SITE INFORMATION AND CORRESPONDENCE_FILE 2

Home Browse Search

Mr.Russell Chapin 1766 W.Monte Diablo Avenue Page 2 of 7 Good cross sections can be utilized as the basis to infer initial values and characteristics for a hydrological model, as is illustrated under point 2 below, but must be augmented with sufficient site-specific data to generate a valid model. The cross sections allow some intuitive estimates of hydrological conductivity differences between the sand unit and the fine-grained units, but give no useful insight to the variation of hydrological conductivity within the sand that may lead to preferred migration pathways - even oblique or counter to the groundwater flow direction inferred from field data. For these reasons, the EHD disagrees with ATC's statement that the cross sections are sufficient to determine groundwater flow characteristics. A model utilized to justify leaving a large mass of contamination in place should utilize all the site-specific hydrological characteristics that are practicable to obtain. Point 2: A one-layer Modflow model could be sufficient to characterize groundwater flow and contaminant migration, if the layer attributes are properly defined for the model, and the model is representative of the site. The EHD does not believe that all attributes for this site are or can be properly characterized with the one-layer model presented, the layer thickness and hydraulic conductivities being examples of poorly characterized attributes. ATC utilized a single layer model with a thickness of 25 feet for the Modflow model. The EHD believes that the characteristics of the sand unit and overlying fine-grained unit require a two- layer model at a minimum. The sand unit can be anticipated to provide the dominant contaminant migration pathway in the subsurface, while the hydraulic conductivity and retardation effects of the overlying fine-grained unit is likely to slow contaminant migration several orders of magnitude compared to the sand unit. Conversely, the upper fine-grained unit more likely provides the predominant contaminant flux from both the vadose and saturated zones. As shown on the cross sections, the sand unit thickness varies from 3 to 17 feet and averages 9.5 feet, based on the sand thickness shown on the boring traces in the cross sections. In the cross sections, ATC shows a laterally continuous sand unit that displays abrupt variations of thickness and depth. Following the common practice utilized in the environmental field, the cross sections were constructed utilizing the ground surface as a datum from which to hang the soil data and were not corrected for elevation differences. The sand unit varies in apparent thickness from approximately 3 feet (VE-1/ASO-3) to 17 feet (MW-2), the depth to the top of the sand unit varies from 7 feet below surface grade (bsg, MW-5) to 20 feet bsg (VE-2), and the depths to the bottom of the sand varies from 17 feet bsg (VE-1/ASO-3) to 26 feet bsg (MW-2). Therefore, the EHD believes the model should consist of a minimum of two layers with the sand layer thickness locally varying from 3 to 17 feet and a variable total model thickness up to 26 feet. Modflow can accommodate layers of variable thickness. Point 3: Selection of the dominant flow direction based on the contaminant distribution does not seem unreasonable to the EHD, but reconciling the observed groundwater elevation data and inferred flow direction with the selected flow direction has implications regarding the hydraulic characteristics of the site. The EHD did not intend to question the flow direction selected for the model, but does not have a clear understanding of the interplay of factors that controls contaminant migration in the subsurface at this site, and did not see the controlling relationship of factors demonstrated in the ATC model. The basic questions are: How does the plume move in a subordinate flow direction oblique or counter to the dominant flow directions inferred from field data, and what are the implications to both the model variables and to possible future plume migration? Also, how does one select with confidence a hydraulic gradient for use in the model?