Melter Feed Research Articles

Melter feed viscosity during conversion to glass: Comparison between low‐activity waste and high‐level waste feeds

AbstractDuring nuclear waste vitrification, a melter feed (a slurry mixture of a nuclear waste and various glass forming and modifying additives) is charged into the melter where undissolved refractory constituents are suspended together with evolved gas bubbles from complex reactions. Knowledge of flow properties of various reacting melter feeds is necessary to understand their unique feed‐to‐glass conversion processes occurring within a floating layer of melter feed called a cold cap. The viscosity of two low‐activity waste (LAW) melter feeds were studied during heating and correlated with volume fractions of undissolved solid phase and gas phase. In contrast to the high‐level waste (HLW) melter feed, the effects of undissolved solid and gas phases play comparable roles and are required to represent the viscosity of LAW melter feeds. This study can help bring physical insights to feed viscosity of reacting melter feeds with different compositions and foaming behavior in nuclear waste vitrification.

Effect of melter feed foaming on heat flux to the cold cap

The glass production rate, which is crucial for the nuclear waste cleanup lifecycle, is influenced by the chemical and mineralogical nature of melter feed constituents. The choice of feed materials affects both the conversion heat and the thickness of the foam layer that forms at the bottom of the cold cap and controls the heat flow from molten glass. We demonstrate this by varying the alumina source, namely, substituting boehmite or corundum for gibbsite, in a high-alumina high-level-waste melter feed. The extent of foaming was determined using the volume expansion test and the conversion heat with differential scanning calorimetry. Evolved gas analysis was used to identify gases responsible for the formation of primary and secondary foam. The foam thickness, a critical factor in the rate of melting, was estimated using known values of heat conductivities and melting rates. The result was in reasonable agreement with the foam thickness experimentally observed in quenched cold caps from the laboratory-scale melter.

Determination of heat conductivity of waste glass feed and its applicability for modeling the batch‐to‐glass conversion

AbstractThe effective heat conductivity (λ) of reacting melter feed affects the heat transfer and conversion process in the cold cap, a layer of reacting feed floating on molten glass. A heat conductivity meter was used to measure λ of samples of a cold cap retrieved from a laboratory‐scale melter, loose dry powder feed samples, and samples cut from fast‐dried slurry blocks. These blocks were formed to simulate the feed conditions in the cold‐cap by rapidly evaporating water from feed slurry poured onto a 200°C surface. Our study indicates that the effective heat conductivity of the feed in the cold cap is significantly higher than that of loose dry powder feed, which is a result of the feed solidification during the water evaporation from the feed slurry. To assess the heat transfer at higher temperatures when feed turns into foam, we developed a theoretical model that predicts the foam heat conductivity based on morphology data from in‐situ X‐ray computed tomography. The implications for the mathematical modeling of the cold cap are discussed.

Balance of oxygen throughout the conversion of a high-level waste melter feed to glass

Abstract Gases evolve from nuclear waste melter feed during conversion to glass in response to heating. This paper is focused on oxygen mass balance based on the stoichiometry of feed melting reactions and evolved-gas analysis data. Whereas O 2 -producing and -consuming batch-melting reactions are complete in the reacting and primary-foam layers of the cold cap, O 2 from redox reactions continues to evolve as long as melt temperature increases, and thus generates secondary foam. Also, we discuss the relationship between the oxygen mass balance and the temperature-dependent iron redox ratio and the O 2 partial pressure, as they evolve during the feed-to-glass conversion.

Effects of alumina sources (gibbsite, boehmite, and corundum) on melting behavior of high-level radioactive waste melter feed

ABSTRACTTypes of melter feed materials affect glass production rates. This study focuses on the effects of alumina sources on melting behavior of high-alumina high-level-waste melter feeds containing different alumina sources, namely, gibbsite, boehmite, and corundum. The heat flow from the glass melt to the cold cap, a floating layer of the reacting feed, is partially hindered by a foam layer at the bottom of the cold cap. Volume expansion tests and thermoanalytical methods revealed that a slow-melting feed with corundum foamed extensively, whereas a fast-melting feed with boehmite had a low reaction heat and produced less stable foam. The foam thickness, a critical factor for the rate of melting, estimated using the relationship between the heat conductivity and foam porosity was in reasonable agreement with experimental observation.

Conversion of nuclear waste to molten glass: Formation of porous amorphous alumina in a high-Al melter feed

The transition of Al phases in a simulated high-Al high-level nuclear waste melter feed heated at 5 K min−1 to 700 °C was investigated with transmission electron microscopy, 27Al nuclear magnetic resonance spectroscopy, the Brunauer-Emmett-Teller method, and X-ray diffraction. At temperatures between 300 and 500 °C, porous amorphous alumina formed from the dehydration of gibbsite, resulting in increased specific surface area of the feed (∼8 m2 g−1). The high-surface-area amorphous alumina formed in this manner could potentially stop salt migration in the cold cap during nuclear waste vitrification.

Effects of heating rate, quartz particle size, viscosity, and form of glass additives on high‐level waste melter feed volume expansion

AbstractNuclear waste can be vitrified by mixing it with glass‐forming and ‐modifying additives. The resulting feed is charged into an electric glass melter. To comprehend melting behavior of a high‐alumina melter feed, we monitored the volume expansion of pellets in response to heating at different heating rates. The feeds were prepared with different particle sizes of quartz (the major additive component) and with varied silica‐to‐fluxes ratio to investigate the glass melt viscosity effects. Also, we used additional melter feeds with additives premelted into glass frit. The volume of pellets was nearly constant at temperatures <600°C. After a short period of volume shrinkage at ~600°C‐700°C, foam generation produced massive volume expansion. The low heat conductivity of foam hinders the transfer of heat from molten glass to the reacting feed. The extent of foaming increased with faster heating and higher melt viscosity, and decreased with increasing size of quartz particles and fritting of the additives. Volume expansion data are needed for the mathematical modeling of the cold cap.

Conversion of Nuclear Waste to Molten Glass: Cold‐Cap Reactions in Crucible Tests

The feed‐to‐glass conversion, which comprises complex chemical reactions and phase transitions, occurs in the cold cap during nuclear waste vitrification. To investigate the conversion process, we analyzed heat‐treated samples of a simulated high‐level waste feed using X‐ray diffraction, electron probe microanalysis, leaching tests, and residual anion analysis. Feed dehydration, gas evolution, and borate phase formation occurred at temperatures below 700°C before the emerging glass‐forming melt was completely connected. Above 700°C, intermediate aluminosilicate phases and quartz particles gradually dissolved in the continuous borosilicate melt, which expanded with transient foam. Knowledge of the chemistry and physics of feed‐to‐glass conversion will help us control the conversion path by changing the melter feed makeup to maximize the glass production rate.

A Method for Determining Bulk Density, Material Density, and Porosity of Melter Feed During Nuclear Waste Vitrification

Glassmelting efficiency largely depends on heat transfer to reacting glass batch (melter feed), which in turn is influenced by the bulk density (ρb) and porosity (ϕ) of the reacting feed as functions of temperature (T). Neither ρb(T) nor ϕ(T) functions are readily accessible from direct measurements. For the determination of ρb, we monitored the profile area of heated feed pellets and calculated the pellet volume using numerical integration. For the determination of ϕ, we measured the material density of feeds quenched at various stages of conversion via pycnometry and then computed the feed density at heat‐treatment temperature using thermal expansion values of basic feed constituents.

Melter Feed Reactions at T ≤ 700°C for Nuclear Waste Vitrification

To understand feed‐to‐glass conversion for the vitrification of nuclear waste, we investigated batch reactions and phase transitions in a simulated nuclear waste glass melter feed heated at 5 K/min up to 700°C using optical microscopy, scanning electron microscopy with energy‐dispersive X‐ray spectroscopy, and X‐ray diffraction. To determine the content and composition of leachable phases, we performed leaching tests; the leachates were analyzed by inductively coupled plasma atomic emission spectroscopy. By 400°C, gibbsite and sodium borates lost water and converted to amorphous phase, whereas other metallic hydroxides dehydrated to oxides. Between 400°C and 700°C, carbonates decomposed before 500°C; amorphous aluminum oxide and calcium oxide reacted with the sodium borate and formed the more durable amorphous borate phase along with intermediate crystalline products; above 500°C, quartz began to dissolve, and hematite started to convert to trevorite.

Rhenium volatilization in waste glasses

We investigated volatilization of rhenium (Re), sulfur, cesium, and iodine during the course of conversion of high-level waste melter feed to glass and compared the results for Re volatilization with those in low-activity waste borosilicate glasses. Whereas Re did not volatilize from high-level waste feed heated at 5Kmin−1 until 1000°C, it began to volatilize from low-activity waste borosilicate glass feeds at ∼600°C, a temperature ∼200°C below the onset temperature of evaporation from pure KReO4. Below 800°C, perrhenate evaporation in low-activity waste melter feeds was enhanced by vigorous foaming and generation of gases from molten salts as they reacted with the glass-forming constituents. At high temperatures, when the glass-forming phase was consolidated, perrhenates were transported to the top surface of glass melt in bubbles, typically together with sulfates and halides. Based on the results of this study (to be considered preliminary at this stage), the high-level waste glass with less foaming and salts appears a promising medium for technetium immobilization.

High Chrome Refractory Characterization: Part I. Impact of Melt Reduction/Oxidation on the Corrosion Mechanism

High Cr2O3 containing Monofrax™ K‐3 is a robust refractory that is used in the fiberglass industry and used in radioactive waste glass melters worldwide. Monofrax™ K‐3 is tolerant of transition metal oxides but contains highly reduced solid solutions of spinels, that is, (Mg,Fe2+)(Al,Cr)2O3. Conversely, many of the waste feeds being processed are highly oxidizing. The K‐3 refractory corrosion was tested in sealed crucibles starting with slurried melter feed instead of prereacted glass called for by ASTM C621. Testing the refractory coupon during the feed‐to‐glass conversion exposes the refractory to the oxidizing and reducing species being released during vitrification, for example, NO3−, NO2−, CO2, CO, O2. Corrosion rates measured in highly oxidizing (high nitrate) feeds were ~1.8–2.8 times higher than those determined using prereacted glass or reduced feeds. Confirmatory corrosion rates were measured on Monofrax™ K‐3 coupons immersed in oxidizing feed in a 1/100th‐scale HLW pilot‐scale melter. Corrosion is heterogeneous or incongruent as Ni and Fe in the waste glass exchange with Mg and Al in the refractory. An insoluble NiFe2O4 spinel corrosion product is formed that can build up a protective layer along the refractory walls or spall and settle to the melter floor depending on melt pool convection/agitation.

The void fraction of melter feed during nuclear waste glass vitrification

ABSTRACTTo efficiently vitrify Hanford waste, the melting process (i.e., melter feed turning into waste glass) must be modeled and optimized. The rate of heat transfer to the melter feed in a waste glass melter, and thus the rate of melting, is strongly affected by the melter feed porosity, especially in the final stages where the glass-forming melt produces foam that insulates the feed from the molten glass. The volume expansion test allows the determination of the melter feed porosity as a function of temperature. This test measures the profile area of the feed pellet as it turns into glass. This contribution presents the calculation of the void fraction (porosity) of the melter feed as a function of temperature, heating rate, and material parameters. The process of finding the void fraction is described as well as results from the application of this process.

Effect of bubbles and silica dissolution on melter feed rheology during conversion to glass.

Nuclear-waste melter feeds are slurry mixtures of wastes with glass-forming and glass-modifying additives (unless prefabricated frits are used), which are converted to molten glass in a continuous electrical glass-melting furnace. The feeds gradually become continuous glass-forming melts. Initially, the melts contain dissolving refractory feed constituents that are suspended together with numerous gas bubbles. Eventually, the bubbles escape, and the melts homogenize and equilibrate. Knowledge of various physicochemical properties of the reacting melter feed is crucial for understanding the feed-to-glass conversion that occurs during melting. We studied the melter feed viscosity during heating and correlated it with the volume fractions of dissolving quartz (SiO2) particles and the gas phase. The measurements were performed with a rotating spindle rheometer on the melter feed heated at 5 K/min, starting at several different temperatures. The effects of undissolved quartz particles, gas bubbles, and compositional inhomogeneity on the melter feed viscosity were determined by fitting a linear relationship between the logarithm of viscosity and the volume fractions of suspended phases.

Toward Understanding the Effect of Low‐Activity Waste Glass Composition on Sulfur Solubility

The concentration of sulfur in Hanford low‐activity waste (LAW) glass melter feed will be maintained below the point where the salt accumulates on the melt surface. The allowable concentrations may range from near zero to over 2.05 wt% (of SO3 on a calcined oxide basis) depending on the composition of the melter feed and processing conditions. If the amount of sulfur exceeds the melt tolerance level, a molten salt will accumulate which may upset melter operations and potentially shorten the useful life of the melter. At the Hanford site, relatively conservative limits have traditionally been placed on sulfur loading in melter feed, which in turn significantly increases the amount of LAW glass that will be produced. Crucible‐scale sulfur solubility data and scaled melter sulfur tolerance data have been collected on simulated Hanford waste glasses over the last 15 years. These data were compiled and analyzed. An empirical model was developed to predict the solubility of SO3 in glass based on 253 simulated Hanford LAW glass compositions. This model represents the data well, accounting for over 85% of the variation in data, and was well validated. The model was also found to accurately predict the maximum amount of sulfur in melter feed that did not form a salt layer in 13 scaled melter tests of simulated LAW glasses. The model can be used to help estimate glass volumes and make informed decisions on process options (e.g., scale of supplemental LAW treatment facility, and pretreatment facility performance requirements). The model also gives quantitative estimates of component concentration effects on sulfur solubility. The components that increase sulfur solubility most are Li2O > V2O5 > CaO ≈ P2O5 > Na2O ≈ B2O3 > K2O. The components that decrease sulfur solubility most are Cl > Cr2O3 > Al2O3 > ZrO2 ≈ SnO2 > Others (i.e., the sum of minor components) ≈SiO2. The order of component effects is similar to previous literature data, in most cases.

Application of evolved gas analysis to cold-cap reactions of melter feeds for nuclear waste vitrification

Abstract In the vitrification of nuclear wastes, the melter feed (a mixture of nuclear waste and glass-forming and modifying additives) experiences multiple gas-evolving reactions in an electrical glass-melting furnace. We employed the thermogravimetry-gas chromatography–mass spectrometry (TGA-GC–MS) combination to perform evolved gas analysis (EGA). Along with identifying the gases evolved, we performed quantitative analysis relating the weighted sum of intensities of individual gases in linear proportion with the differential thermogravimetry. The proportionality coefficients were obtained by three methods based on the stoichiometry, least squares, and calibration. The linearity was shown to be a good first-order approximation, in spite of the complicated overlapping reactions.

Determination of Heat Conductivity and Thermal Diffusivity of Waste Glass Melter Feed: Extension to High Temperatures

The heat conductivity (λ) and the thermal diffusivity (a) of reacting glass batch, or melter feed, control the heat flux into and within the cold cap, a layer of reacting material floating on the pool of molten glass in an all-electric continuous waste glass melter. After previously estimating λ of melter feed at temperatures up to 680°C, we focus in this work on the λ(T) function at T > 680°C, at which the feed material becomes foamy. We used a customized experimental setup consisting of a large cylindrical crucible with an assembly of thermocouples, which monitored the evolution of the temperature field while the crucible with feed was heated at a constant rate from room temperature up to 1100°C. Approximating measured temperature profiles by polynomial functions, we used the energy equation to estimate the λ(T) approximation function, which we subsequently optimized using the finite-volume method combined with least-squares analysis. The heat conductivity increased as the temperature increased until the feed began to expand into foam, at which point the conductivity dropped. It began to increase again as the foam turned into a bubble-free glassmelt. We discuss the implications of this behavior for the mathematical modeling of the cold cap.

Conversion of Nuclear Waste into Nuclear Waste Glass: Experimental Investigation and Mathematical Modeling

The melter feed, slurry, or calcine charged on the top of a pool of molten glass forms a floating layer of reacting material called the cold cap. Between the cold-cap top, which is covered with boiling slurry, and its bottom, where bubbles separate it from molten glass, the temperature changes by up to 1000K. The processes that occur over this temperature interval within the cold cap include liberation of gases, conduction and consumption of heat, dissolution of quartz particles, formation and dissolution of intermediate crystalline phases, and generation of foam and gas cavities. These processes have been investigated using thermal analyses, optical and electronic microscopies, x-ray diffraction, as well as other techniques. Properties of the reacting feed, such as heat conductivity and density, were measured as functions of temperature. Investigating the structure of quenched cold caps produced in a laboratory-scale melter complemented the crucible studies. The cold cap consists of two main layers. The top layer contains solid particles dissolving in the glass-forming melt and open pores through which gases are escaping. The bottom layer contains bubbly melt or foam where bubbles coalesce into larger cavities that move sideways and release the gas to the atmosphere. The feed-to-glass conversion became sufficiently understood for representing the cold-cap processes via mathematical models. These models, which comprise heat transfer, mass transfer, and reaction kinetics models, have been developed with the final goal to relate feed parameters to the rate of glass melting.

Model for the conversion of nuclear waste melter feed to glass

The rate of batch-to-glass conversion is a primary concern for the vitrification of nuclear waste, as it directly influences the life cycle of the cleanup process. This study describes the development of an advanced model of the cold cap, which augments the previous model by further developments on the structure and the dynamics of the foam layer. The foam layer on the bottom of the cold cap consists of the primary foam, cavities, and the secondary foam, and forms an interface through which the heat is transferred to the cold cap. Other model enhancements include the behavior of intermediate crystalline phases and the dissolution of quartz particles. The model relates the melting rate to feed properties and melter conditions, such as the molten glass temperature, foaminess of the melt, or the heat fraction supplied to the cold cap from the plenum space. The model correctly predicts a 25% increase in melting rate when changing the alumina source in the melter feed from Al(OH)3 to AlO(OH). It is expected that this model will be incorporated in the full glass melter model as its integral component.

Determination of Temperature‐Dependent Heat Conductivity and Thermal Diffusivity of Waste Glass Melter Feed

The cold cap is a layer of reacting glass batch floating on the surface of melt in an all‐electric continuous glass melter. The heat needed for the conversion of the melter feed to molten glass must be transferred to and through the cold cap. Since the heat flux into the cold cap influences the rate of melting, the heat conductivity is a key property of the reacting feed. We designed an experimental setup consisting of a large cylindrical crucible with an assembly of thermocouples (TC) that monitors the evolution of the temperature field while the crucible is heated at a constant rate. Then we used two methods to calculate the heat conductivity and thermal diffusivity of the reacting feed: the approximation of the temperature field by polynomial functions and the finite‐volume method (FVM) coupled with least‐squares analysis. Up to 680°C, the heat conductivity of the reacting melter feed was represented by a linear function of temperature.