Thesis (MSc)--Stellenbosch University, 2003. / ENGLISH ABSTRACT: Mine water, that is all forms of water associated with and affected by mining operations is probably the largest area of concern in the environmental geology of mining. This study looks at the inter-relationship between mine water geochemistry and mine water management. The objectives of this study are: 1. To examine major geochemical processes influencing mine water; 2. To apply new national water and environmental legislation to the mining industry and discover how mine water management will be regulated; 3. To examine the approach of Integrated Catchment Management, and discover how this approach can be applied to the mining industry; 4. To conduct two case studies, where environmental geochemistry, general geochemistry and multivariate analyses are used as tools to investigate groundwater contamination problems in mining areas, and therefore suggest mine water management interventions; and 5. To, considering mine water problems discovered in the case studies, develop a newly-patented chemical treatment method for possible application in the mining industry. The changing legal framework - principally the implementation of the National Water Act (Act No 36 of 1998) and the National Environmental Management Act (Act No 107 of 1998) - is leading to the responsibilities of a mine becoming substantially clearer, and responsibilities which in the past could have been ignored until public outcry will now be difficult to escape. Two case studies are investigated. In both cases, general geochemistry and hydrogeochemistry, coupled with factor analysis are used to determine the major signatures in groundwater chemistry and the major sources of contamination. On the basis of this, management interventions are suggested. In the first case study, Sishen Iron Ore Mine of the Northern Cape, three signatures are identified in the groundwater: a clean dolomitic water signature, and a contamination signature from the mine and one from agriculture. The extent of nitrate and diesel contamination is shown to be related to agriculture, the use of explosives and to mine workshops and depots. Surface water controls may help reduce these problems. In the second case study, West Driefontein Gold Mine of the Far West Rand, two signatures are identified in the groundwater: a clean dolomitic water signature and a signature relating to contamination from the mine. Cluster analysis is used to suggest three groundwater zones, the chemistry of one being fairly clean dolomitic aquifer, the chemistry of the second affected by the mine and the chemistry of the third affected by granites. Contamination is shown to be related to mining operations, especially the slimes dams, and agriculture. Rehabilitation of dams and dumps, as well as surface water controls may help reduce these problems. Considering mine water problems discovered in the case studies, a newly-patented chemical treatment method is examined and tested for possible application in the mining industry. It is shown to be extremely effective for the removal of calcium from water, but less effective for the removal of iron and manganese. More broadly, this study shows the inter-dependence between mine water geochemistry and mine water management, and the need to be multi-disciplinary in approach. / AFRIKAANSE OPSOMMING: Mynwater, dit is aile soorte water wat geaffekteer en geassosieer word met mynbou bedrywighede, is moontlik die grootste rede tot kommer in omgewingsgeologie van mynbou. Hierdie studie stel ondersoek in na die verhouding tussen mynwater geochemie en mynwater bestuur. Die doelwitte van die studie is soos volg: 1. Om die hoof geochemiese prosesse te ondersoek wat mynwater beinvloed: 2. Om die nuwe nasionale water- en omgewingswette toe te pas op die mynboubedryf en te ondersoek hoe die bestuur van mynwater gereguleer sal word; 3. Om die benadering van Gerntegreerde Opvangs Bestuur ("Integrated Catchment Managemement") te ondersoek, en te ondersoek hoe hierdie benadering op die mynbou industrie toegepas kan word' 4. Om twee gevallestudies te doen, waar omgewingsgeochemie, algemene geochemie en multivariant analise gebruik word as middels om grondwater besoedeling in mynbougebiede te ondersoek, en daarvolgens om mynwater bestuurswysigings voor te stel; en 5. Om, nemende in ag die mynwater probleme wat ontdek is in die gevallestudies, 'n nuutgepatenteerde chemiese behandelingsmetode, vir moontlike toepassing in die mynbou industrie, te ontwikkel. Die veranderende regtelike raamwerk - hoofsaaklik die implementering van die Nasionale Waterwet (Wet No. 36 van 1998) en die Nasionale Omgewingsbestuur Wet (Wet No.1 07 van 1998) - lei daartoe dat 'n myn verantwoordelik moet wees om opmerklik skoner te word, verantwoordelikhede wat in die verlede maklik gergnoreer kon word maar wat nou te moeilik is om te ignoreer as gevolg van publieke uitroepe. Twee gevalle is ondersoek. In be ide studies word algemene geochemie en hidrogechemie, saam met faktoranalise, gebruik om die hoof kenmerke te bepaal in grondwaterchemie en die hoof bronne van besoedeling. Deur dit as basis te gebruik word bestuurswysigings voorgestel. In die eerste gevallestudie, Sishen Ystererts Myn in die Noordkaap, is drie kenmerkende samestellings qeidentitiseer in die grondwater: 'n skoon dolomitiese samestelling, en een elk van 'n myn en landbou gekontamineerde samestelling. Die omvang van nitraat en diesel kontaminasie word aangedui as geassosieerd met landbou, die gebruik van plofstowwe, mynwerkswinkels en depots. Oppervlak waterbeheer mag help om hierdie probleme te beheer. In die tweede gevallestudie, Wes Driefontein Goudmyn in die Ver Wesrand, is twee kenmerkende samestellings in die grondwater gerdentifiseer: 'n skoon dolomitiese samestelling en 'n samestelling geassosieer met kontaminasie van die myn. "Cluster" analise is gebruik om drie grondwatersones te identifiseer, die eerste een se chemie stem redelik ooreen met 'n skoon akwifer, die tweede een se chemie is bernvloed deur die myn en die derde se chernie is deur granite bemvloed. Kontaminasie word aangedui as geassosieer met mynboubedrywighede, veral die slikdamme, en landbou. Rehabilitasie van damme en afvalhope, asook oppervlak waterbeheer mag help om die probleem te verminder. Deur die mynwater probleme wat in die gevallestudies ontdek is in ag te neem, word 'n nuutgepatenteerde chemiese behandeling ondersoek en getoets vir moontlike toepassing in die mynboubedryf. Dit word aangewys as uiters effektief vir die verwydering van kalsium, maar minder effektief in die verwydering van yster en mangaan. In die algemeen, wys hierdie studie inter-afhanklikheid tussen mynwater geochemie en mynwater bestuur, en 'n behoefte aan 'n rnultidissiplinere benadering.

Current mining methods used to extract coal from underground mine workings disturb the natural environment and the existing stable geological structures. As a result, the ingress of water into the mines increases and the quality of the water passing through the mine workings deteriorates, irrespective of the operational status of the mines. Water ingress is generated by regional aquifers, local aquifers, recharge from the surface through rainfall, natural drainage paths on the surface, and surface water bodies. The quality of water in the mines deteriorates as a result of contact with the remaining coal in the mine workings. Mining can therefore cause an increased influx of water into a mine and the degradation of this water. The solution to reducing the impact of mines on the environment is to prevent, or at least reduce, the amount of water entering the mines, and to manage this water to prevent further degradation in water quality. This study focused on afforestation with Eucalyptus viminalis trees to manage or inhibit ingress of water into underground mine workings. The hypothesis of this study was that a change in vegetation, from grassland to fast-growing and potentially high water-using trees like Eucalyptus. could possibly reduce the drainage of water below the root-zone and into the mine workings. The hypothesis was tested by estimating the components of the soil water balance for a grassland site and a Eucalyptus tree site. The research site was situated in Mpumalanga, (260 36' Sand 290 08' E, 1650 m a.m.s.l.), one of South Africa's major coal bearing areas. Although the Secunda area is a treeless environment and conditions are not optimal for forestry, some Eucalyptus species are suited for conditions (frost and periodic droughts) encountered in this area. The soil water balance of grassland and E. viminalis trees were studied through a field experiment and a long-term (30 years) modelling exercise. Total evaporation of the grassland site was estimated using the Bowen ratio energy balance technique. The transpiration of six representative E. viminalis trees were estimated using the heat pulse velocity technique. The soil water storage changes at both sites were determined from the soil water content, estimated using water content reflectometers. Measurements were performed in a smectic clay soil which resulted in measurements difficulties. Vertical cracks were formed under soil drying. To establish the importance of climate and plant growth on the drainage beyond the root-zone, the soil water balance of a grassland and an E. viminalis site were simulated over a 30-year period with the Soil Water Atmosphere Plant (SWAP) model. It was concluded from the comparative field experiment and modelling, that a change in vegetation from grassland to E. viminalis will reduce the drainage of water below the root-zone, especially under above-average rainfall conditions. The reduction in drainage beyond the root-zone at the E. viminalis sites, compared to the grassland site, was demonstrated in the modelling exercise and can be deduced from the total evaporation and soil water storage estimated at both sites. The results from the field experiment confirmed the modelling results and showed that usually there were higher transpiration rates for the E. viminalis tree site, compared to the grassland site. The higher transpiration rates for E. viminalis trees resulted in lower relative saturation of soil layers and lower profile soil water contents at the E. viminalis site, and higher daily soil water storage changes at the E. viminalis site compared to the grassland site. These differences were more pronounced during winter when the grassland was dormant. The results from the modelling exercise showed that an E. viminalis tree stand, with a closed canopy, reduced drainage below the root-zone compared to a grassland. The drainage at the grassland site contributed to up to 54 % of the rainfall, compared to the 43 % at the E. viminalis site. However, under below-average rainfall conditions the annual drainage at both sites, were similar. Further, the absolute magnitude of the drainage was similar to the total evaporation at the grassland site under certain conditions. The results not only suggest that a change in vegetation, from grassland to E. viminalis trees, would reduce the drainage beyond the root-zone, but that it may delay the onset of drainage. Under above-average rainfall conditions, the modelled drainage at the E. viminalis site only exceeded 20 mm, a month later than at the grassland site. The simulation results also showed that under conditions of aboveaverage rainfall, drainage occurs whenever the rainfall exceeds the long-term average rainfall, irrespective of the existing vegetation. However, when the rainfall is belowaverage drainage at both sites are limited to large rainfall events. This simulation showed that over a period of eight years, E. viminalis trees could potentially reduce the drainage by 1235 mm more than grassland, which is equivalent to 1540 m3 ha- I a-I, or 1.54 Me ha- I a-I. The annual average reduction in drainage below the root-zone caused by E. viminalis trees (1.79 Mf ha-1 a-\ is a small reduction when compared to the influx of water into mineworkings. E.g. the influx of water into a bord-and-pillar mine range between 0.5 and 4 Mt d-I per area mined and up to 17000 Mt d-I per area mined under high extraction mining (Hodgson and Krantz, 1998; Hodgson et aI., 2001). This work gave a comprehensive account of the differences in the soil water relations of grassland and E. viminalis trees overlying coal mine working. Few other studies in South Africa compared the total evaporation and soil water relations of grassland and E. viminalis trees in so much detail. State of the art monitoring techniques were used and produced valuable comparison of their use in expansive clay profiles. The work should contribute to management decisions focussed on limiting ingress of water into mine workings. / Thesis (Ph.D.)-University of Natal, Pietermaritzburg, 2003.

Ph.D. / Mining operations significantly influence the environment due to direct and indirect discharges of waste products into the aquatic systems. The primary aim of this study was to assess the current situation in the platinum mining area and develop a management plan to ensure that existing and potential environmental impacts caused by platinum mining and processing are mitigated. To do this, an assessment was carried out to investigate changes in critical aquatic invertebrate and fish community distributions and assess how they relate to measured environmental factors. Five sites were selected, one reference site which is upstream of heavy mining activities and four sites within heavy mining and processing activities. Standard techniques for water, sediment, invertebrate and fish sampling were used. Macro-invertebrates sampled were identified to family level whereas fish were identified to species level. Multivariate analysis used was cluster analysis by non-metric multidimensional scaling (NMDS) for both macro-invertebrates and fish. Three methods of ordination were used to analyze the biotic and abiotic data namely N-MDS, Correspondence Analysis (CA) and Canonical Correspondence Analysis (CCA). Cluster analysis of macro-invertebrates data revealed three major groups based on sampling period (low flow or high flow) and the last cluster according to the locality. Multidimensional scaling ordination of high and low flow for macro-invertebrate communities confirmed the groupings detected by cluster analysis. Cluster analysis for fish communities revealed two groups at 50% similarity; the first group is the combination of reference and exposure sites for both high and low flow sampling regimes. No fish were sampled at site 4 during both low and high flow regimes. Multidimensional scaling ordination of high and low flow fish communities confirmed the groupings detected by cluster analysis. Analysis using a similarity profile (SIMPROF) test indicated that fish communities are statistically (p=5%) the same. It was found that macro-invertebrates and fish respond differently to environmental variables.