No abstract

Depth‐resolved chemistry samples are critical to a wide range of groundwater investigations. If a well intersects zones of variable concentrations, a pumped sample is a composite of the inflows, which mix in the well. Where discrete concentrations are required, excessive mixing makes samples less useful and potentially misleading. However, installations for depth‐discrete sampling are expensive, particularly for regional studies, so sometimes there is incentive to use existing infrastructure designed for other purposes (e.g., supply wells). This paper shows how the resolution of groundwater chemistry derived from long‐screened and open borehole wells can be improved by measuring and sampling the in‐well vertical flow regimes in ambient (unpumped) and/or pumped conditions. The ambient flow regime, driven by a natural vertical head gradient, is shown to be particularly useful to sample groundwater native to defined inflow zones (head in the zone > head in the well) and avoid zones impacted by the invasion of intraborehole flow (head in the zone < head in the well). Depth‐specific samples are interpreted either as native groundwater from a discrete source, subject only to analytical error, or a mixture from multiple sources that can be deconvolved, incorporating error in both flow and concentration measurements. Depth‐resolved age tracers (chlorofluorocarbons, 14C, and He) in groundwater from three supply wells are verified with samples from a multidepth nest of piezometers. Results show old groundwater at all depths and the simultaneous occurrence of young water at shallower depths in undisturbed dual‐porosity fractured aquifers in the Pilbara region of Western Australia.

Hydraulic head differences across the screened or open interval of a well significantly influence the sampled water mixture. Sample bias can occur due to an insufficient pumping rate and/or due to native groundwater displacement by intraborehole flow (IBF). Proper understanding of the sampled water mixture is crucial for accurate interpretation of environmental tracers and groundwater chemistry data, and hence groundwater characterization. This paper uses numerical modeling to quantify sample bias caused by IBF in an un-pumped high-yield well, and the influence of pumping rate and heterogeneity on the volume of pumpage required to purge an IBF plume. The results show that (1) the pumping rate must be at least an order of magnitude greater than the IBF rate to achieve permeability-weighted yield, (2) purge volume was 2.2 to 20.6 times larger than the IBF plume volume, with the ratio depending on plume location relative to hydraulic conductivity and head distributions, and (3) after an example 1000-day un-pumped period, purging required removal of at least three orders of magnitude more water than the common practice of three to five well volumes. These results highlight the importance of knowing the borehole flow regime to identify IBF inflow and outflow zones, estimate IBF rates, and to develop a strategic sampling approach.