Modeling Results and Conclusions

1. No Leaching Potential

The following paragraphs present the modeling results and conclusions for cPAH concentrations in soil that are protective of the shallow water-bearing unit as well as the deep zone. The results are presented along with the intricacies of the modeling effort and a discussion of the validity of the results from an analytical solution standpoint. Assumptions concerning the mobility of cPAHs are formulated from a review of the respective Mis. Benzo[6]fluoranthene is the most relatively mobile cPAH because of its higher water solubility. It has the greatest po tential to leach and affect water quality at a point of compliance. As stated previously, all modeling efforts will therefore use benzo[b]fluoranthene physical and chemical data in the determination of a total cPAH soil cleanup goal.

2. Shallow Aquifer

The VHS model is used to estimate the dispersion of cPAHs in shallow zone groundwater. The shallow aquifer is defined as the first 3 m of the saturated zone and is hydraulically connected to the intermediate aquifer. The shallow zone and the upper intermediate water-bearing zone communicate through the upper intermediate aquitard, which consists of an 11-30-ft layer of interbedded clays, silty clays, and sandy clays. Since the VHS model assumes isotropic homogeneous conditions, it cannot be extended to describe constituent movement in the upper intermediate aquitard; therefore, vertical dispersion cannot be extended beyond 3 m in the shallow aquifer. Domenico and Palciauskas [19] defined Z by monitoring the contaminant plume at the waste area boundary. This type of monitoring data is not always available or possible, and commenters on the model suggest that a less arbitrary estimation of the Z parameter can be calculated using vertical dispersivity and the width of the source area parallel to groundwater flow directions. Constituent penetration at the downgradient source boundary is estimated with the equation

where Z is the penetration depth, az the vertical dispersivity, and Y' the width of the source area parallel to the groundwater flow direction. The value of a2 is estimated as 0.2 m [13].

Penetration depths of 2.24, 3.02, and 4.71 m were derived for the southeastern, southwestern, and northern source areas, respectively. Penetration depths in the southwestern and northern source areas exceed the effective depth of the shallow aquifer. Therefore, dilution by vertical dispersion is applicable only for the southeastern source area. Furthermore, vertical dispersion for constituents emanating from the southeastern source area is complete after 20 m of travel.

Horizontal spreading provides little dispersion for COIs from the southeastern source area (due to the length of the source area perpendicular to groundwater flow direction) and the southwestern source area (due to its proximity to the APC). Horizontal dispersion provides some dilution for the combined northern source areas. The dilution associated with vertical and horizontal dispersion is applied to the performance standard of 10 (ig/L to determine the COI concentration in leachate from the source area. The leachate concentration is then substituted into the OLM equation to determine the soil concentration that is protective of shallow groundwater at the APC.

Soil cleanup goals for cPAHs calculated for the southeastern, southwestern, and northern source areas are 250, 100, and 125 mg/kg, respectively.

3. Deep Aquifers

The potential impact of the 10-|ig/L individual cPAH performance standard on the deep aquifer can be estimated by assuming that the hypothetical deep well conduit acts as a continuing line source to the deep water-bearing zones.

The approximate location of the abandoned well is not directly downgradient of the southeastern source area, and since horizontal spreading does not occur to any significant extent relative to the southeastern source area's width, this area should not affect groundwater near the hypothetical conduit. Also, the hypothetical conduit is upgradient of the southwestern source area and is not located near the northern source areas; therefore, it should not be af fected by the leachate emanating from either of these source areas. Since 10 (ig/L is the performance standard for shallow groundwater, this concentration is used to represent the cPAH concentrations entering the conduit.

Shallow zone groundwater entering the deep aquifer is limited by availability. As stated previously, the flow rate used in the pumping test should be modified by a factor of 10 to account for the fact that shallow groundwater contours do not reflect an influence in the area of the hypothetical conduit.

Well logs indicate that the well was screened over a 15-m interval. Although the aquifer in the vicinity of the site is thought to be about 200 ft thick, dilution will be considered over only the 15 m that represent the screened interval.

The results of modeling indicate that the 10-|xg/L performance standard for the shallow zone is sufficiently protective of the performance standard of 0.2 (xg/L in the deeper waterbearing unit.

4. Uncertainties

The site conceptual model consists of a series of assumptions and algorithms derived from the 1988 RI report and current EPA databases. As with any model, uncertainties in the output stem from the variability in selected input parameters and the mathematical limitations of describing environmental transport phenomena. Every effort has been made in this modeling effort to ensure that when uncertainties arise any error in the output will be conservative. This section describes the major sources of uncertainty inherent to the site conceptual model for the site and possible implications for the modeling outcome.

The site conceptual model is limited to describing transport phenomena associated with Darcian flow characteristics of the underlying aquifer and subsequent Fickian transport of the cPAHs in the saturated soil matrix. If conditions other than these exist, the assumptions of the VHS model are violated. For example, if preferential flow pathways or zones of high permeability exist in the saturated substrata, dispersion does not occur geometrically with respect to the groundwater flow direction. The applicability of the VHS model is therefore limited to describing constituent movement in the first 3 m of the saturated zone. Although there is an interconnection between the shallow zone and the upper intermediate aquifer, the model is limited in its ability to describe potential dispersion of cPAHs through the upper intermediate aquitard that separates the two zones. To this extent, the model assumes that a no-flow boundary exists between the shallow zone and the deeper aquifers. This is a reasonable assumption given site hydrogeological data and will not lead to over- or underestimation of contaminant transport and subsequent attenuation.

The shallow aquifer is characterized in the RI report as being comprised of predominantly silty materials and clayey sands. This material is relatively homogeneous, and preferential flow pathways do not exist, with the exception of slickenslides, which occur regionally. The RI report also indicates that the shallow zone is continuous beneath the site. The use of the VHS model should provide an adequate quantitative estimate of the dispersive processes at the site in the shallow zone.

Limited dilution is afforded by the use of the VHS model, due to the size of the potential source areas, their proximity to the APCs, and the thickness of the shallow aquifer. This is particularly evident in the southern portion of the site, where no dilution by vertical or horizontal spreading is produced when cPAHs are modeled from the southwestern source area to the western property boundary.

The assumption that cPAHs are chemically conservative and are not retarded by the organic carbon content of the saturated zone soils produces a very significant level of conserva tism in the estimation of cPAH transport in groundwater. Similarly, high estimates of permeability and porosity reduce the potential for underestimating migration potential from the source areas.

The site conceptual model depends upon the size of the potential source area used in the calculations of groundwater at the APC. These areas are approximated, and insufficient chemical data are available to better estimate the time and the size and shape of the areas. The source areas were estimated by a visual inspection of the soils (i.e., if a soil is stained, it is assumed to contain cPAHs) completed in 1988 as part of the RI. Overestimation or underestimation of the size of the potential source area could lead to an over- or underestimation of the source's effect on the underlying groundwater quality. Due to the conservative nature of the modeling efforts, the underestimation would probably not be significant in the determination of the no leaching potential.

The most significant source of uncertainty in the site conceptual model arises from the use of the OLM to estimate leaching potential. Actual leaching of cPAHs could be greater or less than leaching predicted by the OLM. Verification of site-specific leaching potential by a comparison of OLM estimates to actual TCLP analytical results will reduce this source of uncertainty by providing site-specific soil leachate concentrations or a level of certainty to the OLM values.

No site-specific information exists for the deep aquifer. Information on the zone is either gathered from recent literature or inferred from trends occurring in overlying aquifers. The greatest uncertainty in modeling potential constituent migration to the deep aquifer is the estimation of flow through the hypothetical conduit from the shallow zone to the deep unit. The existence of this pathway is questionable, as historical investigations failed to determine whether the conduit exists, and trends in shallow groundwater do not indicate the presence of a significant shallow groundwater sink in the approximate location of the well. If the conduit does exist, in all likelihood it is not open from the shallow zone to the deep zone, and flow will be impeded by subsurface materials inside the former well or by the integrity of the remaining well casing.

5. Validation of the Site Conceptual Model

Examination of the site conceptual model indicates that the total cPAH action level is greatly influenced by each constituent's ability to leach from site soils, not its dilution by dispersion. Vertical and horizontal spreading account for a dilution factor of approximately 1.8 from the southeastern source area to the fence line. Vertical and horizontal spreading do not account for dilution from the southwestern source area to the fence line, suggesting that leaching from unsaturated soils is most influential in determining the soil cleanup goal.

As a result, EPA suggested that samples below the established direct-contact soil cleanup goal of 700 mg/kg should be analyzed for cPAHs and subjected to TCLP analysis. Results obtained from the OLM portion of the site conceptual model could then compare to site-specific soil leachate data. In addition, OLM values will be generated for the samples subjected to TCLP. The OLM value and the TCLP result for each cPAH can then be compared for each sample collected from the site. The average difference between the OLM and TCLP data could then be subjected to a paired t test at the 95% confidence level to determine whether the site-specific TCLP results are less than, equal to, or greater than the OLM results used in the modeling effort for the determination of the soil cleanup goal [22], If the TCLP results are less than or equal to the OLM results, then modification of the model is not required. If the TCLP results exceed the OLM results, then the leaching portion of the site conceptual model will be revisited to include site-specific leaching data, and a new soil cleanup goal will be developed.

0 0

Post a comment