This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Transferability of geoscientific information from various sources (study sites,underground rock laboratories, natural analogues) to support safety cases forradioactive waste repositories in argillaceous formations

Martin Mazurek a,*, Andreas Gautschi b, Paul Marschall b, Georges Vigneron c, Patrick Lebon c,Jacques Delay c

a University of Bern, Rock-Water Interaction, Institute of Geological Sciences, Baltzerstr. 3, CH-3012 Bern, Switzerlandb Nagra, Hardstr. 73, CH-5430 Wettingen, Switzerlandc Andra, Parc de la Croix Blanche, 1/7 rue Jean Monnet, 92298 Châtenay-Malabry, Cedex, France

a r t i c l e i n f o

Article history:Available online 14 October 2008

Keywords:TransferabilityArgillaceous formationGeological disposalCallovo–Oxfordian shaleOpalinus Clay

a b s t r a c t

In studies related to deep geological disposal of radioactive waste, it is current practice to transfer exter-nal information (e.g. from other sites, from underground rock laboratories or from natural analogues) tosafety cases for specific projects. Transferable information most commonly includes parameters, investi-gation techniques, process understanding, conceptual models and high-level conclusions on systembehaviour. Prior to transfer, the basis of transferability needs to be established. In argillaceous rocks,the most relevant common feature is the microstructure of the rocks, essentially determined by the prop-erties of clay–minerals. Examples are shown from the Swiss and French programmes how transfer ofinformation was handled and justified. These examples illustrate how transferability depends on thestage of development of a repository safety case and highlight the need for adequate system understand-ing at all sites involved to support the transfer.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

In several national programmes for the disposal of radioactivewaste, argillaceous formations are considered as potential hostrocks. The stages of development of the various programmes arequite different. Some programmes are at the stage of desk studies,others at the stage of generic investigations from the surface orfrom underground rock laboratories, and some have identified po-tential sites and initiated their characterisation. Given the differ-ences in the degree of development and sophistication, intechnical focus and in project-relevant time scales, exchange ofinformation and technical collaboration among the implementingorganisations is very active. Aspects of common interest are inves-tigated in facilities such as the underground laboratories at MontTerri (Switzerland), Mol (Belgium) and Bure (France). Internationalactivities are also co-ordinated in Framework Programmes of theEuropean Union. For example, the SELFRAC project was targetedat understanding self sealing of fractures in the excavation-dis-turbed zone around excavations in clays and shales (Blümlinget al., 2007; Bastiaens et al., 2007). NEA’s Working Group on Mea-surement and Physical Understanding of Groundwater Flowthrough Argillaceous Media (Clay Club) is another institution tar-

geted at international collaboration. The Clay Club identified spe-cific fields of interest and launched technical initiatives, such as:

� Water, gas and solute transport through argillaceous media(Horseman et al., 1996).

� Catalogue of Characteristics of argillaceous rocks (Boisson,2005).

� Methods of sampling and interpretation of pore-waters in argil-laceous formations (Sacchi and Michelot, 2000; Sacchi et al.,2001).

� Self sealing in argillaceous rocks (Bock, in preparation).� Features, events and processes evaluation catalogue for argilla-

ceous media (FEPCAT, Mazurek et al., 2003).� Natural tracer profiles across argillaceous formations – review

and synthesis (CLAYTRAC, Mazurek et al., 2008).

While such co-ordinated actions increase the general processunderstanding and knowledge of the properties of argillaceous sys-tems, the transfer of information to safety cases1 for specific sites is

1474-7065/$ - see front matter � 2008 Elsevier Ltd. All rights reserved.doi:10.1016/j.pce.2008.10.046

* Corresponding author. Tel.: +41 31 631 87 81; fax: +41 31 631 48 43.E-mail address: mazurek@geo.unibe.ch (M. Mazurek).

1 A safety case is a collection of arguments, at a given stage of repositorydevelopment, in support of the long-term safety of the repository. It comprises thefindings of a safety assessment and a statement of confidence in these findings. Itshould acknowledge the existence of any unresolved issues and provide guidance forwork to resolve these issues in future development stages. A safety case is the endproduct of a safety assessment.

Physics and Chemistry of the Earth 33 (2008) S95–S105

Contents lists available at ScienceDirect

Physics and Chemistry of the Earth

journal homepage: www.elsevier .com/locate /pce

Author's personal copy

not trivial and needs to be done with care, considering possible differ-ences in spatial and temporal scales, formation properties and statesof the system. This paper investigates two well characterised shaleformations, namely the Callovo–Oxfordian at Bure (France) and theOpalinus Clay in northern Switzerland, and identifies properties andprocesses that may (or may not) be transferred from one formationor location to the other. A description of the methodology and the lo-gic reasoning related to transferability is another important objective.

2. Basic site descriptions

2.1. Callovo–Oxfordian at Bure, France

In the eastern part of the Paris Basin, Andra is conducting stud-ies in and around an underground research laboratory near Bure(also called ‘‘Site Meuse–Haute Marne”), mainly targeting at aflat-lying, ca. 130 m thick marine shale formation of Callovo–Oxfordian age at 420–550 m below surface. On a regional scale,the area is characterised by very low seismotectonic activity. Theformation and its embedding units are lithologically homogeneousin the lateral dimensions on a scale of tens of kilometres aroundthe laboratory, which indicates deposition in a tectonically quietenvironment. Given the limited burial (ca. 850 m below surface)that occurred in the Cretaceous, diagenetic effects are weak andessentially limited to compaction and the formation of carbonatecement (Clauer et al., 2007; Andra, 2005a). Clay–mineral reactions,such as the partial illitisation of smectite, have not been identified(Clauer et al., 2007; Trouiller, 2006; Andra, 2005a). Brittle deforma-tion is weak and mostly focussed in known major faults that occurseveral kilometres away from the laboratory. The Callovo–Oxfor-dian is embedded by the Dogger and Oxfordian limestones, whichcontain horizons with aquifer properties (Delay and Distinguin,2004; Andra, 2005a).

Since 1991, the region has been extensively studied from thesurface by 2D seismic campaigns, a series of deep boreholes (witha total of 4300 m of cores in the Callovo–Oxfordian) and a 3D seis-mic campaign covering 4 km2. Since August 2000, the constructionof the underground research laboratory near Bure has led to thecollection of in situ information during the shaft-sinking activitiesand the excavation of galleries at 445 m below surface (upper partof the formation) and 490 m below surface (middle part of the for-mation), augmented by studies conducted from the galleries. Theseinvestigations provided the basis for an extensive report (Référent-iel géologique; Andra, 2005a), which is a part of the feasibilitystudy, called ‘‘Dossier 2005” (Andra, 2005b), for a geological repos-itory in the investigated region. After the evaluation of these doc-uments, the French Government issued a new law in June 2006,requiring Andra to select a disposal site and to design a repositorywithin the studied region as a basis for an authorisation applica-tion in 2015 and a scheduled commissioning in 2025.

2.2. Opalinus Clay in the Zürcher Weinland, Switzerland

In the Zürcher Weinland, Opalinus Clay (a middle Jurassic mar-ine shale formation) is flat lying, essentially undeformed and oc-curs at a depth of 550–650 m below surface (Nagra, 2002a). Itwas subjected to a complex burial history, during which a maxi-mum depth of ca. 1650 m and a maximum temperature of ca.85 �C were reached (Mazurek et al., 2006). It is not directlybounded by aquifers but embedded by lithologically more hetero-geneous (even though often argillaceous) units. Natural tracer pro-files of Cl� and of water isotopes indicate that diffusion is thedominant transport mechanism between the late Jurassic and lateTriassic aquifers, i.e. over a vertical distance of ca. 300 m (Gimmiand Waber, 2004; Gimmi et al., 2007).

In 1994, Opalinus Clay was identified as the priority sedimen-tary host rock option for the disposal of spent fuel, vitrified high-le-vel and long-lived intermediate level waste in Switzerland, and theZürcher Weinland (north-east Switzerland) as the first-priorityarea for site-related investigations. Detailed characterisation ofthe host rock and the potential siting area followed after 1994.The key elements of this research programme were a 3D seismiccampaign in the Zürcher Weinland covering an area of around50 km2, an exploratory borehole at Benken, experiments as partof the international research programme in the Mont Terri under-ground research laboratory, comparative regional studies on Opali-nus Clay including deep boreholes in the near and far vicinity of thesiting area, and comparisons with clay formations that are underinvestigation in other countries in connection with geological dis-posal. Synthesis reports have been published (Nagra, 2002a,b).

2.3. Opalinus Clay at Mont Terri, Switzerland

The Mont Terri underground research laboratory is located innorth-western Switzerland (Canton Jura) and consists of a dedicatedtunnel section that branches off from an existing motorway tunnelacross the Mont Terri anticline in the Folded Jura Mountains. It is lo-cated in Opalinus Clay, ca. 270 m below surface (Thury and Bossart,1999a,b). At Mont Terri, Opalinus Clay is 160 m thick, occurs withinan anticlinal fold and therefore is affected by brittle deformation(Freivogel and Huggenberger, 2003). The formation was subjectedto two burial stages, and maximum burial to ca. 1350 m below sur-face occurred in the late Cretaceous (Mazurek et al., 2006). Thin-skinned folding occurred in the period 10.5–3 Ma (Berger, 1996;Bolliger et al., 1993), followed by partial erosion of the anticline, pro-gressively exposing deeper strata down to the Triassic.

Research managed by the International Mont Terri Consortiumwas started in 1996 and is documented in Thury and Bossart(1999b) and follow-up synthesis reports on geochemistry (Pearsonet al., 2003), hydrogeology (Marschall et al., 2004), excavation-dis-turbed zone/rock mechanics (Martin and Lanyon, 2002) and otherproject-specific reports. In future, the main focus will be on long-term and demonstration experiments. Mont Terri is a genericunderground facility and will not be considered as a disposal site.

3. Basic properties of the considered formations: Analogies anddifferences

The most relevant parameters of the formations under discus-sion are summarised in Table 1, where similarities and differencesbecome evident. Similarities include:

� Qualitatively, the mineralogy of the shales at all sites is similar.The same is true for the chemical type of the pore-water.

� Hydraulic conductivities and diffusion coefficients are in compa-rable ranges.

� Both formations are slightly to moderately over-consolidated.� Geomechanical parameters are similar.

Differences include:

� The clay–mineral content of the Callovo–Oxfordian at Bure issubstantially lower when compared to Opalinus Clay, especiallyat Mont Terri. Also, lithological heterogeneity in the verticaldimension is more pronounced in the Callovo–Oxfordian.

� Opalinus Clay was subjected to a deeper burial than the Callovo–Oxfordian, which is reflected by a higher degree of consolidationand a more pronounced anisotropy of a number of geomechan-ical and transport parameters. As an example, the anisotropyfactor of the diffusion coefficient for HTO is ca. 5 in Opalinus Clay

S96 M. Mazurek et al. / Physics and Chemistry of the Earth 33 (2008) S95–S105

Author's personal copy

(Nagra, 2002a), compared to 62 in the Callovo–Oxfordian (Poca-chard et al., 1997; Homand et al., 2006).

� Given the different burial evolutions, a major difference inporosity would be expected, which is not the case. The reasonis a diagenetic cementation that reduced porosity of the Cal-lovo–Oxfordian, whereas cementation is essentially absent inOpalinus Clay.

4. The basis of transferability

There are various levels on which information can be trans-ferred to a specific safety case from other sources (results from dis-

posal programmes, information from underground rocklaboratories and natural analogues):

� Individual parameters.� Investigation techniques and data evaluation methods.� Process understanding.� Conceptual models.� High-level conclusions (e.g. engineering feasibility, safety

aspects).

In argillaceous formations, the microscopic structure governsmany macroscopic properties, including, among others, transportand geomechanical properties (see, e.g., Wenk et al., 2008 for an

Table 1Basic characteristics of the target formations.

Property/parameter Callovo–Oxfordian at Bure Opalinus Clay in the ZürcherWeinlanda

Opalinus Clay at Mont Terri

Depositional environment Shallow marine Shallow marine Shallow marineAge Middle Callovian – Lower Oxfordian, ca. 163–

158 MaAalenian, ca. 174 Ma Aalenian, ca. 174 Ma

Max. temperature reached duringdiagenesis (�C)

33–38 85 85

Present burial depth (centre (m)) 488 596 275Maximum burial depth (centre (m)) 850 1650 1350Over-consolidation ratio (–) 1.5–2 1.5–2.5 2.5–3.5Thickness (m) 138 113 160Clay–minerals (weight-%) 25–55 54 66Clay–mineral species (in the order of

decreasing abundance)Illite/smectite mixed-layers, illite, (chlorite,kaolinite)

Illite, kaolinite, ill/smec mixed-layers, chlorite

Illite, kaolinite, ill/smec mixed-layers, chlorite

Quartz (weight-%) 20–30 20 14Feldspars (weight-%) 1 3 2Calcite (weight-%) 20–38 16 13Dolomite/ankerite (weight-%) 4 1 b.d.Siderite (weight-%) <1 4 3Pyrite (weight-%) 1–2 1.1 1.1Organic carbon (weight-%) 0–1 0.6 0.8CEC (meq/100 g rock) 11b–22 10.6 11.1Pore-water type Na–Cl–SO4 Na–Cl–SO4 Na–Cl–SO4

Mineralisation (mg/L) 6530 12898 18296Eh (mV) <�150 �170 �227Bulk wet density (Mg/m3) 2.3 2.52 2.45Water content (weight-% rel. to dry weight) 8.6 4.0 7.03Physical porosity (–) 0.14b–0.18 0.12 0.16Anion-accessible porosity (–) 0.09 0.06 0.09Total specific surface (m2/g)c 56–78 (BET H2O) 90 (EGME) 130 (EGME)External specific surface (m2/g) 43 (BET N) 28 (BET N) 31 (BET N)Eff. diffusion coeff. De (HTO) normal to

bedding (m2/s), anisotropy factord2.6E�11, <2e 6.1E�12, 5 1.5E�11, 5

Eff. diffus. coeff. De (Cl, Br, I) normal tobedding (m2/s), anisotropy factord

1E�12–1E�11, Anisotropy not measured(expected small)

6.5E�13, 5 4.1E�12, 6

Hydraulic conductivity K normal to bedding(m/s), anisotropy factord

5E�14–5E�13, 2–10 2.4E�14, 1–10 4.0E�14, ca. 5

Seismic velocity Vp (m/s) 3000–3200 (normal)f, 3200–3400 (parallel)f 3030 (normal), 4030 (parallel) 2620 (normal), 3030 (parallel)Uniaxial compressive strength normal to

bedding (MPa)21f 30 16g

Undrained E (tangent) modulus normal tobedding (MPa), anisotropy factord

4000f, 1.3 5500, 2.1 3600, 1.6

Cohesion normal to bedding at lowconfining pressure (MPa)

6.4f 8.6 5

Friction angle normal to bedding at lowconfining pressure (�)

29f 25 25

Swelling pressure (MPa) 1.0–3.0f (anisotropy not measured) 1.1 (normal), 0.15 (parallel) 1.2 (normal), 0.5 (parallel)

Diffusion coefficients De were measured at ambient laboratory temperature (ca. 22 �C). Definition: De ¼ nv=s2D0 = nDp, with Dp = pore diffusion coefficient, D0 = free waterdiffusion coefficient, n = porosity, v ¼ constrictivity, s = tortuosity; v=s2 = geometry factor G.Data sources: Andra (2005a), Bock (2000), Boisson (2005), Nagra (2002a).

a Including Murchisonae Beds in Opalinus Clay facies (see Nagra, 2002a).b Value for the most carbonate-rich beds (upper part of the formation).c Specific areas derived from BET (H2O) are typically smaller than those measured by EGME on the same samples. Therefore, the values are only comparable when based on

the same methodology.d Anisotropy factor = value parallel/value normal to bedding.e According to Pocachard et al. (1997).f Value at the level of the underground research laboratory.g Recommended value according to Bock (2000).

M. Mazurek et al. / Physics and Chemistry of the Earth 33 (2008) S95–S105 S97

Author's personal copy

example of the latter). Fig. 1 illustrates the microstructure onscales of millimetres to nanometres. The microstructure is essen-tially determined by clay–mineral platelets with grain sizes of afew lm or less, arranged in a network of plate-to-plate andedge-to-plate contacts (see, e.g., Yven et al., 2007). This network re-sults in high surface areas and in an equally complex geometry ofthe interstitial space, i.e. the water-filled porosity, which is charac-terised by apertures in the range of nm. The electrostatic interac-tion between water molecules and clay–mineral surfaces leads tothe distinction of ‘‘bound” and ‘‘free” water – only the latter isessentially unaffected by the presence of clay–minerals. Moreover,the accessible pore space is solute specific, and only a fraction ofthe physical porosity is accessible to anions due to the repulsiveforces near clay–mineral surfaces. Given the generally reducingconditions that prevail in argillaceous units (except for those thatwere deposited under oxidising conditions, or units that are lo-cated in surficial environments today), sorption of solutes onclay–mineral surfaces is an important mechanism retarding trans-port of these solutes.

This type of microstructure is common to all argillaceous for-mations and is the fundamental basis of transferability among dif-ferent sites and formations. In this respect, some differences occuramong argillaceous formations and are mainly due to variations ofthe

� Degree of compaction (and therefore porosity).� Degree of diagenetic cementation.� Contents of coarser-grained clastic material, such as quartz and

carbonate minerals.� In situ stress regime, in particular the magnitude of shear stress.

Thus, good knowledge of mineralogy, of diagenetic evolutionand of tectonic, burial and thermal history is essential for theassessment of transferability. Microstructure, i.e. the nature, size

and shape of mineral grains and the way these grains are spatiallyorganised, is essentially defined by the mineralogy and porosity.Therefore, the latter are among the most important characteristicsto be considered for the transfer of data and processes among dif-ferent argillaceous systems.

4.1. Why do we want to transfer features, processes and know-how,and at which stage?

Specific programmes are at quite different stages of develop-ment (borehole data, underground facilities) and also pursue coun-try-specific strategies of developing a safety case. There are thefollowing major motivations for the transfer of information amongformations/sites:

1. Information from other sites is taken to complement the infor-mation on specific issues in a programme, in order to obtainsuitable data for a preliminary safety case. For example, thebasis for extrapolating laboratory measurements to in situ con-ditions can be justified if both laboratory and in situ data areavailable in other programmes. As an example, comprehensivedata bases such as the Clay Club Catalogue of Characteristics(Boisson, 2005) or the catalogue of features, events and pro-cesses for argillaceous systems (FEPCAT, Mazurek et al., 2003)can be used to transfer findings from other programmes.

2. Information from other sites is taken to highlight that the inves-tigated formation and/or site does not represent an exceptionalsituation but has relevant features in line with other argilla-ceous formations. If independent programmes convergetowards consistent data sets and conclusions, confidence isbuilt in the national programme. Examples: Demonstration ofdiffusion as the dominant transport process, of the efficiencyof fault self sealing, of the hydraulic irrelevance of faults, ofreducing geochemical environments, etc. On the other hand,

Fig. 1. Microstructure of shales and pore-water in argillaceous formations on scales of mm (thin-section), lm (SEM) and nm (schematic representation).

S98 M. Mazurek et al. / Physics and Chemistry of the Earth 33 (2008) S95–S105

Author's personal copy

the identification of evident differences can be used to guidefuture research.

3. Dedicated investigation techniques are often developed andevaluated in underground rock laboratories and then used forsite characterisation elsewhere. Practical examples include,among others, the instrumentation for in situ diffusion experi-ments and the development of geotechnical equipment.

4. Conceptual models on different levels (individual features andprocesses, their coupling and safety relevance) can be trans-ferred from other, similar sites or underground rock laborato-ries. Examples include the identification of relevant transportprocesses in the geosphere and the role of natural and inducedfractures on flow and transport.

In practice, point 1 above is more relevant in the early stages ofdeveloping a safety case, while point 2 becomes more important inmature safety cases. The more mature a safety case, the less infor-mation is used from elsewhere, and the more are external data,general observations and conclusions used as additional, indepen-dent lines of evidence. Points 3 and 4 are important throughout thewhole process of developing a repository.

Natural analogues are used as independent lines of evidence forsafety cases in all stages of development. The strength of naturalanalogues lies in the fact that they record processes that occurredover large scales in space and time, often comparable with thoseneeded for safety assessment, and thus are helpful for the upscal-ing of experimental work conducted in the laboratory or in under-ground research facilities (e.g. Miller et al., 2000). However, thetransferability of information from natural analogues to specificrepository sites is limited by a number of weaknesses, such asthe difficulty to constrain the age, duration and interplay of rele-vant processes, significant differences in the geological settings,the boundary conditions and their evolution with time. Due tothese limitations, a substantial part of the transferable informationis qualitative in nature, even though not exclusively.

In some cases, specific features of the target formations them-selves can be used as ‘‘self analogues”. For example, natural tracerprofiles in pore-water (such as anions, water isotopes, dissolvednoble gases) may provide information on the dominant transportmechanisms (Gimmi and Waber, 2004; Gimmi et al., 2007; Mazur-ek et al., 2008), and anomalous hydraulic-head distributions indi-cate very small hydraulic conductivity on the formation scale.

4.2. What can be transferred, what cannot, and how?

As discussed above, mineralogy and porosity are among themost important macroscopic parameters of argillaceous systems.Many parameters and processes are functions mainly of mineral-ogy and porosity, and the understanding of this dependence is animportant pre-requisite for transferability. In principle, the follow-ing types of features and processes can be distinguished (see alsoTable 2):

� Features and processes that do not strongly depend on known orexpected differences between different argillaceous formations;

e.g. thermodynamic data used to quantify water/rockinteractions.

� Features and processes that essentially depend on mineralogyand porosity but only insignificantly on the state of the system(in situ stress, hydraulic head, temperature, pore-water salinity);e.g. permeability, thermal conductivity and some geomechanicalproperties. The dependence on mineralogy and porosity requiresdetailed characterisation, and this is often linked to a thoroughknowledge of the geological evolution of the sites involved. Intheory, the dependence on mineralogy and porosity could bequantitatively assessed by microscopic models. However, dueto the complex handling and parameterisation of such models,empirical relationships are generally preferred.

� Features and processes that essentially depend on mineralogyand porosity but also on other properties and on the state ofthe system; e.g. sorption properties (also depending on pore-water composition and temperature), geometry and extent ofthe excavation-damaged zone (EDZ) around underground struc-tures. In cases where information on complex system behaviourhas to be transferred into another environment (e.g. the geome-try and the properties of the EDZ around tunnels in the targetarea, where no direct observations are currently available),transferability is more complex and appropriate conceptualmodels, computational tools and parameter sets have to be usedin order to assess the combined effect of rock properties andsite-specific conditions.

� Features and processes that depend substantially on the state ofthe system; e.g. hydraulic head and gradient, salinity of pore andground water. For such data, site-specific information is essen-tial and transfer among sites is generally not possible. In safetyassessment, long-term evolution scenarios have to be consid-ered for such parameters (for example the evolution of porepressure and hydraulic gradient in response to glacial loadingor erosion).

Example 1 (Diffusion coefficients). For argillaceous rocks, there isan empirically defined relationship between the logarithm of theeffective diffusion coefficient and porosity, as shown in Fig. 2 forHTO in the direction normal to bedding. A similar relationship canalso be seen for the logarithm of the pore diffusion coefficient andporosity, consistent with Archie’s law with empirical parametersbetween 2.5 and 3.5 (formalism see Van Loon et al., 2003). Itfollows – as an empirical conclusion – that compaction of anargillaceous formation reduces both porosity and tortuosity in aregular way. If porosity is known, diffusion coefficients can beestimated on the basis of this correlation.

Example 2 (Hydraulic conductivity). Fig. 3 (adapted from Neuzil,1994) shows that there is a positive correlation between porosityand the logarithm of hydraulic conductivity for porosities largerthan ca. 0.05–0.1. However, for highly compacted (and, at thedepth of observation, also strongly over-consolidated) formationswith smaller porosities, the trend is broken, and effective hydraulic

Table 2Transferability of features and parameters among different shale formations.

Dependence on mineralogyand porosity

Dependence on the stateof the system

Transferability Example

None to weak None to weak Directly transferable Thermodynamic and kinetic dataStrong None to weak Transferable using empirical relationships Hydraulic conductivity, strengthsStrong Moderate Transfer is more complex (requires underlying conceptual

models and model calculations)Geometry of the excavation-damaged zone(EDZ), sorption characteristics

Strong Not transferable Hydraulic head, pore-water composition

M. Mazurek et al. / Physics and Chemistry of the Earth 33 (2008) S95–S105 S99

Author's personal copy

conductivity may increase sharply. This is probably due to the factthat matrix permeability becomes insignificant when compared tofracture permeability. Severe over-consolidation and fracturingtypically occur in highly compacted systems that have lost theirplasticity and part of their self-sealing capacity due to the morelimited occurrence of swelling clay–minerals. In such systems,the transferability of hydraulic conductivity and some otherparameters is limited. Comprehensive data sets from highly com-pacted argillaceous formations are only available for a small num-ber of specific sites:

� Toarcian–Domerian at Tournemire, France: maximum burialdepth ca. 2,000 m, maximum temperature P110 �C (Barbarandet al., 2001; Constantin et al., 2004; Peyaud et al., 2005).

� Boda Clay Formation, Hungary: ca. 4,000–5,000 m, 200 �C (Kov-acs, 1999).

� Palfris Formation at Wellenberg, Switzerland: ca. 10,000 m,190–250 �C (Mazurek, 1999).

Whereas permeability of the Toarcian–Domerian at Tournemiremay be locally enhanced but very small on the formation scale

(Boisson et al., 2001; Mazurek et al., 2008, Section 5.6), the Bodaand Palfris Formations are clearly fractured media in whichfracture flow dominates. Due to the limited number of casestudies, it is currently difficult to more accurately specify theformation properties at which fractures become hydraulicallyimportant. In addition to porosity, other characteristics, such asover-consolidation ratio, stress regime, cementation, self-sealingcapacity etc. also play a role.

5. Practical experience with transferability

5.1. Andra’s Dossier 2005 for the site Meuse–Haute Marne (Bure)

Andra gained substantial experience from experimental workcarried out in the underground rock laboratories at Mol (over thelast 20 years) and Mont Terri (since 1996). The main pointsinclude:

� The development of experimental tools and methods.� Questions related to scale issues.

Fig. 2. Effective diffusion coefficient for tritium (normal to bedding) in various argillaceous formations. The shaded area indicates the range consistent with Archie’s law withempirical parameter values m = 2.5–3.5. Data from Andra (2005c), Atkins (1990), Boisson (2005), Bourke et al. (1993) and Nagra (1993, 2002a).

Fig. 3. Hydraulic conductivity in various argillaceous systems. Data sources – Formations 1–10; Neuzil (1994), shown in dashed lines; formations 11–13, 17–19: Boisson(2005); 14: Andra (2005c); 15–16: Nagra (2002a). Note that data without a documented experimental protocol in Neuzil’s (1994) compilation were not considered.

S100 M. Mazurek et al. / Physics and Chemistry of the Earth 33 (2008) S95–S105

Author's personal copy

� The development of conceptual models.� Testing the comprehensiveness of the site characterisation and

safety approach.

5.1.1. Development of experimental tools and methodsThe availability of tested experimental tools and protocols be-

fore and during the development of the experimental programmefor the Meuse–Haute Marne underground research laboratory re-sulted in a substantial gain of time. Examples, all documented inAndra (2005a,b), include:

� Design and dimensioning of in situ diffusion tests (e.g. DIRexperiment, Dewonck et al., 2006) in the Bure underground lab-oratory were based on an analogous experiment (DI) at MontTerri (Palut et al., 2003; Tevissen et al., 2004).

� Geotechnical instruments, such as extensometers and inclinom-eters, were originally developed at Mol (CLIPEX experiment, Ber-nier et al., 2003) and are now used for the REP experiment (shaftsinking response) at Bure.

� The feasibility and hydraulic performance of the engineered bar-rier system considering the cutting of radial trenches into thetunnel walls, thereby removing the excavation-damaged zone(EDZ), and subsequently backfilling them by bentonite bricks,was tested in the EZ-A (‘‘EDZ cut-off”) experiment at Mont Terriin 2003 and 2004 (Fig. 4; Armand, 2004). Focus was placed on(1) the study of possible mechanical damage due to trench exca-vation, (2) feasibility of bentonite emplacement, and (3) testingthe hydraulic performance of the seals by cross-hole packertests. This successful experiment was used to design and dimen-sion the KEY experiment in the Bure underground laboratory,which will run over extended periods of time in order to achievefull saturation of the bentonite. A custom saw has been built tocut trenches up to 2.7 m deep into the tunnel walls.

5.1.2. Scale issuesInformation from Mont Terri has also been used for scale issues.

At Mont Terri, diffusion coefficients for various species (HTO, 36Cl�,I�, 22Na+) were measured both in the laboratory (through-diffusioncells; Van Loon et al., 2002, 2004a; Tevissen et al., 2004) and in situ(DI and DI-A experiments; Palut et al., 2003; Tevissen et al., 2004;Van Loon et al., 2004b; Wersin et al., 2004). The in situ data char-acterise scales in time and space at least 1 order of magnitudegreater than the laboratory data. The resulting coefficients were

very similar for a given solute (Table 3), which was taken as evi-dence that the scale dependence is small. This conclusion was usedto justify the use of laboratory-derived diffusion coefficients for theCallovo–Oxfordian at Bure for larger scales. Use of this conclusionwas made both for the design of the in situ test at Bure (Dewoncket al., 2006) and for preliminary modelling of solute migration onthe site scale (Andra, 2005a).

5.1.3. Conceptual models – examples pertinent to the role of transportalong brittle discontinuities

The general contention that fractures in argillaceous media areinfrequent and/or hydraulically insignificant is one of the mainarguments for considering these rocks for waste isolation (Bock,in preparation). Andra used observations and conceptual modelsfrom Mont Terri to better justify this contention in the safety argu-mentation for a potential disposal site in the Meuse–Haute Marneregion, regarding both artificial fractures in the EDZ as well as nat-ural fractures.

A preliminary conceptual model for the geometry and hydraulicsignificance of the EDZ around the tunnels in the Bure under-ground laboratory was needed at a stage when no site-specificmeasurements were available. Given the different settings and de-grees of compaction of Opalinus Clay at Mont Terri and the Cal-lovo–Oxfordian at Bure, the in situ stress field and the degree ofanisotropy of geomechanical properties are distinctly different(Table 1), and some differences are also identified in strengthsand moduli. Therefore, a direct transfer of observations and mea-surements pertinent to the EDZ at Mont Terri was not feasible,and the current conceptual model of the EDZ at Bure is based ondata derived from geomechanical laboratory experiments and sub-sequent modelling. However, the same modelling approach wassuccessfully used to predict the EDZ geometry at Mont Terri (usinglocal parameters), and this adds confidence to the overall validityof the chosen approach (Andra, 2005a). In the same way, the eval-uation of the initial EDZ permeability was assessed taking into ac-count the information from Mont Terri (Andra, 2005c).

Another conceptual aspect is the time evolution of the EDZ,which has not yet been thoroughly investigated at Bure. Giventhe fact that swelling pressures of Opalinus Clay and the Cal-lovo–Oxfordian are comparable, the self-sealing behaviour can beassumed to be similar. Accordingly, the characterisation of fractureclosure and related reduction of EDZ permeability with ongoingresaturation of the host rock and of the bentonite backfill at Burewas guided by the observations made at Mont Terri (Andra,2005a,c).

Fig. 4. EDZ cut-off experiment at Mont Terri: Opening and filling a trench with bentonite bricks.

M. Mazurek et al. / Physics and Chemistry of the Earth 33 (2008) S95–S105 S101

Author's personal copy

A similar argument can be made regarding self sealing of natu-ral fractures and faults. Various types of brittle discontinuitieswere identified at Mont Terri, while fractures are remarkably rareat Bure. At Mont Terri, observations and measurements indicatethat natural brittle discontinuities of all types are hydraulicallyinsignificant (unless part of the EDZ). This is true for small fracturesas well as for major faults with extents of more than 100 m, indi-cating that self sealing is very efficient (Marschall et al., 2004;Andra, 2005a). The scarcity (or, as at the 490 m level of the under-ground laboratory, the absence) of observed fractures at Bure, to-gether with the efficient self sealing observed at Mont Terri, wasone of the justifications for disregarding fractures as potential flowconduits in solute transport calculations for Bure (Andra, 2005c).

5.2. Gas transport mechanisms in Opalinus Clay: Lessons learned atMont Terri

Gas transfer from the emplacement tunnels through the hostrock has been recognised as an important issue in Nagra’s assess-ment of long-term safety of a repository for radioactive waste inOpalinus Clay of northern Switzerland. High emphasis was there-fore given to the elaboration of a well-founded, defensible database on gas-related properties of the host rock. The investigationsconcentrated on the deep borehole in the candidate repository re-gion in the Zürcher Weinland, on the Mont Terri rock laboratory(Nagra, 2002a; Marschall et al., 2005) and on observations madeelsewhere (e.g. oil/gas industry). The approach consisted of the fol-lowing steps:

� Collect site-specific data bases on gas-related rock parameters.� Develop structural and process models of gas transport in Opali-

nus Clay and check their applicability to both sites.� Check consistency of parameters and concepts with information

from elsewhere and derive generalised empirical relationships.

The opportunities for gas-related investigations in the Benkenborehole (Zürcher Weinland) were quite restricted due to the tech-nical constraints associated with deep borehole investigations(poor borehole stability, limited accessibility of host rock, demand-ing core sampling procedures) and due to the tight time scheduleof the geoscientific investigation programme. One single in situgas injection test was performed, complemented by two long-termgas permeability tests on core specimens and a variety of micro-structural analyses (e.g. nitrogen and water adsorption/desorption,mercury porosimetry). The corresponding programme at MontTerri was more comprehensive both in extent of testing and exper-imental sophistication (including in situ experiments with exten-sive hydro-mechanical crosshole instrumentation and gaspermeability tests on high quality rock specimens, recovered withadvanced core sampling procedures, in order to minimise sample

disturbance). The challenge was to combine the strengths of thetwo data sets by a traceable geoscientific methodology, which linksboth the conceptual understanding of gas transport mechanismsand the gas related parameters from Mont Terri to the geologicalconditions in the Zürcher Weinland, augmented by informationfrom elsewhere.

The development of process models was based on the fact thatgas transport in low-permeability formations is largely controlledby the microstructure of the rock. The pore space of Opalinus Clayat both sites is formed essentially by a network of micro/meso- andmacropores with effective pore throats in the order of 1–100 nm(Fig. 1). Even though the tectonic regime at Mont Terri is distinctand includes fractures and faults, these do not contribute signifi-cantly to the bulk permeability of the rock due to efficient self seal-ing (Marschall et al., 2004). Further, minor differences between thetwo sites include the slightly lower clay–mineral content and high-er degree of compaction in the Zürcher Weinland when comparedto Mont Terri (Table 1), which results in a somewhat different geo-mechanical behaviour.

The phenomenological description of gas transport processes inOpalinus Clay has been driven by the above mentioned microstruc-tural conceptualisation of the rock (Fig. 5a) and suggests that thefollowing basic transport mechanisms may occur, depending onthe gas transport rate and the local stress field:

� Advective–diffusive transport of gas dissolved in the pore-water.� Visco-capillary two-phase flow.� Dilatancy controlled gas flow (‘‘pathway dilation”).� Gas transport along macroscopic tensile fractures (hydro- and

gas fracturing).

The complex hydro-mechanical processes of gas transferthrough the rock can be decomposed into the issues of flow ofimmiscible fluids (Fig. 5b) and of geomechanical behaviour(Fig. 5c). In Nagra’s gas-related studies, experimental evidencewas needed to describe the postulated transport processes andgeomechanical material laws. The corresponding data base wasgained by a targeted hydrogeological and geomechanical investiga-tion programme at Mont Terri (Marschall et al., 2005) and con-firmed by checking the consistency of experimental results withthe available data from the Zürcher Weinland. The comparison ofresults comprised a broad spectrum of aspects such as clay con-tents, pore size distributions, capillary pressure curves, gas perme-ability measurements, stress–strain relationships and rock failuremodes.

Gas permeabilities and gas entry pressures are somewhat dif-ferent in the Benken borehole (k = 1E�21 to 7E�21 m2, pae = 4–10 MPa) and at Mont Terri (k = 1.5E�20 to 6E�20 m2, pae = 0.2–1 MPa). The differences can be satisfactorily explained as mainlydue to slightly different degrees of compaction (and therefore

Table 3Comparison of effective diffusion coefficients for Opalinus Clay from Mont Terri measured in the laboratory and in situ.

Species Through-diffusion cell @ 14 �C DI in situ experiment DI-A in situ experiment (short term)

De (m2 s�1) xapp De (m2 s�1) xapp De (m2 s�1) xapp

Van Loon et al. (2004a) Tevissen et al. (2004) Van Loon et al. (2004b)

HTO 4.9E�11 0.17 5.0E�11 0.15 4.0E�11 0.15I� (36Cl�) 1.2E�11 0.12 1.5E�11 0.13 1.0E�11 0.0922Na+ 6.9E�11 0.66 Not measured 6.0E�11 0.64

The laboratory determinations by Van Loon et al. (2004a) were performed at 22–24 �C, and resulting diffusion coefficients were corrected for the effects of porous filters. Thevalues at 14 �C were obtained using the formalism of Van Loon et al. (2005) and an activation energy of 20 kJ/mol.In situ temperature at Mont Terri is 14 �C.xapp: Apparent porosity (=accessible porosity�retardation factor).All data refer to the direction parallel to bedding.

S102 M. Mazurek et al. / Physics and Chemistry of the Earth 33 (2008) S95–S105

Author's personal copy

porosity and pore-size distribution), of the in situ stress field and ofclay–mineral content. The negative correlation between perme-ability and gas entry pressure as seen in Opalinus Clay at the twoinvestigated sites is what would be expected on the basis of theo-retical considerations. Qualitatively, the correlation corresponds toempiric trends derived from various studies in other formations(Davies, 1991; Ingram et al., 1997), and this adds further confi-dence in the results.

Comparing the gas transport characteristics of Opalinus Claywith the expected production rates of repository gas, it is con-cluded that the dominant gas transport mechanism will be two-phase flow with a possible contribution of dilatancy controlledgas flow, whereas no gas fracs are expected to occur in the poten-tial siting area.

6. Conclusions

6.1. General

Transfer of external data and concepts as input to specific safetycases is an important element of confidence building. The basis oftransferability as well as its limits need to be elaborated and justi-fied in each specific case, and it may vary substantially among thedifferent groups of host rocks (e.g. argillaceous and crystallinerocks, salt formations). Depending on the nature of the feature orprocess in question, this basis is established by the characterisa-tion, understanding and comparison of:

� Relevant formation properties (such as microstructure, mineral-ogy, porosity, redox conditions), and/or

� The states of the system (such as stress state and pore-waterpressure)

in the concerned sites/formations. In many cases, it can beshown empirically that the feature or process to be transferred isessentially a function of a limited number of formation properties(for example, diffusion coefficients in argillaceous rocks dependmainly on porosity). In such cases, the basis of transferability isthe characterisation of the empirical relationships. Further confi-dence in this methodology is gained (i) by including data and in-sights from as many sites and formations as possible, and (ii) byexplicitly modelling these relationships.

Transfer occurs at different levels, depending on (1) the level ofmaturity of the safety case and (2) the quality of the analogy thatcan be made between the sites and formations concerned. In theearly stages of a safety case development, information is trans-ferred from other sites to complement the information gained fromsite characterisation and to obtain suitable and defendable data fora preliminary safety case. At this stage, the basis of transferabilitymay not yet be that well established, and conservative assump-tions may also be needed. In mature safety cases, the role of infor-mation transfer is less to supply data but to contribute to processunderstanding and confidence building, for example by means ofestablishing empirical relationships that include information fromdiverse sites and settings.

In summary, it is concluded that transfer of information,whether explicitly or implicitly, occurs at all stages in the develop-ment of a repository and includes data from specific similar sites(repository sites, underground rock laboratories, natural ana-logues) but also the much larger (even though less similar) bodyof external information documented in the literature.

6.2. Conclusions specific to argillaceous rocks

The main commonalty of argillaceous rocks (and therefore themost commonly applicable basis of transferability) is the presence

Fig. 5. Classification and analysis of gas transport processes in Opalinus Clay after Marschall et al. (2005): (a) phenomenological description based on the microstructuralmodel concept, (b) basic transport mechanisms, (c) geomechanical regime and (d) effect of gas transport on the barrier function of the host rock.

M. Mazurek et al. / Physics and Chemistry of the Earth 33 (2008) S95–S105 S103

Author's personal copy

of clay–minerals with their distinct properties that determine anumber of relevant features and processes, such as:

� Advection and diffusion of water and gas depend on the struc-ture of the pore space (interconnected network on a nanometricscale), which in turn is determined by the spatial arrangement ofclay platelets.

� Sorption is mainly determined by the high specific surfaces andreactivity of clay–minerals.

� Self sealing of fractures depends on the swelling properties ofclay–minerals and on the rheology of the rocks.

� Geomechanical properties, such as strengths and moduli,depend on the nature of the minerals as well as on the micro-structure of the pore space and its water content.

� The degree of anisotropy of transport and geomechanical prop-erties is linked to the bedding-parallel alignment of clayplatelets.

Transferability is more restricted among formations with sub-stantially different mineralogical compositions and porosities be-cause these basic properties affect many other features andprocesses. The same is true for transfer of information among siteswith very contrasting states of the system (e.g. pore-water salinity,stress field). Overall, transferability is feasible at least to some de-gree among weakly to moderately consolidated formations. Thefact that fracture flow and transport may occur in some highlycompacted formations limits the transferability of findingsfrom/or to less compacted formations, as the dominant role ofmicrostructure as a basis for transferability may no longer apply.

In the specific examples illustrating transfer of informationamong Opalinus Clay at Mont Terri, in the Zürcher Weinland andthe Callovo–Oxfordian at Bure, the basis of transferability for anumber of transport-related and geomechanical features and pro-cesses is particularly good due the similarity of the mineralogicalcomposition and of porosity. Differences in the states of the systemexist but are well known and can be accounted for when transfer-ring information to different formations and sites.

Acknowledgements

Helpful comments and reviews by Johan Andersson (JA Stream-flow AB, Sweden), Peter Blümling and Piet Zuidema (both Nagra,Switzerland), Delphine Pellegrini (IRSN, France) and Jean-LucMichelot (Uni. Paris-Sud, France) are gratefully acknowledged.

References

Andra, 2005a. Dossier 2005 – Référentiel du site Meuse/Haute-Marne. Andra Reportno. C.RP.ADS.04.0022, Andra, France.

Andra, 2005b. Dossier 2005 Argile – Evaluation de la faisabilité du stockagegéologique en formation argileuse profonde – Rapport de synthèse, June 2005,Andra, France (<http://www.andra.fr>).

Andra, 2005c. Dossier 2005 Argile – Tome «Evolution phénoménologique dustockage géologique». Andra report no. C.RP.ADS.04.0025, Andra, France(available at <http://www.andra.fr>).

Armand, G. 2004. Premiers enseignements de l’expérimentation EZ-A au Mont-Terri. Laboratoire de recherche souterrain de Meuse/Haute-Marne. Andra reportno. D.RP.ADPE.04.1034, Andra, France.

Atkins, P.W., 1990. Physical Chemistry, fourth ed. Oxford University Press, Berlin.Barbarand, J., Lucazeau, F., Pagel, M., Séranne, M., 2001. Burial and exhumation

history of the south-eastern Massif Central (France) constrained by apatitefission-track thermochronology. Tectonophysics 335, 275–290.

Bastiaens, W., Bernier, F., Li, X.L., 2007. SELFRAC: experiments and conclusions onfracturing, self-healing and self-sealing processes in clay. Phys. Chem. Earth 32,600–615.

Berger, J.P., 1996. Cartes paléogéographiques-palinspastiques du bassin molassiquesuisse (Oligocène inférieur – Miocène moyen). N. Jb. Geol. Paläont. Abh. 202, 1–44.

Bernier, F., Li, X.L., Verstricht, J., Barnichon, J.D., Labiouse, V., Bastiaens, W., Palut,J.M., Ben Slimane, K., Ghoreychi, M., Gaombalet, J., Huertas, F., Galera, J.M.,Merrien, K., Elorza, F.J., Davies, C., 2003. CLIPEX: clay instrumentation

programme for the extension of an underground research laboratory. FinalReport, European Commission, Brussels, EUR 20619 EN.

Blümling, P., Bernier, F., Lebon, P., Martin, C.D., 2007. The excavation damaged zonein clay formations time-dependent behaviour and influence on performanceassessment. Phys. Chem. Earth 32, 588–599.

Bock, H., 2000. RA experiment, Rock mechanics analyses and synthesis: data reporton rock mechanics. Mont Terri Consortium Technical Report TR 2000-02.

Bock, H. (Ed.), in preparation. Potential for Self-healing of Fractures in Plastic Claysand Argillaceous Rocks Under Repository Conditions. Nuclear Energy Agency,OECD, Paris.

Boisson, J.Y., Bertrand, L., Heitz, J.F., Moreau-Le Golvan, Y., 2001. In situ andlaboratory investigations of fluid flow through an argillaceous formation atdifferent scales of space and time, Tournemire tunnel, southern France.Hydrogeol. J. 9, 108–123.

Boisson, J.Y. (Ed.), 2005. Catalogue of Characteristics of Argillaceous Rocks Studiedwith Respect to Radioactive Waste Disposal Issues. OECD/NEA IGSC WorkingGroup on the Characterisation, the Understanding and the Performance ofArgillaceous Rocks as Repository Formations (the ‘‘Clay Club”), OECD/NEA,Paris, France.

Bolliger, T., Engesser, B., Weidmann, M., 1993. Première découverte desmammifères pliocènes dans le Jura neuchâtelois. Eclogae Geol. Helv. 86,1031–1068.

Bourke, P.J., Jefferies, N.L., Lever, D.A., Lineham, P.R., 1993. Mass transfermechanisms in compacted clays. In: Manning, D.A.C., Hall, P.L., Hughes, C.R.(Eds.), Geochemistry of Clay-Pore Fluid Interactions. Chapman & Hall, pp. 331–350.

Clauer, N., Fourcade, S., Cathelineau, M., Girard, J.P., Vincent, B., Elie, M., Buschaert,S., Rousset, D., 2007. A review of studies on the diagenetic evolution of theDogger-to-Tithonian sedimentary sequence in the Eastern Paris Basin – impacton the physical and chemical rock properties. Mém. Soc. Géol. France 178, 59–72.

Constantin, J., Peyaud, J.B., Vergély, P., Pagel, M., Cabrera, J., 2004. Evolution of thestructural fault permeability in argillaceous rocks in a polyphased tectoniccontext. Phys. Chem. Earth 29, 25–41.

Davies, P.B. 1991. Evaluation of the role of threshold pressure in controlling flow ofwaste-generated gas into bedded salt at the waste isolation pilot plant (WIPP).Sandia Rep. SAND 90-3246, Sandia National Laboratories, Albuquerque, NewMexico.

Delay, J., Distinguin, M., 2004. Hydrogeological investigations in deep wells at theMeuse/Haute-Marne underground research laboratory. In: Hack, R., Azzam, R.,Charlier, R. (Eds.), Engineering Geology for Infrastructure Planning in Europe,Lecture Notes in Earth Sciences, vol. 104. Springer, pp. 219–225.

Dewonck, S., Descostes, M., Blin, V., Tevissen, E., Radwan, J., Poinssot, C., Cartalade,A., Brisset, P., Coelho, D., 2006. In situ diffusion experiments in Callovo–Oxfordian mudstone. Geochim. Cosmochim. Acta 70 (Suppl. 1), A140.

Freivogel, M., Huggenberger, P., 2003. Modellierung bilanzierter Profile im GebietMont Terri – La Croix (Kanton Jura). In: Heitzmann, P., Tripet, J.P. (Eds.), MontTerri Project – Geology, Paleohydrogeology and Stress Field of the Mont TerriRegion. Report no. 4 of the Federal Office for Water and Geology, Bern,Switzerland, pp. 7–44.

Gimmi, T., Waber, H.N., 2004. Modelling of Tracer Profiles in Pore-Water ofArgillaceous Rocks in the Benken Borehole: Stable Water Isotopes, Chloride, andChlorine Isotopes. Nagra Technical Report NTB 04-05, Nagra, Wettingen,Switzerland.

Gimmi, T., Waber, H.N., Gautschi, A., Rübel, A., 2007. Stable water isotopes in pore-water of Jurassic argillaceous rocks as tracers for solute transport over largespatial and temporal scales. Water Resour. Res. 43, W04410. doi:10.1029/2005WR004774.

Homand, F., Shao, J.F., Giraud, A., Auvray, C., Hoxna, D., 2006. Pétrofabrique etpropriétés mécaniques des argilites. C. R. Geoscience 338, 882–891.

Horseman, S.T., Higgo, J.J.W., Alexander, J., Harrington, J.F., 1996. Water, Gas andSolute Movement through Argillaceous Media. OECD/NEA/Clay Club Report CC-96/1. OECD/NEA, Paris, France.

Ingram, G.M., Urai, J.L., Naylor, M.A., 1997. Sealing processes and top sealassessment. In: Molle-Pedersen, P., Koestler, A.G. (Eds.), Hydrocarbon Seals:Importance for Petroleum Exploration and Production, Norwegian PetroleumSociety, Special Publication vol. 7. Elsevier, Amsterdam, pp. 165–174.

Kovacs, L. 1999. Overview Volume Short Term Qualification Program of the BodaClaystone Formation, Final Report, vol. 9 (overview volume). Mecsek & PuramReport, Hungary.

Marschall, P., Croisé, J., Schlickenrieder, L., Boisson, J.Y., Vogel, P., Yamamoto, S.,2004. Synthesis of hydrogeological investigations at Mont Terri site (phases 1 to5). In: Heitzmann, P. (Ed.), Mont Terri Project – Hydrogeological Synthesis,Osmotic Flow, Report of the Federal Office for Water and Geology (Bern,Switzerland). Geology Series, vol. 6, pp. 7–92.

Marschall, P., Horseman, S.T., Gimmi, T., 2005. Characterisation of gas transportproperties of the Opalinus Clay, a potential host rock formation for radioactivewaste disposal. Oil Gas Sci.Technol. 60, 121–139.

Martin, C.D., Lanyon, G.W. 2002. EDZ in Clay Shale: Mont Terri. Mont Terri TechnicalReport TR 2001-01, Mont Terri Consortium.

Mazurek, M., 1999. Evolution of gas and aqueous fluid in low-permeabilityargillaceous rocks during uplift and exhumation of the central Swiss Alps.Appl. Geochem. 15, 211–234.

Mazurek, M., Pearson, F.J., Volckaert, G., Bock, H. 2003. FEPCAT Project: Features,Events and Processes Evaluation Catalogue for Argillaceous Media. OECD/NEAReport, Paris, France.

S104 M. Mazurek et al. / Physics and Chemistry of the Earth 33 (2008) S95–S105

Author's personal copy

Mazurek, M., Hurford, A.J., Leu, W., 2006. Unravelling the multi-stage burial historyof the Swiss Molasse Basin: integration of apatite fission track, vitrinitereflectance and biomarker isomerisation analysis. Basin Res. 18, 27–50.

Mazurek, M., Alt-Epping, P., Bath, A., Gimmi, T., Waber, H.N., 2008. CLAYTRACProject: Natural Tracer Profiles across Argillaceous Formations – Review andSynthesis. Nuclear Energy Agency, OECD, Paris.

Miller, W.M., Alexander, W.R., Chapman, N.A., McKinley, I.G., Smellie, J.A.T., 2000.Geological Disposal of Radioactive Wastes and Natural Analogues. WasteManagement Series, vol. 2. Pergamon, Amsterdam, The Netherlands.

Nagra, 1993. Geologische Grundlagen und Datensatz zur Beurteilung derLangzeitsicherheit des Endlagers für schwach-und mittelaktive Abfälle amStandort Wellenberg. Nagra Technical Report NTB 93-28, Nagra, Wettingen,Switzerland.

Nagra, 2002a. Projekt Opalinuston – Synthese der geowissenschaftlichenUntersuchungsergebnisse. Entsorgungsnachweis für abgebrannteBrennelemente, verglaste hochaktive sowie langlebige mittelaktive Abfälle.Nagra Technical Report NTB 02-03, Nagra, Wettingen, Switzerland.

Nagra, 2002b. Project Opalinus Clay – Safety Report. Demonstration of DisposalFeasibility for Spent Fuel, Vitrified High-level Waste and Long-livedIntermediate-level Waste (Entsorgungsnachweis). Nagra Technical Report NTB02-05, Nagra, Wettingen, Switzerland.

Neuzil, C.E., 1994. How permeable are clays and shales? Water Resour. Res. 30, 145–150.

Palut, J.M., Montarnal, P., Gautschi, A., Tevissen, E., Mouche, E., 2003.Characterisation of HTO diffusion properties by an in situ tracer experiment inOpalinus Clay at Mont Terri. J. Contam. Hydrol. 61, 203–218.

Pearson, F.J., Arcos, D., Bath, A., Boisson, J.Y., Fernández, A.Ma., Gäbler, H.E., Gaucher,E., Gautschi, A., Griffault, L., Hernán, P., Waber, H.N., 2003. Geochemistry ofwater in the Opalinus Clay formation at the Mont Terri laboratory. In: Report ofthe Federal Office for Water and Geology (Bern, Switzerland). Geology Series,vol. 5.

Peyaud, J.B., Barbarand, J., Carter, A., Pagel, M., 2005. Mid-Cretaceous uplift anderosion on the northern margin of the Ligurian Tethys deduced from thermalhistory reconstruction. Int. J. Earth Sci. 94, 462–474.

Pocachard, J., Duding, B., Beaudoing, G., Launay, M., Thoby, D., Mourzagh, D., 1997.Coefficients de diffusion et de permeation d’eau à travers 8 échantillons du siteMeuse – Forage EST 104. Unpublished Andra Report D.RP.3CEA.97.004, Andra,Châtenay-Malabry, France.

Sacchi, E., Michelot, J.L., 2000. Pore water Extraction from Argillaceous Rocks forGeochemical Characterisation. Methods and Interpretation. OECD/NEA Report,Paris, France.

Sacchi, E., Michelot, J.L., Pitsch, H., Lalieux, P., Aranyossy, J.F., 2001. Extraction ofwater and solutes from argillaceous rocks for geochemical characterisation:methods, processes and current understanding. Hydrogeol. J. 9, 17–33.

Tevissen, E., Soler, J.M., Montarnal, P., Gautschi, A., Van Loon, L.R., 2004. Comparisonbetween in situ and laboratory diffusion studies of HTO and halides in OpalinusClay from the Mont Terri. Radiochim. Acta 92, 781–786.

Thury, M., Bossart, P., 1999a. The Mont Terri rock laboratory, a new internationalresearch project in a Mesozoic shale formation, in Switzerland. Eng. Geol. 52,347–359.

Thury, M., Bossart, P., 1999b. Mont Terri Rock Laboratory. Results of theHydrogeological, Geochemical and Geotechnical Experiments Performed in1996 and 1997. Swiss National Hydrological and Geological Survey, GeologicalReport 23.

Trouiller, A., 2006. Le Callovo–Oxfordien du bassin de Paris: du contexte géologiqueà la modélisation de ses propriétés. C. R. Geoscience 338, 815–823.

Van Loon, L.R., Soler, J.M., Bradbury, M.H., 2002. Diffusion of HTO, 36Cl� and 125I� inOpalinus Clay samples from Mont Terri: effect of confining pressure. J. Contam.Hydrol. 61, 73–83.

Van Loon, L.R., Soler, J.M., Jakob, A., Bradbury, M.H., 2003. Effect of confiningpressure on the diffusion of HTO, 36Cl� and 125I� in a layered argillaceous rock(Opalinus Clay): diffusion perpendicular to the fabric. Appl. Geochem. 18, 1653–1662.

Van Loon, L.R., Soler, J.M., Müller, W., Bradbury, M.H., 2004a. Anisotropic diffusionin layered argillaceous rocks: a case study with Opalinus Clay. Environ. Sci.Technol. 38, 5721–5728.

Van Loon, L.R., Wersin, P., Soler, J.M., Eikenberg, J., Gimmi, T., Hernan, P., Dewonck,S., Savoye, S., 2004b. In situ diffusion of HTO, Na-22(+), Cs+ and I- in OpalinusClay at the Mont Terri underground rock laboratory. Radiochim. Acta 92, 757–763.

Van Loon, L.R., Müller, W., Iijima, K., 2005. Activation energies off the self-diffusionof HTO, 22Na+ and 36Cl� in a highly compacted argillaceous rock (Opalinus Clay).Appl. Geochem. 20, 961–972.

Wenk, H.R., Voltolini, M., Mazurek, M., Van Loon, L., Vinsot, A., 2008. Preferredorientations and anisotropy in shales: Callovo–Oxfordian shale (France) andOpalinus Clay (Switzerland). Clays Clay Min. 56, 285–306.

Wersin, P., Van Loon, L.R., Soler, J.M., Yllera, A., Eikenberg, J., Gimmi, T., Hernan, P.,Boisson, J.Y., 2004. Long-term diffusion experiment at Mont Terri: first resultsfrom field and laboratory data. Appl. Clay Sci. 26, 123–135.

Yven, B., Sammartino, S., Géraud, Y., Homand, H., Villiéras, F., 2007. Mineralogy,texture and porosity of Callovo–Oxfordian argillites of the Meuse/Haute–Marneregion (Eastern Paris Basin). Mém. Soc. géol. France 178, 73–90.

M. Mazurek et al. / Physics and Chemistry of the Earth 33 (2008) S95–S105 S105