United States
         Environmental Protection
Office of
Research and Development
Washington, DC 20460
May 1992
         Technology Transfer
&EPA   Handbook
         Vitrification Technologies
         for Treatment of Hazardous
         and Radioactive Waste


                                      May 1992
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

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

    1.1  Purpose	1-1
    1.2  Overview	1-1
    1.3  Handbook Organization	1-2

    2.1  Glass Structure	2-1
    2.2  Stabilizing Mechanisms	2-4
    2.3  Chemical Attack Mechanisms	2-5

    3.1  Electric Process Heating	3-1
    3.2  Thermal Process Heating	3-8

    4.1  Applicable Waste Types	4-1
    4.2  Applicable Contaminants	4-3

    5.1  Product Durability	5-1
    5.2  Product Volume Reductions and Densities	5-5
    5.3  Product Use	5-5

    6.1  Off-Gas Components	6-1
    6.2  Constituents of  Concern	6-1
    6.3  Means of Off-Gas Control	6-2

    7.1  Capabilities	7-1
    7.2  Limitations	7-2

    8.1  Physical Tests	8-1
    8.2  Chemical Tests	8-3

    9.1  Selection of Vitrification Processes	9-1
    9.2  Initial Testing and Scaling-Up	9-1
    9.3  Cost	9-5


                                      LIST OF TABLES

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

                                      LIST OF FIGURES
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

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       '

                                  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

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

                                         CHAPTER ONE

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 1600-2000C) than ex situ processes
(typically 1000-1600C) 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

 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
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

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
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)

    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).  |
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

 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 -

        Table 2-1.  Sample Compositions of Soda-Lime Glass, Borosilicate Glass, and ISV Glass

1From McLellan
2From Goldston
3This glass was
Soda-Lime Glass1
(wt %)
and Shand, 1984.
and Plodinec, 1 991 .
produced by ISV of INEL soils. From
SRS Borosilicate
Benchmark Glass2
(wt %)

Farnsworth, Oma, and Reimus, 1990.
ISV Glass3
(wt %)

       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

Certain inorganic species can be immobilized by chemi-
cal bonding with the glass-forming materials, particularly

 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

                  Matrix dissolution

           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
 Under Oxidation
                                                                           Under Reduction
            Cadmium Sulfide
            Cadmium Sulfide, Selenium
            Cobalt Oxide
            Copper Oxide
            Cuprous Oxide
            Cerium Oxide
            Chromic Oxide
            Iron Oxide
            Manganese dioxide
            Noodymium oxide
            Nickel oxide
            Nickel oxide
Greenish blue
Gjreenish blue
T|tania Yellow
Ypllowish green
Yellowish green
Amethyst to purple
Violet in K2O glass
Brown in Na2O glass
Yellow with green fluorescence
Greenish blue
Emerald green

Bluish green
Violet in K2O glass
Brown in Na2O glass
Yellow to amber
Green with fluorescence

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

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 10C 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

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

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

saturation, thus reducing even more the tendency of silica
in the glass to move into solution.               !
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

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

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 100C or higher, network
dissolution can dominate. The exact temperature forthe
shift in mechanism varies with test conditions and glass

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.

                  Table 2-3. Effects of Waste-Glass Components on Processing and
                Product Performance (adapted from Plodinec, Wicks, and Bibler,1982).
 Frit Components
 Product Performance







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

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
 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


                                         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

 2. Thermal Process Heating
Ceramic Melter
In Situ Vitrication
Plasma Furnace
Microwave Melter
Resistance Heating,
Induction Heating,
Electric Arc Heating

Rotary Kiln Incinerator
(operated in slagging
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

               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

processes. However, soils and other materials heated by
joule heating are frequently quite resistant and require
higher voltages.
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 1300-1400C.  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.
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,

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. 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).
1,2 Ordinary glass
3  Glass lamp
4  Neutral glass
5  Borosilicate
6,7 Pyrex
8  Lead crystal
      1100   1200
1 300   1 400

  Figure 3-1. Relationship Between Resistivity and
  Temperature for Selected Glasses ( Orfeuil, 1987)

 Ceramic Melter
                                                 Filter      Heat
                                                         Exchanger   HEPA
                                 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  1000-1600C  (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
                 Incoming  feed     off-gases

                          1__L-     .plenum
                                           cold cap
      refractory wall
       molten glass

      metal precipitate
        molten metal tap
          Figure 3-3. Generalized JHCM Showing
         Components of Melter and Molten Material

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). 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.

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 2000C, well above initial
soil-melting temperatures of 1100C  to  1400C. 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-
                Floating Layer
               (Rocks, Ceramics)
                     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-~
                                           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)

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,

                          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-
                                Off-Gas Hood
                                    Off-Gas Line  V ' V
                                    Support    t
r                                                                         Outdoor
                                                                         Lighting (lyp)
                                                  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)

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., 20,000K 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
2,000K  and 5,000K.  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

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,

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
         Transferred Arc
Non-Transferred Arc
    Figure 3-6. Comparison of a Transferred Arc
    and a Non-Transferred Arc (Source: Plasma
                Energy Corporation)

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 10,000C, heats
the waste material beyond the point of melting to about
1,600C. 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
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

                                    EXIT GAS AND
                                   SLAG REMOVAL
         Figure 3-7. Schematic of the Demonstration PCR Showing the Bottom-Pour Configuration
                    for Exit Gas and Molten Glass (Eschenbach, Hill, and Sears, 1989)

                                                   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.,

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
  2,450 MHz
    Figure 3-9. Microwave Melter (Orfeuil, 1987)

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.  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).  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).  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

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.

                   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 1200C. 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.

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
1260C) 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
                                                   OF FINES
                   MIXING/BLENDING TANKS
      Figure 3-11. Simplified System Schematic of MSP's Process (adapted from Harlow et al., 1989)


                                        CHAPTER FOUR
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

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
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).

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

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

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

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

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

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,
  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,
  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

              Table 4-2.  Metals Retention Efficiency Test Results for ISV (Hansen, 1991)
Class Metal
Volatile Mercury
Semi-Volatile Arsenic
Non-Volatile Americium

(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)
1-2 ft.

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. 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

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).

  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-
                                     La/Nd i
                                              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

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).  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

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 1290C.  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,

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

 evolve as part of the treatment process and thus require
 attention.                                      ; 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 400-900C, 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 1300C.| 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  1100C 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 (900C 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). 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). 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

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).  Inorganic Oxides of Concern

Primary inorganic oxides of concern include  nitrogen
oxides (NOX), sulfur oxides (SOx^ and phosphorous
compounds (such as P25 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. 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.

Table 4-3. ISV Organic Destruction and Removal Efficiencies(77;e Hazardous Waste Consultant, 199Gb)
Fuel Oils
(ppb) :
Total DRE
(including off-gas
        Table 4-4. Demonstrated Organic Destruction Efficiencies for Vitrification Systems1
Hydrocyanic Acid
Formic Acid
Methylene Chloride
Mustard Gas
Nitrogen Mustard
Carbon Tetrachoride
'Data collected from Armstrong and Klingler,
C for 99%
Destruction in
2 Seconds
31 8-368
> NA
i NA
1985; USATHMA, 1,988; Klingler and Abellera,
Measured DE (%)

------- 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

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 100C), 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.


                                         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

                 Table 5-1. TCLP Leach Data for Selected Processes and Selected Metals*
Glass Kiln/Vitrification ISV
Metal Melter1-" Process2-" Glass2-0
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,
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

 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,
 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





           SOXHLET CORROSION RATE, g/cm2-d x 10'5

Figure 5-1. Leach Resistances of Selected Materials
                (Buelt et al., 1937)

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

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)
Unreinforced Concrete
Compressive Strength (psi)
3,000 - 8,000
Tensile Strength (psi)
400 - 600

 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).
 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.,

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).


                                         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

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

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
       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.

6.3 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
         recycling of  contaminants  captured  in the
          off-gas system                      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

Reduction of off-gases forms an important means of off-
gas control. Numerous methods permit control of off-
gases at the source of production.
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

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

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-

poration.  However, repeated recycling may also in-
crease processing complexity, total treatment time, and

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

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.

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
Off-Gas Component
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
charcoal and HEPA filters
remove particulate
cool gases
neutralize acid gases
remove water droplets
re-heat gases above dewpoint
filter remaining particles
HEPA filter (optional)
scrubbers (two in series)
condenser         ;
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
quench water      ',
precipitators and stack assemblies
recover heat
cool gases
remove particulate
    'Harlowetal., 1989
    'Froeman, 1986
    'Battey and Harrsen, 1987
    Hnatetal., 1990a

                      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
1 Currently, a
Surface1 Hood
non-combustible fabric
is placed
HEPA Filters
as a ground
cover inside the
Tank 2
hood to prevent
Stage 1
Stage 2
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
Scrub Solutions
Primary HEPA Filter
Secondary HEPA Filter
Amount of
137Cs (Ci)
Amount per
Unit Area
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).


                                       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

        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

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.

 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
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.,

 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).

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,

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

Na20 & K20
Other oxides
Range in
SRS, South

2adapted from Timmerman, 1989
3Campbell and
Buelt, 1990


 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
 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

 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

 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).

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
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)

 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

        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

        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-

 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).

 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

 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)
         1.0m             2.0m             3.0m

  Figure 7-2. Predicted Versus Achieved Large-Scale
            Melt Shape (Buelt et al., 1987)


                                        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-

 It is also important to note that water is often not the only
 liquid-phase constituent in contaminated materials. The

 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

 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.,

 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

(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

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

 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.,

 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
 and Bibler,  1990).  This test  is being evaluated a|s an
 ASTM standard test.                           !

 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.

                                        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
        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.

 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

/^Modified ^
V^ Commercial^/
HLW Joule
Heated Melter

(stirred Melter)

Raw Materials Cost
Waste Loading

(indicates Preferred System)

             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
Low temperature
thermal separator
Infared furnace
Molten salt furnace
Plasma centrifugal reactor
Plasma arc furnace
Conventional temperature
Microwave discharge
Molten glass furnace
In situ vitrification
Microwave melter






LDA = Level of Development and Availability
VRSL = Versatility
THRU = Throughput
FWF = Final Waste Form



EFLT = Effluents















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. 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

 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

 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. 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

 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

 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
                        Table 9-3. Testing Units For Developing ISV Technology
Equipment Size
Bench Scale
Engineering Scale
Pilot Scale
Large scale
! 0.11 m
OJ23 to 0.36 m
3.5 to 5.5 m
1 to 10 kg
10 to 50 t
400 to 800 1

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

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 processessuch 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,
        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.  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.

                 Table 9-4. Major Components of ISV Costs (adapted from Buelt et al., 1987)

       Transporting equipment to and from site
       Clearing vegetation
       Rough grading
       Removing overburden
       Acquiring and applying backfill material

          Portable generator
          Powar lines
          Power cables
          Electrode frame and hood
          Drilling machinery
          Front-end loader
       Off-gas and monitoring
          Off-gas treatment system
          Radiation and off-gas monitors/alarms

    Process preparations
        Drill holes and place electrodes
        Spread graphite starter material
        Position frame and hood, secure electrodes
        Connect power cables and off-gas line
        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


    Secondary Wastes

 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 generatorsif high-voltage line
 power is not available, a crane, and a front end loader or
 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

 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

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)

Moving Equipment
Large Scale,
5-m Depth,
5% Moisture,
Large Scale,
5-m Depth,
25% Moisture,
  Table 9-6. ISV Electrode Spacing and Vitrification
             Settings (Buelt et al., 1987)
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

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

 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)
                           Manpower Rate,
                      Average Man-Hours/Setting
      Table 9-8. Labor Estimate for ISV Processing Operations at a Radioactive Site (Buelt et al., 1987)
Workers Per Shift
Job Classification
Operator (Technician)
Radiation monitor^3)
(a) Radiation monitoring

personnel would
Swing Graveyard
0 0
0 0
2 2
0.25 0.25

not be required for a hazardous
waste site.

 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). 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
Rate, m3/yr
(@ $0.05/kWh)
                             Table 9-10. Sample ISV Cost Estimates ($/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.
(Buelt etal., 1987)
(Carpenter and Wilson, 1988)
(Koegleretal., 1989)
(USEPA, 1990c)
(Landau Associates, 1991)


                        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
of Total Cost

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

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.

 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

 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
Secondary Waste
Total Consumables
Total Operating Costs
Cost, $1000
(% operational cost)

$ 7,380




 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

 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

              Table 9-13. Equipment Required for JHCM Processing (Koegler et al., 1989)
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
Off-Gas Treatment System
Quench Scrubber
Scrub Solution Recycle System
Roughing Filter
Heat Exchanger
HEPA Filter
Glass Handling System
Glass Quencher
Heat Exchanger
Fritted Glass Transfer System
Total Equipment Cost:
"Value is total cost of JHCM system including transformers






Total Cost, $1000






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.
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.

   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
Building & Facilities
Site Preparation
Fees and Engineering Contingency
Total Costs In Support of Equipment
Total Capital Costs
($1000) Cost

50 <
4500 28%

100 <
$5,500 34%


,800 11%
,200 14%
$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
calcined; HLW
toluene, oil
spent resins
wood, cloth, paper
glass melter - Vitrifix asbestos-contaminant ed soil
rotary kiln - MSP
'Glass production is greater than feed
these types of contaminants.
incinerator ash, soil
input due to the need to add
Feed Input Rate
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*
materials to

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:

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-

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,

 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-

 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,

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,
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-

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.

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. PNL5904-
   4, Pacific  Northwest  Laboratory,  Richland,

Graziano, J.H. and C. Blum.  1991. Lead Exposure from
    Lead Crystal. The Lancet 337:141-142, January 19,

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,

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,

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.

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
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.
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-

 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,

 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.

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

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,

Timmerman, C.L. 1986.  In Situ Vitrification of  PCB-
    Contaminated Soils. EPRI-CS-4839, Electric Power
    Research Institute, Palo Alto,  California.

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.
Timmons, D.M.,  V.  Fitzpatrick, and S.  Liikala.;  1990.
    Vitrification Tested on Hazardous Wastes. Pollution
    Engineering,  June, 1990. pp. 76-81.
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,

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 AssessmentWaste 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,

 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,

 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.

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.

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-

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.




*-* /A
8 8







; >;
0 0

^~ ^
"O "O

c c

1 1
f-< -t-<
"o "o
c -c

1 1
-1 -1


D5 C5
T3 -O
3 3
~ ^
'= 3
.cr g-

v-" E
Institute for NukU
Karlsruhe, West '

r*"" a) > 
O> 2. O) i-
T- Q. T- QL






J) -
O o
"2- l
"= S ro



series of




co to"
ca -c
Industries Co., Y



2 g
.  .






cn cn
cn cn
i  i 
>i >i
o o
^> ^>

"O "O

5 -C <
1 1
"o *o
c -c ^
 CD .-=

CB ^~.

"o -*
Q- cn
O O tj CB
u. u. <">
_J _l CBC-

CO co
.0 'c
3 3  S5
x x o.i

O) D)
o -o
3 ^
'5 '5 tf
CT .Of ^

^ CB
^> nj
*p - ~u
^ Q "
3 5 3

' ^ 'l 0) .Q
T- Q. i- a. -i-


(O ^








co SE
CO i;
CO ~


to .c
X3 'C

^ Q-
Dalzell Glasswoi
demonstration fc







o 'c
co E5
c to
- 1


       S^  en

       =  8
                                 en     :
                                 co      (0
                                 o>     -
 P.      T=f,

                           o     ?
                          c      -y
                       Q. co
                                         CB     -=
                                                                               c:     	 o   	  o
                       _O        g)

                       "VT   *~>   i-i
                       .2   O
               ^      "O
               ^      -a
                                                            CB 3>   ^a
                                                            sO   J3
                     [ .is   CO C   -t^
                                                      2 to
               CO g

               O ?
                                                                      D> ft


                                                                            -22 "
                  c=        -^
to   E .3 y  w
                               "rri     CO
                                               DC CO
'JS    nj_5

      o co
     g  to
                                     to  p
                                   2 3  Q. E   =
                                               o p)
                                                                                      I     ~
                                                                                                     CD CO

                                                                                                        ^   EC
                                                                            CO i
                                                      '^   8
                                                      .=:      ^J      r! U    >>
                                                      co      ipr      c o    5=
                                                                               CO O
II   I

I!   I
                                                                                                            _co co
                           ^^      M/

                           O     a.
                                                           O      T
                                       I      15
                                       CD      ~ C
                                       >      I CO



> a.



c A (continued)
^ "o
= -a
a. S3

u Vitrification


co o
0 "
en W
T ^
c c
^ D>
CB -^

0 ^
_Q Q.


liquid radioactive

Q. 0
CO CO +z


10 [-~
co co
en en
T 1 

1 1
CO ^2
^ .i
3 CD
m cc

en c
r ^


~o co
g m
S 0
E o-

ra "S
> CO

i_ ^"
LU X af
8 11
> CD 3
 J 5 f^
j*^1 QJ 3
Q ~^_ (/)

t**~ CO

1 =
cf c
0) 0)

2- co


organics (petroleu
products), metals



en o
> en
^ en
"co ^T.
* V
1 I
E i"
co to
LU 0
D) O)
~ '^
1 I

o W 'm
spiked with steel ti
test shorting of EF
radioactive and mi

co H>

0 O
1 1
_r _r
Q_ Q.

G) en
co co
05 en




"S E
5. co




CD -.
5 ^

1 1
J3 ^3

> >

_ >> T3
I "O E

:=T ""
"o ^
"B'cfl =
^ = O
co cr co
^ o
o CD"
5 CO
CO c
'5 ^
c >-
O fD
"O 1 1 1 > ^f
5o icL
> Q O LU

en o
co o>
O5 (35



organics, metals,
simulated radioac






mercury, arsenic,
aldrin, and dieldrir








E .
Northwest Transfo
Superfund site, Wi





radioactive and







       : =    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
                           _    =8
                     in     .a
                     1     _Q

                     =     '85-
                                  co '
 Q_   CO     <0 i3

i LU   it     "H .<2
2  -5
   en c     w iz
   Few   N E

s-lli   il1

ooll   li

                                                       o<   P.
                                                        .yj   "a
                                                             CO g
                                                      m Q
                                                                   *- o

                                               CO ^
                                               CD co
                                               r -*-
                                               E Q.
                                                            > ^a
                                                            o o
                                                                                 co a>
                                                                                  LI  I

        T=  W

        S  =




cz "55
followed by
induction m





^ CD
t- Q.
eer/ng International,
5 CJ3
[o "3
induction m






electric arc


< CD
^= Z)
CO p

2 co


'mal Process Hea







rotary kiln
(Marine Shale
some I
inorganics and

incinerator ash an




baghouse dust





thermal proces
cadmium and 1


Gas Research Institute





flame retardan'

simulated waste


Chiyoda R&D Center, .
Chyoda Chemical
and Construction Co.,





modified rotary
organics & inoi

modified MSW

Brisbane, Australia



kiln process/



smelter blast
furnace soda slag


^ i