Search Documents

By Ivo Dreiseitl, Kondratenko. Anatoly

"The second prime area within the Interoceanmetal Joint Organization exploration area named H22 was delineated in 2013 with the goal to ensure sufficient nodule reserves for profitable mining operations and processing.The knowledge base for area H22 has been obtained from archive materials. The geological situation can be characterized as follows: the sea bottom is represented by tectonic blocks oriented in a sub-meridional direction. According to swath mapping, the depressions (troughs) are contoured at depths of 4500-4600 m while the tops of horsts occur at 4300-4350 m. Volcanic relief occupies 6% of the H22 area; in the eastern part a multicone volcano was detected. Its area at the base is 153 km2. The maximum height of a cone above the base is 850 m. The crystalline basement composed by tholeiitic basalts of late Oligocene age is covered by about 100 m of sediments. Semi-consolidated to consolidated calcareous nanofossil sediments of the Marquesas formation are deposited on the basement. Higher up in the profile sediments of the Clipperton formation occur. They are represented by unconsolidated siliceous (Radiolaria, Diatom) ooze. Sediments containing calcium carbonate have not been detected. Where opaline silica (or CaCO3 if any) content is below 5%, the sediments belong to zeolitic clays or red pelagic clays. Grain size composition of the seabed and the near-seabed sediments in H22 classifies them into clayey silts."

Sep 1, 2014

By Wang Xuguang, Zhang Xiaozhi, Li Guozhong, Xiong Daiyu

The emulsion explosive is a new kind of water-based industrial explosives that was developed at the end of 1960s. In the 1970s, BGRIMM invented EL-series emulsion explosive that created a new era of water-based industrial explosives technology and products in China. From single type to varieties, the perfect BGRIMM emulsion explosives technology has been formed, which consists of 9 series, over 30 types explosives and related production equipments and processes such as cap-sensitive rock emulsion explosive, opencast emulsion explosive, coal permitted emulsion explosives, low-temperature-resistant emulsion explosive, high-temperature-resistant emulsion explosive, high power emulsion explosive, bulk emulsion explosive for loading truck and powder emulsion explosive. From the early 1990s, BGRIMM successfully developed opencast emulsion explosives and the computer controlled continuous production line of small-diameter cap-sensitive emulsion explosive which raised emulsion explosive production technology of our country to a new lever. With the successful development of powder emulsion explosive and related computer controlled continuous production line, on-site mixed emulsion explosive and loading truck as well, the category of industrial explosives is getting more complete and reasonable. This paper pres ents the history, achievement and development of BGRIMM emulsion explosive technology, also forecasts the future of linked technology.

Jan 1, 2003

By D. Denton, M. Stickney, R. Langston

The National Institute for Occupational Safety and Health and the Stillwater Mining Company worked cooperatively with the Montana Bureau of Mines and Geology to develop a system that would collect seismic data at the Stillwater and East Boulder mines. The purpose was to obtain baseline information on the magnitude and location of mining-induced seismicity to determine if the mines needed multi-channel in-mine monitoring systems. Seismic data recorded at field sites near the mines are being telemetered via FM radio through a series of repeaters to a central recording site where the seismic signals are digitized using an Earthworm data acquisition system. The Earthworm system performs several data analysis tasks in near-real time and places raw seismic data, preliminary hypocenter locations, and magnitudes on a Website within 5 min of a seismic event. Such rapid access to seismic data allows personnel at the Stillwater Mine, Spokane Research Laboratory, and Earthquake Studies Office to evaluate seismic events quickly and respond in ways that may improve the safety of mine personnel underground. Installation of the system also broadened earthquake coverage to south-central Montana, a region previously not covered by the seismograph network.

By A. Korchevskiy

While Russia continues to be a worldwide leader in asbestos production, the environmental and health status around Russian asbestos mines is not well known and understood, internationally. A literature review was performed to evaluate the potential impact of Russian mining operations on mesothelioma incidence rates in the Ural region where the majority of the asbestos deposits are located. Asbestos that has been mined in this region includes chrysotile, crocidolite, and anthophyllite forms. The Nicholson-Peto (1986) model was applied to published exposure data and mesothelioma mortality statistics in different cities and villages of the Ural region to estimate mesothelioma potency factors (KM) for the various associated fiber types. It was demonstrated that mesothelioma potency factors, as determined by Berman and Crump (2008), could be used to predict mesothelioma risks of non-occupationally exposed individuals in locations where either pure chrysotile, or both crocidolite and chrysotile, are mined and milled. A mesothelioma potency factor was calculated for anthophyllite that appeared to be higher than expected based on previously published studies. The highest incidence of mesothelioma was identified in a community close to a crocidolite mine. The study confirms the validity of utilizing current risk models for non-occupationally exposed populations near asbestos mining activities where asbestos exposures are relatively low and provides methods of characterizing mesothelioma risk in the former Soviet republics (FSRs).

Feb 27, 2013

By Merete Tangstad, Kai Tang

The primary aim of the present investigation is to develop a model which permits to calculate/predict the viscosities of ferromanganese slags, typically focus on the MnO-SiO2-Al2O3-CaO-MgO-FeO multicomponent system. The present viscosity model emphasizes the ability of prediction, i.e. using the binary parameters to extrapolate the viscosities in ternary and higher order systems. To this end, a physical sound model is essential. Temperature dependence of viscosity is described by the Eyring?s formalism. The well-known cell model is used to define the SiO44- tetrahedra in the silicate melts. Composition dependence is expressed in term of viscous activation energy, which is characterized by the binary structural units. The viscosities in ternary and higher order systems are hence predicted using the binary activation parameters. The model has been used to calculate the viscosities in the whole liquid composition range of ferromanganese slag.

Jan 1, 2007

By E. E. Ungar

The ways in which noise is generated and radiated by chutes and screens are described and corresponding noise control concepts are discussed. The amounts of noise reduction produced by applicable noise control means are delineated and generally cost-effective means are recommended. Available noise control materials and their applications are discussed. The results of field measurements, laboratory experiments, theoretical analyses, and studies of data provided by suppliers are summarized.

Jan 1, 1975

By Cyril Wells, Paul E. Busby, Donald P. Hart

EARLY workers in the field have established that the diffusion of nitrogen follows normal diffusion laws. Concentration-penetration data from layer analyses of reasonably pure iron specimens nitrided in an atmosphere of CO-NH, are given in the literature,' and were utilized to calculate diffusion constants for nitrogen. The diffusion of nitrogen in y-iron may be represented approximately by the equation D - 1.07x10-' e-:".'m'"'r The diflusion coefficients, D, given by this relation represent mean values, since carburization occurred simultaneously with nitriding. The rate of diffusion in a-iron has been determined from calculations based on data obtained by measuring the internal friction of samples at each of several frequencies and temperatures.' The expression for the diffusion of nitrogen in a-iron derived from these experiments is D= 1.4x10-1 e-it tudent Increasing carbon content decreases the diffusion of nitrogen in a-iron.3 Oxygen, either dissolved or combined, has been shown' to exert a marked influence on the diffusion of nitrogen from a nitriding atmosphere. D values increase rapidly with increasing amount of dissolved oxide in the metal, but diffusion is completely inhibited if preliminary oxidation has produced the slightest trace of iron oxide on the surface. Diffusion coefficients have also been calculated from permeability and solubility data,' but these values are of questionable accuracy since the boundary conditions assumed may not actually exist. Experimental Method The method selected for determining the diffusion coefficient of nitrogen in a-iron in the present investigation was the so-called fractional saturation method. This method consists of heating the specimen in an atmosphere of H2-NH8 mixtures of known composition for a predetermined time. Following this treatment, the specimen is analyzed for nitrogen; and the diffusion coefficient is calculated from a known relationship of the fractional saturation, time of diffusion, and dimensions of the specimen. This relation is based on a solution of Fick's second law and assumes that the surface of the specimen immediately reaches a given concentration which remains constant during the diffusion anneal. Tabular solutions are available4-5 and provide the relationship between fractional saturation and the quantity \/Dt /a, where D is the diffusion coefficient in cm2 per sec, t is the time of the diffusion anneal in seconds, and a, in the present experiments, is the half-thickness of the slab specimen in cm. The method is adequately discussed elsewhere.9-10 Material The material for this study consisted of 3/32 in. thick (0.24 cm) carbonyl iron plate. The major impurities in this iron are oxygen and carbon, which are readily removed by long time high temperature treatment in an atmosphere of purified hydrogen. In order to reduce the diffusion anneal time for saturation runs, some of this material was rolled to a thickness of 0.015 in. (0.038 cm) prior to the hydrogen treatment. Hydrogen treatments were carried out in a hydrogen-purified Armco iron tube heated in a global

Jan 1, 1957

By Turnbull A. G

Thermodynamic calculations based on the ideal system NiO - Fe203 - H2(g) - CO(g) - H2O(g) - CO2(g) show that it is not possible to selectively reduce nickel oxide in the presence of hematite because ferronickel is the zero-valent nickel-containing product. The calculations show that temperature, reducing atmosphere composition and oxide to reductant ratio have a marked effect on the nickel content of the ferronickel alloy and the amount of nickel reporting jn the alloy. The general trends based on the calculations correlltte well with experimental observations.

Jan 1, 1980

By J. van der Merwe, D. Tharandt

The corrosion behaviour of 304L stainless steel laser-cladded with various amounts of ruthenium (Ru) was evaluated in solutions of sulphuric acid and sulphuric acid plus sodium chloride at 25°C and 45°C by open-circuit potential and cyclic potentiodynamic polarization tests. In general, the addition of Ru to the stainless steel increased its corrosion resistance in 1 M H2SO4, as well as in 1 M H2SO4 plus 1 wt% NaCl. This was observed for a number of parameters such as corrosion rate, corrosion potential, open-circuit potential, and current density. However, increasing the amount of Ru added beyond a certain level did not result in further improvement in corrosion protection. For each environment there is an optimal Ru concentration for the best corrosion protection. For example, in 1 M H2SO4 at 25°C, 2.44 wt% Ru shows the least active surface in terms of corrosion. Further research into ruthenium coatings on stainless steels is recommended.

Jan 1, 2015

The clay slime generatea in many sand washing operations is usually very difficult to de-water. Treatment of these slimes is often limited by economic considerations and the properties of the solids to a gravity-driven process. The moisture characteristics of kaolinitic slimes from two different sand washing operations were measured.From these curves, equilibrium moisture contents for sedimentation and for drainage were calculated using the methods of soil physics. The validity of these calculations were confirmed by column experiments.The results show that the increased de-watering to be gained by drainage, as compared with sedimentation, is significant. Furthermore, observed rates of outflow indicate that water can be removed considerably morerapidly when drainage is possible.The practical application of these results to the dewatering of clay slimes is discussed.

Jan 1, 1978

By Pengfu Tan

Electric arc furnace (EAF) dust from the steel industries is listed by the United Sates EPA as a hazardous waste under the regulations of the Resource Conservation and Recovery Act due to the presence of lead, cadmium and chromium. The disposal of the approximately 650,000 tons of EAF dust per year in the U.S. and Canada is an expensive and unresolved problem for the majority of steel companies. The Waelz process has been considered as the best process for treating EAF dust. A process model, combined thermodynamic modeling with heat transfer calculations, has been developed to simulate the chemical reactions, mass and heat transfer and heat balance in the kiln. The injection of air into the slag and the temperature profile along the kiln has been modeled. The effect of (CaO+MgO)/SiO2 on the solidus temperature of slag has been modeled also and discussed. Some optimizing results have been presented in this paper.

Jan 1, 2005

By C. Guo

The Mildred Lake Settling Basin (MLSB) is the largest disposal site for mature fine tailings (MFT) at the Syncrude Canada Ltd oil sands plant. Over the past years (since 1997), there has been a marked change in the consolidation behaviour of MFT in the MLSB, which is attributed to microbial activity. Methane-producing microorganisms, known as methanogens, have become very active, and large amounts of biogas (mainly methane) have been produced. In certain regions within the MLSB, gas bubbles are released to the water surface of the tailings pond. Continued field monitoring of the MLSB has provided convincing evidence of the rapid densification process (rapid water drainage from the tailings). This phenomenon contradicts the consolidation models for MFT developed over the past 20 years. The above phenomenon has also been reported at the oil sands tailings ponds operated by Suncor Energy Ltd. in Fort McMurray. This rapid densification has caused pumping challenges in the transfer of fine tailings from the Mildred Lake Settling Basin for the creation of composite tailings. The rapid densification of the MFT may also have potential positive effects in accelerating the reclamation of the oil sands fine tailings. A field and laboratory research program is underway to study the mechanism leading to the rapid densification phenomenon. Systematic field investigations were performed to determine the distribution and characteristics of the rapidly densified MFT. Gas bubble distribution on the water surface was mapped to determine the zones of different biological activities in the pond. These field investigations provided solid ground for further experimental and theoretical research. A number of small-scale column tests were carried out to observe the gas evolution and to measure the changes of some geotechnical parameters under different microbial activity. Also, a series of gassy MFT densification tests are being conducted to study the mechanism of the rapid densification of the MFT under microbiological activity. A review and discussion of the research program is given and some preliminary research results are presented in the paper.

Apr 1, 2005

By L. Nichols, H. Keele, K. Czajewski

"The oil sands industry in northern Alberta disposes of the by-products of bitumen extraction into tailings ponds. The mix of tailings sand and mature fines, with gypsum added as a flocculent, constitute a waste stream called consolidated tailings or composite tails (CT). The two main challenges of the tailings disposal process are a chronic shortage of economical containment and the potential mobility of CT deposits in the event of breaching by, for example, long-term erosion, which makes abandonment of tailings ponds more complicated. Thick deposits of CT do not consolidate in adequate time to provide reassurance for abandonment of the CT containment facilities and that is why the containment must be adequately robust or the properties of CT improved for long-term stability. This paper addresses potential solutions to these problems by applying the gravity drainage of CT and quasi-homogeneous construction for containment facilities.Gravity drainage is suggested as a mechanism to facilitate the consolidation of CT in critical areas of the ponds by inserting internal sand layers within the body of the deposit. Critical CT areas are defined as areas around the perimeter of containment. The advantages of this solution are longterm stability of the disposal area by creation of stiff deposits around its perimeter and facilitation of reclamation efforts on its surface.The present solution for containment of CT deposits is not favourable from a materials balance point of view. In general, sand has become a valuable resource as it both forms the primary component of the CT matrix and becomes the containment building material. If sand is required for CT, less is available for dyke construction. This puts additional pressure on the industry as a whole to maximize the use of overburden and interburden for construction of tailings containment structures. Typically, only 50 percent of dry mine waste (overburden + interburden) is incorporated into dyke construction. The remainder of dry waste is stored either in waste dumps or in tailings ponds, consuming valuable in-pond space originally allocated to CT material. Quasi-homogeneous dykes are proposed to contain CT deposits rather than traditional minimum cross-section dykes built of overburden and/or cell sand. The quasi-homogeneous dyke may change the unfavourable material balance and lessen the construction constraints that are presently imposed on the mine/waste disposal process."

Jan 1, 2008

By Jacob Carr, Christopher C. Jobes, Chenming Zhou, LaTasha R. Swanson, Jennica Bellanca, Jingcheng Li

Magnetic proximity detection systems (PDSs) are used with continuous mining machines (CMMs) to protect miners from striking and pinning accidents. Generators are used in a PDS to create magnetic fields covering the space around a CMM. The PDS determines the proximity of a miner relative to the CMM based on the magnetic flux density detected by a miner-wearable component (MWC) and simultaneously alerts the miner and stops the motion of the CMM if the miner is within a proximity that creates a striking hazard. A stable magnetic field is essential to the accuracy of the proximity calculations performed by the PDS. This paper presents the results of a systematic study of the magnetic influence of two types of steel structures found near a CMM— the body of the CMM itself and the wire mesh used for roof and rib control. The results show that the steel of the CMM body can change the magnetic field distribution and also alter electrical parameters of a PDS by changing its generator current. The study also shows that, depending on the distance between the wire mesh and a generator, the magnetic field can also be altered.

By Jacob Carr, Christopher C. Jobes, Chenming Zhou, LaTasha R. Swanson, Jennica Bellanca, Jingcheng Li

Magnetic proximity detection systems (PDSs) are used with continuous mining machines (CMMs) to protect miners from striking and pinning accidents. Generators are used in a PDS to create magnetic fields covering the space around a CMM. The PDS determines the proximity of a miner relative to the CMM based on the magnetic flux density detected by a miner-wearable component (MWC) and simultaneously alerts the miner and stops the motion of the CMM if the miner is within a proximity that creates a striking hazard. A stable magnetic field is essential to the accuracy of the proximity calculations performed by the PDS. This paper presents the results of a systematic study of the magnetic influence of two types of steel structures found near a CMM— the body of the CMM itself and the wire mesh used for roof and rib control. The results show that the steel of the CMM body can change the magnetic field distribution and also alter electrical parameters of a PDS by changing its generator current. The study also shows that, depending on the distance between the wire mesh and a generator, the magnetic field can also be altered.

By L. Salnais

"Aluminum alloy structural components f or aircraft designed prior to the early 1970s, were machined from 2024, 7075 and 7079 plate, bar and forgings in either the T3xxx of the T6xxx tempers. Protective treatments for these parts consisted of chromic acid anodizing and alkyd zinc chromate primer, plus a nitrocellulose lacquer for exterior parts.The major service problem with these parts was the frequent occurrence of exfoliation corrosion and stress corrosion cracking (S. C. C.). Therefore, 2024 and 7079 alloys, with min or exceptions, were not used in aircraft designed after the early 1970s. Replacement materials were 7075 and 7050 alloys in the overaged, exfoliation corrosion and S. C. C. resistant, T73xxx tempers.The use of improved materials together with the upgrading to chromated epoxy primers, polyurethane enamels and the introduction of water displacing corrosion inhibiting compounds have significantly reduced corrosion in aircraft. IntroductionThe prevention of corrosion in aircraft has been a challenge to aircraft designers for many decades. With the design of the first all-metal aircraft, the designer and builder were thrust into the relatively new field of protective treatments as corrosion started to become more prevalent. This resulted from the continually increasing importance and wider utilization of aircraft, and their exposure to a wider variety of ho stile environments such as urban industrial pollution, high humidity tropical areas and salt ladened marine environments.Corrosion advanced another significant step when, after World War II, designs of new and larger aircraft utilized larger and thicker metal sections in the manufacture of structural components. With thicker mill products came higher residual heat treat, fabrication and assembly stresses resulting in an increased number of stress corrosion and exfoliation corrosion problems.Apart from the exterior environment, aircraft interior s can be exposed to a number of fluids which are either directly or indirectly responsible for corrosion. The most common of the fluids is water which is usually introduced as a result of condensation, especially in pressurized fuselages, and spillage from lavatories, galley s and electrolytes from batteries. These fluid s have a degrading effect on protective treatments and are also corrosive to the substrate metal. Fluids such as phosphate ester hydraulic fluids and ester based turbine engine lubricating oil s are not corrosive to the substrate metal, albeit they do contribute to corrosion by the degrading effect on protective treatments."

Jan 1, 1988

By Harry Schnabel

A 6200 m² cement deep soil mixed (CDSM) wall was constructed for an addition to the Museum of Fine Arts in Boston, Massachusetts. The addition was located along the east side of the existing museum within an existing courtyard and in the area of a recently demolished structure. The basement for the new addition extended up to 9.1 m below the floor slabs of the adjacent buildings. The adjacent buildings are supported on a variety of foundations including; shallow spread footings on alluvial sand deposits, low capacity piles bearing in the alluvial sands and Boston blue clay, and caissons bearing on the alluvial sands and clay. Groundwater was less than one meter below the top of the CDSM wall. The new addition was founded on shallow foundations in the desiccated crust of the Boston Blue Clay. The CDSM wall was selected as the most economical means to control deep seated ground movements, limit movements of adjacent structures, and inhibit groundwater flow into the excavation. Braces and a continuous wale were used to support the CDSM wall where it was located adjacent to structures. Tiebacks were used where there were no structures. Other geotechnical construction at the site included: ? Jet grouting (JG) to improve the ground and provide an impervious subgrade layer at the west end of the addition, ? Concrete pier underpinning was used to support an existing structure on the west end of the site, and ? A secant pile wall using alternating drilled shafts and jet grout columns was used for earth support and a barrier to groundwater flow to allow for the installation of a storage tank at the south end of the site. Inclinometers, total station surveying, vibrating wire strain gauges on the braces and wales and load cells on the tiebacks were used to monitor performance of the excavation support system. The excavation support system performed better than required limiting lateral movements of the adjacent structures to less than 13 mm and vertical settlement to less than 8 mm. No settlements or movement were detected during the construction of the wall. Lateral movement of the CDSM wall was typically below 15 mm with the maximum movement occurring near subgrade.

Jan 1, 2011

By A L. Welsh, P S. Newling

Mill operators are entrusted to manage what are normally the largest capital assets on any mine site. Achieving maximum utilisation and lowest costs that drive operating margins are realised by applying well known fundamentals of good housekeeping and proactive (as opposed to reactive) maintenance programmes. Historically, maintenance and operations staff are bought in late during the construction stage and often do not have the time, know how or tools to put in place (let alone train the incumbent staff) a functioning maintenance system before commissioning starts. They then get drawn into a reactive maintenance paradigm as teething problems post commissioning are dealt with and day to day issues envelop the line tradesmen and supervisors. Additionally, even when some semblance of a planned maintenance system is established, it is often the case that over time personnel and cultural issues mask the development of problematic maintenance practices. As has been observed by the authors on many occasions, even when some basic programmed maintenance system is in place, the collection of maintenance data is often inaccurate, incomplete and not subject to analysis. The net result is of course a high incidence of machine failure leading to lower than expected availability, which ultimately translates to lower revenue and higher net maintenance costs. The time taken and the cost to repair a failed machine is always greater than the timely and planned repair of a worn machine.Upon review of a number of processing plants around the world, especially gold and small base metal operations, it is apparent that companies need to implement maintenance systems whereby the operators are not overwhelmed with data and can manage by exception. Inadequate or inaccurate data collection and analysis result in poor decisions being made and misguided actions taken. GR Engineering Services (GRES) has developed a system to enable the custodians of mining plant assets to make more effective and effective use of maintenance data by being able to focus on the indicators of plant deterioration and to programme concurrent maintenance activities to maximise plant availability and utilisation.This paper outlines the structures that can be employed to address the challenges of implementing a proactive asset management system. The best practice strategies described herein have been based on the fundamentals of a sound asset management framework:a phased approach to the development of Asset Management methodology and processeseffective implementation of programmed maintenance routinesmanaging asset condition data collected from various sourcesefficient analysis of asset data and management by exceptioninterfacing asset management data with Computerised Maintenance Management System (CMMS) and Electronic Management Systems (EMS) systemstips to assist with maximising asset performance.CITATION:Welsh, A L and Newling, P S, 2015. Operating strategies for custodians of company assets creating wealth, in Proceedings MetPlant 2015, pp 338–352 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Sep 7, 2015

By Rambabu Pothina, Rajive Ganguli

The economics of a mineral processing circuit is dependent on the numerous sensors that are critical to optimization and control systems. The sheer volume of sensors results in mines not being able to keep the sensors calibrated. When a sensor goes off calibration, it results in errors which can severely impact plant economics. Normally, these errors are not detected until the error grows and becomes a large (“gross”) error. Undetected errors are not remedied until the next calibration, which on average are a year apart Beamex (2019) [1] across all industries. Classical statistical methods that depend on linear relationships between input and response variables are typically used to find gross errors, but are not effective for highly non-linear and non-stationary operations in mining and mineral processing industries. Calibration of sensors is time-consuming and generally requires physical intervention that results in equipment downtimes causing production losses. Therefore, in situ detection methods are warranted that do not require physical removal of sensors to compare with standard measurements or well-calibrated sensors. Data-mining based techniques and algorithms have high success rate in tackling such problems. This paper presents results from a groundbreaking multi-year big data research project whose goal was to develop data-mining-based original techniques in a multi-sensor environment to identify when sensors start to stray, rather than wait for errors to grow. The carbon stripping (gold) circuit in the Pogo Mine of Alaska was selected for this project. Maintaining strip vessel temperatures at optimum levels is crucial for maximizing gold recoveries at Pogo, and hence the focus was on the temperature sensors of the two strip vessels. An automated algorithm to detect errors in strip vessel temperature sensors was developed that exploited data from multiple sensors in the strip circuit. This algorithm was able to detect with a high success rate when bias errors of magnitude as low as ± 2% were artificially induced into strip vessel temperature data streams. The detection times of about 1 month were much less than the industry average calibration frequency of 1 year. Thus, the algorithm helps detect errors early, without stopping the circuit. The algorithm alarms can serve as triggers when calibrations are done. Experimental methodology along with results, limitations of algorithms, and future research are also presented in this paper.

By Rambabu Pothina, Rajive Ganguli

The economics of a mineral processing circuit is dependent on the numerous sensors that are critical to optimization and control systems. The sheer volume of sensors results in mines not being able to keep the sensors calibrated. When a sensor goes off calibration, it results in errors which can severely impact plant economics. Normally, these errors are not detected until the error grows and becomes a large (“gross”) error. Undetected errors are not remedied until the next calibration, which on average are a year apart Beamex (2019) [1] across all industries. Classical statistical methods that depend on linear relationships between input and response variables are typically used to find gross errors, but are not effective for highly non-linear and non-stationary operations in mining and mineral processing industries. Calibration of sensors is time-consuming and generally requires physical intervention that results in equipment downtimes causing production losses. Therefore, in situ detection methods are warranted that do not require physical removal of sensors to compare with standard measurements or well-calibrated sensors. Data-mining based techniques and algorithms have high success rate in tackling such problems. This paper presents results from a groundbreaking multi-year big data research project whose goal was to develop data-mining-based original techniques in a multi-sensor environment to identify when sensors start to stray, rather than wait for errors to grow. The carbon stripping (gold) circuit in the Pogo Mine of Alaska was selected for this project. Maintaining strip vessel temperatures at optimum levels is crucial for maximizing gold recoveries at Pogo, and hence the focus was on the temperature sensors of the two strip vessels. An automated algorithm to detect errors in strip vessel temperature sensors was developed that exploited data from multiple sensors in the strip circuit. This algorithm was able to detect with a high success rate when bias errors of magnitude as low as ± 2% were artificially induced into strip vessel temperature data streams. The detection times of about 1 month were much less than the industry average calibration frequency of 1 year. Thus, the algorithm helps detect errors early, without stopping the circuit. The algorithm alarms can serve as triggers when calibrations are done. Experimental methodology along with results, limitations of algorithms, and future research are also presented in this paper.