Search Documents

By Renato G. Bautista

The solvent extraction separation of nickel, cobalt and their associated metals have been considered for industrial application since the early 1960's. This review of the solvent extraction processes is limited to the industrial applications, pilot plant and the more extensive bench scale studies reported in the literature. The processes with their significant chemical and technological contributions are discussed in detail whenever the information are available. The application of solvent extraction to the nickel-cobalt processing has so far been limited to the secondary separation of nickel-cobalt containing feedstocks derived from other primary nickel-cobalt separation methods.

Jan 1, 1993

By D. B. Dreisinger, V. van der Pas

Zinc sulphate electrolyte used for zinc electrowinning must be purified for cObalt ions. The cobalt is removed in a cementation stage by the addition of zinc dust. Copper arid antimony are frequently used additives which enhance the cementation of cobalt. This paper aims at a better understanding of how copper and antimony promote the removal of cobalt. Initial experiments were done in a batch cementation reactor. Copper and antimony precipitated in the early stages of cementation indicating that a preferential substrate was formed. The individual role of copper and antimony was further investigated in an electrochemical cell. On a microsca1e, the , growth process, morphology and the composition of the precipitates were examined with SEM and XRD. It was found that cobalt could not be deposited in its pure form but as a cobalt-zinc alloy with zinc as the prime constituent. Copper increases the cathodic surface area of zinc dust by precipitating as numerous dendrites. Antimony acts as a cathodic surface onto which a cobalt-zinc alloy with an increased cobalt content is deposited.

Jan 1, 1997

By J. M. Demarthe

The recovery of minor metals from zinc concentrates is dependent on the process used for their treatment. The behaviour of gennanium and indium in pyrometallurgical processes such as the Imperial Smelting furnace or the cupola furnace is of practical importance. The .Anaconda process for the treatment of zinc leach residues is not well suited to the recovery of valuable minor metals, and new technologies must be studied. In this regard, a solvent extraction process, using Kelex 100, developed by Minenet Recherche for the recovery of gennanium from dilute sulphate leach solutions is described. Also, new hydranetallurgical techniques in chloride media can be applied to the purification and concentration of minor metals depending on the stability of their chloro-complexes and the valency of associated inpurities. A synthesis of methods is proposed to address the problems related to the recovery of valuable minor metals in the zinc industry.

Jan 1, 1990

By Michael Sheedy

Recoflo® ion exchange technology utilizes fine mesh resin beads, a fully packed resin bed, and counter-current regeneration. These unique features help to improve exchange kinetics, reduce regenerant consumption, and increase the concentration of strip solutions. Commercially, Recoflo technology has been extensively used in the following areas: the separation of strong mineral acids from dissolved metal salts, the purification, removal and recycle of dissolved metals in the surface finishing industries, and the production of high purity water. In the mining and metallurgical industries Recoflo ion exchange systems have been installed to treat a copper refinery bleed stream at Falconbridge's Kidd Creek plant and to treat a nickel and cobalt bearing ammonium sulfate effluent stream at another primary metal producer's facility. In the first application an APU® Recoflo system is used to separate the sulfuric acid at 200g/L from the dissolved copper and nickel sulfate salts at 80 g/L. The absorbed sulfuric acid is recovered from the resin by elution with water and is then returned to the refinery electrolyte circuit. In the second application a chelating ion exchange resin is used to remove nickel and cobalt at 0.01 - 3.0 g/L from process streams containing 120 - 240 g/L ammonium sulfate. This Recoflo system has been designed to reduce the nickel and cobalt levels to 0.001 - 0.002 g/L while recovering cobalt and nickel at concentrations up to 30 g/L and 15 g/L respectively. Case studies for both these metallurgical applications and other Recoflo processes are presented in this paper.

Jan 1, 1998

By Nagahiro Saito, Takahiro Ishizaki, Daisuke Yata, Akio Fuwa

"Ternary compound-semiconductor, CuInTe2, film was deposited cathodically under potentiostatic conditions on titanium substrate from aqueous solution containing CuCl2, InCl3, TeO and HCI in this study.The deposition parameters such as electrolytic solution composition, potential and temperature were optimized· for electrodeposition of CulnTe2. Structural characterization of the: deposited film was also carried out using XRD and SEM. Electrodeposited films Were prepared and analyzed for their chemical composition using ICP. The ICP analyses showed that the stoichiometry of the films could be controlled by (1) the deposition potential and/or by (2) the electrolytic bath composition.1. IntroductionThe ternary chalcopyrite type compound semiconductors exhibit such electrical, chemical and optical properties that make them promising for photovoltaic devices (Kazmerski and Wagner 1985)(1). Many researches have been devoted to the characterization of these materials for over one and a half decades. High efficiency solar cells based on p-CulnSe2/n-CdS have been reported (Mickelson and Chen 1983)(2).The electrodeposited fil ns of semiconductors have more advantage than other physical and chemical deposition techniques; it is relatively easy and economic technique and the semiconducting properties such as band-gap variation, concentration of dopant stoichiometric composition and so on can be controlled with some accuracy. The electrodepositions of the ternary CuInS2, CulnSe2 have been reported by many workers (Hodes and Cahen 1986, Bhattachaya 1983, Singh et al 1986, Lokhande and can be prepared as both p-type and n-type. Therefore, its potential for application in optical and photovoltaic devices is one of the great interests. Bhattacharaya and Rajeshwar (1986)(9) reported the electrodeposition of CuInTe2 from an acidic bath· on indium tin oxide (ITO) and Mo substrate. Triethanolamine (TEA) was used as a complexing agent for indium. The electrodeposited CuInTe2 films were non-stoichiometric in composition under all their experimental conditions by electron-probe micro analysis (EPMA) and Auger depth profile analysis. In this paper, we report our results on the behavior .of Cu-In-Te electrodeposition from ·acidic baths; the preparation of CulnTe2 and the effect of bath composition on electrodeposited films obtained."

Jan 1, 2000

By John E. Litz

The Denver Process is a chemical/metallurgical process for treating radioactive or mixed wastes by extraction and concentration of the radioactive and/or hazardous component(s). The process was developed at bench scale, evaluated at pilot scale, and will be implemented at full scale to remediate a low-level radioactive sludge. Six unit operations comprise the process: feed handling, leaching, solvent extraction, electrodeposition, precipitation, and tailing preparation. Unit operation design may be modified to tailor the process for varying feed stock composition. Radioactive or mixed waste is reacted with aggressive reagents in an aqueous phase, solubilizing targeted components. These components are then selectively extracted and concentrated in subsequent steps. Waste solids and solution exit the process as non-hazardous non¬radioactive waste. Dehydration of aqueous waste by submerged combustion eliminates aqueous discharge.

Jan 1, 1994

By Alaina Mallard, Katie Schumacher, Jannette Chorney, Grant C. Wallace, Jerome P. Downey

Commercial ceramic carbide operations require large thermal and mechanical energy inputs in order to produce a powder product. A process that could reduce the energy requirements needed to make these materials would allow for these materials to be implemented in a greater number of industrial applications. In this study, silicon carbide (SiC), tungsten carbide (WC), and molybdenum carbide (Mo2C) were synthesized via the carbothermal reduction of activated carbon loaded with silicate, tungstate, and molybdate anions adsorbed from aqueous solutions. Carburization was carried out under reducing and inert gas atmospheres, at temperatures lower than those utilized by commercial operations. Silicon carbide “whiskers” were synthesized under H2 at 1400 °C, while molybdenum carbide and mixed crystals of WC, W2C, and W were synthesized at temperatures below 1000 °C. X-ray diffraction and scanning electron microscopy were used to characterize the carburization products, and inductively coupled plasma-optical emission spectroscopy (ICP-OES) was used to determine the degree of adsorption onto the activated carbon matrix.

Mar 1, 2018

By Shoukun Wei, Cong Wang, Lin Cheng, Zhiyu Qiao, Jingkui Liang, Weidong Zhuang

"The subsolidus phase relations in the Sm2O3-CeO2-CuO ternary system sintered at 1000°C in air have been investigated by X-ray powder diffraction. The system can be divided into two ternary phase regions and three binary phase regions. Solid solution Sm2-xCexCuO4-y is the only ternary phase in the system. The maximum solid solubility was confirmed to be x = 0.2. The crystal structure, oxygen content and superconductivity of the solid solution Sm2-xCexCuO4-y have been investigated.I. INTRODUCTIONAfter the first discovery of the superconducting ceramic system La-Ba -Cu-O[l] with Tc = 30 to 40K, other families of cuprate-based ceramics have been synthesized with higher critical temperature. These oxides include the R-Ba-Cu-O (R: rare earth) systems (Tc=90K or so)[2,3], the Pb-Sr-(R,Ca)-Cu-O systems (T c = 50 to 100K), the Bi-Sr-Ca-Cu-O systems (Tc=80 to 105K)[4], the T1-Ba-Ca-Cu-O systems (Tc=85 to 125K)[5], the Hg-Ba-Ca-Cu-O systems (Tc=90 to 150K)[6,7], and some others. The improvement of Tc is very rapid. The measurement of the Hall and Seebeck coefficients shows that they all are p-type superconductors. In 1989, Tokura et al.[8] discovered the superconductors R2_XCexCU04_y(R = Nd, Pr, Sm) with Tc=20 to 24K (onset) and found that the charge carrier is electron in contrast to their hole doped counterparts. This discovery increases the fundamental understanding of the cuprate-based superconductors by comparing their characteristics and mechanisms to those of the p-type superconductors.One of the outstanding features of these n-type superconductors is the difficulty in obtaining samples in good charactors, such as the difficulty in obtaining zero resistivity state in single crystal or bulk samples; the semi-conductivity of bulk samples in ordinary state (opposite to the superconducting state) and a reducing reaction involved before the superconductivity has been found[9]. Therefore, it is necessary to optimize the sample preparation of the n-type superconductors. All the n-type superconductors have been found· considerable solid solubility. The compositional variations play a governing role in the carrier doping of the superconducting phases. Also, it is vital to the ceramic processing of superconductors to identify the compatible phases in the phase diagram. Finally, there is always the possibility of finding additional compounds, both superconducting and non-superconducting, in an unexplored system."

Jan 1, 1996

By L. E. Murr

Preliminary studies of bacterial attachment to MoSp (molybdenite) have indicated a low correlation between dislocation density in the (0001) basal plane and the density of attached microbes. In the present investigation, poly- crystalline thin sections of CuFeSp (chalcopyrite) have been ion-etched or carefully comminuted to electron transparency and observed in the transmission electron microscope. The thinned test sections were then immersed in a culture solution in which bacterial attachment occurred. The attached bacteria were biologically stained in a uranyl acetate solution and the entire thin CuFeSp sections or tiny particles containing thin, electron transparent edges with bacteria attached were observed in the transmission electron microscope, allowing the coincidence of bacteria and their density to be directly observed in relation to lattice dislocations and crystallographic orientations of the exposed surfaces. This technique has the advantage of exposing crystallographic orientations where dislocation lines intersect the free mineral surface, and the emergence sites of dislocations could in principle be directly investigated as a possible preferential attachment site for bacteria. Bacteria morphologies were also extensively observed in the scanning electron microscope. The implications of the results are that although a strong correlation between dislocation density and distribution and bacterial attachment, density, and distribution has not been established, systematic multiplication of dislocation densities in crystalline mineral particles to be leached could enhance leaching rates by increasing the reaction rates as well as the density of attached bacteria at crystalline, metal-rich phases. Techniques for dislocation production in mineral and ore bodies and the particular implications in chalcopyrite leaching are discussed in relation to in-situ leaching where comminution is attained by explosives and the generation of strong shock waves in an ore body. The present study demonstrates the use of microanalytical and microscopic techniques in the investigation of fundamental microstructural phenomena in hydrometallurgical processes.

Jan 1, 1976

Heap leaching with thiosulfate is only used in one location for the treatment of a refractory preg-robbing gold ore initially pre-treated with bio-oxidation. The chemistry of thiosulfate leaching is complex. Further work is required to develop a better control of the system. A mild refractory gold ore containing pyrite and chalcopyrite was used in this study to investigate the use of thiosulfate as an alternative heap leaching technology. Preliminary bottle rolls tests indicated that similar gold extraction was obtained with cyanide and thiosulfate. In the column leach test, effects of thiosulfate, copper and ammonia concentrations, and their ratio on both gold extraction and reagent consumption, were assessed. The range of reagent concentration were: 0.1 - 1.0M (NH4)2S2O3, 0.03 - 0.10M CuSO4*5H2O, 1.0 - 6.0M NH4OH. The solid-liquid ratio was in the range of 0.83:1 to 5: 1. Best results showed that 72% of gold was extracted in 50 days with 0.3M (NH4)2S2O3, 0.05M CuSO4*5H2O and 6M NH4OH or a combination of 1.0M (NH4)2S2O3, 0.03M CuSO4*5H2O and 3M NH4OH. The reagent consumption at solid/liquid ratio of 3:1 was 37 - 48.6 kg/t (NH4)2S2O3 and 0 - 0.62 kg/t CuSO4*5H2O. The results of thiosulfate leaching were also compared with cyanide leaching.

Jan 1, 1999

By I. A. Ibrahim, A. K. Ismail, A. A. I. Afifi

"Synthetic rutile for paints industry is conventionally produced by a combined smelting and leaching process. Smelting process is an energy intensive step and represents economic burden in view of the increasing energy costs. Egyptian ilmenite ore containing 40% TiO2 and 38% total Fe was thus directly leached with sulphuric acid to produce eventually synthetic rutile. The process-involves digestion of the ore with sulphuric acid whereby both TiO2 and Fe are dissolved. The solution was then treated with iron powder to convert Fe3+ to Fe2+ to avoid co-precipitation of iron during the hydrolysis step. After solid/liquid separation and clarification, Fe was crystallized as ferrous sulphate hydrate by chilling. The titanium pregnant liquor was concentrated by evaporation and then hydrolyzed to precipitate hydrated titanium dioxide which was then calcined to produce synthetic rutile containing 98.6 TiO2 and 0.58% Fe2O 3. The optimum conditions of all the processing stages were determined.IntroductionSynthetic rutile for paints, welding rods and other industries could be produced from ilmenite ore by sulfuric acid leaching process. This process is a batch process and nearly 80% of all TiO2 produced in the world is made by this process [1]. Figure l shows simple block flowsheet of the process. The process involves the following stages.Digestion: In this stage, ilmenite ore is treated with concentrated sulfuric acid where both iron and titanium are converted to their respective sulfates according to the following equations:"

Jan 1, 2000

By P. Louis

Gécamines is the world largest cobalt producer with about 60 % of the Western World production. During the last ten years, its annual production varied from 5,400 to 14,600 t/year. The actual cobalt production capacity of Gécarnines is 18,000 t/year and could be increased in the near future if necessary. Cobalt is presently produced at the Shituru plant by an old in-pulp electrolysis process and in Luilu, by electrolysis of a clear weakly acidic solution. The Shituru Cobalt quality is lower than the Luilu one. For that reason, part of the Shituru Cobalt is fire refined and cast into granules. Large efforts have been succesfully accomplished to improve the cobalt quality. This has been made possible by a better knowledge of the purification processes and of the electrolysis parameters.

Jan 1, 1988

By Mikiya Tanaka, Mikio Kobayashi, Kazuya Koyama, Yoshihiro Nishisu

Preparation of multicomponent powders are very useful and important for producing the materials with complicated and homogeneous structure. Therefore multicomponent powders have a lot of attention in the field of advanced value-added materials. The aqueous processing in hydrometallurgy has a broad potential for producing multicomponent homogeneous precipitates. Furthermore the aqueous processing has a future potential for integrating the separation, precipitation and synthesis processes. In this presentation the precipitation from homogeneous solution including rare-earths and/or common metals is presented. These precipitates probably consist of almost monodispersed, multicomponent and spherical particles by precisely controlling several reaction factors, reaction temperature, concentrations of starting materials, pH and so on. The separation of the nucleation stage and the growth stage of the particles is very important and how to control the reaction factors to realize enough separation is discussed. In this presentation the systems of Y-Eu-0, Y-Fe-0 and Y-Al-0 are mainly presented. In addition the precipitation-stripping method from organic phase is presented focusing the synthesis of multicomponent powders of Ni-La-0 system. These precipitates have multicomponents homogeneously and so they are converted to multicomponent ceramic powders by calcination. They are important compounds as phosphor, opto-magnetic materials, optical materials, hydrogen storage alloy or as the related precursors. In this presentation some characterization of these particulate materials is included.

Jan 1, 1998

By M. J. Soria-Aguilar, F. R. Carrillo-Pedroza, S. Preciado-Nuñez

"Dust originated from the Iron Blast Furnace (BF) and the Basic Oxygen Furnace (BOF) containing alkalis and zinc are not completely recycled because due affects the process efficiency. These types of dust represents an economical lost as a consequence of values contents of iron and carbon. Several processes in aqueous media have been developed to treat the steelmaking dust. This work presents a laboratory study for reduce zinc and alkali containing in BF and BOF dust, using different acid and alkaline solutions and oxidants as ozone. The results obtained show that the elimination of zinc and alkalis can vary from 5 % to 80 % depending on the treatment conditions (the solution, the oxidant and the temperature). The dust obtained from this treatment can to be recycling to the steelmaking processes to recovery iron and carbon values.IntroductionEach one of the steel making industry processes generate several sub-products and residues containing useful units of iron, carbon and other elements that are or could be recycled to the productive process. Some of those sub products are the dust from the gas cleaning system of the blast furnace and the dust from BOF reductive process; such dusts are sent to the sedimentary lagoons. The dust represent a lost of 2 % of the total iron and coal mineral’s content feed to the process.The typical iron, carbon and zinc content of the blast furnace dust is approximately 40, 24 and 0.3 % and the content of the BOF dust are 50, 15 and 0.9 % weight respectively, with physics - chemical characteristics that became them susceptible to be agglomerated and feed to the blast furnace. But the high content of zinc does not allow recycling it into the process. At the blast furnace the presence of zinc propitiate the formation of film in the furnace wall; these elements are vaporized in the low section of the cube of the furnace and they are dragged by the gases to the top and condensed again at the cube. Part of these condensed materials strongly adheres to the furnace cover harming it and reducing the empty space of that section, causing eventually a severe perturbation of the ascendant gas flux [1]."

Jan 1, 2008

By A. S. Rao

In the original roast-leach-electrowin (RLE) circuit at the Chambishi Cobalt Plant, commissioned in 1978, the separation of nickel and zinc from cobalt in aqueous streams was achieved by successive stages of precipitation using lime. However, this technique was not able to meet the increasingly stringent metal quality requirements of customers whilst maintaining acceptable process recoveries. Intensive process development work was carried out and, in July 1988, an ion exchange circuit for nickel removal was commissioned using DOW XFS 4195 resin and for the first time Z.C.C.M. produced high purity cobalt metal containing less than 0.10% nickel. Following further work, a solvent extraction circuit for zinc removal, using DZEHPA, was commissioned in February 1991 in order to enhance cobalt recovery by reducing co-precipitated cobalt losses. This paper outlines the process development, design, commissioning and subsequent operation of these nickel ion exchange and zinc solvent extraction circuits.

Jan 1, 1993

By Jorge Alberto Soares Tenório, Denise Crocce Romano Espinosa, Kellie Provazi

"The purpose of this paper is to study metal separation and recovery from a sample composed of a mixture of the spent household batteries main types, using liquid-liquid extraction and metals electroplating. The electrolytes preparation consisted: grinding the waste of mixed batteries, reduction and volatile metals elimination using electric furnace and acid leaching. From this solution were studied the liquid-liquid extraction using Cyanex 272 [bis(2,4,4-trimethylpentyl) phosphoric acid] with tributyl phosphate as extracting agent. After study of solvent extraction and stripping has been recover and separate the metallic ions and with electroplating was recovered the manganese, copper, nickel and cobalt in their metallic forms.IntroductionCurrently has increased the consumption of electronics devices such as mobile phones, laptops, wireless phones, comcorders, etc., especially portable devices that use batteries as a power source, and consequently has increased the disposal of such material.Increased consumption of this devices is caused by several factors including increased popularity of consumer electronics devices, has become essential of modem life. Therefore, recycling of batteries is important because its reduces the waste of raw materias and nonrenewable resouces used in manufacturing these materias and help prevent potential environmental risks associated with improper disposal of these products together with household garbage.The batteries have a very diverse composition, containing metals such as zinc, lithium, nickel, lead, mercury, cobalt, cadmium, iron, copper, manganese and rare earths. Some components may contain hazardous materials such as cadmium, mercury and lead, and can contain acids or bases.This research line is a pioneer in dealing with a single process for recycling a mixture with all batteries types, for only processes that exist in the world are based on the recycling a mixture with all types of batteries, for only processes that exist in the world are based on the recycling of one or two types of batteries [1-4]. This process necessitates the previous separation of the batteries by type."

Jan 1, 2012

By Annette Revet

Aluminum fluxes have been used for the remelting of secondary aluminum since the beginning of the recycling industry. During the last sixty years, a number of fluxes have been designed, researched and tested at the laboratory and plant scale. Fluxes are primarily a mixture of potassium chloride and sodium chloride with the addition of some other chloride and fluoride containing salts. The conventional Mining methods of potassium chloride and sodium chloride yield products that are not high in purity without costly reprocessing or additional refining. In 1993, Kalium Chemicals, Ltd. began supplying flux mixtures to the aluminum industry. Kalium produces potassium chloride and sodium chloride which are mined and refined by a patented solution-mining process. The salts produced by this process are high in purity with lower concentrations of iron oxide, sulfates, and other impurities. This paper will present the different mining methods used in the production of sodium chloride and potassium chloride. The paper Will address how these mining methods affect the purity of the final product as well as the effect of purity on the performance of the fluxing salt.

Jan 1, 1995

By Michael D. Odekirk

The U.S. Bureau of Mines investigated solvent extraction methods for recovering scandium from water leach filtrates of sulfated tantalum tailings. These complex solutions typically contained 20 or more metals in significant quantities. Two commercially available organics, DEHPA and PC-88A, were identified as effective scandium extractants. The two organics were tested in continuous solvent extraction circuits to compare loading and stripping kinetics, recyclability, and general processibility. Results showed that 95 pct of the scandium, greater than 80 pct of the zirconium, and only small amounts of other solubilized metals were extracted. Scandium was stripped directly as a 3.6 pct Sc precipitate when a 3n NaOH liquor was used as the stripping agent. This precipitate was subsequently upgraded to 94 pct ScE3 via a selective HE leach.

Jan 1, 1993

By M. Nilmani, H. Pham

"A mathematical model was developed for the dissolution of particles in gas-stirred liquid baths, linking the hydrodynamic flow field of the bath, the average particle trajectory and liquid-side mass transfer controlled dissolution kinetics. The model was validated by comparison with experimental data for the dissolution rates of urea particles in gas-stirred aqueous CaCl2 solutions measured during this study. The dissolution rates of urea. particles were measured using an image analysis system.The effects of operating conditions on the dissolution rates of urea particles were investigated. It was found that the key operating parameters influencing the utilisation efficiency of particles were the gas flowrate, particle diameter and lance submergence. An optimisation of the operating conditions for the model urea-aqueous CaCl2 system indicated that the utilisation efficiency increased with decreasing gas flowrate and decreasing particle size for a given lance submergence.IntroductionThe utilisation solid reagents in molten baths varies from one process to another, depending upon the rates and mechanisms of the reactions between the particles and melts, as well as the trajectory and residence time of the particles in the melts. Previous studies have not attempted to predict the utilisation efficiency of injected solids by coupling the motion of the particle with its dissolution rate. Instead, the particles have usually been assumed to remain suspended in the bath until completely dissolved, and the mass transfer coefficients were based on the average specific mechanical energy dissipation rates in the bath. Investigations of this nature relevant to the present study have been undertaken by Apelian et al. [1], Brucato et al. [2] and Langberg and Nilmani[3]. The objective of the present study was to develop a mathematical model, incorporating both the reaction kinetics between the particles and liquid and the particle motion simultaneously, to predict the dissolution rates and times of the particles. The mathematical model was validated by comparison with experimental dissolution rates of a cold model analogue system measured using an image analysis technique, and then applied to optimise the operating conditions in the model system."

Jan 1, 1996

By Farhang Hedjazi

Anglo Asian Mining have recently installed an industrial-scale UF/RO plant to treat process liquors from the integrated heap leach/agitation leach/ resin-in-pulp gold ore treatment facilities at their Gedabek gold-copper mine in Azerbaijan. The UF/RO plant is designed to treat 60 m3/h of leach solution to produce discharge quality water. The RO concentrate, which contains elevated concentrations of copper, silver and cyanide, is sent to the company’s SART plant for the recovery of the metal values as a precipitated sulphide concentrate and the cyanide, which is recycled to leaching. The new water treatment plant improves the sustainability of the operations at Gedabek by enabling replacement of fresh water input with RO permeate and safe discharge of permeate to the environment, during periods of excess water balance. The paper includes pilot plant data and early results from the full-scale plant.