Search Documents

By P. Hanniala, S. Harkki, B. Andersson

"The Outokumpu Flash Smelting Process was put into commercial operation in 1949. The original process was based on the use of preheated air. The adoption of oxygen enrichment of the flash smelting process air in 1971 was a crucial improvement, especially for smelter capacity. Oxygen use according to the Outokumpu method means selection of the most suitable oxygen content for the process air on the basis of the concentrate composition, capacity and matte grade. The furnace is not bound to only one fixed degree of process air oxygen enrichment; the oxygen content may vary from 21% to 100%.The degree of oxygen enrichment and the matte grade are the most important process parameters in reducing the operating costs of the whole system, including smelting, converting and acid production. Raising the matte grade allows more efficient use of the reaction heat in the smelting stage. Sulphur is recovered in a high-strength, constant-flow gas stream. The blowing time in the converters is reduced considerably.Oxygen enrichment and high matte grade improve the energy efficiency of the plant complex as long as the surplus heat from the system is not wasted. When the operating costs and investment expenditure of the total plant complex are taken into account, a reasonably high degree of oxygen enrichment (35-80°/0) is the best alternative in practice. Existing operating smelters have achieved single-stream thoughputs of over one million t/yr of concentrates.HistoryLocal capital and operating costs as well as lower energy consumption per unit of metal produced are now the crucial factors in determining the choice of process. In the late I940s, the same reasons prompted Outokumpu to develop the original flash smelting process in Finland. Between 1900 and 1940, very few technical changes in pyrometallurgical copper production took place. However, the commercial use of the flash smelting technique began a new era in metal extraction from sulphide concentrates. The characteristic feature of the method is that the concentrate is smelted in suspension: this offers the most favourable conditions for heat and mass transfer, and thus high reaction velocities. Outokumpu was the first to exploit the known advantages of flash smelting in a commercial installation at Harjavalta in 1949."

Jan 1, 1982

By Hans Svensson

INTRODUCTION The Alimak method of raising (Fig. 1) was intro¬duced in 1957. Because of its flexibility, economy, and speed, it soon became the most used system of raising in the world. With the same unit, raises of various areas, inclinations, and lengths can be driven. Air-motor¬-driven raise climbers normally are used in mines with short raises (Fig. 2). For longer raises electric or diesel-¬hydraulic drives are used. The Alimak raise climber is designed entirely in ac¬cordance with the instructions of the Swedish Board of Industrial Safety (Arbetarskyddsstyrelsen). There are good margins of safety concerning the breaking strength of the material and the raise climber comprises safety devices and features that practically eliminate the risks of accidents. The drive gear of air-driven climbers is equipped with an air-operated brake which is automatically activated when air for the motors is shut off. There is a safety device which activates automatically at overspeed and a speed-regulated brake for descent by gravity, by which the raise climber can be taken down to the bottom of the raise in case of cutoff of the air supply. The raise climber climbs along a pin rack welded to a guide rail which also comprises pipes for air and water. The guide. rail can be extended by using 1- or 2-m (3.2¬or 6.4-ft) sections as the driving progresses. Each guide rail section is bolted to the rock wall by special expan¬sion bolts of Alimak design. Men ride up to the face in the cage, traveling com¬fortably and easily while material is transported on the platform. There is no dangerous climbing up ladders. The men are well protected during ascent and arrive in good condition at the face. The greatest advantages of raise climber systems are flexibility and versatility. The equipment provides a comfortable working position in vertical as well as in¬clined raises of any area and length. The direction may

Jan 1, 1982

By Yael Miron

ANFO blasting agents may undergo an exothermic reaction when they contact warm ore bodies that contain pyrite and its weathering products. The reaction is very fast at temperatures above 80° C. Urea, when present in sufficient amounts, inhibits this reaction at temperatures below 180° C. But urea-ammonium nitrate mixtures are very hygroscopic and, in certain climates, may be difficult to handle in the field. Reaction-inhibiting characteristics of three compounds were reported by the Bureau of Mines in 1979. Recently, the Bureau investigated the reaction-inhibiting qualities of five additional compounds. Mixtures of ANFO, pyrite, weathering product, and prospective inhibitors were slowly heated, under conditions simulating a hot borehole, to determine prospective inhibitor capability at temperatures up to ~175° C. Small-scale thermal tests of the same mixtures were used for verification and clarification of results. Two organic compounds, phthalimide and xanthine, similar in structure to urea, failed to inhibit the reaction. Zinc oxide and calcium carbonate were found inferior to urea, though zinc oxide raised the onset of reaction from 80° C to about 140° C. Only magnesium oxide was found to be similar to urea in its inhibitory ability at temperatures below 175° C. Based on results from analysis of quenched residues, the use of more than 5 wt-pct magnesium oxide is recommended.

Jan 1, 1982

By A. J. Sinclair, N. Champigny

"The Cinola gold deposit in the northern Queen Charlotte Islands, British Columbia, is in a clastic sequence consisting of a lower shale of the Late Cretaceous Haida Formation and an overlying interbedded sequence of pebble conglomerate and coarse-grained sandstone of the Middle Miocene Skonun Formation. Both formations are intruded by a stock and dykes of porphyritic rhyolite. Two K-Ar model ages indicate mineralization and probably porphyritic-rhyolite intrusion at about 14 Ma (middle Miocene). A splay of the Sandspit fault system constitutes the footwall on the west of the deposit.The Cinola deposit can be classed as Carlin-type based on features such as (1) small particle size for gold, (2) Tertiary age of mineralization, (3) element abundances (e.g. high Hg), (4) dominantly argillic alteration, (5) a structural setting associated with major faults, (6) porosity of the host rock and (7) spatial and possibly genetic association with felsic intrusions. Gold is widespread and occurs mainly as grains of< 0.5 11-m in silicified sedimentary rock, and in quartz veins. Locally, coarse > 100 11-m particles of native gold occur in quartz veins, especially in strongly brecciated porphyritic-rhyolite. Ore minerals are mainly pyrite and marcasite, but include small amounts of chalcopyrite, sphalerite, galena, pyrrhotite, cinnabar, tiemannite (HgSe), rutile, magnetite, hematite and limonite in addition to native gold and electrum. No silver minerals have been found; silver was found in gold particles in amounts varying from 6.2 to 24.2 weight per cent. Alteration products are sericite, illite, kaolinite and chlorite, with abundant quartz of several stages.The host rocks, specifically the Skonun Formation, formed as an alluvial plain facies in a braided river system discharging into a marine basin in early Middle Miocene time. During the Middle Miocene, this sequence was intruded by a rhyolitic stock. The highly porous and permeable Miocene clastic sequence apparently provided an optimum setting for the development of a large geothermal system, the energy for which probably derived from the rhyolite stock. Mineral deposition followed a well-defined paragenesis, which, from oldest to youngest, is: (1) precipitation of iron sulphide minerals, and early quartz; (2) several stqges of quartz veins, with deposition of sphalerite succeeded by galena, chalcopyrite and visible gold. Micron-size gold was precipitated throughout these two stages of mineral deposition. Argillitization of the host rocks probably in part coincided with the mineralization and continued during cooling of the geothermal cell.In-situ reserves have been estimated at 41.1 million metric tons averaging 1.85 g Au/tonne using a cutoff grade of0.86 g Au/ tonne."

Jan 1, 1982

By Loren R. Anderson

This is an Appendix of Volume I. It contains the results of the electric cone penetrometer soundings from the first phase of the testing program.

Jan 1, 1982

By Loren R. Anderson

A probabilistic slope stability computer model was developed at Utah State University and independently evaluated by CH2M Hill Consulting Engineers as a design and analysis tool. Volume I of this report describes the model and Volumes III and IV are users manuals for the computer programs. This volume (Volume II) constitutes the report by the CH2M Hill. It includes a description of a field investigation and stability analysis of an existing tailings dam. Based on the results of the analysis, the tailings dam was evaluated for redesign using probabilistic considerations. Conclusions are made regarding the usefulness of the model as a design tool.

Jan 1, 1982

By G. F. Skala

Additional testing of two prototype instruments for the discrete detec­tion of submicrometer particles in metal and nonmetal underground mines is described. The detectors are designed to utilize properties of submicrometer particles to distinguish between those produced by fires from mine ambient particles primarily resulting from diesel engine exhaust. One instrument is based on the ability to pyrolyze fire produced parti­cles, but not diesel exhaust particles. An alarm is produced when the ratio of pyrolyzed to unpyrolyzed particles exceeds a preset value. The other detector utilizes a continuous flow ion chamber preceded by a denuder to remove charged particles. Discrimination against diesel produced submicrometer particles results because the diesel particles are naturally more highly charged than fire produced particles. In-mine testing was conducted over a period of several months. Both in­struments exhibited an ability to discriminate against diesel exhaust. However, problems with the recording equipment, and with the instruments themselves, due to large ambient dust particles, prevented enough data from being obtained to choose between the two methods. Recommendations to correct the problems are made.

Jan 1, 1982

By John A. Riddle

Some of the pitfalls and problems in selecting and applying corrosion-resistant materials in uranium plants are discussed, together with some suggested solutions. Includes some cost data.

Jan 1, 1982

By T. P. Meloy

Improvement of grade and recovery requires the use of optimum processing circuits. Critical to an optimum circuit is the set point for separation. With the introduction of the concept of circuit analysis (Meloy, 1982), it became possible to determine optimum circuits for a given task. No analysis, however, has yet been made as to the grade requirements of the larger circuits. Williams and Meloy (1982), reviewed the literature and concluded that, within the operating range, most unit operations involved in particle separation behave linearly. For the purposes of this analysis, linearity will be assumed. Grade, G, is defined as the weight of the coal mineral, C, that appears in the product, divided by the weight of the coal mineral plus the weight of the gangue mineral or waste mineral, Q, which appear in the product.

Jan 1, 1982

By R. W. Titshall

WHAT IS SAMPLING? We are all involved in sampling almost every day of our lives, by tasting, feeling, or smelling. Most raw materials, food products or manufactured items are sampled several times at various stages of the production process before a commodity of suitable quality can be produced. Data obtained from sampling is used in establishing price, quality control, allocating production costs, or overall process efficiency. The area of sampling outlined in this article is the collection of a small quantity of material from large continuous bulk flows of solids and slurries. The concern in obtaining the sample is that the material obtained is truly representative of the main flow in every respect. The methods and equipment outlined are for the collection of a manageable sample. The sample has to be analyzed in a laboratory to determine its characteristics and compliance to set standards. SELECTING A SAMPLING SYSTEM When selecting a sampling system, it is essential to define the requirements, as to what the actual sample obtained will be used for and what degree of precision is required. Determine whether or not the system has to comply with any set standards such as A.S.T.M. (American Society for Testing and Materials) which have established methods for collection of a gross sample of coal. Sampling which is done for in-plant or process control is not generally governed by set procedures, but similar care should be taken during the selection process so that when the equipment is put into service it will operate reliably and produce a sample which is truly representative of the total material lot or consignment.

Jan 1, 1982

By Charles E. Semler

The possible utility of selected compositions in the systems SrO-ZrO2-MgO and SrO-ZrO2-Dolomite for commercial refractory products was confirmed in this study. Specifically the compositions of interest are located in the sub-systems SrZrO3-ZrO2-MgO and SrZrO3-ZrO2-Dolomite. Both sub-systems have regions that are not affected by hydration, but the sub-system SrZrO3-ZrO2-MgO has a larger candidate compositional region. Solid state phase relations were determined for the subject systems over the temperature range 1600°C to 1300°C, with compatibility diagrams presented for each. The systems appear to show good refractory character with no apparent low melting compositions. As the compositions of commercial interest lie mostly in the higher ZrO2 portion of each system, information is presented regarding the firing treatment required to develop stabilized cubic ZrO2. A list of suggestions for further study of these systems, to further delineate the characteristics required for possible commercial utilization, is presented.

Jan 1, 1982

By S. P. Girrens

A coupled finite element geomechanical- hydrology code is currently under development for application to the problem of predicting groundwater disturbances associated with mine subsidence. The structural-fluid coupling is addressed by calculating the subsided mine geometry, with emphasis placed on determining the strata disturbance and locating damaged regions, for input into a hydrology code, which determines localized volume flow rates and aquifer fluctuations. Benefits from coupling will be best realized when field measurements, an additional aspect of the study concurrent with analytical investigations, indicating the relationship between increasing rock strain and increasing permeability are incorporated into hydraulic material descriptions. Hydrologic and struc- tural calculations are presented to demonstrate computational capabilities applicable to mine subsidence

Jan 1, 1982

By H. S. Chiang, D. F. Lu, S. S. Peng

INTRODUCTION The most important element in a successful lingual mining is a good roof control. The modern longwall mining employs hydraulic powered supports for roof control at the face area. The application of hydrau¬lic powered support requires the knowledge of over¬burden strata behavior for proper selection of sup¬port type and capacity. Failure to do so could lead so serious loss. There are several methods available for determining the required support capacity (1-3). While these methods are simple for application, they do not include the complicated roof behavior observed in longwall mining. As research progresses and operational experience accumulates (4,5), the concept about the designing and selection of powered support improves. The design of a longwall powered support consists of three major phases: 1. structural integrity and stability of the powered support, 2. external loadings induced by the movements of the overburden strata, and 3. interaction between the support, roof and floor. Phase 1 involves structural analysis (5) and full-sized testing (6) of the supports. Its validity is limited by the accuracy of the assumed external loading because of the uncertainty about the actual loading underground. The third phase includes the reaction of the support and the floor to the movements of the overburden strata and vice versa. Among the three phases, the second phase concer¬ning the external loading seems to be the least known because of the complicated behavior of the roof strata. There are many unresolved problems. For example, does the main roof break periodically and cause periodic roof weighting in the face area? If so, are there any rules governing its behavior? How does the roof load on the support canopy! Finally, how can one determine the required support capacity and select a proper type of support to meet a certain roof behavior? In order to answer those questions, underground instrumentation and observations were performed at 4 longwall panels in 3 separate mines for the past two years. This paper summarizes the current findings. PANEL LAYOUTS AND EQUIPMENT EMPLOYED The three mines selected are all located in West Virginia; two in northern and one in southern West Virginia. As shown in Table 1, seam conditions (i.e. seam, depth and thickness) and panel layouts are different among the three mines. The most significant difference in equipment is the face powered supports. Three mines used three different types of shield; 2-leg caliper, 2-leg lemniscate, and 4-leg lemniscate chock-shield. (Fig. 1) UNDERGROUND INSTRUMENTATION AND OBSERVATION PROGRAM Two events were instrumented in each observed longwall face: one was the hydraulic pressure (resistance) of the powered supports and the other was the canopy load distributions. In addition, the gob caving conditions were visually observed and recorded. Leg and Support Resistances One or two automatic Weksler Pressure Recorders were installed at the designated shield support,. In most cases, the daily charts were used to record the pressure variations in both the front or the rear legs (for the 4-leg shield), or in both the leg and the fore-pole ram (for the 2-leg shield). The recorded pressure w a s then converted to load or resistance by multiplying it by the cross-sectional area of the hydraulic leg or canopy ram piston. Fig. 2 shows the typical pressure-recorded charts for the 4-leg and 2-leg shields in a 23-24 hour period. The support resistance is the summation of the resistance in each of all the legs for that support. Generally, the resistance of the fore-pole ram will not be considered in determining the capacity of the support because of its rather small vertical compo¬nent force at the tip of the fore-pole. Canopy Load Distribution External load distribution on the canopy as exer¬ted by the roof was monitored. The measurements employed 12-14 pieces of pressure cells (6-inch square) that were uniformly arranged in two rows on the canopy. After support setting, the pressure changes in the cells were monitored at various stages of the mining (supporting) cycle while the support leg pressures were recorded continuously by the pressure recorders. Based on the calibration chara¬cteristics of each pressure cell as performed in the laboratory before and after each underground test, the cell pressures were converted to actual loadings. From these load measurements the canopy load distri¬butions and the relations between measured canopy loadings and support leg resistances were determined. Accordingly, the supporting efficiency of the shield support can be determined.

Jan 1, 1982

Develop a lightweight, portable device for use in mines for non-destructive testing of resin grouted roof bolt bond integrity. Approach A unique, intrinsically safe, electronic instrument has been developed that determines the holding quality of a fully grouted roof bolt by testing the integrity of the resin bond to both the bolt and to the surrounding rock. How It Works A measured pulse of energy Is sent Into the bolt by a piezoelectric transducer, and the instrument listens for reflected energy. Where the bolt is properly bonded, the energy passes from the bolt, through the resin, and out into the rock mass. Where proper bending does not exist, a portion of the energy Is reflected back along the bolt. The Instrument measures the reflected energy and compares it to the original pulse. The amount

Jan 1, 1982

By James C. Fowler

The use of ground-probing radar as a reflection mapping tool is gaining acceptance in subsurface investigations. This tool may be used to obtain near-surface information useful to geotechnical engineers. Along with the lightweight field equipment now available, radar is being used to obtain information that may impact the safety and production of mining, tunnelling, and excavating operations.

Jan 1, 1982

By M. V. K. Reddy

A differential flotation process for recovery of molybdenite from chalcopyrite concentrates produced at Rakha Copper Concentrator, Bihar, India has been developed. The flowsheet is simple, and relatively inexpensive and yields a concentrate assaying not less than 45% Mo at a recovery of minimum 65%, starting from chalcopyrite concentrate assaying 0.4% Mo. The consumption of the main depressant used, sodium sulphide is not more than 25 kg/t of copper concentrate processed without steaming; and with steaming the consumption was brought dorm to less than 12 kg/t of concentrate. Further reduction is possible by optimization of steaming parameters. Based on laboratory experimental results, further tests were conducted on concentrator site and process design parameters were established for design of commercial plant.

Jan 1, 1982

By E. H. Kurt

INTRODUCTION The simplest way to drill a hole into rock is to strike a steel chisel or drill bit with a hammer. Early miners used this elementary hand technique so successfully in "single jacking" and "double jacking" that the first mechanical rock-drill designers sought to duplicate it. However, in the early developments, the designers were forced to retreat to a construction known as the "piston drill," wherein the entire drilling element is tied to the piston and reciprocates with it. It took nearly 50 years to devise a method of divorcing the two elements and to achieve the original hammer principle used in hand drilling. The Mont Cenis tunnel, drilled through the French Alps in 1861, usually is considered the birthplace of the mechanical rock drill. The Hoosac tunnel in Massa¬chusetts was drilled at about the same time, and the success of these two ventures paved the way for innova¬tions that produced the Burleigh, Ingersoll, Sergeant, and Waugh piston drills between 1870 and the turn of the century. In 1897, George Leyner of Colorado introduced what has been considered the most significant development in rock-drill history. He devised the hollow drill steel for water flushing; when combined with his free-piston hammer drill, this became the first lightweight and dust¬free underground machine. In addition, Leyner&apos;s drills introduced improvements such as automatic rifle-bar rotation of the drill steel and chuck, automatic lubrica¬tion, and an enclosed throttle control. During the 1920s and 1930s, automatic feeds, centralizers, sliding cones, and the automatic water back head were developed, laying the groundwork for all pneumatic underground drills currently in use. Since these pioneering efforts, rock-drill development has been concerned primarily with refining the designs and improving the metallurgy to make faster, lighter, and more dependable machines. Fig. 1 illustrates the progress in drilling speed during the first 125 years of rock-drill development; that progress is a credit to the persistence and inventiveness of the designers in the rock-drilling industry. ROCK-DRILL CLASSIFICATIONS To meet the variety of conditions encountered in rock drilling, several distinct types of drills have been de¬veloped. In general, rock drills may be classified as either hand-held or mounted, with the hand-held ma¬chines including the jackhammer or sinker, the jackdrill or jackleg, and the stoper. The mounted drills are com¬monly known as "drifters." Table 1 shows that each type of drill is available in several sizes from different manufacturers. Fig. 2 illustrates a typical rock drill, showing the principal components. The jackhammer or sinker, shown in Fig. 3, is used primarily for general mine utility work such as drilling anchor holes (short vertical holes to bolt or anchor machinery), pin holes (short, usually horizontal holes to fasten sheaves, etc. to side walls), popholes (for blasting large boulders), and similar applications. They also are used for shaft sinking. Jackhammers are classified ac¬cording to weight, and they range from 7 to 30 kg (15 to 65 lb). The rock drill originally known as a jackleg was made by clamping a pneumatic cylinder or leg to a jack¬hammer, both to support the weight of the machine and to feed the tool forward in horizontal or uphole work. The more modern jackdrill refined this concept to make the hinged pivot for the leg integral with the drill cylinder and to group all drill and leg controls in the drill back head for convenience. Fig. 4 illustrates a typical jackdrill. Jackdrills are classified according to cylinder bore size, ranging from 60 to 83 mm (2.375 to 3.25 in.). Because of their light weight and versatility, jack drills are very effective in small drifts, small tunnel head¬ings, and stoping. The larger machines are applied to hardrock formations and in applications where drilling speed is a primary consideration. In soft formations or

Jan 1, 1982

By Parviz Montazer, Robert King, William Ubbes, Gideon Chitombo

This paper describes the results of a detailed study that was undertaken to define the spatial distribution of permeability within a five meter thick envelope around the CSM/ONWI room. Detailed fracture orientations and distribution were determined through mapping of the walls of the opening, logging of oriented core, and visual examination of borehole walls with a borescope and a T.V. camera. Over 3,500 measurements were made from which the three dimensional distribution of fractures was determined. Two principally different testing equipment designs and various methods of air injection tests were used in all boreholes to identify and characterize conductive fractures. Through analizing the distribution of permeability and fracture index around the room a zone of blast damage of approximately .5 m thickness was delineated. The effect of stress modification seems to extend farther into the host rock to approximately 3.5 m depth.

Jan 1, 1982

By B. Ehgartner, M. Karmis, C. Haycocks

INTRODUCTION Interaction effects between mining operations developed on different levels of contiguously placed bedded deposits are a major problem in many parts of the United States. Efficient design criteria, derived from careful analysis of field studies as well as theoretical and model analysis, have shown that many problems that may result from seam interaction can be avoided, minimized, or on occasion used with beneficial results. Selection of seam mining sequence is rarely based only on ground control considerations, but rather on factors such as seam ownership, characteristics and accessibility. Therefore all possible mining sequences and methods must be considered in the development of multi-seam design criteria. Increased utilization of the longwall method of mining creates special problems in multi-seam mining due to the inevitability of caving and, therefore, subsidence. Considerable valuable experience has been gained in this direction from British mining operations where interaction problems are extremely common and the longwall method is the primary method used. However, limitations persist in the application of much of this information to U.S. conditions due to variation in layouts between advancing and retreating longwall systems and the inherent difference in the geologic environment. Considerable field data is available for U.S. room and pillar operations, and where successful pillaring operations have been carried out close parallels may be drawn with ground control mechanisms in longwall operations. VARIABLES AFFECTING INTERACTION Analysis of ground control multi-seam problems shows that factors contributing to interaction may be classified into variables that are fixed by the geo- logic environment and those that depend on engineering design. These may be described as follows (Haycocks et al, 1982): Fixed: (i) Depth*, (ii) Innerburden thickness*, (iii) Innerburden physical characteristics*, (iv) Stress field, (v) Seam thickness, (vi) Coal characteristics, (vii) Completed mining operations either above or below the seam to ve mined, and (viii) Age of workings. Mining: (i) Height, (ii) Dimensions and geometry*, (iii) Method*, and (iv) Spatial location of entries* *These variables are regarded as being critical effects. Few attempts have been made to quantify the variables controlling interaction. Dunham&apos;s (1978) linear regression analysis, combined with the National Coal Board&apos;s MRDE Roadway Monogram, is probably the most noteworthy. This study involved eighteen multi- seam longwall mines in the United Kingdom. The following variables were taken into account: seam depth, pillar width, opening shape, floor and roof conditions, rate of face advance, vertical seam interval, and age and geometry of overlying workings. A multi-linear regression analysis of these variables fitted by appropriate field values developed a prediction equation for lower seam stability. Although the study included many variables, it neglected the nature or geology of the innerburden. The prediction equation was found unreliable, but it did show the significance of geometry as a factor controlling lower seam stability. MINING SEQUENCES Irrespective of the type of mining method selected only four possible mining sequences are possible for two seams. Finlay and Winstanley wrote in 1934. .. "Experience has shown that whether the upper or lower seam was worked first, the subsequent working of the other seam in the same area seriously affected the roads in the seam first worked." Those effects depended upon the four basic mining sequences: 1) The upper seam is mined out prior to mining the underlying seam; i.e., the seams are mined sequentially downward and mining of the underlying seam is not commenced until operations above are completed and have ceased.

Jan 1, 1982

By J. E. Pahlman

The Bureau of Mines conducted re-search to determine the resistance of basic refractories to corrosion by ash slags that would result from the burning of coal and lignite in metallurgical operations. Basic refractories are of economic interest because they are from one-third to one-half as expensive as 90-to 99-pct-alumina refractories. Static tests were performed at 1,300° C and dynamic tests at 1,400° to 1,650° C to investigate the stability of refractory specimens in a slag environment. In alkali-containing lignite ash slags, in order of decreasing resistance, magnesite, magnesite-chrome, and chrome-magnesite refractories all had corrosion resistances equal to or better than 90-to 99-pct-alumina refractories. These basic refractories formed slag penetration layers in dynamic tests, which helped protect them from further attack. Chemically bonded basic refractories in general did not have as good a resistance to slag attack as did burned refractories of the same composition. In very acidic coal ash slag, a chrome-magnesite refractory containing 30 pct MgO plus a slag inhibitor and direct-bonded magnesite-chrome refractories with 60 pct MgO had good resistance to attack. Chrome-magnesite refractories (30 to 40 pct MgO) were moderately to severely attacked, and refractories with greater than 70 pct MgO were severely attacked by this same slag.

Jan 1, 1982