Applied Geochemistry (v.27, #7)

► Hysteresis does not play a major role in this case (sandy deposits). ► 2D axi-symmetric model is needed to reproduce field observations, i.e. boundary effects play a role. ► Geological layering needs to be accounted for. ► Optimisation of parameter values carried out with HYDRUS gave sensible results. ► Numerical model validated by a forward 1D model of the test site under natural circumstances.This article describes an infiltration–drainage test carried out in the unsaturated zone (UZ) at the Chernobyl Pilot Site during October 2008; this is an international radioecology study site and is the subject of several papers in this special issue. The test has to be seen in the larger context of radionuclide transport from a waste trench. The conducted experiment consisted of infiltrating a layer of 9.5 cm of water in a circular area of 5.51 m2 over 5 h. Its main objective was to create a larger range of water content values (and hence suction pressure values), not only at the top of the soil profile but also at greater depths, in this case up to 1.50 m. Observations of water content and suction pressure were carried out continuously at seven different depths during infiltration, drainage and during the return to natural conditions over a period of several months. This allowed deriving UZ parameter values with greater confidence than those derived from monitoring small natural water content changes over periods of years.The experiment also shows that a 1D model was incapable of correctly reproducing the observed water balance established on the central axis. A 2D axi-symmetric model was needed showing that lateral boundary effects have to be accounted for. This implies that small scale infiltrometer tests should be analysed in more than one dimension. It further became clear from the drainage curves that soil layering played a significant role and that in the present case hysteresis did not play a major role; i.e. the infiltration and drainage event can be reproduced with the same set of parameters and subsequent natural rain events can be reproduced well enough with two numerical models used for test interpretation, one is module oriented for reactive transfer, and the second with an automatic optimisation procedure. The robustness of the estimated parameter values, of the model discretisation and layer identification was tested over a 2-month period with natural infiltration using a 1D model. The results validated the test outcome.

Groundwater residence time downgradient of Trench No. 22 at the Chernobyl Pilot Site: Constraints on hydrogeological aquifer functioning by C. Le Gal La Salle; L. Aquilina; E. Fourre; P. Jean-Baptiste; J.-L. Michelot; C. Roux; D. Bugai; T. Labasque; C. Simonucci; N. Van Meir; A. Noret; S. Bassot; A. Dapoigny; D. Baumier; P. Verdoux; D. Stammose; J. Lancelot (1304-1319).
► Stratification of the groundwater is evidenced through Na and Cl and 18O and 2H. ► Good agreement is observed between 3H/3He and CFC ages. ► Complex mixing processes are evidenced at intermediate depth. ► Apparent age distribution suggest exponential or piston flow model. ► Extremely high SF6 concentrations are attributed to the nuclear reactor explosion.Following the explosion of reactor 4 at the Chernobyl power plant in northern Ukraine in 1986, contaminated soil and vegetation were buried in shallow trenches dug directly on-site in an Aeolian sand deposit. These trenches are sources of radionuclide (RN) pollution. The objective of the present study is to provide constraints for the Chernobyl flow and RN transport models by characterising groundwater residence time. A radiochronometer 3H/3He method (t 1/2  = 12.3 a) and anthropogenic tracers including CFC and SF6 are investigated along with the water mass natural tracers Na, Cl, 18O and 2H.The groundwater is stratified, as evidenced by Na and Cl concentrations and stable isotopes (18O, 2H). In the upper aeolian layer, the Na–Cl relationship corresponds to evapotranspiration of precipitation, while in the underlying alluvial layer, an increase in Na and Cl with depth suggests both water–rock interactions and mixing processes. The 3H/3He and CFC apparent groundwater ages increase with depth, ranging from ‘recent’ (1–3 a) at a 2 m depth below the groundwater table to much higher apparent ages of 50–60 a at 27 m depth below the groundwater table. Discrepancies in 3H/3He and CFC apparent ages (20–25 a and 3–10 a, respectively) were observed during the 2008 campaign at an intermediate depth immediately below the aeolian/alluvial sand limit, which were attributed to the complex water transfer processes. Extremely high SF6 concentrations, well above equilibrium with the atmosphere and up to 1112 pptv, are attributed to significant contamination of the soils following the nuclear reactor explosion in 1986. The SF6 concentration vs. the apparent groundwater ages agrees with this interpretation, as the high SF6 concentrations are all more recent than 1985. The persistence of the SF6 concentration suggests that SF6 was introduced in the soil atmosphere and slowly integrated in the groundwater moving along the hydraulic gradient. The apparent age distribution in the lumped parameter models suggests an exponential or piston flow model in the upper geological section, followed by more pronounced mixing processes in the lower section.

Geochemical influence of waste trench no. 22T at Chernobyl Pilot Site at the aquifer: Long-term trends, governing processes, and implications for radionuclide migration by D. Bugai; E. Tkachenko; N. Van Meir; C. Simonucci; A. Martin-Garin; C. Roux; C. Le Gal La Salle; Yu. Kubko (1320-1338).
► Groundwater geochemistry data set for a waste trench in Chernobyl zone is analyzed. ► Leaching losses from the trench are governed by degradation of buried organic matter. ► Geochemical impacts decrease in time due to humification of the organic matter. ► Additional factor controlling leaching losses is nutrient element uptake by vegetation. ► Evolution of geochemical regime favors attenuation of groundwater migration of 90Sr.This article discusses and analyzes data for geochemical monitoring of groundwater, collected in 1998–2008 at the international experimental radioecological study site (Chernobyl Pilot Site) located in the Red Forest radioactive waste dump site in the Chernobyl exclusion zone. Groundwater in the zone of influence of the studied waste trench no. 22T was characterized by a specific geochemistry. Along with a high content of 90Sr (of an order of n  × 1000–n  × 10,000 Bq/L) groundwater showed elevated concentrations of Ca, K, NO 3 - , SO 4 2 - and of some trace elements (in particular stable Sr), and had more acidic pH values compared to “background” aquifer conditions. The observed water quality changes are apparently related to degradation of organic matter (pine forest remnants, litter, humus containing topsoil layer) buried inside trench no. 22T, which have lead to acidification of groundwater and leaching of cations adsorbed on the exchange complex of soils buried inside the trench. Regime monitoring data for the project period of 1998–2008 shows a progressive decrease of concentrations of the leached ions accompanied by an increase of pH in the chemical plume emerging from the trench. This can be explained by a combined effect of a gradual decrease of the inventory, humification of the original organic matter inside the trench, and of nutrient element uptake by roots of the newly planted pine forest on top of the trench. The identified trends of evolution of geochemical regime favor attenuation of subsurface migration of 90Sr from Red Forest waste dumps due to an increase of the 90Sr distribution coefficients, and stabilization of plumes of contaminated groundwater.

Migration of transuranic elements in groundwater from the near-surface radioactive waste site by S. Levchuk; V. Kashparov; I. Maloshtan; V. Yoschenko; N. Van Meir (1339-1347).
► Plutonium isotopes spatial distribution in unconfined aquifer near shallow waste dump. ► We found that plutonium in aquifer was originated from the dump. ► Pu mobility in waste and aquifer is characterized by in situ partition coefficients. ► Pu in groundwater is associated mainly with low molecular weight species (<1 kDa).Field experiments and laboratory studies were performed to investigate migration processes of plutonium isotopes from a near-surface radioactive waste trench to the underlying sandy aquifer at the Red Forest waste dump in the Chernobyl zone. The objectives of these experiments were to characterize the spatial distribution and possible migration mechanisms of plutonium in the aquifer. During 2002–2007 experimental investigations were carried out and spatial distributions of plutonium isotopes (239 , 240Pu, 238Pu), 90Sr and major ions in the aquifer in the direction of the groundwater flow were obtained. Specific activities of radionuclides in groundwater depended on the location of the piezometer and varied in the range of 1–360 mBq kg−1 for 239 , 240Pu, 0.5–180 mBq kg−1 for 238Pu and n–n·104 Bq kg−1 for 90Sr. It was found that the spatial features of the distributions of plutonium and strontium specific activities in the upper eolian aquifer were similar, i.e. there was a correlation between the positions of the activity maxima of the radionuclides. The Pu isotopes plume in the aquifer spreads about 15 m downstream of the radionuclides source. Characterization of the initial radionuclide composition of the waste showed that all plutonium in the aquifer originated from the trench. The ratio of plutonium isotopes (239 , 240Pu/238Pu) at the sampling time was the same in waste material and in groundwater samples. In situ ultrafiltration of several groundwater samples was carried out. The size fractionation data obtained suggest that a significant part of plutonium (50–98%) in the groundwater sampled close to the source from the upper part of the aquifer is associated with a very low molecular weight fraction (<1 kDa).

► We describe radionuclide distribution and inventories in the shallow waste dump. ► Radionuclides in the waste dump are associated with fuel particles of three types. ► For each fuel particle type the dissolution parameters have been obtained. ► Biogenic migration fluxes of radionuclides from the waste dump have been determined. ► The effects of radiation on Scots pine in the waste dump territory have been characterized.This review article introduces an experimental site located within the Chernobyl exclusion zone that is equipped to study radionuclide behavior in the environment after disposal of radioactive waste into shallow subsurface storage (trenches). This paper presents how the site is equipped and the methodology that was followed in order to understand and reproduce the observed 90Sr contaminant plume downstream from a shallow waste trench in an area about 2.5 km west of the Chernobyl Nuclear Power Plant (ChNPP), called the Red Forest. The main results include identification of the radionuclide source term (the distribution and inventories of radionuclides in the trench, the description of the physical and chemical properties of the fuel particles encountered in the waste trench) and a model of fuel particle dissolution and subsequent radionuclide leaching into the soil solution. The biogenic migration of radionuclides from the trench to, and effects of radiation on, plants (Scots pine) are also described.

Radionuclide migration at experimental polygon at Red Forest waste site in Chernobyl zone. Part 2: Hydrogeological characterization and groundwater transport modeling by D. Bugai; A. Skalskyy; S. Dzhepo; Yu. Kubko; V. Kashparov; N. Van Meir; D. Stammose; C. Simonucci; A. Martin-Garin (1359-1374).
► Strontium-90 migration from the waste trench in Chernobyl zone is studied. ► Results of the detailed hydrogeological characterization program are presented. ► The 1D and 2D models for 90Sr transport in groundwater are developed. ► Model calibrations indicate a need to account for transient geochemical regime. ► Models served useful tools for risk assessment of the waste site.This article represents the second of two articles, which review the main results of the international radioecological projects: Chernobyl Pilot Site Project (1999–2003) and Experimental Platform in Chernobyl (2004–2008). These projects studied radionuclide migration from the near-surface radioactive waste trench at the Red Forest waste dump in the Chernobyl zone, which contained nuclear fuel particles. This article presents results from the comprehensive hydrogeological site characterization program including the following issues: geological structure of the study site, hydraulic properties of the deposits, tracer tests in the aquifer, results of groundwater monitoring and unsaturated zone regime studies, as well as data on the 90Sr distribution in the unsaturated zone and aquifer, and analyses of 90Sr sorption behavior. The derived parameters were used to develop and calibrate 1D (flow tube) and 2D (cross-section) models describing the migration of 90Sr from the studied waste trench to the unsaturated zone and aquifer over a 16-a period (1986–2002). The models involved the following sub-models: (1) the geostatistical (structural) model for radioactivity distribution in the trench (using GSLIB); and (2) the radionuclide source term model (STERM1D) describing dissolution of fuel particles and a 1D of radionuclide redistribution in the trench body and unsaturated zone. The MODFLOW – MT3D codes were used to model the 2D 90Sr transport in the aquifer cross-section. Calibration of the 1D model with respect to Kd s and dispersivities allowed quite accurate reproduction of 90Sr migration behavior for the early period (1995–1998). The less perfect fit between the 1D and 2D modeling results and monitoring data for the later period (1999–2002) suggests the need to improve the conceptual radionuclide migration model (i.e. to account for transient hydraulic and geochemical regimes of the waste site).

Microbial diversity in contaminated soils along the T22 trench of the Chernobyl experimental platform by Virginie Chapon; Laurie Piette; Marie-Hélène Vesvres; Frédéric Coppin; Claire Le Marrec; Richard Christen; Nicolas Theodorakopoulos; Laureline Février; Sviatoslav Levchuk; Arnaud Martin-Garin; Catherine Berthomieu; Claire Sergeant (1375-1383).
► We examined the diversity of bacterial communities inhabiting Chernobyl soils. ► Highly and weakly RN contaminated samples were collected at the trench T22 site. ► Molecular- and culture-based approaches evidenced a wide diversity of bacteria. ► The presence of RN does not exert a high selection pressure on bacterial communities.The diversity of bacterial communities exposed to radioactive contamination in Chernobyl soils was examined by a combination of molecular and culture-based approaches. A set of six radioactive soil samples, exhibiting high levels of 137Cs contamination, were collected from the T22 trench. Three samples were also collected in nearby soils with low contamination. Complex bacterial community structures were observed in both highly and weakly contaminated samples, using a molecular approach targeting the 16S rRNA gene. However, the presence of specific populations within samples from highly contaminated soils could not be revealed by statistical analysis of the DGGE profiles. More than 200 culturable isolates, representative of dominant morphotypes, were grouped into 83 Operational Taxonomic Units (OTUs) and affiliated to Firmicutes, Actinobacteria, Alpha-, Beta-, Gamma-Proteobacteria and Bacteroïdetes. No specific pattern linked to contamination was observed for these culturable bacteria. The results show that both highly and weakly contaminated soils host a wide diversity of bacteria, suggesting that long term exposure to radionuclides does not lead to the extinction of bacterial diversity.

► Cement/argillite engineered analogue. ► Alkaline fluid flow through an argillite matrix. ► Alkaline fluid flow along argillite discontinuities. ► Mineralogical towards geochemical disturbances.Samples of Toarcian argillite were collected both next to and far from a CEM II cement paste and a CEM II concrete, within the specific context of a 15-a old borehole located in the Tournemire Experimental Platform (Aveyron, France). The objectives were evaluation of the mineralogical and geochemical changes of the claystone at the contact with the cementitious materials and determination of the spatial extent of the interactions. The approach includes the examination of the mineralogical (XRD, SEM, TEM), chemical (major, trace, rare earth elements) and isotopic (Sr, C, O) compositions of argillite whole-rocks and of various soluble phases, at two scales: in the rock matrix (P1 scale) and along micro-cracks (P2 scale). The two study scales outline nearly similar mineralogical modifications, shown by the presence of Ca silicate hydrates (C–S–H) and newly-formed CaCO3 within 10–15 mm of the cement paste and concrete. Chemical data from whole-rock argillites indicate few changes in a slightly thicker zone (18–20 mm), mainly consisting of an increase in the CaO wt.%, and a decrease in Sr contents. The other elementary contents remained quite constant except for MgO, which suggests redistribution with precipitation of a Mg-rich mineral phase at 20 mm from cement paste/concrete interface. Acetic acid leachates had more pronounced variations, including a decrease of the total elementary content in the same ‘geochemical disturbed zone’ (GDZ), together with a significant increase of the Sr isotopic ratios. A combination of Sr and C/O isotopic patterns was used to distinguish the behavior of secondary cementitious phases in the clay-rich rock: (i) calcite dissolution and re-precipitation is supported by C/O isotopic data and (ii) C–S–H neoformation is evidenced by the 87Sr/86Sr ratios; this tool also contributes to determine the origin of the fluids. The proportion of newly-formed C–S–H in the matrix and in the micro-cracks of the argillite is modeled.

► Diffusion properties in an argillaceous formation at the Tournemire site. ► Changes after the excavation of galleries over a long time period up to 123 years. ► Radiogenic helium depth profiles in four galleries of different ages. ► Excesses of Ar, Kr, and Xe suggest diffusion of air from the gallery. ► Model calculations confirm effect the aging of galleries on the EDZ.The purpose of this work is to study the diffusion properties in an argillaceous formation and how these properties change after the excavation of galleries, over a long period. The Tournemire Underground Research Laboratory (URL) located in Aveyron (France) offers the unique opportunity to investigate different stages of the evolution of the so-called Excavation Damaged Zone (EDZ) over a period of about 124 a. Four horizontal boreholes parallel to the bedding and one vertical borehole were drilled in order to study diffusion transport in the EDZ in both directions in galleries excavated 124, 13, 6, and 1 a ago, respectively. Noble gases dissolved in the pore water of 47 core samples were extracted in order to establish precise depth profiles. Mainly produced in situ by decay of U and Th, radiogenic He accumulates in the pore water of rocks and escapes toward the gallery, allowing a characterization of the diffusive transport in the EDZ and its evolution with time. For the oldest gallery (1882–1888), core mapping reveals an EDZ well developed with fractures up to a depth of 1.8 m. In this zone, all radiogenic He has been lost and at the same time heavy noble gases are trapped in the rock with an enrichment of Xe relative to Kr and of Kr relative to Ar. From 1.8 m, He concentration increases with depth to reach a plateau at about 3.5 m corresponding to a concentration normally found in the unaltered zone. Considering diffusive transport of the radiogenic He to the EDZ, model calculations allowed proposing transport properties in the unfractured and undisturbed zone (UZ) with a pore diffusion coefficient Dp of 1 × 10−10  m2/s deduced from the best-fit curve to experimental data and a mean porosity of 0.1 obtained form petrophysical measurements. The most recent gallery (2008) did not show any evidence of EDZ at depths greater than 7.5 cm. A fracture located around 0.75 m, possibly of tectonic origin, induced large advection and diffusion losses of He on both sides in the next 50 cm. Excesses of Ar, Kr and Xe are found, suggesting diffusion of air from the gallery associated with He losses. The 6-a-old gallery is clearly influenced by fractures of tectonic origin increasing the EDZ thickness of 0.7 m up to a Damaged Zone of 1.8 m. Based on the He concentration profile, the 13-a-old gallery (1996) has a net EDZ of 1.2 m, larger than the 0.7 m deduced from the structural analysis by core mapping, followed by a disturbed zone of up to 3.5 m that could represent an intermediate stage before the development of fractures in the EDZ. The vertical borehole drilled in the same gallery shows a less extensive damaged zone (1.4 m) and allows the determination of a Dp of 3 × 10−11  m2/s taking into account the anisotropy relative to the stratification plane. The rare gas analysis has, therefore, enabled confirmation and better description of the net time-dependent evolution of the EDZ around drifts in the argillaceous Tournemire site: no obvious EDZ for the 1-a gallery (2008), 1.2 m width for the 13-a-old gallery (1996), and 2 m width for the century-old tunnel. The occurrence of a disturbed and unfractured zone (dZ) around the 13-a-old gallery undetected by structural studies and geophysical methods clearly demonstrates the power of radiogenic He in assessing the transport properties around drifts excavated at different ages in compacted clay rocks.

Geochemical characterization and modelling of the Toarcian/Domerian porewater at the Tournemire underground research laboratory by J. Tremosa; D. Arcos; J.M. Matray; F. Bensenouci; E.C. Gaucher; C. Tournassat; J. Hadi (1417-1431).
► A suitable model missed to obtain the porewater composition of Tournemire argillite. ► Fine geochemical experimental characterization of water–clay system is carried out. ► Water–clay interactions by thermodynamic modelling using multi-site cation exchange. ► Good data representation with multi-site model, better than previous models. ► Efficient reproduction of K behaviour, sensitive to exchange site proportion.For safety evaluation of hazardous waste repositories in clay-rocks, a thorough assessment of porewater chemistry and water–rock interactions is required. However, this objective is a challenging task due to the low hydraulic conductivity and water content of such rocks, which subsequently renders porewater sampling difficult (without inducing perturbations). For this reason, an indirect approach was developed to determine porewater composition of clay-rocks, by a geochemical model of water–rock interaction using some properties of the rock and the solution. The goal of this paper is to obtain the porewater composition of the Toarcian/Domerian argillaceous formation at Tournemire (South of France), for which a reliable model is still lacking. The following work presents a comprehensive characterization of the geochemical system of the Tournemire clay-rock, including mineralogy, petrology, mobile anions, cation exchange properties, accessible porosity and CO2 partial pressure. Perturbation corrections from fracture water sampling were also computed. These water were found in sealed fractures () and their radiocarbon apparent age is estimated at 20 ka. Their age together with their equilibrium situation allow considering these fracture waters as representative of the formation porewater. The model developed to calculate the Tournemire porewater composition is essentially based on cation exchange by a multi-site approach, but equilibrium with some mineral phases (calcite, quartz and pyrite) is also considered. Different exchange sites of different affinities towards cations are used, which proportions are given by the mineralogy. Exchange on illite is performed with a three-sites model, while one site is considered for smectite phases. Multi-site model results are compared with corrected fracture water data and two other models: a model only based on mineral equilibrium and a model using cation exchange on one global site. The best results were obtained with the models that take into account cation exchange and particularly with the multi-site model. The interest in considering a model with exchange sites of different affinities is particularly obvious for a satisfactory representation of the K+ content in solution. A dependence of K+ content to the amount of high affinity sites was observed, leading to an improvement of its simulation when uncertainty on mineralogical data is considered. Once validated, the multi-site model was applied at different levels of the Tournemire argillaceous formation to obtain a profile of the porewater composition.

► We study diffusion of HTO, 36Cl and 125I in Tournemire argillite samples. ► Effects of initial iodide concentration and ionic strength were investigated. ► Iodide retention increases with the decrease of initial iodide concentration. ► Higher ionic strength solution results in higher anion diffusivity. ► Discrepancy between 125I and 36Cl effective diffusion coefficient was highlighted.Diffusion parameters for HTO, 36Cl , and 125I were determined on Upper Toarcian argillite samples from the Tournemire Underground Research Laboratory (Aveyron, France) using the through diffusion technique. The direction of diffusion was parallel to the bedding plane. The purpose of the present study was 3-fold; it was intended (i) to confirm the I interaction with Upper Toarcian argillite and to verify the effects of initial I concentration on this affinity, as previously observed by means of radial diffusion experiments, (ii) to highlight any discrepancy between Cl and I diffusivity, and (iii) to investigate the effect of an increase of the ionic strength of the solution on the anionic tracers’ diffusive behaviour. The results show that the effective diffusion coefficient (De ) and diffusion accessible porosity (εa ) values obtained with an ionic strength (I.S.) synthetic pore water of 0.01 eq L−1 are: De  = 2.35–2.50 × 10−11  m2  s−1 and εa  = 12.0–15.0% for HTO, and De  = 14.5–15.5 × 10−13  m2  s−1 and εa  = 2.5–2.9% for 36Cl. Because of anionic exclusion effects, anions diffuse slower and exhibit smaller diffusion accessible porosities than HTO, taken as a water tracer. The associated effective diffusion coefficient (De ) and rock capacity factor (α) obtained for 125I are: De  = 7.00–8.60 × 10−13  m2  s−1 and α  = 4.3–7.2%. Such values make it possible to calculate low 125I distribution ratios (0.0057 <  RD  < 0.0192 mL g−1) which confirm the trend indicating that the 125I rock capacity factor increases with the decrease of the initial I concentration. Additional through-diffusion experiments were carried out with a higher ionic strength synthetic pore water (I.S. = 0.11 eq L−1). No evolution of HTO diffusion parameters was observed. The anionic tracers’ effective diffusion coefficient increased by a factor of two but no clear evolution of their accessible porosity was observed. Such a paradox could be related to the particularly small mean pore size of the Upper Toarcian argillite of Tournemire. The most significant finding of this study is the large discrepancy (factor of two) between the values of the effective diffusion coefficient for 125I and 36Cl. Whatever the ionic strength of the synthetic solution used, 125I exhibited De values two times lower than those of 36Cl. A detailed explanation for this difference cannot be given at present even if a hypothesis based on ion-pairing or on steric-exclusion cannot be excluded. This makes questionable the assumption usually made for quantifying 125I sorption and postulating that 36Cl and 125I would diffuse in the same porosity. In other terms, at Tournemire, 125I sorption could be more pronounced than previously indicated.

Microbial diversity of the 180 million-year-old Toarcian argillite from Tournemire, France by Laurent Urios; François Marsal; Delphine Pellegrini; Michel Magot (1442-1450).
► The cultivable microbial diversity of the argillite of Tournemire was characterized. ► The biodiversity of the undisturbed argillite is limited to three Gram-positive genera. ► That of EDZ differed, depending mainly on oxygen availability and moisture content. ► The overall metabolic diversity is limited to a few trophic groups. ► All attempts of DNA extraction for molecular approaches failed.Even though the microbiology of various subsurface environments has been investigated for more than 30 a, the microbial diversity of deep argillaceous media is still poorly known. In the context of radioactive waste disposal in clayey formations, the consequence of microbial activity is of concern regarding e.g. the corrosion of metallic components. The purpose of the present work was to characterise the cultivable microbial diversity of different zones of the Toarcian argillite of Tournemire (France) as a preliminary indication regarding the potential of development of microbes in such subterrestrial environments. Cores were drilled in the Excavation Damaged Zone (EDZ) and in the deeper undisturbed zone of the argillite layer, as well as in a zone intersected by a geological fault. Samples from the wall of the drift were also collected. Microorganisms were cultivated from all samples, but the biodiversity differed depending mainly on the O2 and moisture content. Aerobic bacteria were identified on the wall, in the EDZ and in the wet faulted area, whereas SO4-reducing bacteria were isolated from the wet faulted area only. Anaerobic heterotrophs were cultivated from all zones. One hundred and twelve isolates were identified. Small ribosomal subunit gene sequences showed that bacteria of the undisturbed zone were affiliated to three genera only, whereas the three other sampled zones harbour more diverse microflora, including isolates closely related to taxons characterized from subsurface, deep marine and polar environments.