2.3.4   Work Plan Deviations

The Dunham Carbonate-rock Classification System (1962) was used to describe the limestone and dolomite lithologies encountered in the subsurface. A Munsell Color Chart was employed to consistently express the color and hue of the retrieved core. The core was described with respect to porosity, faults, fractures, fossils, and texture; all characteristics which aid in correlation between wells.

A photoionization detector (PID) was used on the core to check for VOCs at the time of retrieval. The lithologic and well construction logs for each well can be found in Appendix A. Additionally, a data CD with relevant electronic files for the lithologic and well construction logs is included with the appendix.

2.7 - Monitoring Well Development

Monitoring well development was performed by air-lifting, bailing, and pumping. Each well was surged immediately after reaming to the final drilling depth. The wells were bailed after the well screen and 4-inch casing were installed, sealed, and grouted. Following bailing, the wells were developed by pumping for several hours.

2.7.1   Air-Lifting

Air-lifting was performed after a well had been reamed to its final depth, but before casing insertion. At that point, the bottom portion of the well was still an uncased borehole. Compressed air was injected downhole to within 20 feet of the total depth of the well via the drill pipe string. This process flushed out drilling foam remnants and the majority of loose heavy sediments produced during final stages of reaming. At selected intervals, the driller jetted the well by releasing bursts of air pressure in the saturated column, thus causing the sediments to become suspended and airlifted to the surface where they collected in the rig containment pit. This process was also used to clean out well holes before each stage of casing installation in the double-and triple-cased wells.

2.7.2   Bailing

New wells were bailed after a minimum of 48 hours had passed after grouting, and prior to any pumping. The apparatus used was a 6-foot long, steel dart-valve, 3-gallon capacity bailer. Drilling subcontractor personnel operated the bailing apparatus. A pump installation truck (Smeal�) was backed over each well and the bailer was lowered and raised by a motorized cable reel. The bailer was first gently lowered to the bottom of the constructed well, and then raised several feet. The cable was marked at that point so the operator would have indication when the bailer approached the bottom of the screen. This allowed rapid descent of the bailer in the well while preventing the heavy bailer from striking the screen bottom. In general, the bailer was lowered down into the well screen and then quickly raised to surge the screened interval. This draws out fine particulates from the sand filter and surrounding borehole walls.

Bailing helped remove sediments that may have been difficult to remove by pumping. Parsons geologists occasionally monitored color, odor, pH, conductivity, and specific conductivity of bailed groundwater. It was not necessary to achieve stabilization of these field parameters at this time. Bailing time averaged approximately 4 hours per well, as per Work Plan specifications. Groundwater in some wells became clear quickly and were bailed only a few hours; others remained highly turbid and were bailed for more hours over several efforts. In general, the BS wells exhibited greater turbidity during development than LGR or CC wells.

2.7.3   Pumping

Once most sediment had been removed by air-lifting and bailing, well development was completed by pumping. Each completed monitoring well was pumped with a decontaminated, 1-horsepower Grundfos submersible pump rated at 5 gpm. The pump was attached to flexible tubing and hand-lowered by drilling subcontractor personnel. The pump was suspended within the screened section, approximately 2 feet off the bottom of each well. Power for the pump was supplied by a portable generator provided by CSSA.

The development discharge of each well was monitored by flowmeter. Field parameters, including turbidity, odor, temperature, pH, conductivity, and specific conductivity, were monitored every 30 minutes in the first 2 hours of pumping, and hourly thereafter. This process continued until the water removed from the wells was clear and field parameters stabilized, and the volume withdrawn surpassed the estimated volume of water injected during drilling. Stabilization was achieved when the water appeared foam-and sediment-free, turbidity remained within 10 nephelometric turbidity units (NTUs), temperature was +/- 1.0˚C, pH +/- 0.1 units, and conductivity +/- 5%, all for a period of at least 30 minutes. Pumping was normally maintained for at least 8 hours, even after the above conditions had been met. Groundwater from wells that showed trace levels of contamination was discharged into rolloff containment for characterization.

2.8 - Geophysical Borehole Logging

A total of 18 geophysical logging runs were made in open coreholes/boreholes at all 15 drilling locations (multiple runs at CS-MW8 and CS-MW10). These logs yielded information on borehole size changes, shale content, and formation conductivity. The drilling firm contracted GeoCam, Inc. (GeoCam) of San Antonio, Texas to perform the geophysical logging services.

Direct readouts were obtained on site during the logging process. Results of the logging were then used to determine the depth and thickness of stratigraphic units. In addition, information on lithology and stratigraphy was obtained from the borehole geophysical data. The borehole geophysical logs for each well can be found in Appendix B (not available electronically).

Borehole geophysical techniques employed for this project included short and long resistivity (8-inch, 16-inch, 32-inch, and 64-inch), spontaneous potential (SP), natural gamma ray, caliper logging, and a downhole camera. Geophysical logging was performed in boreholes to identify soil/rock types before surface casing was installed and packer tests performed. Geophysical data were stored in both hard-copy and electronic formats. Electronic data from the geophysical surveys can be found on the CD in Appendix A.

2.8.1   Electric Logs (Resistivity and Spontaneous Potential)

Electrical resistivity logging was conducted in conjunction with SP logging. When used together, these methods are commonly referred to as electric logs (�E-logs�). The normal resistivity tool used four electrodes to aid in the correlation of stratigraphic layers. An 8-inch and 16-inch separation of the electrodes was used during logging of the coreholes for the short normal log, and a 32-inch and 64-inch separation was used for the long normal log. The larger spacing delivers deeper penetration but lower bed resolution. Silt, clay, and shale typically have low resistivity, while sandstone and limestone saturated with fresh water have the highest resistivity values.

The SP tool records the electrical potential produced by the interaction of formation water, conductive drilling fluids, and certain ion-selective sediments. Although the SP curve is not a measurement of permeability, a deflection is generated when permeable formations come into contact with the drilling fluids, and a baseline curve (no response) is formed when a nonpermeable formation (shale or clay) is encountered. Individual bed boundaries and thicknesses can be differentiated on the curve, making stratigraphic correlation more accurate and representative.

2.8.2   Natural Gamma Ray Logs

Natural gamma ray logging was used to estimate lithologic characteristics of geologic formations by recording gamma radiation emissions. The gamma ray logging tool contains one or more scintillation detectors which measure natural radioactivity in soil/rock layers adjacent to the borehole. Gamma logging may be used in conjunction with SP, resistivity, and caliper logs in fluid-filled boreholes. Because radioactive elements tend to concentrate in clays and shales, the log normally reflects the shale content of the formations. Low clay intervals such as grainstone usually have a very low level of radioactivity.

2.8.3   Caliper Logs

Caliper logs measure variations in the borehole diameter. The caliper, a spring-loaded mechanical device with one to four adjustable arms that press against the borehole wall, measures the diameter in cased and uncased boreholes. Variations in the borehole diameter, factors such as borehole erosion (washout), and the presence of swelling clays or resistant strata, can be identified with the caliper. The volume of filter pack or grout needed for well completion can also be determined using this tool. Caliper logs are conducted by lowering the device to the bottom of the borehole and recording the measurements as the caliper is raised.

2.9 - Injection Packer Tests

2.9.1   Theory

Hydraulic conductivity (K) of the formation is an integral parameter in estimating the groundwater flow velocity and the formation yield capacity to a well. The methods for deriving hydraulic conductivity can be determined by grain size estimations, permeability testing of core in a geotechnical laboratory (by American Standards for Testing and Materials [ASTM] method 2434), or by field measurements utilizing slug tests, packer tests, specific capacity, and pumping tests. Each method has its advantages and disadvantages. For example, the disposal of a large quantity of contaminated groundwater is of concern in a pumping test even though this is the best method for most accurately estimating K. For this DO task, the drill-stem packer test was specified in the SOW.

A packer test is a permeability test conducted in a single borehole by sealing off one section of the borehole at a time with pneumatic rubber packers and pumping water into the isolated section. Packer tests can be performed at various depth intervals in an open hole. Thus, it may identify vertical distribution of the high and low hydraulic conductivity layers. The permeability of the rock adjacent to the isolated section of the borehole is measured as a function of the pumping head (pressure) and the rate of water loss from the sealed section (Watson and Burnett, 1995). The lower the pumping head, the higher the formation permeability. The higher the water loss, at the same pressure, the higher the formation permeability.

The SAP (Volume 1-4, RL83 Addendum) provided the packer test design and analytical equation for computing hydraulic conductivity. Watson and Burnett (1995) quoted this method from the Ground Water Manual, �Pressure Permeability Tests in Stable Rock� (Bureau of Reclamation, 1977).

Packer tests were performed on each of the new coreholes before they were completed as monitoring wells. Using observations made during drilling and geophysical log data, lithologic intervals were selected for the testing of hydrogeologic characteristics. Depending on the DO, generally three to six packer tests were conducted per hole drilled. The packer test data for each well, along with the water loss and flow readings, are contained in Appendix C. The data are also included in the CD located in Appendix A.

2.9.2   Methodology

Prior to initiating the injection packer tests, the corehole was cleaned of all cuttings and drilling mud. Next, a water level in reference to the ground surface was obtained. The packers consisted of a 24-inch rubber sleeve which expanded against the wall of the borehole when pressure was applied. The testing of isolated sections of a borehole required the use of two pneumatic isolation packers separated by a length of perforated pipe. The spacing of packers (which govern the length of the test section) were generally between 5.5 and 12 feet apart. The minimum spacing of packers was governed by the following guideline:

S_p/D > 5

Where:

S_p = the spacing, or length between packers, and

D = the diameter of the borehole.

The double-packer assembly and inflating line were then lowered into the borehole on a string of pipe so the perforated section was opposite the interval to be tested. The length and diameter of injection tubing and the height of the pressure gauge were measured and recorded.

Prior to each packer test a check was made to ensure that the injection tubing was completely filled with water. The packers were inflated to pressures between 225 pounds per square inch (psi) and 380 psi. With the line valve closed, the pump was started and allowed to recirculate. A starting value for the flow totalizer and timepiece were recorded.

To initiate the test, the line valve was opened and the recirculation valve closed simultaneously. Both pressure and injection rates were measured and recorded over time throughout the test, with measurements being made as frequently as possible during the early moments of the test. The period of injection continued for up to 35 minutes. As the time of injection continued, changes in pressure and injection rate occurred more slowly; therefore, the frequency of measurement was decreased. The maximum measurement interval was normally 5 minutes.

The pumping times and pressures were dependent on the depth, requirements of the test, and nature of the materials being tested. As far as the length of time to run the test was concerned, the general rule was to run the test until an equilibrium condition was established. This was considered to have been reached when four or five readings of pressure and flow taken at approximately 5-minute intervals were essentially constant. Some test intervals were unable to inject water into the formation at the maximum pressure attainable (380 psi) with the pump used. Those test intervals are deemed to be impermeable, and are denoted as zones of �no flow.�

2.9.3   Calculation of Hydraulic Conductivity

An estimated value for K can be ascertained for a geologic media by measuring the pressure at which a discharge (Q) can be injected into a permeable formation. For the purposes of this study, Q and pressure recorded at the conclusion of the test were used for the estimation of K. As described in the Ground Water Manual, �Pressure Permeability Tests in Stable Rock� (Bureau of Reclamation, 1977), three conditions (or zones) can be tested, and each has its own testing method. The Zone 1 method is for testing the interval above the water table in the vadose zone. The Zone 2 method is for testing the saturated interval above the water table (e.g., in perched groundwater or capillary zones). Likewise, the Zone 3 method is for testing saturated intervals below the water table. For each injection packer test performed at CSSA, the test met the assumption that the zone tested was saturated and beneath the water table (Zone 3). With that assumption, K is calculated from the following formula:

K = Q/C_s*r*H

Where:

K = Hydraulic Conductivity (feet per second [ft/sec])

Q = Flow (cubic ft/sec)

A = Length of test section (feet)

r = Radius of the test hole (feet)

C_s = Conductivity coefficient for semi-spherical flow in saturated materials with partially penetrating cylindrical test wells (obtained from Chart B from Appendix C of Volume 1-4: RL83 Addendum using A/r as variables).

H = Effective head in feet (h₁ + h₂ � L) where:

h₁ = distance between the gauge and the water table (feet)

h₂ = Applied pressure at the gauge (1 psi = 2.307 ft of head)

L = Head loss in feet due to friction:

L =10.44*P_L*Q^1.85/c^1.85*d^4.865)

P_L = Length of drill pipe (feet)

Q = Flow (gal/min)

c = 140 (Pipe surface roughness coefficient)

d = Diameter of pipe (inch)

The computation of hydraulic conductivity was performed using a Microsoft Excel spreadsheet through the input of testing variables. The collected field data, test inputs, and computations are presented in Appendix C, and results are discussed in Section 4.2.

2.10 - Well Surveying

Horizontal and vertical control for the 15 new CSSA wells was established by Fisher Engineering (Fisher) of San Antonio, Texas. For each well, Fisher obtained the northing, easting, and elevation for the survey bolt affixed within the well pad, the notch at the top of the PVC casing, and natural ground (NG) elevations. A licensed surveyor performed surveying.

Control was brought in by using existing survey data, and using CS-D, CS-MW1-LGR, and CS-MW2-LGR as benchmarks. Those reference points had been established by prior surveying efforts by Northstar Land Surveying (to National Vertical Geodetic Datum [NVGD] 1983 and horizontal control to North American Datum [NAD] 1983) and Macias and Associates (NVGD 1927 and NAD 1929). All points required to control the survey were occupied as stations within a closed and adjusted traverse. The controls met or exceeded third-order accuracy standards.

Fisher completed the survey using a professional-grade Trimble XRS global positioning system (GPS), and reported all coordinate point data in Universal Transverse Mercator (UTM), Zone 14 North, NAD 1983. The northings and eastings were recorded in meters, and the elevations are reported in U.S. feet above mean sea level (MSL). Surveying data reported by Fisher are presented in Appendix D.

2.11 - Decontamination Procedures

All equipment that could directly or indirectly contact samples was decontaminated in designated decontamination areas. This equipment was prevented from contacting potentially contaminating substances such as motor fluids, engine exhaust, corroded surfaces, and dirt.

2.11.1   Downhole Equipment Decontamination

The drilling rig, bits, collars, drill pipe, core barrel, large bailers, casing, and all other drilling, downhole sampling, and logging equipment were decontaminated prior to insertion into new boreholes. This includes resumption of drilling in the same well after an overlying hydrogeologic or contaminated zone had been cased off. Decontaminated equipment was stored on plastic sheeting or on decontaminated steel racks until ready for use. Equipment was normally put to use as soon as possible after decontamination.

GPI constructed a large decontamination pad at its main staging area on the east side of Building 92, several hundred feet south of the CS-MW7 cluster location. The pad was constructed according to SAP specifications. This main pad was large enough to accommodate the drilling rig or trailer with a load of pipe or bits and collars. The pad area was partially enclosed by plastic sheeting to catch splashed wash waters. Very heavy gauge plastic lined the pad floor. The ground sloped slightly at the pad site and the pad was set up such that wash water accumulated at the rear, enclosed end of the pad. When enough waste water accumulated so it could be effectively pumped, it was transferred to larger IDW containment. Frequently, lesser amounts of decontamination waste water rapidly evaporated due to sunny, hot and dry conditions.

Rig parts and large equipment were positioned inside the decontamination pad to be thoroughly cleaned by high-pressure steam-cleaning. Brushing washed off stubborn dirt, and then steam-cleaning and rinsing were repeated over the scrubbed area. Casing sections were steam-cleaned inside and out after being positioned on pipe racks next to a well in preparation for installation. The steel well casing was delivered in generally clean condition. Wash water from steam-cleaning the casing was therefore not contained. The 4-inch stainless steel screen and inner PVC casing sections were certified clean by the manufacturer and not washed.

2.11.2   Decontamination of Logging and Sampling Equipment

All sampling equipment, split spoons, core barrel sections, trowels, and chisels were decontaminated prior to and during use with an Alconox^� soap scrub wash, potable water rinse, isopropyl alcohol rinse, and a final distilled water rinse. Decontaminated equipment not used immediately was allowed to air dry and then wrapped with oil-free aluminum foil for storage or transport. Smaller equipment was decontaminated at temporary pad sites erected at each cluster well location. Temporary pads consisted of multiple plastic sheeting layers laid on the ground. The edges were bermed to contain used decontamination liquids. Racks were placed in the center of the pad so equipment could be washed and kept off the ground. When enough waste water accumulated to be effectively pumped, it was transferred to larger IDW containment.

GeoCam decontaminated geophysical logging equipment under Parsons oversight before and after each logging run. The delicate equipment was rinsed with isopropyl alcohol and clean deionized water.

2.12 - Management and Sampling of Investigation-Derived Waste

Significant volumes of drill cuttings and water were generated during the course of monitoring well construction for this project. The water and cuttings mud were contained in aboveground mobile containers pending chemical analysis as per the RL83 Work Plan with Addenda. After proper characterization, the contained IDW was disposed accordingly. In the case of the RL83 and DO23 wells, all solid IDW was eventually classified as non-hazardous and, with concurrence of EPA, was disposed on site at CSSA. The location of the cuttings disposal area is shown on Figure 1.1. Generally, water and mud waste was discharged to the ground surface. Surface discharge of IDW was not conducted in areas near streambeds or surface water bodies. Those portions of groundwater exhibiting low concentrations of VOCs were treated at the CSSA GAC unit and subsequently discharged at permitted CSSA Outfall 002 (Texas Pollutant Discharge Elimination System [TPDES] Permit No. 03849). One separate container was dedicated for non-hazardous construction debris, refuse, and general trash.

The majority of containment consisted of rented 20- to 30-yard³ roll-off containers supplied and serviced by Eagle Construction and Environmental Services (Eagle). Between three to eight of these rolloffs were placed at each active drilling location and received cuttings and fluids generated during coring, reaming, and well development. Several containers were placed temporarily at the CSSA GAC unit to hold fluids destined for treatment. Several days were required to allow for settling of solids prior to GAC treatment. Liquid wastes were periodically transported to the GAC unit in the drilling subcontractor's 2,990-gallon vacuum truck and plastic tanks.

Initial containment of drilling wastes occurred at the wellhead in the form of a constructed impoundment surrounding the wellhead and rear of the drilling rig. The impoundment was erected above ground using wood planks and heavy gauge plastic. The cuttings, drilling fluids, and groundwater were expelled into the rig containment structure. When the rig containment filled, the fluids were pumped into adjacent leak-proofed roll-off containers. The roll-offs were prepared by applying silicone sealant to the door seam inside the boxes and then lining the roll-offs with heavy gauge plastic. The doors were tightly sealed across the door gasket. Only after proper preparation were drilling fluids pumped into the containers. The various roll-offs could hold volumes ranging from approximately 4,000 to 6,000 gallons. Once contained in the roll-offs the solids began settling, leaving clear water to accumulate above the sediment. Although mud accumulated below the water the two media only partially separated, with much of the mud remaining extremely fluid. Due to the liquidity of the mud, the roll-offs were prevented from becoming more than 50 percent full by mud to avoid spillage during uploading onto trucks. Mud particle sizes ranged from millimeters to sub micron, as evidenced by cloudiness of water after passing through 1-micron filters at the GAC unit. Heavier, nonpumpable cuttings remained in the rig containment and were periodically shoveled out into roll-offs.

Water samples were collected out of full roll-offs and analyzed for the VOCs of concern. When laboratory results showed the water to be uncontaminated, the water was pumped off and discharged to the ground surface. Mud samples were also collected and likewise analyzed. After drilling was completed at a location and the mud analytical results confirmed the material was not contaminated, the roll-offs were transported to an on-post site selected by CSSA and the mud contents dumped. The roll-off boxes were then relocated to the next drilling site and re-sealed, where the containment process was repeated.

When no contaminants were detected in roll-off water, the groundwater from the corresponding producing zone in a well was considered clean. With this knowledge, the additional water produced by reaming or developing the same intervals was pumped out of the corresponding roll-offs to ground surface after the majority of solids settled out. Once chemical analyses of groundwater showed the screened portions of a well to be clean, development water pumped from those wells was also discharged directly to ground surface.

Discrete and IDW sampling during drilling at CS-MW5-LGR, CS-MW7-LGR, CS-MW8-LGR, and CS-MW10-LGR revealed that groundwater from several zones within the LGR Formation contained low concentrations of TCE and/or PCE. In the case of CS-MW5-LGR and CS-MW8-LGR, the muddy drilling water was transferred via vacuum truck to sealed roll-offs at the GAC unit. The water was treated at the GAC unit to remove contamination and then released at the permitted outfall. The water contained high levels of suspended fine sediments and the 5- and 1-micron filters in the GAC system had to be frequently replaced. Development water from CS-MW8-LGR was treated at the GAC unit prior to discharge at the outfall. Sediments from the roll-offs located at the GAC unit did not exhibit VOC contamination.

When CS-MW7-LGR was pumped for development, this discharge was also contained. The water from CS-MW5-LGR, CS-MW7-LGR and CS-MW10-LGR remained in roll-offs for extended periods of time after sampling due to processing limitations of the GAC, which was originally not designed for large volumes of highly turbid water. Upon re-sampling of these roll-offs approximately 2 to 4 weeks after the initial sampling, laboratory results indicated no detection of VOCs and the water was released to the ground surface.

Mud cuttings contained at the first drilling location, CS-MW9, were treated with a polymer flocculant in an attempt to congeal the mud and separate it from the water. The flocculant did not cause much additional separation of solids, and water beyond that which occurred by gravity alone. As much water as possible had to be removed to prevent the weight of the mud from exceeding the maximum weight capacity of the containers, which could otherwise become too heavy for uploading onto the transport truck. The flocculent succeeded only in producing slightly gelled mud that maintained very low viscosity. After chemical analysis showed the solids to be uncontaminated, the CS-MW9 mud was discarded in the North Pasture at the direction of CSSA.

Ancillary wastes such as plastic sheeting, personal protective equipment (PPE), and disposable sampling equipment were segregated from other IDW and disposed as non-hazardous plant production waste in accordance with the EPA's protocols (Investigation-Derived Wastes During Site Inspections, 1991). Wastewater from decontamination operations collected in the decontamination pad sumps and was pumped into a plastic tank for transport to the GAC roll-offs, or was pumped directly into other nearby roll-off containers. Small amounts of dirt solids from washing equipment were put into roll-offs along with drill cuttings.

2.13 - Sampling

Over the course of the project various samples were collected to detect and monitor for the presence of contaminants. These samples included subsurface soil or rock samples from retrieved core, groundwater samples of formation water using either bailers or pumps, and groundwater samples from off-post wells. Data from soil/rock samples and off-post wells were analyzed to definitive QA levels, and 100 percent of the data were verified and validated by a chemist. Data for grab samples of formation groundwater collected during well installation required basic quality assurance/quality control (QA/QC) for analytical methods used, and were screening level quality to aid in decision-making regarding the progression of fieldwork. Results of all sampling are compiled in Appendix E.

2.13.1   Soil/Rock Sampling

A total of six samples (one surface sample and five subsurface samples) were obtained from each of the five cluster locations for VOCs and inorganic analyses listed on Table 2.1. Sample depths were based on field measurement of volatiles (using a PID), lithologic contacts, zones of fracturing, secondary porosity features, or intervals of saturation. No visible indications of contamination or staining were observed.

Rock samples were collected for chemical analysis using an air rotary drill rig and a core barrel. Coring was performed very slowly to minimize heating of the rock and core barrel. After the core was brought to the surface and removed from the core barrel, rock selected for chemical analysis was broken off from the core with a hammer and placed in the appropriate sample jar.

Table 2.1 - Sampling Parameters for Soil/Rock Samples

2.13.2   Groundwater Sampling

At clusters CS-MW7, CS-MW8, and CS-MW10, discrete interval groundwater sampling was conducted in either a 4-inch corehole or 6-inch borehole using a straddle-packer system. Sampling zones were selected based on observations recorded during the lithologic description, and was aided by review of the geophysical logs. In a 4-inch corehole, a double packer system was used to straddle either a submersible pump or screened inlet section, based on the yield of the formation at that depth. The packer system�s open interval did not exceed 12 feet in length. When the formation freely yielded groundwater as determined by the site geologist, a pump was used to purge and collect discrete groundwater samples. When the formation did not freely yield groundwater, the test zone was purged via air-lifting, followed by sample collection with a �-inch diameter bailer. In wells that yielded sufficient groundwater, the minimum purge volume was 1.5 pore volumes; which is the volume of water to be purged between the upper and lower packer utilized to obtain the discrete interval groundwater sample. However, in most cases, three or more pore volumes were removed before sampling. To reduce turbidity or improve other quality parameters prior to sample collection, the site geologist required greater volumes to be removed.

For the collection of discrete groundwater samples at locations other than the initial corehole drilled at each cluster, the boring was specifically reamed to a 6-inch diameter to accommodate a similarly-sized packer system. The 6-inch packer system consisted of a single packer set above a submersible electric pump with no check valve. The total depth of the borehole served as the lower isolation point for the discrete sample; therefore, the sampling occurred as drilling progressed in depth. Purging and sampling methods and volumes were the same as with the 4-inch packer system, and were entirely dependent on the amount of water yielded by the selected interval.

2.13.3   Off-post Impacts Sampling

Prior to beginning drilling operations at CS-MW10, located approximately 20 feet from the CSSA boundary and about 215 feet from the nearest off-post private well, a monitoring program was initiated at neighboring off-post wells to identify potential effects to their water quality, if any, resulting from the drilling at CSSA. The drilling process involved the high pressure injection of air, water, and occasionally a foaming agent to circulate and remove drill cuttings as the borehole was advanced in depth. While most of the injected media was returned to the surface, a significant portion was lost to the limestone formation. High pressure injection of air and water can have the effect of mobilizing sediments, drilling mud or additives (e.g., foam), and/or contaminants away from the hole being drilled. Coupled with the cyclic removal of groundwater from off-post locations, it was anticipated that those consumers might notice a change in water quality which may include taste, odor, or appearance (turbid, stained, or foamy).

To address this situation, CSSA implemented a proactive program to ensure that the public water supplies were protected. Prior to beginning CS-MW10 drilling, baseline samples from off-post wells were monitored for general water quality, with follow-up sampling for monitoring and comparison during installation of the CS-MW10 well pair. The baseline and follow-up sampling included laboratory analysis for VOCs, total dissolved solids (TDS), total suspended solids (TSS), pH, conductivity, and turbidity. The pH, conductivity, and turbidity measurements were obtained using field monitoring equipment. This sampling consisted of a baseline analysis, three bi-weekly monitoring events, followed by a final round of sampling to check against the baseline event. Based on the monitoring results, drilling at the MW-10 location had minimal impacts on water quality at the nearby drinking water wells. A summary of results is presented in Section 4.3.2.

2.14 - Fracture Analysis

In support of ongoing environmental investigations at CSSA, Parsons collected and submitted core samples to Core Laboratories Reservoir Geology Group (Core Labs) of Houston, Texas for fracture analysis. A total of 80 feet of carbonate rock core from four coreholes was studied to determine type, density, and relative orientation of natural fractures and/or faults within the Middle Trinity Group.

All cores were retrieved by GPI using air rotary methods, and prepared and packaged by Parsons. The core was not oriented with any reference to an azimuthal direction, and the maximum segment length was 10 feet. From each corehole, up to a total of 20 feet of core originated either from the LGR or CC formational members which make up the water-bearing portions of the Middle Trinity Aquifer unit. Core intervals considered representative of the structural features present at the site were submitted for detailed analysis.

Typical petroleum exploration drilling uses tools which allow collection of oriented core; however, this is not common practice in environmental and water well drilling because it is cost prohibitive and not typically required. Because the CSSA cores were not oriented downhole, true fracture strike and azimuth could not be determined. However, the measurement of fracture dip angle is unaffected by the lack of an oriented core. To measure the dip and azimuth of fractures in the cores, a line of known and non-varying position has to be marked on the core surface. Normally, this is the high side or 0� azimuth line based on the orientation survey data. In the present study an arbitrary master orientation line (MOL) was drawn on the surface of the core. This vertical line served as the reference point from which to measure the orientation of the fractures. Naturally, a MOL can only be drawn along a continuous core interval. Rubble zones or other breaks in the core mean that the MOL will be drawn in a different position from one core interval to the next.

An electromagnetic goniometer (EMG-200) was utilized to digitize and record measurements of fracture and slickenside orientations. Further description with regard to the fracture condition, origin, fluid and mineral fill, structure, and aperture was also documented. While the submitted cores were not oriented with respect to a azimuthal direction, the resulting fracture log reports all structural features relative to each other for each core segment. Since there was no way to ensure consistent orientation of the core between successive coring runs, the analysis was considered independent for each 10-foot continuous interval.

The MOL was used as a reference from which to measure the orientation of fractures with the electromagnetic goniometer. For ease of measurement, each core interval was divided arbitrarily into a number of segments, each segment being placed in turn onto the nylon rollers or core holder of the goniometer. The depths of the top and bottom of each core segment were recorded, together with the diameter and position of the MOL. The orientation of all fractures were then measured relative to the axis of the core and this line. This was done by moving a stylus along each fracture and digitizing its attitude relative to a magnetic field generated around the core segment. The stylus is connected to a three-space tracker capable of following the tip of the stylus in three-dimensional space. A best-fit plane is then computed for each feature and the results are automatically recorded and sorted by depth. All measurements can be reproduced with an accuracy of 1-2�, while the rollers allow the segments to be rotated during digitization so that all parts of the core can be examined.

Once each fracture was digitized, the computer prompted a series of qualitative descriptors for the measured fracture. These included attributes such as natural or induced, open or closed, total and effective width, mineralized or non-mineralized, type of cementing mineral, and lithology. A detailed fracture report submitted from Core Labs is presented in Appendix F.

2.15 - Pump Installation

2.15.1   Low-Flow Pumps

A QED Well Wizard bladder pump was installed in each of the new monitoring wells. These stainless steel pumps are equipped with dura-flex Teflon^� bladders, Teflon-lined polyethylene tubing (3/8-inch outside diameter (OD) sample tube with 1/4-inch OD air line), Teflon-covered stainless-steel support cable, and a 4-inch well cap. For wells deeper than 300 feet, the pumps were equipped with a drop tube. The drop tube allowed the pump inlet to be placed in the middle of the well screen while keeping the pump higher in the well, where it functions better. Pump depths are listed in Table 3.1.

All stainless steel components of the pump, including the body, discharge nipple, center discharge rod, fittings, center rod and body cross pins, and inlet screen assembly, were electropolished by the manufacturer to remove all traces of embedded scale, rust, foreign debris, oils, and grinding compounds from manufacturing; and to protect the surface of the metal against corrosion, tarnish, or oxidation that could affect sample chemistry. The pump fittings are type 316 stainless steel, with compression-type design.

The bladder pump allows water to flow through an inlet check valve into the interior of the pump bladder due to the pressure gradient exerted by the hydrostatic head of the water in which it is submerged. After the interior of the bladder is filled with water, compressed nitrogen gas is applied to the exterior of the bladder to force the water to flow through an outlet check valve and out of the pump. The compressed gas is delivered to the pump through a gas supply tube connected to a compressed gas source. A control device located at the wellhead regulates gas and fluid flow. The pump liquid discharge is delivered to the wellhead through a water discharge tube in the wellhead cap. The water is pumped and conveyed in a manner to minimize alteration of water quality in any way. When the pump bladder is squeezed sufficiently to empty it of water, the compressed gas control device stops the flow and vents the excess gas to the atmosphere. This venting allows pressure on the outside of the pump bladder to decrease to less than that of the hydrostatic head present at the pump inlet due to the pump's submergence. The pump bladder can then refill and repeat the cycle as needed to achieve desired flow for purging and sampling the well.

2.16 - Transducer Installation

2.16.1   Transducer Installation

A total of four transducers were purchased by CSSA from In-Situ, Inc. for permanent installation in selected wells. The systems consist of Troll 4000 sensors equipped to measure downhole pressure which can be converted to water level and temperature. Each unit is rated for 100 psi submergence suspended from a polyethylene-coated data cable. Each transducer is automatically compensated for barometric effects through a vented sleeve within the cable jacket. The Troll 4000 is fully programmable, and includes datalogging capability on either linear, logarithmic, or event-based modes. The built-in datalogger on each device is capable of storing over 500,000 data points between data downloads. Data is stored within the on-board datalogger in the form of a "test," and is easily retrieved from the sensor using a remote computer node at the wellhead.

Transducers were installed in each unit of the Middle Trinity Aquifer at the MW9 cluster to record the effects of recharge/discharge within and between the various members of the aquifer. The fourth transducer was installed at MW4-LGR in hopes of gaining insight regarding the interaction between groundwater and surface water recharge along Salado Creek. Each transducer was set to a depth within the water column that was anticipated to fall within the seasonal extremes while still operating within the sensor 100 psi (231 feet of water) rating. During installation, each sensor was referenced to the static water in the respective well as measured from the top of the well casing. To conserve battery power and minimize data points, the transducers were programmed to monitor the head and temperature every 15 minutes, but only record data points that showed a net change of 0.1 feet from the prior reading, or every fourth reading (once per hour), whichever occurred first. The reference level (calibration) was checked and the data downloaded from each transducer on a routine basis, which coincided with the quarterly groundwater sampling event.

[Next Section]