S. M. Serkiz
Westinghouse Savannah River Company
Aiken, SC 20908

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

This report has been reproduced directly from the best available copy.

Available for sale to the public, in paper, from:  U.S. Department of Commerce, National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161,  phone: (800) 553-6847,  fax: (703) 605-6900,  email:  orders@ntis.fedworld.gov   online ordering:  http://www.ntis.gov/support/ordering.htm

Available electronically at  http://www.osti.gov/bridge/

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

Degradation products of cellulosic materials (e.g., paper and wood products) can significantly influence the subsurface transport of metals and radionuclides. Co-disposal of radionuclides with cellulosic materials in the E-Area slit trenches at the SRS is, therefore, expected to influence nuclide fate and transport in the subsurface.

Initial modeling of this system was undertaken in 1996 and utilized published stability constants for reactions between low molecular weight acids (i.e., citric acid and EDTA) and numerous radionuclides. In this approach, the physicochemical properties of these simple organic acids were intended to approximate effects of cellulose degradation products (CDP) on nuclide sorption. Notable differences in the allowable radionuclide inventory of the E-Area Slit Trench were predicted when the presence of CDP was accounted for in this conceptual model (Serkiz and Myers, 1996).

A series of laboratory studies were initiated in April 1998 to validate and/or update the modeling assumptions employed in the 1996 modeling effort. In these studies, laboratory results were generated for ternary systems containing divalent and trivalent radionuclides (uranium (VI) or europium (III)), organic matter (Suwannee River dissolved organic matter or laboratory generated CDP), and a soil phase (SRS Burial Ground soil or Aiken kaolinite) (Serkiz et al., 1998 and Serkiz et al., 1999). Laboratory data on the influence of CDP on metal/nuclide cation sorption show that these organic compounds can significantly reduce sorption, both directly and indirectly. Directly through complexation of metals in the aqueous phase, thus, inhibiting sorption and indirectly by reducing the system pH that, in turn, results in a more positively charged soil surface also inhibiting cation sorption. The direct effects of organic matter at more neutral pH values is several orders of magnitude reduction under conditions of high organic matter content. At low pH values, K_d values are similar regardless of organic matter content. K_d variability due to pH, at constant organic carbon content, is as high as about two orders of magnitude. The degree of this pH effect decreases with increasing organic matter content. Furthermore, these studies showed that dissolved organic matter production from cellulosic source materials was a very dynamic process with concentrations over 1 g per liter of organic carbon and pH values below 4.5 observed in laboratory static leaching tests.

Keywords: Radionuclide Speciation and Transport, Cellulose Degradation, Dissolved Organic Matter,
Europium, Uranium, E-Area Vaults, Performance Assessment

Introduction

This report documents the data analysis of the results of the above described laboratory studies in order to recommend K_d values for use in Performance Assessment modeling of nuclide transport in the presence of CDP.

Because generating laboratory data on the influence of CDP on the mobility of the 29 elements included in the E-Area Vaults PA modeling effort would be cost and time prohibitive, an indexing approach based on chemical analogues was adopted for cationic species. The index approach taken in this work is largely empirical and should be validated with additional studies both in the field and laboratory.

In recommending CDP influenced K_d values for use in PA modeling, both the primary (interactions of CDP with both the soil surface and the dissolved metal/nuclide) and secondary (reactions involving CDP that impact major-ion chemistry (e.g., pH)) influences of the CDP on metal/nuclide sorption were considered. The general conceptual model is:

• CDP has a direct affect on the sorption of multivalent metal/nuclide cations, with a general trend of increasing sorption at low-pH values and decreasing sorption at higher pH-values. This behavior can be explained, at least in part, by the pH-dependent sorption of natural organic matter on SRS soils. Where at higher pH values, more organic matter is in solution and at sufficiently high concentrations can outcompete the soil for reaction with the contaminant metal/nuclide.
• CDP has a secondary influence on the sorption of multivalent metals/nuclides that is related to the ability of CDP to reduce the aqueous-phase pH. In the pH range of interest (below 6.0), increased concentrations of CDP decreased the pH of the system and, concurrently, decreased sorption to soils.

The general approach to generating CDP-influenced K_d values was to:

1. Separate the transport model into five distinct regimes based on dissolved organic carbon content (1000 to 100 mg C/L, 100 to 30 mg C/L, 30 to 10 mg C/L, 10 to 1 mg C/L, and less than 1 mg C/L).
2. The pH of each of the regimes was estimated based on the relationship between organic carbon content and pH obtained in laboratory CDP generation studies. For purposes of conservatism, these pH estimates are based on the highest carbon content in the model regime.
3. The PA for the E-Area Disposal Trenches evaluated inventories for isotopes of 29 elements. Because generating laboratory data on the influence of CDP on the mobility of 29 radionuclides would be cost and time prohibitive, a number of simplifications and assumptions were imposed. These were:

1. Anionic and monovalent cationic species generally exhibit a very weak affinity for natural organic matter and, therefore, to CDP. Sorption of nuclides that are primarily anionic or monovalent cationic were considered to be unaffected by the presence of CDP.
2. In the two cases where laboratory data were generated for the ternary system of multivalent cation (UO₂²⁺ and Eu³⁺), organic matter (Suwannee River DOM and CDP), and SRS soil materials (kaolinite and SRS Burial Ground Soil), K_d values were estimated directly from these data. To facilitate this data analysis, mathematical models were fitted to laboratory sorption data (both for UO₂²⁺ and Eu³⁺) for the soil/metal/organic matter system. Based on this relationship between pH and K_d for a given organic carbon content, CDP-influenced K_d values were then calculated for each model regime (see Item 1) based on total organic matter content and pH.
3. For those multivalent cations that were not investigated in the laboratory, it was assumed that the influence of CDP on K_d values observed for UO₂²⁺ and Eu³⁺ could be used to index the behavior of other divalent and trivalent metals/nuclides. In order to calculate CDP influenced K_d values for nuclides that were not examined in the laboratory, the percentage reduction in K_d for UO₂²⁺ and Eu³⁺ laboratory data were applied to the K_d values used in the E-Area Slit Trench PA (McDowell-Boyer et al., 2000) for each model regime in Item 1. In all laboratory experiments involving CDP and UO₂²⁺ and Eu³⁺, the presence of CDP, primarily due to the associated reduction in pH, resulted in lower K_d values.
4. The effects of pH on monovalent cation sorption in the absence of organic matter are smaller than for multivalent cations. To account for these K_d variations in the PA model, an indexing approach similar to that described in Section c. for the multivalent cations was also applied to monovalent cations. The only difference was that the organic matter only had an indirect influence of lowing the system pH. These pH effects were estimated from the small K_d reductions observed for lower pH regimes observed for Cs sorption to SRS soil in Johnson (1995).
5. For anionic species, K_d values generally increase with decreasing pH and, therefore, ignoring the affect of reduction in pH associated with CDP production is expected to be a conservative assumption. This was the assumption employed for anionic species in developing these recommended Kd values.

This section describes the curve fitting activities employed on laboratory data for the ternary system of organic matter/SRS soil material/nuclide.

Laboratory data for the sorption of divalent uranyl (UO₂²⁺) and trivalent europium (Eu³⁺) to SRS soils or soil components in the presence of both naturally occurring dissolved organic matter and laboratory generated CDP formed the basis for evaluating the direct affect of CDP on divalent/trivalent cation sorption. In order to facilitate K_d calculations, a three-parameter logistic function was fitted to the experimental data in the form of pH versus fraction metal/radionuclide sorbed. To minimize model-fitting biases, data sets were truncated to pH values of less than 6.0 prior to model fitting. The mathematical form of the three-parameter logistic function is:

Where: FS is the fraction metal/nuclide sorbed to the soil; a is the y asymptote and is a fitting parameter with a limit set to less than 1; pH₀ is the pH at the inflection point of the curve and is a fitting parameter; and b is a fitting parameter that relates to the steepness of the rise of the curve. The general shape of the curve is a sigmoid whose fraction of metal/nuclide sorbed ranges between zero and one and, for the cation sorption data analyzed, starts low at low-pH values and increases at higher values.

Where S/L is the solid to liquid ratio under which the data were generated.

This section describes the results of the general approach to estimate CDP-influenced K_d values described above.

In a previous laboratory study, actual cellulose degradation products were generated in the laboratory from materials known to be present in slit trenches at SRS (Serkiz et al., 1999). As expected, the pH of the cellulose leachate slowly decreased with time (Figure 1) from 5.2 at the start to 4.4 after 19 weeks. This is consistent with the production of carboxylic acid and hydroxyl functional groups from wood degradation processes. After an initial large production of dissolved organic carbon (DOC) in the first week, the DOC of the samples steadily increased from 545 mg C/L at week 1, to 1018 mg C/L at week 19 (Figure 1). The slope of the total organic carbon (TOC) data as a function of time (i.e., organic carbon production rate) decreases as the experiment progresses.

Based on these experimental data and soil-pH for the soil used in the laboratory sorption study (Johnson, 1995), a relationship between organic carbon content and pH was developed to facilitate further K_d calculations. This relationship is summarized in Table 1.

Laboratory data for the ternary system containing the trivalent lanthanide europium, organic matter (Suwannee River DOM and CDP), and SRS soil are contained in Serkiz et al. (1998) and Serkiz et al. (1999). Data at 0, 10, 30, 100, and 500 mg C/L were fitted to the three-parameter logistic function as described above. The curve fitting data are summarized in Tables 2 through 6 for 0, 10, 30, 100, and 500 mg C/L laboratory data respectively. Graphical representations of the laboratory data and model fits are shown in Figures 2 through 6 for the 0, 10, 30, 100, and 500 mg C/L laboratory data respectively.

Laboratory data for the ternary system containing the divalent uranyl ion, organic matter (Suwannee River DOM), and SRS soil components (kaolinite and Burial Ground soil) are contained in Serkiz et al. (1998). Uranyl sorption data to kaolinite at 0, 10, and 30 mg C/L and uranyl sorption to an SRS Burial Ground soil in the absence of added organic matter (Johnson, 1995) were fitted to the three parameter logistic function as described above. The curve fitting data are summarized in Tables 7 through 9 for 0, 10, and 30 mg C/L sorption to kaolinite laboratory data respectively and Table 10 for uranyl sorption to SRS soil in the absence of added organic matter. Graphical representations of the laboratory data and model fits are shown in Figures 7 through 9 for the 0, 10, and 30 mg C/L laboratory data respectively and Figure 10 for uranyl sorption to SRS soil in the absence of added organic matter. The kaolinite K_d data were adjusted to a whole soil response by assuming a clay content of 3.8% by weight and that the remaining soil mass was unreactive with respect to to uranyl sorption.

The variability in K_d with respect to changes in both pH and organic carbon content is clearly shown in the laboratory data. The variability due to pH, at a constant organic carbon content, is as high as about two orders of magnitude. The size of this pH effect decreases with increasing organic matter content. Similarly, the effects of organic matter at more neutral pH values is several orders of magnitude reduction at high organic matter content. At low-pH values, however, K_d values are similar regardless of organic matter content.

Recommended K_d values, along with the fractional decrease in K_d with increasing organic matter content, based on laboratory data for ternary systems of Eu or U, natural organic matter, and SRS soil components are summarized in Table 11.

Site-specific data for the sorption of Cs to SRS soils as a function of pH are from work presented in Johnson (1995). These data are summarized in Table 12 for the pH in each of the model regimes. As expected, the fractional K_d reduction with decreasing pH is much less than for the multivalent cations. Because organic matter is not expected to strongly influence monovalent cation sorption directly, only the effect of pH on monovalent cation sorption was considered. Similar to the approach taken with the multivalent cation species, other monocations were indexed to the K_d reductions observed with decreasing pH for Cs sorption to SRS soil.

Based on the approach described above, recommended CDP-influenced K_d values for the 29 elements considered in E-Area Vaults Disposal Facility were calculated. Where site-specific K_d data were available (i.e., U, Cd, Cs, and Eu) for nuclide sorption in the absence of organic matter, they were recommended for use in the less than one mgC/L model regime and in calculating K_d values for all other model regimes. The results from this exercise and summary information from the previous PA modeling effort (McDowell-Boyer, 2000) are presented in Table 13.

Laboratory data on the influence of CDP on metal/nuclide cation sorption show that these organic compounds can significantly reduce sorption, both directly and indirectly. Directly through complexation of metals in the aqueous phase, thus, inhibiting sorption and indirectly by reducing the system pH that, in turn, results in a more positively charged soil surface also inhibiting cation sorption. The direct effects of organic matter at more neutral pH values is several orders of magnitude reduction under conditions of high organic matter content. At low pH values, K_d values are similar regardless of organic matter content. K_d variability due to pH, at constant organic carbon content, is as high as about two orders of magnitude. The degree of this pH effect decreases with increasing organic matter content.

Because generating laboratory data on the influence of CDP on the mobility of the 29 elements included in the E-Area Vaults PA modeling effort would be cost and time prohibitive, an indexing approach based on chemical analogues was adopted for cationic species. As expected, the higher the cationic charge of the specie of interest the greater the influence of the organic matter, over the geochemical range found in the model regimes. Using this indexing approach, K_d reductions of over 99 percent were calculated. Given the large K_d reduction for this index approach, if a high degree of conservatism is incorporated in the PA K_d value, then unrealistically low inventories for certain nuclides could be calculated.

The index approach taken in this work is largely empirical and should be validated with additional studies both in the field and laboratory. Certain mechanisms not accounted for in this work could significantly increase the allowable PA inventory if they are incorporated in the PA models. For example metal/nuclide reduction could result in a very high K_d for U, Pu, and, Tc by the virtue of a low solubility limit on the reduced form of these elements. The highest DOC concentrations are expected in the vicinity of the actual waste. These concentrations, and the resulting effect on K_d values, will decrease with distance from the waste. This effect should be investigated, both in the field and with modeling studies. Although the pH buffering capacities of SRS soils are relatively low, this was not accounted for in estimating the magnitude of the pH reduction due to the formation of CDP. Some credit for this soil-buffering mechanism could be accounted for with additional laboratory studies.

1. Johnson, W.H. (1995). Sorption models for U, Cs, and Cd on Upper Atlantic Coastal Plain Soils. Ph.D. Thesis. Georgia Institute of Technology.
2. McDowell-Boyer L., Yu A. D., Cook J. R., Kocher D. C., Wilhite E. L., Holmes-Burns H., and Young K. E., (2000). Radiological Performance Assessment for the E-Area Low-Level Waste Facility, WSRC-RP-94-218 Rev. 1 (U). Westinghouse Savannah River Company. , WSRC-RP-94-218 Rev. 1, Aiken, SC.
3. Serkiz S. M., Knaub D., and Uhal H., (1998). Phase I Nuclide Partition Laboratory Study Influence of Cellulose Degradation Products On the Transport of Nuclides from SRS Shallow Land Burial Facilities (U). Westinghouse Savannah River Company. WSRC-TC-98-00460. Aiken, SC.
4. Serkiz S. M., Knaub D., and Lee C., (1999). Phase II Nuclide Partition Laboratory Study Influence of Cellulose Degradation Products On the Transport of Nuclides from SRS Shallow Land Burial Facilities (U). Westinghouse Savannah River Company. WSRC-TC-99-00298. Aiken, SC.
5. Serkiz S. M. and Myers J.L., (1996). Additional Information for E-area Vault Performance Assessment, Appendix I "Suspect Soil Performance Analysis" – Results of Modeling the Effects of Organic Matter on the Mobility of Radionuclides as it Relates to the Disposal of Wood Products in E-Area Slit Trenches. Westinghouse Savannah River Company. WSRC-TR-98-00049. Aiken, SC.
6. Turner, D. R.(1995). A uniform approach to surface complexation modeling of radionuclide sorption. Center for Nuclear Waste Regulatory Analyses report CNWRA 95-001 San Antonio, TX .

Figure 2 – Eu Sorption to SRS Soil (24 g/L) Suwannee River DOM at 0 mgC/l

Figure 4 - Eu Sorption to SRS Soil (24 g/L) Suwannee River DOM at 30 mgC/l

Figure 5 - Eu Sorption to SRS Soil (24 g/L) Suwannee River DOM at 100 mgC/l

Figure 6 - Eu Sorption to SRS Soil (24 g/L) Suwannee River DOM at 500 mgC/l

Figure 7 - U Sorption to Kaolinite (4 g/L) Data from Turner (1995) No Added Organic Matter

Figure 8 - U Sorption to Aiken Kaolinite (3g/L) Suwannee River DOM at 10 mgC/l

Figure 9 -U Sorption to Aiken Kaolinite (3g/L) Suwannee River DOM at 30 mgC/l

Figure 10 - U Sorption to SRS Soil (2.38 g/L) From Johnson (1995) No Added Organic Matter