Identifying soils with potential of expanding sulfate mineral formation using electromagnetic induction

Sulfate-bearing soils are a problem in highway construction as they combine with materials used for lime stabilization to form minerals, particularly ettringite, that expand and induce heave in the stabilized soil. This research involves quantifying sulfate in soils that may be potentially used in highway construction using electromagnetic induction. The objectives are to: 1) document electrical conductivity (EC) variability within selected sites that contain sulfate-bearing materials, and 2) determine if electromagnetic induction has potential for locating hazardous levels of sulfate-bearing materials. The 0.43 ha study area is located in the Blackland Prairies and is a Vertisol known to contain gypsum at the time of site selection. Apparent EC using a model EM38 electromagnetic induction instrument was measured at 200 locations in July and November 2003, using a sampling grid with 5-m spacings. Representative rows and columns were selected from the map of apparent electrical conductivity, and soil cores taken to a depth of 1.5 m at 29 points. Soil samples were obtained by dividing cores into depth increments of 0 to 25 cm, 25 to 75 cm, and 75 to 150 cm. Laboratory analyses were run for each sample and included moisture content, EC and soluble cations and anions of the saturated paste extract, and percent gypsum. Elevation measurements were made to determine if changes in elevation related to EC measurements.
Apparent EC proved to be more successful at detecting soluble salts during the dry sampling period (July) when the effect of soil moisture content was less. For July data, EC and gypsum were significantly correlated in the deepest samples (r2 = 0.51 and 0.15, respectively) to apparent EC. Further, soluble sulfate was significantly correlated to apparent EC (r2 = 0.30) at a depth of 25 to 75 cm. Results suggest that the EM38 can be used successfully to map variability of soil salinity across a field, but although correlation exists between apparent EC and sulfate-bearing materials, it is not sufficiently strong to serve as a good predictor for conditions surrounding lime-induced heave in soil.