Experimental investigations of factors controlling the spread of DNAPL contamination in undisturbed columns of fractured saprolite

A series of laboratory experiments were conducted to further investigate the physical and chemical factors influencing the spread of TCE (trichloroethylene) contamination m fractured saprolite formed from sedimentary rock. The experiments involve simulation of a spill by injection of dyed immiscible phase TCE into two undisturbed columns of fractured saprolite, allowing a period of time for dissolution and diffusion of the TCE, and then dismantling them to map the resulting distribution of contamination. The use of dyed TCE allowed for the identification of preferential flow paths, which were associated with fractures and macropores in the two dominant saprolite lithology types, namely clay-rich limestone saprolite and blocky siltstone/shale saprolite, that comprise the columns. Despite evidence of discrete and erratic flow paths, TCE contamination was found throughout almost all of each column, with only small areas below detection limits. Based on observations and supported by calculations, the variable saprolite lithology and the processes of dissolution and diffusion are responsible for rapidly spreading the contamination. Comparisons were made of fracture aperture and porosity values derived from the cubic law and bulk hydraulic conductivity with DNAPL (dense non-aqueous phase liquid) entry pressure apertures and estimates of residual DNAPL volumes. These comparisons indicate that the cubic law can underestimate fracture aperture and may underestimate residual DNAPL volumes.