Search Documents

By W. S. Kirk

World cobalt production in 1987 was estimated to be about 26,600 metric tons (MT) of which Market Economy Country (MEC) production was estimated to be about 22,000 MT. Zaire and Zambia produce about 80 percent of the primary cobalt produced among Market Economy Countries (MEC). Production cutbacks. lowering cobalt grades and mining curtailments in Finland and Zambia in 1987, resulted in lower production than that of 1986. The trend is expected to continue through 1988. Cobalt production from Canadian ores and from Botswana is expected to increase significantly in 1988 as a result of increased nickel production. MEC production in 1988 is estimated to be 22,100 MT, only slightly higher than that of 1987. There are currently no primary cobalt mines. Virtually all cobalt is recovered as a byproduct of copper and/or nickel mining. Zaire, which produces cobalt as a byproduct of copper, is the worlds largest cobalt producer, with over 50 percent of total estimated world cobalt production and nearly 65 percent of MEC production. Zambia follows with about 14 percent and 17 percent, respectively. Zambia's cobalt production also results from copper production.

Jan 1, 1988

By Young-Geun Park

A new decarburization model for the RH system was constructed for calculating the decarburization reaction using a three-dimensional fluid dynamics program. The entire fluid flow in the RH process was calculated by evaluating the shape and volume of the plume zone, formed by Ar gas blown through multiple nozzles on the wall of the snorkel. On the basis of this fluid flow, the decarburizaiton reaction was calculated through establishing the decarburization reaction model, which considers the thermodynamics, reaction kinetics, and the decarburization sites (Ar bubble surface, bath surface, and the CO bubble fomtation at inner sites). The results from this model were compared with a real RH operation results with good agreement. Because it is possible to predict the carbon concentration at various conditions using this model, it may be useful in increasing the decarburizaiton rate. Key words : RH process, plume zone, decarburization, Ar bubble surface, bath surface, inner site, circulation flow rate.

Jan 1, 2001

By Vladimir Panjkovic, Ostrovski. Oieg, Campbell Cripps Clark, John Truelove

"Effects of properties of raw materials and blast on the performance of an iron bath reactor have been examined by a mathematical model. The model consists of three modules: 1) chemical reactions in the gas phase and heat transfer from gas to droplets ejected from the bath; 2) heat losses from the vessel and radiative heat transfer from the freeboard to the bath; and 3) heat and mass balances for the bath. The following parameters of the process are calculated: coal, ore, flux and oxygen consumption, post-combustion ratio and heat-transfer efficiency, temperature and composition of the off-gas, and slag and metal chemistry.IntroductionOne of the anticipated advantages of smelting reduction over conventional ironmaking is the use of lower-grade raw materials. In this work, the effect of-properties of raw materials on the performance of an iron-bath reactor is examined with a numerical model. The model is also used to assess the effect of properties of the blast (flowrate, temperature and oxygen enrichment) and ejected droplets (generation rate, size and ratio of total weights of slag and iron droplets) on the key parameters of the process. In the process coal, oxygen, ore (which may be partly prereduced) and fluxes for specified slag composition are charged to the bath. Slag and iron droplets are ejected from the bath into the gas phase, where they are heated before returning to the bath. Iron droplets undergo decarburisation and reoxidation with CO2 and H2O. Dust is also generated from the bath, and oxygen or preheated air for post combustion is injected from the top of the vessel.The model consists of three modules: 1) reactions in the gas phase and heat transfer from gas to droplets ejected from the bath; 2) heat losses from the vessel and radiative heat transfer from the freeboard to the bath; and 3) heat and mass balances for the bath. The structure of the model and the data exchange between the modules are shown in Fig. 1. The following parameters of the process are calculated: coal, ore, flux and oxygen consumption, postcombustion ratio (PCR), heat transfer efficiency (HTE), off-gas temperature and composition, slag and metal chemistry, and rates of decarburisation and reoxidation of the iron droplets."

Jan 1, 1996

The cleaning of copper containing fayalite slag using a CuFe alloy was studied under an oxygen pressure ranging from 10-7 to 5 X 10-11 atm at 1200º c. The reduction of copper from fayalite slags by Cu-Fe alloys was found to be very effective. For example, at a P 02 of 10-7 atm, the copper content in the slag was reduced from 4.1 to 0.402 wt. percent after 90 min. of reaction time using a Cu- 5.27 wt. percent Fe alloy. The copper reduction rates were found to be independent of P02 in the gas phase when using Cu-Fe alloys which exhibit an iron activity of greater than 0.4. A mechanism for slag c1eaning is proposed based on the concept that the Cu-Fe alloy lowers the oxygen potential of the slag according to Eq.1 1, fol1owed by the reduction of cuprous oxide by wustite as shown in Eq.2. Here X represents element X in the alloy phase while (X) represents the element in the slag phase. The copper and iron transfer behaviour obtained in this study are explained by a ?Stepwise equilibrium model.

Jan 1, 1985

By Mehmet Bu daycı, Onuralp Yücel, Ahmet Turan, Murat Alkan

In this study, iron based alloys (Fe, Fe-Cr-Ni-Mo, etc.) were produced by using metallothermic methods starting from Mill Scales obtained continuous casting processes. During the process, the combustion was realized by ignition and the produced wave propagates throughout the reactant mixture yielding the desired product. In this work, Mill scale used as an iron source, which includes +2, +3 values iron oxides. Aluminum, Magnesium and Silicon were used as reductants. In these metallothermic experiments, different ratios of reductants were examined, and their effects on the metal recovery and alloy compositions were investigated. The samples were characterized by using chemical analysis (AAS), XRD and XRF techniques.

Jan 1, 2016

By Honggang Zhong, Xin Cao, Qijie Zhai, Hongmei Song, Congsen Wu, Cheng Zhang, Jingzheng Ye

"The influence of superheat on the solidification structure and ferrite/austenite phase transformation of a novel lean duplex stainless steel (DSS) was investigated on a setup, which designed to thermally simulate the continuous casting process. The results showed that the macrostructure of the samples all exhibited chill zone, columnar zone and equiaxed zone. The distance from chill zone to position of columnar to equiaxed transition (CET) moved from 51.0 mm to 39.8 mm with an decrease of superheat from 70 °C to 10 °C, and that the grains in the chill zone tended to be finer. In equiaxed zone, the fme and uniform grains were observed when superheat higher than 40 °C; while, coarse grains were found in samples poured at low superheat. Generally, the volume fraction of ferrite increased along the dendritic growth direction, and stabilized in equiaxed zone. The high superheat decreased the volume fraction of ferrite in equiaxed zone, while it had indistinctive influence on chill zone and columnar zone.IntroductionDuplex stainless steel with approximately equal volume of ferrite and austenite, offering an attractive combination of corrosion resistance and mechanical properties. [ l]. The production of DSS, worldwide application in structural engineering[2] and petrochemical industry[3], have increase by more than 100% in the latest decade[4]. But due to the sharp prices increase of nickel and chromium, lean duplex stainless steels which have lower contents of alloy elements were developed. They perform well in both mechanical and corrosion resistance[S, 6] and even have TRlP(Transformation Induced Plasticity) effect[?]. Nonetheless , the mechanical properties of DSS was affected greatly by the solidification structure, such as columnar to equiaxed transition (CET), grain size and phase content[8]. So, the solidification behavior of stainless steel was investigated widely"

Jan 1, 2014

By Shadia J. Ikhmayies

"Tin oxide (SnU2) thin films of thicknesses in the range 100-600 nm were prepared on glass substrates by vacuum evaporation at ambient temperature. The films were characterized by transmittance measurements, X-ray diffraction (XRD), scanning electron microscope (SEM) observations and energy dispersion X-ray analysis (EDAX). It is found that the films have high transmittance and non sharp absorption edges. XRD diffractograms showed that the films are amorphous and the SEM micrographs showed that the surfaces are smooth, uniform and well covered with the material. The EDAX analysis showed that the films are deficient in oxygen. Indirect optical bandgap energy and Urbach tailing in the bandgap was observed, and the width of the tail, which is related with disorder and localized states, was estimated.IntroductionTin oxide is an n-type wide bandgap semiconducting material with a direct bandgap energy of 4 eV and an indirect bandgap of 2.6 eV [1], where inherent oxygen vacancies act as an n-type dopant [2], Tin oxide thin films have been used for transparent electrodes in photoelectric conversion devices namely amorphous silicon solar cells, liquid crystal displays, gas discharge displays, etc. [3]. Perfectly amorphous thin SnO2 films show good electrical response to reducing gases in air and could, hence, be applied to construct semiconductor gas sensors where the metal oxide films function as a monograin-eqivalent active layer [4]. Amorphous SnU2 can also be used in extended gate field-effect transistors (EGFET) as a pH sensor, where the sensitive part is made of the SnO2/Al structure. This device can be used to detect and to quantify any kind of substance that can produce or consume protons, like an enzymatic reaction, therefore showing a large range of applications as biosensors [5].There are different methods to prepare SnO2 films, such as atomic layer deposition (ALD) [6], reactive magnetron sputtering [7], chemical vapor deposition [8, 9], dip coating [10], spray pyrolysis (SP) [3, 11-14], and evaporation [8], The thermal evaporation method was chosen to produce SnO2 thin films in this work because deposition by this technique does not present any compositional problems. In addition, it is a low temperature technique, which increases the probability of producing amorphous films. To increase this probability, the films were deposited at ambient temperature."

Jan 1, 2012

By Robert C. Laudon

Spectral reflectance data for 335 minerals, including 228 minimum-maximum combinations for anisotropic minerals giving a total of 563 data sets, have been collected from the literature and combined into a single database. Quantitative color values for x, y, Y, Ad , and Pe referred to ICI standard i11uminant C have been calculated from the spectral reflectance data and are included in the database. The programs plot spectral reflectance curves together with all quantitative color data for each mineral in the database. The plots and data can be directed to a dot-matrix printer such that a databook of quantitative color and spectral reflectance having one mineral per page can be generated for the ore microscopy laboratory. In addition, by sorting the database and grouping the minerals, groups or individual minerals can be plotted on the chromaticity diagram, and this plot can also be sent to a dot-matrix printer. The programs are written in compiled BASICA for an IBM-PC with graphics capabilities and at least 128KB random access memory (RAM). The databases are accessible either through the BASIC programs or through any commercial database system capable of loading and writing to an ASCII format sequential file. The programs and database are available by sending a blank diskette and self addressed, stamped envelope to the authors.

Jan 1, 1987

By Do-Hyeong Kim, Jong-Jin Pak, Jung-Mock Jang, Min-Kyu Paek

"Thermodynamic relations among Mn, Al and N in liquid iron were studied by measuring the effect of Mn on the solubility product of AlN in liquid Fe-Mn-Al-N alloys in the temperature range from 1773 to 1873 K. The critical Al and N contents for AlN formation were determined by the metal-nitride-gas equilibration technique over wide ranges of melt composition and temperature. The effects of Mn on N solubility in Fe-Mn and Fe-Mn-Al melts containing Mn up to 26 mass% were separately determined in the same temperature range. The experimental results were thermodynamically analyzed using the Wagner's formalism to estimate the critical Al and N contents for the onset of A1N formation. The validity was checked by the observation of inclusion evolution in Fe-20%Mn-1.5%Al-0.6%C-0.3%Si-0.012%N melt during cooling and solidification.IntroductionRecently, high manganese steels such as Twinning Induced Plasticity (TWIP) steels have been developed as next generation steels because of their high formability as well as excellent strength and ductility.[1•2l In these steel grades, a large amount of Mn (15-20 mass%) and Al (1-2 mass%) are added to liquid steels, and such high Mn and Al concentration bring the great challenges to the steelmaking and casting processes. Mn significantly increases the N solubility in liquid iron and the AlN inclusions can be easily formed at such high Al concentrations due to high affinity of Al and Mn with N in liquid steels. Although the AlN inclusion is known to be a dominating and considered as a detrimental phase for the hot ductility of steels,l3l thermodynamics of AlN formation in high Mn-Al alloyed liquid steels was not well known.Moreover, the melting point of these steels decreases down to near 1623 K as increasing Mn content up to 20 mass%. However, most of thermodynamic data has been measured at the steelmaking temperature around 1873 K. Even at steelmaking temperature, there are huge scatters in thermodynamic interactions of Mn and Al on N in liquid iron. The nitrogen solubility in Fe-Mn alloy melts has been measured using the Sieverts' method[4l and the samplinij method[5-7J in the temperature range of 1813 to 1973 K. In the authors' recent study,[8•9 the N solubility in Fe-Al alloy melts was measured using the sampling method in the temperature range of 1823 to 1973 K. However, the simultaneous effect of Mn and Al at high concentrations on Nin liquid Fe-Mn-Al-N alloys is still not known. Also, there remains uncertainty about the temperature dependence of the interaction parameter of Mn on N in liquid Fe-Mn alloys. Therefore, it is worthwhile to check the temperature dependence of this parameter at a low temperature near the melting point of Fe-Mn alloys. There have been few studies on the characterization of non-metallic"

Jan 1, 2014

By Thomas P. Battle

Modern-day direct reduction of iron first developed as a small-scale, low capital and operating cost alternative to the blast furnace. Since commercialization of continuous DR technology in the late 1960s, the market for the products of direct reduction has grown to more than 74 million tonnes in 2012. The initial advantages of DR plants over BF facilities have grown over the years, for several reasons, including the increased size of DR modules, lower energy and emissions (particularly CO2), along with the flexibility to use a number of different reductants. The development of DR technology over the past forty years will be emphasized in this presentation, including recent developments that allow for even more direct sustainability comparisons with the iron blast furnace – and even combine the two technologies for improved synergies. Also briefly discussed will be the possibilities of using direct reduction for non-ferrous ores.

Jan 1, 2014

By James Vaughan

Australia is a major miner of ore that requires hydrometallurgical processing. According to the 2016 US Geological Survey Minerals Commodities Summaries, the country is 1st for aluminium (bauxite) and lithium, 2nd for gold, zinc and cobalt, 4th for nickel and silver, and 6th for copper mining, not to mention its wealth in coal and iron ore. In this paper, examples of recent Australian hydrometallurgical activities are summarised. Then, selected research projects from the University of Queensland hydrometallurgy research group are profiled. The projects profiled are related to fundamental aspects of processing bauxite with organics and reactive silica as well as the development of a synergistic hydro- and pyrometallurgical process for copper. The process context and motivation for the research is introduced, key results are highlighted with the associated relevant references.

By T. A. Venugopalan

The 'Hall-Heroult' Process of reduction of Alumina to Aluminium has remainded basically the same for over 100 years, but the technology has been undergoing drastic changes with respect to selection of materials for Cathode and Anode, and also energy consumption. The latest generation pots have a working life of about 1800-2000 days normally. After the failure of the pots, the Cathode Lining is dug out and disposed of on land. Since the lining contains objectional materials causing pollution, various processes have been tried for re-processing of this waste. However, most of the processes have been unsuccessful from the techno-economic point of view. At Hindalco a new chemical process for converting the fluorine values of the dug out material to Cryolite was developed in 1988 and a" one ton per day Cryolite production plant has been operating since 1989. This process is especially applicable where the integrated Alumina and Smelter Plants exist side by side. The experience in operating this Plant is discussed in detail in the paper.

Jan 1, 1992

By Robert L. Stephens

"It can be argued that one of the most important property of any material is cost, since this property detennines the applications for which a material can be considered. Despite growing demand and production for commodity metals such as copper, nickel, zinc, lead, and tin, the cost of these metals has been steadily decreasing in real terms. The only way that metal miners and producers can remain financially viable in the long term is by continuing to lower their cost curve through the implementation of new metals production technologies. This paper reviews technological developments in the production of copper, nickel, zinc, lead, and tin that have or threaten to redefine the cost curves for these metals.DisclaimerThe opinions expressed in this paper the author's, and are not necessarily those held by the author's employer, Cominco Ltd.IntroductionThis meeting will discuss many novel technologies and processes for producing metals and materials with specific properties that will enhance the applications ranges for the material or will allow entirely new applications to be developed. However, in every real-world case, the cost of the new material or process is an important property that determines which applications the material or process can be used in. Other properties that are important depend on the applications.A recent press report (1) discussing the potential replacement of aluminum radiators with copper radiators in pickup trucks and sport utility vehicles provides an interesting example to reinforce the importance of cost in materials selection. In this case, the displacement of aluminum by copper is being considered because of the more favorable heat dissipation characteristics of copper. Heat dissipation has become more important due to the development of hotter-burning V-8 and V-6 engines. For this application, engineers also consider cost, weight, size, durability, and manufacturing process. The report stated that the development by the International Copper Association (ICA) of the ""CuproBraze"" manufacturing process for heat transfer products was an important development. This technology produces 845 radiators that weigh less, are more uniform in quality, and are potentially more compact. They are also expected to cost less in high production quantities than copper units produced in the past, last three to four times longer than soldered copper radiators, and the process also eliminates the use of lead solder from the manufacturing process."

Jan 1, 2000

By Harsh Deep Chopra, Jason N. Armstrong, Susan Z. Hua

"Existence of anisotropic Curie temperature materials [E. R. Callen, Phys. Rev. 124, 1373 (1961); J. Appl. Phys. 32 S221 (1961)] is a longstanding prediction - materials that become paramagnetic at a lower temperature along certain crystal directions while remaining magnetically ordered in other directions up to a higher temperature. Validating Callen's theory, we show that all directions within the basal plane of monoclinic F e1Ss (pyrrhotite) single crystal remain ordered up to 603 K while the c-axis becomes paramagnetic at 225 K. Results prompt a re-evaluation of existing magnetic materials with a focus on magnetic characteristics along different crystal orientations above instead of below the ordering temperatures. Theoretical guidelines for identifying new materials with large anisotropy of Curie temperature are also given, and analysis protocol to characterize them in a self-consistent manner is discussed. Materials with such a large directional dependence of Curie temperature open the possibility for uniquely new devices and phenomena, including energy storage and spintronics devices.IntroductionThis paper is an overview of our recently published work on anisotropic Curie temperature materials [l]. As a background, ferromagnetic and ferrimagnetic materials are characterized by a critical temperature below which magnetic interactions prevail over thermal agitation, leading to the appearance of a spontaneous magnetization. Above this critical point, called the Curie temperature, thermal energy eventually overcomes magnetic interactions and the material becomes paramagnetic.It is less well known that the Curie temperature varies with the orientation of the crystal. In ordinary ferromagnets this orientation dependence of Curie temperature is negligible because of the vanishingly small ratio of anisotropy to exchange energy, i.e., the spherical or isotropic exchange term dominates. Thus Curie temperature in ordinary ferromagnets appears to be an isotropic property, and assigning it a single characteristic ordering temperature suffices. However, a 'small' difference is not equivalent to being the 'same', and under appropriate set of conditions the anisotropy of Curie temperature is predicted to become appreciable, as discussed in the following.In formulating the theory of 'anisotropic magnetization' [2, 3] E. R. Callen and H. B. Callen first discussed the ""interesting possibility"" of anisotropic Curie temperature [2] or ACT; Callen and Callen noted that it was W. J. Carr, Jr. who initially raised this question, and they discussed it in detail at the Conference on Magnetism and Magnetic Materials (NY, November 1960), see footnote on page 320 of Ref. [2]. Soon thereafter, E. R. Callen developed the theory underlying the anisotropy of Curie temperature and conditions conducive to its observation [4, 5]."

Jan 1, 2013

By Lloyd Robert Nelson

The historical progression in installed electric furnace smelting capacity of some 1650 calcium, chromium, manganese, nickel and silicon ferro-alloys furnaces is reviewed. Key increases in the inherent installed electrical capacity, often achieved through uprating furnaces, are identified. Possible factors facilitating such advances are explored, including: specific process engineering and pyrometallurgical enhancements (e.g., improved control over the consistency, size and chemical form of raw materials, extents of preheating and pre-reduction delivered into the furnace; and patterns of feed distribution); furnace electrical configuration (AC or DC; immersed, submerged-, shielded- or open-arc); form of electrode (graphite, self-baking or composite); furnace configuration (circular or rectangular; closed or open and 1-, 3- or 6- electrodes); better engineering and equipment designs (e.g., high-intensity cooling) and state of furnace electrical and metallurgical control systems. A view is presented on the prevailing state of evolution of the Mega-scale in ferro-alloy smelting and opportunities for still further improvements.

Jan 1, 2014

By Yasuo Ojima

The mainstream of copper smelting processes in the world today is a combination of Outokumpu type Flash smelting and PS (Pierse Smith) converting process. Presently it can be estimated that about 50% of the worldwide copper production is produced by the Flash smelting process and 80% by the PS converting process. PS converting process due to its operational flexibility and despite its batch character of operation has been widely used in many smelters for more than 100 years. However, an environmental issue has been pointed out related to this process because of the difficulty of handling the fugitive gases. Recently several copper smelters in the world have been newly construeted and others have executed major modification and modernization. The continuous converting process has been applied in many of them instead of PS converting process. This paper compares PS to the emerging continuous converting process and makes a prospect of the future of the copper converting process.

Jan 1, 2003

By Chikabumi Yamauchi

The comprehensive scientific development of conventional industrial technologies stimulates the creation of new ones in the next generation. Prof. Akira Yazawa has just accomplished such an important work in the non-ferrous pyrometallurgical field. The results of his successful research are immortal. The research group of the author has conducted since 1985 fundamental studies on the application of sodium carbonate slag to copper fire refining. In the present paper, a calculation method for the distribution ratios of impurity elements between sodium carbonate slag and molten copper will be shown. The specific treatment by sodium carbonate slag will also be presented as an effective way for reducing each impurity content in the molten copper to 0.1 mass ppm, which is the level required for the production of high purity copper. The thinking-out of this evaluation method gets a hint from the Equation (2) in the text which was proposed by Prof. Akira Yazawa.

Jan 1, 2003

By Atul Kurnar

In the present research a two-cell polymer electrolyte membrane fuel cell stack (PEMFCS) with SS-316 bipolar plates has been tested. Different flow field designs were studied and a modified multi-parallel channel design is proposed. The stack is stable within ±3 mV even after 1000 hours of continuous operation, indicating efficient water management and effective removal of any metal ions from the stack. Also no iron is detected in membrane electrode assembly (MEA) by EDAX analysis after 1000 hours of cell operation. Results indicate that high operating pressures beyond 138 kPa didn't significantly increase the performance of cell, however added much to the pressurization cost of the system. Tests at different temperatures showed that at temperatures below 80°C, external humidification could be completely removed. The prototype stack is also tested under variable load conditions simulating highway-driving schedule. The experimental efficiency of PEMFCS is 48%, which is in good agreement with the model predicted efficiency of 53.8 %. Results show that 59% of the total losses are due to ohmic polarization in the cell. Further analysis on applications of the stack in automotive transportation system is in progress.

Jan 1, 2002

By Emmanuel N. Zevgolis

After the Krupp-Renn process was shown to be unsuitable for treatment of the Greek nickeliferous laterites, the LM process was developed. It involved roasting reduction up to metallic iron in a Rotary Kiln (R/K) and smelting in an Electric Furnace (E/F). Ferronickel (with 15% Ni) produced in the E/F, was upgraded to 90-95% Ni in an 02 converter. Raw Ni was then electrolyzed to pure Ni. The LM process was not applied for very long however, due to various problems. So, the Larco process was developed, i.e., pre-reduction in R/Ks, smelting of the calcine in E/Fs for ferronickel production (15% Ni) and enrichment to 20% Ni, in an O2 converter. Until now, many technological improvements have been invented and applied or adjusted to the process, which have made it a fully reliable and competitive one. sohe of the improvements were the invention and application of a new granulation process for liquid ferronickel, the modification of the existing R/Ks, E/Fs and converters and the design of new ones, and others that are described in the paper. Keywords: Ferronickel, metallurgy, laterites, LM process, Larco process

Jan 1, 2004

By C. Kathryn Sole

The close proximity of cobalt and nickel on the Periodic Table and their frequent occurrence together in nature means that their separation in primary hydrometallurgical flowsheets is not only often necessary, but requires the use of innovative chemistry to achieve high selectivity. Today, increasingly stringent demands for the high-purity cobalt and nickel salts required for battery applications in consumer electronics, electric vehicles, and solar applications are again challenging the state of the art. The evolution of cobalt–nickel separations is traced from the basic precipitation technologies employed in 1960s, through commissioning of the first cobalt solvent-extraction plant in South Africa in 1974, to the commercialisation of organophosphinic-acid extractants (which provided a step-change improvement in selectivity) and their eventual widespread use in the 1980s and ’90s. This was followed later by their sulphur analogue, which enabled different possibilities for laterite processing in the new millennium. Separations in chloride media, which take advantage of the different speciation chemistries, and nickel extraction by oximes in ammoniacal systems are also discussed. More recently, chelating and solvent-impregnated resins have added further dimensions to the possibilities for flowsheet design and enabled higher purity products to be attained from both primary and secondary sources. Selected case studies and flowsheets are described.