Groundwater flowing through soil pores and rock fractures accounts for about 97% of the liquid freshwater on earth (excluding snow and ice). Rivers and lakes thus constitute only a very small part of the continental water bodies. During its journey, from its infiltration into the ground, its flow through the rocks and its resurgence in springs, rivers, lakes or the ocean, groundwater undergoes a series of transformations, which determine its chemical composition: the one that can be read on the labels of bottled water, for example.
Tanguy Le Borgne, a teacher-researcher at the Université de Rennes 1, is studying reactive mixing fronts in particular: these are areas where groundwater converges, where slicks of different compositions mix and react with each other. The reaction rates appear to be extremely high compared to the rest of the environment. These localized chemical reactions in the mixing fronts are used in many applications. For example, they help to improve water quality (removal of nitrates) or to trap contaminants. But they can also be harmful to the development of geothermal power plants, when bacterial biofilms clog the boreholes.
"Oxygen-rich rainwater infiltrates the soil until it meets an old water table, which is low in oxygen but high in iron, for example. Where these two waters meet, there is a proliferation of particular bacteria: those that are able to react iron and oxygen to produce energy and fix carbon for other micro-organisms."
An ambitious project
It all started with two important advances obtained with Yves Méheust (Geosciences / OSUR) and Hervé Tabuteau (IPR / OSUR): the researchers succeeded for the first time in visualizing a reactive mixing front at very high resolution within a porous medium.
"It was previously assumed that these waters, once brought into contact, mixed diffusely and homogeneously as they flowed through the pores of the rock. This is not the case at all! On the contrary, we observed that these mixing fronts have filamentary structures similar to those that develop in turbulent flows (smoke, milk in coffee, for example). This means that the models used today for geothermal energy, pollution control or to analyse the prospects for CO2 sequestration in the ground are obsolete: they give predictions that may be largely incorrect. They absolutely must be improved," says Tanguy Le Borgne.
The stakes are high. The ReactiveFronts project led by T. Le Borgne was selected by the European Research Council in 2015. It was allocated a budget of 2 million euros over five years, following a very tough competition. For the researcher, it is the results obtained upstream of the project, the rich interdisciplinary environment of the OSUR where it was set up and the support of the UBL's European project platform that explain its selection. The funds allow the financing of the salaries of three PhD students and two post-doctoral fellows, teaching leave for T. Le Borgne and Y. Méheust, as well as innovative experimental equipment.
The scale of the resources allocated by Europe goes hand in hand with the ambition of the project. Indeed, the study of microbial hotspots is only one of the five pillars of ReactiveFronts.
Field, manoeuvres and numerical models
To set up the ReactiveFronts project, T. Le Borgne combined three approaches that are often used separately: field observation, experimental laboratory imaging and computer modelling. "In the world of hydrology, only at OSUR in Rennes is it possible to combine such a wide range of approaches in a single project," emphasises the researcher.
In the field, the project is based on the H+ network of experimental boreholes, developed over the last 15 years in France, which T. Le Borgne is currently coordinating with two colleagues: Olivier Bour and Cédric Champollion. At the Ploemeur site, for example, using a camera inserted into boreholes about 100 metres deep, he observed that at certain depths bacteria were profiling in very large numbers, while at others nothing was visible. The hypothesis is that these hotspots are located in very active mixing zones, where very high concentrations of dissolved oxygen and iron meet (strong concentration gradients). One of the challenges of the ReactiveFronts project is to predict at what depth and with what intensity these hotspots develop, depending on the flow and the types of soil and rock.
To improve these models, Tanguy Le Borgne is currently exploring the analogy between flows in soils and turbulent flows, in collaboration with Emmanuel Villermaux's team in Marseille.
In the Hydrology Experimentation Laboratory, coordinated by Yves Méheust, the doctoral students and researchers of the ERC team have developed an experimental device that is much more advanced than the initial prototypes. At the centimetre scale, it is possible to make the grains and the fluid transparent: the filamentary structure of the plumes, visualised by coloured tracers, on which the analogy with turbulent flows is based, can be seen for the first time in 3D.
This 3D model does not allow the direct study of bacteria in the flows. For this purpose, the team led by T. Le Borgne built another device, on the micron scale (that of the bacteria). This uses new microfluidic imaging techniques developed with Hervé Tabuteau of the IPR to study bacteria in flows in collaboration with Alexis Dufresne of the ECOBIO laboratory. Under the microscope, it is possible to form a double gradient of oxygen and dissolved iron, which recreates the conditions existing in the microbial hotspot observed in the Ploemeur borehole. The first results are coming in, and they are very promising.
The next objective will be to combine field observations, 3D flow experiments and microfluidic imaging to build a numerical model integrating all these data.
"Finally, we will test our simulation results in the field by implementing reactive tracing experiments: we will inject tracers into the boreholes and measure their dispersion and reactivity either by direct measurement in the boreholes or by new geophysical imaging techniques, where the environment is scanned in a similar way to medical imaging," indicates T. Le Borgne.
A better understanding of the conditions under which these reactions develop will eventually help engineers to manage water and energy resources. Similarly, the models developed in the project could eventually help to better predict the phenomena involved in the geological sequestration of CO2, one of the solutions currently being considered to curb climate change.