Search Documents

By Robert H. Merrill

INTRODUCTION In most cases, the design of an underground mine is based upon the premise that the ground either will cave or will be stable. This chapter concerns the design of a mine in ground that will cave readily or with some as¬sistance, such as by long-hole drilling and blasting. Some of the more widely used caving systems of mining are panel caving, block caving, sublevel caving, and large pillar recovery. Some of the less widely used systems are glory-hole, top slicing, and induction caving. Al¬though the common practice of pillar robbing is not usually considered to be a caving system, this subject will be treated as a part of this chapter. BASICS OF CAVING Caving systems are most successful in ground that will cave in sizes that will flow through openings and grizzlies, and will easily load in cars or on belts for haul¬age. The ground most likely to cave well is highly frac¬tured and contains breaks, flaws, or other discontinui¬ties that form planes of weakness. Also, caving action can be greatly enhanced if the host rock itself is low in compressive, shear, and tensile strength. Ideally, a cav¬ing system of mining is best employed when the criteria for caving is a feature of the ore body and the develop¬ment drifts, haulageways, and drawpoints can be mined in a highly competent rock beneath the mineralized zone. However, the development is often in the same, or similar, fractured rock and the openings require sub¬stantial artificial support to assure stability. Several clues can be assembled to identify potential caving ground; however, for borderline cases, no sure method has been devised to date. The diamond-drill cores taken for exploration can provide an excellent clue provided drilling is performed carefully by experienced drillers. For example, if the ground is cored in such a manner that the breaks in the core are caused more by failure of the rock than by whipping core barrels, plugged drill bits, or other drilling causes, and the intact core lengths are consistently long [say, 0.6 to 3 m (2 to 10 ft) of unbroken core], there is little reason to believe the ground will cave without considerable as¬sistance. This is especially true for rocks with compres¬sive strengths above 34.5 MPa (5000 psi) and tensile strengths above 2.1 MPa (300 psi). On the other hand, if core recovery is low (below 80%) and the recovered ore is broken in small pieces and the breaks are along obvious weaknesses in the rock, the chances are excel¬lent that the ground will cave. This is true even when the rock between the defects has high compressive and tensile strength. Another clue has already been mentioned, that is, the measurement of the physical properties of the rock and the natural planes of weakness or defects in the rock. The planes of weakness in the rock can often be detected from outcrops, cores, or other exposures of the rock under consideration. Some rock types are known to be strong and will sustain large, unsupported open¬ings and would be difficult to cave intentionally. Yet the same rock type can also contain unbonded or weak planes of weakness or fractures, and in these locations the rock would undoubtedly cave with little assistance. Therefore, although the inherent strength of the rock is a factor in caving, the natural defects in the rock are more often the deciding factor. DESIGN CONCEPTS For the most part, the design of openings for caving ground is a problem of the interaction of openings over a relatively large area of the mine. To illustrate, Fig. 1 is a simplified section of a series of openings along the grizzly level or draw level of a block caving or panel caving development, and above this opening is a simpli¬fied section of a room-and-pillar arrangement on the undercut level. At this stage of the development, the stresses around the openings on the grizzly level are only moderately influenced by the openings on the undercut level and vice versa. Therefore, the stresses around the openings are approximated by the stresses around single or multiple openings in rock, the values of which are de¬scribed in the literature (Obert, Duvall, and Merrill, 1960; Obert and Duvall, 1967). Once the pillars on the undercut level are blasted (Fig. 2), the situation changes abruptly. The undercut opening (prior to caving) now can be approximated as an ovaloidal opening above the grizzly drifts and this opening tends to shield the vertical stress field. As the caved stage is drawn the stope approximates a much larger rectangular or square opening filled with rock, and if the rock is not sustaining a major portion of the stress field, this opening can be considered (for en¬gineering purposes) to be empty and the stresses that interact between the larger and the smaller openings take on a totally new perspective (see Fig. 3). Next, let the material cave to the surface, and let the caving ma¬terial sustain some stress, but much less than if the ma¬terial were intact. This condition is similar to a soft inclusion in a rigid body and has been treated in the literature (for example, Donnell, 1941). At this point in time, the grizzly drifts are subjected to the stress con-

Jan 1, 1982

By R. R. Klimpel

Introduction The grinding of coal-water slurries has received increasing industrial attention during the last decade. In particular, there is long-term interest in the use of pulverized coal-water slurries to replace oil in combustion equipment and in the development of coal gasification/liquefaction processes that require coal-water slurries as feed. More specifically, the use of coal-water slurries in gasification requires grinding a high-density slurry containing the smallest amount of water consistent with slurry pumping and spraying. As part of a fundamental engineering research support program aimed at the industrial implementation of dense coal-water slurry grinding, this author has published several papers on how specific rates of breakage vary as a function of slurry rheology (Klimpel, 1982,1982/83). These papers demonstrated that there is a consistent pattern of change in specific rates of breakage of coal in dense slurries with controlled variation in slurry rheology. By matching rheological data with laboratory grinding results, it was possible to identify directly slurry conditions that correspond to: 1) slowing down of breakage rates, 2) the occasional acceleration of breakage of some sizes, and 3) conditions where chemical additives will increase rates of breakage. In brief, these conditions were analyzed using two different criteria: a) the net production rate of material less than some specified size (e.g. kg/min of minus 325 mesh) in a standard batch laboratory mill test as a function of controlled changes in grinding conditions, and b) the use of the one-size-fraction feed method, which consists of following the disappearance of this largest size over grinding time in a batch laboratory mill to arrive at well-known specific rates of breakage (Austin et al., 1984). Detailed references to the methodology used as well as the conclusions are available (Klimpel, 1982, 1982/83) and will not be repeated here. The purpose of this paper is to further demonstrate several additional characteristics of dense coal-water slurry grinding that were shown in a simplified sense in the earlier publications of the author but which have clearly demonstrated themselves as being very important in the industrial simulation and scale-up of such coal-water grinding systems. In particular, this includes the clear and unambiguous demonstration of how the simultaneous acceleration of breakage of some size fractions and slowing down of the breakage of other size fractions is occurring as a function of changes in coal-water slurry density. In the earlier publications (e.g. Klimpel, 1982), it was shown by specially designed experiments that the addition of fine material and/or the use of a chemical thickening agent accelerated the specific rates of breakage of coals of coarser size fractions using the one-size-fraction method. There were also numerous examples given of non first-order breakage (the slowing down of coal breakage rates) using also the one-size-fraction method due to the presence of excessive amounts of fines which corresponded to the development of a rheological yield value. The problem with the simulation and scale-up of any laboratory and/or pilot-scale mill data to an industrial scale using the mechanistic modelling approach involving specific rates of breakage and breakage distribution parameters (e.g. Austin et al., 1984) is the number of assumptions involved in translating the smaller mill breakage parameters to the predicted larger mill breakage parameters. It is apparent, at least to this author, that to accurately simulate and predict larger scale equipment performance from smaller scale data (given that the larger scale data performance is known and hence predictions can be thoroughly checked) requires a better knowledge of breakage parameters than is currently available. More specifically, it was felt that one of the chief problems was the inability of the one-size-fraction method of determining breakage parameters to sufficiently represent the actual magnitude and sometimes even the directions of. the complicated interactions involved with slurry density changes in coal-water slurry grinding. Thus, a special set of experiments was conducted in a somewhat larger batch ball mill (0.457 m diam x 0.610 m length) than the 0.203-m-diam mill used in the original rheology characterization paper (Klimpel, 1982) so as to minimize any unusual effects due to wall-ball interactions (2.54-cm-diam balls used in both mills). More importantly, the measurements of specific rates of breakage were done using a proprietary tracer method on a portion of a given size fraction, which was then remixed into a natural feed size distribution before grinding. The experimental procedure and analysis of subsequent data was done in exactly the same manner as the radioactive tracer technique on coals as originally developed by Gardner (1962). The advantage of such an approach is that it makes no assumptions such as the independence of the specific rate of breakage of any size on the absolute sizes and amounts of other sizes present (both larger and smaller) in the mix of natural feed material. It will be shown that the measured rates of breakage using the direct tracer technique and the one-size fraction method on the same coal are indeed different. In fact, an accurate assessment of what is happening to the rates of breakage as a function of changing slurry density can only he made by measuring particle breakage under grinding conditions approximating the size distributions actually being produced in practice. Experiment procedures and results The pilot mill used was 44 cm diam x 60 cm long with a volume of 91,250 cm3 and was fitted with six 0.5-in. lifter bars.

Jan 1, 1992

By G. W. Riley, D. E. Pickett

HINDERED SETTLING CONCENTRATION In the free settling of mineral particles in a liquid, the falling particles are at a distance from each other so that no particle is affected by its neighbor. In hindered settling, the concentration of particles is sufficiently high so that each particle is affected by its proximity to other particles in the suspension. Richards and Locke86 have described the hindered settling phenomenon as the condition “.. where particles of mixed sizes, shapes and densities in a crowded mass, yet free to move along themselves, are sorted in a rising current of water, the velocity of which is much less than the free-falling velocity of the particles but yet fast enough so the particles are in motion." This is the condition normally encountered in mineral con¬centration processes. The well known Newton equation for free settling of coarse (ap proximately +10-mesh, 1/16-in. or -- 2000-µm) spherical particles is:87 where v," is free settling velocity, cm/sec; p' is density of the fluid; p is density of the particle, g/cm3; d is particle diameter, cm; g is acceleration due to gravity, cm/sec2; and Q is coefficient of resistance, dimensionless, ~0.4. For the settling of fine spheres in water (approximately 150 mesh or 100 µm) the equation of Stokes pertains:87 where µ is viscosity of the fluid in poises and the other symbols have the same meaning as those in Eq. 1. For particles whose size lies between about 10 mesh (--2000 µm) and 150 mesh (100 µm), their settling velocity can be determined from experimental data. These data are available in convenient form in the text by Taggart88 based on the original work of Richards. Alternatively, a Reynolds number-coefficient of resistance plot may be used to determine the settling rate of such particles.87 The settling rate of spherical particles under hindered settling conditions can also be calculated from Eqs. I and 2 by replacing p', the density of the fluid, by p" the apparent density of the suspen¬sion. The concentration of particles in the fluid thus imparts an appar¬ent density to the composite fluid or suspension greater than that of the liquid alone, resulting in a buoyant effect on the larger particles. Particle shape affects the settling rate of both coarse and fine particles. The general effect is to reduce their settling velocities and the effect is greater for coarse particles and for those settling under hindered settling conditions than for fine particles or free settling ones. For two particles of differing densities but settling at the same velocity under Newtonian conditions, the ratio of their diameters from Eq. 1, called the free settling ratio is: where L signifies the lighter particle and H, the heavier particle. Under Stokesian conditions the exponent would be 0.5. For hindered settling conditions the fluid density p' is replaced by the apparent density of the suspension, p", to obtain a generalized equation for the hindered settling ratio: assuming both particles settle in approximately the same regime. The free settling ratio as given by Eq. 3 has been called by Taggart88 the "concentration criterion" and is used to predict the effectiveness of any gravity concentration process (see Introduction to this section). Based on Eqs. 3 and 4, if two particles of densities pH and p,, settle at the same velocity, the diameter of the lighter particle will be larger than that of the heavier particle. For example, in the case of galena (pH = 7.5) and quartz (pL = 2.65) settling in water (p =1.0) under free settling, Newtonian conditions 3.9. Thus, a quartz particle nearly four times as large as a galena particle will settle at the same velocity. Any quartz particle just slightly less than four times the diameter of the largest galena particle may be separated from it. Under hindered settling, Newtonian condi¬tions in a suspension where p" = 1.65, dL/dH = 5.85 or any quartz particle just slightly less than about six times the largest galena particle may be separated from it. Reference to Eq. 4 indicates that a superior separation between two minerals of differing densities is favored by: (1) coarse particles settling under Newtonian conditions, (2) a large difference in (pa - PL), and (3) separation under hindered settling conditions where p" is high. Of course, there are practical limits to increasing p" excessively because at very high percent solids suspen¬sion fluidity would be lost and the hindered settling separation process defeated. Examples of hindered settling separators are the Dorrco-Fahren¬wald sizer,89 the Rheolaveur box 90 the Spitzkasten,89 and the Willoughby washer.91 These devices make a mineral separation on the basis of both specific gravity and size and all of them are essentially obsolete except for the Dorrco-Fahrenwald sizer and similar devices which still find application for the removal of coarse particles from a much finer particle assemblage and for preconcentration ahead of shaking tables. However, nearly all gravity concentration processes (jigs, tables, flowing film concentrators, heavy media separators) and many sizing devices (sizing classifiers, clarifiers, thickeners, hydrosepa¬rators) make use of the hindered settling phenomenon during the separation of particles. JIGGING Introduction In jigging, a mixture of ore particles, supported on a perforated plate or screen in a layer or "bed" with a depth many times the thickness of the largest particle, is subjected to an alternating rising and falling (pulsating) flow of fluid with the objective of causing all

Jan 1, 1985

By A. L. Wilder, S. N. Dixon

Introduction Coeur d'Alene Mines Corp.'s largest precious metals property is located in the historic Rochester Mining District 40 km (25 miles) northeast of Lovelock, NV. The property encountered cold weather operational problems soon after its fall start-up in 1986 due to its elevation of over 1830 m (6000 ft). The problem of ice buildup on the heaps because of sprayed solution application was faced immediately. It was felt that allowing ice to build up all winter long until a spring thaw was impractical due to the large area under leach. Further, the operating cost and delivery schedule for a solution heating system was unacceptable. The development and installation of a leach solution distribution system using drip emitters made efficient, cost-effective winter operation possible. Other benefits of this system have also been observed and are discussed here. General process description 15,422 kt/day (17,000 stpd) of - 1.27-cm (-1 /2-in) crushed ore from the three-stage crushing plant are delivered to the leach pad using 77.1 t (85 st) rear dump haul trucks. The ore is drifted into place with a D-9 bulldozer. Leach panels are contiguous and are approximately 8861 m'(90,000 square ft) in area built in 6-m (20-ft) lifts. New panels are built on top of older areas to a final height of 61 m (200 ft). Each panel is ripped and cross-ripped prior to leaching. Barren solution is distributed to the heap using drip emitters at rates of 0.02 to 0.41 L/min/m2 (0.0005 to 0.01 gpm per sq ft), depending on the age of the panels. The pH of the leach solution is 10.7 with a cyanide concentration of 0.75 kg/t (1.5 lb per st). Approximately 50% of the silver and 80% of the gold are finally recovered. Pregnant solution percolates though the heap and flows by gravity into one of two 9.46 ML (2.5 million gal) pregnant solution ponds. The solution is then pumped to a conventional Merrill-Crowe process plant. Clarification takes place in three 9464 L/min (2,500 gpm) capacity filters. The solution is then pumped to a packed vacuum deareation tower for the removal of dissolved oxygen. Typical deareated solution contains 0.7 parts per million dissolved oxygen. Precipitation of gold and silver is accomplished by adding a zinc dust slurry to the deareated solution at the suction of the filter press feed pump. Precipitated gold and silver are recovered in three recessed plate and frame filter presses. Barren solution is discharged into a 11.7 ML (3.1 million gal) pond where cyanide makeup occurs. This solution is pumped back to the heap for further leaching. The precipitate filter cake, containing approximately 75% dore (Ag + Au), is then fluxed with anhydrous borax, soda ash, sodium nitrate and fluorspar to yield a neutral, bisilicate slag. The fluxed precipitate is then charged into a propane-fired melting furnace and heated to 1150° C (2100° F) for 3 1/2 hours. Slag and dore bullion are poured into conical cast iron pots yielding buttons of 800 to 1000 troy oz. The dore typically contains 98.5% silver and 1 % gold. Slag is crushed and tabled to recover the trapped dore blebs and beads. Concentrate from the table is returned to the furnace. Table tails are sent to the crushing circuit and out to the leach pad. Solution application The area kept under leach at Rochester is approximately 130 000 m2 (1.4 million sq ft). Barren solution is delivered to the pad at 21.2 kL/min (5600 gpm) for a resultant application rate of 0.16 L/min/m2 (0.004 gpm per sq ft). A traditional solution sprinkling system using No. 12 Senninger Wobblers with individual pressure regulators was installed at the onset of leaching activities. The Wobblers were placed at 9.1-m (30¬ft) staggered centers and were fed off of a gridwork of Yellowmine plastic piping. Solution flow rates were moni¬tored to each panel. The onset of cold weather with an average nighttime temperature of -12° C (10° F) made it apparent that continual operation would not be possible with the sprinklers. A significant amount of ice was built up on top of the heap, making maintenance and pipe removal dangerous, if not impossible. Leach solution application was restricted to daylight hours to inhibit ice formation. Process plant flow rates were reduced to maintain steady-state operating conditions. However, as daylight temperatures dropped below freezing, ice continued to accumulate due to the sprays. Besides the obvious operating hazards brought on by the growing icefield, there was also the potential environmental hazard associated with an early thaw melting the ice too rapidly for the solution containment facilities. One other option for preventing ice formation was heating of the barren solution prior to spraying. Initial plant design allowed for expansion of the propane storage and distribution system as well as modification of the barren piping for a solution heater. This option was not exercised because the operating costs for an adequate system would have been prohibitive, and timely delivery of a system was not available. An investigation was conducted on the various drip irriga¬tion products available, since subsurface solution applicators would eliminate ice formation altogether. Systems utilizing external flow emitters were ruled out because of their ten¬dency to clog when buried. Emitter systems using perforated tubing were also eliminated from consideration due to their inability to adequately control flow over required lengths of tubing. An in-line emitter system was finally selected which demonstrated clog resistance and adequate flow control, enabling direct burial.

Jan 1, 1990

By Kathleen Kreiss, A. James Ruttenber

INTRODUCTION Early in the morning of July 16, 1979, there was a breach in the earthen retaining dam of a tailings pond at the United Nuclear Corporation's (UNC's) Church Rock uranium mill. The acidified liquid and tailings slurry spilled through the damaged portion of the retaining wall into an arroyo that is a tributary to the Rio Puerco river system. The Rio Puerco runs through Gallup, New Mexico, and eventually crosses the New Mexico-Arizona border (Fig. 1). On its way to Gallup, the Rio Puerco and its tributaries pass through land with a checkerboard pattern of ownership, with portions owned or leased by the Navajos, individuals, the Bureau of Land Management, and the State. In terms of tailings liquid volume (3.6 x 108L; 94 million gal), the UNC spill ranks as one of the largest. The mass of solids released in the slurry (10.0 x 105 kg; 1 100 tons) appears to be close to the median for accidents of this kind, however [U.S. Nuclear Regulatory Commission (NRC), 1979]. The UNC first opened its Church Rock uranium mill in 1977 on land adjacent to acreage belonging to the Navajo tribe. The mill, which is next to the UNC Church Rock mine, is located approximately 16 km (10 miles) northeast of Gallup, New Mexico (Fig. 1). Gallup, a town of 18 000 people, is the closest population center. The region surrounding the plant site is sparsely populated by Navajos, at a density of approximately 5.8 persons/km2 (15 persons/sq mile). The UNC mill and mines employ approximately 650 persons, and the adjacent Kerr-McGee uranium mine employs about 300. The UNC mill normally processes 3.2 x 106kg/day (3 500 tons/day) of uranium ore, depositing the acidified tailings slurry in a series of three earthen holding ponds. The tailings ponds are located east of the pipeline arroyo that feeds into the Rio Puerco approximately 2.4 km (1.5 miles) from the southernmost tailings dam. The liquid portion of the tailings slurry evaporates in the ponds; hence, under normal conditions, there is no surface flow from the holding ponds to the arroyo. Both runoff from the plant site after heavy rains and possible seepage from the tailings ponds may deliver radionuclides to the arroyo-river system, however. The dam across the southernmost tailings pond was considered to be in keeping with the state of the art. However, the New Mexico Environmental Improvement Division (NMEID) had warned UNC about dangers of locating the pond over a heterogeneous geological formation. The state Engineer's Office approved of the site only after UNC agreed to strict design criteria. Others have pointed to dangers of constructing earthen dams for impoundment of uranium mill tailings (Carter, 1978). Causes of the dam break were multiple: the UNC mill filled the tailings pond to a level that exceeded permit criteria; the tailings pond was lined improperly; the dam was constructed using clay that was compacted excessively, resulting in cracking and subsequent seepage; and the unstable substrate beneath the dam permitted differential settling. The UNC Church Rock mine has continuously released dewatering effluent into the pipeline arroyo at a rate of 88.3 L/sec (1 400 gal/min) since 1968. Before 1975 this effluent was not treated; after 1975 it received precipitation treatment for removal of Ra-226. Radionuclides are also released into the river system through the dewatering of the Kerr-McGee uranium mine 1.6 km (1.0 mile) north of the UNC mill. During usual mining operations, approximately 227 L/sec (3 600 gal/min) are released into the pipeline arroyo and subsequently into the Rio Puerco. The Kerr-McGee mine began continuous release of dewatering effluent in January 1972. In 1974 Kerr-McGee began Ra-226 precipitation treatment of its dewatering releases, but NMEID data indicate that treatment has often been incomplete. The effluent from both mines has been responsible for transforming the downstream portion of the Rio Puerco from a sporadically dry riverbed into a continuously flowing stream and has contributed to the current levels of background radiation along the river system (Table 1). This paper will summarize the postspill monitoring efforts and relate the assessment of this spill to the general question of evaluating the health effects of nuclear fuel-cycle wastes. The data pertaining to the measurement of radionuclides in the Church Rock environment and the radionuclide concentrations in animals will appear in forthcoming reports. CHURCH ROCK HEALTH EFFECTS ASSESSMENT APPROACH The initial health effects evaluation involved identifying the radionuclides that were released into the river system from the tailings pond. Table 1 lists the State of New Mexico maximum permissible radionuclide concentrations for liquids released to unrestricted areas, the typical tailings liquid concentrations, and postspill river water concentrations. The tailings liquid contained comparatively high levels of Th-230, Ra-226, Pb-210, and Po-210--all of which, according to postspill river water samples, had exceeded the state maximum permissible concentrations (MPC) at one time or another. After the radionuclides in the tailings were identified, pathways through which humans could be exposed were clarified. Environmental monitoring data were then used to quantify the important pathways of human exposure. Water samples were collected from the river, from test wells dug near

Jan 1, 1981

By Joyce A. Ober

Minerals for electronic and optical uses divide easily into two sections: 1) quartz and 2) minerals other than quartz. QUARTZ The properties of quartz crystal that make it useful for radio communications were discovered in 1918. Since that time, an in¬dustry for the mining and processing of natural quartz crystal has grown, matured, and been almost entirely replaced by new tech¬nology. The new technology still involves quartz crystal, but ma¬terial that is grown rather than mined. An economic summary of the commercial growing of quartz crystals has a place in a handbook directed to the mineral engi¬neering industry because quartz crystals have long been an impor¬tant commercial mineral, and the raw material for cultured quartz - ¬that is to say, quartz crystals grown through the ingenuity of man - is still natural quartz. Nearly all the natural crystals that have been used for elec¬tronics and optics came from Brazil. The larger pieces which met rigorous standards of quality were used for electronic and, to a lesser extent, optical components. Smaller pieces and fragments were used for vitreous silica. The need for high quality material in quantity led to US government sponsored research and exploration programs in the 1940s. No deposits meeting the very rigid requirements for electronic-grade quartz were found, but other projects resulted in the development of a process for the factory growth of beautiful crystals of prescribed shape, size, and quality. Domestic deposits of appropriate quality were identified to use as raw materials for the quartz culturing process. The development of the cultured quartz crystal illustrates the success that technology can have in adapting a product of the mine to increasingly sophisticated uses. A remarkable achievement per¬haps, but foreshadowed by experiments by Giorgio Spezia (1908), an Italian geologist studying the relative effects of temperature and alkaline environment on the solubility of quartz. Modem radio equipment is most often controlled as to fre¬quency by the presence in the circuit of a separately added crystal¬ - the 1918 discovery responsible for the existence and growth of the quartz industry. The crystal is quartz, but this component is a carefully oriented and prepared slice from a crystal, but not a crystal as recognized by a rock hound or seen in a museum. How quartz operates to control frequencies is not a proper subject for a handbook on industrial minerals, and references should be consulted (Cady, 1964, Mason, 1964). Quartz belongs to a class of materials called dielectrics: those that do not conduct an electric current but permit electric fields to exist and act across them. Quartz shows the piezoelectric effect, which means that when a quartz plate is mechanically deformed against its natural stiffness, one of its surfaces becomes negatively charged, the other positively charged. When the plate is released quickly from the stress, the charges disappear as the plate regains its original shape, but because of mechanical momentum the plate deforms in the opposite direction (to a lesser amount) and the surfaces correspondingly become charged in the opposite direction. By thinly coating the two surfaces with metal and attaching flexible wires, these charges can be brought into an electronic circuit. If the surfaces are suddenly electrically charged by movement of current through the wires, the converse piezoelectric effect occurs and the plate deforms. Carry the thought further and it is realized that an alternating current flowing through the wires responds to the mechanical oscillation. By controlling the thickness of the plate, its mechanical vibration frequency can be varied through a wide range. One type of quartz plate, the AT-cut, has a precisely defined orientation with respect to the crystallographic axes of the crystal and vibrates on a microscopic scale much as a book would deform when placed flat on a table and the top cover moved parallel back and forth with the hand. At least 17 other orientations have been studied, some of which have preferred uses in various applications (Cady, 1964). The quartz crystal industry is composed of three main segments (excluding fused quartz and quartz used for optical purposes): 1. Natural electronic-grade quartz crystals. Mined quartz suitable for fabrication into piezoelectric units. Zlobik (1981a) esti¬mated the waste to ore ratio at 1:1000 to 1000 000, depending upon the deposit. 2. Lasca. Mined quartz usable as feedstock in the production of cultured quartz. Approximately 0.63 kg of lasca are required to produce 0.45 kg of cultured quartz. 3. Cultured quartz. Cultured quartz is produced from lasca feed¬stock in a process of crystal growth in an autoclave under conditions of heat, pressure, and time. It is estimated that 0.45 kg of cultured quartz is equivalent to 1.4 to 4.5 kg of natural quartz crystal in yield of commercial quartz suitable for slicing into piezoelectric units. The chronology of the development of the quartz crystal industry both natural and cultured follows: Date Comment 1918 Discovery of the piezoelectric effects of quartz crystal 1921 Application of the piezoelectric effects of quartz crystal in the circuitry of radios 1948 Establishment of a quartz crystal commodity stockpile by the US Government 1952 US consumption of natural quartz crystal at an all time high of 228 t 1958 First commercial production of cultured quartz crystal 1970 Cultured quartz crystal production exceeds imports of nat¬ural quartz crystal 1971 Cultured quartz crystal consumption surpasses natural quartz crystal consumption

Jan 1, 1994

By James M. Barker, George S. Austin

Stone, one of the oldest building materials, today remains a well-established material throughout the construction industry. The use of natural stone is much less prevalent now than in the past. It is still widely considered to be the most aesthetically pleasing, prestigious, and durable building material. New and re-opened quarries are coming onstream to meet increased demand related to new building technology and increased residential use of stone. CLASSIFICATION No classification can completely eliminate overlap between dimension stone, aggregate, and decorative stone because most stone is multi-purpose. Many used for decorative purposes are not produced specifically for that end use. Rock otherwise considered waste in dimension stone or aggregate quarries can be decorative stone coproducts (Fig. 1). Many uses require a compromise between decorative and structural qualities (Bowles, 1992, written commu¬nication). Shipley (1945) used decorative stone interchangeably with or¬namental stone. Gary et al. (1972) defined decorative stone as that used for architectural decoration, such as mantels, columns, and store fronts, but added that it is sometimes set with silver or gold in jewelry as curio stones. Bates and Jackson (1987) also restricted decorative stone to that used for architectural decoration. Meanings of otherwise identical terms used in the stone industry differ be¬tween geologists, engineers, and quarriers. They often carry a much broader meaning for quarriers and engineers compared to their very specific use by geologists (Makens et al., 1972). Decorative stone, including ornamental stone, is more broadly defined by geologists as any stone used primarily for its color, texture, and general appearance. It is not used primarily for its strength or durability, such as construction stone, or in specific sizes, such as dimension stone. The decorative stone industry uses a much wider range of stone types compared to stone that is dimensioned. Decorative stone usually serves some structural pur¬pose, but is not load-bearing to any great extent. Weak or costly stones serve in decorative, not structural, applications. STATISTICS AND END USES Decorative and dimension stone data are difficult to separate because the US Bureau of Mines keeps statistics only on dimension stone and crushed stone. The value of domestic dimension stone production in 1990, which includes some decorative stone, was about $210 million compared to imports of about $524 million and exports of about $35 million. Production was 1 080 t of which at least one-third was for decorative uses (Taylor, 1992). The principal uses are rough blocks in building construction (23%) and monu¬ments (18%); the remainder is used as ashlar (18%), curbing (12%), and miscellaneous (29%). Major rock types are granite (50%), limestone (30%), sandstone (10%), slate (3%), marble (2%), and other (5%) (Harben, 1990). Crushed stone valued at $5.6 billion was produced in the United States in 1990 by 1700 companies operating 3400 active quarries in 48 states (Tepordei, 1991). About 52% is used in con¬struction, 9% in cement and lime manufacturing, 2% in agricul¬ture, 2% in industrial uses, and 35% for unspecified uses including decorative aggregate. Limestone and dolomite comprise about 71%, granite 14%, and traprock 8% of the stone crushed in the United States. The remaining 7% are, in descending quantity, sandstone, quartzite, miscellaneous rock, marble, shell, calcareous marl, volcanic cinder and scoria, and slate. The basic types of decorative stone are: rough stone, aggregate, cut or dressed stone, and manmade stone [(Table 1)]. Rough Stone Rough stone is used as it is found in nature with very limited processing such as minor hand shaping, edge fitting, and size or quality sorting (Perath, 1992, written communication). This stone type is often marketed locally in relatively small tonnages and includes fieldstone and flagstone. The primary end uses of rough stone are landscaping, edging, paving, or large individual stone landscape or interior accents [(Fig. 2)]. Fieldstone: Fieldstone is picked up or pried out of the ground (gleaned) without extensive quarrying and includes garden or large landscaping boulders (Austin et al., 1990, Hansen, 1969). Boulders and cobbles may be split or roughly trimmed for use in rubble walls and veneers, both interior and exterior. Popular fieldstone rock types include sandstone, basalt, limestone, gneiss, schist, quartzite, and granite, but many others are suitable. Much fieldstone is col¬lected by individuals or small companies because the industry is labor intensive and markets are small. The stone may be sold locally in small quantities from the back of vehicles (Austin et al., 1990). Fieldstone includes many rock types, sizes, and shapes with the only common denominator that it must be set by hand and be durable (Power, 1992, written communication). Moss Rock. Moss rock is fieldstone partially covered by algae, mosses, lichens, and fungi that give the rock an aged and variegated patina (Austin et al., 1990). The plants are supported by moisture and nutrients in the stone. Moss rock is used for landscaping, walls, and fireplaces. Although almost any durable rock can be a moss rock, most are slabby or rounded sandstone and limestone (Fig. 3). Flagstone: Flagstone or flagging consist of thin irregular slabs used for paving, walkways, and wall veneers. Random-shaped flagging is produced widely in the United States. Suitable stone

Jan 1, 1994

By A. L. Frank, Benton

INTRODUCTION The selective sensitivity of plastic nuclear track detectors (PNTD's) to low energy 4He particles in an environment which also contains gamma and beta radiations has made these detectors prime candidates for the dosimetric measurement of the concentrations of radon and its daughter products in mine air. Their passive, integrating mode of measurement, small size, low weight and inexpensiveness are attractive characteristics for large scale personnel dosimetry. The detectors can be used in either active or passive dosimeters. In active devices, the PNTD is placed in close proximity to a sampling filter. The filter collects, from a calibrated air flow, all the daughter nuclei which are in suspension. As the daughter nuclei pass through the decay chain, through 214Po, a fraction of the 4He particles emitted are registered as latent tracks in the plastic. In passive devices, the PNTD is placed in direct contact with the ambient air containing the radio-nuclei concentrations to be determined. The active dosimeters have the advantage of excellent sensitivity, and the measured track densities yield very close approximations to accumulated Working Level (WL)** exposures. They have the disadvantage that the simplicity of the passive PNTD is lost, since a battery-driven constant flow-rate air pump is a necessity. The passive dosimeters are simple in construction and use, but they have the disadvantage that WL exposures are not directly measured and certain assumptions concerning radon and daughter equilibrium conditions must be made in determining WL exposures from the measured track densities. Also the sensitivities, in track densities per Working Level Hour (WLH) exposure, are much less than for active dosimeters. Dosimeters of both types have been investigated. Passive devices have been tested extensively both in the laboratory and in mines. Active devices have been laboratory tested. In our earlier measurements, cellulose nitrate plastic detectors were used exclusively, since this material had the highest sensitivity of the PNTD's then in use. When the properties of CR-39 plastic were discovered (Cartwright, 1978; Cassou and Benton, 1978), this material was used, where possible, to take advantage of its superior sensitivity. The results of our earlier work have previously been published (Frank and Benton, 1977). * Research sponsored by the Bureau of Mines, U.S. Department of Interior, under Contract No. JO-188003. ** A Working Level is defined as any combination of the short-lived radon daughters containing 1.3 x 105 MeV of potential 4He-particle energy per liter of ambient air. PASSIVE DOSIMETRY In the earliest testing of PNTD's for passive WL dosimetry, a single strip of cellulose nitrate plastic was used to determine the total cumulative 4He-particle activity in the air of simulated uranium mine and uranium mine environments (Rock, 1968; Rock [et al], 1969; White, 1969). The measurements yielded a close relationship between track densities and WL exposures in a controlled, static environment, but poor accuracy in the active mine tests. It was assumed that equilibrium differences contributed largely to the variations found. The response of PNTD's to airborne 4He-particle emitters, derived by Lovett (1969), demonstrated that detectors such as cellulose nitrate, which has a sensitivity cutoff at 4He-particle energies below the emission energies of radon and its daughters, are equally sensitive to the activity concentrations of radon, 218po(RaA) and 214Po(RaC'). Since radon does not contribute to WL as it is defined, and since, under normal ventilated uranium mine conditions, the radon activity constitutes about 40% to 70% of the total 4He-particle activity, the WL exposures calculated from measured track densities are very sensitive to the particular radon daughter equilibrium conditions. Also, the detector does not weight the individual daughter activities in proportion to their importance to WL. The equation for WL is WL = 0.00103C2 + 0.00507C3 + 0.00373C4 (1) where C2, C3 and C4 are the activity concentrations of RaA, RaB and RaC-C', respectively, in pCi/[L]. The detector leaves C3 unmeasured and weights C2 and C4, equally. [This problem has been approached by Domanski et al (1975, 1979), for some non-uranium mines, by measuring equilibrium conditions throughout the mines to determine a mean value for A, the inverse of the Working Level Ratio (WLR = 100 WL where C1 is the activity of radon-222 in pCi/[L].) The distribution of values allowed Domanski to determine probable errors in calculating WL exposures from track densities. However, measurements of equilibrium conditions in U.S. uranium mines (Breslin et al, 1969; Holub and Droullard, 1978) indicate that this method would not be accurate enough for uranium miner personal dosimetry.] A two-element dosimeter was designed at our laboratory to compensate for equilibrium variations. The first element is a radon detector; the second is a detector for the total 4He-particle activity in the ambient air, just as in the single element dosimeters cited above. The addition of the radon detector allows the equilibrium conditions for the individual nuclei to be determined, given certain assumptions concerning the interrelationships between the nuclei concentrations in mine air. The two detectors and the problems involved in calculating WL exposures from their measurements are discussed below.

Jan 1, 1981

By Bernard L. Cohen

I.INTRODUCTION The B E I R-III Report (NAS-1980) introduces large increases in the estimated health effects of radon as compared with previous work (NAS-1972). It is the purpose of this paper to point out some important failures of these new BEIR-III estimates, to offer a general critique of the procedures used in obtaining them, and to offer more rational estimates. In Sec. II we use the BEIR-III model to calculate the risk to non-smokers from environmental radon, and show that it predicts more than twice the total lung cancer rate actually experienced by nonsmokers. In Sec. III we review the histological evidence which shows that no more than about 10% of the lung cancers among non-smokers can be due to radiation. In Sec. IV, we discuss alternative causes of lung cancer, which further reduces the fraction that can be caused by radiation, and in Sec. V we summarize and conclude that the BEIR-III model over-estimates the lung cancer rate in nonsmokers due to environmental radon by at least a factor of 40. In Sec. VT we review the evidence on risk of radon exposure to smokers, and conclude that it is probably not more than four times the risk to non-smokers; this means that the BEIR-III model over-estimates the risk of low level radon exposure to smokers by at least a factor of 10. In Sec. VII, we consider the reasons for the large over-estimates in the BEIR-III. Report. II. BEIR-III LUNG CANCER RATES DUE TO ENVIRONMENTAL RADON AND COMPARISON WITH TOTAL LUNG CANCER RATES AMONG NON-SMOKERS The BEIR-III Report gives the following estimates of the lung cancer risk from low-level radon exposure in terms of working-level-months (WLM): age 35-49, risk = 10 x 10-6 /yr-WLM 50-64, 20 >65, 50 where ages refer to age at death. For latent periods between exposure and onset of these risks it gives age 0-14, latent period = 25 years 15-34, 15-20 years (we use 17 yr) >35, 10 years where ages refer to age at exposure. This is a clear and unambiguous model which is readily usable for deriving numerical estimates. We begin by using it to calculate lung cancer rates due to environmental radon. *This is an abridged version of a paper scheduled to appear shortly in Health Physics. The first step in this process is to estimate the environmental exposures; this was done in a recent paper (Cohen-1981) which concluded that these are about 0.22 WLM/year. In Table 1, this is used to calculate the BEIR-111 predictions for radoninduced lung cancer rates in the U.S. (Col. (5)), and by combining these with population statistics, it is shown (Col.(7)), that it predicts about 24,500 fatalities per year, almost one-third of all U.S. lung cancers. The comparison between the age-specific expected rates from Col. (5) of Table 1 and observed rates among non-smokers is shown in Table 2. The recent paper by Garfinkel (1980) presents the results of a 12 year follow-up on one million Americans in a study by the American Cancer Society. The paper by Hammond (1966) gave the results of the first four years of that study. The paper by Kahn (1966) is based on the so-called "Dorn Study" of 293,000 U.S. veterans of World War II who carry government health insurance. It represents 8 and 1/2 years of follow-up. A recent update on that study (Rogot-1980) does not give absolute lung cancer rates, but the age-standardized ratio between smokers and non-smokers has remained the same which indicates that there has probably not been an important change in the rates for either. The paper by Hammond and Horn (Ha-1958) was an early study by American Cancer Society. It is immediately evident from Table 2 that the BEIR-III estimates for lung cancer induced by environmental radon exceed the [total] lung cancer rates due to [all] causes among non-smokers by about a factor of two at every age. It is only fair to point out that this does not represent a direct discrepancy with the BEIR-III Report since the latter states that its estimates for non-smokers may be too high by a factor ranging from 1 to 6, favoring a factor intermediate between these. Comparisons can also be made with total lung cancer incidence for all ages. A paper by Hammond and Seidman (Hammond-1980) gives the rate for ages above 40 to be 177 x 10-6/year for men and 124 x 10-6/year for women, whereas the rate calculated in Table 1 from BEIR-III for ages above 40 is 309 x 10-6, a factor of two higher. For all ages, the rate among women was reported as 36 x 10-6/year (Hammond 1958) as compared with 114 x 10-6/year calculated from BEIR-III in Table 1, a discrepancy of well over a factor of two. All of the data we have presented are basically from three study groups, but in all three cases the BEIR-III estimates for lung cancer induced [by environmental radon alone] are a factor of two higher than actual [tota] lung cancer rates among non-smokers. Another approach to comparing the BEIR-III pre-

Jan 1, 1981

By S. J. Bluhm

INTRODUCTION Lesotho is a mountainous area of southern Africa from which water is to be exported in an extensive tunnel system, to industrial regions inland. The related tunnelling project has involved a num- ber of drives using tunnel boring machines [TBMs] to excavate about 100 km of 5 m diameter water tunnels [von Glehn and Bluhm, 1995). This paper describes the ventilation and cooling of some of the tunnel drives from both the operational and design points-of-view with a particular focus on heat generation. There were many common features in all of the drives but this paper is focused mainly on the Hlotse drive which was 18,4 km long. The drives were ventilated using forced ventilation systems to provide appropriate air flow throughout the tunnels and face zones. In addition, the Hlotse drive required refrigeration equip- ment which provided chilled water to the tunnel. While the sec- ondary ventilation systems play an important role in gas and dust handling, the paper concentrates on the primary ventilation and cooling issues. The ventilation of these tunnels was an exacting exercise be- cause: • Rock temperatures and geothermal heat flow were high. • TBMs with relatively high power ratings were used. • Diesel locomotives were used. • Drives were relatively long. • High altitude meant a low air density. An important feature was the simulation and monitoring of the ventilation and heat flow components and the project was characterised by analysis, monitoring and ongoing tactical decision-making throughout the progress. The thermodynamics of the systems were complex because there were many interactive effects and analyses were carried out using special computer pro- grams. The monitoring confirmed the accuracy of the models, and in this manner it was possible to confidently ensure healthy and safe working conditions and still minimise costs. Local ambient climate conditions range from temperatures higher than 35 "C in summer to below -10 OC in winter. Based on available statistical data and the thermal storage/damping effects in the system, design summer ambient conditions were taken as 15 OC/25 "C wet-bulb/dry-bulb. The barometric pressure was 80 kPa and due to the altitude, the ambient air density was only 0,9 kg/m3. The local Authority specified a maximum in-tunnel wet- bulb temperature [at any point] of 32,O OC and a mean wet-bulb temperature [from all locations] of 27,5 OC maximum. The maxi- mum height of ground cover above the tunnel was 1 200 m and the maximum virgin rock temperature was 41 OC; see Figure 1. Diesel dilution criteria specified by the local Authority was a minimum of 0.1 m3/s per rated kW of diesel engine. Other requirements related to gases such as CO, CO2, NOx and CH4 [and the need for intrinsically safe equipment] but these are not of direct relevance to this paper. The actual average face advance was about 30 m/d with good days achieving 60 m/d and good months achieving 1 000 m [23 working days]. The original design tunnelling rate was 50 m/d. DESCRIPTION OF HLOTSE DRIVE VENTILATION AND COOLING SYSTEM The ventilation requirements in the tunnels were dictated by heat and diesel dilution needs. The best ventilation and cooling policy is generally a balance between using increased quantities of fresh air or refrigeration [or both]. In this particular scenario it turned out that, since the diesel emission criteria required large quantities of air, the refrigeration needs were modest. The drive was ventilated using a ducted, forced ventilation system from fans located at the portal. The maximum ventilation requirement was 51 m3/s when the drive was at 18.4 km. From a heat flow point of view, the worst scenario was a heat load of 3.5 MW when the drive was at 7 km. This was cooled by the ventilation air and a supply of chilled water to the tunnel. Refrigeration and chilled water system In the design phase, a detailed comparison was carried out between two general alternatives for providing refrigeration. First, was a system in which refrigeration sets and air coolers are installed on the TBM train; the refrigeration sets are cooled by condenser water piped to and from cooling towers at the portal. Second, was a system in which refrigeration water chillers are in- stalled at the portal and chilled water is piped into the tunnel. The detailed comparison indicated that the capital and running costs of the second system were at least 60 % lower than the in-tunnel plant. There were also many obvious practical benefits for favouring the portal system. The refrigeration plant supplied 23 11s of cold water at a temperature of 10 OC. After providing the cooling effect in the drive, the water returned to the portal where it was initially cooled in open-circuit evaporative pre-cooling towers, chilled in the refrigeration plant and then returned to the tunnel. The cold water flowed into the tunnel in an insulated supply pipe and returned in an uninsulated pipe; the water was simply circulated to the end of the pipe and returned. The cooling effect in the tunnel was achieved entirely through heat transfer from the pipe [long linear heat exchanger] and no air coils or other heat exchangers were required. The cooling requirements were satisfied by the heat transfer to the returns chilled water steel pipe [200 mm]. The pipes were eventually installed to a maximum distance of 10,8 km in what was considered a very practical and cost effective solution.

Jan 1, 1997

By N. Arbiter

Introduction The theoretical analysis of tumbling mill energetics and performance has largely neglected mill dimensionality, and, in particular, the importance of the length/diameter (L/D) ratio. This is in spite of the fact that practice varies substantially: geographically, for semi-autogenous and autogenous mills, as between North America and South Africa/Scandinavia, and historically, for overflow ball mills, for which the L/D ratio has increased significantly from the earliest small mills to the largest mills currently. The present study is concerned primarily with the influence of the L/D ratio on the design and operation of overflow ball mills, on the occurrence of the overload phenomenon, and on the limits, if any, it may impose on mill capacities because of critical pulp axial velocity limits. It will be shown that the shape factor is of major importance in this area and that its adjustment to the extent that this is practical should remove the diameter limitations previously postulated for this mill type. Dimensionality in mill design Ball mill shape factors in the period prior to 1927 (Taggart, 1927) averaged 1.1/1 for 29 center discharge mills and 1.0/1 for 30 peripheral discharge mills. With the resumption of new plant construction after the 1930s depression, the Morenci concentrator continued the 1/1 ratio with its 3.1 x 3.1 m (10 x 10 ft) mills. The ratio was increased progressively from then on, reaching 1.6 and 1.8/1 for the largest overflow mills currently. As shown recently (Arbiter, 1989), this is in sharp contrast to autogenous and SAG mill shapes, for which the ratio averages 0.4 in North America. On the other hand, South African practice, starting at the turn of the century with autogenous mills having 4/1 ratios, moved toward 1/1 until recently, when a 2.5/ 1 ratio mill was installed. The reasons given for such divergent practice for mill shape factors are in some respects contradictory and generally inconclusive. The most complete discussion from a practical viewpoint (Dor and Bassarear, 1982) is limited to primary SAG and autogenous mills. Considerations of ball mill dimensionality have had a twofold direction. On the one hand, it has been argued that ball mill efficiencies should increase with increasing diameter and that the specific energy for a particular grind should be reduced accordingly. An inadvertent test of this idea at the Bougainville operation (Burns and Erskine, 1983) resulted in drastic underpowering, which led to failure to reach design capacity until additional mills were installed. This can be taken as strong evidence against any increase in efficiency with diameter. In another direction, it has been argued (Arbiter and Harris, 1982, 1983) that there is a limit to ball mill diameters because of the demonstrable limit to axial flow velocities evident in the overload phenomenon, which as a fact is incontrovertible. But that it places a limit on mill diameters overlooks the evidence given below that appropriate variation in the L/D ratio will permit major increases in diameter, limited only by constructional or economic factors. The present study was directed toward quantifying the overload phenomenon through examination of the influence of mill dimensionality variation and mill operating variables on its occurrence. It is shown that varying the operating conditions, specifically the load fraction and the fraction critical speed, can reduce the risk of overload for existing operations; while appropriate decreases in the L/D ratio can minimize the risk in the design of new circuits. Ball mill overload Ball mill overload is a consequence of the approach to a critical velocity with increasing feed rates or circulating loads. Although the effect has been known for over 50 years, there have been no previous attempts to quantify it. The following description of the ball mill as a flow system is the preliminary to a quantitative analysis: 1) In the absence of a ball load, axial flow of pulp through a mill resembles open channel flow, except for disturbances near the shell due to shell/lifter rotation. 2) For a given ball load (Lf), the void fraction available for pulp hold-up (H) for the ascending portion of the load is approximately 0.4 Lf. In the descending portion, it is greater than this and increases with fraction critical speed (Fc) because of load expansion. 3) Increasing the feed rate increases pulp hold-up and progressively fills the available void space. At a critical flow rate, which depends on system geometry, hydraulic head and pulp rheology, void filling reaches its limit; a pool forms rapidly and fills available space outside the ball load and up to the overflow level. Prior to this, pulp discharges mainly along the ascending rim of the overflow. 4) For a small mill (Lo et al., 1990), it has been shown that with increasing feed rates the critical filling is at or near 50% of the mill volume. Power drops rapidly when this level is reached, as required by the torque formula. 5) The transition to overload is associated with the following phenomena: a) The decrease in power draw. b) Damping of mill sound. c) Reduced comminution of coarser feed sizes, probably due to reduced direct impact in the presence of a pool. d) Increased circulating loads, which further intensify the overload, and increased density of cyclone underflows, which can lead to roping. e) Conditions beyond overload are not known because feed rates are not delibelrately increased beyond this point. f) The existence of the phenomenon limits the capacity of a mill with a fixed set of operating conditions and can prevent the balancing of hard ore feed rate decreases by increases in soft ore rates.

Jan 1, 1992

By William T. Jr. Renfroe, Donald C. Gipe

INTRODUCTION Historically, pollution control in the metal-ore mining industry has been very limited. Unless mine water contained large quantities of solids, it was generally discharged without any treatment. If treatment was used to control solids, it was principally the provision of a settling basin in the form of a tailings impoundment used in conjunction with an associated metal ore dressing facility. Recently, however, a growing awareness of the adverse environmental impacts of mine drainage, coupled with strict environmental laws, has prompted the mining industry to look at new technologies and to refine the existing methods to further treat the wastes generated. This industry is unique in that waste loadings are extremely variable, and a "typical facility with typical waste loads" does not exist. Consequently, one waste- water treatment system cannot be utilized on an industry wide basis; rather, each treatment system must be designed specifically for the pollutants in each individual discharge. Public Law 92-500, the Federal Water Pollution Control Act (FWPCA) Amendments of 1972, became effective on Oct. 18, 1972. This law completely restructured Federal laws and philosophies underlying the Federal approach to water pollution control. Prior to the 1972 amendments, the principal Federal regulatory tool had been water-quality standards based on a designated use for a particular body of water. The concept was that waste disposal into water bodies is a desirable and acceptable use of the water body if it does not interfere with other beneficial uses. This had the effect of requiring various degrees of treatment and, consequently, various economic hardships on industries de- pendent upon their location. In many waterways. it is very difficult to quantitatively relate discharges to water quality. The 1972 amendments changed the basic philosophy, as stated in the Senate Committee report on the bill, to ". . . no one has the right to pollute . . . that pollution continues because of technological limits, not because of any inherent right to use the nation's waterways for the purpose of disposing of wastes." Pursuant to Sections 301, 304(b), and 306 of the FWPCA Amendments of 1972, the US Environmental Protection Agency (EPA) was required to establish effluent standards applicable to all industrial discharges. These standards must be based upon the application of the "best practicable control technology currently avail- able" (BPT) and the application of the "best available technology economically achievable" (BAT). The BPT and BAT levels must be achieved industry-wide by July 1, 1977, and July 1, 1983, respectively. WASTE SOURCES The waste-water situation in the mining segment of the ore mining and dressing industry is unlike that encountered in most other industries. Most industries (e.g., the milling segment of this industry) utilize water in the specific processes they employ. This water frequently becomes contaminated during the process and must be treated prior to discharge. However, in the mining segment, process water normally is not utilized in the actual mining of ores (exceptions are hydraulic mining operations and dust control), but it is a natural occurrence that interferes with mining activities and must be removed before mining can commence. Water enters mines by ground-water infiltration and surface runoff, and it comes into contact with materials in the host rock, ore, and overburden. The underground mine must pump large quantities of ground water to prevent flooding of the mine. Water from surface mining operations generally occurs as a result of surface runoff of rainwater. Generally, mining operations control surface runoff through the use of diversion ditching and grading to prevent, as much as possible, excess water from entering the working area. Nevertheless, some surface runoff does come into contact with the working area and may become contaminated. The quantity of water from an .ore mine is unrelated, or only indirectly related, to production quantities. De- pending upon its quality, the mine water may require treatment before it can be discharged into the surface drainage network. The variability of water quality from mines can best be demonstrated by looking at Table 1. This table shows the range of pollutant concentrations in untreated discharges from three different categories of mines (as categorized by EPA in the development of BPT and BAT effluent standards for the metal-mining industry). Data for this table were obtained during EPA's preparation of effluent standards for this industry. The parameters shown on the table are the pollutant parameters of primary interest in this industry; blanks in the table indicate that data were not available, and the parameter is not expected to be present in significant quantities. Other pollutant parameters are present in mining waste water, but they are either incidentally removed in the treatment process or are found only in trace amounts. The three categories comprise more than 90% of the metal production value in the United States and approximately 95% of the total mine discharges. It is important to note that not all parameters are found in significant concentrations at all locations. IMPACT ON WATER QUALITY One of the most troublesome mine-drainage problems is acidity. Although generally associated with coal mining, acid mine drainage frequently occurs from other types of mines. Although the exact mechanism of acid mine drainage is not fully understood, it generally is believed that pyrite (iron sulfide, FeS,) is oxidized by oxygen (Eq. 1) or ferric iron (Eq. 2) to produce ferrous sulfate (FeSO4) and sulfuric acid (H2SO4) . The mining of ores associated with pyritic material exposes the pyrites to water and oxygen and grossly accelerates the natural oxidation processes, resulting in the significant production of acid mine drainage.

Jan 1, 1982

By B. A. Ferguson, P. W. MacMillan

INTRODUCTION Jack Spalding in Deep Mining, Chapter 3, states: "In deep mining, to start stoping an orebody without a definite plan of operations covering the whole extraction from beginning to end is to invite serious trouble. Any stoping without plan is bound to leave, toward the end, a number of pillars, remnants, or promontories, which, as they are reduced in size, are liable to fail and cause a general collapse. The object of stope planning is so to mine an orebody that at no stage in the operation is a remnant left. Definition: `When a block of ore is stoped in such a way that eventually a small piece is left entirely surrounded by stoping, that piece is termed an island remnant.' The design of and adherence to a plan of stoping is of greater importance in preventing rock bursts than the type of ground control adopted." This, of course, is just an example of the old axiom that an ounce of prevention is worth a pound of cure. And in this context, mining does not have to be very deep before poor ground conditions, made worse by lack of good planning, force a mining company to resort to more expensive methods of ore recovery. At all times when rock pressures become excessive, stoping plans must be devised to avoid the creation of pillars or promontories. This is most readily recognized when mining deep narrow ore veins by open methods. It would be a little more general to say that stoping systems should seek to avoid or minimize localized concentrations of rock stress. Ore bodies, and particularly those at Falcon¬bridge, can reach large widths and leaving pillars for extraction at a later time cannot be avoided. However, good planning can reduce the effect of factors which create bad ground conditions. The necessary ingredients of planning are time and information: time to formulate and revise ideas; information, gained in exploration and early development, to provide the fullest knowledge of the ore body. In this way, plans become meaningful. It must always be kept in mind that the long-term results of good planning often require decisions not compatible with reaching earliest full production. PLANNING FOR DEEP MINING At Falconbridge Nickel Mines ore bodies presently being found and prepared for production lie at depths in excess of 610 m (2000 ft) below surface. Current un¬derground exploration is being directed towards favor¬able locations below 914.4 and 1219.2-m (3000 and 4000-ft) depths, at most mines. The Onaping Deep ore body has been traced to a depth of 1508.8 m (4950 ft) and at Falconbridge mine the ore body is under develop¬ment below the 5000 level. Planning for more efficient ground control has therefore become an important con¬sideration of future mining. HYDRAULIC BACK FILLING AND CUT-AND-FILL STOPING The advent of hydraulic back filling has led to the increasing use of flat-back cut-and-fill stoping methods during the last ten years, until it has probably become the commonest method of mining deep ore deposits in Canada. These methods account for almost all the pro¬duction from Falconbridge Mines at this time. The ad¬vantages claimed are: safety-a minimum of open ground; very little supplementary ground support (except rockbolted backs and walls); greater selectivity and flexibility in extraction (horizontal extension of the vein, high grade ore stringers going out into the walls); and high productivity and increased mechanization in larger ore widths. There is some difference of opinion as to whether hydraulic classified tailing back fill provides any great resistance to initial closure of the stope walls. The evi¬dence available suggests that it does not. Addition of portland cement to the back fill just prior to placement has produced a material having the properties of a weak concrete. It is confidently expected that this material will assist greatly in the efficient mining of pillars and generally in giving increased ground support to oppose closure. PRINCIPLES OF STOPE PLANNING BY LONGITUDINAL FLAT-BACK CUT-AND-FILL Again quoting from Spalding: "The practice of con¬trolling output by stopping and starting stopes is bad¬stopes once started should proceed without interruption, and control must therefore be obtained by altering the rate of stoping. It is therefore necessary that faces should normally advance at a medium rate which can be boosted, if required, to give temporarily a greater output. Modern high productivity stoping methods, however, generally aim to achieve and maintain the maximum output possible from a stope. Control of tonnage and grade, therefore, becomes a more complex problem requiring very careful scheduling and organiza¬tion. With the basic method of flat-back overhand stoping using hydraulic fill, it is not usually possible to mine in such a way as to avoid creating pillars." Consider an ore body as illustrated in Fig. 1. It is 274.3 m (900 ft) high by 228.6 m (750 ft) long on strike and averages 4.6 m (15 ft) in ore width. It rep¬resents approximately 907 000 t (1,000,000 st). The ore body could be mined by advancing one long single face through all the levels from bottom to top. A total face length of 228.6 m (750 ft) would accommo¬date five 45.7-m (150-ft) stopes. See Fig. la. Each of these stopes would produce approximately 1814 t/m (2000 stpm) and the ore body would require a period of ten years for complete extraction. A possible im¬provement to this scheme might be to stagger the indi¬vidual stope faces. See Fig. 1b. Problems of hanging wall failure can be reduced by this arrangement. From the point of view of reducing travelway maintenance, it is good practice to employ a stoping system of retreat. Stopes farthest from the shaft are mined first so that on completion, haulage drifts below can be aban

Jan 1, 1982

By W. Ripmeester

Canada is committed to the development of indicators to measure the contribution of its minerals and metals sector to sustainable development. The Minerals and Metals Indicators (MMI) Initiative is described, including drivers for the development of indicators, the MMI theoretical framework, the multi-stakeholder consultation process utilized in the development of indicators, and possible policy and business implications of MMI.

Jan 1, 2003

By M. Phillip Jameson, Paul Sennett

The beneficiation of kaolin by flotation has been practiced on an industrial level since 1961. A brief history of industrial kaolin flotation is given, with a description of the various processing schemes used by the companies that practice kaolin flotation. Some of the areas where improvements could be made in the beneficiation of kaolin are given.

Jan 1, 1989

By J. M. Gambogi

Demand for titanium minerals and pigments was nearly unchanged in 1996 compared with 1995. However, through mine expansions and process optimization by the TiO, pigment industry, the available supply of titanium minerals and pigments increased moderately. At the end of 1996, exploration, development and expansion programs were ongoing in the mining sector. But expansion projects in the pigment sector were delayed pending increased demand.

Jan 1, 1997

By Vljay K. Chhipa

Artificial intelligence is" an emerging technology in the field of computer application. Expert system have been developed to imitate human intelligence and reasoning process. Expert systems have much scope of application in the decision making process in mineral exploration as such decisions are highly subjective and expert opinions are very helpful. This paper presents a small expert system to analyze reasoning process in exploring for uranium deposits in sandstone.

Jan 1, 1989

By Z. M. Zhang, Y. Luo, S. S. Peng

This paper presents the precise definitions of a number of commonly used subsidence parameters. Using an integrated approach, the parameters were determined for 110 subsidence cases collected from major US coal fields. Attempts were made to establish empirical formulae for the parameters. The results of this study indicate that some traditionally held concepts should be changed.

Jan 1, 1997

By W. M. Ma, Daniel W. H. Su, K. Centofanti, Yi Luo, W. L. Zhong, Syd S. Peng

6.1 INTRODUCTION A surface structure will suffer damage when the additional stresses induced by ground deformations associated with surface subsidence, plus the original stress introduced by construction de¬sign, exceed the strength of the structural elements. In this con¬text, there are two methods available for preventing and control¬ling surface structural damage: one is to strengthen the structure and the other is to design the mining operations such that ground deformations at the structure site can be reduced to an acceptable level. Mining operations include panel layout and mining tech¬niques. These methods are detailed in this chapter. It must be noted that most prevention/control methods men¬tioned in this chapter are used in the countries where the reference papers are cited. In the United States, the coal operators are not required to take those measures mentioned in this chapter. Some of the methods described in this chapter cannot be implemented with¬out changes in the current mining practice as permitted by laws. In addition cost of implementing those methods are not considered here. 6.2 PANEL LAYOUT As shown in Figs. 2.9, 2.10, and 2.11, permanent ground deformations in a subsidence basin mainly concentrate near the edges of the underground opening, and can be divided into four zones. A structure located in different zones will be subjected to different types and magnitudes of ground deformations. In laying out the panels, Table 5.1 and Figs. 2.9, 2.10, and 2.11 could be taken into consideration. Attempts could be made to avoid placing the structure on a location where the ground deformation to which that structure is sensitive is at its maximum. Therefore rational design of the panel is the simplest way to reduce structural defor¬mations. Panel design involves the determination of panel dimension, panel edge location, direction of face advance, and use of yield chain pillars. A. Panel Dimension Since longwalls in the US employ a multiple-entry system, where rows of chain pillars are left unmined, subsidence over those chain pillars is usually smaller. Therefore, whenever possi¬ble, the panel dimension could be designed such that a major structure or structures are over those unmined chain pillars, be¬tween adjacent panels, or some distance beyond both ends of the panels. At the center of the supercritical final subsidence basin, a structure will not be subjected to any final or permanent ground deformations. In order to create such a condition, the panel width must be such that the structure will be located beyond the major influence zone of the final subsidence basin, the minimum dimen¬sion of which must be: [ ] where L is the width or length of the final mined out gob, t is the width or length of the structure to be protected, h is mining depth and [ ] s is the angle of full subsidence. B. Panel Edge Location Wherever there is a permanent panel edge, there are large ground deformations induced on the surface on both sides of the permanent panel edge. Therefore whenever possible the panel di¬mension should be designed such that the permanent panel edges could be located in the areas with the least impacts. In terms of permanent edge location, it is best to eliminate any permanent panel edge under a structure or groups of structures. If this cannot be done, the panel should be lengthened to reduce the number of permanent panel edges, or narrower multiple panels advancing in the same direction in a staggered manner could be employed. If the structures are located in Zones II and III, the longer dimension of the structure must be parallel to the nearest perma¬nent edge (Fig. 6.1). But in Zone IV, the longer dimension should be tangential to the corner of the permanent panel edge. If the structure is inclined to the permanent panel edge, it will be sub¬jected to twisting and shearing. C. Direction of Face Advance The direction of face advance should be parallel to the long axis of the structure. But if the structure is to be located at or close to the center of the final subsidence basin, the direction of face advance should be perpendicular to the long axis of the structure. Careful choice of the direction of multiple face advance is the most effective way to reduce structural deformation and thus dam¬age. This applies the principle of overlapping and cancellations of ground deformations, due to multiple face advance, at the right time and at the right intensity, e.g., opposing tilts, concave and convex curvatures, tensile and compressive strains are induced simultaneously on the structure to be protected by two or more faces. D. Use of Yield Chain Pillars In US longwall panels there are generally two or three rows of stiff chain pillars between the panels. The combined width of the chain pillars ranges from 100 to 350 ft(30 to 107m). depending on mining depth. In general, surface movement above the chain pil¬lars after the panels on both sides have been extracted is much smaller, as compared to that in panel center. Thus in order to create critical or supercritical width of opening and eliminate sur¬face bumps over the chain pillars, yield chain pillars may be em¬ployed (Jarosz and Karmis, 1986). However if yield pillars are to be used, it must be designed such that it yields totally right after the panels on both sides have been extracted. Unfortunately cur¬rent yield pillar design techniques cannot predict when and how much it will yield. In summary, whenever possible attempts could be made to lay out the panel in such a way that surface structures are located above chain pillars between panels or above solid coal beyond both ends of the panels. In those areas the surface structures will most likely be unaffected, or if affected, the damage is so minor that no remedial measures are necessary. 6.3 CONTROLLED MINING TECHNIQUES Several mining techniques are available for reducing the sur¬face ground deformations of specific types. Regardless of tech-

Jan 1, 1992

By Jorgen Elbrond, Xiaoli Tang

In this paper we apply continuous-time option pricing theory (Dixit 8 Pindyck 1994) to show how to obtain an optimal investment rule for a mining project which requires sequential in- vestments. The application per se is to some extent very simple. The main purpose of the paper, however, is to draw the attention of mining people to how the option pricing theory may be used to capital investment decisions. This topic in economics is quite new but is attracting a lot of attention.

Jan 1, 1996