|Title||Numerical approach to modelling pulse-mode soil flushing on a Pb-contaminated soil|
|Publication Type||Articolo su Rivista peer-reviewed|
|Year of Publication||2013|
|Authors||Luciano, Antonella, Viotti P., Torretta V., and Mancini G.|
|Journal||Journal of Soils and Sediments|
|Keywords||Artificial Neural Network, Calibration, chelating agent, Complexation, Dissolution, EDTA, feasibility study, flushing, Lead, Model validation, numerical model, parameterization, permeability, Soil pollution, soil remediation, tracer|
Purpose: Soil flushing can represent a suitable technology in remediation of soils, sediments and sludge contaminated by persistent species (e. g. toxic metal). This paper presents a model specifically developed to evaluate the feasibility of chelating agent-enhanced flushing. The model, here applied to the remediation of real Pb-contaminated soils, was conceived also to simulate an innovative pulse-mode soil flushing technique. Materials and methods: The soil flushing application was firstly carried out through columns laboratory experiments. Columns were filled with a real Pb-contaminated soil (3,000 mg kg-1 of dry soil) and flushing was operated in a pulse mode with different chelating agent dosages (3 and 4. 3 mmol kg-1soil). Experimental results were used to calibrate and validate the developed reactive transport model that accounts for transport of ethylenediamine tetraacetic acid (EDTA) and EDTA-Pb chelate complexes, Pb residual concentration on soil and the reduction in permeability by soil dissolution. Determination of hydrodynamic and hydro-dispersive parameters was carried out through a numerical approach incorporating the use of neural network as interpolating function of breakthrough data obtained by a tracer test. Results and discussion: The EDTA dosage strongly influenced the efficiency in Pb extraction and soil permeability. Cumulative extractions of Pb were found to be 20 and 29 % for the EDTA concentrations of 3 and 4. 3 mmol/kg of dry soil, respectively. The soil dissolution caused a significant flow rate decrease, as a consequence of the increase in chelating agent concentration. Therefore the recovery phase duration increased from 738 to 2,080 h. The ability of the model in simulating all the examined phenomena is confirmed by a good fit with experimental results in terms of (a) soil permeability reduction, (b) eluted Pb and (c) residual Pb in the soil. Conclusions: Results highlighted as the model, supported by a preliminary and careful characterization of the soil, can be useful to assess the feasibility of the flushing treatment (avoiding soil clogging) and to address the choice of the operating parameters (flow rate, chelating agent dosage and application method). On the basis of the present research results, a protocol is suggested for in situ soil pulse-flushing application. © 2012 Springer-Verlag.
cited By 9