Sample Collection and Analysis

Groundwater-level measurements, sample collection and processing, and field analyses were in accordance with applicable USGS procedures (U.S. Geological Survey, variously dated), except that samples were collected using a peristaltic pump. Geochemical measurements and concentrations determined for all samples included dissolved hydrogen (H₂), dissolved oxygen (DO), filtered organic carbon (referred to as dissolved organic carbon, DOC), filtered nitrate plus nitrite, filtered manganese, filtered ferrous iron (or iron (II)), filtered sulfate, unfiltered sulfide, dissolved methane, dissolved carbon dioxide, pH, specific conductance, oxidation-reduction potential (ORP), and filtered chloride. Concentrations of 64 VOCs were determined for samples from piezometers in 2007 and 2008, and from 3 wells in 2008. Concentrations of the dissolved gases ethane and ethene were determined for all samples.

After measuring depth to groundwater, all well and piezometer samples were collected with a peristaltic pump and single-use polyurethane tubing. A stainless-steel weight was attached to the bottom of the tubing to accurately collect the sample from the mid-screen altitude in each well. Samples were collected after approximately three casing-volumes of water were purged from the wells and after allowing the analytes pH, specific conductance, and DO to stabilize within 0.1 unit, 5 percent, and 0.3 mg/L, respectively. Those three analytes and ORP were measured in a flow-through chamber using temperature-compensated sensors from a Yellow Springs Instruments (YSI) data sonde. The specific conductance sensor was checked daily with standard reference solutions; the pH sensor was calibrated daily with two pH standards; and the DO sensor was calibrated daily using the water-saturated air method and occasionally verified with zero DO solution. Dissolved-oxygen analyses were confirmed for most samples using 0 to 1 mg/L CHEMets Rhodazine-D colorimetric ampoules (manufactured by CHEMetrics, Inc., Calverton, Virginia). These ampoules were filled directly from the sampling tube after well purging was completed.

Concentrations of iron (II) were measured in the field in samples that had been filtered through a 0.45-μm membrane filter using a colorimetric 1,10 phenanthroline indicator method and a Hach Model 2010 spectrophotometer following Hach Method 8146 (Hach Company, 1998; adapted from American Public Health Association, 1980). Sulfide concentrations were immediately measured in the field using a colorimetric methylene-blue indicator method using the same spectrophotometer according to Hach Method 8131 (Hach Company, 1998; procedure is equivalent to U.S. Environmental Protection Agency (USEPA) method 376.2 (U.S. Environmental Protection Agency, 1983)). Methodologies used to determine iron and sulfide concentrations also are described online at http://www.hach.com/fmmimghach?/CODE%3AIRONFER_AVPP_OTHER_P1873%7C1 and http://www.hach.com/fmmimghach?/CODE%3ASULFIDE_NONE_OTHER_M1965%7C1, respectively, accessed January 6, 2009. Dissolved carbon dioxide (CO₂) concentrations were measured in the field with Titret-Sodium hydroxide titrant with a pH indicator (manufactured by CHEMetrics, Inc., Calverton, Virginia).

Samples for H₂ in groundwater were collected using the bubble-strip method of Chapelle and others (1997) and concentrations were measured in the field using a reduction gas analyzer (Trace Analytical model E-001). Initial gas samples from each well were collected and analyzed after at least 20 minutes of stripping; subsequent samples were collected and analyzed at about 5-minute intervals until consecutive H₂ concentrations stabilized to within 10 percent, a process that often required one hour or more.

Samples for analysis of nitrate plus nitrite, manganese, sulfate, and chloride concentrations were filtered through a 0.45-μm membrane filter into polyethylene bottles, chilled, and sent to the USGS National Water Quality Laboratory (NWQL) in Lakewood, Colorado. Samples for analysis of manganese were acidified in the field with nitric acid to a pH of less than 2 and then analyzed at NWQL by inductively coupled plasma as described by Fishman (1993). Chloride and sulfate were analyzed using ion chromatography as described by Fishman (1993). Nitrate plus nitrite were analyzed colorimetrically by cadmium reduction and diazotization as described by Fishman (1993). The results for the nitrate plus nitrite analyses are referred to simply as “nitrate” in this report because nitrite was not detected historically at the site (Dinicola and others, 2002).

Samples for DOC analysis were filtered through a 0.45-μm filter, collected in amber glass bottles, acidified in the field with sulfuric acid to a pH of less than 2, chilled to less than 4°C, and shipped to NWQL. DOC concentrations were determined using persulfate oxidation and infrared spectrometry as described by Brenton and Arnett (1993).

Samples for VOC analysis were collected in pre-acidified 40-mL glass vials, placed on ice, and shipped to NWQL for subsequent analysis at TestAmerica (TA) Laboratories, Inc. in Denver, Colorado, using purge and trap capillary-column gas chromatography/mass spectrometry according to USEPA Method SW846 8260B (U.S. Environmental Protection Agency, 1996). Samples for analysis of ethane, ethene, and methane were collected in pre-acidified 40-mL glass vials, placed on ice, and shipped to NWQL for subsequent analysis at TA Laboratories, using gas chromatography with a flame-ionization detector according to USEPA Method RSK SOP‑175 (U.S. Environmental Protection Agency, 1994). The reporting limit for a given compound often differed between wells because of different degrees of sample dilution by TA Laboratories. The VOC and dissolved gas samples were collected in pre-acidified vials supplied by TA Laboratories; consequently, the vials could not be overfilled during sampling as is recommended in applicable USGS procedures (U.S. Geological Survey, variously dated) to avoid aeration of the sample.

Quality control of geochemical and contaminant sampling for 2007 included the collection of one duplicate sample for selected redox-sensitive analytes and one duplicate sample for VOCs, and for 2008 quality control included the collection of two duplicate samples for redox-sensitive analytes and one duplicate sample for VOCs. A field blank for redox-sensitive analytes and VOCs was collected in 2007 and a field blank was collected for VOCs and organic carbon in 2008. No substantial quality issues were identified in those samples (appendix A).