Search Documents

By Alfred Jr. Petrick

Modelling of cash flows and calculation of net present value of rate of return over the expected life of a mine/processing unit continues to be the accepted basis for mineral investment decision-making. The computer and spreadsheet programs have greatly improved ability to perform sensitivity and risk analysis, hence our ability to deal with key analytical problems. These problems of cash flow analysis result from the necessity of forecasting an uncertain and risky future. Cash flow models begin with the critical assumption of a "design basis", then integrate all of the variables relevant to the investment decision including forecasts of price, cost, taxation, financial conditions, and company strategy. Alternate "scenarios" portray possible future outcome of the investment decision. The potential for straying from reality is enormous. The computer has eliminated the time consuming hand calculations from the cash flow analysis, but the analyst still struggles with the risk of trying to predict "how the project will really turn out".

Jan 1, 1991

By H. E. Powell

Magnetic susceptibilities were determined by the Bureau of Mines fur 19 cobalt- and nickel-bearing minerals and for a suite for synthetic cobalt-nickel diarsenides analogous to the rammelsbergite and safflorite series. Relations between susceptibility and composition were found to be nonlinear for both impure natural minerals and pure synthetics.

Jan 1, 1967

By M. Buarzaiga

High-pressure acid leaching (HPAL) of nickel laterite ores typically extracts 95% of the nickel and 91% of the cobalt. Impurity elements in the feed, such as iron, aluminium and chromium, also dissolve and can be detrimental to downstream processes. Rejection of impurities within the autoclave is highly desirable. Aluminium and iron hydrolyze, and then precipitate in the autoclave leaving only small amounts that are subsequently removed by precipitation in a neutralization step. The concentration of dissolved chromium can range from 200 to 2000 ppm, depending on the feed composition. Chromium in the liquid phase, usually in the hexavalent state, introduces many problems and can limit refining options. For example, chromium poisons the organics in solvent extraction making SX an expensive option. Also, hexavalent chromium must be converted to the trivalent state before it can be precipitated along with iron and aluminium. Bench-scale work confirmed that mixing small amounts of carbon-base additives, such as coke, with the autoclave feed significantly reduced chromium concentration in the autoclave discharge without any detrimental effects on paymetals recoveries and selectivity.

Jan 1, 2004

Jan 1, 1881

De Beers? Finsch diamond mine began operation in the 1960s as a surface project using the latest technology available. During the 1990s, the operation, located about 160 km (100 miles) west of Kimberley, South Africa, began its move as an underground mine. The earlier stages (Blocks 1, 2 and 3) of underground work at Finsch used open stope mining methods, also using the latest in technology. The mine began developing Block 4 in 2004. How-ever, studies on this level showed that block caving was a more efficient mining method than open stoping. It was during those studies that De Beers determined that Block 4 would be a good candidate for testing Sandvik?s automated truck haulage technology.

Jan 1, 2008

By Randall M. Wheeler

The purpose of this workshop is to illustrate how computer software can be used to evaluate blast design timing and simulate probable vibration characteristics. The workshop will be presented in two parts. Part I will cover the COMPU-BLAST Blast Design and Timing Analysis Program. Part II will focus on the ALPHA-BUST Vibration Analysis and Waveform Simulation Program. The information presented in the workshop will show how these two programs can be used together to help plan blast designs and aid in the control of vibration effects.

Jan 1, 1991

By Mike Treffy

El presente texto describe las características geoquímicas de las rocas ígneas del tipo HHPG (High Heat Producing Granites) en los Andes peruanos. Dichas rocas tienen diferentes edades geológicas y están relacionadas a los depósitos minerales como estaño, plata, cobre, zinc, plomo, uranio, etc y están ubicadas al extremo sur de la Cordillera Oriental y norte del Batolito de la Cordillera Blanca. Asímismo, en base al procesamiento geoquímico de la de ANDESDATA, se demuestra que estos granitos y rocas volcánicas félsicas corresponden a diferentes series magmáticas y que este tipo de rocas pueden relacionarse a altas concentraciones de plata, oro, cobre, zinc, molibdeno, uranio, favoreciendo la circulación de fluidos hidrotermales a grandes distancias de las intrusiones y rocas volcánicas bastante diferenciadas.

Sep 21, 2013

By J. C. Agarwal

Major scientific discoveries cannot be programmed. New or frontier technology, however, can and should be planned. It is the vital force underlying the business strategy of a vigorous and profitable mining and metals company. Generally, a radical departure from existing technology is required to 1) Enhance the competitive value of ore reserves, i.e., to turn low grade or marginal reserves into competitive reserves. 2) To increase yield and decrease costs. 3) To decrease investment. 4) To increase mine life. 5) To improve product quality to meet market demands. The very definition of frontier technology implies that it is so different that the old plant, if already in use, will not be directly suitable to use the new technology. Therefore, a priori, we know that the commercialization of frontier technology would very likely involve new investment. The recommended technique of economic and technical analyses, I propose to call "Success Analysis.'' Simply stated, it forces us, by step- by-step procedure, to ask "what is needed to succeed in this particular project." Therefore, the first step is to identify what is meant by "success." The definition of "success" requires an agreement with top management or sponsors of the project as to what constitutes "success" in their minds as contrast to the technical people perception of "success." In general, top management would define success as meeting: 1) The technical objectives, i.e., technology being developed would produce the desired product at the acceptable quality. 2) Financial objectives, i.e., when commercialized, the project's capital cost, operating costs and return on investment are better than the projects they would normally fund. The return has to be better because new technology is inherently riskier. 3) Marketing objectives in that the project will produce product of acceptable quality in timely manner, i.e., scale-up to commercial plant would not cause undue delays in start-up and that large fix-up expenditure would not be required to protect the projected return on the investment. Many projects are started in the laboratory to develop "Frontier Technology" in mineral processing. Most of these efforts are never commercialized. There are many causes of these failures. The major causes are listed below: 1) Inadequate communication with the sponsors of the project so that project objectives are defined and agreed to by all. 2) Ill-defined technical problems. 3) Incomplete development of technology. This has two major components: a. Incorrect scale-up and engineering. b. Inadequate start-up and maintenance planning. 4) Incomplete and over-optimistic process economics. 5) Lack of understanding of the future behavior of the marketplace. 6) Failure to complete R&D effort by the time it is needed. 7) Incorrect projections of human resources required; that is, technical, managerial and operating. 8) Lack of realistic forecast of the expenditures required for commercialization of the R&D effort. These pitfalls can be avoided by comprehensive planning and the use of success analysis technique. The major elements in "Success Analysis" are: 1) Identification of the "need" for new technology. 2) Unbiased and complete scale-up engineering. 3) Acceptable economic feasibility with enough margin of error in the sale prices and operating costs to justify the inherent risks in frontier technology. 4) Timing for availability of investment capital and markets. Many failures in the development of frontier technology are preordained because the initial selection of the research project has been without adequate attention to the needs of the industry. Many very innovative hydrometallurgical processes for recovery of copper from concentrates were developed in the past twenty years. They have not adapted except to a very minor extent because they did not satisfy the success analysis criteria. The copper industry in North America needed a process to compete against high grade ores and to avoid

Jan 1, 1985

By Roy A. Bell

A 200-foot diameter, 72-foot high petroleum storage tank was built in a hillside ravine where a portion of the site contained soft clay fill. Soil-cement piles, 40 inches in diameter and up to 38 feet long, were installed to support the tank shell and ringwall, and a portion of the tank floor by drilling into the stiff to hard underlying soils and weathered bedrock. The goal was to provide a strong enough foundation to resist the seismic tank shell loading and to minimize differential settlement of the tank shell to avoid out-of-round clearance problems between the shell and floating-roof. A cantilevered retaining wall up to 8 feet high was constructed around a portion of the tank perimeter by inserting steel H-beams into soil-cement piles. The cantilever soldier piles for the retaining wall were installed from a temporary earth embankment on the hillside directly over the top of the wall. The temporary embankment was then removed and the tank pad was cut to grade in front of the soldier piles. Some of the soil-cement piles for both the tank foundation and retaining wall could not be advanced to design depth because the drill/mixing tools could not penetrate to the required depth in the hard soils/bedrock. Smaller diameter borings (22 inches in diameter) were drilled through the centers of the short-of-design soil-cement piles to the design depth and then backfilled with concrete.

Jan 1, 1997

By W. A. Boyer

MINE hoisting ropes can be loaded to capacity only when the strength of each component is exactly known. Characteristic curves provide this information. When load and rate of acceleration are specified for an individual rope, the characteristic curve assumes a predetermined shape. The factor of safety, is the value drived by dividing the breaking strength of the rope by the total stress on the rope where it goes over the sheave wheel. Total stress is made up of the weight of skip and cage, weight of ore, weight of rope between sheave and rope clevise or attachment, friction load, and acceleration force. Bending stresses are neglected in this discussion, as sheaves and drum di- ameters are assumed to be large enough to reduce these stresses to a negligible value. Fig. 1 gives a family of characteristic curves for 13/8-in. improved plow steel hoisting ropes of 6 x 19 or 6x21 construction with hemp core. The curves will hold for all rope sizes of this specified quality when they are subjected to an acceleration rate of 1.608 ft per sec2. As acceleration rates of most deep shaft hoists are close to this value, this family of curves is presented as a standard. Deviation is slight for other rope sizes. Note that for 0, which is the most important, the deviation is within the width of the line on the graph. In a comparison of values for ropes of 5/8-in. and 21/4-in. diam the deviation ranges from a maximum 1.2 pct at 0 ft to 0.6 of 1 pct at 5000 ft. The value of 5.55 is the same for all ropes at 2500 ft. The tabulated columns to the left of the curves are values of total connected load for each size of rope from 5/8-in. to 21/4-in. diam inclusive. With each of these connected loads the safety factor characteristic curve for the rope will be as indicated.

Jan 1, 1957

By Yoginder P. Chugh

Physical mine modeling and computer simulation techniques were used to conduct technical feasibility and economic analyses of selected pit configurations for surface mining a single coal seam dipping at about 15 percent to a maximum depth of 150 feet. The pit orientations along the dip (dipline), strike (strikeline) and along an apparent dip were studied. Physical models of each pit configuration were prepared first to delineate potential operational problem areas, and also to furnish a data base-for the computer simulation models. A dragline micro- simulation model (Fluor), a shovel/truck simulation model and a cost analysis model were used to conduct the study. Sensitivity analyses were conducted on variables such as the pit width, block length, and angle of dip. The productivity (Cu.yds/hour) of the mining equipment (dragline or stripping shovel) under each set of conditions was used to optimize the pit configuration. Both block length and panel width were found to significantly affect the dragline productivity larger size trucks were preferable to smaller size trucks in the pit configuration along an apparent dip. The cost analysis results showed that the pit configuration along the seam dip was more economical than the pit along the strike or along an apparent dip.

Jan 1, 1984

By R. S. C. Rogers

The present study examines the use of two vs. one stage of hydrocyclones in a pilot scale ball mill circuit. The approach used for doing this is based on computer simulation with mathematical models that explicitly account for the grinding mill/classification system interaction. Parameters for the ball mill model are obtained using a short-lived radioactive tracer method in open circuit tests of the study mill. The hydrocyclone model, based on the size selectivity concept, is developed and validated in standalone testing of the study hydrocyclones. Simulation procedures are validated through actual tests of the closed circuit grinding configurations of interest. Results of the evaluation indicate that two stages of hydrocyclones in the study circuit would improve classification, increase productivity and lead to more closely sized circuit products when compared to the conventional one stage hydrocyclone circuit. A computer simulation approach for comparing these circuit configurations in larger scale grinding circuits is also suggested.

Jan 1, 1981

By Milton F. Goodrich

The idea of using water-only cyclones to scalp clean coal from the feed to other devices has recently been gaining in popularity. 1,2,6 An indication of this popularity is that water-only cyclone scalping is already employed in a few commercial plants. Water-only cyclones scalp the feed to a heavy media vessel in the Warwick plant,3 to Deister tables in the Rowland plant,4 and to a heavy media cyclone circuit in the Amherst Coal Company's McGregor plant.5 Despite this trend, we found surprisingly little information in the literature on the performance of water-only cyclone scalping. Therefore, we undertook a study of its technical merits, limiting our investigation to a water-only cyclone scalper in conjunction with heavy-media cyclones. We felt this to be a reasonable combination of equipment, since both processes operate well on the same size coal. First, we developed simple mathematical relationships revealing general characteristics of the system. Next, we compared the clean coal yield and quality obtainable by the system on four different coals with that obtainable by heavymedia cyclone circuit alone and water-only cyclones middlings recirculation circuit alone. This was done by using computer program to simulate the operation of the three systems. Finally, we considered the effect that implementation of the system would have on a typical preparation plant circuit. The System The system chosen for study [(shown schematically in Fig. 1)] comprises two stages: a primary stage (PI) consisting of water-only cyclones, and a secondary stage (P2) consisting of heavy media cyclones. The water-only cyclones separate raw coal at a low specific gravity, producing a relatively small amount of high quality clean coal and a refuse containing much good coal. This refuse is then recleaned in the secondary stage, where the sharper separation of heavy media cyclones, operated at a higher specific gravity, reclaims additional clean coal. Finally, the clean coal from both stages is combined as the system's clean coal, while the refuse from the secondary stage becomes the system's refuse output. Theoretical Analysis Two assumptions lie behind the following theoretical analysis. The first assumption is that partition curves can represent not only the performance of each stage but also the performance of the system as a whole. These curves show partition number as a function of specific gravity, where partition number is either the percent or fraction of a raw coal with a given specific gravity reporting to clean coal. The second assumption is that these curves are independent of the raw coal characteristics. At any specific gravity the system's partition number is solely a function of the partition numbers of each stage. For convenience, we use the following variables in describing this relationship: g = specific gravity, PS(g) = the partition number of the system at specific gravity g, P1(g) = the partition number of the primary stage at specific gravity g, P2(g) = the partition number of the secondary stage at a specific gravity g. For this system, we can show algebraically that the relationship is: [Ps (g) = PI (g) + (1- P1(g)) P2 (9)(1)] With this equation the system's partition curve can be determined if the partition curves of each stage are known. By definition, the partition numbers P1(g) and P2(g) must each be less than, or equal to, one. Placing these constraints on Eq. (1), we arrive at the relations: [Ps(g) % P1(g) and Ps(g) % P2(g)(2)] Thus, at any specific gravity the system's partition number is always greater than that of both stages. [Figure 2] shows typical partition curves. Curve I represents the primary stage, which consists of water-only cyclones operating at a separating gravity of 1.3. (The separating gravity or the

Jan 1, 1979

By Jackson R. Pressley, Bruce Vandenberg

The knowledge of how and when your explosives go off can help you make intelligent decisions regarding future application of explosives thus removing some of the black magic associated with blasting. The net result will be intelligently set-up shots with timing, explosive charges, and placement more accurately done. Lower operational costs can be achieved when the correct combinations are applied.

Jan 1, 1992

By Sam G. Demou

Drag bit cutting tests were conducted chiefly to determine bit forces encountered in rock types representing a range of compressive strength and hardness and thus enable the design engineer to estimate a continuous miner's performance in a range of rock types. Testing was carried out on the Bureau of Mines uniquely designed, instrumented rock cutting device (RCD). This device can make 3-ft-long cuts at various depths and speeds to approximate full-scale drag bit cutting by continuous miners. Point attack and cutter bits were tested on four rock types ranging in compressive strength from 7,200 psi (trona) to 29,600 psi (Valders white rock). Cutting tests were made at depths up to 1 in and speeds up to 70 in/sec. Data included normal and cutting forces, groove depth, groove width, and weight of rock broken. The effect of varying the cutting speed was determined, and specific energies were calculated for the various conditions. The bit wear was also observed. For the rocks tested, compressive strength and hardness are good indicators of forces, specific energy, and wear encountered. Results, especially bit wear observations, show the possibility of extending continuous miners, already being used in trona, to Indiana limestone (10,000 psi). Rapid bit deterioration precludes use in Tennessee marble (16,800 psi) and Valders white rock (29,600 psi).

Jan 1, 1983

By J. S. Walker

Public Law 95-87 enacted by Congress in 1977 and subsequent regulations require coal mine operators to predict the effects of mining on the environment prior to the start of mining. Prediction of the effects of mining on the local groundwater system is a difficult task due to its complex nature and the lack of detailed published information. An investigation of water level fluctuations in five we1 1s positioned over three consecutively mined longwall panels was initiated in conjunction with an intensive subsidence monitoring program. The purpose of this work was to provide detailed information which could be used to predict certain hydrologic effects of longwall mining in the Northern Appalachian Coal Region. The results of this case study indicate that water level fluct- uations in the local ground-water system above longwall panels is associated with subsidence and that the static water level will generally re- establish at or near the pre-mining elevation after mining is completed.

Jan 1, 1986

By Xiaolin Cheng, A. J. Sinclair

"Norm calculation procedures originally designed principally for igneous rocks are very rigid in their application and, in general, do not utilize important alteration minerals. For instance, volatile components are essential constituents of many metasomatic rocks but either are not used or are not used effectively in determining normative minerals by the CIPW norm or Niggli-Barth norm procedures. Our approach to determining norms for metasomatic rocks allots all major, minor and volatile constituents to a set of standard normative minerals which are commonly observed as products of hydrothermal alteration. A useful procedure is to have a metasomatic norm approximate the mode as closely as possible. Petrographic information helps to provide the necessary mineralogical constraints. With the recognition of an immobile component and knowledge of sampling and analytical errors in lithogeo chemical data, it is possible to recast norms into amounts of alteration minerals relative to a given amount (100 grams) of parent rock. Combining these results with the calculations of absolute loss or gain of individual chemical constituents and considering errors (e) for an appropriate confidence level, an overall (net) material exchange equation can be presented as follows:<equation>where all items have extensive units (e.g., grams). The model includes the initial and final mineralogy of rocks (evident in the field) as well as gains and losses of specific elements. Only those constituents whose calculated losses and gains are significantly different from zero are retained in the above model, and only the normative minerals with abundances significantly different from zero are included in the model (equation). One advantage of the chemico-mineralogic model is that parent rock and product are modeled in mineralogical terms that are familiar to and easily usable by exploration geologists."

Jan 1, 1995

By Xiaolin Cheng, A. J. Sinclair

"Abstract - Norm calculation procedures originally designed principally for igneous rocks are very rigid in their application and, in general, do not utilize important alteration minerals. For instance, volatile components are essential constituents of many metasomatic rocks but either are not used or are not used effectively in determining normative minerals by the CIPW norm or Niggli-Barth norm procedures. Our approach to determining norms for metasomatic rocks allots all major, minor and volatile constituents to a set of standard normative minerals which are commonly observed as products of hydrothermal alteration. A useful procedure is to have a metasomatic norm approximate the mode as closely as possible. Petrographic information helps to provide the necessary mineralogical constraints. With the recognition of an immobile component and knowledge of sampling and analytical errors in lithogeochemical data, it is possible to recast norms into amounts of alteration minerals relative to a given amount (100 grams) of parent rock. Combining these results with the calculations of absolute loss or gain of individual chemical constituents and considering errors (e) for an appropriate confidence level, an overall (net) material exchange equation can be presented as follows:SMineralparent rock ± e + SConstituentgained from solution ± e= SMineralaltered rock ± e + SConstituentlost from wall-rock ± ewhere all items have extensive units (e.g., grams). The model includes the initial and final mineralogy of rocks (evident in the field) as well as gains and losses of specific elements. Only those constituents whose calculated losses and gains are significantly different from zero are retained in the above model, and only the normative minerals with abundances significantly different from zero are included in the model (equation). One advantage of the chemicomineralogic model is that parent rock and product are modeled in mineralogical terms that are familiar to and easily usable by exploration geologists."

Jan 1, 1995

By K Baskaran, T McKenzie

Sustainability is becoming increasingly important in the mining and mineral processing industries and must incorporate the associated waste products. Acid mine drainage (AMD) is one such by-product and is one of the most serious environmental problems facing the minerals industry today. The oxidation of sulphidic mine wastes often continues for a substantial period of time after mine closure, resulting in difficult and costly remediation and rehabilitation works. Mining companies are often reluctant to spend increasing amounts of money on waste treatment when the mine life is limited or even finished. Hence a simple, low maintenance and low-cost method of treating AMD is required. Whilst this paper does not address the issue of AMD, it does propose methods for removal of individual species from AMD with potential benefits, including raising AMD pH. A novel concept of using biosolids as a biological adsorbent, or æbiosorbentÆ, of metals from AMD is being investigated at a laboratory/pilot scale level. Biosolids are a by-product resulting from the biological treatment of wastewater, and have been previously shown to adsorb metals from aqueous solutions. This could lead to an environmentally sustainable or ægreenÆ method for treating both AMD discharges and disposing/reusing the biosolids. The result of a laboratory-scale study of the biosorption of Zn(II) is presented in this paper. Physical parameters including reaction kinetics, mixing speed and solution pH were investigated. Solution pH also rose an average of 2 pH units over the 24 hour equilibrium time û a valuable side effect when treating acid mine drainage. The outcome of the study highlights the usefulness of biosolids as a biosorbent for the removal/recovery of metal ions from acid mine drainage. A simple, low-cost treatment technology requiring low maintenance would be beneficial to the mining industry to address some issues relating to AMD and would help integrate environmental and economic considerations into sustainable environmental management.

Jan 1, 2002

By W. A. Sheaffer

Electric furnace installation and tough-pitch copper-casting operation at Canadian Copper Refiners Ltd. are described. General layout, power supply and control, refractories, induction pour hearth, casting equipment, metal temperature control, and oxygen-content control are discussed. E LECTRIC furnace at Canadian Copper Refiners Ltd. was put into operation in August 1949. The installation was designed primarily to melt electrolytic copper cathodes and to produce vertically cast tough-pitch shapes. To meet emergencies during reverberatory furnace shut-downs, provision was made for casting horizontal tough-pitch wire bars. Due to the ease with which metal temperatures, melting rates, oxygen content, and casting hours can be altered to suit production demands, the electric furnace has met the requirements admirably. General Layout The plant has been described by H. S. McKnight&apos; and by J. H. Schloen and E. M. Elkin.2 Since these papers were written, the electric furnace has been installed in a 140 ft extension of the original casting building. The extension, 264 ft in width, is divided into one 24 ft and four 60 ft bays, each a continuation of similar bays in the older building. Fig. l, a partial floor plan of the extension, shows the location of major equipment. Power Supply and Control Power for the electric furnace and its auxiliary equipment is delivered by Quebec Hydro over two 12 kv three-phase, 60 cycle pole lines. One line is in use, while the second serves as an emergency standby. Connections for the electric furnace from these lines, which also feed the refinery power house, are made to a substation on Canadian Copper Refiners Ltd. property. Connections between this substation and the furnace transformer room are in underground conduit. The are-furnace transformer, 550 v auxiliary-equipment transformer, and 12 kv switchgear are in the transformer room near the furnace. Stepdown from 12 kv to are-furnace voltages is done by a 7500 kva oil-immersed water-cooled three-phase 60 cycle transformer. This unusually high kva rating is available because the arc furnace and transformer were adapted from a steel-melting unit. Secondary voltages are available in 16 steps between 251 and 95 v (across phases). A four-position tap changer, operated from the furnace switchboard is connected to taps yielding 199, 164.5, 127, and 115 v, respectively. The transformer primary feeder is equipped with a General Electric Magne-Blast circuit breaker. The pour hearth, casting wheel, bosh conveyor, and other auxiliary equipment are fed by a 12 kv to 550 v, 750 kva three-phase 60 cycle transformer. Graphite electrodes, 14 in., with tapered nipples are held in the are-furnace electrode arms with steel wedges. The arms, fed from the transformer secondary busbars by flexible cable bundles, are actuated by winch cables. Direct-current motors operating the winches are supplied by a 250 v motor-generator set. Power input to the furnace is controlled from the furnace switchboard. The board has the usual combination of remote controls for the 12 kv breaker and transformer tap changer, meters, overload relays, automatic and manual electrode motion controls, switchgear, etc. When power is on the furnace, automatic electrode feed is used; the power draw on a given voltage tap then is governed by current-limiting rheostats. The rheostats work through a Westinghouse bal-lanced-beam control circuit which, in turn, operates reversing contactors in the winch motor circuits. Arc Furnace The arc furnace is a standard-type NT Pittsburgh &apos;Lectromelt with inside shell diameter 12 ft 3¾ in. By hydraulic-lift cylinders, the furnace can be tilted forward to 39 ½ ° and backward to 5" from horizontal. A roof-swing cylinder is also installed but not used. Fig. 2 shows refractory-lining details. With the lining as shown, operating life between repairs is 6 to 9 months of two-shift casting operation. Charge slot, skim door, and roof refractories then are replaced completely, and any necessary repairs made to side walls. Bottom life is longer and the original under courses are still in service, while the top course was replaced in October 1951. The launder is lined with high chrome-magnesite trough brick backed by fireclay and insulating brick. The trough is covered with fireclay brick and asbestos sheet. At intervals, openings 12x6 in. are left in the permanent covering. These are covered with

Jan 1, 1956