Search Documents

By Randal J. Barnes

The nonlinear network optimization algorithm developed in this paper solves the following mine ventilation design problem: find the location and size of all fans and regulators which minimize the overall system power consumption while satisfying all physical, legal, and operational constraints. This work includes the formulation of the problem within the framework of mathematical optimization, the development of the computational theory, and the creation of an efficient solution methodology. The algorithm has been implemented on a microcomputer, and a small but representative example is included. The algorithm presented in this paper has four limitations. First, the optimization process is for a single point in time: the dynamic nature of a mine ventilation system in an evolving mine is not directly addressed. Second, the topology of the ventilation network is required input information. The mathematical algorithm does not optimize the size or location of the mine ventilation airways; however, the post optimality analysis techniques can be used to aid in topological decisions. Third, the problem is formulated using volumetric flow; thus, heat and moisture transfer are ignored. Fourth, the capital cost of ventilation equipment is not included in the formulation.

Jan 1, 1989

By David M. Mchaina, Allan E. Hall, Siegfried Rohr, J. D. Botsford

Incidents associated with secondary dust ignitions in the mining of massive sulphide ore deposits are of major concern in Canadian underground metal mines. During stope or development blasting, a highly concentrated cloud of sulfide dust is created and may be ignited, resulting a secondary explosion which releases a considerable volume of sulphur dioxide and possibly other sulphur gases. The consequent pressure, heat and toxic gases can result in loss of life, significant damage to equipment and loss of production. These losses are expensive and should be prevented. Explosions of this type have occurred at the H-W mine of Westmin Resources mainly in the development headings and cut and fill stapes. This paper presents the results of a study conducted at the H-W mine to compare dust sampling methods for explosivity tests. The associated benefits and pitfalls of each sampling technique are examined. The investigative work conducted to develop and test preventive and control methods against sulphide dust ignitions is also discussed. The techniques which will minimize the risk of sulphide dust ignition in metalliferous mines are outlined. A review of current preventive and control measures practised at the H-W mine is also presented.

Jan 1, 1989

By S. A. F. Murrell, M. Hong, L. J. Bond

This paper investigates the propagation, interaction and scattering of explosively generated seismic waves with discontinuities in rock masses using numerical models which employ a finite difference method. The models consider 2-D systems and give a full solution including mode-conversion in the mid-frequency scattering region where feature dimensions are of the order of the wavelengths involved. The paper has two parts, the first of which considers wave propagation and diffraction by simulating an explosive body source in a half- space containing a step or a slot on the surface which represent a cliff and a rift valley that has dimensions of the order of the wavelengths in the source. The second part of the paper reports a hybrid technique which combines a perturbation method and the finite difference method for wave propagation again caused by an explosive body source in a 2-D half-space system but with a shallow surface obstacle, such as a smooth ridge, with elastic constants slightly different from bulk material and the slope of this surface topography are shallow. This hybrid technique is also reported being applied to a half-space with a horizontal void or insert.

Jan 1, 1986

By C. B. Manula

The total systems model for underground coal mine planning and design (Master Design Simulator) is a set of computer programs which encompasses a realistic reproduction of the interaction among the geological, environmental, equipment, methods, and the health and safety aspects of the mining plan (Manula, Rivell, and Ramani, 1975). The total system includes nine major subsystems: geology and reserves, methane generator, roof support, subsidence, production, rail haulage, cost, water generator, and ventilation (Fig. 1). The subsystems themselves are divided into two categories based on the frequency with which they are executed: (1) event-oriented subsystems are executed either independently or at discrete intervals when certain specific events occur; and (2) dynamic subsystems function interactively during the production cycle, transferring data to or requesting data from event-oriented subsystems only when specific events occur (Fig. 2). Each subsystem has been tested and validated independently (Ramani, White, and Sutton, 1974; Ramani and Owili-Eger, 1974; Manula, Mozumdar, and Jeng, 1974a; Manula, Mozumdar, and Jeng, 1974b; Manula, et al., 1974; Manula, Rivell, and Sanford, 1974; Ramani, McClay, and Burgos, 1975; Manula and Owili-Eger, 1975; and Ramani, Owili-Eger, and Manula, 1975).

Jan 1, 1979

By J. E. Gebhardt, P. E. Richardson

External control of the electrochemical potential of a bed of mineral particles was investigated by the US Bureau of Mines as a method for controlling particle interaction with organic collectors and, thus, flotation response. A cylindrically shaped, 200- to 300-g bed of sulfide mineral particles was used as the working electrode in an electrochemical-flotation cell. In a pH 9.2 sodium tetraborate buffer, chalcocite floated with ethylxanthate at applied potentials [=] - 0.25 V (versus the standard hydrogen electrode) ; pyrite floated at potentials [=] 0.15 V. Differential separation of a chalcocite-pyrite mineral mixture was obtained by controlling the potential, but the flotation response of the mixture did not follow the behavior predicted from measurements on the individual mineral beds. Chalcocite dissolution products had a deleterious effect on chalcocite-pyrite separation by activating pyrite so that it floated below the potential (~ 0.13 V) at which xanthate is oxidized to dixanthogen. A pyrite bed activated with a copper(II) sulfate solution exhibited a flotation response with a potential dependence between that of chalcocite and pyrite. Pyrite dissolution products had no effect on chalcocite flotation.

Jan 1, 1988

By A. P. Schissler, S. Badr, U. Ozbay, M. D. G. Salamon

Yield pillars are used in deep coal longwall mines, where the depth of cover is so great that a conventional chain pillar design tends to be unsuccessful. The design of yield pillars in deep U.S. mines has been based mainly on experience, pressure arch theories, numerical modeling, and empirical methods that originated in the eastern U.S. coal mines. This paper offers field experience based guidelines for the design of yield pillars through analysis and evaluation of numerous case histories. Case histories of yield pillar applications were collated and critically reviewed to identify trends in data that can be used to formulate the design guidelines. Convincing trends observed in the case of successful case histories indicate that yield pillars should have a width to height ratio in the range from 2.8 to 6.0, a stability factor during development between 0.4 and 0.6, and shed between 94% to 98% of the cover and abutment load, as provided by the Carr-Wilson (1982) method. Also, yield pillars should be utilized in coalmines only where the roof and floor rock have a CMRR rating of 50 or better.

Jan 1, 2003

By Byron V. Knelson

Although the concept of centrifugal recovery has been around in one form or another for many years, it has not been until recent times that it has become clearly understood and properly put into perspective. Simplistic in theory, it is when one begins to examine the variables involved that the true complexities of this matter begin to emerge. In creating an enhanced g force concentrator, there are many aspects both positive and negative that are encountered. Many conditions that are inherent to all conventional 1 g concentrators do not apply to centrifugal concentrators and therefore very few cm factors exist between them. In developing this technology, virtually all principles were proven by method of trial and error. Among the topics discussed are the effect of g force enhancement on particulate behavior with respect to size and shape, density, enrichment, fluidization, and the "teeter principle". In addition, we will discuss the physical effects of g force enhancement with respect to manufacturing in the areas of abrasion, bearings, and vibration. In discussing the recovery parameters of centrifugal concentrators we will draw direct comparisons to various other recovery methods, outlining the conditions most favorable to their application.

Jan 1, 1988

By R. A. Metz, George S. Barnett

The Equis and El Roble projects are located in the Cordillera Occidental of Colombia. The Equis vein deposits have sulfide ore reserves of 95,955 mt (105,770 st) grading 8.51 g/mt (0.248 oz/st) gold and 12.4 g/mt (0.36 oz/st) silver with 0.94 % copper, 1.10% lead and 11.06% zinc. Ore reserves at El Roble are estimated as 1.1 Mmt (1.21 million st) grading 4.89 % copper and 3.39 g/mt (0.099 oz/st) gold. Geologically possible reserves at Equis may total 1.6 Mmt (1.76 million st). El Roble has the potential to contain several tens of millions of metric tons. Exploration included diamond drilling, trenching, channel sampling, rock and soil geochemistry, geophysics and tunneling. Erratic high grade samples affect the weighting of channel and drill core samples. Secondary reconcentration of gold is also a factor. Sample maps and graphs show the distribution of values to be highly variable. Geostatistical treatment of some of these data gave important guidelines for parameters such as sampling densities, drill hole spacing and confidence limits.

Jan 1, 1992

By A. M. Al Taweel, V. K. Kasireddy

Intensive coal cleaning usually necessitates fine grinding of the feed in order to liberate inorganic constituents from the surrounding organic matrix. This however adversely affects the ability of flotation techniques to beneficiate coal. Such difficulties are mainly attributed to the increase of particle population density, the decrease in capture efficiency, and the greater tendency of inorganic components to report with the froth. A three-pronged approach was adopted in an attempt to overcome these difficulties. The aggregative flotation technique was used to reduce particle population densities and improve capture, while unconventional spargers were used to generate a wide range of bubble sizes. Use of unagitated flotation columns allowed for the maintenance of countercurrent flows in both the recovery and washing sections. The combined effect of these parameters enabled the recovery of 92% of the carbon content of coal having an average particle size of 20.5 µ m, while rejecting 85% of the ash. This was accomplished using 2 kg of fuel oil per tonne of coal in a column 2 meters high operating at 2-3 tph/m3, a capacity which compares favourably with conventional cells.

Jan 1, 1989

By Richard B. Berg

Until its closure in 1999, the Antler was North America's only active chlorite mine. The mine is located 25 miles (40 km) southeast of Butte in the Silver Star district - an old Montana gold mining district. In 1975, two prospectors searching for talc re-examined a prospect showing a green, fine-grained mineral resembling talc. Following identification and testing of this magnesian chlorite (variety clinochlore), Cyprus Industrial Minerals began processing ore from this deposit in 1976. Chlorite veins ranging up to 30 feet (9 m) in thickness were formed by essentially complete replacement of Precambrian (Archean?) quartzofeldspathic gneiss. Chlorite replacement was controlled by near- vertical faults. These chlorite veins, typically surrounded by envelopes of sericitic alteration, were formed by the introduction of magnesium-rich hydrothermal fluids, probably related to the Proterozoic talc-forming event in southwestern Montana. Luzenac America, Inc. acquired Cyprus' talc operations in 1992, and continued to mine this deposit until the ore body was mined out in March 1999, and the mine was reclaimed. Total sorted ore production from this deposit is about 250,000 short tons (st) (227,000 metric tons [mt]). Chlorite from this deposit was primarily used in ceramic and paint applications.

Jan 1, 2001

By Paula N. Wilson, W. T. Parry

Carbonaceous and metalliferous black shales occur in close proximity to Mercur-type gold deposits in the southern Oquirrh Mountains, Utah. These shale units, the Manning Canyon shale and the Long Trail shale, have undergone regional hydrothermal alteration. Hydrothermal solutions oxidized organic matter and pyrite, replaced shale by hydrothermal phyllosilicates and silica, and replaced fossils by phyllosilicates. The shales are depleted in organic carbon, calcium carbonate, alkali and alkaline earth metals, and enriched in aluminum and hydrogen. Hydrothermal veins containing illite, chlorite, pyrophyllite, kaolinite, Fe-oxides, and sulfides are present throughout the study area.The illite in veins is an ammonium illite that contains up to 2 % NH4 and is correspondingly depleted in K. Correlations among Au, As, Hg, TI, Sc, and rare earth elements indicate that hydrothermal fluids responsible for formation of ammonium illite were gold bearing. K/Ar ages of illite veins and illite-rich, clay-size separates of altered shales indicate hydrothermal activity occurred in the time range 193 to 122 Ma. Several ages from clay-size separates within the Mercur district cluster near 160 Ma. These ages are coeval with thrust faulting recognized to the north and west.

Jan 1, 1990

By J. W. III Leonard, T. E. Wilson

In addition to coalbed thickness and basic coal quality data, coal drill cores can produce design data for coal beneficiation processes. To generate design data, a coal drill core should be reduced from its original cylindrical form to a particle size distribution and composition that represents (with the possible exception of top size) a raw coal process stream. Reducing a coal core with a crushing machine rarely accomplishes this representation. This investigation employs drop-shatter impact breakage followed by dry batch tumbling in a steel cylinder rotated about its transverse axis to characterize the core in terms of first-order and zeroth-order break- age rate constants, which are indices of the propensity of the coal to degrade during handling. To reduce composition variables and conserve difficult-to-obtain coal cores, the initial drop-shatter and dry-tumbling tests were done with synthetic cores. Then, cores from three coal beds, Illinois No. 6, Upper Freeport and Pocahontas No. 5, were tested. Comparison of a crushed drill core and a drill core prepared by drop-shatter and tumbling indicated that the size distribution and size fraction composition of the drop-shattered/tumbled core more closely resembled the plant feed.

Jan 1, 2000

By K. A. Natarajan, Abha Kumari

In this paper, electroleaching and electrobioleaching of ocean manganese nodules are discussed along with the role of galvanic interactions in bioleaching. Polarization studies using a manganese nodule slurry electrode system indicated that the maximum dissolution of iron and manganese due to electrochemical reduction occurred at negative DC potentials of -600 and -1,400 mV(SCE). Electroleaching and electrobioleaching of ocean manganese nodules in the presence of Acidithiobacillus ferrooxidans and Acidithiobacillus thiooxidans at the above negative applied DC potentials resulted insignificant dissolution of copper, nickel and cobalt in IMH2S04 and in sulfuric acid solution at pH 0.5 and 2.0. Mechanisms involved in electrobioleaching of ocean manganese nodules are discussed. Galvanic leaching of ocean manganese nodules in the presence of externally added pyrite and pyrolusite for enhancement of dissolution was also studied. Various electrochemical and biochemical parameters were optimized, and the electroleaching and galvanic processes thus developed are shown to yield almost complete dissolution of all metal values. This electrobioleaching process developed in the laboratory may be cost effective, energy efficient and environmentally friendly. Keywords: Ocean manganese nodules, Electroleaching, Electrobioleaching, Galvanic interactions

Jan 1, 2003

By Morgan M. Sears

In recent years, one large mine collapse and numerous smaller bump events have resulted in a renewed interest in coal bump research. With numerical modeling, the Energy Release Rate (ERR) calculation quantifies the ?release? of the gravitational potential energy of the rock mass into the environment as mining progresses. This release of energy can occur passively in the form of heat and sound or dynamically in the form of pillar bumps or rock bursts. From the initial application of a calculated energy release rate in the deep hard-rock mines of South Africa, the ERR was found to have a significant correlation with the risk or potential of damaging coal bumps or rock bursts. In this research, an ERR calculation is incorporated into the modern LaModel program to facilitate the analysis for potential coal bumps. Initially the ERR calculations in LaModel are verified using a case study of cut sequences originally modeled using the MULSIM/NL program. Then, the ERR calculations are applied to a bump-prone mine in Southern Appalachia were a number of different pillar recovery cut sequences were used in order to minimize the risk of bumps.

Jan 1, 2009

By Nai-piu Tong, Huaren Dou, Sui-shun Poon, Leslie H. Swann

This paper mainly presents geotechnical aspects of a vehicular underpass that was built across under one of the busiest highway in Hong Kong. The underpass, in a section comprising an upper arch and invert slab, is approximately 6.2 m high at crown and 15 m wide at the base level with 40 m in length. The construction inevitably requires the underpass to be built using tunneling techniques to keep the overheadroad traffic uninterrupted. The paper will describe the design of temporary support for the tunnel excavation that was provided by an arch formed with horizontal pipes, which were installed using the pipe ramming method. The pipes arch was then supported by steel ribs founded on rock or mass concrete. Permanent lining design consideration was also discussed. Results of numerical modeling of the soil-structure interaction using the 2-D computer code FLAC will be presented. The predictions have been compared to the monitored results during construction, which indicates the prediction is in reasonable agreement with the measurement in the light of ground movement. The construction challenges will also be discussed in the paper.

Jan 1, 2003

By Stephen C. Tadolini, Thomas M. Barczak

Western mines have had limited choices for standing support. Wood cribs were often too soft and unstable. The introduction of the Can support in the early 1990’s provided a very effective alternative and remains the dominant form of tailgate support. Water-filled prestressing cells are now used to cap the Can and preload it to provide an active roof load. The Cluster Prop, consisting of three timber wedge props bundled together, provides more capacity than an equivalent sized Can support and improves transport efficiency. However, the Cluster Prop is less stable and does not maintain a consistent load throughout its loading profile. Pumpable roof supports are another alternative support, but they have not been proven in high deformation environments. Alternative supports are also being used in longwall recovery operations. These include the Rocprop and Omni Prop, which are extendable support systems. They also provide active roof loading which can be beneficial in these applications. Specialty supports, such as the Sand Prop and Spider Prop, have also been introduced into western longwall operations. This paper compares the performance characteristics of these various support systems.

Jan 1, 2005

By Richard R. Klimpel

The first part of this two-part article summarizes the result of a 10-year, multimillion dollar research and plant testing program involving operating mining companies throughout the world. The program was aimed at developing a better understanding of the influence of slurry rheology on the performance of mineral/coal grinding circuits. Part 1 presents the basic concepts identified, including typical laboratory results showing the influence on grinding behavior of controlled changes in percent solids, particle size, slurry temperature, and viscosity control chemicals. Part 2, in January, will illustrate typical industrial scale test results on both open and closed circuits using the concepts presented in Part 1. Special emphasis will be placed on identifying some industrial operating implications of controlling rheology by different methods. Introduction The industrial-scale practice of wet grinding minerals and coals in tumbling media mills (ball, rod, and pebble) is well-known to operating mineral processing engineers and plant design personnel. Over the last century, considerable experience and data has been collected, correlated, and put into rules or guidelines for successful circuit design and operation. Typical factors influencing circuit throughput that have been analyzed in some depth are mill diameter and length, grinding media size and loading, feed and product particle size, mill power use, and the interaction of classifier type and efficiency with the grinding device. Until recently, knowledge of slurry rheology influence on large scale grinding circuit performance has been limited. Even today slurry rheology effects are essentially ignored in plant design. Indeed, in most plant operations only qualitative guidelines are followed. For example, too thick a slurry causes throughput losses and too thin a slurry uses excessive water or causes water handling problems such as pumping limitations. A search of published literature and numerous discussions with experienced plant operating personnel has shown an awareness that changing the percent of solids in the mill influences grinding or the response of grinding involving fine feeds versus coarse feeds. However, most the work identified relative to rheology effects in mineral/coal grinding was qualitative. Even in the few studies that were more quantitative, the scope of the work was usually too narrow to draw general conclusions. Also, the detailed rheological characterization tied to observed breakage and the combination of laboratory to industrial-scale testing that is required to develop a broad understanding of any factor influencing grinding had not been done. In this environment, this detailed program began in 1967. It had three distinct goals: • To develop a basic understanding of slurry rheology influence on breakage characteristics in the laboratory. • To test general rheology concepts developed in the lab on the industrial scale. • To develop a commercially viable chemical approach to controlling slurry rheology that would lead to the use of chemical grinding additives or grinding aids. The goals have been met, at least in part, this year. A small fraction of the extensive laboratory and plant results have been made available in published literature. Much of the research program dealt with identifying and developing chemical grinding aids, but a lot of useful and practical information on percent solids, particle size, and slurry temperature influence on grinding circuit performance also was collected. The purpose of this article is to present a clear and concise sum¬mary of the results of the entire program. The importance of rhe¬ology control in grinding circuit behavior will be demonstrated using a series of carefully per¬formed industrial scale experi¬ments (both open and closed cir¬cuits) that have not yet been reported in general literature. Mining companies participating in the test program on a proprie¬tary basis have found the informa¬tion in this article to be useful in better understanding and improv¬ing circuit performance. Viscosity Effects in Grinding Studies In this program&apos;s previous labo¬ratory studies, it was shown that there is a consistent pattern of change in specific rates of break¬age of both mineral and coal slur¬ries as slurry fluidity changes. Using the net production rate of material smaller than some spec¬ified size-for example, kg/hr of <75 µm (200 mesh)-as an index of mill production in a standard batch mill test (with a given feed material, feed size, mill, and mill conditions such as constant time of grind), the following facts were established (see Fig. 1): • The normal range of low den¬sity, low viscosity slurry (region A) gave no variation in mill produc¬tion. The measured rates of break¬age exhibited normal first order grinding. • Grinding of a somewhat higher viscosity slurry (region B) could give increased production. The higher viscosity was obtained

Oct 12, 1982

By J. Lovering, H. Wilhelm, P. Siewert

Introduction The Afton property is located 290 km (180 miles) by air east-northeast from Vancouver and 14 km (8.7 miles) west of Kamloops, a city of 60,000 people, in south central British Columbia, Canada. The mine is adjacent to the Trans-Canada Highway at an elevation of 670 m (2198 ft) above sea level. The ore body is a porphyry copper deposit that has undergone supergene alteration. The major economic minerals in the supergene zone are native copper and chalcocite with chalcopyrite and bornite in the hypergene areas. The grade is 1% with an overall copper distribution - 70% native, 25% chalcocite, and 5% chalcopyrite with bornite and covellite. The ore also contains important but variable amounts of gold and silver. The mill was designed to treat 6350 t/d (7000 stpd). Semiautogenous grinding was selected to minimize capital cost and because of the expected high clay content of the ore, which would have caused problems in a conventional crushing and screening plant. Test work indicated that a recovery of 87% was possible in a circuit incorporating both flotation and gravity separation. Flowsheet Run-of-mine ore is crushed in a 1.06 x 1.65-m (3.5 x 5.4-ft) Allis Chalmers gyratory crusher set at 228.6 mm (9 in.), closed side setting. The surge pocket, below the crusher, is emptied by a Hydrastroke feeder onto number one conveyor, which discharges onto a 180,000-t (198,416-st) coarse ore stockpile. Six Hydrastroke feeders on two conveyors withdraw the crushed material from the bottom of the pile. These two conveyors, in turn, discharge onto the belt feeding the semiautogenous mill. The live storage in the stockpile is approximately 22,000 t (24,250 st), sufficient for three days&apos; mill feed. Primary grinding is accomplished in an 8.5-m (28-ft) diam by 3.7-m (12-ft) long Koppers (Hardinge Cascade) mill (Fig. 1) containing a 10% ball charge and driven by a 4000-kW dc variable speed motor. The mill dis¬charge is pumped by a 10 x 12 G.I.W. pump to a 1.22 x 4.88-m (4 x 16-ft) stationary screen sloped at 20°. Screen oversize returns to the semiautogenous mill (SAM), and the undersize flows by gravity to the ball mill discharge pump box. Secondary grinding is performed in a 5-m (16.4-ft) diam by 8.84-m (29-ft) Koppers overflow ball mill driven by a 3430-kW synchronous motor through an air clutch. The mill is in closed circuit with a Krebs Cyclopac containing 10 635-mm (25-in.) cyclones and the cyclone overflow, at 35% solids and 65% to 70% -200 mesh, is flotation feed. In order to limit the buildup of native copper, circulating in the secondary grinding circuit, a portion of the underflow from the cyclones is processed in a circuit containing screens, cyclones, and shaking tables to produce a finished metallic copper concentrate. Primary mill variable speed drive The overall waste to ore ratio at Afton was 4.5:1. The mining was to be done with only three shovels, which meant that it was highly unlikely that more than one of them would be in ore at any one time. The resulting inability to blend the mill feed made it impossible to prevent wide swings in the grade and grindability. The variable speed do drive motor installed on the semiautogenous mill was selected because of the extreme variability of the Afton ore body. This variability has persisted throughout the lifetime of the mine. There are times, however, when due to ore conditions, the mill is operated at full speed (78% of critical) for extended periods of several shifts duration. There are other times when the mill speed may be changed several times in a 12-hour shift due to changing ore conditions. When ore is processed that contains a fairly large proportion of fine native copper, the primary mill speed and, consequently, the tonnage may be reduced to improve the secondary grind and to maintain an acceptable grind and recovery. High clay ores require less mill speed and more dilute grinding densities. In the latter case, the slower primary mill speed also helps to minimize damage to the mill liners. Approximately 57% of the time the mill operates between 90% and 100% of full speed or between 71% and 78% of critical. The variable speed is also used for inching during mill relines.

Jan 1, 1987

By N. J. Peberdy, C. N. Moreton, K. C. Garner

Introduction During the past decade a major objective of the process industry has been to use digital computer technology to improve plant operating efficiencies. This objective implied some form of optimization, a concept that has various interpretations depending on the view of the prospective user. For the purpose of this paper, optimization of a process plant is defined as the establishment and setting of plant operating conditions that maximize some mathematical yield function, i.e. maximum profit, minimum residue, etc. Analysis of these objectives and the available design and implementation techniques led to the conclusion that digital computer and optimization techniques are not the stumbling blocks, but rather the development and derivation of the mathematical models of the unit operations and process plants to be optimized. Such models should not only describe the optimized (steady-state) objective, but also how one steers to this state (control algorithm). Due to the multidisciplinary nature of the skills associated with the design and operation of process plants, the development of suitable models by a single discipline, such as the process control engineer, was found to be not only difficult but often impossible, due to budget and human resource limitations. To over-come these limitations, a computer aided design (CAD) tool has been developed. It aims to provide a productivity tool to the various disciplines, at the same time coordinating the technical input from each. The system described is but the starting point in an evolutionary development of a tool that, with use, is becoming more efficient and cost effective to use. Development has become an application engineering activity rather than the preserve of the computer specialist. Project phasing The development of a mathematical description of a process plant requires coordination of information from conceptual design to operation management. The activities required to build and operate a process plant are divided into four basic chronological activities or phases. These activities are often undertaken by different organizations and disciplines. As a result, continuity is often lost with the resultant loss of critical design data. The major activities are considered to be: conceptual and flowsheeting; detailing around the P & ID; building and commissioning; and plant operation. The CAD system described provides a design tool to be used for each of these activities, as well as providing continuity between the activities and the disciplines involved. The heart of the system is the dynamic simulation of the flowsheet. Each of the activities will be discussed, leading to two simple examples that demonstrate the use of the simulator. Figure 1 shows a schematic format of the various activities and the path followed by the dynamic flowsheet simulator in the life of a project. Flowsheet development The prime requirements in the design and develop¬ment of a process flowsheet are • selection of the correct unit operations to achieve the most economic (capital and operating) beneficiation of the specified reserve ; • the sizing of the unit operations to achieve the desired results, as a function of the projected feed rates etc., to handle the time related (dynamics) of the process; and • the production of a set of engineering documents showing the drawn and labeled flowsheet with an equipment list and process specification for each of the unit operations. The question may well be asked at this stage why dynamic flowsheet simulation should be considered when steady state modeling has been found to be adequate to date. With the increases encountered in the cost of capital, one often cannot afford the luxury of designing around the compounding worst case technique. Further, a more accurate design of the control surges can be achieved. No information is lost in that the steady state solution is in fact a subset of the dynamic model. In generalized state space modeling, the differential equations describing the process dynamics are illustrated in the following matrix notation: XDOT=A.X+B.U(1) Y =C.X+D.U(2) where XDOT describes the set of first order derivatives of the system state Vector, and X- is the system state Vector; A - is the system matrix operator which in the general nonlinear case is both a function of X and time ; U- is the process input vector; B - is the input mapping matrix; Y - is the set of observations; C - is the output mapping matrix which maps X - onto Y; and D- maps the input onto the observations. Thus, by time integration of the system dynamic equations, described in (1), the dynamic trajectory away from any set of initial conditions can be deter¬mined. Further, by finding the conditions at which XDOT = 0, the steady state solution can be determined.

Jan 1, 1987

By M. Measures, W. A. Napier, C. Pomroy, J. M. Jardine

INTRODUCTION In most uranium mines the concentration of uranium in the ore is low enough that inhalation of dust does not present a radiological problem. Conversely in the mills there is a potential inhalation hazard from inhaled uranium in concentrated form, namely "yellowcake" and other compounds intermediate between the ore and the final product of the mill. It was generally assumed that the solubility of yellowcake placed it in the category of Class D materials. This plus the fact that the uranium is non-enriched led health physicists to assume that the potential health hazards were a result of the uranium&apos;s chemical properties rather than its radiological properties. Bioassay programs have thus been limited to routine urine sampling at a weekly or monthly frequency. In most cases when a high reading was obtained, subsequent sampling within one or two days showed a rapid return to low levels, therefore tending to confirm the class D nature of the material. Two developments over the last few years have however suggested that modifications to this simple bioassay program may be necessary and practical. Firstly, a number of studies on the solubility of uranium mill products, from ore to yellow cake, in simulated lung fluid (Eidson A.F., Mewhinney J.A., 1980) have shown that some of the material is not class D, but class W or even Y. This raised the possibility that, while the urine levels rapidly decreased, indicating rapid excretion of the class D component, the W and Y components remained in the lung. Thus an unsuspected lung burden could have been accumulated. The second development, the ready availability of [in vivo] counting facilities with which to make direct measurements of uranium lung burdens, offered a means of confirming the hypothesis of unsuspected lung burdens. A study was therefore initiated to correlate lung burdens with urinary excretion. METHOD A total of 79 subjects from 3 uranium mills participated in the study. They were brought to the Ottawa laboratories of the Radiation Protection Bureau in groups of 4 or 6, over a period of several months. In Ottawa, [in vivo] thorax measurements were performed, and a urine sample collected. Each subject also brought with him a urine sample that had been collected the previous day. All samples were split, for analysis by both R.P.B. and the mill laboratory. [Concentrations of Uranium Dust] Results obtained by mills A, B and C for concentrations of uranium in air at various locations are given in table 1. Samples taken at mill B were obtained using continuous sampling with a midget impinger over a six hour period. Samples taken at mills A and C were 5-minute gravimetric samples. [Subject Selection] The subjects were selected to represent a wide range of work history in terms of both length of employment and type and location of job, as shown in table 2.

Jan 1, 1981