Electrometallurgical Techniques for DOE Spent Fuel Treatment: Status Report on Argonne National Laboratory's R & D Activity as of Fall 1998 (1999)

At its June 25, 1998, meeting at Argonne National Laboratory-West (ANL-W), the committee heard three presentations related to the electrorefiners at ANL-W and Argonne National Laboratory-East (ANL-E). These electrorefiners are described in detail in the committee's previous status report. ¹

At ANL-W, work is in progress to develop uranium product specifications ² in accord with success criterion 2, goal 1 for the demonstration program.³ At the time of the June meeting, 64 driver fuel assemblies had been electrorefined in the Mark-IV electrorefiner.^4,5 It was reported that upwards of 97% of the uranium could be dissolved under conditions where ~70% of the zirconium was left undissolved in the cladding hulls. An anode-cathode module (ACM) for the Mark-V high-throughput electrorefiner (HTER), which does not use a cadmium pool and “scrapes” the electrodeposited uranium metal off the cathode and into a collection basket, was tested in the Mark-IV electrorefiner. Initial results were such as to require modification of the ACM design and thus of the proposed Mark-V ACM modules.

A computational model for the Mark-IV electrorefiner is being developed at ANL-W.⁶ The model's predictions appeared to be in good agreement with the experimental results.

Work at ANL-E on electrorefiner development and testing includes work in support of the Mark-IV HTER in operation in the Fuel Conditioning Facility at ANL-W.⁷ In addition, research and development on the Mark-V HTER at ANL-W are also ongoing.⁸

Initial testing of the 25-inch (inner diameter) HTER at ANL-E revealed several problems that required modification of the 25-inch system. ⁹ The 25-inch ACM employs 20 baskets arranged on four concentric rotating anodes interspersed between eight

stationary cathode surfaces, and it relies on ceramic “scrapers” to dislodge continuously electrodeposited uranium from these surfaces as the uranium accumulates. The dislodged uranium falls to the bottom of the cell and is captured by a bucket with a screen-covered bottom. The ability of these scrapers to remove the developing deposits is crucial to the continuous operation of the HTER because the accumulation of a significant mass of dense uranium deposit on the cathode surfaces leads to binding and stalling of the ACM anode rotor. In addition, this accumulated uranium can cause an electrical short between the anode and cathode, greatly decreasing the efficiency of the uranium recovery process.

In conjunction with work at ANL-W, ANL-E also modified the 10-inch (inner diameter) ACMs for the Mark-V electrorefiner to be used at ANL-W for processing experimental breeder reactor II (EBR-II) blanket fuel. The 10-inch ACM for use at ANL-W operates on the same basis as the 25-inch ACM but has only nine anode baskets and four cathode surfaces. ANL-E has attempted to mimic the operation of the 10-inch Mark-V HTER by using only the baskets in the two inner channels of the 25-inch HTER. ANL was able to achieve excellent uranium dissolution and zirconium retention by controlling the operating conditions of the HTER. However, desired objectives of 98% dissolution of uranium from the fuel segments with 80% retention of zirconium in the cladding hulls were not met when operating at anodic potentials less than 0.4 V.¹⁰

Unfortunately, two major problems were encountered. One was that a very dense deposit of uranium metal adhered to the cathodes in the HTER and made scraping impossible. A second was that the scrapers, which in the initial Mark-V design, and also on the 25-inch HTER at ANL-E, were mounted on the trailing edges of the anode baskets, caused hold-up of uranium metal in the space between the anode baskets. Additional scrapers were added and other modifications were made, including placing the scrapers on the leading edge of the baskets. In addition, current reversal (stripping) was carried out at periodic intervals to remove the high-density uranium metal from the cathode and deposit it on the anode basket. During the stripping cycle, these baskets were used as the cathode.

The problem with the hold-up of uranium has been a significant issue in the development of the ACM module. For example, the Argonne EMT Monthly Highlights for January 1998¹¹ describe efforts to reposition the ceramic scrapers from a trailing position to a leading position in order to provide more efficient removal of the electrodeposited uranium. The April 1998 EMT Monthly Highlights¹² describe continued problems, including jamming of the redesigned anode rotor by accumulated dense, adherent uranium deposits.

Work at ANL-E in the electrorefiner area supports the EBR-II demonstration project. In addition, this work is also used in support of the development of a larger HTER that could be used in the pyrometallurgical processing of other DOE spent fuels,

such as N-reactor fuel, that can be treated by EMT without any preliminary processing.¹³ Work on the 25-inch HTER also uncovered problems with placement of the ceramic scrapers. This HTER utilizes the same concept as the ACM in the Mark-V HTER at ANL-W, in that anode baskets rotate around a steel cathode from which the uranium metal is scraped off, by beryllia scrapers, into a collection basket under the electrodes.¹⁴ Scrapers were placed on the leading edge of the anode basket, rather than the trailing edge, and the number of scrapers was increased. High-density uranium deposits on the cathodes that could not be scraped off were initially removed by stripping (current reversal to dissolve the deposit), or by chemical dissolution. Some modification was also made in the placement of fuel in the anode baskets. The ACM for use in the Mark-V has three inner and six outer anode baskets. The operational mode consisted of one empty anode basket in each of the rings, which would, of course, decrease by two-ninths the amount of material that could be placed in the anode baskets. Also, a redesign of the 10-inch ACMs for the Mark-V has been carried out, in which space between the anode baskets was increased, as was the spacing between the anodes and cathodes.

During the site visit, the committee members were invited to view a 25-inch ACM that had stalled during operation. ANL-E personnel stated that the design modifications described above and changes in operating parameters do not prevent buildup of uranium on the cathodes, which leads to stalling of the anode rotor. This problem seems different from those noted in the past and appears to stem from the formation of uranium deposits on the anode rotor that are not oxidized from the rotor during the operation of the cell, possibly caused by a short in the rotor. If the Mark-V HTER is to achieve sustained operation with the 10-inch ACM, this problem must be solved.

It was noted by the committee during its tour of the research facilities at ANL-E that the salt content of the product from the ACM was significantly higher than that of the Mark-IV HTER cathode product. The Mark-IV HTER cathode product contains 10 wt % or less salt, whereas the cathode deposit produced by the Mark-V HTER contains 40 or more wt % salt. Even though the associated fission products contained in the driver and blanket fuels are quite different, there may be important consequences for downstream processes as a result of this additional salt content.

The current approach to development of the Mark-V ACM appears to be highly empirical, and the committee perceives that there is a lack of understanding about some of the fundamental electrochemical and mechanical factors affecting the ACM.

1. ANL should broaden its perspective regarding the Mark-V ACM by seeking information about the following:

   1. Physical, morphological, and mechanical characteristics (e.g., plasticity) of the uranium/salt mixture produced during HTER operating conditions;

   2. Electrochemical behavior of uranium in molten LiCl-KCl under HTER operating conditions; and

   3. Useful strategies from the metal electrowinning industry that can be applied to the uranium electrometallurgical process.

2. ANL should evaluate the potential impact of the higher salt content of the Mark-V HTER product on the performance of the cathode processor.

3. ANL should evaluate the effects of cathode surface roughness on the adhesion of the uranium deposit; other materials or metallic coatings that might reduce adhesion of uranium on the ACM cathode should be considered.

Efforts continue at ANL-W to evaluate the performance of both the cathode processor and the casting furnace unit processes. The time required for the cathode processor to reach the appropriate temperature had increased from around 8 to about 12 hours, a change believed to be due to increased loss of heat from the graphite liner. A decision was made to replace the liner rather than wait until it failed. After replacement of the liner, the time required for the cathode processor to come to temperature returned to the nominal 8 hours. This result was taken as evidence that the time delay in the heating cycle was in fact due to degradation of the graphite liner.

In order to melt the uranium and free the enclosed salt and/or cadmium, the cathode processor must be operated above the melting point of uranium metal (1132 °C). After running a number of experiments over the temperature range from 1150 to 1300 °C, ANL found that by operating the cathode processor at 1225 °C it was possible to achieve a reasonable separation of the salt while losing nominally 3 to 4% of the uranium. The mechanism for loss of uranium was found to be attributable to the reaction of liquid uranium with the zirconia crucible. The 3 to 4% loss of uranium is above ANL's target of = 1% loss of uranium stated earlier in the project.

To further ensure complete vaporization of the salt, studies were carried out in the range between the initial operating pressure of 10 torr down to 0.1 torr. The optimum pressure was found to be 1 torr. When the initial operating pressure is reduced from 10 to 1 torr, the boiling point of the salt is lowered by about 150 °C, to an estimated 800 °C.

To realize the desired 1-torr pressure during processing, ANL also modified its procedures so as to continue pulling a vacuum throughout the entire run rather than closing the valve to the vacuum system after the initial pump-down. Qualitatively, these changes in operating parameters and procedures resulted in more efficient distillation of fission products from the uranium. Consequently, higher-purity uranium product could be passed on from the cathode processor to the casting furnace. However, quantitative

data were not presented that showed the separation efficiency and mass balances of salt, cadmium, fission products, and uranium.

In terms of distillation effectiveness, ingot consolidation, and uranium loss minimization, both temperature and pressure have now been optimized for the cathode processor. These parameters, i.e., a vacuum of 1 torr and a minimum operating temperature of 1200 ° C, will be used as the operating reference parameters for the cathode processor during an upcoming 3-month repeatability demonstration, planned to address success criterion 1, goal 4.¹⁵

ANL-W is working with an outside vendor to produce larger beryllia crucibles needed to increase the throughput of the cathode processor. The scale-up of beryllia crucibles continues to be a problem at the outside vendor. The larger beryllia crucibles are failing mechanically, apparently due to thermal stresses. The current baseline material for both the cathode processor and the casting furnace is zirconia-coated graphite. The desire to use beryllia as the crucible material in the cathode processor stems from the marginal performance of the baseline material when uranium, process salt, and cadmium are present simultaneously in the cathode processor.¹⁶ Since cadmium is removed prior to the casting furnace step, beryllia crucibles are not required in the casting furnace.

As of October 1998, from irradiated EBR-II fuels, 30 batches of uranium and 7 batches of cladding hulls had been processed through the cathode processor. With the increase in the cathode processor batch size from 12 to 17 kg, it will be possible to process approximately 50% more material within the same cycle time.

Similarly, the batch size in the casting furnace was increased from 36 to 54 kg, representing a 50% increase. At the time of the October 1998 meeting, efforts included casting 29 batches or 692 kg of low-enrichment uranium and producing metal waste forms from six batches of spent cladding hulls.

Electrorefining of blanket material in the high-throughput Mark-V electrorefiner yields significantly more salt carry-over in the uranium product stream relative to that in the product stream of the Mark-IV. As a result, more salt will have to be removed by the cathode processor and potentially more fission products will remain in the uranium. It is not clear if these changes will necessitate new or more stringent handling and/or storage requirements for the uranium after it is cast into an ingot.