Assessment of flushing methods for the removal of heavy chlorinated compounds DNAPL in an alluvial aquifer

Immiscible mobilization and foam flushing were assessed as low surfactant consuming technologies, for the enhanced recovery of dense non-aqueous phase liquid (DNAPL) residual at a site contaminated by heavy chlorinated compounds. Preliminary experiments in well-controlled conditions demonstrated the phenomena involved in these remediation technologies and their limitations. Furthermore, we characterized the technologies according to by their surfactant consumption (per kg of DNAPL recovered) and the final DNAPL saturation reached. Surfactant foam flushing (SFF) produced lower DNAPL saturation than immiscible mobilization, thanks to its higher viscosity. However, its efficiency is strongly correlated to the pressure gradient (▽P) used during injection, and that is limited by risks of soil fracturing. The two technologies were tested in field cells (10m×10m×10m) delimited by cement/bentonite walls anchored in the clayey substratum. The deepest soil layer was the most contaminated. It was composed of silt-sandy soil and had an average hydraulic conductivity of 10−4ms−1. Field results show that we should now model flushing fluid propagation to design efficient set-ups for recovering the displaced DNAPL. [Display omitted] •Several technologies were assessed for enhanced DNAPL recovery.•Foam flushing was the most effective due to higher viscosity.•Geological anisotropy must be considered to optimize DNAPL recovery during field work.•Foam flushing is technical and requires modelling of propagation in anisotropic media.