1.Introduction

Electrorefiner (ER) electrolyte is a product of a hightemperature electrochemical method developed at Idaho National Laboratory (INL)[1] for extracting plutonium and minor actinides from spent nuclear fuel to produce a feedstock for fast breeder reactor fuel.[2] LiCl-KCl eutectic salt electrolyte used in this trial contains active metal fission products and transuranic elements. Processing in the ER continues until the maximum allowable concentration of NaCl, fission products, or plutonium is attained in the salt. When a maximum concentration in the electrolyte is reached, all or a portion of the salt must be removed from the ER and replaced with pure, eutectic LiCl-KCl.[3] Electrolyte removed from the ER becomes a waste product. The ceramic waste form (CWF) process is used to immobilize the spent electrolyte. In the CWF process, zeolite-4A is utilized to absorb the LiCl-KCl eutectic salt in a heated mixing process to form salt-loaded zeolite (SLZ). A borosilicate glass frit is blended with the SLZ to serve as a binder, and the powder mixture is processed in a high temperature furnace to consolidate the mixture.[3] Hot isostatic press treatment has proven effective for consolidation in small-scale tests. For a full (production) scale process capable of producing waste forms weighing more than 400 kg, pressureless consolidation (PC), where the waste is heated in a furnace at atmospheric pressure, was found to be a practical alternative to hot isostatic press treatment for this material.[4]

Initial development of the PC process took place at Argonne National Laboratory – East. Process scale up and demonstrations up to approximately 140 kg continued at Argonne National Laboratory – West, now the Materials and Fuels Complex (MFC) at INL. CWF process development continued at the INL to develop full-scale waste forms that could be used for the treatment of Experimental Breeder Reator(EBR)-II spent fuel.

The CWF process is planned to undergo a qualification procedure before it is employed in the hot cell to treat ER salt. This qualification procedure requires that 5 full-scale (300–400 kg) CWFs are produced utilizing LiCl-KCl salt and must be processed with similar operating parameters. The final portion of the qualification procedure will verify the process in an INL hot cell using ER salt.

To date, three of the five full-scale pre-qualification CWFs have been produced at INL. The material for the first run consisted of LiCl-KCl eutectic salt with the ER simulant blended with zeolite-4A and borosilicate glass frit. The first full-scale CWF run also demonstrated equipment parameters with a basic CWF container design.

The principal focus of the second trial was to evaluate a preliminary multi-section CWF container design. Other areas of attention during the second trial were to finalize material preparation parameters, implement a height measurement gauge to measure the material as it consolidates in the CWF furnace, add audio recording to record when cracking occurs during the CWF cooldown process, and identify process improvements that can be made during future trials. The material for the second run was similar to the first run and consisted of LiCl-KCl eutectic salt, zeolite- 4A, and borosilicate glass frit.

A third batch of material produced consisted of LiCl- KCl eutectic salt, zeolite-4A, and borosilicate glass frit. Due to programmatic changes, this material was never utilized to produce a full-size CWF. Instead, a fourth batch of material was produced to create a nonradioactive salt material that more closely represented actual ER salt by doping LiCl-KCl eutectic salt. The third full-scale CWF trial utilized the LiCl-KCl eutectic salt-doped material. Some slight improvements were made to the canister design, but all other processing parameters were similar to the second CWF trial.

2.Experiment Description

The CWF process consists of material preparation followed by PC in a furnace. For trials completed to date, all processing has occurred outside of the hot cell using nonradioactive surrogates in place of ER salts. To process actual ER salts, radioactive salt handling and preparation occurs in the hot cell, while the preparation of the zeolite remains out of cell. Sized and dried zeolite needs to be transferred into the hot cell for blending and PC steps.

2.1.Material Preparation and Results

Four full-scale, pre-qualification material preparation runs were completed. Material preparation included separate salt and zeolite size-reduction steps, drying of the zeolite, blending of the salt and zeolite, and blending of the SLZ mixture with borosilicate glass frit. The resultant blend then was used as feed material for CWF consolidation. For all trials, a 75wt% SLZ and 25wt% borosilicate glass frit mixture was fed into the PC furnace.[5-7] Material preparation steps (size reduction, salt/zeolite blending, SLZ/borosilicate glass frit blending) were completed in four separate batches. An entire pre-qualification run of the four batches (~380 kg total) was consolidated in the CWF.

For the first three material preparation runs, LiCl-KCl eutectic salt was used as an ER salt simulant. For the fourth material run, the LiCl-KCl salt was doped with surrogate additives to produce a simulant more representative of radioactive ER salt. These surrogate salt additives were certified by Aldrich-APL, Urbana, IL, USA and were weighed using a calibrated Mettler Toledo analytical balance, Model XP1203S. The composition of doped LiCl-KCl salt is shown in Table 1.

To begin material preparation, zeolite and simulant salts were reduced to similar particle size ranges to improve salt occlusion and mixing during the salt/zeolite blending process. Salt, in the form of ~1-kg slugs or ~1-mm-diameter beads, was crushed and ground in an argon atmosphere glovebox to a particle size < 250 μm. Zeolite-4A beads that were ~1.5 mm in diameter were sized using a roller mill and reduced to a target particle size distribution of 45 to 250 μm. A Rotex sieve shaker was then used to separate zeolite particles outside of the 45- to 250-μm range.

Crushed and sized zeolite was dried in a mechanically fluidized dryer (MFD) to prevent moisture from affecting the simulated waste form and salt/zeolite blending. The MFD consists of a rotating retort inside of a hexagonal heater shroud (Figure 1).[8] Zeolite was loaded into the MFD retort with an average batch size of ~67.5 kg and was dried at 550°C, first under a nitrogen purge and then under vacuum.

The crushed dried zeolite and crushed salt simulant were blended at a ratio of 10wt% salt/90wt% zeolite in a V-mixer (Figure 2).[3,9,10] The V-mixer simultaneously rotates (~10 revolutions per minute) while it heats the SLZ mixture to 500°C using two immersion heaters and a mantle heater. As the temperature of the salt and zeolite increased, the salt melted and was occluded into the zeolite, forming SLZ. To estimate the fraction of chloride salt that remained unabsorbed on the surface of the zeolite particles, samples of SLZ were taken after each run and subjected to a free-chloride analysis using methods described previously.[3]

After the salt/zeolite blending, the SLZ mixture was allowed to cool and borosilicate glass frit was added to the V-mixer for a ratio of 75wt% SLZ/25wt% glass frit. The V-mixer was then run for at least 1 hour, without heat, to thoroughly mix the SLZ and the glass frit. The total weight of each batch of SLZ/glass frit was ~95 kg.

Zeolite drying results are given in Table 2. The standard heating cycle included a 12-hour hold at 500°C. A limit of 1.0wt% was selected with the target moisture content of 0.3wt% based on what has been achievable in previous experiments/ runs. Moisture concentrations were determined using a Karl Fisher titration-based moisture analyzer equipped with a sample furnace operating at 600°C.

One source of zeolite moisture variability could be reabsorption of moisture between the time when the sample is taken and the time it is analyzed. The glovebox initially used to prepare samples for moisture analysis had a tendency for high moisture levels. After the third trial, the glovebox was replaced with a glovebox that maintains moisture and oxygen levels below 0.1 ppm.

Free-chloride concentrations of salt/zeolite blending are shown in Table 3. Batch sizes varied from 27 to 101 kg. The reference temperature cycle includes a 15-hour hold at 500°C. The target free-chloride content, 0.5wt%, was based on previous experiments/runs.

Three of the 12 V-mixer batches had free-chloride contents greater than the 0.5wt% target. There seems to be some correlation between large batch sizes and high freechloride concentrations (Run 1, Batch 2 and Run 3, Batch 3), but it is difficult to explain why these three batches yielded free-chloride results higher than 0.5wt% because little else changed in the V-mixer run conditions.

2.2.Pressureless Consolidation

To consolidate the surrogate waste form, a cylindrical CWF container was first loaded into a large pit-type furnace (Figure 3). The material from an entire pre-qualification run (~380 kg of SLZ/glass frit mixture) was loaded into the CWF container for heating and consolidation in an argon atmosphere. The CWF furnace has three zones. Zones 1, 2, and 3 represent lower, middle, and upper heating element banks in the furnace, respectively. Each zone is controlled independently to achieve the desired temperatures in the furnace. For each PC trial, the CWF furnace temperature was initially ramped up until it reached 500°C, at which point the temperature was held constant for a prescribed period. The furnace was further heated until it reached 925°C, at which point the temperature was again held constant. As the furnace heated, the material consolidated to form the CWF.[11] Following the hold at 925°C, the heating elements automatically turned off and the CWF cooled. A forced-cooling system consisting of an argon circulation pump and heat exchanger may be activated to increase the rate of CWF cooling to room temperature.

2.2.1.Single-Section Container

It is important to note the PC process has only been completed for three of the four material qualification runs (material preparation serial numbers CWF_FS_001, CWF_FS_002, and CWF_FS_004). CWF_FS_001 and CWF_FS_002 contained LiCl-KCl eutectic salt as an ER salt simulant. CWF_FS_004, which was consolidated in the third CWF trial, contained the doped LiCl-KCl salt simulant. A single piece CWF container design was used for the first CWF trial, and a multi-piece container was used for the second and third trials. Table 4 summarizes the three CWF PC trials. For the first two trials, the heating and cooling parameters remained the same. For the third trial, the hold time at 500°C was increased, and the CWF was heated more slowly (between 500 and 925°C). Forced cooling was not used in the third CWF trial. Results for completed CWF trials are reported in the following sections.

Figure 4 shows the CWF container used in the first fullscale test run. The container is a one-piece design with a 20-inch-diameter, 89-inch-long, stainless-steel pipe with a 0.188-inch wall thickness as the main container body. A base plate was welded to the bottom of the pipe. A weight/ scraper plate floated on top of simulated waste material, applying pressure to keep the CWF top surface level as it consolidates within the furnace. A one-piece center guide/ lifting rod that attaches to the base plate guides the weight during the consolidation process and was used for lifting operations. An upper guide weldment kept the guide/lifting rod concentric during consolidation and lifting operations. The center guide/lifting rod design has several benefits:

The first CWF run resulted in a 60% volume reduction. The scraper plate/weight design functioned as designed, consolidating the waste CWF at the bottom of the container. The entirety of the 60% volume loss was the result of a reduction in the vertical dimension of the CWF. Following consolidation, the portion of the container that was left empty as a result of the consolidation was removed (Figure 5).

The results of the first CWF showed that there was room for improvement in the CWF container design. The goal of CWF studies is to implement a consolidation for actual radioactive ER salt. To take advantage of the consolidation provided by the PC process, the excess (empty) volume of the one-piece container would need to be removed. Consolidation operations for radioactive materials will take place in hot cell. Cutting the excess outer and lift/guide rod material is difficult and time consuming, particularly in a hot cell environment. Excess material would either be discarded or reused as another CWF container. Reusing the excess portion of the container in a subsequent consolidation run would limit the amount of material that could be loaded into the container initially and would still result in post-consolidation void space. Additionally, sealing a plate to the top and bottom of a sectioned container would prove more difficult in a hot cell when compared with sealing a cover plate to a cylindrical end that was machined outside of a hot cell.

2.2.2.Multiple Section Container

The second and third CWF consolidations used an improved multi-section container design assembled before, and disassembled after, the consolidation process. It was decided that the upper section of the container, which was left empty after the consolidation process, would be reused as the lower container section for the following run. From experience, it has been found that it is important to continue to shift the container sections down after each run because a thin film of the CWF material remains on the inside of the upper container after consolidation. If the upper section was continually reused, this layer of material would probably continue to build up, eventually interfering with the operation of the equipment.

Another design consideration included the final surrogate waste form. With the goal of qualifying the PC for disposal of actual ER salt, it was decided that two consolidated CWFs would need to fit within a cask liner with minimal empty space. The identified cask liner internal dimensions are 21-5/8 inch inner diameter × 86-1/4 inch downpipe, allowing for a maximum CWF container size of ~21.5 inch outside diameter × ~43 inches tall. A single container needs to have the capacity to hold up to 400 kg of CWF material in order to maximize the loading of the cask. To minimize the empty space in the CWF disposal container, it was decided that a three-piece, rather than a two-piece, design would be used. For a two-piece container, the 60% volume/height reduction would result in 10% (of preconsolidation material height) empty space in the bottom section of the CWF container.

The second CWF container is shown in Figure 6. This container consisted of the following components, as numbered in the figure:

Figure 6 also shows that the three-piece container is not filled to its entirety prior to consolidation. Material was loaded to fill the bottom two container pieces and partially fill the top piece. The loading height was chosen so that the consolidation process produces a CWF form that fits into the bottom section of the container with minimal void space.

In an effort to keep the components at a workable height, the container was assembled and disassembled section by section in the CWF furnace (Figure 7). A clamp was used to hold the lower section(s) in place while a center rod and outer section were installed or removed. To install or remove the outer container section, a Morse barrel lifter was modified to accommodate the CWF container size, which is slightly smaller than a standard 55-gallon drum. A standard electromechanical manipulator handle, identical to those used within the hot cell facilities at INL’s MFC, was installed on the center lifting rod to provide a lift attachment point for container assembly, disassembly, and removal of the CWF from the furnace.

3.Results

The first CWF trial served as a baseline for subsequent trials. As a result of lessons learned from this trial, modifications were made to collect additional data during the second and third trials. Table 5 is a summary of the data collected for the three CWF trials.

3.1.Trial #1

The first CWF trial consolidation took place in a singlesection CWF container and resulted in a 60% loss of volume during consolidation. During the cool down process, the sound of cracking could be heard. Following cool down, the CWF was sectioned for visual examination of the interior of the surrogate waste form. Three internal samples, disks ~1.5 inches thick, were obtained from the lower, middle, and upper portions of the CWF (Figure 8). Sectioning confirmed that large cracks had formed in the CWF.

The bulk surrogate CWF density (2.17 g/cc) and an immersion density measurement of a sample of the surrogate CWF (2.15 g/cc) showed good agreement with one another and were comparative to the reference CWF density previously established with laboratory-scale samples (2.34 g/ cc).[12] X-ray diffraction (XRD) analysis confirmed the presence of the expected major phase of sodalite with d spacings indicative of salt-loaded zeolite.

It is desirable to limit cracking within the CWF in order to produce a more durable waste form. Several modifications were made as a result of the first trial. Initially, only the temperatures of the lower, middle, and upper heating element banks of the CWF furnace were recorded. Following the first trial, four auxiliary thermocouples were added to record the temperature of the CWF during heating and cooling. Three of the thermocouples were attached to the lower section of the three-piece CWF container to measure the top, middle, and bottom of the outside of the CWF. The fourth thermocouple was attached halfway up the center rod of the lower CWF section to measure the internal temperature of the CWF.

Additionally, a Micro-Epsilon (ILR1181-30(01)) displacement laser, a type of time-of-flight laser sensor, was installed to measure the height reduction of the CWF during the consolidation process. Knowing exactly when the consolidation process was complete would allow for a shorter hold at 925°C, which would either shorten the overall process or allocate more time for the cool-down process. The laser projected a pulse through a quartz window at the top of the CWF furnace. The distance to the material was calculated by measuring the amount of time it took the reflected light to return to the sensor. A motorized turntable with a small cutout was used to minimize the amount of radiant heat transmitted to the laser during the furnace process (Figure 9).

Finally, it was decided that audio would be recorded real-time during subsequent CWF trials to determine when cracking occurred during cool down. High stresses in the CWF are likely to cause cracks. Sound data from subsequent CWF runs will be evaluated to determine the consistency of the cooling/cracking process. With knowledge of when cracking occurs, the cool down process could be modified to reduce the stress and the amount of cracks that occur in future CWF runs. For audio recording, a laptop with a software program called Loop Recorder Pro was used. This software continuously records sound on userdefined intervals and saves the files with the start/stop time and date in the file name (1-hour time intervals were used). Adobe Audition software was used to analyse the sound data after the cool down process was complete.

3.2.Trial #2

In the second trial, the multi-section container performed as anticipated. Assembly and disassembly took minimal effort. As in the first trial, the scraper weight was successful in consolidating waste in the vertical dimension, and the trial showed that the weight could be reused for subsequent runs. The high-temperature rope contained the CWF material as it consolidated but disintegrated during disassembly (Figure 10).

Figure 11 illustrates the waste consolidation temperature profile for the second furnace test run. Zones 1, 2, and 3 represent lower, middle, and upper heating element banks in the furnace, respectively. Each zone is controlled independently to achieve the desired temperatures in the furnace. The Outer Top thermocouple (TC) was attached to the outside of the lower CWF container section near the top of the consolidated CWF. The Inner Middle TC was attached to the center rod of the lower CWF section near the middle of the consolidated CWFand the Outer Middle TC was attached to the outside of the lower CWF container section near the middle of the consolidated CWF.

The furnace zone temperatures and the temperatures of the outside of the container tracked very closely during the initial ramp to 500°C. As the Aux TC (core temperature) temperature profile illustrates, the 40-hour hold at 500°C allowed the core temperature of the CWF material to get closer to the outer temperature of the material. This provided for a more uniform consolidation of the CWF material during the next temperature increase from 500 to 925°C. The 75-hour hold at 925°C was to ensure that the CWF material had completely consolidated before the cool-down portion of the process begins. The controlled cooling of the furnace utilized a blower that pulled hot argon from the furnace chamber through a heat exchanger and returned the cooled argon back to the furnace. A chiller was used to cool the atmosphere heat exchanger. The cool-down rate of the furnace can be controlled, to an extent, through the use of two temperature controllers. The first temperature controller will attempt to control the cool-down of the furnace according to a user-defined temperature profile. The second temperature controller monitors the inlet of the circulation blower to ensure it does not exceed 100°C, which limits how rapidly the furnace can be cooled.

Figure 12 illustrates the material height as it dropped during the consolidation process. The measurement system operated as expected. However, the laser was not scaled correctly, resulting in the last portion of the CWF material height drop, after ~65 hours, not being recorded. The laser was scaled to measure the entire height reduction in the third trial.

A total of 14 cracks were captured by the audio recording during the cool-down process, and a fifteenth crack was recorded after the CWF was removed from the furnace. As illustrated in Figure 13, the volume of the sound produced in the first eight cracks decreased as the cool-down process progressed. The remaining six cracks, which were recorded with the CWF in the furnace and the furnace lid in place, occurred after 250 hours and are not shown on Figure 13. The downward trend of the volume continued for the cracks not shown on Figure 13.

Each of the 1-hour sound recordings was analyzed graphically to quickly identify when cracks occurred. Figure 14 illustrates the graph of the left and right channels of the 1-hour time recording during which Crack 3 was identified.

The second CWF was sectioned in a manner similar to the first CWF. The density of a sample retrieved from the middle disk measured using helium pycnometry was found to be 2.125 g/cm³ +/- 0.007 g/cm³, which was comparable to density measurements of the first CWF was well as the previously established reference density for the CWF. As with the first trial, XRD analysis showed the expected major phase of sodalite to slightly lower d-spacings, which is indicative of salt-loaded zeolite. The pattern also shows the minor halite phase typical of the CWF (Figure 15).

3.3.Trial #3

In the third CWF trial, nonradioactive doped salt was consolidated in the three-piece container. Figure 16 illustrates the waste consolidation temperature and height reduction profile for the third furnace test run. As was observed for the second CWF, furnace zone temperatures and the outside temperatures of the CWF track very closely during ramp up. For the third CWF, several changes were made to the heating and cooling process as compared with the first and second CWFs:

As a result of these changes, the CWF core temperature tracked more closely to the outer CWF temperatures, resulting in a more uniform CWF temperature throughout heating and most of the cooling process (Figure 16) for the third CWF. Figure 16 shows that most of the CWF height loss occurs as the core CWF temperature is ramped up from 500 to 925°C.

As illustrated in Figure 17, a total of 18 cracks were captured by the audio recording during the cool-down process. The bulk surrogate CWF density (2.09 g/cc) showed good agreement with the previous two trials.

The CWF was sectioned, and a sample was taken from the middle disk (Figure 18). XRD analysis (Figure 19) confirmed the presence of the expected major phase of sodalite and a minor phase of halite. Electron micrograph images show the glass and sodalite in the CWF (Figure 20). A 7-day Product Consistency Test was performed in triplicate for samples from the CWF. Normalized mass losses of Si, Al, and B were less than those in the reference CWF. Li, K, Na, and Cl mass loss exceeded the reference CWF (Table 6). Table 7

4.Modeling

A model was created to simulate and predict the behavior of the CWF throughout the duration of the heat cycle. One key assumption to this model is that the CWF is axisymmetric. If the CWF is uniformly heated in the circumferential direction, the temperature gradient in that direction can be neglected and this model can be applied. Thus, the CWF is modeled in only two dimensions: the radial and axial dimensions.

4.1.Model Description

The CWF model determines the temperature profile of the CWF with respect to time by solving the heat diffusion equation for the temperature at a given node in the CWF for the current time step. This was done using the finite-difference method with six radial nodes and 48 vertical nodes. This results in the following form of the heat diffusion equation.

T represents temperature and q̇ represents the volumetric heat generation. The equation is discretized in the r and z directions and with respect to time (t). Thermal conductivity (k), density (ρ), and heat capacity (C_p) varied with time during the heat cycle. A convective boundary condition was applied to the upper, lower, and outer edges of the CWF. On the centerline of the CWF, an adiabatic boundary condition was applied due to symmetry.

The physical parameters that varied were modeled using empirical formulas. The thermal conductivity of CWF was modeled using the following empirical formula.

The subscripts i and ref represent the initial and reference values for k, p, and T. The reference values used in the model were 1.4 W·m^-1·K^-1 for K_ref , 2 g/cm³ for ρ_ref , and 840 K for T_ref . The density also varied during the heat cycle due to thermal variations and the consolidation reaction of zeolite to sodalite. Both effects were captured in the following equation.

ρ_th is the theoretical density of the CWF (2.15 g/cm³), X is the mass fraction of zeolite, and β is the coefficient of thermals expansion (10^-5 1/K). The change in the height of the CWF was calculated using the calculated density, total mass, and cross-sectional area of the CWF. The heat capacity of the CWF in J·kg^-1·K^-1 was calculated using the following empirical equation.

The volumetric heat generation has two sources: radioactive decay and zeolite-to-sodalite heat of reaction. In these tests, a nonradioactive surrogate material was used. Thus, the heat generated due to radioactive decay was set to zero. The heat of reaction for zeolite-to-sodalite reaction was modeled using the following formula.

The reaction from zeolite to sodalite was assumed to be irreversible, and its rate was found to follow the following empirical rate expression.

k_o is rate constant and was set to 10^-5 s^-1. α is a pseudo rate order and was found to be 5.32. This rate was used to calculate the change in zeolite composition, which would affect the density and other physical parameters.

4.2.Trial #2 Simulation Results

The simulated temperatures of the surface and core of the CWF at the mid-height of the consolidated CWF are plotted in Figure 21 along with the measured temperatures at the same position. For the first 50 hours, the temperature profile for both the surface and the core were a good match. However, the simulated core temperature lagged behind the measured core temperature after the first hold. Then the simulated core temperature accelerated and spiked, meeting the measured core temperature profile.

After the spike, the simulated core temperature decayed and then rose again to meet the measured temperature profiles. This spike was due to the heat of reaction generated when zeolite converted to sodalite. However, the absence of the spike in the measured data and the lag in the simulated core temperature seem to imply that thermal conductivity estimated by the empirical relation developed for the model is too low or some other significant means of heat transfer to the core is absent from the model.

The model was developed on the assumption of an adiabatic boundary condition at the center. However, there is a metal center rod, which allows for substantial heat conduction to the core. The model was modified to capture the effect of a center by adding additional nodes for stainless steel near the radial center. At the CWF-steel material interface, it was assumed that the thermal resistance would be negligible. The addition of the center rod removed the spike in the simulated core temperature but did not resolve the temperature lag. Additionally, the pre-consolidated CWF was assumed to be solid, but it was a powder/air mixture, and some convection could occur and contribute to more rapid heat transfer to the center.

To better understand the results of the model, a visualization program was developed in which temperature contours were plotted for an axisymmetric slice of the CWF over time, resulting in a movie. In this movie, the simulated temperature distribution of the CWF is clearly seen, as well as the consolidation process. Frames from the movie for the Trial #2 simulation without the center rod have been copied from this movie and are included in Figure 22. Each frame is a temperature contour of a crosssection of the CWF and is oriented such that the left-hand side is the center of the CWF, and the right hand side is the outer edge of the CWF. The first four frames from left to right share the same temperature scale, which is shown to the right of the fourth frame. The fifth frame is the same as the fourth frame but with a narrower temperature scale to show more detail.

Frame A shows the CWF at the end of the first ramp. The outer edge is heated, but the center is still quite cool. Frame B shows the CWF just before the second ramp, and the temperature distribution is more even. Frame C is during the second ramp before the spike. The CWF was then shorter due to the reaction to sodalite, which was nearly complete. Frames D and E were captured during the spike. The center temperature was then greater than temperature in the area between the outer edge and the core, which was presumably from the heat of the zeolite-to-sodalite reaction being trapped in the center.

4.3.Trial #3 Simulation Results

The third trial was also simulated. The simulated and measured temperatures for the surface and center of the CWF are plotted in Figure 23. Both the simulated and measured core temperatures were closer to the surface temperature due to the longer hold time, but the simulated temperature still spiked. The simulated and measured profiles during cool-down still matched but not as well as in Figure 21, despite using the same heat convection coefficients. This implies that forced cooling does not have a dramatic effect on the cooling rate.

5.Future Work

The work performed thus far shows that PC is a viable method to consolidate ER salt, but further changes are needed to adapt the process to the remote hot cell environment. The high-temperature rope seal, which disintegrated during disassembly, must either be improved such that it can withstand multiple heating and cooling cycles, or a method to replace the rope seal within the hot cell must be developed. Additionally, the clamp that was used to secure the lower CWF container sections, as it was during assembly and disassembly, needs to be redesigned to be secured within the hot cell. The height gauge has successfully measured the CWF height reduction throughout consolidation but needs to be modified with protective shielding to allow for use in the hot cell.

Modifications must also be made to allow for remote assembly and disassembly of the three-piece container. The J-groove locking mechanisms for the in-the-container design functioned adequately, but there are currently no methods of twisting such a large container in the hot cell. For this purpose, a lifting fixture (Figure 24) similar to the Morse barrel lifter was designed to utilize the twisting motion of the in-cell handling equipment to engage the dowel pins and the J-grooves through the use of a torque multiplier.

Equipment operating parameters, including hold times and heating and cooling rates, will continue to be refined during future test runs to produce more consistent CWF results. The method of monitoring and identifying loud cracks that occurred during the cool-down portion the CWF furnace run worked satisfactorily. However, adjustments to the cool-down rate to minimize cracking needs to be evaluated and future CWF test runs will provide this opportunity.

6.Conclusions

INL has developed a full-scale (~400 kg) CWF process for solidification of ER salts. The process consists of salt sizing and blending with zeolite and glass frit, followed by PC of the mixture by heating in a furnace. To date, three full-scale CWFs have been successfully produced with nonradioactive LiCl-KCl salt surrogates, including one trial using salt doped with additives to more closely represent the contents of actual ER salt. For all trials, a monolithic waste form was produced consisting of a sodalite major phase and a halite minor phase. Density measurements for the three trials are consistent with one another and are comparable to laboratory-scale reference CWF density measurements.

INL has demonstrated that the CWF process is scalable to production-sized batches and that only minor adjustments are needed to make the process suitable for hot cell processing of ER salt waste.

A model was developed to simulate the performance of the furnace to assist in troubleshooting and optimizing the heat cycle. A data visualization tool was also added to assist in the analysis and interpretation of the simulated results. The simulated and measured temperature profiles matched well except for a lag and spike in the core temperature. This may be from inaccuracies in the thermal conductivity or neglected modes of heat transfer that may be significant, such as heat transfer through the center rod or convection in the initial air/powder mixture. Further research is needed to identify and correct these deficiencies in the model.