Experimental Design

As discussed above for low-permeability systems, the analysis of spatial contaminant distributions is an appealing strategy for shortening the duration of column experiments. Again, the use of multiple columns operated over different time periods can facilitate identification of nonequilibrium sorption effects and lead to more robust parameter estimation.

If the retardation factor approach is adopted for performance assessment, the distribution coefficient should be measured under conditions that represent a "worst case" competition scenario. One approach to approximating "conservative" conditions is to pre-wash the zeolite in an effort to saturate the exchange sites with calcium, the chief competitor (e.g., Cantrell, 1996). Alternatively, it may suffice to operate the columns until full breakthrough of the competing solutes is observed, as suggested by Figure 4, followed by the analysis of spatial concentrations.

The primary difficulty with the spatial approach is the need to section and analyze column segments. In particular, saturated zeolite materials are difficult to extrude from a column. In our work, therefore, we have used a "scooping" procedure, in which a stainless steel spoon is used to remove material from one end of the column. This simple approach has yielded smooth contaminant mass profiles, with good reproducibility between replicate columns. A similar procedure was reported by Fuhrmann et al. (1995).

Another experimental issue concerns the choice of Sr compounds. For safety and logistical reasons, most researchers have used surrogates for Sr-90, including Sr-85, a gamma-emitter with a half-life of 65 days, and nonradioactive Sr (e.g., SrCl2). The advantage of Sr-85 is that the sorbed activity can be easily measured to very low levels without the need for an extraction step. A disadvantage is that the rapid decay imposes a practical limit on the duration of the experiments and complicates the data interpretation. These difficulties are avoided by using nonradioactive Sr, but an additional extraction step is necessary to remove Sr from the sorbed phase. Our work utilized microwave-assisted digestion with concentrated nitric acid. Despite the advantages of nonradioactive Sr, two significant disadvantages remain: 1) the limited detection limits associated with conventional analytical procedures, and 2) the presence of Sr in natural zeolite, which complicates data interpretation and obscures the leading edge of the contaminant migration front. For these reasons, the source concentration and column duration should be carefully selected to ensure usable data.

4.3 Boundary Conditions

In interpreting the results of advection-dominated column experiments, it is essential that a mass-conserving entrance boundary condition be utilized (e.g., as discussed by Parker and van Genuchten, 1984):

where Cin is the concentration in the feed solution.

Application of the above condition is especially important for studies of strongly sorbing materials because the reduced penetration of the contaminant results in sharp spatial concentration gradients over the entire duration of the study, which implies a significant dispersive flux at the column entrance. Thus, the calibration of the sorption coefficient is highly sensitive to the estimated dispersion coefficient. This sensitivity is reduced as the duration of the experiment increases, suggesting that the experiments should be performed for as long a period as reasonably possible. Also, an independent estimate for the dispersion coefficient (e.g., through tracer experiments) is needed for a more robust estimate of the sorption parameter(s).

For high-permeability barriers, predictions at the field scale are less sensitive to the form of the specified BCs than low-permeability systems. For most applications, a specified-concentration entrance condition should provide an adequate representation of the source. For the exit condition, the nature of the transition from "sorbing" to "nonsorbing" material suggests the use of a zero-derivative condition, although sensitivity analysis indicates that minimal error is introduced by the use of the common semi-infinite condition, which is more amenable to analytical solution.

4.4 Example: West Valley Demonstration Project

In the fall of 1999, a pilot zeolite barrier was installed at the West Valley Demonstration Project (WVDP) in Western New York. The WVDP is an environmental management project being conducted by the U.S. Department of Energy (DOE) with the cooperation of the New York State Energy Research and Development Authority. Details of the installation are reported by Moore et al. (2000). The clinoptilolite material used in the barrier had previously been studied by Cantrell (1996) for a proposed installation at the DOE Hanford Facility, by Fuhrmann et al. (1995) for use at the WVDP, and by Lee et al. (1998) for the 1998 installation at Chalk River, Ontario. The range of previously estimated distribution coefficients (Kj) was from 650 mL/g (Fuhrmann et al.) to 2600 mL/g (Cantrell). The variation across these studies is most likely attributable to differences in the source water and experimental conditions, although the data interpretation methodology may play a role. For example, Cantrell conducted batch tests using synthetic groundwater and zeolite that had been "prewashed" with a high-calcium solution, while Fuhrmann et al. utilized small columns and site groundwater from the WVDP. For the Fuhrmann et al. study, the unusually high reported dispersion coefficients suggest that the Kd estimation may have been influenced by the boundary effects described above.

In support of the WVDP, eight column tests were conducted at the University at Buffalo using WVDP groundwater spiked with nonradioactive Sr2+, over four durations: 10, 20, 40, and 60 days. A single Kdof 2045 mL/g was calibrated from data from one of the 60-day columns, then used to successively predict the results for the other columns (Figure 5, 10-day data omitted for brevity). The importance of the specified boundary condition was highlighted by comparing results from various calibration schemes. For example, specification of a constant-concentration entrance boundary led to similar model fits but estimated Kd values that were 50% lower. Even when the recommended third-type BC was applied, efforts to simultaneously calibrate both the sorption and dispersion coefficient yielded similar fits for several combinations of parameters. Specification of the dispersion coefficient to a value obtained from an independent tracer test was necessary to obtain a robust estimate of the sorption coefficient.

The reasonable agreement between the Kd values from our study (2045 mL/g) and Cantrell's work (2600 mL/g) suggests that both studies similarly accounted for the effects of calcium competition in the experimental design. Also, the ability of a single to describe data from several column durations (10 to 60 days) indicates that the equilibrium assumption is reasonable. However, the limited number of applicable studies suggests that more research is needed to confirm these conclusions.

Extrapolation of these results to the proposed field scale suggested a barrier life in excess of 25 years, based on the identified performance criteria (effluent Sr-90 < 1000 pCi/L). Work is ongoing to develop a more detailed model based on an explicit consideration of competitive ion-exchange effects using the general framework described above. One advantage of a multi-solute model is that less strongly sorbing solutes (e.g., Ca2\ Mg2+) could be monitored to develop improved models and predictive capability for the breakthrough of Sr-90.

0 211—day column data model prediction

Distance (cm)

9 Jtt-day column dala - model prediction

9 Jtt-day column dala - model prediction

Distance (cm)

\

_

\

0 60-day column d»t«l

-

4

- - - - model calibration J

-

i , , i i ,

Distance (cm)

Figure 5. Data and model simulations for WVDP column tests (parameters summarized in Table 4).

Table 4 Conditions for zeolite column experiments

Parameters

Spatial data (Figure 5)

B cut

20 day

40 day

60 day

Zeolite length (cm)

11.1

12.2

11.7

13.5

Velocity (m/d)

0.43

0.45

0.45

0.42

Porosity

0.62

0.59

0.59

0.61

Dispersion coefficient (m2/d)a

0.22

0.23

0.23

0.21

Bulk density (g/mL)

0.86

0.80

0.80

0.83

Influent Sr2< (mg/L)

1.0

1.0

1.0

1.0

Influent Ca"' (mg/L)

NA

NA

NA

90

Influent Mg2' (mg/L)

NA

NA

NA

16

Influent K+ (mg/L)

NA

NA

NA

38

Influent Naf (mg/L)

NA

NA

NA

60

estimate based on tracer test, all other parameters measured, NA = not analyzed estimate based on tracer test, all other parameters measured, NA = not analyzed

0 0

Post a comment