A thesis submitted to the Faculty of Science in fulfilment of the requirements for the degree of Doctor of Philosophy at the School of Geosciences University of the Witwatersrand, Johannesburg, 2017 / Namibia, the 6th largest producer of uranium globally, has produced uranium from Pan African granite-hosted (primary) deposits since 1976, and from palaeochannel deposits since 2007; exporting 3 472 tonnes U in 2016. The large granite-hosted deposits at the Husab Mine are expected to add over 5 700 tonnes U/year at peak, while three large primary-hosted deposits remain in various stages of development at Goanikontes, the Ida Dome, and Valencia. This study presents a comprehensive geological, geochemical and uranium mineralogical appraisal of four of the major primary-hosted uranium deposits, all situated within the southern Central Zone (sCZ) of the polydeformational (D1-D3) Damara Belt. The sCZ comprises highly deformed Neoproterozoic sediments, unconformably draped over rheologically competent granite-gneiss domes and inliers of a Palaeoproterozoic basement. A suite of fractionated sheeted leucogranites (SLGs) are a characteristic of the final stages of Orogenic deformation; while most SLGs appear to precede D3 deformation and metamorphism (ca. 510 Ma); most of the mineralised SLGs across the region invade reduced-facies sediments in pressure shadows formed in the hinges and limbs of upright D3 antiforms, proximal to basement inliers. A pre-existing, six-fold, alphabetised SLG classification scheme is revised and extended to categorise distinctive and consistent field and petrographic characteristics of the SLGs across the region. Discriminating SLG sub-types is less consistent in standard geochemical diagrams, except where high field-strength (HFS) and rare-earth elements (REE) are concerned. REE profiles in pre-D3 SLGs reflect abundances, or paucities, of characteristic accessory mineral assemblages; while REE profiles show relative REE enrichment, prominent REEfractionation and -ve Eu anomalies in the uraniferous SLGs, reflecting lower-percentage partial melts in the more uraniferous samples. The overwhelming majority of primary uranium mineralisation is in magmatic uraninite, followed by coffinite which predominate as a replacement phase of uraninite, and more rarely as solid solution with thorite. The refractory minerals betafite and brannerite are rare, but are locally abundant in discrete, magmatic textures within uraniferous SLGs of some deposits. Hydrated uranyl silicates predominate in the supergene portions of the orebodies across the region. An electron microprobe study presents the first comprehensive assessment of uraninite compositions in the region, while Husab deposit betafite and brannerite compositions allow for a well-rounded comparison with refractory minerals from the Rössing deposits. Key Words Primary Uranium, Granite, Orogenic, Damara, Namibia, Rare Earth Elements, Mineralisation, Fractionation, High-grade Metamorphism, Economic Geology, Mining, Processing, Uraninite, Coffinite, Etango, Goanikontes, Husab, Ida Dome, Rössing, Valencia / XL2018

A dissertation submitted to the Faculty of Science, University of the Witwatersrand, in fulfilment of the requirements for the degree of Doctor of Philosophy, Johannesburg 2014 / Namibia is renowned for its abundant mineral resources, a large proportion of which are hosted in the metasedimentary lithologies of the Damara Belt, the northeast-trending inland branch of the Neoproterozoic Pan-African Damara Orogen. Deposit types include late- to post-tectonic (~ 523 – 506 Ma) LCT (Li-Be, Sn-, and miarolitic gem-tourmalinebearing) pegmatites, and uraniferous pegmatitic sheeted leucogranites (SLGs), which have an NYF affinity. Fluid inclusion studies reveal that although mineralization differs between the different types of pegmatites located at different geographic locations, and by extension, different stratigraphic levels, the fluid inclusion assemblages present in these pegmatites are similar; thus different types of pegmatites are indistinguishable from each other based on their fluid inclusion assemblages. Thorough fluid inclusion petrography indicated that although fluid inclusions are abundant in the pegmatites, no primary fluid inclusions could be identified, and rather those studied are pseudosecondary and secondary. Fluid inclusions are aqueo-carbonic (± NaCl), carbonic, and aqueous. It is proposed that all of the pegmatites studied share a similar late-stage evolution, with fluids becoming less carbonic and less saline with the progression of crystallisation. Oxygen isotope ratios allow the discrimination of different pegmatites into two groups, Group A (Sn-, Li-Sn-, and gem-tourmaline-bearing LCT pegmatites), and Group B (Li-Bebearing LCT, and U-bearing NYF pegmatites). Group A pegmatites have O-isotope ratios ranging from 11 to 13 ‰ suggesting that they have an I-type affinity. These values are, however, elevated above those of typical I-type granites (7 - 9 ‰), indicating either a postemplacement low-temperature exchange with meteoric fluid, high-temperature hydrothermal exchange with δ18O country rocks during emplacement, or the derivation of these pegmatites from a non-pelitic/S-type metaigneous source. Group B pegmatites have higher δ18O ratios (δ18O = 15 - 16 ‰), indicative of their S-type affinity, and their derivation from metapelitic source rocks. δD values of all the pegmatites range from -40 ‰ to -90 ‰ indicating that the pegmatitic fluids are primary magmatic with a metamorphic fluid component. Trends in the trace element concentrations of both Group A and Group B pegmatites are very similar to each other, making the two groups indistinguishable from each other on this basis. The Damaran pegmatites also share similar geochemical trends with their country rocks. There is, however, no direct field evidence to suggest that the pegmatites were derived from the in situ anatexis of the country rocks. It is more likely that anatexis occurred some distance away from where the pegmatites were ultimately emplaced, and that the melts migrated and were finally emplaced in pre-existing structures, possibly formed during Damaran deformation. O-isotope and Ti-in-quartz geothermometry indicate that Damaran pegmatites can be subdivided into two groups based on their crystallisation temperatures. LCT pegmatites crystallised at temperatures ranging from ~ 450 - 550 ºC, while the NYF pegmatites crystallised at higher temperatures, ranging from 630 - 670 ºC. It is important to note that the subdivision of pegmatites in Groups A and B based on their O-isotope systematics does not correspond with their subdivision into the LCT and NYF pegmatite families according to their crystallisation temperatures. In addition to clarifying aspects of the emplacement and evolution of the Damaran pegmatites, this study points out that there are several discrepancies in the current classification schemes of pegmatites. It shows that in addition to the problems encountered when trying to distinguish between LCT and NYF pegmatites based on their mineralogy, they also cannot truly be distinguished from each other using their geochemical and isotopic characteristics, or their tectonic settings. It is tentatively proposed that crystallisation temperature be considered as an alternative or additional characteristic in the classification of pegmatites, and that it be considered on a regional scale rather than only in the evaluation of the highly evolved end-members of a pegmatite swarm.

A thesis submitted to the Faculty of Science, University of the Witwatersrand, in fulfilment of the requirements for the degree of Doctor of Philosophy University of the Witwatersrand School of Geosciences and Dublin Institute for Advanced Studies School of Cosmic Physics Geophysics Section February 2016 / Archean cratons, and the Proterozoic orogenic belts on their flanks, form an integral part of the Southern Africa tectonic landscape. Of these, virtually nothing is known of the position and thickness of the southern boundary of the composite Congo craton and the Neoproterozoic Pan African orogenic belt due to thick sedimentary cover. In this work I present the first lithospheric-scale geophysical study of that cryptic boundary and define its geometry at depth. The results are derived from two-dimensional (2D) and three-dimensional (3D) inversion of magnetotelluric data acquired along four semi-parallel profiles crossing the Kalahari craton across the Damara-Ghanzi-Chobe belts (DGC) and extending into the Congo craton. Two dimensional and three-dimensional electrical resistivity models show significant lateral variation in the crust and upper mantle across strike from the younger DGC orogen to the older adjacent cratons. The Damara belt lithosphere is found to be more conductive and significantly thinner than that of the adjacent Congo craton. The Congo craton is characterized by very thick (to depths of 250 km) and resistive (i.e. cold) lithosphere. Resistive upper crustal features are interpreted as caused by igneous intrusions emplaced during Pan-African magmatism. Graphite-bearing calcite marbles and sulfides are widespread in the Damara belt and account for the high crustal conductivity in the Central Zone. The resistivity models provide new constraints on the southern extent of the greater Congo craton, and suggest that the current boundary drawn on geological maps needs revision and that the craton should be extended further south. The storage possibilities for the Karoo Basins were found to be poor because of the very low porosity and permeability of the sandstones, the presence of extensive dolerite sills and dykes. The obvious limitation of the above study is the large spacings between the MT stations (> 10km). This is particularly more limiting in resolving the horizontal layers in the Karoo basin. However the 1D models provide layered Earth models that are consistent with the known geology. The resistivity values from the 1D models allowed porosity of the Ecca and Beaufort group lithologies to be calculated. It is inferred that the porosities values are in the range 5-15 % in the region below the profile. This value is considered too low for CO2 storage as the average porosity of rock used for CO2 is generally more than 10 to 12 percent of the total rock unit volume.

A dissertation submitted to the Faculty of Science, University of Witwatersrand in the fulfilment of the requirements for the degree of Master of Science. Johannesburg, 2015. / Northwest Botswana holds a key position for the correlation of the Pan-African mobile belts of southern Africa (i.e. the Damara-Zambezi-Lufilian Orogeny). Phanerozoic cover (Kalahari Group) precludes direct correlation between Proterozoic lithologies of the Damara Belt and thick metasedimentary sequences of northwest Botswana. A combination of new geological and geophysical field observations, interpretation of 50 m resolution aeromagnetic data, and 2.2 km resolution gravity data of Namibia and Botswana, have led to the development of a new sub- Kalahari geological map of the Damara Belt and northwest Botswana. The interpretation of potential field and magnetotelluric (MT) data complemented with both new and published geological data, has improved the identification of the northern and southern margins of the Damara Belt and northwest Botswana, and tectonostratigraphic zones within them. In addition, these correlations have established that the northern margin of the Kalahari Craton on geological maps extends further north than previously noted. The northeast trending Damara Belt is confidently traced into northwest Botswana (Ngamiland) to ~19.5°S, 22.0°E. At this location, in map view, aeromagnetically interpreted structures follow a radial distribution from northwest-striking in the west to northeast-striking in the east. The lithostratigraphic units to the north of this location cannot be confidently correlated with lithostratigraphic units of the Damara Belt. Instead, these units are better correlated with lithostratigraphic units in southern Angola and/or Zambia. The southeastern margin of the Damara Belt is in tectonic contact with the northern margin of the Ghanzi-Chobe Belt as identified in the aeromagnetic images. The Ghanzi-Chobe Belt is correlated with the Sinclair Supergroup in the Rehoboth Subprovince in Namibia. The basal Kgwebe volcanics are correlated with the Oorlogsende Porphyry Member and Langberg Formation and the unconformably overlying metasediments of the Ghanzi Group are correlated with the metasediments of the Tsumis Group. The correlations are based on similar aeromagnetic signatures, lithologies, mineralisation and age dates constrained by carbon isotope chemostratigraphy. Physical property measurements were collected on Meso- to Neoproterozoic lithologies of the Damara Belt, northwest Botswana and Zambia. The measurements included hand held magnetic susceptibility measurements on 303 samples and density measurements on 174 samples. The measurements provide one of the largest physical property databases for Namibia, Botswana and Zambia. In general, the sedimentary units have the lowest magnetic susceptibility values of ~0.207 x 10-3 SI units, respectively. The exceptions are the iron formation and diamictite of the Chuos Formation and conglomerate of the Naauwpoort Formation of 15.2 x 10-3 SI units. The iron iii formation ranges in magnetic susceptibility from 3.34 x 10-3 SI units to 92.0 x 10-3 SI units and the diamictite has a magnetic susceptibility of 7.68 x 10-3 SI units. The igneous lithologies have a density and magnetic susceptibility range from 2.58 g.cm-3 to 3.26 g.cm-3 and 0.001 x 10-3 SI units to 11.6 x 10-3 SI units, respectively. The lower values are associated with pegmatites and rhyolites and the higher values are associated with mafic lithologies and magnetite bearing granites (Omangambo, Salem, Sorris-Sorris and Red Granites). The metamorphic lithologies have the widest range of density and magnetic susceptibility values, between 2.61 g.cm-3 and 3.37 g.cm-3, and -0.299 x 10-3 SI units and 49.5 x 10-3 SI units, respectively. The lower values are associated with low grade metamorphic facies of sedimentary origin, and the higher values are associated with high-grade metamorphic facies of an igneous origin. The first upper crustal-scale interpretation of the Southern African MagnetoTelluric EXperiment (SAMTEX) was developed. The results were derived from 1D Occam inversion models, at depth intervals of 1 – 5 km, 1 – 15 km and 1 – 35 km. The MT data were acquired across the semiparallel, north-south striking DMB, NEN and OKA-CAM profiles in the vicinity of the Namibia – Botswana border between 2006 and 2009. Beneath the MT profiles are two zones of enhanced conductivity, a northern and southern zone. The enhanced conductivity of the northern zone (> 100 Ωm) is associated with individual geological bodies. The southern zone forms an elongated belt of enhanced conductivity (> 300 Ωm) at a depth of less than 5 km. This zone of enhanced conductivity is associated with Proterozoic plate boundaries and subduction zones. Three ~350 km long, north-south trending magnetic profiles were 2D forward modelled to investigate the proposed northward subduction of oceanic crust and subsequently a portion of the Kalahari Plate beneath the Congo Craton. Additionally, the folding pattern of the Ghanzi- Chobe Belt was developed. The interpretation of the magnetic models suggests a northward subduction is a possible cause for the evolution of the Damara Orogen with the regionally eastwest striking negative aeromagnetic anomaly, in northern Namibia, being caused by a thick package (~12 km to 20 km) of metasediments with a modelled magnetic susceptibility of 0. 829 x 10-3 SI units. The Damara Orogen has passed through the subduction-collisional transition but did not evolve into a large-hot orogen. Evidence suggests that the Damara Orogen has gone through the transition of subduction of oceanic crust to terrane accretion (speculated to be represented by the Deep-Level Southern Zone and Chihabadum Complex) and continental collision. However, the doubly vergent wedges did not evolve into an orogenic plateau completing the transition from a small-cold orogen to a large-hot orogen. This is similarly observed in the Alps Orogeny.