The "A" -Reef occurs within the Aandenk Formation of the Central Rand Group, Witwatersrand Supergroup, in the Welkom Goldfield. It comprises the Witpan and Uitsig Reefs which are both oligomictic conglomerates, and are exploited for their gold content by Anglo American Corporation. The main Witpan channel complex is orientated in a NW-SE direction and occurs close to No.3 Shaft of Western Holdings Mine where it is currently being mined. The Witpan Reef varies in thickness from 7-220cm, with lateral facies changes controlling the thickness. Within the reef, gold is associated with degradation surfaces, carbonaceous material, increase in pebble sphericity, and channel edges. Two channel edges have yielded the best gold values on No.3 Shaft although the potential for further payable gold lies in the recognition of sieve conditions. The potential also exists for extensions of the "carbon"-bearing Uitsig channel currently being mined on President Steyn Mine. The palaeo-environment proposed for formation of the Witpan Reef is a braidplain that was partly reworked by a brief transgression. Ore evaluation using geostatistics was considered a valid technique as the dataset is sufficiently well structured. Semi-variograms in the channel and across-channel directions differ markedly. It was found that variograms of gold in cmg/t lacked sufficient structure for modelling, however, log semi-variogram modelling followed by simple log-kriging and back-transformation, proved to be the most successful method. Owing to the morphology and distribution of gold within the reef, a geologically based geostatistical valuation method is proposed. The potential for further exploration of "A" -Reef depends on a substantially higher gold price. In this event, exploration of Uitsig Reef to the southwest of the current mining area is recommended as well as a new exploration strategy for Witpan Reef.

The distribution of gold in igneous rocks and minerals is described and discussed. Not all the gold in igneous rocks is contained within early formed crystal lattices. Evidence that gold can be associated with late stage crystallizing phases is described. It is concluded that some of the gold in hydrothermal and volcanogenic deposits may have come from a primary magmatic source rather than having been leached from solid country rock. Gold is probably transported as chloride complexes at temperatures greater than about 300°C. At lower temperatures it is probably transported with other metals as sulphide and thio-sulphide complexes. The precipitation of gold from the transporting medium is brought about by changes in the physico-chemical conditions within that medium. Decrease in pressure is probably not a major cause of precipitation in volcanogenic environments. The geology of volcanogenic iron-formations is described and discussed, relative to the development of greenstone belts. Oxide facies iron-formations were formed in shallow oxidizing environments. They are associated with volcanogenic and clastic sediments. Sulphide facies iron-formations were precipitated in the deeper parts of geosynclinal structures. They are associated with mafic and ultramafic rocks similar to modern oceanic volcanic assemblages. Carbonate facies iron-formations were deposited in the regions between oxide facies and sulphide facies. Other banded iron-formations are found associated with base metal massive sulphide deposits related to arctype volcanic centres. These deposits are found in the regions where carbonate facies iron-formations were formed. Exploration for and exploitation of gold deposits in Archaean iron-formations are discussed. Geochemical exploration programmes are aided by the association of gold with trace amounts of base metals. Geophysical exploration methods that can be employed include magnetometer, I.P. and E.M. surveys. The metallurgical treatment of the ores should include "roasting" because a large proportion of the gold occurs as submicroscopic grains within sulphide mineral crystals.

The general characteristics of gold deposits are reviewed, and a classification of gold deposits based on mineability is proposed. Evaluation, mlnlng and beneficiation methods are briefly discussed. It is concluded that the most viable targets for small-scale companies comprise deposits that require the least pre-production time and expense. Great potential exists for the small-scale reclamation of gold from tailings dumps and abandoned mines in Southern Africa. There is also potential for developing new smallscale gold mines in the Archaean greenstone terranes of the Zimbabwean and Kaapvaal cratons.

The first Witwatersrand gold deposits in the Orange Free State were discovered under younger cover rocks in the 1930's with the aid of drilling and geophysics. The Welkom gold deposits are found in the sedimentary rock sequences of the Central Rand Group, which represent unconformity bounded genetic packages. The structural configuration of the goldfield is one of a north to south trending synform that is split near it's axis by the De Bron and Homestead faults. The "B" Reef is a highly variable, erratically mineralised reef that lies on an unconformity at the base of the Spes Bona Formation. The "A" Reef Zone consists of several placers 1 including the "Reworked BPM" 1 the Witpan, the Uitsig, the Hanging Wall Grits and the Upper "A" Reef. The Beatrix Reef lies at the base of the Eldorado Formation on an unconformity surface overlying the Virginia Formation in the southern part of the Welkom Goldfield. The origin of gold in the Witwatersrand basin can be classified into the modified placer theory, the syngenetic theory and the epigenetic theory. From the distribution of basin edge unconformities it can be deduced that the Welkom fan depository was tectonically active on the western, southern and eastern margins during sedimentation. Palaeocurrent studies indicate that sediment was transported predominantly from the south and west. It is thought that the "B", "A" and Beatrix Reefs were all deposited in a braided stream environment. A multidisciplinary approach to ore evaluation of Witwatersrand deposits is considered to be the best method, where sedimentology, geostatistics and structural geology are used.

A vast amount and variety of mine tailings are produced around the world each day. In the gold¬mining industry in South Africa the residue of crushed are is disposed of in large tailings or hydraulic fill dams. The outer walls of these dams are built up of layers of material, each of which is allowed to dry before the next layer is placed. In order to study the stability of these walls, the appropriate engineering properties of the tailings must be ascertained. Due to the construction technique used in tailings dam construction, the outer walls are in an unsaturated state, which also means that suctions are generated within the tailings. Various techniques exist to measure suction, most of which are indirect methods. The recent development of the mid-¬plane suction probe at the University of Pretoria created the opportunity of measuring suctions directly on desiccating samples of gold mine tailings. A test method has been developed from which soil mechanics parameters can be derived from suction measurements. The experimental programme consisted of a series of these newly developed tests on fine and coarse samples of gold mine tailings, as well as on different particle size ranges. The experimental results were used in the development of a new method of predicting the air-entry value, with only the grading of the tailings known. A new method of predicting the soil-water characteristic curve up to the air-entry value was also proposed. The results of the research showed that the tailings remain saturated up to the air-entry value. The clay, fine silt and medium silt sized tailings was found to be the controlling particle size ranges in the development of suctions. The vast amount of parameters and information gained through the use of the proposed test method clearly indicates its effectiveness in studying the performance and characteristics of a material drying from saturation. The results also indicated the effectiveness of the mid-plane suction probe for the direct measurement of suction. / Dissertation (M Eng (Geotechnical Engineering))--University of Pretoria, 2006. / Civil Engineering / unrestricted

Gullies have been regarded as the "artery" in mining because they provide the only access route for material, people and ore in stopes. It thus becomes vital to provide the industry with suitable guidelines for gully layouts, geometry and the support required at all depths to keep the gully safe at all times. The research has indicated that best practices for gully layouts have been well recognised, but often poorly applied for many years. To address the issue of best gully practices, research was based on a review of past practices, underground visits, mine standards, codes of practice and the use of numerical modelling as a tool to back- analyse the underground conditions observed. The recommendations provided do not attempt to develop any new techniques for gully protection. They try to provide a guide for best practice under various geotechnical conditions. Based on depth, or stress environment, a broad based recommendation for gully geometry is provided. Using numerical modelling calibrated to underground observations, optimum widths and spans for each mining layout used at different depths are provided as a prescriptive guideline. • Sidings can only be omitted where stress damage does not occur. • A minimum siding width of 2m is recommended wherever sidings are cut. • Lagging sidings should be avoided and used only if absolutely essential. • In high stress areas gullies should be footwall lifted behind the stope faces or within wide headings. • All sidings must be cut on reef. Off reef sidings are not acceptable. • Correct blasting practice is essential to ensure stability of gully shoulders. • Gully width and span between support over gullies should be minimised. • Gullies must be kept straight. Excavation and cleaning of downdip sidings remains problematical from a practical mining point of view and future research is recommended in this area. / Dissertation (MSc (Mine Strata Control))--University of Pretoria, 2006. / Mining Engineering / unrestricted

M.Sc. (Geology) / The Main Reef Complex (MRC) section is developed in the Zwartkoppie and Sheba Formations of the Onverwacht and Fig Tree Group respectively. The section is represented by three shear zones, associated with a prominent anticlinal structure, locally known as the Birthday No.2 anticline. This tight isoclinal structure is marked by a chert unit, with a core of green quartz-carbonate schist, and is overlain and surrounded by greywackes and shales of the Sheba Formation. At depth the main mineralized fracture occurs to the south of, and parallel to the Birthday No.2 chert bar. Above the crest of the anticlinal structure gold mineralization is associated with three shear zones (which transgress the bedding of the greywackes and shales) known as the No .. I, No. 2 and No. 3 Fractures. Pyrite and arsenopyrite are the main ore minerals, with minor amounts of chalcopyrite, sphalerite, tetrahedrite and pyrrhotite. Pyrite mineralization is ubiquitously developed along the shear zones, while arsenopyrite mineralization is restricted to certain areas. Assay data of the greywackes and shales show that there is a correlation between gold contents and the abundance of arsenopyrite. Gold has been observed in association with all the sulphide minerals present, and as inclusions within pyrite and arsenopyrite. Electron microprobe analyses have revealed significant levels of gold in arsenopyrite, probably as submicroscopic inclusions.

M.Comm. / The South African gold mining industry has in the last decade increasingly come under pressure. The radical drop in the dollar price of gold, the lowest it has been in 16 years, has depressed revenues to the point where drastic action is needed. Furthermore, costs have been escalating. Where the South African Industry was the international leader in low cost production ten years ago, it has now become the most expensive. The following statement by the Chamber of Mines President, Bobby Godsell (1998: Godsell) , underlines the problem: "The gold mining industry world-wide is in a state of crisis, with the price of its product at a 16-year low. Few industries would not be in crisis if they were receiving 1982 prices for their product. In South Africa, two other features contribute to an even more acute crisis. Over the last decade South Africa has moved from being the lowest cost major gold producer to being the highest. Secondly, many South African gold companies are facing physical depletion of their ore reserves. Many companies have now been in operation for more than half a century. Eventually every mine must come to the end of its mineable reserves. Members of the Chamber are deeply committed to meeting the challenges of low prices, high costs and one depletion constructively and effectively. To this end the Chamber proposed a productivity-linked wage agreement last year. Since that agreement we have seen a further $30-drop in the gold price, with a third of Chamber member gold mines making losses, aggregating about R12-million a week. The Chamber shares the National Union of Mineworkers' concerns about job losses. Retrenchments are the consequences of the industry's problems - not their causes. Without addressing the causes, a moratorium on retrenchments will, in our view,achieve nothing, and possibly create expectations which would not be fulfilled. There is only one way to reduce job losses and this is to find ways to return companies to profitability and to make them cost-competitive with producers of gold elsewhere in the world. Much as it understands the National Union of Mineworkers' anger, the industry will not find a way forward through marches, protests or strikes. The combined wisdom, and then the combined effort of management, labour and government will be required to find a way forward for the South African gold industry."

D.Ing. (Civil Engineering) / This thesis describes an Investigation Into the nature of the fracture and deformation mechanism which occur at the edges of tabular gold mining excavation. Published Information on these phenomena Is reviewed, and the necessary underground Investigation required to consolidate the previous work Is described. It Is concluded that the rock near the reef plane at the edges of these mining excavations Is subject to stresses sufficiently high to cause It to fracture through the formation of regular patterns of shear planes. These fractures can form In the solid rock some distance ahead of the mining excavation. Nearer the mining face, extension fractures form which result In slabbing or splitting of the exposed rock. An Idealization of the observed rock behavior Is proposed, which Is then incorporated with conventional boundary element techniques Into a numerical model (SEAMS) which Is capable of analyzing two Dimensional tabular mining excavations where the rock near the reef plane at the edge of the mining excavation fractures, deforms and sheds load. A Sensitivity analysis of the numerical model Is described which Identifies those mining parameters capable of being used to advantage In controlling the size of the fracture zone.

D.Phil. (Geology) / The structural history and mechanisms of bedding-parallel faulting associated with the Ventersdorp Contact Reef (VCR) on the Kloof Gold Mine are discussed. Pre-VCR deformation is determined by folding of the Booysens Shale Formation and the Turffontein Subgroup. Extension of Ventersdorp age (± 2700 Ma) probably took place by the simultaneous operation of normal, strike-slip and oblique slip faults (mixed-mode extension). Because the strikeslip and oblique-slip faults were extensional, they could be utilised as conduits for dykes, forming "Fault and Dyke Zones". Northwestward directed thrusting, which postdates the normal faulting, resulted in positive inversion. The thrust faults have a ramp-flat symmetry with ramps forming in the VCR horizon and the overlying Alberton Formation and flats forming in the underlying Booysens Shale Formation and along the contact between the VCR and the Alberton Formation. The thrust faults follow complex deformation paths, indicated by out-of-sequence thrusting, simultaneous folding and thrusting, underthrusting and compressed boudins. Hanging-wall ripouts, which can be related to this thrust event, are introduced as a shear sense indicator. No northwestward verging thrust faults of post-Transvaal age (post 2430 Ma) could be identified, indicating that this compressional event occurred prior to the deposition of the Black Reef Quartzite Formation. The minimum amount of shortening due to northwestward directed thrusting was estimated as 37 % with the help of restored sections. However, due to the out-of-sequence propagation of thrust faults, a control of the strain estimation was possible and true shortening probably exceeded the calculated amount of shortening considerably. The restoration of sections with out-of sequence thrusts is discussed in detail. Sheath fold-like structures in pseudotachylyte can be related to northwestward directed thrusting and are introduced as a new shear sense indicator.