Well C-20 Transport Paths

Given results of the calibrated steady-state flow model, particle paths were calculated for the city wells using the program MODPATH [Pollock, 1994]. A 9x9x5 array of particles was released at the site of each well, and their paths and travel times were tracked backward from the well. To better-define the 3-D margins of the capture zones, additional arrays of particles were released in the cells surrounding the backward-tracked capture zone, and their final destinations determined by forward-tracking. Typically some 2500 particles were tracked for each well, resulting in 50,000 space-time positions. The resulting 3-D volume dataset of travel-time to each well was then processed using standard scientific visualization software (IBM Data Explorer) to determine 3-D isosurfaces of travel time (e.g. 5 and 10 year capture zones), 2-D projections of which are typically used directly as management zones in WHPA efforts.

As is clear from Figure 11, the connectivity of the most productive facies (Qal) is limited beneath Hays. This is borne out by anecdotal evidence of seemingly random distribution of productive vs. ``dry'' wells in Big Creek valley. Visual estimation of this connectivity suggests it is greatest in a NE-SW orientation (perpendicular to Big Creek), and least parallel to Big Creek. This should severely affect calculated particle paths and capture zones for the city wells. Following the procedure outlined above, particle paths to C-20 were found to align primarily NE from the well, indicating significant recharge to C-20 from the Chetolah Creek area (Fig. 15). Primarily owing to interference from other wells, few particles arrived from the vicinity of Big Creek to the southwest of C-20.

Figure 15: Backward-tracked particle paths to well C-20. Paths up to 100 years long are shown, although the bulk of the particles traverse the aquifer in 30 years or less. 85% of the calculated particle paths omitted for clarity. Qal aquifer thickness shown in color, city wells indicated by magenta circles. Select image to view full-sized version.

Most significantly, the calculated paths reproduced the dogleg shape of the observed PCE plume, although offset slightly to the east owing to coarseness of the finite-difference discretization. The particle paths leave a blank area NNE of well C-20. This area is the location of a bedrock high, and the reported and interpolated thickness of the Qal aquifer is very thin to zero. Total Quaternary sediment thickness is also small. The close correspondence between calculated and observed migration paths at this site give considerable confidence in the model approach and results, since no ``tuning'' of the model was carried out to improve this match. This validation also supports the contention that connectivity sand facies (Qal) bodies be the dominant influence on travel paths, and therefore should be accounted for in determining the most accurate possible transport and capture zone models at Hays. Although unmodeled, detection of PCE in only the easternmost (C-28A) of two closely spaced wells (C-27 is the other, see Fig. 15) at the southern end of the plume can be qualitatively attributed to connectivity effects by the Qal facies.

For planning purposes, transport predictions are commonly summarized as capture zones for drinking water wells. The 3-D pathlines can be projected to the surface, and 2-D capture zones delineated. Such zones for 5, 10, and 20-year travel times to well C-20 show an isolated area in the vicinity of the bedrock high where travel times are calculated to be greater than 30 years (Fig. 16). Note that paths across this rectangular zone lie in the silt facies (Layer 1), or the upper portion of the Kc bedrock (assumed to be slightly weathered and permeable).

Figure 16: 2-D capture zones for well C-20 shown at 5, 10 and 20 years. Note the hole in the capture zones with travel time greater than 30 years, where the Qal aquifer is absent. Select image to view full-sized version.