Tracer tests for quantifying the distribution of transfer times

People involved:Peter Kang (MIT, USA), Tanguy LE BORGNE (UR1), Olivier BOUR (UR1), Nicolas LAVENANT (CNRS), Rebecca HOCHREUTENER (CNRS), Christophe PETTON (CNRS), Vincent BOSCHERO (CNRS), Pascal Goderniaux (UR1), Maria KLEPIKOVA (CNRS)

Flow in fractured media is known to be very heterogeneous, both at the fracture scale where aperture fluctuations can imply significant flow channeling, and at the scale of fracture networks, which are characterized by a large distribution of fracture lengths and transmissivities. The coexistence of fast pathways and stagnation zones, where transport is dominated by diffusion, implies broad residence time distributions.

In May and June 2011, we conducted a series of field tracer tests in the Ploemeur site in order to quantify the distribution of residence times in the aquifer. We observe that the shape of the breakthrough curves (which represent the resident time distribution) depends on the flow configuration and the injection fracture. Specifically the tailing (which represents the longest residence times) disappears under push-pull geometry, and when we injected at a fracture with high flux. This indicates that for this fractured granite, the longest residence times are controlled by heterogeneous advection and not by matrix diffusion.

To explain the change in tailing behavior for different flow configurations, we employed a simple lattice network model with heterogeneous conductivity distribution. The model assigns random conductivities to the fractures and solves the Darcy equation for an incompressible fluid, enforcing mass conservation at fracture intersections. The mass conservation constraint yields a correlated random flow through the fracture system. We investigated whether breakthrough curve tailing can be explained by the spatial distribution of preferential flow paths and stagnation zones, which is controlled by the conductivity variance and correlation length.

By combining the results from the field tests and numerical modeling, we show that the reversibility of spreading is the key mechanism that needs to be captured. We also demonstrate the dominant role of the injection fracture on the tailing behavior: where we inject makes the difference in the tailing.

Figure.Blue line is a tracer breakthrough curve where we injected tracer into slow velocity zone under convergent flow configuration. The late-time tailing observed in blue line diminished for push-pull experiment (red line). Black line is a breakthrough curve where we injected tracer into high velocity zone under convergent flow configuration. We can observe late-time tailing is much less compared to injecting into slow velocity zone. This is an indication of lack of matrix diffusion, and importance of injection fracture. The full tracer tests dataset indicates an origin of anomalous transport related to heterogeneous advection. Insets: illustration of convergent and push-pull tracer tests using a double packer system.


Thermal tracer tests

People involved:Tanguy LE BORGNE (UR1), Olivier BOUR (UR1), P. Goderniaux (UR1), T. READ (East Anglia), V. Bense (East Anglia), Nicolas LAVENANT (CNRS), Rebecca HOCHREUTENER (CNRS), Christophe PETTON (CNRS), Vincent BOSCHERO (CNRS), Maria KLEPIKOVA (CNRS)

Published articles:

Read, T., Bour, O., Bense, V., Le Borgne, T., Goderniaux, P., Klepikova, M. V., Hochreutener, R., Lavenant, N., and Boschero, V. (2013), Characterizing groundwater flow and heat transport in fractured rock using fiber-optic distributed temperature sensing, Geophysical Research Letters, 40(10), 2055-2059.

The coupling of several types of data is a possible way to reduce the uncertainty associated with the interpretation of tracer tests to assess the transport mechanisms. Temperature is a parameter that may have a strong potential for providing new constraints on flow heterogeneity, as discussed in recent reviews (Anderson, 2005, Saar, 2011). Furthermore, recent technical developments, such as distributed temperature sensing, allow measuring temperature with high accuracy and fine spatial resolution.

We conducted a series of thermal and solute tracer tests at the Ploemeur site in collaboration with V. Bense and T. Read (East Anglia University). Thermal tracer tests are performed by injecting continuously 50 degrees Celsius water in a fracture located at 50 meters depth. The breakthrough curves measured in an adjacent borehole show a significant time lag between the thermal and solute breakthrough curves due to the large coefficient of heat diffusion compared to molecular diffusion. Combining heat and solute tracer tests allows measuring tracer dispersion, with Peclet numbers varying over orders of magnitude, thus providing important constraints on the effective transport behavior.

Figure. Illustration of field work. left: Heater for preparing hot water for the thermal tracer test, center: installation of packer for tracer injection, right : injection of fluorescent tracer


Figure. Left: illustration of the experimental setup with injection of hot water in the left borehole and pumping the right borehole. Center : detection of the warm tracer arrival in the pumping borehole with the Fiber optic. Right: temporal monitoring of temperature in the injection and pumping wells.

Laboratory experiments for assessing dispersion, mixing and reaction in saturated and unsaturated porous media

People involved:Joaquin Jimenez Martinez (CNRS), Pietro de Anna (CNRS), Yves Meheust (UR1), Hervé Tabuteau (CNRS), Regis Turuban (UR1), Tanguy Le Borgne (UR1)

Published articles:

De Anna, P., Jimenez-Martinez, J., Tabuteau, H., Turuban, R., Le Borgne, T., Derrien, M., Méheust, Y. (2013), Mixing and reaction kinetics in porous media: an experimental pore scale quantification, Environ. Sci. Technol.

We have developed a new setup to study transport, mixing and reactions in a two-dimensional heterogeneous porous medium at laboratory scale. To fulfill our purpose, we have merged different techniques applied in other scientific fields. The analogous porous medium consists of two parallel glass plates, between which cylinders representing the soil grains are positioned with their axes perpendicular to the glass plates. Two facing sides of the cell are sealed, the two others being used as inlet and outlet for fluids to flow through the synthetic porous medium. The radii and position of the grains along the plates are fully controlled: by use of a lithography technique commonly used in the field of micro fluidics, we reproduce a numerical model defined at will. The principle of the lithography is explained in the caption of Figure 1. A high resolution camera allows for the determination of fluid velocity fields as well as spatially-resolved quantification of solute concentration fields (see Figure 2). Velocity fields are inferred from particle tracking of passive solid tracers. When studying non-reactive transport, the concentration fields are obtained by measuring the light intensity emitted by  a fluorescent tracer. When studying reactive transport, we resort to the use of a chemiluminescent reaction: the measure of the number of emitted photons allows for direct quantification of the concentration for the reaction product. This complete set of measurements allows us to characterize incomplete mixing, concentration probability density functions, and reaction kinetics, at the pore scale.

Figure 1: A: a schematic view of the lithography process. Two glass place are superposed and spaced with solid spacers of thickness a.The mask, which contains the negative image of the geometry desired for the pores, is placed on top of the plates. The space between them is then filled with an UV sensible glue. Where photons coming from the collimated 365nm UV source can pass through the mask, the glue polymerize, giving rise to a solid grain. B: an example of pore geometry printed on the mask

Figure 2: A schematic view of the experimental setup. Clear water and a Fluorescein solution are consecutively injected in the porous medium by suction through a syringe pump. A homogeneous white light irradiates the transported solution within the pores. An optical filter allows only photons with a wavelength l = 521nm to reach the sensor of a CCD camera positioned on top of the experimental set up. The intensity of the light detected by the camera is stored in image files at constant time increment. The camera is fully controlled by a computer.

Figure 3: Example of a concentration field of passive scalar as it is being injected in the heterogeneous porous medium. The concentration field has been rescaled by the maximum local concentration.