Deterministic model for rock domains

Bolundsfjärden

5 Bedrock geology

5.4 Deterministic model for rock domains

5.4.1 Data input

The deterministic modelling of rock domains has primarily made use of the following data and interpretations:

• The distribution of rock units at the ground surface as presented in the bedrock geological map (section 5.2.1).

• The rock units identified in the single-hole interpretation of c. 15 km of bedrock extracted from twenty-one cored boreholes, complemented with the single-hole interpretation of thirty-three shallow percussion boreholes (section 5.2.2).

• The character of rock types in each rock unit as inferred in data both from the ground surface and from boreholes (section 5.2.3).

• The character and variation in the style of ductile deformation as inferred in data both from the ground surface and from boreholes (section 5.2.4).

5.4.2 Conceptual model

The bedrock inside the Forsmark tectonic lens is dominated by medium-grained granite. Between 1.87 and 1.86 Ga, both this granite and the surrounding rocks were affected by penetrative ductile deformation at mid-crustal depths and under high-temperature metamorphic conditions (/Söderbäck (ed) 2008/ and section 3.1). Amphibolite and fine- to medium-grained granitoid intruded syntectoni-cally as dykes and minor bodies during this time interval and, losyntectoni-cally, at least the amphibolites gave rise to conspicuous alteration (albitization) in the older granitic rocks. Ductile deformation with fold-ing continued to affect the younger intrusive rocks, includfold-ing amphibolite, under lower metamorphic conditions, prior to 1.85 Ga. Subsequently, until at least 1.8 Ga, the ductile strain continued to affect the bedrock, predominantly along the margins of the tectonic lens along discrete zones (/Söderbäck (ed) 2008/ and section 3.1). Borehole data indicate that the tectonic lens is a major geological structure that can be traced from the surface down to at least 1,000 m depth.

Folding of a ductile fabric and a generally lower degree of ductile strain are inferred to be present inside the lens, including the target volume in its north-western part. By contrast, the surrounding rocks are affected by a higher degree of ductile strain and a conspicuous WNW to NW structural trend. The ductile structures at the site are characteristic of regions where variable degrees of ductile strain, folding and stretching are intimately related during strong, progressive, non-coaxial deformation.

Compressive deformation in the Forsmark region was absorbed initially by dextral strike-slip shear along high-strain belts with WNW to NW trend, combined with shortening across these belts expressed by grain-shape fabric development. As the ductile strain progressed, these structures were folded at different scales. This transpressive deformation was accompanied by continuous extrusion of material in a south-easterly direction. Folding developed initially with a normal cylindrical shape, but, to variable extent, the folds were progressively drawn out in the stretching direction into tubular-shaped structures, i.e. sheath folds. A major sheath fold is inferred to be present inside the tectonic lens (Figure 5-23). This fold is synformal in character in the part of the lens occupied by the target volume (Figure 5-23).

The Group B metatonalite that is situated outside the local model volume in the south-eastern part of the Forsmark lens (Figure 5-2) is inferred to be a mega-xenolith inside the metagranite. The geochronological data (/Hermansson et al. 2008/ and section 3.1) support this concept. A younger

Group D granite body, which is situated in the archipelago north of the candidate area and outside the local model volume (see inset map in Figure 5-2), has been treated conceptually as a laccolith.

It shows a broad extension at the surface, but a rapidly decreasing extension at depth. This geological feature has not been drilled and remains as an unconfirmed concept.

The bedrock anisotropy at Forsmark, which was established at an early stage in the geological evolu-tion in the high-temperature ductile regime, has important implicaevolu-tions for an understanding of the spatial distribution of younger deformation zones at the site and for uncertainties in the modelling of such zones. These implications are discussed further in sections 5.5 and 5.9.

5.4.3 Methodology, assumptions and feedback from other disciplines

The conceptual model has provided the basis for the modelling of rock domains inside and around the target volume. Furthermore, geological data generated in connection with the surface mapping has formed the keystone in the modelling work.

Singö DZ Eckarfjärden

Forsmark DZ

Strongly curved (>90°), fold hinge line – eroded away

Surface intersection – eye-shaped structure

Mineral stretching lineation – shallower to south-east

Fold axis parallel to mineral stretching lineation

Figure 5‑23. Conceptual model for the development of sheath folding inside the tectonic lens at Forsmark, with fold axes sub-parallel to the mineral stretching lineation.

orientation of geological objects in the boreholes, only to the actual orientation of the boreholes, i.e. to the borehole deviation measurements /Munier and Stigsson 2007/. The absence of linear grain-shape fabric data at depth is an intrinsic weakness in the modelling work.

The following assumptions are inherent in the modelling procedure:

• The mean values of the orientation of planar ductile structures (tectonic foliation and tectonic banding) in each rock domain are assumed to provide an estimate of the orientation of the contacts between the domains.

• Rock domains that form isolated bodies of metamorphosed intrusive rocks at the ground surface are assumed to plunge downwards in approximately the direction of the mineral stretching lineation as flattened rod-shaped entities – these formed in a setting dominated by constrictional strain.

• The major rock domains at the ground surface with steeply dipping contacts, which are included in the regional model volume, are assumed to extend downwards to, at least, the base of this volume (–2,100 m).

• The rock domain composed of partly albitized and metamorphosed granite close to

Asphällsfjärden, which is restricted to the local model volume, is assumed to extend downwards to, at least, the base of this volume (–1,100 m).

• The rock domains outside the local model volume, which are conceptually handled as a mega-xenolith or laccolith, are modelled as gently dipping rock sheets and do not extend to the base of the regional model volume (Figure 4-15 in /Stephens et al. 2007/).

As in previous models, judgements concerning the confidence of existence of a rock domain both at the surface and at depth (–2,100 m in the regional model, –1,100 m in the local model) are provided for each domain. The assessment of confidence at the surface is coupled to the confidence in the geological map of the bedrock in the regional model area. The assessment of confidence at depth is coupled to the number and depth of intersections of a particular domain in the boreholes.

The documented properties of each rock domain include the character of the dominant and subordinate rock types, the degree of homogeneity, the character of alteration and metamorphism, and the character and orientation of the ductile structures. The properties of the dominant rock type in a domain include, for example, quartz content from modal analyses, density, and natural exposure rate from in situ gamma-ray spectrometry measurements. Quantitative estimates of the uncertainty in some of the properties, the source of the primary data used in the assignment of a particular property and the level of confidence in the assignment (high, medium or low) are all presented.

Integration with the modelling of thermal properties (see chapter 6) recognised the need for a better understanding of the thickness, the frequency of different thickness classes, and the orientation of subordinate rock types in the target volume, in particular the metamorphosed dykes and lenses that occur as amphibolite (/Stephens et al. 2007/ and section 5.2.3). This work has contributed to the stochastic simulations of the subordinate rock types in the target volume (see chapter 6). The defined rock domains were judged appropriate for the purposes of safety assessment and no feedback was obtained from SR-Can on this subject (see /SKB 2006b/).

5.4.4 Division into rock domains, geometries and property assignment The identification of rock domains was initiated at the surface and made use of the division of the bedrock during the bedrock mapping work into two different types of rock units (see Table 4-3 in /Stephens et al. 2007/ and section 5.2.1). These are defined on the basis of:

• The composition and to, some extent, the grain size of the dominant rock type.

• The degree of bedrock homogeneity in combination with the style and degree of ductile deformation.

The inferred distributions of rock domains at the surface in both the regional and local model areas are shown in Figure 5-24. Rock domains have been identified on the basis of all the combinations in the two types of rock unit described above. For example, the composition of the rocks in domains

RFM012 and RFM044 are similar to those in domain RFM029, but the rocks in domains RFM012 and RFM044 are affected by a higher degree of ductile strain with the development of a conspicuous SL-fabric compared with domain RFM029.

Apart from the identification of some rock units on the basis of the character and frequency of fractures, the principles for the division of rock units in the boreholes during the single-hole

interpretation followed those used during the surface mapping work. This permits a correlation with the rock domains as defined on the basis of the surface data (Table 4-4 in /Stephens et al. 2007/).

Rock domains along all the twenty-one cored boreholes used in model stage 2.2 are presented in Appendix 13 in /Stephens et al. 2007/.

Fourteen rock domains have been recognised inside the local model volume. Most of these domains are situated outside the tectonic lens and target volume, on the south-western or north-eastern limbs of the synform that is part of the inferred major sheath fold (Figure 5-25). All the domains in these marginal volumes dip steeply towards the south-west. Key domains RFM029 and RFM045 occur inside the tectonic lens and target volume, in the hinge of the synform, which plunges moderately to steeply (55–60°) to the south-east (Figure 5-25). The properties of the rock domains have been presented in Appendix 14 in /Stephens et al. 2007/. A description of the key domains RFM029 and RFM045 follows below.

26 23

31 34

17 20

44 18

29 21

18 34

12 20

17 26

26 1630000

1630000

1632000

1634000

6698000 6698000

6700000 6700000

6702000 6702000

±

40 33

29R

23 32

25 34

1 37 18

2 5

36 17

6 3

8 12

11 41

4 14

35 39 44

1630000

1635000

1640000

6693000 6693000

6698000 6698000

6703000 6703000

6708000 6708000

±

0 0 0,5 1 2km

Coast line Candidate area Local model area Local model area, stage 2.2 Regional model area Rock domain boundary

Aplitic granite, medium-grained granite and felsic volcanic rock, metamorphic and, in part, albitised

Granite to granodiorite, metamorphic, medium-grained Tonalite and granodiorite, metamorphic

Diorite, quartz diorite, gabbro and ultramafic rock, metamorphic Felsic to intermediate volcanic rock, metamorphic and, in part, albitised

2 4

1 km

Coast line Candidate area Local model area Local model area, stage 2.2 Regional model area Rock domain boundary Granite, fine- to medium-grained

Granite, granodiorite and tonalite, metamorphic, fine- to medium-grained Granite, metamorphic, veined to migmatitic

Granitoid, metamorphic

Aplitic granite, medium-grained granite and felsic volcanic rock, metamorphic and, in part, albitised

Granite to granodiorite, metamorphic, medium-grained Tonalite and granodiorite, metamorphic

Diorite, quartz diorite, gabbro and ultramafic rock, metamorphic Felsic to intermediate volcanic rock, metamorphic and, in part, albitised Sedimentary rock, metamorphic, veined to migmatitic

a b

Figure 5‑25. Rock domains (numbered) included in the three dimensional local model, stage 2.2. The model is viewed to the west towards SFR and the nuclear power station from a position above the local model volume and approximately 2 km to the south-east of the SFR repository. The target volume consists of domains RFM029 and RFM045 in the hinge of the major synform. Boreholes marked by larger cylinders constrain the boundaries between different domains with the help of fixed-point intersections. Quantitative estimates of the proportions of different rock types in rock domains RFM029 and RFM045 are shown in the histograms and the orientation of ductile structures inside these two rock domains are shown in the stereographic projections (equal-area, lower hemisphere). Planar structures are plotted as poles to planes.

An estimation of the degree of point, girdle or random distribution pattern (Vollmer fabric index, PGR) is provided. No Terzaghi correction has been applied. The data come from the surface and from several boreholes with different orientations.

44 18

16 Regional deformation zone

Aplitic granite, medium-grained granite and felsic volcanic rock, metamorphic and, in part, albitized Granite to granodiorite, metamorphic, medium-grained Tonalite and granodiorite, metamorphic

Diorite, quartz diorite, gabbro and ultramafic rock, metamorphic

Granite to granodiorite, metamorphic, medium-grained (101057) Pegmatite, pegmatitic granite (101061) Granitoid, metamorphic, fine- to medium-grained (101051)

Amphibolite (102017), minor mafic to intermediate rock (101033) Granite, fine- to medium-grained (111058) Other rock types

Granite, commonly affected by albitization and fine-grained, metamorphic (101057_101058_104) Pegmatite, pegmatitic

granite (101061) Granitoid, metamorphic, fine- to medium-grained (101051)

Amphibolite (102017), minor mafic to intermediate rock (101033) Granite, fine- to medium-grained (111058) Other rock types

% 70

In document Site description of Forsmark at completion of the site (Page 131-135)