Search Documents

By M. G. Ayat, B. C. Scott

H. P. Ferri I refer to the article "Degradation Process in Coal Slurry Pipelines" by M.G. Ayat and B.C. Scott in the September 1988 issue of Mining Engineering. I would like to congratulate Dr. Ayat and his colleague on a very interesting and valuable piece of work. This article is an excellent one, describing a real phenomenon that has been ignored for so long. I do, however, have an objection. The article fails to address the degradation process when coal is processed or transported in other ways than slurry form, particularly those employed in this work. Could, for example, screening (wet or dry) have contributed to the extent of degradation? Perhaps scanning electron micrographs of the samples examined and after screening would have helped to clarify this. I would like to point out that degradation is also observed in pneumatic conveying systems, mechanical conveying, and other means of transportations employed by the mineral industry. Reply by M. G. Ayat We gratefully acknowledge Dr. Fern's comments on the subject paper. As he so rightly points out, the degradation process is a real problem, not only in coal slurry pipelines, but also in many handling and processing equipment employed by the coal and mineral industry. Scanning Electron Microscopic examination of samples may offer very little information about the extent of degradation for several reasons. One is that preparing a representative sample that has undergone degradation will be extremely difficult if not impossible. This problem is exacerbated by the fact that SEM requires only very small amounts of material for testing. Secondly, there is a sample preparation step involved in the SEM examination and, as such, it would be very difficult to associate the extent of degradation to either sample preparation or handling and processing operations. Finally, it is probably worth mentioning that degradation is a multi-faceted phenomenon involving many natural and man-made factors. The logical approach to this problem would, therefore, seem to be a "one-step-at-a-time" approach.

Jan 2, 1989

By F. Mielli

Over the last three years there has been an increase in mining congress and conferences focusing on water management. Water in mining is a key concern for mine managers and operations, and we are now seeing this operational dilemma escalated to the level it deserves. According to Frost & Sullivan research, mining water and wastewater treatment and its related technologies will be a U$3.6 Billion global market in 2016, which translates the importance of this issue. Looking at the current mining trends, some factors increased the importance of water in mining operations, including: ? Future mining investments will be located in geographical locations where water is scarce (water stressed areas) ? A decline in grade will increase the intensity of mineral processes and therefore increase the use of the water ? Increased sustainability and regulations will extend the control/monitoring of water during the entire mine life cycle representation of the mine water flow system. A water balance has several benefits/objectives in a mine operation: ? Mine planning/simulation of mining scenarios/climate change Production optimization ? Minimizing strategic and operational risks ? Social license to operate ? Water-smart mine planning ? Target: water neutral mines ? Benchmarking within and between sites ? Sharing knowledge/technology transfer ? Reporting/communications ? Permitting/Compliance

Feb 23, 2014

By Mehmet Mesutoglu, Ihsan Ozkan

This study investigates the feasibility of rock bolting support in an underground coal mine gallery with a thick coal seam. The Ömerler underground coal mine working area, owned by West Lignite Enterprise of Turkish Coal Enterprises, was selected for this purpose. Longwall top coal caving is implemented as the production method in the Ömerler underground coal mine. Field and laboratory studies were conducted to determine rock mass and rock material properties, followed by experimental, empirical and numerical analyses based on the acquired data. The obtained design results were evaluated using a resin-grouted rebar rock bolting (RBR) system and steel arch support (SAS) pilot application areas. The numerical modeling results conducted using the fast Lagrangian analysis of the Continua 3D (FLAC 3D v6.0) program indicated less displacement and secondary stress change in the RBR-supported zone compared to the SAS-supported zone. In situ measurements also demonstrated that RBR provided more successful support to the roof during coal production activities. The findings suggest that RBR is a more effective solution when evaluating the feasibility of rock bolting support systems in underground galleries with thick coal seams at the Ömerler Mine. This study emphasizes the importance of more sustainable and safe support systems to enhance operational efficiency in the coal mining industry.

Mar 1, 2025

By Donald W. Gentry

Introduction The mining industry is perhaps best characterized by the numerous and abnormally high risks associated with the business of mining and by its most basic assets - ore deposits, people and technology. At the most fundamental level, it is the quality and effective utilization of these basic assets that determine a mining company's ability to compete efficiently in the international market place. The gold industry has received much attention in recent years and especially in recent months, for good reason. This commodity serves as an excellent platform for the subsequent discussion of these three basic assets and their role in mining's future. Facets of the gold industry The fate of this noble metal in recent times illustrates some of the various risks associated with the mining industry and the importance of high-quality assets, whether they be ore deposits, people or technologies. The following comments pertaining to the gold industry help place proper focus on subsequent discussions. All monetary amounts given are in US dollars. Throughout 1997, falling gold prices broke record after record: first hitting three-year lows, then a 12-year low in July and finally an 18-year low at year's end. From Jan. 1 through Dec. 31, 1997, the price of gold fell by more than $2.57/g ($80/oz), or some 22%. The decline in gold-mining profitability and stock prices was even more dramatic. At the beginning of 1997, the 20 largest gold producers in North America had a market capitalization of approximately $44.5 billion. At year's end, this market cap had fallen to $26.2 billion, a reduction of 41%. The total market cap of all the world's gold companies fell from $71 billion to $46 billion, a 35% drop (Serwer, 1998). Analyses suggest that, at a spot gold price of $9.65/g ($300/oz), only one, or possibly two, North American gold-producing companies can make money when break-even or "total costs of production" are considered. While strong hedge positions mitigate this situation in the short term, clearly, this is an industry in desperate straits. For example, write-downs for North American producers in 1997 were approximately $2.5 billion. Speculative short positions in 1997 were near historical record highs and generally exceeded long positions by factors of 9:1 or 10:1. Much of this is driven by some US commercial banks, hedge fund managers and commodity speculators who can significantly move gold prices sharply over short periods of time by selling gold they do not own in the expectation they can buy it back at a lower price before they have to deliver. These signifcant short positions are likely to continue as long as central banks are willing to sell or loan gold to financial institutions that in turn make the gold available to speculators who hope to capitalize on price fluctuations. The Far East accounts for approximately 26% of the annual demand for gold. Including India, the region accounts for 43% of total demand. More than 85% of the physical demand for gold is for investment jewelry in Asia and the Middle East. This demand is extraordinarily price sensitive. The fairly recent turmoil in Asian markets and economies, combined with currency devaluations throughout the region, have made gold more expensive (30% to 50% in local currencies) for those peoples most likely to treat gold as a safe investment haven. Higher interest rates also have reduced the amount of money they have to spend on consumer goods such as jewelry. All of these recent Asian events are deflationary in nature. Deflation, if coincident with rising unemployment, could result in significant dishoarding of gold. There is one key fact to remember when analyzing

Jan 1, 1999

By Albert Hofstra, Joel Leventhal

The gray, dark gray and black colors of the sediments and the presence of pyrite in the Carlin, Jerritt Canyon, Horse Canyon, Betze, and Gold Acres sediment-hosted disseminated gold (SHDG) deposits indicate that these rocks are not oxidized with respect to carbon and iron sulfide. Both ore and non-ore host rocks generally are dark colored. Analyses for noncarbonate carbon (usually called organic carbon) in samples from SHDG deposits and related unmineralized rocks indicate that these rocks contain from a few tenths to a few percent carbon. We use the term noncarbonate carbon because, strictly speaking, organic carbon denotes carbon that is bound to hydrogen. However, the carbon present in SHDG deposits and related rocks has very little hydrogen chemically bound to it. Rocks with less than 2% organic carbon are usually gray in color. The dark colors in our samples are due to the relatively high temperatures that these rocks have experienced relative to organic matter maturation. We confirm the results of Nelson et al. (1982) that previous reports (for example, Radtke and Scheiner, 1970) were incorrect in reporting significant amounts of humic acids, extractable hydrocarbons (bitumen) or "carbonaceous" material. We believe these reports had problems with contamination or misdirected emphasis. Our samples consisted of unweathered rocks, collected mainly from mine cuts (benches, trenches, cores) and put into glass containers or wrapped in aluminum foil to avoid contamination. Only a few samples were collected from natural outcrops. We were careful to avoid samples in proximity to blast holes that were packed with ammonium nitrate and diesel fuel ("ANFO"). Contamination by ANFO may explain some of the previous reports of hydrocarbons, pyrobitumens, humic acids and anomalous nitrogen. When we were collecting samples, it was apparent that in addition to lithologic and color differences, the surface appearance of the rocks was distinguishable: "preg robbing" material that had a sooty surface; "graphitic" rock with a shiny surface; and more ordinary rocks that had a flat gray or black appearance. [Preg robbing is a term for ores that adsorb gold from the milling circuit; the graphitic material had not been x-rayed to actually identify graphite.] We hoped that laboratory study of these samples would show some fundamental differences that we could relate to ore genesis. Even though these sample types appeared different in the field, we could not distinguish any differences in thermal maturity (see below). Normal petroleum/organic geochemical laboratory procedures (solvent extraction, Rock-eval, pyrolysis-gas chromatography) failed to yield much information because the carbon was over-mature relative to oil and gas generation. Rock-eval results show less than 20 (± 20) mg of pyrolizable hydrocarbons per gram organic carbon (and an indeterminate Tmax) indicating that the organic matter in these deposits is burned-out relative to petroleum generation [with an H/C (atomic) ratio of less than 0.41. Because of the non-definitive results from standard organic geochemical methods, x-ray diffraction (XRD) and Raman spectroscopy were used to characterize the carbon. Prior to XRD, samples were treated with HCl and HF acids (1) to remove quartz that has a peak near 26° [20] at the position of the 002 peak of graphite and (2) to concentrate the carbon to a level detectable by XRD. The residue was put on a colodion mount and analyzed by XRD. The XRD analyses yielded a broad peak at 26° [20]. The shape of this peak indicates that the carbon is more thermally mature than anthracite rank coal but less mature than crystallized graphite. Based on XRD results from the literature (Landis, 1971) that are calibrated to metamorphic facies, results from our samples correspond to pumpellyite-actinolite to lowermost greenschist facies conditions, probably

Jan 1, 1990

By C. Y. Chen

In order to put an end to the uneven environmental standards and enforcement established by various coal mining States in the United States, the Congress passed and the President signed the Surface Mining Control and Reclamation Act (Public Law 95-87 or the Act) on August 3, 1977. Within one year after passage, the Secretary of the Interior was directed to promulgate permanent program regulations implementing Title V provisions of the Act. These regulations were to form the basis for State regulatory programs under the Act and to allow States to assume primary regulatory responsibility. In order to implement the underground mining provision of the Act, the Secretary is directed in section 516(a) of the Act to promulgate rules, but ones which take into consideration the distinct differences between surface coal mining and underground coal mining. The regulations developed by the Secretary and promulgated on March 13, 1979, with respect to subsidence control, included (1) a subsidence control plan as part of the permitting process, (2) substantive provisions establishing performance standards for subsidence control, and (3) provisions providing for the prohibition of mining under perennial streams, impoundments larger than 20 acre feet, and certain public facilities and urban areas.

Jan 1, 1985

By R. T. Hukki

The paper describes a two-stage industrial hydraulic classification system including hydrocyclones for primary classification and newly developed cone classifiers for secondary classification. Performance data based on large scale grinding and classification tests are presented.

Jan 1, 1981

By Tuncel M. Yegulalp

This article summarizes some of the results of the continuing research activity at the Henry Krumb School of Mines, Columbia University, directed towards a better understanding of the stochastic behavior of slurry flows in underground coal-slurry pipeline systems. The slurry system under consideration here replaces all conventional haulage systems between the coal face and the coal preparation plant on the surface. At the beginning of the pipeline system, flow rate and frequency of coal slurry is very closely related to the activities of production equipment at the coal face. Due to several scheduled and unscheduled interruptions in continuous miner's coal mining activity, resulting slurry flows possess stochastic characteristics. Proper understanding of these characteristics is believed to lead to a more realissic and economic design of slurry handling equipment. Fundamental ideas supported by field observations are presented and a Semi-Markov model of the slurry flows is developed. As a result of this formulation, behavior of combined slurry flows from N independent coal faces can be expressed in analytical form. The contents of a slurry surge tank receiving slurry flows with certain stochastic characteristics is investigated as a function of discharge pump characteristics and time.

Jan 1, 1981

By G. C. Presley

With plant construction scheduled for the Spring of 1997, U. S. Zeolites will become the new source for quality in natural zeolites. The company will selectively mine, dry process, package, and market a very high quality clinoptilolite possessing many unusual properties. These unusual chemical and physical properties provide a multitude of industrial, commercial, and consumer applications. Extensive testing has determined many potential applications.

Jan 1, 1997

By J. G. Pippenger

The Coteau Properties Company's Freedom Mine is a surface lignite mining operation located near Beulah, North Dakota. Coteau, a wholly owned subsidiary of The North American Coal Corporation, produces more than 13.5 million tonnes per year, ranking it among the largest mines in North America. In 1993, Coteau ranked seventh in terms of annual coal sales, becoming the largest mine outside the Powder River Basin of northeastern Wyoming. However, for the Freedom Mine to stand among these Powder River heavyweights, Coteau over-came several obstacles, including a coal seam with three quarters the heating value and overburden to coal ratios three times higher than the Powder River mines. Coteau's ability to compete has been built around a program of continuous productivity improvement, cost efficiency, and innovation. This report highlights a few of the major improvements implemented in recent years.

Jan 1, 1995

By RALPH BALTES, AMIR EHSAN KIANFAR, FABIAN UTH, ELISABETH CLAUSEN

As indoor positioning provides particular challenges due to the unavailability of GPS signals, various systems such as ultrawideband (UWB), radio frequency identification (RFID), ultrasound, and wireless local area network (WLAN) have been proposed in recent years. Some of these technologies are currently being marketed and some are still being developed. UWB technology allows for higher precision while also reducing power consumption. Hence, the underground automation and localization systems can use this technology for more accuracy and robustness. This article discusses new robust UWB modules used for underground positioning and collision avoidance with regard to human safety in underground mining operations.

Jul 31, 2020

By Ritika Kundu, Moussa Wone, Amanda M. Morgan, Tatiana K. Proctor

The Piney Branch CSO 049 Storage Tunnel (PBST) will be the last major project for the DC Clean Rivers Program. The project includes construction of a 2,200-foot-long, 18-foot-internal-diameter storage tunnel; two connections to combined sewer infrastructure; and two shafts. Major challenges include excavation underneath an existing live outfall that must remain in service; excavation of the tunnel through several areas of low rock cover and mixed-face conditions; and negotiations with a variety of federal, regional, and local stakeholders to address project impacts. This paper describes the overall project scope; geotechnical conditions along the alignment; design and construction challenges; and construction procurement.

Jun 23, 2024

By A. Jennings, J. Cornet, P. Perrone

Controlling long overland conveyors with multiple drive stations is a complex problem. The methods for controlling these belts in a steady state are well-covered in existing literature, but a discussion of starting and stopping strategies is absent. This paper outlines different control strategies for starting, stopping and running conveyors with multiple drive stations. To illustrate the benefits of applying the methods described in this paper, the authors show how their adoption corrected many of the control system problems at Essroc Cement’s overland conveyor, which features three drive stations separated from each other by long lengths of continuous conveyor belt.

Jan 1, 2013

By M. J. Raffaldi, L. A. Sandbak, K. K. Dehn, S. N. Warren, J. B. Seymour

"Confidence in the design and stability of an undercut cemented rockfill (CRF) span requires a thorough understanding of its emplaced material properties. The unconfined compressive strength (UCS) of cast 6×12-inch QA/QC cylinders is typically used as an index test to track the performance of a backfill plant. However, these strengths need to be adjusted for oversize aggregate removal and sample size effects if used to estimate the emplaced UCS. UCS testing of 6×12, 12×24, and 18×36-inch cast CRF cylinders was used to quantify the relationship between readily available 6×12-inch QA/QC strength data and the emplaced UCS. Test results indicate that the UCS of 18×36-inch cylinders is roughly 40% of the 6×12-inch UCS for similarly compacted samples. Mine QA/QC data indicate that the UCS of CRF is highly dependent on the degree of compaction as measured by bulk density. Based on these results, knowledge of the emplaced CRF density compared to 6×12-inch QA/QC density is also necessary to estimate the emplaced UCS. The scale effect reported in this study is more pronounced, compared to current understanding of backfill properties discussed in the literature, and underscores the need for further site-specific testing. INTRODUCTION The National Institute for Occupational Safety and Health (NIOSH) Spokane Mining Research Division (SRMD), in cooperation with Barrick Gold Corporation, is conducting research with the goal of determining a relationship between 6-inch QA/QC cylinder strengths and the in-place strength of cemented rockfill (CRF). CRF cylinders with dimensions (diameter × length) of 6×12, 12×24, and 18×36 inches were cast onsite at the Turquoise Ridge Joint Venture (TRJV) and Cortez Hills Underground (CHUG) mines using CRF mixed at the mines’ respective batch plants. UCS tests were performed after 28 days of curing to estimate the size correction factor for each mine’s CRF. This paper presents an overview of the mix designs, sample collections, testing methods, test results, and an estimate of the UCS size correction factor for CRF at these mines. BACKGROUND Underhand cut and fill mining methods utilizing cemented backfill are common in underground metal mines in the United States. These methods are advantageous for controlling weak or highly stressed rock mass conditions by creating an engineered material that can be placed in openings, forming a beam that provides a safe, stable back for the next undercut. Increased use of underhand methods in the last 30 years has greatly improved safety in many cut-and-fill mines. However, with the continual push for more efficient and cost effective mining, there is a need to thoroughly understand the geotechnical properties of cemented backfill to assist in designing safe, stable openings while optimizing costs and production schedules."

Jan 1, 2018

By P. T. Bissert, L. Yan, M. D. Klein

"Federal regulations require the installation of refuge alternatives (RAs) in underground coal mines. Mobile RAs have a limited ability to dissipate heat, and heat buildup can lead to a life-threatening condition as the RA internal air temperature and relative humidity increase. The National Institute for Occupational Safety and Health Office of Mine Safety and Health Research performed heat testing on a 10-person tent-type training RA and contracted ThermoAnalytics, Inc. to develop a validated thermal simulation model (TSM) of the tested RA. The TSM was used to examine the effects of the constant mine strata temperature assumption, initial mine air temperature (MAT), initial mine strata surface temperature (MSST), initial mine strata temperature at depth (MSTD), and mine strata thermal behavior on RA internal air temperature using 117 W of sensible heat input per simulated miner. For the studied RA, when the mine strata temperature was treated as a constant, the final predicted RA internal air temperature was 12.8°F lower than it was when the mine strata thermal behavior was included in the model. A 10°F increase in the initial MSST resulted in a 7.1°F increase in the final RA internal air temperature, whereas a 10°F increase in the initial MSTD yielded a 2.5°F increase in the final RA internal air temperature. INTRODUCTION Since 2008, the Mine Safety and Health Administration (MSHA) has required the installation of refuge alternatives (RAs) in underground coal mines (1). MSHA requires that RAs provide an environment with breathable air for entrapped miners for a 96-hour period. Heat buildup inside an occupied RA is a serious concern. Without a means to dissipate the heat and humidity generated by the occupants and the carbon dioxide scrubbing system, the temperature and humidity inside RAs could lead to severe discomfort or heat stress depending on the mine ambient temperature before and during occupation of the RA. NIOSH recommended an apparent temperature limit of 95°F in its 2007 report to Congress (2). In 30 CFR 7.504, MSHA has specified a maximum apparent temperature of 95°F inside an occupied RA (3). The apparent temperature is calculated using both air temperature and relative humidity (4). Calculations could be used to determine if the air within an occupied RA will reach the apparent temperature limit. In these calculations, in order to simplify the problem, the mine temperature could be assumed to remain constant. With this assumption, the effects of the mine strata thermal mass and thermal conductivity on the heat buildup in an RA are ignored. OMSHR research has shown that the mine air and strata temperatures surrounding an RA increase when subjected to the heat of an occupied RA (5). Thus, calculations performed with the assumption that the mine temperature is constant would under-predict the resulting dry-bulb temperature and apparent temperature for a given RA occupancy. The mine air temperature has been used as the basis for RA temperature calculations. However, research has not been used to confirm that the mine air temperature is the appropriate basis."

Jan 1, 2016

By McCarthy. Richard E.

The prime movers of the Centralia Project, the surface coal mine and the power plant, private utilities are very aware of public attitudes. They have established and maintain the reputation for leadership in public service and are genuinely concerned that this project initiate the pro-grams and design the hardware to answer the question :What will be the impact of the mine and the plant on the physical environment, i.e., the air, soil and the water." So in the beginning almost as soon as the decision had been made to go ahead with the project, officials of the two private utilities, Washington Water Power and Pacific Power & Light went on record and publicly stated time and again that the overall approach of the Operators would be to do all that was reasonably possible to safeguard the environment. This attitude became part of the thinking of the project planners, part of the design of the power plant and part of the plan for the surface mine. The coal to be mined is from the Skookumchuck formation. The formation is Late Eocene and is made up of poorly consolidated marine and non-marine deposits. The major overburden removal equipment is a bucket-wheel excavator -conveyor system and a 56 yard dragline. Mining of the coal will be done with truck and shovel. The coal is processed, washed and crushed to 1 ¼ " for delivery to the power plant. The muddy water created by the washing process is chemically treated and the solids removed by a thickener-vacuum filter system.

Jan 1, 1971

By Elmer H. Citron

The Monterey Coal Company, a wholly owned subsidiary of The Carter Oil Company of Houston, Texas, recently awarded a contract to build the Monterey No. 1 Preparation Plant, which will be located near Carlinville, Illinois, to the McNally Pittsburg Manufacturing Corporation. The preparation plant will be one of the most modern jig preparation plants in the State of Illinois. Illinois No. 6 coal will be delivered by a 48" slope belt having 35-degree carrying idlers, operating at 650 FPM and operating on an incline of nearly 17 degrees. The slope belt is driven by two 450 BP motors and will deliver the mined coal at the rate of 1800 TPH to a stationary step-type ' screen having six-inch openings. The overproduct from the stationary screen will discharge into a 10' x 24' McNally Rotary Breaker and the underproduct will bypass the rotary breaker. The rotary breaker is equipped with heavy manganese screens with 6 ½" diameter openings. The sized 6" x 0 coal from the stationary step screen and the rotary breaker will be discharged by gravity to a 4%" raw coal incline storage belt. The 48" raw coal storage belt will deliver the 6" x 0 sized coal at the rate of 1800 T?H to an 8000 ton concrete raw coal silo, 70 feet in diameter and 1x7 feet high. The 6" x 0 sized coal will be drawn from the concrete storage silo by seven (7) 36" wide reciprocating plate feeders. Five feeders are to have constant speed drives at a rated capacity of 145 TPH each. Two (2) feeders will have remote con- trolled variable speed drives, to provide a varying feed.

Jan 1, 1970

By Jon C. Yingling

A control system for reducing peak loads on mine conveyor networks is proposed. The system involves continuous monitoring of flows into the conveyor network and section feeders with variable discharge capability. The control strategy uses differences in material transit times from various production units to the transfer points in the network to forecast load levels that will be intersected by the discharge of the controlled feeders over future time intervals. Feeder setpoint adjustments are made to prevent excessive peaks in the combined flows at the transfer points, while maintaining adequate feeder throughflow to avoid production bottlenecks.

Jan 1, 1992

By Jens Classen, Martino Gatti

"GENERAL PROJECT DATAThe Sparvo Tunnel is part of the “Variante di Valico” project, a new 65km (40mi) highway link between Bologna and Florence, designed to renew the existing A1 route which connects Bologna and Florence. It is composed of 2 x 2,5km (1,55mi) twin tunnels excavated inside a complex geological context, characterized by the presence of a series of surface landslides and methane gas.The inclination is 2,6 % and the minimum radius is 1.400m (4.600ft). The two tunnel tubes were excavated by a 15,6m (51ft) diameter EPB-TBM from summer 2011 to summer 2013. The internal diameter of the concrete lining is 13,6m (44,5ft), made up of precast concrete segments 2m (6,5ft) long and 70cm (28”) thick, being the largest and heaviest segments ever produced. Excavation went smooth apart from initial problems concerning clogging problems and a main bearing damage of the TBM, which are discussed in a further paper during the congress. No structural damage was observed until the end of the TBM drives.SITE OBSERVATIONS DURING THE POST-EXCAVATION PHASEThe first evidence of the stress state of the segmental lining appeared on November 20, 2013 inside the northern tube (1st drive). Ring No. 884 showed obvious signs of overstress, resulting in spalling and crushing of the concrete, and the subsequent bending of the reinforcement, as visible in figure 1.At this time the tunnel invert was already partially backfilled, and construction traffic was running through the tunnel. First safety measures included closing of the tunnel, and initial measurements of the displacements."

Jan 1, 2016

By C. Cho

One problem with the application of neural networks to the multivariable control of mineral and extractive processing is deciding whether and how best to use them. The objective was to compare neural network control to more conventional strategies and find any advantages in terms of set point tracking, rise time, settling time, disturbance rejection, and other criteria. The procedure was to develop neural network controllers using both historic plant data and simulation models. We tried various neural network control patterns including both inverse and direct neural network plant models. These were compared to state space controllers that were, by nature, linear. For grinding and leaching circuits, a nonlinear neural network-based model predictive control strategy was superior to a state space-based linear quadratic Gaussian controller. We also learned the importance of incorporating state space into neural networks by making them recurrent--feeding certain output state variables into input nodes in the neural network. We concluded that neural network controllers can have better disturbance rejection, set point tracking, rise time, settling time, lower set point overshoot, and can be more foolproof and easy to implement in complex, multivariable plants.

Jan 1, 1996