A dissertation submitted to the Faculty of Science of the University of the Witwatersrand for the degree of Master of Science / Detailed high precision geochronological studies have been performed on the 2054 Ma old Bushveld Complex, in an attempt to unravel its tectonic and thermal evolution in the period immediately following intrusion and crystallisation. The geochronological techniques used have been specifically chosen to sample specific temperature episodes in the cooling of the Complex, rather than to necessarily provide an accurate emplacement age, The Bushveld Complex is seen in this study as part of the Bushveld Magmatic Province, rather than as an isolated intrusion, The geochronological data are therefore interpreted in the context of the current understanding of the Proterozoic tectonic and thermal history of the Kaapvaal Craton. The development of clean chemical methods and accurate geochronological methods are essential to this type of study. The reduction of laboratory blanks, especially for lead and the development of laboratory techniques for the analysis of small samples therefore played an important part in this study. It has been possible to lower analytical blanks, especially lead blanks to levels where the analysis of small samples is possible. In addition, the zircon evaporation technique was attempted. Phlogopite micas from the Critical Zone of the Bushveld Complex give a wlde range of Rb-Sr model ages, some almost 100Ma older than the preferred age. This indicates a period of hydrothermal alteration of considerable duration at the same time as the intrusion. The slightly young Rb-Sr age recorded for all the mica and whole rock data collected for this study indicates the alteration of the micas which is evident from petrographic and electron microprobe studies. U-Pb and Pb-Pb zircon ages are also Significantly younger than the preferred age, indicating a degree of alteration. This is also seen in the discordance of the zircons seen in the U-Pb data. / AC2017

A Dissertation submitted to the Faculty of Science, University of the Witwatersrand; in fulfilment of the requirements for the degree of Master of Science. Johannesburg August 2016. / Rocks of the Neoarchaean Witwatersrand Supergroup exposed in the collar of the impact-induced 2.02 Ga Vredefort Dome exhibit complex geological structures. These structures are generally considered to have been formed by the Vredefort impact event, through rapid deformations on time scales of seconds to minutes associated with the relatively brief impact processes. However, geological mapping of the structures and petrographic analysis from the northern collar of the dome show that the collar hosts at least three generations of pre-impact structures. In contrast to impact-induced structures, these pre-impact structures indicate slow and progressive deformations that are uncharacteristic of impacts. The pre-impact deformations comprise: (a) an extensional D1 deformation characterised by listric faults up to kilometre-scale; (b) Syn-metamorphic (M2(NC)) D2 ductile deformation characterised by regional S2 foliation, which locally indicates northwest-directed vergence; and (c) D3 deformation that crenulated the pre-existing S2 foliation (S3). Pre-impact structures can be distinguished from impact-induced structures by: (1) difference in the geometry and sense of slip between D1 faults and D4 impact-induced faults; and (2) crosscutting relationships between impact-induced D4 features and D2 and D3 pre-impact features. In their present (rotated) orientation, the D1 faults exhibit an apparent strike-slip separation, which translates to normal-slip fault geometries when impact-induced overturning of strata is undone. Displacement affects the Witwatersrand and Ventersdorp Supergroup rocks but no offset is observed of the base of the Transvaal Supergroup. The faults also exhibit a listric geometry, curving into parallelism with bedding in the lower West Rand Group. In their restored orientation, faults define half-graben and horst blocks, synthetic and antithetic faults, and rollover and drag folds, which are typical for extensional tectonics. These geometries and crosscutting relationships of the D1 faults are similar to that of the Neoarchaean listric faults described in the Witwatersrand goldfields and the wider Kaapvaal craton, that exhibit a general west-side-down sense of slip (2.70-2.64 Ga Hlukana-Platberg extensional event). Metamorphic grade in the study area decreases from amphibolite- to greenschist-facies away from the centre of the dome. These are largely M2(NC) metamorphic assemblages that are attributed to elevated regional heat flow related to 2.06 Ga Bushveld magmatism. There is some evidence that M2(NC) metamorphic mineral assemblages developed along the same stratigraphic units differ across the large D1 faults, indicating the pre-impact nature of the D1 faults and implying that the M2(NC) metamorphism occurred after the Hlukana-Platberg event. Also, M2(NC) assemblages are syn-tectonic to the S2 foliation hosted in metapelite units of the West Rand Group and knotted quartzite horizons of the Central Rand Group. The S2 foliation is attributed to the post-Transvaal Supergroup, compressional, Ukubambana Event. Crosscutting relationships in the study area indicate a deformational period of 2.06 Ga to no less than 2.02 Ga. The northwest-directed vergence exhibited by the S2 foliation is broadly consistent with the regional, general north-directed, vergence exhibited by post-Transvaal Supergroup foliation developed in the northeastern collar and the Johannesburg Dome. The S2 foliation and M2(NC) mineral assemblages are crosscut by D4 pseudotachylitic breccia, micro-faults and kinks, and M4(NC) metamorphic features associated with the impact. / LG2017

Deformation of the western part of the Table Mountain Group rocks during the Cape Orogeny created a series of folds and associated fractures. The subsequent continental break-up of Gondwana led to the development of large fault systems. These exert a major influence on deep and shallow groundwater flow. There are 3 main types of structures that are investigated. The geological contacts between hydraulically different lithologies, the primary characteristics of the sediments comprising the main geological units and the secondary structures developed from the tectonic events. These inter-alia include lithological boundaries, bedding and conjugate joints and large faults. Compartmentalisation of the aquifers by lithological and fault boundaries are the main regional level controls on flow in the study area. Joints are important for local control of flow, but cumulatively exert a regional effect as well. These controls exert a strong 3 dimensional impact on flow patterns within the area. Geological cross sections and detailed fieldwork combined with the conceptual models proposed are used to determine groundwater flow and the extent of the flow constraints. There is heterogeneity in the fault characteristics whilst there isconsistence in the impermeable aquitards. These effect boundaries at the base of the aquifer, divide the aquifer into upper and lower units and cap the top of the aquifer. Using water level data, EC and pH an attempt is made to establish patterns created by structures, mainly faults. There appears to be some control of these shown by patterns seen on contour plots of the data. Understanding of the structures can significantly alter the way the available data could be interpreted. The integration of all available data into the conceptual model provides an effective research tool, which opens up further avenues for new approaches and methods for continued research in this area.

A thesis submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy, / The Ventersdorp Contact Reef(VCR) is a major gold-bearing reef in the Witwatersrand Basin. It occurs between the overlying Klipriviersberg Group lavas and the underlying Central Rand Group sediments, and was strongly altered by hydrothermal fluids circulating in the Witwatersrand Basin. A detailed study of the mineralogy, geochemistry of rocks and minerals, physicochemical conditions, stable isotopes and ages of hydrothermal alteration zones associated with the VCR were carried out at Western Deep Levels South Mine, South Africa. ( Abbreviation abstract) / Andrew Chakane 2019

The Ngoye Granite Gneiss Formation is located in the Natal sector of the Proterozoic Namaqua-Natal Mobile Belt, about 10 km southwest of Empangeni. It forms a prominent east-west trending elongate whalebacked massif some 30 km in length, within amphibolitic gneisses and schists of the Tugela Group. A suite of twelve different, gneissic granitoids has been-recognised within the Ngoye Formation on the basis of field relationships, mineralogy and supportive geochemistry. They range in composition from peraluminous syenite to peralkaline granite. Peraluminous varieties are typically muscovite and garnet-bearing whereas metaluminous granites in the formation contain olivegreen biotite and/or hornblende and sphene. Riebeckite, aegerine and yellow-brown biotite, with accessory fluorite and zircon are characteristic of the peralkaline granites. Geochemically, the samples analysed display a range in SiO₂ from 63,79 - 78,47∞, are extremely depleted in CaO and MgO, while being enriched in Na₂O and K₂O. Depletion of CaO relative to alkalis is shown by an alkali-lime index of only 36, suggestive of an alkalic character. The agpaitic index (A. I. = mole Na₂O + K₂O/AL₂O₃) of the peralkaline samples ranges between 1,02 and 1,16; which classifies them as granites of comenditic affinity. Various chemical classification schemes have been tested and evaluated, of which the RI - R2 multicationic diagram provides results most similar to modally-derived terminology. Accordingly, the Ngoye granitoids are shown to range from minor syenites and alkali granites to predominant monzo - and syeno-granites. Trace element data indicate that the peralkaline granites are enriched in Nb, Zr and Zn relative to the other, non-peralkaline, granites in the formation. In addition, radioactive, magnetite-bearing quartz-rich rocks associated with the peralkaline granites, have extremely enhanced contents of Nb, Zr, Y, Zn, U, Th and to a lesser extent Sn and W. Peraluminous and near-peraluminous granites have the highst Rb/Sr and Rb/Ba ratios of all samples analysed, as well as enhanced Sn, U and Th contents while Zr is notably depleted. Small, muscovite-rich pods associated with muscovite-bearing granites are highly enriched in Sn. The application of certain discriminants based on modal and geochemical parameters has shown the Ngoye Formation to comprise typical "A" - type granites. "A" - type granites are characteristically intruded as ring complexes into anorogenic or post-orogenic tectonic settings in attenuated or epiorogenically-domed continental crust. Comparison of the Ngoye Formation wi th the well-known "younger granite" complexes of Nigeria and Saudi Arabia reveals marked similarities. The inference is therefore that the Ngoye Formation represents a metamorphosed "postorogenic" granite complex with most of the hallmarks of "A" type or "within-plate" magmatism. Four phases of deformation (D₁ to D₄) are recognised within the area mapped. Evidence of D₁ deformation is rare, but rootless folds within the transposed layering in the amphibolitic country rocks reflect the intensity of this prograde metamorphic event, M₁, during which upper amphibolite grades were achieved. Field evidence shows that the Ngoye granites were intruded after the D₁ event and prior to D₂. This latter event caused widespread folding about east-west F₂ axes, with the development of a pervasive S₂ planar fabric within the antiformally folded Ngoye Formation. S₂ is locally developed in the amphibolitic country rocks. The D₂ event culminated in the development of northward-directed overthrusting and retrogressive ,M₂, metamorphism of mylonitic thrust planes. Lateral shearing characterizes D₃, with development of macroscopic mylonites and mesoscopic conjugate shear zones. This was in response to a sinistral sense of movement, as indicated by prominent sub-horizontal extension lineations (L₃) and microscopic asymmetric augen structures. D₄ is deduced from stereograms and is indicated as cross-folding of F₃ fold axes. / Thesis (M.Sc.)-University of Durban-Westville, 1985.

Deformation of the western part of the Table Mountain Group rocks during the Cape Orogeny created a series of folds and associated fractures. The subsequent continental break-up of Gondwana led to the development of large fault systems. These exert a major influence on deep and shallow groundwater flow. There are 3 main types of structures that are investigated. The geological contacts between hydraulically different lithologies, the primary characteristics of the sediments comprising the main geological units and the secondary structures developed from the tectonic events. These inter-alia include lithological boundaries, bedding and conjugate joints and large faults. Compartmentalisation of the aquifers by lithological and fault boundaries are the main regional level controls on flow in the study area. Joints are important for local control of flow, but cumulatively exert a regional effect as well. These controls exert a strong 3 dimensional impact on flow patterns within the area. Geological cross sections and detailed fieldwork combined with the conceptual models proposed are used to determine groundwater flow and the extent of the flow constraints. There is heterogeneity in the fault characteristics whilst there isconsistence in the impermeable aquitards. These effect boundaries at the base of the aquifer, divide the aquifer into upper and lower units and cap the top of the aquifer. Using water level data, EC and pH an attempt is made to establish patterns created by structures, mainly faults. There appears to be some control of these shown by patterns seen on contour plots of the data. Understanding of the structures can significantly alter the way the available data could be interpreted. The integration of all available data into the conceptual model provides an effective research tool, which opens up further avenues for new approaches and methods for continued research in this area.

D.Phil. (Geology) / The tectonic evolution of the "western margin" of the Witwatersrand Basin is examined and indications are that it has undergone a long and complex history. In order to examine the nature of Witwatersrand-age structures, structures in both pre- and post-Witwatersrand sequences are also examined. Rocks of the ±3074 Ma Dominion Group were subjected to a tectono-metamorphic event prior to the deposition of Witwatersrand strata on an angular unconformity. An oligomictic conglomerate is sporadically developed at the base of the Witwatersrand Supergroup. PreVentersdorp structures in Witwatersrand strata are developed in two distinct trends, north-south and northeast-southwest. The relationship between the two directions of folds and thrust faults are best explained within a regional, sinistral transpressive shear couple; the north-south faults are sinistral strike-slip faults and the northeast-southwest trending folds and thrust faults are secondary structures associated with the strikeslip faults. The implications of this model are that Witwatersrand sedimentation was probably controlled by lateral movements on north-south trending faults and not by thrust faults in a foreland system as suggested by the most recent models of Witwatersrand basin development. Post-Witwatersrand deformation is complex. Southeastward verging, pre-Ventersdorp, thrust faults were reactivated as normal faults during Platberg times and the resultant half-grabens were infilled by conglomerates of the Kameeldoorns Formation. Later deformational events include eastward verging post-Ventersdorp thrust faults and post-Transvaal normal and strike-slip faults. It can be demonstrated that the majority of this later fault movements took place along pre-existing fault planes and therefore tectonic inversion is a fundamental process in the evolution of the Witwatersrand Basin. Clearly therefore, the present distribution of Witwatersrand strata does not reflect the original basin geometry, it is the result of several periods of basin inversion and no basin margins can be defined.

M.Sc. (Geology) / Please refer to full text to view abstract

D.Phil. (Geology) / This study examines the nature, distribution and origin of a distinctive chert-like fault rock in the West Rand Goldfield of the Witwatersrand Basin in South Africa. These fault rocks, termed pseudotachylites, are characterized by an aphanitic groundmass enclosing subangular to rounded clasts of the host rocks. No glass has been observed in the matrix but features such as spherulites, coronas and altered margins to the host rocks as well as geochemical evidence, suggest that the pseudotachylite formed as a result of melting of the host rocks due to the heat generated by friction on faults. The colour of the pseudotachylite is a function of its chemical composition and parentage. The pseudotachylite has abrupt contacts with the host rocks which comprise a lower Proterozoic to Archaean succession of rocks belonging to the predominantly sedimentary Transvaal Sequence, the predominantly volcanic Ventersdorp Supergroup and the predominantly . sedimentary Witwatersrand Supergroup. The orientation of many of the pseudotachylite fault veins parallels a pre-existing set of mylonitic faults. These pseudotachylite fault veins most commonly occur in sub parallel southward dipping pairs and are accompanied by injection veins. If treated on a statistical basis, the vergence concept can be extended to injection veins to give the approximate movement direction of the fault system. The pseudotachylite is thought to be genetically related to brittle or semi-brittle extensional faulting of post-Transvaal age.

This study describes the bedrock lithologies and structure of the Ordovician to early Devonian (485-419 Ma) Table Mountain Group (TMG), the Devonian (419-358 Ma) lower Bokkeveld Group, and the Miocene to Holocene (<23 Ma) overburden sediments of the Algoa Group within an area identified by Eskom for the potential construction of South Africa’s second proposed nuclear power plant (NPP), ‘Nuclear-1’. The study area is located along the southern coastal margin of the Eastern Cape Province, South Africa, between Oyster Bay and St. Francis (approximately 88 km west of Port Elizabeth), and encompasses the Thyspunt site where the proposed NPP will be built. The study aims to supplement existing information about the Thyspunt area, related to the geoscientific topic ‘Geological Setting’, as outlined in section 2.5.1.1 of the US Nuclear Regulatory Commission (USNRC) Standard Review Plan NUREG-800, which details the geological information required for review of a proposed NPP. The results obtained from geoscientific studies are used to determine geological factors that may potentially affect site specific design. Factors considered include: bedrock lithology, stratigraphic bedrock contacts, bedrock palaeotopography, thickness of overburden sediments and structural geology. Work by previous authors is combined with new data to create a GIS based 2½D model of the study area’s geology (geomodel) and on which future research or interpretations can be based. Field mapping and petrographic analyses of the TMG, comprising the Peninsula, Cedarberg, Goudini, Skurweberg and Baviaanskloof Formations as well as the lower undifferentiated Bokkeveld Group were undertaken to define the study area’s lithologies and structure. Interpretation of geophysical results and the integration of existing borehole data aided in defining the variability in overburden sediments, the identification of contacts between TMG formations beneath overburden, and the palaeotopography of bedrock. Borehole data indicates a clear N-S trend in the thickness distribution of Algoa Group aeolian and marine related sediments. Four coast-parallel trending thickness zones (zones A – D) are recognized within the study area. At Thyspunt overburden thickness reaches a maximum of 61 m, approximately 1200 m from the coastline, in areas underlain by the argillaceous Goudini and Cedarberg Formations. Overburden thickness is influenced by a combination of dune relief, bedrock lithology, palaeotopography and the area’s sediment supply. Interpolation of bedrock elevation points and detailed cross sections across bedrock reveals four NW-SE trending palaeovalleys at Thyspunt, Tony’s Bay, Cape St. Francis and St. Francis, where bedrock relief (beneath overburden) is formed to be below present day sea-level. Approximately 450 m NW of Thys Bay, a 1050 m2 (area below sea-level) palaeovalley, gently sloping SE to a depth of -15.5 m asl, is cut into strata of the Goudini Formation resulting in thicker overburden fill in that area. Structural analysis of the TMG confirms that NE-SW striking strata form part of the regional SE plunging, north verging Cape St. Francis anticline. Bedding inclination is controlled by the distance away from the fold axis, varying from a 5° SE dip along the broad fold hinge to 65° along its moderately steeper SE limb. Folds within the study area plunge gently southeastward at shallow angles, with axial planes dipping steeply SW or NE. Fold axes orientated perpendicular to the fold axis of the Cape St. Francis anticline indicate a secondary stress orientation oblique to the main palaeostress direction. The previously identified 40 km long, NW-SE trending Cape St. Francis fault occurring offshore within 17.5 km of Thyspunt show no onshore continuation within the bounds of the study area. Late jointing is pervasive within the study area and four joint systems are identified. The dominant joint set J1, trends N-S to NNE - SSW; perpendicular to bedding and has a subvertical dip. Normal right-lateral and left-lateral micro-faults dip subvertically, with a displacement that ranges from a few centimetres to <3 m. Micro-faults trend parallel to joints sets J1 and J4 (ESE-WSW). Inferred faults, identified by the Atomic Energy Co-operation (AEC), are interpreted as zones of closely spaced jointing (shatter zones), and show little to no recognizable displacement. Faults and joints do not extend into the younger cover deposits of the Algoa Group and are therefore older than 23 Ma years.