Comparison of passive and pumped sampling methods for analysis of groundwater quality, Kirtland Air Force Base, Albuquerque, New Mexico, 2019

A plume of ethylene dibromide (EDB) dissolved in groundwater extends northeast from the Bulk Fuels Facility on Kirtland Air Force Base, New Mexico. The leading edge of the EDB plume is upgradient from several water-supply wells. In 2013, the U.S. Geological Survey (USGS), in cooperation with the Albuquerque Bernalillo County Water Utility Authority and the U.S. Air Force, installed four sentinel well nests and two aquifer-test pumping wells between the EDB plume and the water-supply wells to serve as an early warning if the plume travels toward the water-supply wells. Since 2015, the USGS has used submersible pumps to sample the sentinel wells quarterly. In February 2017, the USGS began using dual-membrane passive diffusion bag samplers for quarterly sampling in the wells. To ensure that the passive samplers are obtaining representative samples of the groundwater contaminants, the USGS, in cooperation with the U.S. Air Force, initiated a study in 2019 to compare results from pump sampling and passive samplers and to use vertical profiling to determine the optimal depth for passive sampler placement in the screened interval to better inform long-term monitoring of the site.

Vertical profiling included deploying passive samplers throughout the submerged screened interval of four shallow sentinel wells. After retrieval of the passive samplers, pump samples were collected. The results of analyses of both types of samples were compared. Volatile organic compound results for this study were all below the raised reporting levels, which is a level five times the maximum concentration detected in a blank and determined by an in-depth quality assessment; therefore, this study focused on inorganic constituent results, including major ions, trace elements, and stable isotopes of water, to calculate the relative percent difference (RPD) between the pump and passive sampling method results as a way to determine where passive samplers would be best placed in each of the wells. Several analytes had an RPD of more than plus or minus 50 percent, and several analytes were not within the estimated variability for each sampling method. Additionally, the variability within each sampling method was quantified and compared. Factors that likely contributed to the lack of comparison between each sampling method included temporal variability, flow regime, volume of sample integrated through different aquifer intervals, and reduction/oxidation processes. RPD and method variability were used to determine the intervals within each well with the greatest agreement between sampling methods. Optimal sampling depths for each well were then correlated to the intervals where quarterly sampling has been occurring.