Search

Stormwater Infiltration in Highway Embankments - Saturated Hydraulic Conductivity Estimation for Uncompacted and Compacted Soils

T. M. Allen

Abstract

The estimation of hydraulic conductivity (K(sub sat)) is a key step to assess the rate of infiltration, whether that estimate is for an infiltration pond or trench, if it is for a highway embankment, or if it is for natural dispersion in general. The focus of this research is to assess available methods for estimating K(sub sat), especially with regard to the ability of various methods to assess K(sub sat) in both a loose, uncompacted state as well as in a compacted state for embankments. K(sub sat) prediction for natural soils is also considered. To accomplish this, a series of relatively large diameter (i.e., 6 to 9 inch) saturated hydraulic conductivity tests were conducted both in a loose state and in a compacted state. Existing K(sub sat) prediction equations such as those developed by Hazen (1892), Slichter (1898), Terzaghi (1925), Chapuis (2004), and Massmann (2003) were evaluated and, using the K(sub sat) laboratory measurements gathered in this study, were empirically optimized to improve prediction performance. Those equations that included soil porosity, eta, or void ratio, e, were given preference for further development, since eta or e were determined to be the best parameters to address the effects of compaction on K(sub sat). The empirically optimized Slichter, Terzaghi, and Chapuis equations were found to provide the most accurate prediction performance. Since it may be difficult to obtain a measured porosity or void ratio at design time, a method to estimate the soil porosity using grain size parameters plus degree of soil compaction, or for natural soils, degree of over-consolidation, was developed, and could be used in the optimized equations with only minimal reduction in K(sub sat) prediction accuracy. The optimized Slichter Equation was used for several example infiltration facilities as was done by Massmann (2003) to determine what effect the use of this new equation would have on infiltration design and infiltration rate prediction accuracy.