Geochemistry of the Madison and Minnelusa aquifers in the Black Hills area, South Dakota

The Madison and Minnelusa aquifers are two of the most important aquifers in the Black Hills area because of utilization for water supplies and important influences on surface-water resources resulting from large springs and streamflow- loss zones. Examination of geochemical information provides a better understanding of the complex flow systems within these aquifers and interactions between the aquifers.

Major-ion chemistry in both aquifers is dominated by calcium and bicarbonate near outcrop areas, with basinward evolution towards various other water types. The most notable differences in major-ion chemistry between the Madison and Minnelusa aquifers are in concentrations of sulfate within the Minnelusa aquifer. Sulfate concentrations increase dramatically near a transition zone where dissolution of anhydrite is actively occurring.

Water chemistry for the Madison and Minnelusa aquifers is controlled by reactions among calcite, dolomite, and anhydrite. Saturation indices for gypsum, calcite, and dolomite for most samples in both the Madison and Minnelusa aquifers are indicative of the occurrence of dedolomitization. Because water in the Madison aquifer remains undersaturated with respect to gypsum, even at the highest sulfate concentrations, upward leakage into the overlying Minnelusa aquifer has potential to drive increased dissolution of anhydrite in the Minnelusa Formation.

Isotopic information is used to evaluate ground-water flowpaths, ages, and mixing conditions for the Madison and Minnelusa aquifers. Distinctive patterns exist in the distribution of stable isotopes of oxygen and hydrogen in precipitation for the Black Hills area, with isotopically lighter precipitation generally occurring at higher elevations and latitudes. Distributions of 18O in ground water are consistent with spatial patterns in recharge areas, with isotopically lighter 18O values in the Madison aquifer resulting from generally higher elevation recharge sources, relative to the Minnelusa aquifer.

Three conceptual models, which are simplifications of lumped-parameter models, are considered for evaluation of mixing conditions and general ground-water ages. For a simple slug-flow model, which assumes no mixing, measured tritium concentrations in ground water can be related through a first-order decay equation to estimated concentrations at the time of recharge. Two simplified mixing models that assume equal proportions of annual recharge over a range of years also are considered. An ?immediate-arrival? model is used to conceptually represent conditions in outcrop areas and a ?time-delay? model is used for locations removed from outcrops, where delay times for earliest arrival of ground water generally would be expected. Because of limitations associated with estimating tritium input and gross simplifying assumptions of equal annual recharge and thorough mixing conditions, the conceptual models are used only for general evaluation of mixing conditions and approximation of age ranges.

Headwater springs, which are located in or near outcrop areas, have the highest tritium concentrations, which is consistent with the immediate-arrival mixing model. Tritium concentrations for many wells are very low, or nondetectable, indicating general applicability of the timedelay conceptual model for locations beyond outcrop areas, where artesian conditions generally occur. Concentrations for artesian springs generally are higher than for wells, which indicates generally shorter delay times resulting from preferential flowpaths that typically are associated with artesian springs.

In the Rapid City area, a distinct division of isotopic values for the Madison aquifer corresponds with distinguishing 18O signatures for nearby streams, where large streamflow recharge occurs. Previous dye testing in this area documented rapid ground-water flow (timeframe of weeks) from a streamflow loss zone to sites located several miles away. These results are used to ill