Other research programmes

Yüklə 2,78 Mb.

səhifə	6/42
tarix	16.01.2019
ölçüsü	2,78 Mb.
	#97487

1 2 3 4 5 6 7 8 9 ... 42

The scope of the DeepMine, FutureMine, Coaltech 2020, and PlatMine collaborative research programs are described in substantial detail, as the reports are not generally in the public domain, unlike the SIMRAC reports.

3.2.1Chamber of Mines Research Organisation (1964 - 1993)

Prior to the establishment of COMRO, mining research in South Africa had been mostly reactive, directed at finding solutions to problems as they arose.

Research began in 1889, three years after the discovery of gold near present-day Johannesburg, when surface deposits of gold were exhausted and underground mining commenced. The MacArthur-Forrest cyanide process was developed (in Scotland) to extract gold from unoxidised ore.
Research into rockbursts and the effects of rock pressure started in 1910 and continued on an ad hoc basis thereafter.
The serious effects of silica dust on the health of mineworkers led the Transvaal Chamber of Mines to establish the Dust Research Laboratory in 1914.
By 1924 mines reached depths at which wet-bulb temperatures exceeded 30°C and heat stroke emerged as a hazard. Systematic studies of heat tolerance and heat acclimatisation were then initiated.

The Chamber of Mines Research Organisation (COMRO) was established following the Coalbrook Colliery disaster in January 1960 in which 435 men died. It was the worst accident in South Africa’s mining history. The official enquiry into the disaster found that no scientific basis was available for the design of pillar workings in coal mines, and highlighted the need for a far more systematic approach to research.

COMRO was established in 1964 to address issues such as pillar design in coal mines and the threats to the gold mining industry of increasing depth and working costs and a stagnant gold price. It was funded on a cooperative basis by the six major mining houses operating in South Africa at that time. By 1986 COMRO employed nearly 700 people. Rock engineering work focused on three main areas:

Mine layout, aimed at minimising the effect of rock pressure at the design stage. The MINSIM computer program is an example of one of the research products. Another significant output was the publication of Guidelines for Ameliorating the Hazards of Rockfalls and Rockbursts (first edition 1977; second edition 1988). This book has been superceded by SIMRAC’s A Handbook on Rock Engineering Practice for Tabular Hard Rock Mines, (A.J. Jager and J.A. Ryder, 1999).
Support units and systems, aimed at reducing falls of ground and the extent of rockburst damage. The development of the rapid-yielding hydraulic props represented the first modern breakthrough in deep stopes. Backfilling was another major theme.
Rockburst control, which was concerned with developing seismic and engineering techniques to control the rockburst hazard. Seismic research at the Bernard Price Institute of Geophysical Research, led by Dr Art McGarr, was also supported.

In 1993 COMRO merged with the CSIR, and the CSIR Division of Mining Technology (CSIR Miningtek) was established. In 2005 the staff complement of CSIR Miningtek was about 150, including about 70 graduate engineers and scientists.

3.2.2DeepMine (1998 - 2002)

In 1996, Dr Güner Gürtunca, Director of CSIR Miningtek, advanced the concept of a collaborative research programme to create the technological and human resources platform for mining gold safely and profitably at depths of 3 to 5 km. The proposal was presented to the South African Institute of Mining and Metallurgy and the Foundation for Research Development (now the National Research Foundation). A forum of interested parties appointed a steering committee to review the proposal, formulate a research strategy and draft a business plan. It was decided that the initial focus should be on the extension of existing technology rather than on the development of radically new technology. The business plan was presented to the DeepMine Forum in February 1998 and AngloGold (now AngloGold Ashanti), Durban Roodepoort Deep, Gold Fields, the Chamber of Mines of South Africa, and CSIR confirmed their support. Research work commenced in July 1998.

The following process was used to structure the research effort:

The ultra-deep environment was defined, e.g. rock types, stress, virgin rock temperature.
Systems criteria were established, e.g. maximum Energy Release Rate, maximum temperature in working places, desired face advance rate.
It was established whether currently available technology could meet the systems criteria.
If the ultra-deep environment or systems criteria were not known, or technology did not meet the criteria, research was then conducted to fill the knowledge or technology gaps.

In the first year (1998/99), emphasis was placed on those issues considered to be serious obstacles to mining at ultra-depth, e.g. the effect of barometric pressure changes on health, the risk of mining-induced seismicity, and the cost of refrigeration and ventilation. During the second year, a new focus on the transport of men, material and rock was introduced. During the third year, the emphasis was on the integration of the research findings and the formulation of guidelines for the designers and operators of future ultra-deep mines. During the fourth and final year, emphasis was placed on the transfer of knowledge and technology through activities such as schools, consultancy reviews on mines operated by the industrial partners, and the commercialisation of services and technologies.

The DeepMine Program had a total budget of R66 million, a duration of four years, and covered a wide range of disciplines, including industrial sociology, physiology, mining and mechanical engineering, and the earth sciences. Over 250 researchers were involved. The rock-related research work was conducted within the ambit of several projects. It should be noted that SIMRAC work formed the basis for many of these projects.

3.2.2.1Mapping of geological structures ahead of mining

Objective: A foreknowledge of geological structures is crucial for safe and productive mining at ultra-depth. This technology project set out to develop tools to detect all structures that disrupt the reef by more than 2 m at distances of up to 200 m ahead of stoping.

The first step was the compilation of a catalogue describing the geotechnical and geophysical characteristics of the potential ultra-deep mining areas. The capability to integrate and visualize geological, geophysical, rock engineering and mining data for mine design and planning applications was then developed (Drummond et al., 2001). As boreholes can be used to deploy geophysical tools to investigate the orebody ahead of mining, the stability of boreholes drilled in rock subjected to stress levels similar to those expected at ultra-depth was investigated. It was found that cover and exploration holes generally stay open for several months, though the collar of the borehole is frequently blocked when tunnels are supported and equipped. Mine Seismic Profiling (MSP) was identified as the seismic technique with the greatest potential for mapping discontinuities in the reef at the desired range and resolution. However, tremendous difficulties were experienced in acquiring data due to seismic noise generated by mining activities, and it was concluded that the method was not suitable for routine application (Stevenson et al., 2002). Borehole Radar, which had been identified as the electromagnetic imaging technique offering the greatest potential, gave encouraging results. It proved possible to detect the Ventersdorp Contact Reef at a range of up to 80 m from the borehole (Trickett et al., 2000; Vogt, 2002).

3.2.2.2Mining layouts and methods

Objective: To establish criteria for optimal mining layouts and to develop appropriate in-stope processes for each geotechnical area expected at ultra-depth, taking into consideration rock engineering, resource productivity and environmental criteria.

The critical systems criteria for ultra-deep level mining were identified and quantified:

Rock engineering parameters for exploration, mining, support and monitoring;
Environmental conditions;
Mining and rock breaking; and
Transport of men, materials and rock.

It was noted that some criteria, such as Energy Release Rate (ERR) and peak particle velocity (ppv), are poorly defined. It was found that established technologies could meet the systems criteria, except in areas of exceptionally high stress.

In order to evaluate mining layouts and sequences, the mining of a realistic ultra-deep ore body was simulated (Vieira et al., 2001). The advantages and disadvantages of various layouts were explored, with detailed assessments of rock engineering (Vieira & Durrheim, 2001), refrigeration and ventilation (Bluhm & Biffi, 2001), transport, and economics. Guidelines for ultra-deep mining were formulated. No single layout emerged as the best; rather, the characteristics of an ore body (e.g. nature of faults and dykes, variability of the gold grade) were viewed as dictating the preferred layout and mining sequence.

3.2.2.3Stope support

Objective: To design and develop cost-effective and user-friendly stope support systems that will enable safe and economic mining at depths between 3 and 5 km under static and dynamic loading conditions, whilst minimizing material transportation requirements.

The likely geotechnical environments at ultra-depth were established. No new rock types are expected, although the thicknesses of the various strata change, with the ore body becoming increasingly quartzitic. Although the intensity of rock fracturing and deformation increase significantly with depth, it is expected that the fracture envelope will remain much the same as at current mining depths (Güler et al., 2000). Stress was successfully measured at a depth of 3352 m, which is 700 m deeper than any previous stress measurement in a South African gold mine. The vertical stress was measured to be 91,0 MPa, in close agreement with the calculated overburden stress of 88,8 MPa.

The rock engineering criteria and the consequent support requirements were established. It was concluded that, in general, present support technologies were adequate to cater for rockfall and rockburst conditions expected at ultra-depth, although areal coverage becomes increasingly important as depth increases. It was found that commonly used rockburst-resistant support systems (e.g. yielding timber elongates or timber packs with rapid yielding hydraulic props or backfill and rapid yielding hydraulic props) were all capable of meeting the criteria for the face area support of narrow stopes. An abundance of suitable supports existed for gully edges. There was, however, an evident need to improve support systems for gully hangingwalls subjected to seismic shaking.

The role of backfill in back area support was examined and a new backfill design criterion for ultra-deep level mining was formulated (Gürtunca et al., 2001). This criterion specifies that backfill should be placed in 60 to 70 per cent of the area mined with a maximum fill-to-face distance of 3 m. The vertical stress generated in backfill should be between 2-3 MPa at a distance no greater than 25 m from the face at a corresponding strain value of 4-6 per cent. The face area should be supported by hydraulic props or elongates with sufficient yielding capacity. It was also found that backfill comprising aggregates/tails offers high potential for good performance, but that these systems are expensive and difficult to control. Comminuted waste was found to be easier to control and only marginally less effective and was thus considered the most attractive option for ultra-depth (Ilgner, 2001).

3.2.2.4Seismic management

Objective: To acquire the necessary understanding of seismicity and develop techniques to manage seismicity so that rockbursts do not prevent productivity and safety objectives from being met.

Studies were conducted to establish the influence of depth, mining method and face advance rate on seismicity. Mining in the Carletonville district had taken place at depths ranging from 1500 m to 3400 m below surface. Analysis of the data did not reveal a sufficiently clear dependence on depth for allowing the seismic hazard at ultra-depth to be estimated with confidence. A new methodology, based on the concept of Volumetric Energy Release and Potential Damage Area, was developed to estimate the relationship between seismicity and depth. An assessment of the relative effects on seismicity of mining by blasting or by continuous non-explosive processes concluded that the overall seismic hazard, which is dominated by the larger seismic events, is independent of the rock breaking process. In determining the effect of face advance rate on seismicity, it was found that the maximum safe mining rate was dependent on the geotechnical area. Slightly higher mining rates may be sustained at ultra-depth without an increase in the hazard of face bursting because time-dependent fracture processes take place more rapidly.

The integration of seismic monitoring and numerical modelling as an aid to seismic management was investigated and prototype software developed (Spottiswoode, 2001; Hofman et al., 2001; Lachenicht et al., 2001; Wiles et al., 2001). In a review of current seismic prediction and hazard assessment practice and capability, it was found that predictions might be 2-3 times better than random. However, prediction is not sufficiently reliable to be used as a management tool.

The management of seismicity was found to be one of the major challenges to mining at ultra-depth, and a holistic strategy is required to limit seismic events and protect excavations from shaking. The proposed strategy had three main components: a reduction in seismicity by using geophysical methods to identify seismogenic structures (dykes, faults) ahead of mining and designing layouts accordingly; the use of appropriate support systems to create rockburst-resistant excavations; and the continuous monitoring of seismic hazard. It was concluded that, with foreknowledge of potentially seismogenic structures, it is possible to adapt rock engineering technologies (mine layouts, stope and tunnel support systems, etc.) to manage the levels of seismicity expected at ultra-depth (Durrheim, 2001).

3.2.2.5Access development and support

Objective: To evaluate and develop techniques for safe, cost-effective and rapid access to orebodies, such that these access-ways are conducive to the rapid transport of men, materials and rock and remain functional for their required life.

Fracturing and associated damage around circular openings in highly stressed rock at depths of 3 to 5 km were investigated and it was concluded that it is feasible to excavate tunnels at ultra-depth, unless adverse rock conditions are encountered (Kuijpers, 2000; Sellers & Klerck, 2000; Diering, 2000). Investigations into systems criteria for access development found that drill and blast was more cost-effective than tunnel boring in most scenarios anticipated in the development of an ultra-deep mine (Willis & Rupprecht, 2002). An optimum mining cycle was devised which was designed to advance a single tunnel 150 m a month using a 3 m advance per blast, increasing to 180 m per month with a 5 m round. In a search for equipment to remove broken rock from the face in both flat and inclined development ends, it was established that loading rates are not the limiting factor in obtaining high rates of face advance. The entire cleaning system should be addressed, and the components of the system matched so that all operate close to maximum performance.

It was concluded that support components and systems are available for most ultra-deep tunnels, and the cost of a yieldable support system, desirable at ultra-depth, should not differ significantly from conventional systems currently in use. Surface support liners such as shotcrete were concluded to be mandatory (Durrheim, 2002). A methodology was developed to design the most cost effective tunnel support system under dynamic loading conditions. It was found that the most cost effective system makes use of relatively low strength tendons with a high yielding capacity, as this enables the shotcrete thickness to be minimised. A survey of international and local shotcrete usage showed that wetcrete offers more scope for optimisation, in terms of both transportability and quality control. Underground comminution of waste rock, batching and distribution to multi-end developments for use in shotcreting becomes increasingly viable as depth increases.

3.2.2.6Transport of men, material and rock

Objective: To identify and develop those technologies that will ensure the rapid and safe transport of men and the efficient transport of materials and rock between surface and the working place.

Issues relating to shaft sinking and operation were investigated. It was decided that ultra-deep shafts would need to be of high output (±400 000 ton/month) to offset the high initial capital costs. An audit of current shaft system operations, efficiencies, design, and technologies was carried out, and it was established that this criterion could be met with the technology currently available. The only major concern is that no ropes have been proven capable of meeting the required duties of the main shaft. Multi-level in-shaft loading installations were shown to be a viable alternative to internal shaft ore passes for rock transportation and to offer lower capital and operating lifecycle costs than internal shaft ore pass installations for the layouts and systems considered.

A review of current material transport systems showed that these technologies could meet the quantitative systems criteria for ultra-deep mines. However, there was concern that these systems might not meet the requirements pertaining to safety, and Guidelines for Design and Best Practice were formulated. Introducing lightweight monorail systems for crosscut to face transportation and chain-driven material handling devices in panels was found to make the greatest improvements in material handling. Substitute materials (e.g. piped grouts) were identified as possible options for reducing the quantity of materials used in stopes, with the largest savings likely to arise from a reduction in consumption and handling of timber. With regard to rock transport, various technologies for rock comminution and lateral hydraulic transportation of rock were investigated, but it was concluded that conventional rail-bound transportation was more cost-effective and provided greater flexibility. Underground milling and hydraulic distribution of development waste for use in backfill could, however, prove viable in some situations.

Best practices for the design, support, and systematic maintenance of shaft rock passes at ultra-depth were identified (Hagan & Acheampong, 1999). Investigations into the lining of ore passes concurrently with boring concluded that the design and engineering of a suitable system does not pose any major problems, as most components are either available or currently undergoing prototype development. An investigation into stope rock passes showed that they will fail at depth as currently configured (Rupprecht, 2001). It was found that the number of ore passes may be reduced significantly and still deliver the required output, while mechanised methods of developing ore passes were preferred as they allow practical, safe and timely support.

3.2.2.7Knowledge transfer

A range of vehicles was used to transfer the knowledge gained during the course of the DeepMine Program to practitioners (Durrheim & Diering, 2002). Aspects of the knowledge transfer process that were novel to the South African mining industry include:

Summary reports and guideline handbooks that seek to consolidate the huge body of knowledge comprehensively and concisely, highlighting the interactions between the different systems that make up a deep-level mine. The publication of these documents on CD, cross-linked to related topics and down-linked to source documents, was particularly effective.
A series of DeepMine Schools designed to enable industry practitioners to apply the knowledge to their current work situation.
Consultancy reviews, where DeepMine researchers sought to apply new knowledge to specific real problems on mines.

3.2.3FutureMine (2001 - 2004)

The FutureMine Programme was the successor to the DeepMine Programme. In an environment characterised by a volatile gold price, mature ore bodies and aging infrastructure, the gold mining companies recognised that the major challenges facing the industry were the reduction of working costs (from US$250 per oz to US$170 per oz or lower) and the improvement of health and safety. Meeting these challenges would not only result in an increase in profit, but the reduced pay limit would convert lower grade resources into mineable reserves, thereby increasing total output and extending the life of the mines. All major gold mining companies operating in South Africa were invited to the inaugural meeting, held in October 2000. African Rainbow Minerals, AngloGold, CSIR Miningtek, Gold Fields, Harmony, and Placer Dome Western Areas JV nominated representatives to serve on a steering committee, and workshops were held to formulate the research tasks. The business plan was presented to stakeholders in the gold mining industry in June 2001, and research work commenced in August 2001. The budget for the first three-year phase was about R50 million.

FutureMine complemented and extended DeepMine in the following respects:

DeepMine addressed potential future ultra-deep operations, while FutureMine addressed current underground gold mining operations. Consequently, DeepMine tended to be knowledge-driven, while FutureMine was technology-driven, with a strong emphasis on the implementation of solutions within a three-year horizon. FutureMine research teams worked closely with manufacturers and technology suppliers.
DeepMine worked on the assumption that a new part of an ore body would be developed, providing the opportunity to install new technologies and systems from the outset. In contrast, FutureMine generally worked within the constraints of existing infrastructure, and the costs and benefits of modernising or replacing technologies and systems had to be evaluated.

All rock-related research work was carried out in tasks included in the Stoping Technology project, whose objective was defined as follows: To provide implementable technologies that will radically improve the safety and efficiency of the stoping process.

3.2.3.1Rock-breaking methods and systems

A survey of rock-breaking methods and systems was conducted. The findings were fed back to the manufacturers and suppliers of technology to enable improvements to be made, and were also used to identify research priorities.

Conventional drill and blast: It was found that hand-held drills are widely used, but are associated with low mining efficiencies and safety. Numerous drilling jigs and rigs have been developed. (The term “jig” is used to describe a drill that is mounted on a bar between temporary roof supports, has only one drill and feed per unit in a stope width of up to 1,2 m, and is fully operator dependent for the accuracy of drilling as well as for the movement required to traverse the face. The term “rig” is used to describe a system with at least two booms with drills and feeds that are ‘fixed ‘ at preset angles to the face, with a self-powered unit to traverse the face. The rigs are usually mounted on sleds or wheels running on rails, tracks, or the footwall.) It was found that the use of drill jigs was increasing, despite their current limitation to stoping heights in excess of 90 cm. While mines had acquired more than 100 drill rigs, fewer than 20 were in regular use owing to their mechanical complexity and the requirement that the hangingwall be supported by roofbolts. Nevertheless, rigs are believed to have specialised applications where high mining and labour productivity are required and the geology enables their use. It was concluded that the longhole drilling system has the potential to increase productivity to the point that previously uneconomic low-grade ore could be profitably mined.
Preconditioning: It was found that preconditioning could reduce the risk of face-bursting in high stress conditions, and bring about significant improvements in productivity and safety. The experience gained on numerous mines was consolidated and best practice guidelines were formulated.
Systems to pump explosives for in-stope use were investigated, and pump reliability was found to be the weak link.
A survey of fume-free rock breaking methods identified four promising technologies: activated rock cutting, the water pulse gun, the Impact Mining System (IMS), and the Mini-Disc™ cutter. It was decided to concentrate development work on the IMS, as this technology was the most advanced. As the IMS depends on the presence of pre-existing fractures, it is suited for work in environments where stresses are high enough to induce fracturing. However, work on the IMS project was terminated as no mining company was prepared to offer a test site.
Low-energy blasting techniques, which use either low energy explosive packages (‘tailored energy packs’) or propellants, have the potential to radically increase face advance rates as they enable mining to take place continuously. The evaluation of current performance revealed that significant problems remain to be overcome before these techniques can be implemented on a wide scale, notably the accumulation of gas and dust to levels above the legal limits in production stopes.

3.2.3.2In-stope processes

A survey of various in-stope processes was conducted, best practice guidelines were formulated, and research priorities identified.

Face, strike, and dip gully cleaning: It was found that the cleaning of strike gullies is a serious bottleneck on many mines. The development of an automated strike gully scraper winch that can be controlled remotely by an operator or run autonomously was identified as an opportunity to improve productivity and safety.
In-stope water management: The most recent and comprehensive work on gold losses suggests that the concern that water-jetting results in increased gold losses is unjustified. However, a concern remains that reefs with high carbon content should not be cleaned with water jets.
Shaft pillar extraction: FutureMine partners were in the process of extracting five shaft pillars at the time of the study, and a further ten extractions were scheduled within the following decade. The experience gained during many past pillar extractions was consolidated, and Best Practice Guidelines formulated. It is believed that improved efficiency and safety has the potential to increase revenue by 10%, equating to a total of R600 million for these fifteen shafts.
In-stope material handling and transport: The vision for material handling is a “hands-free” system that can carry relatively heavy objects such as winches from the crosscut to the point of use. The only technology currently capable of achieving this vision is the monorail, although it is expensive and time-consuming to install. In the short term to medium term the use of the monowinch appeared to offer the most acceptable compromise in terms of capacity, flexibility and cost. A system comprising a monowinch operating to the centre raise, combined with in-line monowinches operating to the face, was recommended.

3.2.4Coaltech 2020 (1999 - present)

Coaltech 2020 is a collaborative research programme that was formed in 1999 by the major coal companies, the national electricity generating company (ESKOM), universities, the CSIR, organised labour, and the state. Its objective is to develop technology and apply research findings that will enable the South African coal industry to remain competitive, sustainable, and safe. The potential impact of Coaltech 2020 is vast:

For every year the life of the Witbank-Highveld coalfield can be extended, a potential income of ±R30 billion will be realised.
Approximately 2.25 million people are dependent on the mines in the Witbank-Highveld coalfield.
To replace a power station the size of Kriel or Matla, plus a mine supplying it with coal, would cost approximately R16 billion (2002 prices).
If the superfine coal, which is currently being discarded, can be dried and agglomerated, it could earn an additional income of approximately R200 million per annum.

The Coaltech programme is independently managed by a Coaltech Management Team (CMT) which consists of representatives from Ingwe Coal Corporation (Billiton), Anglo Coal, Total SA, Xstrata Coal, Eyesizwe, Eskom, Sasol Mining, Kumba, Department of Minerals & Energy, National Union of Mineworkers, Mine Labour Caucus, the CSIR, the Chamber of Mines, the universities of Pretoria and the Witwatersrand, and the National Research Foundation (NRF). A key objective of the programme is to foster collaboration among and within each of the main parties involved in the programme. Another aim is the building of the human resources capacity that will result in an adequate supply of practising mining engineers and other required disciplines, with a high level of competence and commitment to the coal mining industry. The coal mining industry, the CSIR and the state fund the programme in equal proportions (i.e. one-third each).

Each research task in the Coaltech 2020 programme must address the following issues in addition to satisfying its specified research and technology needs:

Health and safety;
Risk assessment and economic evaluation;
Productivity and cost benefits;
Technology transfer and implementation;
Impediments to implementation (e.g. legislation); and
Education and training.

The research tasks or “needs” have been grouped into six project areas with a common technological theme. Rock-related research has been conducted in five of these areas. The exception is “coal processing”, which addresses the beneficiation and briquetting of fine coal. The rock-related research tasks are described below.

3.2.4.1Optimal reserve utilization

Task 1.1.1: Create sedimentological and structural models to assist in identifying geological features that impact on mining

Many of the problems experienced during the extraction of coal are directly related to geological disruptions in the continuity and quality of the seam. These disruptions may be both sedimentological and structural in nature. An understanding of the occurrence and distribution of these seam disruptions is important for the safe and efficient extraction of the remaining coal reserves in the Northern Witbank-Highveld Coalfield. To this end, a three-dimensional model of the regional geology was constructed.

Task 1.3.1a: The optimum methodology for conducting high-resolution, shallow reflection seismic surveys on South African collieries

The objective of this task was to improve the resolution capabilities of reflection seismic surveys such that they can become an integral aspect of coal mine planning, thereby leading to the selection of optimal mine layouts. The focus of this research is on the depth range between 30 and 100 m with a vertical and horizontal resolution of 2 m and 10 m, respectively. A technology audit and a number of surveys were conducted, and the minimum acceptable standards for field procedures were determined. It was found that the principal reason for reflection seismic surveys failing to provide results of an acceptable quality is insufficient preliminary data (vertical seismic profiles, noise spreads, etc.). It was concluded that reflection seismic surveys are unlikely to be used on a routine basis, as exploration drilling is more cost effective for mapping coal seams shallower than 150 m.

Task 1.3.1b: Detection of old workings, sinkholes, underground fires and acid mine pollution using geophysics and airborne thermal infrared imagery

An audit of various geophysical techniques was conducted to assess their applicability to solving typical coal-related problems such as the detection of old workings, sinkholes, underground fires and acid mine pollution. A number of promising techniques were identified and geophysical trials were conducted at Kromdraai, New Largo, Arthur Taylor, and Bank collieries. The following ground geophysical techniques were tested: micro-gravity, electrical resistivity methods (including resistivity tomography), magnetics, ground penetrating radar, and frequency-domain and time-domain electromagnetics (FDEM and TDEM). An airborne geophysical technique was also tested, namely the thermal infrared imaging (TIR) method. These surveys resulted in the compilation of a collection of case studies and a geophysical applications guideline for the Witbank-Highveld collieries. Micro-gravity, electrical resistivity and TDEM proved to be the most promising ground geophysical techniques in terms of locating the edge of old workings and cavity detection. The accurate mapping of individual bords and pillars, however, does not seem possible with any of the conventional techniques tested. Although it was not demonstrated as part of Task 1.3.1b, the electrical resistivity and TDEM techniques are also known to be well suited to mapping pollution plumes.

The TIR method is a very versatile and cost-effective tool for regional and mine-scale mapping. Despite a relatively high mobilisation cost, the overall cost per square metre is only a fraction of the cost of applying other conventional geophysical methods. The TIR surveys yielded some promising results. It appears to be the ideal reconnaissance tool for the regional detection of shallow sinkholes, fires, and possible pollution plumes. The TIR technique does, however, appear incapable of accurately mapping the exact positions of individual bords and pillars. Seismic techniques were evaluated by a number of experts. It was concluded that seismic methods are not well suited to applications such as the detection of cavities and workings in the South African coal environment. Limited experimentation with, for example, fan shooting refraction techniques was proposed. Novel and unconventional techniques were also investigated. One such technique involves the surface tracking of a miniature “submarine-like” electromagnetic tag that is deployed in flooded workings. Techniques such as this may be regarded as impractical, but a need for some new, innovative ideas to solving age-old problems was highlighted.

Task 1.4: The impact of geotechnical factors on high and low wall stability

This task investigates the impact of geotechnical factors on high wall and low wall stability with the objective of developing a methodology to predict failures. New Vaal Colliery, which suffered a number of sloughing events in 2002, was used as a test site. A set of intersecting joint planes, seepage water from an adjacent spoil pile, change in paleotopography and removal of the overlying burden were identified as possible factors.

Task 1.8.2: Critical geotechnical factors impacting on reserves in previously mined areas in the Northern Witbank-Highveld Coalfield

Coaltech 2020 aims to extend the life of the Northern Witbank-Highveld Coalfield beyond 2020. Large areas of this coalfield have been mined in the past using the bord and pillar method, leaving significant amounts of coal in pillars and as floor or roof coal. Improvements in machinery and technology now allow extraction of previously unmineable coal. However, the constraints on secondary extraction imposed by the geotechnical factors impacting on seam extraction and related safety aspects had not been previously investigated. The aim of this project was to devise a methodology to identify and quantify these factors for both open cast and underground mining methods. A literature review identified general geotechnical factors that impact on mining efficiency and safety during secondary extraction (underground and open cast). The review also revealed that very little publicly available literature exists with regard to the factors impacting on secondary coal mining. The general geotechnical factors were then categorised into nine classes. These classes were considered in relation to their impact on secondary underground and open cast extraction with regard to rock mass behaviour (roof caving characteristics, rockburst potential, rib and pillar stability, floor heave, roof/surface subsidence, slope stability, high wall stability), roof support (mine- and panel-scale roof support), and explosions and/or ignitions (gas conditions, spontaneous combustion).

3.2.4.2Underground mining

Task 2.4: Underground trials of seismic-counter and goafing characteristics in pillar extraction

A prototype stand-alone warning device, called GoafWarn, was developed to provide early warning of pending goafing. The warning algorithm is based on the temporal behaviour of the micro-seismicity in the immediate roof area. The algorithm was initially developed for the automatic detection of impending goafing at longwall mining at AusMine Colliery in Australia. The method was adapted to applications involving pillar extraction and roof stability. It has been shown that the primary mode of roof layer failure is by bending-induced tension. Direct tension can be excluded in most practical situations, and only in exceptional cases will the first failure be in shear mode. In roof layers with a high joint frequency, failure will occur at spans equal to the joint spacing. A fundamental procedure to evaluate the likely roof layer failure in a stacked beam system has been developed. This procedure can be used to estimate the loading condition on snooks and also to predict the positions of cavities in the overburden under the conditions where a full goaf has not developed.

Task 2.5.2: Economic and safe extraction of pillars and associated reserves for each category identified in Task 1.8.1 using underground mining methods

The task incorporates a literature survey of previous work done and various site visits, with the objective of identifying potentially suitable mining methods as well as limitations, shortcomings and factors constraining these methods (e.g. environmental impacts, strata control, ventilation, health and safety features). Although it was generally agreed that the risk associated with pillar-extraction mining is greater than for normal bord and pillar development, the risks appear to be similar if continuous miners are used for pillar extraction. The Nevid pillar extraction method was identified as a potential method for pillar extraction, and its application at shallow depths was investigated and the dimensions of the cutting widths and snook sizes established. It was also shown that panel widths and abutment angles are important factors in pillar extraction. Design flow sheets for rock engineering, environmental assessment, mining, beneficiation and financial evaluation were developed and integrated in an analytical hierarchichy process for selecting the most suitable mining approach.

Task 2.7.1: Determine and quantify the benefits of ashfilling (and slurries) in coal mining

The aim of the task is to increase the extraction recovery of coal, while reducing mining costs and improving the environment. Extensive laboratory tests were performed on fly ash from different power stations, and these tests showed that the achievable slurry concentrations varied between the different sources with the result that the amount of drainage water would vary considerably. Rock mechanics tests indicated that uncemented ash only increases the strength of “hard” coal if it is mechanically confined. The results for “soft” coal are awaited. These results will be coupled with future underground trials to quantify the benefits of ashfilling on coal extraction and surface stability.

Task 2.13.1 / 2: Verification of the new coal pillar strength formula and investigations into the prediction of pillar life in the Witbank and Highveld coalfields

The Salamon and Munro formula, empirically derived following the Coalbrook disaster, has been used to design coal pillars in South Africa for the past three decades. The database was updated and a new analysis method used to refine the constants in the formula. The new Van der Merwe formula predicted significantly greater strength for pillars with larger width / height ratios, and lower strength for smaller pillars. The formula also allowed a higher extraction ratio, depending on the depth of workings and the required safety factor. In the case of the Witbank and Highveld Coalfields, reserve utilisation could potentially be increased by 11 per cent. This task aims to verify the new Van der Merwe linear formula through a programme of underground observations and monitoring.

Task 2.14: Evaluation of design procedures for thin-rectangular pillars

New mining methods are being developed in South Africa involving long slender pillars, which have the potential to increase the productivity and coal utilisation in the Witbank and Highveld coalfield. The strength of rectangular pillars has been investigated in the past by means of laboratory testing and numerical modelling. Currently, the effective pillar width formula developed by Wagner is being used in the design of rectangular pillars. Although this formula has been successfully used since its introduction in South Africa, a detailed study is required to determine its applicability to long thin rectangular pillars.

Task 2.15: A preliminary investigation into the geotechnical interpretation of geophysical logs

This task involves an international review of geophysical techniques that can be used to identify the critical parameters controlling the stability of coalmine roofs in the Witbank and Highveld coalfields. Underground trials will then be performed to evaluate the most promising techniques. Features that currently affect roof stability include variable stiffnesses over short distances, high concentrations of mica, low-strength materials, low cohesion on critical bedding planes, open vertical joints, coal / roof modification by intrusions, and the presence of water. It is hoped that the geophysical techniques will be able to identify these features.

3.2.4.3Surface mining

Task 3.2.2: Economic extraction of pillar reserves using opencast methods

In the past a large number of shallow coal seams were mined by underground mining methods. These coal deposits can be further exploited using surface mining methods, provided the risks associated with the old underground workings can be determined and managed. The following issues were investigated. The risks and hazards were identified, quantified, modelled and incorporated into a feasibility model.

Geology – Where and how extensive are the reserves of coal contained in shallow pillars?

Mining methods – How can the reserves be mined?
Mine planning and design – What factors need special attention, and can appropriate planning improve the risk exposure and economic viability of an operation?
Highwall stability – How will the high wall stability change if heavy surface mining equipment exposes previous mine workings?
Stability of underground workings – What will happen to the stability of bords and pillars if the extra weight of heavy surface mining equipment is added?
Pillar identification – Can the size and position of pillars be determined?
Surface subsidence – How will it influence the mining operation?
Spontaneous combustion – To what extent will this occur and how can it be controlled?
Water and hydrology – What will happen to the quantity and quality of mine water stored or collected in the underground workings over many years?
Equipment selection – Do the practical challenges require new equipment selection criteria?
Drilling and blasting – To what extent can drilling and blasting reduce the risks associated with mining these reserves?
Processing – To what extent will factors such as oxidised and/or hot coal, higher dilution factors and foreign objects affect processing?

3.2.4.4Surface environment

Task 6.9.1: The development of techniques to predict and manage the impact of surface subsidence

The purpose of this task is to provide the coal mining industry with a modelling tool for use in the determination of the investment required in post-closure funds to cater for subsidence associated with bord and pillar mines. Site-specific input data on subsidence, time-related pillar failure data obtained from historical data, and the experience and judgement of mining engineers were analysed through the use of mainly engineering design tools and statistical methods. The following parameters were found to be fundamental and were incorporated in the subsidence prediction model: depth of workings, age of pillars, immediate roof thickness, geology (competent and incompetent strata), bord width, and pillar width. Three mathematical models were developed:

Potential for sinkhole development based on available space underground (bords and intersections), bulking factor, and overburden strata;
Seam-specific expected life of pillars; and
Maximum depth of scaling, expected maximum vertical subsidence and maximum tilt.

3.2.4.5Human and social aspects

Task 7.14.1: National and sector specific social and economic implications of selecting capital or labour intensive methods of coal production

This study evaluated the sector-specific and national outcomes of various strategic choices relating to the technique of production used in coal mining. Terms of reference for the study were derived through multi-stakeholder workshops at which the scenarios around techniques of production were defined. In essence, it was agreed to examine labour-intensive mining represented by a mechanised drill and blast method and capital-intensive mining represented by bord and pillar mining with continuous miners, as these were at the extremes of a reasonable continuum of labour and capital intensiveness. All stakeholders regarded extreme labour-intensive methods, such as hand-got coal winning, as beyond the realms of practically implementable techniques. It was also agreed that limited choices on capital and labour intensity exist for open cast mining, and the scope of this study therefore was restricted to underground coal mining operations.

3.2.5PlatMine (2003 - present)

The PlatMine collaborative research programme, involving Anglo Platinum, CSIR Miningtek, Impala Platinum, Lonmin and Northam Platinum was launched in April 2003 with an initial funding level of R8 million per annum. The primary goals of the programme are to:

Ensure the long-term sustainability of the platinum industry in South Africa;
Develop technologies and competencies to improve overall safety and health;
Increase productivity;
Facilitate mechanisation by solving common technological problems; and
Improve the underground working environment.

The outputs of the research program were not in the public domain at the time of writing this report. Consequently, only the objectives and scope of each rock-related project could be described.

3.2.5.1Stope support

Task 1.1: Alternative support systems for mechanised stopes

Stoping on the UG2 and Merensky Reefs is currently undertaken with small in-stope pillars as support. This is common practice in other shallow tabular orebodies and ensures that stope backbreak does not occur. If, however, the mining of stope faces were mechanised, the presence of pillars would greatly hinder the stoping operations and reduce mining efficiencies. This project will determine alternative layout and support options to allow mechanised stoping without in-stope pillars. In order to meet these objectives, the support resistance requirement to ensure stability in Merensky and UG2 stopes in current pillar stopes must be determined. Methods of achieving the required support resistance then need to be considered. This will entail the evaluation of mine layouts, stiff backfill, and conventional support. This will result in the conceptualisation of a number of mining layouts with suitable support systems that will allow long faces to be mined without pillars. There will be a constant reference to the practicality of the proposed layouts. Risk assessments on the proposed mining layouts will then be undertaken.

Task 1.2: Platinum reef pillar strength formulae and pillar design methodology

It is well known that pillar strength is a function of many parameters. However, the fundamental parameter, the actual strength of a hard rock pillar, is not known with any degree of certainty. In this study the strength of the pillars was calculated through back analysis of failed, crushed and stable pillars using the "maximum likelihood method". The advantage of this method is that because it is based on actual pillar collapses, it takes the effect of all the parameters into account when the strength is calculated. The other advantage of this system is that it allows the quantification of the probability of safety. However, it is very important to note that the confidence level of this type of formula is determined by the standard deviation of the data. A preliminary investigation conducted using this method indicated that for a probability of stability of 1.0, a safety factor of 3.5 is required. This is due to the high standard deviation of the data, indicating that other factors affect the pillar strength. It is therefore suggested that these factors should be identified, investigated and their variation determined, not only using numerical modelling and laboratory experiments, but mainly by conducting underground investigations. The main objective of this study will therefore be to determine the strength of pillars through analysis of failed, crushed and stable pillars and the determination of variations of different parameters underground and the effect of those parameters. This will result in an improved pillar design methodology. The project has the potential to increase extraction ratios as conservative pillar designs may be in use in some cases.

Task 1.3: Engineering hydraulic fills (and/or pastes) for local and regional underground support considering Bushveld tailings

Increased seismicity is becoming a key issue as the depths at which platinum mining occurs increase. The application of hydraulic fill concurrent with mining not only offers the potential to alleviate the seismicity problem and enable safer mining at depth, but also provides opportunities for reducing, or eliminating, the use of reef stabilising pillars underground. Platinum group metals recovery from underground may therefore be considerably increased, and significant economic benefits may be gained. By maximising the use of development waste crushed underground, the technology also offers a potential benefit to substantially increasing reef hoisting capacity by leaving an equivalent tonnage of development waste underground. After conducting an audit complemented by rigorous monitoring and detailed evaluation of the performance of current hydraulic fill as used at Northam Platinum Mine, the project will identify key issues and risks involved in the application of paste or hydraulic fill technology in future deeper level platinum mining. Further, relevant techno-economic information and guidelines for practical application of fill technology will be developed for typical platinum mining operations.

Task 1.4: Micro-seismic monitoring of pillar systems in the Bushveld Complex to evaluate pillar component failure

The prevention of pillar failure is an important concern of any pillar design strategy. There are a number of completed studies (e.g. SIMRAC projects GAP 334 and GAP 617) and others currently in progress (PlatMine projects 1.2 and 3.1) that address various aspects of pillar design practice. However, the failure mechanism of pillar systems is not well understood and further research is strongly motivated in the Bushveld Complex, where limited work has been carried out. The pillar failure mechanism is known to be related to the following factors: (i) the mining method and amount of mining; (ii) the ambient stress and effect of various k-ratios; (iii) the pillar width to length ratio; (iv) geological structures and type of host rock; (v) the fracture formation around the hangingwall and footwall contacts; (vi) foundation failure; and (vii) the magnitude of dynamic loading during the interaction with the seismic waves propagating through the excavations. These factors may vary significantly in different geotechnical areas. Most of the pillar failure models are associated with micro-seismicity occurring before and during the failure. Detailed study of this seismic activity should indicate the state of stability of pillar systems. The seismicity around different pillar systems using data recorded underground by close-in micro-seismic networks will be measured and analysed. The study of the spatial distribution and peak particle velocities of seismic events, supported by numerical modelling of the source mechanism of pillar failure and the generated wave-field, will assist the interpretion of the rockmass behaviour and failure mechanism. CSIR Miningtek has made good progress during the last decade in the field of static and dynamic behaviour of the rockmass surrounding excavations, using data from deep level gold mines (SIMRAC projects GAP 201, GAP 530 and GAP 615). However, the complexity of the problem and the wide difference in rockmass conditions between the reefs in the Witwatersrand gold fields and the Bushveld Complex do not allow a straightforward application of the findings. Conversion of the methods and additional underground observations are needed for the mines in the Bushveld Complex.

3.2.5.2Rockmass characterisation

Task 2.2: Enhanced application and implementation of borehole radar

This project aims to improve safety and productivity in the platinum mining environment by identifying orebody geometry and associated features ahead of mining. This will be achieved by implementing borehole radar for detecting changes in reef geometry, for example faulting and potholing, and associated features such as the location of domes and the Triplets or Leaders. The most important features specific to the UG2 and Merensky Reefs that can be mapped using radar will be identified, and routine availability of suitable boreholes will be determined. A feasibility study will determine the ability of borehole radar to map the desired targets. The study will include the measurement of electromagnetic properties of typical Bushveld rocks to determine propagation parameters. It will also include synthetic forward modelling of selected scenarios to determine feasibility. A number of surveys will be undertaken on typical targets and target geometries to ascertain the performance of borehole radar, to determine the accuracy of the synthetic models, and to investigate the feasibility of introducing borehole radar into routine operations. The project will also focus on the development of borehole radar for routine application in platinum mining, thereby allowing the optimisation of productivity, safety, ore reserves, and mining layouts.

Task 2.3: Implement and optimise Ground Penetrating Radar (GPR)

GPR has been identified as a promising geophysical technique for application in the platinum environment. However, current GPR systems are not particularly user-friendly and cannot be routinely applied. GPR is currently mainly used by rock engineering practitioners and mining engineers. The aim of this project is to evaluate the capabilities and applications of current GPR technologies and to match these with the most important GPR needs. The most suitable system(s) will then be tested in the platinum environment. Finally, recommendations regarding the implementation of, and improvements to, the best system(s) will be made.

Task 2.4: Using resistivity tomography to image mafic pegmatoids and potholes

Geological features such as iron-rich replacement pegmatoids and potholes often disrupt the continuity of the UG2 and Merensky reef horizons and adversely affect mining operations. Resistivity Tomography (RESTOM) could potentially be used to detect large-scale disruptive features in un-mined blocks of ground between, for example, raise lines. The outputs of such RESTOM surveys are cross-sectional colour-coded images mapping variations in electrical conductivity. Potholes and pegmatoids should manifest as conductivity anomalies on such images. The aim of this project was to evaluate the feasibility of applying RESTOM to the above-mentioned problems. The first phase of the project will involve defining the problem parameters. Benchmarking of the technique will be done through further trial surveys and modelling. In the process, the optimum system, survey, and processing parameters are to be determined.

3.2.5.3Mine design

Task 3.1: Strategies for the mining of the UG2 below partially extracted Merensky reef

Extensive UG2 reef reserves underlie the partially mined Merensky reef, with the middling between the two varying from as little as 18 m to as much as 400 m. Unmined ground in the form of pillars, potholes and remnants were left during the mining of the Merensky reef. The adverse influence of this unmined ground left on the Merensky reef horizon is a major concern during the mining of the UG2 reef. This is more pronounced when the two reefs are in close proximity to each other. A key question that needs to be addressed is whether mining under these circumstances will lead to rock bursting. This project proposes to study the influence of partially extracted Merensky reef on the mining of the UG2 reef, and aims to provide recommendations for mining strategies, including regional and local support, to achieve efficient and safe extraction of the UG2 below the Merensky reef for various middling thicknesses.

3.2.6Shotcrete working group

A working group was convened during the mid-1990s to investigate issues related to shotcrete. Only limited information regarding the scope and outputs of the study could be obtained (personal communications, Prof TR Stacey, University of the Witswatersrand, 2005 and Mr Johan Wesseloo, SRK, 2005). Apparently Mr Gerhard Keyter, who was an employee of SRK at the time, led the initiative. It was co-funded by the Department of Trade and Industry’s Technology and Human Resources for Industry Programme (THRIP). The working group was disbanded owing to difficulties in obtaining sustained support.

3.2.7International collaborative research projects

Several international collaborative projects are being conducted in South African mines.

3.2.7.1Semi-controlled earthquake generation experiments

South African and Japanese seismologists have been collaborating since 1992 to understand the entire lifespan of an earthquake, using modern technology developed in Japan and by ISSI International of South Africa. The first experimental site was established in 1996 in a mine near Carletonville (Ogasawara et al., 2001). The site was 200 m x 200 m in extent, and undisturbed by faults or dykes. More than 20 000 events (M<2) were recorded by triaxial borehole accelerometers in a 10-month period. An earthquake sequence associated with an M2 event in a remnant area about 150 m from the site was detected. Significant changes in seismic parameters such as stress drop, b-value, and energy index were observed. A second site was established in 2000 on a mine in the Free State gold field, close to a fault where events in the M=2-3 range were anticipated. A strainmeter, sensitive enough to monitor earth tides of the order of 10^-8, was installed in a 14 m long borehole drilled parallel to the strike of the fault.

In 2003 to 2004, two experimental arrays were deployed to monitor faults stabilised by bracket pillars in the Tautona and Mponeng mines near Carletonville (Ogasawara et al., 2005a). The arrays consisted of multiple strainmeters, arrays of strong ground motion meters, fault displacement meters, and sensitive thermometers to monitor seismic heat generation. For the second phase of the study, a sensitive, stable, and robust strainmeter was installed in a potential seismic source area at the Bambanani mine near Welkom (Ogasawara et al., 2005b). Strain changes up to 10^-4 were measured within 100 m of two M>2 events, corresponding to a stress change of about 7 MPa in the hypocentral area. For both events, relaxation in the maximum principal stress at a rate of 10^-6/week was observed for several days prior to the main shock. However, no acceleration in deformation was observed.

3.2.7.2Drilling active faults in South African mines

The deep gold mines of South Africa are natural laboratories that offer unique opportunities to study seismic event processes and fault properties. Drilling into active faults add significant new dimensions to the ongoing investigations elsewhere in the world. These include:

Systematic structural analysis of active faults;
Intra-fault zone observations at the focal depth of in situ stress, creep activity, triggered slip, and frictional heat;
Investigation of dynamic changes of normal stress during brittle failures, mode of rupture, and wave propagation effects in low-velocity damaged fault zone layers;
Comparison between drilling and core observations and detailed mapping in nearby tunnels to verify the accuracy of both techniques; and
Validation of inferences from seismic data recorded by in-mine geophone arrays.

The project ‘Drilling Active Faults in South African Mines’ (DAFSAM) is a five-year research project, coordinated by Dr Ze’ev Reches of the University of Oklahoma, USA. The International Continental Drilling Program (ICDP) covers only the drilling costs, while each participating country covers the costs of instrumentation and researcher manpower. The American, German, and Japanese researchers are funded by the National Science Foundation, the GeoforschungsZentrum Potsdam, and Ritsumeikan University, respectively. The South African researchers are funded by a variety of agencies.

The scientific objectives of the project concentrate on investigating the near-field behaviour of active faults during the entire seismic cycle. Instead of drilling a single deep hole from surface, as was done for a study of the San Andreas fault in California, the DAFSAM project relies on drilling many short holes into faults that are likely to be activated by mining. In addition, a consortium of Japanese research institutes and universities will perform detailed studies of active faults, monitoring in situ stresses, near-field seismic data, and strain. The dense array of near-field instruments, coupled with a mine-wide seismic system and additional microseismic systems installed in the stopes, will permit the researchers to track the initiation and propagation of seismic events through the rock mass. This will lead to the development of criteria for quantification of the damage effects of the seismic events on the strength and stiffness of the rock mass and, hence, the stability of excavations and structures. This knowledge will contribute to improved methods for designing excavation layouts that are inherently resistant to unstable rock failure. The drilling programme started early in 2005.