Free and residual phases are the most abundant phases of LNAPL (light non aqueous phase liquid) in the subsoil; therefore the free phase volume estimation and the understanding of residual phase behaviour are fundamental for an effective remediation. The volume estimation of free phase present in a contaminated site was carried out through the application of the Pancake Model and the Vertical Equilibrium Model. This estimation shows a remarkable difference between the two models and between the different delimitation area methods employed. In particular, the estimated volumes are lower for the Vertical Equilibrium Model and the higher difference is observed for a 200x200 mesh. The results underline also some critical points as the amount of the effective porosity; indeed the sensitivity analysis shows that a reduction of this parameter produce a variation of the estimated volume until to 20%. The behaviour of residual phase was analysed through lab-scale column tests carried out using three different porous media and toluene as contaminant. The higher residual saturation and lower dissolution in finer materials reveal the influence of porous media on toluene behaviour. The lower dissolution is confirmed also by the modelling realized using a traditional approach and an approach based on experimental results. The risk analysis, carried out applying a traditional approach and a new experimental approach, shows that the last one allow to have a more reliable hazard index which contemplates the site specific conditions.
LNAPL, free phase volume, dissolution column test, risk assessment