Modeling of the injection and long term fate of nanoparticles in groundwater systems

The understanding of fate and transport of nanoparticles (NPs) in groundwater has become a significant environmental issue in the last decades. NPs can be found in groundwater as natural colloids or anthropogenic NPs released from industrial processes. Moreover, engineered NPs can be injected on purpose into the subsurface for the remediation of contaminated aquifers. The aim of this work is to develop tools and methods to assess, optimize and predict the NP transport at field scale. The modelling tools MNMs and MNM3D were developed to deal with NP transport at the different scales. MNMs was developed for the quantitative analysis of laboratory-scale column tests and preliminary design of pilot injections in 1D radial geometries. MNMs is embedded in a user-friendly graphical interface to be easily applied by final users. MNM3D is a full 3D model for the simulation of NP transport at the field-scale. MNM3D represents the most important innovation of this study, extending for the first time the ionic strength and velocity dependent transport of NPs to 3D geometries and complex scenarios. MNM3D is currently being implemented in the Visual Modflow interface to make it easily available to practitioners and to promote its diffusion in the remediation field. MNM3D was successfully applied for the interpretation of column transport tests of ferrihydrite NPs under transients of ionic strength and for the simulation of the injection of Carbo-Iron® NPs in a 2D large scale experiment. The good results obtained suggest that MNM3D may be a useful tool to simulate the long term fate of NPs in fieldscale conditions and to support the design of a nanoremediation, thus promoting the definitive establishment of this technology. Experimental tests were carried out to study the transport of humic acids-coated goethite NPs under transient geochemical conditions and to optimize their delivery into the subsurface. NP stability and mobility were found to be strongly dependent on the dose of calcium in the suspension. The concentration of the slurry was the key parameter controlling the NP delivery: high concentrations provide additional stability and mobility to the slurry. This result was surprisingly in contrast to the common practice of keeping low the concentration to reduce the NP aggregation phenomena. These findings were also used to develop an injection strategy to control the delivery of goethite NPs in sand columns. A reactive zone can be created in the desired portion of the column by tuning the NP deposition in space and time. If up-scaled, this method can be used to control the delivery of NPs in field applications, improving the effectiveness of the remediation and allowing a more rational use of the slurry, with a consequent reduction of the overall costs of the technology. This protocol is part of a more general patent application about methods for a controlled deposition and immobilization of colloids in porous media. Finally, an integrated experimental-modelling approach was presented to foresee the NP behavior at the field scale by characterizing the NP transport at the laboratory scale. Targeted laboratory experiments are performed to identify the main driving forces influencing the NP mobility and to derive site specific transport parameters by modelling interpretation of the results with MNMs. The results are then up-scaled to foresee the NP transport at larger scales using MNM3D. The approach allows to get a reliable estimation of several operative parameters (e.g. NP distribution, radius of influence, number of wells, etc.) and to easily span a wide range of conditions requiring only a limited number of laboratory and pilot tests. In this work, the combined experimental and modelling approach was applied to the design of an iron oxide-based nanoremediation and for the prediction of the long term fate of graphene oxides NPs. This work was co-founded by the European research projects NanoRem (FP7, G.A. 309517) and Reground (H2020, G.A. 641768).