Variability of Hydraulic Conductivity in Sandy Till: True Variation Versus Method

The sandy till deposited by the Green Bay Lobe in Wisconsin is included in the Horicon Formation, a lithostratigraphic term that implies that certain properties of the unit can be recognized everywhere it occurs. A compilation of hydrogeological studies of the Horicon Formation showed that hydraulic conductivity ranges over four orders of magnitude in a medium that appears texturally and lithologically homogeneous. The overall objectives of this study were (I) to determine if this apparent heterogeneity is real or a result of different testing methods at different scales, and (2) to examine the effects of the scale of measurement on different methods of determining hydraulic conductivity in these materials.

Two field sites in till of the Horicon Formation were instrumented with piezometers. Each site is in an area of thick, uniform till, away from drumlins and moraines. At Site 1, the till aquifer is unconfined, with a saturated thickness of 8 m. At Site 2, the till aquifer is confined by locally occurring lake silt and clay. The saturated thickness at Site 2 is also about 8 m. The sites were instrumented with 25 and 26 piezometers, respectively. Each piezometer is 5 cm in diameter and has a screen length of 30 cm. The piezometer array is roughly square, with dimensions of about 10 m by 10m.

The results of piezometer tests and pumping tests performed at the sites showed that hydraulic conductivity ranges over nearly two orders of magnitude, from about 4xl 0-5 to about 2xl 0-3 cm/s. The results of borehole dilution tests to determine hydraulic conductivity were inconclusive. In general, larger-scale tests yield larger values of hydraulic conductivity. Repeated tests of individual piezometers gave consistent values of hydraulic conductivity. Textural analyses of samples of the till from the screened intervals showed little variability, and there was no correlation between simple textural characteristics and hydraulic conductivity. Results of the testing indicate that most of the variability is attributable to different types of tests that test different volumes of aquifer. The till aquifer can be considered homogeneous for a single type of test at this scale of study.