Groundwater quality in the 1,850-square-mile North San Francisco Bay Shallow Aquifer (NSF-SA) study unit was investigated by the U.S. Geological Survey (USGS) from April to August 2012, as part of the California State Water Resources Control Board (SWRCB) Groundwater Ambient Monitoring and Assessment (GAMA) Program’s Priority Basin Project (PBP). The GAMA-PBP was developed in response to the California Groundwater Quality Monitoring Act of 2001 and is being conducted in collaboration with the SWRCB and Lawrence Livermore National Laboratory (LLNL). The NSF-SA study unit was the first study unit to be sampled as part of the second phase of the GAMA-PBP, which focuses on the shallow aquifer system.

The GAMA NSF-SA study was designed to provide a spatially unbiased assessment of untreated-groundwater quality in the shallow aquifer systems and to facilitate statistically consistent comparisons of untreated-groundwater quality throughout California. The shallow aquifer system in the NSF-SA study unit was defined as the part of the aquifer system that is used by many private domestic wells and is shallower than the primary aquifer system used by many public-supply wells.

In the NSF-SA study unit located in Marin, Mendocino, Napa, Solano, and Sonoma Counties, groundwater samples were collected from 71 wells. Seventy of the wells were selected by using a spatially distributed, randomized grid-based method to provide statistical representation of the study unit (grid wells), and one well was selected to aid in evaluation of water-quality issues (understanding well).

The groundwater samples were analyzed for organic constituents (volatile organic compounds [VOCs], pesticides, and pesticide degradates); constituents of special interest (perchlorate and 1,2,3-trichloropropane [1,2,3-TCP]); naturally occurring inorganic constituents (trace elements, nutrients, major and minor ions, silica, and total dissolved solids [TDS]); and radioactive constituents (radon-222 and gross alpha and gross beta radioactivity). Naturally occurring isotopes (stable isotopes of hydrogen, oxygen, boron, strontium, and inorganic carbon in water, tritium activities, and carbon-14 abundances) were measured to help identify the sources and ages of the sampled groundwater. In total, 207 constituents and water-quality indicators were measured.

Three types of quality-control samples (blanks, replicates, and matrix spikes) were collected at up to 13 percent of the wells in the NSF-SA study unit, and the results for these samples were used to evaluate the quality of the data for the groundwater samples. Blanks rarely contained detectable concentrations of any constituent, suggesting that contamination from sample-collection procedures was not a significant source of bias in the data for the groundwater samples. Replicate samples generally were within the limits of acceptable analytical reproducibility. Matrix-spike recoveries were within the acceptable range (70 to 130 percent) for approximately 91 percent of the compounds.

Most of the wells sampled for this study were private domestic wells. Private domestic wells are not regulated in California, and groundwater from these wells is rarely analyzed for water-quality constituents. Although regulatory benchmarks for drinking-water quality do not apply to private domestic wells, to provide some context for the results, concentrations of constituents measured in the untreated groundwater were compared with regulatory and non-regulatory health-based benchmarks established by the U.S. Environmental Protection Agency (USEPA) and California Department of Public Health (CDPH), to non-regulatory health-based benchmarks established by the USGS in cooperation with the USEPA, and to non-regulatory benchmarks established for aesthetic concerns by the CDPH. Comparisons between data collected for this study and benchmarks for drinking water are for illustrative purposes only and are not indicative of compliance or non-compliance with those benchmarks. Most of the organic and inorganic constituents that were detected in groundwater samples from the 70 grid wells in the NSF-SA study unit were detected at concentrations less than drinking-water benchmarks.

Of the 149 organic and special-interest constituents analyzed for in groundwater samples, 31 were detected; concentrations of most detected constituents were less than regulatory and non-regulatory health-based benchmarks. One VOC, benzene, and one insecticide, dieldrin, were detected at concentrations above their respective health-based benchmarks. In total, VOCs were detected in 40 percent of the grid wells sampled, pesticides and pesticide degradates were detected in 13 percent, and perchlorate was detected in 27 percent of the 70 grid wells sampled.

Groundwater samples from 70 grid wells were analyzed for trace elements, major and minor ions, nutrients, and radioactive constituents; most detected concentrations were less than health-based benchmarks. Exceptions are 12 detections of manganese greater than the USGS Health-Based Screening Level (HBSL), 7 detections of arsenic greater than the USEPA maximum contaminant level (MCL-US) of 10 micrograms per liter (µg/L), 2 detections of boron greater than the HBSL of 6,000 µg/L, 2 detections of fluoride greater than the CDPH maximum contaminant level (MCL-CA) of 2 milligrams per liter (mg/L), 2 detections of nitrate greater than the MCL-US of 10 mg/L, and two detections of radon-222 greater than the proposed MCL-US of 4,000 picocuries per liter.

Results for constituents with non-regulatory benchmarks set for aesthetic concerns from the grid wells showed that iron concentrations greater than the CDPH secondary maximum contaminant level (SMCL-CA) of 300 µg/L were detected in 13 grid wells. Chloride was detected at a concentration greater than the SMCL-CA recommended benchmark of 250 mg/L in two grid wells. Sulfate concentrations greater than the SMCL-CA recommended benchmark of 250 mg/L were measured in two grid wells, and the concentration in one of these wells was also greater than the SMCL-CA upper benchmark of 500 mg/L. TDS concentrations greater than the SMCL-CA recommended benchmark of 500 mg/L were measured in 15 grid wells, and concentrations in 4 of these wells were also greater than the SMCL-CA upper benchmark of 1,000 mg/L.