M. E. Stone and J. W. DuVall
Westinghouse Savannah River Company
Aiken, SC 29808

Available for a processing fee to U.S. Department of Energy and its contractors, in paper, from: U.S. Department of Energy, Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN 37831-0062,  phone: (865 ) 576-8401,  fax: (865) 576-5728,  email:  reports@adonis.osti.gov

Proposed flowsheets were developed to recover from credible processing errors as defined by NMSS Technical Task Request 99-NMSS/SE-006. Specifically, the following flowsheets were developed:

• A recovery method for processing a diluted feed.
• A recovery method from the addition of 50 wt% nitric acid in lieu of 8 wt% oxalic acid
• A recovery method from the addition of 0.1M oxalic acid in lieu of 8 wt% oxalic acid

All flowsheets were successfully demonstrated at the Cylindrical Induction Melter Pilot Facility, although the recovery from 0.1M oxalic acid addition was only marginally successful. Three recommendations were made as a result of the tests:

• Implement administrative controls to prevent inadvertent chemical addition in place of 8 wt % oxalic acid.
• Locate the decant tube as close to the center of the precipitator as possible.
• Revise the oxalic acid feed pump flow requirement to 100 to 500 ml/min with an accuracy of ± 5% of flowrate.

Keywords: Americium, Curium, Precipitation, Vitrification, Am/Cm

The americium and curium currently stored in F-canyon will be vitrified using a batch process that uses an oxalate precipitation process to remove nitrate from the melter feed. This process has been demonstrated and tested on lab-scale (1/70, 1/7, and 2/7 scale) as well as full scale processes. The feed will be a one molar nitric acid solution containing 100 grams per liter of solids (oxide basis) with a lanthanide/actinide content of 89.9 grams per liter (oxide basis). The steps in the baseline feed preparation process are:

Precipitation and Wash Addition

1. Add 6.91 liters of feed to precipitator
2. Start agitation at 450 RPM
3. Allow feed solution temperature to reach 30 degrees Celsius
4. Meter in 12.02 liters of 8 wt% oxalic acid at 350 milliliters per minute
5. Digest at 450 RPM for 10 minutes
6. Allow solids to settle for 30 minutes
7. Decant supernate to 2.5 liters remaining in precipitator
8. Add 6.91 liters of 0.1M oxalic acid wash solution
9. Mix for 10 minutes at 450 RPM
10. Allow solids to settle until ready to transfer batch to melter (minimum 30 minutes)

Decant, Drain, and Rinse Precipitator

1. Prime DI water rinse pump
2. Decant supernate to 2.5 liters remaining in precipitator
3. Position vessel over melter
4. Begin agitation at 450 RPM
5. Wait one minute
6. Open drain valve and start DI water rinse
7. Rinse for 20 seconds at 800 ml/min
8. Stop agitator one minute after drain valve opened
9. Close drain valve when dripping stops
10. Move precipitator from over melter

The proposed flowsheets for recovering from the upsets conditions are documented in SRT-AMC-99-0169 and have been previously evaluated with 1/14 scale tests (WSRC-TR-99-461). Acceptability of the proposed flowsheets was determined by the same requirements used during the Demonstration Runs (SRT-AMC-98-0300): no spillage or overflow during operation, recovery of lanthanides > 95%, holdup in the precipitator < 1%, and no spillage or overflow during transfer of oxalate to the melter.

Tests were conducted using a non-radioactive surrogate feed made up per SRT-AMC-98-0088. Except where noted, the parameters for the baseline process outlined above were used to perform the precipitation in the Coupled Precipitator One (CP-1) system. The baseline vitrification process was used to vitrify all runs in the 5" Cylindrical Induction Melter (CIM) pilot facility. The material balance for the melter is shown in Attachment 1.

Three full-scale tests were successfully conducted with diluted surrogate feeds. Processing dilute feed can be accomplished by producing one small batch for the melter or multiple batches can be processed in the precipitator to produce the equivalent of one baseline batch. If multiple batches are to be performed, the feed should be adjusted to 50,33, or 25 percent of the baseline concentration so that multiple batches combine to form the equivalent of one batch. Both processing options were successfully demonstrated during the testing. Feed diluted to 50% and 25% were processed using multiple batches while feed diluted to 77% was processed to produce one small batch.

The oxalic acid feed pump must be capable of accurately metering oxalic acid at flowrates other than the 350 ml/min baseline flowrate in order for multiple batch processing to be successful. SRTC recommends that the oxalic acid feed pump requirement be revised to 100 to 500 ml/min flowrates with an accuracy of ± 5% of flowrate.

Equal amounts of surrogate feed (SURR-1532) and 1M nitric acid were combined in the laboratory to provide the dilute feed. A "multiple batch" precipitation process that provided the equivalent of one baseline precipitation batch to the melter for vitrification was used to process the dilute feed. A precipitation was performed on 6.91 liters of surrogate feed, then a second batch of 6.91 liters of surrogate feed was added on top of the first batch prior to washing with 0.1M oxalic acid. The second batch was precipitated and the combined solids formed the equivalent of one baseline precipitation batch. The combined solids were then washed with 0.1M oxalic acid.

The amount of oxalic acid required per batch to precipitate the lanthanide/actinides from the solution is less than a baseline run. Therefore, the oxalic acid addition amount was reduced to maintain 0.275M excess oxalic acid concentration at the end of each batch and the oxalic acid rate of addition was slowed to 220 ml/min to maintain a precipitation time of 34 minutes and 20 seconds.

The settled solids volume after the second precipitation matched the settled solids volume from baseline runs: the imprint of the lower impeller was just visible and the solids left ~1" of the cone visible. No entrainment of solids was noted during the decant process and a sample of the combined decant solution indicated a process yield of greater than 99%. The amount of holdup in the vessel was visually determined to be similar to a baseline run. The amount of splashing during the precipitator rinse phase of the transfer to the melter was similar to baseline process runs. Vitrification of the material in the 5" CIM proceeded smoothly (CIM5 TTR Run #26) and resulted in a uniform glass product similar to baseline runs. A material balance on the melter indicated an overall yield of 101%.

Surrogate feed (SURR-1532) was mixed with 1M nitric acid in the laboratory to prepare the dilute feed. The baseline precipitation process was performed without adjustments in the oxalic acid addition rate or volume. No process upsets were noted during the precipitation process. The settled solids volume was below the lower agitator impeller. The transfer to the melter went smoothly with no differences in the amount of splashing noted during the rinse.

Vitrification in the 5" CIM went smoothly (CIM5 TTR Run #27) and produced a uniform glass with a nominal 45% lanthanide content (oxide basis). Drying time was increased slightly by the increased amount of free solution, but no adverse impact on the calcination or vitrification processes were noted. The material balance in the melter indicated an overall yield of 96%.

The 25% dilute feed was processed in the same manner as the 50% dilute feed, except that four precipitation batches were performed instead of two. The oxalic acid addition rate was slowed to 155 ml/min to maintain the 34 minute, 20 second addition time and 0.275M excess oxalic acid. The settled solids volume was slightly higher than baseline runs, as indicated by the imprint of the lower impeller not being visible after settling, but no entrainment was noted during the decant process and the transfer to the melter was normal. A sample of the combined decant indicated a precipitation yield of greater than 99%. Holdup in the precipitator was not increased by the process. Vitrification of the dilute feed proceeded smoothly (CIM5 TTR Run #28) and the glass that was produced was uniform in appearance. The melter material balance indicated an overall yield of 96%.

Addition of 0.1M oxalic acid in place of 8 wt% oxalic acid during the precipitation process resulted in partial precipitation of the lanthanides/actinides with a large number of fines during labscale tests. The process vessels (except the precipitator) will not be designed for slurry transfers, since a flowsheet has been developed to utilize solid oxalic acid crystals to precipitate the lanthanides/actinides and increase particle size to acceptable levels without transferring the batch from the precipitator. The first step in the recovery method is to add 50% nitric acid to the precipitator to produce a 1M nitric acid slurry. Solid oxalic acid dihydrate is then slowly added (~30 g/min) over 43 minutes to precipitate the lanthanides/actinides and provide 0.275M excess oxalic acid. After digesting for ten minutes, the solids are allowed to settle and the supernate decanted. The nitric acid addition results in higher nitrate levels in the supernate, therefore two 0.1M oxalic acid washes are performed to lower the nitrate amount in the melter feed.

Run CP-821 was conducted with surrogate feed SURR-1615 to demonstrate the recovery method. The settled solids volume during the run was higher than a baseline run and resulted in some entrainment of solids during the decanting process. A small amount of solids were pulled into the decant tank during the initial decant of the spent precipitator as well as the during each wash step. The combined decant was filtered to determine the amount of entrainment. The filtered solids weighed 38.5 grams wet, 26.8 grams after drying at 110 ° C, and 9.9 grams after calcination at 700 ° C, representing 1.6% of the amount of lanthanides fed to the precipitator. The precipitator walls hinder settling, producing a bowl shaped surface with the settled solids higher along the walls than in the center of the tank. The decant tube on CP-1 is positioned near the edge of the solids bed increasing the chances for entrainment, as shown in Figure 1. SRTC recommends positioning the decant tube as close to the center of the precipitator vessel as possible to reduce the risk of entrainment.

Transfer of the oxalate to the melter proceeded smoothly, but the increased solids volume caused splattering out of the melter during the drying process during CIM5A-21. A small amount of splatter is noted during all runs, but this splatter typically is caught by the melter walls and falls back into the bed during the calcination process. The solids height during the vitrification of CP-821 was approximately 1.5 inches higher in the melter than a typical process run, and a large portion of the splatter during the drying process was not contained in the melter. The amount of material associated with the splatter exiting the melter was small and did not affect the remainder of the vitrification process. The amount of splatter during the drying process is directly related to the amount of power applied to the melter. Reducing the drying power could be used to minimize the amount of splatter that is not contained in the melter.

The glass that was produced was uniform in appearance and poured smoothly from the melter. The melter material balance indicated a overall yield of 95%. Precipitation yield based on soluble losses to the decant solution was > 99%.

Observation of the precipitation process with the camera installed in CP-1 showed dramatic differences between a baseline run and CP-821. During a baseline run, large white clumps of precipitated lanthanides can be seen within 30 seconds of the start of the precipitation cycle. During CP-821, these clumps were not noted, providing an early indication of the upset condition. If the run was stopped prior to addition of a large amount of the 0.1M oxalic acid, recovery could be performed with addition of 8 wt% oxalic acid, likely dramatically decreasing the impacts on the process.

Although the recovery process only marginally met the requirements specified, no further tests are planned. Implementation of administrative controls (such as verification of 8 wt% oxalic acid density) must be utilized to prevent this upset condition.

Addition of concentrated nitric acid to the precipitator in place of 8 wt% oxalic acid would result in a nominal 6M nitric acid solution with a solids content of 35 grams per liter (oxide basis). The nitric acid concentration of the solution is too high for an oxalate precipitation process to be performed. Neutralization of the nitric acid with a metal hydroxide (such as sodium hydroxide) would result in impurity levels in the solution that would interfere with the oxalate precipitation process. A combination of dilution water and lanthanum hydroxide can be used to create a 1M nitric acid solution with the same lanthanide/actinide concentration as the typical feed solution. The adjusted solution can then be processed using the baseline flowsheet.

Run CP-822 was performed to demonstrate the recovery from a 50 wt% nitric acid addition. In order to reduce the amount of waste generated, 1.1 liters of surrogate feed were used in place of 6.91 liters and the initial process steps were performed in a laboratory beaker. 1.9 liters of 50 wt% nitric acid was added along with 5.3 liters of dilution water and the beaker contents mixed. 647.9 grams of lanthanum oxide were then added to neutralize the excess nitric acid. The resulting solution was only 74 grams per liter lanthanide oxides and was 1.7M nitric acid. Subsequent analysis revealed that the lanthanum oxide had absorbed atmospheric moisture and had mostly converted to lanthanum hydroxide prior to addition to the feed solution. An additional 107 grams of the lanthanum oxide/hydroxide were added to adjust the batch to proper concentrations. The neutralization reaction is exothermic and the solution temperature increased from 30 to 45 degrees Celsius during the initial addition of lanthanum.

The sample results on SURR-1619A indicated a nitric acid molarity of 1.185M and a feed concentration of 98.87 grams per liter of lanthanide oxides. The laboratory process resulted in approximately 8 liters of surrogate feed, which was transferred to the CIM/CP-1 pilot facility and processed as normal feed. All processes went smoothly with a precipitation yield of >99%, no entrainment during decanting, no holdup in the precipitator, and no splatter during the transfer to the melter. The vitrification process (CIM5A-22) did not experience the typical bed expansion due to the lower cerium content of the recovered surrogate solution and resulted in a uniform glass that poured well. A material balance on the melter indicated an overall yield of 97%.

The demonstration test was conducted with 1.1 liters of surrogate feed to limit the amount of adjusted solution created to one batch. If this process upset occurred in MPPF with 6.91 liters of surrogate and 12.02 liters of 50 wt% nitric acid, 32 liters of dilution water and 4.74 kilograms of lanthanum hydroxide would be required. The resulting solution would be approximately 51 liters and, therefore, adjustment cannot be accomplished in the precipitator vessel. After adjustment in the Sump Receipt Tank or Feed Tank, eight process runs would be required to vitrify the solution.

The camera system in CP-1 is capable of showing clumps being formed by the precipitation process within 30 seconds of starting the addition of 8 wt% oxalic acid. This system would provide an early indication that a process upset was occurring and allow the addition to be stopped and a recovery method developed that would result in a significantly reduced volume of solution after recovery. Implementation of adminstrative controls must be utilized to prevent the inadvertent addition of nitric acid in place of 8 wt% oxalic acid.

The flowsheets developed for recovery from upset conditions were successfully demonstrated. The dilute feed and 50 wt% nitric acid recovery runs met all requirements and no process abnormalities were noted during the runs. SRTC recommends that the flow requirement for the oxalic acid feed pump be revised to 100 to 500 ml/min with an accuracy of ± 5% of flowrate..

The recovery from 0.1M oxalic acid met all process requirements, but entrainment of solids during the decanting process and splatter during the drying process were noted during the run. SRTC recommends adminstrative controls be used to minimize the risk of inadvertently adding 0.1M oxalic acid or nitric acid in place of 8 wt% oxalic acid and that the decant tube be located as close to the center of the vessel as possible.

1. Laboratory Notebook WSRC-NB-2000-00068, ITS Research and Development 5" Cylindrical Induction Melter Development (CIM5A).
2. Laboratory Notebook WSRC-NB-99-00173, ITS Research and Development 5" Cylindrical Induction Melter Development 11/1/99 – 3/13/2000
3. Laboratory Notebook WSRC-NB-99-00135, Coupled Precipitator One.


Run Plans for 5" CIM Operation

4. T. M. Jones and D. C. Witt, CIM5A-21 TTR Item 2.05 – 0.1 Molar Oxalic Acid Precipitation Recovery Run (U), March 23, 2000.
5. T. M. Jones and D. C. Witt, CIM5A-22 TTR Item 2.06 – 50% Nitric Acid Precipitation Recovery Run (U), March 27, 2000.
6. T. M. Jones and D. C. Witt, CIM5 TTR Run # 26 Task 2.03 –50% Dilute Feed From Coupled Precipitator (U), December 8, 1999.
7. T. M. Jones and D. C. Witt, CIM5 TTR Run # 27 Task 3.06 –Off- normal Ln Loading (45 wt%) From 77 g/L Feed (U), December 10, 1999.
8. T. M. Jones and D. C. Witt, CIM5 TTR Run # 28 Task 2.03 –25% Dilute Feed From Coupled Precipitator (U), December 14, 1999.


Run Plans for CP-1 Operation

9. M. E. Stone, Coupled Precipitator Run Plan for Recovery from 50% Dilute Feed (U), SRT-AMC-99-0223, December 3, 1999.
10. M. E. Stone, Coupled Precipitator 77 g/L Feed for 45% Glass Run (U), SRT-AMC-99-0231, December 7, 1999.
11. M. E. Stone, Coupled Precipitator Run Plan for 25% Dilute Feed (U), SRT-AMC-99-0224, December 3, 1999.
12. M. E. Stone, Coupled Precipitator Run Plan for Recovery from 50% Nitric Acid Addition (U), SRT-AMC-99-0228, December 3, 1999.
13. M. E. Stone, Coupled Precipitator Run Plan for Recovery from 0.1M Oxalic Acid Addition (U), SRT-AMC-99-0227, December 3, 1999.


Other References

14. Task Technical and QA Plan, WSRC-RP-99-0610, July 12, 1999.
15. M. E. Stone, Calculations for Labscale Precipitation Tests (U), SRT-AMC-99-00169, August 25, 1999.
16. M. E. Stone, Run Plan for 0.1M Oxalic Acid Addition with Nitric Acid and Solid Oxalic Acid Addition (U), SRT-AMC-99-00207, October 25, 1999.
17. M. E. Stone and J. W. DuVall, Preliminary Evaluation of Am/Cm Melter Feed Preparation Process Upset Recovery Flowsheets (U), WSRC-TR-99-00461, December 7, 1999.
18. M. E. Stone, AM/CM Surrogate Makeup Sheet (U), SRT-AMC-98-00088, Revision 1, October 28, 1998.
19. J. E. Marra, Completion of 5" Cylindrical Induction Melter (CIM5) Integrated Demonstration Runs (U), SRT-AMC-98-0300, December 10, 1998.
20. Technical Task Request 99-NMSS/SE-006, Revision 0, April 28, 1999.