Preferential flow and transport in variably saturated fractured media

A significant number of subsurface environmental problems involve preferential flow and solute transport in variably saturated fractured media. The presence of large connected void spaces such as fractures and channel-like openings in the vadose zone have been widely documented to provide rapid pathways for spreading of contaminant from the surface to deeper aquifers.

The objective of this Ph.D.-study is to characterise and estimate preferential flow and solute transport in three different variably saturated fractured media: chalk, till, and granite. All three media consist of relatively low permeable matrix with fractures and macropores introducing fast preferential flow and transport paths. In these media, not all the fractures and macropores are registered to be of equal importance for the flow and transport regime. The most important hydraulic pathways and hereby the preferred pathways are thus identified through detailed fracture mapping, an identification of the sources or mechanisms that are capable of producing the stress field requisite for initiating and propagating the fractures, and detailed in situ tracer-experiments.

The standard procedure for monitoring leaching processes is based on indirect methods such as pumping tests, well interference studies and tracer tests, which include extraction of soil solutions using suction cells or extraction of soil samples in situ. The limitations by using these procedures are that they only provide aggregated information on the effects of the fractures and macropores at scales significantly larger than the fractures and macropores because of the data uncertainty on fracture and macropore locations, properties, connectivity, boundary conditions, fracture- and macropore-matrix interactions, and host matrix properties. To circumvent these limitations, detailed tracer tests combined with detailed characterisation of the fractured vadose chalk and till were performed.

For the chalk, infrared thermography (IRT) provided the refined spatial information to clarify the hydraulic activity of the five fracture-systems registered by analysis of the regional setting and fracture mapping along quarry walls and in wells. By making use of the contrast between the constant temperature of the deeper groundwater and the temperature of the exposed wall in the quarry on a cold winter and a hot summer day, zones of groundwater discharge were delineated.

For the till, tracer experiments with two fluorescent tracers Acid Yellow 7 (AY7) and Sulforhodamine B (SB) were performed with three different rain events in a fall and summer season. The movement of both tracers in exposed profiles was delineated simultaneously by high spatial resolution concentration maps obtained with an imaging device. Additionally, a detailed description of the fractured vadose media including fracture and macropore mapping and estimation of structural properties of the different domains (biopores, fractures, coating, oxidized and reduced matrix) was performed. The two-dimensional concentration distribution profiles of the tracers showed that: (a) biopores dominated the tracer migration in the upper 1.2 meter, (b) dead-end biopores were not activated in the fall season, (c) only tectonic fractures connected to hydraulically active biopores contributed to the migration, and (d) the water content in the upper 20 cm of the irrigated till had a pronounced retardation effect on the migration of AY7 but no effect on the migration of SB.

To perform risk assessment analysis on these naturally fractured variably saturated media, a better understanding of the often-controlling effects of the media is vital. A conceptual model based on detailed field-scale information concerning the different domains and a numerical model tool capable of representing the processes taking place in the media are in this connection crucial.

For the till, the three-dimensional numerical model, HydroSphere, describing fully integrated subsurface and solute transport was used to interpret the fluorescent tracer experiments. Special emphasis was given to address the different tracer migration pathways given the initial water content, rain intensities, tracer characteristics, geometry of the structure system, and domain properties.

For the granite, a description of the micromorphology of the fractures and faults and measurement of the fracture aperture (scale~0.01-1 mm) was conducted based on detailed analysis of impregnated fractured granite-samples in a Scanning Electron Microscope (SEM). Others used this information as input to the three-dimensional numerical model, SIMUSCOPP, for the simulation of a transient NAPL migration scenario based on a multi scale geological investigation (multi-scale fracture properties).