Search Documents

By Stephen Mansfield

Analysing how accurately a blast design is implemented is one of the cornerstones of a successful operation, and critical for mine optimisation. This paper presents a concept that involves the collection and analysis of data for the five stages through which a blast transitions, from budget through to implementation.

Feb 1, 2020

By T. L. Joseph

A DEMAND for specialized knowledge has directed engineering curricula towards competency in some particular field or occupation. Preparation for life in a broad sense of completeness has received little consideration. Ways and means are needed for weaving some of the threads of liberal education into the fabric of engineering education. Any effort to develop a humanistic-social stem of engineering education will depend largely upon the content of the courses and the type of instruction offered. Many of the courses in college curricula are intended to lay a foundation for more advanced courses to be taken later by those. who specialize in some particular field. Students in other fields lack the background necessary to take the advanced courses and have too much detailed information, soon forgotten, forced upon them if they pursue the foundation courses. To develop a humanistic-social stem for engineering education existing catalogue courses selected should follow some general plan aimed towards broad educational objectives. Courses packed with detailed factual information are necessary for the stu-

Jan 1, 1947

By Ken Musgrave

"A program is described that helps managers and supervisors control maintenance more effectively. Work- order generation, preventive maintenance, repair parts inventory, planning and scheduling, reporting, analyzing costs, printing purchase orders and other functions are covered. Flexibility often permits integration with other programs such as payroll, cost accounting and materials management.The Role of MaintenanceMaintenance provides a service to allow the rest of the operation to do its job effectively. Its success should not be measured by maintenance cost alone, but also by its effectiveness in reducing downtime and maximizing production. It can make sense to spend more maintenance dollars if this results in an even greater contribution from increased production.The Cost of MaintenanceIn mining companies, maintenance costs can be as much as half of operating costs. Maintenance has generally not been given the management attention that an expenditure of this magnitude deserves."

Jan 1, 1986

By Yuusaku Tominaga

A flexible expert system for a personal computer has been developed to take countermeasures for prevention of emergency in underground coal mines. The system makes a diagram to estimate maximum airflow rate of a specified branch such as coal face or advancing face, etc. under the conditions of opening regulators set in any two different branches, by computer simulation. The system approximates the ventilating condition underground by using both measured values and safety factor derived from the rules in the system. The probability of happening gas explosion, spontaneous combustion and mine fire underground is shown on a display and countermeasures in a data base in the system are selected according to a confidence factor. When it is recognized in the system that either safety factor or confidence factor of countermeasure has a small value, the system sends information relating to reasoning in the system to registered human experts by computer communi¬cation and facsimile. The replies in file transfer from the experts are stored in the system and reconsultations for estimation of underground condition are started automatically. As the replies in facsimile are shown on the display, user inputs new rules and, values for reconsultation. It is shown that the system is available to take better countermeasures as the condition' of a network in coal mine is changed by mining day by day.

Jan 1, 1992

By Takeshi Tanaka, Yasuhiro Komamura, Yuusaku Tominaga, Takayoshi Iwamoto

In order to assist better countermeasures for ventilation control in normal/abnormal condition of mine ventilation network in real time, branch relation table derived from a macroscopic characteristic of ventilation network is introduced in this study. The main results in this study are as follows; 1) One of macro characteristics of a ventilation network, as a relation between airflow rates in two branches chosen arbitrarily, is specified under the change of aerodynamic resistances in a branch. Airflow rate in a branch can be estimated by airflow rate in other branch changing aerodynamic resistance by using empirical formula. The empirical formula by computer simulation of the Horonai coal mine, Japan is considered to be a one-valued function with monotone or extremums. 2) A branch relation table shows the macro characteristics, mentioned above, of all the set of two-branch combinations in a network. Elements in the table give both information of control possibility range of airflow rate and total number of extremums in an empirical formula. 3) It is important that when aerodynamic resistance change in a branch is estimated by measuring airflow rates in two typical branches, the two branches should be chosen from branch combinations with monotonous relation in the branch relation table. 4) A flexible expert system to assist ventilation control in the specified branch in real time has been developed by using a branch relation table for a ventilation transfer diagram of the specified branch, such a coal face and an advancing face. A fuzzy reasoning is utilized in the recognition of the safety factor, methane concentration, in the system. And when conviction degree in recognition for methane concentration at the specified branch is too low, the system searches for other branch combinations in the branch relation table stored in the system, to use other ventilation transfer diagram of the specified branch. When a needed ventilation transfer diagram is not stored in the system, the expert system asks automatically the ventilation simulation from other computers linked to the main computer automatically. The system repeats consultation until conviction value of safety in the specified branch approached

Jan 1, 1993

By Takiuchi T, Kurita M, Owada S

This paper describes the possibility of the application of flotation to the separation of monazite from xenotime and of columbite from ferberite, which have been said to be difficult with generally used physical beneficiation methods. For Malaysian monazite and xenotime samples, a great difference is observed between the flotation recoveries at pH 7.5 using 1.0 x 10-4mol/1 dodecylammonium chloride (DAC). The grade and recovery of columbite in the binary flotation with both minerals result in 85 per cent and 67 per cent in the cleaning froth and 91 per cent and 51 per cent in the re-cleaning froth, without adding any scavenging stages. For monazite samples from various localities, the flotability is dominated by the electrostatic interaction between mineral surface and collector ion using DAC as collector, but the singular decrease of flotation is observed in the sample of which the crystallinity is low and the surface structure is rugged and complex. Otherwise, ferric ion in the flotation system probably dissolved from mineral is found to activate the flotation using sodium dodecylsulfate as collector. For columbite and ferberite, the separation is judged possible at pH 3.0 using 1.0 x 10'3mo1/l sodium oleate (NaOl) from the single flotation results of each mineral. In the binary flotation, however, columbite recovery in the froth is greatly diminished below that expected. This would be because NaOl is prevented from adsorbing on the columbite surface of which the charge is transformed more negative side by the specific adsorption of the tungstate ion dissolved from ferberite.

Jan 1, 1993

By Geoffrey Purcell

Improvements which have been made in commericial flotation systems during the past fifty years have resulted more from trial and error methods in the mill than from research in the laboratory. Research has often provided explanations of why mill changes have improved the metallurgical process. Hopefully, in the next fifty years, changes in mill flowsheets will come about, primarily, as a result of research work. However, since minerals, in their infinite variety, cannot be reduced to a standard state, research on commercial ores will have only a local utility, and research on pure minerals will have little commercial application. What will emerge from research on pure systems are mechanisms of adsorption, and this in turn may throw some light on the very much more complex adsorption processes taking place in the pulps of commercial ores. Quartz, without a doubt, has been the subject of more research effort than any other mineral, but we are still a long way from completely understanding its behavior in flo¬tation. Hematite and calcite have received considerably less attention. Ores containing a mixture of all three minerals, however, are commercially important, and the iron deposits in Alabama have been extensively studied by the United States Bureau of Mines (1). The literature reveals no reports of research on synthetic mixtures of all three.

Jan 1, 1967

By A. G. Roach

MILLING of Nova Scotia gold ores began in 1862 with arrastras and wooden stamps and has progressed to heavier stamps followed by tables and other forms of concentration for the recovery of the sulphides. Chlorination and cyanide leaching of the concentrates has been practised, and in 1936 the combined amalgamation and all-sliming cyanidation plant of Seal Harbor Gold Mines was built. It has since been found that it is not necessary to cyanide the full tonnage and now only the concentrates are being so treated. From his experience in this enterprise, the writer presents the accompanying flow-sheet and supporting comments. While he has not the temerity to maintain that it is the best or only flow-sheet for Nova Scotia gold ores, he feels that it warrants thoughtful consideration by members of the milling fraternity of the Province. THE ORE The gold in Nova Scotia ores occurs both free and intimately associated with sulphides-chiefly arsenopyrite. Much of the free gold is quite coarse and crystalline. This is especially true when it occurs in white quartz. When in the slate, it is usually in thin plates and scales, frequently coated and discoloured. The values bound up in the sulphides are not amenable to amalgamation and must be recovered by other methods-concentration, followed by smelting or fine grinding and cyanidation. While high freight and treatment charges apply on concentrates shipped from the Province to smelters, a considerable quantity has been so disposed of in the past. However, it is hoped that a customs cyanidation plant will soon become available to producers of these auriferous concentrates in Nova Scotia.

Jan 1, 1938

Operating CIP/CIL gold recovery plants in the arid regions of Western Australia face unique challenges due to water and ore quality issues. The presence of cations of Mg, Fe and Cu metals in water supplies and leached sulphide ores can have considerable effect on plant consumption of lime and cyanide, due to pH buffering and complexation effects. Reagent costs are critical to many of these operations. In this paper, computational modelling is used to analyse a number of process options to mitigate such issues. This requires careful consideration of the chemistry of the process system, and the ability to predict accurately the response of speciation and precipitation reactions to changing process conditions. The Mg-Fe-Cu-CN-H20 system is specifically modelled in this application. The Gibbs Free Energy Minimisation thermodynamic framework has been selected to model equilibrium phenomena. Furthermore, this approach has been extended to the process flowsheet level, incorporating multiple unit operations and recycle streams in a steady-state mass balance model. The Gibbs equilibrium composition module of HSC Chemistry software has been successfully integrated into the commercial flowsheeting package SysCAD. Consideration of the Gibbs Energy Minimisation capabilities of other flowsheeting packages has also been given. This approach has been successful in simulating the aqueous process chemistry typical of such plants. It has resulted in a predictive capability where equilibrium states respond appropriately to process changes. Potential cost savings may be identified by examining alternative process conditions to minimise overall reagent consumption.This methodology may be extended to other hydrometallurgical plants where issues of solution ideality, temperature and other equilibrium processes are sufficiently understood.

Jan 1, 2005

By M. Rosales

Using the k-e turbulence model, a mathematical fluid dynamics simulation of the flow characteristics in a Teniente converter has been conducted. Two cases were studied: - Transversal sections in the smelting-converting zone; and - A three-dimensional simulation that includes the walls of the reactor The calculations give the velocities field, the gravity waves amplitude, the splashing behaviour, the phase distribution and the shear stress distribution. The latter helps to explain the pattern of refractory lining wear. The calculation includes the approximate shape and dimensions, and the eventual coalescence of bubbles.

Jan 1, 1999

By Z. Lin, M. Isac, R. I. L. Guthrie

The kinetic and thermodynamic efficiencies of chromium recovery from the slag in an AOD process during the slag de-oxidation period with ferrosilicon is a complex process affected by slag basicity, temperature, the dissolution kinetics of the reducing agents, and intermixing of the slag/metal/alloy addition phases within the vessel. The present work focused on a study and simulation of the fluid dynamics and mixing behaviour of ferrosilicon additions in the AOD process for the production of stainless steel. In order to visualize the mixing behaviour, a two dimensional slice model of an AOD vessel was first built and filled with ZnCl2solution and then covered with silicone oil in order to represent the bulk liquid steel and slag phases, respectively. The injection of argon/oxygen process gases was simulated using air or helium. Particle trajectories and intermixing of the phases was monitored with a high-speed camera. To simulate the 75FeSi additions, glass spheres were used. A three-dimensional version of the AOD model was then constructed in order to confirm the adequacy of the slice model. Similar particle mixing behaviour was observed. Both models demonstrated that the ascending gas/liquid plume created two asymmetric mixing zones: a smaller rapidly re-circulating mixing zone adjacent to the side-wall locating the tuyere(s), and a larger mixing zone in the main part of the vessel. The process efficiency during the reduction period is governed by the thermal behaviour of ferroalloy as well as the mixing efficiency in the main large mixing zone. To enhance the recovery efficiency for chromium, small Fe-Si particles are suggested, given their wide distribution within the slag phase, and their large interfacial areas and rapid melting times for Cr2O3reduction from the slag. Keywords; modelling, ferrosilicon, chromium oxidation, AOD steelmaking.

Jan 1, 2004

By S Corbel, P J. T Donchak, W Gorczyk

The Hodgkinson Province numerical modelling project aims to develop an improved understanding of the physical conditions that lead to the localisation of goldbearing fluids within known systems and apply this understanding to the prediction of potential new sites of gold mineralisation. The project follows a series of stages from identifying the key parameters required to construct and test simulations through to reverse engineering of known deposits and ultimately applying the conditions identified in the preliminary stages to predictive simulations. Results of the modelling demonstrate strong geometric controls of the localisation of fluids during mineralisation. An EXTENDED ABSTRACT is available for download. A full-length paper was not prepared for this presentation.

Jan 1, 2008

By K. Muehlenbachs, J. U. Stemler

"Abstract - The Boston deposit is a recently discovered Archean lode-gold deposit located approximately 700 km northeast of Yellowknife, in the southern portion of the Hope Bay volcanic belt. Primary H2O-CO2-NaCl fluid inclusions associated with gold mineralization at Boston are of low salinity (4.6 ± 1.4 eq. wt.% NaCl, n = 182) and contain pure CO2. Homogenization temperatures of liquid-phase inclusions average 270° ± 33°C (n = 141), but some fluid inclusion assemblages contain coexisting liquid- and vapor-phase inclusions that homogenize near critical conditions, suggesting phase separation. In these groups, liquid-phase inclusions homogenized at 299° ± 25°C (n = 24), and vapor-phase inclusions homogenized at 338° ± 28°C (n = 18). These temperatures overlap within error, and a median temperature of ~320°C is taken as representative of trapping conditions within the veins. An approximate pressure of 2-3 kbar was estimated from fluid inclusion isochores at this temperature. Stable isotopic compositions of carbonate minerals from the mineralized zones range from -4.7‰ to -3.1‰ in d13C, and from 11.2‰ to 12.8‰ in d18O; quartz ranges from d18O = 12.8‰ to 14.9‰; and pyrite ranges from d34S = 2.1‰ to 3.5‰. The d18OH2O, d13CCO2, and d34SH2S values of the fluid calculated from these minerals at 320°C are 6‰-9‰, -2.4‰ ± 0.6‰, and 1.5‰ ± 0.5‰, respectively, and are intermediate between metamorphic and magmatic compositions. A unique source for the fluids cannot, therefore, be constrained from these data. Petrographic and geochemical data for the Boston deposit are consistent with data from other Archean lode-gold deposits worldwide, implying that the Boston deposit formed in a similar environment and from similar types of fluids. These results suggest that the Hope Bay volcanic belt may represent a major new mesothermal gold belt in this part of Nunavut. Sommaire - Le gîte Boston a été récemment découvert à environ 700 km au nord-est de Yellowknife dans la partie sud de la ceinture volcanique de Hope Bay, il s’agit d’un gîte d’or filonien d’âge Archéen. Les inclusions fluides primaires de H2O-CO2-NaCl associées à la minéralisation aurifère du gîte Boston présentent une faible salinité (4.6 ± 1.4 eq. pds.% NaCl, n = 182) et contiennent du CO2 pur. La température d’homogénéisation des inclusions à phase liquide est en moyenne de 270° ± 33°C (n = 141), mais certains assemblages d’inclusions contiennent des phases liquides et gazeuses coexistantes qui s’homogénéisent dans des conditions quasi critiques, suggérant qu’elles résultent d’une séparation de phase. Dans ces groupes, les inclusions de phase liquide s’homogénéisent à 299° ± 25°C (n = 24) et les inclusions de phase gazeuse s’homogénéisent à 338° ± 28°C (n = 18). Ces températures se recoupent à l’intérieur des limites de l'erreur de mesure, et nous considérons qu’une température médiane de ~320°C est représentative des conditions de piégeage dans ces veines. Les isochores permettent d’estimer une pression d’environ 2-3 kbar à cette température. La composition des isotopes stables des carbonates dans les zones minéralisées s’étend de -4.7‰ to -3.1‰ pour d13C et pour de 11.2‰ à 12.8‰ pour d18O; les valeurs de d18O pour le quartz varient de 12.8‰ à 14.9; et les valeurs de d34S dans la pyrite vont de 2.1‰ à 3.5‰. Les valeurs de d18OH2O, d13CCO2, et d34SH2S du fluide calculé à partir de ces minéraux à une température de 320°C est de 6-9‰, -2.4‰ ± 0.6‰, et 1.5‰ ± 0.5‰ respectivement, reflétant des conditions intermédiaires entre des compositions métamorphiques et magmatiques. Il n’est donc pas possible d’identifier une source unique pour ces fluides à partir de ces données. Les données pétrographiques et géochimiques du gîte Boston sont similaires à celles d’autres gîtes d’or filoniens ailleurs dans le monde, ce qui implique que le gîte Boston s’est formé dans un environnement similaire à partir des mêmes types de fluides. Ces résultats suggèrent que la ceinture volcanique de Hope"

Jan 1, 2006

By Pandey H. D.

The paper describes a fluidised bed ion exchange method for simultaneous removal of phenols and cyanide from the effluent. Use of an eluant enables regeneration of the resin in the continuous operation mode. The breakthrough experiments conducted both on the laboratory scale and on a pilot plant scale have shown that while the level of phenol can be brought down from 462 ppm to about 6 ppm, the free cyanide is almost totally removed. An examination of data indicates that the method can serve the twin objectives of pollution control and phenolic compounds recovery.

Jan 1, 1992

By James P. Kimball

AMONG the curious results of the recent advance of prices in the iron trade of the United States, one of them at least is to be regarded as of great importance. I allude to the utilization of mill cinder as a cheap and almost costless substitute for high-priced iron ores in many blast furnaces, situated within easy reach of such raw material. It has been, with but few exceptions, the practice in this country, based upon tradition rather than upon authority or European precedents, to waste this material on 'account of its high proportion of silicic and phosphoric acids, although known to contain more iron than most ores, and although also known to be sometimes applied to the blast furnace in greater or less quantities without serious detriment to the iron for specific purposes. While in use to a limited extent in the Hocking Valley and Hanging Rock iron regions of Ohio for several years, its use in Western Pennsylvania has only become general of late, under the exigencies of high prices and a short supply of ore. Especially has mill cinder largely entered into the stock of furnaces attached to rolling mills in and about Pittsburgh. Under such circumstances the cinder pig has been paddled on the premises, and the malleable iron produced has been unexpectedly good. Cautiously as at first mill cinder has been applied, it is evident that the temptation to use it freely is a strong one whenever the commercial rates for high-class ores are out of proportion to the general conditions of the iron trade, as sometimes happens in this country, especially as the free use of it is apparently justified by the quality of the iron produced from this cheap and abundant material. Indeed the consumption of cinder pig has., not been wholly confined to mills using up the production of furnaces belonging to the same plant. One furnace at least on the Ohio, at Covington, for several' months has been making foundry pig from "all cinder." The iron has gone into the general trade. It is probable that malleable iron of good quality can he made from the poorest cinder pig, but with what degree of economy will

Jan 1, 1881

By W. J. Taylor

In making heating gas with anthracite coal for roasting ore during the past few years, I have tried many forms of gas generators. So far, the most successful and satisfactory one has been what we call a fluxing producer, which is simply a small cupola or blast furnace, in which we charge with the coal, say, 30 or 40 per cent. of basic blast-furnace cinder, to unite, when melted, with the ash or earthy matter in the coal, and carry it off as cinder. The producer we have in use at Chester, N. J., has more of the lines of a blast furnace than a cupola, the hearth being 24 inches diameter and 24 inches high. It then enlarges on an angle of about 25 degrees from a perpendicular to 4 feet, and then is drawn into 3 feet on top ; total height, 12 feet. We use one water coil-tuyere 12 inches above the bottom, and blow through a 1 1/2 - inch nozzle. The depth of coal above the tuyere averages about 6 feet. The producer-is blown with

Jan 1, 1881

By Glenn Kosick, Roger Amelunxen, Glenn Dobby

"The variability in flotation response within an orebody is often a significant problem to address during the design phase of a flotation circuit. In addition, for existing operating plants, the reasons for such a variability need to be clearly understood and accounted for if plant performance is to be optimized. MinnovEX Technologies, in collaboration with major mining companies, is developing methodologies and tools to address the issue of variability in ore characteristics (such as ore hardness and flotation response). The tool is used for design and scale-up of flotation circuits from lab flotation data, and for prediction of recovery and final concentrate grades throughout the mine block model for any given circuit design.INTRODUCTIOTNH: EI SSUEB EINGA DDRESSEDThis paper describes the current development of a methodology and software tool for improved flotation plant design and enhanced prediction of froth flotation performance in an operating plant. The need for this tool is attributed to two primary factors: variability of flotation response properties throughout an orebody and related difficulties with scaleup from laboratory tests. Consider these in more detail.It is well recognized that separation performance via froth flotation will vary significantly within a single mineral deposit. This has been enhanced through the wide application of SAG milling. Throughput with SAG mills is more susceptible to ore hardness changes than is the case with crushing-rod-ball circuits; it is not uncommon to see SAG mill throughput to vary by up to 100% on a regular basis, e.g. 1000 tph to 2000 tph.The implication of this on laboratory testing for prediction of metallurgical performance is as follows: it is unlikely that either the plant grind (size distribution) feeding froth flotation or the flotation residence time will be known at the time of laboratory testing for a given block of ore. Hence, the grind used for the laboratory test is not likely to be that which actually occurs once the ore is processed, and the laboratory retention time may not represent the plant conditions."

Jan 1, 2002

By M. F. G. Bruce

Two backfill systems have been developed on West Driefontein, namely the centrifuge dewatered tailings system and the cyclone-classified tailings system. Modifications to the two systems since the publication of the original paper (Close and Klokow, 1985) are discussed in addition to the results obtained. Considerable experience has been gained from the operation of the centrifuge dewatering units as well as the placement of dewatered tailings in a difficult mine environment. The centrifuge dewatered tailings method has proved successful in the extraction of a highly stressed block of ground at No. 5A Sub-vertical Shaft. The utilization of backfill support in this area made it possible to dispose of a high-grade barrier pillar. This pillar was considered necessary prior to the introduction of backfill support. The recovery of the gold from this pillar made the use of dewatered tailings backfill cost- effective, despite high labour and maintenance costs. Current experience with the bagfilling method of placing cyclone- classified tailings indicates that it is viable and requires considerably less capital investment. Although expensive at present, it is envisaged that costs will decrease once cheaper bags and other forms of stope support are used. The placement of backfill in small-diameter sausage-type bags requires no change in mining cycle and allows the backfill to be placed close to the stope face. Both backfill systems are being expanded but preference is to be given to the cyclone-classified tailings system.

Jan 1, 1988

By M. J. Deats

INTRODUCTION To prepare the reader adequately, reference should be made to the paper 'The Pioneering of Fully Mechanized Longwall Coal Mining in South Africa' by R. T. Naude and M. J. Deats, published in the February, 1967, edition of the Journal. At the time mechanized longwalling had been proved operationally feasible with encouraging results since its inception in South Africa in mid 1965. Subsequently, further top seam panels were successfully extracted with improved productivity and with some modification to equipment. In 1968, after an overseas visit by mine officials, equipment specific to suit local conditions was acquired and lower seam trials commenced. Despite initial obstacles and difficulties the first panel was successfully mined. Modification and improvements to equipment and techniques resulted in the second lower seam panel being extracted economically during 1969-1970 and with productivity exceeding even that of top seam faces. During extraction of the second lower seam panel, gradual and complete surface subsidence occurred over the panel without affecting face operations. Rock mechanics investigations had originally indicated some uncertainty regarding adverse dolerite sill behaviour which could have caused longwall mining to be hazardous. All such doubts have now been removed. RESUMÉ AS AT FEBRUARY, 1967 Mechanized longwalling was introduced in 1965 on an experimental basis in an effort to improve productivity in the newer area of the mine and to increase the quality of the comparatively dirty run of mine feed to the washing plant. In addition advantages could be predicted in the long term by better utilizing reserves of straight coking coal and increasing the life of the mine. The top seam unit was obtained on a rental basis and by the beginning of 1967, two panels, the first having a face length of 215 m, a panel length of 345 m, seam section of 1 219 mm, and the second the same face length, a panel length of 453 m and a seam section of 1 067 mm had been successfully longwalled. At this stage a best month of 24 602 metric tons with an average of 17 364 metric tons per full working month could be reported. Results were sufficiently encouraging for the company to exercise its right to acquire the equipment as its own asset and from 1967 onward, the unit has operated on this basis. From the point of view of rock mechanics, the dolerite sill over the first two panels was known to have 'bridged', only the material below the base of the sill having truly 'goafed'. No adverse pressure effects had significantly affected face operations except for two 'bumps' towards the end of operations in each panel. Gate road maintenance presented no problem. The strong inflows of water experienced on the first face were now known to be associated with an isolated water bearing fault plane and were not experienced on the second face. TOP SEAM EXPERIENCE - MAY, 1966, TO DATE Panel 2 (February, 1966, to November, 1966) The operations in panel 2 (see Fig. 1 for location of panels) continued uneventfully until the final month in this panel. At this stage, the chocks behaved most inconsistently often lowering under the weight of the canopies alone. Roof trouble became so severe that the unit was unable to produce adequately and blasting of chocks became a daily occurrence. Finally 24 m short of the planned limit it was decided that a complete overhaul was the only solution and the extraction operations commenced. At this time, the only fatal accident associated with longwalling occurred when a Non-White inadvertently fell into the panzer conveyor transfer point at the main gate. Some gate road difficulties in this panel were overcome by bolting tapes to the roof as decking and allowing them to pass over the gate chocks and collapse into the goaf. Two cribbed roof falls were also negotiated by the gate chocks at the face ends without undue difficulties. It was decided in the light of maintenance problems encountered on this face that the best solution would be to appoint a foreman solely in charge of face maintenance crews. This step has to a large extent contributed to success achieved subsequently. Panel 2-face length 213 m panel length 456 m cutting height 1 092 mm average monthly output 16 443 t best month 19 130 t Panel 3 (December, 1966, to June, 1967) By the end of 1966 results were sufficiently encouraging for the company to exercise its right to acquire the equipment. This was negotiated and the unit has operated as such to the present date. Prior to installation in panel 3, all face equipment was thoroughly overhauled underground and the chocks boost-tested to yield-load. The costs of overhauling, including almost complete hose replacement, were much higher than anticipated. Advice had been received that complete hose replacement was common practice after +/- 18 months' operation. Shortly after commencement of operations in panel 3 the chocks again failed as they had towards the end of panel 2. At this stage, the suppliers advised that the valve gear should have been repaired during the overhaul as the needle and seat seals could no longer be relied upon. This operation was impossible to carry out on the face and would have been e1ttremely time consuming. Fortunately a new type of valve gear was then available from the suppliers, termed the capsule-type valve, and a complete change over could be made on the face. The encapsulated valve has a delrin pad sealing over a raised

By E. J. Maunders

This paper describes briefly a computerised model of the South African electricity system which has been developed jointly by the Atomic Energy Board and the Electricity Supply Commission. The results of an analysis of this system have been subjected to a fairly extensive sensitivity analysis which reveals that up to the year 1990 between 680 and 700 million tons of coal will be required for electric power generation, together with between 7 000 and 10 000 tons of uranium. This result is fairly insensitive to data variations, but the prediction of a cumulative requirement up to the year 2000, of between 1 300 and 1 800 million tons of coal and 25000 to 70000 tons of uranium, proves to be very data-sensitive.