Three-Dimensional Capture Tubes: Well C-20

The need for a 3-D view arises from the requirement for highly accurate prediction in contaminant monitoring or remediation. Two-dimensional capture zones are conservative, since they depict the maximum transport rate over a vertical column, and can therefore be inaccurate in the presence of vertical heterogeneity. Three-dimensional travel time volumes do not suffer from these limitations. Three-dimensional isosurfaces (``capture tubes'')of travel time to well C-20 were computed using the particle tracking results described above. These show the combined effects of sand facies discontinuity, and predominance at depth. The capture tubes are generally narrow near the water table (especially in the silt facies) and are wide at depth in the sand facies. Consequently, areas of rapid transport from the water table are highly discontinuous (e.g. red areas, Fig. 17). The 3-D capture tubes are useful in indicating the true (rather than maximum possible) transport rate for an actual surface spill, and where monitoring is most likely to intercept (or avoid) a known contaminant plume.

Horizontal and vertical sections through the 3-D travel-time volume for well C-20 show tremendous discontinuity in travel time zones (e.g. red 5-year travel time, Figs. 17-18). The boundaries between successive time periods (i.e. 5, 10 and 20 years) can be considered 3-D capture zones, and would be suitable for designation as 3-D WHPA's. The 10-year 3-D capture volume (tube) is included in the 3-D interactive model accompanying this report (see button following Fig. 18). Travel time from the water table (red areas, Fig. 17) shows that 3-D paths terminating at C-20 within 5 years reach the water table at points a kilometer or more from the well, but do not reach the water table in the intervening area. In the 3-D view, the water table near C-20 contributes little water to the well, because of the intervening low-permeability silt facies (Qt). In addition, the 10-year travel time zone reaches to the water-table near the northern end of the PCE plume (Fig. 17), but is not present near the dogleg bend in the plume, again owing to the presence of silt facies at the water table.

Figure 17: Map view of travel time from water table to well C-20. Colors show time as indicated in color scale. PCE 100ppb outline (1995) shown for comparison. Select image to view full-sized version.

Figure 18: Side view, 3D volume showing travel time from water table to well C-20 (SW-NE section, looking NW at Hays) and hydrostratigraphic units. Location of city wells shown by blue spheres, hydrostratigraphic contacts shown by brown lines.

Relative to the 2-D model, the 3-D model provides enhanced information on several key aspects of the system. Apparent in the 3-D model are locations within the 2-D capture zone that are poor choices for monitoring (e.g. the sand-absent window in Fig. 15), sites that may not contribute to C-20 in 5 years despite being inside the 5 year 2-D capture zone, and unforeseen directions of transport (displacement of particle paths around the sand-absent zone). Again for pro-active applications (e.g. planning and land-use or WHPA designation) the 2-D view is sufficient. For responsive applications (monitoring and remediation of existing contamination), the 3-D view is extremely valuable.

The severe discontinuity of travel time zones at the water table results from two phenomena: variable connectivity and the semi-confined status of the sand facies. Where the confining layer (thick silt facies) is present between the water table and the Qal aquifer (thick sand facies), travel times are increased. Within the 2-D capture zones depicted in Fig. 15, the water-table terminations of the 5 and 10 year 3-D capture zones are found only where the water table lies in the sand facies (i.e. unconfined Qal, Fig. 18). The spatial relationship between the water table and the Qt-Qal contact is the primary control on the location of the water-table terminus of the 3-D capture zones. While not readily apparent before modeling began, this relationship is reasonable, since fastest-paths to the C-20 well should be primarily horizontal in the sand facies (Qal aquifer), and primarily vertical in the silt facies (Qt aquitard).

Traditional two-dimensional capture zones (i.e. projection of particle pathlines to the ground surface) for other Hays city wells are similarly affected by sand facies distribution. In particular, pathlines for particles tracked backward from wells C-17, C-29, C-30 and C-32 show gaps where the sand facies is thin (Fig. 19). Temporal changes in computed two-dimensional capture zones at Hays also reflect the influence of estimated Qal thickness (Fig. 20). For example, particles entering well C-17 within 5 years are separated by a zone of thin Qal, while at 10 years the capture zone completely surrounds the area of thin sand facies.

Figure 19: Ten-year capture zones and particle paths for all Hays city wells. Capture zones calculated by backward tracking in modpath, shown in magenta. Representative particle pathlines to 100 years shown in white, showing complete path of particle (i.e. extending beyond 10-year travel time). Hays 1987 city limit, Big Creek, and PCE 100ppb contour shown by black lines. sand facies thickness shown by color fill. UTM coordinates (meters) shown on axes.

Figure 20: Five- (magenta) and Ten- (black) year capture zones, Hays city wells. shown in magenta. Digital map of Hays city streets circa 1993 and location of Big Creek shown in brown. UTM coordinates (meters) shown on axes.