The behaviour of elemental mercury in porous media: A modelling approach on the multi-phase-flow of elemental mercury

Abstract

Elemental mercury is known for its high density, viscosity and surface tension, making it suitable for a variety of industrial processes, particularly in the past. Nowadays, elemental mercury is a contaminant of concern and occurs in the subsurface at numerous locations. To date however, its behaviour in porous media is not yet fully understood. Therefore, this research focusses on elemental mercury in porous media to establish an understanding of processes characterizing its multi-phase-flow. Such an understanding would enable field simulations of elemental mercury contamination and would enable development of remediation methods. Elemental mercury is a dense-non-aqueous-phase-liquid (DNAPL), given its immiscibility and high density and therefore it is comparable to the better known DNAPL PCE (PerChlorinatedEthylene). In this study, the multi-phase-flow of elemental mercury was modelled for several scenarios by using the numerical simulator STOMP. Modelling results were compared to that of PCE, whilst the use of field observations and literature enabled a close link to reality. In homogeneous saturated porous media, the exceptional high density of elemental mercury governs its flow behaviour, whereas viscosity has a limited influence. The low residual saturation allows elemental mercury to infiltrate substantially deeper than PCE, with the same volume of DNAPL spill. In saturated heterogeneous porous media both elemental mercury and PCE flow via preferential pathways, where the ability of infiltrating less-permeable layers is similar for both DNAPLs. However, elemental mercury migration extends further and is faster than that of PCE, due to the low residual saturation and high density. In the unsaturated zone the behaviour of elemental mercury also deviates from that of PCE, since PCE is wetting to air whereas mercury remains non-wetting. Moreover, when assuming PCE and elemental mercury to be spreading and intermediate wetting, mercury does not infiltrate fine sand whereas PCE does. Finally, this research suggests to validate multi-phase-flow models for the saturated zone and to determine fundamental principles governing three-phase-flow in the unsaturated zone. This could be achieved by experimentally investigating the behaviour of elemental mercury in porous media.