Simulations of Ground-Water Flow, Transport, Age, and Particle Tracking near York, Nebraska, for a Study of Transport of Anthropogenic and Natural Contaminants (TANC) to Public-Supply Wells

Abstract

Contamination of public-supply wells has resulted in public-health threats and negative economic effects for communities that must treat contaminated water or find alternative water supplies. To investigate factors controlling vulnerability of public-supply wells to anthropogenic and natural contaminants using consistent and systematic data collected in a variety of principal aquifer settings in the United States, a study of Transport of Anthropogenic and Natural Contaminants to public-supply wells was begun in 2001 as part of the U.S. Geological Survey National Water-Quality Assessment Program. The area simulated by the ground-water flow model described in this report was selected for a study of processes influencing contaminant distribution and transport along the direction of ground-water flow towards a public-supply well in southeastern York, Nebraska. Ground-water flow is simulated for a 60-year period from September 1, 1944, to August 31, 2004. Steady-state conditions are simulated prior to September 1, 1944, and represent conditions prior to use of ground water for irrigation. Irrigation, municipal, and industrial wells were simulated using the Multi-Node Well package of the modular three-dimensional ground-water flow model code, MODFLOW-2000, which allows simulation of flow and solutes through wells that are simulated in multiple nodes or layers. Ground-water flow, age, and transport of selected tracers were simulated using the Ground-Water Transport process of MODFLOW-2000. Simulated ground-water age was compared to interpreted ground-water age in six monitoring wells in the unconfined aquifer. The tracer chlorofluorocarbon-11 was simulated directly using Ground-Water Transport for comparison with concentrations measured in six monitoring wells and one public supply well screened in the upper confined aquifer. Three alternative model simulations indicate that simulation results are highly sensitive to the distribution of multilayer well bores where leakage can occur and that the calibrated model resulted in smaller differences than the alternative models between simulated and interpreted ages and measured tracer concentrations in most, but not all, wells. Results of the first alternative model indicate that the distribution of young water in the upper confined aquifer is substantially different when well-bore leakage at known abandoned wells and test holes is removed from the model. In the second alternative model, simulated age near the bottom of the unconfined aquifer was younger than interpreted ages and simulated chlorofluorocarbon-11 concentrations in the upper confined aquifer were zero in five out of six wells because the conventional Well Package fails to account for flow between model layers though well bores. The third alternative model produced differences between simulated and interpreted ground-water ages and measured chlorofluorocarbon-11 concentrations that were comparable to the calibrated model. However, simulated hydraulic heads deviated from measured hydraulic heads by a greater amount than for the calibrated model. Even so, because the third alternative model simulates steady-state flow, additional analysis was possible using steady-state particle tracking to assess the contributing recharge area to a public supply well selected for analysis of factors contributing to well vulnerability. Results from particle-tracking software (MODPATH) using the third alternative model indicates that the contributing recharge area of the study public-supply well is a composite of elongated, seemingly isolated areas associated with wells that are screened in multiple aquifers. The simulated age distribution of particles at the study public-supply well indicates that all water younger than 58 years travels through well bores of wells screened in multiple aquifers. The age distribution from the steady-state model using MODPATH estimates the youngest 7 percent of the water to have a flow-weighted mean age