[en] Glass formulation development depends on an understanding of the effects of glass composition on its processibility and product quality. Such compositional effects on properties in turn depend on the microscopic structure of the glass. Historically, compositional effects on macroscopic properties have been explored empirically, e.g., by measuring viscosity at various glass compositions. The relationship of composition to structure has been studied by microstructural experimental methods. More recently, computer simulation has proved a fruitful complement to these more traditional methods of study. By simulating atomic interaction over a period of time using the molecular dynamics method, a direct picture of the glass structure and dynamics is obtained which can verify existing concepts as well as permit ''measurement'' of quantities inaccessible to experiment. Atomistic simulation can be of particular benefit in the development of waste glasses. As vitrification is being considered for an increasing variety of waste streams, process and product models are needed to formulate compositions for an extremely wide variety of elemental species and composition ranges. The demand for process and product models which can predict over such a diverse composition space requires mechanistic understanding of glass behavior; atomistic simulation is ideally suited for providing this understanding. Moreover, while simulation cannot completely eliminate the need for treatability studies, it can play a role in minimizing the experimentation on (and therefore contact handling of) such materials. This paper briefly reviews the molecular dynamics method, which is the primary atomistic simulation tool for studying glass structure. We then summarize the current state of glass simulation, emphasizing areas of importance for waste glass process/product modeling. At SRS, glass process and product models have been formulated in terms of glass structural concepts

[en] Whereas a prior conception of glass dissolution assumed a relatively rapid initial dissolution which then slowed to a smaller, fairly constant longer-term rate, some recent work suggests that these two stages are followed by a third phase of dissolution, in which the dissolution rate is accelerated with respect to what had previously been thought of as the final long-term rate. The goals of the present study are to compile experimental data which may have a bearing on this phenomena, and to provide an initial assessment of these data. The Savannah River Technology Center (SRTC) is contracted to develop glass formulation models for vitrification of Hanford low-level waste (LLW), in support of the Hanford Tank Waste Remediation System Technology Development Program. The phenomenon of an increase in corrosion rate, following a period characterized by a low corrosion rate, has been observed by a number of researchers on a number of waste glass compositions. Despite inherent ambiguities arising from SA/V (glass surface area to solution volume ratio) and other effects, valid comparisons can be made in which accelerated corrosion was observed in one test, but not in another. Some glass compositions do not appear to attain a plateau region; it may be that the observation of continued, non-negligible corrosion in these glasses represents a passage from the initial rate to the accelerated rate. The long-term corrosion is a function of the interaction between the glass and its environment, including the leaching solution and the surrounding materials. Reaction path modeling and stability field considerations have been used with some success to predict the changes in corrosion rate over time, due to these interactions. The accelerated corrosion phenomenon highlights the need for such integrated corrosion modeling and the scenario-specific nature of a particular glass composition's durability

[en] A variety of processing operations for high level radioactive waste occur in the High Level Waste Tanks in the H-Area of the Savannah River Site. Thermal design constraints exist on these processes, principally to limit the amount of corrosion inhibitor which must be added to protect the tank and cooling coil materials. The required amount of corrosion inhibitor, which must subsequently be removed prior to trapping the waste in borosilicate glass, increases exponentially with temperature over a fairly narrow range (some tens of degrees Celsius). For this reason, there is a need to model the thermal-hydraulic processes occurring in the waste tanks. A FORTRAN computer code, called ITP.FOR, was written to provide a simple but reasonably accurate analysis tool for plant operation design. The code was specifically written to model Tank 48, in which the In-Tank Precipitation (ITP) process of precipitating radioactive cesium will be initiated. Although the ITP.FOR code was written as personal-use software for scoping design calculations for Tank 48, the current intent is to extend the code's applicability to other H-Area waste tanks, and to certify the code in accordance with the NRTSC Quality Assurance requirements for critical-use software (1Q-34, 1991). Since the code's capabilities have generated some interest to date, the present report is presented as interim documentation of the code's mathematical models. This documentation will eventually be supplanted by the formal documentation of the expanded and benchmarked code

[en] Thermal design constraints exist on the processing operations in the High Level Waste (HLW) tanks of the Savannah River Site (SRS). A FORTRAN computer code was developed to provide a simple, fast, and reasonably accurate analysis tool for plant operation design. The code computes a lumped transient temperature for the liquid contents of a waste tank by modeling the liquid (slurry), the vapor space above it, the tank wall, and the cooling air outside of the tank. Results for a typical processing cycle of several months' duration can be obtained in 2--4 minutes CPU time on a VAX computer. This paper discusses the code's mathematical models, presents model results for a typical HLW process schedule, and compares the code predictions with operations data

[en] Two-phase flow regimes have been mapped for vertical, cocurrent downflow in a narrow annulus which is partially segmented by the presence of longitudinal ribs. This geometry and flow condition has application to the analysis of a Large-Break Loss of Coolant Accident (LB-LOCA) in the production K-Reactor at the Savannah River Site (SRS). The ribbed annular geometry, particularly the presence of non-sealing ribs, gives rise to some unique phenomenological features. The flow behavior is influenced by the partial segmentation of the annulus into four quadrants or subchannels. A random element is induced by the natural bowing of the slender tubes; the width of the azimuthal flow path between two subchannels at a given axial location is indeterminate, and can take on any value between zero and the maximum clearance of 7.6 x l0^-4 m. When the rib gap is zero at a given location, it is at a maximum 180P away at the same axial location. The range of rib gaps is spanned in a single test section, as it would be also in a reactor assembly. As a result of these effects, flow regime maps obtained by other researchers for downflow in annuli are not accurate for defining flow regimes in a ribbed annulus. Flow regime transitions similar to those noted by, e.g., Bamea, were observed; the locations of these transitions were displaced with respect to the transition equations derived by Bamea. Experimental bubble rise velocity measurements were also obtained in the same test section. The bubble rise velocities were much higher than expected from the theory developed for slug bubbles in tubes, unribbed annuli, and rectangular channels. An elliptical-cap bubble rises faster than a slug bubble of the same area. Large, slug-shaped bubbles injected into the test section were observed to reduce in size as they rose, due to interaction with a longitudinal rib. They thereby adopted a shape more like an elliptical-cap bubble, hence rising faster than the original slug bubble

[en] Processing in the Defense Waste Processing Facility (DWPF) is controlled by constraints on predicted properties of the product glass. One of these properties is chemical durability, which is measured as the response of various glass constituents to the seven-day Product Consistency Test (PCT) [1]. As currently implemented into the DWPF's Product Composition Control System (PCCS) the response of boron is taken as representative of all of the constituent responses, and control is in terms of the boron response. This response, in normalized units and in log scale, is taken to be a linear function of the glass's free energy of hydration, ΔG. ΔG is a parameter which represents the sum of influences on durability of the various glass oxide components. A generalized relationship between these two variables is documented in [2]. This relationship appears to underpredict releases for glasses in the so-called ''Purex'' range of compositions which comprises a worst-case DWPF operating range. Using a similar methodology as in [2], a linear regression specific to Purex compositions is developed herein

[en] A variety of vitrification technologies are being evaluated to immobilize radioactive and hazardous wastes following years of nuclear materials production throughout the Department of Energy (DOE) complex. The compositions and physical forms of these wastes are diverse ranging from inorganic sludges to organic liquids to heterogeneous debris. Melt and off-gas products can be very corrosive at the high temperatures required to melt many of these waste streams. Ensuring material durability is required to develop viable treatment processes. Corrosion testing of materials in some of the anticipated severe environments is an important aspect of the materials identification and selection process. Corrosion coupon tests on typical materials used in Joule heated melters were completed using glass compositions with high salt contents. The presence of chloride in the melts caused the most severe attack. In the metal alloys, oxidation was the predominant corrosion mechanism, while in the tested refractory material enhanced dissolution of the refractory into the glass was observed. Corrosion testing of numerous different refractory materials was performed in a plasma vitrification system using a surrogate heterogeneous debris waste. Extensive corrosion was observed in all tested materials

[en] The GAM-HEAT code was developed for heat transfer analyses associated with postulated Double Ended Guillotine Break Loss Of Coolant Accidents (DEGB LOCA) resulting in a drained reactor vessel. In these analyses the gamma radiation resulting from fission product decay constitutes the primary source of energy as a function of time. This energy is deposited into the various reactor components and is re- radiated as thermal energy. The code accounts for all radiant heat exchanges within and leaving the reactor enclosure. The SRS reactors constitute complex radiant exchange enclosures since there are many assemblies of various types within the primary enclosure and most of the assemblies themselves constitute enclosures. GAM-HEAT accounts for this complexity by processing externally generated view factors and connectivity matrices, and also accounts for convective, conductive, and advective heat exchanges. The code is applicable for many situations involving heat exchange between surfaces within a radiatively passive medium. The GAM-HEAT code has been exercised extensively for computing transient temperatures in SRS reactors with specific charges and control components. Results from these computations have been used to establish the need for and to evaluate hardware modifications designed to mitigate results of postulated accident scenarios, and to assist in the specification of safe reactor operating power limits. The code utilizes temperature dependence on material properties. The efficiency of the code has been enhanced by the use of an iterative equation solver. Verification of the code to date consists of comparisons with parallel efforts at Los Alamos National Laboratory and with similar efforts at Westinghouse Science and Technology Center in Pittsburgh, PA, and benchmarked using problems with known analytical or iterated solutions. All comparisons and tests yield results that indicate the GAM-HEAT code performs as intended

[en] Chemical process simulators have the potential to combine, in a single calculation, thermodynamic modeling of glass melters with vitrification system material and energy balances. These commercially-available computer codes have been developed for process simulation in the chemical and petroleum industries, i.e., processes which occur primarily in the vapor and liquid phases. As such, they are not oriented towards solid-liquid transition phenomena. They are capable of tracking solids, however, and contain algorithms for complex phase and chemical equilibrium similar to those which underlie codes used in metallurgical thermodynamics. One such chemical process simulator, ASPEN+trademark, is being used to model melter thermodynamics for low-level mixed waste (LLMW) vitrification systems. Such systems are now under development by the Savannah River Technology Center (SRTC) in cooperation with the DOE/Industrial Center for Vitrification Research at Clemson University. This paper reports glass melter model calculations using ASPEN+trademark and compares this tool to STGSOL, a modeling tool which is widely used for high-temperature thermodynamic equilibrium calculations for smelters and glass melters

[en] The US Department of Energy (DOE) weapons complex has numerous radioactive waste streams which cannot be easily treated with joule-heated vitrification systems. However, it appears these streams could be treated With certain robust, high-temperature, melter technologies. These technologies are based on the use of plasma torch, graphite arc, and induction heating sources. The Savannah River Technology Center (SRTC), with financial support from the Department of Energy, Office of Technology Development (OTD) and in conjunction with the sites within the DOE weapons complex, has been investigating high-temperature vitrification technologies for several years. This program has been a cooperative effort between a number of nearby Universities, specific sites within the DOE complex, commercial equipment suppliers and the All-Russian Research Institute of Chemical Technology. These robust vitrification systems appear to have advantages for the waste streams containing inorganic materials in combination with significant quantities of metals, organics, salts, or high temperature materials. Several high-temperature technologies were selected and will be evaluated and employed to develop supporting technology. A general overview of the SRTC ''High-Temperature Program'' will be provided