Swelling experiments on mudstones

Transcription

1 Journal of Rok Mehanis and Geotehnial Engineering. 21, 2 (1): 51 Swelling experiments on mudstones C. L. Zhang, K. Wiezorek, M. L. Xie Repository Safety Researh ivision Company for Reator Safety (GRS) mbh, Braunshweig, 3122, Germany Reeived 21 May 29; reeived in revised form 23 November 29; aepted 15 eember 29 Abstrat: This paper studies the swelling of highly onsolidated mudstones by theoretial onsiderations and laboratory experiments. A key assumption was made that saturated and unemented lays behave as heavily dense olloid without diret ontats among solid partiles. It leads to an important onlusion that the swelling pressure ating on adsorbed interpartile water-films is equivalent to the effetive stress. This so-alled lay-olloid onept is validated by various swelling experiments on two kinds of mudstones, the Callovo-Oxfordian argillite in Frane and the Opalinus lay in Switzerland. In the tests, water adsorption-desorption, swelling pressure and strain were measured on the samples at various sutions and load-ontrolled onditions. Results suggest that: (1) the mudstones an take up great amounts of water from the humid environment, muh more than the water ontent in the natural and saturated states; (2) the swelling pressure inreases with water uptake to high levels of the overburden stresses at the sampling depths of 23 to 5 m, indiating that the adsorbed water-films are apable of arrying the lithostati stress; and (3) the large amount of water uptake auses a signifiant expansion of mudstones even under the lithostati stresses. Key words: lay; mudstone; adsorption; swelling pressure; expansion; sution; stress analysis 1 Introdution Clay formations are being investigated worldwide as host medium for the disposal of radioative waste beause of the favorable properties suh as very slow transport of fluids and high sorption apaity for most radionulides. However, exavation of an underground repository produes fratures and raks around openings, whih may form pathways for fluid transport. uring operation phase of a repository, the ventilation of the underground openings will lead to de- and re-saturation of surrounding rok, depending on hanges in humidity of the ventilating air. The de- and re-saturation may result in swelling and shrinking and even fraturing in the lay. After bakfilling and losing of repository, a gradual sealing proess of the exavation-indued fratures is to be expeted due to the ompression by the rok deformation and the expansion of the fratured lay matrix by taking up water from the saturated far-field. The swelling may play a dominant role in the sealing proess, depending on availability of water. Whereas swelling as a signifiant proess ourring in lays suh as bentonite- oi: 1.372/SP.J Corresponding author. Tel: ; hun-liang zhang@grs.de Supported by the German Federal Ministry of Eonomis and Tehnology (BMWi) (2E1377) based buffer materials is broadly investigated, the swelling behavior of highly-indurated mudstones has been less haraterized and was not well understood. Based on the disjoining pressure onept that the swelling pressure is equivalent to the average disjoining pressure of water interlayers in lay, Horseman et al. [1] proposed that the isotropi effetive stress in a ompated and unemented lay, whih behaves as perfet olloid without diret ontats among solid partiles, was equal to the swelling pressure in the interpartile water-films. This also means that the adsorbed water-films in ompated lays are apable of bearing external loads and arrying the lithostati stress. Obviously, this so-alled lay-olloid model differs from the lassi porous medium model, in whih pore water is generally assumed to be free and physio-hemial interations of water with lay minerals are not expliitly taken into aount. So the appliation of the lassi porous model enounters diffiulties in interpreting some anomalous phenomena observed in mudstones [1]. However, the olloidal onept may be helpful for understanding of strongly-oupled hydro-mehanial proesses in ompated lays and mudstones. Unfortunately, there are almost no diret experimental evidenes for support of the olloidal onept.

2 C. L. Zhang et al. /Journal of Rok Mehanis and Geotehnial Engineering. 21, 2 (1): 51 5 Reently, GRS has arried out various swelling experiments to validate the olloidal onept for mudstones. The Callovo-Oxfordian (COx) argillite in Frane and the Opalinus (OPA) lay in Switzerland have been tested. Both are the potential host roks for the disposal of radioative waste. This paper analyzes the relationship among total stress, swelling pressure and matri sution in ompated lays, and disusses the swelling experiments designed under the layolloid onept. 2 Effetive stress in ompated lays The natural and saturated lay may be regarded as a part of olloidal system [1, 2]. Most of the pores are interonneted and oupied with water and solutes. In highly onsolidated lays, the pore size is mostly so small that a very signifiant portion of the water ontent is strongly bound on mineral surfaes. Only small amounts of free water may be inluded in relatively large pores [3]. In addition to lay minerals, there also exist other mineral partiles and diageneti bonds in natural lays. The properties and responses of the lay olloidal system are signifiantly determined by very omplex interations among water, solutes and lay surfaes. A oneptual model for the inner struture of a ompated lay may be assumed as illustrated in Fig.1. Bond Clay partile Other minerals Free pore water Adsorbed water-film Fig.1 Coneptual model for onsolidated lay. Physio-hemial interations of water with lay minerals ause adsorption of water on the internal and external surfaes of lay partiles, forming eletrostati double-layer. In the narrow spaes between lay partiles in ompated lays, the double-layer is overlapping. This generates loal exess pressures to develop in interpartile water-films when the volume of the lay is kept onstant by enlosing it in a rigid permeable box. This exess pressure is termed as the disjoining pressure [1] when equilibrated with the external bulk water. The disjoining pressure is the sum of a number of ontributions whih stem from eletrostati double-layer interations, Van der Waal s dispersion fores, strutural fores and solvation fores. The loal disjoining pressure,, is related to the ratio of the vapor pressure of water-film, p vf, to that of the bulk water p vp by RT vf ln p (1) v w p vp where R is the gas onstant, T is the absolute temperature, and v w is the partial molar volume of the water-film. The average disjoining pressure over a wavy ross-setion passing through the midplanes between partiles is equivalent to the swelling pressure ating on the walls of the box. For unemented lays whih behave as perfet olloid so that there are no diret ontats between solid partiles, the average disjoining pressure or the swelling pressure,, is equivalent to the onventional isotropi effetive stress eff. Horseman et al. [1] suggested that an isotropi total stress applied to the lay-water system an be expressed by eff pw po pw po (2) where p o is the referene atmospheri pressure, and pw po is the gauge pressure in the absene of osmoti effets. This equation learly indiates that the total loal pressure p fm ating on an adsorbed water-film, is the sum of the disjoining pressure and the pressure p w in the free bulk pore water. The negative measurable gauge pressure is defined as matri sution p [2]: p po pw (3) Substituting Eq.(3) into Eq.(2), the total stress an also be expressed as a funtion of matri sution: eff p p () This equation implies that the matri sution of a saturated lay is simply the differene between the effetive stress (or the swelling pressure) and the total stress. Rodwell et al. [2] illustrated the fore balane ating on a water-film between two lay platelets, as shown in Fig.2. The analogue model (Fig.2()) demonstrates the oupling between the fores. (a) Total stress, (b) Fig.2 Relationships between the total stress, the matri sution and the disjoining pressure on a water-film among two lay platelets [2]. ()

3 6 C. L. Zhang et al. / Journal of Rok Mehanis and Geotehnial Engineering. 21, 2 (1): 51 uring undrained loading on a saturated lay, the water ontent or the thikness of the adsorbed water-films remains nearly onstant, keeping the swelling pressure unhanged. So the matri sution inversely or the bulk pore water pressure proportionally varies with the total stress: p pw 1 (5) uring drained loading at onstant pressure in the external water reservoir, i.e. p w = onst or p = onst, the effetive stress or the swelling pressure in the lay varies proportionally with the total stress: eff 1 (6) p p This implies that the interpartile water-films in ompated lays bear external loads and are apable of arrying the lithostati stress. When p =, the externally applied total stress or the lithostati stress will be fully supported by the adsorbed water-films: eff (7) Appliation of the total stress to a saturated lay ompresses the adsorbed water-films, leading to desorption of water moleules from mineral surfaes. Redution of the film thikness inreases the disjoining pressure in it. If the total stress falls below the effetive stress (or the swelling pressure), for instane, during rapid erosional unloading or destressing by exavation of underground openings or extration of samples from a saturated lay formation, sution exerts an internal tensile stress among all pairs of neighboring mineral partiles against the differene of the internal disjoining pressure and the external boundary load. Maintaining the total stress onstant, the swelling pressure and the sution must hange simultaneously: eff p p 1 () When =, the matri sution of a saturated lay is equal to the average swelling pressure but not zero: p eff (9) Thus, the matri sution and the average swelling pressure of ompated lays may be expressed by the same mathematial form as Eq.(1). Both are related to the ratio of the vapor pressure of the adsorbed water in narrow pores p vm to that of the pure bulk water p vp : p RT ln p vm v w p vp (1) It is to be distinguished here that the disjoining (swelling) pressure is a generalized quantity that inorporates the full range of mehanisms affeting the hemial potential of water in the thin films of a ompated lay, while the matri sution is the reation to the externally-applied boundary stress and the internally-developed exess pressure in the narrow gaps between partiles [1, 2]. Matri sution is atually determined by the intermoleular strain of the pore water. When the matri sution is positive, the adsorbed water-films are in tension. As a sution threshold is exeeded, the water moleules evaporate from the lay and the hydrogen bonds of the interpartile water-films break down, resulting in ollapse of pores. When the matri sution is negative, the water-films are in ompression. Up to now, ompated lays and mudstones have still been treated as lassi porous media, in whih all the pore water is assumed to be free without distinguishing the different effets of adsorbed and free waters. The appliation of the lassi porous model to ompated lays and mudstones finds diffiulties in interpreting some anomalous phenomena, for instane, anomalous pore water pressures observed in field mudstones [1] and in mudstone samples during loading []. In ontrast, the above-mentioned olloidal onept may be suitable for the interpretation of suh phenomena appearing in ompated lays and mudstones. However, this model must first be validated by experiments. 3 Studied mudstones Both the COx argillite and the OPA lay are sedimentary mudstones and over-onsolidated. For laboratory tests, COx ores were extrated from the Meuse-Haute-Marne under- ground researh laboratory in Frane, while OPA ores from the Mont Terri underground rok laboratory in Switzerland. The COx argillite ontains % 5% lay minerals, 2% 3% arbonates and 2% 3% quartz [5, 6], whereas the OPA lay has 5% 76% lay minerals, 6% 2% arbonates and 5% 2% quartz [6]. Both mudstones do not ontain signifiant quantities of expansive lay minerals suh as smetite, 13% 23% illite-smetite in COx argillite [6] and 5% 2% illite-smetite in OPA lay [7]. The basi properties of the mudstones were determined on the samples and are summarized in Table 1. The studied mudstones have similar properties. It is to be pointed out that the samples are more or less desaturated during sampling, storage and preparation.

4 C. L. Zhang et al. /Journal of Rok Mehanis and Geotehnial Engineering. 21, 2 (1): 51 7 Table 1 Basi properties of the studied mudstones. Mudstone Clay ontent (%) Carbonate ontent (%) Quartz ontent (%) Water ontent (%) Grain density (g/m 3 ) ry density (g/m 3 ) Porosity (%) COx OPA Measurements of swelling pressure.1 In volume-onstraint onditions As usual, the onventional method is adopted to measure swelling pressures of the COx and OPA mudstones in volume-onstraint onditions. Figure 3 shows the shemati setup of a swelling ell in whih a disk sample of 1 mm in thikness and 5 mm in diameter is inserted. The onstrained sample is allowed to be wetted from the top and bottom with liquid water or water vapor at the ontrolled humidity. Fore sensor Temperature / air humidity sensors Fig.3 Shemati setup of a swelling ell. esiator Sample 1 Humidity Sample 2 regulator Air pump Figure illustrates the evolution of the swelling pressure measured on a partly-saturated COx sample whih was preloaded to.5 MPa axially. The first wetting was performed by irulating wetted air at a relative humidity of 9%, resulting in an inrease of the swelling pressure to about 1 MPa and then a slight flutuation due to the temperature hange. The following wetting by introduing syntheti formation water into the sample ended when the atmospheri pressure gave a quik rise of the swelling pressure up to 2.5 MPa and a gradual inrease to 3.3 MPa over months. Similar tests made by others on the same mudstones provided lower values below 2 MPa [5, ]. Obviously, these values are not omparable with the expeted ones aording to Eq.(7), i.e. the swelling pressure should be omparable to the overburden stress of about 12 MPa at the COx sampling depth of 9 m [5] and the mean lithostati stress of 6 MPa at the OPA sampling loation [7]. One of the reasons for the measured low values of swelling pressure might be that lay minerals lose to the entering water expanded Fig. Swelling pressure measured on a COx sample in a ell with wetted air and syntheti formation water. so highly that the loal pore spaes were rapidly losed, making more water enter into the onstraint samples more diffiultly and even impossibly. Thus the swelling pressure in the samples ould not homogenously develop..2 In axially-fixed and laterally-unonstrained onditions To minimize effets of rapid losure of the entry pores during wetting and ahieve a homogeneous distribution of entering water within a sample, a new test method has been developed by the authors [9, 1] for determination of swelling pressure of hard lays. Figure 5 illustrates the priniple of the so-alled uniaxial swelling test. A lay sample is axially-fixed and laterally-unonstraint in a ell. Water moleules in vapor an easily aess and move into the narrow spaes deeply, so that rehydration of the lay sample may be better made by irulating wetted air around Fore sensor Temperature / air humidity sensors Humidity regulator Air pump Fig.5 Priniple of uniaxial swelling test. esiator Sample 1 Sample 2

5 C. L. Zhang et al. / Journal of Rok Mehanis and Geotehnial Engineering. 21, 2 (1): 51 the peripheral surfae at the ontrolled relative humidity (RH = p vm / p vp ) or sution ( p ). Aording to Eq.(1), the sution is related to the relative humidity as RT p RT p v p v vm ln ln( RH) w vp w (11) Variation of the air humidity leads to hanges in water ontent (see Setion 6) and thus in thikness of interpartile water-films, whih determines the swelling pressure. A suffiiently high stiffness of a lay sample makes it possible to keep its stability without need of any radial onfinement. Whereas the buildup of swelling pressure in the sample is measured in axial diretion by the reation of the rigid piston, it is also possible to monitor the swelling strain in radial diretion. Figures 6 presents the results of uniaxial swelling pressure measured on COx and OPA samples, respetively. The samples were prepared to a size of mm in diameter and 5 mm in length. The initially saturated samples were more or less desaturated during preparation. They were axially preloaded to 2 MPa and then the axial strain was fixed. =1.5 MPa (a) COx samples inrease in relative humidity leads to a rising axial stress. At 95% relative humidity, the reating stress in axial diretion of the COx samples reahed a maximum of MPa while a peak value of 5.5 MPa was reorded for the OPA sample. Higher swelling pressures an be expeted by elevating the boundary humidity. For instane, wetting another COx sample with water vapor at RH = 1% (or p = ) led to a larger swelling pressure of 11 to 12 MPa, as shown in Fig.7. These maximum values of swelling pressure are almost equal to the overburden stresses of 12 and 6 MPa at the sampling positions of COx and OPA mudstones, respetively. This test result learly onfirms the theoretial stress onept (Eq.(7) or Eq.() at p = ), whih is indeed developed for unemented lays behaving as perfet olloid. It is evident that the olloidal model with the assoiated stress onept is also true for the diageneti mudstones. Beause the studied mudstones do not ontain signifiant quantities of expansive lay minerals suh as smetite (13%-23% in COx argillite [6] and 5% 2% in OPA lay [7]), it seems likely that the swelling fores are developed between the external surfaes of losely-paked platy lay minerals suh as illite (3%-2% in COx argillite [6] and 16%-% in OPA lay [7]) by mehanisms similar to those operating during interlayer swelling [1]. The observed phenomenon of stress relief by drying and stress rising by wetting reflets that the adsorbed interpartile water-films in ompated lays and mudstones support the externally-applied loads and even the lithostati stress. Additionally, the buildup of swelling pressure in the fixed axial diretion without any lateral onfinement may suggest that the pressure ating on the water-films among lay platelets is probably not a salar quantity and should be represented by a seond-rank tensor [1, 2]. = 5.5 MPa = 11 MPa (b) OPA samples Fig.6 Uniaxial swelling pressure measured on COx and OPA samples as a funtion of relative humidity of surrounding air. From Fig.6(b), it an be reognized that drying by irulating air of % relative humidity aused a quik drop of the axial stress tending to zero. The subsequent Fig.7 Response of uniaxial swelling pressure to drying and wetting reorded on a COx sample. It is to be noted that the flutuation of the measured swelling pressure may be aused by the loal hanges in the stiffness of lays during wetting, whih dereases with the hange of water ontent.

6 C. L. Zhang et al. /Journal of Rok Mehanis and Geotehnial Engineering. 21, 2 (1): Measurements of swelling strain In addition to the measurements of swelling pressure, the swelling potential of the mudstones has also been examined by measurements of swelling strain under various onditions. Free swelling tests were performed by wetting unonstrained samples with water vapor, during whih the deformation and water ontent hange were reorded. Figure illustrates the evolution of measured strains on COx and OPA samples at a low relative humidity of RH = 23% (orresponding to the sution p = 2 MPa) for two months and 1% relative humidity (p = ) for eight months, respetively. At the high sution applied externally, the water adsorbed on mineral surfaes evaporates and moves out from the pores, indiated by the redution of the water ontent. The release of the pore water as stress-supporting element results in ollapse of the pore struture and thus a marosopi shrinkage. After reahing equilibrium during drying at RH = 23%, the remaining water ontents in both samples are nearly the same (1.7%), but the shrinkages are different. A larger shrinkage of 1.6% was observed on the OPA sample with more lay ontent, about 5 times that of COx sample with less lay ontent. uring the following wetting phase with water vapor at zero sution around the sample surfae, the high sution potential of p = 2 MPa in the inner pores, whih have been indued by the previous drying, drives water moleules to move into the pores, inreasing the water ontent and hene the thikness of the water-films. The tests showed that the mudstones an take up great amounts of water, over 1% 1%, muh more than that of about 7% in natural and saturated states. The inrease of water ontent yielded a large volume expansion up to % 12%. More water uptake and volume expansion ould be expeted if wetting ontinued to reah equilibrium. This finding onfirms that the matri sution dominated by the adsorption or swelling potential in the saturated mudstones is not zero after unloading (Eq.(9)). Additionally, it is to be pointed out that the swelling of the sedimentary mudstones is pronounedly anisotropi due to bedding planes. A larger swelling strain ours in diretion perpendiular to the bedding plane (Fig.). The anisotropi swelling is so strong that it is able to break down the mudstones along the bedding planes. Figure 9 shows the pitures of the samples after swelling.. Water ontent (%) COx sample (diameter is 5 mm, length is 56 mm) RH=1% w o = 7% in saturated RH=23% Time (day) (a) COx sample 16 OPA sample Water ontent 1 (diameter is 5 mm, 1 length is 56 mm) 12 Volumetri strain w o =7% RH=1% in saturated Axial strain 6 6 RH=23% 2 2 Radial strain Time (day) (b) OPA sample Fig. Free swelling strains measured on COx and OPA samples at various humidities (sample axis perpendiular to bedding plane). Water ontent (%) (a) COx sample Water ontent Volumetri strain Radial strain Axial strain (b) OPA sample Fig.9 Pitures of mudstones after free swelling. On another COx sample, axial swelling was measured in a swelling ell by introduing syntheti formation water into the top and bottom at atmospheri pressure. The swelling urve is depited in Fig.1. As the sample ontated with the water, a rapid expansion took plae and then gradually inreased with time up to 7% over months. The swelling urve obtained during wetting with liquid water is quite similar to those observed during wetting with vapor (Fig.). The high swelling apaity of the mudstones was also onfirmed by other tests under onfining stresses. Figure 11 shows a triaxial swelling test on a COx sample under the originally lithostati stress state of 1 = Swelling strain (%) Swelling strain (%)

7 5 C. L. Zhang et al. / Journal of Rok Mehanis and Geotehnial Engineering. 21, 2 (1): 51 Fig.1 Axial free swelling strain measured on a COx sample with syntheti formation water. Fig.11 Swelling strains of a COx sample under high triaxial onfining stresses. 15 MPa and 2 = 3 = 12 MPa. The in-situ water pressure p w =.5 MPa was applied by injeting syntheti water to both ends of the sample. The strain urves show that a ontinuous expansion rather than ompression ourred in axial diretion even though the sample was subjeted to the high effetive ompressive stress of 11.5 MPa in this diretion. The gradual expansion suggests that the swelling potential (pressure) stored in the mudstone is larger than the differene between the externally applied total stress and the water pressure. The supporting apaity of the bound pore water, however, an be redued by elevating the temperature. At high temperatures, the adsorbed water-films beome thin due to the thermal-indued desorption of water. In drained onditions, the thermal-mobilized pore water is expelled from the sample under onfining stresses, ausing ompation of pores. 6 Water retention The water retention funtion is a relation of matri sution with water ontent. Matri sution has two omponents, the first is assoiated with apillary fores ating gas-water interfaes, and the seond with the adsorption fores of water mineral surfaes [1, 2]. The Temperature (C) adsorption potential is an extremely important omponent of the total potential in ompated lays and mudstones. The water retention urves of the studied mudstones were determined by measurements of water adsorptiondesorption at various ontrolled relative humidities or sutions around samples. Figure 12 summarizes the relationship between the applied sution and the water ontent ahieved after equilibrium on COx and OPA samples. By effet of high externally applied sution, the adsorbed pore water desorbs from the mineral surfaes and esapes from the pores. Conversely, when the external sution is lower than that of the inner pores, water migrates into the lay-water system. From Fig.12, it an be found that, at zero sution, the mudstones an take up large amounts of water up to 1% for COx sample and 1% for OPA sample, whih were obtained over a period of months when equilibrium was not yet ahieved. The amounts of water uptake are signifiantly more than 7% around in the natural state, suggesting that if not all, most of the water ontent in the natural mudstones is bound on the surfaes of lay minerals. This reflets that the lay-water system is really a heavily dense olloid. It is also interesting to point out that the intersetion of the natural water ontent of about 7% with the water retention urves gives the sution of about 6 MPa for OPA samples and about 15 MPa for COx samples, whih are very lose to the measured swelling pressures (Figs.6, 7) and also to the lithostati stresses at sampling depths. This finding reveals again that the matri sution of the saturated mudstones is not zero but equal to the swelling pressure (Eq.(9)). In fat, the water retention funtion of mudstones is governed by the same adsorption-desorption phenomena that determine the swelling pressure. Water ontent (%) OPA: wetting path OPA: drying path Natural water ontent w 7% COx: wetting path COx: drying path Sution (MPa) Fig.12 Water retention urves of COx and OPA samples. Finally, it is to be noted that the adsorbed water an beome free under ertain onditions. The pore water in a saturated lay an be extrated by squeezing samples in laboratory [7]. In the underground

8 C. L. Zhang et al. /Journal of Rok Mehanis and Geotehnial Engineering. 21, 2 (1): laboratories at Bure in Frane and at Mont Terri in Switzerland, a small amount of water outflow from boreholes was observed, whih may result from onsolidation due to loally inreasing stresses after borehole drilling. Another example is the observation of thermal-indued water flow during a heating experiment in the Opalinus lay [11]. 7 Conlusions Various swelling experiments were performed on the Callovo-Oxfordian and Opalinus mudstones. The tests inluded the measurements of water adsorptiondesorption, swelling pressure and swelling strain on the samples in various sution-ontrolled and loadontrolled onditions, from whih the following onlusions an be drawn: (1) The natural mudstones exhibit high adsorption potentials, under whih a great amount of water an be taken up from the humid environment, muh more than the water ontent in the natural and saturated states. (2) The adsorbed water-films in the mudstones are apable of bearing external loads and arrying the lithostati stress, indiated by the measurements of suh high swelling pressures on the samples. (3) The large amount of water uptake results in signifiant expansion of mudstones even under high lithostati stresses and in-situ water pressures. () All the test observations validate the olloid model and the assoiated stress onept, i.e. the swelling pressure ating on interpartile water-films is equal to the effetive stress in ompated lays and mudstones. Wide appliations of the heavily dense lay-olloid model and the assoiated stress onept may be expeted for improvement of understanding and modeling of strongly oupled hydro-mehanial proesses in ompated lays and mudstones, for instane, water and gas transport, rheologial deformation, self-sealing and healing of fratures in mudstones, et.. Aknowledgements The authors gratefully aknowledge the funding by the German Federal Ministry of Eonomis and Tehnology (BMWi). The support from ANRA by providing samples is also gratefully aknowledged. Referenes [1] Horseman S T, Higgo J J W, Alexander J, et al. Water, gas and solute movement through argillaeous media. Paris: Nulear Energy Ageny, [2] Rodwell W R, Harris A W, Horseman S T, et al. Gas migration and two-phase flow through engineered and geologial barriers for a deep repository for radioative waste. Paris: Nulear Energy Ageny, [3] Mazurek M, Gautshi A, Marshall P, et al. Transferability of geosientifi information from various soures (study sites, underground rok laboratories, natural analogues) to support safety ases for radioative waste repositories in argillaeous formations. Physis and Chemistry of the Earth, 2, 33 (Supp.1): S95 S15. [] Zhang C L, Rothfuhs T, Su K, et al. Experimental study of the thermo-hydro-mehanial behavior of indurated lays. Physis and Chemistry of the Earth, 27, 32 ( 1): [5] Lebon P, Ghoreyhi M. Frenh underground researh laboratory of Meuse/Haute-Marne: THM aspets of argillite formation. In: EUROCK2 Symposium. Aahen: [s. n.], 2: [6] Tournassat C, Gauher E C, Fattahi M, et al. On the mobility and potential retention of iodine in the Callovian-Oxfordian formation. Physis and Chemistry of the Earth, 27, 32 ( 1): [7] Pearson F J, Aros, Bath A, et al. Mont Terri Projet geohemistry of water in the Opalinus lay formation at the Mont Terri rok laboratory. [S. l.]: Federal Offie for Water and Geology (FOWG), 23. [] Thury M, Bossart P. Mont Terri rok laboratory, results of the hydrogeologial, geohemial and geotehnial experiments performed in 1996 and Bern: Swiss National Hydrologial and Geologial Survey, [9] Zhang C L, Rothfuhs T, Jokwer N, et al. Results of laboratory investigations on lays. In: Proeedings of the International Conferene on Radioative Waste isposal iste 2. Berlin: [s. n.]: 2: [1] Zhang C L, Rothfuhs T. Moisture effets on argillaeous roks. In: Shanz T ed. Proeedings of the 2nd International Conferene of Mehanis of Unsaturated Soils, Springer Proeedings in Physis 112. Berlin-Heidelberg: Springer-Verlag, 27: [11] Zhang C L, Rothfuhs T, Jokwer N, et al. Thermal effets on the Opalinus lay a joint heating experiment of ANRA and GRS at the Mont Terri URL (HE- Projet). Braunshweig: Gesellshaft für Anlagen-und Reaktorsiherheit (GRS), 27.