium, cadmium, chromium,
lead, mercury, selenium, silver (Campbell, Timmerman,
and Bonner, 1990).
Toxicity Tests of ISV Waste Product. EPA compared the
teachability of chemically-stabilized soil and vitrified (ISV)
soil in treatability studies for the Western Processing
PYREX
VITRIFIED
HANFORD
SOIL
GRANITE
MARBLE
BOTTLE
GLASS
0123456
SOXHLET CORROSION RATE, g/cm2-d x 10'5
Figure 5-1. Leach Resistances of Selected Materials
(Buelt et al., 1937)
5-2
-------
Superfund Site. In this study, eiuates were prepared from
untreated soil, soil stabilized by three chemical stabilization
processes, and vitrified soil. They were then compared
for metal releases and toxicity effects on algae and an
invertebrate. Results indicated that chemical stabilization
increased toxicity to both organisms. In contrast, testing
of the vitrified soil indicated that vitrification reduced
toxicity to the algae, with no toxicity to the invertebrate.
The researchers concluded that the stabilization tech-
niques were inappropriate for the Western Processing
Superfund Site, but that vitrification was appropriate (Green
etal.,1988).
WIPP In Situ Testing Program. The durability of nuclear
waste glasses over time in the storage repository is an
important consideration in decisions concerning glass
composition, packing materials, and container materials.
The first in situ tests involving burial of simulated HLW
forms conducted in the United States were started on July
22,1986. These tests are being conducted at the Waste
Isolation Pilot Plant (WIPP) near Carlsbad, New Mexico.
This effort is known asthe Materials Interface Interactions
Tests (MIIT) and is international in scope. The MIIT
program is a joint effort managed by Sandia National
Laboratories in Albuquerque, New Mexico and the Sa-
vannah River Laboratory (SRL) in Aiken, South Carolina
and sponsored by DOE. In the MIIT tests, multiple
nonradioactive waste glass samples were placed in brine
and salt in an underground test facility under conditions
simulating those in a salt repository. Included in these
tests were over 900 waste forms comprising 15 different
systems supplied by 7 different countries. In addition to
the waste glass samples, 300 potential canister (or
overpack) metal specimens and 600 backfill and salt
geologic samples are being tested (Wicks and Molecke,
1986; Wicks etal., 1990).
While the total program was scheduled to run 5 years,
samples and aliquots of solution were removed and
studied after 0.5, 1, and 2 years. Preliminary results
indicated that the SRL waste glass system SRL 165/TDS
performed well in the salt environment at WIPP and was
not significantly affected by proposed canisteror overpack
metals. Leach data indicated that the rate of dissolution
decreased with increasing time. In addition, selective
leaching appeared to be the main leaching mechanism.
The leaching process was characterized by the formation
of two precipitate layers over three glass interaction
zones that contribute to protecting the glass from further
leaching with time. Additional results will correlate the
results of the SRL system with differing glass systems of
other countries and provide additional details on the
leaching mechanisms involved (Wicks et al., 1990).
5.1.2 Physical Durability
Waste glasses produced by staged ISV and JHCM vitri-
fication were compared for their compressive and tensile
strengths and found to be very resistant to fracture into
smaller pieces (Koegler etal., 1989). The waste for these
tests came from the raffinate sludges at Weldon Springs,
Missouri. These glasses were produced from predicted
waste stream compositions given bench-scale test results,
and their strength performances are presented in Table
5-2. Two formulations of ISV waste glass were tested
based on alternative ISV implementation plans. As the
data shows, the vitrification processes tested had similar
strength characteristics. The compressive and tensile
strength of concrete is included for comparison. Fromthis
comparison, it can be seen that the strengths of waste
glasses range from 5 to 20 times that of concrete.
The impact resistance of partially devitrified and glassy
waste glass systems was studied at various temperatures
and impact velocities for both small and larger scale
samples at PNL and SRS. Results indicated that for
extreme-case scenarios simulating a high-speed train
impact at 80 mph, the fracture of glass was localized to the
area of impact. Furthermore, the surf ace area of the glass
in the canisters was limited to an increase of about a factor
of 40. No major differences were observed between small
and large samples, nor between glassy and partially
devitrified products. Finally, the amount of particles
smaller than 10 u.m produced after impact was small.
Minimizing the production of particles of this size is
important because they are potentially dispersible via air
currents (Wicks, 1985).
Table 5-2. Strength Comparisons of Waste Glasses Produced by ISV and a JHCM (Koegler et al., 1989)
Source of Waste Glass
ISV (50% sludge/50% soil)
ISV (20% sludge/10% soil/70% liner)
JHCM
Unreinforced Concrete
Compressive Strength (psi)
59,350
43,210
43,210
3,000 - 8,000
Tensile Strength (psi)
4,410
4,309
4,300
400 - 600
5-3
-------
Impact resistance studies were also conducted at Argonne
National Laboratory for impact energies up to 10 J/cm3.
The amount of dispersible fines produced was very
similar to the amount of fines produced after impacting
common industrial glasses such as Pyrex (Wicks, 1905).
i
5.1.3 Devitrification
Devitrification is the formation of a non-glassy, crystalline
structure in the waste product. Devitrification may occur
during cooling of the molten glass. Devitrification may
also occur after the glass has cooled if, for some reason,
the amorphous glass structure crystallizes. The degVee
of crystallinity and crystalline phases that may be produced
depend on factors such as the specific cooling rates,
sizes of the forms, and the physical and chemical compo-
sition of the waste-glass.
Assuming that the melt has a chemical composition
appropriate for glass formation, devitrification during
cooling may be caused by slow cooling and subsequent
nuclealion growth. Slow cooling increases the likelihood
that a crystalline structure can form before the amor-
phous structure "freezes." Fast cooling minimizes devit-
rification by "freezing" the amorphous structure of the
molten glass into a solid. As a result, because large SV
melts cool more slowly, they may have higher concentra-
tions of crystalline structure than monoliths arising from
smaller melts. Furthermore, devitrification will most likely
occur in the center of the waste glass due to slower heat
tosses there (Means et al., 1987; Jantzen and Bickford,
1985). Rapid cooling of the molten glass past the anneal-
ing range may reduce devitrification quantities in wastes
where this is a concern (Wicks, 1985,1986). i
If reheated, glass may also devitrify after it has hardened.
This is particularly a concern for nuclear glasses, as ^he
radioactive decay process may generate additional h|eat
in the glass. Temperatures in the waste glass may be
reduced by adding less radioactive waste to the glass or
by using aged waste. However, the storage temperature
of defense waste glass will generally be less than 100|°C,
and devitrification occurs at temperatures above 500° C.
Furthermore, as indicated above, even if devitrification
increases waste glass leaching, data indicates this yvill
still be at acceptable levels (Wicks, 1985).
5.1.4 Estimation of Long-Term Durability
Because the long-term durability of waste glasses has not
been directly measured, estimates of long-term durability
are very important. Natural glasses, such as obsidian and
basalt, and durable synthetic glasses, such as Roman
glasses, give some idea of the potential durability! of
waste glasses. However, there is a wide range in their
measured durability, from millions of years for the natural
glasses to the several centuries demonstrated so far for
the synthetic glasses. Furthermore, the fact that glasses
in general may last a long time does not mean that a
particular waste glass will last as long. Nor does it permit
quantitative predictions of the expected durability of the
specific waste glass (Jantzen, 1988).
To address these concerns, two different methods have
been used to predict waste glass performance: kinetic
models and thermodynamic models. Kinetic models
mathematically describe the processes that affect the
leaching behaviorof a glass: ion exchange, diffusion, and
the formation of protective layers. Kinetic models describe
the leaching behavior of a glass over time and indicate
that waste glasses should be very durable. Some of the
components of a kinetic model are described in Chapter
Two. However, kinetic models cannot identify which of
several glasses is predicted to be most durable (Jantzen,
1988; Wicks, 1985).
The thermodynamic approach is based on the work of
Newton and Paul (1980). They found a logarithmic
relationship between the free energy of hydration of
glasses and measures of reaction progress, such as K2
-------
The long-term durability of ISV glass has also been
calculated. This was accomplished by estimating the
leaching behavior of ISV for a 1,000-year period based on
the leach results of the MCC-1 test. These results were
then compared with the leaching of obsidian over 1,000
years. Given the similarity in leaching behavior over
1,000 years, and given the structural similarity of ISV
glass and obsidian, the ISV glass was estimated to have
a durability similar to obsidian, on the order of 1,000,000
years (FitzPatrick, 1986; Buelt et al., 1987).
5.2 Product Volume Reductions
and Densities
During vitrification, the incoming waste is generally re-
duced in volume and increased in density. This reduction
in volume is the result of the vaporization of void gases,
the vaporization of water in the feed, and the combustion
of organic materials present in the feed.
Volume reductions include: 25-45% for ISV; 70 - 80% for
glass melter vitrification of incinerator ashes; 90% for
glass melter vitrification of asbestos wastes; and 98-
99.5% for microwave melter vitrification of liquid and
sludge wastes. Obviously, volume reduction values will
vary widely with waste feed. As water is vaporized and
organic components are destroyed, waste feeds with
high moisture contents and/or high organic content will
likely have greater volume reductions than those without.
The volume reduction during an ISV melt results in a
depression at the treatment site which may be filled with
clean soil or other fill.
The density of vitrified products ranges from 2.3 to 3.0 g/
cm3. The ISV product has been measured at 2.3 to 2.65
g/cm3 (Buelt et al., 1987), while ex situ vitrification products
have been measured at 2.7 to 3.0 g/cm3 (Komatsu et al.,
1990). Differences in the densities appear to be due to the
increased control which operators have over the ex situ
methods. However, all values are well above the densi-
ties of 0.7-2.2 g/cm3 measured for stabilized/solidified
products (Stegman, Cote, and Hannak, 1988).
5.3 Product Use
Potential uses for melter glass include aggregate, glass
wool, and other ceramic products. In general, given the
variation present in most waste streams, making a con-
sistent, sophisticated glass product from waste glass may
prove difficult (Roberts, 1989).
Depending on how the molten glass is treated, different
products may be formed from the product. If the molten
slag is poured into water, the glass shatters as it cools and
a glass frit is formed. This frit may possibly be used as is
for aggregate in road building, or for abrasive materials
such as sandpaper, shot blast, or grinding media (GRI,
1989; The Hazardous Waste Consultant, 1990a). The
size of aggregate pellets may be controlled by varying the
speed of the screw conveyor into which the molten glass
is poured (Harlow et al., 1989).
If the molten glass is spun as it cools, glass fibers will be
formed which can be used as mineral wool or glass wool
for insulation or in other ways (Vaux, 1988; Hnat et al.,
1990a).
Potentially, the waste glass may also be molded or
reformed and used in the production of architectural
materials (such as wall and floor tiles, patio stones,
mosaics, sinks, tubs, and countertops) or refractory ma-
terials (such as high-temperature bricks and other insu-
lating materials) (Roberts, 1989; The Hazardous Waste
Consultant, 1990a).
The ISV monolith has several potential uses, including
building foundations and subsurface barriers. Subsur-
face barriers would prevent groundwater and biological
organisms from moving into or out of the contaminated
area. Thus, the contaminants would be isolated from the
ecosystem and the site spared the need for further
treatment (Shelley, 1990; Buelt et al., 1987).
5-5
-------
-------
CHAPTER SIX
OFF-GAS TREATMENT
Off-gases from the melt may contain volatilized metals
and other inorganics, undestroyed organics, by-products
from the pyrolysis of organics, and other chemicals of
concern. Discussion of off-gases addresses three areas
in this chapter: potential off-gases of concern, potential
impacts of off-gas constituents, and potential controls
aimed at minimizing undesired off-gases.
6.1 Off-Gas Components
Off-gases are composed of inleakage air, water vapor,
chemical decomposition products (e.g., CO2, H2O, and
HCI), and entrained particulates, as well as volatilized
inorganics.
Inleakage aircomesfrom leaks inthe melterthat permit air
to move into the melter. The primary source of inleakage
air is the air that enters the furnace with the feed material.
Other factors, such as age of the equipment, may also
contribute to leaks. Inleakage air is a concern because it
may create convection currents in the plenum that may
entrain particles and contaminants from the cold crust.
These particles complicate off-gas treatment (Holton et
al., 1988).
Water is vaporized in the plenum by contact with the melt
as the feed is rapidly heated by the high temperatures
there.
Depending on the feed material, the products of decom-
position may form a significant component of the off-gas
system. Feed material with high quantities of combus-
tibles, concrete, and/or other gas-producing materials
may produce significant amounts of gas in the melt (Buelt
etal., 1987). The significance of these gases is that they
may form an important pathway for the movement of
inorganics out of the melt (carrier gas transport). There-
fore, high quantities of gas-producing materials may re-
sult in the need for an effective off-gas system.
Entrained particles may be produced from the feed dust
(Koegler et al., 1989). They also may be produced at the
high temperatures of vitrification from volatile glass com-
ponents (Bonnioud et al., 1986). Again, these may serve
as carriersfora variety of contaminants such as inorganics
and PCBs (Battey and Harrsen, 1987). Entrained losses
represent a physical loss mechanism (Goles and Ander-
son, 1987).
Volatilized inorganics are a concern because they are
often represented by the contaminants that are sought to
be controlled. The difficulty in incorporating these in the
melt has already been described (Chapter Four) and will
not be repeated here.
6.2 Constituents of Concern
Classes of constituents that are commonly controlled
during vitrification include the following:
metals - volatile and semi-volatile
organics
particulates
sulfates and sulfur oxides (SOX)
nitrogen compounds (NOX)
carbon monoxide (CO)
hydrogen halides and halogens
These compounds are of concern for a variety of reasons.
Metals and organics represent the very contaminants
which treatment by vitrification is attempting to control.
Particulates function as an important pathway in carrying
contaminants from the melt and through the off-gas
system. Sulfates and sulfur oxides, nitrogen compounds,
and CO may be produced during the vitrification process
and represent air pollutants which must be controlled.
These compounds also may cause corrosion of the melter.
CO is characteristic of incomplete combustion and may
indicate that greater processing controls need to be
exercised in the plenum or secondary combustion
chamber. Halogens such as flourine and chlorine are
difficult to incorporate into the melt and may also corrode
the melter if they evolve as off-gases.
6-1
-------
6.3 Means of Off-Gas Control
There are a number of methods available to control off-
gas emissions. These may be broken down into two
strategies: reduction of emissions to the off-gas system,
andtreatment of evolved off-gases. These strategies can
be grouped as follows:
Reduction of emissions:
• modification of the feed
• presence of a cold cap
• control of the environment in the secondary
combustion chambers j
i
• recycling of contaminants captured in the
off-gas system i
I
Treatment of evolved off-gases:
• operating with negative pressure
• design of the off-gas treatment systems
6.3.1 Reduction of Off-Gases to Off-Gas System
i
Reduction of off-gases forms an important means of off-
gas control. Numerous methods permit control of off-
gases at the source of production.
i
Modification of the Feed. Feed modifications may inblude
adding materials to or removing materials from the waste
before melting. For example, fluxes may reduce volatil-
ization of inorganics by permitting melting at lower( tem-
peratures. Melting at lower temperatures decreases the
percentage of particles that receive enough energy to
volatilize. These particles may then be removed as the
glass is tapped before they volatilize. However, addition
of fluxes may change the character of the melt and
increase volatilization in other ways. Forexample, bench-
scale test for ISV application showed that sodiunji car-
bonate additions reduced melt temperature, but caused
no net decrease in cesium volatilization because in-
creased gas evolution resulted in increased cesium en-
trainment (Spalding and Jacobs, 1989).
Modification of feed may also change the chemistry of the
melt and potentially increase retention efficiencies. For
example, research has indicated that high mercury^ con-
centrations and/or high halogen contents may increase
cesium volatilization, although by different mechanisms
(Goles and Anderson, 1986). Reducing agents such as
carbon and ferrous salts may reduce arsenates and
selenates to lower valence compounds that are more
volatile, and thus reduce incorporation efficiencies of
these metals (Schreiber, 1988). If possible, constituents
that decrease incorporation may be removed, or, at least,
not added.
Presence of a Cold Cap. The cold cap helps minimize the
volatilization of contaminants because it holds them iri
contact with the melt until they dissolve or decompose. As
mentioned in Chapter Four, studies done with ISV have
shown that increased cap widths over the molten soil
increase metal retention.
Control of the Environment in the Secondary Combustion
Chambers. Combustion of non-pyrolyzed organics and
pyrolysis by-products occurs in the plenum or in the
secondary combustion chamber. Control of conditions in
these locations helps ensure complete combustion and
thus reduces CO emissions and other products of in-
complete combustion. One common method of modifying
the combustion environment is to increase the flow of air
or oxygen to these locations, ensuring the presence of
adequate oxygen supplies for the combustion of all
combustibles.
Recycling of Contaminants Captured in the Off-Gas
System. A number of methods exist to recycle off-gas
constituents to the melt. In one sense, these could be
considered a treatment method for off-gases produced,
because most recycling methods involve capturing the
constituent of concern in the off-gas and then returning it
to the melt. However, in this chapter recycling will be
considered a type of off-gas reduction method, because
it reduces total emissions of a particular off-gas constituent
by increasing retention efficiencies.
The cold cap ("cold" relative to melttemperatu res) is afirst
line of recycling. Here, volatilized contaminants and other
materials may condense and fall back into the melt, thus
increasing retention efficiencies. This method has been
used in the glass melting industry to avoid losses of
expensive materials, but also has potential for waste
treatment.
Contaminants that leave the melt chamber and enter the
off-gas system may be recycled afterthey are captured in
the off-gas system. Captured contaminants may be
recycled by removal from the off-gas system component
in which they were captured (e.g., by passingthe scrubber
solution through af ilter aid and activated carbon) (Hansen,
1991), or by placing the off-gas system component in
which they were collected directly into the melt (e.g., a
contaminated HEPA filter). This form of recycling the off-
gases is a very powerful tool for increasing retention
efficiencies. Repeated recycling of off-gases may even-
tually drive retention efficiencies close to complete incor-
6-2
-------
poration. However, repeated recycling may also in-
crease processing complexity, total treatment time, and
costs.
All vitrification processes are amenable to recycling in
some way or another. Ex situ processes may be more
flexible to recycling options. However, ISV is also ame-
nable to recycling. For example, at the 1991 ORNL pilot-
scale test, a HEPA filter was placed at the junction of the
hood and the ductwork carrying off-gases to the off-gas
system. This HEPA filter reduced Cs emissions to the off-
gas system and simplified off-gas treatment (Spalding et
a!., 1991). This filter could potentially be dropped into the
melt close to power termination. Recycling may also be
achieved at a subsequent ISV melt by placing secondary
wastes from the previous melt into the soil at the second
site before starting the melt. Forexample, pulsed periodic
backflow through the HEPA filter could be used to flush
contaminants from the filter. The flushed particulates
could be dropped directly into the melt, or they could be
deposited in a shielded container and buried at the next
ISV site. The filter itself could also be unloaded, placed
in a shielded container, and buried at the next ISV site.
6.3.2 Treatment of Evolved Off-Gases
Contaminants that have entered the off-gas stream will
need to be removed. These contaminants may be
recycled as described above, or disposed of as secondary
wastes. If the contaminants contain valuable metals
(such as mercury), they may possibly be recovered and
sold. This section is concerned only with ensuring that
gases in the off-gas system do not enterthe environment
before contaminants have been removed.
Operating with Negative Pressure. Operating the entire
process at negative pressure is the first step for ensuring
that contaminants do not enter the atmosphere. How-
ever, treating wastes with high amounts of organicwastes
may produce enough gases to overwhelm the negative
pressure in the off-gas system.
Two intermediate field tests of ISV were conducted at
INEL in the summer of 1990 to examine the applicability
of ISV to buried waste, potentially a major source of
combustibles (Callow, Thompson, and Weidner, 1991).
For these tests, pits were dug and filled with drums and
boxes to simulate waste burial sites at INEL.
In the first of these tests, gas releases from containers
resulted in 14 separate events, characterized by sharp
temperature increases and/or pressure spikes in the
hood. The pressure spikes were the result of either
relatively slow gas releases from the melt, or relatively
slow expansions of gases in the hood that occurred over
a 10- to 30-second period. The pressure spikes were not
rapid and therefore not characteristic of an explosive
reaction. The intermediate-scale ISV system was unable
to contain transient pressure spikes on several occa-
sions. In cases where the pressure did not exceed 1 in.
of water, the gasses were contained within the surge
volume of the hood and subsequently drawn out to the
off-gas treatment systems. When the pressure exceeded
1 in. of water, the hood was able to handle the pressure
spike by relieving a portion of the excess gas through the
HEPA-filtered pressure relief system. In extreme cases,
when the pressure significantly exceeded 1 in. of water,
the gas overcame the surge and pressure relief capacity
of the hood and was released through any available point,
including the base of the hood and through unsealed
panel seams.
In the second intermediate field test, overburden was
placed over the melt and the electrodes and EFS were
slightly modified. These changes reduced the strength of
the transient pressure spikes and the second test suc-
cessfully avoided the problems associated with the first
test.
Design of the Off-Gas Treatment System. Off-gas sys-
tems may remove particulates, recover heat and cool off-
gases, neutralize acid gas, and remove water vapor.
Components used to achieve these objectives include
scrubbers, filters, spray chambers, spray channels,
baghouses, and others. The off-gas systems of selected
vitrification processes are presented in Table 6-1. Table
6-1 does not evaluate the efficiency of selected pro-
cesses; it simply represents off-gas systems used in
vitrification. Depending on site conditions and treatment
goals, selection of an off-gas system may vary.
One concern of efficient off-gas treatment is the location
of secondary waste in the off-gas system. The result of an
ISV pilot-scale radioactive test will give some idea of how
contaminants may behave in the off-gas system. This test
was performed on transuranic-contaminated soil from a
storage crib at Hanford (Timmerman and Oma, 1984). In
addition to the transuranic contaminants, mixed fission
products of 137Cs, 106Ru, 90Sr, and 60Co were added
to the contaminated soil in order to study their behavior
during an ISV melt. Table 6-2 presents the distribution of
the material released from the melt to the off-gas system.
While a small fraction of the off-gas fission products (3%)
reached the primary HEPA filter, there were no fission
products on the second-stage HEPA filter or in the stack
samples taken downstream of the filters. The scrubbers
accounted for removal of 65% to 92% of the radionuclides
released to the off-gas, as indicated by the distribution of
the radionuclides in the two scrub solutions.
6-3
-------
Resultsfromprevious non-radioactive tests hadindicated
that the average mass-mean diameter of particles exiting
the ISV hood was 0.7 pirn, while the scrubbing efficiency
of venturi scrubbers dropped off for particles smallerthan
about 0.5 urn. Given these characteristics of the off-gas
system in the pilot-scale radioactive test, certain hypoth-
eses concerning contaminant behavior as an off-gas
could be made. Rrst, the elevated distribution of cesium
and ruthenium between the venturi and hydrosonic scrub
solutions indicates that more of these volatile radionu-
clides were being released as very small particles. The
distribution of cobalt indicates that these particles were
also very small. Secondly, the transuranic elements were
collected primarily in the venturi scrub solution, indicating
that these elements were released as larger particles.
Possible release of the transuranics as larger particles
indicates that release of the transuranics may have been
increased by the combustion of the test package containing
the contaminated soil andf ission products. This hypothesis
is also consistent with the time that the transuranic levels
started to increase in the scrub solution.
An important concern in waste treatment via ISV is sec-
ondary contamination. A pilot-scale ISV test revealed the
location of the secondary contamination upon completion
of the melt. In this test less than 1 % of the radionuclides
that escaped the melt settled on the ground. Less than
20% were deposited on the hood and off-gas piping. The
20% figure was for cobalt. The other nuclides were all
less than 10%.
Contamination on the ground can be fixed in place before
moving the hood and can be pushed into the subsidence
zone when backfill is added. Deposits on the hood and
off-gas piping can be fixed in place by spraying strippable
fixatives so that the hood and off-gas line can be moved
without concern for loose contamination. The fixative will
then combust during subsequent operations. Spraying
techniques forthese procedures have been demonstrated
with both pilot- and large-scale systems.
ISV was field demonstrated on a simulated radioactive
liquid waste disposal trench at ORNL in July, 1987
Table 6-1. Off-Gas Systems for Selected Processes
Process
Off-Gas Component
Function
kiln1
t
oxidizer (three in series)
semi-dry caustic scrubber
fabric filter baghouses
combust organics
neutralize acid gases
remove dust and paniculate
glass melter2
ceramic fiber filters
gas-to-water heat exchanger
water spray chambers (two in series)
demisting chamber
heater
charcoal and HEPA filters
remove particulate
cool gases
neutralize acid gases
remove water droplets
re-heat gases above dewpoint
filter remaining particles
ISV3
HEPA filter (optional)
scrubbers (two in series)
condenser ;
heater
HEPA filters (two in series)
initial filtering depending on treatment goals
cool gases and remove particulate
remove water vapor
re-heat gases above dewpoint
filter remaining particles
coal-fired vitrification
furnace4
recuperator
quench water ',
precipitators and stack assemblies
recover heat
cool gases
remove particulate
'Harlowetal., 1989
'Froeman, 1986
'Battey and Harrsen, 1987
••Hnatetal., 1990a
6-4
-------
Table 6-2. Radionuclide Distribution in the Off-Gas System
During an ISV Pilot-Scale Test (Timmerman and Oma, 1984)
Percent (%) of Total Radionuclides Released to Off-Gas
Nuclide
Pu-239
Co-60
Sr-90
Ru-106
Cs-137
1 Currently, a
Ground
Surface1 Hood
0.09
0.05
0.05
0.3
0.2
non-combustible fabric
4
4
4
8
2
is placed
HEPA Filters
Piping
<1
14
<1
2
5
as a ground
Tankl
92
24
91
21
17
cover inside the
Tank 2
3
55
5
66
73
hood to prevent
Stage 1
1
3
0.1
3
3
Stage 2
0
0
0
0
0
surface contamination.
Table 6-3. Hypothetical Distribution of 137Cs Activity in ISV Off-Gas System After
Vitrifying 10,000 Ci (Spalding and Jacobs, 1989)
Off-Gas System
Component
Hood
Ducting
Scrub Solutions
Primary HEPA Filter
Secondary HEPA Filter
Amount of
137Cs (Ci)
0.079
0.401
11.74
0.293
0.0008
Amount per
Unit Area
0.24u,Ci/cm2
5.5 u.Ci/cm2
0.5 u,Ci/mL
1 .46 u.Ci/cm2
0.04 u,Ci/cm2
(Spalding and Jacobs, 1989). At this test, samples were
taken at various points in the off-gas treatment system to
determine what was happening to 137Cs that was not
incorporated into the melt. Based on the results of this
test, hypothesized results for the treatment of a trench
requiring 10 ISV settings is presented in Table 6-3.
While retention efficiencies for 137Cs were measured at
99.88%, the amount to be treated by the off-gas system
would generate significant secondary wastes over the
course of treatment for an entire trench. Therefore, ways
to reduce release to the off-gas system were examined.
One recommendation involved placing a HEPA filter that
would filter off-gas before it entered the system. This
HEPA filter would thus filter incoming air and minimize
entry of contaminants to the off-gas system and genera-
tion of secondary wastes. A pilot-scale ISV test at ORNL
in May, 1991, indicated that a pre-filter successfully
captured 137Cs from the off-gases before they entered
the wet scrubbing system (Spalding et al., 1991).
6-5
-------
-------
CHAPTER SEVEN
CAPABILITIES AND LIMITATIONS
This chapter describes the capabilities/advantages and
limitations of vitrification that have been identified in the
literature. As always, for specific site conditions, pro-
cessing goals, and vitrification processes, these factors
may or may not apply.
7.1 Capabilities
Potential capabilities and advantages of vitrification
include:
Thermal destruction of organics
Reduced leachability of immobilized inorganics
Long-term durability of the product
Application to a wide variety of waste streams
Application to a wide variety of contaminants
Volume reduction
Potential re-use of product
Avoidance of excavation, processing, and
reburial of product (ISV).
The first four of these capabilities are the direct result of
the product. Capabilities five and six are the result of the
flexibility of glass and the high temperatures under which
vitrification is conducted. The last two capabilities are
highly site-specific and will depend on site characteristics
and treatment objectives.
Thermal Destruction of Organics. As described in Chap-
ter Four, organics may be destroyed by pyrolysis and
combustion during vitrification.
Reduced Leachabilitv of Immobilized Inorganics. As
Chapter Five pointed out, waste glasses of many types
have shown reduced leachability of inorganics.
Long-term Durability of the Product. Long-term durability
indicates a product that reduces leaching for long periods
of time. It is possible to have a material that currently
reduces leaching but that may not perform well over
many years. Although not measured directly, the long-
term durability of waste glass appears to be excellent and
may extend to geologic time periods, as indicated by
natural glass systems that have been stable for millions of
years.
Application to a Wide Variety of Waste Streams. A rep-
resentation of the waste streams to which vitrification may
apply has been presented in Chapter Four. These will not
be re-listed, but a review of that chapter will reveal that
vitrification is potentially applicable to a wide variety of
waste streams.
Application to a Wide Variety of Contaminants. Because
vitrification is both a high temperature process and an
immobilization process, it can apply to both organics
(thermal destruction) and inorganics (immobilization).
Vitrification may, therefore, be preferred at sites that
present a complex mixture of hazardous and/or radioac-
tive contaminants.
Volume Reduction. Not only does vitrification produce a
long-term, chemically durable product, but it can reduce
waste volume during processing. Thus, vitrification sim-
plifies waste management.
Potential Re-use of Product. As described in Chapter
Five, vitrified waste glass may potentially be re-used in
various ways. Re-use may depend upon whether the
product can be delisted according to EPA regulations and
on whether the public will accept re-use of a product
formed from hazardous wastes.
Avoidance of Excavation. Processing, and Reburial of
Product. This applies only to ISV and may be important
in two respect: worker safety and costs. If the site being
remediated is highly contaminated, worker safety may be
an over-riding concern. Potential worker contamination is
minimized with ISVbecause contaminants are notbrought
to the surface. Costs may also be reduced by ISV
because it avoids the costs of excavation, material han-
dling, and disposal. If ISV is to be applied in staged
application it will lose some of the in situ benefits.
7-1
-------
Conversely, when compared with ISV, ex situ vitrification
processes permit greater control over processing pa-
rameters. Included in control over processing parameters
would be greaterease of feed modification, greater control
of melt parameters, greater control of product character-
istics, and greater control of organic combustion and off-
gas treatment. The trade-off in capabilities and disadvan-
tages between ISV and ex situ vitrification processes are
similarto those that exist in general between in situ and ex
situ processes. Their relative importance will depend to
a large extent on site conditions and treatment goals.
7.2 Limitations !
The following may limit the effectiveness of vitrification:
• Feed moisture content
• Feed material composition
Feed compatibility
• Presence of combustible material ;
• Presence of process-limiting materials
• Potential volatilization of contaminants
• Potential shorting caused by metals
High cost of energy
High cost of trained operators
Depth limitations (ISV).
Feed Moisture Content. Feed moisture content has im-
portant impacts on vitrification economics, but in (itself
may nottechnically limit vitrification applicability. Vitrifica-
tion may potentially drive off water during treatment, but
in so doing requires more time, more energy, and, [thus,
drives up costs. '•
Limits of moisture content will depend on the process, but
limits of 25 wt% and 20 wt% have been identified for some
ISV processes (USEPA, 1987; USEPA, 1988). It is also
possible that at greaterthan 5% free water, the water may
react vigorously with the melt as it rapidly vapojrizes
(USEPA, 1990a). However, the DWPF has a much hjgher
moisture content (>50 wt%) without a violent reaction
occurring. One alternative to increase process ability to
handlefeedswith high moisture contents is to use heaters
in the plenum. Plenum heaters may increase the speed
with which water is vaporized and thereby the incorpora-
tion of feed material into the melt. By increasing incorpo-
ration speed in this way, treatment time and costs Will be
lowered.
i
ISV also may be able to drive off high concentrations of
water. Most important in limiting ISV in areas of high
moisture is soil permeability. As a general rule, soils
having low permeabilities do not inhibit the ISV process,
even in the water table, because the recharge rate is not
significant in terms of the processing rate. The ISV melt
advances at about 7 to 15 cm/h and soils with permeabilities
of 10'5 cm/s or lower are thus considered to be verifiable,
even in the presence of grou nd water or in the watertable.
Soils with permeabilities of 10"£i to 10'4 cm/s are con-
sidered marginally verifiable. Soils with permeabilities
higher than 10'4 cm/s may require additional steps, such
as drawing the local water table clown by pumping and/or
installing underground barriers, prior to ISV (Buelt et al.,
1987).
Feed Material Composition. Feed material composition is
defined here as the chemical composition of the material
that is fed into the furnace or melter. As addressed in
Chapters Two and Five, feed composition may impactthe
abi lity of the vitrification process to form a durable product.
Ex situ processes have an advantage in treating feeds
with difficult compositions because additives can more
easily be added to address feed difficulties. For example,
IRI prepares its feed in batches. Samples of an incoming
batch are taken and additives varied according to kiln
requirements before the batch is fed to the kiln for pro-
cessing (The Hazardous Waste Consultant, I990a).
Potential problems resulting from difficult feed composi-
tions are compounded with HLW because of the very
hazardous (radioactive) nature of the waste. In many
other vitrification applications, glass of poor quality may
be remelted and reformed to improve quality. However,
because of the hazard of HLW, the product glass cannot
be re-melted once it is made. For this reason, the
remediation process at SRS has included the develop-
ment of the statistics-based quality assurance program
described in Chapter Five.
For application of ISV, soils should contain adequate
quantities of glass-forming materials (i.e., SiO2and A^Oa)
and fluxes (i.e., NaaO, K2O, and CaO). The glass forming
compounds in the soil provide the elements which form
the skeleton of the amorphous glass product. Higher
levels of these materials tend to increase the chemical
durability of the resulting glass, but have the negative
effects of increasing its viscosity and decreasing its
electrical conductivity (Buelt etal., 1987). The flux agents
are all alkali elements, such as sodium and potassium,
that carry the charge that conducts the electric current
generating the soil-melting heat during the ISV process.
Therefore, soils with low alkaline contents may be unable
to effectively carry a charge and thereby diminish the
applicability of ISV (Campbell and Buelt, 1990). The
minimum combined alkali concentration determined to be
necessary for vitrification of soils using ISV has been
variously identified as 1.4 wt% (Buelt et al., 1987) and 5
wt% (Lominac, Edwards, and Timmerman, 1989).
7-2
-------
Two sites considered for ISV have soils which proved
difficult for ISV to vitrify because of soil composition:
Arnold Engineering Development Center (AEDC),
Tullahoma, Tennessee and SRS, Aiken, South Carolina.
Table 7-1 compares the compositions of 2 easily vitrif iable
soils with soils from AEDC and SRS.
Bench-scale tests with soils from the AEDC site i ndicated
problems vitrifying a soil which contained only 1% alkali
elements. A flux of 5 wt% to 10 wt% sodium carbonate
was judged necessaryto vitrify the AEDC soil (Timmerman,
1989). The pilot-scale test added 27 wt% sodium car-
bonate to the top three feet of cover soil. Results indicated
that the addition of flux would permit vitrification of AEDC
soil. However, failure to reach target depths (the melt
reached 5 ft.) indicated that treatment would require
injection of the flux agent, rather than just surface
placement, and thus some disruption of contaminated
material would be required (Timmerman and Peterson,
1990).
Bench-scale tests were conducted on SRS soils having
alkali elements (NaaO and KaO) of less than 0.2% wt.
These tests evaluated two enhancement techniques to
enable the vitrification of alkali-depleted soils. The first
was the placement of a starter layer of sand overthe SRS
soil. This technique met with limited success; the process
continued to melt preferentially outward without achiev-
ing a significant downward penetration into the SRS soil.
The second technique required pre-mixing of alkali mate-
rials into the soil, and was highly successful. However,
pre-mixing the soils with alkali constituents detracts from
many advantages of the process of vitrifying the con-
taminants in place without pretreatment. Consequently,
Campbell and Buelt (1990) recommended developing
and testing alternative ways of vitrifying SRS soil in place
without prior removal for alkali mixing.
In summary, some soils may not lend themselves to ISV,
but techniques exist that may address this problem. The
primary alternative is to inject soluble alkaline fluxes into
the soil before vitrifying it. However, the injection of af lux
for the enhanced vitrification of soils has not been field
demonstrated (Campbell and Buelt, 1990). Furthermore,
research on ORNL soils indicated that sodium fluxes may
transport contaminants (such as cesium) as gases evolve
(Spalding and Jacobs, 1989).
Feed Compatibility. Feed compatibility refers to the
physical compatibility of the feed with the vitrification
process: can the process handle all the sizes and types of
material in the feed? For example, at the Weldon Springs
site, Missouri, crushed drums, structural building iron,
process equipment, and a fork-lift truck are among the
debris buried in the quarry (Koegler, Oma, and Perez,
1988). At INEL, buried wastes range from steel drums, to
plywood boxes, to cardboard and fiberboard containers,
to vehicles and large pieces of equipment (Callow,
Thompson, and Weidner, 1991). Locations similar to
these sites represent challenges to most remediation
processes, including vitrification.
Feed materials are fed to various vitrification processes in
a variety of ways, including slurries, calcined powders,
shredded and chopped, bagged, boxed, drummed (as
described in Chapter Four), as well as others. Ability to
handle heterogeneity in the field material also varies with
vitrification process. Pre-treatment by particle classifica-
tion and/or other methods of feed preparation may be
required at many sites prior to vitrification.
Table 7-1. Comparison of Soil Composition (wt%) from Selected Sites
SITE
Oxide
Si02
ALO,
Fe.O,
CaO
MgO
Na20 & K20
Other oxides
Range in
USA1
60-93
5-17
1-11
<1-10
<1-3
<1-9
<1-2
Hanford,
Washington1
60.9
13.6
9.6
6.0
2.9
4.8
2.0
INEL
Idaho1
69.6
11.4
4.1
10.0
-
3.9
1.0
AEDC,
Tennessee2
76.0
9.0
5.6
6.7
0.6
1.0
1.2
SRS, South
Carolina3
92.5
4.8
0.8
0.4
—
0.2
1.6
1Bueltetal.,1987
2adapted from Timmerman, 1989
3Campbell and
Buelt, 1990
7-3
-------
Combustible Material. Combustible materials generate
gases and may include combustible solids, liquids, |and
packages, void spaces, and organics. For example,
buried, combustible wastes at INEL include wood [and
cardboard packaging, wood pallets, and cans and drums
containing combustible organic liquids (Callow, Thomp-
son, and Weidner, 1991). Gas-generating situations may
result from the intrusion of the molten glass into yoid
spaces and the release of entrapped air. Finally, natural
organics, such as the humus in soil, may also be a source
of combustible material. However, gas generation from
the decomposition of humus and other natural chemipals
within the soil is generally considered insignificant (Buelt,
Timmerman, and Westsik, 1989). The main concern with
combustible materials is that the gases they generates will
carry contaminants to the glass surface and away from
the melt. With ex situ processes, if combustibles present
a problem, pre-treatment processes may remove much of
this material and thus minimize this problem. Processes
and batch compositions may also be adjusted to minimize
these effects. I
For ISV, combustibles are not removable if the process is
to proceed in situ. Therefore, production of gases must be
controlled by controlling processing conditions. Further-
more, the production rate of off-gases during the burning
of combustibles must not be high enough to overwhelm
the off-gas system's capacity to maintain a negative
pressure during processing. If this were to occur, [the
fugitive emissions could possibly spread contamination.
Maximum processing events that the off-gas system is
capable of handling are as follows: combustible liquids
(4800 kg/m of depth or 7 wt%); void volumes (4.3 m3/
combustion event); combustible packages (0.9 fa3/
combustion event) ; and combustible solids (3200 kg/m of
depth) (Buelt, Timmerman, and Westsik, 1989).
Two intermediate field tests on buried waste at
indicated some of ISV's abilities to handle buried combus-
tibles. Numerous pressure spikes occurred in the first
test. These pressure spikes lasted from ten to th'irty
seconds and were not instantaneous or characteristic of
an explosion. Furthermore, pressure spikes decreased
with increased depth, indicating the potential for addi-
tional soil to be placed over the buried waste to buff er the
effects of the transient temperature spikes (Callow,
Thompson, and Weidner, 1991).
Presence of Limiting Constituents. Limiting materials affect
processing or product quality and may be present in the
incoming feed. These include halogenated compounds,
reducing agents, and metals of difficult types or High
concentrations.
Halogenated compounds affect product durability be-
cause incorporation into the glass in high enough concen-
trations may produce an undesirable, porous product
(USEPA, I990a). Reducing agents such as carbon and
ferrous salts may reduce arsenates and selenates to
lower valence compounds that are more volatile and thus
reduce incorporation efficiencies of these metals (USEPA,
1990a). Certain metals such as mercury and cadmium
may be undesirable because of theirdifficulty to incorporate
into the melt, their reduction of product quality, and/or
because their volatility requires treatment in the off-gas
system. Metals in high enough concentrations may also
be insoluble in the glass, as all metals have solubility
limitations in glasses (USEPA 1987; USEPA, 1990a).
Finally, with the microwave vitrification of incinerator
ashes, unburned carbon was found to affect processing.
In excess of 5 wt% unburned carbon, the carbon would
rapidly heat and cause arcing, thereby affecting process
performance (Komatsu et al., 1990).
Methods to overcome the presence of limiting materials
includes pre-treatment to reduce concentration levels
and, alternatively, to increase the glass-forming additives
and thereby dilute the difficult materials (USEPA, 1990a).
Treatment of materials with limiting constituents may
therefore be economically limited and not technically
limited.
Potential Volatilization of Contaminants Volatilization of
contaminants refers primarily to inorganics, although or-
ganics may potentially volatilize before pyrolysis. By-
products of incomplete organic pyrolysis may also volatil-
ize. The potential migration of contaminants into the
ambient soil during ISV could also be considered a type
of contaminant volatilization. However, this issue has
already been addressed in Chapter Four and will not be
discussed again here.
Volatilization of contaminants increases the quantity of
secondary contamination and thereby complicates treat-
ment. At high enough concentrations, contaminants may
thereby make vitrification cost prohibitive. Volatilized
contaminants may be recycled to the feed to increase
retention efficiencies, but this complicates treatment pro-
cesses and may drive up costs. Finally, volatilized metals
may potentially be recovered from the off-gas system and
re-used.
Volatilized metals of concern include mercury, lead, and
cadmium. Cesium volatilization during ISV treatment has
been a concern at ORNL, but this problem appears to
have been solved in recent tests (Spalding et al., 1991).
Radium may also be a concern, although tests on Fernald
K-65 residue indicate potential successful treatment
(Janke, Chapman, and Vogel, 1991).
7-4
-------
Potential Shorting Caused by Metals. The presence of
metals in the feed may present another problem: shorting
of the electrodes used in joule heating. Metal induced
shorting is primarily a problem when the feed material has
a high iron content or similar metal. These metals may
sink to the bottom of the melt, concentrate there, and
possibly create a conduction path that may lead to elec-
trical shorting between the electrodes. This problem may
be solved rather easily by electric melters by adding a
bottom tap to remove the accumulated metals and ac-
companying slag. Modification of melter geometry may
improve the efficiency of metal draining (Bickford, Propst,
and Plodinec, 1988). Published metal limits for ISV have
been 90% of th e li near distance between the electrodes or
5wt%ofthemelt(BueIt,Timmerman,andWestsik,1989).
However, the recent development of the EPS has made
these limitations superfluous (Figure 7-1).
The EPS was developed to treat soils characterized by a
high content of metals. With electrode feeding, the four
electrodes that are used to initiate the ISV process are
independently fed to the molten soil as the melt proceeds
downward instead of being placed in the soil prior to test
startup. Upon encountering a full or partial shorting
condition, the affected electrodes are simply raised and
held above the molten metal pool at the bottom of the melt.
During this time, the melt (and molten metal pool) contin-
ues to grow downward. The affected electrodes can then
be reinserted into the melt to their original depth, and all
four electrodes can resume electrode feeding operations.
Electrode feeding is expected to eliminate many other
potential problems that can develop when processing
soils containing high concentrations of metals (i.e., power
limitations, void formation, electrode preplacement)
(Farnsworth, Oma, and Bigelow, 1990).
The EPS has been extensively tested, including field tests
at both INEL and ORNL.
At the first INEL test, some problems with electrical
instabilities occurred. Electrical instabilities appeared for
a variety of reasons, but under test conditions, the EPS
was not able to respond aggressively to the instabilities.
This was because the silicon-based coating applied to
reduce electrode corrosion through oxidation would tend
to stick to the glass. At times the electrodes became
frozen to the cold cap and thus unable to be moved
(inserted or retracted) to respond to electrical imbalances
(Callow, Thompson, and Weidner, 1991).
The second INEL test was conducted without the
silicon-based coating and the EPS performed well: no
sticking was observed and oxidation losses were accept-
able (Callow, Thompson, and Weidner, 1991).
The primary conclusion concerning the EPS from INEL
tests was that uncoated graphite electrodes appear pref-
erable to silicon-based coated graphite electrodes. Oth-
erwise, the EPS seemed to perform well (Callow, Thomp-
son, and Weidner, 1991).
High Cost of Energy. Generally, vitrification does require
large amounts of energy to process wastes. Increased
energy costs drives up process costs. Therefore, anyway
in which vitrification can be used efficiently will help
control energy consumption, drive down costs and help
15.2 cm (6 in.) Graphite electrodes
Access door
Figure 7-1. Schematic of a Pilot-Scale ISV Hood Assembly (adapted from Callow, Thompson,
and Weidner, 1991)
7-5
-------
make vitrification competitive with other treatment
processes. !
A primary way in which vitrification can be used efficiently
is to use it at highly contaminated sites where the contami-
nation is not diluted. For example, IRI felt that its process
would not be cost effective for waste streams from small
generators unless it served as a regional site for a number
of generators (The Hazardous Waste Consultant, 1990a).
Evaluation of ISV for use at ORNL indicated that it would
not be applicable where groundwater had dispersed
contaminants (Spalding, Jacobs, and Davis, 1989)! At
Faslane, in England, vitrification was used at sites highly
contaminated with asbestos, but not at sites with rela-
tively low asbestos contamination (Denner, Langridge,
andAffleck, 1988). Thus, vitrification could be considered
a process for treating the "hot spots."
Vitrification has also been considered for treatment of
incinerator ash. This also could be considered treatment
of hot spots. This is because, in one sense, the incin'era-
tionof a waste stream, MSW for example, destroys many
of the organic contaminants and produces a concentrated
waste stream containing inorganic contaminants. I Re-
ported costs for vitrification of incinerator ash make it a
cost competitive option for immobilization of inorganics in
ash. \
Finally, vitrification feasibility studies should consider
availability of electricity and unit costs in treatment
evaluations for a specific site. Energy costs may also be
reduced by vitrifying during non-peak hours or seasons.
High cost of Trained Operators. Because of the com-
plexity of vitrification, trained operators are required for
both ex situ processes and ISV. This, of course, does not
limit vitrification technically, but may drive up costs
(USEPA, 1990b).
Depth Limitations. Depth limitations apply only to ISV&nd
are a primary limitation of ISV at present. Currentlyjthe
greatest depth achieved has been 5 m by PNL (5.8 rr) by
Geosafe). Sixty percent of DOD contaminated soil sites
extend deeperthan 5 m. If ISV could be extended to 9 m,
then 90% of DOD sites would fall within ISV depths.
The primary problem appears to be heterogeneous power
distributions within the melt: half of the delivered power
is held in the upper third of the melt, and power decreases
as depth increases. This results in a slowing of the melt
advance as the melt reaches an equilibrium and finally
melt advance stops. The result is a melt that spreads.out
more and remains shallower than predicted by early ISV
modeling (see Figure 7-2). The primary need, therefore,
in increasing melt depth is to increase heat near the melt
floor. If this can be addressed, the present depth limit may
well be doubled. Of course, deeper melt penetration will
make ISV applicable to an even greater range of sites.
Possible solutions to increasing heat near the melt floor
include:
Hot-tipped electrodes
Use of passive electrodes (EFS)
Start melt at depth and moves upwards
Vertical thermal barriers (walls or floors)
Hot-tipped electrodes could concentrate current at the
bottom of the melt in a number of ways. These methods
include:
Attaching a molybdenum tip at the bottom of the
electrode. The greaterconductivity of Mo directs
current through this tip.
Covering the upper portion of electrodes with an
electrically insulating material which would then
funnel electricity through the tips of the elec-
trodes.
Introduction of passive electrodes involves the intentional
placement of iron-based metals in the startup layer. The
metal will melt and remain at the bottom of the molten
vitrified zone. This has the effect of diverting the electrical
current near the bottom of the molten mass, as shown in
Figure 7-3. The molten metal thus acts as a "passive"
electrode that diverts electrical current and power near
the bottom of the melt by providing a path of lower
resistance to the electrodes. This creates higher melting
temperatures, which may enhance the downward melting
rate. In addition, the molten metal layeris in direct contact
with the soil being vitrified. The greaterthermal conductivity
of the molten metal may enhance heat transferto the soil,
thereby assisting in the downward melting process. The
use of passive electrodes has been shown to enhance
downward melting rates (Buelt and Farnsworth, 1990).
The development of the electrode feed technique made
the introduction of passive electrodes possible (Buelt and
Farnsworth, 1990). As metals are encountered, the
self-feeding electrodes can be withdrawn slightly from the
bottom of the melt to avoid a direct electrical short. The
influence of the passive electrodes can be controlled by
the separation between the bottom of the graphite elec-
trodes and the molten metal pool; decreased separation
will increase the concentration of current nearthe bottom
and increase the downward melting rate (Campbell and
Buelt, 1990).
7-6
-------
Starting the melt below ground and melting upward may
be another way to increase melt depth. However, this
may result in a cavity below ground level which may
eventually cave in and splash molten glass into the hood
area.
Thermal barriers could be placed alongside the site to be
melted and prevent the movement of glass and heat into
adjacent areas. Thus, the glass and heat energy would be
forced downward and melt depths increased. Apilot-scale
melt at PNL tested the applicability of thermal barriers.
The thermal barriers were placed 6 inches from the
electrodes and proved capable of withstanding melt
conditions for 72 hours without evidence of degradation.
Because these barriers were designed to reshape the
melt and not to prohibit heat transfer, about 1 inch of sand
was fused to the side of the barrier opposite the melt.
Drive mechanism
Molten metal pool
Figure 7-3. The Effect of a Molten Metal "Passive"
Electrode on Electrical Current Distribution
in the Melt (Campbell and Buelt, 1990)
.OmO
0
1.0m 2.0m 3.0m
DISTANCE FROM ELECTRODES (CROSS-SECTIONAL VIEW), m
Figure 7-2. Predicted Versus Achieved Large-Scale
Melt Shape (Buelt et al., 1987)
7-7
-------
-------
CHAPTER EIGHT
PHYSICAL AND CHEMICAL TESTING
Performance tests include both physical and chemical
tests and may be performed before treating the material
to be vitrified and after treating the vitrified waste glass.
For additional discussion of performance tests, please
see Stabilization/Solidification of CERCLA and RCRA
Wastes (USEPA, 1989a).
8.1 Physical Tests
Physical testing is conducted to characterize and contrast
waste before and after vitrification. It provides basic
information on the physical characteristics of the waste
material and allows some estimate to be made on the cost
of waste treatment and handling. Physical property
characterization of untreated waste focuses on excava-
tion, transport, storage, mixing, hydraulic conductivity,
strength, and physical durability considerations. Physical
testing of the vitrified product is one part of demonstrating
the relative success orf ailure of the process. The physical
testing methods described in this chapter may apply to
both untreated hazardous wastes and treated hazardous
wastes; however, the tests were not developed for use on
these wastes.
This section describes some of the more common physi-
cal tests used to evaluate waste vitrification processes.
These physical tests include:
Index Property Tests: provide data that are used
to relate general physical characteristics of a
material (e.g., moisture content) to process op-
erational parameters (e.g., costs).
Density Tests: are used to determine weight-to-
volume relationships of materials.
Hydraulic Conductivlitv Tests: measure the rela-
tive ease with which fluids (water) will pass through
a material that is to be vitrified.
Strength Tests: provide a means for judging the
effectiveness of a vitrification process by stress-
ing the product mechanically.
Durability Tests: determine how well a waste
glass withstands environmental exposure.
Individual values of waste properties derived from specific
tests are used along with other available data to make
informed engineering decisions.
It is important to note that many of these tests were
originally developed for testing soils and cement-like
materials for physical stability for construction projects.
Extreme caution must be exercised when applying these
tests to untreated and vitrified hazardous wastes, and in
the subsequent data interpretation. Many of the tests
involve frequent handling of the waste materials; there-
fore, due consideration must be given to personnel pro-
tection, sample handling and disposal requirements, and
other factors associated with the presence of hazardous
constituents in the samples.
8.1.1 Index Property Tests
Index property tests provide data that are used to relate
general physical characteristics of a material to process
operational parameters. These tests are most frequently
performed on untreated waste to determine the feasibility
of vitrification.
One important index property for ISV is the moisture
content of the material to be vitrified. The Moisture
Content Test (ASTM D2216-80) determines the amount
of free water (orf luid) in agiven amount of material. In this
test method, the term "water" refers to "free" or "pore"
water, not waters of hydration. Alsowaterindiscontinuous
pores is not measured by this test. The results of this test
are usually expressed as fluid representing a percentage
of total mass. This test is often used to determine if pre-
treatment is necessary in the design of the vitrification
process. An example of waste pre-treatment would be
sludge drying, dewatering, or consolidation prior to vitrifi-
cation.
It is also important to note that water is often not the only
liquid-phase constituent in contaminated materials. The
8-1
-------
fluids may also include a broad range of liquid wastes
present in solution or as nonaqueous phase liquids. This
can have several effects on the performance and results
of moisture content determinations. For example, if
volatile organic compounds (VOCs) are present, samples
should be aerated to allow volatilization of flammable
VOCs before samples are oven-dried. (Of course, VOC
content itself may be an important parameter to measure
at some sites.) The type and level of contamination may
also influence the relationship between "free" and adsorbed
water.
In addition to the physical properties measured by the
index property tests described above, chemical compo-
sition, thermal conductivity, fusion temperature, specific
heat, electrical conductivity, and viscosity are all physical
properties of the material to be vitrified that may influence
process applicability and economics (Buelt et al., 1987).
8.1.2 Density Tests
Bulk density is the ratio of the total weight (solids^ and
water) to the total volume. Bulk density, along with
specific gravity and moisture content measurements, can
be used to calculate a material's porosity. More com-
monly, bulk density values are used to convert weight to
volume for materials-handling calculations and are es-
sential for characterizing the rates at which a soil can be
excavated. In addition, bulk density data provide a
comparison between untreated waste and the vitrified
product. !
Methods of bulk density measurement include) the
Drive-Cylinder Method (ASTMD2037-83), the Sand-Cone
Method (ASTM D1556-82), and the Nuclear Method
(ASTM D2922-81). The data from each are sufficiently
accurate for calculating densities. Selection of a method
is usually based on ease of use. Laboratory determina-
tion of specific gravity can supplement these measure-
ments. ;
8.1.3 Hydraulic Conductivity Tests
Hydraulic conductivity, often referred to as permeability,
Is a measure of the resistance of a material to the passage
of fluids. Permeabilitytestsareperformedtoestimatethe
quantity and flow rates of water through a material under
saturated conditions. Laboratory permeability testing
consists of applying a hydraulic head of water to onejend
of a specimen and measuring the flow through | the
specimen. >
There are two basic types of permeability tests:
constant-head and falling-head. The constant-head|test
allows relatively large quantities or water to flow through
the sample and be measured. This test is suitable for;
materials with a permeability greaterthan 10"6cm/s. The
falling-head test, which allows for more accurate mea-
surement of small quantities of water, is more suitable for
materials with a permeability of less than 10~6 cm/s
(Carter, 1983).
Sand, a highly permeable material, has a permeability on
the orderof 10'2 cm/s. Clay, a material that is used to line
lagoons and surface impoundments, can have perme-
ability on the order of 10'6 cm/s or less and is considered
relatively impermeable.
As described in Chapter Seven, the permeability of a soil
may bean important factor in estimating the effectiveness
of ISV for a particular site. At high moisture contents, ISV
will have to first boil off the water at the vitrified site before
a glass melt will be formed. Thus, at high moisture
content, if the permeability of the soil is too high (above
10"4 cm/s), water will move into the melt site faster than
it can be boiled off and the melt will not form (Buelt et al.,
1987).
8.1.4 Strength Tests
Strength-test values indicate how well a material will hold
up under mechanical stresses created by over-burden
and earth-moving equipment. It can also be used to give
some idea of how well a waste glass will resist fracturing
and thus increasing its surface area. This test, along with
the chemical leach test (see below), helps form an esti-
mate of the product durability.
A common strength-test is Unconfined Compressive
Strength of Cylindrical Cement Specimens (ASTM
D1633-84) However, several other strength tests may be
performed in addition to or in place of this test, depending
on the intended use of the data.
The EPA considers a solidified waste material with a
strength of 50 psi to have a satisfactory Unconfined
Compressive Strength (USEPA OSWER Directive, No.
9437.00-2A). This minimum guideline of 50 psi has been
suggested to provide a stable foundation for materials
placed upon it, including construction equipment and
impermeable caps and cover material.
The minimum required Unconfined Compressive Strength
for a treated material should be evaluated on the basis of
the design loads to which the material will be subjected.
The anticipated over-burden pressure and other loads,
along with appropriate safety factors, can be used to
calculate this.
Typical construction and compaction equipment can gen-
erate very high contact pressures of 1000 psi or more
8-2
-------
(e.g., sheepsfoot rollers), but surface contact pressures
on the order of 50 to 100 psi are more common. This
surface load is attenuated with depth so that bearing
pressures are reduced to values on the order of 10 to 20
psi at a depth of 2 feet and 3 to 7 psi at a depth of 5 feet
below grade. Overburden pressures will usually be on the
order of 0.75 to 1.0 psi perfoot of depth. If guidelines such
as these are used, the stresses to which the waste glass
will be subjected can be predicted, and design criteria can
be selected accordingly.
8.1.5 Durability Tests
Durability Testing evaluates the resistance of a waste
glass to degradation due to external environmental
stresses. The tests are designed to mimic natural con-
ditions by stressing the sample through: (1) freezing and
thawing; and (2) wetting and drying. The waste glass
specimens may undergo repeated cycling during the
testing. Unconfined Compressive Strength, flexural
strength, permeability, or other performance-based tests
may be conducted on the glass samples after each cycle
to determine how the physical properties of the waste
change as a result of simulated climatic stresses. The
number of cycles a material can withstand without failing
can be used to judge the mechanical integrity of the
material.
These tests relate to the long-term stability of the sample.
If the results show low loss of materials and retention of
physical integrity aftertesting, then the chemical compo-
sition of the vitrified product is adequate. If the test results
show a large loss of material and loss of physical integrity,
then various chemicals may have to be added to the feed
material to provide the long-term stability needed.
Poor durability results often can be addressed by a
change in design and should not be used as automatic
grounds for exclusion. For example, materials that fail
freeze-thaw durability testing can be placed below the
frost line to mitigate their poor durability property.
8.2 Chemical Tests
This section discusses leachingtests, the tests most often
used to evaluate the performance of vitrification as a
treatment process for hazardous waste.
In the field, leaching of hazardous constituents from waste
glass is a function of both the intrinsic properties of the
waste form and the hydrologic and geochemical proper-
ties of the site. Although laboratory physical and chemical
tests can be used to define the waste form's intrinsic
properties, the controlled conditions of the laboratory
environment are usually not equivalent to changing field
conditions. At best, laboratory leaching data can simulate
the behavior of waste forms under "ideal", static (condi-
tions atone point in time), or "worst-case" field conditions.
Presently, leach tests can be used to compare the ef-
fectiveness of various waste glasses, but they have not
been verified for determining the long-term teachability of
the waste.
8.2.1 Toxicity Characteristic Leaching Procedure
(TCLP) (Federal Register, 1986)
This test involves the definition of a toxicity characteristic
waste under the RCRA hazardous waste regulations.
The test is defined in 40 CFR 261 as follows. Waste
samples are prepared by crushing the wastes to pass
through a9.5-mm screen, and liquids are separated from
the solid phase by filtration through a 0.6 to 0.8 urn bo-
rosilicate glass-fiber filter under 50 psi pressure. Two
choices of buffered acidic leaching solutions are offered
underTCLP, depending on the alkalinity and the buffering
capacity of the wastes. Both are acetate buffer solutions.
Solution No. 1 has a pH of about 5; Solution No. 2 has a
pH of about 3. The leaching solution is added to a Zero
Headspace Extractor (ZHE) at a liquid:solid ratio of 20:1,
and the sample is agitated with a National Bureau of
Standards (NBS) rotary tumbler at 30 rpm for 18 hours.
The leaching solution is filtered, combined with the sepa-
rated liquid waste fraction, and analyzed for specific
organics and metals.
8.2.2 Materials Characterization Center Static
Leach Test (MCC-1P) (MCC, 1984)
This static leaching test was developed for HLW. It
involves leaching of a monolithic waste form with water
(ASTM Type I or II) at a volume of leaching solution to
surface area of solids (V/S) ratio of between 10 and 200
cm. The period and the temperature of extraction vary,
depending on the schedule selected. MCC-1P test results
can be combined with those from extraction tests (e.g.,
TCLP) to determine a range of leachate concentrations in
the short term (well-managed site with waste form intact)
and the long run (waste matrix has been subjected to
many years of environmental stress and is fractured).
8.2.3 Materials Characterization Center MCC-3 Test
The MCC-3 agitated powder leach test is very similar to
the MCC-1 test procedure with two exceptions: the glass
is in a powdered form and the glass powder and leachant
are agitated by rotating the container. This produces an
elemental leachate concentration that may be more rep-
resentative of dissolution under saturated conditions.
Leachate saturation is achieved more rapidly in the MCC-
3 test because higher surface area to volume ratios are
8-3
-------
used than in the MCC-1 test. The powder MCC-3 test is
also very useful in cases where multiple phases are
present in the waste form. Because the MCC-1 test uses
a cut monolith fortesting, results are often affected by the
representation of the different phases on the surface of
the monolith. The MCC-3 uses powdered samples thereby
allowing all phases to contact the leachate (Koegleretal.,
1989).
8.2.4 Product Consistency Test (PCT)
Th e PCT evo Ived from the M CC-3 test and was developed
for evaluating high-level vitrified waste forms from the
DWPF at SRS. The test can be performed remotely and
is reproducible. Leachate is monitored for metal concen-
tration and pH. The glass is crushed, sized, rinsedi and
submergedindeionizedwaterat90°Cfor7days(Jantzen
and Bibler, 1990). This test is being evaluated a|s an
ASTM standard test. !
i
8.2.5 American Nuclear Society Leach Test
(ANS-16.1,1986) (ANS, 1986)
A "quasi-dynamic" leach test, ANS-16.1, can be applied to
vitrified low-level and hazardous wastes. A monolithic
cylinder (length rdiameter of 0.2 to 5.0) is leached | with
demineralized water applied at a V/S ratio of 10 cm under
ambient temperatures. At the start of the experiment, the
sample is rinsed to obtain zero contaminant concentration
at the surface of the sample. Afterwards, the sample is
immersed in water, which is replaced after 2 houj-s, 7
hours, 24 hours, 48 hours, 72 hours, 4 days, 5 days, 14
days, 28 days, 43 days, and 90 days.
The results of the leaching test are recorded in terrps of
cumulative fraction leached over the total mass in the
waste form, against time. Calculations are then used to
derive an effective diffusion coefficient, De (cm2/s), and a
teachability index (LX = -log De). The LX values range
from 5 (De ^ 5-10, rapid diffusion) to 15 (De = 10-15, very
slow diffusion).
8.2.6 Leaching Test Selection and Interpretation
As mentioned in the preceding discussions, leaching
tests produce results that are not directly applicable to
leaching behaviorin the field. Nevertheless, the results of
several leaching tests or of leaching tests combined with
physical tests or microscopic techniques can be used as
indicators of field performance and environmental impact.
When used for comparative purposes, results from sev-
eral leaching tests can help identify field conditions 'that
may result in different concentrations of waste leaching.
Therefore, these data may be used to select or design
waste facilities that will minimize the leaching of hazard-
ous constituents from the wastes. The data also may be
used to predict the leaching of waste glass at different
stages in time. For example, a closed facility that has a
cover which is maintained (i.e., a 30 year post-closure
period) and minimizes precipitation infiltration, leaching
conditions may be similar to those of the MCC-1 P test
(i.e., static hydraulic conditions).
In the few cases where the actual field leaching solution
is well known, use of this solution in the laboratory tests
may yield more representative results. When the site
leaching solution is used, however, the results may be
relevant only to field leaching conditions in the short term
because the site hydrogeochemistry may change over
the long run.
8-4
-------
CHAPTER NINE
PROCESS EVALUATION
Technology screening is an important part of evaluating
any technology. In this chapter, examples of vitrification
comparison studies are presented, scaling-up problems
are discussed, and cost categories for in situ and furnace
vitrification are described.
9.1 Selection of Vitrification Processes
In this section, two studies evaluating vitrification pro-
cesses are described. These studies are presented as
examples of ways in which vitrification processes may be
evaluated. Because of the differing goals and identified
waste streams inthetwo studies, the questions asked and
the conclusions drawn differed. In addition, the pro-
cesses evaluated differed between the two studies.
Evaluation of vitrification processes should proceed with
site or waste specific goals in mind: the questions asked
will shape the conclusions drawn.
Bickfordetal. (1991) evaluatedfourvitrificationprocesses:
calcination followed by pot melting, a modified commer-
cial melter, a HLW joule-heated melter, and a stirred
melter. Each melterwas evaluated for process character-
istics in treating two groups of waste. The first step in this
study was to identify a series of desired process charac-
teristics (Table 9-1). (Please note that the evaluation for
only one of the waste groups is presented in Table 9-1.)
These characteristics were ranked on a scale of 0 to 10for
their relative importance in treating each of the two waste
groups. Then, each melter system was rated for its ability
to meet each characteristic on a scale of 1 to 10 and
assigned a decimal value accordingly (i.e., a "5" became
a "0.5"). The score of each melter system for each of the
desired characteristics was multiplied by the relative
importance value of that characteristic. These values
were summed to establish the total rating.
In the second study (Haz Answers, 1991), conducted for
INEL, a variety of thermal processes were evaluated for
Radioactive Waste Management Center (RWMC) waste.
A total of 27 technologies made up the preliminary list of
thermal technologies. Only those technologies which
applied to sludge, solid combustible, or solid inert wastes
passed the initial screening of the preliminary list and
were evaluated in detail. Each of the remaining 16
technologies were scored on the basis of 7 identified
evaluation criteria. These criteria were chosen as the
most important site-specific aspects to be considered in
making decisions concerning the treatment of RWMC
wastes. A relative index was applied to the evaluation
criteria in order to weight the evaluation criteria according
to perceived importance in treating RWMC wastes (from
a high of 0.25 for "Final Waste Form" and "Versatility" to
a low of 0.05 for "Cost"). Finally, each technology was
scored on a scale of 1 to 5 for each of the evaluation
criteria and then multiplied by the weighted factor of that
evaluation criteria. Scores were summed. The results of
this summary are presented in Table 9-2.
9.2 Initial Testing and Scaling-Up
The question of testing a process and then scaling up to
a full-scale operation is a common problem in dealing with
hazardous material. This section describes some of the
general issues in this procedure and address some
specific questions that may be important at the
bench-scale. Variations in site and waste characteristics
drive the development of objectives for a particular site or
waste, and these objectives may influence the nature of
the treatability study.
9.2.1 Treatability/Bench-Scale Testing
Treatability/bench-scale testing involves the performance
of various physical and chemical tests on actual contami-
nated materialsfrom the site,followed byengineering-scale
melt testing on the materials. Treatability testing is used
to:
Demonstrate that the technology is applicable to
the specific soil/waste combinations at the site.
Produce contaminant-related performance data
necessary to support permitting activities.
Develop design data necessary to support cost
estimates and quotes.
9-1
-------
Table 9-1. Determination of Preferred Melter System for Beta-Gamma, Low-Level Mixed, Inorganics (Heavy
Metals), Asbestos, Organics, and Soils Wastes (Bickford et ai., 1991)
Melter Type
Relative
Importance
Calciner/Pot
Welters
/^Modified ^
V^ Commercial^/
HLW Joule
Heated Melter
(stirred Melter)
Characteristics
Raw Materials Cost
6
3
6
3
5
Waste Loading
6
5
6
5
6
(indicates Preferred System)
2.
'5
c
a>
X
8
4
6
6
8
to
3
1
-------
Table 9-2. Criteria Raw Scores and Weighted Overall Scores for INEL Thermal
Process Evaluation Study (Haz Answers, 1991)
Technology Name
Slagging kiln
Fluidized bed
Multiple hearth
Rotary kiln
Controlled air
Cyclone
Low temperature
thermal separator
Infared furnace
Molten salt furnace
Plasma centrifugal reactor
Plasma arc furnace
Conventional temperature
pyrolysis
Microwave discharge
Molten glass furnace
In situ vitrification
Microwave melter
LEGEND:
LDA
5
5
5
5
5
5
2
4
3
2
2
4
3
4
3
2
VRSL THRU
5
3
1
5
1
1
2
2
1
4
4
1
2
4
4
1
LDA = Level of Development and Availability
VRSL = Versatility
THRU = Throughput
FWF = Final Waste Form
3
5
3
5
3
5
5
5
3
4
4
3
1
5
3
1
EFLT = Effluents
FWF
3
1
1
1
1
1
1
1
1
3
3
1
1
2
5
3
EFLT
3
2
3
1
5
5
5
5
5
3
3
4
5
5
4
4
COMP
2
3
5
4
5
5
4
3
2
1
2
4
5
3
5
5
COST
1
5
4
5
4
5
5
5
4
4
2
4
3
4
5
2
Overall
Weighted
Score
3.4
2.9
2.6
3.5
2.8
3.0
2.8
2.9
2.1
3.0
3.2
2.4
2.7
3.6
4.3
2.6
COMP = Complexity
COST = Cost
tive of pilot-scale tests is to confirm that bench-scale test
results will be applicable on a larger scale. This is
accomplished by testing a portion of the actual site or a
simulation of an actual site. The testing typically evalu-
ates process operations effectiveness, off-gas behavior
of volatile or entrained materials, potential costs, and
potential processing problems (Lominac, Edwards, and
Timmerman, 1989).
9.2.3 Scaling-Up Case Studies
The scaling-up process of Geosafe's ISV and Retech's
PCR (both described in Chapter Three) will be presented
in this section to give some idea of the steps involved in
this process.
9.2.3.1 Scaling up of ISV
The scaling up process of ISV used at PNL is described
in this section. Development and deployment of the
large-scale ISV system is the ultimate goal of the ISV
program, because it is less costly to operate, and it is more
adaptable to numerous types of waste sites than the
pilot-scale system. The cost of vitrifying a given waste
volume with the large-scale systemis lessthan that of the
pilot-scale system. The large-scale system is more
adaptable because of its high-capacity off-gas system,
which can process off-gas at a rate of 104 standard m3/
min, is better equipped to contain sudden gaseous re-
leases from combustible and othergas-generating wastes.
Nevertheless, the pilot-, engineering-, and bench-scale
systems provide important data that are used to deter-
mine the performance of the large-scale system at a
significantly reduced development cost. The scale of all
four development units for DOE is described in Table 9-3.
Descriptions of these development units follow in the text.
Bench-Scale System. The bench-scale system is used
primarily to verify ISV processability and off-gas charac-
teristics from alternative types of soil and waste inclu-
sions. The bench-scale unit has been used on many
9-3
-------
occasions for new soil and waste types as a precursor to
larger-scale testing or onsite demonstrations. It is rela-
tively inexpensive to operate, and it is instrumental in
determining the applicability of ISV to various waste
types,
Engineering-Scale Tests. One of the primary develop-
mental tools for ISV has been the engineering-scale
laboratory test, which is operated in the PNL developmen-
tal laboratory. The engineering-scale system has many
flexible design features for testing new concepts. Be-
cause of its smaller scale, the engineering-scale system
can test new concepts at a reduced cost while maintaining
a high level of confidence in its predictive capabilities for
larger-scale operations. Many of the analyses of] ISV
process limits have been based on tests conducted^ith
the engineering-scale unit (Buelt et al., 1987). i
Pilot-Scale Tests. The pilot-scale system is a portable
field system. The pilot-scale system is intermediate in
scale between the engineering-scale tests and the larger
scale tests and performs an important linking step itji the
scaling-up of ISV. The pilot-scale ISV system has trav-
elled to ORNL twice, INEL twice, and AEDC once for on-
site testing in contaminated areas.
Large-Scale Tests. Large-scale ISV tests are usdd to
verify that designs indicated at smaller-scale tests do
indeed work effectively with the large-scale system. De-
velopment of the large-scale ISV system has proceeded
in two steps: large-scale operational acceptance tests
and large-scale verification tests. The operational ac-
ceptance tests verified conformance of processing c|har-
acteristics to the established functional design criteria
relevant to the large-scale tests on actual waste streams.
The verification tests verified the effectiveness of process
modifications identified during the operational accep-
tance tests in readying the process for actual testing.
9.2.3.2 Scaling-up ofRetech's Plasma Centrifuge
Reactor (PCR)
Retech went through a similar scale-up process in the
development of its PCR (Eschenbach, Hill, and Sears,
1989). Theirscale-upprocesswentthroughthree phases.
Phase I (1985-86) consisted of a series of tests conducted
with a transferred-arc plasma on materials (metals, glass,
rubber, plastics, filter elements, etc.) typical of materials
which may get contaminated with radioactivity. These
tests demonstrated the feasibility of a transferred arc
plasma as a volume reduction process. As a result of
these tests, it was concluded that the addition of oxygen
or air as an oxidant in the plasma gas (argon in Phase I
tests) would be desirable in order to convert hydrocar-
bons to CO2 and water instead of soot, CO, and
hydrogen.
Phase II (1986-88) saw the development of the first PCR.
This quarter-scale reactor had a 0.46m (1.5 ft.) reactor
well and a 150 kw transferred-arc plasma torch operating
on air or an oxygen-argon mix. Tests were conducted on
dirt spiked with water and organics. These tests showed
that the product passed standard leach tests for the
non-volatile components retained in the glass. Difficulty
with pouring the glass indicated that the melt was too
viscous.
Phase III (1988-1989) focused on the operations of a
larger PCR. This PCR had a 1.8 m (6ft.) reactor well and
a 600 kw plasma torch. It was tested with a soil spiked with
15% oil. Tests with this PCR showed that air proved to be
the most satisfactory plasma gas: argon proved to be
unstable and the O2/argon mix corroded the electrodes.
Results indicated that DRE's of 99.99% to 99.999% were
obtained. These tests also confirmed that the changes in
the reactor well permitted pouring of the molten glass.
Waste feeder problems were also identified during these
tests.
Table 9-3. Testing Units For Developing ISV Technology
Equipment Size
Bench Scale
Engineering Scale
Pilot Scale
Large scale
i
Electrode
Separation
! 0.11 m
OJ23 to 0.36 m
1.2m
3.5 to 5.5 m
Glass
Block
Size
1 to 10 kg
0.05to1.0t
10 to 50 t
400 to 800 1
9-4
-------
By testing the PCR in increasing-scale tests, Retech was
able to establish that their system could potentially treat
heavy metals and organic waste with favorable results.
Furthermore, problems encountered with smaller-scale
systems enabled Retech to refine their process design
before encountering these problems in a large-scale
PCR.
The PCR is now being furthertested and developed at the
Component Development and Integration Facility (which
is underthe administration of INEL) for testing in the EPA
Superfund Innovative Technology Evaluation (SITE) pro-
gram (Viall, Sears, and Eschenbach, 1990). A larger
PCR is located in Basel, Switzerland (Schlienger and
Eschenbach, 1991).
9.3 Cost
In addition to technical feasibility and questions of scaling
up, costs are also an important component in the evalu-
ation of the applicability of any remediation process. The
following section is intended to identify key cost variables
of vitrification processes—such as site preparation, mo-
bilization/demobilization, energy costs, etc.—and to
summarize cost information developed to date. The
reader is cautioned that the cost information is presented
for summary and evaluation purposes only, and should
not be used for feasibility study cost estimates nor for
comparative purposes. Furthermore, comparisons among
various cost studies may be misleading because of
variations in:
cost variables included in cost estimates (such
as overhead or profit for commercial vendors,
etc.)
values assumed in estimating cost variables
(variation in location and extent of contamination,
site clean-up objectives, etc.)
type of vitrification process evaluated (in situ vs.
ex situ vs. ex situ process type)
• waste type (radioactive vs. hazardous waste)
Cost estimates also can vary with time, and cost estimates
made for one technology in one year may not be com-
parable with cost estimates made fo,r another technology
in another year. Furthermore, costs estimates for the
same technology may have been developed in different
ways by different researchers. Forexample, in estimating
costs for the furnace melter vitrification of contaminated
soils, researchers may or may not have included the cost
of excavation of the soils in their estimate. Finally, as
vitrification is in its early development stages as a waste
treatment, costs are not established on actual projects
and are often estimates.
In an attempt to clarify the major sources of costs, this
section will discuss ISV and furnace melters separately.
Under each discussion, categories of cost and estimated
costs will be described. In addition, the furnace melter
discussion will include a description of several methods
considered at a DOE site to reduce overall treatment
costs. The intent of this section is that the reader gain an
understanding of cost categories which are significant
areas of concern in managing costs, and a general idea
of actual vitrification costs.
9.3.1 ISV Costs
This section describes cost categories for generic ISV
application.
9.3.1.1 Cost Categories for ISV
The main costs for ISV vary depending upon electrical
costs fora particular geographic region and soil moisture
content. The moisture within the soil must be driven from
the melt zone before vitrification can begin. ISV costs can
be categorized into five subsections (Liikala, 1991):
1. Site Activities
2. Equipment
3. Operations
4. Expendables
5. Electrical Power
Cost items that fall within these categories are identified
in Table 9-4. Categories are briefly summarized below.
In addition to the costs identified above, another area of
cost is treatability testing. Treatability testing includes the
performance of various physical and chemical tests on
actual contaminated materials from the site and
engineering-scale ISV melt testing on the materials. The
cost of treatability testing is in the range of $40,000 to
$70,000 or more, depending on application. Unusual
analytical requirements, such as those posed by dioxin
analyses, may increase the costs (Timmons, FitzPatrick,
and Liikala, 1990).
The cost of equipment mobilization and demobilization
depends on transport distance to and from the site. The
combined total of these costs may be estimated at $50,000
plus $50 per transport mile. Typical total mobilization/
demobilization costs fall in the range of greater than
$100,000 to as much as $200,000 (Timmons, FitzPatrick,
andLiikaia, 1990).
Finally, the reader should be aware there is profit involved
when buying commercial services.
9-5
-------
Table 9-4. Major Components of ISV Costs (adapted from Buelt et al., 1987)
SITE ACTIVITIES
Transporting equipment to and from site
Clearing vegetation
Rough grading
Removing overburden
Acquiring and applying backfill material
EQUIPMENT
Power
Portable generator
Powar lines
Substation
Power cables
Mechanical
Electrode frame and hood
Drilling machinery
Crane
Front-end loader
Off-gas and monitoring
Off-gas treatment system
Radiation and off-gas monitors/alarms
OPERATIONS
Process preparations
Drill holes and place electrodes
Spread graphite starter material
Position frame and hood, secure electrodes
Connect power cables and off-gas line
Vitrify
Disconnect power cables and off-gas line
Hood fixation
Remove frame and hood
Backfill vitrified area
Move power cables for next setting
Process operations
Off-gas treatment system
Power system
Radiation or toxic chemical monitoring
Melt verification
Off-gas secondary waste disposal
EXPENDABLES
Electrodes
Secondary Wastes
ELECTRICAL POWER
Site Activities. Activities included in site preparation
include soil staging (if necessary), electrode placement,
set-up of ISV process trailers, electrical connections, and
subsidence backfilling. Site activities include transport-
ing equipment to and from the site, clearing vegetation,
grading the ground, removing overburden, and acquiring
and applying backfill material as needed' If
uncontaminated overburden could be removed safely, it
would always be advantageous to do so from a cost
standpoint. For example, removal of the top meter of
clean soil from a 2700-m2 site would cost less than
$10,000, compared to the hundreds of thousands of
dollars needed for labor and power charges to vitrify the
same area to a 1-m depth. In short, site activity costs will
be insignificant when compared to equipment, labor, and
electrical power, for the majority of potential ISV applica-
tions (Buelt etal., 1987). ;
Equipment. The ISV process trailers are the major
equipment required on-site. The only additional equip-
ment required are diesel generators—if high-voltage line
power is not available, a crane, and a front end loader or
dozer.
Electrical equipment requirements are determined by the
voltage and current demands of ISV: the high voltage at
the beginning of processing requires sufficient insulation
while the high current at the end of processing requires
sufficient conduction capacity. For example, the high
level of current (4000 A) requires that six 750-rhcm power
cables be used for each of the four electrodes (Buelt et al
1987).
Two pieces of heavy equipment are necessary for ISV
operations: a crane for transporting the electrode frame
and hood from one setting to the next and a front-end
loader for backfilling and site preparation. Purchase of
this equipment may be more cost effective than rental
because of the higher cost of renting these types of
equipment (typically several hundred dollars per day) for
the duration of projects that Iastfrom9 months to 10years
(Buelt et al., 1987). If the EFS is not used at a particular
site, then a drilling or auguring machine for placing the
electrodes in the ground may also have to be purchased.
Operations. The time required for each setting of the
electrode frame and hood is the sum of the time required
to vitrify the soil to the predetermined depth plus the time
required to move the off-gas equipment to the next
9-6
-------
setting. The ISV processing rates are generally 4 to 5
tons/hour. Typically, less than 24 hours are required to
restage the hood and the ISV trailers between subse-
quent melts. Calculated time per setting forthe large-scale
systems as a function of moisture content is shown in
Table 9-5. The effects of moisture content on vitrification
rate and operating time are evident from these vitrification
times (Buelt et al., 1987).
Total project time is equal to the time per setting multiplied
by the number of settings. The number of settings
Table 9-5. Time Requirements for Each ISV Setting
(Buelt etal., 1987)
Vitrification
Moving Equipment
Total
Large Scale,
5-m Depth,
5% Moisture,
h/setting
90
16
106
Large Scale,
5-m Depth,
25% Moisture,
h/setting
117
16
133
Table 9-6. ISV Electrode Spacing and Vitrification
Settings (Buelt et al., 1987)
Parameter
Large Scale,
5-m Depth
Electrode spacing 4.5m
Separation between electrodes
of adjacent set 3.0m
Width vitrified per set 7.8m
Area to be vitrified 90m x 30m
Set matrix (rows x colulmns) 4x12
Number of settings 48
depends on the dimensions of the site to be vitrified and
the area vitrified per setting. This latter characteristic is a
function of electrode spacing and acceptable allowances
for overlap between vitrified blocks. For example, given
the parameters estimated in Table 9-6, a contaminated
area 90m x 30m is estimated to require 48 separate ISV
settings.
Personnel and their estimated hours for large-scale pro-
cess preparation are identified in Table 9-7. These
personnel are required at scheduled intervals: once per
setting of the off-gas containment hood.
Personnel and their estimated hours for large-scale pro-
cess operations are identified in Table 9-8. A typical ISV
melt requires two operators per shift: a shift engineer and
an ISV technician dedicated to operating the system while
power is supplied to the electrodes. An engineer is
included on day shift as an operator (see Table 9-8) and
to provide technical resolution of any operational prob-
lems. Maintenance and radiation monitoring personnel
(required during the vitrification of radioactive wastes) are
included in the operations on an estimated part-time
basis.
Expendables. Electrodes have been considered one of
the major expenses of ISV because the cost of the
molybdenum used in the electrodes is around $20 per
Table 9-7. Manpower Requirements for ISV
Process Preparation (Buelt et al., 1987)
Job
Classification
Manpower Rate,
Average Man-Hours/Setting
Electrician
Laborer
Operator
4
34
19
Table 9-8. Labor Estimate for ISV Processing Operations at a Radioactive Site (Buelt et al., 1987)
Workers Per Shift
Job Classification
Engineer
Maintenance
Operator (Technician)
Radiation monitor^3)
Total
(a) Radiation monitoring
Day
1
0.5
1
0.25
personnel would
Swing Graveyard
0 0
0 0
2 2
0.25 0.25
not be required for a hazardous
Total
Man-Hours/Day
8
4
40
6
58
waste site.
9-7
-------
pound. After a single melt, the electrodes are not reusable
due to the large crystalline growth of the molybdenum
(Liikala, 1991). However, the development of the EPS
permits the use of all-graphite electrodes instead of the
molybdenum/graphite electrodes previously used. Thus,
electrode costs will be substantially reduced if the EPS is
used. !
In addition to electrode costs, a cost must be included for
disposing of the secondary liquid wastes that are col-
lected in the off-gas system. Approximately 200 li (530
gal) per large-scale setting must be disposed of at a cost
of $0.26/L ($1.00/gal). Forthe site configuration giyen in
Table 9-9 this results in additional charges of $25,000
($1.85/m3) (Buelt et al., 1987).
Electrical Power. Electrical power requirements ^re a
significant portion of the operating costs. Energy costs for
high voltage line power varies greatly with location, rang-
ing from as low as 2.5 cents per kwhr on the west co'ast to
8 cents per kwhr in the Midwest and east coast. A diesel
generator may conceivably be used in locations inacces-
sible to power lines or where electrical power is prohibi-
tive. Cost for use of diesel generators is equivalent to
about 8.25 to 13 cents per kwhr. '
The power requirements and estimated costs for the ISV
vitrification of low (5%) and high (25%) moisture cpntent
materials are given in Table 9-9 fora specific hypothetical
application. Also shown is the annual vitrification rate
which is based on an 80% operating capacity of the ISV
model prediction (Buelt et al., 1987).
9.3.1.2 Estimated Cost for ISV
As indicated in Table 9-10, the on-site service cost of ISV
processing may range from $96 to $390 per ton of
material processed for the references cited. These esti-
mates should not be considered as firm estimates ap-
propriate for all sites and all applications. Rather, they
serve as rough ISV cost estimates. Site characteristics
and clean-up goals will play important roles in modifying
these estimates. Furthermore, the parameters consid-
ered have not always been identified, or they may differ
from study to study. For example, Buelt et al. (1987)
included elements of direct and indirect cost, such as
labor, materials, energy, equipment amortization, and
contractor overhead and profit, but ignored treatability
costs. Carpenter and Wilson (1988) calculated their
estimates from the following formula:
Cost, $/ton = ($13 dredging) + ($8-80 transportation)
+ ($96-210 treatment) + ($46 redeposition)
= $163-349
Thus, great care should be used when examining these
cost estimates.
Table 9-9. Power Requirements for ISV Rate as a Function of
Moisture Content (Buelt et al., 1987)
Moisture Content
5%
25%
Energy
Requirement
kWh/setting
302,000
392,000
Annual
Vitrification
Rate, m3/yr
15,300
12,200
Cost/Setting
(@ $0.05/kWh)
$15,100
$19,600
Table 9-10. Sample ISV Cost Estimates ($/ton)
Year Cost range ($/ton)
1985 $117-165a '
1986 $ 96-21 Oa
1988 $163-349a
1989 $166-175a
1990 $103-382a
1991 $360-390 ,
Calculated from reported figures assuming 11.2 tons/yd3.
Reference
(Buelt etal., 1987)
(USEPA.1986)
(Carpenter and Wilson, 1988)
(Koegleretal., 1989)
(USEPA, 1990c)
(Landau Associates, 1991)
9-8
-------
Table 9-11. ISV Equipment Costs (Koegler et al., 1989)
Estimated Costs
Equipment ($1000)
Engineering and Design
Equipment Mobilization (6 systems)
Transformers (6 required)
Off-Gas Hood and Line (6 required)
Off-Gas System (3 required)
Backup Blower System (3 required)
Power Lines (6 systems)
Electrode Power Cables (6 systems)
Portable Generators (3 systems)
Equipment Demobilization (6 systems)
Electrode Placement Machinery (1 system)
Crane (1)
Front End Loader (1)
Total Equipment Costs
500
540
1,500
3,600
9,000
600
120
240
300
780
120
130
80
17,500
Percentage
of Total Cost
3
3
9
21
51
3
<1
1
2
4
<1
<1
<1
Weidon Spring Site. If cost estimates are examined on a
percentage basis, those factors contributing most greatly
to costs can be identified. Costs were carefully broken
down in evaluating the treatability of ISV to the Weldon
Spring site in Missouri (Koegler, Oma, and Perez, 1988;
Koegler et al., 1989). Examination of this data permits a
more detailed discussion of the relative importance of ISV
cost categories.
The Weldon Spring site comprises a 9-acre former lime-
stone quarry, a 52-acre disposal area for raffinate waste
(the less soluble residue remaining afterchemical extrac-
tion), and a 169-acre mothbalied uranium-feed materials
plant. The quarry, about 4 miles south of the main site,
contains an estimated 95,000 cubic yards of rubble and
soil contaminated with trinitrotoluene (TNT), dinitrotoluene
(DNT), uranium, thorium, and their decay products. The
waste material is piled 40 feet above the floor of the
quarry, and most of the waste is covered by several feet
of soil. Vegetation covers the quarry surface and the
lowest area is covered by water. Where a cross section
is visible, a large amount of metal (e.g., crushed drums,
sheet metal, structural building iron, and process equip-
ment) protrudes from the soil. Large pieces of equipment
such as tanks, a fork-lift truck, and up to 3000 drums are
also buried, although ground-penetrating radar or similar
techniques have not been used to locate these large
items. The water table is about 15 feet above the floor of
the quarry, and the standing water level is about 6 feet
above the water table (Koegler, Omar, and Perez, 1988).
Table 9-11 gives an itemized capital cost breakdown of
the site equipment estimated to be required forthe ISV of
the Weldon Springs site. The equipment listed includes
six electrical transformers and six off-gas hoods, with
three off-gas treatment systems and three backup blower
systems. Each off-gas system and backup blower syste m
would treat the off-gas from two ISV operations, thus
reducing capital costs. The equipment costs include the
costs for engineering and designing the equipment and
mobilizing and demobilizing it atthe site. Equipment costs
for waste excavation and transport are not included in this
itemized list, however. In addition, the equipment costs
for filtration of the Weldon Spring sludge have not been
estimated.
Site operating costs are listed in Table 9-12. Energy
consumption is clearly the single largest item contributing
to treatment costs. If equipment costs (from Table 9-11)
are added to treatment costs, energy consumption still
accounts for 49% of total costs. The costs forthe 3 off-gas
systems, on the other hand, drops to only 9% of total
costs. Clearly, any effects to reduce treatment costs at
Welden Spring should be targeted primarily at reducing
energy consumption.
Actual methods considered to reduce ISV costs at Welden
Springs is included the following:
1. Use three off-gas systems instead of six. This
would net a $9 M savings.
9-9
-------
2. Dewater the raffinate sludge from 24 wt% solids
to 35 wt% solids prior to ISV. By reducing the
energy required to vaporize the excess water, as
much as $10 M may be saved. :
3. Combine contaminated materials prior to treat-
ment. Because the sludge required the addition
of soil or clay prior to vitrification, contaminated
soil and contaminated clay-liner could be substi-
tuted for clean material netting a calculated net
savings of $8.8M. i
9.3.2 Estimation of Melter Vitrification Costs
The JHCM was also evaluated as a possible remediation
process for the Weldon Spring site. In this section cost
estimates from this evaluation will be presented. Icost
estimates are broken down by capital equipment costs,
capital costs, and operating costs. These are described
sequentially.
Capital Equipment Costs. The equipment list in Table
9-13 is complete for preparing and vitrifying the wastes.
Included in the list are equipment needed for size reduc-
tion and blending, vitrification, glass product handling,
and off-gas treatment. Not included are equipment asso-
ciated with excavation of raffinate pit materials. The
JHCM represents over 82% of equipment costs by itself.
Capital Costs. Capital costs include capital equipment
costs as well as costs in support of capital equipment, etc.
These are itemized in Table 9-14. The costs assum^ that
the facility can be built using standard practices for
chemical plant structures with additional requiremenjts for
ventilation, filtration, and monitoring equipment given that
the site contains low-level radioactive wastes. Table 9-14
shows that the melter is the most expensive single capital
Table 9-12. ISV Site Operating Costs
(Koegleretal., 1989)
Cost Breakdown
Labor Costs
Vitrification Crew
Heavy Equipment
Total Labor
Consumable Costs
Electrodes
Energy
Secondary Waste
Total Consumables
Total Operating Costs
Cost, $1000
(% operational cost)
$ 7,380
1,480
8,860
21,900
46,100
439
68,500
$77,400
(10%)
(2%)
(28%)
60%)
(0.5%)
expenditure, but that purchased-equipment installation
and building and facilities also contribute greater than
10% of capital costs.
Operating Costs. Costs associated with the operation of
the vitrification facility include the cost of bulk chemicals
added to the feed to improve product quality (borax and
soda ash), utilities, and labor. The labor costs include
operating personnel for three shifts per day, a plant
manager, maintenance personnel, clerical staff, and
overheads. It was assumed for this evaluation that the
vitrification facility would operate seven days per week,
365 days per year, with an on-line efficiency of 80%.
Costs associated with the start-up of the facility were not
included but assumed to be insignificant. Electrical costs
included as part of the utilities costs are based on an
electricity rate of $.06/kWh. Treatment would be com-
pleted in about four years at a total operating cost of
$60M.
Given these operating parameters the total operating
costs break down as follows:
Labor cost
Cost of chemical additives
Utilities cost
$4.2 M (7% of total costs)
$21.7 M (36%)
$34.4 M (57%)
Summary. If equipment costs, costs in support of capital
equipment and operating costs are combined. The total
remediation cost becomes almost $77M. This breaks
down as presented in Table 9-15.
From these comparisons, it can be seen that utilities still
are the primary remediation cost, followed by the cost of
the chemical additives. Melter costs, the single item
dominating capital costs, is only 5.8% ($4.5M) of the total
remediation costs.
Costs Reduction. In the process of cost analysis, certain
decisions were made to reduce the cost of remediation.
Several answers were pursued in these reductions. Ex-
amination of these will give some idea of similar reduc-
tions that can be made at other sites. Specifically the
costs estimated included the following actions:
1. Selection of a fluxing agent based on a compromise
between desired processing characteristics and costs.
Li2O or B2O3 were mixed with NaaO and these mix-
tures were compared as potential fluxing agents. It
was estimated that use of the LJ2O mixture would
reduce the total quantity of glass produced and the
totalprocessingtimeforthesite. However, LiaOproved
to be a more expensive additive (by a 350% increase
in additives cost) than B^Os and these costs out-
weighed the other savings. Therefore, the best
9-10
-------
Table 9-13. Equipment Required for JHCM Processing (Koegler et al., 1989)
Equipment
Bulk Materials Handling
Raffinate Sludge Transfer Pump
Crush, Delump Unit
Clay Liner/Vicinity Soil Transfer System
Chemical Additive Unloading Station
Chemical Additive Transfer System
Mechanical Mixer
Melter Feed Transfer System
Melter feed Storage Silo
Melter Feed Transfer System
Dust Abatement System
Melter Feed System
Melter Feed Storage Hopper
Rotary Valve
Joule-Heated Ceramic Melter
Melter
Off-Gas Treatment System
Quench Scrubber
Scrub Solution Recycle System
Roughing Filter
Heat Exchanger
HEPA Filter
Concentrator
Blower
Glass Handling System
Glass Quencher
Heat Exchanger
Fritted Glass Transfer System
Total Equipment Cost:
"Value is total cost of JHCM system including transformers
Quantity
2
1
1
1
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Total Cost, $1000
30
50
25
50
75
50
25
20
25
150
25
25
4500"
50
35
25
25
15
50
150
50
25
25
$5,500
3.
fluxing agent proved to be a mixture of NaaO and BaOs
in a ratio of 7:1.
Blending of waste streams. The waste material to be
treated at Weldon Springs consisted of three matri-
ces: sludge, soil, and linerf rom the sludge pits. Alone,
each matrix had chemical composition defects which
would require the addition of additives to create a
durable glass. However, if mixed prior to vitrification,
the blending would remediate some of the deficien-
cies of each matrix. Additives required and costs
would therefore be reduced.
Dewatering the sludge. As described in Chapter
Seven, dewatering would increase the efficiency of
the melter and could prove cost effective by reducing
total treatment time.
5.
Use of a high-temperature melter. Waste matrices at
the Weldon Springs site require a temperature of
1450° C to melt. By permitting a higher operating
temperature, a high-temperature melter reduces the
need forfluxes. This reduces additive costs and may
make the increased cost of a high-temperature melter
pay for itself.
Fritting the waste glass instead of pouring it into
canisters. This option reduces the complexity of
waste material handling after vitrification and thus
reduces costs. This option is dependent on the
quality of the fritted glass product.
9-11
-------
Table 9-14. Capital Cost Summary for JHCM
(Koegler et a!., 1989)
Costs Total
Capital Equipment Costs
Bulk Materials Handling Equipment
Melter Feed System
Joule-Heated Ceramic Melter
Off-Gas Treatment System
Glass Handling System
Total Equipment Costs
Purchased-Equipment Installation
Instrumentation & Control
Process Piping
Electrical
Auxiliaries
Building & Facilities
Site Preparation
Contingency
Fees and Engineering Contingency
Total Costs In Support of Equipment
Total Capital Costs
($1000) Cost
500
50 <
3%
:1%
4500 28%
350
100 <
2%
=1%
$5,500 34%
1
1
1
2
1
,800 11%
850
650
,100
,100
5%
4%
7%
7%
,200 14%
550
,100
1,400
3%
7%
9%
$10,750 66%
$16,250 100%
Table 9-15. Comparison of Capital Costs and
Operating Costs for a JHCM (Koegler et al., 1989)
Capital Costs $16.3M (21 %)
Equipment costs $5.5M (7%)
Costs In support of capital equipment $10.8M (14%)
Operating Costs $60.3M (79%)
Labor costs $4.2M (5%)
Chemical additives cost $21.7M (28%)
Utilities cost $34.4M (45%)
Total Cost $76.6M
9.3.3 Additional Cost Factors
Two additional factors play an important role in remediation
costs: throughput rate and energy costs.
Throughput rate is the amount of material that can be
processed per unit time. High throughput rates generally
decrease costs because of economy of scale. Because of
the reduction in volume during vitrification, throughput is
often expressed both in terms of feed material treated per
unit time and glass produced per unit time. Selected
throughputs are presented in Table 9-16. The values
presented represent process results under a variety of
conditions and do not necessarily represent maximum
throughput or expected throughput. For example, when
using a glass melter to vitrify, process rate can be ad-
justed by varying the size of the melter. For solutions and
concentrated slurries, the process rate is between 36 and
85 gallons/hour/square foot. For contaminated soils and
other inorganic feeds, the process rate ranges from 400-
600 pounds/day/square foot. Obviously, for increased
process rates the melter must be increased in size. In
addition to melter size, processing rate will be affected by
water content, inherent energy content, particle size, etc,
Energy demands for vitrifying a waste (kwh/ ton of waste
or soil) will also vary with a variety of factors, but will
depend primarily upon water content and exothermic:
energy present in the feed.
Table 9-16. Throughput Rates for Selected Vitrification Processes
Melter Type
LFCM- West Valley
AVM - France
Glass Melter - Penberthy
Glass Melter - Penberthy
Glass Melter - Penberthy
coal-fired melter - Vortec
Feed Type
HLLyv
calcined; HLW
toluene, oil
spent resins
wood, cloth, paper
glass-making
glass melter - Vitrifix asbestos-contaminant ed soil
rotary kiln - MSP
ISV
'Glass production is greater than feed
these types of contaminants.
incinerator ash, soil
soil
input due to the need to add
Feed Input Rate
150L/h
60L/h
125-1 000 Ib/h*
250-1 000 Ib/h*
400-4000 Ib/h*
20 tons/day
5 tons/day
1 00 tons/day
--
substantial glass forming
Glass Production Hate
45 kg/h
25 kg/h
500-4000 Ib/h*
500-4000 Ib/h*
500-4000 Ib/h*
NA
NA
NA
3.5to4tph
materials to
9-12
-------
REFERENCES
Armstrong, K.M. and L.M. Klingler. 1985. Evaluation of
a Unique System for the Thermal Processing of
Radioactive and Mixed Wastes. MLM-3340, EG&G
Mound Applied Technologies, Miamisburg, Ohio.
Associated Press. 1991. Device Blasts Asbestos to
Rubble. The Atlanta Journal/The Atlanta Constitution,
Sept. 1, 1991, p. H4.
Baehr, W. 1989. Industrial Vitrification Processes for
High-Level Liquid Waste Solutions. IAES Bulletin, 4:
43-46.
Battey, R.F., and J.T. Harrsen. 1987. In Situ Vitrification
for Decontamination of Soils Containing PCBs. in
ST. Kolaczkowski and B.D. Crittendon, eds. Pro-
ceed. Manage. Hazard. Toxic Process Ind. [Int.
Congr.], Elsevier Appl. Sc., London, U.K. pp. 263-
272.
Berkey, E. 1989. Vitrification of Asbestos Waste. Haz-
ardous Materials Management Magazine, Nov/Dec,
1989, pp. 42-43.
Bickford, D.F., P. Hrma, andB.W. Bowan. 1990. Control
of Radioactive Waste Glass Melters: II, Residence
Time and Melt Rate Limitations. J. Am. Ceram. Soc.,
73(10): 2903-15.
Bickford, D.F., R.C. Propst, and MJ. Plodinec. 1988.
Control of Radioactive Waste-Glass Melters: Part 3 -
Glass Electrical Stability, in D.F. Bickford et al. (eds)
Advances in the Fusion of Glass. The American
Ceramic Society, Inc., Westerville, Ohio.
Bickford, D.F., M.E. Smith, P.M. Allen, J.P. Faraci, C.A.
Langston, and K.Z. Wolf. 1991. Application of High
Level Waste - Glass Technology to the Volume Re-
duction and Immobilization of TRU, Low Level, and
Mixed Wastes(u). inR.G. Post (editor), Proceedings
of Waste Management '91. American Nuclear Soci-
ety (ANS), Tucson, AZ.
Bjorklund, W.J., G.B. Mellinger, and J.M. Pope. 1984.
Vitrification of Simulated West Valley Waste Using a
Liquid-fed Ceramic Melter. in Wicks, G.G., W.A. Ross,
eds. Advances in Ceramics. Volume 8: Nuclear Waste
Management. The American Ceramic Society, Inc.,
Columbus, Ohio.
Bonniaud, R., R. Demay, R. Richter, and L. Roznad.
1986. Glass Melter Materials: Technical Options for
the French Vitrification Process and Operations Ex-
perience. Presented at the Technical Committee
Meetingon Materials Reliability inthe Back-end of the
Nuclear Fuel Cycle, Vienna, Austria, September 2-5,
1986. CEA-CONF-8632, Commissariat a L'Energie
Atomique, France.
Bromberg, L., D.R. Cohn, C.H. Titus, and U.K. Wittle.
1991. Waste Detoxification Using Transfer Arc and
Electron-Beam Driven Plasma Reactors. Presented
at Workshop on Plasma Applications to Waste
Treatment, Idaho National Engineering Laboratory,
Idaho Falls, Idaho, January 16, 1991.
Bromberg, L., D.R. Cohn, C.H. Titus, and U.K. Wittle.
1991. Waste Detoxification Using Transfer Arc and
Electron-Beam Driven Plasma Reactors. Workshop
on Plasma Applications to Waste Treatment, Idaho
National Engineering Laboratory, Idaho Falls, Idaho,
January 16,1991.
Brunner.C.R. 1984. Incineration Systems: Selection and
Design. Van Nostrand Reinhold Company, New
York, New York.
BueltandWestsik. 1988. In Situ Vitrification-Preliminary
Results From the First Large-Scale Radioactive Test.
PNL-SA-15277, Pacific Northwest Laboratory,
Richland, Washington.
Buelt, J.L., and R.K. Farnsworth. 1990. In Situ Vitrifica-
tion of Soils Containing Various Metals. PNL-SA-
17242, Pacific Northwest Laboratory, Richland,
Washington.
R-1
-------
Buelt, J.L., C.L. Timmerman, and J.H. Westsik, JrJ 1989.
In Situ Vitrification: Test Results for a Contaminated
Soil-melting Process. PNL-SA-15767, Supplement
1, Pacific Northwest Laboratory, Richland, Washing-
ton.
Bueit, J.L, C.L. Timmerman, K.H. Oma, V.F. FitzPatrick,
and J.G. Carter. 1987. In Situ Vitrification of Tran-
suranic Waste: An Updated Systems Evaluation and
Applications Assessment. PNL-4800 Suppl.[l, Pa-
cific Northwest Laboratory, Richland, Washington,
March, 1987. !
Callow,R.A.,L.E.Thompson,andJ.R.Weidner. 1991. In
Situ Vitrification Application to Buried Waste: interim
Report of Intermediate Field Tests at Idaho National
Engineering Laboratory. EGG-WTD-9422, JEG&G
Idaho, Inc., Idaho Falls, Idaho. ;
Campbell, B.E. and J.L Buelt. 1990. In Situ Vitrification
of Soil from the Savannah River Site. PHli-7421,
Pacific Northwest Laboratory, Richland, Washington,
August, 1990.
Campbell, B.E., C.L. Timmerman, and W.F. Bonner.
1990. Underground Tank Vitrification:
Engineering-Scale Test Results. PNL-7295, Pacific
Northwest Laboratory, Richland, Washington.
Carpenter, B., and D.Wilson. 1988. Technical/Economic
Assessment of Selected PCB Decontamination Pro-
cesses. Journal of Hazardous Materials 17:125-148.
Carter, G.W., A.V. Tsangaris, C. Simpson, and N.C.
Burnett. 1990. Municipal Solid Waste Feasibility of
Gasification with Plasma Arc. EPRI Symposium,
Industrial Application of Plasma, Palo Alto, California,
March 7-9,1990.
Carter, M. 1983. Geotechnical Engineering Handbook.
Chapman and Hill, New York, New York. '
Chapman, C.C. 1984. Design Preferences for a Slurry-
fed Ceramic Melter Suitable for Vitrifying West; Valley
Wastes, in G.G. Wicks and W.A. Ross, eds. Ad-
vances in Ceramics, Vol. 8: Nuclear Waste Manage-
ment. The American Ceramic Society, Columbus,
Ohio.
Chapman, C.C. 1991. Evaluation of Vitrifying Municipal
Incinerator Ash. Presented at the Fifth International
Symposium on Ceramics in Nuclear and Hazardous
Waste Management, American Ceramic Society,
Cincinnati, Ohio, April 29-May 3, 1991. Ph|L-SA-
18990, Pacific Northwest Laboratory, Richland,
Washington.
Chapman, C.C. and J.L. McEilroy. 1989. Slurry-Fed
Ceramic Melter — A Broadly Accepted System to
Vitrify High-Level Waste, in S.C. Slate, R. Kohout,
and A. Suzuki. High Level Radioactive Waste and
Spent Fuel Management. Book No. 10292B, The
American Society of Mechanical'Engineers.
Clery, D. 1990. Where There's Muck, There's Glass.
New Scientist, June 16,1990, p. 34.
Darnell, G.R. 1990. Intensive Volume Reduction of Mixed
and Low-level Waste, Solidification in Sulphur Poly-
mer Concrete, and Excellent Disposal at Minimum
Cost. U.S. Department of Energy, Proceedings of
DOE Model Conference on Waste Management and
Environmental Restoration, October 29 - November
2,1990, Oak Ridge, Tennessee.
Denner, J.M., R.E. Langridge, and M.J. Affleck. 1988.
Development of an Asbestos-contaminated Site - the
Faslane Project. Journal of the Institute of Water and
Environment Management, June 2, 1988. pp. 300-
304.
Dragun, J. 1991. Geochemistry and Soil Chemistry
Reactions Occurring During In Situ Vitrification. Jour-
nal of Hazardous Materials 26: 343-364,1991.
Eisenstatt, L.R. and C.C. Chapman. 1986. The West
Valley Vitrification Process Model: A Method for
Providing Glass Composition, in D.E. Clark, W.B.
White, and A.J. Machiels, eds. Advances in Ceram-
ics, Vol. 20: Nuclear Waste Management II. The
American Ceramic Society, Westerville, Ohio.
Eschenbach, R.C., R.A. Hill, and J.W. Sears. 1989.
Process Description and Initial Test Results with the
Plasma Centrifugal Reactor. In: Forum on Innovative
Waste Treatment Technologies: Domestic and Inter-
national, Atlanta, GA, June 19-21, 1989, EPA-540/
2-89-056, U.S. Environmental Protection Agency,
Office of Research and Development, Cincinnati,
Ohio, 1989. pp. 263-279.
Farnsworth, R.K., K.H. Oma, and C.E. Bigelow. 1990.
Initial Tests on In Situ Vitrification Using Electrode
Feeding Technique. PNL-7355, Pacific Northwest
Laboratory, Richland, Washington.
Farnsworth, R.K., K.H. Oma, and M.A.H. Reimus. 1990.
Crucible Melts and Bench-Scale ISV Tests on Simu-
lated Wastes in INEL Soils. PNL-7344, Pacific North-
west Laboratory, Richland, Washington.
R-2
-------
Federal Register. June 1, 1990. Vol. 55, No. 106, pp
22560-22561, 22573-22574.
FitzPatrick.V.F. 1986. In Situ Vitrification-A Status of the
Technology. Paper presented at SPECTRUM '86,
West Valley, New York, September 14-18, 1986.
PNL-SA-14281, Pacific Northwest Laboratory,
Richland, Washington.
Freeman, H. 1986. Innovative Thermal Processes for
Treating Hazardous Wastes. Technomic Publishing
Co., Inc., Lancaster, Pennsylvania.
Geosafe. 1991. Engineering-scale Test Report for Appli-
cation of In Situ Vitrification Technology to Soils
Contaminated with Polychlorinated Biphenyis at the
Northwest Transformer Superfund Site. GSC 1006,
Geosafe Corporation, Richland, Washington.
Gilges.K. 1991. Europe's Hazardous Waste Dilemma.
Chemical Engineering, February, 1991. pp. 30-39.
Goles, R.W. and C.M. Andersen. 1986. Off-Gas Sys-
tems, in LFCM Vitrification Technology, Quarterly
Progress Report, July-September, 1986. PNL—5904-
4, Pacific Northwest Laboratory, Richland,
Washington.
Graziano, J.H. and C. Blum. 1991. Lead Exposure from
Lead Crystal. The Lancet 337:141-142, January 19,
1991.
Green, J.C. et al. 1990. Comparison of Toxicity Results
Obtained from Eluates Prepared from Non-Stabilized
and Stabilized Waste Site Soils. Hazardous Waste
Treatment :77-80.
GRI. 1989. Experimental Study of Natural Gas Fueled
Technologies for Reduction of Heavy Metal Leachate
from Solid Wastes. GRI-89/0078, Gas Research
Institute, Chicago, Illinois.
Hansen.J.E. 1991. Treatment of Heavy Metal Contami-
nated Soils by In Situ Vitrification. Presented at the
4th Annual Hazardous Materials Management Con-
ference, Rosemont, Illinois, April 3-5,1991.
Hansen, J.E. and V. FitzPatrick. 1989. In Situ Vitrifica-
tion: Heat and immobilization Are Combined for Soil
Remediation. Hazmat World, December, 1989.
Harlow, G.L., C.A. Whitehurst, H. Robards, D. Deville,
and B.V.Rao. 1989. Ash Vitrification-A Technology
Ready for Transfer. Presented to the American
Society of Mechanical Engineers.
Haz Answers. 1991. Thermal Processes Evaluation for
RWMC Wastes. EGG-WTD-9449, EG&G Idaho,
Inc., Idaho Falls, Idaho. Prepared for EG&G Idaho,
Inc. and U.S. Department of Energy.
Hazardous Waste Consultant, The. 1990a. Electroplat-
ing Wastes Recycled by New Vitrification Process.
The Hazardous Waste Consultant, July/August, 1990.
Hazardous Waste Consultant, The. 1990b. In-Situ Vitri-
fication Update. The Hazardous Waste Consultant,
September/October, 1990. pp. 1-18.
Hnat, J.G., W.F. Olix, L.M. Bartone, W.F. Talley, and A.
Magley. 1990a. Recycling of Insulation Fiberglass
Waste with a New Vitrification Process. Presented at
22nd Mid-Atlantic Industrial Waste Conference, Drexel
University, Philadelphia, Pennsylvania, July 24-27,
1990.
Hnat, J.G., W.F. Talley, L.M. Bartone, P.T. Myles. 1990b.
Cyclone Melting System Adds Options for Glass
Producers. Ceramic Industry, April, 1990.
Holton, Jr., L.K., Dierks, R.D., Goles, R.W., Y.B. Katayama,
J.E. Surma, and N.M. Thomas. 1988. Operating
Experience in a Radioactive Liquid-fed Ceramic Melter
Vitrification Facility. PNL-SA-15243, Pacific North-
west Laboratory, Richland, Washington, February,
1988.
Jacob, A. 1991. Processing with Plasma. The Chemical
Engineer, pp. 33-35, May 30, 1991.
Jacobs, G.K., B.P. Spalding, J.G. Carter, and S.S. Koegler.
1988. In Situ Vitrification Demonstration for the
Stabilization of Buried Wastes at the Oak Ridge
National Laboratory. Nuclear and Chemical Waste
Management, 8: 249-259, 1988.
Janke, D.S., C.C. Chapman, and R.A. Vogel. 1991.
Results of Vitrifying Fernald K-65 Residue. Pre-
sented at the Fifth International Symposium on Ce-
ramics in Nuclear and Hazardous Waste Manage-
ment, American Ceramic Society, Cincinnati, Ohio,
April 29-May 3,1991. PNL-SA-1860, Pacific North-
west Laboratory, Richland, Washington.
Jantzen, C.M. 1988. Prediction of Glass Durability as a
Function of Environmental Conditions. Materials
Research Society, Mat. Res. Soc. Symp. Proc. Vol.
125.
R-3
-------
Jantzen, C.M., and N.E. Bibler. 1990. The Product
Consistency Test forthe DWPF Wasteform. WSRC-
MS-90-149, Westinghouse Savannah River Com-
pany, Savannah River Laboratory, Aiken South
Carolina. Presented atthe 2nd International Seminar
on Radioactive Waste Products, Julich, Germany,
May 28-June 1,1990.
Jantzen, C.M., and D.F. Bickford. 1985. Leading of Devit-
rified Glass Containing Simulated SRP Nuclear Waste.
Materials Research Society, Vol. 44, pp. 135-145.
Johnson, N.P. and M.G. Cosmos. 1989. Thermal Treat-
ment Technologies for Hazardous [Waste
Remediation. Pollution Engineering, October, 1989,
pp. 66-85. !
Jouan, A., C. Ladirat, and J.P. Moncouyoux. | 1986.
Present Status of the French Continuous !Fission
Product Vitrification Process, in Advances! in Ce-
ramics, Vol. 20: Nuclear Waste Management II, The
American Ceramic Society. i
I
Klingler, L.M., and P.L. Abellera. 1989. Joule-Heated
Glass Furnace Processing of a Highly Aqueous Haz-
ardous Waste Stream. MLM-3577, EG&G Mound
Applied Technologies, Miamisburg, Ohio.
Koegler, S.S., K.H. Oma, and J.M. Perez, Jri 1988.
Vitrification Technologies for Weldon Springs Raff inate
Sludges and Contaminated Soils Phase 1 Report:
Development of Alternatives. PNL-6704, Pacific
Northwest Laboratory, Richland, Washington.
Koegler, S.S., R.K. Nakaoka, R.K. Farnsworth, and S.O.
Bates. 1989. Vitrification Technologies for yVeldon
Springs Raffinate Sludges and Contaminated Soils
Phase 2 Report: Screening of Alternatives. PNL-7125,
Pacific Northwest Laboratory, Richland, Washington.
I
Komatsu, F., A.Takusagawa, R. Wada, and J.K. Asahina.
1990. Application of Microwave Treatment Technol-
ogy for Radioactive Wastes. Waste Management,
Vol. 10, pp. 211-215. |
Landau Associates. 1991. Report of Engineering con-
sultation Vitrification Treatability Study Northwest
Transformer (Mission/Pole) Site, Everson, Washing-
ton, Final. Landau Associates, Inc., Edmonds,
Washington. Prepared for Northwest Transformer
Steering Committee.
Liikala, S.C. 1991. Applications of In Situ Vitrification to
PCB-contaminated Soil. Presented at the 3r|d Inter-
national Conference for the Remediation of PCB
Contamination, Houston, Texas, March 25-26,1991.
Loewenstein, K.L. 1983. The Manufacturing Technology
of Continuous Glass Fibres, Second Edition. Elsevier,
Amsterdam, The Netherlands.
Lominac, J.K., R.C. Edwards, and C.L. Timmerman.
1989. Pilot Scale In Situ Vitrification at Arnold Engi-
neering Development Center, Arnold AFB, TenneV
see. Superfund '89, Proceedings of the Tenth Na-
tional Conference, November 27-29, Washington,
D.C., Hazardous Materials Control Research Insti-
tute.
Marcus, A.D. and S.C. Falud. July 8-12, 1991. Federal
Conflict of Interest Claimed in Waste Cleanup.
Wall-Street Journal, July 8-12,1991.
McLellan, G.W. and E.B. Shand. 1984. Glass Engineer-
ing Handbook, Third Edition. McGraw-Hill Book Com-
pany, New York, New York.
Means, J.L.etal. 1987. Long-Term Performance of High-
Level Glass Waste Forms. NUREG/CR-4795, BMI-
2143, Battelle, Columbus, Ohio.
Mendel, J. 1973. A Review of Leaching Test Methods
and the Leachability of Various Solid Media Contain-
ing Radioactive Wastes. BNWL-1765, Pacific North-
west Laboratory, Richland, Washington.
Miyata, K., J. Ohuchi, E. Inada, and N. Tsunoda. 1989.
Incineration and Ash Melting for Plutonium-contami-
nated Wastes at PWTF. Presented atthe Incineration
Conference, Knoxville, Tennessee, May 1-5,1989.
Newton, R.G., and A. Paul. 1980. Glass Technology, 21,
307.
Nuclear Engineering International. 1990. Assessing
Sellafield's Vitrification Nuclear Engineering Interna-
tional, March, 1990. pp. 43-46.
Ohuchi, M.,K. Miyata, J. Ohuchi, E. Inada andN.Tsunoda.
1989. Melting of Pu-contaminated Residue by Pow-
ering Microwave. Presented ay the 4th International
Symposium on Ceramics in Nuclear Waste Manage-
ment, Indianapolis, Indiana, April 23-27,1989.
Orfeuil, M. 1987. Electric Process Heating. Battelle
Press, Columbus, Richland, 1987.
Petersen, R.D. 1990. Microwave Vitrification of Rocky
Flats TRU Sludge. RFP-4414, U.S. Department of
Energy, Rocky Flats Plant, Colorado.
R-4
-------
Pincus, A.G. and G.M. Diken. 1976. Electric Melting in
the Glass Industry. Books for Industry and the Glass
Industry Magazine, N.Y., N.Y.
Plodinec, M.J., G.G. Wicks, and N.E. Bibler. 1982. An
Assessment of Savannah River Borosilicate Glass in
the Repository Environment. DP-1629, Savannah
River Laboratory, Aiken, South Carolina.
Postles, R.L. and K.G. Brown. 1992. The DWPF Product
Composition Control System at Savannah River:
Statistical Process Control Algorithm. Presented at
the Fifth International Symosium on Ceramics in
Nuclear and Hazardous Waste Management, to be
published in Ceramic Transactions.
Rankin.W.D. and G.G. Wicks. 1992. Chemical Durability
of SRP Waste Glass as a Function of Waste Loading.
J. Am. Ceram. Soc., 66(6): 417-420.
Reimus, M.A.H. 1988. Feasibility Testing of In Situ
Vitrification of New Bedford Harbor Sediments.
PNWD-1310, Battelle, Pacific Northwest Laborato-
ries, Richland, Washington.
Richards, R.S. and J. W. Lacksonen. 1991. Stir-Melter
Vitrification of Simulated Radioactive Waste, Fiber
Glass Scrap, and Municipal Waste Combustor Fly
Ash.
Roberts, D. 1989. The Vitrifix Process. IEEE Transac-
tions on Industry Applications, 25(3): 451-455, May/
June, 1989.
Roy, K.A. 1991. Molten Metal Technology Destroys
Wastes, Recovers Valuable Constituents. Hazmat
World, July, 1991. pp.62-64.
Schlegel, R. 1989. Residues from High Temperature
Rotary Kilns and Their Leachability. In: Forum on
Innovative Waste Treatment Technologies: Domes-
tic and International, Atlanta, GA, June 19-21,1989,
EPA-540/2-89-056, U.S. Environmental Protection
Agency, Office of Research and Development, Cin-
cinnati, Ohio. pp. 195-230.
Schlienger, M.P. and R.C. Eschenbach. 1991. Latest
Developments with Plasma Centrifugal Furnace.
Presented at IN EL Plasma Waste Workshop, Idaho
Falls, Idaho, January 16-17,1991.
Schreiber,H.D.,etal. 1988. Oxidation-Reduction Chem-
istry of Nonmetals in a Reference Borosilicate Melt, in
D.F. Bickford et al., eds., Advances in the Fusion of
Glass, Proceedings of the 1st International Confer-
ence on Advances in the Fusion of Glass. American
Ceramic Society, Westerville, Ohio. .
Shelley, S. October, 1990. Turning Up the Heat on
Hazardous Waste. Chemical Engineering, October,
1990. pp. 47-54.
Sing, K.S. and D. Swallow. 1960.
Chemistry, 10:171.
Journal of Applied
Smith, R.A., D.H. Nyman, and B.N. Anderson. March,
1990. Getting Ready to Build the Hanford Waste
Vitrification Plant, Nuclear Engineering International,
March, 1990. pp. 40-43.
Spalding, B.P. and G.K. Jacobs. 1989. Evaluation of an
In Situ Vitrification Field Demonstration of a Simu-
lated Radioactive Liquid Waste Disposal Trench.
ORNL/TM-10992, Oak Ridge National Laboratory,
Oak Ridge, Tennessee.
Spalding, B.P., G.K. Jacobs, and E.G. Davis. 1989.
Demonstrations of Technology for Remediation and
Closure of Oak Ridge National laboratory Waste
Disposal Sites. ORNL/TM-11286, Oak Ridge Na-
tional Laboratory, Oak Ridge, Tennessee.
Spalding, B.P., G.K. Jacobs, N.W. Dunbar, M.T. Naney,
J.S. Tixier, T.D. Powell, and N.E. Peterson. 1991.
Tracer-Level Radioactive Pilot-Scale Test of In Situ
Vitrification Technology for Stabilization of Contami-
nated Soil Sites at ORNL. Oak Ridge National
Laboratory, Oak Ridge, Tennessee.
Stegman, J., P.L. Cote, and P. Hannak. 1988. Preliminary
Results of an International Government/Industry Co-
operative Study of Waste Stabilization/Solidification.
In: R. Abbou, ed. Hazardous Waste: Detection, Con-
trol, Treatment. Elsevier Science Publishers B.V.,
Amsterdam, The Netherlands, pp. 1539-1548.
Timmerman, C.L and M.E. Peterson. 1990. Pilot-Scale
Testing of In Situ Vitrification of Arnold Engineering
Development Center Site 10 Contaminated Soils.
PNL-7211, Pacific Northwest Laboratory, Richland,
Washington.
Timmerman, C.L. 1986. In Situ Vitrification of PCB-
Contaminated Soils. EPRI-CS-4839, Electric Power
Research Institute, Palo Alto, California.
R-5
-------
Timmerman, C.L. 1989. Feasibility Testing of In Situ
Vitrification of Arnold Engineering Development Cen-
terContaminated Soils. PNL-6780, Pacific Northwest
Laboratory, Richland, Washington. '
Timmerman, C.L. and K.H. Oma. 1984. An| In Situ
Vitrification Pilot-Scale Radioactive Test. PNL-5738,
Pacific Northwest Laboratory, Richland, Washington.
I
Timmons, D.M., V. Fitzpatrick, and S. Liikala.; 1990.
Vitrification Tested on Hazardous Wastes. Pollution
Engineering, June, 1990. pp. 76-81.
i
Tooley, F.V. 1984. The Handbook of Glass Manufacture,
Vol 1, Third Edition. Ashlee Publishing Co. |
Trier, W. 1987. Glass Furnaces. Society of Glass
Technology, Sheffield, UK.
Twidwell, L.G., and A.K. Mehta. 1985. Disposal of Ar-
senic Bearing Copper Smelter Flue Dust, in Nuclear
and Chemical Waste Management, Vol. 5, pp. 297-
303. Butte, Montana: Montana College of (Mineral
Science and Technology.
USATHMA. 1987. Development of a New Process for
Treatment of Paint Sludge Wastes, Final Report.
AMXTH-TE-CR-87141, U.S. Army Toxic and Haz-
ardous Materials Agency, Aberdeen Proving Grounds,
Maryland.
USATHMA. 1988. Bench Scale Glassification Test on
Rocky Mountain Arsenal Soil. AMXTH-TE-CR-88015,
U.S. Army and Toxic and Hazardous Materials Agency.
USDOE. 1982. Environmental Assessment—Waste Form
Selection for SRP High Level Waste. DOE/EA-0179,
U.S. Department of Energy, Washington, D.b.
USEPA. 1986. PCS Sediment Decontamination-Techni-
cal/Economic Assessment of Selected Alternative
Treatments. EPA/600-2-86/112, U.S. Environmental
Protection Agency, Office of Research and Develop-
ment, Cincinnati, Ohio. '
USEPA. 1987. Technical Resource Document Treat-
ment Technologies for Halogenated Organic Con-
taining Waste. PB88-131271, U.S. Environmental
Protection Agency. !
USEPA. 1988. Technology Screening Guide for Treat-
ment of CERCLA Soils and Sludges. EPA/540/2-88/
004, U.S. Environmental Protection Agency; Wash-
ington, D.C.
USEPA. 1989a. Stabilization/Solidification of CERCLA
and RCRA Wastes: Physical Tests, Chemical Test-
ing Procedures, Technology Screening, and Field
Activities. EPA/625/6-89/022, U.S. Environmental
Protection Agency, Center for Environmental Re-
search Information, Cincinnati, Ohio.
USEPA. 1989b. The Superfund Innovative Technology
Evaluation Program: Progress and Accomplishments
Fiscal Year 1988. EPA/540/5-89/009, U.S. Environ-
mental Protection Agency, Office of Solid Waste and
Emergency Response, Washington, D.C.
USEPA. 1989c. Vitrification Technologies. Immobiliza-
tion Technology Seminar. CERI-89-222, U.S. Envi-
ronmental Protection Agency, Center for Environ-
mental Research Information, Cincinnati, Ohio.
October, 1989.
USEPA. 1990a. Final Best Demonstrated Available
Technology (BOAT) for K031, K084, K101, K102,
Characteristic Arsenic Wastes (D004), Characteristic
Selenium Wastes (D010), and P and U Wastes Con-
taining Arsenic and Selenium Listing Constituents.
EPA/530-SW-90-059A, U.S. Environmental Protec-
tion Agency, Office of Solid Wastes, Washington,
D.C.
USEPA. 1990b. Handbook on In Situ Treatment of
Hazardous Waste-Contaminated Soils. EPA 540/
2-90/002, U.S. Environmental Protection Agency,
Risk Reduction Emergency Laboratory, Cincinnati,
Ohio.
USEPA. 1990c. Remedial Action, Treatment and Dis-
posal of Hazardous Waste: Proceedings of the Fif-
teenth Annual Research Symposium. EPA/600/9-
90/006, U.S. Environmental Protection Agency, Risk
Reduction Engineering Laboratory, Cincinnati, Ohio.
USEPA. 1991. Demonstration Bulletin. Flame Reactor.
EPA/540/M5-91/005, U.S. Environmental Protection
Agency, Center for Environmental Research Infor-
mation, Cincinnati, Ohio.
Vaux.W.G. 1988. Today's Challenge of MSW Incinera-
tion - The Management of Incinerator Ash. in Pro-
ceedings of APCA Annual Meeting, 81st(2), 88/26.12,
pp. 2-9.
Viall, A.J..J.W. Sears, and R.C. Eschenbach. 1990. Test
Results with the Plasma Centrifugal Furnace. Pre-
sented at The Incineration Conference, May 16-18,
1990, San Diego, California.
R-6
-------
Volf, M.B. 1984. Chemical Approach to Glass. Elsevier
Science Pub., New York, New York.
Wakui, H., O. Yamamoto, M. Matsui, and M. Kawakami.
1988. Development of an Advanced Off-Gas Scrub-
ber for an Incinerating Melter. Proc. Symp. Waste
Manage., Waste Management '88, Vol. 1, pp. 537-41.
Wicks, C.G. 1985. Nuclear Waste Glasses. Treatise on
Materials Science and Technology, Vol. 26. ISBN
0-12-341826-7, Academic Press, Inc., 1985. pp.
57-117.
Wicks, G.G. 1986. Structures of Glasses. mM.B.Bever,
ed. Encyclopedia of Materials and Science and
Engineering, Vol. 3. Pergamon Press Ltd., pp.2020-
2025.
Wicks, C.G. and D.F. Bickford. 1989. Doing Something
About High-Level Radioactive Waste. Technology
Review, November/December, 1989.
Wicks, G.G. and M.A. Molecke. 1986. WIPP/SRL In-Situ
Testing Program, in D.E. Clark, W.B. White, and A.J.
Machiels, eds. Advances in Ceramics, Vol. 20: Nuclear
Waste Management II. The American Ceramic Soci-
ety, Westerville, Ohio.
Wicks, G.G., P.B. Macedo, M.A. Molecke, A.R. Lodding,
and D.E. Clark. 1991. MIIT; International In-Situ
Testing of Simulated HLW Forms - Preliminary Analy-
ses of SRL/TDS Waste Glass and Metal Systems.
R-7
-------
-------
*^
Q
O
«
•*-* /A
8 8
|s
>£
•g
_c
E
H
8
0)
Is
Site/Contractor
•Ss,
is
0)
"o
0
E
co
o>
O
•o
a
CO
0)
s
3
O
-3
m
O
CO
en
5^
o
LU
O
-o
c
CO
CO
E
1
6
to
2
"o
; >;
O O
LU LU
0 0
^~ ^™
"O "O
c c
CO CO
1 1
•f-< -t-<
"o "o
CO CO
CB CB
•c -c
1 1
-1 -1
5 §
I X
CB CB
D5 C5
T3 -O
3 3
CO CO
~° ^
'= 3
.cr g-
c
v-" E
ro'c,!
Institute for NukU
Entosogungstecf
Karlsruhe, West '
SRS
r*"" a) £>
O> 2. O) i-
T- Q. T- QL
§
en
-£
o
LU
"5
^^
"O
CO
E
CO
6
CO
"o
CO
CB
•c
1
-1
3
I
CB
C5
T3
3
CO
^
.0-
J) -
O o
"2- °l
"= S ro
!-e"jf
^1
O)
CO
en
>;
2
LU
o
~^f
^^
"O
c
CO
E
6
1
series of
|
3
X
CB
O3
13
3
CO
T3
'3
a"
co to"
CB E
ca -c
•-•§
Ishikawajima-Ha
Industries Co., Y
Japan
_
2
3T
1
2 g
*s
E^
CO D-
. — .
I
2
CB
cn
co
2
^
CO
CO
CB
to
CO
>
CO
_a
CO
CO
'5
CO
o
CO
cn
T—
cn cn
CO CO
cn cn
i — i —
>i >i
£ 2
LU LU
o o
^> ^>
"O "O
CO CO
CO CO
E E
5 -C <
O O Z
1 1
"o *o
CO CO
CB CB ^
•c -c ^
® CD .-=
CO CO O.
CB ^~.
N CB
"o -*
Q- cn
O O tj CB
u. u. <">§
_J _l CBC-
•a
CO co
.0 'c
3 3 « S5
x x o.i
CB CB
O) D)
•o -o
3 ^
CO CO
'5 '5 tf
CT .Of ^
-1
"55
^ CB
^> nj
*p - ~u
i §
^ Q "®
3 5 3
LU ? CO
' ^ 'l 0) .Q
§£ §52 §
T- Q. i- a. -i-
1
,_r
_c
CB
CO
m
,
0
'1
CB
O-
O
T3
|
^
(O ^
1^
°-°
5
=
5
1
"^**
-a
1
IT
CO
co
cc.
CO
cn
S
CO*
•c
0}
X)
D?
"5
««_
2
CB
co SE
CO i;
CO ~
OJ^-
co
•to
^
CO
CO
cS
to
1
Q)
"nt
to .c
X3 'C
CO O
CO O
>
^ Q-
co'LU
Dalzell Glasswoi
demonstration fc
i-^
co
cn
^
co
cn
t—
<
CO
=)
-s
-§
J3
^
1
0
-o
CO CO
o 'c
co E5
0.1
c
"ll
c to
•- 1
-------
I
•s
cc
•q
S^ en
= 8
— £
O)
CO
o>
oo
en —:
co (0
o> »-
P. T=f,
CO
I"
6
o •?
o
if
o
•5
c -y
Q. co
CB -=
c: o o
_O g)
"VT *~> i-i
.2 O
CO
^ "O
^ -a
O
CB 3> ^a
sO J3
[ .is CO C -t^
2 to
Q
CO g
O ?
• D> ft
2-5-=
1
LU
OS
-22 "£
c= -^
!••
to E .3 y w
CO •£
"rri CO
^^
DC CO
'JS nj_5
o co
.
•g to
to p
2 3 Q. E =
o p)
I ~
CD CO
1!
i
IS
1
^ EC
CO £i
'^ 8
to-
.=: ^J r! U >>
co ipr c o 5=
01
D)
co
LLI
CO O
u
0)
o?
II I
I! I
LU
CO
_co co
o:
CO
8
;8
^^ M/
O a.
O T—
LU
g
CO
I 15
CD ~ C
> I CO
cn
-------
§
£
£
0)
CC
CD
i
o
•a
c
c5
.0>,
Is
£ 2
> a.
Contaminant
0)
I!
c A (continued)
Site/Contractor
^ "o
= -a
a. S3
u Vitrification
CO
—
E
en
CO
en
co"
co o
0° "
en W
T— ^
-o
c c
CO CO
^ D>
CB -^
F CO"
0 ^
S J
_Q Q.
CO CO
PCBs
liquid radioactive
waste
-o
CO
CO
"Q_
CO -C
Q. 0
— CD C
O CD CD
CO CO +z
LU
o
£ z
Q. CC
LU O
10 [-~
co co
en en
T— 1 —
co
CO
o>
1
1 1
CO ^2
^ .i
3 CD
m cc
en c
r« ^
CO CO
~o co
g m
•S 0
E o-
_Q
"1 «
ra "S
> CO
Q.
LU
i_ ^"
LU X af
8 11
CO T3 CZ
> CD 3
— J 5 f^
j*^1 QJ 3
Q ~^_ (/)
t**~ CO
CO CO
o
o>
0>
„
o
CO
Q)
CD
CD Q_
1 =
cf c
CO CO
E E
0) 0)
pp
•i'o
»2- co
CO
E
organics (petroleu
products), metals
"8
Q LU
en
CO
CO
co
T-
r~i
en o
°> en
^ en
"co ^T.
* V
.C CD
1 I
E i"
co to
LU 0
D) O)
C C
~ '^
CD CD
CD CD
1 I
CO CO
s
o W 'm
spiked with steel ti
test shorting of EF
radioactive and mi
contaminants
_j
LU
^
2
=5 «
co H>
LU LU
0 O
Q Q
1 1
_r _r
Q_ Q.
G) en
co co
05 en
o
CO
O)
"^
0)
tn
•o
CO
I
CO
O
.c
O
c
0)
_Q
CO
CD
!
"S E
5. co
1
LU
6
CO
co"
cc
co
en
CO
en
en
co
O5
T—
"CO
0)
CD
O5
CD -.
5 ^
1 1
CD CD
J3 ^3
> >
CO CO
JT
5j
_ >> T3
I— "O E
CD
:=T ""
"o ^
CO CD
"B'cfl =
^ = O
co cr co
^ o
o CD"
5 CO
CO c
05 §
'5 ^
CX CO
CO CC
c >-
O fD
"O 1 1 1 > ^f
5o icL
> Q O LU
en o
co o>
O5 (35
m
en
"c5
1
1
'5.
CO
CD
.s
organics, metals,
simulated radioac
wastes
CD
CO
1
13
INEL, Idaho/DOE
§
en
o
1
"co
'ft
CO
o
E
F
D)
C
"l_
CD
CD
_g
CD
>
mercury, arsenic,
aldrin, and dieldrir
"o
CO
c
CO
CD
D>
"55
z
o
en
en
§
i—
CD"
•co
S
CD
O
C
CD
CD
C
'01
0)
CO
CO
m
o
a.
1
CD
E .
Northwest Transfo
Superfund site, Wi
o
T—
^™
en
O5
"co
O5
_c
"co
Q.
CO
'5.
CO
radioactive and
inorganic
contaminants
CO
CD
•s
c
0)
•*-*
-&
CO
CO
"5.
1
ORNL, TnVDOE
en
-------
•g
§1
11
II
cu
en
c
X
TO
at
CO
LU
LU
•g
: •=» £ to
5 <5 IS-m
.-
E o
C (!)
XJ IVJ w^ ;;r w C/J r— *"
JS CD tOj;; COCDC CO
O.DC "5.S. 15.CCri CLI
o
_ =8
in .a
1— _Q
1
= '85-
co '
Q_ CO <0 i3
i LU it "H .<2
•§2 -5
II
en c w iz
Few N E
s-lli il1
ooll li
LU
CO
Q.
.SO
O
o< P.
.yj "a
CO g
m Q
co
*- o
Q.E
£0
CO ^
CD co
r— •-*-
E Q.
J5
co
C3J
"CO
08
> ^a
ey)
o o
CO
CC
co a>
LU
LI I
-------
•i
cc
S
CO
co
0
c
8
Q.
u
S
S
o
O
a
CO
T= W
S =
I
_
UJ
V)
o
0
8
in
i
o
co
en
..
1
m
reduction
Q.
D5
C
cz "55
calcinatatio
followed by
induction m
>
i
X
_I
X
73
CD
'o
8
*o5
o
OS
LL
0)"
^3
^f
CO LU
20
CO «
^ CD
t- Q.
eer/ng International,
.^
%
Hi
^
TO
|§
S?
"CO
cz
g
to
1.
o
73
CD
5 CJ3
— S
[o "3
calcination •
induction m
>
3
z
5
—i
-C
"S
|g
<3
^
CO
en
c
LU
-C,"
"5
M—
co
"3
CO
i —
C5
O5
•z.
hakedown
co
electric arc
to
.g
'cz
CO
0
08
CO
_g
§
cp
o
•—
j£
<« CD
CO Q)
^= Z)
^-co
CO p
CO §
"p
2 co
o
m
c
o
D)
CD
O
^
cz
CO
.0,
CM
en
en
c
.g
s!
CD
cz
"cj
D)
C
'mal Process Hea
0>
x:
^^
en
»
T-
"cfl
to
s
_o
CO
I
'cfl
"S
0
75
CD
Q.
0
^
CD
CO
CD
Q.
rotary kiln
(Marine Shale
Processors)
some I
inorganics and
T3
incinerator ash an
Louisiana/vendor
CO
en
organics
baghouse dust
t:
CD
CO
CD
Q.
0
CO
O)
E
o
o
c
CD
CO
thermal proces
heating
"8
CD
cadmium and 1
JZ
CO
CO
Gas Research Institute
en
CO
co
?
"co
•s
._
1
c:
CO
Q.
'5.
_c
eo"
combustibles,
flame retardan'
simulated waste
c
1
non-combustib
±3
Chiyoda R&D Center, .
Chyoda Chemical
Engineering
and Construction Co.,
CO
co
en
2
io"
33
CD
cz
c:
o
i
modified rotary
1
CO
D5
organics & inoi
modified MSW
Brisbane, Australia
CO
co
en
V-
.
(5
a.
'5.
kiln process/
Nentralysis
Industries
•E
CD
CO
CD
Q.
C
w
as
a
to
c:
o
08
CO
ns
00
E
smelter blast
furnace soda slag
Q.
co"
o
ico
Is
^ i —
-------
-------
-------
------- |