[Home]
[Table of Contents] [Next Section]
1995 Technical Memorandum on Soil Boring Investigation
Section 2 - Investigative Procedures
2.1 - Drilling and Sampling ProceduresProcedures for performing the surface geophysical surveys are discussed in the June 1995 technical memorandum on geophysical surveys (Parsons ES, 1995b). Field observations of surface indications, such as waste, metal scraps, and topographic expression, for each potential source area were recorded in the field logbook.
Prior to this drilling task of the source characterization, soil borings were located based on geophysical anomalies, when present. When an anomaly was not present at a site, topographic lows or surface features, such as metal debris or man-made mounds, were chosen as locations. When an anomaly was present, locations were selected near the anomaly. This approach was used to collect representative samples, but not expose the field team to undue health and safety risk.
Drilling was performed using hollow stem augers (HSA) where soil ws present and air-coring when rock was encountered. Cores were continuously collected every 5 feet for classification and potential sampling. Monitoring of the breathing zone was performed with an Hnu® and a MINIRAM®. Due to the heat and moisture created from air coring, the Hnu sometimes indicated false positives while scanning the cores. While drilling with HSAs, no difficulties were noted in taking readings with the Hnu, because heat and moisture were not being generated by the samples. Because the Hnu appeared to be reading accurately when not exposed to high moisture, exchanging the instrument did not seem to be an adequate method for resolving the situation.
The soil and rock core were described by a qualified geologist. The core described the Unified Soil Classification System (USCS). Core samples were described according to lithology, color (using a Munsell color chart), fossil content, textural features (e.g., bedding, bioturbation), structural features (e.g., fractures, solution cavities), hardness, and moisture content. Lithologic descriptions were recorded on soil boring logs which can be found in Appendix A. these logs also contain sample recovery intervals, analytical sample intervals, a graphic log, and comments concerning drilling and sampling.
Samples were identified by site, soil boring number, and depth of sampling. This was accomplished by the first part of the ID being the site, such as B1. The second part was the boring number, such as SB2. Next as the depth in feet (ft) or GW for groundwater. The resulting ID is B1-SB2 (29-30 ft) or B1-SB2 (GW). Duplicate samples were indicated by DUP following the sample depth. Equipment blanks were identified with (EB) following the boring location and number. Matrix spike/matrix spike duplicate (MS/MSD) samples were indicated as such and were collected in the same jar as the duplicate sample.
Soil samples were primarily collected at three depths: 1) within the first 5 feet or near the bottom of the trench; 2) near the bottom of the trench or between 10 and 20 feet; 3) at total depth (TD) or approximately 30 feet. The majority of the soil borings were drilled to 30 feet TD. Samples were also taken to characterize any waste encountered; when photoionization detector (Hnu) scans of the core indicated potential volatile organics; or if a discernible odor was present. Samples were collected with decontaminated sampling equipment. Sampling tools included a hammer and chisel for soil samples, and a Teflon® bailer for groundwater samples.
Samples were placed into 8-oz. soil jars to be analyzed for volatile organic compounds (VOCs) (SW5330/SW8240), semivolatile organic compounds (SVOCs) (SW3510/SW8270), and ICP metals (SW3010/SW6010). ICP metals included cadmium, calcium, chromium, copper, iron, lead, magnesium, manganese, nickel, and potassium. Of these metals, calcium, copper, iron, magnesium, manganese, and potassium were considered geochemical parameters, rather than chemcials of concern (i.e. potential contaminants). Samples were packed on ice and sealed in coolers for daily pickup by the laboratory. Chain-of-custody (COC) procedures were maintained.
Boreholes were grouted within 24 hours or less of being drilled, with the exception of those not completed within one day. When fractures and/or solution cavities were encountered, bentonite pellets were used to plug the void so that grout could be emplaced to surface level. The bentonite pellets were installed, hydrated and allowed to expand before additional grout was added.
The soil boring locations have not been surveyed at the present time. Appendix A contains all boring logs for each site.
The boreholes were left open a sufficient amount of time to determine if groundwater would accumulate. Water levels were measured with a decontaminated water level indicator and recorded to the nearest hundredth of a foot. If sufficient volume was present, a groundwater sample was collected and analyzed for the same parameters as the soil. Groundwater samples were collected with a decontaminated Teflon bailer and nylon rope, placed into two VOC vials, one plastic metals bottle, and one 1-liter SVOC bottle, and kept on ice until transferred to the laboratory.
2.2 - Site GeologyIn most areas on site, a thin layer of soil overlies the Cretaceous-age Glen Rose Formation. During this investigation, the soil profile reached a maximum thickness of 5.5 feet at SWMU B-28. The soil consists of a dark grayish brown clay with little to some silt and occasional limestone gravel.
The limestone rock formation exposed at the surface in various places at CSSA is the upper member of the Glen Rose Formation. The thickness of this member is estimated, from outcrop data, to range from less than 10feet to a maximum of about 150 feet. The outcrop is at its thinnest in the Salado Creek channel which traverses the site from north to south. The lower member of the Glen Rose Formation, which is approximately 300 feet thick at CSSA, is not exposed at the installation.
There are two fossil zones which mark the uppermost part of the lower Glen Rose Formation (Whitney, 1952 and Atnipp, 1986). The Corbula zone, which is considered the contact between the upper and lower Glen Rose, is comprised of the small clam Corbula martinae. The second zone is highly fossiliferous with an abundant fauna, including bivalves (pelecypods), echinoids, gastropods, and foraminifera. The distinctive fossil of this unit is the echinoid Salenia texana. The two zones together comprise a unit that ranges from less than 3 feet to 10 feet thick (Atnipp, 1986). The Salenia zone was observed exposed in the Salado Creek bed south of SWMU B-28. Common fossils include several different pelecypod and gastropod steinkerns, Salenia texana, Hemiaster possibly Heteraster (echinoid), worm burrows, and foraminifera. The Corbula bed was not identified during this drilling investigation at the SWMUs, but has been identified in a gravel pit on the southern part of CSSA.
Appendix B contains the data validation report for the samples collected as part of the potential sourcce characterization. Appendix C contains the validated laboratory data and analytical summary tables for all SWMUs. Qualified data are discussed below.
Twenty VOC samples had holding times exceeded by the laboratory. The data were qualified as UJ, meaning analytes were not detected, but the detection limits were estimated values and may or may not represent the actual limit of quantitation necessary to accurately and precisely measure analytes in the sample.
Base surrogate recoveries for SVOCs were outside of control limits on two soil samples and one groundwater sample. For each SVOC sample, the sample was spiked with six surrogate compounds. If two or more surrogates for each fraction were outside quality control criteria, the data was qualified accordingly. Data was qualified as rejected (R), when the analyte was not detected and the surrogate recovery was less than 10% of the lower limit. As a result, several of the SVOC analyte concentrations were rejected. Additional samples were qualified as (J) and are further discussed in the data validation report in appendix B. These samples and those whose holding times were exceeded will be recollected for VOC and/or SVOC analysis.
False positive identifications of cadmium resulted from interference due to high iron content in the soil. Therefore, ninety-eight associated cadmium results were qualified as nondetect.
2.4 - Statistical Analysis of Background Soil/Rock Samples
Ten background Glen Rose Limestone samples were collected for chemical analysis as part of a previous investigation (Parsons ES, 1994). Figure 2.1 shows the locations of the background samples and Table 2.1 summarizes the analytical results. With the exception of nickel, the statistical evaluations were initially performed as part of the CSSA B-20 remedial investigation (Parsons ES, 1995c).
The background results were statistically evaluated to determine the upper tolerance limits (UTL) for the background metals concentrations of the Glen Rose Formation. These values were evaluated with regard to risk-based concentrations to establish comparison criteria for the potential source characterization analytical results (Section 2.5). Appendix D describes the statistical methodology used. Calculated background levels are shown in Table 2.2. Statistical evaluations of the background samples were then performed for cadmium, chromium, lead and nickel, which were considered to be possible contaminants of concern. Statistical evaluations of calcium, copper, iron, magnesium, manganese, and potassium were not performed because these metals are not considered to be contaminants. The analytical results of these six metals are not included with the analytical summary tables for each site, but can be found in appendix C.
The statistical evaluation included determining the percentage of non-detects, the distribution of the data (normal and lognormal), adjustments to the mean and standard deviation (when appropriate), and finally, the 95% UTL. Appendix D also shows the statistical calculations.
Because of the possibility of being contaminants of concern, the following metals are discussed:
Cadmium was not detected in the background samples. Therefore, the parametric tolerance limit test could not be performed on the data. Accordingly, the highest concentration, including half the detection limit, was established as the background concentration. The non-parametric tolerance limit was 0.55 milligrams per kilogram (mg/kg).
Chromium was detected in eight of the ten background samples. Therefore, the parametric tolerance limit test was performed after the data distribution was determined. The Glen Rose had a stronger normal distribution than lognormal.
Since 20% of the Glen Rose chromium data was nondetect, the mean and standard deviation had to be adjusted before determining the upper tolerance limit. Based on the correlation coefficients for censored and detects-only data, the data was adjusted using Cohen's method.
The tolerance limit for chromium was 3.2 mg/kg.
Lead was detected in the ten background samples. The distribution of lead data in the Glen Rose Formation was normal.
The tolerance limit for lead was 67.8 mg/kg.
Nickel was detected in all ten samples. The parametric tolerance limit was performed on the data. The distribution of nickel was lognormal.
The tolerance limit for nickel was 29.4 mg/kg.
2.5 - TNRCC Comparison Criteria
The results of the statistical evaluation are compared to Texas Natural Resource Conservation Commission (TNRCC) Standard 2 risk-based concentrations for chemicals of concern for health-based closures of solid waste management units. The concentrations are specified in the State of Texas risk reduction rules (30 TAC 335 Subchapter S, Appendix I). The primary chemicals of concern for this investigation are those associated with the contamination in well 16: tetrachloroethene (PCE), trichloroethene (TCE), and cis- and trans-1,2-dichloroethene (DCE). Secondary contaminants of concern include cadmium, chromium, lead, and nickel. In accordance with 30 TAC 335.555(d)(1), if the practical quantitation limit (PQL) and/or the background concentration for a contaminant is greater than the cleanup level, the greater of the PQL or background level shall be used as cleanup levels for SWMUs at that facility. Table 2.3 presents the statistical background concentrations and TNRCC medium-specific concentrations (MSCs) for soil.air and ingestion at an industrial site (soils less than 2 feet deep) and groundwater protection standards for industrial sites (soils greater than 2 feet deep). These sites are not considered to be residential because of the remote locations on a federal facility and the sites are not currently being used for human habitation or for other purposes with a similar potential for human exposure (30 TAC 335.552).
Lead levels were compared to the statistically determined background concentrations in accordance with the section cited above (30 TAC 335.555(d)(1)). Cadmium concentrations were compared to the PQL, since the PQL was greater than the groundwater protection standard. For copper and nickel, the groundwater protection standard was used for comparison.
These concentrations are only used for comparison and are not intended as recommendations for closure actions.
Groundwater analytical results, in accordance with 30 TAC 335.559(d)(2), were compared to the Safe Drinking Water Act (SDWA) maximum contaminant levels (MCLS), if promulgated. Otherwise, the analytical results are compared to water MSCs for ingestion, as determined pursuant 30 TAC 335.556. The water MSCs are listed in Table 2.2.
The MSCs for lead in groundwater is 0.015 mg/L. The detection limit for SW6010 was 0.03 mg/L. Because of the difference between the action level and the detection limit, for groundwater samples where lead was not detected it cannot be stated with confidence that the levels of lead do not exceed the action level. For the samples were lead was detected the action level was exceeded.
These concentrations are only used for comparison and are not intended as recommendations for closure actions.