The Hyrkkbla Native Copper Mineralization As a Natural Analogue for Copper Canisters
FI9700011
POSIVA-96-1 5
The Hyrkkbla native copper mineralization as a natural analogue for copper canisters
Nuria Marcos
October 1 996
POSIVA OY Annankalu 42 D. FIN-OO1OO HELSINKI. FINLAND *UL 2 8 IK? Q 5 Phone (09) 228 030 (nat ). ( + 358-9-) 228 030 (int.) Fax (09) 2280 3719 (nat.). ( + 358-9-) 2280 3719 (int ) POSIVA-96-1 5
The Hyrkkola native copper mineralization as a natural analogue for copper canisters
Nuria Marcos Helsinki University of Technology Laboratory of Engineering Geology and Geophysics
October 1996
POSIVA OY
Annankatu 42 D. FIN-OO1OO HELSINKI. FINLAND
Phone (09) 228 O3O (nat.). ( + 358-9-) 228 030 (int.)
Fax (09) 2280 3719 (nat.), ( + 358-9-) 2280 3719 (int.) ISBN 951-652-014-6 ISSN 1239-3096
The conclusions and viewpoints presented in the report are those of author(s) and do not necessarily coincide with those of Posiva. Posiva-raportti - Posiva report Raportinmnnus-Report co* POSIVA-96-15
Annankatu 42 D, FIN-00100 HELSINKI, FINLAND Juikasuaka-Date Puh. (09) 2280 30 - Int. Tel. +358 9 2280 30 October 1996
Tekija(t) - Author(s) Toimeksiantaja(t) - Commissioned by Nuria Marcos Helsinki University of Technology Posiva Oy Laboratory of Engineering Geology and Geophysics
Nimeke - Title
THE HYRKKOLA NATIVE COPPER MINERALIZATION AS A NATURAL ANALOGUE FOR COPPER CANISTERS
Tiivistelma - Abstract
The Hyrkkola U-Cu mineralization is located in southwestern Finland, near the Palmottu analogue site. The age of the mineralization is estimated to be between 1.8 and 1.7 Ga. Petrological and mineralogical studies have demonstrated that this mineralization has many geological features that parallel those of the sites being considered for nuclear waste disposal in Finland. A particular feature is the existence of native copper and copper sulfides in open fractures in the near-surface zone. This allows us to study the native copper corrosion process in analogous conditions as expected to dominate in the nuclear fuel waste repository. The occurrence of uranyl compounds at these fractures permits also considerations about the sorption properties of the engineered barrier material (metallic copper) and its corrosion products.
From the study of mineral assemblages or paragenesis, it appears that the formation of copper sulfide (djurleite, CU1.934S) after native copper (Cu°) under anoxic (reducing) conditions is enhanced by the availability of dissolved HS" in the groundwater circulating in open fractures in the near-surface zone. The minimum concentration of HS" in the groundwater is estimated to be of the order of 105 M (~1(H g/1) and the minimum pH value not lower than about 7.8 as indicated by the presence of calcite crystals in the same fracture.
The present study is the first one that has been performed on findings of native copper in reducing, neutral to slightly alkaline groundwaters. Thus, the data obtained is of most relevance in improving models of anoxic corrosion of copper canisters.
ISBN ISSN ISBN 951-652-014-6 ISSN 1239-3096
Sivumaara - Number of pages Kieli - Language 39 + Appendices English Posiva-raportti - Posiva report POSIVA-96-15 Annankatu 42 D, FIN-00100 HELSINKI, FINLAND Juikasuaika- Date Puh. (09) 2280 30 - Int. Tel. +358 9 2280 30 Lokakuu 1996
Tekijä(t) - Author(s) Toimeksiantaja(t) - Commissioned by Nuria Marcos Teknillinen korkeakoulu Posiva Oy Insinöörigeologian ja geofysiikan laboratorio
Nimeke - Title
HYRKKÖLÄN LUONNONKUPARIESIINTYMÄ KUPARIKANISTERIN LUONNONANALOGIANA
Tiivistelmä - Abstract
Hyrkkölän luonnonkupariesiintymä sijaitsee Lounais-Suomessa, lähellä Palmottua, jossa tutkitaan uraaniesiintymää analogiana kallioon loppusijoitettavalle käytetylle uraanipolttoaineelle. Minerali- saation ikä on 1800 - 1700 miljoonaa vuotta. Mineralogiset ja petrografiset tutkimukset osoittavat, että luonnonkupariesiintymä vastaa niitä olosuhteita, jotka vallitsevat loppusijoituksessa syvällä (300 - 800 metrin syvyydellä) graniittisessa peruskalliossa. Eräs ilmiö on luonnonkuparin ja kupari- sulfidin esiintyminen avoimissa raoissa lähellä maanpintaa. Tämä mahdollistaa luonnonkuparin korroosioilmiöiden tutkimisen samoissa olosuhteissa, joiden odotetaan vallitsevan loppusijoitus- paikassa. Myös U+6:n esiintyminen niissä raoissa antaa mahdollisuuden selvittää tarkemmin niin luonnonkuparin kuin sen korroosiotuotteiden sorptio-ominaisuuksia.
Mineraaliparageneesien perusteella käy ilmi, että kuparisulfidi (djurleite, CU1.934S) muodostuu luonnonkuparista pelkistävissä olosuhteissa, joissa pohjaveteen liuennut HS' kulkee avoimissa raoissa lähellä maanpintaa. Pohjaveden HS":n minimipitoisuuden arvioitiin olevan noin 105 M (~ 10'4 g/l) ja minimi-pH-arvon noin 7,8, perustuen raoissa esiintyviin kalsiittikiteisiin.
Tämä tutkimus on ensimmäinen, joka on suoritettu sellaisessa paikassa, jossa luonnonkupari esiin- tyy pelkistävissä olosuhteissa. Saatua dataa voitaisiin käyttää päivittämään kuparikapselin korroosio- malleja.
ISBN ISSN ISBN 951-652-014-6 ISSN 1239-3096
Sivumäärä - Number of pages Kieli - Language 39 + liitteet Englanti CONTENTS
ABSTRACT
THVISTELMA
PREFACE
1 INTRODUCTION 1 1.1 Background 1 1.2 Objectives 2
2 GEOLOGY 3 2.1 Geological setting 3 2.2 Petrology of the rock association at Hyrkkola 6 2.2.1 Amphibolites 6 2.2.2 Gneisses 7 2.2.3 Granite pegmatites 7 2.3 Bedrock fracturing 7 2.4 U-Cu mineralization 8 2.4.1 Native copper 8 2.4.2 Uranium minerals 10 2.4.3 Associated minerals 10 2.4.4 Genesis and age of mineralization 12
3 SUMMARY OF THE HYRKKOLA MINERALIZATION 13
4 FRACTURE MINERALOGY 15 4.1 Experimental techniques 17 4.2 Cuprite composition 17 4.3 Copper sulfide composition 18 4.4 Native copper composition 20 4.5 Fracture surface and uranium distribution 21 5 THERMODYNAMICS OF COPPER CORROSION AT HYRKKOLA 28 5.1 Hydro thermal corrosion, cuprite, sealed fractures 28 5.2 Corrosion in open fractures: djurleite 29
6 FEATURES IN COMMON WITH PROPOSED
REPOSITORY SITES 32
7 CONCLUSIONS 35
8 REFERENCES 36
APPENDICES PREFACE
This work was carried out at the Helsinki University of Technology (HUT) on contract for Posiva Oy. The contact persons were Juhani Vira and Marjo Matikainen at Posiva and Nuria Marcos at HUT. The author thanks Lassi Pakkanen and Bo Johanson (GSF, Geological Survey of Finland) for performing the SEM analyses and Lasse Ahonen for valuable discussions. 1 INTRODUCTION
1.1 Background
In Finland and Sweden spent nuclear fuel is planned to be disposed of in a repository to be constructed at a depth of about 300 - 800 m in crystalline bedrock. The nuclear waste will be isolated from the biosphere by multiple technical and natural barriers. In assessing the performance of the system, the corrosion of a copper canister (technical barrier) is considered.
Numerous laboratory studies have investigated the corrosion rates and mechanisms of metallic copper and its alloys (SKI Report 94:6, 1994, contains more than 150 references). Whilst these experiments may provide useful indications of the relative durability of the metallic copper, few are conducted in conditions that adequately simulate those expected in a repository (Miller et al., 1994). The authors of the most recent laboratory study dealing with the long-term predictions of copper canisters corrosion under oxic and anoxic conditions (Wersin et al., 1994) state that chemical corrosion of copper canisters does not appear to constitute a problem for repository safety but other factors such as increased temperature have not been explicitely included in their model. Miller et al., 1994, write: "One of the investigations of copper in a more repository relevant environment is that of Marcos (1989), which shows that copper is stable, and has remained stable, under a broad range of hydro- geochemical environments, but none of these were totally representative of the repository environment."
To cover this gap, a native copper mineralization in crystalline bedrock at Hyrkkola (SW Finland), is herein evaluated as a natural analogue for copper canister corrosion behaviour under repository conditions. 1.2 Objectives
The main objectives of this study on the HyrkkOla Cu mineralization were:
to clarify how the Hyrkkbla Cu mineralization corresponds to the circumstances prevailing in the final repository planned by Posiva in crystalline bedrock to a 300 - 800 m depth, and
to estimate the chemical conditions of the groundwater in fractures in order to evaluate the existing models for copper corrosion (mainly sufldization) in the deep bedrock environment.
To achieve the first objective, geological, petrographical and mineralogical studies are presented. The estimation of the geochemical conditions under which native copper has been preserved is based on these studies.
To address the second objective detailed quantitative and qualitative fracture mineral studies are presented. The estimation of the geochemical condition under which the fracture mineral assemblage has been formed and preserved, and an evaluation of the models for copper corrosion (mainly sulfidization) are based on these studies. 2 GEOLOGY
2.1 Geological setting
The native copper mineralization at Hyrkkola is related to a main uranium mineral- ization and thus in the following it is referred to as the Hyrkkbla U-Cu-mineral- ization. It was found in the 1980's by car-borne surveys utilizing a Scintrex BGS-3 radiometric instrument mounted in a vehicle.
The Hyrkkola U-Cu-mineralization is situated about 10 km NE from the Palmottu U-Th deposit, SW Finland (Fig. 1). The investigation area is within a zone of metamorphosed supracrustal volcanic and sedimentary rocks, which extends from SW Finland to Central Sweden. The Hyrkkola U-Cu-mineralization has been related to the late stages of the Svecokarelidic orogenic events about 1.8 - 1.7 Ga ago (Raisanen, 1986). The surrounding rocks were metamorphosed under low pressure amphibolite-facies conditions (3-5 kbar) and a temperature about 550 - 650° C (Schreus and Westra, 1986).
The rocks present are amphibolites, quartz-feldspar gneisses, and granite pegmatites (Appendices 1 to 4). They are all strongly deformated as would be expected considering their structural situation (Fig. 2). The rocks are characterized by a dense network of micro fractures. GEOLOGICAL MAP NUMM1 - PUSULA
GRANODIORITE QUARTZ FELDSPAR GNEISS MICA GNEISS
AMPHIBOLITE
URANIUM
0 1 2 3 < S km
GEOLOGICAL SURVEY OF FINLANO
FIG. 1. Geological map of the study site (after Raisanen, 1986). PALMOTTU REGION
TECTONIC MAP
LEGENO
II cltsi fracture zone
Itl cuts
possible III class
strike ol schistosiiy Sft / S(
•no '"".CV P*rnio granite
U-miner alixa lion
Palmollu study site
TH€ MAP IS BAS60 ON THE INTERPRETATION OF THE LOW MTITUOE AEROMAGNETlC utf
( mkOe br Geologic*! Swrv«r «l Finland 1
FIG. 2. The tectonic interpretation of low altitude aeromagnetic map around Palmottu and Hyrkkola U-deposits (after Kuivamaki et al., 1991). 2.2 Petrology of the rock association at Hyrkkola
The drill cores (Dh 301 to Dh 305; Fig. 3) used in this work were available at the central store of the Geological Survey of Finland. At this preliminary stage of investigation, more than 20 samples were studied with the stereomicroscope and about a dozen of thin and polished sections were examined. Appendices 2 to 4 show the location of the studied samples. a-Autoradiography was used to found the spatial relation between native copper and uranium minerals.
6715
6714 I FIG. 3. Situation of drill holes 301 to 305. 496
2.2.1 Amphibolites
Amphibolites are dark gray, homogeneous or banded, fine to medium-grained rocks.
The main minerals are hornblende, plagioclase (An32j,3) and minor diopside. Accessory and secondary minerals are quartz, opaques (magnetite, hematite, pyrite, pyrrhotite, graphite), epidote, chlorite, carbonates and clay minerals. Hematite, epidote and carbonates are typical fracture-filling minerals. 2.2.2 Quartz-feldspar gneisses
These rocks are light to dark gray, homogeneous to highly fractured and fine to medium grained. Medium grained gneisses often show primary volcanic textures and
notable porosity. The main minerals are quartz and plagioclase (An20.33). Hornblende is a major accessory mineral where gneisses are in contact with amphibolites; other accessory minerals are garnet, titanite, green apatite, zirkon and opaques (magnetite, graphite, pyrite). Epidote, chlorite and hematite are secondary minerals. Epidote and hematite occur in joint fillings.
2.2.3 Granite pegmatites
The granite pegmatites occur as veins (0.05 -5m thick) in quartz-feldspar gneisses and amphibolites (Appendices 2 to 4.3). They are heterogeneous, medium to coarse grained, and light to red. The main minerals are quartz, feldspar (mostly microcline), and black tourmaline. Black tourmaline may be an accessory in the light granites. Other accessory minerals are green apatite, titanite, zirkon, graphite, uraninite, and native copper. Epidote, hematite, cuprite, chalcocite, calcite, uranophane, and gummite are secondary minerals.
2.3 Bedrock fracturing
In order to evaluate the stability of native copper in fractured, water-saturated rock, the main fracture features were examined. The open fractures may or may not contain surface coating minerals, but closed fractures are commonly sealed with secondary minerals.
The surface coating minerals on open fractures are mostly hematite, epidote, clay minerals, calcite, and Fe-hydroxide. Epidote and calcite are the typical filling minerals in sealed fractures.
The drill cores were not orientated when drilled, so the real dips and dip directions of the fractures are not known. Fractures and fracture zones are difficult to connect from one drillhole to another at this stage of investigation. The fracture density (number of fractures/m) varies between 1 and 8. It decreases with depth in drillhole Dh 305 but no significant variations were observed down drillholes Dh 303 and Dh 304 (Appendixes 2 to 4).
2.4 U-Cu-mineralization
U-Cu-mineralization occurs in granite pegmatite veins. Copper-bearing veins in amphibolites and quartz-feldspar gneisses were intersected at varying depths from about 8.3 m in drillhole Dh 304 to 62 m in drillhole Dh 303 (Fig. 4).
'80 [-"-"- Ampmool ite "9U . Native copper m
FIG. 4. Sketch of Dh 303 and Dh 304.
2.4.1 Native copper
Native copper occurs as thin flakes (0<1.5 mm) in the granite pegmatites in and between grains of feldspar (Fig. 5a), and in and around sheared tourmaline grains (Fig. 5b). Copper also fills microfractures in granite pegmatites and in the associated rocks where these are in or near the contact with intrusive rocks. FIG. 5. a) Native copper slates in and around a feldspar grain. Sample S36, Dh 303. b) Native copper slates in and around a strongly deformated tourmaline grain. Sample S36, Dh 303. 10
2.4.2 Uranium minerals
Primary uraninite occurs as disseminations in the host-rocks (Raisanen, 1986). However uraninite was not observed by the eye of the author during the present study, but it was distinguised by a-autoradiography. Other uranium minerals are secondary uranyl (U6*) compounds like gummite and uranyl silicates (uranophane
Ca(UO2)2Si2O7-6H2O; Fig. 6). These secondary minerals are disseminated in the rock matrix, void fillings, microfractures, and fracture surfaces.
FIG. 6. Uranophane (yellow crystals).
Sample S37 in Dh 303.
2.4.3 Associated minerals
Feldspar (microcline), and black tourmaline are the primary rock-forming minerals associated with native copper occurrence. Hematite and minor cuprite are common where copper fills microfractures. It was observed from autoradiographs that a common mineral assemblage in filled microfractures is native copper, hematite, and secondary uranium phases. Native copper, chalcocite (Cu2S), and gummite form an interesting mineral assemblage in sample S40 (Fig 7a). Cuprite (Cu2O) replaces native copper from the borders to the center of the fissure in sample S30 (Fig. 7b). These samples are from depths of 8.3 m and 11.3 m respectively (Fig. 4). 11
FIG. 7. (a) Native copper(bright pink-yellow) and chalcocite (blue) in open fracture. Reflected light, parallel nicols. Sample S40, Dh 304. (b) Cuprite (light blue) replaces native copper from the borders to the center of the fissure. Reflected light, parallel nicols. Sample S30, Dh 303. Explanation of numbers and profile in Chapter 4. 12
2.4.4 Genesis and age of mineralization
The host-rocks of the U-Cu-mineralization are granite pegmatites. These rocks are considered to be the products of crystallization of residual melts from granitic mag- mas rich in volatiles and elements that are not readily incorporated into the lattices of major granite minerals (Evans, 1980). Raisanen (1986) relates the source of the uranium to the Pernio granite, a late-kinematic intrusion situated about 14 km SW of the mineralization. The uraninite of the Hyrkkola mineralization has been dated to 1.774 Ga (Kouvo, 1983), which correlates with the late stages of Svecokarelidic orogenic events about 1.8 - 1.7 Ga ago (see above).
Copper is not a typical element enriched or present in pegmatitic phases. The source of the copper is more likely to be from the country rocks. It may have been mobil- ized and transportated by pegmatitic fluids migrating through the amphibolites and quartz-feldspar gneisses. Copper solubilizes readily at pH values of 6 to 9 to cuprous (Cu+) chloride complexes at [Cl] = 1 M (Rose and Bianchi-Mosquera, 1993). Most hydrothermal solutions have pH values within 2 units either side of neutral pH, due to the buffering capacity of silicate minerals in the surrounding rocks. Decrease in temperature or pressure may decrease the activity of chloride (Cl") and/or the activity of dissolved copper. Also an increase in pH at constant Eh may cause the precipitation of metallic copper.
Uranium is transported as soluble uranyl (U6*) complexes. In a discussion of the genesis of uraninite, Romberger (1984) suggests also that an increase in pH under constant oxygen fugacity (log/<92 —35) may reduce the uranyl ion to the uranous (U4*) ion. The electrons necessary for the reduction may be derived from ferrous (Fe2+) iron during its oxidation to the ferric (Fe3*) state. Hematite, uranium minerals, and native copper form, in Hyrkkola, a common mineral assemblage.
Temperatures of crystallization of feldspars, in which native copper is included, may vary from about 250 to 300 °C, and temperatures of crystallization of tourma- line, to which copper is related, may vary from 350 to 550 °C. Without further studies it is hard to say more about the temperature of crystallization of the granite pegmatite or the mineralization itself. 13 3 SUMMARY OF THE HYRKKOLA MINERALIZATION
Figure 8 is a simplified representation of the evolution of the Hyrkkola mineral- ization. As mentioned above, Raisa"nen (1986) relates the granite pegmatites in- trusions to the late stages of orogenic events occuring about 1.8 - 1.7 Ga ago. The data of temperatures, presures, and Eh-pH at the time of native copper (Cu°) and uraninite (UO2) precipitation follows from the preceding text.
Soon (in a geological sense) after the crystallization of the granite pegmatites, fracturing occurred and hydrothermal fluids penetrated indistinctly the host-rocks and wall-rocks of the mineralization. Hydrothermal fluids introduced the metals (native copper) and minerals (e.g., hematite, uraninite) into the open fractures where they subsequently precipitated. In a later stage, residual hydrothermal fluids with a high oxidation potential corroded native copper in fissures to cuprite. The tempera- ture and Eh-pH conditions have been estimated from the mineral assemblage present in these sealed fractures.
The Svecokarelidic orogeny (2.0 - 1.75 Ga), to which the Hyrkkola' mineralization is related, was followed by a long period of erosion and cratonization, and the present surface level was almost reached about 0.57 Ga ago (Simonen, 1980). The history of the geological events during this time and up to the last glaciation (10 000 years) is full of gaps and consequently it is difficult to fix the periods when the Cu-U mineralization was exposed to corrosion processes. The minimum age esti- mated is that of the end of the last glaciation, 10 000 years. This is the minimum estimate for the start of the corrosion processes because during glaciation the flow of groundwater (and corrodants in it) is limited. Furthermore the land uplift in Finland since the last Quaternary glaciation may have caused recent fracturing and the activation of old fractures (Niini, 1968, 1987). The data of copper corrosion scenario in Figure 8 comes from Chapter 5. The minimum depth estimation is based on data of Okko (1964). 14
Cu-U Hyrkkola mineralization:
Genesis scenario
T 250 - 550 °C Age 1.8 - 1.7 Ga P 300 - 500 MPa Depth 3 - 5 Km Q-FG Quartz-feldspar pH - 6 - 9 gneiss Eh? AM Amphibolites GP Granite pegmatites
Hydrothermal alteration scenario
Agem». 1.75 - 1.6 Ga
Ageroi, 1.2 Ga ?
T 100 - 150 °C p ? Cu° + Fe2O3/Cu2O + UO2 Eh - +200 to -200 mV filled veins pH - 6.5 to 9
Copper corrosion scenario
Tra,, - 72 . P
15 - 20 m Agerall. ? Eh - -300 to -400 mV Age,™. 10 000 years pH - 7.8, from Figures 18 and 19. Open fractures, where Cu° and Cu,,,,S coexist
Corrosion process: Groundwater flow ? Eq(3) 1.934Cu° + HS + H* = CuIV34S + H2(aq) Groundwater composition: Bacterial induced corrosion? [HSU = 10" - 105 g/1 Reaction kinetics? [Ca21 = 0 Other compounds ?
FIG. 8. Scheme of the geological and geochemical events at the Hyrkkola mineralization. Data from Chapters 2 and 5 and from Okko (1964). 15 4 FRACTURE MINERALOGY
Understanding the evolution of the Hyrkkbla U-Cu mineralization, its mineral assemblages and the prevalent geochemical conditions, is of relevance in verifying the existing geochemical models of native copper corrosion and the modelling of the long-term stability of native copper in granitic environment.
Korzhinskii (1959) states that a mineral assemblage at a given point may have been fixed by an externally controlled medium such as a moving fluid or pervasive at- mosphere. Therefore a detailed study of the fracture mineral assemblages may reveal the geochemical conditions in which they formed and the chemistry of the ground- water in which they have been preserved. The hydrogeochemical conditions (water- rock interactions) that control the stability of the native copper and related fracture minerals are herein evaluated.
The fracture mineralogy assemblages are native copper - cuprite/hematite and native copper - copper sulfide. The first assemblage is present in sealed microfractures (e.g. Fig. 7b, p. 11) and the second one is present in an open fracture at about 8.3 m depth in Dh 304 (Fig. 3 and Appendix 4.1). So far, this fracture is the only one where, among other minerals, native copper and copper sulfides are present. Second- ary uranium phases are present in both types of fractures, but gummites are only eye-visible in open fractures (Figs. 9a and 9b).
The outer part of steel-copper canisters, engineered barriers of spent nuclear fuel, is composed of metallic copper. The occurrence of uranyl compounds at both type of fracture leads to considerations about the sorption properties of the engineered barrier (native/metallic copper) and engineered barrier corrosion products under possible repository conditions; cuprite may be the corrosion product of copper in oxidizing environments and copper sulfide may be the corrosion product of copper in reducing environments. 16
FIG. 9. (a) Fracture surface, a) Djurleite, b) Gummites, c) Ferric hydroxides. Sample S40, Dh 304. (b) Fracture surface. Electronprobe microanalyses from the square in Section 4.5. Sample from 8.2 - 8.3 m depht, closed to S40. 17 4.1 Experimental techniques
Fracture surfaces, polished sections and polished thin sections were re-examinated using a stereomicroscope and a polarizing microscope; a-particle autoradiography was used to determine the occurrence of radionuclides and its spatial relation to other fracture minerals. This same relation and quantitative and qualitative analyses of the native copper and copper sulfide in fractures were obtained using an electronprobe microanalyser (EPMA) CAMECA SX 50 at the GSF, Espoo. The ideal minimum size of the particles to be analysed is 0 = 5 urn. The detection limit of quantitative analyses is of the order of 10 ppm. X-ray semiquantitative analyses are also possible using energy dispersive spectrometry. The characteristics and properties of the EPMA have been described by Johanson and Kojonen (1995).
4.2 Cuprite composition
Cuprite occurs as an infill of a closed fracture at 11.3 m depth (Fig. 7b, p. 11). The data was normalized to 100%, the atomic weight of copper was taken as 63.54 and the one of oxygen as 16.00. Thus, the stoichiometric relation of copper and oxygen is that of Cu/O = 2/1 (Table 1).
Table 1. Chemical composition (weight-%) of cuprite at sample S30.
Point no. Cu Fe Ca S U O Si Zn Total
1 85.07 0.00 0.02 0.02 0.58 10.50 0.36 0.05 96.60
2 80.82 0.00 0.01 0.01 0.63 11.76 0.42 0.04 93.69
3 83.02 0.02 0.01 0.00 1.21 11.25 0.16 0.01 95.69
The data of microprobe analyses (Table 2) through the profile marked in Figure 7b shows that uranium concentration increases with increasing content of oxygen and towards the boundaries of the fracture, where cuprite is most abundant. The migra- 18 tion of uranium along fractures has been observed before (e.g., Jaakkola et al., 1989). Cuprite acts in this case as an adsorbent of uranium.
Table 2. Profile at sample S30 (weight %).
Point no. Si Cu U O Total
1 13.65 36.46 1.78 19.23 71.11
2 5.44 76.61 1.41 10.26 93.72
3 0.67 92.23 0.20 3.17 96.27
4 0.12 93.68 0.00 2.24 96.04
5 0.07 93.89 0.00 2.16 96.12
6 0.09 94.07 0.00 2.28 96.04
7 0.10 93.10 0.02 2.62 96.12
8 0.12 91.04 0.24 5.70 96.44
9 0.16 85.77 0.74 10.87 95.84
10 0.19 83.17 0.69 12.21 97.10
11 0.56 83.09 0.79 12.95 97.39
12 3.41 80.54 0.84 14.33 99.32
4.3 Copper sulfide composition
The copper sulfide at sample S40 (8.3 m depth) was named chalcocite after its study under a polarizing microscope. Chalcocite (Cu2S) is one of the end members of the copper-sulfur system; the other end member is covellite (CuS). The end members can be easily distinguished one from each other by polarizing microscopy. The other members (Table 3) can be distinguished easier by quantitative analytical techniques.
Table 4 shows the results of microprobe analyses of copper sulfide at sample S40 (Fig. 10). The data was normalized to 100%, the atomic weight of sulfur was taken as 32.06 and the one of copper as 63.54. The calculated average of the Cu/S ratio is 1.934 and the standard deviation on_! = 0.033. This result indicates that the copper sulfide phase is djurleite, Cu, 9J4S, the upper stability of which varies between 72 ± 2 and 93 ± 2 °C. Snellgrove and Barnes (1974) found that djurleite is the solid phase precipitated at low temperatures under reducing conditions (Eh < 0) and alkaline pH's (pH - 8). 19
Table 3. Stable phases of the system copper-sulfur. Compositions are in Cu/S ratio. After Potter, 1977.
T^ °C Covellite Anilite High digenitt Djurleite Low chalcocite High chalcocite
507 ±2 1.000 ± 0.001 - 1.732 ±0.005 - - -
43S ±8 - - 2.000 ± 0.0038 - - 2.000 ± 0.0038
103.5 ± 0.5 - - - - 2.000 ± 0.002 2.000 ± 0.002
93 ±2 - - 1.834 ±0.002 1.942 ±0.002 - 1.988 ±0.005
90 ±2 - - - 1.960 ± 0.002 1.993 ±0.002 1.990 ± 0.005
75 ±2 1.000 ± 0.001 1.750 ± 0.003 1.765 ± 0.002 - - -
72 ±2 - 1.750 ± 0.003 1.805 ± 0.002 1.934 ± 0.002 - -
Table 4. Chemical composition (weight-%) of copper sulfide at sample S40. The values are from random points at the bulk of the biggest grains of copper sulfide in Figure 10.
Cu Fe Ca S U O Si Zn Total
78.85 0.03 0.01 20.68 0.13 0.37 0.03 0.04 100.14
78.90 0.02 0.00 20.66 0.18 0.28 0.03 0.04 100.10
79.27 0.02 0.02 20.13 0.10 0.34 0.02 0.05 99.93
79.34 0.01 0.02 20.60 0.00 0.27 0.01 0.03 100.27
79.42 0.00 0.01 20.29 0.00 0.34 0.05 0.04 100.15
79.18 0.04 0.02 20.68 0.00 0.21 0.02 0.07 100.23
79.85 0.01 0.01 20.75 0.00 0.47 0.01 0.01 100.11
77.38 0.04 0.02 20.75 0.00 0.93 0.09 0.01 99.22 20
FIG. 10. Sample S40, Dh 304. Reflected light. Parallel nicols. Native copper (pink) and copper sulfide, djurleite (blue).
4.4 Native copper composition
The data of microprobe analyses of native copper (Table 5) was normalized to 100%. Thus, the maximum Cu content obtained is 99.35%. These results are affected by the the background 'noise' produced by the surrounding silicate minerals, and the small grain size of the analysed copper (Fig. 10). Otherway, the composition of native copper is estimated to be about 99.99% Cu.
Table 5. Chemical composition (weight-%) of metallic copper at sample S40. The values are from random points at native copper grains.
Cu Fe Ca S U O Si Zn Total
98.63 0.05 0.01 0.00 0.00 0.60 0.04 0.04 99.37
98.20 0.03 0.02 0.02 0.08 0.45 0.04 0.00 98.84
96.72 0.09 0.02 0.00 0.06 2.35 1.01 0.07 100.32
95.57 0.08 0.04 0.02 0.00 2.45 1.21 0.04 99.41 21 4.5 Fracture surface mineralogy and uranium distribution
Figure 9b shows the surface of an open fracture at about 8.2 - 8.3 m depth in Dh 304.
The situation of this sample is very close to that of S40 (Figs. 7a and 9a), and in the same fracture zone. Almost all the fracture surface is covered by dust-like material (gummites and clay minerals). A study of this sample under stereomicroscope revealed the existence of native copper and copper sulfides at least in the zone marked by the square. Further information of this zone was obtained using the electronprobe micro- analyser.
The x-ray spectrum in Figure lib shows that the main mineral at the fracture surface is a silicate, rich in aluminium and magnesium, most possibly a clay mineral of the mont- morillonite group (smectites). Calcite, CaCO3 (Fig. 13a) is also a common mineral in the fracture surface. The lighter zones in the backscattered secondary electron (BSE) image (Fig. lla) correspond to concentration of heavy materials, mostly metals. The x-ray spectrum of a lighter zone (Fig. 11) reveals the presence of metallic copper Cu and uranium. The latter is strongly adsorbed in a silicate phase (a clay of the illite group), and may form a mineral of its own, which probably could be amorphous uranophane (Ca-U silicate). Crystalline uranophane occurs in Dh 303 as a void filling (e.g. Fig. 6).
Figures 13b and 13c show copper sulfide grains. The Cu/S ratio was found to be greater than 1.9, indicating that the mineral phase could be djurleite as in sample S^. Figure 12c also shows a grain or aggregate of a uranium rich calcium-alumine silicate. Calcite occurs in the darker zone between the copper sulfide grain and the U-rich silicate. The x-ray spectrum at the center of the copper sulfide grains showed a small peak of uranium, indicating that uranium is not significantly absorbed in the bulk of copper sulfide grains, but adsorbed around the grain boundaries, as it is also shown by the results of cx-particle autoradiography (Fig. 14).
Figure 15a shows the BSE image of a heavy mineral phase. The corresponding x-ray spectrum (Fig. 15b) shows relative high peaks of sulfur and copper, indicating that copper sulfide is probably underlying the U-rich silicate and that the maximum thickness of the U-rich clay layer is about 0.5 um (depth from which x-rays emerge).
The x-ray spectrum in Figure 16b shows the composition of a disturbed (Fig. 16a) zone of the main silicate at the fracture. The occurrence of uranium compound/s causes high a-track density, which disturbs the otherway regular appearance (dark grey to black) of the main silicate in the BSE image. 22
• o.o 5.0 10.0 15.0 20.0 keV
FIG. 11. a) Fracture surface (see Fig. 9b). Backscattered electron (BSE) image, b) X-ray spectrum at the center of the image. 23
o.o 15.0 20.0
FIG. 12. a) Backscattered electron (BSE) image. Left side of Fig. lla. b) X-ray spectrum at the center of the image. 24
FIG. 13. a) BSE images, a) Calcite crystals at the fracture surface, b) Grain of copper sulfide. c) U-Ca-rich silicate at the center of the image. Dj (Djurleite) Ca (Calcite). 25
FIG. 14. Sample S40, Dh 304. a) Transmitted light. Parallel nicols. Dj (Djurleite). b) The corresponding a-particle autoradiography. Width of field = 7 mm. 26
i o.o 15.0 20. C
FIG. 15. a) A heavy mineral phase, BSE image, b) X-ray spectrum at the center of the image. 27
Si
15.0 20.0
- b
FIG. 16. a) BSE image, b) X-ray spectrum at the center of the image. 28 5 THERMODYNAMICS OF COPPER CORROSION AT HYRKKOLA
Two different corrosion processes formed cuprite (Cu2O) and djurleite (Cu1]934S) after native copper at Hyrkkola. The processes are neither related in space nor in time, and may be considered separately.
5.1 Corrosion in sealed fractures: hydrothermal cuprite.
Cuprite is a mineral of the assemblage native copper — secondary uranium phases — cuprite. It occurs in sealed microfractures that have not been in contact with recent groundwater. The occurrence of this assemblage in the microfractures of a microbrec- ciated granite pegmatite suggests the precipitation of the mineral phases at two stages. At a first stage, soon after crystallization and fracturing, native copper might fill the fractures. At a second stage, residual hydrothermal fluids with a minimum oxidation potential of about +150 mV, and under acidic pH (Fig. 17), oxidized native copper to 2+ cupric ion, Cu , then to cuprite (Cu2O) from the borders to the centre of the fractures. Secondary uranium compounds are present as a sorbed phase in cuprite.
1.2 ^\^ SYSTEM Cu-O-H 1.0 25°C, 1 bar -
0.8 - -
0.6 \ \^ 0.4 - CuO
I _ -0.2 \ a 0.0 CU -0.2 - ^J\ \ \ -0.4
-0.6 - - \
_r> a Ill | t 1 2 4 6 8 10 12 14 PH Fig. 17. Eh-pH diagram for part of the system Cu-O-H after Brookins (1988). The assumed activity for dissolved Cu = 10"6. 29 5.2 Corrosion in open fractures: djurleite
The mineral assemblage native copper — djurleite — uranyl compounds occurs only in
Dh 304 at an open fracture in a zone at a depht of 8.1 - 8.3 m. Djurleite (Cu, 934 S) is a copper sulfide of low temperature (Potter, 1977) of supergene origin after sulfidization of native copper.
From the various corrosion processes, which may affect the stability of metallic copper in the deep bedrock (e.g. Ahonen, 1995), sulfidization, i.e., a process in which metallic copper reacts with dissolved sulfide forming a solid sulfide phase, is considered as the most important one at Hyrkkola. The secondary sulfidization of native copper in frac- tures is dependent on the availability of hydrogen sulfide (HS) in the groundwater. Water acts as an electron acceptor (oxidant) in the corrosion reaction (3) through (1) and (2):
+ 1.934Cu° + HS" = Cu1934S + 2e + H (1)
2H+ +2e = H,(aq) _____ (2)
+ 1.934Cu° + HS + H = Cu1934S + H2(aq) (3)
The algorithm in Appendix 5 allows the calculation and representation of the data in Figures 18 and 19. Calcite tends to dissolve at pH < 7.8 (25 °C) (Mason, 1966). Thus, the occurrence of calcite crystals at the fracture surface (Fig. 13) may indicate a mini- mum pH value of about 7.8 for circulating groundwaters in the fracture. At 25 °C, and a pH ~ 7.8 the minimum concentration of HS for the stability of djurleite is of the order of 105 M - 10"4 g/1 (Fig. 18a). At these same values of temperature and pH, and a [HS] = 106 M, the stability limits of djurleite lie out of the range at which water can act as oxidant (Fig. 19a). At lower temperatures (Figs. 18b and 19b) and the same pH, the stability limits of djurleite lie in the range at which water can act as oxidant even at a concentrations of HS" of the order of 106 M ~ 10"5 g/1. 5 : 2S*C 1 a) CS(aq)l= .00001 \^ 0 •: Cu° i | 1 PE -5 -.
I ^~^~;^ Cu0 15 J -15 3 1 i l i • i e 10 12 14 PH 5 1 5 ; b) 5*C j [S(aq)3= .000001 0 •: PE -5 -. ^^^ Cu1Ms 10 -_ cU2s ' ^>^^ ^^ -^ -15 J 15 J • i • i 8 10 12 14 PH C FIG. 18. Stability fields of metallic copper and different copper sulfides. a) T = 25 °C, FIG. 19. Stability fields of metallic copper and different copper sulfides. a) T = 25 C, 5 6 b) T = 5 °C. Activity of dissolved sulfur [S] = 10 M (either as HS or SO4"). Dotted lines b) T = 5 °C. Activity of dissolved sulfur [S] = W M (either as HS or SO/). Dotted lines indicate the pE-range at which water can act as an oxidant, upper limit ([H2] = 10"*) indicate the pE-range at which water can act as an oxidant, upper limit ([HJ = 10"*) corresponds to the estimated lowest concentration of H2(aq), lower limit corresponds to the corresponds to the estimated lowest concentration of H2(aq), lower limit correspond to the Hj(aq) activity of 1. H2(aq) activity of 1. 31 The relationship between pE and Eh in Figures 18 and 19 is given by: Eh = 2.3*R*T*pE/nF, where F is the Faraday constant, R the gas constant, and T the absolute temperature. At 25 °C, Eh = 59.155*pE (mV), and at 5 °C, Eh = 55.164*pE (mV). Ferric iron has been also considered as a possible oxidizer of metallic copper (e.g., Wersin et al., 1994, Ahonen, 1995). At Hyrkkola, iron compounds do not form part of the mineral assemblage at the fracture surface. Thus, the secondary sulfidization of metallic copper has not been dependent on the small-scale availability of ferric oxy- hydroxides. On the contrary, the previous coexistence (coprecipitation of hematite and native copper at a primary mineralization or at a hydrothermal stage) of hematitic pigments and native copper may prevent the latter from sulfidization at the near fracture surface in the following way: Fe2O3 + 3H2O = 2Fe(OH)3 (4) + 2+ Fe(OH)3 + 15/8H + 1/8HS = Fe + 1/8SO/ + 5/2H2O (5) Ferric oxyhydroxide, which may be formed after hematite (reaction 4) readily oxidizes dissolved hydrogen sulfide to sulfate (reaction 5). The most common sulfide minerals, pyrite and pyrrhotite, have been evaluated as a source of hydrogen sulfide (Ahonen, 1995). At Hyrkkola, pyrite is present in the bulk of amphibolites and gneisses in a maximum amount of 3% and pyrrhotite in amounts no greater than 1%. These minerals do not occur in the studied fracture surfaces. So far, they have not been observed in other possible water-conducting fractures near the fractures, where djurleite is present. Moreover, the minimum amount of sulfide (10'5M at 25 °C and 10"6M at 5 °C) needed, for copper corrosion to happen, fits well with the average content of sulfide in natural waters (e.g., Pitkanen et al., 1992), and no addi- tional source of sulfide would be necessary to call for. 32 6 FEATURES IN COMMON WITH PROPOSED REPOSITORY SITES The Hyrkkola U-Cu mineralization is located in a crystalline, metamorphic bedrock (Fig. 1, Appendices 2 to 4) comparable to the bedrock in the site(s) planned by Posiva Oy for nuclear waste disposal (e.g., Okko et al., 1990). The repository conditions at a 300 - 800 depth are expected to be reducing (e.g., Werme, 1990, Werme et al., 1992, Pitkanen et al., 1992, 1996). At Hyrkkola, the unique copper corrosion product observed is copper sulfide, djurleite (Cu1934S) formed presumably under reducing conditions. Anoxic conditions may also be encountered at near-surface environments (Sato, 1992). So far, the mineral associa- tion of native copper and its corrosion product, djurleite (Cu, 934S), is present only in open fractures. The presence of the assemblage native copper-copper sulfide in an open fracture surface (probably in contact with continously or intermittently flowing groundwater) offers the possibility to study the water-rock interaction affecting directly the stability of native copper and its corrosion products. Moreover, a study on the occurrence of uranyl compounds at the fracture surface at 8.3 m depth shows that the sorption properties of the engineered barrier (native/metallic copper) are negligible with respect to the sorption properties of the engineered barrier corrosion products (copper sulfide). Figure 20 shows a copper canister scenario in the spent fuel repository. For some features it is comparable to the copper corrosion scenario at Hyrkkola (Fig. 8). The composition of groundwater in the copper canister scenario is shown in Table 6. Groundwater conditions at Hyrkkbla are not precisely known but they can be est- imated from the mineral assemblies at the fracture surface. Some groundwater data from the Palmottu area about 10 km SW from Hyrkkola (Fig. 1) is given in Appendix 6. Eh values from Palmottu hint at slightly reducing conditions, but data on [HS] is not available. The presence of secondary uranium phases (uranyl com- pounds) indicates alkaline and basic conditions (Finch and Ewing, 1990). The formation of clays of the montmorillonite and/or illite group after alteration of feldspars at low temperatures and pressures is favoured also by alkaline and basic conditions (Deer et al., 1962). Both types of clay occur at the open fracture surfaces. The formation of djurleite after native copper is also promoted in reducing and basic conditions (Figs. 17 and 18), and the calculated minimum average content of hydro- gen sulfide HS" was between 10s and 10"6 g/1. These groundwater conditions are not exceptional in granitic groundwaters (e.g. Pitkanen et al., 1992, Blomqvist et al., 1992, p. 44). Tmax=IOO°C Pmax= !5MPa Eh= -350..-200mV pH=7.3..IO.O [S2tot]max= 3 mg/I Corrosion processes: + 500 m 2Cu°(»)+HS"+H = CU2S(s) + HJ2(g) Canister life: > 100 000 a Water flow: u> 0,1 dm3m'2a'' Water chemistry: in Table 6 Filling material Bentonite Copper canister Steel canister FIG. 20. Spent fuel repository concept of Posiva Oy (TVO, 1992)(Figure not to scale). Chemical data from Posiva detailed site investigation phase (Ruotsalainen and Snellman, 1996). EhmM value based on the interpretation of sulphidic redox conditions (e.g. Pitkanen et al., 1996) and Ehmin value based on available data from the detailed site investigations. Canister data from Posiva (Salo, 1996). 34 Table 6. Chemical composition of natural reference groundwaters and bentonite water compiled by M. Snellman (Appendix 4 of Pitkanen et al., 1992). Fresh gw Brackish gw Saline gw Synthetic bentonite Sample Kivetty KR1 Olkiluoto KR1 Olkiluoto KR1 water Depth m 169 140 613 Eh mV -350 -270 -300 PH 7,8 8,1 8,9 8,5 Conductivity mS/m 19,4 265 3000 Alkalinity mekv/l 2 4,18 0,4 9,8 SiO2 mg/l 8 11,8 3,1 16 TOC 5,8 n.a. 5,9 Ca mg/l 22,2 110 3275 18 Na mg/l 9,5 565 3900 272 Mg mg/l 5,4 37,4 49,3 4,3 K mg/l 2,4 6,7 19 3,9 Fe(tot) mg/l 0,08* 0,05* 0,38* 0,3 Fe(ll) mg/l o,r 0,06* 0,3* Al mg/l 0,15 0,057 0,18 Mn mg/l 0,18 0,1 0,39 Ba mg/l 0,17 <0,001 0,545 Sr mg/l 0,14 1,14 29,9 NH4 mg/l <0,03 0,07 <0,03 a mg/l 1,46 910 13000 80 F mg/l 2,32 0,55 1.1 7,5 Br mg/l 0,05 3,4 95 NO3 mg/l <0,1 <0,1 <0,1 NO2 mg/l <0,1 <0,1 0,2 PCM mg/l <0,1 <0,1 0,14 0,4 S(-ll)tot mg/l 0,07 - 0,12 SO4 mg/l 0,05 198 0,14 59,6 H-3 TU <6,1 n.a. <8 U(tot) ug/l 1,9 - 0,34 Flush.water % 0,3 9,4 1,6 Charge Bal. % -2,67 -1,03 -4,29 -0,53 TDS mg/l 200 2000 20000 1000 n.a.= not analyzed *Fe(ll) analyzed in the field, Fe(tot) analyzed in the laboratory 35 7 CONCLUSIONS The Hyrkkola U-Cu-mineralization is evaluated as natural analogue for copper canisters. Petrological and mineralogical studies show that the circumstances (rock types, mineralogy and its geochemical implications) in which the native copper has been preserved and/or changed are similar to the circumstances predicted for the final nuclear waste repository planned by Posiva Oy in the Finnish bedrock. The presence of native copper and copper sulfide at open fractures in the near- surface (8.3 m depth) implies the existence of reducing conditions, the which ones are also predicted to prevail between 300 and 800 m depth in granitic bedrock. This finding allowed to the study of the native copper corrosion process in anoxic condi- tions. It has been estimated that the formation of copper sulfide (djurleite, Cux 934S) after native copper (Cu°) under anoxic (reducing) conditions is enhanced by the availabili- ty of dissolved HS' in continously or intermittently circulating groundwaters. The minimum quantity of dissolved HS" has been calculated to be of the order of 10"4 to 10"5 g/1 (Eh of about -350 to -248 mV), at temperatures of 25 °C and 5 °C respec- tively, and a pH of about 7.8. It was also found that the intimate relation of native copper and ferric compounds (hematitic pigment) may inhibit the oxidation of the former under reducing condi- tions. That is, ferric oxides and native copper in contact form a galvanic pair where hematite act as anode, and is preferentially corroded. Secondary uranium compounds were observed to be significantly adsorbed at the djurleite grain boundaries. Soption phenomena was not significant in the bulk of copper sulfide grains nor in the bulk or boundaries of native copper grains. Cuprite does not act as a notable adsorbent of uranium. The present study is the first one that has been performed on findings of native copper in reducing, neutral to slightly alkaline groundwater conditions. Thus, the data obtained is of most relevance in improving models of anoxic corrosion of copper canisters. At Hyrkkola, further hydrogeochemical studies would provide an unique opportunity to test the thermodynamic models (and their associated databases) used for copper corrosion analysis in performance assessment. An important co-product of these studies could be the determination of aqueous speciation of radionuclides and the chemical and physical properties of the mineral-water system. 36 REFERENCES Ahonen L., 1995. Chemical stability of copper canisters in deep repository. Hel- sinki, Nuclear Waste Commission of Finnish Power Companies, Report YJT-95- 19. Blomqvist R., Pilvio R., Ahonen L., and Ruskeeniemi T., 1992. Groundwater conditions and uranium mobility in Palmottu analogue study site in Finland. In The Palmottu Analogue Project, Progress Report 1991, (eds. J. Suksi, L. Aho- nen, and H. Niini). Geological Survey of Finland, Nuclear Waste Disposal Research, Rep. YST-78, 32 - 67. Brookins D.G., 1988. Eh-pH diagrams for geochemistry. Springer-Verlag. Deer, W.A., R.A. Howie, and J. Zussman 1962: "Rock-Forming Minerals," vol. 3 "Sheet Silicates," Longmans, Geen &Co., Ltd., London. Engman U. and Hermansson H-P., 1994. Corrosion of copper materials for en- capsulation of radioactive waste - a literature study (in Swedish). SKI Report 94:6, Stockholm. April 1994. Evans A.M., 1980. An introduction to ore geology. Elsevier. Finch R. and Ewing R., 1990. Uraninite alteration in an oxidizing environment and its relevance to the disposal of spent nuclear fuel. Stockholm, Swedish Nuclear Fuel and Waste Management Co (SKB), Technical Report 91-15. Jaakkola T., Suksi J., Suutarinen R., Niini H., Ruskeeniemi T., Soderholm B., Vesterinen M., Blomqvist R., Halonen S., and Lindberg A., 1989. The behaviour of natural radionuclides in and around uranium deposits. 2. Results of investiga- tions at the Palmottu analogue study site, SW Finland. Geological Survey of Finland, Nuclear Waste Disposal Research, Report YST-64, 60 p. Johanson B. and Kojonen K., 1995. Improved electron probe microanalysis serv- ices at Geological Survey of Finland. Geological Survey of Finland, Special Paper 20, 181 - 184. Korzhinskii D.S., 1959. Physicochemical Basis of the Analysis of the Paragenesis of Minerals, New York: Consultants Bureau. 37 Kouvo O., 1983. The radiometric age of Hyrkkola uraninite. Internal report. Geo- logical Survey of Finland, 3 p. (in Finnish). Kuivamaki A., Paananen M., and Kurimo M., 1991. Structural modelling of bedrock around the Palmottu U-Th-deposit, progress report 1990. Geological Survey of Finland, Nuclear Waste Disposal Research, Report YST-72. Espoo, 1991. Marcos N., 1989. Native copper as a natural analogue for copper canisters. Hel- sinki, Nuclear Waste Commission of Finnish Power Companies, Report YJT-89- 18. Mason B., 1966. Principles of geochemistry. John Wiley & Sons, Inc. New York. Miller W., Alexander R., Chapman N., McKinley I., and Smellie J., 1994. Natural Analogue Studies in the Geological Disposal of Radioactive Wastes. NAGRA, Technical Report 93-03, January 1994. Niini H., 1968. A study of rock fracturing in valleys of Precambrian bedrock. Fennia 97(6), 60 p. Niini H., 1987. Bedrock fractures affecting land uplift in Finland. Geological Survey of Finland, Special Paper 2. p. 51 - 54. Okko V., 1964. Maapera. In Suomen geologia, (ed. K. Rankama), Kirjayhtyma, Helsinki, p. 239 - 332. Okko O., Front K., Hassinen P., and Vaittinen T., 1990. Geofysikaalisten rei- kamittausten tulkinta Olkiluodon kairanreiat KR1, KR2 ja KR3. TVO/Paik- katutkimukset/Site investigations, TyoraportuVWork Report 90-08. (Abstract in English). Pitkanen P., Snellman M., and Leino-Forsman H., 1992. Modelling of water rock interaction at TVO investigation sites - Summary report. Nuclear Waste Com- mission of Finnish Power Companies, Report YJT-92-30. Pitkanen P., Snellman M., and Vuorinen U., 1996. On the origin and chemical evolution of groundwater at the Olkiluoto site. Posiva report, POSIVA-96-04. 38 Potter R.W., 1977. An electrochemical investigation of the system copper-sulfur. Econ. Geol. 72, p. 1524 - 1542. Romberger S.B., 1984. Transport and deposition of uranium in hydrothermal systems at the temperatures up to 300°C: geological implications. In Uranium geochemistry, mineralogy, geology, exploration and resources (eds. B. DeVivo, F.I.G. Capaldi and P.R. Simpson). Institution of mining and metallurgy, 12 - 17. Rose A.W. and Bianchi-Mosquera G.C., 1993. Adsorption of Cu, Pb, Zn, Co, Ni, and Ag on goethite and hematite: A control on metal mobilization from red beds into stratiform copper deposits. Econ. Geol. 88: 1226 - 1236. Ruotsalainen P. and Snellman M., 1996. Hydrogeochemical baseline character- ization at Romuvaara, Kivetty and Olkiluoto by Posiva Oy, Finland (to be pub- lished). Raisanen E., 1986. Uraniferous granitic veins in the Svecofennian schist belt in Nummi-Pusula, Southern Finland. Technical Committee Meeting on Uranium Deposits in Magmatic and Metamorphic rocks. Report IAEA-TC-571, 12 p. Salo J-P., 1996. Canister design report (to be published). Sato M., 1992. Persistency-field Eh-pH diagrams for sulfides and their application to supergene oxidation and enrichment of sufide ore bodies. Geochimica et Cos- mochimica Acta, 56: 3133 - 3156. Schreurs J. and Westra L., 1986. The thermotectonic evolution of a Proterozoic, low pressure, granulite dome, West Uusimaa, SW Finland. Contrib. Mineral. Petrol. 93: 236 - 250. Simonen A., 1980. The Precambrian of Finland. Geol. Surv. Finland, Bull. 304, 58 p. Snellgrove R.A. and Barnes L.H., 1974. Low temperatures solid phases and aqueous species in the copper sulphide system (abst.): Trans. Am. Geophys. Union, v. 55, 484. 39 TVO, 1992. Final disposal of spent nuclear fuel in Finnish bedrock - Technical plans and safety assessment. Nuclear Waste Commision of Finnish Power Companies, Report YJT-92-31E, 1992. Werme L., 1990. Near-field performance of the advanced cold process canister. Helsinki, Nuclear Waste Commission of Finnish Power Companies, Report YJT-90-20. Werme L., Sellin P. and Kjellbert N., 1992. Copper canisters for nuclear high level waste. Corrosion aspects. SKB Technical Report 92-26, Stockholm, Octo- ber 1994. Wersin P., Spahiu K. and Bruno J., 1994. Kinetic modelling of bentonite-canister interaction. Long-term predictions of copper canister corrosion under oxic and anoxic conditions. SKB Technical Report 94-25, Stockholm, September 1994. Appendix Explanation to the Appendices 2 to 4.3 Rock type Fracture density (1 - 8 fractures/m) 120 m - CO c Cu o U _ Cu • 11 * 1111 •\V\\\\\X\V Cu Observation of native copper S31 Location of sample Scale 1 : 250 Amphibolite Q-F Gneiss Granite Quartz Chi or. •p+p Soil Appendix 2 Hyrkkbla DH 301 0 m 20 m - 40 m - 60 m - I I I I I I I I I I I I I I I I I I I I I I I I i i' i' i' i' i' vi i 'i • r ii (Vr- Dh 303 (continuation) B0 m - 100 m Cu 120 m - - Ss, and S, Dh 305 (continuation) Appendix 3.2 II I \v 80 m - t \l 100 m - ^^^1 120 m - ••. •«•'*. N. N. Dh 305 (cont.) T/9' . s \ V\ •\Y\v\v \ w \ \ V ' \V\\\ \\\J R KV x \> 160 m -i vV 180 m A C. Dh 303 (cont.) Appendix 3.3 Cu S3, and SM 140 m XV\X\NXXXXj 160 m - ^xxxxxxxN] Dh 303 (cont.) 200 m - 180 m - x\ 220 m - -fiESL. 2L Appendix 4. Hyrkkbla DH 304 0 m 20 m - 40 m - «m 60 m - Appendix 4.2 Dh 304 (continuation) iilis 80 m -h •:^i ..V 100 m -K 120 m -t Dh 304 (cont.) Appendix 4.3 140 m 160 m - 180 m - Appendix 5(1) 10 REM APP5 2.5.96 20 REM Figures 18 and 19 30 KEY OFF:CLS 40 REM This code was used in construction of diagrams 18 to 19 50 R=8.314OOOE.O3' kJ/mol 60 INPUT "Name of output file:";FTLE$ 70 OPEN FILES FOR OUTPUT AS 1 80 INPUT "Temperature in Celsius:";TEMP 90 T=TEMP+2.731499E+02 100 E=2.302999*R*T/9.648699E+01*1000>Eh in mV 110PRINT"EhinmV=";E 120 INPUT "Total dissolved sulfide in moles/liter:";STOT 130 HS=LOG(STOT)/LOG(10) 140 PRINT "Logarithm of dissolved S =";HS 150 H2S=HS ' each of these is the only and predominant 160 SO4=HS ' form considered, except if 170 EQSULF=HS-LOG(2)/LOG(10) ' two species have equal activities 180 REM Reading data of table A5 190 FOR K=l TO 4:READ N(K),M(K),LGK(K),DH(K):NEXT K 200 READ LGK(5),DH(5) 210 READ LGK(6),DH(6) 220 READ LGK(7),DH(7) 230 REM table A5 240 REM n hs-, m e-, stability constants, and Enthalpies 250 DATA 1, 2, 17.05, -63.1 260 DATA 0.034, 0.068, 0.27, -1.18 270 DATA 0.105, 0.21, 0.78, -2.34 280 DATA 0.75, 1.5, 4.36, -3.73 290 DATA -7, 22.3 300 DATA 33.7, -168 310 DATA -3.17, -7.4 320 REM Temperature correction of stability constants 330 T0=2.981499E+02 340 FOR K=l TO 7 350LGK(K)=-DH(K)/(2.302999*R)*(l/T-l/T0)+LGK(K) 360 NEXT K 370 REM Four boundaries between five phase 380 REM Cu metal/chalcocite/djurleite/anilite/covellite are calculated 390 FOR K=l TO 4 400 REM Equation (phasel + nHS" = phase2 + nH* + me, log K) 410 PH=-LGK(5): PH(3,K)=PH 420PE(3,K)=-LGK(K)/M(K)-N(K)/M(K)*PH-N(K)/M(K)*HS 430 REM sulfide speciation at low pH, equation pE=-(LGK+n*pH+n*LG[HS])/m 440 PH=4: PH(4,K)=PH: PH(1,K)=PH 450PE(4,K)=-LGK(K)/M(K)-N(K)/M(K)*PH-N(K)/M(K)*(LGK(5)+PH+H2S) - 460 REM Equation (phasel + nHS = phase2 + nH* + me, log K) & H2S(aq) = H* + US 470 TERM=M(K)-8*N(K)'just to simplify coding 480PE(l,K)=(-LGK(K)-N(K)*PH-N(K)*LGK(6)+9*N(K)*PH-N(K)*SO4)/(M(K)-8*N(K)) 490 REM print 60*pe(l,k),60*pe 500PH=(LGK(K)/M(K)+N(K)/M(K)*EQSULF-LGK(K)/TERM-N(K)/TERM*LGK(6)- N(K)/TERM*EQSULF)/N(K)/TERM-9*N(K)/TERM-N(K)/M(K)) 510EFPH>14THEN550 520PE(2,K)=(-LGK(K)-N(K)*PH-N(K)*LGK(6)+9*N(K)*PH- N(K)*SO4)/M(K)-8*N(K)) 530 PH(2,K)=PH 540 GOTO 640 550 PH=14 560PE(2,K)=(-LGK(K)-N(K)*PH-N(K)*LGK(6)+9*N(K)*PH-N(K)*SO4)/M(K)-8*N(K)) 570 PE(3,K)=-LGK(K)/M(K)-N(K)/M(K)*PH-N(K)/M(K)*HS Appendix 5(2) 580 PH(2,K)=PH 590 PH(3,K)=PH 600 PH=-LGK(5):PH(4,K)=PH 610PE(4,K)=-LGK(K)/M(K)-N(K)/M(K)*PH-N(K)/M(K)*HS 620 PH=4:PH(5,K)=PH 630PE(5,K)=-LGK(K)/M(K)-N(K)/M(K)*PH-N(K)/M(K)*(LGK(5)+PH+H2S) 640 NEXT K 650 REM printing results to a file 660 FOR K=l TO 4 670 FOR L=l TO 5 680 "IF PH(L,K)>0 THEN PRINT#l,USING "#####.###";PH(L,K),PE(L,K)(E*PE(L,K) 690 PRINT#1,USING "#####.###";PH(L,K),PE(L,K),E*PE(L,K) 700 NEXT L 710 NEXT K 720 SCREEN 9 730 WINDOW (0,-20)-(18,10) 740 LINE (4,-16)-(14,5),,B 750 FOR 1=4 TO 14 STEP 2 760 LINE (I,-16)-(I,-1.650000E+01) 770 LOCATE 32,I/18*80:PRINT USING"##",I; 780 NEXT I 790 FOR 1=5 TO 13:LINE(l,-16)-(I,-1.625000E+01):NEXT I 800 FOR 1=15 TO 5 STEP 5 810 LINE(4,I)-3.799999,1) 820 LOCATE 2.550000E+01-25*(I+20)/30,13:PRINT USING"###":I: 830 NEXT I 840 FOR I=-1.450000E+01 TO 4.500000 STEP 5.000000E-01:LINE(4,I)-(3.900000,I):NEXT I 850 LOCATE 24,40:PRINT"pH";:LOCATE 12,7:PRINT"pE";:LOCATE 1,1 860 LOCATE 6,48:PRINT USING "###0C";TEMP; 870 LOCATE 7,48:PRINT "fS(aq)]=";STOT; 880 FOR K=l TO 4 890 PSET (PH(1,K),PE(1,K)) 900 FOR L=2 TO 5 910 IF PH(L,K)>0 THEN LINE -(PH(L,K),PE(L,K)) 920 NEXT L 930 NEXT K 940HPH(l)=4:HPH(2)=14:HPH(3)=14:HPH(4)=4 950H2(l)=-9:H2(2)=H2(l):H2(3)=l:H2(4)=H2(3) 960 PRINT* 1, "Lower stability of water" 970 FOR K=l TO 4 980HPE(K)=LGK(7)/2-HPH(K)-H2(K)/2 990 PRINT#1,USING"#####.###";HPH(K),HPE(K),E*HPE(L,K) 1000 NEXT K 1010 LINE (HPH(1),HPE(1))-(HPH(2),HPE(2)),,,9 1020 LINE (HPH(3),HPE(3))-(HPH(4),HPE(4)),,,9 1020 A$=INKEY$:IF A$="" THEN1030 1031 SCREEN 0,0,0 1040 END Appendix 6 Table A6.1. Uranium in groundwater from four different drill holes. Uranium concentrations are expressed as weight/vol; for paniculate fraction (0 > 0.45 jam) uranium content in one liter of groundwater is given. The column (%) indicates the percentage proportion of U(VI) of the total uranium in sample. The given error in activity result is one sigma and is due to the standard deviation from radiactive maesurement. The last column indicates the Eh value calculated from U(IV)/U(VI) pair. Drillhole, Eh PH »u at,U % depth (mV) (mBq/1) 0*g/l1) (mV) 302, -55 8.42 U(VD 2282 ± 130 185 ± 11 2.12 ± 0.17 91 -57 90-95m U(IV) 66 ± 4 5.3 ± 0.3 2.28 ± 0.17 3 > 0.45pm 45 ± 3 3.6 ± 0.2 2.25 ± 0.17 324, +55 6.87 U(VI) 743 ± 12 60 ± 1 1.31 ± 0.03 95 +40 95-101m U(TV) 43 ± 1 3.5 ± 0.1 1.26 ± 0.04 5 > 0.45pm 10 ± 1 0.82 ± 0.02 1.20 ± 0.05 346, -11 8.05 U(VD 1183 ± 36 96 ± 3 1.12 ± 0.05 93 -41 65-71m U(IV) 83 ± 3 6.8 ± 0.2 1.13 ± 0.05 7 > 0.45pm 1.74 ± 0.14 0.14 ± 0.01 1.42 ± 0.16 346, -70 8.40 U(VI) 116 ± 7 9.4 ± 0.6 4.25 ± 0.33 94 -65 122-128m U(IV) 7.6 ± 0.6 0.62 ± 0.05 3.30 ± 0.29 6 > 0.45pm 0.31 ± 0.10 0.03 ± 0.01 4.03 ± 1.50 346, -80 8.40 U(VI) 45 ± 3 3.7 ± 0.2 3.83 ± 0.29 40 -106 122-128m U(IV) 68 ± 31 5.5 ± 2.5 4.04 ± 2.41 60 > 0.45pm 0.56 ± 0.13 0.05 ± 0.01 2.50 ± 0.67 346, -92 9.05 U(VI) 1.33 ± 0.12 0.11 ± 0.01 1.72 ± 0.87 3 -198 240-246m U(IV) 18 ± 1 1.43 ± 0.05 3.46 ± 0.14 97 > 0.45pm 0.42 ± 0.12 0.03 ± 0.01 2.23 ± 0.75 348, -70 8.82 U(VI) 20 ± 1 1.62 ± 0.06 2.30 ±0.11 93 -100 200-225m U(IV) 1.55 ± 0.12 0.13 ± 0.01 1.42 ± 0.15 7 > 0.45pm 0.45 ± 0.08 0.04 ± 0.01 1.39 ± 0.34 Table A6.2. Chemical composition (mM) of fracture waters. DH Depth (m) Na K Ca Mg Cl SO4 HCO, F Si Fe 302 90-95 0.97 0.03 0.45 0.18 0.04 0.31 1.4 0.019 0.16 <0.001 324 95-101 0.6 0.04 0.77 0.19 0.12 0.18 1.8 0.007 0.19 0.048 346 240-246 17 0.11 0.65 0.19 2.1 8.0 1.0 0.06 0.13 <0.001 348 200-225 20 0.08 0.70 0.22 1.9 8.7 1.0 0.05 0.10 0.001 Reference: Ervanne H., Ahonen L., Jaakkola T., and Blomqvist, R., 1994. Redox chemistry of uranium in groundwater of Palmottu uranium deposit, Finland, In The Palmottu Analogue Project, Progress Report 1993, (eds. T. Ruskeeniemi, J. Suksi, H. Niini, and R. Blomqvist). Geological survey of Finland, Nuclear Waste Disposal Research, Report YST-85, 37 - 48. LIST OF REPORTS 1(3) LIST OF POSIVA REPORTS PUBLISHED IN 1996 POSIVA-96-01 Determination of U oxidation state in anoxic (N2) aqueous solutions method development and testing Kaija Ollila VTT Chemical Technology June 1996 ISBN 951-652-000-6 POSIVA-96-02 Fault plane solutions of microearthquakes in the Loviisa region in south-eastern Finland Jouni Saari IVO International Ltd Ragnar Slunga Fftrsvarets Forskningsanstalt, Stockholm, Sweden June 1996 ISBN 951-652-001-4 POSIVA-96-03 Thermal optimisation of the final disposal of spent nuclear fuel Heikki Raiko VTT Energy June 1996 (In Finnish) ISBN 951-652-002-2 POSIVA-96-04 On the origin and chemical evolution of groundwater at the Olkiluoto site Petteri Pitkanen Technical Research Centre of Finland Margit Snellman Imatran Voima Oy Ulla Vuorinen Technical Research Centre of Finland June 1996 ISBN 951-652-003-0 POSIVA-96-05 Seismic emissions induced by the excavations of the rock repository in Loviisa Jouni Saari IVO International Ltd June 1996 ISBN 951-652-004-9 LIST OF REPORTS 2(3) POSIVA-96-06 Geochemical modelling study on the age and evolution of the groundwater at the Romuvaara site Petted PitkSnen Technical Research Centre of Finland Margit Snellman Imatran Voima Oy Ulla Vuorinen Hilkka Leino-Forsman Technical Research Centre of Finland September 1996 ISBN 951-652-005-7 POSIVA-96-07 Boring of full scale deposition holes using a novel dry blind boring method Jorma Autio Timo KirkkomSki Saanio & Riekkola Consulting Engineers November 1996 (to be published) ISBN 951-652-006-5 POSIVA-96-08 Production methods and costs of oxygen free copper canisters for nuclear waste disposal Harri Aalto Hannu RajainmSki Lenni Laakso Outokumpu Poricopper Oy November 1996 (to be published) ISBN 951-652-007-3 POSIVA-96-09 Characterization of the excavation disturbance caused by boring of the experimental full scale deposition holes in the Research Tunnel at Olkiluoto Jorma Autio Saanio & Riekkola Consulting Engineers December 1996 (to be published) ISBN 951-652-008-1 POSIVA-96-10 Gamma and neutron dose rates on the outer surface of the nuclear waste disposal canisters Markku Anttila VTT Energy December 1996 (to be published) ISBN 951-652-009-X POSIVA-96-11 Criticality safety calculations for the nuclear waste disposal canisters Markku Anttila VTT Energy December 1996 (to be published) ISBN 951-652-010-3 LIST OF REPORTS 3(3) POSIVA-96-12 Assessment of alternative disposal concepts Jorma Autio Timo Saanio Pasi Tolppanen Saanio & Riekkola Consulting Engineers Heikki Raiko Timo Vieno VTT Energy Jukka-Pekka Salo Posiva Oy December 1996 (to be published) ISBN 951-652-011-1 POSIVA-96-13 Design report of the canister for nuclear fuel disposal Heikki Raiko VTT Energy Jukka-Pekka Salo Posiva Oy December 1996 (to be published) ISBN 951-652-012-X POSIVA-96-14 Final disposal of spent nuclear fuel in the Finnish bedrock. Technology project 1993-1996. Posiva Oy December 1996 (to be published) (In Finnish) ISBN 951-652-013-8 POSIVA-96-15 The Hyrkkola' native copper mineralization as a natural analogue for copper canisters Nuria Marcos Helsinki University of Technology Laboratory of Engineering Geology and Geophysics October 1996 ISBN 951-652-014-6