Search Documents

By M. E. Ryan, C. L. Hoebbel, E. J. Haas

DISCLAIMER The findings and conclusions in this paper are those of the authors and do not necessarily represent the official position of the National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention. Mention of any company or product does not imply endorsement does not constitute endorsement by NIOSH. ABSTRACT It has been suggested that effective disaster prevention and response in mining requires that elements of emergency management be incorporated into everyday operations, planning, and decision-making, much in the same way that the industry approaches routine risk management. Since Congress passed the MINER Act in 2006, the newly mandated structural and technological advancements in mine emergency response preparation have been made and are apparent, but measureable improvement in mineworker emergency preparedness is difficult to ascertain. To begin exploring whether measurable indicators of mine workforce preparedness can or should be addressed within a risk management framework, NIOSH researchers examined the datasets from two separate research efforts—one which measured the organizational safety climate and self-reported safety performance, and the other which measured individuals’ perceptions of their own self-escape competency. INTRODUCTION In response to the mine disasters in 2006, the United States Congress passed the Mine Improvement and New Emergency Response Act (MINER Act), which strengthened existing regulations [30 CFR, Part 48] and introduced new provisions to improve mine emergency response. Since that time, new technological and structural features, such as mandated redundant communication systems, personnel tracking systems, and lifelines in primary and secondary escapeways, have been installed in underground coal mines across the country. Additionally, mine operators are required to develop emergency response plans, provide adequate supplies of breathable air throughout the mine, provide more frequent and realistic expectations training, and are required to immediately notify the Mine Safety and Health Administration (MSHA) in the event of an accident. While the structural and technological advancements in emergency preparedness are readily observable and easily measured, levels of individual mineworker preparedness remain difficult to assess. There is a general consensus that deficiencies in competence in the non-routine tasks required for individual miners to effectively respond to a mine emergency likely remain (e.g., National Research Council [NRC], 2013), and NIOSH research continues address this concern (Haas, et al., 2015; Hoebbel et al., 2018). To maximize an operation’s capacity to effectively prevent disasters and/or respond to emergency events, it has been suggested that individual and emergency management efforts be incorporated into everyday operations, planning, and decision-making in much the same way organizations manage routine risks to health and safety (Haddow et al., 2017). To the authors’ knowledge, the mindful intersection of routine risk management policies and practices and emergency prevention and response readiness have yet to be empirically studied within the mining industry.

Jan 1, 2019

By E. Kapilashrami, S. Seetharaman

Refractory/metal interactions and associated reactions are significant during casting as they can lead to inclusion formation. High-temperature interactions between aluminosilicate refractories and molten steel were studied in the present investigation using the sessile drop method. The changes in the contact angles between molten iron with well-defined oxygen levels and substrates of silica, mullite (3Al2O3?2SiO2) as well as commercial refractories were measured in the dynamic mode. In all cases, the reaction led to slag formation. The experiments were performed under isothermal conditions at 1823 and 1873 K. A known and constant oxygen partial pressure was imposed above the iron drop, through appropriate CO-CO2-Ar mixtures. For all substrates, the contact angles started decreasing due to the lowering of the surface tension of iron, as oxygen was imposed on to the system. When a critical level of oxygen was reached, slag started forming at the drop/substrate interface and at this stage, the contact angle dropped suddenly. Subsequent behaviour of silica and mullite was observed to be different, with SiO2showing a non-wetting behaviour. In the case of SiO2substrate, there were also deep erosions along the periphery of the drops and there was slag accumulation at the sides of the metal drop, probably due to Marangoni flow and/or SiO(g) formation. SEM and EDS analysis of the substrates revealed that the formed slag was fayalite. The thermodynamic criteria for slag formation and possible mechanisms of the reaction are discussed in the light of the experiments. Key words: silica, mullite, fayalite, corundum, slag formation, sessile drop.

Jan 1, 2004

By William H. Schiffbauer, Timothy Lutz, Michael R. Yenchek

The National Institute for Occupational Safety and Health (NIOSH) Pittsburgh Research Laboratory (PRL) participated in an experiment to facilitate the remote control of dozers used on coal stockpiles. The experiment consisted of supplementing the normal sensory cues an operator requires to manipulate the dozer from a remote location. NIOSH provided and tested a vision systems intended to give the operator the ability to operate the dozer safely and efficiently from a remote position. This paper highlights the genesis of the experiment, identifies the contributions of Consol Energy and Caterpillar, and details the NIOSH efforts and successful test results. Keywords-remote-control; vision system; stockpiles; safety

A Simulation Study on Two-Stage Grinding with High-Pressure Roller Mills By the application of high-pressure roller mill (HPRM) in the field of mineral processing, computer simulation can be used to predict overall performance of a grinding circuit based on data obtained from HPRM comminution test in pilot scale or industrial scale. Using the model developed by Klymowsky and Liu for HPRM comminution and an extended Plitt's model for classification, steady-state of a closed circuit grinding can be calculated by means of a virtual locked-cycle-test conducted on computer. This paper presents the result of a simulation study on two-stage grinding with HPRM as the comminution device. The emphasis is put on the influences of circuit-configuration and way of recirculating classifier's coarse-product on the overall circuit performance. The results show that two-stage grinding with HPRM has its merit and flexibility in meeting demands of various grinding tasks in mineral processing.

Sep 13, 2010

By Frederick M. Fowkes

Dusts can be filtered out of air streams onto polymer surfaces by electrostatic trapping resulting from the spontaneous transfer of electrons upon contact between dust particles and the polymer surfaces. Such spontaneous contact electrification has been studied quantitatively in this contract with the aid of modern semiconductor charge-sensors with the aim of determining what basic principles govern the sign and magnitude of contact electrification. Information obtained by capacitance-voltage measurements of thin polymer films on silicon showed that charges as large as 4 x 1012 electrons per cm2 develop spontaneously, and that such a charge gives adhesional forces of 165 pounds per square inch (11 atmospheres). Direct measurement of dust polymer contact electrification, using silicon surface channel conductance sensors, also showed spontaneous charges of more than 1012 electrons per cm2. Some polymers can either donate or accept electrons. These included chlorinated polyethylene, chlorinated polypropylene, polyethylmethacrylate, and a copolymer of maleic anhydride with n-octadecylvinylether. Other polymers never accepted electrons but could donate electrons; these included polystyrene, pulychlorostyrune, polyvinylchloride, polyvinylacetate, a polycarbonate (Lexan), and cellulose tributyrate. Contact electrification studies were made with dusts (coal dust, a channel black, or a silica powder) contacting thin films of various polymers solvent-cast onto the semiconductor charge-sensor. Sometimes charging was instantaneous and sometimes slower, but in all but one case (coal dust on a copolymer of maleic anhydride with n-octadecylvinyl ether) the charge trunsfer was a donation of electrons by the polymer to the dust. Some

Jan 1, 1973

By V. Velicka, J. Ptacek, R. Snuparek, P. Konecny, G. Takla

Coal mining in Ostrava-Karvina coalfield is being accompanied by rockbursts. 82 rockbursts occurred within the period between 1990 and 2002. Natural and technological factors from these rockbursts were collected and analysed. They confirmed that rockbursts are complex geomechanical events and that the relations between co-acting technological and natural factors are very complicated. Der Abbau in dem Ostrava-Karvina Steinkohlenrevier wurde von Anfang an durch Gebirgsschläge begleitet. 82 Gebirgsschläge in dem Zeitabshnitt von 1990 bis 2002 erfolgten. Die Angaben über natürliche und technische Faktoren der Gebirgsschläge wurden gesammelt und analysiert. Es wurde bestätigt, dass die Gebirgsschläge komplexe geomechanische Phänomene sind und dass die Beziehungen zwischen den zusammenwirkenden Faktoren sehr kompliziert sind. L´abattage d´houille dans le Bassin houiller d´Ostrava-Karvina est accompagnée par les coups de terrains. 82 coups de terrains s´étaient produit dans la période entre 1990-2002. Les données sur les facteurs naturels et techniques ont été ramassées et analysées. Les résultats ont corroboré que les coups de terrains sont des phénomènes géoméchaniques complexes. Les relations entre les facteurs techniques et naturels coopérants sont très compliqués.

Jan 1, 2003

By D. W. Alexander

Researchers from the Office of Mine Safety and Health Research (OMSHR) of the National Institute for Occupational Safety and Health (NIOSH) completed field trials during coal mine rescue contests using simulated gas detectors in place of placards. The gas monitor simulator (GMS) devices replaced static, paper placards to deliver gas concentration data. Eleven teams were observed during the field contest, which was approved by the Mine Safety and Health Administration (MSHA). GMS devices were given to three team members, the captain, and both gasmen. Observation of teams revealed that team members changed behavior by acknowledging and sharing information more often while using the GMS. As a result of this research, it has been demonstrated that under the existing contest format, the newly developed GMS device may be used with no adverse impact on team contest performance. Furthermore, when GMS technology advantages are used, contest realism is improved. In the future, competency assessment may be incorporated.

Jan 1, 2011

By S. Yalcin, S. Chang, M. Oliazadeh

"Present day ferrochrome demand is on the rise, as more steel products require this material as an alloying element. It is projected that the stainless steel demands will grow by about 6% per year to 2020; as such, the supply and demand for ferrochrome appears promising. Chromite is the only economical source of chromium, which is commonly upgraded by gravity techniques and used to produce ferrochrome. The gravity separation methods are not able to treat particles finer than 0.1 mm efficiently. Therefore, in processing chromite ore, a large quantity of fine tailings; some 25% of the mineral values have been reported as lost in slimes during processing. The chromite content of these tailings is typically 10-20% Cr2O3 and accumulates into a large stockpile/pond. A chromite concentrator project in Canada, considers a cut-off grade of 20% Cr2O3. This indicates further processing might be required to recover chromite from low-grade stockpile when the high-grade deposits are depleted. Low-grade ores normally indicates disseminated chromite ores with low liberation and consequently fine comminution. This paper reviews the techniques for treating chromite from fine material and provides some examples of successful operations in fine chromite recovery. Beneficiation techniques for fine particles include gravity separation, magnetic separation, and flotation. Several case studies are presented here that used one of the aforementioned methods or a combination of them.INTRODUCTIONChromium is one of the essential element and important material for production of ferrochrome alloys and special steel. Chromium occurs as chromium spinel, a complex mineral with chemical composition of (Fe, Mg) O (Cr, Al, Fe)2 . Chromite ore is exclusively found in ultramafic igneous rocks. Major associated gangue minerals are talc, quartz, serpentine and iron oxides.Chromite is a heavy mineral, with a specific gravity in the range of 4-5, relative to associated gangue minerals in a chromite ore, which makes gravity concentration a suitable process. This is generally true for coarse particles, ranging from 1 mm to 100 mm. Typically, the ore is prepared by crushing (and grinding if necessary) to a suitable size and the material is sized for the appropriate beneficiation techniques. It is common to use heavy medium separation (e.g. dense media drums, dense media cyclones, or Dyna Whirlpool) to treat coarse size and intermediate size particles (1 – 100 mm). Intermediate and fine particles (20 – 0.1 mm) could be treated by jigs. Spirals, shaking table, sluices, and Reichert cones can upgrade fine particles. Figure 1 shows a typical chromite beneficiation process flowsheet. As can be seen, it consists of three major sections: feed preparation (crushing, screening), de-sliming, and concentration. Beneficiation recoveries range from 80 to 85%, although the ratio of upgrading is only 2 to 3."

Jan 1, 2013

By J. Sliwinski, J. Oliveira

"The application of geometallurgy and corresponding sample selection can take on several formats within industry. Current practice can range from a strictly geostatistical approach to one in which geological and metallurgical parameters are used in combination to create a representative sample selection strategy. The stage of a project and the level of information available at the time of sampling influences the methodology used to create a representative sample.A geometallurgical study of Carpathian Gold’s Rovina Valley Project in Romania is used to demonstrate a sample selection strategy that integrates geological, mineralogical and metallurgical data to maximize the relevance of the geometallurgical unit definitions. This approach, practiced by Xstrata Process Support for its’ clients, ensures that the results of the metallurgical program can be interpreted in the context of these units.INTRODUCTIONGeometallurgy is a cross disciplinary approach in which metallurgical performance of an ore is linked to intrinsic geological and mineralogical characteristics. The objective of a geometallurgical study is to characterize the range of metallurgical performance of an orebody. This allows the geoscientist or metallurgist to create a robust flowsheet able to treat the full range in variability and to develop production strategies to maximize financial performance. The standard approach is to individually sample and test a series of ore types or geometallurgical units. Several different approaches are used in industry to define what constitutes geometallurgical units and how to sample them.One current methodology is the use of multivariate statistics. Geochemical data along with available geological data are used to define multiple variability domains (Williams, 2004; Walters, 2009). There are risks associated with this methodology since the extent of the data set itself may not support a purely statistical approach at the time that metallurgical units need to be defined. The assay suite collected by exploration programs is often limited to pay metals while metallurgical performance is often affected by ore gangue components. Furthermore, in many cases metallurgical scoping and pre-feasibility studies take place prior to the availability of extensive geochemical data sets. Purely statistical methods can produce multiple populations which require extensive sampling, testing and validation prior to consolidation into a reasonable number of composites for metallurgical testing. Methodologies are therefore required that allow representative sampling and testing of consolidated geometallurgical composites to occur at various phases throughout the development of a mining project."

Jan 1, 2013

By J S. Dunlop

At its most basic level, minerals project management may be broken down into a series of logical chronological phases, described in this paper as: exploration, discovery; scoping and pre-feasibility; feasibility and approval; project financing and commissioning; operation; expansion and closure. Each phase has its own peculiarities and challenges, any one of which can potentially upset the forward progress of a minerals project.Reference is made to each of these project phases based on the authorÆs many years of first-hand experience, in an attempt to identify issues within the project management spectrum where additional caution may be warranted. Left un-addressed, these issues can cause undue delays in the project timeline, ultimately delaying approval at significant project cost.Finally, reference is made to the Mine ManagerÆs Handbook recently published by the AusIMM which is an excellent project management resource, considered to be very relevant for minerals treatment plant managers, whether the context be constructing, operating, expanding or shutting down.CITATION:Dunlop, J S, 2013. Mineral project management - A perspective from four decades in the industry, in Proceedings MetPlant 2013 , pp 13-27 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Jul 15, 2013

By C. D. Thompson

Pile load test programs were carried out in 1984 and 1985 at Stelco's and Dofasco's Steel Making Plants in the Hamilton Bayfront area. 323.9 mm outside diameter by 9.5 mm wall thickness Grade 3 steel closed-end pipe piles were driven to practical refusal at depths of 21 to 30 m in very dense silt till with Delmag D30 open-end diesel hammers. Ultimate bearing capacities of up to 5330 kN were proved by dynamic monitoring and static load tests. Allowable bearing capacities of 1780 kN were achievable on production piles without undue difficulty. It was found that the impact stress at the top of the pile was controlled by the yield strength of the steel rather than a useable upper limit based on the modulus of elasticity. A bearing to impact stress ratio of 1.2 at practical refusal was observed to be reasonable but conservative at this site, and is felt to be generally applicable for piles driven to practical refusal elsewhere. Dynamic monitoring during production pile driving was able to show that apparent relaxation was caused by pre-ignition of the hammer when visual observations indicated that pile driving was proceeding satisfactorily.

Jan 1, 1987

By B. N. Das, P. A. Beck

THE o phases have a complex tetragonal structure with Lo/ao = 0.52 and 30 atoms per unit cell.1 A large number of such phases are now known to occur in various binary systems of the transition elements and they exhibit the following typical features: 2,3 The composition of binary o phases is variable, ranging from approximately A4B to ABI. With only a few exceptions, in the o phases the A-ele-ment comes from the V- or Cr- group of the periodic table, while the B-elements are from the Mn-, Fe-, Co-, or Ni- group. In the first long period the corresponding o phases of the V- and Cr-series have approximately the same compositions. However, as the B-element changes from Mn to Fe to Co to Ni, the mean composition shifts monotonously toward higher A-contents.2-4 For example, if the mean compositions of the various o phases with vanadium are compared with each other, the composition shift mentioned above becomes plainly visible: V 30Mn0 80 Vo.45Feo..,, Vo .,Coo. 45 and Vo.5Nio.3. Similiar composition shifts, although not the same composition values, were observed in u phases formed by second and by third long-period transition elements.3'4 This behavior, very characteristic of o phases, is in rather sharp contrast to the usually quite well-behaved stoichi-ometry of the Laves phases. In the latter, the AB2 stoichiometry predominates, with A the large atom and B the small atom.= The p phase occurs in the four binary systems of Mo or W with Fe or Co. On the basis of their crystal structure determination, Arnfelt and westgren6 ascribed to the p phase in the Co-W system the ideal composition W,Co7. Since all four binary pphases have nearly the same composition, it would seem quite plausible that some definite stoichiometry such as A6B7, might be characteristic of this phase, as AB2 is of the Laves phases. On the other hand, it has been noted,7 that the phases occur at average electron concentrations near to those at which the various cr phases are found. Also, it has been pointed out by Shoemaker and associates8 that structurally the a phases and the p phases are somewhat related to each other; both of these structures may be described in terms of suitable stackings of quasi-hexagonal layers'of atoms. By studying a suitably chosen ter- nary phase diagram, one might expect to be able to find out which one of the two conditions, the analogy with the o phase or the constant ratio between the number of large atoms and the number of small atoms in effect governs the behavior of the p phase. Several isothermal sections of ternary systems containing o phases have been investigated. In all cases where two binary o phases occur, these were found to be connected b a more or less straight ternary o phase field.7' '10 Such elongated o phase fields were also found in ternary systems in which only one of the limiting binaries is known to contain o phase. In this manner a latent o forming tendency could be detected and the virtual composition of a "binary o phase", which does not actually occur as such was approximately determined by extrapolation. The behavior of the phases, including their composition shift, is analogous to that of the classical electron compounds, as described by Hume-Rothery.11 In order to decide whether the p phase is essentially a "size compound" like the Laves phases, where a fixed stoichiometric ratio is quite generally maintained between the "large atoms" and the "small atoms", or whether the p phase is typically an "electron compound" in the same sense as ois, it was decided to investigate an isothermal section of the Mo-Mn-Co system. The Mo-Co system12 and the Mo-Mn system13 each are known to have a binary o phase. It could be therefore expected that a ternary o phase field will connect these two binary cr-phase fields. The p phase is known to occur in the Mo-Co system. If the p phase should turn out to form an elongated phase field extending deep into the ternary system, the following considerations mighto apply. The atomic radius of manganese (1.314for C.N.12) is between the va,lues for Mo (1.40Afor C.N.12) and for Co (1.26Afor C.N.12). If it is assumed that p is a "size compound", and that Mn is substituting for the smaller Co atoms in the p-structure in the ternary system, the y-phase field might be expected to extend parallel to the Mn-Co binary line. If Mn is substituting for the large Mo atoms in the p-structure, the p-phase field may be expected to extend parallel to the Mn-Mo binary line. On the other hand, if the p-phase turns out to be an "electron compound" like o, the ternary p-phase field should extend parallel to the ternary o -phase field connecting (Mo,Mn)owith (Mo, Co) a

Jan 1, 1961

By R. W. Fedors, S. T. Green, G. R. Adams, C. Manepally, L. B. Browning

This paper provides estimates of large-scale temperature gradients that can be used to support modeling of natural convection and cold-trap processes in thermally perturbed drifts. Temperature influences estimates of the quantity and chemistry of water contacting engineered components (e.g., drip shields and waste packages) in a potential high-level waste repository at Yucca Mountain, Nevada. Temperature gradients drive natural convection and cold-trap processes, which could affect the distribution, quantity, and chemistry of liquid-phase water contacting drip shields (waste packages).The confluence of high temperatures, liquid-phase water, and high chloride may lead to corrosion of drip shields and waste packages. An overview of the evolution of environmental conditions in thermally perturbed drifts is presented to provide a context for the estimates of temperature gradients. A mountain-scale conduction-only model is used to estimate temperature gradients and temperature differences for a drift located at the center of a potential repository. Thermohydrologic models and heat transfer algorithms are used to evaluate the reasonableness of the conduction-only results. Estimates of temperature using an in-drift heat transfer algorithm account for the possibility of drift degradation and formation of a rubble pile on the drip shield. Temperature differences between the center and ends of the drift are estimated to rapidly increase to 80 °C soon after repository closure. That temperature difference decreases over time to 15 °C after 10,000 years. Local temperature gradients are generally estimated to be less than 0.2°C/m, though the gradients can be much larger near changes in lithology and near the end of the drift. The estimates of temperature gradients do not account for intermediate- and small-scale variations, such as those influenced by differences in heat loads between waste packages, intra-layer variations in thermal properties of the surrounding fractured rock, or localized drift degradation.

Jan 1, 2004

By P. L. Crouse, S. J. Lubbe, M. J. Kabangu

A selective extraction and separation of zirconium by solvent extraction with 1-octanol is proposed. The process is based on the dissolution of the metal basic carbonates in hydrochloric acid in the presence of potassium fluoride, using 1-octanol as extractant. The effect of several process parameters, including hydrochloric acid and potassium fluoride concentrations, reaction time, and phase ratio between the aqueous and the organic phase was investigated. Zirconium is enriched in the organic phase. The best extraction and separation results were obtained with the basic carbonate dissolved in 10% hydrochloric acid and 1.5 M potassium fluoride. The optimum reaction time was 15 min with an organic to aqueous phase ratio of 2:1. Based on the selectivity of zirconium over hafnium in the organic phase, a McCabe-Thiele diagram was constructed. A near complete stripping of zirconium from the organic phase was achieved using a 3 M sulphuric acid solution. Additional studies are required in order to determine the reaction mechanism and the chemical speciation of the current investigation.

By M. K. Seguin

The study described in this paper is first centered on the determination of the physical properties (gravimetric, magnetic, electrical, and electromagnetic) o f metamorphosed iron ores of the Carol Lake-Labrador City District. Different geophysical surveys available in the Carol Lake-Wabush area were interpreted quantitatively and the dimensions, attitude, and tonnage of the ore bodies investigated were determined. A geophysical-statistical study was undertaken which permitted the determination of the grade of iron, magnetite, and, subsequently, specularite in the ore bodies. A discussion of the theory and logical progression of the study and the application of the results to actual geophysical data is presented. The conclusions, which are of real significance from the economic standpoint, may be predicted from the data obtained from geophysical surveys before any diamond drilling is undertaken. An integrated interpretation of this type is useful in planning future exploration work and selecting the most favorable geological structures. It also serves as a guide to minimize and determine tonnage drilling patterns. The general conclusion that a restricted range of residual gravity and magnetic anomalies are a key to exploration is valid. Indeed large depths to the top of deposits are absent since all the ore bodies are overturned synclinal structures with tops located close to the surface. Certain irregular geometries do occur but they are not common and the alignment of the iron ore bodies is sub parallel so that magnetic anomalies are evenly reduced in the earth's magnetic field vector.

Jan 1, 1975

By Loring A. Wyllie

Preparation, understanding and use of soils reports must be a team effort to be fully effective. The soils engineer must know what is to be built and exactly where. The structural engineer and owner must be fully informed on the foundation problems of the site. There must be frequent team review of information obtained from subsurface investigation and of design as details develop. The soils engineer should be available on a continuing team basis as design proceeds and should be directly involved with inspection and interpretation of site conditions as construction continues.

Jan 1, 1986

Real Time and In Situ Measurement of Elemental Composition Using Lasers

Sep 13, 2010

By Louis Schaffer

The purposes of the research were to assess the applicability of new accident investigation methods to mining accidents, and to adapt those found to be applicable in the development of a new methodology for use in mining. The development included field testing, evaluation and modification of the new methodology. The report is in two volumes. This, volume one, discusses primarily the Phase I work: background research, design of a methodology for field evaluation, and preliminary field tests in Utah mines. Phase I was completed in June, 1980. Also discussed are some activities to prepare for a field evaluation during the period May 1980 -October 1981. Volume 2 discusses the subsequent field tests, modifications of the initial methodology and the results of use of the methodology in two mining companies for a period of approximately eighteen months, and in other companies for selected loss incidents during that period.

Jan 1, 1986

By C. Hellar

A new dual-boom remotely controlled roof bolter was designed, fabricated, and then tested in an underground coal mine. Remote-control functions allowed complete bolting cycles to be performed on two booms with both operators located within protective cabs. Capabilities included 3-to 6-foot-long bolts in 5-to 8-foot seams, limited to bolts 20 inches shorter than seam height. Machine was in compliance with all regulations including dust collection, machine lighting, and permissibility.

Jan 1, 1979

Paper written on project work carried out in partial fulfilment of B.Sc. Eng. (Extractive Metallurgy)degree The concepts behind support vector machines are very interesting both in theory and in practice, as they are based on a universal constructive learning procedure derived from the statistical learning theory developed by Vapnik. In this paper, their application to time series modelling is considered by means of simulated data from an autocatalytic reactor. In particular, reconstruction methods from non-linear dynamics are used to define a state space model for the process. Multivariate embedding techniques are compared to scalar embedding with respect to modelling. Keywords: System identification, Support vector machines, Multilayer perceptrons, Non-linear dynamics

Jan 1, 2003