Chlorinated solvent dense non-aqueous phase liquids (DNAPLs) can be consistent long-term sources of groundwater contamination. Remediation is costly and poses significant technical challenges. Nanoscale zero-valent iron (NZVI) is a proposed in situ remediation agent for DNAPL source zones, but the effectiveness of NZVI treatment relies on the ability to emplace and retain the NZVI near or within the DNAPL source area. The dominant physical and chemical processes controlling the reactivity and migration and distribution of NZVI in the subsurface are poorly understood, however, making it difficult to ensure that NZVI will be delivered where it is needed and determine whether or not it can be an effective treatment alternative for DNAPL source zones.
The overall objective of this project was to evaluate if and where NZVI may provide a rapid cost-effective method to diminish the mass and “strength” of DNAPL sources. Specific research objectives were to (1) obtain a fundamental understanding of the physical and geochemical processes governing the migration and distribution of polymer modified NZVI in DNAPL contaminated zones of a naturally heterogeneous subsurface where the free phase is entrapped in a complex architecture and (2) understand in situ treatment efficiency of DNAPL and dissolved chlorinated organics by polymer modified NZVI.
Controlled investigations in small 1-dimensional and 2-dimensional laboratory test systems, large intermediate-scale 2-dimensional test tanks, and a field-scale 3-dimensional tank, as well as mathematical models were used to determine the primary physical and chemical principles controlling colloid transport (e.g., particle-particle interactions, fluid velocity, grain/pore size). Particle and environmental factors affecting NZVI reactivity and the ability to provide DNAPL targeting were also evaluated experimentally. The effect of porous media heterogeneity on nanoiron transport and the ability of emplaced NZVI to decrease the source mass and mass emission were evaluated in 2-dimensional small- and intermediate-scale tanks. Several combinations of nanoiron types and polymeric surface modifiers were used to identify optimal nanoiron surface properties. Finally, transport of the two types of NZVI was evaluated in a field-scale 3-dimensional tank study to determine transport distances in realistic heterogeneous porous media.
The mobility of NZVI in porous media depends on the surface modifier properties, seepage velocity, ionic strength and composition, pH, heterogeneity in the hydraulic conductivity field, and the presence of silica fines and clay fines. A small amount (e.g., 2 wt%) of fine particles and clay particles added to the sand matrix was found to limit NZVI transport. The presence of excess free polymer in the injection solution partially alleviates this problem.
Transport was limited by aggregation of NZVI and by deposition and pore plugging near the injection well. Magnetic attraction between NZVI particles caused rapid aggregation, but could be slowed through surface modification with polymers. NZVI suspensions of 6 g/L could move through model porous media; however, the mobility decreases as NZVI polydispersity and concentration increases, primarily due to enhanced aggregation and straining. NZVI transport decreased with decreasing seepage velocity; however, the extent depends on the type of polymer modifier. Mobility decreases with increasing ionic strength. In heterogeneous porous media, NZVI transports preferentially through high conductivity regions where fluid velocities are greatest. NZVI deposits in regions of low flow and stagnation points. Injecting too much NZVI leads to pore plugging and diversion of dissolved contaminants around the reactive zone. A model coupling both the fundamental particle properties and flow fields can successfully model the transport and deposition of polymer modified NZVI at 6 g/L in 2-dimensional heterogeneous porous media. Injection methods that enhance fluid shear and disaggregate surface modified particles may improve mobility in situ.
The mobility significantly decreased as the solution pH was lowered from 8 to 6 due to enhanced aggregation and deposition. The presence of excess free polymer (approximately 0.1 wt %) in the injection solution increases mobility compared to cases where excess free polymer is not present. To maximize mobility, injection schemes should use low concentrations of excess free polymer (approximately 0.1 wt%) in solution. Importantly, the presence of more than a few wt% of silica fines or clay (kaolinite) fines decreased mobility. Clay (kaolinite) had a greater effect than silica at the same wt%. NZVI transport in a field-scale tank was less than what is predicted according to bench-scale column experiments using model porous media. The low mobility was a result of the fines particles in the sand matrix, consistent with laboratory findings. The decreased transport was also a result of near-well plugging for one type of NZVI.
Polymer modified NZVI can be targeted to the NAPL-water interface in situ. Targeting is a result of the affinity of the polymer coated NZVI for the NAPL-water interface. The targetability increases as NAPL saturation decreases because of greater NAPL-water interface at low NAPL saturation. However, NZVI placed at the NAPL-water interface is rapidly oxidized and does not enhance NAPL dissolution.
Surface modification required for mobility decreases the reactivity of the Reactive Nanoscale Iron Particles (RNIP) by a factor of 5 to 24 depending on the type of surface modifier used. However, the rate of hydrogen evolution is unchanged by the surface coatings. The reactivity is sufficient, however, for NZVI emplaced just down-gradient of the source to decrease mass flux measured down-gradient by up to 85%. Emplaced NZVI acts primarily as a reactive barrier treating dissolved phase contaminants dissolved from the entrapped NAPL.
These results indicate that emplaced NZVI can decrease the flux of contaminants emanating from entrapped DNAPL; however, they also suggest that NZVI available commercially today still needs to be optimized to work as an effective reactive barrier. This will require mobility of a greater fraction of the injected iron and innovative injection/emplacement methods. The lessons learned in this study can be used to determine the most promising innovative solutions to attempt in field trials (e.g., inclusions of free polymer in the injection solution) and be used by project managers to determine site conditions that are favorable or unfavorable for NZVI application. These results also provide guidance on the properties of NZVI with the greatest potential for success, i.e., those having sufficient reactivity to attenuate the concentrations of dissolved contaminants emanating from the source zone, but also having greater selectivity for the dissolved contaminants such that the NZVI reactive lifetime can be extended.