Robert A. Fjeld, John T. Coates, and Alan W. Elzerman
Department of Environmental Systems Engineering
342 Computer Court
Anderson, SC 29625

James D. Navratil
Lockheed Martin Idaho Technologies Company
P.O.Box 1625
Idaho Falls, ID 83415-3921

Experiments were performed to determine if high mobility forms of plutonium, such as soluble complexes or colloids, are possible in sedimentary interbed below the Radioactive Waste Management Complex at the Idaho National Engineering and Environmental Laboratory. Column studies were performed using Pu(III+IV) and Pu(IV+V) spikes in a synthetic perched water containing potential complexants. The constituents of the synthetic perched water were selected because (1) they might have been present in perched water during normal conditions or during flood events and (2) they might contribute to high mobility transport of plutonium either through compexation or colloid formation. The results suggest the presence of multiple physical/chemical forms having different mobility's depending on the oxidation state of the plutonium and the geohydrological history of the interbed. The implication of the results, if they are representative of processes that occur under actual field conditions, is that transport models based on a single retardation factor inferred from batch distribution coefficient measurements are not appropriate.

There are considerable inventories of plutonium buried in portions of the Radioactive Waste Management Complex (RWMC) at the Idaho National Engineering and Environmental Laboratory(INEEL). The RWMC is separated from the Snake River Plain aquifer by approximately 180 m (590 feet) of alternating layers of basalt and sedimentary interbed. Based on batch distribution coefficient (K_D) measurements, plutonium retardation in interbed is very high, and interbed is expected to be an effective barrier to plutonium migration. However, field measurements below the RWMC suggest that plutonium may have migrated through at least one interbed layer. The purpose of this study was to determine if plutonium might be present in one or more forms (such as colloids or soluble complexes) which moves relatively quickly through interbed material.

Laboratory column studies were conducted to determine if high mobility fractions of plutonium are observed for a synthetic perched water containing a number of potential complexants. The experiments were performed with columns containing natural colloidal material ("unwashed" columns) and columns treated to remove colloids ("washed" columns). Plutonium was introduced into the columns as a finite step spike which was followed by unspiked synthetic perched water. Plutonium concentrations were measured in the column effluent. The data are presented in the form of column breakthrough curves. Following completion of the tests, the spatial distribution of plutonium remaining in the columns was also measured.

The sedimentary interbed was obtained from an uncontaminated region adjacent to the RWMC. The material was composited from well cores collected at 5 different depths ranging from 50 to 120 m and was sieved to the size fraction smaller than 250 m m. The interbed was composed primarily of silt and sand particles with varying amounts of clay, and it had a cation exchange capacity of 17 meq/100 g.

The composition of the synthetic perched water (SPW) is given in Table 1. The constituents were selected because (1) they might have been present in perched water at the RWMC either during normal conditions or during flood events and (2) might contribute to high mobility transport of plutonium through either complexation or colloid formation. The concentrations were selected to be representative of maxima that might occur.

The column specifications (Table 2) were based on the recommendations of Relyea [1982] for minimum flow velocities and column dimensions. The columns were dry packed 1 cm at a time and compacted by lightly tapping the column on the bench top.

The column apparatus (Figure 1) consisted of reservoirs for the spiked and unspiked synthetic perched water, a peristaltic pump (Masterflex Model 7550-90), a column packed with interbed, and a fraction collector (Eldex Universal Fraction Collector). Experiments were performed using spikes that were to be Pu(IV) and Pu(V). The spikes also contained strontium-85. The Pu(IV) spike was produced by reduction to plutonium nitrate. The Pu(V) spike was produced by oxidation to Pu(VI) followed by photoreduction to Pu(V). The oxidation state distribution in each spike was measured using a modification of the TTA/HDEHP extraction technique described by Neu et al. (1994). Based on the TTA analyses, there were sizeable amounts of Pu(IV) in the Pu(V) spike and Pu(III) in the Pu(IV) spike. For this reason, the spikes are hereafter designated as Pu(IV+V) and Pu(III+IV). The plutonium concentrations in the spikes were approximately 100 Bq/mL and 350 Bq/mL for Pu(III+IV) and Pu(IV+V), respectively. The strontium-85 concentration was approximately 225 Bq/mL. The test procedure for the unwashed column was to fill the column with the synthetic perched water until the pore spaces were filled. The spike was introduced as a finite step of approximately one pore volume and followed by 1000 - 2000 displaced pore volumes (DPV) of the synthetic perched water. The washed columns were first washed or "eluted" at a flowrate of 0.3 ml/min for 24 hours at pH 8, 24 hours at pH 10.7, and 24 hours at pH 12. The columns were again eluted with synthetic perched water at pH 8 for forty-eight hours immediately prior to introducing the spike. For both columns, fraction collection was initiated with the emergence of the leading edge of the spike from the column (determined visually for the unwashed column and based on calculated travel times for the washed column). The column effluent was collected in 1 to 10 mL fractions from which 1 mL aliquots were analyzed with the liquid scintillation counter. Following completion of a test, the spatial distribution of plutonium remaining in the column was determined qualitatively. This was accomplished by separating the soil 1 cm sections and contacting the sections with 5 mL of 5 M HCl and 5 mL of 5 M HNO₃ for 48 hours. The supernatant was separated and counted by liquid scintillation. The aqueous effluent concentrations, C(t), were normalized to the influent concentration, C(0), and were plotted as a function of displaced pore volume (determined by normalizing the run time of the experiment to the hydraulic residence time for the column).

Presented in Figure 2 is an overlay of the Pu(IV+V) breakthrough curves for the unwashed and washed columns. The concentration profile for each column is characterized by a small peak between 1 and 2 DPV (recoveries were 0.07% and 0.03% for the unwashed and washed columns, respectively) followed by a low, non-zero residual for the remainder of the test. The total recoveries were 4.3% and 2.7% for the unwashed and washed columns, respectively. The oxidation state analysis is given in the upper left hand corner of the figure.

Presented in Figure 3 is an overlay of the Pu(III+IV) breakthrough curves for the unwashed and washed columns. In contrast to the results for Pu(IV+V), there is a large difference in the two concentration profiles. For the unwashed column, the concentration exhibits a plateau from approximately 1 to 10 DPV that is almost two orders of magnitude above background. This is immediately followed by a large peak that reaches a maximum at approximately 20 DPV, extends to approximately 1000 DPV, and has a retardation factor of 40. The total recovery is almost 60%. For the unwashed column, there is a non-zero plateau from the beginning of the test until about 20 DPV that is approximately one order of magnitude above background. This is followed by no detectable concentration from 30 to 90 DPV, followed by a broad peak that extends from 100 DPV to past 1000 DPV, has a maximum between 500 and 800 DPV, and has a retardation factor of 560. The total recovery is approximately 16%.

The results of the analyses to determine the distribution of plutonium remaining in the columns following the test are presented in Figure 4 and Figure 5 Pu(IV+V) was more strongly retained than Pu(III+IV), and the retention was stronger in the washed columns than in the unwashed columns. For example, over 90% of Pu(IV+V) was retained in the first two centimeters of the washed column compared to 70% in the unwashed column. The corresponding values for Pu(III+IV) were 55% and 15%.

Breakthrough curves for ⁸⁵Sr in the Pu(IV+V) columns are presented in Figure 6. Similar curves were obtained for the Pu(III+IV) columns. Essentially all of the strontium emerged in a single peak between 50 and 500 DPV, and there was no evidence of a high mobility fraction. In addition, there was very little difference between the washed and unwashed columns. The retardation factors for the four tests were between 225 and 278.

The plutonium breakthrough curves in Figures 2 and 3 clearly suggest the presence of multiple physical/chemical forms having different mobility's in interbed. For Pu (III+IV) in the unwashed column, there could be as many as three different forms: a small, high mobility fraction from 1 to 10 DPV; a large fraction of moderate mobility from 10 to 1000 DPV; and a large low mobility fraction represented by the plutonium retained in the column. For Pu(III+IV) in the washed column there could also be as many as three forms: a very small, high mobility fraction for the first 20 DPV, a larger fraction of moderate mobility from 100 to more than 1000 DPV; and largest fraction retained in the column. Although Pu(IV+V) exhibited much greater retention than Pu(III+IV), small, high mobility fractions were also observed.

The implications of the results presented here, if they are representative of processes that occur under actual field conditions, are that transport models based on a single retardation factor inferred from batch K_D measurements are not appropriate. A retardation factor so determined would represent a weighted average for the various fractions that exist. When used in a transport model a single retardation factor would overpredict the mobility of some fractions and underpredict the mobility of others. This could be corrected by characterizing retardation as a distribution rather than as a single value. This is illustrated in Figure 7 using data from the unwashed, Pu(III+IV) column. This histogram is produced using both the column effluent (i.e. breakthrough) data and the data from the soil analyses. It is intended solely to illustrate the approach because the retardation factor bins are not optimum (they correspond to the 1 cm sections that were analyzed and they include the plateau from 1 to 10 DPV under the R=42 bin). In modeling the approach would be to calculate the contaminant concentration as the superposition of the contributions from each of the discrete retardation factors, weighted by the frequency from the histogram.

1. Neu, M.P., Hoffman, D.C., Roberts, K.E., Nitsche, H., Silva, R.J., (1994) "Comparison of Chemical Extractions and Laser Spectroscopy for the Determination of Plutonium Species in Near-Neutral Carbonate Solutions," Radiochem. Acta, 66, 265-272.
2. Relyea, J.F., (1982) "Theoretical and Experimental Considerations for the use of the Column Method for Determining Retardation Factors," Radioactive Waste Management and the Nuclear Fuel Cycle, v.3, pp. 151-166.

BACK