[Home] [Master Table of Contents]

[Report Table of Contents] [Next Section]

Groundwater Investigation and Associated Source Characterization

Section 8 - Hydrogeologic Investigation

The purpose of the hydrogeologic investigation was to characterize the geology and hydrogeology at CSSA to understand possible causes of contamination and migration prior to performing any remedial work, if necessary. The hydrogeologic investigation described below was based on the preliminary assessment work done in October through December 1992 (Section 4), and on published data and information pertinent to the investigation. Investigative tasks included geophysical well logging, well upgrades, groundwater monitoring over 2 years, a geologic investigation, and installation of two groundwater monitoring wells.

8.1 - Initial Evaluation

After groundwater monitoring began in May 1994, CSSA decided that the following three tasks would be appropriate to begin the second phase of the hydrogeologic investigation. These tasks were downhole geophysical well logging in inactive CSSA water wells, discrete interval groundwater sampling in uncased portions of the wells, and well abandonment and plugging. These tasks were discussed in a technical memorandum (Parsons ES, 1995a) and are summarized below.

8.1.1   Geophysical Well Logging

The December 1992 downhole camera survey demonstrated that the inactive water wells in the area of contamination had only 4 to 45 feet of surface casing, and that the wells were generally unobstructed for geophysical logging and sampling. Three of the wells (2, 16, and D) were selected for logging by natural gamma, gamma-gamma, neutron, and caliper tools to yield information about shale content, bulk density/porosity, moisture content, and borehole size changes, respectively. Logging was performed by Century Geophysical, Inc., in August 1994.

Stratigraphic information included determination of depth to the Bexar Shale in wells 2 and 16. Well 16 also penetrates into the Cow Creek Limestone below the Bexar Shale. Depths to the shale bed were interpreted at 310 feet bgl in well 2 and at 314 to 384 feet bgl in well 16. Well D is not deep enough to penetrate the Bexar Shale. No structural displacement between wells 2 and 16 was interpreted. The density tool indicated a relatively uniform porosity (bulk density) throughout the logged interval except where the caliper tool indicated larger hole diameters, suggesting increased size of the hole due to solution enhanced fractures or bedding planes. As determined by neutron logging, response to water levels in all three wells were equivalent, indicating a horizontal correlation between wells and perched water zones. Zones interpreted from the neutron logs were roughly 0 to 70 feet, 70 to 130 feet, and 130 to 180 feet bgl. The results were then used to perform discrete interval sampling.

8.1.2   Discrete Interval Sampling in Perched Zones

Because little surface casing existed in the inactive CSSA wells 2, 3, 16, and D (located in the area of groundwater contamination), discrete interval sampling of the uncased perched zones above the aquifer water level was conducted in September 1994. The zones to be sampled were determined from the August 1994 geophysical logging. Data from the downhole camera survey and the geophysical logging were used to determine optimum placement of the packer below each zone.

Due to the amount of suspected fractured or solution-enhanced zones and the seasonal decrease in water levels, few perched zones were encountered during this task. Only the 0 to 180-foot interval was sampled in each of the four wells, and one additional sample from 0 to 95 feet was obtained in well D. All samples were analyzed for chlorinated hydrocarbons using EPA SW-846 method 8010 (EPA, 1994).

Table 8.1-1 presents the well sample identifications, depths to static water, depths of perched zone sampling intervals, and analytical results. The results range from less than 1.0 to 15 ug/l in the 0 to 185-foot well D sample. The September 1994 quarterly groundwater monitoring results are PCE and TCE at 110 and 130 ug/l in well D, and 81 ug/l for each compound in well 16 (see table in Appendix G for 1994-1995 groundwater analytical results). Comparison of the discrete interval sample results to quarterly groundwater monitoring conducted in the same month, with no significant rainfall, shows that the discrete interval sampling was not conclusive as to the depth that contaminants might be entering wells D and 16. The contaminants may enter the wells below the water level as well as through perched zones intersecting the wells.

8.1.3   Well Upgrades and Plugging

8.1.3.1   Water Well Upgrades

Well upgrades were performed on wells 2, 3, 4, 16, and D. The upgrading of the wells consisted of setting approximately 200 feet of surface casing in each well. The depth of casing installation was based on measured water levels and geophysical logging of the wells. Two centralizers were set in each well near the bottom and top of each casing.

8.1.3.2   Water Well Plugging

The need for plugging was determined during the 1992 water well inventory. Wells which were obstructed or covered over were scheduled for plugging and recordation with the TNRCC. Plugging methods are described in Section 3.6.1. Plugging was performed by Layne Environmental Drilling, Inc., a state licensed driller, under supervision of a Parsons ES geologist.

CSSA well A was covered with asphalt near the intersection of North Outer Drive and Central Road; when uncovered, it was found to be filled in at the surface. Well C was also filled in with a concrete cap. Guidance from the TWDB indicated that covered wells with a permanent cap such as concrete were to be left undisturbed, and the plugging reported with the state. Wells 5 and 6 were plugged from their blockage depths of 158 and 190 feet bgl, respectively, to ground level. Wells B and F were plugged from 8 and 20 feet bgl, respectively. Layne submitted the water well plugging records to TNRCC in September 1994.

8.1.3.3   Tank Monitoring Wells Plugging

Fifteen monitoring wells located around removed underground storage tanks at CSSA were plugged in accordance with TNRCC requirements for final closure of the tank sites. Plugging, which took place in December 1995, was performed in accordance with the procedures discussed in Section 3.6.2. Jones Environmental Drilling, Inc., a state licensed driller, conducted the well plugging under Parsons ES supervision. Casings from the wells were first pulled, and the borehole then tremied grouted to ground level. To fulfill DoD guidance which requires reclamation and reuse of materials when possible, the casings were tripled rinsed and stored at the Building 5 site for future CSSA use. The well caps and locks were also saved for reuse. Plugging records, submitted to TNRCC in February 1996, are presented in Appendix D.

8.2 - Groundwater Monitoring

Quarterly groundwater monitoring has been conducted since May 1994. Quarterly monitoring consists of water level measurements, sampling, and analysis followed by data evaluation and report preparation. Other actions performed under this action include a packer test at well 16, pump and piping removal from well 11, weather station installation, and continuous water level monitoring. The following sections summarize the data collected during these activities.

8.2.1   Summary of Quarterly Groundwater Monitoring Events

CSSA wells were initially sampled on November 3-5, 1992 to determine the extent and impact of contaminant migration. Since 1994, groundwater monitoring has been performed every quarter. A map of well locations as of April 1996 is shown in Figure 8.2-1. Current well status and surveyed positions are provided in Table 8.2-1.

Groundwater flow direction has typically been to the southeast except for September 1994, and August and December 1995 during which flow was to the southwest (Figure 8.2-2, Figure 8.2-3, Figure 8.2-4, Figure 8.2-5, and Figure 8.2-6). Groundwater flow gradients have remained consistent throughout monitoring, ranging from 0.003 to 0.1 foot per foot. Flow to the southwest correlates with periods of low water level. This may reflect the influence of water table elevation upon groundwater pathways such as fractures and karstic zones. For instance, if water levels drop below zones dominated by such features and into areas controlled by primary porosity and permeability, a change in flow direction could result.

The potentiometric map for April 1996 (Figure 8.2-7) may reflect such an influence. A significant change in flow gradient to the east-southeast in the vicinity of wells 2, 3, 4, D, 16, MW1, and MW2 was observed. This flow direction indicates that well 16 is upgradient of source areas B-3 and O-1. April 1996 water level measurements represent historic lows for CSSA since quarterly monitoring has begun. It is important to note that April 1996 is the first month to employ data from monitoring wells 1 and 2, thus filling a data void and revealing a potentiometric feature previously not detectable because of the CSSA well configuration. Prior maps indicated flow gradients southwest and southeast of B-3 and O-1. The depicted flow gradient changes the direction of contaminant transport from SWMUs B-3 and O-1 from southeast or southwest to the east based on the assumption that the aquifer is isotropic. Most likely the aquifer is anisotropic, due to the presence of fractures, faults, and solution enhanced channels.

The lowest average water level observed was 960.45 feet above mean sea level (MSL) and occurred in April 1996 during monitoring well installation. The highest average water level was 1,090.44 feet MSL measured in October 1992. Water levels measured to date are listed in Table 8.2-2. Pumping from CSSA wells 9, 10, and 11 prior to water level measurements affects water levels that are expressed as a sink in the vicinity of the wells.

Rainfall events appear to induce periodic flushing of the contaminants into the water table and well bores or introduction of the rising groundwater table into contaminated, unsaturated strata (Parsons ES, 1995f). The time lag involved between rainfall events and changes in halogenated volatile organic (HVO) concentrations may range from a couple of days to a month which is likely the result of rainfall proximity and volume. HVO concentrations in monitoring wells D and 16 have been above MCLs for the majority of groundwater monitoring, with peaks in August 1991 in well 16 and from May 1994 to June 1995 in wells 16 and D (Figure 8.2-8 and Figure 8.2-9). HVO levels for the past year (April 1995 through April 1996) are shown in Table 8.2-3; results from 1992 through 1994 are summarized in Appendix G.

Groundwater now appears to flow from well 16 and D towards the source areas B-3 and O-1, not vice versa. Even though the formation is anisotropic, the April 1996 potentiometric maps show that wells 16 and D are still upgradient from B-3 and O-1. The most contaminated well is well 16, not monitoring well MW2 which is sidegradient to O-1 and much closer to O-1 than well 16. Possible explanation of the wells 16 and D contamination is the periodic flushing of contaminants below O-1 and B-3 after rainfall events. A slug of solvents could migrate through the 100- to 260-foot-thick unsaturated zone to the water table or to the uncased well bores. The downhole camera survey indicated water seeped into well bores far above the water table. One conceptual model scenario suggests that slug flow follows along bedding planes and fractures descending to the water table northwest towards toward wells 16 and D, allowing some contaminants to enter the water table in the vicinity of well 16 and D. A well immediately southwest of B-3 between well 16 and O-1 and a well northeast of B-3 may prove or disprove this scenario.

Metals and HVO analyses were performed on CSSA and four off-site wells in December 1995. Wells exhibiting high metals values were re-tested in January 1996. Well 1 was the only CSSA drinking water supply well that exceeded MCLs for any of the metals tested. However, two subsequent sampling events indicated the lead concentration was at 0.015 ug/l. Further sampling, analysis, and statistical evaluation may prove that the lead in well 1 may not be statistically significant above the drinking water action level for lead (0.015 ug/l). Agriculture wells G and I exceeded the MCLs for lead, and well 2 was at the MCL for copper. The metals data for CSSA wells are listed in Table 8.2-4. Results for the offsite private wells are listed in Table 8.2-5 and Table 8.2-6.

8.2.2   Well 16 Packer Test

8.2.2.1   Summary of Test and Results

Packer tests were conducted at well 16 before the April 1995 monitoring event in an attempt to determine the hydraulic relationship between the two production zones (lower Glen Rose and the Cow Creek limestones) of the middle Trinity aquifer. The tests coincided with quarterly groundwater monitoring to aid in the characterization of the groundwater from each zone. A technical memorandum of the packer test and results was submitted in summer 1995 (Parsons ES, 1995h) and is summarized in Appendix G.

The well 16 packer test results were inconclusive. Fluctuations in water level were attributed to a combination of test pumping and groundwater recharge from recent rainfall (Figure 8.2-10 and Figure 8.2-11). Equipment problems combined with a recent rain event prevented the use of recovery data to evaluate hydraulic communication between the upper and lower aquifer zones. The Cow Creek was isolated from the overlying saturated Bexar Shale and lower Glen Rose by setting the packer at 380 feet bgl, and the lower Glen Rose isolated from the Cow Creek by setting the packer at 365 feet bgl. Analytical results for samples taken from the isolated zones exhibited similar levels of PCE, TCE, and cis-1,2-DCE concentrations, indicating that the solvents may enter into the upper zone first. This may be interpreted as evidence of hydrologic interconnectedness between the two zones. However a complete set of pumping and recovery data may establish this interpretation.

8.2.2.2   IDW Characterization and Disposal

The IDW generated from the well 16 packer test includes the purge water associated with development of well 16. The purge waters were collected and containerized in a temporary tank "Baker Tank". The characterization of these water was made by the use of the interval sampling analytical results. The results of analyses for these waters are presented in Table 8.2-7 and indicates that the purge waters are nonhazardous as defined in 30 TAC 335 Subchapter R. Waste manifests for the IDW generated during this investigation are included in Appendix H.

Table 8.2-7 - Analytical Results of Purge Water at Well 16

8.2.3    Well 11 Pump and Piping Removal

Pump and piping were removed by a Parsons ES subcontractor from well 11 on December 14, 1995. This action has permitted direct water level measurement and sampling from the well 11 borehole since the removal date. During pipe removal two pipe sections were cracked and several exhibited extensive pitting, i.e., corrosion. The pipe sections and pump are currently stored at the well 11 shed. CSSA intends to replace the piping and put the well back on line in 1996.

8.2.4   Well 16 Transducer and Weather Station

Water level and rainfall monitoring to correlate the influence of rainfall upon groundwater levels is critical to interpretation of hydrogeology at CSSA. Quarterly water level measurements have been taken by electric water level indicator at well 16 since May 1994. However, characterization of the middle Trinity aquifer is most effectively addressed through continuous water level monitoring. Consistent precipitation readings have not been taken at CSSA. Rainfall measurements near Boerne, about 10 miles north of CSSA may vary significantly from those measured at the San Antonio International Airport.

Based on these observations and pursuant to the goal of site specific hydrologic characterization, an In-Situ LTM 3000 datalogger, PXD 260 pressure transducer was installed in well 16, and a Texas Weather Instruments (TWI) WPS-32 weather station was installed near the well 16 pump house on August 22, 1995. Installation of the equipment is discussed in Appendix G. Weather station equipment difficulties prevented station operation until April 15, 1996. Time constraints and limited meteorological data prevented transducer and weather station data correlation at the time of this report.

The transducer data and regional rainfall data for September 1995 through March 1996 are shown graphically on Figure 8.2-12. The sharp increase in the early November 1995 water level may indicate rapid local recharge. The much smaller increase which follows in mid-November may indicate regional recharge. In both cases water table fluctuations correlate with precipitation events measured at the San Antonio International Airport. The inverted relationship in observed magnitude between water table response at CSSA and rainfall at San Antonio is most likely the result of variation in rainfall over distance. The summer thunderstorm is sporadic and local, while the winter frontal storm covers a wide region. A gradual decrease in the water table is observable until data collection on March 6, 1996. The two water level "dips" on January 8 and January 16, 1996 are the result of transducer disturbance during the well 16 downhole velocity test of the seismic reflection survey (Section 3.1.3.2 and Section 8.3.3).

Well 16 transducer data indicated a discernible correlation between San Antonio precipitation and water levels at well 16. As previously mentioned, such a correlation was initially identified in the 1994-1995 groundwater monitoring report (Parsons ES, 1995f) in which water levels taken during quarterly monitoring were compared with San Antonio rainfall data from May 25, 1994 to April 6, 1995. The ability to correlate the data on a much shorter time scale as demonstrated by the transducer data indicates that a more detailed evaluation employing site specific rainfall data will provide meaningful precipitation, water level, and contaminant level relationship. The effects of and time lag associated with local precipitation events may be characterized once a suitable amount of rainfall data is collected.

8.3 - Geologic Investigation

Investigation into the geology at CSSA occurred at several times. Chronologically, the 1992 preliminary evaluation included a brief field survey and a more in-depth evaluation performed in 1995 and 1996. The 1995 drilling and sampling program provided limited geologic data by logging of 30-foot boreholes (Section 5.2). Soil gas survey results strongly indicated fractures or faults in the vicinity of well 16 and associated source areas B-3 and O-1 (Section 5.3). Because of the suspected faulting and the possible effects such geologic structures might have on contaminant migration, in-depth geologic mapping was performed in the fall of 1995. Several faults were found, but the results were not considered totally conclusive about all possible faultings, and therefore, subsurface data were generated during the GPR profilings and a seismic reflection survey in January 1996. Periodically throughout the project, area well logs were also compiled and reviewed for information about the stratigraphy and structure in and near CSSA.

All the data was compiled and correlated during the preparation of a new geologic map of CSSA. With the exception of the preliminary geologic survey (Section 4.5) and the drilling at potential source areas (Section 5.2), the following sections describe the geologic findings. The resulting geologic map is presented in Section 8.3.5.

8.3.1   Geologic Mapping

8.3.1.1   Field Mapping

During the 1995 drilling and sampling at potential source areas, a potential fault was discovered along Salado Creek southwest of well 16, indicating the need for in-depth mapping. Fault and fracture systems are hydraulic conduits for contaminant migration and as such accurate delineation of these structures through observation of direct evidence, stratigraphic position, and bed orientation is imperative.

Parsons ES mapped creek beds and outcrops in the inner cantonment August 29 to September 19, 1995. The initial mapping showed that there are two beds of Corbula at CSSA, separated by a marl layer, and that at least two minor faults exist as determined by 1-foot displacements of the uppermost Corbula bed. Strike and dip orientations of the surface limestone beds indicated that strike of the beds trends either generally northeast or northwest, with gentle dips ranging from 0 to 15 degrees southeast, southwest, and northwest. The draft results were discussed with EPA and TNRCC at the October 19, 1995, CSSA technical presentation meeting of the groundwater investigation.

An extended surface geologic mapping was conducted in 1996 following the initial survey of the inner cantonment. Outcrop mapping was carried into the northern and eastern sections (pastures) of the outer cantonment with the goal of refining structural interpretations through surface observation. Limited mapping was continued in the inner cantonment to confirm the historical findings.

Direct Observations

Outcrop at CSSA is minimal in area, and creek beds provide the greatest lateral exposure. Hillsides and quarries also provided areas of exposed bedrock. A map of survey coverage is shown on Figure 8.3-1.

Due to the limitations of direct observation, stratigraphic continuity is a crucial tool in identifying and delineating structural disturbances. Faulting was directly observed at three localities in CSSA. Two faults, mentioned above, were detected in the Salado Creek southwest of well 16. A third fault, exhibiting slickensides, stair-step displacement, and fault breccia, is located several hundred yards north of the southern CSSA boundary road and was discovered during the initial survey.

Stratigraphy and Bed Orientation

The initial mapping survey and research of published geologic work indicated that lithologies would primarily consist of the upper and lower Glen Rose Formation and occasional Edwards Limestone. However, no Edwards Formation is found at CSSA, though there are outcrops on hilltops to the southeast on Camp Bullis. Morphologically the upper Glen Rose is characterized by a weathered "stair-step" pattern. The upper and lower Glen Rose are separated by a thin (1 to 4 feet in thickness) bed of weather resistant limestone abundant in the foraminifera Corbula martinae. The triangular, millimeter sized foram is characteristic of this laterally continuous marker bed. The upper portion of the lower Glen Rose is characterized by a 2 to 8-foot thick marly limestone zone containing the echinoid Salenia texana. Bed identification relied heavily upon the occurrence of these stratigraphic index fossils due to the similarity of the upper and lower Glen Rose limestone beds.

Index fossil assemblages were present at fourteen separate localities. The majority of outcrops did not exhibit characteristics unique to either Glen Rose member, such as evaporite beds known to exist in the upper Glen Rose type location. Stratigraphic position in the sequence was determined by relating outcrop positions with respect to the nearest index fossil occurrence, and incorporating structural data. Bed orientation measurements contributed to the structural interpretation, especially in the north and east pastures of the outer cantonment where stratigraphic information was minimal.

8.3.1.2   Results

Stratigraphy

A significant portion of the inner cantonment is interpreted as lower Glen Rose. The north pasture is entirely upper Glen Rose, and the east pasture is predominately upper Glen Rose. The southeastern section of the north pasture and north half of the east pasture contain echinoid assemblages similar to those found in the lower Glen Rose Salenia zone. These occur in two distinct groupings at 100 to 160 feet above the Salenia zones in Salado Creek. The groupings indicate that significant stratigraphic breaks (such as those associated with faulting) do not occur within these groups. None of the assemblages contain the index fossil Salenia texana, nor was the Corbula bed observed at elevation above the assemblages. Thus, the stratigraphic units to which these assemblages belong could not be identified. Only members of the Cretaceous-age Glen Rose Formation were found and mapped at CSSA. A summary of strike and dip measurements for the upper and lower Glen Rose members and faults is shown in Appendix K.

Structure

No major displacements are evident at CSSA though a series of small fractures are prevalent in the northern and western areas of the inner cantonment. These fractures represent potential contaminant pathways though limited surface exposure prevents a more detailed interpretation of fracture continuity. Large displacements were originally thought to be present in the fault zones of the northern and southern areas of the inner cantonment, but were re-interpreted upon re-evaluation of geophysical well logs (Section 8.3.4) and information from the April 1996 logging of two pilot holes (Section 8.4.3.2).

Structural features found through field mapping are faults, fractures, two locations of slickenslides, and several small scale folds. Strike and dip of faults, summarized in Appendix K, show that the bedding orientation was found to vary throughout the study area. Average fracture strikes of N47E and N40W were identified from a stereonet projection. Dips of the beds are low and average 5 degrees.

Two northeast-southwest trending fault zones are observed - one through the central portion of CSSA near well 16, and the other near the southern boundary of CSSA. Fault measurements also show two trends of strike and dip on a stereonet. One set has an average orientation of N68W, N50E, and the other averages N74E, 68SE. These fault measurements compare favorably with the regional trends (Simpson and Guyton, 1993).

8.3.2   Ground Penetrating Radar Profiling

Ground penetrating radar and EM surveys were initially conducted by Parsons ES in March and May 1995 to locate potential source areas related to waste disposal activities. Geophysical anomalies included trenches at B-3 and a subsurface disturbance at O-1. The geophysical survey results were presented in a technical memorandum (Parsons ES, 1995b), and are summarized in Section 5.1. Copies of the GPR profiling report were submitted to AL/OEB, CSSA, EPA, and TNRCC in February 1996.

Additional GPR profiling was conducted in December 1995 and January 1996 to locate near surface expressions of faulting that could potentially serve as contaminant migration pathways. These surveys were conducted by Dr. Charles Young, Geophysical Consultant, Michigan Technological University, with the support of Parsons ES. Approximately 35,000 feet of GPR profiling was performed along nine transects during the extended survey. A 25- to 50-foot depth of penetration is estimated for the GPR surveys. Detailed information about the survey methods and results is in the GPR profiling report (Young, 1996).

GPR profiling identified interpreted offsets or subsurface disturbances along the western boundary road of the inner cantonment, the northern section of the inner cantonment area intersecting Central Road, and between well 16 and SWMU B-3 (Figure 8.3-2). Isolated interpreted offsets were also identified farther south along the transects. The majority of offsets interpreted from GPR data fall within the northern fault zone identified prior to the GPR profiling from surface observation and well log data.

GPR lines 1, 5, and 8 were located at B-3 and near O-1, well 16, and well D. Interpreted faults for this area exhibit less than 10 feet of displacement per fault and roughly coincide with those interpreted from surface mapping data and soil gas trends, results which are consistent with a fault zone. Throws on the faults are both normal and reverse, again consistent with a fault zone. The small faults identified on GPR line 11 are consistent with field mapped observations of faulting along Salado Creek in the same area. Displacements identified along this transect are less than 10 feet and reflect the scale of offset observed at surface. Results from line 7 along the western boundary indicated a series of anomalies interpreted as offsets with no surface expression, but may also represent fracturing or other subsurface disturbances such as channels. Due to relief along the road and absorption of the radar signal, no evidence of faulting was observed along the East Outer Drive (line 4). A chain link fence west of line 7 caused a false depth reflection at 70 feet.

The identified structural features represent potential contaminant pathways, particularly displacements in the vicinity of source areas B-3 and O-1. The GPR survey provided significant near surface data, but conclusions regarding fault versus fracture zone continuity and interconnectedness are limited by areal coverage because the data are limited by inherent responses to surface and subsurface effects and discontinuities in the limestone or clay-infilled beds. The GPR results helped to limit the amount of seismic reflection line to be surveyed (Section 8.3.2) and were a significant factor in structural interpretations (Section 8.3.5).

8.3.3   Seismic Reflection Survey

A seismic reflection survey was performed by Blackhawk Geosciences, under the supervision of Parsons ES. The primary objective was to locate offsets of the Bexar Shale at depths of about 310 feet bgl. The target was the contact between the lower Glen Rose Limestone and the Bexar Shale. This contact was believed to be a good seismic reflector because of expected acoustic velocity and density contrasts between these units. Approximately 10,000 linear feet of common depth point seismic reflection data were collected along two lines oriented in a generally north-northwest to south-southeast trend (Figure 8.3-3). Terrain permitting, seismic lines were run at approximately 90 degrees to the anticipated trend of the fault zone. In addition, a downhole velocity survey was performed in well 16 to establish one-way travel times to the contact. The seismic reflection report (Blackhawk Geosciences, 1996) and Section 3.1.3 discuss seismic reflection methodologies and data processing.

The interpreted reflector from the top of the Bexar Shale is not evident in a large portion of the data from Lines 1 and 2 due to a loss of reflectivity. This loss is believed to be caused by changes in the near surface, such as thickness of the unconsolidated layer, karst features, and fractures, rather than changes within the Glen Rose Limestone or Bexar Shale. Although the reflector is difficult to interpret in some areas, it is relatively flat where present. Faults with offsets greater than 20 feet would have a deflection of 4 milliseconds (ms) or more. It is difficult to discern others less than 20 feet.

The seismic reflection survey data is quite extensive and are not included in this report. However, copies of the seismic reflection lines can be found in the seismic reflection report which was submitted to AL/OEB, CSSA, EPA, and TNRCC in February 1996 (Blackhawk Geosciences, 1996). Figure 8.3-3 shows the three surface areas of offset determined from the seismic data. The offsets are discussed below.

A reflector interpreted as the top of the Bexar Shale is identified on Line 1 at a depth of approximately 314 feet (65 ms) bgl. Between shot points (SP) 165 and 180, approximately 30 feet (6 ms) of displacement is observed. This is interpreted as a normal fault with the southern block downthrown. A segment between SP 310 and 325 appears to be offset approximately 20 feet (4 ms) higher relative to the reflector observed on both sides of this segment.

A reflector interpreted as the top of the Bexar Shale is identified on Line 2 at a depth of approximately 380 feet (78 ms). Between SP 420 and 455, approximately 65 feet (13 ms) of displacement is observed. This is interpreted as a normal fault with the southern block downthrown. A reflector was not detected along the last 1,800 feet (SP 485 to 665) of Line 2 due to the previously discussed loss of reflectivity in this area. The three offsets identified by evaluation of the seismic reflection data were further assessed in light of other data before preparation of the geologic map of CSSA.

8.3.4   Interpretation of Area Well Logs

Throughout the project, driller�s records and logs were requested from various sources, including files at TWDB, TNRCC, and EUWD. These logs provided significant data and allowed evaluation of the area around CSSA. An index of over 100 well logs and a location map are presented in Appendix L. The index lists the wells by owner names and state codes, and notes if a lithologic or geophysical well log is on file. If only a well location is known, then only the well name is shown in the index.

The majority of logs on record were lithologic driller�s logs and natural gamma logs. For a few wells, resistivity and spontaneous potential were also logged. The primary reason for drilling and logging through the years was installation of water wells, lithologic logs contributed information required by state regulations, and gamma ray was used to determine depth to shale formations. In the Travis Peak Formation of the area around CSSA, no major sand bodies that might contain naturally occurring radioactive potassium are suspected. However, the gamma logs typically showed characteristic increases in gamma counts throughout shale layers in response to the naturally occurring radioactive material in shales. Therefore, gamma log characteristic increases in gamma count are used for delineation of the top and bottom of shale layers.

With regard to the CSSA groundwater investigation, data of interest included depths to major lithologic boundaries (e.g., the boundary between lower Glen Rose Limestone and Bexar Shale) from gamma logs, characteristic "packets" interpreted as shale beds for comparison between logs, interpreted thicknesses of strata for comparison to regional average thicknesses, and where present, depths of casing from resistivity logs. Most scales of gamma counts were 0 to 200. Inconsistency in the well log scales made detailed evaluation of the logs critical to interpretations.

This project relied upon gamma log data from various well logs to compare top and bottom elevations of the Bexar Shale for delineation of possible faults. As some wells were cased to below the top of the Bexar, a secondary pick was the contact between the bottom of the Bexar and the top of the Cow Creek Limestone. During preliminary assessment of the logs in preparation for a technical presentation to EPA and TNRCC on October 19, 1996, the data was interpreted as showing an offset of 70 feet in the Bexar Shale between the Poetchke well just west of CSSA, and CSSA wells 9, 10, and 11 on the upthrown side of a north-south fault (see well location index map in Appendix L). Though the GPR profiling indicated subsurface disturbances on the west side of CSSA, and the seismic reflection data suggested an offset in the central portion, no surface expression of a major fault had been detected. Therefore, pertinent logs were digitized to a common scale and then compared again. Subsequent review of the logs by several geologists and geophysicists resulted in interpretations that did not indicate any faulting between the Poetchke well and CSSA wells.

Interpreted depths to major lithologic boundaries are shown in Table 8.3-1. This information was used to prepare cross-sections and to aid in interpretations for the geologic map (Section 8.3.5) and cross-sections discussed in the summary of the hydrogeologic investigation (Section 8.6).

8.3.5   Geologic Investigation Results

The geologic survey culminated in a surface geologic map based upon interpretations of all the surface and sub-surface data collected up to that point (Figure 8.3-4). Seismic data and resolution indicated all displacements were 20 feet or less. The GPR data provided more detailed near surface resolution, although in many cases the underlying cause of a feature was not clear. Geophysical data was used to support surface observation and well log interpretation.

No major displacements are evident at CSSA though a series of small fractures are prevalent in the northern and western areas of the inner cantonment. Data from independent investigative techniques match fairly well (and in several cases coincide):

This correlation between four independent data sets provides strong evidence that a zone of small displacement faults and fractures exists in the vicinity of well 16 and SWMUs B-3 and O-1. This zone is expressed as a variety of offset relationships. Two fault traces 300 and 700 feet south of well 16 are downthrown on the south side. Based on seismic data, the fault closest to well 16 exhibits approximately 30 feet of displacement where it crosses Moyer Road and greater than 13 feet of displacement about 1,700 feet to the southwest where it is interpreted to intersect Salado Creek. This displacement along Salado Creek is based upon bed orientation and index fossil occurrence which indicate a reversal of throw, most accurately described as a scissor fault. The fault trace, located 700 feet south of well 16, was interpreted from a GPR offset recorded along Moyer Road and index fossil positions at Salado Creek. A small fault segment was observed 1,400 feet southwest of well 16 along Salado Creek and just south east of a Corbula martinae outcrop. The two fault traces that run south of O-1 were identified by seismic displacement at Salado Creek and GPR (exhibiting a 2-foot displacement down to the south) at 200 feet east of Moyer Road.

Table 8.3-1 - Geologic Unit Thickness Determined by Geophysical Log Interpretation

Additional corroborating evidence for a fault zone across the northern inner cantonment is found in two master�s theses from University of Texas at San Antonio, which indicate the possibility of faulting through mapping west of CSSA (Hefner, 1993) or through interpretation of lineaments on aerial photographs and topographic maps (Waterreus, 1992).

The southern fault zone extending across CSSA is based on a 65-foot seismic displacement along the southern part of Barnard Road, the absence of index fossil occurrence, and a clear surface expression of a fault along a creek drainage in the southwestern most corner of the inner cantonment. All the interpreted faults at CSSA occur in the north and south zones discussed above. The only exception is one small 4-foot displacement directly in between these zones and 400 feet west of East Outer Drive, which was interpreted from GPR. The fault zone interpreted from surface observation, well log, and seismic in the southern portion of the inner cantonment is geologically significant. Additional data exists in the form of mapped units and structure that show southwest-northeast displacement of the Glen Rose in the southern area of CSSA (Geologic Atlas of Texas, 1983). However, the distance of the zone south of SWMUs B-3 and O-1, in addition to lack of evidence for communication with the fault zone in the north inner cantonment, limits the need for the well 16 investigation to focus on this zone at this time.

Surface exposures are areally limited and geophysical data provides information that is laterally limited. As a result the degree and extent to which the faults are interconnected is unclear. However, their presence is significant in that they represent optimum pathways for groundwater flow and contaminant migration. Also, regional flow may alter dramatically in the vicinity of these zones (Figure 8.2-7 and Section 8.2 discussion).

8.4 - Lower Glen Rose Monitoring Well Installation

The only deep drilling performed at CSSA was in April 1996 for the purpose of lower Glen Rose monitoring well installation. Previous drilling of soil borings to a maximum depth of 30 feet bgl had generated little stratigraphic, structural, or hydrogeologic data (Section 5.2). Thus, collection of additional geologic and hydrogeologic information during this task was considered as significant as completion of the groundwater monitoring wells. The geologic and hydrogeologic data was compiled by use of geophysical borehole logging with natural gamma and caliper tools, geologic logging of drill core and cuttings, and collection of water samples for chemical and aqueous geochemical parameters. This data was used in interpretations of site-specific geology and hydrogeology.

Monitoring well installation at CSSA began on April 1, 1996 with the commencement of drilling at PH1. Two pilot holes were drilled and two monitoring wells were subsequently installed. Methods used to install the monitoring wells are described in Section 3.5. Drilling and monitoring installation and development was performed by Geoprojects International, Inc., under supervision of the on-site Parsons ES geologist. The task was completed on April 15, 1996 when the monitoring wells were developed. Observations made during performance of this work are described below.

8.4.1   Pilot Hole Drilling

Two pilot holes were drilled in the inner cantonment area at locations downgradient of the source areas and where intersection of faults within the basal saturated portion of the Glen Rose Formation was expected based on previous nonintrusive investigations. PH1 and PH2 are located in the northeast area of the inner cantonment as shown on Figure 8.3-4.

PH1 drilling initiated with a 6-inch diameter tricone rock bit. Lithologies encountered were logged from cuttings as shown on the drilling log in Appendix B. Interbedded limestone and marl of the lower Glen Rose Formation was found to overlie the Bexar Shale. No perched water was encountered at PH1, though moisture was encountered at various depths at and above 107 feet bgl. These moist zones included intervals at 7.5 to 14 feet bgl, at 65 to 70 feet, and at 107 feet bgl where there appeared to have been a fracture. Drilling at PH1 was halted at a depth of 140 feet bgl on April 1, 1996 to see if any water would come into the hole overnight. A weighted measuring tape was dropped down the hole to check for water the following day. The end of the tape, when retrieved to the surface, was observed to be muddy on one side and dry on the other. No perched water was present at PH1 on April 2, 1996, but there had no been a significant rainfall at CSSA since November 1995. Slightly moist weathered zones encountered at and above 107 feet bgl would probably be zones of perched water during a rainy season. Since no perched water was encountered, a temporary surface casing was not installed at PH1 before drilling deeper.

Drilling at PH1 from 140 feet bgl to total depth encountered no moisture until 255 feet bgl. Cuttings from 270 feet bgl were moist to wet, and those below 275 feet bgl were saturated. As logged from cuttings, the Bexar Shale was encountered at a depth of 326 feet bgl. Drilling continued to a depth of 361 feet bgl to allow for ample depth for the 20-foot long sonics tool to sense the top 15 feet of the Bexar Shale. Drilling of PH1 was complete at 3 p.m. on April 2, 1996.

Pilot hole 2 was drilled on April 4, 1996 with a 6-inch diameter tricone rock bit. Lithologies encountered were logged from cuttings and are shown on the drilling log in Appendix B. Interbedded limestone and marl of the lower Glen Rose Formation was encountered. No moisture was observed at PH2 until the aquifer was encountered at a depth of 297.5 feet bgl. However, in order to collect rock samples from the saturated basal portion of the Glen Rose for physical analysis, core was collected from 300.5 to 310.5 feet bgl with a 10-foot long, 2 1/2-inch ID split core barrel. To sample the contact between the lower Glen Rose and the underlying Bexar Shale, core samples were collected from 300.5 to 360.0 feet bgl. The behavior of the rig during drilling suggested that the contact between the lower Glen Rose and the Bexar Shale at PH2 is at 257 bgl. However, examination of PH2 core samples shows the rock below 257 feet bgl to be a marly limestone without the fissile texture or high mud content indicative of shale.

8.4.2   Logging of Drill Core and Cuttings

The drilling logs for PH1 and PH2 are presented in Appendix A. A summary of observations from the logs of the drill core and cuttings is as follows.

With the exception of about 3 feet of brown to black clay at the surface, the predominant lithology logged in each borehole is marly limestone. More competent and hard limestone is logged at 3-32 feet, 37-51, and 270-361 feet in PH1, and at 3-17, 19-3, 32-50, 66-104, 174-184, and 191-348 feet bgl in PH2. Evidence of fractures was limited to observations of drill rig response, such as sudden drops in the drill bit. These drops were noted at about 7.5-9.5, 107 and 261-265 feet bgl in PH1, and at 20-23, 207-208, 294, and 350 feet bgl in PH2. Slightly moist to moist zones were detected in both pilot holes, zones that may be perched zones during rainy seasons. Moist zones were detected in PH1 at 7.5-9.5, 17.5, 23-25, 27, 37-51, 107, and 255 feet bgl, and in PH2 at 0-3 (clay soil), 14-17, 23-32 (moist clay seams), 66.5, 69, 84.5, and 102 feet bgl. The saturated portion of the lower Glen Rose was encountered in PH1 and PH2 at 275 and 297.6 feet bgl, respectively. Bluish-gray shale was identified from cuttings at 326 to total depth (360 feet bgl) in PH1, and a medium gray marly limestone was logged from core samples at 348 to total depth (361 feet bgl) in PH2.

The core samples collected from 300.5 to 360.5 feet bgl in PH2 provided insight into the upper 50 feet of the Bexar Shale at that locality. Based on drill cuttings and geophysical logging of PH1 which indicated that the top of the Bexar Shale is at 296 feet bgl, the cored interval in PH2 was selected to span the contact between the base of the lower Glen Rose and the top of the Bexar Shale. Logging of core samples from PH2 showed that the predominant lithology was limestone. The gray color became increasingly darker from 348 to 360.5 feet bgl, with increasing amounts of shale-like material observed in the core. However, the major lithology was limestone to total depth. The interval of 300 to 310 feet bgl was very fossiliferous, with coral, bryozoans, and pelecypods, and very porous. From 311 to 348 feet bgl, the limestone was light gray with signs of recrystallization such as stylolite and infilling of vugs. The core showed increasing amounts of marl from 348 to 357 feet bgl, becoming limey shale from 357 to 360 feet.

Based on these observations, the top of the Bexar Shale in PH1 and PH2 appears to be beneath the total depth of the boreholes at about 360 feet bgl. However, the Corbula bed (top of the lower Glen Rose) is known to outcrop at the surface about 300 feet northeast of PH1, and the dip of the beds measured at the surface is typically 1 to 2 degrees southeast to southwest. If the average lower Glen Rose thickness of 300 feet is used, then the top of the Bexar Shale should be about 320 feet bgl, or a fault exists between the Corbula bed outcrops and the location of PH1 and PH2. To address this issue, geophysical logging was next evaluated.

8.4.3   Downhole Geophysical Logging

8.4.3.1   Field Actions

The methodology for geophysical logging and interpretation is presented in Section 3.3. Parsons ES arranged geophysical logging of the two pilot holes, and CSSA well 11 as its pump was temporarily removed, with EUWD.

On April 2, 1996, EUWD logged CSSA well 11 with an analog recorder, but the peaks were truncated. On April 3, 1996, Geoprojects personnel decontaminated EUWD�s tools and cables and then allowed EUWD to use the rig to lower the natural gamma and caliper tools. EUWD ran natural gamma and caliper tools down three holes (PH1, MW2, and well 11).

8.4.3.2   Interpretation of Logs

The gamma logs for well 11 and pilot holes PH1 and PH2 were comparable to logs reviewed from other wells in the area discussed in Section 8.6. Typical increases of 50 to 100 gamma counts are observed at about 300 and 326 feet bgl in PH1 and PH2, respectively, indicating the top of the Bexar Shale. These pilot holes extend about 50 feet into the Bexar Shale. In well 11, the gamma count significantly increases at 422 feet and decreases at 504 feet. These depths are interpreted as the top and bottom of the Bexar Shale, and the 504-foot depth as top of the Cow Creek Limestone.

Depth to the contact between upper and lower members of the Glen Rose, the Corbula bed, is difficult to determine. Using the top of the Bexar Shale determined from increases in gamma counts, and an average thickness of 300 feet for the lower Glen Rose, a small increase in gamma can be observed. However, without the 300-foot thickness to measure from the definitive Bexar Shale packet, the small increase in gamma can also be interpreted as part of the limestone and thus is not considered an absolute pick for correlations between wells.

Caliper logging of well 11 showed total depth of casing to be 287 feet bgl. A void space at 8 to 10 feet bgl was observed in PH1 but not in PH2, while a void of over 4 inches was observed at 146 to 148 feet bgl in PH2 but not PH1. A ledge inside the PH1 borehole was indicated at about 274 feet bgl. Neither caliper log indicated large voids in the saturated zones, while in well 11 in the void space was 2 to at least 4 inches beyond the borehole throughout the saturated zone. Caliper logging of wells 16 and D in 1994 by Century Geophysical, Inc., indicated voids up to 2 inches beyond the borehole at 280-foot depths and at total depth in well 16. The caliper log results suggest that wells 11 and 16 are in solution-enhanced channels, while the PH1 and PH2 boreholes are not in solution enhanced areas, or at least those areas are not discernible at a scale of inches.

8.4.4   Monitoring Well Completion

The depth of 140 feet was selected for total depth of surface casings because previous downhole camera surveys had detected perched waters above that depth, and because the depth was shallower than the highest historic water levels measured in the lower Glen Rose member of the middle Trinity aquifer at CSSA.

Constructing an open hole well is a simple matter of setting a surface casing, drilling down to water, constructing the surface pad, and developing the well. Completion of the MW2 involved cleaning cuttings and sand out of the tail hole, and then after development, constructing the surface completion as described in Section 3.5.3.4. MW1 was drilled and completed as a new hole, which made it easier to set the surface casing since there was no tail hole.

8.4.5   Monitoring Well Development

Monitoring wells were developed as described in Section 3.5, and the well development records are in Appendix C. Airlifting started immediately after the wells were drilled (MW1) or cleaned out (MW2). Development by pumping was performed until the water was clear and measured parameters had stabilized. Both wells were pumped. From MW1, approximately 575 gallons were removed by airlifting and 461 gallons were removed by pumping at a rate of 5.66 gpm. At least 490 gallons were airlifted from MW2 and 578 gallons were pumped at a rate of 5.45 gpm. The final stable parameters were temperature of about 70�F, pH that ranged from 7.1 to 7.38, and specific conductivity values ranging from 5,260 to 5,690 mS/cm.

8.4.6   Monitoring Well Sampling

The two new monitoring wells were sampled on April 16, 1996, in accordance with the procedures discussed in Section 3.9.4. Because the wells were developed to stable parameters on April 15, and the wells are constructed open borehole that is considered to be in equilibrium with the surrounding aquifer and thus not in need of traditional purging used to purge stagnant water from filter packs, the 90- to 95-foot water columns were not purged. An initial sample was also collected from pilot hole PH1 for analysis of VOCs. Sampling was through use of a clean Teflon bailer for all samples.

Table 8.4-2 - Aqueous Phase Constituents in Well Samples

8.4.7   IDW Characterization

The IDW generated from the monitoring wells 1 and 2 installation include waters associated with development of the monitoring wells. A composite sample water produced from PH1 during drilling was collected for IDW characterization (Table 8.4-3). No volatile organics were detected by SW8260 in the IDW sample, so 770 gallons of PH1 groundwater contained in fourteen 55-gallon drums was discharged to the ground on April 3, 1996. A second sample was composited from five 55-gallon drums of development water from MW2. The results show that the waters did not contain any chemical constituents of concern, and thus were disposed of on the ground near well MW2.

Table 8.4-3 - Analytical Results of MW1 and MW2 Purge Water and Decontamination Water

8.5 - Future Waste Water Disposal

CSSA amended their NPDES and TNRCC wastewater discharge permits (Parsons ES, 1995g). The amendments included the addition of a groundwater treatment plant to treat groundwater in the future. The existing wastewater treatment plant at CSSA discharges to a tributary of Leon Creek and is identified administratively as Outfall 001. The proposed groundwater treatment plant for CSSA would discharge to Salado Creek and is identified administratively as Outfall 002. Both Outfall 001 and Outfall 002 have an allowable discharge of 30,000 gallons per day, i.e., a total of 60,000 gpd.

The NPDES permit amendment application was submitted in April 1995 and is presently being reviewed by EPA Region VI in Dallas, Texas. The TNRCC Industrial Low Potential Impact Permit application was submitted in July 1995. A draft permit was issued in January 1996. The permit was signed by the executive director of the TNRCC on June 7, 1996, and the final TNRCC permit number 03849 is expected to be issued by June 14, 1996.

8.6 - Summary of Hydrogeologic Investigation

The hydrogeologic investigation at CSSA encompassed use of several disciplines in order to better define not only the groundwater contamination but also the physical regime in which the contamination is located. The investigation evolved alongside the source characterizations, and all are considered integral to an understanding of the problem and what actions are most appropriate to resolve the issues involved. The primary sources of data used to determine not only the hydrogeology at CSSA but also to help delineate contaminant migration trends, were groundwater monitoring, soil gas concentration trends, field mapping, geophysical surveys, and downhole geophysical well logs. Data from the different tasks of the investigation correlate fairly well, and the following sections sum up significant information achieved from the investigation.

8.6.1   Site Geology

8.6.1.1   Stratigraphy

This report presents data from the first in-depth field mapping at CSSA. Previous geologic interpretations had only regional information, which indicated that the upper Glen Rose was the primary surface formation, with minor surface exposures of the lower Glen Rose at the northern and southern areas of the facility. The mapped index fossil localities clearly show that the boundary between the upper and lower Glen Rose members exists at CSSA, and beds located higher in elevation above the boundary are interpreted as upper Glen Rose where they are not associated with Holocene alluvial deposits. Figure 8.3-4, the geologic map, therefore depicts interpreted boundaries between the upper and lower members of the Glen Rose, with the lower Glen Rose found in the south-central area of CSSA.

Surface geologic features at CSSA are determined through location of key stratigraphic fossils, strike and dip measurements, and faults versus fractures. Stratigraphically, only three formations are found at the facility - isolated Holocene alluvial stream deposits (not depicted on Figure 8.3-4) and the upper and lower members of the Cretaceous-age Glen Rose Formation. Interpretation of the data shows that at the top of the lower Glen Rose, the 1- to 3-foot thick Corbula bed was delineated in the central portion of the facility, and that the upper Glen Rose is interpreted above this marker bed. The Corbula bed was recognizable in stream beds and outcrop only. On hillsides, the bed apparently weathers to the point that it is not recognizable except as occasional broken pieces at the surface (float), found mostly in streambeds. A second key stratigraphic fossil, Salenia texana, is found below the Corbula bed. Large sequences of limestone seen in outcrop were not necessarily recognizable as upper or lower Glen Rose if one of the two key index fossils were not known. Therefore, the interpretation of surface geologic outcrops relies on the location and orientation of the index fossil beds.

Strike and dip of bedding planes in the upper and lower members of the Glen Rose Formation, summarized in Appendix K, show that the bedding orientation was found to vary throughout the study area. Average strikes of N47E and N40W were identified from a stereonet projection. Dips of the beds are low and average 5 degrees.

Lithologic and geophysical logs for wells 2, 16, D, PH1 and PH2 show that the top of the Bexar Shale is found in the north-central portion of CSSA at depths of 314 to 326 feet bgl. The average thickness of the lower Glen Rose is 300 feet. Thus, subsurface geologic formations known to exist beneath CSSA include the Glen Rose Formation, the Bexar Shale member of the Travis Peak Formation, and by inference of horizontal superposition, the underlying Travis Peak Formation members (Cow Creek Limestone, Hammett Shale, Sligo Limestone, and Hosston Sand) which overlie pre-Cretaceous rocks (see Table 2.5-1).

8.6.1.2   Structure

Structural features mapped in the field are faults, fractures, slickenslides, and several small scale folds. Three quarries were also inspected whose presence can imply but not confirm areas of structural weakness in the limestone beds. Two northeast-southwest trending fault zones are observed - one through the central portion of CSSA near well 16, and the other near the southern boundary of CSSA (Figure 8.3-4). One set of fault strike and dips has an average orientation of N68W, 50NE, and the other averages N74E, 68SE. These fault measurements compare favorably with the regional trends (Simpson Company and Guyton Associates, 1993).

In order to summarize the subsurface geology at CSSA, it was necessary to evaluate formation thicknesses interpreted from geophysical well logs and from published data. Table 8.6-1 lists thicknesses of geologic units of interest at CSSA. At the surface, the Corbula bed and Salenia zone provide the best marker beds for the top of the lower Glen Rose member. In geophysical well logs, the Bexar Shale unit served as a common marker on geophysical logs from which other stratigraphic units could be "picked" using typical thicknesses. Because published thicknesses for these units vary, the most recent publication that also evaluated other previous data (Simpson Company and Guyton Associates, 1993) was utilized in interpretations of subsurface geology. Other critical data from CSSA, Camp Bullis, and private and municipal area wells were used in conjunction with surface mapping data to develop the structural interpretations upon which cross-sections and the geologic map of CSSA (Figure 8.3-4) are based. Appendix K presents pertinent geologic data (table of strike and dip measurements at CSSA, bed thicknesses determined from geophysical log interpretations, and a comprehensive table of well data used to prepare cross sections), and Appendix L contains an index of well logs and location map.

Two cross sections summarize the subsurface geology at and around CSSA. Figure 8.6-1 (cross section A-A�) is a northwest to southeast-and-south section shown generally perpendicular to the regional dip of the beds, thus noted as the "dip section." Figure 8.6-2 is a southwest to northeast section (B-B�) shown along the general strike of the beds and is the "strike section." The lines of section are shown in the upper right corners of the diagram, and the well locations noted. The color schematic for the upper and lower members of the Glen Rose is the same used in Figure 8.3-4. Along the vertical line indicating wells to total depths, interpreted top and bottom depths and elevations of the major formations are shown, as well as the highest and lowest water levels from CSSA wells. Because the fault zones at CSSA have displacements ranging from surface measurements of 1 foot to interpretation of geophysical well log data indicating displacements of less than 20 feet in a general area, the fault zones are shown as a gray area with the beds offset on either side. Particularly within the northern fault zone, the data generated from the hydrogeologic investigation, such as 1-foot offsets of the Corbula bed mapped in the field, suggests that several faults exist in the zone but are not discernible at the scale of the diagram.

Both sections extend horizontally through CSSA, and the vertical exaggeration is 20x. The cross sections extend vertically from the upper member of the Glen Rose Formation to the Hammett Shale member of the Travis Peak Formation. The bottoms of the sections depicted are intended to show the lowest member of the middle Trinity aquifer. The lower Glen Rose member is interpreted with a 300-foot thickness, underlain by the 70- to 84-foot-thick Bexar Shale, the 66- to 70-foot-thick Cow Creek Limestone, and the Hammett Shale, estimated at 50 feet thick based on published information. The elevation of the upper Glen Rose member is based on either mapped locations of the Corbula bed or interpreted as 300 feet above the contact between the Bexar Shale and the lower Glen Rose member.

Cross section A-A� shows the regional dip of beds interpreted between well logs at 10 to 12 degrees to the south-southeast. The two fault zones shown in the geologic map (Figure 8.3-4) are depicted in this section, as is the small graben-type offset interpreted from seismic reflection survey data (shot points 310 to 325) about 1,100 feet south of well 16 (Section 8.3.3). A fourth area of faulting at the south end of the section is a fault or fault zone located offsite in the area of The Dominion, about 1.7 miles south of CSSA. This fault correlates with one of the faults shown on a cross section extending north and south of CSSA over 8 miles along I-10 (Simpson Company and Guyton Associates, 1993). Because this fault is interpreted between two wells which show an offset in elevations of the top of the Bexar Shale, the fault is shown on this diagram as a zone in which the exact surface location is not known. The fault is included in the CSSA cross section A-A� to verify the presence of faults with displacement up to the southeast as well as many faults along the Balcones fault zone which have displacement down to the southeast.

Of particular interest is the area between and beneath CSSA wells 16, MW1, MW2, and CSSA well 1. The northern fault zone shown in the geologic map is represented in this area, as are the two source areas of groundwater contamination, SWMUs B-3 and O-1. The faulting in this area, though not of great magnitude such as 100 feet of throw that would displace a relatively impermeable shale against a more permeable limestone, is significant in that pathways for groundwater contamination are more complex than in unfaulted areas. Even with the vertical exaggeration of 20x horizontal, the diagram shows the short horizontal distance (400 to 1,200 feet) compared to the depths (at least 150 feet to groundwater) over which PCE, TCE and DCE probably have traveled between SWMUs O-1 and B-3 to well 16. When the karstic geology regime is taken into account, in which contaminants can migrate along solution enhanced fractures as well as faults, joints, and bedding planes, determination of the exact pathways that the dense nonaqueous phase liquid contaminants (DNAPLs) PCE and TCE have migrated becomes more difficult if not impossible.

As discussed in Section 8.3.5, draft versions of a southwest-to-northeast cross section (B-B�) indicated a displacement between the Poetchke well and CSSA well 11. More detailed examination of geophysical well logs did not find such displacement, and Figure 8.6-2 shows the resulting interpretation.

Looking at the subsurface geology along strike on cross section B-B� shows a slight dip of bedding planes to the southwest. The section shows one fault between well 2 and PH1, which correlates to an interpreted fault from surface mapping and the GPR profiling. A minor displacement of about 4 feet is noted between elevations of the Bexar Shale. The northern fault zone is interpreted as southeast of well 2 on the line of section and just northeast of well 16 (see Figure 8.3-4 for reference). Because the line of section is about 35 degrees from parallel along the northern fault zone, the width of the fault zone looks much wider in Figure 8.6-2 than in Figure 8.6-1. Interestingly, the April 1996 water level measurements between wells 2 and MW1 also show a decrease in potentiometric surfaces. When viewed from well 2 to MW1 along the projection, the direction of this decrease is to the southeast.

8.6.2   Site Hydrogeology

[Next Section]