Detecting the near-surface redox front in crystalline bedrock using fracture mineral distribution, geochemistry and U-series disequilibrium

Applied Geochemistry 24 (2009) 1023–1039

Contents lists available at ScienceDirect

Applied Geochemistry journal homepage: www.elsevier.com/locate/apgeochem

Detecting the near-surface redox front in crystalline bedrock using fracture mineral distribution, geochemistry and U-series disequilibrium Henrik Drake a,*, Eva-Lena Tullborg b, Angus B. MacKenzie c a

Department of Earth Sciences, University of Gothenburg, Box 460, SE-405 30 Göteborg, Sweden Terralogica AB, Box 4140, SE-443 14 Gråbo, Sweden c Scottish Universities Environmental Research Centre, Rankine Avenue, East Kilbride, Glasgow G75 0QF, Scotland, United Kingdom b

a r t i c l e

i n f o

Article history: Received 4 June 2008 Accepted 9 March 2009 Available online 17 March 2009 Editorial handling by M. Gascoyne

a b s t r a c t Oxidizing conditions normally prevail in surface waters and near-surface groundwaters, but there is usually a change to reducing conditions in groundwater at greater depth. Dissolved O2 originally present is consumed through biogenic and inorganic reactions along the flow paths. Fracture minerals participate in these reactions and the fracture mineralogy and geochemistry can be used to trace the redox front. An important task in the safety assessment of a potential repository for the disposal of nuclear waste in crystalline bedrock, at an approximate depth of 500 m in Sweden, is to demonstrate that reducing conditions can be maintained for a long period of time. Oxygen may damage the Cu canisters that host nuclear waste; additionally, in the event of a canister failure, oxidizing conditions may increase the mobility of some radionuclides. The present study of the near-surface redox front is based on mineralogical (redox-sensitive minerals), geochemical (redox-sensitive elements) and U-series disequilibrium investigations of mineral coatings along open fractures. The fractures have been sampled along drill cores from closely spaced, 100 m deep boreholes, which were drilled during the site investigation work in the Laxemar area, south-eastern Sweden, carried out by the Swedish Nuclear Fuel and Waste Management Co. (SKB). The distribution of the redox-sensitive minerals pyrite and goethite in open fractures shows that the redox front (switch from mainly goethite to mainly pyrite in the fractures) generally occurs at about 15–20 m depth. Calcite leaching by recharging water is indicated in the upper 20–30 m and positive Ceanomalies suggest oxidation of Ce down to 20 m depth. The U-series radionuclides show disequilibrium in most of the samples, indicating mobility of U during the last 1 Ma. In the upper 20 m, U is mainly removed (due to oxidation) or has experienced complex removal and/or deposition. At depths of 35– 55 m, both deposition and removal of U are indicated. Below 55 m, recent deposition of U is generally indicated which suggests removal of U near surface (oxidation) and deposition of U below the redox front. Scattered goethite occurrences below the general redox front (down to ca 80 m) and signs of U removal at 35–55 m mostly correlate with sections of high transmissivity (and/or high fracture frequencies). This shows that highly transmissive fractures are generally required to allow oxygenated groundwaters at depth greater than ca 30 m. Removal of U (oxidation) below 55 m within the last 300 ka is not observed. Although penetration of glacial waters to great depths has been confirmed in the study area, their potential O2 load seems to have been reduced near the surface. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction It is important to demonstrate that reducing conditions can be maintained for a long period of time at the depths (500 m) of a high-level radioactive waste (HLW) repository (e.g. Gascoyne, 1999; Guimerà et al., 1999; Puigdomenech et al., 2001). This is because (1) O2 may harm the Cu canisters in which the used fuel is contained and (2) in the case of canister leakage, oxidizing conditions may increase the mobility of several radionuclides in the HLW, especially U. Oxidizing conditions in surface waters and

* Corresponding author. Fax: +46 31 7861986. E-mail address: [email protected] (H. Drake). 0883-2927/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.apgeochem.2009.03.004

near-surface groundwaters generally change to reducing conditions in deeper groundwaters, by consumption of the O2 by organic and inorganic reactions along the flow paths. These reactions often involve fracture minerals (Rivas-Perez et al., 2003), and the fracture coating mineralogy and geochemistry can be used to trace the redox front (cf. Landström et al., 2001; Tullborg et al., 2008) to depths where the reducing capacity in the fractures is significantly decreased. For example, complete oxidation of pyrite and formation of goethite marks the position of the redox front within the fracture-flow system. Below the redox front there is a transition zone where oxidizing and reducing conditions have varied over time (cf. Perez del Villar et al., 2002; Arcos et al., 2008). The exact position of this zone is most easily observed in sedimentary

1024

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

rocks, (e.g. sandstones) which often show distinct colour differences. However, in granitic environments such a front is usually less well developed, but has been possible to detect in some cases (e.g. Akagawa et al., 2006). Such oxidation-related colour differences are not common in the crystalline basement of Sweden and potential oxidation effects are usually limited to water-conducting fractures (or fracture zones). In these cases, the only way to recognise a redox front is by using microscopy to evaluate mineralogical effects due to oxidation, and to examine the behaviour of redox-sensitive elements, in fracture coatings and in the adjacent wall-rock. Information about the recent redox front position in the shallow bedrock and its stability over time is important for the understanding of the maximum penetration depth of descending oxygenated waters. A scenario which may introduce oxygenated water to great depth in the bedrock is glacial melt water intrusion triggered by high hydraulic heads beneath a continental ice sheet (cf. Boulton et al., 2001). Such water can be assumed to contain more dissolved O2 than the present recharge water, mainly due to dissolution of air bubbles trapped in ice into glacial melt water (cf. Glynn et al., 1999) and elevated pressure beneath the melting ice (e.g. Guimerà et al., 1999). Furthermore, the organic buffer may be severely reduced below the ice (Puigdomenech et al., 2001). The extent and potential for dissolved oxygen to migrate to great depth in such a glacial scenario has been debated and different model approaches give very different results (cf. Glynn and Voss, 1999; Glynn et al., 1999; Gascoyne, 1999; Guimerà et al., 1999). The present study of the redox front location has been carried out in the Laxemar area on the east coast of Sweden (Fig. 1), an area currently investigated as a potential site for a deep HLW repository by the Swedish Nuclear Fuel and Waste Management Co. (SKB). The site investigations have resulted in detailed information about the present hydrogeology and hydrochemistry (e.g. SKB, 2006) as well as about the evolution of the bedrock geology (e.g. Wahlgren et al., 2006), glacial and post-glacial deposits (e.g. Rudmark et al., 2005) and paleohydrogeology (fracture fillings and wall-rock alter-

ation) (e.g. Drake and Tullborg, 2007). The present study has been carried out on samples of fracture coatings from 11 closely spaced, 100 m deep boreholes from two different drill sites located about 20–27 m above sea level. The main focus has been to detect the redox front using mineralogical (distribution of redox-sensitive minerals), geochemical (redox-sensitive elements, e.g. Ce, Mn and Fe) and U-series isotope investigations of mineral coatings along open fractures (and to a lesser degree the migration of the redox front into the wall-rock adjacent to near-surface fractures). Detailed analyzes of Fe(III)/Fe(II) and Fe isotopes on fracture coating Fe-oxides, are presently being carried out on samples from the same boreholes and will be reported in Dideriksen et al. (in press). Redox-sensitive Fe-minerals such as pyrite, Fe-oxides and Feoxyhydroxides (e.g. goethite) can be used to detect the redox front (cf. Landström et al., 1989, 2001), as shown by detailed mapping of the distribution of these minerals in redox front studies at Äspö close to Laxemar, at other sites in southern Sweden (Eliasson et al., 1989; Tullborg, 1997), and in Spain (Perez del Villar et al., 2002). In these studies, the redox front position was recognised at various depths (down to 100 m) depending on local topography (affecting recharge/discharge), as well as fracture frequencies and orientations. Oxidation of pyrite, precipitation of goethite, and associated dissolution of calcite were observed near the surface. In areas with a thick soil cover, signs of recent oxidative alteration may not be seen in the bedrock because the groundwater may be reduced by the time it enters the bedrock. Actual measurements of Eh (potentiometric) boreholes can be technically very difficult and are usually completely unreliable in the near-surface region. Moreover, sampling of groundwater close to the surface usually results in mixing of surface waters and shallow groundwaters and the interpretation of the redox-sensitive element concentrations in the groundwater may therefore be jeopardised (SKB, 2006). Instead, redox conditions can be evaluated with better depth control using redox-sensitive elements like Ce and U in minerals deposited along the fracture-flow paths. Since oxidized Ce(IV) is less mobile than Ce(III), lake and stream waters

Fig. 1. (a) Geological map of the SKB site investigation area at Laxemar and Simpevarp with the drill sites indicated. (b–c) detailed maps of the drill sites KLX11 and KLX09 with projections of the boreholes shown. (d) Map of southern Sweden with the location of the study area indicated.

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

usually show negative Ce-anomalies (cf. Dia et al., 2000; Rönnback et al., 2008). Both positive and negative Ce-anomalies can be found in minerals close to the surface, e.g. due to in situ oxidation of Ce on mineral surfaces (Landström and Tullborg, 1995; Bau, 1999; Takahashi et al., 2000) and due to precipitation from water with negative Ce-anomalies (Landström and Tullborg, 1989; Vaniman and Chipera, 1996; Milodowski et al., 2005), respectively. Although the position of the redox front can be inferred from the distribution of redox-sensitive minerals and elements, this information provides no time constraints to whether the redox front corresponds to the present conditions or not. However, measurements of radionuclides in the U-series decay chain (238U–234U–230Th), can be used to derive time constraints for oxidizing and reducing conditions in the bedrock fractures, as shown at the adjacent Äspö site (Tullborg et al., 2003), but also in other areas (e.g. Smellie et al., 1986; Gascoyne and Cramer, 1987; MacKenzie et al., 1992; Griffault et al., 1993; Suksi and Rasilainen, 2002; Min et al., 2005). Disequilibria in the decay chain 238 U–234U–230Th indicate redistribution (removal or deposition) of U within the last 1 Ma and after such perturbation, the system will gradually return towards a state of secular equilibrium in which the activity ratios (AR) equal unity (Osmond and Ivanovich, 1992; Scott et al., 1992; Gascoyne et al., 2002). In the Laxemar area, radioactive disequilibrium in the fracture coatings requires that fracture fillings have experienced open system conditions in the last 1 Ma, even though most fractures probably formed much earlier. Oxidation of U(IV) to the more mobile U(VI) usually results in leaching of bulk U at shallow depths (e.g. Landström et al., 2001) and since both 238U and 234U are leached by oxidative fluids and are much more mobile than Th, recent removal of U (oxidizing conditions) results in 234U/238U AR  1 and 230Th/234U AR > 1 in the solid phase. Under reducing conditions at greater depths, 234U, is more likely present in the oxidized state of U(VI), relative to 238 U, and is leached more easily than 238U (e.g. Suksi and Rasilainen, 2002). In addition, 234U is generated by alpha decay and thus occupies a radiation-damaged site that is more susceptible to leaching than the site of 238U (e.g. Gascoyne et al., 2002). The degree of disequilibrium tends to decrease with depth due to more stagnant conditions (e.g. Suksi, 2001; Tullborg et al., 2003). The redox zone position varies due to short and long-term variations in groundwater conditions, e.g. annual variations in groundwater chemistry and land uplift, respectively. Therefore, deposition and removal of U from fluids with different redox conditions may have occurred at a specific depth over very long time periods. The fractured bedrock constitutes a very heterogeneous system with large variations in the amount of fracture coating material and fracture transmissivities. Furthermore, flow paths have probably varied over time and it is not expected that different elements and minerals should respond uniformly to changes in redox conditions (e.g. Ce-anomalies cannot automatically be assumed to correspond to the U mobilisation or deposition). Furthermore, the results from the different methods may represent signatures from different time periods each represented by a specific depth of O2 penetration. Short periods with weakly oxidizing conditions may not be detected when analyzing fracture coatings, in contrast to signs of long time periods of stable conditions, either reducing or oxidizing.

2. Setting 2.1. Geology The bedrock in the Laxemar area is dominated by ca 1.80 Ga granitoids ranging from diorite to granite in composition (Wahlgren et al., 2006). The two dominant rock types in the area are ‘‘Ävrö granite” (generally granite to quartz monzodiorite) and

1025

‘‘quartz monzodiorite”. They have overlapping compositions and are mainly distinguished by texture (Ävrö granite is commonly porphyritic and quartz monzodiorite is commonly equigranular). The rocks in the Laxemar area are generally well preserved and more or less undeformed, although low-grade metamorphism and some local shear zones occur. At about 1.45–1.44 Ga two granites intruded near the Laxemar area (Åhäll, 2001). The Sveconorwegian orogeny affected western Sweden at 1.1–0.9 Ga (e.g. Bingen et al., 2006), succeeded by development of a foreland basin with sediments covering eastern Sweden (Larson et al., 1999). After erosion of these sediments in the late Proterozoic the sub-Cambrian peneplain was created (Lidmar-Bergström, 1996) and the presently exposed bedrock surface largely corresponds to this erosion surface. Cambrian to Early Silurian transgression and marine sedimentation resulted in deposition of Paleozoic sequences and remnants of these are currently mainly found off shore (Lindström et al., 1991; Koistinen et al., 2001). Various thermal indicators, including apatite fission track analyzes show that the sedimentary cover (mainly Caledonide erosion products) had a thickness of several kilometres during the Late Paleozoic (Larson et al., 1999; Cederbom et al., 2000; Cederbom, 2001). Subsequent erosion, mainly in the Permian to Triassic, reduced the thickness of the sedimentary cover considerably (Cederbom, 2001; Söderlund et al., 2005). Transgressions in the Cretaceous changed the conditions to marine. Finally, the sedimentary cover was eroded away during uplift in the Tertiary and the present sub-Cambrian surface was re-exposed (Lidmar-Bergström, 1996). Repeated deformation (e.g. early mylonites and later brittle fractures) have affected the area (Munier and Talbot, 1993; Drake and Tullborg, 2007). Most of the fractures and shear zones are older than 1.4 Ga (Wahlgren et al., 2006; Drake et al., 2007) and are usually associated with intense wall-rock alteration, i.e. red-staining; alteration of biotite, plagioclase and magnetite and formation of chlorite, albite, adularia and hematite (Drake et al., 2008). However, the rock affected by this alteration has largely the same reducing capacity as unaltered rock (Drake et al., 2008). Reactivation of these fracture sets and formation of new fractures occurred at later events (e.g. related to the Sveconorwegian and Caledonian orogenies), at successively lower temperatures (Drake et al., 2007). Stable isotope signatures of several generations of fracture calcite indicate that many of the fractures formed during the Paleozoic have probably been intermittently water conducting ever since (Tullborg et al., 1999, 2008; Drake and Tullborg, 2007). However, the detailed configuration of the flow paths has changed through time and past groundwaters have left precipitates from which the different groundwater regimes can now be traced (Tullborg et al., 1999). Earlier investigations in the Laxemar area have shown that precipitation of calcite, pyrite, clay minerals and goethite (near surface) might have occurred as late as in the Quaternary (Tullborg, 1997; Drake and Tullborg, 2007). Recent precipitation is expected to be dominantly localized to shallow depths where the groundwaters have short residence times, and possibly faster flow rates (Drake and Tullborg, 2007). During the Quaternary, several glaciations influenced the Laxemar area. The ice sculptured the bedrock surface and removed weathered surface layers. After the latest glaciation, most of the Laxemar area was only covered by sea water for a very short time and areas at higher elevation, including the investigated drill sites, have been above sea level for at least 10.7 ka (Söderbäck, 2008). Outcrops are common in these areas. The appearance of striation directions coinciding with the latest ice progression implies that the last continental ice has scraped the outcrops (Landström et al., 2001; Rudmark et al., 2005). The present surface is thus influenced by mechanical erosion and low-temperature chemical weathering during the Weichselian glaciation and the last postglacial period. The dominance of outcropping rocks in areas of

1026

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

higher elevations allows rapid transport of surface water to areas of lower elevation, either by direct runoff or by fracture infiltration. Due to aggressive humic acids, chemical weathering is more active below the thin soils present (Landström et al., 2001). Because the Laxemar site is situated close to the Baltic Sea, hydrogeological and hydrochemical changes during a glacial cycle involve possible injection of glacial water into the bedrock fractures (due to high hydraulic heads beneath the ice) followed by later marine-brackish conditions (Baltic Sea influence and density turnover), and subsequent interplay between meteoric recharge and glacial rebound (SKB, 2006). A combination of meteoric, brackish sea water, glacial melt water and old saline waters is found with varying depths at the site. 2.2. The drill sites and the boreholes Core was examined from eleven 100 to 130 m deep boreholes from two drill sites (KLX09 and KLX11) (Fig. 1). Both sites are characterized by exposed bedrock or less than 0.5 m of Quaternary deposits (Rudmark et al., 2005). The bedrock surface was generally exposed at the starting point of each borehole, with the exceptions of about 30–40 cm deep soil pockets at some locations. Boreholes were drilled using a triple-tube technique (e.g. Ask, 2007a) which provides high-quality core with good preservation of minerals in open fractures. Drill-core diameter is about 50 mm. For both of the KLX09 and KLX11 drill sites, boreholes C through F were drilled with a dip of 60° from horizontal and emplaced in a box-like geometry around the centrally located, vertical borehole B (Fig. 1). At drill site KLX09, an additional borehole (G) was drilled dipping 60° to the east starting ca 50 m east of KLX09D. Drill site KLX09 is located closer to the Baltic Sea at a slightly lower elevation (borehole collars at 19.63–23.75 m; Fig. 4) relative to drill site KLX11 (22.65–27.27 m). Core from KLX09 boreholes is dominated by Ävrö granite, with occasional 5 m sections of mafic rock in some boreholes (Forssberg et al., 2005; Ask, 2007a,b). Core from KLX11 boreholes is dominated by quartz monzodiorite, with some exceptions, e.g. Ävrö granite in KLX11D: 0–50 m (Ask, 2006; Cronquist et al., 2006).

3. Methods Hundred and nine drill core samples were collected from the surface down to 115 m vertical depth. Sampling focused on open fractures with large amounts of coatings that provided enough material for complete mineralogical, geochemical, and U-series analyzes. Special attention was given to coatings containing pyrite and goethite. The majority of samples were collected from the upper 30 m of the boreholes and the uppermost occurrences of pyrite and calcite were sampled in most of the boreholes. In contrast, the samples from below 30 m are generally from shear zones. Consequently, the samples from below 30 m contain higher amounts of fracture coatings and relatively higher amounts of chlorite and clay minerals. Samples are designated according to their depth from the bedrock surface and not to their core length. Fracture surfaces from 70 samples were removed by sawing and investigated using a Hitachi S-3400N scanning electron microscope equipped with an Oxford Instruments energy dispersive system (SEM-EDS) in order to identify minerals and their weathering features. After SEM-EDS characterization, the fracture coatings were scraped off using a knife. The scraped off material was then sieved and grains smaller than 0.125 mm were analyzed for 238U, 234U and 230 Th (22 samples), X-ray diffraction (XRD; 15) and chemistry (35), if the amount was sufficient. Sample weights were 0.125–5 g for XRD and chemical analyzes and 1–5 g for USD analyzes (up to 20 g). Since the U content is usually low in the fracture coatings,

U-series analyzes required use of the entire filling. Therefore, isotope analyses include U both from stable minerals that are probably not reactive with migrating solutions (e.g. zircon, apatite and titanite) and easily accessible U adsorbed and/or precipitated on fracture mineral surfaces (dominantly goethite and clay minerals). Samples analyzed for USD are generally from more transmissive sections with higher amounts of fracture coating than the samples analyzed with ICP-MS only. Thin sections were prepared from the wall-rock and the fracture coatings of four near-surface fractures in order to investigate the weathering and alteration in the wall-rock and the fracture coatings, using optical microscopy and SEM-EDS. SEM-EDS analyzes were carried at the Department of Earth Sciences, University of Gothenburg. The instrument was calibrated twice every hour using a Co standard linked to simple oxide and mineral standards. The acceleration voltage was 20 kV and the specimen current was about 1 nA. X-ray spectrometric corrections were made by an on-line Oxford INCA software. The detection limit was 0.1 oxide% and Fe(II) and Fe(III) were not distinguished. Quantitative chemical analyzes were not achieved for fracture surface samples due to the uneven surface of the samples. Minerals from these samples were identified by visual inspection and interpretation of X-ray spectra or element ratios. The XRD analyzes were carried out by the Swedish Geological Survey, Uppsala. The samples were separated into a coarse fraction and a fine fraction. The coarse fraction was ground in an agate mortar under acetone and the powder was randomly orientated in the sample holder. The fine fraction was dispersed in distilled water, filtered and oriented according to Drever (1973) and measurements were carried out in three steps; (1) dried sample (2) after saturation with ethylene glycol for 2 h and (3) after heating to 400 °C for 2 h. A Siemens D5000 theta-theta-diffractometer with Cu-radiation, Cu Ka and graphite monochromator at 40 kV and 40 mA. Coarse fraction scans were run from 2° to 65° (2-theta) without orientation of the samples being used. Fine fraction scans were run from 2° to 35° (2-theta) with a step size of 0.02° (2-theta) and counting time 1 s/step. The analyzes were performed with a fixed divergence slit, a 2 mm antiscatter slit and a 0.1 mm receiving slit. Bruker AXS DIFFRACPLUS 2.2 and EVA software were used for raw data evaluation. The minerals were identified using the PDF database (version 1994) and clay mineral data from Brindley and Brown (1984) and Jasmund and Lagaly (1993). The fracture coatings were analyzed for major, minor and trace element contents by Analytica AB, Sweden. 0.125 g samples were fused with 0.375 g LiBO2 and dissolved in dilute HNO3. LOI was not carried out due to small sample volumes. Concentrations of the elements were determined by ICP-AES and ICP-QMS. Analyzes were carried out according to EPA methods (modified) 200.7 (ICPAES) and 200.8 (ICP-QMS). The accuracy was about 5–10%. The analyzes do not distinguish between Fe(II) and Fe(III). Analysis of 238U, 234U and 230Th were made accordingly. The samples were heated in a furnace at 450 °C to remove any organic matter, followed by acid digestion using aqua regia and HF to achieve complete sample dissolution. Following digestion, samples were dissolved in 9 M HCl and extraction of U and Th was effected by anion exchange separation. At the start of the procedure, a known activity of a yield tracer solution of 232U in transient equilibrium with 228Th was added to the sample. Thin sources for alpha spectroscopy were prepared by electrodeposition onto stainless steel planchettes after which alpha spectroscopy was performed using an Ametec surface barrier detector system (MacKenzie et al., 1986, 1992; Smellie et al., 1986). d13C, d18O and 87Sr/86Sr in calcite and d34S in pyrite from the samples of the present study are presented in Drake and Tullborg (2007) and are described briefly in Section 5. Information on fracture mineral occurrences and fracture frequencies in the boreholes was obtained from the mapping data

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

stored in the SKB database (Sicada). The mapping crew paid extra attention to redox-sensitive minerals, e.g. pyrite, goethite and hematite, as well as calcite. The uppermost occurrences of the redox-sensitive minerals in each borehole were examined in detail and the mapping data was updated with these observations. 4. Results 4.1. Fracture frequencies The total fracture frequency is relatively constant with depth and is elevated in the KLX09-boreholes (generally 7–11 fractures/ m, of which 2–3 are open) compared to in the KLX11-boreholes (generally 4–7 in total, open: 1–2). However, the frequency of open fractures is greatest (1–2 fractures/m higher) in the upper 10 m and at 70–80 m depth. The latter is due to intersection of minor shear zones and sections of crushed rock in some of the boreholes. The total number of water-conducting fractures in each borehole is generally between 37 and 43, except for in KLX11F: 24 water-con-

1027

ducting fractures and KLX11D: 49 water-conducting fractures (Sokolnicki and Väisäsvaara, 2006; Väisäsvaara et al., 2006; Sokolnicki and Kristiansson, 2007). 4.2. Mineralogy The fracture coatings consist of a mixture of minerals, usually including chlorite, clay minerals (e.g. illite Fig. 2i, smectite or mixed-layer clay; corrensite or smectite/illite), calcite, quartz, Kfeldspar, plagioclase, hematite, goethite and pyrite. Quartz, K-feldspar and plagioclase are mainly wall-rock fragments included in the fracture coating. Accessory fracture minerals include REE-carbonate (probably bastnäsite), barite, zeolites (harmotome and laumontite), chalcopyrite, epidote and trace amounts of a U-silicate (possibly coffinite). Near the surface, Ce- and Mn-oxide (Fig. 2j) are identified in a few samples. SEM investigations show that chlorite and clay minerals as well as calcite, hematite, goethite, pyrite and REE-carbonate are generally the outermost minerals of the fracture coating, and therefore constitute the surface in contact

Fig. 2. Microphotographs (reflective light), (a–d) and back-scattered SEM-images (e–l). (a–d) Wall-rock pyrite increasingly replaced by goethite with decreased distance from the fracture (a) 0.4 mm from the fracture, (b) 4 mm, (c) 10 mm, and (d) 17 mm, sample KLX11F: 2.0 m – thin section). (e–g) Altered pyrite crystals on fracture surfaces; (e) KLX09B: 10.5 m, (f) KLX09D: 17.3 m and (g) KLX11F: 17.7 m. (h) A fairly fresh pyrite crystal (right) and a slightly more altered pyrite crystal (left) on a fracture surface at 19.9 m depth (KLX09D). (i) Illite on a fracture surface (sample KLX09D: 6.0 m). (j) Mn-oxide or oxyhydroxide on a fracture surface (KLX09D: 3.0 m). (k) Fe-oxyhydroxide (‘‘Fe”) penetrating chlorite (‘‘chl”, which has replaced biotite – ‘‘bt”) close to and at the fracture surface (left side of image, sample KLX11F: 5.5 m – thin section). (l) Fresh biotite (1), replaced by chlorite (2, 3 and 4) close to the fracture surface (EDS analyzes of 1–4 are shown in Table 1), in sample KLX09D: 0.15 m (thin section).

1028

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

with the recharging water. The mineralogy of the samples analyzed for chemistry and U-series isotopes is listed in Table 2. Separation of hematite and goethite (including other FeOOH minerals) grown on fracture surfaces is problematic using SEMEDS because the crystals are very fine-grained and the surface is irregular, which compromises EDS analysis. In this study goethite is identified by its typical colour (in drill core samples and in thin sections) and by XRD (only diffuse peaks due to low amounts). Because most of the fracture filling hematite in the Laxemar area is interpreted to have been formed under different conditions than those at present, based on its paragenesis of relatively high-temperature minerals (Drake and Tullborg, 2005) and its Fe-isotope values and degree of crystallinity (Dideriksen et al., 2007, in press), interpretation of recent redox conditions based on Fe(III)-minerals is mainly based on goethite. Drill core mapping indicates that minerals in open fractures are dominated by calcite (present in 76% of the fractures), chlorite (76%), clay minerals (28%), hematite (23%), pyrite (17%), epidote (5.5%) and goethite (3.5%). Mapping data and SEM-EDS investigations reveal that the distribution of the redox-sensitive minerals goethite and pyrite, as well as calcite, varies with depth. The distribution of pyrite and goethite for each borehole is shown in Fig. 3 (vs. vertical depth) and Fig. 4 (vs. elevation). Goethite is preferen-

Fig. 4. The uppermost occurrences of goethite and pyrite in KLX09B-G and KLX11BF relative to elevation. Mapping data is from the SKB database Sicada and from this study.

tially found in the upper 10–25 m of the boreholes although depths vary between the boreholes and it may be absent in some boreholes (KLX09G). Occasionally goethite is found as deep as ca 80 m, where it is associated with highly transmissive fractures (flow rates of >1  10 6 m2/s) and/or highly fractured intervals with more than 20 fractures/m. Goethite occurrences below 80 m are rare. In contrast, pyrite is absent in the fractures shallower than 10–25 m (Fig. 3). In most boreholes the near-surface shift from mainly goethite to mainly pyrite in the fractures is distinct, although boreholes KLX09B, KLX09F and KLX11D have alternating goethite and pyrite bearing fractures between 10 and 25 m. Both goethite and pyrite are absent between 10 and 25 m in KLX09E which complicates locating the redox front in this borehole, although chalcopyrite is present in a fracture at 22 m. Pyrite crystals in fractures above 20 m depth are usually altered and some-

Table 1 SEM-EDS mineral analyses.

1 2 3 4

MgO

Al2O3

SiO2

K2O

CaO

TiO2

MnO

FeO

Total

12.6 12.4 8.6 8.6

14.5 14.2 10.9 12.0

36.1 34.5 21.9 26.1

9.3 3.3 1.2 3.8

<0.1 0.9 0.9 0.7

1.6 1.4 0.4 1.1

0.5 0.3 0.2 0.3

16.7 18.5 37.2 30.6

91.3 85.5 81.3 83.2

Fresh and altered (chloritized) biotite; sample 09D: 0.15 m (Fig. 2l) FeO represents the total iron content.

Fig. 3. Occurrence of goethite and pyrite in open fractures in boreholes KLX09B through KLX09G and KLX11B through KLX11F, shown as vertical depth from the ground surface (each symbol equals one fracture). Shaded areas are intervals with highly transmissive fractures of >1  10 6 m2/s and/or highly fractured sections (>20 fractures/m). Mapping data is from the SKB database Sicada and from this study.

Fig. 5. Percent of open fractures containing calcite vs. depth. The depth intervals are 10 m (except in the upper 20 m: 5 m intervals). The trend of 80–95% calcite-bearing fractures continuous down to 100 m (not included in the graph). No mapping was carried out on the uppermost 2.8 m of the KLX11 boreholes and these sections are therefore absent from the graph. Mapping data is from the SKB database Sicada and from this study.

Table 2 Chemistry, U-series (errors in 1r) and mineralogy of bulk fracture coatings. Mineralogy is determined by XRD and/or SEM-EDS and is presented in order of abundance (brackets indicate minor amounts). Chemical composition of Ävrö granite and quartz monzodiorite (average values and the range) is also given (values from Drake et al., 2006). 09B

09B

09C

09D

09D

09D

09D

09D

09D

09E

09E

09E

09F

09F

09F

09F

09F

09F

09G

Depth

4.7

8.7

72.3

3.0

7.8

11.0

71.2

75.9

89.6

0.7

7.5

9.8

10.9

50.4

58.7

68.2

69.9

115.0

35.0

51.5 14.9 3.76 12.0 4.52 4.06 0.15 1.95 0.31 0.77 93.9 1610 14 14.4 34.7 31.2 <6 22.5 7.70 <2 14.4 17.2 259 10.8 5.62 508 1.48 9.75 4.72 105 1.93 22.6 607 275 47.3 109 13.4 50.8 8.21 1.69 5.68 0.71 4.10 0.81 2.22 0.32 2.18 0.33

31.6 11.3 19.8 8.51 4.39 2.21 0.24 0.64 0.21 0.47 79.4 1410 13.7 7.83 30.0 17.0 20.3 15.8 4.26 <2 7.43 <10 169 8.51 4.2 488 0.69 5.21 43.0 132 4.15 18.3 150 148 120 183 19.8 74.6 8.85 1.86 5.67 0.57 3.09 0.60 1.56 0.21 1.53 0.24

SiO2 Al2O3 CaO Fe2O3 K2O MgO MnO Na2O P2 O 5 TiO2 Total Ba Be Co Cr Cs Cu Ga Hf Mo Nb Ni Rb Sc Sn Sr Ta Th U V W Y Zn Zr La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

% % % % % % % % % % % mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

43.3 12.7 13.1 5.37 3.60 2.99 0.26 2.48 0.22 0.67 84.7 20300 7.91 <6 24.9 4.22 14.2 17 5.71 <2 9.18 20.4 139 8.45 3.67 419 1.06 4.4 4.09 90.5 1.8 17.5 142 189 108 152 16.1 58.7 8.47 2.11 5.7 0.62 3.52 0.64 1.93 0.27 1.84 0.28

45.6 16.0 5.16 11.2 8.02 4.51 0.17 0.44 0.24 0.87 92.2 589 46.2 22.5 94.6 16.4 6.17 29.2 4.91 <2 9.34 35.1 331 24.9 6.18 114 0.99 9.32 5.92 85.9 12.3 23.8 276 168 65.5 121 12.9 50.4 9.8 2.42 7.22 0.94 5.20 0.99 2.55 0.38 2.60 0.37

46.2 18.2 5.14 7.51 2.95 3.82 0.84 0.38 0.08 0.14 85.3 5920 19.3 14.0 13.9 8.56 69.7 25.1 3.22 3.43 3.47 <10 78.1 3.52 13 120 0.28 2.3 2.60 80.8 2.96 7.73 229 107 32.9 94.3 6.8 24.3 3.49 0.92 2.28 0.26 1.48 0.29 0.87 0.13 0.98 0.19

56.6 16.3 2.55 7.28 6.81 2.38 0.10 1.31 0.38 0.81 94.5 1970 24.1 11.8 42.9 15.6 317 18.2 8.84 <2 16.1 20 210 12.1 25.2 312 1.74 14.8 8.11 141 5.74 39.6 811 294 263 423 39.4 146 18.8 3.80 12.0 1.34 7.09 1.35 3.81 0.53 3.63 0.54

44.8 13.9 2.58 12.6 4.74 6.88 0.15 0.32 0.08 0.17 86.2 1770 34.9 10.4 17.8 5.93 55.5 20.3 2.58 <2 4.16 <10 124 2.34 5.66 135 0.59 2.63 5.63 167 1.2 7.24 214 77.4 47.1 116 7.45 25.5 3.29 0.66 2.27 0.26 1.45 0.28 0.74 <0.1 0.83 0.16

37.4 16.8 6.09 14.8 3.87 7.95 0.34 0.07 0.22 1.21 88.7 212 62.9 39.5 223 41.3 32.9 48.3 5.21 <2 17 45.2 430 19.6 6.93 64.2 1.09 8.66 25.3 125 12.8 75.5 237 169 59.0 177 27.9 126 24.5 5.66 18.2 2.51 14.2 2.73 7.90 1.13 8.22 1.26

49.5 21.6 1.66 8.77 8.41 2.98 0.12 0.68 0.09 0.35 94.2 1150 45.6 10.1 61.9 55.1 65.1 28.4 3.81 <2 8.62 17.4 990 15.7 10.7 67.8 0.72 11.3 8.76 152 12.8 18.5 121 104 15.1 30.2 3.91 19.3 5.42 1.39 5.1 0.66 3.79 0.73 1.86 0.26 2.00 0.29

56.6 8.95 10.2 6.63 2.10 3.29 0.20 1.84 0.11 0.41 90.3 407 22.3 17.9 90.0 13.9 35 14.1 2.91 <2 5.46 39 148 14.4 4.51 267 0.38 10.9 3.58 76.4 2.39 18.7 212 105 70.9 138 14.6 48.9 7.97 1.57 5.69 0.78 4.28 0.75 1.82 0.24 1.60 0.24

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

46.9 13.8 2.66 8.42 3.79 3.59 6.27 2.70 0.45 0.79 89.4 6090 35.1 107 114 10.7 458 18.8 6.73 11.6 11.7 36.9 174 15.7 8.24 675 1.11 8.31 9.59 141 38.9 31.6 286 271 148 341 28.4 94.6 14.0 3.40 10.0 1.11 5.75 1.05 3.08 0.45 3.21 0.51

50.8 14.8 1.85 9.53 7.18 7.1 0.19 0.71 0.34 0.82 93.3 1770 20.4 33.4 40.3 17.4 60.8 23.4 5.79 <2 14.8 29.2 283 12.0 3.56 203 1.25 6.66 5.24 138 5.52 24.0 360 222 76.6 150 16.3 56.8 8.82 1.93 6.0 0.77 4.35 0.76 2.13 0.35 2.27 0.35

53.5 17.7 2.02 5.38 4.96 6.08 0.16 1.77 0.19 0.49 92.2 2060 26.0 <6 22.2 20.9 54.6 26.3 4.85 <2 11.4 14.6 177 8.81 9.49 205 1.07 20.7 15.0 104 3.91 28.3 376 155 345 434 44.7 147 15.3 2.66 9.18 0.89 4.82 0.93 2.52 0.35 2.31 0.33

11.6 3.40 40.5 3.82 0.895 1.67 0.51 0.41 0.04 0.07 62.9 351 15.0 <10 <20 1.11 <10 6.63 4.01 <4 2.53 <20 28.9 <2 2.42 192 0.59 1.29 3.07 22.4 1.55 5.05 196 82.4 173 142 9.9 30.3 2.34 0.47 1.27 <0.2 0.67 0.14 0.37 <0.2 0.49 0.09

48.7 14.9 7.88 8.09 5.03 2.48 0.11 1.56 0.15 0.34 89.2 11400 11.6 <6 15.6 18.9 10.3 21.3 4.59 <2 9.13 <10 212 4.27 4.64 213 0.92 6.79 3.13 55.2 6.01 10.7 92.7 134 83.4 117 11.3 37.6 4.5 1.02 2.79 0.37 1.9 0.36 1.14 0.19 1.22 0.22

57.7 16.9 2.26 5.47 4.75 4.22 0.19 2.62 0.19 0.47 94.8 937 41.5 9.86 23.8 17.8 <6 30.5 6.28 <2 9.32 13.2 250 7.33 4.98 706 1.08 9.18 13.1 64.5 3.02 18.4 158 213 75.9 121 13.0 45.2 6.67 1.27 4.54 0.60 3.55 0.69 1.94 0.29 2.03 0.30

39.1 16.8 4.31 16.3 5.50 7.04 0.20 0.82 0.28 1.10 91.4 409 19.5 26.0 247 25.6 <6 31.4 4.39 2.59 10 48.1 413 34.3 6 115 0.60 7 29.0 186 11.4 22.8 333 151 121 157 17.8 68.1 10.5 2.19 7.55 0.84 4.31 0.82 2.02 0.28 2.08 0.31

238

U

234

U

Bq/kg ± Bq/kg

117 2 155

76 1 88

75 2 87

36 1 42

n.a. n.a. n.a.

n.a. n.a. n.a.

321 1 396

119 2 128

53 1 60

76 1 79

n.a. n.a. n.a.

60 1 67

n.a. n.a. n.a.

n.a. n.a. n.a.

39 1 48

180 2 239

460 n.a. 2 n.a. 677 n.a. (continued on next

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

Borehole

592 6 964 page) 1029

Borehole

09B

09B

09C

09D

09D

09D

09D

09D

09D

09E

09E

09E

09F

09F

09F

09F

09F

09G

Depth

4.7

8.7

72.3

3.0

7.8

11.0

71.2

75.9

89.6

0.7

7.5

9.8

10.9

50.4 58.7

68.2

69.9

115.0

35.0

3 210 10 68 5 1.32 0.04 1.36 0.07 1.79 0.07

1 96 5 53 3 1.15 0.02 1.09 0.06 1.26 0.07

2 86 6 56 5 1.17 0.03 0.99 0.08 1.15 0.10

1 85 3 19 1 1.18 0.03 2.01 0.08 2.36 0.06

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

2 293 13 53 4 1.23 0.01 0.74 0.03 0.91 0.05

2 123 5 61 3 1.07 0.02 0.96 0.04 1.03 0.06

1 62 2 57 2 1.13 0.02 1.03 0.03 1.17 0.05

1 96 9 61 7 1.03 0.02 1.21 0.12 1.26 0.11

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

1 78 3 45 2 1.12 0.03 1.16 0.04 1.30 0.06

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

1 45 1 31 1 1.21 0.04 0.95 0.04 1.15 0.05

3 201 5 59 2 1.32 0.02 0.84 0.02 1.12 0.04

3 656 21 45 4 1.47 0.01 0.97 0.03 1.43 0.04

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

8 794 10 36 1 1.63 0.02 0.82 0.01 1.34 0.02

cm,il, hm/go, kf,qz, (us, zn)

il,cc, kf,ch,cc,qz,il,pl, la,ch,pl, ha hm,apo(mlc, cc,kf,co, go*) apo,qz,an,mt, mn,hm/ go,ep,rc, (ha)

cc,il, cm,ch, hm,qz, kf,rc, (ba, ha)

kf,ch, ch,qz, cm,il, cc,il, hm,qz kf,pl, hm

qz,il, kf,pl, cc, ch, (hm)

ch,il, Il,ch, qz,(py, hm/go, hm) apa,qz, ka,kf

il,ch, cm,al, hm,rc, ba,ep, ha

cc, ch, cm

qz,kf, cc,qz, kf,pl, pl,ch, ch,mlcs * (sm),il,hm,go il,cc, mlcs (sm/il)

ch,il, cc, pl, kf,qz, hm

ch, cm, il,qz, hm, qz

230

Th

232

Th

234

U/238U

± Bq/kg ± Bq/kg ±

± 230

Th/234U

± 230

Th/238U

± Mineralogy

cc,cm,il,kf,mn, cc,il, cm, hm/go, ch,qz,kf,rc, (ba,zn,qz,rc, hm us)

09F

11B

11B

11C

11D

11D

11D

11D

11D

11D

11E

11E

11E

11E

11F

11F

11F

11F

11F

Depth

2.4

12.1

11.6

0.6

1.5

25.3

54.5A

54.5B

91.0

1.4

3.7

16.9

36.7

1.9

8.2

13.2

45.4

60.8

43.9 15.0 1.17 9.72 3.61 1.59 0.06 1.26 0.73 0.52 77.6 706 1.86 14.5 48.4 4.56 27 13.6 4.71 3.09 11.0 19.9 117 12 2.75 257 1.50 26.8 9.96 90.3 3.77 48.3 114 155 348

50.5 15.7 7.32 9.01 4.08 3.7 0.24 1.44 0.30 0.89 93.2 1840 6.72 17.1 29.1 5.42 43.8 20.4 3.28 <2 15.3 16.7 119 11.2 4.06 232 1.10 8.96 3.53 121 4.06 27.2 195 113 69.1

52.4 16.9 3.88 9.12 6.4 3.75 0.31 1.58 0.10 0.41 94.8 1150 3.27 22.2 23.5 1.94 10.8 19.0 4.53 <2 7.01 23.3 189 10.4 2.49 344 0.56 2.26 2.73 110 6.45 16.5 331 169 185

47.0 11.8 2.49 11.5 4.0 9.44 0.42 0.82 0.42 1.26 89.2 409 1.62 19.7 53.6 16.9 21.0 19.6 12.2 <3 16.2 17.4 221 16.6 3.61 119 1.16 3.81 2.92 153 63.9 33.7 206 454 164

60.0 15.5 3.58 6.37 4.26 2.66 0.12 2.83 0.30 1.16 96.8 1100 2.39 11.8 43.2 2.86 24.1 17.2 4.46 <2 14.4 16.3 146 14.8 3.96 525 1.40 7.15 4.20 111 2.43 28.7 93 140 57.6

57.6 11.7 3.2 9.08 5.78 4.89 0.25 0.65 0.22 0.73 94.1 841 20.4 22.3 94.9 6.9 8.03 18.1 12.6 <2 9.19 18.8 115 33.4 3.33 117 1.09 7.06 3.55 145 9.88 104 451 463 245

49.4 18.8 6.95 9.68 5.63 2.19 0.09 0.90 0.14 0.59 94.4 485 14.1 9.33 27.8 63.0 23.7 38.3 3.09 <2 7.3 <10 380 25.1 5.87 2700 0.64 4.53 3.35 206 5.36 32.1 101 99.9 264

46.6 15.7 3.35 14.3 6.05 5.23 0.19 1.36 0.41 0.95 94.1 717 6.97 33.4 76.0 36.8 18.6 28.9 14.0 <2 11.8 30.9 232 25.5 7.44 403 0.82 16.1 6.48 184 10.4 34.1 262 524 88.7

43.6 14.1 8.47 12.3 4.64 3.61 0.14 2.00 0.24 0.82 89.9 646 14.1 23.4 34.2 15.0 156 20.8 6.81 <2 13.2 19.4 206 26.5 6.33 118 1.30 10.6 13.6 156 12 50.9 4700 207 35.7

40.2 15.4 2.21 10.2 1.81 2.49 0.27 1.31 0.78 1.21 75.9 492 2.19 26.8 110 4.65 70.4 13.9 9.74 3.26 17.0 30 110 14.2 6.71 230 1.59 15.2 9.36 136 3.87 45.7 118 320 91.7

40.2 15.9 1.5 23.1 3.43 4.25 0.36 0.86 0.13 0.45 90.2 1100 5.39 65.8 26.4 3.13 58.8 38.2 2.99 <2 6.29 33.8 128 44.1 6.06 93.6 0.58 4.45 1.39 206 4.07 51.5 584 92.5 45.6

32.6 14.1 7.15 19.1 1.96 9.06 0.34 0.50 1.24 4.21 90.3 250 3.05 38.8 28.0 2.21 14.8 50.1 55.4 2.62 59.5 26 69.1 20.1 7.6 536 5.28 263 23.2 121 4.32 120 524 1860 766

35.0 11.8 9.1 15.2 0.64 13.1 0.27 0.09 0.01 0.04 85.3 170 6.61 25.2 15.7 1.59 15.9 41.8 1.41 <2 1.13 30.1 15.9 <1 2.08 1100 0.25 1.85 0.77 103 2.63 8.55 292 10.1 64.8

55.1 14 2.52 8.83 5.62 4.59 0.23 2.15 0.42 0.87 94.3 688 2.03 15.5 39.2 3.79 74.9 20.4 6.99 <2 10.9 16 150 17.1 9.26 152 0.86 8.97 7.24 111 3.59 22.1 134 233 77.6

46.0 11 2.24 12.5 2.34 9.98 0.24 1.01 0.15 0.50 86.0 569 2.35 15.1 32.7 6.92 58.2 18.2 3.08 <2 10.2 <10 70.4 8.36 2.82 117 0.76 15.1 5.97 117 7.99 25.1 576 93.9 507

52.5 14.6 2.64 8.72 7.8 4.83 0.16 0.47 0.06 0.42 92.2 388 5.94 9.73 86.2 14.5 113 22.9 1.43 <2 8.98 17 218 22.5 5.93 284 0.49 4.3 3.64 197 7.48 18 290 36.5 131

39.1 11.8 1.68 13.8 1.91 18.2 0.20 0.34 0.03 0.09 87.2 114 1.14 26.3 17.6 0.67 31.7 17.5 0.858 <2 2.29 12.9 32.6 6.05 2.37 106 0.28 1.53 1.13 90.3 1.18 4.59 389 16.7 9.62

26.0 10.8 15.7 18.6 2.31 5.07 0.27 0.62 0.12 0.53 80.0 293 13.9 24.5 42.2 19.2 48.8 20.4 5.23 <2 8.16 21.9 106 37.9 8.28 89 0.70 10.2 6.12 250 15.4 76.6 981 175 94.7

SiO2 Al2O3 CaO Fe2O3 K2O MgO MnO Na2O P2O5 TiO2 Total Ba Be Co Cr Cs Cu Ga Hf Mo Nb Ni Rb Sc Sn Sr Ta Th U V W Y Zn Zr La

% % % % % % % % % % % mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg

cc,cm, il,hm, (py,sp, cp) H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

Borehole

1030

Table 2 (continued)

mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg

711 61.4 201 27.2 3.70 17.2 2.25 12.1 2.06 5.37 0.74 4.47 0.56

132 13.2 47.6 7.44 2.04 5.87 0.78 4.87 0.96 2.83 0.44 3.05 0.44

275 18.7 56.2 5.98 1.47 4.13 0.57 3.10 0.60 1.74 0.22 1.67 0.25

297 27.9 95.5 11.7 2.17 8.53 1.12 6.01 1.19 3.44 0.48 3.17 0.48

126 13.2 47.3 7.86 1.78 6.42 0.85 5.23 1.11 3.13 0.42 3.09 0.42

425 45.7 172 29.2 6.39 29.6 3.89 21.3 4.20 10.8 1.29 7.71 1.06

508 58.9 212 25.6 4.17 13.0 1.24 6.33 1.16 3.18 0.54 4.05 0.70

174 20.2 73.3 11.8 2.54 9.46 1.21 7.12 1.38 3.93 0.60 4.29 0.69

66.6 9.15 44.9 13.0 3.46 14.7 1.79 10.2 1.79 4.93 0.65 4.35 0.57

267 23.0 86.4 15.6 3.45 11.3 1.77 10.9 1.95 5.68 0.78 5.76 0.78

93.3 13.2 56.9 12.7 2.82 12.4 2.01 12.0 2.44 7.07 1.16 9.37 1.40

1540 150 465 57.8 4.27 37.2 4.51 24.2 4.82 13.4 1.94 12.9 1.94

62.6 4.95 15.5 1.93 0.41 1.82 0.24 1.34 0.27 0.77 <0.1 0.72 0.12

147 13.4 44.5 6.59 1.52 5.39 0.70 4.15 0.83 2.44 0.34 2.42 0.37

860 73.6 212 19.7 1.90 9.2 0.94 4.59 0.84 2.47 0.32 2.19 0.32

332 29.1 91 12.6 2.15 7.51 0.88 4.76 0.81 2.16 0.31 2.08 0.37

16.5 1.74 6.44 1.22 0.26 1.09 0.15 0.76 0.17 0.43 <0.1 0.39 0.07

156 18.0 69.9 15.0 3.82 16.0 2.54 14.5 2.62 7.21 0.94 5.79 0.76

238

Bq/kg ± Bq/kg ± Bq/kg ± Bq/kg ±

126 2 136 2 152 3 130 2 1.08 0.02 1.11 0.03 1.21 0.04

38 1 41 1 70 2 40 1 1.08 0.03 1.68 0.05 1.84 0.06

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

71 1 74 1 80 2 48 2 1.04 0.02 1.08 0.03 1.13 0.04

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

43 1 51 1 76 5 63 4 1.21 0.03 1.48 0.1 1.77 0.09

248 2 296 2 324 8 93 4 1.19 0.01 1.09 0.03 1.31 0.03

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

23 1 23 1 37 3 48 3 1.02 0.03 1.57 0.12 1.61 0.13

292 3 281 3 447 34 1115 67 0.96 0.01 1.59 0.12 1.53 0.09

10 1 12 1 28 1 15 1 1.25 0.06 2.23 0.09 2.80 0.15

108 1 111 1 107 9 43 5 1.03 0.01 0.96 0.08 0.99 0.09

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.

qz,kf, pl, ch, mlc (sm/il),rc, go*

ch,cm, il,hm/go, pr, kf,qz, cc

cm,il, hm, go,kf

il,qz, ch,cm,hm/ go,kf, (w, apa)

ch,cm,il, ch,cm, ch,kf, ch,kf,al, qz,kf, kf,qz, al, ep, qz, ep, hm/ hm,rc,il cm, hm qz,il, mlcs (sm/ go, pl il),hm, (ka,go,ba)

ch,cm,il, cc, py,hm, cpy

ch,cm, ka,qz, hem/ go,kf,pl

ch,qz, kf,pl, mlcs (sm/ il), (ka), go*

ch,kf, ch, al,cc,rc, cm, (ru, cc go*)

ch, cm, go

il,ch, cm, kf, rc,hm/ go,al, ce,qz

qz,pl, kf,cc, ch,kf, co,qz, ch,cc, rc,ch, mlc pl,cc, (go*) mlc(sm,il), qz, (v, sm),go* kf,hm, go*

U

234

U

230

Th

232

Th

234

U/238U

± 230

Th/234U

± 230

Th/238U

± Mineralogy

SiO2 Al2O3 CaO Fe2O3 K2O MgO MnO Na2O P2O5 TiO2 Total Ba Be Co Cr Cs Cu Ga Hf Mo Nb

% % % % % % % % % % % mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg

Ävrö Granite Average

Ävrö granite Range

Quartz monzodiorite Average

64.9 16.0 3.26 4.43 3.87 1.68 0.07 3.98 0.25 0.62 99.0 1370 2.61 6.85 31.9 2.51 29.1 33.6 4.84 <2 9.37

57.6–69.7 14.4–18.3 1.34–5.39 2.94–6.44 2.57–4.61 0.92–2.79 0.046–0.131 3.35–4.76 0.135–0.417 0.36–0.93 96.8–100.6 924–2190 1.56–3.99 3–14.1 13.4–77.8 2.11–2.92 6.61–203 7.21–124 2.56–7.77 <2–6.67 5.14–17.4

56.6 16.4 5.64 8.15 3.10 3.44 0.14 3.38 0.38 1.06 98.3 1040 2.24 16.0 86.0 1.96 38.6 38.0 5.00 <2 10.4

Quartz monzodiorite Range 50.7–60.7 15.1–17.4 4.03–7.23 6.28–10 1.65–3.87 2.11–4.89 0.106–0.195 2.99–3.76 0.264–0.487 0.859–1.34 96.2–100.9 700–1310 1.79–2.65 10–26.2 <10–178 0.993–2.64 12.2–121 1–154 1.92–9.09 <2–3.45 8.02–13.9 (continued on next page)

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

1031

1032

Table 2 (continued)

mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg

Ävrö granite Range

Quartz monzodiorite Average

Quartz monzodiorite Range

19.6 94.7 6.83 4.09 934 0.96 8.83 3.45 61.0 0.72 19.6 77.7 222 41.8 92.1 10.7 40.2 5.79 1.07 3.90 0.61 2.96 0.59 1.48 0.29 1.63 0.25

<10–40.8 52.7–145 2.74–14 1.29–20.4 422–1530 0.33–1.64 0.42–23.1 1.3–7.93 29.7–96.2 <0.4–3.07 9.34–35.6 41.9–204 147–335 18.6–61.1 36.6–142 4.2–16.5 15.7–62.4 1.82–9.3 0.13–1.98 0.85–7.04 0.17–1.1 0.86–5.35 0.09–1.06 0.16–2.77 0.12–0.48 0.61–3.11 0.09–0.47

24.3 84.1 18.5 1.95 734 0.83 5.90 2.22 136 0.66 31.1 94.8 233 37.2 82.8 8.65 39.6 5.83 1.45 4.96 0.73 4.36 0.86 1.91 0.33 2.24 0.31

5.69–45.2 39.5–131 12.5–24.5 1–3.06 528–1150 0.37–1.19 2.06–9.87 0.7–4 81.9–200 <0.4–1.77 25.2–36 59.7–146 78.3–394 29.2–46.1 59.9–104 1–12.9 30.4–49.5 3.5–8.1 0.493–2.03 0.62–6.53 0.1–1.05 2.41–5.7 0.4–1.22 0.1–3.06 0.1–0.54 0.83–3.46 0.08–0.47

Abbreviations: al = albite, an = antigorite, apa = apatite, apo = apophyllite, ba = barite, bt = biotite, cc = calcite, ce = Ce-oxide, ch = chlorite, cm = clay minerals, co = corrensite, cp = chalcopyrite, ep = epidote, go = goethite, ha = harmotome, hm = hematite, il = illite, ka = kaolinite, kf = K-feldspar, la = laumontite, mlc = mixed-layer clay, mn = Mn-oxide, mt = magnetite, pr = prehnite, qz = quartz, rc = REE-carbonate, ru = rutile, sm = smectite, sp = sphalerite, us = U-silicate (possibly coffinite), v = vermiculite, zn = Zn-rich mineral, possibly Zn-oxyhydroxide, w = W-rich mineral, probably wolframite, s = swelling clay,  = very diffuse XRD peak, n.a. = not analyzed.

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

Ni Rb Sc Sn Sr Ta Th U V W Y Zn Zr La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Ävrö Granite Average

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

times partially replaced by goethite and clay minerals (Fig. 2e–h). This alteration decreases in intensity with increased depth and pyrite below 20 m is generally fresh. Pyrite present in the host rock is also partially to completely replaced by goethite with increasing intensity over a ca 20-mm-thick zone adjacent to near-surface fractures (Fig. 2a–d). Similarly, wall-rock biotite is replaced by secondary minerals; e.g. chlorite and goethite, which are richer in Fe and OH but depleted in Mg, Al, Si and K compared to biotite (Fig. 2k–l, with EDS analyzes in Table 1). Penetration of oxygenated fluids into the wall-rock adjacent to near-surface fractures is evidenced by red-coloured micro-fractures filled with goethite, chlorite and clay minerals. Traces of oxidizing water usually reach <2 cm into the wall-rock, in most cases along grain boundaries and the 0 0 1-planes of biotite crystals. The percentage of open fractures coated with calcite is shown in Fig. 5. Calcite coats 80–95% of the fractures below ca 20–30 m depth but is much less abundant in fractures above 20–30 m. Only 10–20% of open fractures contain calcite in the uppermost 10 m. The zone of calcite depletion extends deeper in KLX11 boreholes (to 30–35 m) compared to KLX09 boreholes (to 20–25 m). Textural evidence of calcite dissolution has been observed in samples from the upper 25 m. 4.3. Chemistry Most of the analyzed fracture coatings (Table 2) have higher concentrations of e.g. Fe, K, Mg, Mn, Be, Cs and Rb than the wallrock (i.e. Ävrö granite and quartz monzodiorite from outcrops and other boreholes in the Laxemar area, Drake et al., 2006). Iron, Mg and Mn are mainly associated with the fracture minerals chlorite, illite and mixed-layer clay. Excess Fe is also associated with hematite, goethite, pyrite and epidote. Manganese is also associated with Mn-oxide near the surface. Calcium and Sr are mainly related to calcite and wall-rock fragments of plagioclase but also to epidote. Furthermore, the abundant mixed-layer clay corrensite may contain up to 3 wt.% CaO (Drake et al., 2006). Potassium is mainly associated with K-feldspar, illite and clay minerals. Rubidium and Cs are mainly associated with mixed-layer clay and illite but also with K-feldspar. Barium is mainly associated with K-feldspar, harmotome and barite (Drake et al., 2006). Sodium is mainly related to albite but also to mixed-layer clay. Uranium is related to U-rich silicate and wall-rock derived apatite, zircon and titanite

1033

but may also be incorporated in amorphous phases in fracture coatings and in altered wall-rock (cf. Drake et al., 2008). Rare earth elements are mainly associated to REE-rich carbonate, calcite and wall-rock derived titanite and zircon and rarely to Ce-oxide. Fifty percent of the samples have higher U contents than the wall-rock and a couple of samples, rich in barite and/or harmotome, are significantly enriched in Ba. Most bulk fracture samples have lower contents of Si, Na and Sr than the wall-rock. Other elements in the bulk fracture coatings have concentrations within the range observed in wall-rocks. Most element concentrations in bulk fracture coatings do not show obvious trends with depth (including the redox-sensitive Fe). Exceptions are Ca and U which are depleted near surface (and U enriched at great depths), and Mn and the light rare earth elements (LREE, Fig. 7a), which show elevated concentrations in a number of fracture coating samples near the surface (<20 m), compared to the deeper samples (>20 m). Furthermore, Cs concentrations are elevated at depths greater than 50 m and Ba, Co, Cu, P and Ti are partly enriched near the surface. U and Th concentrations are both usually 3–10 ppm (Table 2) and U concentrations of above 10 ppm are only found in samples below 11 m depth. Barium, Co, Cu and P concentrations are mainly within the values of the wall-rock but a minority of the samples shows highly elevated concentrations of these elements (also shown for Ti in one sample) and these are, as a rule, from the upper 20 m. A small number of samples (mostly below 30 m) have much higher Ca concentrations than in the wall-rock, while most fracture coating samples above 10 m have Ca contents that are lower than or similar to the wall-rock, in accordance with the apparent calcite dissolution near the surface, although CaO is related to other phases as well, which makes the depletion in CaO less distinct than the calcite distribution with depth (Fig. 6a). Manganese concentrations in the bulk fracture coatings are highly variable near the surface and there is a slight decrease with increasing depth. The most striking feature is the very high Mn contents found in two fracture coating samples containing Mn-oxide or -oxyhydroxide and Mnrich clay minerals/chlorite above 10 m depth (Fig. 6b), indicating the presence of Mn(IV). Most of the fracture coating samples have much higher Cs concentrations than in the wall-rock and the highest concentrations are found in samples below 50 m depth (Fig. 6c), mainly in clay mineral- (e.g. mixed-layer clay of illite/smectitetype) and chlorite-rich samples. LREE are within the range of the wall-rock for most samples. However, a couple of near-surface

Fig. 6. Concentrations vs. depth for (a) CaO, (b) MnO (note the broken x-axis), (c) Cs, for bulk fracture coating samples from boreholes KLX09B-G and KLX11B-F.

1034

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

Fig. 7. (a) La/Yb vs. depth, (b) Ce-anomalies (CeCN/CeCN*) vs. depth (CeCN* = (LaCN)1/2(PrCN)1/2) for bulk fracture coating samples from KLX09B-G and KLX11B-F. CN = Chondrite normalized, using chondrite data from Evansen et al. (1978).

4.4. U-series Results of the U decay series analyzes are presented in Table 2 and Figs. 7–10 (for comparison, 1 Bq/kg is equivalent to a U concentration of 0.0810 mg/kg). Comparison of the U and Th values shows reasonable agreement between the isotope dilution alpha spectroscopy and ICP results, suggesting that sample heterogeneity is negligible compared with other uncertainties (cf. MacKenzie et al., 1986). Uranium activities are generally lower in samples from the upper 15 m than in samples from greater depths (Fig. 8). The samples with 238U activities above 240 Bq/kg from depths greater than 35 m are from highly transmissive fractures. 234 U/238U- and 230Th/234U activity ratios (AR) deviate from radioactive secular equilibrium values of unity in most of the samples. The general trends for 234U/238U AR (Fig. 9a) and 230Th/234U AR (Fig. 9b) with depth are: 230 Th/234U AR decreases with depth; >1 above 20 m (median: 1.21, range is 0.96–2.01) and <1 but close to unity at 55–80 m (median: 0.96) and close to unity below 80 m (median 1.06, note that N = 2). At 35–55 m the AR varies widely (three samples with AR: 0.82, 1.48 and 2.23).  Slight excess of 234U (234U/238U AR > 1) is measured at most depths and the AR for samples above 20 m are generally somewhat lower (median: 1.08) than samples from below 20 m; 35– 55 m: median 1.25; 55–80 m: median 1.21 and below 80 m: median 1.16 (N = 2).  Two of the fracture coating samples in the depth interval 35– 55 m (KLX11D: 54.5 m and KLX11E: 36.7 m) have excess 230Th and are from fractures currently flowing at rates of 5  10 7 and 1  10 6 m2/s.



Fig. 8. Activity of 238U vs. depth for bulk fracture coating samples from boreholes KLX09B-G and KLX11B-F. Errors (1r) are within the size of the symbols.

samples (<20 m depth) have higher LREE concentrations (La, Ce, Pr, Nd and Sm) than the wall-rock and these samples possibly contain Paleozoic, REE-carbonate coatings, rich in Ce, La, Nd and Y (cf. Drake and Tullborg, 2005), and Ce-oxide in one of the samples. In the upper part of the boreholes the La/Yb-ratio in the fracture coatings are occasionally very high (Fig. 7a). Most samples from 0 to 20 m depth show positive Ce-anomalies (up to 1.52 CeCN/CeCN*, Fig. 7b), in contrast to samples from 20 to 70 m depth, which have slightly negative Ce-anomalies and samples from greater depths than 70 m, which generally lack Ce-anomalies.

The samples showing the largest disequilibrium 230Th/234U AR (>1.5) are generally low in 238U (Fig. 10). The 238U, 234U and 230Th data are summarized in a Thiel plot (Thiel et al., 1983) in Fig. 10. The result of bulk dissolution of U from a system, initially in secular equilibrium, would be to generate a distribution of samples with constant 234U/238U = 1 and 230 Th/238U > 1. Some of the near-surface samples (KLX11: 1.5, 3.7 and 16.9 and KLX09: 0.7) plot in positions that could be interpreted

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

Fig. 9. Activity ratios of

234

U/238U (a) and

230

1035

Th/234U (b) (with errors, 1r) vs. depth for bulk fracture coating samples from boreholes KLX09B-G and KLX11B-F.

high U content (KLX09: 35.0, 68.2, 69.9, 71.2 and 75.9) and all but one (KLX09: 91.0 m) of the other samples from depths greater than 60 m all plot in positions indicating previous deposition of U followed by a return towards equilibrium over a time of the order of the 230Th half life.

5. Discussion Earlier studies in the area and from these boreholes indicate that d13C, d18O, d34S and 87Sr/86Sr compositions in pyrite and calcite of the youngest fracture mineral paragenesis are similar to what is expected for precipitates from modern water or paleowater of similar composition as the modern water at ambient temperatures (e.g. Tullborg, 2003; Drake and Tullborg, 2007; Tullborg et al., 2008). The isotope characteristics are distinctly different from that of Precambrian calcite and pyrite, also present as fracture coatings and fracture fillings. The youngest fracture minerals in the coatings in the studied fractures may therefore possibly be precipitated recently, i.e. during the Quaternary and the characteristics of these outermost minerals indicate that Quaternary waters have descended to at least 100 m depth and have precipitated calcite at low temperatures. 5.1. Mineralogy

Fig. 10. Activity ratios of 234U/238U (a) and 230Th/234U (b) vs. activity of 238U (with errors, 1r) for bulk fracture coating samples from boreholes KLX09B-G and KLX11BF.

as indicating such U dissolution, consistent with oxidizing conditions at these depths. Importantly, none of the deep samples plot in the U-removal sector in Fig. 10, consistent with maintenance of reducing conditions at depth in the cores over a time period of the order of the 230Th half life of 7.54  104 a. The samples with

The distribution of the redox-sensitive minerals pyrite and goethite in open fractures shows that the redox front is generally located at about 15–20 depth below the surface (minimum 10 m depth, Fig. 3), which is in accordance with results of detailed Feoxide studies (Mössbauer spectroscopy, SEM and XRD) in these boreholes, which indicate (mainly based on the low degree of crystallinity, the small particle size and Fe-isotope ratios) that the uppermost Fe-oxides (0–20 m) are recent and of low-temperature origin, while deeper Fe-oxides are either older low-temperature precipitates or hydrothermal (Dideriksen et al., in press). In boreholes KLX09C and KLX09E the uppermost pyrite occurrences are found at ca 25 m. Overlapping occurrences of pyrite and goethite slightly deeper than 15–20 m in some boreholes indicate a redox

1036

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

transition zone over which the redox front has varied over time, but also indicate flow path heterogeneity in the fracture system. Above the redox front the wall-rock is affected by oxygenated water as shown by replacement of pyrite by goethite and alteration of biotite (Fig. 2a–d, k–l). This alteration is most common within a distance of 1 mm from the fracture but may locally extend up to 2 cm into the wall-rock. Goethite present in fracture coatings deeper in the boreholes (from 40 m down to ca 80 m) are uncommon, but occurs in some highly transmissive (flow rates > 1  106 m2/s) or highly fractured (>20 fractures/m) intervals. Furthermore, the uppermost occurrences of pyrite (at 10–20 m, Fig. 2e–h) are generally partly altered, which indicates an ongoing slow downward migration of the redox front and/or a zone where the conditions change with seasonal and annual variations. Fracture frequencies and orientations, soil cover and differences in reducing capacity of the fracture minerals and the wall-rock may be partly responsible for variations in redox front elevation/depth (Figs. 3 and 4). Redox fronts in KLX09 boreholes are slightly deeper relative to KLX11 boreholes and are considered a consequence of greater fracture frequencies in KLX09 host rock. The smaller percentage of open fractures containing calcite in the upper 20–25 m (Fig. 5) is interpreted as a result of descending oxidizing and weakly acidic fluids which have dissolved the calcite. Under reducing conditions deeper in the bedrock, the pH values of fracture water are higher and solutions become saturated to oversaturated with respect to calcite (SKB, 2006) allowing precipitation of recent calcite. Another mineralogical indication is the absence of chalcopyrite in fractures above the redox front defined by the pyrite distribution in the fractures (except in one sample).

5.2. Chemistry The positive Ce-anomalies observed in fracture minerals within the upper 20 m (Fig. 7b) suggest that Ce(III) has been oxidized to Ce(IV) and scavenged by goethite and/or Mn-oxide as described by Landström and Tullborg (1995), Vaniman and Chipera (1996), Bau (1999) and Takahashi et al. (2000). These positive Ce-anomalies compliment strong negative Ce-anomalies measured in nearsurface groundwater in the area (Rönnback et al., 2008). Positive Ce-anomalies in the fracture coatings are created through nearsurface weathering which results in dissolution of Ce(III) and the other REE. Subsequent accumulation of Ce(IV) in the solid phase is either due to sorption of the REE onto fracture coating goethite, at which Ce(III) is partly oxidized to Ce(IV) leading to preferential desorption of Ce(III) and the other REE compared to Ce(IV) as suggested by Bau (1999), or due to oxidation of Ce(III) to Ce(IV) in the aqueous phase followed by preferred sorption of highly adsorptive Ce(IV) relative to other REE onto fracture coatings (e.g. goethite), as shown by Takahashi et al. (2000). The depth distribution of the positive Ce-anomalies in analyzes of bulk fracture coatings (Fig. 7b) is similar to the distribution of goethite (Fig. 3), and is thus in good agreement with the redox front position determined by redox-sensitive minerals. As a consequence of the preferential precipitation of Ce(IV) above 20 m, fractures in the interval 20–70 m, where the conditions change to reducing, show slightly negative Ce-anomalies. At depths greater than 70 m this effect is no longer visible. High LREE contents (high La/Yb-ratios, Fig. 7a) and high Mn contents in the upper 20 m indicate influence of descending organic-rich waters and possibly also microbial activity, e.g. due to REE-complexation with organic material and due to subsurface microbial Mn reduction of organic material, respectively, in accordance with earlier studies in this area (Landström and Tullborg, 1995; Pedersen et al., 1997; Tullborg et al., 1999; Bath et al., 2000; Drake and Tullborg, 2007; Rönnback et al., 2008).

The elevated U contents in some samples at depths greater than 11 m (>15 ppm, down to 71 m, cf. 238U vs. depth in Fig. 8) might be due to deposition of U, mobilised at or near the surface and transported into the fracture system as mobile U(VI) mostly stabilised as carbonate complexes. At greater depths and reducing conditions this U is deposited either due to reduction and/or due to changes in HCO3 concentration (cf. Laaksoharju et al., 2008). The higher amount of e.g. Fe, Mg, Mn, K, Rb and Cs in the fracture coatings compared to the wall-rock is probably partly due to the extensive alteration of biotite in the wall-rock (cf. Drake et al., 2008) and that these elements are enriched in chlorite and clay minerals which dominate most of the fracture coatings. The much higher Cs contents (and Cs/K-ratios) in most samples below 50 m relative to samples above 50 m (Fig. 6c) is probably due to the presence of crush-zones containing increased amounts of chlorite and clay minerals with relatively high sorption affinities for Cs (Grütter et al., 1986; Landström and Tullborg, 1995). Depletion of CaO in near-surface fracture samples (Fig. 6a) is attributed to dissolution of calcite at these depths (Fig. 5), although Ca is also incorporated in other secondary phases such as mixed-layer clay. Occasional enrichments of Ba, P, Co, Cu and Ti in fracture coatings above 20 m may be related to the presence of Fe- and/or Mn-oxyhydroxides, since they are effective scavengers of these elements (Quejido et al., 2005, cf. Landström and Tullborg, 1995). However, Ba is probably related to harmotome and barite. 5.3. U-series disequilibrium U-series disequilibrium in most of the samples suggests extensive water–rock interaction at least within the last 300 ka to 1 Ma. However, a low number of samples show AR near secular equilibrium which indicates insignificant water–rock interaction during a long time period. Only two of the samples (KLX09D: 75.9 m, KLX11F: 1.9 m) show secular equilibrium, indicating negligible water–rock interaction within the last 1 Ma, in these specific fractures. Although there is scatter in the data set, mainly because of the heterogeneity of the fracture system (variation in transmissivity and amount of fracture coating material) some depth related trends of recent U mobilisation can be discerned. Based on

Fig. 11. Thiel diagram of U decay series for bulk fracture surface samples from KLX09B-G and KLX11B-F, separated by depths (after Thiel et al., 1983). Horizontal and vertical lines are shown for 230Th/238U = 1 and the 234U/238U = 1, respectively. The box of 1.0 ± 0.1 indicates secular equilibrium of 230Th/238U and 234U/238U within the conservative analytical error of 10%. Errors are in 1r. A and B are forbidden sectors for any single continuous process. Complex process sector is forbidden for any single process. The ‘‘forbidden sectors” indicate samples that have suffered complicated U transport (Thiel et al., 1983).

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039 234

U/238U and 230Th/238U AR (cf. the Thiel diagram in Fig. 11) most of the samples from the upper 20 m in the bedrock indicate a complex pattern of deposition and mobilisation during the last 300 ka (possibly as a result of both oxidizing and reducing conditions). Only two samples are consistent with simple removal of bulk U (KLX11E: 16.9 m and KLX11E: 3.7 m). Some of the samples originating from levels shallower than 20 m are close to 230Th/234U unity and do not contain any goethite, which suggests that these fractures have not been open to oxidizing water recently. A majority of the samples from depth greater than 35 m show deposition of U either as one single continuous process or as deposition during several occasions. However, elevated values of both 234U/238U and 230 Th/238U AR in two of the deeper samples (KLX11E: 36.7 m and 54.5 m, Fig. 11) indicate that fracture coatings have experienced both deposition and removal of U. A higher degree of bulk dissolution of U, probably from oxidizing groundwater circulation, in the upper 20 m is indicated by 234U/238U AR close to 1 and 230 Th/234U > 1. Fracture coatings from below 55 m tend to have 230 Th/234U < 1 and 234U/238U > 1, which is consistent with the concept of long-term deposition rather than removal of U under reducing hydrologic conditions (Landström et al., 1989; Tullborg et al., 2003). Furthermore, bulk removal of U during oxidizing conditions near the surface and redeposition of U at greater depths is also possibly indicated by the somewhat increasing 238U with increased depth, especially in highly transmissive fractures (Fig. 8), and by the correspondence between low 238U activities and high 230 Th/234U AR (>1.5, Fig. 10). Excess of 234U in the deeper samples may be due to preferential leaching of 234U relative to 238U, possibly at shallower depths, due to 234U being more likely to be present in the oxidized and more easily leached state of U(VI), relative to 238 U (e.g. Suksi and Rasilainen, 2002), and/or because of alpha recoil effects that makes more 234U susceptible to leaching than 238 U in the site (e.g. Gascoyne et al., 2002). The results indicate that recent U oxidation (within 300 ka) has preferentially taken place in the upper 20 m of the boreholes and locally down to 55 m along fractures with high transmissivity. At depths below 55 m recent (within 1 Ma) U deposition under reducing conditions is evident while more stable reducing conditions prevail below 80 m. However, a couple of samples between 55

1037

and 80 m show 230Th/234U  1, suggesting rather stable recent conditions in this interval as well. Water samples are unfortunately not available from these boreholes, but a few samples of near-surface groundwater (sampled in soil pipes) and shallow groundwater (sampled in percussion and cored boreholes) from the Laxemar area have been analyzed for 238U and 234U (Laaksoharju et al., 2009). The near-surface groundwaters show generally 234U/238U AR  1 and large variation in U contents (1–100 mBq/kg 238U), indicating oxidizing or mildly reducing conditions, whereas the shallow groundwaters sampled at depth from 50 to 250 m show 234 U/238U AR larger than 2 and U contents <50 mBq/kg 238U, indicating reducing conditions.

6. Conclusions A combination of different methods, such as detailed mapping and investigation of redox-sensitive minerals, geochemical analyzes (especially redox-sensitive elements) and U-series, of fracture coatings in several closely spaced cored boreholes is an excellent method to detect the redox front in crystalline rocks. The location of the redox front shows at what depths oxygenated groundwater associated with varying Quaternary hydrologic conditions (including subglacial or periglacial environments) was able to reach before O2 was consumed by fracture minerals reactions and organic redox buffering. These hydraulic conditions have been residing in bedrock fractures in the area, as shown by modelling (e.g. Puigdomenech et al., 2001, and references therein), groundwater chemistry (SKB, 2006) and by characteristics of the youngest fracture minerals (Drake and Tullborg, 2007). These investigations show that the redox front at two drill sites in the Laxemar area is generally positioned at about 15–20 m depth. The main features of the redox front are shown in Fig. 12. These are generally shown by a shift from: (1) mainly goethite to mainly pyrite in the fractures, (2) positive Ce-anomalies to slightly negative or insignificant Ce-anomalies, (3) mainly removal of U to mainly deposition of U.

Fig. 12. Tentative sketch model of the near-surface redox front in the Laxemar area, Sweden. The different fields represent the depth intervals where mineralogical, geochemical and U-series analyzes of fracture coatings indicate Quaternary oxidizing conditions, reducing conditions or a transition zone between these.

1038

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039

Leaching of calcite in open fractures in the upper 20–30 m further support infiltration of very dilute recharge waters which probably have relatively low pH and may at least partly be loaded with O2 (SKB, 2006). Scattered goethite occurrences down to ca 80 m and occasional signs of U removal at depths between 35 and 55 m is indicated by 230Th/234U AR > 1 in intervals of high transmissivity that facilitate localized percolation of oxygenated groundwater to depths that otherwise experience reduced conditions. Values of 230Th/234U AR that are near or below 1.0 in the sampled fracture coatings below 55 m depth imply that U was not removed preferentially with respect to Th during the last 300 ka and, therefore, that U was likely present in its insoluble reduced form. Although penetration of glacial waters to great depths has been confirmed in the area (SKB, 2006) this study supports the modelling carried out by Guimerà et al. (1999) and conclusions drawn by Gascoyne (1999) that these glacial waters were not oxidizing at repository depth. Instead, O2 in the recharge water is generally consumed within the upper 20 m (±5 m) and to slightly greater depth in fractures or shear zones with increased transmissivities (P1  10 7 m2/s). These observations are highly important for the demonstration of stable reducing conditions during the life time of a repository for spent nuclear fuel. The methodology used in this paper provides useful information for the safety assessment of a potential repository site. Acknowledgements Financial support by the Swedish Nuclear Fuel and Waste Management Co. (SKB) is acknowledged. Professor Sven Åke Larson, University of Gothenburg, provided useful comments and improved the manuscript. The authors would like to thank the reviewers Dr. Pierre Glynn and Dr. Jim Paces for their constructive comments. Fredrik Hartz, SKB, provided the map of the Simpevarp area. References Åhäll, K.-I., 2001. Åldersbestämning av svårdaterade bergarter i sydöstra Sverige. Swedish Nuclear Fuel and Waste Management Co. Report R-01-60. Akagawa, F., Yoshida, H., Yogo, S., Yamomoto, K., 2006. Redox front formation in fractured crystalline rock; an analogue of matrix diffusion in an oxidizing front along water-conducting fractures. Geochem. Explor. Environ. Anal. 6, 49– 56. Arcos, D., Pérez del Villar, L., Bruno, J., Domènech, C., 2008. Geochemical modelling of the weathering zone of the ‘‘Mina Fe” U deposit (Spain): a natural analogue for nuclear spent fuel alteration and stability processes in radwaste disposal. Appl. Geochem. 23, 807–821. Ask, H., 2006. Oskarshamn site investigation. Core drilling of short boreholes KLX11B, KLX11C, KLX11D, KLX11E and KLX11F for discrete fracture network investigation (DFN). Swedish Nuclear Fuel and Waste Management Co. Report P-06-283. Ask, H., 2007a. Oskarshamn site investigation. Core drilling of 13 short boreholes (KLX09G, KLX10B, KLX10C, KLX22A, KLX22B, KLX23A, KLX23B, KLX24A, KLX25A, KLX26A, KLX26B, KLX28A and KLX29A) for investigation of minor deformation zones (MDZ). Swedish Nuclear Fuel and Waste Management Co. Report P-06-297. Ask, H., 2007b. Oskarshamn site investigation. Core drilling of short boreholes KLX09B, KLX09C, KLX09D, KLX09E and KLX09F for discrete fracture network investigation (DFN). Swedish Nuclear Fuel and Waste Management Co. Report P-06-265. Bath, A., Milodowski, A., Ruotsalainen, P., Tullborg, E.-L., Cortés Ruiz, A., Aranyossy, J.-F., 2000. Evidences from mineralogy and geochemistry for the evolution of groundwater systems during the quaternary for use in radioactive waste repository safety assessment (EQUIP project). EUR report 19613. Bau, M., 1999. Scavenging of dissolved yttrium and rare earths by precipitating iron oxyhydroxide; experimental evidence for Ce oxidation, Y–Ho fractionation, and lanthanide tetrad effect. Geochim. Cosmochim. Acta 63, 67–77. Bingen, B., Skår, O., Marker, M., Sigmod, E.M.O., Nordgulen, O., Ragnhildstveit, J., Mansfeld, J., Tucker, R.D., Liégeois, J.-P., 2006. Timing of continental building in the Sveconorwegian orogen, SW Scandinavia. Norw. J. Geol. 85, 87–116. Boulton, G.S., Zatsepin, S., Maillot, B., 2001, Analysis of groundwater flow beneath ice sheets. Swedish Nuclear Fuel and Waste Management Co. Report TR-01-06. Brindley, G.W., Brown, G. (Eds.), 1984. Crystal structures of clay minerals and their X-ray identification. Mineral. Soc. Great Britain and Ireland. London, UK.

Cederbom, C., 2001. Phanerozoic, pre-cretaceous thermotectonic events in southern Sweden revealed by fission track thermochronology. Earth Planet. Sci. Lett. 188, 199–209. Cederbom, C., Larson, S.Å., Tullborg, E.-L., Stiberg, J.P., 2000. Fission track thermochronology applied to phanerozoic thermotectonic events in central and southern Sweden. Tectonophysics 316, 153–167. Cronquist, T., Forssberg, O., Hansen, L.M., Koyi, S., Vestgård, J., Wikholm, M., 2006. Oskarshamn site investigation. Detailed outcrop mapping on drillsite KLX11. Swedish Nuclear Fuel and Waste Management Co. Report P-06-06. Dia, A., Gruau, G., Olivie, L.G., Riou, C., Molenat, J., Curmi, P., 2000. The distribution of rare earth elements in groundwaters; assessing the role of source-rock composition, redox changes and colloidal particles. Geochim. Cosmochim. Acta 64, 4131–4151. Dideriksen, K., Christiansen, B.C., Baker, J.A., Frandsen, C., Balic-Zunic, T., Tullborg, E.-L., Mørup, S., Stipp, S.L.S., 2007. Fe-oxide fracture-fillings as a palotemperature and -redox indicator: structure, crystal form, REE content and Fe isotope composition. Chem. Geol. 244, 330–343. Dideriksen, K., Christiansen, B.C., Frandsen, C., Balic-Zunic, T., Mørup, S., Stipp, S.L.S., in press. Paleo-redox boundaries in fractured granite. Geochim. Cosmochim. Acta. Drake, H., Tullborg, E.-L., 2005. Oskarshamn site investigation. Fracture mineralogy and wall rock alteration, results from drill cores KAS04, KA1755A and KLX02. Swedish Nuclear Fuel and Waste Management Co. Report P-05-174. Drake, H., Tullborg, E.-L., 2007. Paleohydrogeology of the Simpevarp area, southeastern Sweden, as evidenced by stable isotopes in fracture minerals. In: Bullen, T.D., Wang, Y. (Eds.), Water–rock Interaction: Proc. 12th Internat. Symp. Water–Rock Interaction, Kunming, China. Taylor & Francis, London, pp. 723– 726. Drake, H., Page, L., Tullborg, E.-L., 2007. Oskarshamn site investigation,40Ar/39Ar dating of fracture minerals. Swedish Nuclear Fuel and Waste Management Co. Report P-07-27. Drake, H., Sandström, B., Tullborg, E.-L., 2006. Mineralogy and geochemistry of rocks and fracture fillings from Forsmark and Oskarshamn: compilation of data for SR-Can. Swedish Nuclear Fuel and Waste Management Co. Report R-06-109. Drake, H., Tullborg, E.-L., Annersten, H., 2008. Red-staining of the wall rock and its influence on the reducing capacity around water conducting fractures. Appl. Geochem. 23, 1898–1920. Drever, S.I., 1973. The preparation of oriented clay mineral specimens for X-ray diffraction analysis by a filter-membrane peel technique. Am. Mineral. 58, 553– 554. Eliasson, T., Tullborg, E.-L., Landström, O., 1989. Fracture filling mineralogy and geochemistry at the Swedish HDR research site. In: Baria, R. (Ed.), Camborne School of Mines Internat. Hot Dry rock Conf. Robertson Science Publ., London, pp. 425–435. Evansen, N.M., Hamilton, P.J., O’Nions, R.K., 1978. Rare earth abundances in chondritic meteorites. Geochim. Cosmochim. Acta 42, 1199–1212. Forssberg, O., Cronquist, T., Hansen, L., Vestgård, J., 2005. Oskarshamn site investigation. Detailed outcrop mapping at the drill site of KLX09 in Laxemar. Swedish Nuclear Fuel and Waste Management Co. (SKB). Report P-05-260. Gascoyne, M., 1999. Long-term maintenance of reducing conditions in a spent nuclear fuel repository. A re-examination of critical factors. Swedish Nuclear Fuel and Waste Management Co. (SKB). Report R-99-41. Gascoyne, M., Cramer, J.J., 1987. History of actinide and minor element mobility in an archean granitic batholith in Manitoba, Canada. Appl. Geochem. 2, 37–53. Gascoyne, M., Miller, N.H., Neymark, L.A., 2002. Uranium-series disequilibrium in tuffs from Yucca Mountain, Nevada, as evidence of pore-fluid flow over the last million years. Appl. Geochem. 17, 781–792. Glynn, P.D., Voss, C.I., 1999. Geochemical characterization of Simpevarp ground waters near the Äspö Hard Rock Laboratory. Swedish Nuclear Power Inspectorate (SKI). SKI Report 96:29. Glynn, P.D., Voss, C.I., Provost, A.M., 1999. Deep penetration of oxygenated meltwaters from warm based ice-sheets into the Fennoscandian shield. In: Proc. NEA/OECD Workshop ‘‘Use of Hydrochemical Information in Testing Groundwater Flow Models”, Borgholm, Sweden, September 1–3, 1997. Griffault, L.Y., Gascoyne, M., Kamineni, C., Kerrich, R., Vandergraaf, T.T., 1993. Actinide and rare earth element characteristics of deep fracture zones in the Lac du Bonnet granitic batholith, Manitoba, Canada. Geochim. Cosmochim. Acta 57, 1181–1202. Grütter, A., von Gunten, H.R., Rössler, E., 1986. Sorption, desorption, and isotope exchange of cesium (10 9–10 3 M) on chlorite. Clay Clay Min. 34, 677–680. Guimerà, J., Duro, L., Jordana, S., Bruno, J., 1999. Effects of ice melting and redox front migration in fractured rocks of low permeability. Swedish Nuclear Fuel and Waste Management Co. Report TR-99-19. Jasmund, K., Lagaly, G. (Eds), 1993. Tonminerale und Tone. D. Steinkopff, Darmstadt. Koistinen, T., Stephens, M.B., Bogatchev, V., Nordgulen, Ø., Wennerström, M.K.J. (Eds.), 2001. Geological map of the Fennoscandian shield scale 1:2000000. Geol. Surv. of Finland, Norway and Sweden and Ministry of Natural resources of Russia. Laaksoharju, M., Smellie, J., Tullborg, E.-L., Gimeno, M., Hallbeck, L., Molinero, J., Waber, N., 2008. Bedrock hydrogeochemistry Forsmark, site descriptive modelling, SDM-Site Forsmark. Swedish Nuclear Fuel and Waste Management Co. (SKB). Report R-08-47. Laaksoharju, M., Smellie, J., Tullborg, E.-L., Gimeno, M., Hallbeck, L., Molinero, J., Waber, N., 2009. Bedrock hydrogeochemistry Laxemar, site descriptive modelling, SDM-Site Laxemar. Swedish Nuclear Fuel and Waste Management Co. (SKB). Report R-08-93.

H. Drake et al. / Applied Geochemistry 24 (2009) 1023–1039 Landström, O., Tullborg, E.-L., 1989. The influence of fracture minerals/groundwater interaction on the mobility of U, Th, REE and other trace elements. Swedish Nuclear Fuel and Waste Management Co. Report TR-90-37. Landström, O., Tullborg, E.-L., 1995. Interactions of trace elements with fracture filling minerals from the Äspö Hard Rock Laboratory. Swedish Nuclear Fuel and Waste Management Co. Report TR-95-13. Landström, O., Smellie, J., Tullborg, E.-L., 1989. Mineralogical and geochemical studies of fracture-infillings in drillcore KLJ 01. In: Bäckblom, G., Stanfors, R. (Eds.), Interdisciplinary Study of Post-glacial Faulting in the Lansjärv Area Northern Sweden. Swedish Nuclear Fuel and Waste Management Co. (SKB). Report TR-89-31 (Chapter 9). Landström, O., Tullborg, E.-L., Eriksson, G., Sandell, Y., 2001. Effects of glacial/postglacial weathering compared with hydrothermal alteration – implications for matrix diffusion. Results from drillcore studies in porphyritic quartz monzodiorite from Äspö SE Sweden. Swedish Nuclear Fuel and Waste Management Co. Report R-01-37. Larson, S.Å., Tullborg, E.-L., Cederbom, C., Stiberg, J.P., 1999. Sveconorwegian and caledonian foreland basins in the Baltic Shield revealed by fission-track thermochronology. Terra Nova 11, 210–215. Lidmar-Bergström, K., 1996. Long term morphotectonic evolution in Sweden. Geomorphology 16, 33–59. Lindström, M., Lundqvist, J., Lundqvist, T., 1991. Sveriges geologi från urtid till nutid. Studentlitteratur, Lund. MacKenzie, A.B., Scott, R.D., Smellie, J.A.T., 1986. II: a comparison of neutron activation and alpha spectroscopy analyzes of thorium in crystalline rocks. Swedish Nuclear Fuel and Waste Management Co. Report TR-86-01. MacKenzie, A.B., Scott, R.D., Linsalata, P., Miekeley, N., 1992. Natural decay series studies of the redox front system in the Pocos de Caldas uranium mineralization. J. Geochem. Explor. 17, 289–322. Milodowski, A.E., Tullborg, E.-L., Buil, B., Gomez, P., Turrero, M.-J., Haszeldine, S., England, G., Gillespie, M.R., Torres, T., Ortiz, J.E., Zacharias, J., Silar, J., Chvatal, M., Strnad, L., Sebek, O., Bouch, J.E., Chenery, S.R., Chenery, C., Shepherd, T.J., McKervey, J.A., 2005. Application of mineralogical, petrological and geochemical tools for evaluating the palaeohydrogeological evolution of the PADAMOT study sites. PADAMOT Project Technical Report WP2. Min, M., Peng, X., Wang, J., Osmond, J.K., 2005. Uranium-series disequilibria as a means to study recent migration of uranium in a sandstone-hosted uranium deposit, NW China. Appl. Radiat. Isotopes 63, 115–125. Munier, R., Talbot, C.J., 1993. Segmentation, fragmentation and jostling of cratonic basement in and near Äspo, Southeast Sweden. Tectonics 12, 713–727. Osmond, J.K., Ivanovich, M., 1992. Uranium-series mobilization and surface hydrology. In: Ivanovich, M., Harmon, R.S. (Eds.), Uranium-series Disequilibrium; Application to Earth, Marine, and Environmental Sciences, second ed. Oxford Scientific Publ., Oxford. Pedersen, K., Ekendahl, S., Tullborg, E.-L., Furnes, H., Thorseth, I., Tumyr, O., 1997. Evidence of ancient life at 207 m depth in a granitic aquifer. Geology 25, 827– 830. Perez del Villar, L., Bruno, J., Campos, R., Gomez, P., Cozar, J.S., Garralon, A., Buil, B., Arcos, D., Carretero, G., Ruiz Sanchez Porro, J., Hernan, P., 2002. The uranium ore from Mina Fe (Salamanca, Spain) as a natural analogue of processes in a spent fuel repository. Chem. Geol. 190, 395–415. Puigdomenech, I., Ambrosi, J.-P., Eisenlohr, L., Lartigue, J.-E., Banwart, S.A., Bateman, K., Milodowski, A.E., West, J.M., Griffault, L., Gustafsson, E., Hama, K., Yoshida, H., Kotelnikova, S., Pedersen, K., Michaud, V., Trotignon, L., Rivas Perez, J., 2001. O2 depletion in granitic media. The REX project. SKB Report TR-01-05. Quejido, A.J., Perez del Villar, L., Cozar, J.S., Fernandez Diaz, M., Crespo, M.T., 2005. Distribution of trace elements in fracture fillings from the Mina Fe uranium deposit (Spain) by sequential leaching; implications for the retention processes. Appl. Geochem. 20, 487–506. Rivas-Perez, J., Tullborg, E.-L., Banwart, S.A., 2003. The kinetics of O2 (aq) reduction during oxidative weathering of naturally occurring fracture minerals in groundwater. Mineral. Mag. 67, 399–414. Rönnback, P., Åström, M., Gustafsson, J.-P., 2008. Comparison of the behaviour of rare earth elements in surface waters, overburden groundwaters and bedrock

1039

groundwaters in two granitoidic settings, Eastern Sweden. Appl. Geochem. 23, 1862–1880. Rudmark, L., Malmberg-Persson, K., Mikko, H., 2005. Oskarshamn site investigation. Investigation of quaternary deposits 2003–2004. SKB Report P-05-49. Scott, R.D., MacKenzie, A.B., Alexander, W.R., 1992. The interpretation of 238 U–234U–230Th–226Ra disequilibria produced by rock–water interactions. J. Geochem. Explor. 45, 323–343. SKB, 2006. Hydrogeochemical evaluation, Preliminary site description Laxemar subarea – version 2.1. Swedish Nuclear Fuel and Waste Management Co. (SKB). Report R-06-70. Smellie, J.A.T., Mackenzie, A.B., Scott, R.D., 1986. An analogue validation study of natural radionuclide migration in crystalline rocks using uranium-series disequilibrium studies. Chem. Geol. 55, 233–254. Söderbäck, B. (Ed.), 2008. Geological evolution, palaeoclimate and historic development of the Forsmark and Laxemar/Simpevarp areas, site descriptive modelling, SDM-Site. Swedish Nuclear Fuel and Waste Management Co. (SKB). Report R-08-19. Söderlund, P., Juez, L.J., Page, L.M., Dunai, T.J., 2005. Extending the time range of apatite (U–Th)/He thermochronometry in slowly cooled terranes; palaeozoic to cenozoic exhumation history of southeast Sweden. Earth Planet. Sci. Lett. 239, 266–275. Sokolnicki, M., Kristiansson, S., 2007. Oskarshamn site investigation. Difference flow logging of boreholes KLX11B-F Subarea Laxemar. Swedish Nuclear Fuel and Waste Management Co. Report P-07-74. Sokolnicki, M., Väisäsvaara, J., 2006. Oskarshamn site investigation. Difference flow logging of boreholes KLX09B-F Subarea Laxemar. Swedish Nuclear Fuel and Waste Management Co. Report P-06-199. Suksi, J., 2001. Natural uranium as a tracer in radionuclide geosphere transport studies. PhD Thesis, report series in radiochemistry 16/2001, Univ. Helsinki. Suksi, J., Rasilainen, K., 2002. Isotope fractionation of U in rocks reflecting redox conditions around a groundwater flow route. Mater. Res. Soc. Symp. 663, 961– 969. Takahashi, Y., Shimizu, H., Kagi, H., Yoshida, H., Usui, A., Nomura, M., 2000. A new method for the determination of CeIII/CeIV ratios in geological materials; application for weathering, sedimentary and diagenetic processes. Earth Planet. Sci. Lett. 182, 201–207. Thiel, K., Vorwerk, R., Saager, R., Stupp, H.D., 1983. 235U fission tracks and 238Useries disequilibria as a means to study recent mobilization of uranium in archaean pyritic conglomerates. Earth Planet. Sci. Lett. 65, 249–262. Tullborg, E.-L., 1997. Recognition of low-temperature processes in the Fennoscandian Shield. PhD Thesis, Göteborg Univ., Sweden. Tullborg, E.-L., 2003. Palaeohydrogeological evidences from fracture filling minerals – results from the Äspö/Laxemar area. Mater. Res. Soc. Symp. 807, 873–878. Tullborg, E.-L., Drake, H., Sandström, B., 2008. Palaeohydrogeology: a methodology based on fracture mineral studies. Appl. Geochem. 23, 1881–1897. Tullborg, E.-L., Landström, O., Wallin, B., 1999. Low-temperature trace element mobility influenced by microbial activity; indications from fracture calcite and pyrite in crystalline basement. Chem. Geol. 157, 199–218. Tullborg, E.-L., Smellie, J., Mackenzie, A.B., 2003. The use of natural uranium decay series studies in support of understanding redox conditions at potential radioactive waste disposal sites. Mater. Res. Soc. Symp. 807, 571–576. Väisäsvaara, J., Leppänen, H., Kristiansson, S., Pöllänen, J., 2006. Oskarshamn site investigation. Difference flow logging of boreholes KLX09G, KLX10B and KLX10C Subarea Laxemar. Swedish Nuclear Fuel and Waste Management Co. Report P-06-229. Vaniman, D.T., Chipera, S.J., 1996. Paleotransport of lanthanides and strontium recorded in calcite compositions from tuffs at Yucca Mountain, Nevada, USA. Geochim. Cosmochim. Acta 60, 4417–4433. Wahlgren, C.-H., Hermanson, J., Forssberg, O., Curtis, P., Triumf, C.-A., Drake, H., Tullborg, E.-L., 2006. Geological description of rock domains and deformation zones in the Simpevarp and Laxemar subareas. Preliminary site description Laxemar subarea – version 1.2. Swedish Nuclear Fuel and Waste Management Co. Report R-05-69.