Hanford Tank Waste Chemistry

Transcription

1 TASK 3 Hanford Tank Waste Chemistry Process Chemistry Support for Hanford Waste Operations Planning J. S. Lindner, A. Antonyraj, L. T. Smith, T. Durve, and R. K. Toghiani Introduction Previous studies in these laboratories in support of Hanford tank farm operations have included extensive work on improving the understanding of salt cake waste thermodynamics, on evaluating the transport phenomena associated with high ionic strength supernatant (interstitial liquid) transfers, and on investigating the chemistry and physical parameters inherent to salt cake dissolution Additional calculations were used for the campaign to remediate 241-SY-101 and for calculations on 241-S ,24 The efforts have all involved the use, evaluation, and subsequent improvement of the primary site thermodynamic tool, the Environmental Simulation Program (ESP). Software validation has been a step wise process including an examination of the data called by the code, a comparison of ESP results with predictions of other thermodynamic models and a formal evaluation against actual salt cake dissolution experiments. Herting performed these later studies on core samples from twelve different DIAL

2 single-shell tanks. 25 Ranges in the chemical compositions of the different samples have provided for a stringent evaluation of the capability of the code to accurately model the thermodynamics of the wastes. Data deficiencies within the code have been identified and appropriate remedies were initiated including the development of a laboratory program to obtain needed solid-liquid equilibrium (solubility) data and the use of additional databases that better described the associated chemistry. The ultimate goal of this work is the development of a database specific to the Hanford salt cake wastes. Initial solubility experiments focused on chemical sub-systems present within the waste that lead to the formation of difficult to dissolve double salts such as natrophosphate (Na 7 F(PO 4 ) 2 19H 2 O, burkite (Na 6 (SO 4 ) 2 CO 3, and sodium fluoride sulfate, Na 3 FSO 4 ). A history of these experiments along with additional efforts related to this work is presented in Table 1. TABLE 1. Status of solubility experiments. Completed Reviewing In Progress FY 03 Subtask 1.1 Subtask 1.2 Subtask 1.3 Na-F-NO 3 -OH Na-NO 2 -CO 3 -OH Na-SO 4 -PO 4 -OH Na-Al-OH Na-F-SO 4 -OH Na-NO 3 -CO 3 -OH Na-SO 4 -NO 2 -OH Na-Al-XX-OH (1) Na-F-PO 4 -OH Na-F-PO 4 -NO 3 -OH Na-PO 4 -NO 3 -OH Na-SO 4 -CO 3 -OH Na-SO 4 -NO 3 -OH (1) where XX is a second anion typically present in the waste Additional partitioning discrepancies, noted in previous reports in this series, have been observed during DIAL experiments associated with saltwell transfers and simulant salt cake dissolution. Problems DIAL

3 with aluminum partitioning have been noted by Reynolds. 26 Results from the DIAL dissolution experiments are presented in Table 2. TABLE 2. Comparison of aqueous phase aluminum concentrations for simulant saltcakes. Salt Cake Simulant Hanford SC-1 Hanford SC-2 SRS -1 Al aq, (ICP) Al aq. (ESP) ESP calculations have also been used to support pilot-scale experiments at Florida International University. In addition, the DIAL waste chemistry program has received a request from Hanford to evaluate the application of ESP to evaporator operation. The FY 2003 program has been divided into two tasks with subsections. Task 1 concerns completion of the ESP database. Task 2 continues the use of the model to support pilot-scale experiments at Florida International University (FIU) and begins the study of the Hanford 242-S evaporator. Both of these tasks are described below and serve as a generic roadmap for all FY 2003 Results/Work Accomplished report sections. Task 1. ESP database enhancements The solubility measurements identified in Table 1 form the path for the continued development, testing, and distribution of the MSU- DIAL double salt database. Subtask 1.1 involves the work necessary to produce the database. A beta version of the database is scheduled for release in January 2003 and will contain the results of the solubility measurements and the associated equilibrium expression coefficients for those systems listed in column 1 of Table 1. The database will be checked against the experimental data and the results of selected dissolution experiments with the actual Hanford salt cakes. The database, the associated solubility measurement results, and test DIAL

4 streams will be distributed to Hanford engineers for evaluation and comment. Additional database expansion will continue with the solubilities in the remaining columns of the table. Work has been completed for the systems listed in column 2 of Table 1. These data will be analyzed and regressed and the results will be used to update the database. Additional chemistry objectives are contained with Subtasks 1.2 and 1.3. Subtask 1.2 is intended to allow completion of solubility needs identified during FY 2002 for quantification of sulfate interactions. Subtask 1.3 is aimed at developing a more rigorous understanding of aluminum chemistry. Part of the disagreement between the calculated and predicted aluminum anion concentrations cited in Table 2 is based on the direct use of the Gibbs free energy within ESP as opposed to an equilibrium description. Some work, by Reynolds, has resulted in the development of the Weslow database. 27 Calculations will be performed using the standard PUBLIC database and the private database to allow an assessment of the improved agreement. Initial calculations will begin with the Na-Al-OH system followed by increasing the complexity of the system to include other anion species that could affect the partitioning. In this regard, it is noted that solubilities for the Na-Al-OH-NO 3 system have recently been obtained. 28 These data will be examined and compared to laboratory and literature results obtained previously. If additional literature data are not available for aluminum with the other anions of interest, solubility envelopes will be developed based on pure component solubilities. Prioritization will be based on an attempt to determine the strength of the anion-anion interactions as indicated from recent laboratory experiments on simulant saltcake dissolution. The MSU-DIAL double salt database will be upgraded at intervals associated with user feedback and with the completion of series solubility studies. The first stage of the database will include those systems in column 1 of Table 1. The next upgrade will be from user DIAL

5 feedback and from the fitting of the envelopes for column 2. The process will be repeated for those systems in columns 3 and finally for the aluminum chemistry in column 4. Task 2. Application of ESP to pilot-scale simulant studies and evaporator operations Extensive use of the ESP model and involved comparisons to experimental results has indicated that the solid liquid equilibria of a number of waste constituents is properly described by ESP. Very good agreement (within 5%) has been observed for nitrate anion during salt cake dissolution experiments with the core samples and with the simulant saltcakes. Similar comments apply for chloride and nitrite anion. Comparison of the model results with experiments on surrogate compositions is therefore expected to be satisfactory in many regards. In other instances, the limitations of the model, and thus, the improvements that are required for wider use, have not been established. This task is separated into two subtasks that will support on-going pilot-scale experiments and will also address the applicability of the model for waste stream evaporators. Subtask 2.1 provides thermodynamic calculation support for pilot-scale experiments to be conducted at FIU. Here ESP calculations will be used to determine parameters of interest to large-scale salt cake dissolution experiments. The calculations will track all of the chemical constituents of the simulants employed from the preparation of the salt cake through the entire dissolution process. Physical properties such as volume fractions, densities, and supernatant viscosities will also be obtained and permit direct comparisons with the results from the pilot-scale tests. The best available databases will be used and the calculations will represent the diluent volumes employed. Calculations will be performed on an as-requested basis and discussions will be held to enable FIU personnel to take full advantage of the data and for aiding comparisons and results interpretation. DIAL

6 Subtask 2.2 will evaluate the use of ESP for modeling of Hanford s 242-A evaporator/crystallizer. Hanford engineers are interested in modeling the processes occurring, and streams resulting from the 242-A evaporator. 26 The work will commence with a literature study on the conditions employed within the evaporator and on assessing the results of previous campaigns. ESP calculations will be made on simple systems. Efforts will then be aimed at comparing model predictions with experimental data provided by the site. Sequential evaluation of the use of ESP will also include assessing the numerous unit operations within the code. Data gaps will be identified and slurry compositions, solids loading, and waste volume reductions will be compared. Work Accomplished Task 1. ESP database enhancements Subtask 1.1. Preparation and distribution of the MSU-DIAL double salt database During this quarter, significant progress was made towards the release of the beta test version of the double salt database. With the significant changes made to sodium nitrate and sodium sulfate species in the PUBLIC database accompanying Version 6.6 of ESP, it was decided, in consultation with users from the Hanford and SRS sites, to continue development of the double salt database using Version 6.5. This version is most widely used by the engineers and scientists at both Hanford and SRS. The PUBLIC database of version 6.5 is used as the platform for all regressions of experimental data gathered in our laboratory and from the literature. The changes in pure component information for sodium sulfate species (anhydrous and decahydrate) and for sodium nitrate are detailed in Subtask 1.2. Previous calculations comparing ESP V6.5 with the lattice model of M. Ally and with data from the International DIAL

7 Critical Tables were updated using ESP V6.6. These comparisons demonstrated the significant changes that have occurred in the description of pure sodium nitrate within the PUBLIC databases from V6.5 and V6.6. Shown in Figure 21 is a comparison at 25 C. At low Vapor Pressure above NaNO3 (aq), (mm Hg) Comparison at 298 K ESP Prediction, V6.5 Experimental Data from International Critical Tables M. Ally Lattice Model Data from Properties of Aqueous Solutions of Electrolytes ESP Prediction, V Molality or Ionic Strength FIGURE 21. Vapor pressure over sodium nitrate solutions at 25 C. Comparison of ESP PUBLIC databases V6.5 and V6.6. ionic strength, the agreement between both versions of ESP (V6.5 and V6.6), the literature data and the lattice model, is quite good. Differences are evident, however, once the ionic strength reaches a value greater than four. At higher ionic strengths, the predictions from V6.6 show significant deviation from those of V6.5, the literature data, and the lattice model. An additional difficulty was noted when using V6.6, in that prediction of the vapor pressure over solution at ionic strengths greater than 8.5 were not possible, because of the appearance of solids. With V6.5, it was possible to predict the vapor pressure curve DIAL

8 over the full range of literature data (up to approximate 12 m). The region in which discrepancies become significant is the region that is of interest to the Hanford site. At 100 C, similar difficulties are encountered. The comparison between database versions from ESP V6.5 and V6.6 are shown in Figure 22 for 100 C. Again, the V6.6 predictions show significant deviation from the literature data, the lattice model and the V6.5 predictions, particularly at molalities greater than five. Predictions at molalities greater that approximately 15 m were not possible, although the literature data extend past 20 m nitrate. These comparisons have been submitted to OLI Systems, Inc., so that they are aware of the discrepancy between V6.5 and V Comparison at 373 K Vapor Pressure above NaN03 (aq), (mm Hg) ESP Prediction, V6.5 Experimental Data from International Critical Tables M. Ally Lattice Model Data from Properties of Aqueous Solutions of Electrolytes ESP Prediction, V Molality or Ionic Strength FIGURE 22. Vapor pressure over sodium nitrate solutions at 100 C. Comparison of ESP PUBLIC databases V6.5 and V6.6. The development of the double salt database is comprised of two primary tasks: updating the solubility information for the double salts DIAL

9 and identification of those interactions that are not properly described by the existing database. The PUBLIC database contains only cationanion interactions. Thus, the focus in regression of our data is on like ion interactions (anion-anion) to improve the fit of experimental data. The solubility information currently available in the ESP PUBLIC database is based on literature data over a limited temperature range. The experimental data gathered in our laboratory for both the sodium fluoride sulfate double salt and for the sodium fluoride phosphate double salt extend the temperature range from 25 C to 50 C. In the case of the sodium fluoride sulfate double salt, previous literature data span a temperature range from 20 to 35 C, and the sodium fluoride sulfate solubility does not currently carry temperature dependence in the PUBLIC database fit. The temperature dependence in this system as evidenced in our experimental data is weak, but our experimental data demonstrate that the influence of hydroxide loading in the equilibrated solution is significant. The presence of hydroxide significantly suppresses the solubility of the double salt. Common to these systems is the fluoride ion. Available literature data for the common ion ternary system, NaF-NaOH-H 2 O, were used to establish the binary interaction for the fluoride ion with the hydroxide ion (F-OH interaction). This interaction is of the BROMLEY type (as coded within ESP). Up to nine parameters may be fit to describe the interaction. Literature data over the temperature range from 0 to 100 C and hydroxide loadings ranging from 0 m to 5.6 m were regressed (40 data points). Regression using three parameters (b1, c1, d1) was accomplished and predicted results compared to the literature data. This comparison is shown in Figure 23. Series 2 on the figure represents a perfect fit (experimental fluoride molality equal to predicted fluoride molality). The predicted points are plotted as a function of the experimental fluoride molality, and are designated by temperature. It is evident from this comparison that the temperature functionality of the data is not properly described using only three parameters. This is not surprising because these terms carry no temperature dependence. Addition of the parameters (b2, c2, d2) allows DIAL

10 better representation of the temperature dependence in the experimental data. Regression with six parameters was conducted (b1, b2, c1, c2, d1, d2) and the comparison of predicted fluoride molality to experimental molality is shown in Figure 24 (Series 6 in this figure represents a perfect fit of the data). The majority of the temperature dependence is described through the addition of these three parameters. Figure 25 provides a residual plot (experimental - predicted) showing the improved prediction possible through the use of six parameters to describe the fluoride-hydroxide interaction. Figure 26 provides a comparison of the ESP predictions for the common ion ternary system, NaF-NaOH-H 2 O, as a function of temperature and hydroxide loading. These predictions were performed using the results from the regression of the fluoride-hydroxide interaction. The temperature and hydroxide dependence of the available experimental data are reproduced by this model Parameters (b1,c1,d1) Fluoride in Solution (m), Calculated Series Fluoride in Solution (m), Experimental FIGURE 23. Regression results (three parameters) - fitting of fluoride-hydroxide interaction from common ion ternary data. Series 2 represents a 45 line. DIAL

11 Fluoride in Solution (m), Calculated Series 6 6 Parameters (b1,b2,c1,c2,d1,d2) Fluoride in Solution (m), Experimental FIGURE 24. Regression results (six parameters) - fitting of fluoride-hydroxide interaction from common ion ternary data. Series 6 represents a 45 line Fluoride Molality Residual Plot (experimental - predicted) 3 Parameters 6 Parameters 0.10 Residual Fluoride in Solution (m) FIGURE 25. Residual plot comparing three parameter and six parameter description of the fluoride-hydroxide interaction. DIAL

12 Fluoride in Solution (m) Temperature, T ( C) Experimental, OH = 0.0 m Experimental, OH = m Experimental, OH = m Experimental, OH = m ESP, OH = 0.0 m ESP, OH = m ESP, OH = m ESP, OH = m FIGURE 26. Predicted solubility for sodium fluoride as a function of temperature and of hydroxide loading using regression results. Using this description of the fluoride-hydroxide interaction, available experimental data from our laboratory were combined with available literature data for the sodium fluoride sulfate double salt. The experimental data in our laboratory were taken as a function of both temperature and hydroxide loading. The solubility for the double salt was fit as a function of temperature. The KFIT expression used within ESP is of the form A + B/T + CT + DT 2. Three parameters (A,B,C) were obtained through regression of the data. Preliminary regression results are shown in Figure 27 for the experimental and predicted fluoride concentrations in solution. A total of 33 data points were included in the regression. As evidenced by the strong correlation between the experimental and predicted fluoride concentrations, the majority of the experimental data are well described by the threeparameter KFIT expression coupled with the fluoride-hydroxide interaction. Non-convergence was noted for four data points. These DIAL

13 data points were a subset of the solubility data for the double salt in aqueous solution, as measured in our laboratory. An investigation into the reasons for the non-convergence is underway. 0.8 Fluoride in Solution (m), Calculated Converged Points Non-Converged Points Fluoride in Solution (m), Experimental FIGURE 27. Regression results for sodium fluoride sulfate double salt. A beta test version of the double salt database was released to ESP users at the Hanford and SRS sites. This test version contained updated solubility information for the sodium fluoride phosphate and sodium fluoride sulfate double salts. Comments have been received regarding the database from individuals at the Hanford site and these comments are being considered in our revisions of the double salt database. Subtask 1.2. Additional solubility measurements on sulfate and phosphate systems ESP 6.6 modeling of the sulfate-nitrate, sulfate-phosphate, and sulfate-nitrite systems was completed. In the V6.6 PUBLIC database, DIAL

14 significant changes were made to the following species: sodium sulfate decahydrate, anhydrous sodium sulfate, and sodium nitrate. For these species, both the enthalpy of formation at 25 C and the entropy in the reference state (25 C) were changed, with the entropy values changing by approximately 20%. These changes caused prediction of much larger solubilities for these pure component species than observed with the V6.5 PUBLIC database or found in existing literature data. OLI was notified of these problems and is in the process of preparing necessary adjustments. Solubility experiments have begun with each of these systems. A sample plot of experimental values versus ESP 6.6 predictions is shown in Figure Experimental Data ESP SO (m) NO (m) 3 FIGURE 28. Experimental values versus ESP predictions for the NO 3 /SO 4 System in 1-m NaOH. These preliminary results indicated a problem with the pure component sodium sulfate solubility in aqueous solution at 25 C and in 1- DIAL

15 m NaOH at 25 C. Using data obtained from literature values, a phase diagram was then constructed and is shown in Figure Temperature, T ( C) Weight % Na2SO4 FIGURE 29. Phase diagram of sodium sulfate and water. The first region (starting at ~ 5% wt. Na 2 SO 4 ) represents those overall compositions where the saturated solution (indicated by the curve with literature data marked as triangles) and solid phase Na 2 SO 4 10H 2 O are in equilibrium. The region below the first region represents those overall compositions where saturated solution (indicated by the curve with literature data marked as squares) and solid phase Na 2 SO 4 7H 2 O are in equilibrium. The sodium sulfate heptahydrate (Na 2 SO 4 7H 2 O) is a metastable species. At higher temperatures, the equilibrium of solid and liquid phases is between saturated solution (indicated by the black curve with literature data marked by triangles) and anhydrous sodium sulfate. The transition from sodium sulfate decahydrate to anhydrous sodium sulfate occurs at a temperature of 32.4 C in aqueous solution. This transition temperature is sup- DIAL

16 pressed in higher ionic strength solutions. At 25 C, the stable solid phase formed is dependent on the overall weight percent of sodium sulfate in the prepared solution. Depending on this overall weight percentage, the solid phase can be either sodium sulfate decahydrate (lower weight percentage of sodium sulfate) or sodium sulfate heptahydrate (higher weight percentages of sodium sulfate). The diagram also demonstrates how the equilibrium points for solid Na 2 SO 4 10H 2 O, solid Na 2 SO 4 7H 2 O and solid anhydrous Na 2 SO 4 have only slight differences in weight percent over a very small ( 8 C) temperature range difference. A subtle change in temperature and weight percent affects the solid phase formed. When the solubility experiments were performed at room temperature, the actual room temperature ranged from 23 C to 26 C. Preliminary results for the pure component formed the metastable solid Na 2 SO 4 7H 2 O. These experiments are being repeated keeping the temperature maintained at 25 C and the overall composition of sodium sulfate maintained at or below 25% by weight. This will insure that the solid phase formed is the sodium sulfate decahydrate species. Subtask 1.3. Improvement of ESP model predictions for aluminum species ESP severely under-predicts the experimental aqueous phase concentrations obtained as demonstrated in Table 2. Efforts have begun by examining the literature and beginning calculations for the Na-Al- OH system. Some of these calculations have already been performed. Figure 30 provides a comparison of the predicted aqueous phase concentrations of Al aq and OH- species to the experimental data of Russell. 21 Here version 6.5 of the ESP model using the PUBLIC and Weslow databases were examined. Changes in the specific interactions for Na, Al, and OH were not made between versions 6.5 and 6.6. DIAL

17 and Weslow 40 C 6.5 and Weslow 70 C Russell 40 C Russell 70 C 5 Al (m) OH -(m) FIGURE 30. Comparison of model predictions with the experimental data of Russell. 21 The results suggest that the model using the PUBLIC and Weslow databases does an adequate job at typical waste temperatures and at low to moderate hydroxide loadings of 5 m. Differences between the results given in Table 2 and those shown in Figure 30 suggest that additional interaction terms are needed for properly describing salt cake wastes. Ongoing studies include Na-Al-XX-OH systems (where XX is CO 3 2-, SO 4 2-, NO 2-, or PO 4-3 ) modeling and laboratory experiments for comparison. DIAL

18 Task 2. Application of ESP to pilot-scale simulant studies and evaporator operations Subtask 2.1. Perform ESP calculations in support of pilot-scale experiments ESP calculations in support of FIU pilot-scale experiments have been reported previously. These calculations were performed on both Hanford saltcake simulants and SRS saltcake simulants. No requests for additional calculations have been received at this time. Subtask 2.2. Evaluation of the use of ESP for waste evaporation modeling Background information has been gathered. Flow sheeting options are being evaluated within the ESP code to support evaporator 242-A operations. Project Status Significant progress was made during this quarter. Discrepancies between the PUBLIC database information for sodium nitrate, sodium sulfate decahydrate and anhydrous sodium sulfate were identified between V6.5 and V6.6 of ESP. Discrepancies for sodium nitrate have been forwarded to OLI Systems for consideration. V6.5 will serve as the basis for continued development of the double salt database and for predictive calculations using ESP. Regressions of the experimental data obtained in the laboratory continue and will provide the basis for updating the double salt database. A beta test version of the double salt database was released. Solubility measurements for the sodium sulfate nitrate system, sodium sulfate nitrite system and sodium sulfate phosphate systems are underway. A revised experimental procedure to ensure equilibration at 25 C has been implemented and equipment ordered to provide constant temperature conditions. Comparison of aluminum speciation using currently available databases has been accomplished. DIAL

19 Conclusions Identification of the discrepancies between V6.5 and V6.6 database descriptions of species important at Hanford and Savannah River established the need to use V6.5 as the platform for further development and deployment of the double salt database. All project tasks are on schedule. Work Planned Regressions of available experimental data for inclusion in the double salt database will continue. The laboratory solubility studies for the sulfate-nitrate, sulfate-nitrite, and sulfate-phosphate systems should be complete by the end of the first quarter of A request for calculations to support FIU has been received and will be addressed during the coming quarter. References 20. R.K. Toghiani and J.S. Lindner Saltcake Dissolution Modeling, FY 2000 Status Report. DIAL TR00-1. Diagnostic Instrumentation and Analysis Laboratory, Mississippi State University, Mississippi State, MS J.S. Lindner, T. Durve, V. Raju, and R.K. Toghiani Feed Stability and Transport. DIAL Technical Progress Report 40395R13. Diagnostic Instrumentation and Analysis Laboratory, Mississippi State University, Mississippi State, MS. 22. J.S. Lindner, A. Antonyraj, T. Durve, and R.K. Toghiani Solids Formation. DIAL Technical Progress Report 40395R16. Diagnostic Instrumentation and Analysis Laboratory, Mississippi State University, Mississippi State, MS. 23. J.S. Lindner and R.K. Toghiani Thermodynamic Simulation of Tank 241-SY-101 Dissolution, Part 3: Crust Solids Dissolution Modeling and Associated Gas Release. DIAL TR Diagnostic Instru- DIAL

20 mentation and Analysis laboratory, Mississippi State University, Mississippi State, MS. 24. R.K. Toghiani and J.S. Lindner Thermodynamic Simulations of Saltcake Dissolution Experiments for Hanford Tank 241-S-112. DIAL TR Diagnostic Instrumentation and Analysis laboratory, Mississippi State University, Mississippi State, MS. 25. D.L. Herting Saltcake Dissolution FY 2002 Status Report. HNF- EDC Rev. 0. Fluor Hanford, Richland WA, and references therein. 26. D.A. Reynolds. December 3, Personal communication. 27. D.A. Reynolds. December 12, 2002.Personal communication, Weslow Database. 28. R.D. Hunt. December, Unpublished data. Support of Alternatives for Disposition of High Level SRS Waste J. S. Lindner, R. K. Toghiani, P. Hill, L. Smith, and A. Antonyraj Introduction Previous DIAL efforts have focused on the salt cake waste contained within the Hanford single-shell tanks. Work centered on the evaluation, application, and improvement of the Environmental Simulation Program (ESP). 29 This effort involved comparing the data called by the software with other thermodynamic compilations, contrasting model predictions with those of other programs, and determining discrepancies predicted by the model when evaluated against salt cake dissolution experiments on tank core samples. The work was extended to include solubility studies on specific systems and the DIAL