3H/3He age data in assessing the susceptibility of wells to contamination
Article first published online: 9 MAY 2005
Volume 43, Issue 3, pages 353–367, May 2005
How to Cite
Manning, A. H., Kip Solomon, D. and Thiros, S. A. (2005), 3H/3He age data in assessing the susceptibility of wells to contamination. Groundwater, 43: 353–367. doi: 10.1111/j.1745-6584.2005.0028.x
- Issue published online: 9 MAY 2005
- Article first published online: 9 MAY 2005
- Received January 30, 2004; accepted July 4, 2004.
Regulatory agencies are becoming increasingly interested in using young–ground water dating techniques, such as the 3H/3He method, in assessing the susceptibility of public supply wells (PSWs) to contamination. However, recent studies emphasize that ground water samples of mixed age may be the norm, particularly from long-screened PSWs, and tracer-based “apparent” ages can differ substantially from actual mean ages for mixed-age samples. We present age and contaminant data from PSWs in Salt Lake Valley, Utah, that demonstrate the utility of 3H and 3He measurements in evaluating well susceptibility, despite potential age mixing. Initial 3H concentrations (measured 3H + measured tritiogenic 3He) are compared to those expected based on the apparent 3H/3He age and the local precipitation 3H record. This comparison is used to determine the amount of modern water (recharged after ∼1950) vs. prebomb water (recharged before ∼1950) samples might contain. Concentrations of common contaminants were also measured using detection limits generally lower than those used for regulatory purposes. A clear correlation exists between the potential magnitude of the modern water fraction and both the occurrence and concentration of contaminants. For samples containing dominantly modern water based on their initial 3H concentrations, potential discrepancies between apparent 3H/3He ages and mean ages are explored using synthetic samples that are random mixtures of different modern waters. Apparent ages can exceed mean ages by up to 13 years for these samples, with an exponential age distribution resulting in the greatest discrepancies.