Copper Volume 7.pdf

Proceedings

Volume 7

Plenary

Mineral Processing

Recycling

Posters

Authors Index

Keywords Index

Conference organized by GDMB, IIMCh, MetSoc, MMIJ, SME and TMS

Editor

GDMB

Paul-Ernst-Straße 10, D-38678 Clausthal-Zellerfeld

Internet: www.GDMB.de

Volume 1 ISBN 978-3-940276-25-4

Volume 2 ISBN 978-3-940276-26-1

Volume 3 ISBN 978-3-940276-27-8

Volume 4 ISBN 978-3-940276-28-5

Volume 5 ISBN 978-3-940276-29-2

Volume 6 ISBN 978-3-940276-30-8

Volume 7 ISBN 978-3-940276-31-5

Set (Volume 1+2+3+4+5+6+7) ISBN 978-3-940276-32-2

All rights reserved. No part of this publication may be reproduced or electronically processed,

copied or distributed without the prior consent by the editor.

The content of the papers is the sole responsibility of the authors. All papers were peer reviewed by

the corresponding members of the technical groups of the organizing societies.

Editorial staff: Dipl.-Ing. Jens Harre

Production and marketing: GDMB Informationsgesellschaft mbH

Printed by: Papierflieger

© GDMB Clausthal-Zellerfeld 2010

Bibliographische Information Der Deutschen Bibliothek

Die Deutsche Bibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliographie;

detaillierte bibliographische Daten sind im Internet über http://dnb.ddb.de abrufbar.

Bibliographic information published by Die Deutsche Bibliothek

Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliographie; detailed

bibliografic data is available in the internet at http://dnb.ddb.de.

Proceedings

Volume 7

Plenary

Mineral Processing

Recycling

Posters

Authors Index

Keywords Index

The Copper 2010-Proceedings are friendly supported by

Proceedings of Copper 2010 IV

The Organizing Society:

GDMB Society for Mining, Metallurgy, Resource and

Environmental Technology

The GDMB is a non-profit organization. Its activities focus on combining science with practical

experience in the fields of mining, engineering, tunnelling, mineral processing, extraction, recycling

and refining of metals, as well as on the manufacturing of semi and finished products. There is an

increasing emphasis on associated environmental issues. The GDMB is internationally active with a

European basis and covers a wide variety of topics from applied geology via processing to recy-

cling. These include many important areas of chemistry, especially the complex metallurgical

chemistry and, last not least, also analytical chemistry. As a consequence of their increasing impor-

tance, aspects of industrial minerals are addressed in addition to the traditional fields of metals and

alloys. In order to remain a vibrant and attractive professional society, the GDMB draws on the ex-

perience and interests of its worldwide members.

The Co-Organizing Societies and their Representatives

Institutos de Ingenieros de Minas de Chile (IIMCh)

Enrique Miranda S., Gerente IIMCh, Chile

The Metallurgical Society of the Canadian Institute of Mining, Metallurgy,

and Petroleum (MetSoc)

Joël Kapusta, Ph.D., Air Liquide Canada Inc., Canada

Dr. Phillip Mackey, Xstrata Process Support Centre, Canada

The Mining and Materials Processing Institute of Japan (MMIJ)

Dr. Takahiko Okura, The University of Tokyo, Institute of Industrial Science, Japan

Yasuo Tamura, Japan Mining Industry Association, Japan

Society for Mining, Metallurgy, and Exploration (SME)

Dr. John L. Uhrie, Newmont Mining Corporation, USA

The Minerals, Metals & Materials Society (TMS)

Dr.-Ing. Andreas Siegmund, LanMetCon, USA

Proceedings of Copper 2010 V

Conference Chairman

Dipl.-Ing. Michael Kopke Aurubis AG, Germany

Technical Programme Chair

Dipl.-Ing. Jo Rogiers Aurubis AG, Belgium

Conference Chair Assistance

Dipl.-Ing. Jürgen Zuchowski GDMB Gesellschaft für Bergbau, Metallurgie,

Rohstoff- und Umwelttechnik e. V.

Session Chairs

1 Plenary lessons of general interest Dipl.-Ing. Norbert L. Piret,

for all conference members Piret & Stolberg Partners, Germany

2 Economics Dr. Patricio Barrios, Aurubis AG, Germany

3 Downstream Fabrication, Application Dr.-Ing. Hans Achim Kuhn,

and New Products Wieland Werke AG, Germany

4 Mineral Processing Assoc. Prof. Sadan Kelebek,

Queen’s University Canada

5 Pyrometallurgy David B. George, Rio Tinto, USA

6 Hydrometallurgy Dr.-Ing. Andreas Siegmund, LanMetCon, USA

7 Electrowinning and -refining Dr.-Ing. Heinrich Traulsen, Germany

Mike Murphy, Xstrata Technology, Australia

Mike Hourn, Xstrata Technology, Australia

8 Process Control, Automatization Prof. Dr. Markus Andreas Reuter,

and Optimization Outotec Ausmelt, Australia

9 Recycling Dipl.-Ing. Jörg Wallner, Austria

10 Sustainable Development / Health, Dipl.-Ing. Miguel Palacios

Safety and Environmental Control Atlantic Copper S.A., Spain

Proceedings of Copper 2010 VI

Technical Group Chairs

GDMB (region: Europe, Russia, near Orient) Dipl.-Ing. Jo Rogiers, Aurubis AG, Belgium

IIMCh (region: South America) Sergio Demetrio, IIMCh, Chile

MetSoc (region: Canada, Australia, Africa) Ass. Prof. Edouard Asselin,

University of British Columbia, Canada

MMIJ (region: Japan, China, South East Asia) Dr. Takahiko Okura,

The University of Tokyo, Japan

SME / TMS (region: USA, Mexico) Dr. Andreas Siegmund, LanMetCon, USA

Technical Group Members

Full information you will find in the internet at: www.Cu2010.GDMB.de

Short Course Organizing Committee

Dipl.-Ing. Michael Kopke (Chair), Aurubis AG, Germany

Dipl.-Ing. Miguel Palacios, Atlantic Copper S.A., Spain

Prof. Dr. mont. Peter Paschen, Austria

Dipl.-Ing. Norbert L. Piret, Piret & Stolberg Partners, Germany

Organizing Committee

Dipl.-Ing. Jürgen Zuchowski GDMB Gesellschaft für Bergbau, Metallurgie, Rohstoff-

und Umwelttechnik e. V.

(Copper2010 Organizing Committee Chairman)

Mareike Hahn GDMB Gesellschaft für Bergbau, Metallurgie, Rohstoff-


Thomas Marbach GDMB Gesellschaft für Bergbau, Metallurgie, Rohstoff-


Dipl.-Ing. Jens Harre GDMB Informationsgesellschaft mbH

Mareike Müller GDMB Informationsgesellschaft mbH

Ulrich Waschki GDMB Informationsgesellschaft mbH

Proceedings of Copper 2010 VII

Preface

Copper – Indicator of the progress of civilization

This is the motto for the 7th

international copper conference, the most important copper seminar in the

world, which has been organized by the GDMB, the German based Society for Mining, Metallurgy,

Resource and Environmental Technology, together with IIMCh from Chile, MetSoc from Canada,

TMS, SME from USA and MMIJ from Japan.

The copper conferences bring together the highest level of science and technology: universities, metal

producers, manufacturing companies, suppliers and finally the people who work with copper: scien-

tists, technicians, engineers, traders and many more.

An extensive programme has been arranged for this conference and an abundance of contributions from

all over the world dealing with the different aspects of copper making and its use are registered already,

for which we gratefully thank the authors. Apart from plenary addresses, separate sessions will be held

for economics, mineral processing, pyrometallurgy, hydrometallurgy, electrowinning and -refining,

downstream fabrication and application, process control and automation, recycling and sustainable de-

velopment, environmental control, health and safety. Copper, one of the oldest metals used by mankind,

is still today one of the most important industrial metals and indispensable for modern life. It is the indi-

cator of industrialization and progress in every country. It is used everywhere, where electricity flows

and thus is still valued so highly today. The increased economic potential of newly industrialized coun-

tries, above all East Asia and China, has increased the significance of the red metal once again. More

recent technologies in production, processing and application often provide new answers to old ques-

tions. From the middle of the last century there was another innovation surge resulting in totally new

technologies, a trend still going on. This has made each copper conference into an exciting adventure. It

is positive and reassuring that particularly the high industrialized countries have become the vanguard,

not just in technical innovation, but also protection of environment and nature and preserving resources.

They repeatedly prove that ecology and economy may go hand in hand.

Many sponsors have contributed to the conference’s success, for which I would like to express my

sincere thanks!

Hamburg is expecting its guests! Hamburg, the old Hanseatic city with a 1200 year long tradition, one

of the biggest and most beautiful cities in Germany, combines a wonderful mixture of industry, com-

merce, nature and culture. Not only the “Copper 2010” in the Congress Centre awaits you, but rich

offerings of sightseeing and shopping in a cosmopolitan city, one of the largest harbours in Europe,

the Alster lake in the city centre, green parks and plenty of cultural events, some of which we hope to

show you in the companions programme.

We are delighted that you will join us and look forward to a highly interesting conference!

Michael Kopke

Chairman Copper 2010

Proceedings of Copper 2010 VIII

Structure of the Proceedings

• Volume 1:

Downstream Fabrication, Application and New Products

Sustainable Development / Health, Safety and Environmental Control

• Volume 2:

Pyrometallurgy I

• Volume 3:

Pyrometallurgy II

• Volume 4:

Electrowinning and -refining

• Volume 5:

Hydrometallurgy

• Volume 6:

Economics

Process Control, Automatization and Optimization

• Volume 7:

Plenary lessons of general interest for all conference members

Mineral Processing

Recycling

Posters

Authors Index

Keywords Index

Proceedings of Copper 2010 IX

Plenary Lectures (Abstracts)

Some of the full papers will be published in World of Metallurgy – ERZMETALL.

Is the Copper Industry Fit for the Future?

Dr.-Ing. Bernd Drouven, CEO, Aurubis AG, Hamburg, Germany

The copper world has already changed a great deal during recent years. But what still lies ahead of

us? What are the changes in conditions that we have to cope with and how will our solutions look?

The rise in the demand for raw materials is unrelenting. At the same time, both the primary and sec-

ondary feed materials have a specific complexity, the customers’ needs are becoming increasingly

differentiated and their orders are placed at increasingly short notice. Production times for metals

have to be faster and inventories minimised.

The volatility of the metal prices has increased significantly in recent years and we will probably

have to live with that in the future as well. The LME functioned well in the crisis, but the copper

price is being influenced more and more by funds.

How can the value added chain change to adapt to this? Which consequences will that have for

process technology, production planning and logistics? How will the consolidation of our industry

continue?

This keynote will go into the various effects and challenges – in particular from the perspective of a

European custom smelter and fabricator – and present possible approaches for solutions.

Sustainable Growth Strategy for Japanese Copper Business

Toshinori Kato, Managing Director, Mitsubishi Materials Corporation, Tokyo, Japan

Business environment, for any industry, has changed dramatically over the last decade. Copper in-

dustry is no exception and those involved are experiencing an unprecedented period of upheaval.

The landscape of the market has completely altered, with large-scale M&As taking place among

miners - creating an oligopoly situation - and a rapid expansion of smelting capacity within de-

veloping countries, driven by strong economic growth. Traditional copper smelters and fabricators

have been facing challenges, but a long-term sustainability of Japanese copper business is achiev-

able.

Firstly, an immense amount of effort has been made over the years to develop the clean and one of

the most environmentally-friendly processes in the industry. Mitsubishi Materials (MMC) particu-

larly plays a great role in establishing Japanese smelters’ reputation as the most energy-efficient

operations in the world. It is our strong commitment to be a leader in this expertise by sharing our

technologies with the industry.

Secondly, infrastructures for copper smelting have been utilized in developing the recycling busi-

ness. Exemplified by the operations at Onahama smelter, which has the world’s largest furnace for

Proceedings of Copper 2010 X

treatment of shredder residue, the industry has worked closely with other sectors and municipalities.

Building an effective structure to make the best use of our facilities is a key for success in this field.

It is fair to say that Japanese copper smelters, including MMC, are now regarded as an indispensa-

ble part of national environmental policy.

Finally, progression of integration in the copper fabrication sector has strengthened the industry’s

capability of providing a variety of high-value-added products to end-users. Tight relationships with

the consumers have been beneficial in developing new use of copper. Evidence of a promising fu-

ture can be seen in increasing use of copper in the growing sectors such as hybrid automobiles and

renewable energy.

Copper Sulphide Smelting: Past Achievements and Current Challenges

Dr. Carlos M. Díaz, Adjunct Professor, University of Toronto, Private Consultant, Toronto, Canada

In the last three decades, increased oxygen consumption in copper sulphide smelting and converting

and the implementation of computerized process control, among other factors, have led to higher

process intensity and smelter concentrate processing capacity, decreased smelting energy consump-

tion and improved SO2 capture from process gas streams. An annual primary smelting furnace

throughput of one million tonnes of concentrate is the new industry standard. Top submerged lance

smelting has become an important processing route in recent years. Two new continuous converting

routes were commercialized in the 1990s. However, due to substantially improved converting prac-

tice and larger converters, Peirce-Smith converting still maintains its position as the dominant tech-

nology.

A major realignment of world copper smelting has taken place in the last 30 years. Spurred by rapid

economic expansion and the resulting huge increase in demand for basic materials, a number of

modern, large capacity copper smelters have been built in China, India and other Asian countries.

Only moderate growth in smelting and electrorefining capacity has taken place elsewhere. More-

over, ER cathode output in the USA has substantially decreased.

In this paper, the author examines recent technological advances and industry changes and high-

lights issues, such as energy consumption and the corresponding greenhouse gas emissions that will

become the focus of future discussion.

Energy as a Key Factor of Sustainability

Javier Targhetta, Vice President, Freeport McMoRan, Phoenix, Arizona, USA

Energy must be seen as the prime mover for development and is therefore vital for economic equi-

librium and social welfare. Energy will continue to play a key role in the coming decades being not

only an environmental challenge but also a fundamental issue in terms of the progress of humanity.

Nevertheless, harmonizing several aspects related to energy management will be of essence for the

near future and will make energy one of the greatest challenges of the 21st century. It is becoming

increasingly important to maintain the appropriate equilibrium between environmental issues

Proceedings of Copper 2010 XI

(global warming, the future of nuclear waste, the integration of renewable energies…), coordinated

governments policies to ensure that global energy costs are kept competitive and companies finding

a balanced, rational and fair model of energy utilization based on the review of ethical business as-

pects and the establishment of social responsibility principles.

This paper develops the idea that the rules and procedures for future energy use must be based on a

reasonable equilibrium between technical aspects mainly associated to environmental issues, politi-

cal decisions made by governments and public administrations and social responsibility pro-

grammes that companies themselves must assume and implement.

The Supply and Copper Producer Response to a Growing Demand Scenario

Ricardo Alvarez, General Manager, Codelco El Teniente Division, Santiago, Chile

The prospects for medium and long-term consumption of copper are promising. A recovery pro-

jected for developed economies, starting 2010, joins the growing impact on demand generated from

the process of development and urbanization of emerging countries. The intensity of copper use will

also maintain the positive growth started a decade ago based on factors such as providing solutions

to combat global warming.

The objective of this presentation is to analyze how supply may react and adjust to envisaged de-

mand scenarios, bearing in mind some distinctive elements of the copper industry like:

- Historical supply reaction to demand and price levels

- Availability of resources and copper ore reserves, incorporating a dynamic analysis and the ef-

fects of geological exploration and possible technological changes.

- Pipeline of probable and possible projects, analysing the effects of projects in less developed and

riskier geographical locations

- Technological changes under development that may positively impact the amount of reserves

available and the competitiveness of projects.

- The growing role of scrap as a supply source for copper.

The analysis of these points will allow us to confirm the capability of supply response to growing

demand, ruling out revisited hypothesis of insufficient reserves

Implementation of Recent Global Copper Projects

Tim J. A. Smith, Vice President, Copper, SNC-Lavalin UK Limited, Croydon, Surrey, UK

Despite the recent global recession, a relatively strong copper price combined with continuing sup-

ply shortfalls continues to drive the implementation of a number of important new worldwide cop-

per projects.

Along with base metals projects in general, the scale and complexity of such projects has increased

such that multibillion dollar projects are increasingly common.

Proceedings of Copper 2010 XII

Together with expansion projects in the older traditional copper producing regions, new geographic

areas are still being opened up. These frequently require major infrastructural development, envi-

ronmental and global procurement capabilities as major components of such projects.

This plenary session address will examine and discuss the many project management challenges

and skills needed to deliver successful projects worldwide, as viewed from the perspective of one of

the world’s leading metallurgical plant engineers and constructors.

Proceedings of Copper 2010 XIII

Table of Contents – Volume 7

Plenary

Copper Sulphide Smelting: Past Achievements and Current Challenges 2543

Carlos M. Díaz

Sustainable Growth Strategy for Japanese Copper Business 2563

Toshinori Kato

Mineral Processing

Optimization of Copper Concentrate Bioleaching by 2575

Mixed Moderate Thermophile Bacteria

A. Ahmadi, M. Ranjbar, M. Schaffie, Z. Manafi

Industrial NSC Pressure Oxidation of Combined Copper and 2589

Molybdenum Concentrates

Corby G. Anderson, Todd S. Fayram, Larry G. Twidwell

HPGRs in Copper Ore Comminution – A Technology Broke Barriers 2621

E. Burchardt, N. Patzelt, J. Knecht, R. Klymowsky

Evaluation of Copper Losses in the Slag Cleaning Circuits from 2637

Two Chilean Smelters

N. Cardona, L. Hernandez, E. Araneda, R. Parra, L. Bahamondes, R. Parada,

J. Vargas, M. Artigas

Leaching of Gangue in Technological Flotation Circuits of 2655

Polish Copper Ores

Tomasz Chmielewski, Andrzej Luszczkiewicz

Proceedings of Copper 2010 XIV

Pressure Leaching of Shale Middlings from Lubin Concentrator in 2673

Oxygenated Sulphuric Acid

Tomasz Chmielewski, Jerzy Wódka

Mine to Heap in Mantoverde Anglo American Division 2693

Manuel Díaz, Cristian Salgado, Carlos Pérez, Cristian Alvayai,

Leonardo Herrera, Gabriel Zárate

Predicting the Effects of Locked, Partially Locked, and 2703

Liberated Minerals in Copper Leaching

Michael L. Free, Abraham L. Jurovitzki

Predicting Leaching Solution Acid Consumption as a 2711

Function of pH in Copper Ore Leaching

Michael L. Free

Copper from Pyrite – A Short History 2721

Fathi Habashi

Copper Crud Treatment, Concentration – Dependent Pond Depth 2737

Adjustment for Decanter Centrifuges, DControl®

Dipl.-Ing.Tore Hartmann, Dr. Ulrich Horbach, Jens Kramer

A Specific Electrode for “On-line” pH Measurement in 2747

SAG Cleaner Flotation Circuits

Christian Hecker C., Alejandro González S., Alejandra Mejías J., Claudia Rodríguez F.

Redox Potential Control in Column Leaching of Chalcopyrite 2753

Naoki Hiroyoshi, Takenari Kuwazawa, Yuki Takehara, Masami Tsunekawa

Separation Characteristics of Chalcopyrite and Pyrite via 2765

Bench Scale Flotation Investigations

S. Kelebek, Z. El Jundi, S. Reeves, H. Özdeniz

Proceedings of Copper 2010 XV

Bioleaching of Crude Chalcopyrite Ores by the Thermophilic 2781

Archaean Acidianus brierleyi in a Batch Reactor

Y. Konishi, N. Saitoh, M. Shuto, T. Ogi, K. Kawakita, T. Kamiya

Hybrid Flotation – Newly Developed Flotation Technology for 2793

Increased Recovery – Especially in the Finest Particle Fractions

W. Krieglstein, L. Grossmann

Perspectives of Copper Mining Industry Development in Poland 2807

Ph.D. Eng. Jan Kudełko, Ph.D. Jacek Pyra, Ph.D. Eng. Jerzy Sobociński

Control of Bubble Size in a Laboratory Flotation Column 2829

Miguel Maldonado, Dr. André Desbiens, Dr. Éric Poulin, Dr. René del Villar,

Alberto Riquelme

Flow Process in the Aerator of the Flotation Machine – 2845

Preliminary Simulations

Adam Mańka, Adam Fic, Andrzej Sachajdak, Ireneusz Szczygieł

Analysis of Fine Particles Behaviour in Flotation of Polish Copper Ores 2859

Ph.D. Eng. A. Potulska

HPGR versus SAG Milling Technology in Hard-Rock Mining – 2873

Review and Analysis

Irshad Rana, Kris Chandrasekaran, Ken Wood

Selective Leaching of Arsenic from Copper Ores and 2883

Concentrates Containing Enargite in NaHS Media

William Tongamp, Yasushi Takasaki, Atsushi Shibayama

Copper Leaching from Molybdenite in Acidic FeCl3 Solutions with FeCl2 2897

Yan Zhang, Narangarav Tumen-Ulzii, Zhibao Li

Proceedings of Copper 2010 XVI

Recycling

Shifting Core Business Vision: From Copper to Polymetallics – 2913

A Recycling Point of View

Juan Ignacio Barturen Zabala

Solubility of Scorodite Synthesized by Oxidation of Ferrous Ions 2923

Tetsuo Fujita, Etsuro Shibata, Takashi Nakamura

Biosolubilization of Copper from Waste Electric Cables 2935

Asst. Prof. S. Gaydardzhiev, D. Bastin, Dr. P. F. Bareel, Eng. F. Goffinet

Dowa Mining Scorodite Process® – Application to Copper Hydrometallurgy 2947

H. Kubo, M. Abumiya, M. Matsumoto

Recycling of Electric Home Appliances in Minamata 2959

Yoshihiro Watanabe

New Standards in Environmental Protection for Copper Recycling 2971

Dr. Franz-Josef Westhoff, Dr. Claus Meyer-Wulf

Posters

Recovery of Copper from Copper Smelter Wastewater by Electrodialysis 2986

Dr. Henrik K. Hansen, Dipl.-Ing. Claudia Gutiérrez, Dipl.-Ing. Jorge Ferreiro

Removal of Arsenic from Copper Smelter Wastewaters by 2988

Airlift Electrocoagulation

Dr. Henrik K. Hansen, Dipl.-Ing. Claudia Gutiérrez, M.Sc. Patricio Nuñez

Quantitative Mineralogy: X-ray Analytics of Copper Ores and Concentrates 2990

Dr. Karsten Knorr

Proceedings of Copper 2010 XVII

Continuously Cast Copper Alloys in Thin Dimensions and Their Applications 2992

Dr. rer. nat. Eberhard E. Schmid

Authors Index

Keywords Index

Proceedings of Copper 2010 2541

Plenary



Copper Sulphide Smelting:

Past Achievements and Current Challenges

Carlos M. Díaz

University of Toronto; Private Consultant

Adjunct Professor, Department of Materials Science and Engineering

210 Radley Rd., Mississauga

Ontario, L5G 2R7, Canada

Keywords: Copper, smelting, converting, oxygen, productivity, energy, environment, markets

Abstract

In the last three decades, increased oxygen consumption in copper sulphide smelting and converting

and the implementation of computerized process control, among other factors, have led to higher

process intensity and smelter concentrate processing capacity, decreased smelting energy consump-

tion and improved SO2 capture from process gas streams. An annual primary smelting furnace

throughput of one million tonnes of concentrate is the new industry standard. Top submerged lance

smelting has become an important processing route in recent years. Two new continuous converting

routes were commercialized in the 1990s. However, due to substantially improved converting prac-

tice and larger converters, Peirce-Smith converting still maintains its position as the dominant tech-

nology.

A major realignment of world copper smelting has taken place in the last 30 years. Spurred by rapid

economic expansion and the resulting huge increase in demand for basic materials, a number of

modern, large capacity copper smelters have been built in China, India and other Asian countries.

Only moderate growth in smelting and electrorefining capacity has taken place elsewhere. More-

over, ER cathode output in the USA has substantially decreased.

In this paper, the author examines recent technological advances and industry changes and high-

lights issues, such as energy consumption and the corresponding greenhouse gas emissions that will

become the focus of future discussion.

Díaz


1 Introduction

In August 1973, at a conference on copper held in Chile, Prof. H. H. Kellogg commented to partici-

pants at the copper smelting symposium: “I just wonder how many of you realize how fortunate you

are to be sitting at a meeting where you are discussing four new smelting processes for copper. I

have been attending meetings on extractive metallurgy for more than thirty years and this is a very

recent phenomenon. I can assure you, that year after year we would hear very little new; there would

be new ways of heating reverbs or new ways of punching converter tuyeres, but nothing in the way

of radically new process design. I believe that the present state of affairs holds great interest and

excitement” [1].

In the almost 40 years since Kellogg’s prophetic words, we pyrometallurgists have certainly lived

through most interesting and exciting times. During this period, the practice of anode copper pro-

duction from sulphide concentrates has undergone profound changes. The reverb furnace, still the

dominant copper smelting technology in the early 1970s, has been almost totally replaced by far

more energy efficient and environmentally sound flash and bath smelting technologies. Top sub-

merged lance smelting has become an important processing route in recent years. Continuous con-

verting has been adopted in new smelters. Tonnage oxygen consumption in both smelting and con-

verting has substantially increased, thus fulfilling Paul Queneau’s dream of “lifting the dead hands

of nitrogen from oxidation reactions which utilize the oxygen in air” [2]. Throughout the world,

increasingly stringent environmental regulations are being imposed on the industry, and its carbon

footprint is being examined as part of worldwide efforts to slow down climate change.

More recently, a profound realignment of primary copper smelting has taken place in the world.

Rapid economic expansion has spurred the construction of a number of modern, large capacity

smelters in China and India in the last 25 years. In 2008, the combined blister and anode copper

production of these two countries amounted to about 25 % of world’s copper smelter output. Indica-

tions are that this proportion increased further in 2009 and will continue to do so for the immediate

future.

Reduction of greenhouse gas emissions in primary copper production is possible only if significant

reductions in energy consumption in mining, milling, smelting and refining are achieved. This chal-

lenge offers new exciting opportunities to researchers and engineers.

The discussion in this overview paper focuses mainly on the following topics: a) Impact of increas-

ing consumption of tonnage oxygen on copper smelting process intensity and productivity;

b) Trends in smelting technology and smelter capacity; c) Realignment of world copper smelter out-

put; and d) Consumption of energy and greenhouse gas emissions in copper sulphide smelting as an

integral step of the chain of production of electrorefined copper from ore.



2 The Oxygen Age

The commercialization of Inco’s oxygen flash smelting at the Copper Cliff Smelter on January 2,

1952 [2] signalled the advent of the oxygen age in nonferrous pyrometallurgy. By the late 1950s,

oxygen-enriched air was being used in the Copper Cliff Smelter nickel and copper converters to

increase cold dope digesting capacity [3].

Preheated air had been chosen as the reacting gas in the Outokumpu flash smelting process that was

commercialized in 1949. Both the Inco and the Outokumpu furnaces generated SO2 continuous,

strong gas streams amenable to sulphur fixation in acid plants, thus offering an environmentally

friendly alternative to reverb smelting. However, only Outokumpu decided to offer its technology to

other copper producers. Japanese and European smelters, located close to densely populated areas,

were under social and political pressure to reduce pollutant emission and were the first to substitute

Outokumpu flash smelting for reverb furnaces. Furukawa’s Ashio Smelter commissioned its flash

furnace in 1956. A number of other Japanese smelters followed suit in the 1960s and early 1970s. In

Europe, Norddeutsche Affinerie commissioned a flash furnace in 1972.

Oxygen enrichment in an Outokumpu flash furnace was first adopted at Outokumpu’s own Har-

javalta Smelter in 1971. The smelter operators commented: “The leap to oxygen technology opened

completely new visions to the flash furnace designers as well. Oxygen enrichment obviously low-

ered the amount of nitrogen in the combustion air, allowing the possibility to build smaller furnaces

or make older ones more efficient” [4].

Oxygen-enriched air was also used as the reacting gas in the bath smelting processes that were

commercialized in the 1970s, namely the Noranda Reactor in Canada (1973), the Mitsubishi process

in Japan (1974), the Teniente Converter in Chile (1977) and the Vanyukov furnace in the former

Soviet Union (1977). The commissioning of the Mitsubishi process at the Naoshima Smelter in

1974 marked the dawn of continuous copper smelting-converting.

Besides the use of oxygen-enriched air or just tonnage oxygen as reacting gas, these new processes

had the following common features:

• Utilization of the heat of reaction of the sulphide minerals of the feed, FeS in particular, to satisfy

most, if not all, process heat requirements;

• High specific smelting rates;

• Production of high grade matte; and

• Production of steady, low-volume, SO2 strong process off-gas streams amenable to sulphur fixa-

tion in acid or sulphur dioxide liquefaction plants.

Díaz


3 Technology trends

The combined impact of tightening environmental regulations and a dramatic increase in the price

of oil in the 1970s triggered a major adjustment in the primary copper industry in the USA. Among

other measures, smelters were modernized in the late 1970s and in the 1980s. Outokumpu and Inco

flash furnaces, and Noranda Reactors were the technologies of choice [5]. The modernized plants,

most of them operating with a single primary smelting unit, had substantially larger capacities than

the extinct multi-reverb smelters. The plants that could not justify the investment required to mod-

ernize were closed. The number of American smelters was reduced from 15 in 1981 to 8 in 1987.

During the same period, in Chile, the need to increase smelting capacity, urgency to reduce con-

sumption of expensive imported oil, relatively low capital costs, and expediency in transferring lo-

cally developed technology contributed to the rapid substitution of Teniente converters for reverbs

in the state owned Codelco and Enami smelters [6]. The technology was later adopted in smelters in

Mexico, Peru, Zambia and Thailand.

In almost 20 years following the commercialisation of Mitsubishi Copper Continuous Smelting, the

process was selected for only one new smelter, Kidd Creek in Canada, which began operation in

1981. More recently, in the late 1990s and early 2000s, Mitsubishi smelters have been commis-

sioned in Korea, Indonesia and India.

At present, apart from the Horne Smelter Noranda Reactor which is fed with concentrate and pre-

cious metals containing electronic scrap, there is only one other Noranda Reactor, operating in

China, and a hybrid Noranda-Teniente Reactor operating in Chile. Another Noranda Reactor was

installed in Australia at the Port Kembla Smelter, but it ceased to operate when the smelter was shut

down in 2003.

Outokumpu flash smelting spread rapidly from Japan and Europe to other corners of the world.

Flash furnaces were substituted for reverbs in existing smelters (e.g. Chuquicamata and Chagres in

Chile) and were chosen as smelting units for green-field plants (e.g. Caraiba in Brazil, PASAR in

the Philippines, La Caridad in Mexico, Khatoon Abad in Iran, Birla in India). Outokumpu’s direct-

to-copper flash smelting technology made its debut at the Glogow II Smelter in Poland in 1978 [7].

In 1988, the process was adopted at Olympic Dam in Australia [8]. It was also adopted at the re-

cently commissioned Konkola Smelter in Zambia [9]. This technology is an attractive one-stage

copper smelting route. However, it is limited to the processing of concentrates with a high Cu/Fe(S)

ratio.

Top submerged lance (TSL) smelting has become a prominent copper concentrate processing tech-

nology in the last two decades. Developed in the 1970s at CSIRO, under the leadership of John

Floyd [10], TSL later evolved as two similar but separate technologies. John Floyd established

Ausmelt in 1981 to further develop and commercialize his invention. Zhong Tiao Shan, a Chinese

primary copper producer, started-up the first main stream Ausmelt copper smelter at its Houma



smelter in 1999 [11]. This plant also featured an Ausmelt converter. In 2003, a second Ausmelt

smelter was commissioned in China and another in India. More Ausmelt smelters have been built in

recent years, and more will come on stream in the near future.

Mount Isa Mines pursued the commercialization of TSL independently [12]. ISASMELTTM

fur-

naces for lead smelting were commissioned in 1983. Extensive piloting demonstrated the applicabil-

ity of the technology to copper smelting. Commercial copper furnaces were started-up almost simul-

taneously at Mount Isa and Cyprus Miami [13] in 1992. Following successive ISASMELT copper

furnace capacity increases, the last Mount Isa operating reverb was shut down in 1998. In the last

few years, additional ISASMELT copper furnaces have been commissioned in India, China, Peru

and Zambia.

The 2008 world copper concentrate smelting capacities for Outokumpu flash smelting, TSL (Aus-

melt and ISASMELT), Mitsubishi Copper Continuous Smelting, and combined Noranda-Teniente

bath smelting are summarized in Table 1.

Table 1: Annual overall copper concentrate smelting capacity of main technologies

Technology Number of smelters Annual concentrate

smelting capacity [kt]

Outokumpu Flash Smelting 27 20,720

Ausmelt 5 1920

ISASMELT 7 6075

Mitsubishi Copper Continuous Smelting 5 3300

Noranda/Teniente Bath Smelting 9 5900

The following sources of data were used for estimating the numbers presented in this table:

• Outokumpu – 2003 Copper Smelter Survey [14]; ICSG reports and statistics; private communi-

cations with industry representatives. Not included among the Outokumpu smelters are Konkola

(Zambia), commissioned in 2009, and smelters processing nickel-copper concentrates.

• Ausmelt and ISASMELT – Technical papers [15, 16]; private communications with owners of

respective technologies [17, 18]. Not included among the ISASMELT smelters is Chambishi

(Zambia), commissioned in 2009.

• Mitsubishi – 2003 Copper Smelter Survey [14]; technical monograph [19]. Kidd Creek, that it is

included among the Mitsubishi smelters, will be closed by mid 2010.

• Noranda/Teniente – Technical paper [20]; private communication with owners of respective

technologies.

Outokumpu flash smelting has been the dominant copper smelting technology in the last three dec-

ades. However, the TSL technologies have made an impressive advance in recent years. At present,

Díaz


there are four Ausmelt smelters under construction in China, with a combined concentrate process-

ing capacity of 2,950,000 tonnes. As well, new ISASMELT smelters will come on stream in the

near future in Russia, India and Peru, adding close to 2,000,000 tonnes to world copper concentrate

smelter capacity.

Using the sources of data listed in the notes to Table 1, the author has estimated that the combined

copper output of the 53 smelters included in the table amounted to about 80 % of the 2008 world

new copper smelter output of 12,565 kilotonnes (ICSG). These 53 smelters practice SO2 capture

from both smelting and converting process gas streams, usually as sulphuric acid. In most of the

other copper sulphide smelters, at least converter gases are also treated in acid plants. These obser-

vations suggest that the worldwide proportion of copper smelter sulphur input that it is currently

captured exceeds 85 %. This proportion should increase steadily in the near future, as modern tech-

nologies substitute for remaining reverbs and blast furnaces.

4 The quest for productivity increase

For the last two decades, copper smelting R&D has focused mainly on increasing productivity, a

trend that started in the mid 1980s. These efforts led to a new industry standard: the primary copper

smelter processing over one million tonnes of concentrate per year through a single smelting fur-

nace. Outokumpu and its licensees led the way to this new standard. The key factors that have con-

tributed to increasing flash furnace capacity are:

• high O2 enrichment of the reaction gas,

• improved solids feed system and concentrate burner design,

• water-cooling protection of furnace integrity,

• advances in process modeling and control,

• higher furnace operating factor.

High oxygen enrichment has been the most economic means of dramatically increasing flash fur-

nace capacity, while maintaining the dimensions not only of the furnace but also of the process gas

handling system. An impressive example in this regard is the conversion of the Sagoneski Smelter

from a two-furnace operation, each with a feed rate of 70 tonnes of dry solid charge per hour, to a

single furnace operation with twice that throughput, by substituting room temperature 75-85 % O2

enriched air for preheated low-oxygen enriched air (21-30 % O2, 1000 ºC), in March 1996 [21, 22].

Today, high oxygen enrichment of concentrate combustion gas is common practice not only in flash

smelting but also in other copper smelting technologies. Fourteen out of seventeen Outokumpu li-

censees, that responded the 2003 Copper Smelter Survey [14] questionnaire, were practicing flash

furnace oxygen enrichment above 50 %, and a few of them up to 80 % O2. The three ISASMELT

smelters that participated in the survey reported oxygen enrichment of the combustion gas in the 50



to 80 % range. Typical oxygen enrichment of the combustion gas in the Mitsubishi smelting furnace

is 55 % [14]. In bath smelting tuyere equipped vessels, such as the Noranda Reactor and the

Teniente Converter, oxygen enrichment is limited to about 40 %. Higher oxygen enrichment levels

could be achieved though by using shrouded tuyeres [23].

The author has estimated typical oxygen consumptions per tonne of anode copper for various copper

smelting and converting processes (see Table 2). These numbers are based on mass and heat bal-

ances that were run to determine energy consumption in copper sulphide smelting, using a typical

chalcopyrite copper concentrate analyzing 30.4 % Cu. In each case, the reacting gas consisted of a

mixture of air and 97 % O2 tonnage oxygen [24]. The Mitsubishi process operating numbers corre-

spond to the Gresik smelter that processes a 30.9 % Cu concentrate, using 99.5 % O2 tonnage oxy-

gen [19].

Table 2: Typical tonnage oxygen consumptions in copper smelting and converting

Process Concentrate feed

rate [t/h]

Matte feed rate

[t/h]

Reacting gas

[O2 %]

Tonnage oxygen

[kg/t Cu]

Flash Furnace 161 65 795

Flash Converter 74.4 60 272

Isasmelt 161 59 783

Mitsubishi S Fce 100 53.2 697

Mitsubishi C Fce 47.8 32.5 185

Nor/Tte Reactor 114 30.5 460

In European smelters processing lower grade concentrates (26-28 % Cu), the consumption of ton-

nage oxygen (95 % pure) in flash smelting exceeds 1 t/t of smelter product copper [25].

As discussed earlier, in flash smelting, the adoption of high oxygen enrichment of the reaction gas

has led to substantial increases in furnace concentrate throughput. This has required reengineering

the solids feeding system and the concentrate burner to adjust their operation to higher solids/gas

ratios. The long term flash furnace R&D program implemented by Sumitomo at their Toyo Smelter

is well documented [26-29] and provides a good example of the path to high levels of process inten-

sity. The Sumitomo scientists and engineers have made judicious use of physical-mathematical

modeling to improve burner design, with continuous validation and revision of model against pilot

plant and commercial plant data, a practice preached by Frank Jorgensen and the late Julian Szekely

[30]. It should be noted that all papers relating to the flash furnace capacity increase at Toyo empha-

size the importance of close, continuous cooperation between researchers and operators in achieving

R&D program goals.

The Toyo Smelter story can be examined from another angle by using a process intensity parameter,

the specific furnace volume smelting rate (SFVSR), developed by S. W. Marcuson in 1991 for

Díaz


assessing the specific capacity of copper converters [31] and employed by H. H. Kellogg and the

present author in 1992 for evaluating the intensity of a variety of bath smelting processes [32]. The

SFVSR is defined as follows:

Specific Furnace Volume Smelting Rate: Nm3 O2 consumed/h/m

3 of furnace volume (1)

The value of SFVSR for flash smelting for three different periods is presented in Table 3. Furnace

volume in the case of the flash furnace is the volume of the reaction shaft. The first line of the table

represents typical furnace operating conditions for the period 1949 to mid 1970s. The second line

reflects the situation up to the mid 1990s, while line 3 corresponds to the Toyo furnace operating at

73 % O2 enrichment. The data show that flash smelting intensity has more than tripled since the

early 1970s.

Table 3: Evolution of flash smelting intensity as a function of oxygen enrichment of reaction gas

Period Reaction gas [O2 %] Nm3 O2 consumed/h/m

3 of reaction shaft

1949–mid 1970s (Av 4 fces) Preheated air 55

Mid 1970s–mid 1990s

(Av 4 fces)

35 – 50 90

Toyo furnace (2006) 73 175

It is interesting to note that TSL smelting has achieved similar process intensity. Using data pro-

vided by the owners of the technology [17], the present author estimated that the Mt. Isa

ISASMELT furnace has a “specific furnace volume smelting rate” of 173.

In the last two decades, substantial advances have also been made in converting. Two new continu-

ous converting processes were commercialized in the 1990s, Kennecott-Outokumpu flash convert-

ing [33], and Noranda continuous converting [34]. The flash smelting-flash converting technology

was adopted at the Shandong Fengxiang Smelter in China. Ausmelt and ISASMELT have been de-

veloping their own continuous converting processes, C3 and ISACONVERT respectively. However,

larger vessels, higher matte grades, and improvements in engineering and operating practice have

allowed the PSC to continue as the dominant industry converting technology.

5 The giant smelter

To date, eight Outokumpu flash furnaces and three ISASMELT furnaces have an annual copper

concentrate smelting capacity in excess of one million tonnes. A word of caution is called for here.

In copper sulphide smelting, furnace concentrate processing capacity depends on the concentrate

Cu/S(Fe) wt ratio. At times, it is tonnage oxygen supply or process gas handling capacity that limits

the concentrate processing capacity of a smelter. Nevertheless, the fact is that copper sulphide

smelting has definitely entered the “age of the giant smelter”. The new copper content of the annual



product matte of large capacity Outokumpu flash furnaces or TSL furnaces generally exceeds

300,000 tonnes.

Using mainly ICSG data [35], the present author estimated that, in 2008, the combined new copper

production of smelters each with annual production capacity above 200,000 tonnes amounted to

about 75 % of world primary copper smelter output. In 1995, this proportion was about 60 %. Also,

rapidly increasing is the proportion of world copper output from smelters with an annual production

capacity in excess of 300,000 tonnes of copper. This is already above 20 %. This trend will un-

doubtedly continue in the future. In fact, two of the TSL smelters coming on stream in the immedi-

ate future have annual concentrate smelting capacities in excess of 1,000,000 tonnes. Larger plants

have lower unit production costs, an essential factor for the successful operation of existing and new

custom smelters.

6 The changing geography of copper smelting

The 1970s world recession had a profound impact on the copper industry. As discussed earlier, this

situation triggered a major realignment of the USA copper industry in the second half of the 1980s

and in the 1990s; a number of producers abandoned the field. More American smelters were shut

down during this period. At present, there are only three operating smelters in the USA.

Primary copper production changed little from 1975 to 1985 but started to rise again by the mid

1980s. In traditional copper smelting regions – Chile, Japan, Europe – modernization and expansion

of existing smelters had already more than compensated for the loss of USA smelter output. This

process has continued in the last 25 years. However, the awakening of the Chinese economy in the

early 1980s has had a more dramatic impact on the global geography of copper smelting. In fact, the

profound political and economic reforms implemented in China in the late 1970s and early 1980s

triggered a rapid expansion of the local economy and created an appetite for the essential materials,

copper among them, required for the modernization of the country’s infrastructure. Commissioning

of modern copper smelters, and guaranteeing concentrate feed for these plants, became part of the

country’s economic development program. Years later, through a different path, India’s economy

also began expanding at high annual rates. As in the case of China, access to basic materials became

a key element of India’s economic development. The huge expansion of Chinese and Indian copper

smelting capacities has been one of the most dramatic changes in world production of primary cop-

per in the last 25 years.

The evolution of smelter output in traditional copper smelting regions since 1987, first year of a new

period of steady increase of primary copper production, to 2008 [36] is shown in Table 4.

Díaz


Table 4: Evolution of smelter output in traditional copper smelting regions, 1987-2008 [kt]

Region 1987 1998 2004 2008

Chile 1107 1403 1518 1369

Europe 944 1462 1647 1645

Japan 871 1172 1270 1366

USA 972 1490 542 574

World 7580 10,100 12,000 12,400

The numbers in the table show the dramatic decline of the USA copper smelting output in the last

decade. During the period 1987–2008, the combined smelter production of Chile, Europe and the

USA, as a proportion of world production, fell from 51 % to 40 %.

The copper smelter outputs of China and India for the same period [36] are presented in Table 5.

Table 5: Evolution of copper smelter output in China and India, 1987-2008 [kt]

Region 1987 1998 2004 2008

China 300 839 1500 2500

India 30 108 401 651

World 7580 10,100 12,000 12,400

The proportion of world copper smelter output of the combined production of these two countries of

only 4.4 % in 1987 reached slightly over 25 % in 2008. Current TSL projects will add almost 3 mil-

lion tonnes to the present Chinese concentrate smelting capacity, and 1.35 million tonnes concen-

trate smelting capacity in India [17, 18]. In China, some new TSL furnaces will substitute for exist-

ing reverbs and blast furnaces, thus further increasing capture of smelter sulphur input.

A comparison of the numbers in Tables 4 and 5 with mine-mill copper production in the corre-

sponding regions and countries shows that huge amounts of concentrate are currently crossing the

oceans from regions with insufficient smelting capacity to regions with insufficient production of

smelter feed. The North and South Americans copper producers, in particular Chile and Peru, some

African countries, and Australia are in the first group, while Europe and Asia, in particular China

and India, are in the second group. At present, some 4.5 million tonnes of copper per year in the

form of concentrate are shipped from mostly mining-milling regions to essentially smelting regions.

This amount is well above one third of world copper smelter output.

A review of current projects indicates that, in the next few years, increases in world copper mine-

mill production will take place mainly in Chile and Peru, while most of the smelting capacity in-

crease will occur in China and India. As a result, even greater amounts of copper concentrate will

travel across the Pacific.



7 Copper smelting energy consumption and greenhouse gas

emissions

In this section, consumption of energy in copper sulphide smelting and the corresponding green-

house gas emissions will be discussed, considering smelting as one of four steps in the chain of pro-

duction of electrorefined (ER) copper cathode from ore.

Interest in this subject has increased in recent years. In Chile, Cochilco (Chilean Copper Commis-

sion), a government agency, conducted a comprehensive study of energy consumption and green-

house gas emissions in the local copper mining industry [37, 38]. Thirty-eight mining companies,

with a combined production accounting for 99 % of Chile’s copper output, responded a question-

naire that had been developed in consultation with industry. The Cochilco study determined annual

energy consumption numbers for the period 2004 – 2008. In another study, J. Marsden calculated

average energy consumption values for Freeport-McMoran primary copper operations in the USA,

Chile and Peru [39]. His study encompassed fourteen different operations; five of them producing

concentrate for smelting, and nine consisting of leaching, SX and EW. Cochilco’s and Marsden’s

numbers for mining and milling are presented in Table 6. For electric energy, the original Cochilco

and Marsden numbers are given in brackets. The other numbers in the same cells correspond to the

electric energy fuel equivalent, assuming that 100 % of the electric energy is produced in thermal

power plants with a conversion factor of 38 %.

Table 6: Energy Consumption Numbers in Mining and Milling Copper Ore [MJ/t ore]

Cochilco (2008) [37, 38] Marsden [39]

Unit Operation Electric Energy Fuel Total Total

Open Pit Mining 14.2 (1.5 kWh) 46.6 60.8 68.2

Milling 267.1 (28.2 kWh) 1.5 268.6 (104) 273.7

Cochilco’s and Marsden’s numbers are in reasonable agreement. The milling energy consumption

numbers include the energy equivalent of wear steel. In the ensuing discussion, average values, i.e.

64 MJ/tonne of ore for mining and 270 MJ/tonne of ore (28.5 kWh) for milling, are used. It should

be noted that while fossil fuel (diesel) is the main source of energy for mining, milling consumes

almost exclusively electric energy.

Available data on copper smelting energy requirements are presented in Table 7. For electric energy,

the original numbers are in brackets and side-by-side are the corresponding fuel equivalents. In his

paper, Piret already gives the fuel equivalent of his estimated smelting electric energy requirement

[40]. Energy consumptions in Table 7 are per tonne of ER cathode.

Díaz


Table 7: Energy consumption in copper smelting [MJ/t ER cathode]

Reference Electric Energy Fossil Fuel Total

Cochilco [37, 38] 13,395 (1413.9 kWh) 3990 17,385

Marsden [39] 13,279 (1401.7 kWh) 6296 19,575

Piret [40] 10,400 (1097.8 kWh) 5170 15,570

Cochilco’s numbers are averages for the seven Chilean smelters. Four of these smelters are

equipped with Teniente Converters, one with a hybrid Noranda/Teniente Reactor, another with an

Outokumpu flash furnace and Teniente Converters, and one with an Outokumpu flash furnace. All

of them use PSCs for processing high-grade matte to blister copper but with different slag cleaning

technologies. Waste heat recovery from process gases is the exception in Chilean smelters. The

Cochilco numbers presented in Table 7 were slightly adjusted to account for copper losses in smelt-

ing. Marsden’s numbers are based on Miami ISASMELT Smelter operating data. Piret’s numbers

correspond to a flash smelting-PS converting operation with an annual copper production capacity

of 150,000 tonnes. There is insufficient information available to explain the substantial differences

between the three sets of numbers given in Table 7.

In another paper presented to this conference [24], the authors discuss their own calculated energy

consumption numbers for the processing routes shown in Table 8.

Table 8: Energy requirements in copper smelting [MJ/t of ER cathode]

Processing route Electric energy Fossil fuel Total

Flash smelting + Flash converting 9266 (978 kWh) 1518 10,784

Isasmelt smelting + PS converting 6903 (729 kWh) 4175 11,078

Mitsubishi Continuous Smelting Process 8508 (898 kWh) 2498 11,006

Noranda/Teniente bath smelting + PS converting 10,088 (1065 kWh) 2657 12,746

The electric energy numbers in the table correspond to the fuel equivalents of the kWh numbers in

brackets. With the exception of Noranda/Teniente bath smelting, the difference in energy consump-

tion between processes is within the margin of error of assumptions plus model calculations. How-

ever, the electric energy and fuel consumptions vary within a wider range.

The calculated numbers in Table 8 are lower than the industrial numbers presented in Table 7. In

this regard, it should be noted that the smelting model reflects an “ideal operation”. In real life, there

are extra, sometimes unexpected, heating requirements (launders and ladles, maintaining furnaces

hot during short partial or total smelter shut downs, melting of larger than normal amounts of

reverts, etc.). In addition, the model does not include maintenance energy requirements.

For this discussion, the present author decided to use the average of Piret’s energy requirement for

flash smelting-PS converting given in Table 7 and the calculated energy consumption for Isasmelt



smelting-PS converting presented in Table 8. The smelting energy consumption is then

13,320 MJ/tonne of ER cathode, with an electric energy fuel equivalent component of 8650 MJ

(913 kWh), and a fuel component of 4670 MJ.

For the last production step, electrorefining, the writer selected the Cochilco number of 4710 MJ per

tonne of cathode, with an electric energy fuel equivalent component of 3210 MJ (339 kWh) and a

fuel component of 1500 MJ. Cochilco’s number is the average for the three Chilean copper refi-

neries [37, 38]. Cochilco’s electric energy component is in line with numbers reported in a 1995

world tankhouse operating data survey [41].

Mining and milling energy consumptions per tonne of ER cathode depend on ore grade and on mill-

ing and smelting metal recoveries. In Chile, where more than one third of world primary copper is

currently mined, the 2009 average Cu contents of mined copper sulphide ore and mined copper ox-

ide ore were respectively 0.9 % and 0.6 % [42]. These grades, though, have been declining steadily

over the years, in particular for copper oxide ore. Due to ore grade decline and mine stripping ratio

increase, Cochilco estimates that by 2020 mining energy consumption in Chilean copper operations

will be 20 % higher than in 2009 [42]. Smelting energy consumption is only sensitive to concentrate

grade that does not vary as much as ore grade.

The author has estimated mining and milling energy consumptions per tonne of ER cathode for an

ore analyzing 0.75 % Cu, assuming 90 % and 97.5 % metal recoveries in milling and smelting re-

spectively. Although 0.75 % is well below the current average grade of ore feeding Chilean mills, it

is certainly above the average grade of ore mined elsewhere. The estimated mining and milling en-

ergy consumptions as well as those for smelting and refining are presented in Table 9.

Table 9: Energy consumption for producing ER copper cathode from ore [MJ/t of product]

Electric energy fuel

equivalent Fossil fuel Total % of total energy

Mining 2090 (220.6 kWh) 7635 9725 14.1

Milling 41,000 (4,328 kWh) 41,000 59.6

Smelting 8650 (913 kWh) 4670 13,320 19.4

Electrorefining 3210 (339 kWh) 1500 4710 6.9

Total 54,950 13,805 68,755 100.0

The numbers in the table show that fuel consumption to generate electric energy is four times higher

than the fuel directly consumed in mining (mostly Diesel) plus smelting and refining (heating fuel).

Three quarters of the electric energy is consumed in milling.

In smelting, the most electric energy intensive operation is sulphur fixation from process gas

streams in acid plants. Copper smelting energy requirement calculations [24] show that acid produc-

tion in double-contact acid plants accounts for 35 % to 48 % of total electric energy consumption. It

Díaz


should be noted that sulphuric acid is used in a broad variety of industrial applications. In Chile, for

instance, practically all copper smelting generated acid is consumed in leach-SX-EW plants that

process copper oxide ore. Copper smelter acid is used for similar purposes in Peru and the USA. In

India, sulphuric acid is used for producing fertilizers in plants generally owned by companies that

also own smelters. The present author suggests, then, that the practice of attributing to copper smelt-

ing 100 % of the electric energy consumption in acid production and the corresponding greenhouse

gas emissions begs revision. Material life cycle assessment techniques should be used in this regard.

Turning now to the carbon footprint of primary copper production, and smelting in particular, as

shown in Table 10, direct (heating fuel) and indirect (burning fuel for power generation) CO2 emis-

sions are highly dependant on the type of fuel used [40, 43].

Table 10: Direct (heating fuel) and indirect (electric energy generation) CO2 emissions for various

types of fuel.

Type of fuel Direct CO2 emission [t/GJ] Indirect CO2 emission* [t/MWh]

Natural gas 0.056 0.53

Light oil 0.072 0.68

Heavy oil 0.078 0.74

Anthracite 0.092 0.87

Lignite 0.111 1.05

*Assuming power plant generation efficiency of 38 %.

After selecting a heating fuel and a power plant fuel or fuel mix, the numbers in Tables 9 and 10 can

be used to calculate the amount of CO2 emissions per tonne of ER copper cathode for each of the

production steps listed in Table 9. In smelting, using heavy oil for both direct heating and electric

energy generation, the CO2 emissions per tonne of ER cathode would amount to 1.04 tonnes.

The impact of the particular fuel mix used for power generation on greenhouse gas emissions per

MWh is illustrated by data on average CO2 equivalent emissions from the seven Chilean smelters

[37, 38]. In 2007, five of these smelters received electricity from a grid with a power generation

feedstock consisting of 47.1 % fossil fuel (less than one third was coal) and 52.9 % hydro. The other

two smelters were fed with almost 100 % thermally generated electricity (57.6 % coal, 22.6 % NG,

16.4 % diesel). Based on operating data for the two grids, Cochilco estimated that, in 2007, the av-

erage CO2 equivalent emitted by the Chilean smelters per tonne of copper was 0.86 tonnes.

In China and India, where copper smelting capacity is rapidly expanding, thermal power (over 70 %

coal) accounted for about 80 % of electric energy generation in 2006 – 2007. This situation is ex-

pected to change little before 2020.



The fact is that reducing the carbon footprint of primary copper smelting is directly related to de-

creasing energy consumption in this particular ER cathode production step. The following opportu-

nities could be explored to achieve this goal [24]:

• Recovering heat from a variety of combustion gases (anode furnaces, melt holding furnaces,

etc.), rare practice in smelters today.

• Substituting continuous fire refining for the current batch process. A continuous process has

already been piloted in Chile [44].

• In acid plants fed with high strength gas, utilizing the excess heat from the converter system for

a variety of purposes [45].

Conducting smelter energy audits would be an important first step in understanding and reducing

energy consumption [46].

8 Concluding remarks

It is most relevant to reiterate here some of the concluding thoughts about the future from the paper

that Phillip Mackey and the writer presented at Copper 2007 [47]. First, a few comments about

technology trends:

• The size of smelters will continue increasing. The average plant annual new copper production

capacity will reach about 250,000 tonnes in the next few years.

• The proportion of large custom smelters will also increase. High smelter productivity is essential

in this very competitive business.

• Flash smelting and TSL will compete for additional territory. In this regard, it remains to be seen

if Ausmelt, now owned by Outotech, will continue actively looking for opportunities in primary

nonferrous smelting.

• New green-field smelters and probably expanded/modernized smelters will incorporate continu-

ous converting.

• However, the Peirce-Smith converter will continue to have an important place in a number of

existing copper smelters.

• Progress will be made towards continuous anode copper refining.

• Advanced process control and automation will be introduced in all areas of the smelter.

• In the next few years, average world SO2 capture will exceed 95 %.

• There will be further migration of techniques from plants where fugitive emissions are largely

under control to those that lag behind.

• As energy prices continue rising and, eventually, carbon emission taxes are applied, the harness-

ing of waste heat from operations such as fire refining and holding furnaces will become in-

creasingly attractive.

Díaz


The changing geography of copper smelting also deserves a few words in these concluding com-

ments. Over the last 20 years, the sustained expansion of copper smelting-refining operations in

China and India has been progressively changing the role of the copper-rich South American coun-

tries to, mainly, that of concentrate suppliers. This process appears likely to continue in the near

future. In the case of Chile, for instance, Cochilco anticipates that the country’s copper production

will increase from 5.4 million tonnes in 2009 to 7.4 million tonnes in 2020. During that same pe-

riod, the proportion of Chilean copper exported as concentrate will increase from 35 % to

53 % [42]. Is this course of events in the best interests of Chile and its people? Although, this is a

legitimate question, the author understands that the answer requires in-depth analysis of an intricate

set of economic, social, environmental, political and international trade factors. A simplistic ap-

proach to this question suggests that expanding its existing copper smelting-refining capacity is one

option available to Chile to speed up its industrial development. It remains to be seen what type of

action, if any, may be taken in the future by the copper-rich countries to increase the value of their

copper exports.

Finally, it is the present author’s belief that pyrometallurgical processes will continue to have an

important place in the production of copper from sulphide feeds in the foreseeable future. The au-

thor’s message to his mentor, Herb Kellogg, is that nearly forty years after he enthusiastically com-

mented on the early 1970s copper smelting innovations, it is still most appropriate to say that “the

present state of affairs in copper pyrometallurgy holds great interest and excitement”.

Acknowledgements

The author expresses his appreciation to the Metallurgical Society of CIM and Atlas Copco for sup-

porting his participation in Cu2010. Thanks are also due to the many colleagues and friends in the

industry who contributed valuable information on various topics covered in this paper. Hans

Göpfert and Carlos Landolt were instrumental in uncovering comprehensive studies conducted by

Cochilco’s researchers on energy consumption and greenhouse emissions in Chile’s primary copper

production. The author expresses his profound gratitude to his friends Patricio Barrios, Pascal Cour-

sol, Phillip Mackey and Sam Marcuson for offering incisive comments on the manuscript. The edit-

ing assistance of Lucille Green is also greatly appreciated.

References

[1] DÍAZ, C. (Ed.) (1973): The Future of Copper Pyrometallurgy. – Proceedings of the Pyrometal-

lurgy Symposium, First Latin American Congress of Mining and Extractive Metallurgy, Chi-

lean Institute of Mining Engineers, Aug. 27th

-Sept. 1st 1973, p. II.

[2] INCO Staff (QUENEAU, P.E. & SPROULE, W.K.) (1955): Oxygen Flash Smelting. – JOM 7, 7:

742-750.



[3] SADDINGTON, R., CURLOOK, W. & QUENEAU, P. (1966): Tonnage Oxygen for Nickel and

Copper Smelting at Copper Cliff. – JOM 18, 4: 440-452.

[4] SARIKOSKI, T. (1999): A flash of knowledge; How an Outokumpu innovation became a cul-

ture. – Outokumpu Oij, Espoo, Finland, p. 247.

[5] DÍAZ, C., SCHWARZE, H. & TAYLOR, J. (1995): The changing landscape of copper smelting in

the Americas. – COPPER 95-COBRE 95, Vol. IV – Pyrometallurgy of Copper, Ed. W.J.

(PETE) CHEN et al., MetSoc of CIM: 3-28.

[6] SCHWARZE, H., ACHURRA, J. & DÍAZ, C. (2003): Development of the El Teniente Converter

technology. – COPPER 2003-COBRE 2003, VOL. IV – Pyrometallurgy of Copper (Book 2),

Ed. C. DÍAZ et al., MetSoc of CIM: 3-12.

[7] CZERNECKI, J. et al. (1999): Copper metallurgy at the KGHM Polska Miedz S.A. – present

state and perspectives. – COPPER 99-COBRE 99, Vol. V – Smelting Operations and Ad-

vances, Ed. D.B. GEORGE et al., TMS: 189-203.

[8] HUNT, A.G. et al. (1999): R.C., Developments in direct-to-blister smelting at Olympic Dam,

Ibid.: 239-253.

[9] SYAMUJULU, M. (2007): Opportunities, problems and survival strategies from recent devel-

opments in the copper concentrate treatment and smelting practices at Vedanta’s Konkola

Copper Mines in the Zambian copperbelt. – Cu2007, Vol. III (Book 1), Ed. A.E.M. WARNER

et al., MetSoc of CIM: 155-166.

[10] FLOYD, J.M. (2005): Converting an idea into a worldwide business – Commercializing smelt-

ing technology. – Met. Trans. B, Vol. 36B: 557-575.

[11] MOUNSEY, E.N., LI, H. & FLOYD, J.M. (1999): The design of the Ausmelt Technology smelter

at Zhong Tiao Shan’s Houma smelter, People’s Republic of China. – COPPER 99-COBRE

99, Vol. V – Smelting Operations and Advances, Ed. D.B. GEORGE et al., TMS: 357-370.

[12] PLAYER, R.L. et al. (1992) Top-entry submerged injection and the ISASMELT technology. –

Savard/Lee International Symposium on Bath Smelting, Ed. J.K. BRIMACOMBE et al., TMS:

215-229.

[13] BINEGAR, A.H. (1995): Cyprus ISASMELT start-up and operating experience. – COPPER 95-

COBRE 95, Vol. IV –Pyrometallurgy of Copper, Ed. W.J. (PETE) CHEN et al., MetSoc of

CIM: 117-131.

[14] RAMACHANDRAN, V. et al. (2003): Primary copper production – A survey of operating world

copper smelters. – COPPER 2003-COBRE 2003, VOL. IV –Pyrometallurgy of Copper (Book

1), Ed. C. DÍAZ et al., MetSoc of CIM: 3-106.

[15] SOFRA, J. & MATUSEWICZ, R. (2003): Ausmelt technology – copper production technology for

the 21st century, Ibid.: 157-172.

[16] ARTHUR, P.S. & PARTINGTON, P.J. (2007): Latest developments with copper ISASMELTTM

. –

Cu2007, Vol. III (Book 2), Ed. A.E.M. WARNER et al., MetSoc of CIM: 3-15.

Díaz


[17] ARTHUR, P. (Xstrata Technology), private communication.

[18] MATUSEWICZ, R. (Ausmelt), private communication.

[19] GOTO, M. & HAYASHI, M. (2002): The Mitsubishi Continuous Process. – Mitsubishi Mate-

rials Corporation, Tokyo, Japan.

[20] CABALLERO, C. (2009): Chilean copper smelting and refining – An update. – Molten 2009,

Gecamin, Santiago, Chile.

[21] SUZUKI, Y. et al. (1998): Productivity increase in flash smelting furnace operation at Saganoseki

Smelter and Refinery. – Sulfide Smelting 98: Current and Future Practices, Ed. J.A. ASTELJOKI

et al., TMS: 587-595.

[22] ISHIKAWA, M. (1998): High-intensive operation and productivity increase of flash smelting

furnace at Saganoseki Smelter and Refinery. – Metallurgical Review of MMIJ, 15, No. 2:

139-158.

[23] KAPUSTA, J., STICKLING, H. & TAI, W. (2005): High oxygen shrouded injection at Falcon-

bridge. – Converter and Fire Refining Practices, Ed. A. ROSS et al., TMS: 47-60.

[24] MACKEY, P.J., COURSOL, P. & DÍAZ, C.M. (2010): Energy consumption in copper sulphide

smelting. – Proceedings of Copper 2010, Volume 2 – Pyrometallurgy I, Ed. GDMB: 649-668.

[25] BARRIOS, P. (Aurubis): Private communication.

[26] INAMI, T. et al. (1991): Modification of concentrate burner for a copper flash smelting furnace.

– Copper 91-Cobre 91, Vol. IV: Pyrometallurgy of Copper, Ed. C. DÍAZ et al.; Pergamon

Press: 49-63.

[27] MORIYAMA, K., KEMORI, N. & KUROKAWA, H. (1995): Recent operation at the Sumitomo

Toyo Smelter. – COPPER 95-COBRE 95, Vol. IV – Pyrometallurgy of Copper, Ed. W.J.

(PETE) CHEN et al., MetSoc of CIM: 53-65.

[28] HATTORI, Y. et al. (2002): Developments of the Sumitomo Toyo FSF concentrate burner using

computational fluodynamics. – Sulfide Smelting 2002, Ed. R.L. STEPHENS et al., TMS: 305-

314.

[29] OJIMA, Y. et al.: (2003): Expansion project and advances at the Sumitomo Metal Mining Toyo

Smelter and Refinery. – COPPER 2003-COBRE 2003, VOL. IV – Pyrometallurgy of Copper

(Book 1), Ed. C. DÍAZ et al., MetSoc of CIM: 189-201.

[30] JORGENSEN, F.R.A. (2000): Flash smelting modeling. – The Brimacombe Memorial Sympo-

sium, Ed. G.A. IRONS et al., MetSoc of CIM: 333-347.

[31] MARCUSON, S.W. (1991): Copper converting: A historical Perspective. – Laurentian Universi-

ty Seminar Series on Copper and Nickel Extraction, Sudbury, Ontario, December 11.

[32] KELLOGG, H.H. & DÍAZ, C. (1992): Bath smelting processes in non-ferrous pyrometallurgy. –

Savard/Lee International Symposium on Bath Smelting, Ed. J.K. BRIMACOMBE et al., TMS:

39-65.



[33] GEORGE, D.B., GOTTLING, R.J. & NEWMAN, C.J. (1995): Modernization of the Kennecott

Utah copper smelter. – COPPER 95-COBRE 95, Vol. IV – Pyrometallurgy of Copper, Ed.

W.J. (PETE) CHEN et al., MetSoc of CIM: 41-52.

[34] PREVOST, Y. et al. (1999): First year of operation of the Noranda continuous converter. –

COPPER 99-COBRE 99, Vol. V – Smelting Operations and Advances, Ed. D.B. GEORGE et

al., TMS: 269-282.

[35] ICSG (2009): World Copper Smelters Capacity 2007 to 2012.

[36] USGS Copper Yearbooks 1987, 2002, 2008.

[37] Consumo de Energía y Emisiones de gases de Efecto Invernadero de la Minería del Cobre de

Chile. Año 2008 (Energy Consumption and Greenhouse Gas Emissions in Chile’s Copper

Mining – 2008), July 2009.

[38] PIMENTEL, S. (2009): The Chilean Copper Mining Sector: Energy Consumption and Green-

house Gas Emission Profile (2001-2007). – ENVIROMINE 2009, Santiago, Chile, Sept. 30-

Oct. 2, 2009.

[39] MARSDEN, J.O. (2008): Energy Efficiency and Copper Hydrometallurgy. – Hydrometallurgy

2008, Ed. C.A. YOUNG et al., SME, Phoenix, AZ: 29-42.

[40] PIRET, N.L. (2009): Will Today’s Needs Promote Copper Concentrate Hydroprocessing? Up-

date and Perspectives. – World of Metallurgy – ERZMETALL 62, No. 6: 344-365.

[41] SCHLOEN, J.M. & DAVENPORT, W.G. (1995): Electrolytic copper refining – world takhouse

operating data. – COPPER 95-COBRE 95, Vol. III – Electrorefining and Hydrometallurgy of

Copper, Ed. W.C. COOPER et al., MetSoc of CIM: 3-25.

[42] ZUNIGA, A.I. & PIMENTEL, S. (2009): La industria chilena del cobre frente al cambio climático

(The Chilean copper industry addresses climate change). – Seminario Internacional “Minería

del Cobre: Apostando al Futuro”, Cochilco, Santiago, Chile.

[43] OPFERMANN, A. et al. (2009): Improvement in Energy Efficiency. – Stahl und Eisen, 129, 9:

530-536.

[44] RIVEROS, G., et al. (2007): A new Paipote process of continuous fire refining of copper. –

Cu2007, Vol. III (Book 2), Ed. A.E.M. WARNER et al., MetSoc of CIM: 633-645.

[45] FRIEDMAN, L.J. & FRIEDMAN, S.J. (2007): The metallurgical sulfuric acid plant: Design, oper-

ating and materials considerations 2007 update. – Ibid.: 545-566.

[46] MACKEY, P.J. & COURSOL, P.: Private discussion with the writer.

[47] DÍAZ, C.M. & MACKEY, P.J. (2007): The Copper-Cobre series of conferences: A prime forum

for active discussion of copper smelting technology practice and innovation. – Cu2007, Vol.

III (Book 1), Ed. A.E.M. WARNER et al., MetSoc of CIM: 3-37.



Sustainable Growth Strategy for Japanese Copper

Business

Toshinori Kato

Mitsubishi Materials Corp.

1-3-2 Otemachi, Chiyoda-ku, 100-8117

Tokyo, Japan

Keywords: Copper fabrication, Mitsubishi Process, ASR treatment, environmental preservation,

copper business

Abstract

Business environment, for any industry, has changed dramatically over the last decade. Copper in-

dustry is no exception and those involved are experiencing an unprecedented period of upheaval.

The landscape of the market has completely altered, with large-scale M&As taking place among

miners – creating an oligopoly situation – and a rapid expansion of smelting capacity within devel-

oping countries, driven by strong economic growth. Traditional copper smelters and fabricators

have been facing challenges, but a long-term sustainability of Japanese copper business is achiev-

able.

Firstly, an immense amount of effort has been made over the years to develop the clean and one of

the most environmentally-friendly processes in the industry. Mitsubishi Materials (MMC) particu-

larly plays a great role in establishing Japanese smelters’ reputation as the most energy-efficient

operations in the world. It is our strong commitment to be a leader in this expertise by sharing our

technologies with the industry.

Secondly, infrastructures for copper smelting have been utilized in developing the recycling busi-

ness. Exemplified by the operations at Onahama smelter, which has the world’s largest furnace for

treatment of shredder residue, the industry has worked closely with other sectors and municipalities.

Building an effective structure to make the best use of our facilities is a key for success in this field.

It is fair to say that Japanese copper smelters, including MMC, are now regarded as an indispensable

part of national environmental policy.

Finally, progression of integration in the copper fabrication sector has strengthened the industry’s

capability of providing a variety of high-value-added products to end-users. Tight relationships with

the consumers have been beneficial in developing new use of copper. Evidence of a promising

Kato


future can be seen in increasing use of copper in the growing sectors such as hybrid automobiles and

renewable energy.

1 Introduction

Business environment of the copper industry – whether you are a miner, smelter or fabricator - has

changed dramatically over the past few years. Looking at the supply side, large-scale M&As have

taken place among miners – increasing miners’ strengths and creating an oligopoly situation in the

concentrates market. The global financial crisis may have caused delays in projects, in some cases,

but it is anticipated that the firm commodities markets will support a sustainable growth of mine

development. On the other hand, the demand for copper has surged in recent years, driven by strong

economic growth in developing countries especially China and India. This has led to a rapid expan-

sion in smelting capacity within these countries. In particular, within a very short period of time

China’s smelting capacity has surpassed that of Japan. Putting aside the discussion of a subsidized

growth, China’s increasing appetite for copper concentrates itself is no surprise, taking into account

of its potential for a strong economic growth.

In spite of the global financial crisis, demand for copper seems not to have been impacted. Govern-

ment fiscal stimulus plans implemented in developing countries have buoyed public spending, sup-

porting economic growth. Questions may arise about the sustainability of these initiatives, however,

as some see the growth as “artificial” and “strained”. Concerns remain about social instability and

growing awareness by the general public of the negative impact of key industry sectors on the envi-

ronment quality and human health.

Economic policies in some countries promote growth without taking into account other factors such

as security of raw materials. Policy goals such as maintaining employment and achieving target lev-

els of GDP growth – to be achieved by all possible means – are driving rapid expansion in business

without taking into account of profitability and the broader environmental and social impacts. Evi-

dences of various protectionist mechanisms have been observed to support these measures. The re-

sult of this is clear in so far the market has now been distorted by these companies operating in these

markets and also by the opportunistic players who are looking to benefit from market distortion with

little consideration given to operation of a fair market. As demonstrated by the recent financial crisis

and history, any distorted market will be subject to a correction, the level of correction reflecting the

level of distortion.

Partly in response to culture and the regulatory framework in which they operate Japanese smelters,

including MMC, recognize the importance of a sustainable approach and its importance to a fair

market place. Japanese smelters have a history of being conscious of their environmental impacts,

and have traditionally made significant investments in R&D and capital improvements in new tech-

nology to the benefit of the industry as a whole. In addition, high levels of awareness of reputational

risk and the potential financial impacts have resulted in development of environmental conscious-



ness along the supply-chain. Our commitment to environmental performance, managing risk along

the supply chain and contribution to the development of the industry comes at a cost which is unfor-

tunately not shared by companies in developing countries who typically benefit from these advances

without taking responsibility for their social and/or environmental impact. This paper will highlight

the significance of Japanese smelters, exemplified by MMC, by presenting various business compe-

tence cultivated through pioneering challenges for years.

2 Our commitment to environmental preservation

Japanese smelters have long been committed to preserving the environment. Along with Japan Min-

ing Industry Association, major Japanese players, including MMC, have been members of Interna-

tional Council on Mining & Metals (ICMM). This demonstrates our very clear and obvious com-

mitment to ensuring the sustainable operation of our business, encompassing partners in our supply

chain. As well as contributing to the overall perception of the mining and metals sector as being

socially and environmentally responsible, our investment and initiatives in this area benefit our

partners by minimizing the potential for reputational risk that they may be exposed to as a result of

operations. We have not been subject to any complaints or targeted by Non-Governmental Organi-

zations. MMC, for example, has received Dow Jones Sustainability Index (DJSI) listing because of

its investment and commitment to environmental performance. In addition, MMC has received the

following rewards from Japan Mining Industry Association in terms of environmental protection:

• Soil stabilization and tree planting activities in relation to forest development at Hosokura Mine,

2006,

• Technical development of anti-corrosion for the conbustible waste treatment at Naoshima copper

smelter, 2008.

It is the responsibility of ICMM member companies to minimize social and environmental impacts

associated with operations. The usage of our products by consumers also has to be taken into con-

sideration. Recognizing these factors will consequently contribute to minimizing the potential repu-

tational risks to our operations.

The key to a long-term success of the industry as a whole is to: make investments and continually

improve performance to minimize social and environmental impacts, as opposed to exploiting re-

sources, people and the environment to minimize our costs of production. This is the essential ap-

proach shared by all members of ICMM and it is in line with the commitments made by the mem-

bers. As an example of our investment and commitment to improving our environmental

performance, MMC has invested a significant amount into our Naoshima smelter to reduce atmos-

pheric emissions.

It is no coincidence that Japanese smelters have long kept the reputation of being the leader in clean

technology. As shown in Figure 1, Japanese smelters are by far the most energy-efficient in the

Kato


world. MMC is no exception and prides itself on its strong determination to contribute to society

through utilization of its cutting-edge technology.

Figure 1: Comparison of specific energy efficiency among copper smelters (by area)

(Source: Japan Mining Industry Association)

3 Expansive recycling business

Existing copper smelting plants have infrastructures and technologies to recover valuable metals,

such as Cu, Au, Ag, Pt, and to safely treat the hazardous impurities in concentrates without pollut-

ing the air, water and soil. In addition, pyrometallurgical copper smelting process utilizes the reac-

tion heat of Fe and S in concentrates. Therefore, utilization of copper smelting process is the most

appropriate choice from the point of view of energy efficiency and environment safety to recycle

valuable metals contained in scraps and also to treat industrial waste. MMC’s activity for recycling

business is introduced in this chapter.

Reverberatory furnace, as shown in Figure 2, is operated at Onahama smelter – actually two of them

in operation – to treat copper concentrates. Using the furnace is generally acknowledged as an old-

fashioned smelting process due to the fact that it requires a lot of fossil fuel to melt concentrates.

However, it has a huge combustion space, which can be utilized for treatment of industrial waste. In

1990’s, Onahama commenced treatment of combustible waste, such as automobile shredder residue

(ASR) and shredder residue from electrical home appliances(SR). Increasing the amount of treat-

0

50

100

150

200

250

Japan

EU AsiaN Am

erica

S America

En

erg

y-E

ffic

ien

cy



ment has resulted a reduction of fossil fuel consumption, as shown in Figure 3, and an increment of

valuable metals recycling.

Figure 2: Schematic figure of reverberatory furnace

Figure 3: Treatment amount of shredder residue and coal consumption

4ｍ

33ｍ11ｍ

Copper Conc.

Coal Burner

ASR / SR

To Boi

ler

Matte

Slag

Cu conc. Cu conc.

ASR / SR4ｍ

33ｍ11ｍ

Copper Conc.

Coal Burner

ASR / SR

To Boi

ler

To Boi

ler

Matte

Slag

Cu conc. Cu conc.

ASR / SR

Kato


Figure 4: Schematic flow sheet of the Mitsubishi Process

The Mitsubishi continuous copper smelting process, shown in Figure 4, is operated in the Naoshima

Smelter. It is the ideal smelting process: it requires low investment/operating cost, it is highly

energy-efficient, and pollution-free. Though the Mitsubishi Process is not suitable to treat industrial

waste due to its limited combustion space, it is suitable to treat solid particles containing valuable

metals thanks to its adoption of an injection system. In order to make efficient use of the system,

incinerating and melting plant for industrial waste, as shown in Figure 5, was introduced in 2004.

Slag/metal – containing valuable metals – from industrial waste are treated at the Mitsubishi

process. Combination of the processes has greatly expanded the potential of Naoshima’s recycling

business.

MMC takes pride in that it is one of the pioneers to be engaged in recycling business. Most impor-

tantly, with regard to ASR/SR treatment business, MMC holds a 20 % share in Japan. Main cus-

tomers in this industry include major Japanese automobile and home appliance manufacturers.

ASR/SR treatment enables the low-cost operation of Onahama’s reverberatory furnace – now the

world’s largest furnace for the treatment of SR.

MMC, with our copper and cement facilities shown in Figure 6, can recycle 100 % of various types

of industrial waste without any emissions or landfill, thereby contributing to the realization of a

recycling-oriented society.

Dried copper concentrate

Pulverized coal

Silica & C-slagLance air (50-60 % O

2)

Lance air (30-40 % O

2)

Lift tanks

S furnace

CL furnaceC furnace

Anode furnace

Lime, sludge & coolant (C-slag)

Pressurized tanks

Matte, slag

Matte

Scraps

Anode scrap

Casting machine

Blister copper

O� gasElectrodes

Granulated slag Granulated C-slag

Return to S, C furnaceSale



Figure 5: Schematic flow sheet of the incinerating and melting plant

Figure 6: Schematic flow of MMC’s recycling business

Slag and metalTrack scale

QuencherRotary kiln

Copper smelting and converting process

line of Naoshima Smelter&Refinery

Crane

Bag filter Catalyst tower

Sludge

supply - hopper

Combustible

supply - hopper

Waste reception pit

Calcium hydro-oxide

Scrubber

Fly ash

SmokestackInduced

draft Fan

Onerous recycling

materials

Automobile

shredder residue

“Washing treatment plant

for fly ash from incinerators”

SteamPower plant of

Naoshima

Smelter&Refinery

Caustic soda

Secondary

combustion

chamber

Waste heat boiler

Copper bearing

sludge

Printed circuit

board scrap

Household

appliance residue

CementPlants

CopperSmelters

Cement

Construction

materials and

concerete

Rolled copper

products

(Electric wire,

pipes and lead frames)

Ingots

Creating infrastructure

ThermalPower plants

Automobiles AppliancesLifestyles

Industrial

wastes

(Coal ash, sludge,and excavated soil)

Shredder dust,

battery and

copper scrap Discardedappliances

Wastetires

Appliance

recycling

plants

WasteplasticsWasteplastics

Metal-bearing

residue

Metal-bearingresidue

No!

Landfill

100%

Recycle

CementPlants

CopperSmelters

Cement

Construction

materials and

concerete

Rolled copper

products

(Electric wire,


IngotsCement

Construction

materials and

concerete

Rolled copper

products

(Electric wire,


Ingots

Creating infrastructure

ThermalPower plants

Automobiles AppliancesLifestyles

Industrial

wastes


Shredder dust,

battery and


Wastetires

Appliance

recycling

plants


Metal-bearing

residue


Industrial

wastes


Shredder dust,

battery and


Wastetires

Appliance

recycling

plants


Metal-bearing

residue


No!

Landfill

100%

RecycleNo!

Landfill

100%

Recycle

Kato


The industry has worked closely with other sectors and municipalities. Building an effective struc-

ture to make the best use of our facilities is a key for success in this field. It is fair to say that Japa-

nese copper smelters, including MMC, are now regarded as an indispensable part of national envi-

ronmental policy.

4 Integration in the copper fabrication sector

The copper business of the MMC group has established an integrated operating structure from min-

ing to making fabricated copper products and recycling. In particular, Mitsubishi Shindoh Co., Ltd

(MSC), a wholly owned subsidiary of MMC, produces high-value-added products in collaboration

with MMC from the development stage of copper alloy materials. MSC’s wide range of precision

terminal/connector materials and lead frame materials bound for the automobile, information tech-

nology and electronics markets are of first-tier quality and performance.

Furthermore, Eco Brass®

is highly regarded in Japan and overseas as an eco-friendly new material.

Eco Brass is a new lead-free copper alloy that ensures excellent cutting performance. It excels in

machining performance, such as cutting and forging with high strength, and has resolved issues such

as stress corrosion cracking and dezincification corrosion. Customer needs are increasingly diversi-

fied, thereby requiring the unique, demanding characteristics of thinner and lighter materials oper-

able under the harsh operating environments of the automobile and electronics industries. Moreover,

a more stable and swifter global supply system has become an essential requirement of customers

along with the expansion of emerging markets, such as China.

Furthermore, on November 27, 2009, MMC issued a news release on execution of share exchange

agreement with Mitsubishi Cable Industries (MCI), planning the conversion of MCI into our wholly

owned subsidiary. Over many years, MMC has built a close relationship with MCI through product

development and sales of copper wire rods, which are used for cable. MMC believes that, amid

growing efforts to realize a low-carbon society, the use of electricity as a form of green energy will

increase and, as a consequence, the demand for higher quality copper products with superior con-

ductivity and workability will grow. While deepening the relationship between the two companies,

it will enable the development and sales of new products to meet diverse market needs by combin-

ing MCI’s various technologies and its solid customer base with MMC’s oxygen-free copper and

alloy technologies, resulting in an overall strengthening of the MMC Group’s copper business. MSC

and MCI, belonging to MMC group, consume a large portion of refined copper produced by

Naoshima & Onahama. This means MMC holds a fairly stable market of refined copper. MMC’s

integrated supply chain is shown in Figure 7.



Figure 7: MMC group’s integrated supply chain

Moreover, MMC group has a tight relationship with final consumers of copper products (e.g., home

appliance and automobile industries) and the relationship is beneficial in developing new use of

copper. An example is shown in Figure 8. Our copper alloy tube is used for the heat exchanger in

the hot-water supply system utilizing a heat pump unit.

Figure 8: Hot-water supply system utilizing a heat pump unit

Vertical value chain of MMC metals businessCopper mine → Smelting & refining → Fabrication & alloys → Users

Smelting & refining Shapes • Wire rod Rolled sheets • Wire & cable

copper 350 kt

Shapes 190 kt Wire rod 130 kt

MSC 120 kt MCI + Copper anodes 80 kt

Naoshima re!ned 230 kt

Sakai Plant Cakes & Billets

130 kt

Onahama Copper Casting Div.

60 kt

Sakai Plant SCR ( *) 130 kt

(*) South Wire Continuous Rod system

MSC 120 kt

Anode for plating 20 kt

MCI etc. 60 kt

Anode for platingWire rod

Onahama re!ned 230 kt

75 % of 260 kt

Scraps 60 kt

Re!ned sales to others

90 kt

Shower

Bath

Kitchen

＜CO2 Heat Pump Unit＞

＜Water Tank Unit＞

Heat Heat

ExchangerExchangerAir

Water

High Copper Alloy Tubes are used for

heat exchanger in the heat pump unit

CO2

Cycle

Shower

Bath

Kitchen

Shower

Bath

Kitchen

＜CO2 Heat Pump Unit＞

＜Water Tank Unit＞

Heat Heat

ExchangerExchangerAir

Water

High Copper Alloy Tubes are used for

heat exchanger in the heat pump unit

CO2

Cycle

Kato


It reduces energy consumption by one-third and reduces CO2 emissions by approximately 50 %,

compared to previous hot-water supply equipment using fuel. Electric power utilities and air-

conditioner manufacturers are making significant efforts to market this highly efficient product.

Ministry of Economy, Trade and Industry has set a cumulative sales target of 5.2 million units by

2010; this product was only introduced in 2001.

Progression of integration in the copper fabrication sector has strengthened the industry’s capability

of providing a variety of high-value-added products to end-users. Tight relationships with the con-

sumers have been beneficial in developing new use of copper. Evidence of a promising future can

be seen in increasing use of copper in the growing sectors such as hybrid automobiles and renew-

able energy.

5 Concluding Remarks

Japanese smelters, including MMC, are and will be competitive and sustainable as copper smelters.

Many smelters in developing countries, sooner or later, will be faced with rising costs: increasing

labour costs and a significant amount of investment to be required in environmental measures. In

this context, continuous and pioneering efforts are paying off; Japanese smelters have developed

compatible – both energy-efficient and recycling-oriented – smelting processes, enabling the smelt-

ers to accumulate the know-how to maintain environmentally-friendly operations.

The physical characteristics of copper, its conductivity and workability, are what make copper valu-

able. Sustainable demand for the metal is anticipated for many years to come and, therefore, copper

smelters will continue to serve as important parts of the supply chain.

It is the mission of today’s copper smelters to utilize the valuable resources as effectively as possi-

ble without doing damage to the environment; it is obvious that smelters without sufficient tech-

nologies will not be able to survive in today’s environmentally-conscious world. An achievement of

such mission will greatly contribute to sustainable growth of global economy, and improvement in

living standards. Japanese smelters will continue to take the leading role in the industry by making a

contribution to developing and implementing a global solution to accomplish such mission.


Mineral Processing



Optimization of Copper Concentrate Bioleaching

by Mixed Moderate Thermophile Bacteria

A. Ahmadi a, b

, M. Ranjbar a, b

, M. Schaffie c, d

Z. Manafi

Shahid Bahonar University National Iranian a Mineral Industries Research Centre (MIRC) Copper Industry Company

b Mining Engineering Department

c Energy and Environmental Research Centre

d Chemical Engineering Department

Kerman, Iran Kerman, Iran

Keywords: Bioleaching, moderate thermophile, copper concentrate, optimization

Abstract

In this study, the bioleaching of Sarcheshmeh chalcopyrite concentrate by mixed iron- and sulphur

oxidizing bacteria was studied and the effects of some parameters, namely; temperature, pH, nutri-

ent medium and silver ions have been investigated on the iron and copper extraction from chalcopy-

rite concentrate by using a full factorial design. The experiments were done in shake flasks in 2 in-

cubator shakers at 150 rpm and pulp density 10 % (w/v) for 30 days. The chosen experimental

parameter levels were as follows: temperature, 44-50 °C; pH, 1.2-1.8; nutrient medium, 9K- Norris;

silver concentration, 0-30 mg/l. To predict optimum conditions, several models have been devel-

oped between the response variables (copper and iron recovery in leaching times 14 and 30 days as

well as cells number) and relevant parameters by means of variance analysis using the Design-

Expert software. The influences of these parameters and their interactions on response variables

were investigated. The optimum conditions for copper recovery were found to be as follows: Tem-

perature, 50 °C; initial pH 1.8; nutrient medium, Norris; silver concentration, 30 mg/l. The calcu-

lated copper recovery from concentrate was approximately 65 % under the optimum conditions. On

the other hand, , the maximums copper recovery (~70 %) and iron recovery (~40 %) were obtained

at an experiment that its redox potential during the run was lower than 420 mV (vs. Ag/AgCl) and

silver was added, while the minimum copper recovery was achieved when the redox potential was

maximum (700 mV). This behaviour was attributed to jarosite formation.

1 Introduction

Bioleaching is a novel technology that economically, technically and environmentally has a high

potential to extract base metals from sulphide recourses especially in developing countries. The

most abundant copper bearing mineral is chalcopyrite which compared with many other sulphide

Ahmadi, Ranjbar, Schaffie, Manafi


minerals of base metals; it is relatively recalcitrant to the chemical and bacterial leaching processes.

This recalcitrance has been attributed to both the formation of a passive layer at high ferric sulphate

concentration and its strong crystal lattice [1-3].

Using both iron and sulphur-Poxidizing microorganisms is very effective to improve the extraction

of metals from sulphide minerals [4]. Among the various microorganisms using in this process,

moderate and extreme thermophile microorganisms are more appropriate to dissolve copper from

chalcopyrite [5, 6]. Bioleaching of chalcopyrite by mesophile microorganisms always leads to a low

dissolution rate.

From the industry point of view, moderate thermophile bacteria are being preferred, because they

are more resistant to higher pulp densities and higher heavy metal concentrations than extreme

thermophiles [7, 8]. Moreover, leaching reactions raise the temperature of stirred tanks to their op-

timum range (40-50 °C), so process-cooling requirements are reduced [4]. Furthermore, the use of

extreme thermophile microorganisms has difficulties such as: lower oxygen solubility in water and

lower mechanical resistance than mesophilic and moderate thermophilic microorganisms [9]. Nev-

ertheless, bioleaching by moderate thermophile bacteria without additives hasn't a considerable

leaching rate. Several authors [10-14] have shown that chalcopyrite dissolution in bioleaching sys-

tems improves by adding silver ions as a catalyst. The interaction of silver ions with other relevant

parameters investigated in this study namely: temperature, type of nutrient medium and pH hasn't

been well investigated yet. Hence this research was done to elucidate the influence of the mentioned

parameters in the presence and absence of silver catalyst on copper and iron extraction and as well

as activity and growth of bacteria. A 24 full factorial design was used to conducting experiments in

shake flasks. The optimum condition for copper extraction was obtained and the relationship among

copper and iron recovery, redox potential and growth and activity of bacteria were investigated.

2 Materials and methods

2.1 Mineral

A copper flotation concentrate of the Sarcheshmeh Copper Mine in the region of Kerman located in

the south of Iran was used in all experiments. The concentrate had a size distribution of 80 % pass-

ing 75 micron. Chemical analysis of the concentrate by X-ray fluorescence has been shown in

Table 1. Mineralogical analysis was performed by optical microscopy of polished specimen and X-

ray diffraction (see Table 2). It shows that chalcopyrite is 76.5 % of copper bearing minerals of the

concentrate.

Table 1: Chemical analysis of the copper sulphide concentrate

Components Cu Fe S 2SiO

2 3Al O . .LO I Zn MgO 2

K O

Contents [wt. %] 27.50 23.03 14.82 8.29 2.72 17.44 0.99 0.66 0.57



Table 2: Mineralogical analysis of the copper sulphide concentrate

Mineral Chalcopyrite pyrite Covellite Chalcocite Non-metallic

minerals Oxide minerals

Contents [wt. %] 45 24 6.78 5.84 13.61 4.79

2.2 Microorganisms

A mixed culture of iron- and sulphur-oxidizing moderate thermophile bacteria was used in all ex-

periments. The mixture obtained from Sarcheshmeh copper complex (isolated by Mintek Company,

South Africa). These bacteria were grown in a mineral salts medium (9K or Norris, see Table 3) and

initially were adapted to pulp density of 2 to 10 % (w/v).

Table 3: The amount of basalt salts in nutrient media [gram/Litre]

Nutrient

Medium ( )4 42NH SO

4 2.7MgSO H O 2 4K HPO KCl

( )3 22.Ca NO H O

9K 3 0.5 0.63 0.1 0.014

Norris 0.2 0.2 0.2 - -

2.3 Bioleaching experiments

Bioleaching experiments were carried out in 500 mL Erlenmeyer flasks containing 250 mL suspen-

sions (nutrient medium and solid) in two orbital shakers agitating at 150 rpm. The flasks were in-

oculated with 10 % of adapted bacterial solution (v/v) (initial cells number 2.4×108 cells/ml).

The pH and ORP values in the leach solution were respectively measured with a pH meter (model:

Jenway 3540) and a Pt electrode in reference to an Ag/AgCl electrode (+207 mV vs. SHE at 25 °C)

respectively.

Periodically, both pH and ORP (vs. Ag/AgCl) values of the leach solution were measured during

the experiments with a pH meter (model: Jenway 3540). If pH was greater than desired value it was

reduced to the value with 6 M H2SO4.

Evaporation losses was measured by weighting and compensated by adding distilled water to the

flasks before sampling. Periodically, 10 mL of sample removed from the leach solutions and ana-

lyzed for copper and iron by atomic absorption spectrophotometery (AAS). The sample removed

was centrifuged for 5 min at 2500 rpm (model:SIGMA 3-16) to separate the residual solids and re-

placed with nutrient medium. The solid recovered was returned to the flasks. Finally, solids were

filtered from leach solution samples through Whatman No. 42 cellulose filter paper. The residues

were washed with distilled water and dried at 60 °C for sending to analyses of copper and iron by

AAS. Microbial populations were determined using a counting chamber and an optical microscope.



2.4 Design of experiments

In the present research, a two-level full factorial 24 design was chosen for conducting the bioleach-

ing experiments when the investigated variables were temperature, pH, nutrient medium and pres-

ence of silver ion (see Table 5) . The levels of variables are given in Table 2. Analysis of variance

method was used to investigate the effect of principal factors and their interactions on response

variables i.e. copper recovery, iron recovery and cells number.

Table 4: Factors and levels investigated in the full factorial 24 design

Factors

levels

A

Temperature [°C]

B

pH

C

Nutrient medium

D

Silver ion [mg/l]

Low (-) 44 1.2 Norris 0

High (+) 50 1.8 9K 30

Table 5: Experimental design matrix and results in bioleaching experiments

Run

No.

A

Temperature

[°C]

B

pH

C

Nutrient

medium

D

Silver ion

[mg/l]

Response variables

Copper

recovery %

Iron

recovery %

Cells

number 7( 10 )×

1 - - - - 46.57 29.01 32

2 + - - - 52.28 27.91 25

3 - + - - 55.23 38.82 29

4 + + - - 43.8 20.97 48

5 - - + - 51.16 31.29 20

6 + - + - 51.64 17.97 8

7 - + + - 50.67 29.77 56

8 + + + - 53.77 9.50 32

9 - - - + 68.71 37.89 32

10 + - - + 56.01 25.88 2.8

11 - + - + 53.36 20.76 32

12 + + - + 63.48 9.51 1.2

13 - - + + 55.12 27.99 52

14 + - + + 59.97 26.43 2.4

15 - + + + 64.27 16.91 56

16 + + + + 63.37 18.86 28



3 Results and discussion

3.1 Modelling

Significant factors on regression models are determined by analysis of variance. Using Design Ex-

pert software, first order regression models were obtained to predict the copper and iron recoveries

as well as cells number (Equations 1-5). Models of copper and iron recovery in a period of 14 days

were obtained to investigate the effect of time on the responses.

Statistical parameters of sum of squares, degree of freedom and mean square as well as P and F

values for the obtained models has been shown in Table 6. By considering this note that P values in

all of the presented models are less than 0.05, all of them are significant at the confidence level of

more than 95 %.

Copper and iron recovery and variation of oxidation reduction potential (ORP) during bioleaching

have been shown in Figure 1. It shows that the maximum copper recovery was achieved in run 9 at

temperature 44 °C, pH 1.2, nutrient medium Norris and presence of 30 mg/l silver ions.

Table 6: Regression models obtained to predict copper recovery, iron recovery and cells number

Prob > F

(P Value)

F

Value

Mean

Square df

Sum of

Squares Model

0.0415 3.56 33.58 5 167.92 Recov (%)(14 ) 41.93 2.45

1.56 0.48 0.29 1.33

Cu ery days A

B C D A C

= + ×

− × + × + × − × × (1)

0.0001 19.02 54.17 6 325.01 Recov (%)(14 ) 15.03 2.89 1.84

0.81 0.74 2.03 1.81

Fe ery days A B

C D A B A C

= − × − ×

− × + × − × × − × × (2)

0.0300 4.02 86.13 4 344.52 Re cov (%)(30 ) 55.99 0.35

0.80 0.24 4.55

Cu ery days A

B C D

= − ×

+ × + × + × (3)

0.0462 3.85 113.60 8 908.81

Recov (%)(30 ) 24.34 4.71

3.70 2 1.31 1.21

1.85 2.81 1.52

Fe ery days A

B C D A B

A D B D C D

= − ×

− × − × − × − × ×

+ × × − × × + × ×

(4)

0.0196 7.41 411.28 1

0 4112.78

( )( )30 29.70 8.93

7.93 2.10 1.55 3.30 5.28

5.92 3.27 2.08 4.35

Number of Cells Cell mL days A

B C D A B A C

A D B C B D C D

= − ×

+ × + × − × + × × − × ×

− × × + × × − × × + × ×

(5)

3.2 Influence of ORP

The ORP of a solution is a measure of its tendency to be oxidizing or reducing and is based on the

relationship between the solution Eh and the ratio of dissolved ferric to ferrous ions as given by

Equation 6 (the Nernst Equation) [15]. It is used as an indicator for bacterial activity in bioleaching

systems.



(6)

Where Eh or ORP is the solution redox potential (mV), [i] is the concentration of i species (g/l).

ORP changes observed during bioleaching are presented in Figure 1. It shows an increasing trend in

which its value increases from about 300 mV (vs. Ag/AgCl) to around 700 mV in some tests. This

increase is mainly due to oxidation of ferrous to ferric ions by iron oxidizing bacteria. In some ex-

periments specially those which done in the presence of silver ions, ORP increases slowly that is

due to the negative effect of silver ions on iron oxidation ability of bacteria. As Figure 1 shows, the

maximums copper recovery (~70 %) and iron recovery (~40 %) were obtained at run 9 which its

ORP was lower than 420 mV during the experiment and silver was added, while the minimum cop-

per recovery was achieved at run 3 which its ORP reached to around 700 mV. The least iron recov-

ery (~10 %) was obtained at runs 8 and 12 which their ORP reached to maximum values

(~700 mV). The obtained results confirm the results of other investigations in which the rate of

copper dissolution from chalcopyrite and the formation of passive layer strongly depend upon the

ORP of solution and its dissolution rate is maximum at a certain range of solution potential [16-21].

Figure 1: Cu and Fe recovery as well as ORP (redox potential) as a function of time during

bioleaching experiments at different conditions (see Table 5)

3

'

2

.ln

.H H

FeR TORP or E E

n F Fe

+

+

= +

�



3.3 Influence of temperature

Figures 2-A and 3 show the effect of temperature on copper and iron recovery. As can be seen in

Figure 2-A, by increasing temperature from 44 to 50 °C, copper recovery significantly increases in

a period of 14 days while by prolonging the time to 30 days, its influence is diminished. Further-

more, the effect of temperature on iron recovery has shown at different levels of temperature and

pH in both media (see Figure 3). It shows that increasing both temperature to 50 °C and pH to 1.8 is

led to a remarkable decrease in copper recovery especially in 9K medium. On the other hand, at low

level of pH, rising temperature in 9K medium has a negative influence on iron extraction while it

hasn't influence on its value in Norris medium. Low iron extraction is attributed to the precipitation

of a part of dissolved iron as jarosite (Equation 7) on chalcopyrite and reduces copper recovery.

3

4 2 3 4 2 63 2 6 ( ) ( ) 8Fe X HSO H O X Fe SO OH H+ + − +

+ + + → +, (7)

where 4 3, ,X K Na NH and H O

+ + + +

= .

Moreover, by prolonging leaching time to 30 days, high ORP (high ferric concentration) and high

temperature are two suitable factors to form jarosite. This justification is confirmed by remarkable

decrease of iron recovery at higher temperature and more leaching time (30 days). The influence of

temperature on decreasing iron extraction is more significant at high level of pH (see Figure 3).

Figure 2-B shows that increasing temperature in 9K medium is led to decreasing of cells number. It

seems that essential nutrient elements co-precipitates with ferric irons (Equation 7) leading to an

unsuitable culture medium to growth and activity of bacteria. High concentration basalt salts

(9K medium) is suitable for bacteria at low level of temperature, but when temperature is risen to

50 °C, iron-basalt salt precipitates is increased leading to the depletion of needed components in

medium.

Figure 2: A) Influence of temperature on copper recovery in leaching times 14 and 30 days,

B) Interaction of temperature and nutrient medium on cells number in the presence of

silver at leaching time 30 days

Num

ber

of

Bac

teri

a (C

ells

/mL

1*

10^7

)

A B



Figure 3: Interaction of temperature and pH on iron recovery at A) Norris medium and

B) 9K medium in leaching time 14 days

3.4 Influence of pH

pH of a bioleaching system is one of the most important operating parameters that influences on

metal dissolution and the growth and activity of acidophilic microorganisms [22, 23].

Effect of initial pH on copper recovery can be seen in Figure 4-A. It shows that increasing pH has a

little positive effect on copper recovery in leaching time 14 days, while in the leaching time of 30

days, changing pH doesn't influence on it significantly. pH values in the bioleaching system have a

transient increase then decrease quickly. The transient increase is due to occurring acid consuming

reactions which decreases to desired value by addition 6M H2SO4.

There are two reasons for strong decrease in the pH: one of them is the activity of sulphure oxidiz-

ing bacteria to convert sulphure and sulphide species to sulphate that produces acid according to

Equation (8):

2 2 2 43 2 Sulphur oxidizing microorganismsS H O O H SO+ + →�

(8)

And the other reason is the precipitation of jarosites according to Equation (7) that is an acid pro-

ducing reaction. The probability of jarosite formation at low level of initial pH is remarkably dimin-

ished.

The excess low pH may be harmful for growth and activity of bacteria. The negative effect of very

low pH value on the microorganisms and its positive effect on chemical leaching (as a leaching

agent) may neutralize each other. Figure 4-B shows that low level of pH at 9K medium is led to

decreasing cells number. More increase of pH suits the condition for jarosite precipitation which is

led to chalcopyrite passivation.



Figure 4: A) Influence of pH on copper recovery in leaching times 14 and 30 days,

B) Interaction of pH and temperature on cells number in leaching time 30 days

3.5 Influence of nutrient medium type

Microorganisms require nutrients for their metabolism and biosynthesis [15]. To investigate the

effect of nutrient medium on chalcopyrite concentrate dissolution, experiments were carried out in

9K (high amount of basalt salts) and Norris media (diluted medium) (see Table 3).

Figure 4-A shows the influence of temperature on copper recovery at leaching times 14 and

30 days. It exhibits that at lower leaching time, increasing temperature has a positive effect on cop-

per recovery while it would be diminished at higher leaching time (30 days). On the other hand, as

can be seen in Figure 3, rising temperature at 9K medium significantly decreases iron recovery. The

decrease is attributed to the precipitation of a part of dissolved iron as jarosite caused at high levels

of temperature and nutrient medium. As mentioned before, increasing temperature and basalt salts

is led to increasing iron precipitates. Considering the insignificancy of nutrient medium on copper

recovery (the target response) and the cheapness of Norris medium respect to 9K, using Norris me-

dium is chosen as a more suitable nutrient medium for the moderate thermophile mixed culture.

3.6 Influence of silver addition

Figure 5-A shows that at lower leaching time (14 days), the addition of 30 mg/l silver ion hasn't a

significant influence on copper recovery, while by increasing leaching time to 30 days, silver addi-

tion is the most effective factor. The important role of silver on copper recovery is exhibited at high

coefficient of D variable in Equation 3. The reason for negligible effect of silver addition on copper

recovery at lower leaching time is toxic effect of silver ions on the metabolism of bacteria leading

to prolonging of lag phase. While by increasing leaching time, bacteria are adapted with silver ions,

then because of the catalytic role of silver ions in the presence of bacteria, the copper dissolution

increases. Nevertheless, as Figure 5-B shows that silver addition decreases cells number even at

longer leaching time that is caused by toxic effect of silver on bacteria. On the other hand, Figure 6

Num

ber

of

Bac

teri

a (C

ells

/mL

1*

10^7

) A B



shows that silver addition has decreased iron recovery at the high level of pH. This decrease is at-

tributed to the co-precipitation of silver and iron and formation of silver-jarosite.

It was explained that silver forms an Ag2S film on the mineral surface according Equation 9 and

acts as a channel to transfer electrons leading to activation of the passive layer. Ferric ions oxidizes

the Ag2S (Equation 10). In bioleaching systems, bacteria oxidize sulphure and ferrous ions pro-

duced in Equation 10 and accelerate the rate of reaction [12].

SAgFeCuAgCuFeS 2

22

2 24 ++→++++

(9)

+++

++→+203

2 222 FeSAgFeSAg (10)

It was also found that Ag2S as a cathode and chalcopyrite as an anode form a galvanic couple, as a

consequence, the dissolution of chalcopyrite increases [12].

Figure 5: Influence of silver addition on A) copper recovery in leaching times 14 and 30 days,

B) cells number in leaching time 30 days

Figure 6: Interaction of pH and silver addition on iron recovery in leaching time 30 days

Num

ber

of

Bac

teri

a (C

ells

/mL

1*

10^7

)

A B



4 Conclusion

In this research, the influences of four important factors in bioleaching of chalcopyrite concentrate

were investigated and the main following results were obtained:

The optimum condition obtained from modelling of the experimental results using ANOVA was:

temperature 50 °C, initial pH 1.8, nutrient medium Norris and silver concentration 30 mg/l in which

its calculated copper recovery was approximately 65 %. Meanwhile, the maximums copper recov-

ery (~70 %) and iron recovery (~40 %) from experiments were obtained at an experiment (run 9)

which its ORP was lower than 420 mV during the test and silver was added, while the minimum

copper recovery was achieved at an experiment (run 3) which its ORP reached to around 700 mV.

The least iron recovery (~10 %) was obtained in experiments (runs 8 and 12) which their ORP

reached to maximum values (~700 mV). It seems that this behaviour is due to the passivation of

chalcopyrite by jarosite.

It was found that despite of significant negative effect of silver addition on the cells number, when

bacteria are adapted to silver ions, its addition is the most effective parameter to extract copper from

the concentrate.

Both increasing temperature and using 9K medium (high amount of basalt salts) especially in the

presence of silver ions are led to an intensive decrease in iron recovery that are attributed to the

formation of jarosite.

Acknowledgment

This research was a part of Electro-bioleaching Program for Iranian copper sulphide concentrates.

The financial support of the National Iranian Copper Industry Company is gratefully acknowledged.

References

[1] HABASHI, F. (1978): Chalcopyrite; its Chemistry and Metallurgy. Mac-Graw Hill, New

York.

[2] HABASHI, F. (2006): Chalcopyrite: Bioleaching versus pressure hydrometallurgy. Interna-

tional Conference of Metallurgy of the XXI Century, Almaty.

[3] AHONEN, L. & TUOVINEN, O. H. (1990): Catalytic effects of silver in the microbiological

leaching of finely ground chalcopyrite containing ore materials in shake flasks: Hydrometal-

lurgy, 24, 219-236.

[4] OLSON, G.J., BRIERLEY, J.A. & BRIERLEY, C.L. (2003): Bioleaching review part B: Pro-

gress in bioleaching: applications of microbial processes by the minerals industries. Appl Mi-

crobiol Biotechnol, 63, 249–257.



[5] NORRIS, P.R., BURTON, N.P. & FOULIS, N.A.M. (2000): Acidophiles in bioreactor min-

eral processing- Extremophiles, 4, 71–76.

[6] RODRIGUEZ, Y., BALLESTER, A., BLAZQUEZ, M.L., GONZALEZ, F. & MUNOZ, J.A.

(2003): New information on the chalcopyrite bioleaching mechanism at low and high tem-

peratures. Hydrometallurgy, 71, 47–56.

[7] OKIBE, N., GERICKE, M., HALLBERG, K.B. & JOHNSON, D.B. (2003): Enumeration and

characterization of acidophilic microorganisms isolated from a pilot plant stirred-tank

bioleaching operation. Applied and Environmental Microbiology, 69, 1936–1943.

[8] OLSON, G.J. & CLARK, T.R. (2004): Fundamentals of metal sulfide biooxidation. Mining

Engineering, 56, 40–46.

[9] CANCHO, L., BLAZQUEZ, M.L., BALLESTER, A., GONZALEZ, F. & MUNOZ, J.A.

(2007): Bioleaching of a chalcopyrite concentrate with moderate thermophilic microorgan-

isms in a continuous reactor system. Hydrometallurgy, 87, 100–111.

[10] BLAZQUEZ, M. L., ALVAREZ, A., BALLESTER, A., GONZALEZ, F. & MUNOZE, J.A.

(1999): Bioleaching behaviour of chalcopyrite in the presence of silver at 35° and 68°C. Bio-

hydrometallurgy and environment, PATR A, 127-137.

[11] PRICE, D. W. & WARREN, G. W. (1986): The influence of silver ion on the electrochemical

response of chalcopyrite and other mineral sulfide electrodes in sulfuric acid. Hydrometal-

lurgy, 15, 303-324.

[12] BALLESTER, A., BLAZQUEZ, L.M., GONZALEZ, F. & MUNOZ, J.A. (2007): Catalytic

role of silver and other ions on the mechanism of chemical and biological leaching. In: Do-

nati, E.R. and Sand, W. (Ed.), Microbial Processing of Metal Sulfides, 77–101, Springer,

Netherlands.

[13] HIROYOSHI, N., ARAI, M., MIKI, H., TSUNEKAWA, M. & HIRAJIMA, T. (2002): A new

reaction model for the catalytic effect of silver iIons on chalcopyrite leaching in sulfuric acid

solutions. Hydrometallurgy, 63, 257– 267.

[14] LÓPEZ-JUÁREZ, A., GUTIÉRREZ-ARENAS, N. & RIVERA-SANTILLÁN, R.E. (2006):

Electrochemical behavior of massive chalcopyrite bioleached electrodes in presence of silver

at 35 °C. Hydrometallurgy, 83, 63–68.

[15] ROSSI, G. (1990): Biohydrometallurgy. McGraw-Hill, Hamburg.

[16] HIROYOSHI, N., HIROTA, M., HIRAJIMA, T. & TSUNEKAWA, M. (1997): A case of

ferrous sulfate addition enhancing chalcopyrite leaching. Hydrometallurgy 47, 37-45.

[17] HIROYOSHI, N., MIKI, H., HIRAJIMA, T. & TSUNEKAWA, M. (2000): A model for fer-

rous-promoted chalcopyrite leaching. Hydrometallurgy, 57, 31–38.



[18] HIROYOSHI N., MIKI, H., HIRAJIMA, T. & TSUNEKAWA, M. (2001): Enhancement of

chalcopyrite leaching by ferrous ions in acidic ferric sulfate solutions. Hydrometallurgy, 60,

185–197.

[19] PINCHES, A., MYBURGH, P.J. & MERWE, C. (2001): Process for the rapid leaching of

chalcopyrite in the absence of catalysis. Patent No: US 6,277,341 B1.

[20] THIRD, K.A., CORD-RUWISCH, R., WATLING, H. R. (2002): Control of the redox poten-

tial by oxygen limitation improves bacterial leaching of chalcopyrite. Biotechnology and Bio-

engineering 78(4), 433-441.

[21] CORDOBA, E.M., MUNOZ, J.A., BLAZQUEZ, M.L., GONZALEZ, F. & BALLESTER, A.

(2008): Leaching of chalcopyrite with ferric ion. Part III: Effect of redox potential on the sil-

ver catalyzed process. 93 (3-4), 97-105.

[22] NARESH-KUMAR, A. & NAGENDRAN, R. (2007): Influence of initial pH on bioleaching

of heavy metals from contaminated soil employing indigenous Acidithiobacillus thiooxidans:

Chemosphere 66, 1775–1781.

[23] DAS, T., AYYAPPAN, S. & CHAUDHURY, G.R. (1999): Factors affecting bioleaching ki-

netics of sulfide ores using acidophilic micro-organisms. BioMetals, 12, 1–10.



Industrial NSC Pressure Oxidation of Combined

Copper and Molybdenum Concentrates

Corby G. Andersona, Todd S. Fayram

b, Larry G. Twidwell

c

aAllihies Engineering Incorporated

bContinental Metallurgical Services

cMontana Enviromet

P. O. Box 123

Butte, Montana, 59703, USA

Keywords: Copper, molybdenum, pressure oxidation, NSC, tails leach, flotation

Abstract

The need for copper and molybdenum continues to grow as does the need for clean efficient metal-

lurgical technologies capable of treating mixed metal concentrates. Currently, the use of hydrome-

tallurgical pressure oxidation for copper concentrate treatment is growing. Conversely, with limited

molybdenum roasting capacity, stringent industrial molybdenum concentrate roaster feed specifica-

tions, poor rhenium recoveries, and the inherent environmental issues associated with pyrometallur-

gical treatments, hydrometallurgical options are also now being pursued for molybdenum concen-

trates. Moreover, given the inherent grade, recovery and cost inefficiencies in the differential

flotation process normally employed for molybdenum concentrates produced as a by product of

copper mining, there is a growing need to directly treat combined bulk copper and molybdenum

concentrates. This would minimize molybdenum concentrate roasting limitations, specifications and

requirements while allowing simplification of and cost reductions in upstream mineral processing

circuits producing separate copper and molybdenum concentrates by differential flotation. It would

also allow more direct and efficient recovery of rhenium. Finally, hydrometallurgical technology

will also limit the need for costly final molybdenum concentrate impurity treatment circuits thereby

allowing for lower grade mixed metal molybdenum concentrates to be treated directly for a fuller

metal value realization. In summary, industrial Nitrogen Species Catalyzed (i.e. NSC) hydrometal-

lurgical pressure oxidation has many advantages over conventional pressure oxidation systems and

offers a tangible process route. This paper illustrates the fundamanetal concepts, confirmatory test-

ing, application, proposed design concepts and cost estimates for the development and industrial

implementation of this proven mode of hydrometallurgical processing.

Anderson, Fayram, Twidwell


1 Introduction

As a first step, the basics of Nitrogen Species Catalyzed (i.e. NSC) hydrometallurgical pressure oxi-

dation will be outlined. The commonly reported leach reaction of a sulfide mineral with nitric acid

in conjunction with sulfuric acid is shown below.

3MeS (s)+ 2HNO3 (aq) + 3H2SO4 (aq) � 3MeSO4 + 3S° (s) + 2NO (g) + 4H2O (1)

However, it has been postulated and confirmed that the actual reaction species is NO+ and not NO3

-

(Anderson 1992, Anderson 1996, Baldwin 1996, Gok 2009). The addition of or presence of NO2-

instead of NO3– accelerates the formation of NO

+. As shown in Table 1, the NO

+/NO couple is ca-

pable of an extremely high redox potential (Peters 1992). So, NO+ is readily formed from nitrous

rather than nitric acid. For example, a convenient source of nitrous acid can be sodium nitrite (An-

derson 1992, Anderson 1996). When it is added to an acidic solution, nitrous acid is readily formed.

NaNO2 (aq) + H+ � HNO2 (aq) + Na

+ (2)

Nitrous acid further reacts to form NO+.

HNO2 (aq) + H+ � NO

+ (aq) + H2O (3)

The NO+ then reacts with the mineral and oxidizes the sulfide to sulfur.

MeS (s) + 2NO+ (aq) � Me

2+ (aq) + S° + 2NO (g) (4)

Of course, at higher temperatures and/or nitrogen species concentrations the sulfide can be fully

oxidized to sulfate.

Table 1: Relative potentials of hydrometallurgical oxidizers.

Oxidant Redox equation E°h (pH = 0, H2 ref.)

Fe3+

Fe3+

+ e- � Fe

2+ 0.770 V

HNO3 NO3- + 4H+ +3e

- � NO + 2H2O 0.957 V

HNO2 NO2- + 2H

+ + e

- � NO + H2O 1.202 V

O2 (g) O2 + 4H+ + 4e

- � 2H2O 1.230 V

Cl2 (g) Cl2 (g) + 2e- � 2Cl

- 1.358 V

NO+ NO

+ + e

- � NO 1.450 V

As can be seen, nitric oxide gas, NO, is produced from the oxidation of sulfides. As this gas has a

limited solubility in aqueous solutions, it tends to transfer out of solution. In the pressure leach sys-

tem, a closed vessel with an oxygen overpressure is used. The nitric oxide gas emanating from the

leach slurry accumulates in the headspace of the reactor where it reacts with the supplied oxygen to

form nitrogen dioxide gas. The NO is then regenerated to NO+. Overall this can be viewed as:

Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates


NO (g) + O2 (g) � 2NO2 (g) (5)

2NO2 (g) � 2NO2 (aq) (6)

2NO2 (aq) + 2NO (aq) + 4H+ � 4NO

+ (aq) + 2H2O (7)

Since the nitrogen species is continuously regenerated, its role in the overall reaction as the actual oxidiz-

er is not obvious. The net overall reaction has the sulfide mineral reacting with the acid solution and oxy-

gen to solubilize the metal value into the sulfate solution and form some elemental sulfur.

2MeS (g) + 4H+ + O2 (g) � 2Me

2+ (aq) + 2S° + 2H2O (8)

Of course, at higher temperatures and/or nitrous acid concentrations the sulfide would be fully oxi-

dized to sulfate.

Overall, the nitrogen intermediates serve as an expedient means to transport oxygen to the surface of

the solid particle and allow the resulting reaction to take place at a heightened redox potential. This

inherent asset of the unique system eliminates the need for the use of high temperatures and high

pressures, which lead to higher costs in other pressure leach processes. For example, commonly

available stainless steel can be used for the reactor vessel. And, complete oxidation of sulfide to

sulfate can be achieved without the excessive conditions found in other pressure leach systems.

Thus, the rapid kinetics of the system leads to smaller reactor volumes and higher unit throughputs.

Finally, 99.9 % of the nitrogen species utilized in the leach system report to the gas phase when the

pressure vessel is flashed and they are readily destroyed and contained by commercially available

scrubber systems. So, environmental impacts are minimized and the NSC leach plant solutions con-

tain little or no nitrogen species.

2 Confirmatory NSC pressure oxidation combined copper and

molybdenum concentrate testing

Confirmation testing of the application of industrial NSC pressure oxidation to a mixed chalcoprite

and molybdenite concentrate was undertaken. Previous studies and plant operating data had con-

firmed the applicability of NSC to chalcopyrite concentrates. (Anderson 2003c). Further, the appli-

cations of nitrogen species acid leaching systems for treatment of molybdenum concentrates is well

documented and has been industrially piloted. (Peters 1976, Kerfoot et. al. 1976, Weber and

Borrmann, 1975). In addition recent work suggests that the application of industrial hydrometallur-

gical pressure oxidation of molybdenum concentrates is currently being implemented (Olsen, 2008,

Freeport McMoRan Copper and Gold, 2010). Hence, a low grade, out of specification molybdenite

concentrate with appreciable chalcopyrite, pyrite and rhenium content was selected for confirmatory

testing. Table 2 illustrates the elemental analysis of the chosen concentrate while Figure 1 denotes

the inherent quantitative mineralogical composition as determined by SEM Mineral Liberation

Analysis.



Table 2: NSC tested combined molybdenum copper concentrate [%];

Concentrate particle size = 80 % passing 75 microns.

Mo Cu Re Fe TS TC

34.50 1.59 0.159 2.42 33.23 1.90

Figure 1: MLA combined concentrate mineralogy

2.1 Design of experimentation based NSC pressure oxidation mixed

concentrate confirmatory testing

Using Stat Ease Design Expert computer software, a statisticaly valid set of tests for NSC pressure

leaching of the combined copper and molybdenum concentrate were designed as shown in Table 3.

The key variable parameters studied included grind time (i.e. 0, 5 and 10 minutes) slurry solids

concentration (i.e. 10, 30 and 50 g/L), initial sulfuric acidity (i.e. 25, 50 and 75 g/L), reaction time

(i.e. 30, 60 and 90 minutes). For all these tests, total pressure was fixed at 90 psig, sodium nitrite

addition at 2.0 g/L and autoclave stirring was done at 700 RPM.

The testing was carried out and solids and liquids were analyzed by ICP for Re, Cu and Mo.

Recoveries were then calculated and the data inserted into the Stat Ease program for optimization.

The recovery data is also shown in Table 3.



Table 3: Stat Ease Design of Experimentation Matrix and NSC POX Mo, Re & Cu Recovery Data.

Factors Responses

Std A:Grind B:Solids C:Initial D:Temp E:Reaction Mo Rec Re Rec Cu Rec

Time [min] [g/L] Acidity

[g/L]

[°C] Time [min] [%] [%] [%]

1 10 50 25 150 90 17.6 69.4 87.7

2 0 10 25 150 90 84.3 80.2 97.9

3 0 50 75 150 30 14.6 7.9 37.4

4 0 50 75 130 90 16.7 31.3 48.5

5 10 10 75 150 90 82.9 87.9 95.7

6 0 10 75 130 30 10.2 0.0 22.6

7 10 50 75 130 30 15.7 9.7 23.9

8 10 10 25 130 90 70.2 31.9 63.1

9 10 10 25 150 30 66.8 30.0 59.9

10 0 50 25 130 30 13.2 5.4 23.5

11 5 30 50 140 60 23.6 25.8 48.2

12 5 30 50 140 60 23.9 29.6 53.5

13 5 30 50 140 60 23.4 27.7 45.3

3 NSC confirmatory testing results and discussion

Figures 2, 3 and 4 summarize the Stat Ease optimization of modelling of the confirmatory NSC

testing. In essence, the model fit for Mo, Cu and Re pressure leaching with NSC was excellent. For

good molydenum recovery, grinding, higher temperatures and longer reaction times help at low

molybdenum solids content. This is probably a function of molybdenum solubility limits. For good

copper recovery, higher temperatuess and longer reaction times lead to better recoveries. Finally for

rhenium, lower slurry solids densities and longer reaction times are the key elements for enhanced

reovery. In addition, per the published literature and industrial operating practices, using both ion

exchange and activated carbon, selective separation and concentration of both rhenium and

molybdenum from copper, iron and other metal cations present in acidic mixed metal solutions

produced in NSC testing was successfully carried out.



Figure 2: Copper recovery from chalcopyrite with optimized NSC pressure oxidation of combined

copper and molybdenum concentrates

Figure 3: Molybdenum recovery from molybdenite with optimized NSC pressure oxidation of

combined copper and molybdenum concentrates

Design-Expert® SoftwareOriginal ScaleCu Recovery, %

97.85

22.6

X1 = E: Reaction TimeX2 = D: Temp

Actual Factors

A: Grind Time = 10B: Solids = 10C: Initial Acidity = 25

30

45

60

75

90

130

135

140

145

150

32.0

49.5

67.0

84.5

102.0

C

u R

ecovery

, %

E: Reaction Time D: Temp

Design-Expert® SoftwareOriginal ScaleMo Recovery, %

84.25

10.24

X1 = A: Grind TimeX2 = D: Temp

Actual Factors

B: Solids = 10C: Initial Acidity = 25.00E: Reaction Time = 90

0.00

2.50

5.00

7.50

10.00

130.00

135.00

140.00

145.00

150.00

12.0

32.0

52.0

72.0

92.0

M

o R

ec

ov

ery

, %

A: Grind Time D: Temp



Figure 4: Rhenium recovery with optimized NSC pressure oxidation of combined copper and mo-

lybdenum concentrates

4 Industrial application and design considerations

Based on industrial comminution, flotation and NSC pressure oxidation testing, industrial operating

plant data and pertinent design criteria a comprehensive flowsheet was formulated for treating a

copper and molybdenum ore body. The resultant flowsheet is split in half for better clarity and illu-

strated as Figures 5A and 5B. A detailed description of the process and its detailed development

follows.

Based on a review of the available geologic resources and various contained mineral types, both an

oxide and sulfide treatment scheme was required for the orebody. A literature review was underta-

ken of various copper deposits around the world with the intent of maximizing the copper recovery

using the smallest footprint possible. The resource proposed to be treated is listed in Table 4.

Design-Expert® SoftwareOriginal ScaleRe Recovery, %

87.92

0

X1 = E: Reaction TimeX2 = B: Solids

Actual FactorsA: Grind Time = 10.00C: Initial Acidity = 75.00D: Temp = 150.00

30.00

45.00

60.00

75.00

90.00

10.00

20.00

30.00

40.00

50.00

26.00

41.50

57.00

72.50

88.00

R

e R

ecovery

, %

E: Reaction Time B: Solids



Table 4: Illustrative resource proposed to be mined and processed via open pit mining, bulk con-

centrate flotation, tailings agitated leaching and NSC pressure oxidation of combined

copper and molybdenum concentrates.

Deposit Indicated Resources [106 t] Grade

Cu % Mo %

North 69.258 0.370 0.005

South 45.148 0.377 No Data

Total 114.406 0.373 ---

Deposit Inferred Resources [106 t] Grade

Cu % Mo %

North 18.166 0.271 0.005

South 25.593 0.278 No Data

Total 41.759 0.275 ---



Figure 5A: Flowsheet for comminution, flotation concentrate production, and flotation tails agitated

leaching and industrial NSC pressure oxidation treatment of combined copper and mo-

lybdenum moncentrates



Figure 5B: Flowsheet for comminution, flotation concentrate production, and flotation tails agitated

leaching and industrial NSC pressure oxidation treatment of combined copper and mo-

lybdenum moncentrates

A site visit identified an undetermined amount of sulfides located within the oxide portions of the

ore body. Based on this find, an oxide heap leach scenario was dismissed as leaving an excessive

amount of unleached sulfides on the pad that would potentially create environmental issues and ul-

timately not maximize the value of the deposit. Based on this, a flotation/tails leach/pressure leach



scenario (i.e. a flowsheet similar to that of First Quantum Minerals – Kansanshi Operation in Afri-

ca) was developed.

Based on information gleaned in part from the previous metallurgical testing, adjacent operations,

and past experience, the following parameters were used to develop a realistic flowsheet based on

proven industrial concepts.

4.1 Comminution

Previous work undertaken consisted of six Bond Ball Mill Work Indices, indicating a range of 13.4

to 16.2 kWh/t (14.8 to 17.8 kWh/t), and the results reflect a final screen size of 65 % passing 200

mesh. Overall, the data, together with inspection of the drill cores, indicates that the ore is generally

hard and highly competent.

A detailed program is recommended to identify and confirm the range of ball mill work indices.

This work should be expanded to include the testing of Rod Mill Work Indices, Unconfined Com-

pressive Strength (UCS), Abrasion Indices, and JK Drop Weight Tests.

Since no SAG Mill testing has been completed on the project, a review was performed to determine

a basis point for the prefeasibility comminution design of a SAG Mill grinding circuit. The review

included the investigation of the comminution circuits at both adjacent mills at as well as several

other North American copper projects. The investigation examined grinding conditions with similar

work indexes and rock types. Based on this review, the basic design of the Asarco South Mill (Pi-

ma) was chosen as a base case for the flow sheet development.

4.2 Sulfide Flotation

Previous testwork undertaken has provided useful background data for the development of the cur-

rent flowsheet. Key findings of the work include the following:

• Flotation of sulfide material is relatively simple and conventional with high recovery and good

concentrate grades.

• Separate sands and slimes circuits do not appear to be necessary.

• Oxide and mixed ores respond poorly to flotation. As a consequence, oxide and mixed ores will

be treated in a tails leach system with oxide ore by-passing the flotation circuit.

Most of the testwork was carried out on drill core with separate testing being completed on leaching

and flotation, but not on both together. Further testwork will be required to confirm whether or not

there is an effect on the leaching circuit due to excess organic carry over from the flotation circuit.



4.3 Oxide Leaching

Preliminary bottle roll testwork undertaken identified the following:

• Potentially long residence times of 24 hours at a pH range of 1.0 to 1.2

• Estimated recovery of 85 to 95 % of the soluble copper in 24 hrs at 61 microns

• Major improvement in extraction rate through decreasing grind size

• Gangue acid consumption has varied between testwork programs and is dependent on carbonate

content, but seems to be insensitive to grind size

• Gangue acid consumption is estimated at 17 kg/t between samples and needs to be modeled

across the orebody as a component of the mine schedule

During testing of the vat leaching, an agitated leaching test of the 38 micron thickener underflow

material was undertaken to identify why there were unleached fines. The testing of the material

identified the material leached well at a pH of 1.9 for 48 hours and obtained a total extraction of

approximately 73 % of unleached fines.

4.4 Solvent Extraction

Solvent extraction testwork was completed on resource material for the development of the vat

leaching. The SX circuit consisted of two-stage extraction and two-stage stripping. The organic

phase employed in this circuit was 25 % v/v ACORGA®

M5774 (Cytec) extractant in Exxsol®

D80

(Exxon Mobil aliphatic solvent) diluents matrix. Prior to introducing the PLS from the vat leach to

the SX feed, the solution was filtered through 1-micron string-wound cartridge filters. The PLS was

collected in batches using 1000 L carboys. Each batch was filtered, stirred to homogeneous, sam-

pled, and then fed to the SX circuit.

The SX circuit was fed PLS solution with a copper grade averaging 2.75 g/L Cu and it generated a

raffinate bearing an average 0.038 g/L Cu during the final two days of operation (during steady-state

conditions).

The copper grade in the aqueous solution in the preceding two stages of extraction averaged 0.046

and 0.181 g/L, respectively. This equates to a calculated recovery of 93.4%, 98.3 %, and 98.6 % as

the SX feed solution was sequentially processed through the 3-stage extraction circuit.

The pH of the SX feed and raffinate solution during this same period averaged 1.74 and 1.50, re-

spectively. The copper concentration in the loaded and stripped organic averaged 6.52 and 3.17 g/L.

Some crud formation was observed in the extraction and strip mixer-settlers. During the short test-

ing period, the crud did not appear to interfere with the SX operation. The strip solution employed at

start-up was a 235 g/L sulfuric acid solution. When the strip solution reached 45 g/L copper the EW

process was initiated.



During EW operation, the copper grade in the rich and lean electrolyte averaged 47.5 and 35.4 g/L

Cu, respectively. The free acid averaged 169 and 189 g/L H2SO4, respectively. Prior to starting the

EW process, the strip solution was dosed with sufficient guar gum to achieve 10 mg/L in the strip

solution inventory. Cobalt sulfate salt was also added to the strip solution, which yielded ~25 mg/L

Co in solution.

The target electrolyte temperature and current density were 40 °C and 270 A/m2. The actual meas-

ured temperature of the electrolyte and current density were 31.1 ºC and 313 A/m2. Three copper

plates weighing a total of 35.2 kg were harvested. The deposit was dense and the appearance re-

flected typical cathode surface morphology. The nominal purity of the metal was 99.97 %. This

could be improved with the use of different anodes and fine tuning of reagent additions to the EW

cell. Overall copper recovery from the circuit was 98 %. The basic design parameters noted above

were used in the flowsheet design along with a design review by Cytec below in Table 5 and 6.

Table 5: Vat leaching flowsheet design parameters.

Circuit configuration 2 Extract +2 Strip

Organic 30 v/o Acorga M5640

PLS 10-15 g/L Cu, pH 1.8

Lean electrolyte 25 g/L Cu, 185 g/L H2SO4

O:A Ratio- Extract 1.0-1.4:1 Strip 1.5: 1

Stage efficiency Extract 95 % Strip 95 %



Table 6: Cytec vat leaching flowsheet design review.

File Org.

Flow

PLS

Flow

L.E.

Flow

PLS O:A

Ext.

O:A

Strip

M5640 Raffi-

nate

Rich E. Recovery Produc-

tion

Produc-

tion

[g/m] [g/m] [g/m] Cu pH (v/o) [g/l] [g/l] [%] [t/day] [t/year]

CAB 1,500 1,500 1,000 10.0 1.8 1.00 1.50 30 0.73 48.90 92.7 80.9 27,650

1A 1,650 1,500 1,100 10.0 1.8 1.10 1.50 30 0.60 47.82 94.0 82.9 28,068

2 1,500 1,500 1,000 11.0 1.8 1.00 1.50 30 1.09 49.86 90.1 84.9 29,558

2A 1,650 1,500 1,100 11.0 1.8 1.10 1.50 30 0.85 48.83 92.2 88.1 30,273

3 1,500 1,500 1,000 12.0 1.8 1.00 1.50 30 1.59 50.62 86.8 84.9 30,988

3A 1,650 1,500 1,100 12.0 1.8 1.10 1.50 30 1.22 49.69 89.8 88.1 32,156

4 1,500 1,500 1,000 13.0 1.8 1.00 1.50 30 2.21 51.19 83.0 80.0 32,193

4A 1,650 1,500 1,100 13.0 1.8 1.10 1.50 30 1.71 50.40 86.9 92.4 33,689

4B 1,800 1,500 1,200 13.0 1.8 1.20 1.50 30 1.36 89.5 95.2 34,748

5 1,500 1,500 1,000 14.0 1.8 1.00 1.50 30 2.93 51.60 79.1 90.5 33,033

5A 1,650 1,500 1,100 14.0 1.8 1.10 1.50 30 2.30 50.95 83.5 93.8 34,237

5B 1,800 1,500 1,200 14.0 1.8 1.20 1.50 30 1.84 50.20 86.9 99.5 36,301

5C 1,950 1,500 1,300 14.0 1.8 1.30 1.50 30 1.50 49.41 89.3 102.2 37,295

6C 1,950 1,500 1,300 15.0 1.8 1.30 1.50 30 1.98 50.03 86.8 106.5 38,852

6D 2,100 1,500 1,500 15.0 1.8 1.40 1.50 30 1.66 49.30 89.0 109.1 39,822

No significant concerns were noted. A laboratory-scale pilot plant run will be required to confirm

the operating parameters and examine the build-up of deleterious elements such as silica and crud

under a flotation/leach setting.

A relic copper cementation circuit will be set-up on the raffinate to maximize copper recovery. The

cementation process will be similar to that of Kennecott’s Spedden process and used to ensure mi-

nimal copper is sent back to the tails leach and potentially lost. The cemented copper product will

be recycled back to the pressure oxidation circuit. Subsequent, testwork will be required to optimize

the cementation circuit design, recovery, and economics.



4.5 Thickening

Limited thickening or rheological testing has been completed. Only the slimes from the Vat leach-

ing tests had thickening and rheological testing. Industry averages were used to size the thickeners

and other necessary settling equipment in the flotation and leaching circuits. The following design

parameters were used in developing the thickener characteristics:

4.5.1 Thickener design criteria

Flotation Tails

Thickener U/F density – 50 % solids

Flux rate t/m2/h – 0.52

The current design uses conventional thickening and assumes a 50 % underflow density. Further

testing will be required to ensure that high compression thickening will not be required to make the

necessary 50 % underflow prior to leaching.

Post Leach



Flocculent may be required to meet the specific design criteria noted above.

The current design uses conventional thickening and assumes a 50 % underflow density. A case

study comparing filtration with counter-current decantation (CCD) for the leach residue application

will need to be reviewed. Based on the available data, CCD typically has slightly better overall eco-

nomics, with significant performance advantage and reliability in the cleaning of the leach residues.

The process design will need to review the use of high compression thickeners to maximize under-

flow densities and minimize losses of copper through the CCD thickeners.

NSC pressure oxidation (POX) residue



Flocculent may be required to meet the specific design criteria noted above.

The current design uses conventional thickening and assumes a 50 % underflow density. A case

study comparing thickening to filtration will be required to optimize this part of the circuit. Further

thickening and rheological testing will be required on all aspects of the operation.



4.6 Filtration

No vacuum and pressure filtration tests were carried out in conjunction with the Pregnant Feed So-

lution thickening tests. Industry averages for chalcopyrite concentrate filtration were used. The fol-

lowing design parameters were used in developing the filtration characteristics:

4.7 Filtration design criteria

Copper Concentrate (Includes concentrate thickener)

(Assumes 65 % solids, product has a P80 of 325 mesh)


NSC POX Material (includes any vertical mill ground material to include leach residues)

(Assumes 65 % solids, product has a P80 of 10 micron)


Further filtration test data will be required for the finalization of the flowsheet.

4.8 Mixed copper and molybdenum concentrate leaching

The sulfide flotation concentrate will be treated using industrially proven Nitrogen Species Cata-

lyzed Pressure Oxidation Leaching (NSC). A vertical grinding mill will be fed concentrate contain-

ing a nominal 29 % copper and approximately 0.9 % molybdenum and ground to approximately 10

microns. The ground concentrate is then fed to one of two pressure oxidation autoclaves using the

following leaching parameters listed in Table 7.

Table 7: NSC pressure oxidation combined concentrate leaching parameters.

Parameter Unit Value

Solids flowrate t/d 202

Slurry solids content g/l 50

Initial free sulfuirc acid g/l 25

Reactor working pressure kPag 620

Maximum temperature °C 155

Total leach time min 45



From the autoclave, a Cu/Mo bearing solution is recovered from the autoclave effluent via a high-

rate thickener and polishing filtration, and then through a series of activated carbon columns for

molybdenum recovery. Then an EcoTec Acid Purification Unit (APU) is used to recover excess

sulfuric acid which is recycled back to the autoclave feed system. After molybdenum and acid re-

covery, the copper-laden leach solution is sent to a conventional SX/EW circuit and the copper re-

covered as “electrowon” cathode.

From industrial operating data, the NSC process results in near total oxidation (+99 %) of the sul-

fide concentrate and is estimated to convert a majority of the sulfide sulfur to sulfuric acid (60 %)

based on operating experience, and some to elemental sulfur (40 %). The elemental sulfur is col-

lected in the POX residue through thickening and filtration. The POX residue is then leached for the

sulfur content with sodium hydroxide to make a soluble sodium sulfate and a final leach residue.

The NSC POX residue leaching parameters are listed in Table 8.

Table 8: NSC POX residue leaching parameters.

Parameter Unit Value

Solids flowrate t/d 126

Slurry solids content g/l 50

Reactor working pressure kPag 620

Maximum temperature °C 130

Total leach time min 45

Approximately 50 % of the POX residue will be recycled back to the NSC leach for enhanced cop-

per and molybdenum recovery and to provide inert heat sink materials to the NSC POX system.

This helps to optimize the concentrate pressure oxidation system by providing more heat capacity.

The other estimated 50 % of residue will be sent to tails. The amount recycled and sent to tails will

be optimized to maximize throughput and metal recoveries.

In the overall POX residue reaction, sodium sulfate is formed. After thickening and polishing filtra-

tion, the sodium sulfate solution is passed through a set of carbon columns to recover relic molyb-

denum. The purified sodium sulfate solution will be further processed via electrodialysis (EDU) and

evaporation to recover a 70 % sulfuric acid product and a 100-g/l sodium hydroxide product. These

will be recycled back into the process plant. Any excess sodium sulfate will be sent to a multiple

effect crystallizer, crystallized into flake sodium sulfate, filtered, dried, bagged, and sold. The relic

molybdenum recovered will be processed as identified below.

The molybdenum-laden activated carbon is atmospherically stripped with ammonium hydroxide to

make an ammonium molybdate solution. The pH of the stripped solution is then decreased to 2.0

using sulfuric acid causing ammonium molybdate to precipitate. The precipitated ammonium mo-

lybdate is then filtered with the clear solution being recycled back to the process. The filtered am-



monium molybdate solid is then calcined at 550 °C, converted to molybdenum trioxide, bagged, and

sold. The ammonia driven off by the calcination process is scrubbed, collected, and reused in the

system. The stripped carbon will be recycled back to the molybdenum recovery circuit. A portion of

the carbon will also be reactivated in a conventional kiln operation.

Both the NSC concentrate and leached residue pressure oxidation processes can provide significant

heat. For efficient process, heat, or energy economy, the resultant heat will be used in the process

plant to increase the atmospheric flotation tailings leach temperatures, thereby improving copper

leaching kinetics and ultimately increasing copper recovery levels. This waste heat will also be uti-

lized in the EDU sulfuric acid recovery and sodium sulfate multiple effect evaporation system.

The production rate for the NSC concentrate pressure oxidation circuit is designed with excess ca-

pacity so outside copper and molybdenum containing concentrates can be processed through the

plant on a toll or purchased basis. Further testing will be required to ensure the optimized parame-

ters of the NSC pressure oxidation system are identified and quantified for concentrate treatment

and for final sizing of the various items associated in the metal recovery systems.

4.9 Crushing and ore storage and reclaim

Initial ore to the plant is expected to be relatively easy to treat. A small portion is highly weathered

and the remainder is highly competent with low moisture content values reported at approximately

3%. The proposed crushing circuit utilizes a C160 Nordberg Jaw Crusher that would crush run of

mine rock to a nominal 152 mm. The jaw crusher feeds a live-surge capacity stockpile which is no-

minally designed for 24 hours of live capacity, and six reclaim feeders will feed the two proposed

grinding circuits.

4.10 Grinding and classification

The two milling circuits have been designed on the basis of the use of a SAG mill and ball mill

combination. The design is based on the Asarco South Mill (Pima Mill). The ore milling circuits

have been designed using a SAG Mill Work Index of 5.58 kWh/t to 1200 microns, and a Ball Mill

Work Index of 16.5 kWh/t. The Bond Ball Mill Work Index for the ore is reported to vary from 15.4

to 17.6 kWh/t, and there is no significant difference between the sulfide and oxide material. On this

basis, the installed power for the SAG mill grinding 5.2 Mt/yr of ore is calculated at 8.5 MW.

The designed steel charge for the SAG Mill consists of a nominal ball load of 8 % by volume, with

a design load of 12 %. Abrasion indices are based on the nearby Highmont Mill which are consi-

dered moderate and have been incorporated into the operating costs. Values for the abrasion indices

vary from 0.5 to 0.7. The design cyclone overflow product size is at a P65 of 74 micrometer for the

downstream processes of leaching and/or flotation.



4.11 Flotation

Oxide and sulfide ores will be campaigned separately with all of the ores initially treated by flota-

tion and leaching. As oxide ore is depleted, the leaching circuit will be bypassed. The flotation of

the sulfide ore is based on a conventional rougher/scavenger and two-stage cleaning circuit. Rough-

er concentrate potentially will be reground to liberate sulfide copper minerals from gangue. Testing

will be required to confirm the need for a regrind mill. The selected reagent regime is relatively

simple, using only xanthate as collector combined with a frother. The plant has been designed to

maximize recovery of sulfide mineralization.

Recovery of acid soluble copper mineralization is not critical as the flotation tailings will be acid

leached. General reagent consumption and residence times have been based largely on recent test-

work performed with flotation prior to leaching. Locked cycle bench testing has been performed on

the different ore types and locations.

4.12 Pre-leach dewatering

The pre-leach dewatering (thickening) step is designed to maintain the solution balance in the leach

SX/EW circuit. The leach circuit is designed for a 50 % underflow.

4.13 Copper leach-oxide/mixed ore only

Testwork results indicate relatively slow copper leach kinetics for the flotation tail streams. The

leach circuit design is based on a 24 hour residence time at 50 % solids. Gangue acid consumption

will be based on the mine schedule, with a total estimated acid consumption of 17 kg/t, allowing for

losses and bleed streams.

4.14 Leach residue counter current decantation (CCD)

Literature data has been used to design a CCD circuit comprising four stages of 35 m diameter stan-

dard thickeners. Testing will be required to evaluate the CCD circuit design.

4.15 Solvent extraction

The mass balance indicates a maximum PLS flowrate of approximately 617 m3/h. The mixer-

settlers have then been sized on the basis of residence time of two minutes in each of the two mixer

stages, and a settler specific area of 4.5 m3/h/m

2. The configuration of the solvent extraction circuit

is comprised of two extract mixer/settler stages, a wash stage, and a single strip stage.



Maximum flow will occur during mining of the predominantly oxide leach cap. Flow rates to the

Pregnant Leach Solution pond will consistently drop during the first three years of the project to

approximately 143 m3/hr as sulfides become the predominant species mined. The solvent extraction

system is designed for maximum expected flows and copper grades.

4.16 Electrowinning

The tank house will use stainless steel cathode technology, with anodes consisting of a lead-

calcium-tin alloy. Electrode sizes will be standard for copper recovery, with a submerged area of

1.1 m2, and the cathodes will be pulled on a seven-day stripping cycle. An average current density of

258 (maximum 275) A/m2 will be used and which should be readily achievable in the operating

environment, resulting in a requirement for 149 cells of 45 cathodes each. This cell size results in a

design current of 24,000 Amps divided between two rectifiers. The cells will be arranged in two

blocks of 75 cells, each with its own rectifier. Cathode stripping will typically be manual with au-

tomatic presentation of the cathodes to the stripping station, and will operate on at least two shifts

per day. Total design cathode copper production is 30,000 tonnes per year.

4.17 Concentrate handling

The concentrate handling system will consist of a Larox pressure filter and storage system. The La-

rox filter is designed for 37 m2 of filtration capacity. Dewatered material from the Larox filter will

either be repulped and fed to the Nitrogen Species Catalyzed (NSC) concentrate leach system, or

stored and shipped to a smelter

The NSC oxidative pressure leaching system consists of a 300 HP Tower Mill to grind to a P80 of

10 µm. From there, the material will be diluted to 50 g/L with recycled acidic process solution and

pressure leached in one of two unlined Grade 316 Stainless Steel pressure leach vessels. Pressure

and temperature will be maintained accordingly for full leaching of the copper and molybdenum,

while maintaining some of the sulfur as elemental sulfur.

Upon discharge from the leach vessels, the pulp will be flashed and the leach residue thickened and

filtered. The pregnant leach solution (PLS) will pass through secondary filtration and on to molybde-

num recovery. The filtered pressure leached residue, or POX residue, will report to sulfur recovery.

4.18 Molybdenum recovery

The PLS grade will be maintained at approximately 15 gram/liter copper and 0.28 g/L molybdenum.

The PLS will be passed through a series of five 2-ton carbon columns for molybdenum recovery.

Based on industrial practices, approximately 99 % of the molybdenum is expected to be recovered

in the carbon columns. After molybdenum recovery, the PLS will pass through an industrial EcoTec



Acid Purification Unit to recover free acid from the PLS. The recovered sulfuric acid solution prod-

uct will be recycled back to the NSC pressure oxidation system. The resultant PLS free acidity will

be dropped from about 30 g/L to approximately 5 g/L. This molybdenum-free PLS solution is then

sent to the PLS pond where the copper is recovered through a conventional SX/EW circuit.

Upon loading the carbon with molybdenum, the carbon is transferred to an atmospheric strip vessel.

The pH is raised with ammonium hydroxide to strip the molybdenum from the carbon, and a con-

centrated solution of ammonium molybdate is then formed. The ammonium molybdate solution is

treated by pH adjustment to 2.0 using sulfuric acid, and then solid ammonium molybdate precipi-

tates. The ammonium molybdate slurry is then filtered in a 3.7 m2 filter, dewatered, and fed to a

dryer heated to 550 °C. The ammonium molybdate is calcined to form molybdenum trioxide. The

ammonia driven off in the calcining process is captured in a water scrubber and reused in the

process. The molybdenum trioxide product is dried, bagged and sold.

4.19 NSC POX leached residue sulfur recovery system

Due to the high cost of sulfur based products, another POX system will be implemented to recover

sulfur from the NSC leached residue. The NSC POX residue is settled in a 35-meter thickener and

the underflow is filtered in a Larox filter. The POX material is stored until ready for sulfur recovery.

During sulfur recovery, the POX residue is fed into a leach vessel and ultimately pressure leached

with sodium hydroxide and oxygen. The sulfur is converted to sodium sulfate and a sulfur free POX

residue is formed. The POX residue is thickened and approximately 50 % of this residue is recycled

back to NSC, with the remaining 50 % of the material going to final tails.

The sodium sulfate solution is filtered and polished, and then fed to a relic molybdenum recovery

circuit similar to that discussed above. The clarified and purified sodium sulfate solution is then fed

into an Electrodialysis Unit (EDU) and evaporator system. The EDU will convert approximately

50 t/d of sodium sulfate into about 25 t/d of caustic. The caustic grade will be 100 g/L sodium hy-

droxide. Also, 35 t/d of sulfuric acid will be produced at a concentration of 70 % sulfuric acid.

These by-product solutions will be recycled and utilized in the process where required for leaching,

sulfur recovery, and neutralization. Excess sodium sulfate solution from this process will also be

treated in a conventional crystallizer system and solid sodium sulfate will be produced, bagged, and

sold based on market prices and demand.

4.20 Operating and capital cost estimate

Annual production is based on mining activities operating at assumed 90 % efficiency. The pre-

production stripping numbers are also assumed and will be defined when the final geologic and

mine reserve models have been completed. Haul road distances are also assumed and will be deter-



mined once a location for the processing plant has been selected and waste rock storage areas

designated. The Development and Operational key process indicators are:

Pre-production Stripping (North Pit) 300,000 t

Pre-production Stripping (South Pit) 200,000 t

Haul Road Construction (North Pit) 2500 m

Haul Road Construction (South Pit) 2500 m

Initial Production

Ore Production 15,000 t/d

Waste Production 37,000 t/d

Concentration Production 200 t/d

Annual Production – Ore 5.25 million t

Annual Production – Waste 13.0 million t

The milling plan estimates capital and operating costs assume a nominal 15,000 tonnes per day min-

ing operation. The milling plan includes:

• All labor, material, supply and equipment operating costs for the mill and associated concentrate

leach plant

• Supervision, administration and on-site management

• Benefits and employment taxes

• All on-site development for start-up and production

• Mill equipment and facilities purchase and installation or construction.

• Engineering and construction management fees



Pre-production development, installation and construction of all equipment and facilities necessary

to operate the mill at a nominal 15,000 t/d are included. Costs associated with the following facili-

ties and operations are included. However final locations and design details are pending:

• Crushing and conveyance of ore to the grinding circuit

• Grinding and flotation

• Flotation Tails leaching

• Copper and molybdenum concentrate pressure leaching

• Molybdenum trioxide recovery

• Copper solvent extraction and copper cathode electrowinning

• Eco Tec APU sulfuric acid recovery

• Sodium sulfate production

• Electrodialysis salt splitting of sodium sulfate for caustic and sulfuric acid production

• Tailings facility

• Basic access roads, power lines, and pipelines

• Construction, installation and operation of facilities and equipment necessary for equipment

maintenance and repair, electrical system, fuel distribution, water storage and drainage, sanitation

facilities, offices, labs, storage, and equipment parts and supply storage.

The mill and metallurgical plan does not include:

• Permitting and environmental assessment costs

• Home office overhead

• Taxes (except sales taxes and employment taxes)

• Insurance

• Depreciation

• Off-site transportation of products

• Incentive bonus premiums

• Overtime labor costs

• Sales expenses

• Interest expenses

• Start-up costs (except for working capital)

• Depletion rates

• Environmental costs, including reclamation

• Reclamation on-going assuming a two-year start-up period



Tables 9 and 10 summarize the operating and capital cost estimates done to an accuracy of +/- 30 %.

Table 9: Operating cost estimate USD +/- 30 %.

Cost summary

Category USD $/t

Mine 1.41

Plant – average 12.94

Plant – Oxide 10.13

Plant – Sulfide 13.91

Administration 0.54

Total operating cost 14.76

Table 10: Capital cost estimate, USD +/- 30 %.

Cost summary

Category USD $

Mine (less EPCM) 42,638,050

Mill & hydromet plant 230,770,238

Indirects 45,291,677

Contingency 39,541,957

Working capital 19,727,557

Reclamation bond 7,277,278

Total capital 381,920,231

5 Summary

This paper has demonstrated the successful application of industrial NSC pressure oxidation to

combined concentrates containing both copper and molybdenum as chalcopyrite and molybdenite

respectively. Confirmatory optimized NSC pressure leaching was successfully undertaken to leach

copper, molybdenum and rhenium. Further, as practiced industrially, selective separation and con-

centration of NSC leached rhenium and molybdenum using ion exchange and activated carbon was

successfully tested and confirmed. Finally, using this confirmatory test data along with industrial

operating data and experience, design criteria flow sheets and capital and operating costs estimates

were produced.



Acknowledgements

Thank you to Dr. Paul J. Miranda for all the analysis and mineralogical assessment for the NSC

confirmatory testing of the bulk molybdenite and chalcopyrite concentrates. In addition, as a cour-

tesy to the reader, both pertinent references and a bibliography of related work are included in the

paper.

References and bibliography

[1] Abel, E., and Schmid, H., 1928, Z. Physc. Chem. (Leipzig), 132: 56-64.

[2] Abel, E., Schmid, H. and Pollack, F. 1936, Monatsh, Chem., 69: 125-143.

[3] Ackerman, J.B., Anderson C.G., S.M. Nordwick and L.E. Krys. 1993. Hydrometallurgy at the

Sunshine Metallurgical Complex. Proceedings of the Fourth International Symposium on

Hydrometallurgy. Littleton, Colorado, SME.

[4] Anderson, C.G. and K.D. Harrison. 1990. Optimization of Nitric-Sulfuric Acid Pressure

Leaching of Silver from Refractory Sulfide Concentrate. Precious Metals 1990. International

Precious Metals Institute, Allentown, Pennsylvania, 119-151.

[5] Anderson, C. G. and Harrison, K. D., "Improvements in Plant Scale Nitric-Sulfuric Acid Pres

sure Leaching of Refractory Sulfide Concentrates", Preprint No. 91-183, AIME-SME Annual

Meeting, Denver, Colorado, 1991.

[6] Anderson, C.G., L.E. Krys, and K.D. Harrison. 1992. Treatment of Metal Bearing MineralMa-

terial. US Patent 5,096,486.

[7] Anderson, C.G., K.D. Harrison, and L.E. Krys, edited by R.K. Mishra. 1993a. Process Integra-

tion of Sodium Nitrite Oxidation and Fine Grinding in Refractory Precious Metal Concentrate

Pressure Leaching. Precious Metals 1993. International Precious Metals Institute, Allentown,

Pennsylvania, 19-45.

[8] Anderson, C.G., and L.E. Krys. 1993b. Leaching of Antimony from a Refractory Precious

Metals Concentrate. Proceedings Of The Fourth International Symposium On

Hydrometallurgy. Salt Lake City, Utah, August 1-5.

[9] Anderson, C.G. and S.M. Nordwick. 1994. The Application of Sunshine Nitrous-Sulfuric

Acid Pressure Leaching to Sulfide Materials Containing Platinum Group Metals. Precious

Metals 1994. Proceedings of the 18th Annual IPMI Conference, Vancouver, B.C., 223-234,

June.



[10] Anderson, C.G., K.D. Harrison, and L.E. Krys. 1996. Theoretical Considerations of Sodium

Nitrite Oxidation and Fine Grinding in Refractory Precious Metals Concentrate Pressure

Leaching. Minerals and Metallurgical Processing, AIME-SME, Volume 13, Number 1,

February.

[11] Anderson, C. G., Harrison, K. J., and Krys, L. E., "The Application of Sodium Nitrite Oxida-

tion and Fine Grinding In Refractory Precious Metals Concentrate Pressure Leaching", Mine-

rals and Metallurgical Processing, AIME-SME, Volume 13, Number 1, February, 1996.

[12] Anderson, C.G. 2000a. Nitrogen Species Catalyzed Pressure Leaching of Copper Ores and

Concentrates. ALTA Copper 2000, Adelaide, South Australia, October.

[13] Anderson, C.G. 2000b. The Design, Optimization and Operation of an Industrial Copper Sol-

vent Extraction and Electrowinning Circuit at a Commercial Nitrogen Species Catalyzed

Pressure Leaching Plant. ALTA SX/IX 2000, Adelaide, South Australia, October.

[14] Anderson, C.G. 2001a. The Industrial Non-Cyanide Hydrometallurgical Recovery of Silver

and Gold Utilizing Nitrogen Species Catalyzed Pressure Oxidation. Cyanide Social, Industrial

and Economic Aspects. Symposium Proceedings, TMS Annual Meeting, New Orleans, Loui-

siana, February.

[15] Anderson, C.G. 2001b. Industrial Nitrogen Species Catalyzed Pressure Leaching and Alkaline

Sulfide Gold Recovery from Refractory Gold Concentrates. Precious Metals 2001, 25th An-

nual IPMI Meeting, Tucson, Arizona. June.

[16] Anderson, C.G. April 2002a. The Chemical Analysis of Industrial Alkaline Sulfide

Hydrometallurgical Processes. The Society of Mineral Analysts and the Canadian Mineral

Analysts Annual Meeting, Spokane, Washington.

[17] Anderson, C.G. May 2002b. The Mineral Processing and Industrial Nitrogen Species

Catalyzed Pressure Leaching of Formation Capital's Cobaltite and Chalcopyrite Concentrates.

ALTA Ni/Co and Cu International Conference, Perth, W.A. Australia.

[18] Anderson, C.G. June 2002c. The Application of Industrial NSC Pressure Leaching in the Re-

cycle and Recovery of Secondary Precious and Base Metals. TMS Fall 2002, Recycling and

Waste Treatment Meeting. Lulea, Sweden.

[19] Anderson, C.G. 2003a. Alkaline Sulfide Recovery of Gold Utilizing Nitrogen Species

Catalyzed Pressure Leaching. Hydrometallurgy 2003. Vancouver, B.C., October.

[20] Anderson, C.G., and E. Rosenberg. 2003b. Single Step Separation and Recovery of Palladium

Using Nitrogen Species Catalyzed Pressure Leaching and Silica Polyamine Composites.

Hydrometallurgy 2003. Vancouver, B.C., October.

[21] Anderson, C.G. 2003c. The Application and Economics of Industrial NSC Pressure Leaching

to Copper Ores and Concentrates. COBRE 2003, Santiago, Chile, December.

[22] Anderson, C.G., E. Dahlgren and D. Stacey. 2004. Unpublished research.



[23] Anderson, C.G. and L.G. Twidwell. 2006. The Control of Iron and Arsenic in the Treatment

of Gold Bearing Enargite Concentrates, Iron Control In Hydrometallurgy, Proceedings of the

36th Annual Hydrometallurgy Meeting of CIM, October 2006, Montreal, Quebec

[24] Anderson, C.G. and L.G. Twidwell. 2007. Hydrometallurgical Processing of Gold-Bearing

Copper Enargite Concentrates, 6th COPPER/COBRE Conference, John Dutrizac

Hydrometallurgy Symposium, 37th Annual Hydrometallurgy Meeting of CIM, Toronto, Onta-

rio, August.

[25] Anderson, C.G. and L.G. Twidwell. 2008. Hydrometallurgical Processing of Gold-Bearing

Copper Enargite Concentrates, Canadian Metallurgical Quarterly, John Dutrizac

Hydrometallurgy Symposium Special Issue.

[26] Anderson, C.G. 2008. NSC Pressure Leaching: Industrial and Potential Applications,

Hydrometallurgy 2008 Proceedings of the Sixth International Symposium, SME, 858-885

[27] Awad, H.H., and Stanbury, D.M., 1993, Autooxidation of NO in aqueous solution. Int. J.

Chem. Kinet. 25: 375-381.

[28] Baldwin, S.A. and G.V. Van Weert. 1996. On the Catalysis of Ferrous Sulfate Oxidation in

Autoclaves by Nitrates and Nitrites. Hydrometallurgy, Elsevier Science, B.V., Vol. 42.

[29] Balachandran P., Kamath, A.K. Mitra, S. Radhakrishnan and Kiram Shetty, 2003. Electrolyte

Impurity Control at Chinpada Refinery of Sterlite Industries (India) Limited, Eco-Tec SM

Technical Paper 167.

[30] Bardt. 1919-1923. Recovering Metals Contained in Metalliferous Ore, Waste Residues, and

Alloys. German Patent 353,795, Canadian Patent 233,566, Chemical Abstracts 173,154.

[31] Barth, T.R., A.T.C. Hair, and T.P. Meier, edited by J.E. Dutrizac et al. 1998. The Operation of

the HBM&S Zinc Pressure Leach Plant. Zinc and Lead Processing, The Metallurgical Society

of CIM.

[32] Beake, B.D., Constantine, J. and Modie, R. B., 1992, The kinetics and mechanism of the

reaction of nitrous acid with 4-substituted phenols in aqeous solution. J. Cem. Soc. Perkin

Trans. 2: 1653-1654.

[33] Beattie, M.J.V., R. Randsepp, and A. Ismay, edited by G.S. Dobby and S.R. Rao. 1989.

Arseno/Redox Process for Refractory Gold Ores. Proc. Intern. Symp. Processing Complex

Ores. Pergamon, Oxford, 431-439.

[34] Bjorling, C. and G.A. Koltz. 1964. Oxidizing Leaching of Sulfide Concentrates and Other

Materials Catalyzed by Nitric Acid. Proc. 7th In. Mineral Processing Cong., New York, NY.

[35] Bjorling, G. and G.A. Kolta. 1966a. Oxidizing Leach of Sulfide Concentrates and Other Mate-

rials Catalyzed by Nitric Acid. Proceedings of 7th International Mineral Processing Congress.

Gordon & Breach, New York, NY, 127-138.



[36] Bjorling, G. and G.A. Kolta. 1966b. Wet Oxidation of Iron Sulfide Concentrates Catalyzed by

Nitric Acid. J. Chemistry, U.A.R., 9(2):187-203.

[37] Bjorling, G. and P. Lesidrenski. 1968. Hydrometallurgical Production of Copper from

Activated Chalcopyrite. AIME Annual Meeting, New York, NY.

[38] Bjorling, G. and G.A. Kolta. 1969. Wet Oxidation as a Method of Utilization of Chalcopyrite,

Sphalerite, and Molybdenite. Chemistry U.A.R., 12(3):423-435.

[39] Bjorling, G., edited by D.J.I Evans and R.S. Shoemaker. 1973. Leaching of Mineral Sulfides

by Selective Oxidation at Normal Pressure. International Symposium on Hydrometallurgy.

AIME, New York, NY, 701-707.

[40] Bjorling, G. and W. Mulak. June 1976a. Kinetics of NiS Leaching in Nitric Acid Solutions.

Trans. Inst. Min. & Met. C98-C101.

[41] Bjorling, G. et al., edited by J.C. Yannopoulos and J.C. Agarwal. 1976b. A Nitric Acid Route

in Combination with Solvent Extraction for Hydrometallurgical Treatment of Chalcopyrite.

Extractive Metallurgy of Copper. Vol. 2. AIME, New York, NY, 725-737.

[42] Bosio, S., Ravella, A., Saracco, G.B., and Genon G., 1985. NOX Absorption by Ferrous Sul-

fate Solutions, Ind. Chem. Process Res. Dev., 24:149-152

[43] Brennecke, H.M. 1975. Recovery of Metal Values from Ore Concentrates. US Patent

3,888,748.

[44] Brennecke, H.M. et al. August 1981. Nitric-Sulfuric Leach Process for Recovery of Copper

from Concentrates. Min. Engineering. 1259-1266.

[45] Bunton, C.A., Stedman, G., 1959, Mechanism of the Azide-Nitrite Reaction. Part III Reaction

in [18O] water. J. Chem. Soc. 3466-3474.

[46] Butler, A.R., and Ridd, J.H., 2004, Formation of nitric oxide from nitrous acid in ischemic

tissue ands skin, Nitric Oxide, 10: 20-24.

[47] Caldon, F. 1978. Treatment of Metal Bearing Mineral Material. US Patent 4,084,961.

[48] Chmielewski, T. and Charewicz, W.A., 1984, The oxidation of ferrous iron in aqueous

sulfuric acid under oxygen pressure, Hydrometallurgy, 12:21.

[49] Daugherty, E.W., A.F., Erhard and J.M. Drobnick. 1973. Process of Recovery of Rhenium

and Molybdenum Values from Molybdenite Concentrate. U.S. Patent 3,739,057.

[50] Dresher, W.H., Wadsworth, M. E., Fassel, W. M., 1956, A Kinetic Study of the Leaching of

Molybdenite, Mining Engineering, Transactions of AIME, Littleton, Colorado, June 738-744.

[51] Edwards, C.R., edited by A.D. Zunkel. 1985. Engineering the Equity Concentrate Leach Pro-

cess. Complex Sulfides: Processing of Ores, Concentrates and By-Products. Proceedings of a

symposium sponsored by the Metallurgical Society of AIME and the CIMM, TMS-AIME Fall

Extractive Meeting, San Diego, CA, 197-219, November 10-13.



[52] Epstein, I.R.,Kustin, K. And Simoyi, R.H., 1982, Nitrous acid decomposition catalyzed by and

iron (II) complex: Tris (3,4,7,8-tetramethyl-1, 10-phenanthroline) iron)II). American Chemical

Society. 104: 712-717.

[53] Eugene, S.A, edited by A. Sutulov. 1979. Cymoly Process, in International Molybdenum

Encyclopedia, Vol. 2, Processing and Metallurgy, Antermet Publications, Santiago, Chile,

105.

[54] Fair, K.J., J.C. Schneider, and G. Van Weert, edited by R.D. Salter. 1987. Options in the

Nitrox Process. M. Proc. Intern. Symp. Gold Metallurgy. Pergamon, Toronto, 279-291.

[55] Fair, K.J. and F.J. Basa, edited by G.S. Dobby and S.R. Rao. 1989a. Treatment of Agnico

Eagle's Silver-Bearing Flotation Concentrate by the Nitrox Process. Proc. Intern. Symp. Pro-

cessing Complex Ores. Pergamon, Oxford, 411-420.

[56] Fair, K.J. and G. Van Weert, edited by B. Harris. 1989b. Optimizing the NITROX PROCESS

through Elemental Sulfur Formation. Precious Metals 1989. The International Precious Metals

Institute, Montreal, 305-317.

[57] Fayram, T. and C.G. Anderson. October 2003. The Development and Implementation of In-

dustrial Hydrometallurgical Gallium and Germanium Recovery. Hydrometallurgy 2003, Van-

couver, B.C.

[58] Fedulov, O.V. et al. 1966. Oxidation of Molybdenite by Nitric Acid Solutions. Sb. Statei

Aspir. Soiskatelei, Min. Vyssh. Sredn. Spets. Obrazov Kaz. SSR, Met. Obogashch. (2) 86-94,

Chemical Abstracts 69,3260 In.

[59] Fedulov, O.V. et al. 1967a. End Products of Nitric Acid Decomposition of Molybdenite. Sb.

Statei Aspir. Soiskatelei Min. Vyssh. Sredn. Spets. Obrazov Kaz. SSR, Met. Obogashch. (3)

170-177, Chemical Abstracts 70,89848ym.

[60] Fedulov, O.V. and V.D. Ponomarev. 1967b. Nitric Acid Decomposition Kinetics of

Molybdenite. Sb. Stratei Aspir. Soiskatelei. Min. Vyssh. Sredn. Spets. Obrazov, Kaz. SSR.

Met. Obogashch. (3), 129-139, Chemical Abstracts 70,9llSOg.

[61] Fedulov, O.V. and V.D. Ponomarev. 1967c. Oxidation Process Mechanism of Molybdenite by

Aqueous Nitric Acid Solutions. Sb. Statei Aspir. Soiskatelei Min. Vyssh. Sredn. Spets.

Obrazov. Kaz. SSR, Met. Obogashch. (3), 12208, Chemical Abstracts 70,91182v.

[62] Fossi, P., L. Gandon, C. Bozec, and J.M. Demarthe. July 1977. Refining of High-Nickel

Concentrates. CIM Bull. 188-196.

[63] Freeport McMoRan Copper and Gold, 2010, Bagdad Mine Molybdenum Concentrate Pressure

Oxidation, Personal Communication.

[64] Gok, O.S., 2009. On The Role of Low-Concentration Nitrite In Oxidative Leaching with

Oxygen, PhD Thesis, Colorado School of Mines, USA.



[65] Habashi, F. 1973a. The Action of Nitric Acid on Chalcopyrite. Trans. Soc. Min. Eng. AIME

254, 224-228.

[66] Habashi, F. 1973b. Treatment of a Low-Grade Nickel Copper Sulfide Concentrate by Nitric

Acid. Trans. Soc. Min. Eng. AIME 254, 228-230.

[67] Habashi, F., edited by B. Mishra. 1999. Nitric Acid in the Hydrometallurgy of Sulfides. EPD

Congress 1999. TMS-AIME, Warrendale, PA, 357-364.

[68] Haver, F.P., and Wong, M.M., 1972, Making Copper Without Pollution, Mining Engineering,

Littleton, Colorado, June, 52-53.

[69] Heckner, H.N., 1973, Potentiostatic switching experiments for the cathodic reduction of

nitrous acid in perchloric acid with te addition of nitric acid. Electroanalytcal Chemistry and

Intefacial Electrochemistry. 44: 9-20.

[70] Himmelblau, D.M., 1960, Solubilities of inert gases in water 0O C to near the critical point of

water, Journal of Chemical and Engineering Data, 5: 10-15.

[71] Huffman, R.E. and Davidson, N.J., 1956, Kinetics of the ferrous iron-oxygen reaction in

sulfuric acid solution. Amer. Chem. Soc. 78: 4836.

[72] Huang, H.H. 2002. StabCal Modeling Software, September.

[73] Jeffrey, M. and C.G. Anderson. 2002. A Fundamental Study of the Alkaline Sulfide Leaching

of Gold. The European Journal of Mineral Processing and Environmental Protection, October.

[74] John, C.I.A., R.C. Sathe, and V.S. Kasongamulilo. 1991. Improving Flotation Performance at

the Nchanga Concentrator of Zambia Consolidated Copper Mines Limited. Copper 91/Cobre

91, Volume 3, Mineral Processing and Process Control. Pergamon Press, New York, NY, 19-

33.

[75] Joseph, T.B. 1916. Metal Leaching Process. Canadian Patent 173,452. Chemical Abstracts

12,130.

[76] Kerfoot, D.G.E. and R.W. Stanley. 1976. Hydrometallurgical Production of Technical Grade

Molybdic Oxide from Molybdenite. U.S. Patent 3,988,418.

[77] Kingsley, G.E. 1915. Process for Treating Complex Sulfide Ores. US Patent 1,144,480.

[78] Krysa, B., B. Barlin, and D. Wittleton. May 1988. The Application of Zinc Pressure Leaching

at the Hudson Bay Mining and Smelting Co. Limited. Projects ’88, Paper #8. 18th

Hydrometallurgical Meeting CIM.

[79] Kunda, W. 1984a. Hydrometallurgical Processing of Silver Concentrate. Precious Metals:

Mining, Extraction, and Processing. Proc. Joint Symp. AIME-IMS/IPMI, 397-423.

[80] Kunda, W., edited by V. Kudryk et al. 1984b. Hydrometallurgical Processing of Silver

Concentrate. Precious Metals: Mining, Extraction, and Processing, Proc. Joint Symp. AIME-

TMS/IPMI, 397-423.



[81] Lane, J.W., Bender, F.N., and Ronzio, R.A., 1972, Recovery of Molybdenum from Oxidized

Ore at Climax, Colorado, AIME Transactions Volume 252, 77-82..

[82] Le Nickel. 1968. Recovery of Nickel and Other Secondary Metals from Nickel Mattes. French

Patent 1,597,569.

[83] Lunt, R.R., D.K. Modrow and G.K. Roset. October 2003. Adaption of Dilute Mold Lime Dual

Alkali Scrubbing at Stillwater Mining Company’s PGM Smelter. Hydrometallurgy 2003,

Vancouver, B.C.

[84] Mahmoud, M.H., C.G. Anderson, and C.A. Young. June 2002. Sulfuric Acid-Chloride

Leaching of Platinum, Palladium and Rhodium From Catalyst Residue. Recycling and Waste

Treatment in Mineral and Metal Processing. TMS Fall 2002 Recycling and Waste Treatment

Meeting, Lulea, Sweden.

[85] Markovits, G.Y., Schwartz, S.E., and Newman, L., 1981, Hydrolysis equilibrium of dinitrogen

trioxide in dilute acid solution. Inorganic Chemistry. 20: 445-450.

[86] Molybdenum Corporation of America (Molycorp). 1973. Solvent Extraction Process for

Recovery of Molybdenum and Rhenium from Molybdenite. U.S. Patent 3,751,555.

[87] Mulak, W. 1987. Silver Ion Catalysis in Nitric Acid Dissolution of Ni3S2. Hydrometallurgy

18, 195-205.

[88] Ouellet, R., A.E. Torma, and J.P. Bolduc. 1975. Extraction du Nickel d'un Concentre de

Pentlandite par HNO3 - H2SO4. Can. Met. Quart. 14(4):339-343.

[89] Olsen, D., 2008, Kennecott to invest $ 270 million in molybdenum processing facility, The

Enterprise.

[90] Pauling, E.. 1940, Nassmetallurgische Aufarbeitung schwer aufzubereitender Komplexerze.

Metall und Erz 35, 451-455.

[91] Park, J.Y. and Lee, Y.N. 11988. Solubility and Decomposition Kinetics of Nitrous Acid in

Aqueous Solution. J. Phys. Chem. 92: 6294-6302.

[92] Peters, E. 1976. Direct Leaching of Sulfides: Chemistry and Applications. Metall. Trans., 7B,

505.

[93] Peters, E. 1992. Hydrometallurgical Process Innovation. Hydrometallurgy 29, 431-459.

[94] Prasad, M.S. 1991. Concentration Capacity in Treating Copper-Cobalt and Copper-Zinc Ores

at Gecamines. Zaire. Mining Engineering, 43(1): 129-133.

[95] Prater, J.C., P.B. Queneau, and T.J. Hudson. 1973. Nitric Acid Route to Processing Copper

Concentrates. Trans. Soc. Min. Eng. AIME 254(2):117-122.

[96] Queneau, P.B. and Prater, J.D., 1974, Nitric acid process for recovering metal values from

sulfide ore materials containing iron sulfides. US Patent 3793429, Chemical Abstracts.



[97] Rankin, H.D. 1915. Method of Treating Metalliferous Materials and Recovering Solvents

Used. US Patent 1,150,787.

[98] Shukla, P.P., T.K. Mukherjee, and C.K. Gupta. 1978. A Nitric Acid Route for Processing a

Nickel-Copper Sulfide Concentrate. Hydrometallurgy 3. 55-56.

[99] Stumpf, A. and Berube, Y. 1973, Aqueous Oxidation of Molybdenite in Chalcopyrite

Concentrates, Transactions of AIME, Vol 254, December, 305-309.

[100] Topol, L.E., Osteryoung, R.A. and Christie, J.H. 1965. Electrochemical Studies of NO+ and

NO2+ in Concentrated H2SO4 (1965). 112: 861-864.

[101] Turney, T.A. and Wright, G.A. 1959. Nitrous Acid and Nitrosation. Chem. Rev. 59: 497-513.

[102] Van Weert, G., K.J. Fair, and J.C. Schneider. 1986a. Prochem's Nitrox Process. CIM Bull. 79,

84-85, November.

[103] Van Weert, G., K.J. Fair, and J.C. Schneider. 1986b. The Nitrox Process for Treating Gold-

Bearing Arsenopyrites. Annual Meeting TSM-AIME, Denver, Colorado.

[104] Van Weert, G. 1988a. An Update on the Nitrox Process. Randol Gold Forum, 209-210.

[105] Van Weert, G., K..J. Fair, and V.H. Aprahamian. 1988b. Design and Operating Results of the

Nitrox Process. Gold Mining 88, Society of Mining Engineers of AIME, Littleton, Colorado,

286-302.

[106] Van Weert, G. and K.J. Fair. 1989. Capital and Operating Costs of the Nitrox Process for

Auriferous Arsenopyrites. Extraction Metallurgy '89. Institution of Mining & Metallurgy,

London.

[107] Vizsolyi and E. Peters. 1980. Nitric Acid Leaching of Molybdenite Concentrates.

Hydrometallurgy 6, 103-119.

[108] Weber, F.W. 1925. Silver from Sulfides. US Patent 1,555,615. Chemical Abstracts 193,473.

[109] Westby, G.C. 1918a. Nitric Acid and Copper Ores. Metallurgical & Chemical Engineering.

18(6):290-296.

[110] Westby, G.C. 1918b. Treating Sulfide Ores of Copper, Zinc, or Other Metals. US Patent

1,244,811, Chemical Abstracts 12,131.

[111] Yurkevich, Y.N. and K.Y. Shapiro. 1967. Decomposition of Molybdenite by Nitric Acid. Met.

Vol'frama, Molibdena, Niobiya, 53-56. Chemical Abstracts 69,98546e.

[112] Zelikman, A.N., L.V. Belyaevskaya, and T.E. Prosenkova, 1969, Decomposition of

Molybdenite by Nitric Acid, lzvestiya Vysshikh Uchebnykh, Zavedenii, Tsvetnaya

Metallurgiya 12 (6) 43-48, Chemical Abstracts 72,81734d.


HPGRs in Copper Ore Comminution –

A Technology Broke Barriers

E. Burchardt, N. Patzelt, J. Knecht, R. Klymowsky

Polysius AG

Graf-Galen-Straße 17

Beckum, Germany

Keywords: HPGR, grinding, minerals references, minerals applications

Abstract

The growth of High Pressure Grinding Rolls has been steady and relentless since their introduction

into the diamond industry in 1986. The fastest growth was in iron ores, but the breakthrough came

with the first installations of HPGRs in the copper industry at the Freeport’s “Cerro Verde Mine” in

Peru (2006) and “Grasberg Mine” in Indonesia (2007). The recent fall in commodity prices in 2008

has temporarily interrupted this growth, but has not dampened the interest in the technology.

Many of the newest copper and gold deposits are large, low-grade and hard. Large tonnages and

especially hard rock favour HPGRs. Typically HPGRs offer the lowest operating costs. However,

HPGRs are also able to reduce investment cost for very hard ores.

HPGR technology broke barriers in a two fold way. Starting as a newcomer some years ago, it has

established itself as a matured and accepted technology in the precious and base metals industry. In

addition, HPGRs have proven their capability of being the key to make some projects economical

viable at all.

This paper reviews some further projects that have come on-stream since Cerro Verde and Gras-

berg:

Boddington Gold Mines, July 2009, 35 million t/a; Anglo Platinum’s Mogalakwena, March 2008,

600 ktpm; Assmang’s BKM Iron Ore Project, July 2008; Northam Platinum’s UG2 plant, May

2008.

The first commercial HPGR for heap leaching in gold started in Goldfields’ Tarkwa Mine in Ghana,

October 2009.

The paper includes some of the lessons learnt from operating the rolls at various plants and reviews

on-going development programmes that have been initiated to improve wear life of the tyres, reduce

energy consumption and overall operating costs.

Burchardt, Patzelt, Knecht, Klymowsky


In addition, the paper provides an outlook on how HPGR technology may further advance the

“Concentrator of the Future”.

There are signs of recovery in several commodities. This gives confidence that when the recovery

takes hold, HPGRs will be in a strong position to secure a leading role in mineral comminution.

1 Introduction

In the middle of the 80th

, HPGR (High Pressure Grinding Roll) technology was firstly adopted by

the cement industry due to the high potential for energy savings. Energy requirements for grinding

cement clinker, slag and limestone were in the order of 50 % lower than those consumed in conven-

tional ball mills.

Nevertheless, the minerals industry outside of diamonds and iron ore was initially very reluctant to

take over this technology due to concerns about wear and consequently availability.

It took a lot of effort to improve the design of HPGRs to provide acceptable lifetimes of the wear

components for high wear applications and to make these units maintenance friendly enough to

reach availabilities of more than 94 %. These were the preconditions for the final acceptance of

HPGRs in the base and precious metals industry. The wear issue has been successfully resolved,

although there will be a continuous development process in the next couple of years, to further

reduce wear cost. After the first successful installations in copper and other minerals applications,

today’s focus is more on issues like:

Under which conditions are HPGR circuits more economical than conventional circuits?

Are HPGRs able to make low grade and high cost deposits viable from an economical standpoint?

What are the most economical HPGR based flow sheet configurations and plant layouts?

How to match highest grinding circuit energy with lowest investment cost?

What are the optimum interfaces between secondary crushing, high pressure grinding and ball

milling in terms of transfer size and energy distribution?

What has to be done to reduce investment cost of HPGR based flow sheets and obtain a further wear

and energy cost reduction?

Which lessons can be learnt from major installations in terms of future sizing and design criteria,

optimum plant layout, as well as best operating and maintenance practices?

This paper addresses at least some of these issues, providing a basic overview on to the current

status of this technology.

HPGRs in Copper Ore Comminution


2 Grinding Circuits Incorporating HPGRs

The first serious approach of HPGR technology to the copper industry was the installation at Cyprus

Sierrita in the Middle of the 90th

. At that time, this installation was considered as a failure despite a

lot of positive achievements [1, 2]. The first successful installations of HPGRs in the copper

industry started at the Freeport’s “Cerro Verde Mine” in Peru 2006 [3] and “Grasberg Mine” in

Indonesia 2007 [4].

Major applications for HPGRs in copper and other minerals are

• for pebble crushing in SAG mill circuits of concentrators (greenfield, brownfield) (Figure 1),

• for tertiary crushing in concentrators (greenfield) (Figure 2),

• for quaternary crushing in concentrators (brownfield) (Figure 4),

• for tertiary and / or quaternary crushing in heap leach operations (greenfield, brownfield).

The potential of HPGR for pebble crushing has been studied for a long time. A large variety of pos-

sible flow sheets have been investigated. Figure 1 includes two of these concepts. The most efficient

way to implement HPGRs is to treat the pebbles separately in a secondary crushing (SC) – HPGR

circuit directly feeding the screen undersize product to the ball mills. HPGRs can be retrofitted effi-

ciently into existing SAG/AG mill circuits (brownfield) for capacity increases. First HPGRs for

pebble or SAG oversize crushing were applied in iron ore applications in the 90th

. In 2008, a gold

producer [5] decided to build a SAG mill based concentrator with a HPGR for pebble crushing

(Figure 1: “SAG – Option 2”).

Figure 1: HPGR re-grinding crushed pebbles (left); SC/HPGR circuit for pebble crushing (right)

HPGR based circuits for new concentrators (greenfield) would typically use HPGRs in closed

circuit for tertiary crushing, either with wet or dry screening. The HPGR circuit is normally fed with

a full feed in a normal closed circuit (Figure 2 – left hand) or with a truncated feed in a reversed

closed circuit (Figure 2 – right hand). There have been objections in the gold/copper industry



against the reversed closed circuit mainly due to wear issues although the truncation of the feed is

standard practice in diamonds. Plants in operation like Freeport – Cerro Verde (copper – since

2006), Anglo Platinum – Mogalakwena (platinum – since 2008) and Newmont - Boddington (gold –

since 2009) are based on a normal closed circuit with wet screening (Figure 2 – left hand). So far,

only Anglo Platinum has opted for a dry screening plant [6, 7].

Figure 2: HPGR in tertiary crushing duty – normal closed circuit (left); reversed closed circuit (right)

Figure 3: HPGRs at Cerro Verde (Freeport) – first concentrator in copper fully HPGR based [4]

Other operations to be commissioned in the future, like Vale – Salobo, MolyMines – Spinnefex

Ridge and Gindalbie Metals - Mt. Karara are based on a full feed, wet screening circuit (Figure 2 –

left hand).

A reversed HPGR circuit with wet screening was selected for the BKM Iron Ore Project of

Assmang which went into operation in July 2008. Major reason for this decision was the need to

minimise fines generation which are production losses in this application. Any concerns about



accelerated wear due the missing fines in the feed have not been confirmed. On-going feed back

from this application may be the basis for a re-thinking of the wide spread objections against feed

truncation in copper/gold or any other hard rock applications - especially in low wear applications.

HPGRs have proven themselves as powerful tools for plant up-grades in existing crushing and

grinding circuits (brownfield). First successful installation as a quaternary crushing stage has been

in the Grasberg Mine of Freeport in Indonesia (gold/copper since 2007) [Figure 4].

Figure 4: HPGR application at Grasberg (Freeport) – first one as a quaternary stage in copper

HPGRs at Grasberg [4] are receiving the product of a 3-stage crushing plant. Re-grinding of this

relative fine feed of minus 15 mm results in a finer ball mill feed. This provides the opportunity to

increase the plant throughput at same flotation fineness and recovery or to keep the throughput at

the same level and increase flotation fineness and consequently recovery. Both cases increase the

production value, often at lower overall operating cost.

Reduction of operating cost by replacing existing tertiary cone crushers or rod mills without any

throughput or recovery increase does not justify a HPGR retrofit typically.

A HPGR in similar application has been operated at Northam Platinum on a UG2 ore since 2008,

re-grinding a fine – bi-modal – feed from a crushing plant. The HPGR application allowed the

installation of an intermediate flash flotation circuit and the conversion of the original rod mill to a

ball mill. The complete re-design project has been reported as an overwhelming success [7].

In October 2009, the first commercial HPGR for a heap leach operation in gold (Goldfields –

Tarkwa Mine, Ghana) has been commissioned. This installation is considered as the first full scale

industrial test in order to prove the general suitability of HPGRs for heap leaching as well as the

achievement of faster and higher recoveries of HPGR treated ores, which have been found in the lab

again and again. Positive results from this operation could also influence the SX/EW heap leach

operations in the copper industry.



3 Industrial Operating Experience

Basic design data of recent copper and minerals installations are summarised in Figure 5. Operating

experience gathered from those installations have provided a large basis for design modifications

and improvements, wear life improvements, verification of scale-up and sizing criteria as well as for

the evolution of best operating and maintenance practices. New projects as well as projects coming

on stream in the near future will be able to benefit from the experience continuously collected.

Figure 5: Basic design & operating data from recent applications on copper and other mineral ores

Major findings from these installations are summarised in the following.

In general, sizing of the early minerals HPGRs in terms of throughput, required circulating loads

and energy input in closed circuit operation as well as the lifetime predictions have been quite con-

servative initially due to a lack of industrial references in similar applications.

Typically, lifetimes of industrial HPGRs rolls have been higher than the conservatively provided

figures for warranties. HPGR wear testing (“ATWAL abrasion index”) gave a realistic basis for

lifetime projections. Maximum achieved lifetime has been more than 20,000 hours in a copper in-

stallation on finer feed. There has been only one case where the predicted lifetime (6000 hours) was

significantly lower than the lifetime initially reached (3000 hours). This difference was mainly at-

tributable to changes of ore properties (higher ATWAL abrasion index) compared to the design

phase and in addition to instantaneously higher than design capacities. An optimisation program on

the wear protection tyres, improving stud design and grade, resulted in an increase of the lifetime to

recently more than 5000 hours with a potential to finally exceed 6000 hours with the next design

change. Other high wear operations will benefit from these developments.



The major driver for abrasion on HPGR rolls is the mineralogy of the ore to be treated. Typically the

mineralogy has a higher influence on the wear rate than the feed size, grinding pressure and mois-

ture content. The second dominant factor for the wear rate is the feed size or “feed size to gap ratio”

respectively. A feed size, significantly smaller than the gap, reduces the wear rate. The hardness or

competency of an ore has only a minor impact on the wear rate according to ATWAL abrasion test-

ing and subsequent industrial experience.

Top feed size control to HPGR circuits is essential for long roll life since this is a precondition to

minimise the risk of stud breakage. Competent ores have a higher inherent risk of breaking studs.

Excessive oversize resulted in severe stud breakage and finally resulted in surface damage in one

application. A reduction or limitation of the feed size to the HPGRS treating very competent ore

would increase tyre lifetime and is a precondition to apply stud grades with improved wear resis-

tance on those ores in the future

Roll unit change outs were conducted within less than 36 hours after establishing optimised working

procedures.

Initial problems with roll skewing in some applications were successfully resolved. In general, roll

skewing is caused by insufficient feed rates and/or segregation in the feed. In some applications, roll

skewing was mainly caused by insufficient feed rates during feed material start-ups as well as in

continuous operation resulting in feeding only to the roll centre and preventing the build up of

choke feed conditions. Non choke feed conditions could be overcome by adjusting the roll speed to

the actual available feed quantities in normal operation and a ramp up of the roll speed during feed

material start-ups.

The specific throughput rates of industrial HPGRs have been usually higher than those obtained in

batch type pilot test units, which were the basis for initial sizing. Consequently the achieved

throughput capacities of industrial units often exceeded the conservative design figures.

Circulating loads in closed circuit operation with screens were found to be on the levels as predicted

by the supplier. Initial concerns especially of engineers and consultants about supplier predicted

circulating loads seem to vanish and the tendency to oversize HPGR circuits may be brought down

to reasonable levels. A common understanding of the supplier, engineer and consultant about realis-

tic circulating loads in HPGR circuits as well as of the achievable throughput rates would provide

an opportunity to reduce investment cost.

Oversizing of HPGRs due to underestimating actually achievable capacities and overestimating re-

quired capacities created sometimes operational problems due to a lack of feed material - especially

in case of fixed speed drives. Variable speed HPGR drives should be installed for proper process

control and for the handling of any fluctuations in required throughput.

Variation of the grinding pressure turned out to be a powerful approach to either optimise a HPGR

circuit in terms of maximum throughput at lower circulating loads or maximum energy efficiency at

higher circulating loads.



Cake formation in the HPGR discharge material was found to be less critical than initially feared.

Current cake stability testing provides a good assessment of cake competency. Screen efficiency

was higher in wet than in dry screening. Moisture control of the wet screen oversize could be man-

aged even for screening at 5 to 6 mm. Efficient pulping of the screen feed ahead of the screens,

proper material distribution along the screen width as well as sufficient screening area are precondi-

tions to limit the moisture content in the screen oversize (S/O) to reasonable levels. Typically, the

moisture content of a plus 6 mm S/O is less than 4 % under proper operating conditions. These low

moisture levels may give room to a further reduction of the cut size between HPGR and ball mill

circuit.

Under good operating conditions, specific energy input for producing < 6 mm copper ore product

was lower than 3 kWh/t referred to circuit product.

HPGR performance was found to be quite insensitive against changes in feed material properties.

Even treating a variable ore body, specific throughput, circulating load, power absorption and spe-

cific energy input did not vary significantly. Ball mill feed size distribution was very consistent and

mainly determined by the mesh size of the wet screens rather than influenced by material properties

like competency. This feature allows a smooth ball mill operation at nearly constant throughput and

product fineness and subsequently a controlled operation of the flotation circuit. Throughput varia-

tions required due to operational requirements or caused by minor feed characteristic changes were

easily managed by roll speed adjustments.

SAG mill circuit performance is far more feed material dependent. Feed size distribution and ore

competency has a much larger impact on SAG mill throughput and the transfer size between the

SAG and ball mill circuit. In summary, a consistent circuit balance is much easier to achieve be-

tween a HPGR and a ball mill circuit

4 HPGR vs. SAG Mill Circuits

HPGR based circuits usually compete with traditional SAG mill circuits, although there seem to be

some recent signs of a revival of conventional 3-stage crushing circuits – especially on competent

ores at smaller scale operations.

Large scale concentrators with a capacity of 100,000 tpd would require 5 to 6 secondary crushers

and about 10 to 15 tertiary crushers depending on the ore competency and targeted product fineness.

Lifetime of secondary crusher liners can be less than 30 days in high wear applications with an even

shorter lifetime to be expected in tertiary duty. This maintenance intensity besides the high capital

and operating cost usually avoids that conventional crushing circuits are considered for world class

projects.

Trade off studies on larger scale operations compare SAG and HPGR concepts in most cases. The

economical differences between a conventional SABC circuit and a HPGR based circuit are mainly



determined by the behaviour of the applicable ore(s) in a SAG mill. Originally, SAG mill circuits

were sized on the basis of the Bond Theory using the Bond work index from test work. However,

the Bond Theory frequently failed on SAG mills in the past.

In recent years, SAG mills have been sized on either on the drop weight (DWI) test or the SPI test.

Outcome of the drop weight test is the so-called “A*b” parameter or the DWI of an ore.

Newmont (Veilette & Parker, 2005) [8] published a diagram covering the relation between the re-

quired SAG mill energy input and the ore competency expressed by the A*b parameter for a number

of their operating plants and projects [Figure 6]. The conclusion from this diagram is that the energy

requirements in a SAG mill may range from something like 4 kWh/t (for A*B = 80) to 16 kWh/t

(for A*b = 25) or in other words, the behaviour of individual ores could differ dramatically in SAG

mills – even in one deposit.

Figure 6: Newmont database of SAG power as a function of ore competency – A* b parameter [8]

A significant relation between the “A*b” parameter of an ore and the Bond work index (propor-

tional to the ball mill energy requirements) or breakage characteristic in a HPGR could not be

found.

Figure 7 includes a diagram (left hand side), showing the fineness vs. grinding force relation in a

pilot scale HPGR for ores with “A*b” parameters between 18 and 36. In fact, no significant differ-

ence in product fineness was found. This result suggests that the circulating load in a HPGR circuit

would not vary significantly even for ores within a wide range of “A*b” parameters or competency

respectively.

The correlation between the Bond work index and the “A*b” parameter of more than 30 ores is

shown in a second diagram (right hand side) included in Figure 7.



The Bond index varied only between 16 ± 2 kWh/t even though the A*b parameters were in the

range of between 18 and 75 indicating SAG mill energy requirements between about 5 kWh/t and

more than 16 kWh/t.

Figure 7: HPGR fineness vs. grinding force (left), Bond Index vs. „A*b“ ore competency (right)

The conclusions from these findings are that

• variation of ore competency is mainly an issue in SAG mills,

• ore competency is not a critical issue once ore particles are smaller than the feed size (< 3.15 mm)

in Bond testing, and

• HPGRs are able to efficiently equalize variations in ore competency (“The Big Equalizer”).

The implications of ore competency on the specific circuit energy as well as on required equipment

sizes and numbers are summarised in Figure 8. This comparison - on a very general desktop level -

compares three different ores with 4, 10 and 16 kWh/t energy input in a SAG mill and an equivalent

HPGR circuit producing a minus 6 mm ball mill feed. Any variation of the HPGR performance

treating these different ores is ignored due to the minor influence on the results. The ball mill energy

is mainly a function of the feed and required product fineness which may significantly vary for

different ore bodies and was therefore ignored for simplification. Ball mill energy may be typically

up to 10 % higher in the HPGR case depending on the required ball mill fineness. Nevertheless,

such a correction will not change the general conclusions from Figure 8.



Figure 8: Desktop study – SAG energy for ores of different competency vs. HPGR circuit energy

The conclusions from the desktop study are that a 100,000 tpd plant

• can be built with only one large SAG mill – provided the ore is low competent,

• would require at least 2 SAG mills for medium competent ores and up to 4 SAG mills for highly

competent ores,

• would require most typically 4 HPGRs independent of the ore competency,

• could be built with 3 HPGRs in case of applying the largest POLYCOM models available and/or

considering a coarser transfer size to the ball mills than 6 mm as often used,

• need to compare the installation cost of 2 to 3 SAG mills against the cost for 4 HPGR circuits in

case of medium competent ores for capital expenditure comparison,

• need to compare the installation cost of 4 SAG mills (38”) against the cost for 4 HPGR circuits

in case of extremely competent ores for capital expenditure comparison.

In general, it is concluded that hard and competent ores at large scale definitely favour HPGR based

circuits.

Wear cost, which are a major fraction of the operating cost, will be a function of the required energy

in the SAG and ball mill and will be dependent on the abrasiveness of the ore itself. Wear cost in

HPGRs are very much ore dependent. Wear cost for a SAG vs. HPGR comparison need to be evalu-

ated on a case by case basis.

5 Future Trends for HPGRs in Copper and for Other Minerals

A need for further reduction of total operating cost and in particular of the energy consumption in

tomorrow’s comminution circuits is obvious. This is a result of lower grade and harder ore bodies to

be treated in the future. Restriction of CO2 emissions to avoid global warming has become an im-



portant issue on the world wide political agenda and may result in a limitation of available energy

(CO2 certificates) or at least in higher energy cost than today. Only highly efficient comminution

concepts may make in particular “low grade/hard ore“ projects feasible at all, keeping capital

(capex) and operating (opex) expenditures at economical levels. Finally, availability of water is be-

coming more and more of an issue.

The introduction of HPGRs as the tertiary crushing stage in hard rock applications was a first mile-

stone to achieve energy and wear cost reductions for energy intensive ore bodies. Nevertheless,

there is the demand to utilise the potential of HPGRs to a greater extent than today.

There are three major focus areas for optimisation of HPGR – Ball Mill plants are:

• to reduce the investment cost of secondary crushing - HPGR – ball milling plants (SC-HPGR-BM)

which are percepted to be generally higher than those of traditional SAG – Pebble Crushing - Ball

Milling (SABC) circuits,

• to shift more energy from the ball mills and secondary crushers to the more efficient HPGRs in

order to improve the overall energy efficiency of comminution circuits,

• to further reduce the wear cost in the crushing, HPGR and milling sections.

Trade off studies often resulted in operating cost advantages of HPGR based circuits at higher capi-

tal expenditure as compared to traditional SABC circuits. In fact, this perception is not valid in gen-

eral. The economical differences of both concepts are mainly determined by the energy require-

ments of an ore in SAG mills. The required number of SAG mills – as already shown above – may

range from one to four “20 MW” mills for a 100,000 tpd plant whereas the number of HPGRs

would be four for the majority of ores. An investment cost disadvantage of HPGR circuits is ques-

tionable in case that four large SAG mills would be required. Mayor cost driver in HPGR based

circuits is the material handling equipment and, in particular, today’s practice to operate the secon-

dary crushers and HPGRs in closed circuit with screens.

Open circuit secondary crushing and multiple pass, open circuit grinding in HPGRs is sometimes

viewed as a potential approach for investment cost reductions due to expected plant simplifications

and – hopefully also – operating cost reductions.

On the paper, open circuit secondary crushing might be a viable option if ideal operating conditions

could be permanently guaranteed on the crushers. However, open circuit operation would not pro-

vide any protection against oversize ore entering the HPGRs circuit. Such oversize material could

pass the crushers under certain conditions, especially after gap releases in overload situations. Per-

manent as well as temporary oversize feed would reduce the lifetime of the HPGR wear protection.

For a long time there was or even still is the perception that HPGRs should be fed as coarse as pos-

sible to release the secondary crushers and ensure highest circuit efficiency. Although the lifetime of

secondary crusher’s wear parts may be quite short in some applications, the wear cost for the secon-

dary crushing duty is still relatively low. Wear cost typically increases disproportional in each com-



minution step towards the final product. From this point of view it should be more economical to

install additional secondary crushing capability rather than to tantalize the tertiary HPGRs with a

large or even oversized feed.

HPGRs would definitely benefit from a finer feed size. Firstly, the lifetime of the HPGR wear pro-

tection is significantly longer with finer feed and thus reducing wear cost. Secondly, a finer feed

could become the precondition to apply the next generation of “stud grades”. Those stud grates are

more wear resistant (longer lifetime) but also more brittle at the same time (higher risk of stud

breakage). Earlier recommendations with regard to the “top feed size to gap” ratio have to be re-

defined taking into account particularly the hardness or better ductility of the ore. The risk of stud

breakage has to be mitigated by matching ore abrasiveness, ore hardness, top feed size and applica-

ble stud grades. A comprehensive understanding of these interactions is the precondition to optimise

the interface between secondary crushing and HPGRs in order to ensure the application of the best

suited wear protection design for a particular application and to minimise wear cost.

Furthermore, a frequently discussed issue is: “How to shift more energy from the ball mills to the

HPGRs while simplifying the plant at the same time?” A distinction is to be made between two

principal approaches despite the fact that there is wide variety of possible flow sheet configurations

which may also combine the two said principles. The first option considered for efficiency im-

provements is to produce a finer HPGR circuit product by fine screening and accepting higher circu-

lating loads. Such an approach would not change the current plant design and layout significantly.

The second option is to use HPGRs in multiple pass, open circuit mode without intermediate classi-

fication. Supporters of this concept additionally hope to achieve a simplification of the plant result-

ing in lower investment cost. Further engineering studies are required to prove this expectation.

Four typical flow sheets deviating from the “standard flow” sheet are shown in Figure 9.



Figure 9: Alternative HPGR based circuits

Extensive testwork has been conducted using pilot scale HPGRs and lab scale ball mills to

investigate the difference of various flow sheet configurations in terms of overall energy efficiency

including the subsequent ball mill stage.



The general conclusions from this testwork were that

• screening at finer than current industrial screen sizes of 5 to 10 mm would improve the circuit

efficiency as long as the moisture content of the wet screen oversize could be maintained at

reasonable levels,

• fines removal between subsequent HPGR stages is essential for highest efficiency,

• any multiple pressing without prior fines removal resulted in reduced circuit efficiencies.

These findings are in line with industrial practise outside of minerals applications. Today, the Ce-

ment Industry still has the widest experience with regard to the most efficient utilisation of HPGRs.

Two important lessons can be learnt from the various cement applications. Firstly, HPGRs are 40 to

60 % more energy efficient than normal ball mills and secondly, high efficiency separation is the

key to actually obtain these potential energy savings. Any multiple pressing without intermediate

classification reduced the efficiency of the overall grinding circuits in cement. Based on these ex-

periences is obvious that highest energy savings in minerals would also be achievable if HPGRs

could eliminate the ball mills at all – as several applications on cement clinker, blast furnace slag

and limestones have proven in the past.

So far, the largest HPGR finish grinding applications in limestone are currently operated in an In-

dian cement plant. Two grinding plants, each incorporating one POLYCOM 21/16, are treating a

limestone with a Bond work index in the range of from 15 to 16 kWh/t. Each grinding plant pro-

duces between 320 and 370 tph of product at a P80 of 65 to 90 µm. The energy consumption for the

HPGR, fan, separator and bucket elevators is between 10.7 and 11.5 kWh/t depending on the prod-

uct fineness.

A Bond work index of 15 to 16 kWh/t is not unusual for typical copper ores. Assuming an identical

HPGR size and same grinding and classification efficiencies, the said Indian plant would have the

potential to grind in the order of 450 tph of copper ore to flotation fineness of 150 µm and up to

650 tph with a larger POLYCOM 24/18. Finish grinding with a HPGR could reduce the energy re-

quirements by in the order of 40 % compared to HPGR/wet ball mill circuits as applied today.

Definitely, there are questions and challenges around such a dry finish grinding system for mineral

applications, like moisture, wear, maintenance, etc. Nevertheless, this would be the ultimate ap-

proach to minimise energy requirements for grinding ferrous and non ferrous ores.

The evolution to expand efficient HPGR utilisation in today’s crushing and milling circuits will be

an on-going development in the near future. A revolution would be to apply HPGRs for finish

grinding.



6 Summary

HPGRs have established themselves as an accepted and well matured technology which is able to

significantly reduce the operating cost of copper concentrators and other minerals applications.

The barrier “I do not want to be the first using a new technology” has been broken successfully by

installations such as Cerro Verde, Grasberg, Boddington and Mogalakwena. Feasibility studies of

further world class projects in Chile and Pakistan are solely based on HPGR technology.

HPGRs broke the barrier for some recent and current projects to become viable at all from an eco-

nomical standpoint, especially in the case of competent, low grade ore bodies which would require

high energy inputs in conventional SAG mills.

Nevertheless, HPGRs are a relatively young technology compared to well known conventional

crushing and grinding circuits which have been operated for decades. There is a focus on opportuni-

ties to optimise the first generation of HPGR circuits in terms of operating and capital cost. This

will be an ongoing process in the next couple of years.

Application of HPGRs in heap leach installations as well as for finish grinding to flotation fineness

could be the next milestones in the copper industry.

References

[1] Patzelt N., J. Knecht, E. Burchardt, and R. Klymowsky: “Challenges for High Pressure Grind-

ing in the New Millenium”; 7th

Mill Operators Conference 2000, Kalgoorlie, WA. 47-55.

[2] Patzelt N., R. Klymowsky, E. Burchardt, and J. Knecht: “High Pressure Grinding Rolls in

AG/SAG Mill Circuits. Proceedings”; SAG 2001, Vol. III. 107-123.

[3] J.L. Vanderbeek, T.B. Linde, W.S. Brack, and J.O. Marsden: “HPGR Implementation at Cerro

Verde”; SAG 2006, IV 45 – 61

[4] M. Mular, John Mosher: ”A Pre-production Review of PT Freeport Indonesia’s High Pressure

Grinding Roll Project”; SAG 2006, IV 62 - 79

[5] S. Dixon, B. Olson, E. Wipf: “Squeezing an extra 30% of a typical SABC circuit for 4.8 kWh/t”;

SME 2010

[6] C. Rule: “Development of a Process Flow Sheet for the New Anglo Platinum Concentrator In-

corporating HPGR Technolgy”; SAG 2006, IV 94 – 109

[7] C. Rule, D. Minnaar, G. Sauermann: ”HPGR – Revolution in Platinum”; Third International

Platinum Conference, The Southern African Institute of Mining and Metallurgy, 2008.

[8] B. Parker, G. Veillette: “Boddington Expansion Project”; Randol Conference, Perth, 2005


Evaluation of Copper Losses in the Slag Cleaning

Circuits from Two Chilean Smelters

N. Cardona, L. Hernandez, E. Araneda, R. Parra L. Bahamondes, R. Parada

Universidad de Concepción Chagres Smelter, Anglo American

Edmundo Larenas 285 Camino Troncal s/n

Concepción, Chile Catemu, Chile

J. Vargas, M. Artigas

Paipote Smelter, ENAMI

Camino Público s/n

Copiapó, Chile

Keywords: Pyrometallurgy, copper losses, slag cleaning

Abstract

The slags from two cleaning processes circuits were analyzed in order to evaluate the copper losses

and identify options for its control. The first is a bath smelting unit (Teniente Converter) with an

electric slag cleaning furnace (TC-EAF) and the second is a flash smelting (Outokumpu) with a

tilting furnace (FSF-SCF). A sampling campaign was done in each circuit during the furnace opera-

tions and two sampling devices were tested. Bulk chemical analyses of the slags samples were

combined with the characterization of the solid phases using optic microscopy and electron probe

microanalysis (EMPA). The quantification of phases and distribution of particle sizes were obtained

using Qemscan®

and image analysis techniques with special attention to copper containing phases.

Characteristics phases as “solid matrix”, “fayalite”, spinels and sulfides in matte droplets were iden-

tified in all the slags. Differences appear on the proportions of these phases depending of the slag

chemical composition, oxidizing/reducing condition of the process and the sample cooling. Some

distinctive characteristics of the copper losses and the presence of Fe-spinel equiaxial grains were

used to explain the behavior of the matte droplets during the settling and to estimate the reducing

condition of the melt slag.

1 Introduction

The copper losses in slag are of chemical and mechanical nature; the chemically dissolved cationic

copper (Cu+) is stable in the slag and depends on the matte grade (oxygen-sulphur potential ratio),

temperature and slag composition [1, 2, 4]; the mechanical losses include entrapped or floated un-

settled matte drops and metallic inclusions. The copper content in these phases may be higher or

Cardona, Hernandez, Araneda, Parra, Bahamondes, Parada, Vargas, Artigas


lower than the furnace matte grade [3] in the smelting unit. The majority of copper in industrial

slags is found as mechanically entrained matte particles with sizes varying from several millimeters

to a few microns [2], its recuperation is accomplished in the slag cleaning process.

The pyrometallurgical slag cleaning processes are based on slag reduction and settling of matte par-

ticles. It uses a reductant (coke, graphite, oil or natural gas) for minimize the content of solid mag-

netite and improve the settling of matte droplets toward the matte phase. Many factors influence the

copper recovery and the quality of the final slag: Chemical composition, temperature and opera-

tional times, the nature of the copper losses and type of reductant and reduction mechanism. The

operational control is usually based in the analysis of total copper content and the magnetite content

by the Satmagan balance in the final slag. This information is limited to establish the nature and

distribution of the copper losses in the different slags and to evaluate the effect of the slag composi-

tion in the copper recovery.

In order to evaluate the slag quality in the cleaning circuits a complete characterization of the slag

currents should be done, obtaining the final copper losses distribution that can be related to the con-

trol process parameters and chemical characteristics. The knowledge of the chemical and minera-

logical composition of the solidified samples allows an approximation of the liquid condition of the

slag from a thermodynamic description of the equilibrium between theirs components. In this per-

spective we have developed a methodology for the systematic study of copper losses based in the

bulk chemical and mineralogical characterization of phases in slag samples obtained during the tap-

ping and from the bath into the furnaces. The characterization and sizes distribution of slag phases

associated with copper losses were obtained using analysis on mirror polished surfaces. This paper

presents the results of this analysis on slags from cleaning circuit of two Chilean smelters: The Her-

nan Videla Lira (HVL) Smelter and The Chagres Smelter.

1.1 HVL Slag Cleaning Circuit (TC-EAF)

The HVL smelter of ENAMI (Empresa Nacional de Minería) is situated near Copiapó city in Chile

and has 340 kt of concentrate processing capacity. The process flow sheet is showed at Figure 1. In

2008 the HVL smelter generated 184.5 kt of dump slag with 0.8 % of average copper content. A

Teniente Converter (TC) operates as a unique smelting unit, it produces highly oxidized slag (high

magnetite and dissolved copper contents) that are cleaned in an electric arc furnace (EAF) with ca-

pacity for the treatment of 720 t of slag and 120 t of solid reverts. The TC slags is tapped each 0.8-

1 h and transported by ladles to the electric arc furnace (EAF), then is transferred by launder and

charged through the furnace wall. The tapping time is 0.5 h aproximately and produces 4 ladles of

15 t each.

Evaluation of Copper Losses in Slag Cleaning Circuits from Two Chilean Smelters


Figure 1: Current flow sheet of the Hernán Videla Lira Smelter [7].

The EAF furnace operates in semi-continuous mode. It uses solid coke as reductant (12-14 kg/ton of

slag) and the reduction occur both on the surface of the carbon electrodes and at the slag/floating coke

interface. The average consumption of energy is 134 kWh/ton (slag +reverts) and the temperature of

molten phases is between 1230-1270 °C. The maximum level of molten bath is 1.7 m. The copper

matte from the EAF is tapped (fourth times in turn) and returned to the TC. The final slag is tapped

from the EAF, it flows by a launder to charge 5 or 6 ladles of 6 ton each and finally is transported

toward the dump yard.

1.2 Chagres Slag Cleaning Circuit (FSF-SCF)

The Chagres smelter, part of the Anglo Base Metals Division, Anglo American Plc is situated

80 km north of Santiago in Chile and has 600 kta of concentrate processing capacity. In 2008 pro-

duced 315 kt of slag with 1.1 % Cu content. Process flow sheet is showed in Figure 2. The smelter

unit is an Outokumpu-Flash Furnace (FSF) and operates with instantaneous rate of 78-80 t/h dry

feed. The slag is tapped each 1.5 h after the matte tapping and flows through a refrigerated launder

to one of the Teniente slag cleaning furnace (SCF), tapping time varies from 0.3 to 0.5 h.

Escoria

Inyección

Conc. Seco

Metal Blanco

Anodos

Inyección

Mezcla CPS

Plantas de Acido Sulfúrico

SecadorSecador

Escorias a BotaderoEscorias a Botadero

CT

CPSCPS

RAFRAF

Gases

Gases

Blister

POXPOX

HELEHELEHELE

Precipitador

Electrostático

Gases

Metal

Blanco

CirculantesCirculantesPrecipitador

Electrostático

Gases

RecepciónRecepción

Mezcla RAMMezcla RAM

Gases

Polvos HE

CNU - sílice



The two slag cleaning furnaces with 90 slag capacity each operates in discontinuous mode with 1.3-

1.5 h per cycle. The converter slag is charged with ladles through the mouth before the FSF slag

was charged. The reduction step uses fuel oil (API Grade 6) at a rate of 6.6-7 l/min with an air ra-

tion of 1.6-1.7 Nm3 air/l fuel oil through two tuyers during 0.5-0.7 h. The reduction step is followed

by the slag settling for 0.7-1 h. The final slag is skimmed through the mouth in two ladles of 33 ton

each and transported to the dump yard. The temperatures are between the 1230-1250 °C and there is

a burner air-oil for thermal support. The metal is tapped each 4-5 cleaning cycles and returned to the

Pierce Smith converter in the Fe-blowing stage.

Figure 2: Current flow sheet of the Chagres Smelter [6].

RECEPCION

TOLVAS

ESCORIA

FINAL

ESCORIA

FLASH

POLVO

Concentrado Seco

R. Radiación R. Convección R. Lurgi

EJE ESCORIA CONVERSION

Bomba Moller

Cobre

Blister

Gases Combustión, Quemador

y Reducción

ANODOS

BLISTER

Camara de Mezcla

VTI Conversión

VTI Flash

C Rad

4CPS 4

CPS 3

CPS 2

Lurgi

Horno Flash

C Rad

3

C Rad

2

HLE 2

HR 2

HR 1Rueda

de

Moldeo

TOLVA

400

BroncesBotadero Conchos

Fondos

Botadero Escoria Descarte

PLANTA DE

ACIDO

HLE 2

Tolva

50

Secador 1 Secador 2

Soldado +

Terceros Silice

Caldera

Recuperadora

POLVO

CPS 1

C Rad

1



2 Experimental

2.1 Slag sampling campaigns

The slags from TC-EAF circuit were sampling during three consecutive TC tapping following by

EAF tapping in a period of 6 h. Two samples of the TC slag (1st and 3

rd ladles) were taken before

the charge to the EAF and two samples (1st and 3

rd ladles) of the corresponding final slag were tak-

ing at the end of the cleaning cycle. The furnaces operational conditions are showed in the Table 1.

The samples were taking with a steel ladle of 500 cm3 and air cooled.

Table 1: TC-EAF parameters

Parameters/turn [8 h] TC EAF

Dry concentrate [t/h] 42.3

Enrichment oxygen [%] 37

Reverts charge [t] 32.3 38

Slag Temperature [°C] 1200 1230

Coke charge [t] 4.3

Matte Grade [%] 74.8 71.3

The sampling campaign in the FSF-SCF circuit was done during a complete slag cleaning cycle in

the SCF #2 in a period of 1.5 h. The FSF slag was the only feed and no reduction was done. The

operational conditions are showed in the Table 2. A sample of the FSF slag was taken in the launder

with a steel ladle and cooled with water. Three samples of the SCF slag were taken with a steel bar

grooved through one tuyere in three times: at the beginning, in the middle and at the end of the set-

tling period. The ECF samples collected with the steel bar were cooled by air.

Table 2: FSF-SCF parameters

Parameters FSF SCF

Smelting rate [t/h] 76

Matte Grade [%] 60 76

Settling time [t] 60

Slag Temperature [°C] 1296 1250

Maximum Bath level [m] 1.8

Bath level sampled with steel bar [m] 1.5

2.2 Methodology and experimental techniques

The samples obtained from both circuits were milled and submitted for chemical analysis and bri-

quettes were appropriated prepared for mineralogical characterization. The bulk chemical elemental

composition was determined using atomic absorption spectroscopy for Cu, Fe, Ca, Al, Mg and Zn;

gravimetric analysis for SiO2 and LECO titration for S. The magnetite contents were obtained by



Satmagan™. Polished surfaces on briquettes of samples were examined under an Olympus CK40M

microscopy in reflected light and photomicrographs were taken.

The first step of the identification and microanalysis of phases in industrial slags consisted in the

definition of standards that will be used to identify the phases in different slag samples. The phases

were identified by X-ray powder diffractometry (DRX), analysis of energy dispersion spectrums

(EDS) and backscatter electron image (BSE). The chemical composition of the sulfides, oxides and

silicates phases was determined by electron probe microanalysis (EPMA) in a JXA JEOL JXA-

8600 Electron Probe Microanalyser with Wavelength Dispersive Detectors (WDS) using ZAF cor-

rection. The common phases in the different slags were grouping according to the chemical compo-

sition, association mode and microstructure characteristics developed during the cooling (Table 3).

The slag samples were analyzed comparing with the standards and characteristics or new found

phases didn't allow a complete match between the sample and the standards, new analysis were per-

formed.

After identifying phases, the slag phases’ composition and copper phases’ sizes distribution were

obtained means two procedures: 1) Image analysis on microphotographs with specialized metal-

lographic software (Scentis by Struers) for the TC –EAF slag samples; and 2) Analysis of BSE im-

ages and EDS spectrums with Qemscan®

Tescan system in mode PMA for the FSF-SCF slag sam-

ples.

Table 3: Phases obtained by XRD, EPMA in Copper industrial slags

Slag phase Characteristics (composition, morphology and associations)

Cu-Phase

Metallic copper on large spherical inclusions of matte (matte droplets).

Rich- Copper alloy (Cu>80 wt. %, As, Sb, Pb) with Cu-sulfide pherifery on inclusions of

matte with high Copper content (high grade matte droplets).

Cu-Sulfides Chalcocite or Digenite on large spherical inclusions of matte (matte droplets).

Low Cu-

Sulfides

Cu-(low Fe) sulfides: Bornite solid solution (Cu1-xS-FeS) on matte inclusions.

Mixed sulfides: Pyrrotite (FeS)-Bornite solid solution (Cu1-xS-FeS) on mixed inclusions

(This phase appears on solid matrix and grows during the cooling).

Fe-silicates

"Fayalite"

Fe-silicates (Fe2SiO4) on acicular or dendrites grains (It appears and grows during the cool-

ing and is a major phase in the slags).

Fe-spinel

“Magnetite”

Fe - (Al, Ti, Zn, Cr) oxides phase, on equiaxial grains.

Fe-O phase dendritic grains (This phases appears and grow during the cooling)

Cr-Spinel Cr-Al-Fe-Mg oxides phase on large equiaxial grains with degraded texture.

Silicates

“Solid matrix”

Heterogeneous phase with Al, Ca, Mg, Fe Silicates surrounding the "Fayalite" grains.

Vitreoux phase with Fe (<54 wt. %), Si(>30 wt. %), Al, Ca, Mg, and Cu, S.



3 Results

3.1 TC-EAF slag cleaning circuit

The chemical composition of the samples from the TC-EAF circuit presented in Table 4.

Table 4: TC-EAF slag chemical composition

Slag Courrent TC tapping Slag Final Slag

Tapping 1st 2

nd 3

rd 1

st 2

nd 3

rd

Ladle/total ladles 1/4 3/4 1/4 3/3 2/4 1/4 3/6 1/6 3/6 1/6 3/6

wt. %/Specimen

TCS-

1

TCS-

2

TCS-

3

TCS-

4

TCS-

6

EFS-

1

EFS-

3

EFS-

4

EFS-

6

EFS-

7

EFS-

8

Cu 7.03 8.60 8.20 8.05 6.30 0.75 0.85 0.70 0.65 0.58 0.58

Fe 40.40 38.20 37.60 39.60 39.80 46.2 46.2 46.2 46.4 46.00 46.20

S 1.98 2.59 2.58 2.44 1.66 1.00 1.07 0.93 0.8 1.0 1.06

SiO2 24.19 23.72 31.36 24.36 24.36 26.18 26.74 27.17 28.12 28.43 28.17

Fe3O4 22.85 20.2 19.7 22.15 23.75 6.10 6.00 3.90 4.10 5.00 5.40

Al2O3 3.56 3.53 4.24 3.69 3.77 1.26 0.67 0.90 3.74 3.56 3.68

CaO 1.26 0.91 0.85 0.93 1.2 0.46 0.50 0.52 0.58 0.43 0.71

MgO 0.76 0.78 1.46 0.74 0.76 0.87 0.86 0.89 0.76 0.70 0.48

ZnO 2.54 2.61 2.58 2.69 2.6 0.8 2.18 2.27 2.17 2.07 2.20

BSE images and microscopic images of selected slag samples are shown in Figure 3.



Figure 3: BSE images and microphotographs of slag samples from TC-EAF circuit. a) view of

TCS-6 slag surface; b) view of EFS-6 slag surface; c) matte drop in TCS-6; d) mixed

sulfide on solid matrix of EFS-1; e) matte drop in EFS-3; f) rich Cu alloy phase with

Cu-sulfide on a high copper matte inclusion of EFS-6. (Mte=Matte, H-Mte = High

copper grade matte; Sm=solid matrix; S-mix=Mixed sulfides; Sp=Spinel phase).

In some EAF slag samples (EFS-1, EFS-6, EFS-8), large matte inclusions with a rich copper alloy

phase (high copper grade matte) were identified (Figure 3f) with sizes between 80 to 500 µm. Those

particles were not present in the TC tapping slag and could not be formed in the EAF under

reductant condition, probably they came from the solid reverts charged into the furnace, which is a

material with high oxidized slag (total Cu>30 wt. %).

The slag phases mass distribution was obtained by image analysis from monochromatic micropho-

tographs using densities of pure mineral phases and calculated density of solid matrix [9]. The re-

sults are showed in Table 5. The sizes distribution of matte inclusions (Cu-Sulfides and Cu-(Low

Fe) sulfide phases) in TC and EAF slags for particles with Feret´s diameters bigger than 10 µm are

shown in Figures 4 and 5.



Table 5: TC-EAF slag’s phases distribution

Slag Courrent TC tapping Slag Final Slag

Tapping 1st 2

nd 3

rd 1

st 2

nd 3

rd

wt. % phase/Sample TCS-1 TCS-2 TCS-3 TCS-4 TCS-6 EFS-1 EFS3 EFS-4 EFS-6 EFS-7 EFS-8

Cu- phase 0.02 0.29 0.05 0.03 0.03 0.02 0.06 0.03 0.3 0.02 0.03

Cu-Sulfides 5.47 12.64 10.37 12.26 12.48 1.23 1.37 2.87 1.16 0.70 0.78

Cu-(Low Fe) sulfides 5.80 3.82 3.82 5.75 4.38 0.89 1.03 1.64 0.62 0.58 0.62

Mixed sulfides 0.27 0.35 0.32 0.28 0.46 0.09 0.61 0.63 0.31 0.8 0.65

Fe-Spinel 23.77 19.8 15.32 16.38 21.86 7.39 7.18 11.1 6.23 8.04 6.2

“Fayalite” 45.67 44.25 45.52 48.19 47.82 63.57 53.69 51.31 54.42 61.94 70.8

Solid Matrix 19.46 21.99 27.83 21.85 16.59 27.64 37.02 33.93 37.54 28.48 21.5

Figure 4: Particle sizes distribution of matte inclusions in TC tapping slag samples.

0.1

1.0

10.0

100.0

0

39

79

120

160

200

240

280

320

360

400

440

480

520

560

600

640

680

720

760

800

840

880

920

960

980

Rel

ativ

e fr

equen

cy, %

Diameter, µm

TCS-1 TCS-2 TCS-3 TCS-4 TCS-6



Figure 5: Particle sizes distribution of matte inclusions in final slag samples.

The evaluation of the chemical transformation of species in the slags considered the calculation of

Fe3O4 reduction using “magnetite” contents (by Satmagan balance) from chemical analysis and also

Fe-Spinel phase contents from the phase distribution. This information was related with the total

copper recovery and the Cu-sulfide recovery in the Figure 6.

Figure 6: Cu-recovery for three slag cleaning cycles.

0.1

1.0

10.0

100.0

0

39

79

120

160

200

240

280

320

360

400

440

480

520

560

600

640

680

720

760

800

860

900

940

980

Rel

ativ

e fr

equen

cy,

%

Diameter, µm

EFS-1 EFS-2 EFS-3 EFS-4 EFS-5 EFS-6

0

20

40

60

80

100

0 1 2 3 4

%

Slag cleaning cycle

Cu-Recovery Sulfide-Recovery

Fe3O4 Reduction Fe-Spinel Reduction



According to the Figure 6, the total copper recovery and Cu-sulfide recovery in the EAF are inti-

mately related as is expected. When the “magnetite” or Fe -Spinel reduction is increased, the copper

recovery follows the same trend. It can suppose that there is an opportunity to improve copper re-

covery increasing the Fe-Spinel reduction.

3.2 FSF-SCF slag cleaning circuit

The chemical compositions of the slag samples taken from the SCF bath during the settling period

are showed in the Table 6. The varying of the total Copper and “magnetite” (by Satmagan™ analy-

sis) contents for each level during the settling period are showed in the Figure 7.

Table 6: FSF-SCF slag chemical composition

Time Beginning of settling Middle time of settling End of settling In- Slag

Sample H1-1 H1-2 H1-3 H1-4 H1-5 H2-1 H2-2 H2-3 H2-4 H2-5 H3-1 H3-2 H3-3 H3-4 H3-5 EF-2

Level *[cm] 127 106 85 64 43 127 106 85 64 43 127 106 85 64 43 0.0

% Cu 0.86 0.92 1.69 2.51 10.40 0.73 0.72 0.92 1.36 2.18 0.74 0.67 0.72 1.50 1.89 1.86

% Fe 39.90 39.80 39.20 41.70 37.10 39.40 39.70 40.00 40.40 42.80 40.10 40.00 40.00 40.00 42.20 39.00

% S 0.56 <0.3 0.80 0.60 2.99 <0.5 <0.5 <0.5 0.89 0.83 0.64 <0.5 <0.5 0.68 0.80

% SiO2 30.20 30.20 29.40 27.10 24.30 30.50 30.40 30.10 29.80 26.40 30.80 30.50 30.40 30.00 27.70 29.60

% Fe3O4 10.00 10.50 11.50 16.50 14.70 11.00 10.80 11.00 11.50 17.80 10.50 10.10 9.50 11.50 15.60 8.70

% Al2O3 6.59 6.70 6.63 5.81 5.34 6.73 6.33 6.31 6.44 5.87 7.16 6.89 6.74 6.64 6.15

*The reference is the bottom of the furnace (with Level=0).

Figure 7: Copper and magnetite content during the settling period at different height from bottom

in the SCF furnace.

As shown in Figure 7, the copper contents are lower in the slag from the upper part of bath (127 cm

to 85 cm) with magnetite contents of 10-11.5 wt. %. The slag near the matte furnace (64 cm and

43 cm) has high copper and magnetite contents at the beginning of settling. The copper content

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0 10 20 30 40 50 60

Cu, wt%

Time (min)

h=64cm h=43cm h=127cm h=106cm h=85cm

5.0

7.5

10.0

12.5

15.0

17.5

20.0

0 10 20 30 40 50 60

Fe3

O4, w

t%

Time (min)

h=127cm h=106cm h=85cm h=64cm h=43cm



decreases faster during the first 30 min, and slowly after. Unlike other levels, an important magnet-

ite reduction was observed in the slag in 64 cm at the beginning of settling (30 min). After that, the

magnetite contents decreases in almost all the bath. In the slag near the matte (height = 43 cm), the

magnetite contents increases during the first 30 min and then diminishes. This behavior can be ex-

plained due to the settling of solid magnetite at the beginning of settling period and subsequent re-

duction by interaction with the bottom matte in the furnace.

BSE images of some selected samples are showed in the Figure 9. The outlined sector in Figure 9-c

and BSE image in Figure 9-b show a vitreous aspect of slag that was in contact with the sampler

wall, that means that this part of the slag was effectively quenched avoiding any fayalite cristaliza-

tion during solidifcation.

Figure 8: BSE microphotograph of slag samples from FSF-SCF circuit. a) View of center sector

and b) view of border sector, both of H3-5 sample, c) BSE map on polished surfice of

H3-5 sample; d) Fe-Spinel and “fayalite” phases of H2-2 slag sample, e) solid matrix,

“Fayalite” and Fe-spinel’s grains and f) matte inclusion, both in EF-2 sample. (Sp=

Spinel phase; “Fy”: “Fayalite” phase; Sm=solid matrix; Mte=Matte).

The slag phase’s distributions and sizes distribution of the Cu-sulfide inclusions (matte droplets)

were obtained using Qemscan®

-Tescan system in mode PMA with 2000 µm of field phase, 25 kV

voltage acceleration. The slag phase’s distributions at 106 cm height at the end of the settling are

showed in the Table 7.



Table 7: Phase distribution in SCF samples

Settling time [min] 5 30 60

Level from the bottom of furnace [cm] 106 106 106 85 43

Sample slag H-1-2 H-2-2 H-3-2 H-3-3 H-3-5

Cu-metal 0.00 0.00 0.00 0.00 0.03

Cu-s 0.52 0.49 0.49 0.59 1.92

Cu-S-(low Fe) phase 0.74 0.45 0.45 0.33 0.92

Fe-Spinel [oct] 10.98 5.66 7.04 5.64 19.97

Cr-Spinel 0.00 0.00 0.00 0.00 1.28

Solid Phase (Matrix) 85.94 87.74 90.52 85.41 64.89

Solid slag: Solid matrix and "fayalite" 87.76 93.40 92.03 93.44 75.89

The spherical inclusions of matte (matte droplets) were formed by Cu-sulfides and Cu-(Low Fe)

sulfide phases. Inclusions with mixed sulfides were not identified in the FSF-SCF samples unlike

the TC-EAF slag samples, because the cooling was faster in first ones. The majority of the “faya-

lite” appears on needles or dendrites with composition within the range of iron-silica glass, this

phase could not differenced clearly by Qesmscan®

analysis, then were incorporated with the solid

matrix in a phase called "solid slag".

The sizes distribution of the Cu-sulphides inclusions (matte droplets) for three sampled bath height

and at the final of the settling period was obtained considering particles with diameter bigger than

20 µm. In the Figure 9 are showed the histograms and three images with scale that allows the identi-

fication of the phases according to the color standards used by Qemscan®

equipment.



Figure 9: Sizes distribution of Cu-Cu-sulphides inclusions at the end of settling period in the SCF

obtained by Qemscan with diameters higher than 20 µm. Sulfides and copper in matte

inclusions at different heights: h=106 cm, b) 86 cm, c) 43 cm.

a

0

0

1

10

100

200195185175165155145135125115105958575655545352414

Re

lative

fre

qu

en

cy %

Diameter, µm

5 min 30 min 60 min

Background

Cu-metal

Chalcocite/Digenite

Covellite

Cu-S-(low Fe) phase

Fe-O phases

a b

c

Colour Standard



The sizes distribution obtained in the slag samples of the upper zone of the bath are almost similar:

All the particles are below 75 µm. There are 1 % of 100-190 µm droplets in the slag in the zone

near the matte furnace. Images with “False color” were obtained by Qemscan®

allow visualizing the

different sulfides and copper in matte inclusions in the Figure a), b), c). Those images show that the

proportion of the Cu-Sulphide phase in inclusions in the slag near the bottom matte is higher than

the inclusions in the upper zone of the bath.

4 Discussions

During the cooling of the samples, iron oxides like Fe3O4 or Fe-spinel phase precipitates. This con-

tent is part of the “magnetite” measured with the Satmagan balance. This measurement has a prob-

lem of representativeness such the amount of precipitated magnetite grains varies with the Fe as

Fe3+

content in the slag and the cooling rate of the slag sample. The classical report of the Fe3O4

content therefore depends of the sampling method [1, 3, 4]. Only the magnetite determination on

granulated slag is in good agreement with laboratory values [4] and should be used for the opera-

tional control.

This study two “magnetite” phases with different morphology has identified (see Fe-Spinels phases

in Table 1), it was only possible the quantification of the total content because the brightness and

magnetic properties are similar for both phases, therefore the analytic techniques available are lim-

ited. “Magnetite” such dendrites appear during the cooling slag sample depending of the Fe3+

dis-

solved in the slag and the cooling rate. If the cooling rate is slow enough to approach pseudo equi-

librium in solidification, this phase can grow to form equiaxial crystals and is not possible to

discriminate between the magnetite that came from the melt and the one formed in situ during cool-

ing.

In this study, as well in other published reports [2, 9], both phases are commonlly found in indus-

trial slags, due to a non homogenous and slow cooling of samples. This situation explains why no

relation was found between magnetite contents on bulk samples analysis by Satmagan and Fe-

Spinel phases quantification using image analysis or Qemscan®

.

The settling rate of matte droplets (particles with sulfides phases) can be calculated by the modified

Stokes-Law (Hadamard-Rybeczynski formula) as a first approximation, supposing spherical and

constant diameter matte particles, laminar and homogeneous flow, constant temperature and chemi-

cal composition.

( )2

m e

e

g ρ -ρ du=

12µ (1)

Where: u=settling rate [m/s], ρm=density of matte drops [kg/m3], ρe= density of slag [kg/m

3],

d=diameter of matte drops [m], µe=slag viscosity [Pa·s].

The viscosities were calculated by the Kv method [8] based in the chemical composition of slags

(Table 4 and 6). The values 0.18 Pa·s and 0.12 Pa·s were used for the SCF and EAF, respectively.



Considering 60 minutes of settling in the SCF (see Table 2), can be recovered all the particles with

a diameter higher than 135 µm. The normal operation is 40 min of settling, for this condition, the

calculated critical diameter is 145 µm. The high SiO2 and Al2O3 contents affect the basicity of final

slag and the viscosity, also solids of Fe-Spinel and Cr-Spinel in the zone near the matte furnace may

obstruct the settling of matte drops.

In the EAF the critical diameter is 125 µm for an operational residence time of 60 min. As it is

shown in Figure 5, there are particles with sizes bigger than this critical size; these particles could

be recovering with a specific strategy for tapping practice. This strategy must consider increasing

the time between tap to tap, a specific amount of liquid slag tap for each tapping and an appropriate

control of reverts addition.

The copper enrichment in matte drops from SCF slags and the reduction of magnetite contents in

zones near the matte slag interface could be explained by the interaction of FeS in matte droplets

with the magnetite. A similar phenomenon was studied by other authors [3] in FSF slags with two

matte ranges (60-63 wt. % Cu and 72-75 wt. %Cu). They explained such phenomena as an equali-

zation of the composition of low grade droplets (matte droplets proceeding from FSF) with the fur-

nace matte by the Fe3O4 reduction for slag formation:

FeS (matte) + 3Fe3O4 (slag) + 5SiO2(slag) = 5(2FeO·SiO2)+SO2 (2)

The FeS content in the matte droplets decreases and magnetite is reduced. The inversion tempera-

ture in standard conditions is 1167 °C. The FSF matte grade coexisting with the FSF slag is

60 wt. % of copper; the sulfide phase tapped from the SCF was about 76 % Cu after 4 cycles of slag

charging and tapping. Considering this condition and the Cu-sulfide phase contents in the slag

(Table 6), it is feasible to suppose the copper contents in the matte droplets varying between 60 and

76 %.

The PSO2 for the SCF can be determined in function of the equilibrium constant KI of Reaction (2)

from the FeS content and the slag composition at Temperatures of 1210-1250 °C. Some primary

solids of Fe-Spinel are presents in the SCF (see the Figure 9b), then the a(Fe3O4) could be consid-

ered near unity.

( )3 4

2

2 2

3

5

2

(́ )( / )

I Fe O FeS

SO FeS

FeOSiO SiO

K a ap K T a

a a= =

(3)

Supposing a constant temperature (1250 °C), total pressure (1 atm), the a(FeS)matte were calculated

by FactSagetm

for the range of copper contents in matte droplets (60-76 % Cu) and the values are

between 0.149-0.027. The ratio a(2FeO·SiO2)/a(SiO2)≈1.93 (activities were calculated by Fact-

SageTM

). The pSO2 obtained for the SCF are in the range of 0.025-0.002 atm. Those values are ac-

cording with other thermodynamic calculations for the SCF (0.1-1 atm, 1250 °C and 60 % Cu in

matte) [2].



5 Conclusions

The copper losses in TC-EAF are mainly in the form of mechanical matte spherical inclusions

(matte droplets) constituted by Cu-sulfides and Cu-(Low Fe)-sulfides. The high grade copper matte

particles in the EAF containing a rich copper alloy phase and Cu-sulfide phases in the final slag is

explained by a insufficient settling time of the reverts that are continuously added to the furnace.

The reduction process in the EAF can be improve and major control of reverts rate addition must be

done in order to improve the copper recovery.

The nature of copper losses in slag from the FSF-SCF circuit has been clarified by sampling the

slag from different height of the bath of the SCF and doing mineral analysis of the samples to cam-

pare it with the mineral analysis of the slag charged. The sizes distribution of copper losses at dif-

ferent levels was obtained and a range for pSO2 was determined as function of the a(FeS). In order

to improve the copper recovery, a higher settling time can be used and the chemical composition of

slag should be controlled by reduction of oxides that the increase viscosity (e.g. diminishing the

Al2O3 content increase the slag basicity), an optimizing of the Fe/SiO2 ratio and temperatures are

convenient for the control of the Fe-Spinel formation in the SCF.

Especial care should be taking with the magnetite analysis in industrial slag samples. The origin of

different sources of Fe-Spinel must be considered. Some equiaxials grains were detected in a

quenched surface portion of an industrial slag sample in this study; it´s an evidence of the solid

phases in the bath, that can be affect the settling of matte droplets. More slag samples with appro-

priate cooling should be analyzed, and EPMA measures of copper dissolved in the solid matrix

phase must be done to understand better the complex phenomena of the copper losses in slag.

References

[1] JALKANEN H., VEHVILAINEN J., POIJARVI J (2003). Copper in Solidified copper

smelter slag. Scandinavian Journal of Metallurgy. P 65-70.

[2] IMRIS, I., REBOLLEDO S., SANCHEZ M. , CASTRO, G. ACHURRA G. & HERNANDEZ

F. (2000). The copper losses in the slags from the el teniente Process. Canadian Metallurgical

Quarterly, Vol 39, No 3, pg 281-290.

[3] GENEVSKI, K. & STEFANOVA, V. (2008). Dispersed droplets in industrial slag melts from

flash smelting furnace. Canadian Metallurgical Quarterly, Vol 47, No 1 pp51-58.

[4] SRIDHAR, R. TOGURI, J.M. AND SIMEONOV. S. (1997). Copper Losses and Thermody-

namic considerations in Copper Smelting. Metallurgical and materials transactions B. Volume

28b. Pg.191-200.

[5] WARCZOK, A., RIVEROS, G., ECHEVERRÍA, P., DÍAZ, C.M. (2002). Factors governing

Slag Cleaning in an Electric Furnace. Canadian Metallurgical Quarterly. Vol. 41, No 4 . pp.

465-474.



[6] PARADA, R., CHACANA, P.(2008). Chagres Evolution after Debottnecking project.

12th

International Flash Smelting Congress. October 26-31. China.

[7] NAVARRO, V., VARGAS J. C., PARRA, R.(2007). Minor elements behavior and develop-

ment of predictive model at The Teniente Converter- Electric Furnace System in the HVL

Smelter. First Meeting on Minor Element Contaminants in Copper Metallurgy. Chile.

[8] BAZAN, V., GOÑI, C., CASTELLÁ L., BRANDALEZE E. VERDEJA L. F. Y PARRA, R.

(2006). Estimación de la viscosidad de escorias fayalíticas utilizando el método kv y el

método experimental del plano inclinado. Revista de Metalurgia, 42(2). Pg. 84-90.

[9] SLAG ATLAS (1995). Edited by Verein Deutscher Eisenhüttenleute(VDEh). Germany.

[10] HERREROS, O., QUIROZ R., MANZANO, E. BOU C., VIÑALS, J. (1998). Copper extrac-

tion from reverberatory and flash furnace slags by chlorine leaching. Hydrometallurgy 49. Pp

87-101.


Leaching of Gangue in Technological Flotation

Circuits of Polish Copper Ores

Tomasz Chmielewski Andrzej Luszczkiewicz

Wroclaw University of Technology Wroclaw University of Technology

Faculty of Chemistry Laboratory of Mineral Processing

Division of Chemical Metallurgy Institute of Mining Engineering

Wybrzeze Wyspianskiego 27 Wybrzeze Wyspianskiego 27

50-307 Wroclaw, Poland 50-307 Wroclaw, Poland

Keywords: Copper ores, black shales, flotation, leaching

Abstract

The by-products flotation circuit at the copper concentrator was modified by application of an addi-

tional acidic leaching operation for this hard-to-upgrade material. The operation is based on separa-

tion of a part of the carbonate-clay material, which creates difficulties in the copper ore beneficia-

tion. The main reason for this is a very fine dissemination of valuable sulphidic minerals in the

carbonate gangue matrix. According to the presented concept, the separated byproduct is leached

with sulfuric acid in stirred tanks in order to decompose the carbonate matter and to liberate sulfide

minerals. The leached material is then returned to the flotation circuit for separation of precipitated

gypsum and for cleaning of concentrate after gypsum removal.

The role of chemical modification was evaluated by comparison of commercial flotation results

when the process was performed with and without chemical leaching of the byproducts. It was con-

firmed, both on the commercial scale and in earlier laboratory tests that in order to achieve the

highest possible flotation indices (both recovery and grade) it is necessary to decompose above

70 % of carbonates in processed middlings.

From the analysis of flotation results and mineralogical examination of solids it was found that

acidic leaching results in a high liberation of metal sulfides and leads to considerable improvement

of flotation parameters. After several months of industrial application of the chemical modification

process, a remarkable increase in both flotation recovery and concentrate grade was reported. The

evaluation of the new, chemically modified flotation process, and its effect on the flotation and at-

mospheric leaching were presented in the paper.

Chmielewski, Luszczkiewicz


1 Introduction

Copper ores mined in the LGOM (Legnica-Glogow Copper Basin, SW Poland) deposit are of sedi-

mentary origin and consists of three types of rocks of different grade and different flotation proper-

ties: sandstone and carbonates which are rather of lean grade and easy-to-upgrade and black shale

which is of higher grade but difficult to upgrade. The shale fraction from the Polish copper ores is

specific and unique in terms of lithology, geochemistry, mineralogy, chemical composition as well

as technology of processing.

The shale series in the copper ores of the Foresudetic Monocline deposit, called copper-bearing

shale (Kupferschiefer), forms from 0.3 to 1.7 m thick layers. It is a non-uniform material which

comprises clay minerals, carbonates (dolomite and calcite), organic matter and dendritic compo-

nents. Dark or black color of the shale is a result of presence of carbonaceous matter [1, 2].

The shale ore is considered as a most valuable fraction among lithological layers in terms of high

copper content and considerably elevated content of base (Co, Ni, Zn, Pb) and precious metals (Ag,

Au, PGE). Tomaszewski [3] considers the shale as a “natural polymetallic concentrate” and esti-

mated that about 25 % of total copper and from 30 to 40 % of base metals are present in the shale.

Kijewski and Jarosz [4] estimated that from 5 to 8 % of copper resources are in the form of shale.

Table 1: Composition of lithological copper ore layers and content of copper, silver, and organic

carbon in the feed of the KGHM copper concentrators in 90’s of 20th

century and in

2004 [7, 8]

Ore component Rudna Polkowice –Sieroszowice Lubin

90’s 2004 90’s 2004 90’s 2004

Carbonate ore, % 51.0 33 84.0 75 38.0 25

Shale ore, % 5.0 11 6.0 17 8.0 15

Sandstone ore, % 44.0 56 10.0 8 54.0 60

Cu content, % 2.05 2.23 1.81 2.03 1.36 1.28

Ag content, g/Mg 47 53 34 40 68 67

Corg content, % 0.64 1.49 1.14 1.66 1.76 1.62

In currently applied techniques of ore exploitation, the shale fraction, which content has been in-

creasing in recent years, is a part of a feed to copper concentrators as a mixture with sandstone and

carbonate fraction. Table 1 exhibits an average lithological composition of the copper ore, which

was used as a feed in the processing plants in the 90’s of 20th

century and in 2004. This clearly em-

phasizes that the shale fraction content in the feed of the KGHM Polish Copper S.A. processing

plants has almost doubled. A remarkable increase of the black shale content (up to 27 % for the

Lubin Mine) in the mined ore is currently observed [5].

In Table 2, results of investigations of ore samples taken from one of the mining areas of the Lubin

Mine in 2004 are set together [6]. As can be seen, when the volumetric distribution of the shale in

the ore is 15 %, the distribution of copper and silver originating from this shale in the working face

Leaching of Gangue Matter in Technological Flotation Circuit of Polish Copper Ores


is over 45 %. For organic carbon (Corg) this distribution is about 80 %. The organic carbon distribu-

tion in the ore is fully related to the shale fraction content. Organic carbon in the ore mainly origi-

nates from the shale. Therefore, under particular conditions of ore upgrading, the product enriched

in organic carbon will be of similar in character to the shale fraction.

Table 2: Lithological composition and distribution of copper and organic carbon of working face

in the Lubin Mine (region G1, Malomice I) [6]

Name

of the sample

Layers of

the thick-

ness, m

Volume-

tric dis-

tribution,

%

Cu, % Ag, g/Mg Corg , %

Content distri-

bution Content

distri-

bution,

%

Content distri-

bution

Dolomitic-clay shale 0.10 5.32 2.96 10.09 134 8.77 6.40 26.05

Clay shale 0.18 9.57 5.74 35.21 293 34.51 7.19 52.67

Boundary dolomite 0.10 5.32 0.75 2.56 42 2.75 0.73 2.97

Grey sandstone 1.50 79.79 1.02 52.14 55 53.98 0.30 18.31

Mixed ore (calculated) 1.88 100.00 1.56 100.00 81 100.00 1.31 100.00

Mixture assay 1.64 83 1.36

Chemical analyses of the ore samples of different lithological layers, collected from different min-

ing areas of KGHM Polish Copper S.A., provide a clear correlation between organic carbon (Corg)

content and metal concentration. The correlation between metal and Corg content for copper and

silver is shown in Figures 1 and 2. Figures 3 and 4 exhibit a correlations for cobalt and nickel, re-

spectively. Organic carbon concentration has been found a measure of shale content in the ore.

According to the claims of the KGHM technologists, a growing content of the shale fractions in the

flotation feed results in severe difficulties in flotation of the mixed ores. The flotation flowsheet

became very complex and unit operations very expensive, even though technological ore processing

systems have recently been considerably improved and modernized. The growing contribution of

the difficult-to-upgrade shale fraction is also a major reason for hard-to-accept metal losses in the

flotation tailings [7]. This material is considered as the most troublesome material in the existing

flotation circuit at the Lubin, Polkowice and Rudna concentrators due to the presence of clay and

organic matter and fine dissemination of sulfide minerals.



y = 0.0523x2 + 0.2947x + 0.988

R2 = 0.929

0

5

10

15

20

25

0 2 4 6 8 10 12 14 16 18

Content of organic carbon, %

Conte

nt

of

Cu,

%

.

y = 46.379x0.6661

R2 = 0.6279

0

50

100

150

200

250

300

350

400

450

500

0 2 4 6 8 10 12 14 16


Conte

nt

of

Ag ,

%

.

Figure 1: Cu vs. organic carbon content in Figure 2: Ag vs. organic carbon content in

shale ore samples from shale ore samples from

KGHM deposits. KGHM deposits.

0

50

100

150

200

250

300

350

0 1 2 3 4 5


Conte

nt

of

Ni,

g/t

.

0

50

100

150

200

250

300

0 1 2 3 4 5


Conte

nt

of

Co,

g/t

.

Figure 3: Ni vs. organic carbon content in Figure 4: Co vs. organic carbon content in

shale ore samples from various shale ore samples from various

KGHM deposits. KGHM deposits.

Considerable difficulties to liberate effectively sulphide minerals from the shale matter appear dur-

ing ore comminution. Simultaneously, a part of clay and organic components, which are easily libe-

rated from the shale, create severe difficulties in the flotation of copper sulphide minerals. The

easy-floating fines of clay minerals, liberated during grinding, impregnated with organic matter,



form slime coating on the air bubbles and sulphide minerals, hindering remarkably flotation. Un-

doubtedly, apart from very fine granulation and dissemination of sulphide minerals in the shale, this

is one the most significant reasons of low efficiency of flotation process of the feed containing the

shale fraction.

Figure 5: General balance of organic carbon in flotation circuits of Lubin Concentrator [8].

Figure 5 shows general copper and organic carbon mass balances in the flotation process at the Lu-

bin concentrator. It can be clearly found that above 70 % of organic carbon is transferred do final

tailings and the remaining 30 % becomes a concentrate component [8].

Figure 6 exhibits a simplified flowsheet for the Lubin concentrator, whereas Figure 7 illustrates a

mass balance for the 1st cleaning flotation [8]. According to these figures, almost 70 % of the organ-

ic carbon and some 30 % of copper circulate between 1st cleaning flotation and additional milling

operations. The shale by-product (middlings) from 1st cleaning flotation corresponds to about 60 %

of mass of the flotation plant feed. These data became a base of a novel concept of separation of a

part of the hard-to-upgrade shale fraction from the flotation circuit by means of individual

processing of this by-product (middlings) from 1st cleaning in a separate chemical-flotation process,

followed by oxidative leaching (atmospheric or pressure leaching) as it was shown in Figure 8 [9].

100,0

1,2 100,0

1,6 100,0

93,6 6,4

0,2 14,0 16,0 86,0

1,3 72,2 7,0 27,8

Yield, %

Cu, % Cu recovery, %

Corg, % Corg recovery,%

FINAL CONCENTRATE

LUBIN CONCENTRATOR

FINAL TAILINGS

FEED

(Run of Mine ore)



Figure 6: Simplified flowsheet of first technological line of Lubin Concentrator.

Figure 7: Flowsheet and mass balance of copper and organic carbon for 1st cleaning flotation [8]

(as in Figure 6).

C - concentrate

M - middlings

BM - ball mill

HC - hydrocyclone

final

tailings

final concentrate

HC

M

M

C

3rd cleaning

1st cleaning

1st rougher

BM

C

C

regrind

M

C

from & to

grinding circuit

M2nd cleaning

Rougher-scavenger

2nd rougher

100,0 25,7

100,0 3,8 75,7

100,0 6,0 95,0

40,2 10,3 59,8 15,4

60,3 5,7 45,7 39,7 2,5 30,1

27,0 4,0 25,7 73,0 7,3 69,3

Total mass balance Operational balance

Yield, %

Cu, % Cu recovery, %

Corg, % Corg recovery,%

Cu recovery, %

Corg recovery,%

Yield, %

1st CLEANING FLOTATION

Feed: concentrate from

rougher flotation

Concentrate

to 2nd cleaning flotation

Middlings to rougher

flotation after cycloning

and regrinding



Figure 8: General concept of separate processing of shale-enriched by-product from Lubin Con-

centrator using hydro- and bio-metallurgy [9].

2 Concept and objectives of examinations

The main purpose of the presented investigations was to analyse the concept of separation of flota-

tion by-product enriched in organic carbon (called shale product) from the flotation feed, which was

a natural mixture of three lithological fractions of the ore, and subsequent processing it by means of

chemically-assisted flotation. In general, the method of processing presented in this paper assumes

separation of the hard-to-upgrade shale fractions from the existing flotation circuits and their com-

plex hydrometallurgical processing. This concept was schematically presented in Figure 8. The im-

portant base of the presented concept is the necessity of keeping unchanged the main technological

circuit, producing concentrates for metallurgical plants.

COPPERBASE AND

NOBLE METALS

FLOTATION

TAILINGS

Technological process of copper recovery

at the KGHM Polish Copper

Shale-enriched

by-product

Non-oxidative leaching

with H2SO4 for removing

carbonate gangue

Bio- and hydrometallurgical

processes for metals recovery

Proposed and investigated process of

beneficiation of black shale fraction

SLAGS

FLOTATION

SMELTING

REFINING

CRUSHING

& GRINDING

MINING

FLOTATION GYPSUM



Figure 9: Comparison of results of standard flotation and shale flotation using various doses of

fuel oil as a non-polar collector [10].

Luszczkiewicz and co-authors [10] analysed results of flotation experiments in terms of separation

of shale fraction from a natural mixture of three ore fractions – flotation feed of Lubin concentrator.

Non-polar collector was applied for this purpose. The results of the experiments are collected in

Figure 9 as organic carbon recovery vs. copper recovery plots. The diagonal line of the diagram

represents a lack of differences in floatability of both feed components – copper and organic carbon.

Above the diagonal line and on the left side of the diagram, better organic carbon upgrading was

observed. Below the diagonal line and on the right side of the graph, copper was found to better

upgradeable. The best separation of copper from organic carbon was observed when low doses

(100 g/Mg) of collector (fuel oil) were applied. The differences in the flotation of two components

are rather small.

It is hard to evaluate the process performed on the industrial scale, but separation of the shale

enriched fractions from the rest of feed is already observed, for example in 1st cleaning flotation in

Lubin. The tailings from this process, containing about 2-3 % of Cu and evidently elevated content

of other metals, exhibit some 7-8 % of organic carbon – very similarly to the final concentrate (Ta-

ble 3). The material can be, therefore, considered as a shale concentrate and can be separated from

the flotation circuit for individual treatment, to recover all metals. Hydrometallurgical or biometal-

0

20

40

60

80

100

0 20 40 60 80 100

Recovery of copper in concentrate, Cu, %

Rec

overy

of

org

anic

carb

on i

n c

on

cen

trat

e, %

.

lack of floatability differencesStandard Xant.Fuel oil 100g/MgFuel oil 200g/MgFuel oil 400g/MgFuel oil 600g/Mg



lurgical processes of leaching were preliminary tested for this purposes. This approach can be addi-

tionally justified by mass balances presented in Figure 7.

Table 3: Chemical composition of the Lubin Concentrator shale middlings (tailings of 1st cleaning)

in comparison with a processing feed (run-of-mine ore) and final concentrate produced in

2004

Material Cu, % Ag, g/t Corg, % Ni, g/t Co, g/t Pb, % Zn, g/t Fe, % Sc, % V, g/t Mo, g/t

Lubin flotation feed,

2004 1.28 67 1.62 65 126 0.27 0.04 0.85 1.68 141 42

Tailings from 1st

cleaning 2.30 151 7.71 396 429 1.18 1.23 2.14 2.96 765 291

Lubin final concentrate,

2004 16.94 830 8.35 527 1287 3.12 0.54 5.79 11.68 630 269

The concept of separation of hard-to-upgrade shale fraction and its individual, hydrometallurgical

processing should result in significant simplification of the flotation process. Moreover, metals re-

covery in flotation and hydrometallurgical processes are expected to be remarkably higher (cobalt is

not recovered) with regard to existing systems.

The shale product from the Lubin Concentrator contains carbonate matter and has to be initially

subjected to non-oxidative leaching with sulphuric acid in order to totally decompose acid-

consuming components prior to the atmospheric or pressure leaching. In this paper, additional flo-

tation of leached by-product for separation of gypsum, which is being formed in leaching, was con-

sidered. This process is already applied on the industrial scale at KGHM [11].

3 Non-oxidative leaching

Non-oxidative leaching of the shale-enriched by-product from the commercial circuits was used as

a unit operation which aim was to liberate metal-bearing minerals from carbonate particles. This

kind of leaching consists of selective (without breaking sulphides) chemical calcium and magne-

sium carbonates decomposition by means of sulphuric acid, according to the reactions:

CaCO3 + H2SO4 + H2O = CaSO4⋅2H2O↓ + CO2↑ (1)

MgCO3 + H2SO4 = MgSO4 + CO2↑ + H2O (2)

Hydrated calcium sulfate (gypsum) precipitates as a solid reaction product, whereas water-soluble

magnesium sulfate and gaseous carbon dioxide are other reaction products. Since particles of mid-

dlings are fine, leaching of the carbonate gangue with H2SO4 is very rapid and can be performed at

ambient temperatures in reactors with mechanical stirring of a simple construction.

The amount of H2SO4 applied in the non-oxidative leaching operation directly corresponds to the

content of carbonates and has to be precisely controlled to maintain the final pH of the pulp at a

level enabling its direct transfer either to the flotation circuit without a need for pH correction or to

the leaching and bioleaching. Therefore, for further flotation, the amount of sulfuric acid introduced



to the leaching operation should always be kept below the analytically determined maximum

amount of acid required for the total carbonates decomposition.

0,5

1,5

2,5

3,5

4,5

5,5

6,5

7,5

0 10 20 30 40 50 60 70 80 90

Leaching time, min.

pH 20%

40%

50%

70%

90%

100%

1,0

1,5

2,0

2,5

3,0

3,5

4,0

4,5

5,0

5,5

6,0

0 1 2 3 4 5 6 7 8 9 10

Leaching time, min.p

H

20%

40%

50%

70%

90%

100%

Figure 10: pH vs. leaching time plots for Figure 11: pH vs. leaching time plots for

non-oxidative leaching of Lubin initial stage of non-oxidative

shale middlings. leaching of Lubin shale

middlings.

Kinetics of non-oxidative leaching of the middlings from the Lubin Concentrator was investigated

for different degrees of carbonates decomposition varying from 20 to 100 % (Figures 10 and 11).

The key parameter determining the amount of H2SO4 required for carbonates leaching is the maxi-

mum demand for acid (max

SOH 42z ), which is the mass of pure H2SO4 necessary for a total decomposi-

tion of carbonates in 1 kg of dry solid feed. The max

SOH 42z parameter should be determined analytically

from laboratory tests. H2SO4 is introduced to the reactor containing the shale slurry at a rate that

either assures its total utilization or only a selected grade. The maximum demand for sulphuric acid

for the examined Lubin middlings was 494 g H2SO4/kg of dry material.

The process control of non-oxidative leaching is based on pH measurement of the leached mid-

dlings suspension after introduction of required amount of acid. On the basis of kinetic results for

the Lubin middlings it was found that the process is very rapid (Figures 10 and 11) and after about

5 minutes almost entire amount of acid is already used up. The observed further pH changes (up to

about 40-60 minutes) correspond only to the saturation of the slurry with CO2.

Non-oxidative leaching is not only very fast but also selective. It does not cause chemical decompo-

sition of metal sulphides under non-oxidative conditions created by the carbon dioxide.



The samples of middlings being the tailings from the 1st cleaning commercial operation at the Lubin

Concentrator before and after non-oxidative leaching with sulfuric acid were the material for expe-

riments. The total time of carbonate decomposition in the examined samples was 60 minutes.

4 Flotation of the shale middlings after non-oxidative leaching

Flotation feeds with either 50, 70 or 90 % of carbonate decomposition were used in experiments

performed according to the flow-sheet shown in Figure 12. For comparison, the flotation of raw

middling (i.e. non-leached with H2SO4) was also carried out. For the flotation experiments a Mek-

hanobr sub-aeration type laboratory flotation machine equipped with a 1 dm3 cell was applied. pH

measured during the flotation for the non-leached material was from 7.6 to 8.3, and for the leached

samples from 5.5 to 7.0.

Figure 12: Scheme of flotation experiments.

In the standard xanthate flotation of sulfides, 50-60 g/Mg of collector (mixture of sodium

ethyl + amyl xanthate, 1:1) and 10-20 g/Mg of frother (Corflot) were used. The rougher flotation

was conducted up to the moment when the flotation froth did not contain useful minerals. The ob-

tained concentrate was subsequently subjected to the cleaning flotation where the following prod-

ucts of flotation were collected at suitable time intervals. The procedures of the flotation experi-

ments were the same, but quantities of the reagents and times intervals of the products collection

were different for various examined feeds.

Due to strong frothing properties of the suspension after leaching, the experiments were carried out

with a minimal air-flow. This caused low yield of the products and resulted in higher selectivity of

flotation.

FEED

CLEANING FLOTATION

K1 K2 ... Middlings

(cleaning tail)

Sodium ethyl+amyl xanthate (1:1) + frother

Tailing

( Sulfide concentrates)

ROUGHER FLOTATION



Figure 13: Copper upgrading curves for various Figure 14: Copper upgrading curves –

degree of carbonate decomposition. enlarged part of Figure 13.

Figure 15: Silver upgrading curves for various Figure 16: Upgrading curves for analyzed

degree of carbonate decomposition. components for 70 % of

carbonate decomposition.

Shale middlings samples were taken from the Lubin Concentrator commercial circuit and initially

subjected to determination of the maximum consumption of sulphuric acid for total decomposition

of carbonates [11]. Subsequently, non-oxidative leaching with controlled grade of carbonates de-

composition was conducted followed by laboratory flotation tests. Figures 13 to 16 exhibit the re-

sults of flotation experiments in the form of Fuerstenau curves, drawn according to procedure de-

scribed by Drzymala and Ahmed [12].

0

20

40

60

80

100

0 20 40 60 80 100

Recovery of copper in concentrate, %

Reco

ver

y o

f b

arr

en p

art

in t

ail

ing

, %

.

Lack of upgradingStandard flotation

50% of decomposition70% of decomposition90% of decomposition

Cu

0

10

20

30

40

50

50 60 70 80 90 100

Recovery of copper in concentrate, %

Rec

ov

ery

of

barr

en p

art

of

tail

ing

, %

.

Lack of upgradingStandard flotat ion50% of decomposition70% of decomposition90% of decomposition

Cu

0

10

20

30

40

50

50 60 70 80 90 100

Recovery of silver in concentrate, %

Rec

ov

ery

of

barr

en p

art

of

tail

ing

, %

.

Lack of upgradingStandard flotation50% of decomposit ion70% of decomposit ion90% of decomposit ion

Ag

0

10

20

30

40

50

50 60 70 80 90 100

Recovery of component in concentrate, %

Reco

very

of

barr

en p

art

of

tail

ing

, %

.

Lack of upgradingCuAgCoZnFeCorg

70% of carbonate

decomposition



Figure 13 shows the floatability curves for copper at various carbonate decomposition grades (50,

70 and 90 %), whereas Figure 14 presents an enlarged part of the relationship for more adequate

evaluation. The floatability curves for silver are given in Figure 15 and Figure 16 shows the effect

of 70 % carbonate decomposition on the flotation results of Cu, Ag, Co, Zn, Fe and organic carbon.

It is well seen that the higher is the degree of carbonates decomposition, the better is floatability of

valuable components. From Figure 16 we can find that upgradeability of copper and organic carbon

exhibit the highest values with lowest for iron. Silver, cobalt and zinc are found to be similarly

floatable after chemical decomposition of 70 % of carbonates from the feed.

5 Atmospheric leaching in oxygenated sulphuric acid

The shale by-product from the Lubin Concentrator was subjected to atmospheric leaching after ini-

tial total non-oxidative decomposition of carbonates. Leaching was performed at various tempera-

tures (25, 60, 80 and 90 oC) and various concentration of iron(III) (7, 15 and 30 g/l). The effect of

sulphuric acid was also examined at various temperatures (25-90 oC) at a solid/liquid ratio from 1:5

to 1:4 and at the oxygen flow rate of 30 l/h.

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

0 60 120 180 240 300

Leaching tim e, m in.

Cu

co

nce

ntr

atio

n, g

/l .

25oC

60oC

80oC

90oC

0,0

1,0

2,0

3,0

4,0

5,0

6,0

7,0

8,0

9,0

10,0

0 60 120 180 240 300

Leaching time, m in.

Cu

co

nce

ntr

atio

n, g

/l .

0

30 g/l

7 g/l

15 g/l

Figure 17: The effect of temperature on the Figure 18: The effect of Fe(III) on the

atmospheric leaching of copper atmospheric leaching of copper

from Lubin middlings in from Lubin shale middlings.

oxygenated sulphuric acid.

A minor effect of concentration of sulphuric acid was observed for metals leaching in preliminary

tests. Therefore, the H2SO4 concentration was kept at the level of 50 g/dm3 for all leaching experi-

ments. Concentration of Cu, Fe, Ni, Co and As was analysed in the pregnant leaching solutions vs.

leaching time. Ag and Pb content was determined in solid residue samples.



Solubility of metals being in the sulphidic form requires the presence of oxidation agent: gaseous

oxygen or/and iron(III). Copper, which is present in the feed predominantly as chalcocite (Cu2S)

and bornite (Cu5FeS4) can be leached very rapidly by oxygenated H2SO4 solution and the leaching

rate increases remarkably in the presence of iron(III).

Selected kinetic curves for atmospheric leaching are presented in Figures 17 to 20 for copper, cobalt

and nickel. Both temperature and concentration of iron(III) appeared to be a key parameters in at-

mospheric leaching. It was determined that at 90 oC and at s/l ratio of 1:4 it was possible to reach

concentration of copper of about 9 g/l after 5 hours leaching. Immense effect of Fe(III) was also

observed for leaching cobalt and nickel.

0

10

20

30

40

50

60

70

80

90

100

110

0 60 120 180 240 300

Leaching time, min.

Co

co

nce

ntr

atio

n, m

g/l .

0

7

15

30

0

5

10

15

20

25

30

0 60 120 180 240 300

Leaching time, min.

Ni c

on

centr

atio

n,

mg

/l

.

0

7

15

30

Figure 19: The effect of Fe(III) on the Figure 20: The effect of Fe(III) on the

atmospheric leaching of cobalt atmospheric leaching of nickel

from Lubin middlings. from Lubin middlings.

Particle size analyses for the Lubin middlings indicated that the applied middlings were rather to

coarse in terms of leaching rate. Parameter d80 exceeded 100 µm even after non-oxidative leaching

with H2SO4 (carbonates decomposition Rw = 30-90 %). Additional microscopic SEM (Figure 21)

observations lead to a significant conclusion, that the most coarse particles are formed by the shale

material, which can not be chemically decomposed during treatment with acid, even in the presence

of oxygen (Figure 21D). Therefore, separation of +50/60 µm particle size fraction and its additional

regrinding seems to be very desirable to facilitate further leaching.



Figure 21: SEM mineralogical analyses of Lubin shale middlings before leaching (A, B, C) and

after atmospheric leaching in acidic consitions (D).

6 Conclusions

1. Shale middlings, being the tailings from the 1st cleaning flotation operation at the Lubin Con-

centrator, is recognized as the most troublesome product in the existing flotation circuit due to a

high amount of the difficult-to-treat carbonate and organic fraction. Fine dissemination of sul-

fide minerals in gangue minerals is the main reason of difficulties in upgrading of the shale frac-

tion. It was the major reason for choosing the middlings as the material for alternative process-

ing by means of non-oxidative and atmospheric leaching.

2. Examined middlings can be considered as a ready-to-use shale concentrate. This is due to the

high content of organic carbon (from 5 to 10 %). Moreover, the content of Cu (2-3 %) is close

to the copper concentration in a natural shale. Recovery of organic carbon, directly correlated

with quantity of shale, indicates, that most of the shale fraction is in this by-product. Shale mid-

dlings should be the principal feed for the bio- and hydro-metallurgical processes.

D C

B A



3. Flotation experiments demonstrated that upgrading the middlings was nearly impossible without

a chemical pretreatment. Non-oxidative leaching of Lubin middlings with H2SO4 appeared to be

useful chemical operation for selective liberation of sulphide minerals prior to flotation.

4. Partial decomposition of carbonates from 50 to 90 % remarkably improves the liberation degree

of the valuable minerals from the hydrophilic intergrowths with carbonate matter.

5. Evidently improved flotation of Cu, Ag, Zn, Co, Fe and Corg occurs for leached materials in

comparison with direct flotation of untreated flotation feed.

6. An increase of the carbonates decomposition grade of the feed results in enhanced recovery and

content of the useful components in the concentrate. The best results of the flotation parameters

(recovery and concentrate grade) were obtained for the feed with the 90 % of carbonate decom-

position.

7. Atmospheric leaching of shale middlings after total carbonate decomposition can be effectively

used as alternative process for recovering of metal values from this troublesome feed

References

[1] Konstantynowicz-Zielińska J. 1990, Petrography and genesis of copper-bearing shales from

Foresudetic Monocline. Rudy i Metale Nieżelazne. R.35, Nr 5-6, 128-133, in Polish

[2] Rydzewski A., 1996, Lithology of the deposit rocks. In: Monografia KGHM Polska Miedź

S.A., A. Piestrzynski (Ed). Publ. CPBM “Cuprum” Sp. z O.O., Lubin, 137-141, in Polish

[3] Tomaszewski J., 1985, Problems of a rational utilization of copper-polymetallic ores from the

Foresudetic Monocline deposits. Physicochemical Problems of Mineral Processing, Nr 17,

131-141, in Polish

[4] Kijewski P., Jarosz J., Ore mineralization and form of occurrence of accompanying elements

in the copper deposit. Proceedings of the Conference: “Accompanying metals in the copper

ore deposit, the state and prospects for further use", Rydzyna, May 1987, P. Kijewski (Ed),

Publ. NOT/SITG/Cuprum, Wroclaw 1987, 21-48, in Polish

[5] Kubacz N., Skorupska B., Evaluation of influence of organic carbon on concentration and

smelting. processes. Proc. VIII International Conference on Non-ferrous Ore Processing,

Wojcieszyce (Poland), May 21-23, KGHM Cuprum Wroclaw 2007, 157-166, in Polish

[6] Luszczkiewicz A., Evaluation of upgradebility of the ore with elevated content of black shale.

Report of Investigations, Archive of Laboratory of Mineral Processing, Wroclaw University

of Technology, Wroclaw, October 2004, in Polish



[7] Luszczkiewicza A., Beneficiation of copper black shale from ores of Lubin-Glogow region.

Conference Proceedings on: Current Problems of Mineral Processing of Copper Ore in Pol-

and, Polkowice, November 16, 2000, Publ. Committee of Mining Polish Academy of

Sciences and KGHM Polish Copper, 137-156, in Polish

[8] Luszczkiewicz A., Separation of black shale fraction from the polish copper deposit and a

general concept of change of technology at the Lubin Concentrator. Presentation at Bioshale

Final Meeting, Orleans, November 27-29, 2007

[9] Chmielewski T., Luszczkiewicz A., Konopacka Z., Separation and concept of processing of

black shale copper ore from Lubin mine, Proc. VIII International Conference on Non-ferrous

Ore Processing, Wojcieszyce (Poland), May 21-23, KGHM Cuprum Wroclaw 2007, 171-184,

in Polish

[10] Luszczkiewicz A.., Konopacka Z., Drzymala J., Flotation of black shale of Polish copper ore

from Lubin. Proceedings: Perspectives for applying bioleaching technology to process shale-

bearing copper ores, BIOPPROCOP ’06, Lubin, June 19, 2006, Publ. KGHM Cuprum Sp. z

O.O., Wroclaw 2006, 29-47, in Polish

[11] Luszczkiewicz A., Chmielewski T., Acid treatment of copper sulfide middlings and rougher

concentrates in the flotation circuit of carbonate ores. International Journal of Mineral

Processing, Vol. 88, Issue: 1-2, 2008, 45-52

[12] Drzymala J, Ahmed H.A.M., Mathematical equations for approximation of separation results

using the Fuerstenau upgrading curves. International Journal of Mineral Processing, Vol. 76,

Issue: 1-2, 2005, 55-65



Pressure Leaching of Shale Middlings from Lubin

Concentrator in Oxygenated Sulphuric Acid

Tomasz Chmielewski, Jerzy Wódka

Wroclaw University of Technology

Faculty of Chemistry, Division of Chemical Metallurgy

Wybrzeze Wyspianskiego 27

50-307 Wroclaw, Poland

Keywords: Pressure leaching, shale ore, copper

Abstract

The effect of initial temperature, concentration of sulphuric acid, and oxygen partial pressure on

acidic pressure leaching of shale middlings-tailings from 1st cleaning flotation of Lubin Concentra-

tor (ZWR Lubin, Poland) - have been investigated. Lubin middlings were selected for hydrometal-

lurgical treatment as a troublesome shale material exhibiting elevated content of metals (Cu, Co, Ni,

Zn, As, Ag, Pb) and of organic carbon. Leaching was performed in a stirred autoclave at tempera-

tures within the range of 100-180 oC. Oxygen at partial pressure from 2.5 to 10.0 at was used as an

oxidizing agent. Leaching process was performed at initial sulphuric acid concentration from 20 to

50 g/dm3. Pressure leaching appeared to be very efficient process for recovering of Cu, Co, Fe, Zn

and As from examined shale middlings. Leaching recovery of Ni was remarkably lower. Ag and Pb

remained in the solid residue and will be recovered in separate processes.

1 Introduction

Pressure hydrometallurgy has an extensive and growing application in processing of zinc [1, 2],

nickel – in particular from laterite ores [3-5], copper sulphides [6-12] and in pretreatment of refrac-

tory gold ores [13-14], in which gold is finely disseminated in a lattice of sulphidic minerals. Re-

cently, numerous investigations are undertaken to apply pressure hydrometallurgy to process bypro-

ducts and wastes of copper refining (anode slimes, non-ferrous smelter slags) [16] and for recovery

of metals from ores and raw materials of the specific properties and composition, which cause diffi-

culties in their treatment by standard methods (black shale ores, tailings). Pressure hydrometallurgy

exhibits numerous advantages, creating intensive investigation of it at laboratory and industrial

scale. The main advantages of application of pressure leaching processes for sulphidic, polymetallic

raw materials are:

Chmielewski, Wodka


- high rate of leaching reactions,

- elimination of SO2, other gases and dust emission,

- high selectivity of pressure leaching,

- possibility of arsenic utilization or stabilization (as a low – soluble scorodite),

- total recovery of base and noble metals,

- no restriction of the scale production.

Pressure leaching can be applied in a large scale – to treat millions ton per year of refractory gold

ores, or to produce several tons of nickel and cobalt from their sulphidic or lateritic resources. The

methods of hydrometallurgical treatment of copper sulphide concentrates, as an alternative to tradi-

tional smelter-refinery processes, were compared and discussed in details [17]. Pressure leaching

was recently successfully applied in industrial scale by Phelps Dodge (Freeport McMoran Copper &

Gold) for production of copper (80 kt/y) from copper sulphide concentrates at Morenci plant [18]. It

is the first in the World full-scale application of pressure leaching process for copper sulphide con-

centrates.

Sedimentary, sulphidic copper ores from Polish deposits (LGOM - Legnica - Glogow Basin, SW

Poland) exhibit the elevated content of black shale fraction which has been systematically increas-

ing, particularly in recent years [19]. The black shale ores reveal unique both advantageous and de-

trimental properties. They contain evidently more copper, base metals and noble metals than sand-

stone or carbonate fractions. However, the elevated carbonate and organic coal content as well as

dissemination of the metals-bearing minerals occurring in the black shale ore create significant dif-

ficulties in the flotation and cause remarkable and unacceptable metals losses, which has to be li-

mited for technical and economical reasons [20]. Therefore, hydrometallurgy has been declared as a

major strategy objective for coming years at KGHM.

During the comprehensive investigations on application of bio- and hydrometallurgy for alternative

processing of black shale fraction from Lubin Mine, the non-oxidative leaching, atmospheric leach-

ing in oxygenated sulphuric acid, acidic leaching under oxygen pressure and bioleaching have been

considered as optional processes for recovering of copper and base metals from shale by-product

(middlings) from Lubin Concentrator [21]. Middlings-tailings of 1st cleaning were separated from

the industrial flotation circuit as a feed for further alternative, hydrometallurgical treatment.

Tailings from the first cleaning flotation (middlings) of the first circuit at Lubin Concentrator (ZWR

Lubin) were primarily selected for the laboratory investigations as a shale concentrate. This by-

product can not be effectively upgraded by flotation and creates serious decline of flotation indices,

both recovery and concentrate grade, primarily at Lubin Concentrator. It appeared, that the separa-

tion of the middlings from Lubin flotation circuit and further processing by means of leaching pur-

poses is quite simple. Moreover, the middlings have an enormously advantageous composition with

regard to both atmospheric and pressure leaching. Bornite and chalcocite, the easiest leachable cop-

Pressure Leaching of Shale Middlings in Oxygenated Sulphuric Acid


per sulphides, are found to be dominating copper-bearing minerals in this by-product. Moreover, the

solid contains up to 9 % of organic carbon and some 30 % of carbonates that must be decomposed

by sulphuric acid prior to pressure leaching.

Pressure leaching in oxygenated H2SO4 solution, at elevated temperatures, can be considered as a

mostly recommended way of copper manufacturing. Sulphuric acid is produced at KGHM smelters

(about 650 kt) as a by-product or rather troublesome waste during the processing of copper sulphide

concentrate and is considered as the most suitable, cheap and easy-accessible leaching agent.

The effect of initial temperature, sulphuric acid concentration and oxygen partial pressure on pres-

sure leaching of shale fraction of copper ore has been investigated to evaluate the leaching ability of

the shale fraction (middlings) separated as tailing of 1st cleaning from Lubin Concentrator. Pressure

leaching examinations were performed in the temperature range of 100-180 oC using oxygen as an

oxidizing agent.

2 Experimental

2.1 Lubin middlings characterization

Selection of black shale feed and experimental determination of conditions for effective recovering

of copper and other valuable metals (Fe, Ni, Co, Pb, Au, PGM…) from shale-containing materials

(ore, middlings, shale concentrates) was the general task at the starting point of the BIOSHALE

research project, co-financed by European Commission within the range of VII Frame Pro-

gramme [22]. The research program was initially focused on the geological shale samples collected

for preliminary laboratory examinations (chemical and mineralogical analyses, non-oxidative, at-

mospheric and acid pressure leaching). After the extended period of experimental work, the shale

middlings-tailings from the 1st cleaning flotation at Lubin Concentrator (ZWR Lubin) were finally

selected and accepted by a technical, managing, and engineering staff at KGHM as the only shale

containing solid for hydrometallurgical laboratory tests and for prospective full scale alternative

processing. Simultaneously, this material was unanimously recognized as the most troublesome

solid in existing flotation circuit at Lubin Concentrator due to the high content of shale fraction

(clay and organic matter) and fine dissemination of sulfide minerals. It was fully considered as a

shale concentrate, which can be separated from the flotation circuit for hydrometallurgical treat-

ment. Initial samples were collected of numerous shale ore fractions from various LGOM (Poland)

deposits.

Chemical analyses exhibited varying metal content in the shale. For further laboratory studies the

process alternatives were selected including chemical pretreatment with H2SO4 for flotation, non-

oxidative leaching for carbonates decomposition, atmospheric leaching and pressure leaching inves-

tigations as BIOSHALE project WP4 tasks. The laboratory tests undoubtedly confirmed that the

assumptions made in the first stage of the work program were correct.

Chmielewski, Wodka


Acidic leaching of shale middlings appeared to be very effective from mineralogical and technical

viewpoint. Favorable mineralogical composition of Lubin middlings (dominating content of copper

in the form of chalcocite and bornite) and easy access to sulfuric acid, being – in fact – a waste ma-

terial from copper smelters, makes this approach very attractive for future, necessary technological

alterations [23-27].

One of the main reasons, that black shale fraction became a serious problem in flotation of Polish

copper ores is a very fine mineralization of metal-bearing sulfides and their dissemination in hydro-

philic carbonate and in organic matter. It is impossible to liberate these minerals only by means of

grinding. Consequently, a great amount of metal values remain in flotation tailings and leads to

hardly accepted metals loss. Very complex and expensive alterations in flotation and grinding cir-

cuits at Lubin Concentrator appeared to be ineffective in terms of flotation indices. Both the copper

and silver recovery and concentrate grade have been reported to decrease remarkably.

Table 1: Content of shale fraction (in percent) in Polish copper ores mined from various LGOM

(Legnica – Glogow Copper Basin) deposits [19].

Mine - Deposit Year

2002 2004 2006

Lubin 18.8 15.0 27.0

Polkowice-Sieroszowice 11.3 13.3 11.2

Rudna 2.2 3.9 4.0

An additional unfavorable effect of increasing shale content has been lately observed during the

flash smelting at Głogów II smelter. The growing content of shale fraction (Table 1) and the in-

crease in organic carbon content – exceeding 9 % for Lubin and Polkowice – in the concentrate

smelter feed, results in a notable diminishing of the process efficiency [19]. Therefore, separation of

shale fraction from the flotation circuits (e.g. Lubin middlings) and its individual, hydrometallurgic-

al processing, postulated within the frame of BIOSHALE project, can be accepted as a key alterna-

tive for existing technologies. Such alteration can result in the following advantages:

• Integration of new technology with existing processes.

• Simplification of existing flotation circuits and enhancement of flotation efficiency after separa-

tion of shale fraction.

• Increase of metals recovery in total (flotation concentrate + hydrometallurgy) – remarkable de-

crease of metals loses and environmental impact.

• Stabilization of organic carbon content in flotation concentrates on the proper level, accepted for

flash smelting process.

• Effective utilization of an excess of sulfuric acid – a by-product or waste from smelting.

It seems to be rather obvious that due to the observed apparent decline of copper ores quality and

upgradeability, particularly from Lubin and Polkowice deposits (Table 1), present beneficiation



technologies are not able to guarantee the acceptable metals recovery and concentrate grade without

significant technology alteration and urgent introduction of new alternative concepts. This appeared

to be urgent for Polish copper industry. Acidic pressure leaching can be considered as an alternative

process.

Shale middlings from Lubin Concentrator were collected and characterized in details in D.4.2. Deli-

verables within BIOSHALE European project, on the basis of 1st sampling campaign in 2006 and

2nd

sampling campaign in 2007. Middlings from 1st and 2

nd sampling campaigns were used as a feed

for pressure leaching. It is well seen from Table 2, that chemical composition of material examined

in 2007 was very similar to those used in previous investigations. Copper content was about 2.60 %,

organic carbon was 6.30 % – lower than in 2006 campaign. The content of accompanying metals

was observed to be almost identical, except of zinc. The comparison of chemical analyses of Lubin

middlings samples from 2006 and 2007 evidently shows, that examined feed is fairly stable in terms

of its chemical composition.

Table 2: Chemical characterization of Lubin shale middlings

(1st and 2

nd sampling campaigns – 2006 and 2007).

CONTENT

Cu, % Fe, % Ni, g/t Co, g/t Pb, % As, %

1st campaign 2.72 1.76 374 572 1.51 0.090

2nd

campaign 2.60 1.89 328 613 - 0.085

CONTENT

Ag, g/t Zn, g/t Sc, % Sso4, % Ctotal, % Corg, %

1st campaign 190 1200 2.95 1.45 14.30 8.96

2nd

campaign 168 740 2.62 2.47 10.4 6.30

Figure 1 exhibits the mineralogical composition of Lubin middlings and compares it with composi-

tion of Lubin final concentrate, currently produced as a feed for smelting. This analysis was made

by BRGM (Orleans, France) laboratories and became a very important justification in discussion on

considered technology alterations. The mineralogical composition of flotation concentrate appeared

to be very similar to the composition of the ore. Chalcocite (Cu2S) and bornite (Cu5FeS4), easiest to

leach copper sulfides, are dominating in the material. The content of chalcopyrite (CuFeS2), most

refractory copper sulfide, was only about 20.6 %.

In contrary to concentrate, mineralogical composition of shale middlings appeared to be enormously

favorable for hydrometallurgical treatment. Chalcocite was reported to be dominating (90.5 %) with

some content of bornite (9 %) (Figure 1). Chalcopyrite is a negligible component in this by-product.

Such a mineralogical composition can be considered as an ideal for hydrometallurgical or biometal-

lurgical treatment. The observed mineralogical composition of shale middlings exhibits evidently,

Chmielewski, Wodka


that beneficial mineralogical segregation of minerals takes place during the 1st cleaning flotation

process at Lubin Concentrator.

Figure 1: Mineralogical composition of Lubin final concentrate and Lubin shale middlings

(BRGM data).

It is very obvious that shale fraction has to be processed not only as a copper-bearing raw material

but mainly as a polymetallic one. Separation of this fraction from flotation circuit and its processing

for recovering of Cu and other metals is therefore fully justified.

Particle size analyses for Lubin middlings (Figure 2) indicate that applied middlings was rather to

coarse in terms of leaching rate. Parameter d80 exceeded 100 µm even after non-oxidative leaching

with H2SO4 (carbonates decomposition, Rw = 30-90 %).

Figure 2: Particle size analyses of Lubin middlings raw and subjected to the non-oxidative acidic

leaching at various degree of carbonate decomposition, Rw (within 30-90 %).

20

30

40

50

60

70

80

90

100

10 100 1000

Particle size, µm

Cum

ula

tive r

ecove

ry, %

.

0

Rw - 30%

Rw - 50%

Rw - 70%

Rw - 90%



Additional microscopic SEM (Figure 3) observations lead to the significant conclusion, that most

coarse particles are formed by shale material, which can not be chemically decomposed during

treatment with acid, even in the presence of oxygen (Figure 3D). Therefore, separation of particle

fraction above 50-60 µm and its additional regrinding seem to be very desirable to facilitate further

leaching.

Figure 3: SEM mineralogical analyses of Lubin shale middlings before leaching (A, B, C) and

after atmospheric leaching in acidic conditions (D).

2.2 Pressure leaching procedure

The effect of temperature, oxygen partial pressure and sulphuric acid concentration on the kinetics

and efficiency of pressure leaching of Lubin middlings have been investigated. All experiments

were carried out at temperatures between 100 and 180 oC and under oxygen pressure within the

range of 2.5-10 atm. The pressure leaching was always preceded with non-oxidative acidic pre-

treatment of the feed in order to totally decompose the acid-consuming componets, predominantly

carbonates.

A B

D C

Chmielewski, Wodka


The experiments were performed in 2.0 dm3 autoclave having the stirrer and sampling pipe made of

teflon. The following experimental procedure was used: A teflon beaker containing 1.0 dm3 of solu-

tion of the required concentration of sulphuric acid and the required amount of the middlings was

introduced into the autoclave. The liquid to solid phase ratio (s/l) = 1:10 was kept in all experi-

ments. The stirring was switched on when the non-oxidative decomposition of carbonates with

H2SO4 started prior to the oxidative leaching. After carbonates decomposition (ca. 60 min.) auto-

clave lid was installed and the slurry was purged two times with nitrogen. The heating was switched

on and as the temperature rose to 100 oC the nitrogen was removed from the autoclave.

After the temperature reached the require level the solution “zero sample” was drawn off for chemi-

cal analyses. Subsequently, the mixture of 50 % of oxygen and 50 % of nitrogen was introduced into

the autoclave to establish the required oxygen partial pressure. During the experiments samples of

the solution were taken periodically to determine metals concentration (Cu, Zn, Ni, Co, As) by

AAS.

Solid samples of middlings before and after each pressure leaching test were examined by minera-

logical SEM microscopy to evaluate the composition of the solid and to assess the effectiveness of

applied leaching parameters range.


3.1 Effect of temperature on pressure leaching of the shale middlings

Pressure leaching of Lubin middlings was examined at elevated temperatures from 100 to 180 oC

while concentrations of Cu, Fe, Ni, Co and As were analysed in the liquid samples taken during the

leaching. Experimental results are shown in Figures 4 for Cu, Fe, and Co, respectively. The ob-

served effect of temperature on the leaching of metals was quite complex and the following obser-

vations have been done:

• Leaching of copper in oxygenated sulphuric acid at 100 oC required about 120 min. of activation

time. This was most likely due to the formation of H2S observed at the initial stage of the process.

Remarkable acceleration of the process was subsequently detected after 120 min. of activation.

• Unexpectedly, at 120 oC no copper leaching was detected. This might be most likely explained as a

hindering effect produced by the presence of elemental sulphur at its melting point.

• At temperatures exceeding 140 oC the leaching appeared to be very rapid, although some decrease

of Cu concentration was observed at 160 and 180 oC versus 140

oC. Co-precipitation of iron com-

pounds or sorption on organic matter are the only explanation of observed effects.

• The rate of leaching of Fe from Lubin middlings increases with temperature up to 120 oC, then iron

concentration slightly drops as the result of precipitation of iron(III) oxide (Fe2O3) or goethite

(FeOOH). The appearance of soluble iron in leaching solution results from leaching of Cu-Fe sul-

phides, mainly bornite – Cu5FeS4 and chalcopyrite – CuFeS2.



0

500

1000

1500

2000

2500

3000

0 30 60 90 120 150 180 210

Leaching time, min

Cu

conce

ntr

atio

n,

mg

/dm3

.

100

120

140

160

180

Cu

800

1000

1200

1400

1600

1800

2000

2200

2400

2600

0 30 60 90 120 150 180 210

Leaching time, minF

e co

nce

ntr

atio

n,

mg/d

m3 .

100

120

140

160

180

Fe

0

10

20

30

40

50

0 30 60 90 120 150 180 210

Leaching time, min

Co

co

ncen

trati

on

, m

g/d

m3

100

120

140

160

180

Co

Figure 4: Effect of temperature on the pressure

leaching of copper, iron and cobalt

from Lubin middlings.

Oxygen partial pressure: 0.5 MPa,

concentration of H2SO4: 5 %,

solid to liquid ratio, s/l: 1:10.

Pressure leaching of cobalt was practically not observed at temperatures 100 and 120 oC with detected

concentration of cobalt about 5 mg/dm3. Remarkable acceleration of leaching of Co was recorded at

temperatures above 140 oC and reported concentration of cobalt increased to 40-50 mg/dm

3.

Quite complex kinetic curves for nickel (not presented here) indicated that at 120 oC the highest

leaching recovery was observed with the final Ni concentration of about 180 mg /dm3. At tempera-

tures exceeding 140 oC the leaching rate of nickel decreases, which are hard to explain at this stage

Chmielewski, Wodka


of investigation and require detailed analysis of leaching residues, particularly determination of the

mineralogical forms of metals and gangue.

The concentration vs. leaching time relationship for leaching of arsenic was similar to those of co-

balt. Pressure leaching of arsenic was nearly not observed at temperatures 100 and 120 oC when

detected concentration of As was about 10 mg/dm3. Significant increase in leaching rate of As was

observed at temperatures above 140 oC while concentration of arsenic increased to 70-90 mg/dm

3.

From results of pressure leaching it is well seen, that very effective recovery of metals from Lubin

middlings can be observed for experiments performed at temperatures exceeding 140 oC. At 140

oC

observed were the highest Cu, Co and As recoveries. It was also detected that at temperatures ex-

ceeding 180 oC iron control became possible as a result of precipitation of FeOOH or Fe2O3. Arsen-

ic, a harmful contaminant, will require a removal or stabilisation prior to recovering process of cop-

per, nickel and cobalt. Solvent extraction or precipitation of crystalline scorodite – FeAsO4·2H2O at

temperatures above 180 oC can be taken into consideration.

The relationship presented in Figure 4 are characterized by rapid copper, iron and cobalt leaching

during the initial 60 minutes. Subsequently, a slower increase in metals concentration is observed.

The first, 30 to 60 minutes, very rapid step of leaching, most likely due to initial fast leaching of the

easiest leachable copper minerals – bornite and chalcocite, dominating in the shale fraction from

Lubin Concentrator. The second, slower step, is possibly due to leaching of chalcopyrite and

covellite the most refractory copper minerals.

The observed concentration of iron(II) ions in the solution before introduction of oxygen is about

1.0 g/dm3. It suggests that iron(III) partly occurred as an oxide in the shale dissolves in acidic condi-

tions. Maximum recovery of copper (97 %) and cobalt (80 %) was observed at 140 oC and above

90 % for Fe at 160 oC. However, in the investigated range of temperatures nickel did not leach well.

Nickel leaching recovery rarely exceeded 30-40 % in most experiments. Neither lead no silver was

apparently detected in the solution.

3.2 Effect of initial sulphuric acid concentration on Cu, Fe, and Co

leaching

The effect of sulphuric acid concentration on the metals leaching was investigated in the range of 20

to 50 g/dm3 at temperature of 140

oC and oxygen partial pressure 5.0 MPa. Figure 4 presents con-

centration – leaching time relationships for copper and cobalt at different sulphuric acid concentra-

tions. Copper doesn’t leach before introduction of oxygen and course of copper leaching resembles

those observed during investigations of the effect of temperature.



0

500

1000

1500

2000

2500

3000

0 30 60 90 120 150 180 210

Leaching time, min. .

Co

pp

er c

on

cen

trat

ion

, m

g/d

m3 .

2%

3%

4%

5%

0

10

20

30

40

50

60

0 30 60 90 120 150 180 210

Leaching time, min.

Cobal

t co

nce

ntr

atio

n,

mg/d

m3

.

2%

3%

4%

5%

Figure 5: Effect of sulphuric acid concentration on copper and cobalt pressure leaching.

(Temperature: 140 oC, oxygen partial pressure: 5.0 atm, liquid-solid phase ratio: 10:1,

rate of mixing: 400 rpm).

Figure 6: Effect of sulphuric acid concentration on metals recovery by pressure leaching from

Lubin middlings.

Only for concentration of sulphuric acid of 30 g/dm3 the highest leaching rate and the maximum

concentration of copper in the solution were observed. There are two steps of pressure leaching. The

ferric ion starts to leach sulphidic minerals before the oxygen introduction. There were also ob-

served some differences in ferric concentration depending on sulphuric acid concentration.

0

10

20

30

40

50

60

70

80

90

100

2,0 2,5 3,0 3,5 4,0 4,5 5,0

Lea

chin

g r

eco

ver

y,

%

.

H2SO4 concentration, %

Cu

Fe

Ni

Co

As

Chmielewski, Wodka


For 20 g/dm3 of H2SO4 the concentration of ferric ions in solutions was 680 mg/dm

3 and for

50 g/dm3 of H2SO4 the concentration of Fe(III) increased to 940 mg/dm

3.

Generally, sulphuric acid concentration was found to have an effect on middlings ferric leaching.

For concentration of sulphuric acid below 20 g/dm3 the precipitation of ferric compounds have oc-

curred after 60 minutes of leaching. There were also two leaching steps for 30, 40, and 50 g/dm3

concentrations of sulphuric acid, similarly to copper leaching. The highest concentration of ferric

ions in solution has been reached for sulphuric acid exceeding 30 g/dm3. Kinetic curves observed

for Cu and Co leaching at acid concentrations above 30 g/dm3 were almost identical.

There was no cobalt leaching detected before introduction of oxygen. The course of cobalt leaching

resembles the copper leaching and there were observed two steps of it. The first step with the high

rate of cobalt dissolution and the second one with evidently slow cobalt dissolution rate.

In general, a little effect of sulphuric acid concentration on cobalt and copper leaching has been

found from H2SO4 concentrations above 3 % (Figure 7). A 97 % of leaching recovery of copper and

80 % of leaching recovery of cobalt recovery was reached regardless to sulphuric acid concentra-

tion. The highest iron recovery occurred for 30 g/dm3 of H2SO4. Presented results suggest that high

recovery of copper with simultaneous partial precipitation of ferric compounds is possible. How-

ever, it was evident, that at the lowest H2SO4 concentration (2 %) the precipitation of Fe(III) from

the solution was observed, most likely as FeOOH and Fe2O3 and the concentration of Fe in solution

decreased to about 1.3 g/dm3. This precipitation of Fe was not detected at higher acid concentration

(3-5 %). Iron control in the leaching solution at elevated temperature can be useful as a method of

purification from the excesses of Fe prior the recovering of metals. His effect will be further inves-

tigated in details.

3.3 Effect of oxygen partial pressure on Cu, Fe and Co leaching

The effect of oxygen partial pressure was investigated in the range from 2.5 to 10 MPa. The leach-

ing temperature (140 oC), sulphuric acid concentration (50 g/dm

3) and liquid/solid ratio ratio (10:1)

were kept constant. Figure 7 presents concentration – leaching time relationships for copper and

cobalt. Two steps of leaching courses were found, similarly to the other experiments. Effect of oxy-

gen partial pressure on leaching recovery of Cu, Fe, Ni, Co and As were collected in Figure 7. Lead

and silver were not solubilised during pressure leaching.

On the basis of results presented in Figure 8 it can be seen that pressure leaching is a very efficient

process. Oxygen partial pressure has a noticeable effect on copper, cobalt, arsenic and ferric solubi-

lisation. The highest recovery of metals was detected at oxygen partial pressure of 7.5 MPa. Solubi-

lisation of nickel was highly reduced most likely due to the dissemination of Ni in pyritic phase,

according to mineralogical examinations by BRGM, Orleans [28].



0

500

1000

1500

2000

2500

3000

3500

0 30 60 90 120 150 180 210

Leaching time, min.

Cu c

on

cen

trat

ion

, m

g/d

m3

.

10 atm

7,5 atm

5 atm

2,5 atm

0

10

20

30

40

50

60

0 30 60 90 120 150 180 210

Leaching time, minC

o c

on

cen

trat

ion

, m

g/d

m3

.

10 atm

7,5 atm

5 atm

2,5 atm

Figure 7: Effect of oxygen partial pressure on Cu and Co leaching rate for the pressure leaching of

Lubin middlings. Temperature: 140 oC, concentration of H2SO4: 5 %, liquid/solid ratio: 10:1.

Figure 8: Effect of oxygen partial pressure on metals recovery from Lubin middlings.

Temperature: 140 oC, H2SO4: 50 g/l, s/l: 1:10.

Table 3 summarizes the pressure leaching experiments for Cu, Co, Ni, As and Zn on the basis of

solid residue analysis. According to the presented data, only nickel leaching recovery was observed

to be below 40 %. Copper (98.9 %), cobalt (82.9 %) and zinc (98.2 %) can be easily leached out

0

10

20

30

40

50

60

70

80

90

100

2 3 4 5 6 7 8 9 10

Lea

chin

g re

cov

ery

, %

.

Oxygen partia l pressure, atm .

Cu

Fe

Ni

Co

As

Chmielewski, Wodka


from the examined Lubin middlings. Pressure leaching can be therefore considered as an efficient,

potential alternative for hydrometallurgical processing of Lubin shale middlings.

Table 3: Recovery of metals in pressure leaching of Lubin middlings in oxygenated sulphuric acid.

Sample

Tempe-

rature, oC

oxygen par-

tial pressure,

atm

H2SO4

concen-

tration,

%

Leaching recovery, %

Cu Co Fe Ni As Zn

1. 100 5 5 68.8 8.1 6.6 6.2 8.1 87.5

2. 120 5 5 38.1 33.6 12.6 27.2 30.7 88.3

3. 140 5 5 98.8 80.3 70.3 28.2 50.3 97.5

4. 140 10 5 98.7 81.7 77.3 22.7 41.9 97.4

5. 140 7,5 5 98.9 82.5 83.6 36.2 50.5 97.4

6. 140 2,5 5 97.9 77.5 82.6 30.4 48.5 96.8

7. 140 5 2 98.2 82.9 38.5 39.3 44.0 96.7

8. 140 5 3 98.4 81.2 70.0 33.2 50.4 97.4

9. 140 5 4 98.6 79.3 79.5 32.8 51.5 97.2

10. 160 5 5 83.4 71.7 86.4 25.5 48.4 98.1

11. 180 5 5 89.8 78.6 69.4 29.1 45.8 98.2

12. 140 5 5 98.8 81.2 61.2 31.3 47.1 97.2

13. 140 5 5 98.3 80.2 78.5 26.8 52.8 96.6

Solid residue after pressure leaching was examined by SEM (Figure 9). It can be seen that un-

leached metal-bearing minerals (mostly covellite, chalcopyrite, pyrite and galena) are finely disse-

minated in shale organic matter (Figures 9B, C and D), which was practically not decomposed nei-

ther during non-oxidative nor during pressure leaching. To liberate these particles for further metals

recovering required is either further intensification of pressure leaching or additional milling to re-

duce particle size of unleached solid grains prior the leaching.

Some unleached minerals (CuS, CuFeS2) remain in the solid as quite coarse, spongy structure re-

quiring either longer leaching time or more intensive leaching parameters (Figure 9 A). Reducing of

particle size in the leaching feed is also strongly recommended.



3.4 Characterization of pressure leaching residue by SEM examinations

Figure 9: SEM pictures of pressure leached samples of Lubin middlings exhibiting fine dissemi-

nation of metals-bearing minerals in shale matter.

4 Conclusions

Shale-containing by-product (middlings)-tailings of 1st cleaning from flotation circuits at Lubin

Concentrator, which exhibited remarkably elevated content of organic carbon, can be efficiently

processed hydrometallurgically by non-oxidative leaching followed by pressure leaching. High car-

bonate content, a specific and unique feature of Polish copper ores, and fine dissemination of metal-

bearing minerals (predominantly sulphides) unquestionably require a chemical pretreatment of the

shale feed with H2SO4 prior the atmospheric and pressure leaching in oxygenated solutions of sul-

phuric acid.

Non-oxidative leaching of Lubin middlings (Cu – 2.7 %, Pb – 1.52 %, Ni – 374 ppm, Co – 572 ppm,

Ag – 170 ppm) is a very rapid, selective, and relatively simply-controlled process. Selective liberation

of metal sulphides improves their flotatability while total decomposition of carbonates makes further

pressure leaching more efficient.

B

C

D

A

Chmielewski, Wodka


Pressure leaching of Lubin middlings revealed, that the process is remarkably fast and efficient for

recovering of Cu, Fe, Co and As with much lower recovery of Ni. To accomplish maximum recovery

of metal, pressure leaching has to be conducted at temperatures above 140 oC and under oxygen pres-

sure exceeding 5 atm. From pressure leaching experiments performed at temperatures 120-180 oC,

under oxygen pressure 2.5-10 atm, at H2SO4 concentrations 2-5 %, and at solid/liquid ratio 1:10 it re-

sults that about 96-97 % of Cu, 96 % of Fe 96 % of As, and 82 % of Co can be recovered after about

2 hours of leaching. Much lower leachability of nickel (30-40 %) can be explained in terms of its dis-

semination in pyrite. It was also observed that at temperatures exceeding 160 oC precipitation of Fe as

goethite or hematite commences, which can be applied as an iron control process for solution purifica-

tion.

Pressure leaching with oxidized H2SO4 solutions was not efficient to leach of Ag, Pb, and precious

metals. The solid residue after leaching must be either upgraded by flotation or subjected to leaching

in chloride solutions to recover remaining metals.

Acknowledgements

This work was carried out in the frame of Bioshale (European project contract NMP2-CT-2004

505710). Author acknowledges the financial support given to this project by the European Commis-

sion under the Sixth Framework Programme for Research and Development. We also wish to thank

our various partners on the project for their contributions to the work reported in this paper.

References

[1] COLLINS M.J., STIKMA J., BUBAN K.R., MASTERS J.M., (2000): Pressure acid leacing

of zinc and copper concentrates by Dynatec, EPD Congress, Ed. P.R Taylor, TMS, Nashville,

TN: USA, 2000, pp. 597-605.

[2] JANKOLA W.A., (1995): Zinc pressure leaching at Cominco. The principles and practice of

leaching. Proc. Int. Symp.: Principles and Practice of Leaching., The 25th Annual Hydrome-

tal-lurgical Meeting of the Metallurgical Society of CIM, Winnipeg, Canada, October 15-18,

1995, Ed. W.C. Cooper and D. B. Dreisinger, pp. 63-71.

[3] RUBISOV. D.H., PAPANGELAKIS V.G., (2000): Sulphuric acid pressure leaching of later-

ites – speciation and prediction of metals solubilities “at temperature”, Hydrometallurgy, 58,

pp. 13 – 28.

[4] RUBISOV D.H., KROWINKIEL J.M., PAPANGELAKIS V.G., (2000): Sulphuric acid pres-

sure leaching of laterites - universal kinetics of nickel dissolution for limonites and limonit-

ic/saprolitic blends, Hydrometallurgy, 58, pp. 1-11.



[5] NICE R.W., (2004): The Western Australian nickel laterites project – What we have learned?,

Pressure Hydrometallurgy 2004, Proc. 34th

Annual Hydrometallurgy Meeting, Banff, Alberta,

Canada, Ed. By M.J. Collins and V.G. Papangelakis, 155-180.

[6] KING J.A. AND DREISINGER D.B, (1995): Autoclaving of Copper Concentrates, Proc. Int.

Conf. Copper-Cobre 95, ed. Cooper W.C. et al., The Metallurgical Society of CIM, Montreal,

1995, 511-532.

[7] JONES D.L. AND HESTRIN J., (1998): CESL Process for Copper Sulphides: Operation of

the Demonstration Plant, The Alta 1998 Copper Sulphides Symposium, Brisbane, Australia,

October 19, 1998.

[8] BARTA L.A., BUBAN K.R., STIKSMA J. AND COLLINS M. J., (1999): Pressure Leaching

of Chalcopyrite Concentrates by Dynatec,” Proc. Int. Conf. Copper – Cobre 99, Ed. Young

S.K et al., vol. IV, 167 – 180.

[9] JONES D.L. AND HESTRIN J., (1998): CESL Process for Copper Sulphides: Operation of

the Demonstration Plant, The Alta 1998 Copper Sulphides Symposium, Brisbane, Australia,

October 19, 1998.

[10] DREISINGER D.B., STEYEL J. D.T., SOLE K.C., GNOIMSKI J., DEMPSEY P., (2003):

The Anglo-American Corporation/University of British Columbia (AAC/UBC) Chalcopiryte

process: An integrated pilot – plant evaluation, Int. Copper-Cobre 2003 Conf., Santiago,

Chile, Nov. 30 – Dec. 3, Volume VI – Hydrometallurgy of Copper (Book 1), 2003, 223 - 238.

[11] ANDERSON C.G., (2003): The application and economics of NSC catalyzed pressure leach-

ing of copper ores and concentrates., Copper-Cobre 2003 Conf., Santiago, Chile, Nov. 30 –

Dec. 3, vol. VI – Hydrometallurgy of Copper (Book 1), 2003, 289 – 306.

[12] FLEMING C.A., FERRON C.J., DREISINGER D.B., O KANE., (2000): A Process for the

Simultaneous Leaching and Recovery of Gold, Platinum Group Metals and Base Metals From

Ores and Concentrates, EPD Congress 2000, The Minerals, Metals and Materials Soci-

ety(TMS), 2000 TMS Annual Meeting, March 12 -1 6, Ed.: Taylor P.R, 419 – 431.

[13] THOMAS G.K., (1991): Alkaline and Acidic Autoclaving of Refractory Gold Ores, J. Metals,

43(2), pp. 16-19.

[14] THOMAS G.K., WILLIAMS R.A., (2000): Alkaline and Acid Autoclaves at Barrick Gold, A

Review, EPD Congress 2000, Ed.: P.R., Taylor, p. 433 – 449.

[15] MARSDEN J.O., IAIN HOUSE C., (2006): The Chemistry of Gold Extraction, 2nd Edition,

Soc. Min. Met. Expl. Inc.

Chmielewski, Wodka


[16] CURLOOK W., PAPANGELAKIS V.G., BAGHALHA M., (2004): Pressure acid leaching of

non-ferrous smelter slags for the recovery their base metal values, Pressure Hydrometallurgy

2004, Proc. 34th

Annual Hydrometallurgy Meeting of CIM, October 23-27, 2004, Banf, Cana-

da, (Collins M.J., and Papangelakis, Eds.), MetSoc, pp. 823- 838.

[17] RAMACHANDRAN V., LAKSHMANAN V.I., KONDOS P.D., (2007): Hydrometallurgy of

copper sulphide concentrates: An update, Proc. 6th

Int. Conf. Copper-Cobre 2007, Toronto,

Canada, (Eds.: Riveros P.A., Dixon D.G., Dreisinger D.B.) vol. IV (book 1), Can. Inst. Min.

Metal. Petrol., pp. 101 – 128.

[18] MARSDEN J.O., WILMOT J.C., (2007): Sulphate-based process flowsheet options for hydro-

metallurgical treatment of copper sulphide concentrates, Proc. 6th

Int. Conf. Copper-Cobre

2007, Toronto, Canada, (Eds.: Riveros P.A., Dixon D.G., Dreisinger D.B.) vol. IV (book 1),

Can. Inst. Min. Metal. Petrol., pp. 77 – 100.

[19] KUBACZ N., SKORUPSKA B., (2007): Evaluation of the effect of organic carbon on flota-

tion and smelting processes. Conf. Proc.: VIII International Conference On Non-ferrous Ore

Processing, ICNOP’07, 21-23 May 2007, Wojcieszyce, Poland, KGHM, IMN, pp. 157-166 (in

Polish).

[20] KONIECZNY A., KASINSKA – PILUT E., PILUT R., (2009): Technological and technical

problems in mineral processing of Polish copper ores at KGHM Polska Miedz SA, Proceed-

ings: IX International Conference on Nonferrous Ore Processing, 18-20 May 2009, Ladek

Zdroj Trzebieszowice (Poland), KGHM Cuprum, pp. 11-21 (in Polish).

[21] CHMIELEWSKI T., CHAREWICZ W., (2006): Hydrometallurgical processing of shale by-

products from beneficiation circuits of Lubin Concentrator, In: Perspectives for applying bo-

leaching Technology to process shale-bearing copper ores, BIOPROCOP’06, Lubin 2006,

KGHM Cuprum, Wroclaw 2006, 125-145 (in Polish).

[22] D’HUGUES P., NORRIS P.R., HALLBERG K.B., SÁNCHEZ F., LANGWALDT J.,

GROTOWSKI A., CHMIELEWSKI T., T. GROUDEV T., (2008): Bioshale consortium, Bio-

shale FP6 European project: Exploiting black shale ores using biotechnologies?, Minerals En-

gineering 21, pp. 111–120.

[23] CHMIELEWSKI T., (2007): Non-oxidative Leaching of Black Shale Copper Ore from Lubin

Mine, Physicochemical Problems of Mineral Processing, 41, pp. 323-348.

[24] CHMIELEWSKI T., LUSZCZKIEWICZ A., KONOPACKA Z., (2007): Separation and Con-

cept of Processing of Black Shale Copper Ore from Lubin Mine, Proc. VIII International Con-

ference on Non-ferrous Ore Processing, Wojcieszyce (Poland), May 21-23, KGHM Cuprum,

Wroclaw, pp. 171-184 (in Polish).



[25] KONOPACKA Z., LUSZCZKIEWICZ A., CHMIELEWSKI T., (2007): Effect of non-

oxidative leaching on flotation efficiency of Lubin Concentrator Middlings. Physicochem.

Probl. Miner. Process., 41, pp. 275-289.

[26] CHMIELEWSKI T., (2007): Atmospheric leaching of shale by-product from Lubin Concen-

tra-tor., Physicochem. Probl. Min. Process., 41, pp. 337-348.

[27] WODKA J., CHMIELEWSKI T., ZIOLKOWSKI B., (2007): Pressure leaching of shale ore in

oxygenated sulfuric acid, Physicochemical Problems of Mineral Processing, 41, pp. 349-364.

[28] AUGE T., et al., (2007): BIOSHALE, Deliverable D.5.2., Product behavior during the hydro-

and bioprocesess (T12-30), March 2007.



Mine to Heap in Mantoverde Anglo American

Division

Manuel Díaz, Cristian Salgado, Carlos Pérez, Cristian Alvayai, Leonardo Herrera

Gabriel Zárate

Anglo American Chile

VicePresidency of Metallurgy

Pedro de Valdivia 291

Santiago, Chile

Keywords: Hydrometallurgy, copper, heap leaching, acid consumption

Abstract

Mantoverde is located in Chile, at 900 m above sea level, 53 km SE of Chañaral city. It currently

processes two ore types: heap ore (75 million ton, 0.73 % total copper, 0.61 % acid soluble copper

and 1.23 % calcium carbonate) and dump ore (40 million ton, 0.48 % total copper, 0.37 % acid so-

luble copper and 2.24 % calcium carbonate). In 2008, copper production was 62,500 tons.

During 2007, the following problems were experienced in heap leaching and solvent extraction:

• Lower copper recovery, being the predicted recovery lower than the current one.

• Higher concentration of aluminum, iron, magnesium and manganese in the pregnant leach solu-

tion and as result higher solution viscosity affecting the leach and solvent extraction operation

significantly.

• Higher acid consumption, being the predicted one lower than the current acid consumption and

higher calcium carbonate content.

A multidisciplinary team was formed to investigate the reasons and to set up a working program in

order to optimize these performance indicators. The main metallurgical improvements were:

• An increase in the acid soluble copper recovery by 3.2 points and a decrease in acid consumption

by 20 %.

• A decrease in leaching solution viscosity by 20 %.

The evaluation of the operational parameters and the heap results obtained, after the recommenda-

tions were implemented, are discussed in this paper.

Díaz, Salgado, Pérez, Alvayai, Herrera, Zárate


1 Introduction

The Mantoverde mine is located in the Chañaral province, Region III, in Northern Chile, 1000 km

north of Santiago and 40 km from the coast. The altitude of the mine is 900 m above sea level. Man-

toverde was designed to produce 42,130 ton of copper per year at a treatment rate of 5.4 million ton

of ore per year by using a heap leach-SX-EW process. Mantoverde uses a two stage leach approach,

where fresh ore is sprinkled with an intermediate solution (ILS) to produce PLS in a first stage and

then is irrigated with raffinate solution to produce ILS in a second stage. A dump leach process was

scheduled to be started 5 years after the plant start-up, at a treatment rate of 2 million ton of margin-

al ore per year.

Mantoverde (MVN) was commissioned in December 1995. Copper production in 1996 was

40,539 ton and it has steadily been increased to reach 62,500 ton in 2008, due to the start-up of a

similar size orebody, called Mantoverde Sur (MVS) and of another smaller orebody called Manto

Ruso (MR). In 2002, heap leach operation was changed from a permanent pad to a dynamic pad.

Forecast copper production for 2010 is 62,000 ton for an ore treatment rate of 9.6 million ton. A

first dump leach operation was started in 1999 with a copper production of 1663 ton and a second

dump leach operation was started early in 2002. Copper production from this source will increase

over time to as much as 10,000 ton. The mine statistics for the first years of operation is shown in

Table 1.

Table 1: Mantoverde production statistics.

Year Throughput (ktpy) Production (tpy)

Heap Dump Heap Dump

1996 5.436 ---------- 40.539 ----------

1997 6.156 ---------- 47.627 ----------

1998 6.214 ---------- 48.032 ----------

1999 6.367 1.891 50.364 1.663

2000 7.144 3.233 49.456 4.152

2001 8.001 4.135 50.365 5.205

2002 8.397 4.573 51.448 5.841

2003 9.001 6.048 53.250 6.707

2004 9.017 7.028 52.016 8.095

2005 9.439 3.625 56.435 5.565

2006 9.502 4.880 53.713 6.608

2007 9.281 5.511 54.951 6.050

2008 9.557 4.300 56.618 5.883

Mine to Heap in Mantoverde Anglo American Division


During 2007, the following problems were experienced in heap leaching and solvent extraction:

• Lower copper recovery, being the predicted recovery lower than the current one.

• Higher concentration of aluminum, iron, magnesium and manganese in the pregnant leach solu-

tion and as result, an increase in the solution viscosity affecting the leach and solvent extraction

operation significantly.

• Higher acid consumption, being the predicted one lower than the current acid consumption and

higher ore calcium carbonate content.

A team was formed to investigate the reasons and to set up a working program in order to optimize

these performance indicators. For that purpose, the operating data of 74 heaps was reviewed and a

column test program was carried out. Results were as follows.

2 Analysis of Operating Data

The operating data of 74 heaps, covering several years of operation, was reviewed and the effect on

copper extraction of variables such as total copper (Cut), acid soluble copper (Cusol) and CaCO3

grades, heap height, net acid consumption, acid dosage in curing stage, particle size, flow rate, acid

concentration in leaching solutions and ore feed distribution was assessed.

It was concluded that heap height was one of the most important variables affecting negatively the

copper extraction, as it can be seen in Figure 1. Acid addition in the curing stage was another impor-

tant variable.

Figure 1: Cusol recovery and heap height. Average values.

70,0

72,0

74,0

76,0

78,0

80,0

82,0

84,0

86,0

88,0

90,0

5,5 6,0 6,5 7,0 7,5 8,0

Height (m)

Cu

so

l R

ec

ov

ery

(%

)



It was also found that heap height and CaCO3 grade were the most important variables increasing

acid consumption. Accordingly, it was recommended to decrease gradually heap height from 7.5 m

to 5 m and to optimize acid addition distribution by reducing acid dosage in curing stage and in-

creasing proportionally acid concentration in leaching solutions [1].

3 Column Test Work

A number of column tests have been conducted using ore samples taken from the MVN, MVS, MR

and a new ore body called Kuroki, and ILS and raffinate solution samples. The main goal of this

program was to update copper recovery equations for different Cusol and CaCO3 grades and to op-

timize acid addition management.

The results of this work were reported elsewhere [2]. It has been demonstrated that a 50 % decrease

in the acid added to the curing stage, combined with an increase in the acid concentration of ILS and

raffinate solution, had a significant reduction in acid consumption without affecting negatively the

copper extraction nor the leach cycle, as it can be seen for example in the following figures, where

column 31 was cured and leached in the standard way while column 33 was cured with 50 % less

acid and leached with solutions having 50 % more acid, as compared to standard solutions.

Figure 2: Copper extraction kinetics of columns 31 and 33.

0,0

10,0

20,0

30,0

40,0

50,0

60,0

70,0

80,0

90,0

100,0

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5

m3/t ore

Cu

so

l R

ec

ov

ery

(%

)

Col 31

Col 33



As a consequence, it was recommended to decrease gradually the acid addition to the curing step

from 90 % to 70 % and then to 60 %, being 50 % the final target, and to adjust the acid concentra-

tion of ILS and raffinate solution from 3 to 8 g/l and from 13 to 15 g/l respectively.

Figure 3: Cusol extraction and net acid consumption of columns 31 and 33.

0,0

10,0

20,0

30,0

40,0

50,0

60,0

70,0

80,0

90,0

100,0

0,0 10,0 20,0 30,0 40,0 50,0 60,0

Net Acid Consumption (kg/t ore)

Cu

so

l R

eco

very

(%

)

Col 31

Col 33



4 Commercial Heap Results

The heap height was reduced from 7.5 to 6.5 m by the end of August 2007, resulting in a Cusol re-

covery increase as it is shown in Figure 4, where predicted and actual recoveries are graphed.

Figure 4: Predicted and actual heap Cusol recovery.

Concerning acid addition, it was established in the design criteria that 80 % of the net acid consump-

tion was going to be added in the curing stage with the remaining 20 % through the leach solutions,

namely intermediate leach solution (ILS) and raffinate solution. At Mantoverde, the acid consump-

tion is related to the ore calcium carbonate content which is measured every three hours and the acid

addition to the curing stage is adjusted accordingly. Since the start up and until 2007, the acid was

added following this methodology. Acid addition in the curing stage was increased to 90 % and dur-

ing some period of time it reached 100 %.

The recommendations on acid addition were implemented during the first semester of 2008 at the

heap leach operation with excellent results, at it is shown in Figures 5 and 6. In the first one, the net

acid consumption, actual and estimated, and calcium carbonate grades are graphed for years 2001

through 2007, while in the second one is done for year 2008, from January through August.

It can be seen that the actual net acid consumption started to be higher than the estimated one in

2004 and that this trend has been reversed since June 2008, once the recommendations were fully

implemented. The reduction in net acid consumption was on average 5.8 kg/t, about 20 %, equiva-

70,0

72,0

74,0

76,0

78,0

80,0

82,0

84,0

86,0

88,0

90,0

Oct-06 Abr-07 Nov-07 Jun-08 Dic-08 Jul-09

Month

Cu

so

l R

eco

very

(%

)

Predicted

Actual



lent to US$ 11.4 million as a projected annual figure. The Cusol recovery was not affected, as it can

be seen in the previous figure.

Figure 5: Net Acid Consumption and Calcium Carbonate Grades for 2001-2007.

Figure 6: Net Acid Consumption and Calcium Carbonate Grades for 2008.

0

5

10

15

20

25

30

35

40

2000 2001 2002 2003 2004 2005 2006 2007 2008

Year

Net

Acid

Co

nsu

mp

tio

n (

kg

/t o

re)

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

Actual

Equation

CaCO3

0

5

10

15

20

25

30

35

40

Nov-07 Feb-08 Jun-08 Sep-08 Dic-08

Month

Net

Acid

Co

nsu

mp

tio

n (

kg

/t o

re)

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

Predicted

Actual

CaCO3



The modification in the acid addition strategy resulted not only in a significant reduction in

acid consumption but also in PLS viscosity, as it is shown in Figure 7. Viscosity reduction was

around 20 %.

Figure 7: PLS viscosity.

Due to the success obtained with these modifications, the application to other leach processes such

as dump and vat leaching is being evaluated.

5 Conclusions and Recommendations

Mantoverde is a heap leach-SX-EW plant that was commissioned in December 1995 at a production

rate of 42,130 ton of copper per year and 5.4 million ton of ore per year. Forecast copper production

for 2010 is 62,000 ton for an ore treatment rate of 9.6 million ton.

During 2007, a number of problems were experienced in heap leaching and solvent extraction,

among them a decrease in copper recovery and an increase in acid consumption, impurities concen-

tration and PLS viscosity. A team was formed to investigate the reasons and to set up a working

program in order to optimize these performance indicators. For that purpose, the operating data of

74 heaps was reviewed and a column test program was carried out.

It was concluded that heap height and acid addition in the curing stage were the most important va-

riables affecting negatively the copper extraction. It was also found that heap height, CaCO3 grade

0,0

1,0

2,0

3,0

4,0

5,0

6,0

7,0

Nov-07 Feb-08 Jun-08 Sep-08 Dic-08 Mar-09 Jul-09

Month

PL

S v

isc

os

ity

(c

p)



and acid addition strategy were the most important variables increasing acid consumption. Accor-

dingly, it was recommended to decrease gradually heap height from 7.5 m to 5 m and to optimize

acid addition distribution by reducing acid dosage in curing stage and increasing proportionally acid

concentration in leaching solutions.

The heap height was reduced from 7.5 to 6.5 m by the end of August 2007, resulting in a Cusol re-

covery increase of 3.2 points, while the recommendations on acid addition were implemented during

the first semester of 2008 at the heap leach operation with an average reduction in net acid consump-

tion of 5.8 kg/t, about 20 %, equivalent to US$ 11.4 million, as a projected annual figure. The modifi-

cation in the acid addition strategy resulted not only in a significant reduction in acid consumption

but also in PLS viscosity by 20 %.

Due to the success obtained with these modifications, the application to other leach processes such

as dump and vat leaching is being evaluated.

Acknowledgements

The authors wish to thank the General Management of Anglo American Chile for the permission to

publish this paper. The technical contributions to the work made by a number of process engineers

at Mantoverde Division are also acknowledged.

References

[1] G. Zárate. Análisis Operación Pilas Mantoverde. Julio 2007. Interim Report.

[2] C. Salgado, C. Pérez, C. Alvayai, G. Zárate, Acid management in heap leaching. Are we doing

right? HydroCopper 2009, E. Domic, J. Casas Eds. Gecamin, Santiago, Chile. pp. 47-56.



Predicting the Effects of Locked, Partially Locked,

and Liberated Minerals in Copper Leaching

Michael L. Free, Abraham L. Jurovitzki

University of Utah

Department of Metallurgical Engineering

Salt Lake City, 84112, USA

Keywords: Liberation, heap leach modeling

Abstract

Valuable mineral grains are associated with gangue material in an ore. The association of the valua-

ble mineral grains with the gangue material can be characterized as liberated, partially-locked, or

fully-locked. Valuable mineral grain leaching kinetics depends on the association with the gangue.

However, traditional heap leaching modeling typically assumes all valuable mineral grains are fully-

locked within the gangue. This paper presents modeling results obtained by including the effects of

liberated, partially-locked, and fully-locked valuable mineral grains on leaching.

1 Introduction

Accurate modeling of heap leaching requires accurate quantification of the rate and magnitude of

resulting valuable mineral dissolution. However, within an ore particle, valuable mineral grains are

interspersed with gangue particles. The valuable mineral grain can occupy three different states at

various ore sizes. The valuable mineral grains can be fully locked in the interior of the ore, partially

locked within the ore, or completely liberated from the ore as depicted in Figure 1 [1].

Free, Jurovitzki


Figure 1: Illustration of the three valuable mineral particle states and their relationship with an ore

particle.

Valuable mineral particles that are locked inside of the host rock particles require pore diffusion of

reactant in order for leaching to occur. In contrast, liberated particles have excellent access to reac-

tant in surrounding solution. Partially-locked valuable mineral particles have partial access to reac-

tant in the environment. Each type of relationship between the host rock and valuable mineral par-

ticles requires a different leaching model. The ability to quantify these relationships provides the

foundation for this modelling approach, which is described in greater detail in other references

[1-2].

1.1 Estimating host rock and valuable mineral particle associations

The estimated probability of liberation can be calculated using the expression [2]:

))exp(1)(exp(*

**

*

**

.

hrp

hrpvmp

vmp

hrpvmp

estlibr

rr

r

rrP

−

−

−

= (1)

Plib.est is the estimated probability of particle liberation, r*

vmp is the reference radius of valuable min-

eral particles, and r*

hrp is the reference radius of host rock particles. The probability of valuable

mineral that is partially-locked can be determined using the mathematical expressions [2]:

)...]...()(1))][exp(1)(exp(1[ **3

*

**

*

**

*

**

. vmphrp

hrp

vmphro

hrp

hrpvmp

vmp

hrpvmp

estLockedPartially rrforr

rr

r

rr

r

rrP >

−

−

−

−

−

−=−

(2)

)...))]...(exp(1)(exp(1[ **

*

**

*

**

. vmphrp

hrp

hrpvmp

vmp

hrpvmp


rr

r

rrP <

−

−

−

−=−

(3)

Liberated valuable

mineral particle

Partially-locked valuable

mineral particle

Fully-locked valuable

mineral particle

Predicting the Effect of Locked, Partially Locked, and Liberated Minerals


Correspondingly, the probability of fully-locked valuable mineral particles is [2]:

)...]...()))][(exp(1)(exp(1[ **3

*

**

*

**

*

**

. vmphrp

hrp

vmphro

hrp

hrpvmp

vmp

hrpvmp

estLockedFully rrforr

rr

r

rr

r

rrP >

−−

−

−

−=−

(4)

)......(0 **

. vmphrpestLockedFully rrforP <=−

(5)

1.2 Liberated Particle Leaching

Liberated particle leaching can be described mathematically by the expression:

])1(1[)( y

libvmpliblibvmp rkrt α−−= (6)

klib is a reaction constant (sec/cm), α is the fraction reacted, and y is 1/3,1/2, or 2/3 for reaction con-

trol, rapid flow/fine particle, and slow flow/larger particle leaching conditions, respectively. The

subscript lib is added to indicate these terms apply to the liberated fraction of valuable mineral par-

ticles. This equation assumes leaching without product layer formation. If a porous product layer

forms during leaching, a pore diffusion/shrinking core model can be applied:

])1(1[)( 3/2

,,32

2

,

, PDlibPDlib

effx

vmplibPD

PDlibvmpDC

rkrt αα −−−= (7)

Cx is the concentration of reactant, Deff is the the effective diffusivity, and kPD,lib is a constant.

1.3 Partially-Locked Particle Leaching

A useful model for evaluating partially-locked valuable mineral particle leaching is [2]:

])1(1[]

)3

4(

4[)(

3/2

y

libvmpliblibvmp rkrt α

π

π−−≈ (8)

If a porous product film forms during leaching, the pore diffusion model (equation 7) can be simi-

larly applied using the liberated particle leaching case by multiplying by 4π/(4π/3)2/3

[2].

1.4 Fully-Locked Particle Leaching

A common model for diffusion controlled leaching through a porous reaction product layer or

shrinking core is:

])1(1[)( 3/2

,,32

2

,

, PDlockPDlock

effx

hrplockPD

PDlibvmpDC

rkrt αα −−−= (9)

Free, Jurovitzki


This equation is generally the same as given for the liberated particle leaching for product layer

formation. However, in this case the host rock particle diameter is used. Other constants are also

different due to the general change in application.

1.5 Ore Leaching

Leaching of an ore that consists of liberated, partially-locked, and fully-locked valuable mineral par-

ticles is the sum of the leaching of the valuable mineral grain distributed in these three categories.

Thus, overall ore leaching is determined by summing the contributions of leaching in each category

that are weighted by their relative proportions. The overall leaching was determined using an Excel®

spreadsheet and the accompanying goal seek function. The goal seek function was used to find the

fraction reacted that corresponded with a specified time. The resulting fraction reacted values were

weighted based on the probability of occurrence and summed to determine the overall fraction at a

specified time. The resulting data was used to construct overall fraction reacted versus time plots.

2 Method

2.1 Ore sample and reagents

Copper ore was dry screened into appropriate size classes. Subsequently, samples were prepared for

hydrometallurgical processing by being split into desired quantities.

Leaching solutions were prepared with reagent grade sulfuric acid (H2SO4, MW ~98.08, assay

95.0 -98.0 %) which, was obtained from EMD Chemicals. In the case of chalcopyrite ore leaching

10 g/l of Ferric sulfate (Fe(SO4)3,· nH2O MW ~399.88, purity 73.0 %) was added, which was ob-

tained from Mallinckrodt Chemicals and 15 g/l of Sodium chloride (NaCl, MW ~58.44, purity

99.0 %), which was obtained from Mallinckrodt Chemicals, were added.

Chemical analysis solutions consisted of Hydrochloric acid (HCl, MW ~36.46, assay 36.5-38 %)

which, was obtained from EMD Chemicals, and Nitric acid (HNO3, MW ~63.01, purity 68.0 to

70.0 %) which, was obtained from Mallinckrodt Chemicals.

All of these chemicals were used without any further purification. All of solutions were prepared

using ASTM Type I water.

2.2 Column leaching experiments

Leaching was performed in saturated flow columns packed with glass beads at the bottom and a

layer of copper ore on top. The two layers were separated first by a rigid permeable plastic separator

and on top of that was a flexible permeable plastic barrier. The leaching solution was contained

within a large Erlenmeyer flask. The leaching solution was pumped from the large Erlenmeyer flask



via Masterflex pump to a connection at the bottom of the enclosed column and exited the top back

to the beaker containing the original solution via another Masterflex tube. The solution with the dis-

solved metal ions was continuously recirculated at a flow rate of 1-2 ml/sec. A pH of 1.5 was main-

tained via an AccuTipH glass body rugged blub electrode probe which was connected to an Eutech

Instruments pH 2000 series pH/ORP controller, which in turn activated an acid control pump (Mas-

terflex pump) that was connected to an acid reservoir burette of 0.5-1 M Sulfuric acid solution de-

pending on the test.

Five milliliter samples were withdrawn from the Erlenmeyer flask solution reservoir at regular in-

tervals. One milliliter of each of these samples was diluted up to a volume of 100 milliliters with

two percent Nitric acid to safeguard against possible precipitation of the metal ions.

2.3 Sample testing and chemical analysis

The dilute samples obtained during column leaching were analyzed via Inductively Coupled Plasma

Mass Spectrometry for dissolved copper content.

Chemical analysis tests of the various ore sizes were conducted to obtain a baseline for total copper

and iron content within the various size gradients. A split sample of each ore was comminuted in a

small ball mill under dry cascading conditions until it was a fine powder. Subsequently, a one gram

sample of the fine particle was further crushed via mortar and pestle for 3 minutes. 0.5 grams of this

powder was weighed out and added to a beaker containing twenty milliliters of aqua regia for diges-

tion. The aqua regia consisted of fifteen milliliters of Hydrochloric acid and five milliliters of Nitric

acid. The temperature was raised to 125 degrees Celsius on a hot plate and digested for twenty mi-

nutes. Afterwards, the beaker of digested ore was removed from the hot plate and allowed to cool

for fifteen minutes. After, the cooling period the ore was vacuum filter clear of all remaining par-

ticles using P4 Fisher filter paper. The strongly acidic filtrate was first diluted with 200 milliliters of

ATSM Type I water. One milliliter of this diluted solution was withdrawn and diluted further with

one hundred milliliters of ATSM Type I water. These chemical analysis samples were analyzed for

total iron and copper via inductively coupled plasma mass spectrometry. This process was carried

out for leached and unleached ore to verify the amount of copper that was being leached in the col-

umn leachings. Analysis of digested ore samples before and after leaching were in agreement (with-

in 20 %) with ICP evaluations of the relevant leaching solutions.


Test results from copper oxide ore column leaching are presented in Figure 2 with the associated

model fit shown with the measured data. The data in Figure 1 show the model provides a very good

estimate of the copper leaching performance.

Free, Jurovitzki


Figure 2: Comparison of measured and calculated fraction reacted and time for copper oxide ore

with an average size of 0.4 cm. The model parameters were r*vmp = 0.008 cm,

klib = 1.07×104 sec/cm, kPD,lock/Deff = 4.6×10

9 cm·sec/mol.

Results from chalcopyrite ore leaching are presented in Figure 3. The measured leaching results pre-

sented in Figure 3 are fit reasonably well by the model.

Figure 3: Comparison of measured and calculated fraction reacted and time for chalcopyrite ore

with an average size of 0.284 cm. The model parameters were r*vmp = 0.0045 cm,

klib = 5.76×105 sec/cm, kPD,lock/Deff = 3.2×10

9 cm·sec/mol.

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

0 500000 1000000

Fra

ctio

n R

ea

cte

d

Time (sec)

Model Measured

0

0.02

0.04

0.06

0.08

0.1

0.12

0 500000 1000000 1500000

Fra

ctio

n R

ea

cte

d

Time (sec)

Model

Measured



Leaching results from the mixed oxide/sulphide ore are presented in Figure 3. These results are sim-

ilar to those presented for the other types of ore. However, the model does not fit the mixed ore

leaching data as well.

Figure 4: Comparison of measured and calculated fraction reacted and time for copper

oxide/sulfide ore with an average size of 0.284 cm. The model parameters were

r*vmp = 0.008 cm, klib = 1.92×10

11 sec/cm, kPD,lock/Deff = 6.4×10

9 cm·sec/mol.

Although the model fits the copper leaching data from three very different ore samples, there are

several issues that need to be addressed to provide more confidence in this modelling approach. The

ore that was used has not been analyzed by a mineralogical assessment to evaluate the extent to

which other minerals may affect the modeling results, which have assumed a single mineral or a

group of similar minerals is responsible for acid consumption. Thus, further work is needed, and the

work presented in this paper represents a work in progress in early stages of evaluation.

4 Summary

The data presented in this paper show that a relatively new leaching modeling approach that ac-

counts for the relationship between host rock and valuable mineral particles can be used to predict

leaching performance. The new modeling approach will help to fill an important need to more accu-

rately model leaching of valuable entities from heterogeneous materials. Additional mineralogy as-

sessments before and after leaching are needed to further evaluate the performance of the model.

Thus, further work is needed, and the work presented in this paper represents a work in progress in

early stages of evaluation.

0

0.05

0.1

0.15

0.2

0.25

0 1000000 2000000 3000000

Fra

ctio

n R

ea

cte

d

Time (sec)

Model Measured

Free, Jurovitzki


Acknowledgements

The laboratory work of Abraham Jurovitzki, Ravindra Bhide, and Taylor Bird who performed the

experiments needed to provide the data in this paper are gratefully acknowledged along with partial

funding by the U. S. Department of Energy Center for Advanced Separation Technology and the

University of Utah Research Opportunities Program for undergraduate students.

List of symbols

r* particle reference size at which 62.3 % of the material passes as undersize when s = 1,

s index of the Weibull or Rosin-Rammler distribution function

P probability or fraction of particles smaller than size

r radius of particle

k constant

α fraction reacted,

y coefficient

C concentration of reactant

D diffusivity

subscripts

vmp valuable mineral particle

hrp host rock particle

x reactant

Eff effective

PD pore diffusion

Lib liberated

Lock indicates locked mineral particle

References

[1] FREE, M. (2008). Canadian Metallurgical Quarterly, 47(3), 277-284.

[2] FREE, M. (2010). Predicting Leaching Solution Acid Consumption as a Function of pH in

Copper Ore Leaching, 7th

International Copper 2010 – Cobre 2010 Conference at Hamburg,

Germany, Proceedings of Copper 2010, Volume 7, pp. 2711-2719.


Predicting Leaching Solution Acid Consumption

as a Function of pH in Copper Ore Leaching

Michael L. Free

University of Utah

Department of Metallurgical Engineering

Salt Lake City, 84112, USA

Keywords: Heap leach, acid consumption, modeling

Abstract

pH control and acid consumption are critical issues in many copper leaching operations. A mod-

erately low pH is a prerequisite to effective copper leaching. The quantity of acid needed to achieve

desired pH levels is one of the most important parameters in determining leaching costs. Thus, the

capability to accurately predict leaching solution pH and acid consumption during copper ore leach-

ing is important to the evaluation of leaching performance and cost. This paper will present experi-

mental and modeling results for acid consumption as a function of pH for copper leaching from

chalcopyrite ore in sodium chloride media.

1 Introduction

Acid consumption during copper leaching is a function of acid consuming minerals in the ore and

their leaching properties as well as their association with the host rock. Common acid consuming

minerals that are associated with copper ores include copper minerals such as chrysocolla, mala-

chite, azurite, tenorite, atacamite, and brochantite as well as gangue minerals such as goethite, limo-

nite, biotite, chlorite, and various feldspar minerals. Selected reactions include [1]:

CuSiO32H2O + 12H+ = 6Cu

2+ + 8H4SiO4 (1)

CuO + 2H+ = Cu

2+ + H2O (2)

CuSO43Cu(OH)2 + 6H+ = 4Cu

2+ + 6H2O (3)

KAlSi3O8 + 4H+ = K

+ + Al

3+ + 3H4SiO4 (4)

(H,K)2(Mg,Fe)2Al2(SiO4)3 + 10 H+ = 2K

+ + 2Al

3+ + 3H4SiO4 + 2(Fe,Mg)

2+ (5)

Fe2O3 + 6H+ = 2Fe

3+ + 3H2O (6)

FeO(OH) + 3H+ = Fe

3+ + H2O (7)

Free


Although the stoichiometry of these reactions is fixed, the actual consumption of acid by each of

these minerals is related to mineral size and distribution as well as by the formation of passive lay-

ers that can restrict access to acid consuming minerals.

As the acid consuming minerals dissolve, by-products can form that can alter the acid balance. Not-

able examples include jarosite and alunite formation [1]:

3Fe3+

+ K+ +2SO4

2- + 6H2O = KFe3(SO4)2(OH)6 + 6 H

+ (8)

3H2O + K+ + 2SO4

2- + 3Al

3+ = KAl3(OH)6(SO4)2 + 6 H

+ (9)

Each acid consuming mineral will eventually come to equilibrium with a surrounding solution envi-

ronment, although the time needed to reach equilibrium may exceed a practical time period. Often, a

near equilibrium condition can exist.

Reaction kinetics often temporarily overshadow the influence of equilibrium thermodynamics. In

some cases the dissolution of acid consuming minerals is very slow due to low rate constants or the

formation of secondary reaction products that passivate the mineral. Acid consuming minerals are

often found as locked, partially-locked, and liberated mineral grains. Consequently, their dissolution

behavior in acid follows several kinetic models. The rate of dissolution is also influenced by the

accumulation of ions in the leaching solution.

1.1 Estimating Liberated, Partially-Locked, and Locked Valuable Mineral

Grains

The fraction of particles that are liberated can be estimated using the size distributions of the host

rock particles and valuable mineral grains. Although a variety of size distribution functions can be

used for this purpose, the Rosin-Rammler distribution will be used here to demonstrate the process.

The Rosin-Rammler cumulative passing size distribution function is expressed as [2]:

s

r

rrP

−−=

*exp1)( (10)

in which s is an experimentally determined constant, r* is the particle reference size at which 62.3 %

of the material passes as undersize when s = 1, and P is the probability or fraction of particles

smaller than size r. As r becomes very large, the probability of undersize approaches one. The frac-

tion of liberated particles can be estimated using the size distribution functions for the valuable

mineral grains or particles and the host rock particles. Valuable mineral particles can be categorized

as liberated when they are larger than host rock particles and cannot, therefore, be incorporated in-

side of host rock particles. When size distributions for valuable mineral particles and host rock par-

ticles are considered, the liberated fraction of valuable mineral grains is determined by comparing

small size ranges of these distributions. If the small size range of valuable mineral particles under

examination is larger than the size range of host rock particles selected, that combination is consi-

Predicting Leaching Solution Acid Consumption as a Function of pH


dered to have liberated valuable mineral particles. Accordingly, the probability of that combination

of size ranges is used to appropriately weight that combination. This process is completed by com-

paring each small size range of valuable mineral particles to each host rock particle size range to

determine the overall fraction of valuable mineral particles that are considered to be liberated. This

process is described in more detail elsewhere [3].

The fraction of liberated particles can be estimated in a simple way using the geometric mean size

between the associated distributions. The estimated probability of liberation using the Rosin-

Rammler function assuming s is unity is [3]:

))exp(1)(exp(*

**

*

**

.

hrp

hrpvmp

vmp

hrpvmp

estlibr

rr

r

rrP

−

−

−

= (11)

Plib.est is the estimated probability of particle liberation, r*

vmp is the reference radius of valuable min-

eral particles, and r*

hrp is the reference radius of host rock particles. Particles that are not liberated

are either locked or partially locked and their corresponding joint probability is the remainder of the

probability sum of unity (1- Plib.est). The fraction of valuable mineral particles that are partially

locked can be determined using a similar analysis in which different small size ranges are compared

over the entire distributions and a specific formula to determine locking is used as described else-

where. [3] A simplified use of this approach which uses the estimated probability of liberation (Eq-

uation (11)) with the particle locking estimate can be used to estimate the probability of valuable

mineral that is partially-locked [3]:

)...]...()(1))][exp(1)(exp(1[ **3

*

**

*

**

*

**

. vmphrp

hrp

vmphro

hrp

hrpvmp

vmp

hrpvmp


rr

r

rr

r

rrP >

−

−

−

−

−

−=−

(12)

)...))]...(exp(1)(exp(1[ **

*

**

*

**

. vmphrp

hrp

hrpvmp

vmp

hrpvmp


rr

r

rrP <

−

−

−

−=−

(13)

Correspondingly, the fraction of valuable mineral particles that are fully locked in the host rock par-

ticles can be determined as [3]:

)...]...()))][(exp(1)(exp(1[ **3

*

**

*

**

*

**

. vmphrp

hrp

vmphro

hrp

hrpvmp

vmp

hrpvmp

estLockedFully rrforr

rr

r

rr

r

rrP >

−−

−

−

−=−

(14)

)......(0 **

. vmphrpestLockedFully rrforP <=−

(15)

These new simplified formulas were evaluated by comparing the associated results for valuable

mineral exposure measured by Miller et al. [4], and the results are presented in Figure 1. The results

in Figure 1 indicate that the predicted exposure (liberated and partially-locked valuable mineral par-

ticles) is very similar to the measured exposure obtain using X-ray tomography at most sizes eva-

luated, although the fit is best in the middle and small host rock particle sizes.

Free


Figure 1: Comparison of measured [4] and calculated valuable mineral exposure using the sum of

Equations (11) and (12) based on data from reference 4.

1.2 Liberated Particle Leaching

Liberated particle leaching can involve a variety of scenarios that each require a different kinetic

leaching model. Several leaching models for liberated particles can be described mathematically by

the expression:

])1(1[)( y

libvmpliblibvmp rkrt α−−= (16)

klib is a reaction constant (sec/cm), α is the fraction reacted, and y is 1/3,1/2, or 2/3 for reaction con-

trol, rapid flow/fine particle, and slow flow/larger particle leaching conditions, respectively. The

subscript lib is added to indicate these terms apply to the liberated fraction of valuable mineral par-

ticles. This equation assumes the particle dissolves without forming a product layer. If a porous

product layer forms while leaching takes place, the traditional pore diffusion/shrinking core kinetic

model can be applied:

])1(1[)( 3/2

,,32

2

,

, PDlibPDlib

effx

vmplibPD

PDlibvmpDC

rkrt αα −−−= (17)

Cx is the concentration of reactant, Deff is the the effective diffusivity, and kPD,lib is a constant.

1.3 Partially-Locked Particle Leaching

Leaching of partially-locked mineral particles can be described in different ways, some of which are

not simple, and none of which are rigorous for ore leaching. Valuable mineral particles exposed to

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30Valu

ab

le M

inera

l E

xp

osu

re (

%)

Size (mm)

Simplified Model

Miller et al. (2003)



the surface will leach similarly to liberated particles except that only a portion of the particle is di-

rectly exposed to the solution. A simplified approach to evaluate partially-locked particle leaching

considers the partially locked particles as cubes with one face exposed to solution. The ratio of ex-

posed leaching area of a fully exposed sphere to one side of a cube of equal mass is 4π/(4π/3)2/3

.

Use of this ratio allows for a simplified conversion of liberated particle leaching to partially-locked

particle leaching provided that the particle dissolves and does not form a leaching product layer.

Thus, the associated general leaching model is:

])1(1[]

)3

4(

4[)(

3/2

y

libvmpliblibvmp rkrt α

π

π−−≈ (18)

In scenarios involving the formation of a product film, the pore diffusion model can be similarly

applied from the liberated particle leaching case with the same conversion factor.

1.4 Fully-Locked Particle Leaching

A common model for diffusion controlled leaching through a porous reaction product layer or

shrinking core is:

])1(1[)( 3/2

,,32

2

,

, PDlockPDlock

effx

hrplockPD

PDlibvmpDC

rkrt αα −−−= (19)

This equation is the same as given for the liberated particle leaching through a porous product layer

except that the radius of host rock particle is used instead of the valuable mineral particle radius,

and the other constants are specific to the fully-locked leaching scenario.

1.5 Ore Leaching

Leaching of an ore that consists of liberated, partially-locked, and fully-locked valuable mineral

particles is the sum of the leaching of the valuable mineral grain distributed in these three catego-

ries. Thus, overall ore leaching is determined by summing the contributions of leaching in each cat-

egory that are weighted by their relative proportions. The overall leaching was determined using an

Excel®

spreadsheet and the accompanying goal seek function. The goal seek function was used to

find the fraction reacted that corresponded with a specified time. The resulting fraction reacted val-

ues were weighted based on the probability of occurrence and summed to determine the overall frac-

tion at a specified time. The resulting data was used to construct overall fraction reacted versus time

plots.

Free


2 Experiments

Leaching experiments were performed using columns approximately 10 inches in height and 2 inch-

es in inner diameter. Monosize fractions of chalcopyrite ore (150-500 grams) were used together

with a solution volume approximately twice the weight of the ore sample. Solution was maintained

at specified pH levels using a pH controller and an acid pump using a reservoir of 1 M sulfuric acid.

Reagent grade chemicals and ASTM Type I water were used in the experiments. All column leach-

ing experiments were performed by flowing solution up from the bottom of the column in saturated

flow at a rate of approximately 100 ml/min. All experiments were performed at room temperature.

Acid consumption was measured based on acid reservoir depletion as a function of time. Note that

because monosize fractions of particles were used and the valuable mineral particle size is less than

the host rock particle size, liberated particles were presumed to have passed through the screens

during sample preparation and were therefore not included in this analysis.


3.1 Effect of pH

Leaching of an ore that consists of liberated, partially-locked, and fully-locked valuable mineral par-

ticles is the sum of the leaching of the valuable mineral grain distributed in these three categories.

Thus, overall ore leaching is determined by summing the contributions of leaching in each category.

Figure 2 shows the fit of the model to measured data at pH 1.5 and pH 2.5. The fit of the model data to

the measured data is reasonable but not excellent. Because the ore contains a variety of minerals and

the model fit assumes one predominant mineral, the fit of the data was not expected to be excellent.

Figure 2: Comparison of measured and calculated fraction reacted and time for ore samples

leached at pH 1.5 and 2.5 as indicated. The ore particles were -14 + 20 Mesh.

The model parameters were r*vmp = 0.004 cm, klib = 3.84×10

6 sec/cm,

kPD,lock/Deff = 3.2×109 cm·sec/mol.

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 500000 1000000

Fra

cti

on

Re

acte

d

Time (sec)

Model pH 1.5

Measured pH 1.5

Model pH 2.5

Measured pH 2.5



3.2 Effect of size

The ability of the model to predict the effect of size is presented in Figure 3. Figure 3 shows the

model generally fits the data. The fit of the data to the smaller size fraction is significantly better

than the fit for the larger size fraction.

Figure 3: Comparison of measured and calculated fraction reacted and time for ore samples of the

average size indicated in microns (µm). The ore particles were – 14 + 20 Mesh.

The model parameters were r*vmp = 0.004 cm, klib = 1.35×10

6 sec/cm,

kPD,lock/Deff = 1×109 cm·sec/mol.

Although the model fits the data generally, there are several issues that need to be addressed to pro-

vide more confidence in this modeling approach. The ore that was used has not been analyzed by a

mineralogical assessment to evaluate the extent to which other minerals may affect the results which

have assumed a single mineral or a group of similar minerals is responsible for acid consumption.

Thus, further work is needed, and the work presented in this paper is a work in progress in its early

stages of evaluation.

4 Summary

The model discussed in this paper that can be used to classify the association of acid consuming par-

ticles into liberated, partially-locked, and fully-locked particles relative to host-rock particles provides

a valuable framework for evaluating leaching. The model provides new equations that can be used to

easily estimate the fraction of liberated, partially-locked, and fully-locked valuable mineral particles.

The information from the model can be used to weight the relative influence of leaching models, one

of which is new, to determine overall leaching performance. Results indicate this approach shows po-

tential for improving acid consumption predictions for leaching operations. Additional mineralogy

assessments before and after leaching are needed to further evaluate the performance of the model.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 500000 1000000

Fra

cti

on

Re

acte

d

Time (sec)

Model 820 microns

Measured 820 microns

Model 1420 microns

Measured 1420 microns

Free


Acknowledgements

The laboratory work of Abraham Jurovitzki, Alysha Bhide, and Prashant Saraswat who performed

the experiments needed to provide the data in this paper are gratefully acknowledged along with

partial funding by the U. S. Department of Energy Center for Advanced Separation Technology and

the University of Utah Research Opportunities Program for undergraduate students.

List of symbols

r* particle reference size at which 62.3 % of the material passes as undersize when s = 1,

s index of the Weibull or Rosin-Rammler distribution function

P probability or fraction of particles smaller than size

r radius of particle

k constant

α fraction reacted

y coefficient

C concentration of reactant

D diffusivity

subscripts

vmp valuable mineral particle

hrp host rock particle

x reactant

Eff effective

PD pore diffusion

Lib liberated

Lock indicates locked mineral particle



References

[1] JANSEN, M. & TAYLOR, A. (2003). Overview of gangue mineralogy issues in oxide copper

Heap leaching,

http://www.altamet.com.au/Technical%20Papers%20and%20Articles/ALTA%20Copper/

Overview%20of%20Gangue%20Mineralogy.pdf, Accessed October 16, 2009.

[2] KELLY, E. & SPOTTISWOOD, D. (1982). Introduction to Mineral Processing, John Wiley,

New York, 26.

[3] FREE, M. (2008). Canadian Metallurgical Quarterly, 47(3), 277-284.

[4] MILLER, J., LIN, C., GARCIA, C., & ARIAS, H. (2003). International Journal of Mineral

Processing, 2003, 72, pp. 331–340.



Copper from Pyrite – A Short History

Fathi Habashi

Laval University

Department of Mining, Metallurgical, and Materials Engineering

Quebec City, G1V 0A6, Canada

Keywords: Pyrite, roasting, leaching, extraction, history

Abstract

Pyrite, known since antiquity, had at one time a great strategic importance because it was the main

raw material for making elemental sulfur for gunpowder manufacture and for making SO2 for sul-

furic acid production. Pyrite-bearing ores usually contains appreciable amounts of copper that was

recovered by a variety of methods, now obsolete. A historical review will be given for the RioTinto,

Orkla, and Duisburger Kupferhütte processes. However, when other sources of cheap elemental

sulfur became available, pyrite not only lost its importance but also became a nuisance for the me-

tallurgical industry because of problems associated with its disposal.

1 Introduction

Pyrite (Figure 1) was mentioned by the Greek and Roman writers. It is the most common sulfide

mineral and is widely associated with other metal sulfide deposits. It had at one time a great

strategic importance because it contains approximately 53 % sulfur and was the main raw material

for making elemental sulfur for gun powder and for making SO2 for sulfuric acid manufacture. In

the sixteenth century, pyrite was heap leached in the Harz mountains in Germany and in Río Tinto

mines in Spain to recover its copper content and in the nineteenth century pyrometallurgical

processes were developed.

Habashi


Figure 1: Naturally-occurring pyrite cubes Figure 2: French translation of Johann

Friedrich Henckel’s book on py-

rite published in Paris in 1760

The importance of the mineral can be judged from a book published in 1725 in Leipzig by Johann

Friedrich Henckel (1679-1744) on the mineralogy of sulfide minerals entitled Pyritologia, oder

Kiess-Historie. The book was translated in English in 1757 and in French in 1760 (Figure 2). Full

title: Pyritologia, oder Kiess-Historie (Leipzig 1725) English translation: Pyritologia or a History

of the Pyrites, the principal body in the Mineral Kingdom, in which are considered its names, spe-

cies beds, and origin; its iron, copper, unmetallic earth, sulphur, arsenic, silver, gold, original par-

ticles, vitriol, and use in smelting. In 1907 the French chemist P. Turchot published another com-

prehensive book on pyrite.

2 Copper from Pyrite by Heap Leaching

In these operations, the pyrite-bearing ores were piled in the open air and left for months to the ac-

tion of rain and air whereby oxidation and dissolution of copper took place. A solution containing

copper sulfate was drained from the heap and collected in a basin. Metallic copper was then precipi-

tated from this solution by scrap iron, a process that became known as “cementation process”. Mi-

croorganisms were certainly active in catalyzing the leaching process but this became known only in

the 1960s. When pyrite was fully leached it was loaded on trucks and shipped to sulfuric acid manu-

facturers.

3 Copper from Pyrite Cinder

Pyrite was a major source for generating SO2 needed for sulfuric acid manufacture:

2 FeS2 + 11

/2 O2 → 4 SO2 + Fe2O3 (1)

A variety of roasters were specially developed for this purpose. The first design in 1850 by the Brit-

ish inventor Alexander Parkes (1813-1890) (Figures 3 and 6). A later design by the American



engineer John Brown Francis Herreshoff (1851-1932) (Figure 4) is shown in Figure 7. Multi-stage

roasters were later replaced by the more efficient fluidized bed reactors invented in Germany in

1922 by Fritz Winkler (1888-1950) (Figures 5 and 8).

Figure 3: Alexander Parkes

(1813-1890)

Figure 4: John Herreshoff

(1851-1932)

Figure 5: Fritz Winkler

(1888-1950)

Table 1 shows typical analysis of cinder. A method was patented in England in 1844 by W. Long-

maid to purify the cinder for shipping to the steel industry and at the same time to recover the non-

ferrous metals present. It was first applied by William Henderson of Scotland in 1859. Lower cop-

per content in the pyrite especially since World War I compelled the firm to extract other products

from pyrite. In the process developed, known as Longmaid-Henderson process, the pyrite cinder

was roasted with sodium chloride then leached with water to recover the nonferrous metal chlorides.

The technology was adapted in Germany for over a century at the Duisburger Kupferhütte in Duis-

burg from 1876 to 1982 (Figure 9).

Figure 6: The first pyrite

roaster, 1850

Figure 7: Herreshoff

multi-hearth

furnace

Figure 8: Fluidized bed reactors for

roasting pyrite concen-

trates

Habashi


Table 1: Typical analysis of pyrite cinder

% ppm

Iron 54-58 Cobalt 300-1500

Gangue 6-10 Nickel 10-1500

Sulfur 2.5-4 Manganese 300-3000

Copper 0.8-1.5 Silver 25-50

Zinc 2.0-3.2 Gold 0.5-1.5

Lead 0.3-0.5 Cadmium 40-100

Thallium 15-45

Indium 1-50

Figure 9: Duisburger Kupferhütte plant in Duisburg, Germany operated from 1876 to 1982 to

process pyrite cinder

In this plant, pyrite was imported from all over the world by the company, sold to acid manufactur-

ers in Germany on the agreement that the iron oxide resulting from roasting, called cinder, is

shipped back to Duisburg for further treatment to recover nonferrous metals, precious metals, and

metallic iron (Figure 10).



Figure 10: Importing pyrite from worldwide suppliers to Duisburger Kupferhütte plant

The pyrite cinder was mixed with NaCl and heated continuously in a multiple hearth furnace at

800 °C to transform nonferrous metals into water-soluble chlorides (Figure 11). Each batch requires

about 2 days for leaching in vats (Figure 12). Copper was precipitated from solution in two steps

(Figure 13):

Habashi


Figure 11: Salt roasting of pyrite cinder Figure 12: Leaching plant for purple ore

Cu + CuCl2 → Cu2Cl2(ppt) (2)

Cu2+

+ Fe → Cu + Fe2+

(3)

Copper for the first step was obtained from the second step. It is interesting to note that the blue

CuCl2 solution turns pink after reduction and precipitation of Cu2Cl2 due to the presence of Co2+

in

the remaining solution. Also the residual Cu+ is colorless. Cuprous chloride recovered is then

treated with calcium hydroxide to precipitate copper (I) oxide which was reduced with coal in a fur-

nace to black copper:

Cu2Cl2 + Ca(OH)2 → Cu2O + CaCl2 + H2O (4)

2Cu2O + C → 4Cu + CO2 (5)

The black copper was cast into anodes and refined electrolytically; the precious metals were col-

lected in the anodic slimes. The solution obtained after cementation is evaporated under vacuum to

recover Na2SO4·10H2O (Figure 14). In 1975, the production of black copper was abandoned in fa-

vor of a hydro-electowinning process: cuprous oxide was leached in recycle acid and the CuSO4

solution obtained was electrolyzed to produce copper cathodes.



Figure 13: Cementation plant

Habashi


Figure 14: Sodium sulfate plant



The residue, called “purple ore”, now a high-grade iron ore (61–63 % Fe), is sintered (Figure 15)

and delivered to the blast furnace (Figure 16). Since lead and silver form chlorides during roasting

which are insoluble in the leaching step, they remained in the purple ore. When the sintered purple

ore was charged in the blast furnace, lead-silver alloy is formed. Being insoluble in iron and has a

higher density it settles at the bottom of the hearth. The furnace was provided with an opening be-

low the iron notch to tap the lead-silver alloy once a week. Figure 17 gives a general flowsheet for

the recovery processes. Some plant production data are given in Table 2.

Table 2: Data on Duisburger Kupferhütte plant

■ Imported 3 million tonnes of pyrite annually ■ 60,000 tonnes Zn

■ Processed 2 million tonnes of cinder - 50 tonnes Ag

■ Employed 4100 people - 70 tonnes Cd

■ Produced annually: - 10 tonnes Tl

- 1.2 million tonnes pig iron - Minor amounts of Co, Au, In, Pt

- 6000 tonnes Pb - 170,000 tonnes sodium sulfate

- 24,000 tonnes Cu ■ Consumed 200,000 tonnes NaCl

Figure 15: Sintering plant

Habashi


Figure 16: Blast furnace plant producing pig iron and lead-silver alloy



Figure 17: General flowsheet of operations at Duisburger Kupferhütte

4 Pyrite and the steel industry

Noting the success of Duisburger Kupferhütte, producers of sulfuric acid in Europe, Japan, and

USA became interested to market their finely ground iron oxide produced by roasting pyrite to the

steel industry. Although iron oxide pelletization was invented in Sweden in 1912, it was not intro-

duced in the iron ore industry until thirty years later. This took place when the electric arc furnace

was introduced in the steel industry and the need arose for a palletized feed. This gave an incentive

to the sulfuric acid manufacturers to upgrade their pyrite cinder for the steel industry by removing

copper and other nonferrous metals. As a result, more processes were developed to deal with this

problem.

■ Kowa-Seiko Process. This is a Japanese process developed at Kitakyushu in which the cinder is

mixed with calcium chloride, pelletized, then heated in a rotary kiln at 1100 °C to volatilize non-

ferrous metal chlorides. These are scrubbed in water from the exit gases and the solution treated for

metal recovery. The process was also used in Portugal and in other countries.

■ Bethlehem Steel process. At the plant at Sparrows Point in Maryland, cobalt from the cinder was

recovered by a process based on a careful temperature control during the roasting of pyrite. If the

temperature is kept at 550 °C, cobalt in the pyrite will be converted to sulfate and therefore can be

leached directly from the cinder with water. The hot pyrite cinder was quenched with water to give a

Habashi


slurry containing 6–8 % solids. When the solids are filtered off, the solution contains 20–25 g/L Co;

it is processed further for metal recovery. In the cinder, the Fe:Co ratio is 50:1; in solution it is 1:1.

This plant supplied the only domestic source of cobalt in USA.

■ Outokumpu process. In Finland at the Outokumpu plant in Pori a similar process is in operation in

1979; the sulfated pyrite cinder contains 0.8-0.9 % Co and other nonferrous metals. It is leached

with water to get a solution at pH 1.5 analyzing 20 g/L Co, 6-8 Ni, 7-8 Cu, 10-12 Zn, and trace

amounts of iron, which is treated for metal recovery.

5 Copper from Pyrite by Smelting

Pyrite smelting was developed to melt massive sulfide ore to form matte, and at the same time to

recover the excess sulfur in the elemental form. It was first successfully operated in 1928 by Orkla

Grube in Norway (Figure 13). Similar operations were in Sweden, Portugal, Spain, and Russia. Py-

rite containing about 2 % Cu is mixed with coke, quartz, and limestone and heated in a blast furnace

(Figure 14). In the upper part of the furnace, one atom of sulfur in pyrite is distilled as elemental

sulfur. In the oxidizing zone, FeS formed is oxidized to ferrous oxide and SO2. In the middle part of

the furnace, the reduction zone, SO2 is reduced by coke to elemental sulfur which is volatilized as

vapor. The reactions taking place can be represented by the following equations:

Upper zone: FeS2 → FeS + 1/2 S2 (6)

Oxidation zone: FeS + 3/2 O2 → FeO + SO2 (7)

Middle zone: SO2 + C → CO2 + 1/2 S2 (8)

Carbon disulfide and carbon oxysulfide are formed in the furnace; they are converted to elemental

sulfur on catalytic beds. The matte produced contains 6-8 % Cu and is usually re-smelted with coke,

silica, and limestone to 40 % Cu.



Figure 18: Location map of Orkla plant Figure 19: Orkla process for treating pyrite

concentrates to recover copper and

elemental sulfur

6 The Decline in the Pyrite Industry

During the Napoleonic wars, Spanish pyrite entered into competition with Sicilian sulfur in many

markets. The production of pyrite, however, declined gradually towards the end of the nineteenth

century after the discovery of an economic method for the recovery of sulfur from the sulfur domes

in the Gulf of Mexico. There was also a great rise in petroleum refining activities after World War II

that resulted in large amounts of sulfur-containing refinery gases and the need to recover this sulfur

to avoid polluting the environment with SO2. Also the availability of large volumes of natural gas

containing hydrogen sulfide at Lacq in southern France in 1950s and in Alberta, Canada in 1970s

contributed to the decline in pyrite demand.

Elemental sulfur replaced pyrite as a source for SO2 for sulfuric acid manufacture because of the

purity of the gas generated and the elimination of dust recovery equipment in the plant. Sulfur be-

came available by the following processes:

• Sulfur deposits discovered in the Gulf of Mexico were exploited economically since 1895 by

Herman Frasch (1851-1914) (Figure 20) using superheated water to melt the sulfur and float it to

the surface by compressed air (Figures 21 and 22). The process applies only when sulfur is stratified

between impervious rock formation which was not the case for Sicily. The success of Frasch

process ruined the Sicilian sulfur industry.

Habashi


Figure 20: Herman Frasch

(1851-1914)

Figure 21: Extracting sulfur

by Frasch process

Figure 22: Storage of

Frasch sulfur

• H2S-containing natural gas or petroleum refining gases not only became a source of elemental sul-

fur but recovering this sulfur solved an environmental problem since the gases were burned generat-

ing large amounts of SO2. Hydrogen sulfide must first be separated from the gases by an absorption

– desorption process (then oxidized by a controlled amount of oxygen at 400 oC on aluminum oxide

or bauxite bed using Claus reaction discovered in 1883 by the German chemist Carl Friedrich Claus:

H2S + ½ O2 → S + H2O (9)

As soon as these new sources of elemental sulfur became available, pyrite roasters for sulfuric acid

manufacture were dismantled and replaced by sulfur burners and the decline in pyrite production

started to decline rapidly. Acid plants based on pyrite roasting were expensive because it included

bulky equipment for dust separation.

Suggested Readings

[1] Anonymous, “Upgrading Pyrite Cinders for Iron and Steel Production. Montedison Chlorina-

tion Process Removes and Recovers Nonferrous Metals and Arsenic”, Sulphur 106, 52-57

(1973) May/June [The British Sulphur Corporation]

[2] W. Greiling, “75 Jahre Duisburger Kupferhütte, 1876-1951”, Chemiker-Zeitung p. 577 (1951)

[3] E. Guccione, “The Recovery of Elemental Sulfur from Pyrite in Finland”, Chemical Engineer-

ing (New York), February 1966

[4] F. Habashi, Researches on Copper. History & Metallurgy, Métallurgie Extractive Québec,

Québec City 2009. Distributed by Laval University Bookstore « Zone »



[5] F. Habashi, “The Recovery of Elemental Sulfur from Sulfide Ores”. Proceedings of “XXXVI

Congrés international de chimie industrielle”, Brussel, Belgium, 1966 (pub. 1967). Ind.

Chim. Belge 32 (Special Issue) (Part 2), 250-258 (1967). Also available as Bulletin 59 (1967)

from Montana Bureau of Mines & Geology, Butte, Montana

[6] F. Habashi, “The Recovery Empire Built on Fool’s Gold,” Eng. & Ming. J. 170 (12), 59–64

(1969)

[7] W. Haynes, Brimstone. The Stone That Burns, Van Nostrand, Princeton, New Jersey 1959

[8] H.Kudelka, R.M. Dobbener, and N.L. Piret, “Copper Electrowinning at Duisburger Kup-

ferhütte”, CIM Bulletin 186-197 (1977) August

[9] I. Maeshiro, „Recovery of Valuable Metals from Black Ore“, pp. 315 – 324 in Proceedings of

the Eleventh Commonwealth Mining and Metallurgical Congress, Hong Kong, published by

the Institution of Mining and Metallurgy, London 1978

[10] H.J. Nowacki, “Die Aufbereitung kupferhaltiger Schlämme und Fällprodukte im Schachto-

fen”, Z. Erzbergbau u. Metallhüttenw. 22(1), 22 – 28 (1969)

[11] C.Nuñez et al., „A Non Integral Process for the Recovery pf Gold, Silver, Copper and Zinc

Values Contained in Spanish Pyrite Cinders”, pp. 328 – 337 in XV Congrés International de

Minéralurgie, volume 3 (1985)

[12] W. Teworte, Duisburger Kupferhütte, company brochure published in Duisburg, W. Germany

1957

[13] E. Wiklund, „Die Nutzung von Pyrit als Grundlage für die Produktions von Schwefelsäure

und Eisenoxyd”, Aufbreitungs- Technik (6),285-289 (1977)



Copper Crud Treatment, Concentration –

Dependent Pond Depth Adjustment for Decanter

Centrifuges, DControl®

Dipl.-Ing.Tore Hartmann, Dr. Ulrich Horbach, Jens Kramer

GEA Westfalia Separator Process GmbH

Werner-Habig-Straße 1

D-59302 Oelde, Germany

Keywords: Hyrdometallurgy, copper, centrifuge, crud treatment

Abstract

In the hydrometal process, the presence of the so-called “crud” is a constant challenge in solvent

extraction. Crud is a stable emulsion which forms along the interface between the aqueous and or-

ganic phase. The spread is influenced by the following parameters: first, wind blows the dust and

impurities into the open sedimentation tanks. Second, the undissolved solids such as sand trans-

ported in the PLS cause problems in conjunction with incorrect agitator design.

The crud fraction can decisively impact the efficiency of the hydrometal extraction because the

phase interface can constitute a large fraction and the sedimentation tanks cannot react flexibly to it.

In the downstream process of the series-connected sedimentation tanks, they are thus all contami-

nated with crud. At the same time, the necessary mass transport is significantly impeded at the phase

boundary between organic phase and aqueous phase due to the crud formation.

The transfer of the organic phase into the electrolysis cell can result in a “burnout” of the cathode.

The carry-over of this electrolyte into the extraction can cause problems with the pH regulation. The

carry-over of the organic components into the raffinate also leads to contamination of the leaching

circuit.

The continuous treatment of the crud with a 3-phase decanter centrifuge is extremely effective in

combating this. This technology splits all three phases from each other and they are consequently

continuously separated. All subsequent process steps exhibit a stable, uniform effectiveness. The

main advantage for the customer is that fluctuations in the process are eliminated and the organic

phase can be recirculated back to extraction. The recycling of the solvent alone justifies the invest-

ment in less than 6 months as examples from South Africa and Chile show.

Hartmann, Horbach, Kramer


To operate solvent recovery in daily operation fully automated online at the optimum limit, the de-

canter centrifuge is equipped with a concentration-dependent pond adjustment called DControl®

. By

this means, we can guarantee our customers the maximum possible solvent recovery and minimise

the solvent costs in this process step. This system is presented in detail in this paper.

1 Introduction

The number of applications for continuous centrifugation in the hydrometal process is rising con-

tinuously. This development is characterized by a number of clear customer demands in which de-

canters are clearly superior to other competing technologies.

This customer demand relating to productivity, process specifications, reliability and profitability

can be achieved and even surpassed. The demands of different production facilities involved in the

hydrometal process always vary because the respective composition of the ore requires a unique

process sequence. Consequently, application-specific demands are realized in close cooperation

with the customer. Preventive service concepts are a basic pre-condition for constant plant availabil-

ity which ensures the customer the leading position in hydrometallurgy.

Figure 1: Process schematic of crud formation

Copper Crud Treatment with 3 Phase Centrifuge


One example of this is the recovery of copper in an SX installation. Discharged insoluble solids lead

to crud formation.

Figure 2: Phase distribution in copper recovery

This can lead to a situation where aqueous components are transported into the recovery process

during recycling or the valuable organic phase. This can result in increased chloride contents when

the ores contain Cl– or if seawater is used for the process. This, in turn, can lead to pitting on the

higher storage tank or on the stainless steel agitator.

For this reason, the crud phase cannot be recycled back into the process. Disposal is, however, like-

wise uneconomic because on the one hand the organic phase is a valuable operating fluid and, on

the other hand, dissolved hydrometals are to be recovered in the aqueous phase. The crud phase

must therefore be split up into the three phases solids, aqueous and organic phase.

Previously, different chemical processes were used for this with which the emulsion was split.

However, it has been demonstrated that continuous centrifugation is the most effective and eco-

nomical technique.

The crud from a SX installation is discharged into the crud tank. From there it is conveyed into the

centrifuge which separates the aqueous components and the solids from the organic phase. The liq-

uid and solids are recycled into the raffinate basin, the organic material is recycled back into the SX



installation. The main purpose of this operating mode is to relieve the installation from the emulsion

and to recover the expensive organic solution.

To a certain extent, the formation of crud depends on how far the particles exhibit hydrophobic

properties.

2 Centrifugal technology in the three-phase decanter

During centrifugation, centrifugal force is produced through the rotation of the bowl which is nor-

mally a multiple of the force of gravity. This is expressed in terms of the so-called g-force which

represents the ratio between centrifugal acceleration and acceleration due to gravity. A relative ve-

locity is applied to the heavy components (particles, drops) of the feed suspension as a function of

the g-force.

The decanter is generally used to separate solids from a liquid (two phases). Three-phase systems

can also be processed with a special version. In addition to separating the solids, the liquid is sepa-

rated into a light and a heavy phase.

Figure 3: Design of a three-phase decanter

Figure 3 is a schematic representation of the design of a three-phase decanter. The separation and

dewatering of the suspended particles, which generally exhibit the highest density of the individual

components, is analogue to other decanter centrifuges by sedimentation on the inner bowl wall. The

two liquid phases are separated simultaneously under the action of centrifugal force. The fluids

build up layers in the bowl in accordance with their density. The phases are separated independently



from one another at the cylindrical end of the bowl. The respective phases are either discharged un-

der gravity over a weir or under pressure by means of a centripetal pump. In the case of the variant

illustrated in Figure 1, the heavy phase discharges under gravity, the light phase by means of a cen-

tripetal pump [1].

In the system presented here, the overflow diameter of the aqueous phase is firmly defined and can-

not be altered online. The diameter of the organic phase corresponds to the pond depth and is nar-

rower than the overflow diameter of the aqueous phase due to the difference in density between the

two liquids. Between the two layers is a boundary layer at the liquid-side end of the bowl, the so-

called separating zone. The position of the separating zone within the sedimentation pond depends

on the liquids to be processed and the overflow diameters. This is illustrated in Figure 4 where the

position of the separating zone is plotted over the pond depth. The parameter is the overflow diame-

ter of the heavy phase.

Figure 4: Position of the separating zone as a function of the pond depth

To realize a separating process that is as effective as possible, the position of the separating zone

must be adapted to the given phase distribution in the feed suspension which may be subject to tem-

poral variations depending on the raw materials and the process management. A system is therefore

desirable that permits the online displacement of the separating zone as a function of a suitable con-

trol parameter.



3 Density dependence of the organic phase

In the process under consideration here, the organic phase is primarily a valuable substance which

should be recycled into the process without impunities if possible. A potential contamination origi-

nates largely from the aqueous phase. There is a density difference between these two liquids. A

contaminated light phase accordingly has a higher density compared to the pure organic liquid. This

is graphically represented in Figure 4.

Figure 5: Density of the contaminated organic phase

The density of the valuable substance therefore serves as an indicator for the contamination of the

organic phase.

4 System description

The system illustrated here is illustrated in Figure 6. The decanter is configured as a three-phase

decanter as described above. The aqueous phase is discharged via the bowl outer diameter into a

catcher and flows off via the decanter frame. The organic phase is discharged under pressure by

means of a centripetal pump.



Figure 6: Schematic illustration of the concentration-dependent pond adjustment

In addition to the standard pressure gauge, a density measuring device is also integrated in this dis-

charge. In the control unit the measured density signal is evaluated in accordance with Figure 5 to

determine contamination through the aqueous phase and is compared with a limit value.

By way of example, Figure 7 shows a possible phase distribution in the decanter bowl. The organic

phase (yellow) accounts for a very small proportion so that there is a short retention time in the cen-

trifugal field. As a result, heightened contamination of the organic phase can occur which is de-

tected with the density sensor. If the admissible limit value is exceeded, the valve is actuated ac-

cordingly resulting in a higher discharge pressure. As the discharge pressure rises, the immersion

depth of the centripetal pump in the pond increases, which means that the pond depth likewise in-

creases. In accordance with the relation described above, the separating zone is displaced outwards.

This state is illustrated in Figure 8.



Figure 7: Position of the separating zone without increased discharge pressure

Figure 8: Position of the separating zone with increased discharge pressure

As a result of the displacement of the separating zone outwards, the volume of the organic phase in

the bowl increases and the retention time in the centrifugal field is longer. The result is an im-

provement in the degree of purity of the organic phase due to the externally induced increase of the

discharge pressure.



Figure 9: Separating zone diameter as a function of the discharge pressure

If the discharge pressure is increased too sharply, a particularly pure organic phase is recovered but

there is an increased loss of the organic phase into the aqueous phase and possibly also the solids. In

this case, the operator can optimally adjust the parameters throughput capacity, yield of organic

phase and degree of purity of the organic phase to the specific operating situation.

5 Summary

By means of the DControl®

system described, we can guarantee our customers the maximum possi-

ble solvent recovery and reduce the solvent costs to a minimum in this process step. The economy

of the entire process is increased since the transfer of the organic phase into the electrolysis cell and

also the contamination of the leaching circuit is minimised. The separated aqueous phase and the

solids can be recycled into the raffinate basin by which means the installation can be optimally re-

lieved from the emulsion and the valuable recovered organic solution from the SX installation can

be fed in again. The process demands of the customer with regard to productivity, reliability and

profitability can be achieved and even surpassed with this system.

References

[1] STAHL W (2004): Industrie Zentrifugen, Fest-Flüssig Trennung Band II 2004, page 707

ISBN 3-9522794-0-4

[2] HARTMANN T., CORBELLA J.(2007): Tailor made crud treatment with 3 phase separating centri-

fuge, Copper , Toronto, Canada, COM 2007



A Specific Electrode for “On-line” pH

Measurement in SAG Cleaner Flotation Circuits

Christian Hecker C., Alejandro González S., Alejandra Mejías J. Claudia Rodríguez F.

Hecker Electroquímica Industrial S.A Codelco Chile

Sargento Bernardo Cuervas 093 División El Teniente

Rancagua, Chile Gerencia de Plantas, Chile

Keywords: Alkaline media, electrochemical potential, copper concentrate, wireless measurement

Abstract

Commercial pH electrodes do not allow an on-line control of lime milk dosing in high flow pulp

minerals circuits, when pH of the slurry must be above 12. Glass membrane conventional pH elec-

trodes do not work in strongly alkaline high flow mineral pulp media.

Nevertheless, experiments developed in a SAG cleaner flotation circuits have confirmed the capa-

bility of a new specific electrode to follow on-line, changes of pH in high alkaline slurry as a func-

tion of copper concentrate pulp flow. The behavior of this specific electrode has shown a linear re-

sponse with volumetric titration on strongly alkaline pulps.

The aim of this paper is to show the performance of the specific electrode and his capability to con-

trol on-line lime milk dosage as a function of the copper concentrate mineral pulp that are fed into

the column cell.

1 Introduction

Lime slurry is a widely chemical reagent used in ore flotation processes to control pH in sulphide

copper mineral processing. It is used as a pyrite depressant agent in the cleaning flotation circuits.

In cleaner flotation circuit where it is necessary to work in strong alkaline media, (pH > 12), plants

don´t have an on-line dosage of lime milk, leading to an over-consumption of this reagent. In fact,

glass electrodes do not work to measure on-line pH on strongly alkaline high flow mineral pulp, for

on-line measuring pH.

The main purpose of this paper is to present the behaviour of a new specific electrode and his capa-

bility to measure “on-line”, the electrochemical potentials in strongly alkaline high flow mineral

Hecker, González, Mejías, Rodríguez


pulp media. The electrochemical potentials are proportional to pH changes in the mineral pulps and

on-line control of the selective flotation process would be possible.

2 Materials and methods

2.1 Titration assays

The assays were conducted by taking 100 ml of vacuum filtered solution from alkaline mineral pulp

and phenolphthalein as an indicator reagent. Solution HCl 0.1 N was used as titration reagent. This

solution is added drop by drop into the filtered sample solution and mixed by using a magnetic stir-

ring device.

2.2 Electrode assemble

A metallic rod electrode was manufactured and assembled with an Ag/AgCl reference electrode,

and a ground electrode, into a cylindrical isolate body. Electrical signals were conducted by com-

mercial coaxial wires [2].

2.3 Field installation

The assembled electrode was placed into a high flow mineral pulp open conduit, by a suitable pivot

gear system. The gear system allows to exposing permanently the electrode assembled device into

the high flow mineral pulp ensuring continuous measurements. Additionally, the assembled elec-

trode device was connected to an ORP analyzer and potentiometric measurements were sent by

GPRS system to a PC. Figure 1 shows a flow sheet of the mineral pulp open conduit and the whole

pH measurement system.

A Specific Electrode for “On-line” pH Measurement in SAG Cleaner Flotation Circuits


Figure 1: A flow sheet of the mineral pulp open conduit and the whole pH measurements system

2.4 Data acquisition

Mineral pulp and lime slurry addition rate were acquired from PI-System server. Electrochemical

potential measurements of the specific electrode were sent by GPRS wireless system and

downloaded to a PC. All information was processed using an appropriated worksheet.


Potential measurements on strongly alkaline pulps had a linear response that is proportional to

volumetric titration, as it is shown in Figure 2 [1]. Characteristic values of process titration in SAG

cleaner flotation circuit are 28 ± 4. The specific electrode potential gap reached in this work was

-400 to -350 mV referred to the Ag/AgCl reference electrode. This result provides a real tool for on-

line equivalent titration control by using specific electrode potential measurements.

Hecker, González, Mejías, Rodríguez


Figure 2: Correlation between electrode potential and volumetric titration

In order to validate the specific electrode in the SAG high flow concentrate mineral open conduit,

the electrochemical potential of the electrode and the mineral pulp volumetric flow/lime milk volu-

metric flow ratio against time, is presented in Figure 3. On-line electrochemical potential measure-

ments (dark blue line), properly trends the volumetric ratio of mineral pulp phase/lime milk slurry

rate, as it showed with sky blue line.

Our last plants results point towards that it exist a close response between both variables. Hence, it

is possible to ensure a capability of the specific electrode to measure on-line the electrochemical

potential in the SAG high flow concentrate mineral open conduit, with a very high accurately.

R2 = 0,89

-500

-450

-400

-350

-300

-250

8 11 13 16 18 21 23 26 28 31 33

E /m

V/(

Ag/A

gC

l)

Volumetric Titration / ml

A Specific Electrode for “On-line” pH Measurement in SAG Cleaner Flotation Circuits


Figure 3: Potential electrode and volumetric ratio trends against time

4 Conclusion

1. Experiments developed in the electrochemical laboratory and in copper cleaner flotation plant

has demonstrate the capability of a new specific electrode to measure on-line the electrochemical

potential in the SAG high flow concentrate mineral open conduit, with a very high accurately.

2. New experiences are developing in order to establish an integrate mechanism control for on-line

dosing lime milk as a function of copper concentrate mineral rate flow feed in column flotation

cells, maintaining a stable titration level.

3. Improve in the metallurgical recovery of Cu and Mo sulphide mineral species in cleaner flotation

circuits are expected with an on-line control of lime milk dosage by using a metallic specific

electrode.

References

[1] HECKER C.; GONZALEZ A.; MEJIAS A.; BUSTAMENTE A.; ALARCÓN E.; ZAPATA V.

Patente de Invención. Presentada en el INAPI – Chile. Solicitud Número 2035-2009.

[2] BUSTAMANTE A., GONZALEZ A. (2009): “Desarrollo de un sensor de pH para circuitos de

limpieza en medio fuertemente alcalino, en flotación colectiva“.



Redox Potential Control in Column Leaching of

Chalcopyrite

Naoki Hiroyoshi, Yuki Takehara, Masami Tsunekawa Takenari Kuwazawa

Hokkaido University Hokkaido University

Graduate School of Engineering Graduate School of Engineering

(Present: Nippon Mining & Metals Co.,Ltd.)

Keywords: Chalcopyrite, leaching, redox potential, oxidation, reduction

Abstract

Column leaching experiments for a chalcopyrite/silica rock aggregate (1 wt. % Cu) were performed

using a 0.1 m x 1 m height cylindrical column at 303 K. The sulfuric acid solutions (pH 1.5) con-

taining known concentrations of Fe3+

and Fe2+

was supplied to the top of the column and the effect

of redox potential on copper extraction was investigated.

When 0.04 M Fe3+

/ 0.06 M Fe2+

solution (the redox potential, about 0.65 V vs. SHE) was used,

more than 25 % of copper was extracted after 25 days and the extraction reached 75 % at 90 days.

The extraction with 0.1 M Fe3+

solution (the redox potential, > 0.8 V), on the other hand, was less

than 10 % at 25 days. This indicates that lower redox potential is better for leaching chalcopyrite,

and that the redox potential control is effective to enhance the leaching rate. The effect of redox

potential on chalcopyrite leaching was also confirmed in the depth profile of copper concentration

and redox potential in the column.

1 Introduction

Chalcopyrite, CuFeS2, is a common copper mineral and the development of a hydrometallugical

route to produce copper from this mineral has been an important object in copper extractive metal-

lurgy [1-3]. A major problem in the development of chalcopyrite hydrometallurgy is extremely slow

leaching kinetics of this mineral. Elevated temperature, high pressure, special brine or catalyst like

silver, may be available to improve the leaching rate in a small-scale tank reactor operation for chal-

copyrite concentrate, but it is difficult to use such extreme conditions or expensive chemicals in a

large-scale heap leaching operation for low grade copper ores.

Heap leaching is operated with a huge ore piles (typically several km2 in area and several m in

height) using dilute sulfuric acid at ambient temperature. Operating conditions of heap leaching are

Hiroyoshi, Kuwazawa, Takehara, Tsunekawa


relatively moderate: temperature inside the ore pile increases due to the exothermal leaching reac-

tion but it is not over 70 oC; the solution pH range is limited within 1.5-2.5 to maximize the effi-

ciency of the succeeding solvent extraction-electrowinning process; an oxidant Fe3+

is generated in-

situ by microbiological oxidation of Fe2+

extracted from the ore, but the Fe3+

concentration is not so

high (usually less than 0.1 kmol m-3

). Heap leaching for chalcopyrite must be developed under such

limited conditions.

The leaching rate of chalcopyrite in sulfuric acid solutions depends on the redox potential deter-

mined by the concentration ratio of Fe3+

to Fe2+

in the solution [4-19]: The rate increases with in-

creasing the redox potential and reaches the maximum at a certain values of the potential, then de-

creases at higher potentials. This indicates that chalcopyrite leaching can be optimized by

controlling the redox potential.

The simplest way to control the redox potential in heap leaching is to regulate the chemical compo-

sition of the solution used to irrigate the ore pile. In this method, the concentration ratio of Fe3+

/Fe2+

is maintained at a suitable value for chalcopyrite leaching, and the solution is supplied to the ore

pile. In this study, the effects of the redox potential control on chalcopyrite leaching were demon-

strated in column leaching experiments.

2 Materials and Methods

2.1 Mineral sample and solution preparation

A ground chalcopyrite sample (80 % passing size, 74 µm) was prepared from a massive specimen

(Copper Queen Mine, Arizona, USA). Silica (SiO2), sphalerite (ZnS), and molybdenite (MoS2) were

detected as the major impurity minerals in a X-ray diffraction patern of the sample. Chemical analy-

sis of the ground sample indicated that the sample contains 24.7 % Cu, 25.4 % Fe, 29.5 % S,

2.54 % Ca, 1.91 % Si, 1.04 % Zn, 0.76 % Mg, and 0.16 % Al.

A high purity silica specimen (Nellore, Andhra Pradesh, India) was crushed, and the -13.2 mm

+2.8 mm size fraction obtained by sieving was used as a support rock for preparing the chalcopy-

rite/silica aggregate used in the column leaching experiments.

Solutions used in this study were prepared with pure water (ion-exchange/distilled water) and re-

agent grade chamicals such as H2SO4, Fe(SO4)·7H2O, Fe2(SO4)3·nH2O, and CuSO4·5H2O obtained

from Wako chemical Co. Ltd.

2.2 Shaking flask leaching experiments

To evaluate the redox potential dependence of the chalcopyrite sample, shaking flask leaching ex-

periments were conducted using the ground chalcopyrite sample and the solution containing

0.01 kmol m-3

H2SO4 and 0.1 kmol m-3

Fe3+

. A 500 cm3 Erlenmeyer flask containing 2 g of the

Redox Potential Control in Column Leaching of Chalcopyrite


ground chalcopyrite and 200 cm3 of the solution were shaken reciprocally at 298 K under atmos-

phere in a thermostat water bath shaker (shaking rate and amplitude, 120 min-1

and 40 mm). At

regular time intervals, 2 cm3 of the solution was obtained as sample for Cu and Fe concentration

analysis by an inductively coupled plasma atomic emission spectroscopy (ICP-AES). The solution

pH and solution redox potential (ORP) were also measured using a pH electrode and a combination

ORP electrode (Pt working electrode combined with a KCl saturated Ag/AgCl reference electrode).

The measured value of ORP was converted to the value against standard hydrogen electrode, and

the converted value (E) is used in the results and discussion part. The concentrations of Fe3+

and

Fe2+

were calculated from the total Fe concentration and the redox potential value E, assuming the

following material balance equation and Nernst equation for the Fe3+

/Fe2+

redox pair.

[Total Fe] = [Fe3+

] + [Fe2+

] (1)

E = 0.67 + 0.059 log ([Fe3+

]/[Fe2+

]) (2)


Column leaching experiments were performed on the chalcopyrite concentrate agglomerated on

supporting silica rock in a thermo-controlled room (temperature, 303 K) using an apparatus shown

in Figure 1. A 462 g sample of the ground chalcopyrite was agglomerated on 11.1 kg of the silica

rock particles in a rolling drum using 150 cm3 of pure water as a binder. The aggregate was charged

in 0.1 m diameter x 1 m height cylindrical PVC column. The H2SO4 solution (pH 1.5), containing

known concentrations of Fe3+

, Fe2+

, and Cu2+

,was stored in a reservoir tank and was supplied to the

column top at a flow rate of 2.0 dm3 day

-1 using a tube pump. The solution discharged from the bot-

tom of the column was collected in a drainage tank. At 3 or 4 day intervals, the solution in the

drainage tank was collected and the solution volume, the Cu and Fe concentrations, pH, and ORP

were measured. From the Cu concentration and the solution volume, the fraction of Cu extracted

from chalcopyrite was calculated.

Aside from the solution in the drainage tank, several cm3 of the solution discharged from the col-

umn was also collected to evaluate the solution composition at the bottom of the column.

The solutions at different height levels of the column were also collected from the sampling ports

located at the column side: a sponge block (size, about 1 cm x 1 cm x 2 cm) was fixed at one end of

the plastic bar and it was inserted into the chalcopyrite/silica aggregate ore layer through the sam-

pling port. After 1h, the sponge and the bar were taken out of the column and the solution collected

by the sponge was recovered by pressing it in a piston syringe. The solution was analyzed for Cu

and Fe concentration, pH, and ORP.



Figure 1: Leaching column.

3 Results and Discussion

3.1 Redox potential dependence of chalcopyrite leaching rate

To demonstrate the redox potential dependence of the leaching rate for the chalcopyrite concentrate

sample used, batch-wise leaching experiments (shaking flask experiments) were conducted. Figure

2 shows the copper extraction and redox potential as a function of time in the shaking flask leaching

experiments. The experiments were repeated three times and all data Showed that the redox poten-

tial decreased and Cu extraction increased with time.

In the batch experiments, Fe3+

in the solution is consumed and Fe2+

is formed by the following reac-

tion.

CuFeS2 + 4 Fe3+

= Cu2+

+ 5Fe2+

+ 2S (3)

Therefore, the Fe3+

concentration decreases and the Fe2+

concentration increases with time, resulting

in a reduction of the redox potential, E, determined by the Nernst equation in Equation 2. If we can

simply assume that the leaching rate of chalcopyrite at a given time is determined by the redox po-

tential at that time, the relation between Cu extraction rate and the redox potential can be deter-

mined from the redox potential vs. time plot and the Cu extraction vs. time plot.

2.0L / day

Pum

p

Drainage tank

Solution reservoir

Column with sampling port（0.1m interval）



Figure 2: Cu extraction and redox potential as a function of time in flask shaking experiments for

a ground chalcopyrite sample.

From the data shown in Figure 2, the relation between Cu extraction rate, K, and redox potential, E,

was evaluated. As shown in Figure 3, Cu extraction rate increases with increasing the redox poten-

tial, and reaches a maximum at around 0.6 V, and then the rate decreases at higher potential.

This result is in line with the literature that indicates that there is an optimum redox potential for

chalcopyrite leaching [4-6, 8, 12-15, 18-19]. However, chalcopyrite leaching rate is not only deter-

mined by the redox potential; i.e. the optimum redox potential shifts when Cu2+

concentration

changes [5-6, 12-13]. The authors have proposed a normalized redox potential, Enormal, as an indica-

tor to compensate for the effect of Cu2+

concentration [5-6, 12-13]. The normalized redox potential,

Enormal, is calculated from the concentrations of Cu2+

, Fe3+

, and Fe2+

using the following equation.

(4)

As shown in Figure 3, the leaching rates are higher at Enormal between 0 ~ 1, and the maximum rate

is achieved at Enormal around 0.5.

0.00

0.01

0.02

0.03

0.50

0.55

0.60

0.65

0.70

0.75

0.80

0 10 20 30 40 50

Red

ox p

ote

ntial,

E/ V

vs. S

HE

Time / h

Cu e

xtr

acte

d/k

mo

l m-3



Figure 3: Cu extraction rate as a function of redox potential and normalized redox potential.


The result shown in section 3.1 implies that chalcopyrite leaching would be improved by controlling

the redox potential, E, or the normalized redox potential, Enormal. In heap leaching, the simplest way

to control these potentials is to regulate the composition of the solution supplied to the surface of

the ore pile. To confirm the effects of solution composition and redox potential control on Cu ex-

traction during chalcopyrite leaching, column leaching experiments were carried out.

In the most of the column leaching experiments reported in literatures, the solution discharged from

the bottom of the column is re-circulated back to the top of the column directly. This “circulating

method” may have been selected due to its similarity with the actual heap leaching operation except

for the absence of SX/EW process to recover the extracted Cu from the solutions. In the circulating

method, however, the solution composition and redox potential change with time due to the leach-

ing reaction in the column.

To maintain the constant composition of the solution without using a special apparatus, a non-

circulating method was used in the present study (Figure 1). In this method, the solution is simply

supplied from a reservoir tank and the discharge from the bottom of the column is not charged back

to column.

In the first series of the experiments, Fe3+

/ total Fe ratio ([Fe3+

]/[T.Fe]) in the solution supplied to

the column was varied, and the effects of the ratio on chalcopyrite leaching were investigated (Fig-

ure 4). The solution supplied to the column contained 0.1 kmol m-3

Fe3+

(without Fe2+

) in Experi-

ment 1, and 0.04 kmol m-3

Fe3+

and 0.06 kmol m-3

Fe2+

in Experiment 2.

The fraction of extracted Cu was much higher in Experiment 2 ([Fe3+

]/[T.Fe] = 0.4) than in Ex-

periment1 ([Fe3+

]/[T.Fe] = 1); Cu was extracted continuously and reached over 25 % after 25 days

in Experiment 2, but the Cu extraction stopped at around 5 % after 10 days in Experiment 1. This

indicates that Cu extraction from chalcopyrite is strongly affected by the composition of the solution

supplied to the ore, and that high Fe3+

/total Fe concentration ratio suppresses chalcopyrite leaching.

0.00000

0.00001

0.00002

0.00003

0.5 0.55 0.6 0.65 0.7

Cu e

xtr

actio

n rate

, K/ m

ol h

-1m

-2

Redox potential, E / V vs.SHE

0.00000

0.00001

0.00002

0.00003

-0.50 0.00 0.50 1.00 1.50

Cu e

xtr

actio

n rate

, K/ m

ol h

-1m

-2

Normalized redox potential, Enormal / -



The values of the redox potentials, E and Enormal, were much higher in Experiment 1 (E > 0.7 V,

Enormal > 1.5) than in Experiment 2 (E = 0.60~ 0.65 V, Enormal = 0.5 ~ 1.0). Referring to Figures 3

and 4, the values of E and Enormal in Experiment 1 are excessively high for leaching chalcopyrite at a

fast rate, but the values in Experiment 2 are at a range suitable for leaching. This may be the reason

why larger Cu extraction was obtained in Experiment 2 than in Experiment 1.

The column leaching experiments with Fe3+

/total Fe ratio of 0.4 (sulfuric acid solution containing

0.04 kmol m-3

Fe3+

and 0.06 kmol m-3

Fe2+

) was continued until 90 days. The results are shown in

Figures 5 and 6.

To evaluate the effect of Cu2+

addition on the leaching, after day 27, 0.02 kmol m-3

Cu2+

was added

in the solution supplied to the column top.

Figure 4: The effect of the Fe3+

/total Fe concentration ratio on Cu extraction, redox potential, and

normalized redox potential in the column leaching experiments for a ground chalcopy-

rite/silica rock aggregate using sulfuric acid solutions containing 0.1 kmol m-3

total

soluble Fe (pH 1.5). The Fe3+

/total Fe ratio was 1 in Experiment 1 and 0.4 in Experi-

ment 2. The values of redox potential, E, were measured using the solutions obtained

from the top and bottom parts of the column. The values of Enormal were shown only for

the solution discharged from the bottom of the column, because the values cannot be de-

fined in the solution supplied to the column top due to the absence of Cu2+

.

The fraction of extracted Fe was larger than that of Cu, indicating that Fe extraction is preferred

under the operating condition used in these experiments. Fraction of Cu extraction vs. time plot can

be fitted with a straight line in the period of 0 ~ 50 days, indicating that Cu extraction rate is almost

constant (~1 % / day) in this period, and that the Cu2+

addition started at day 27 did not affect the Cu

extraction rate. On the other hand, Fe extraction rate decreased clearly after the 27th

day, implying

that Cu2+

addition suppresses Fe extraction. As a result, the Fe2+

concentration at the bottom of the

column began to decrease at the 27th

day and it caused increases in the redox potential E and the

normalized redox potential Enormal at the bottom. At day 27, the values of E and Enormal were near the

optimum values for chalcopyrite leaching (E = 0.6 V, Enormal = 0.5), but they deviate from the opti-

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0 10 20 30

Fra

ctio

n o

f Cu e

xtr

acte

d /

-

Time / day

Exp.1 Exp.2

0.50

0.60

0.70

0.80

0.90

1.00

1.10

1.20

0 10 20 30

Red

ox p

ote

ntial,

E /

V v

s. S

HE

Time / day

Exp.1 (column top)

Exp.1 (column bottom)

Exp.2(column top)

Exp.2(column bottom)

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

0 10 20 30

No

rmalize

d redox p

ote

ntial,

En

orm

al/ -

Time / day





mum values after day 27, and Cu extraction rate slowed down after 50 days. The Cu extraction was

about 75 % at day 90. This value is almost the same as that reported by Petersen et al. (2002) for the

bio-column leaching of chalcopyrite using a thermophile at elevated temperature (~70 oC), indicat-

ing that the redox potential control by the solution composition regulation is effective to improve

the chalcopyrite leaching.

Figure 5: The result of a column leaching experiment for a ground chalcopyrite/silica rock aggre-

gate using sulfuric acid solutions containing 0.04 kmol m-3

Fe3+

and 0.06 kmol m-3

Fe2+

.

The concentration of Cu2+

in the solution supplied to the column top was 0 kmol m-3

at

day 0 ~ 27, and 0.02 kmol m-3

at day 28 ~ 90. The vertical broken lines in each panel

mark the 27th

day when Cu2+

addition was started.

Figure 6 shows the vertical distribution of Cu concentration, redox potential, E, and normalized

redox potential, Enormal, in the leaching column at day 22 and day 72. Because Cu2+

was not added at

day 22, the concentration of Cu at the column top (height 1.0 m) was zero, and the concentration at

day 72 was 0.02 kmol m-3

due to the Cu2+

addition in the solution. At both day 22th

and 72th

, the Cu

concentration increases with decreasing height. Neglecting the irregular data at the column top

(1.0 m), bottom (0 m) and at 0.8 m, distributions of Cu concentration at upper part of column

(height 0.5 ~ 1.0 m) was different from that of the lower part of the column (0 ~ 0.5 m): the gradient

of Cu concentration against height was smaller at the upper part but was larger at the lower part.

This indicates that Cu extraction rate is higher at the lower part of the column (0 ~ 0.5 m). The re-

dox potential, E, and the normalized redox potential, Enormal, at the lower part were closer to the

0.00

0.20

0.40

0.60

0.80

1.00

1.20

0 50 100

No

rmalize

d redox p

ote

ntial,

En

orm

al / -

Time / day

Column bottom

0.55

0.60

0.65

0.70

0 50 100

Red

ox p

ote

ntial,

E/ V

vs. S

HE

Time / day

Column top

Column bottom

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 50 100

Fra

ctio

n o

f Cu e

xtr

acte

d /

-

Time / day

Cu

Fe

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0 50 100

Fe

2+co

ncentration /

km

ol m

-3

Time / day

Column top

Column bottom

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0 50 100

Fe

3+co

ncentration /

km

ol m

-3

Time / day

Column top

Column bottom

1.2

1.3

1.4

1.5

1.6

1.7

1.8

0 50 100

pH

Time / day

Column top

Column bottom



optimum values for chalcopyrite leaching (E = 0.6 V, Enormal = 0.5) than that at the upper part. This

might be the reason why higher Cu extraction rate was obtained at the lower part of the column.

Figure 6: Cu concentration, redox potential, and normalized redox potential as a function of

height in the column leaching experiments for a ground chalcopyrite/silica rock aggre-

gate using sulfuric acid solution containing 0.04 kmol m-3

Fe3+

and 0.06 kmol m-3

Fe2+

.

Dotted vertical lines in the figure indicate the optimum values of redox potential and

normalized redox potential for chalcopyrite leaching.

As demonstrated in this study, the control of the chemical composition and redox potential of solu-

tion used for irrigation is effective to improve chalcopyrite leaching. However, the solution compo-

sition changes due to the leaching reaction in ore pile and this makes it difficult to maintain the re-

dox potential at an optimum value for the leaching in all positions of the ore pile. If we can develop

the methods to maintain the redox potential at any depth of the ore pile, further improvement in

chalcopyrite leaching may be achieved.

4 Conclusion

The effects of the control of chemical composition and redox potential of the solution used for irri-

gation in the column leaching of chalcopyrite/silica rock aggregate were investigated. The results of

the preliminary batch leaching experiments for ground chalcopyrite showed that copper extraction

rate from chalcopyrite depends on the redox potential determined by Fe3+

/Fe2+

concentration ratio,

and that the maximum rate is achieved near the potential of 0.6 V vs. SHE, and the rate decreased at

higher redox potential. The results of column leaching experiments showed that higher copper ex-

traction rate was achieved when using 0.04 kmol m-3

Fe3+

/ 0.06 kmol m-3

Fe2+

solution in compared

with the Cu extraction rate obtained using 0.1 kmol m-3

Fe3+

. This is in line with the results of the

batch leaching experiments, and over 75 % of copper was extracted after 90 days when the solution

containing 0.04 kmol m-3

Fe3+

/ 0.06 kmol m-3

Fe2+

solution was used. The effect of redox potential

on chalcopyrite leaching was also confirmed in the depth profile of copper concentration and the

redox potential in the column.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.0 0.5 1.0 1.5

Heig

ht /

m

Normalized redox potential, Enormal/-

Day 22th

Day 72th

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.55 0.60 0.65 0.70

Heig

ht /

m

Redox potential, E / V

Day 22th

Day 72th

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.00 0.01 0.02 0.03 0.04

Heig

ht /

m

Cu concentration / kmol m-3

Day 22th

Day 72th



References

[1] N. Pradhan, K.C. Nathsarma, K. Srinivasa Rao, L.B. Sukla, B.K. Mishra ,” Heap bioleaching

of chalcopyrite: A review”, Minerals Engineering, Vol.21, No. 5, 2008, 355-365

[2] C. Klauber,”A critical review of the surface chemistry of acidic ferric sulphate dissolution of

chalcopyrite with regards to hindered dissolution”, International Journal of Mineral Process-

ing, Vol. 86, No.1-4, 2008, 1-17

[3] H.R. Watling The Bioleaching of Sulphide Minerals with Emphasis on Copper Sulphides —

A Review, Hydrometallurgy, Vol. 84, No. 1-2, 2006, 81-108.

[4] H. Kametani and A. Aoki, “Effects of Suspension Potential on the Oxidation Rate of Copper

Concentrate in a Sulfuric Acid Solutions”, Metallurgical Transactions B, 16B, 1985, 695-705.

[5] N. Hiroyoshi, M.Tsunekawa, H.Okamoto, R.Nakayama,S.Kuroiwa,“Improved chalcopyrite

leaching through optimization of redox potential“, Canadian Metallugical Quartery, Vol.47,

No. 3, 2008, pp.253-258.

[6] Naoki Hiroyoshi, Hiroko Kitagawa, Masami Tsunekawa, ”Effect of solution composition on

the optimum redox potential for chalcopyrite leaching in sulfuric acid solutions”, Hydrometal-

lurgy, Volume 91, Issues 1-4, 2008, Pages 144-149

[7] N. Hiroyoshi, S. Kuroiwa, H. Miki, M. Tsunekawa, T. Hirajima,”Effects of coexisting metal

ions on the redox potential dependence of chalcopyrite leaching in sulfuric acid solutions”,

Hydrometallurgy, Volume 87, Issues 1-2, June 2007, Pages 1-10

[8] N. Hiroyoshi, S. Kuroiwa, H.Miki, M. Tsunekawa and T. Hirajima, “Synergistic Effect of

Cupric and Ferrous Ions on Active-Passive Behavior in Anodic Dissolution of Chalcopyrite in

Sulfuric Acid Solutions, Hydrometallurgy, Vol.74, No. 1-2, 2004, 103-116.

[9] N. Hiroyoshi, H. Miki, T. Hirajima and M. Tsunekawa, ”Enhancement of Chalcopyrite Leach-

ing by Ferrous Ions in Acidic Ferric Sulfate Solutions”, Hydrometallurgy, Vol. 60, No. 3,

2001, 185-197.

[10] N. Hiroyoshi, H. Miki, T. Hirajima and M. Tsunekawa, “A Model for Ferrous-Promoted

Chalcopyrite Leaching”, Hydrometallurgy, Vol.57, No.1, 2000, 31-38.

[11] N. Hiroyoshi, M. Hirota, T. Hirajima and M. Tsunekawa, “ A Case of Ferrous Sulfate Addi-

tion Enhancing Chalcopyrite Leaching”, Hydrometallurgy, Vol. 47, No. 1, 1997, 37-45.

[12] H. Okamoto, R. Nakayama, N.Hiroyoshi, and M.Tsunekawa, “Redox Potential Dependence

and Optimum Potential of Chalcopyrite Leaching in Sulfuric Acid Solutions”,Journal of

MMIJ, Vol.120, No.10, 2004, 592-599.



[13] H. Okamoto, R. Nakayama, S. Kuroiwa, N.Hiroyoshi, and M.Tsunekawa, “Normalized Redox

Potential Used to Assess Chalcopyrite Column Leaching”, Journal of MMIJ, Vol.121, No.6,

2005, 246-254.

[14] A. Pinches, P. J. Myburgh, and C. Merwe: Process for the rapid leaching of chalcopyrite in the

absence of catalysis. Patent No: US 6,277,341 B1, 2001.

[15] K. A. Third, R. Cord-Ruwisch, and H. R. Watling,”Control of the redox potential by oxygen

limitation improves bacterial leaching of chalcopyrite. Biotechnology and Bioengineer-

ing,Vol. 78, No.4, 2002, pp.433-441.

[16] E. M. Cordoba, J. A. Munoz, M. L. Blazquez, F. Gonzalez and A. Ballester, “Leaching of

chalcopyrite with ferric ion. Part III: Effect of redox potential on the silver catalyzed

process”,Hydrometallurgy, Vol.93, No.3-4, 2008, pp.97-105.

[17] J. Vilcaez, R. Yamada, C. Inoue,”Effect of pH reduction and ferric ion addition on the leach-

ing of chalcopyrite at thermophilic temperatures”, Hydrometallurgy, Vol. 96, No.1-2,2009, pp.

62-71.

[18] J. Petersen and D.G.Dixon,” Competitive bioleaching of pyrite and chalcopyrite”, Hydro-

metallurgy, Vol. 83,No. 1-4,2006,pp. 40-49.

[19] G.Viramontes-Gamboa, B.F. Rivera-Vasquez and D.G. Dixon,”The active-passive behavior

of chalcopyrite - Comparative study between electrochemical and leaching responses”, Jounal

of the Electrochemistry Society, Vol. 154, No. 6, 2007, pp. C299-C311.



Separation Characteristics of Chalcopyrite and

Pyrite via Bench Scale Flotation Investigations

S. Kelebek, S. Reeves Z. El Jundi

Queen’s University Suncor Energy

The Robert M. Buchan Department of Mining Oil Sands Operations

Goodwin Hall, Kingston P.O. Box 4001, Fort McMurray

ON, K7L 3N6, Canada Alberta, T9H 3E3, Canada

H. Özdeniz

Selçuk University

Department of Mining Engineering

42079 Kampus, Konya, Turkey

Keywords: Flotation, chalcopyrite, pyrite, ethyl xanthate, lime, sensitivity to process variables

Abstract

Separability of copper bearing minerals, primarily chalcopyrite, from pyrite has been investigated

using bench scale flotation studies on relatively high grade ore samples up to 5 % Cu. The effects of

grind size, pH and ethyl xanthate dosage have been examined. In general, the ore samples were

found to be a “good behaving” type in the sense that their metallurgical response to levels of these

variables indicated expected variations. For example, the separability was improved by a finer grind

size, a higher pH and a lower xanthate dosage. In general, it has been found that the recovery of

chalcopyrite was not sensitive to differences in these variables. This is in sharp contrast with the

behaviour of pyrite which indicated substantial variations in flotation kinetics depending on the ex-

perimental conditions. It has been demonstrated that the separation of the two minerals can be

achieved by taking advantage of this difference, i.e., by proper control of the variables affecting py-

rite flotation the most. The level of potassium ethyl xanthate has been found to be the most impor-

tant aspect in the chemical control of pyrite flotation. The low grade sample at 1.88 % Cu demon-

strated a superior upgrading ratio in the process, yielding a rougher-scavenger concentrate of 22 %

Cu at 95 % recovery. The concentrate grade reaches 32 % Cu with the high grade sample of 5 % Cu

at the same recovery level. An increase in the pyrite/chalcopyrite ratio did not appear to have any

negative impact on separation characteristics of chalcopyrite from pyrite.

Kelebek, El Jundi, Reeves, Özdeniz


1 Introduction

Chalcopyrite as the most abundant copper-bearing mineral, is part of many sulphide ores whether

the latter is a porphyry type of ore or a polymetallic massive sulphide type. This mineral is almost

always accompanied by some pyrite as an iron sulphide gangue. In majority of the complex sulphide

ores containing sphalerite and/or galena, chalcopyrite occurs together with pyrite [1]. The amount of

pyrite can vary from less than 1 % in some of the porphyry copper ores to a level greater than 20 %

in massive sulphides and usually results in separation problems depending on the degree of libera-

tion and dissemination.

The most common method of mineral separation in processing of sulphide ores is flotation. Regard-

less of the type of sulphide ore in which it is present, chalcopyrite is recovered as the first sulphide

mineral in many differential flotation circuits. This is because chalcopyrite is highly floatable even

without a collector. The collectorless flotation of chalcopyrite has been well documented in various

cases [2-5]. Since other sulphide minerals share these characteristics, this can sometimes represent a

selectivity issue as well. In order to show floatability, chalcopyrite requires a threshold pulp poten-

tial that is slightly oxidizing above 0 mV, SHE. This fundamental dependence of flotation on pulp

potential is also observed in collector-induced conditions [6]. Fortunately, in most cases, the redox

potential range observed under natural air entrainment of slurry does not present a serious problem

that requires a significant adjustment. The most common collectors used for copper sulphide flota-

tion are xanthates [7], usually with low hydrocarbon chains due to its high tendency to float. At high

pulp potentials (e.g., > 300 mV), the floatability chalcopyrite remains stable while other sulphides

such as pyrite start to show depression characteristics unless inadvertently activated by copper ions

originating from the copper-bearing minerals. A great majority of the published work on chalcopy-

rite and pyrite has been carried out on a small scale as a single mineral study as a model system to

find out their fundamental surface properties. Results of these studies need to be interpreted careful-

ly as they can sometimes be misleading. It is generally known that there are interactions among in-

dividual mineral components in the ore as well as the grinding media and that the differences in

head grade and/or in the pyrite/chalcopyrite ratio may affect the separation characteristics of these

minerals in the industrial ores. Given the complexity of flotation pulp conditions involving such ore

samples, a need for bench scale studies is quite obvious in order to attain meaningful information

indicative of separation characteristics in industrial practice.

Significant increase in copper prices 6-7 years ago encouraged development of new ore bodies. This

gave rise to bench scale characterization of processing behaviour for many ores. The current work

represents such a study on a pyrite-rich copper sulphide ore. The characterization involved assess-

ment of major variables such as particle size, pH level, collector dosage. Specific attention has been

given to separation efficiency between chalcopyrite and pyrite that was assessed through their rela-

tive flotation recoveries and the grade-recovery performance obtained.

Separation Characteristics of Chalcopyrite and Pyrite Using Flotation



2.1 Samples and preparation

Ore samples originated from the Canakkale region of Turkey (courtesy of Okyanus Madencilik Ltd.)

and had a top size of about 15 cm. They were put through a series of laboratory crushers, namely, a

jaw crusher, a gyratory crusher and finally a roll crusher that was followed by a vibratory Sweco

screen arranged in a closed circuit to reduce the particle size to nominally pass 10 mesh (< 1.7 mm).

The samples were kept in deep freezer overnight to avoid oxidation of sulphides. The minus

10 mesh sample was rotary split into charges of about 900 g. A representative subsample generated

from each was also taken and pulverized to pass 150 mesh (100 µm) as the head sample for chemi-

cal analysis.

Further size reduction was carried out using a laboratory rod mill (Denver). Test charges were

ground at about 60 % solids by weight with local tap water (plus lime) for various periods of time to

allow an estimation of grinding time to produce flotation feeds of various fineness of grind from an

F80 of 135 µm down to 75 µm.

2.2 Flotation test procedure

The ground slurry was transferred into a 2-l Denver laboratory flotation cell. Initial pH level was

measured, and when required, it was raised by additional use of lime. In selected tests, soda ash as a

pH regulator was used during both grinding and flotation for a comparison with the performance of

lime. Potassium ethyl xanthate (PEX) was introduced as a collector at the required dosage and con-

ditioned for two minutes. The initial addition of frother was made during the last thirty seconds of

the conditioning period. Subsequent additions of xanthate and frother were made stagewise between

the collection of individual concentrates. A total of five concentrates were skimmed off over a pe-

riod of 11 min.

2.3 Chemical analysis

The head samples were digested in aqua-regia and analyzed for copper, iron and sulphur. An atomic

absorption spectrophotometer (Perkin-Elmer, Model 2380) was used for metal analysis. The ana-

lytical weight for the concentrate products was about 1 g and 2-3 g for the tails. The amount of sul-

phur in flotation products was determined using a sulphur analyzer from LECO (Carbon and Dual

Range Sulphur Analyzer, Model SC-444DR) using appropriate standards consisting of pulverized

zinc sulphide.



Table 1: Results of Head Analysis

Sample % Cu % Fe % S

Comp 1 1.88 10.90 13.0

Comp 2 3.35 12.51 14.2

Comp 3 5.0 14.77 14.9


3.1 Behaviour of ore samples with 1.88 % Cu

This sample was estimated to have a chalcopyrite/pyrite ratio of 3.8 using results of chemical analy-

sis and mineralogical information. Results obtained on the grade-recovery of a rougher-scavenger

performance are plotted in Figure 1, which shows the impact of particle size and pH. It can be seen

that the concentrate copper grades and recoveries are distinctly higher at a finer grind size and high-

er pH level. The copper grade that is about 8 % at an F80 of 123 µm and pH 10 nearly doubles at a

60 % recovery when the particle size (F80) is reduced to 106 µm. This trend continues as the copper

grade drops with an increase in recovery. The improved grade and recovery performance is related

to more effective liberation of pyrite from chalcopyrite achieved with a finer grind. When the par-

ticle size is further reduced to an F80 of 89 µm, the concentrate grade goes up from 15 % Cu to 25 %

Cu at an equivalent recovery of 70 %. It should be noted that this significant increase in the copper

concentrate grade is coupled with the effect of increased pH as well.

Contribution of a higher pH (i.e., 11) to the improved grade and recovery is related to increased

concentration of OH- ions [8], which inhibits the action of collector due to competition as defined

by Barksky relation as well as transformation of the pyrite surface to hydroxides, primarily through

formation of ferric hydroxide. In addition, use of increased amount of lime increases the concentra-

tion of calcium ions in the flotation slurry, which tend to block active sites by adsorption on pyrite

contributing to its enhanced depression [9].

The differences in the grade-recovery performance are directly related to different levels of pyrite

rejection in flotation experiments. This aspect can be better appreciated by plotting the chalcopyrite

recoveries as a function of estimated pyrite recoveries. Figure 2 shows the chalcopyrite-pyrite flota-

tion selectivity corresponding to the results presented in Figure 1. Compared to a reference line for

non-selectivity (corresponding to equal recovery levels for both chalcopyrite and pyrite), all three

experiments show flotation selectivities in favour of chalcopyrite, at least for the initial stages (i.e.,

roughers) relative to subsequent stages (i.e., scavengers).



Figure 1: Grade-Recovery performance of the sample with 1.88 % Cu under various conditions of

particle size (F80 89-123 µm), pH (10-11) and PEX (11.1 g/tonne), MIBC (16.7 g/tonne)

Figure 2: Chalcopyrite recovery versus pyrite recovery corresponding to data in Figure 1

0

5

10

15

20

25

30

0 10 20 30 40 50 60 70 80 90 100

Cu

Gra

de

(%)

Cu Recovery (%)

F80 123 um, pH 10

F80 106 um, pH 10

F80 89 um, pH 11

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100

Ch

alc

op

yri

te

reco

very

(%

)

Pyrite recovery (%)

F80 123 um, pH 10

F80 106, pH 10

F80 89 um, pH 11

No selectivity line



However, it is clear that best selectivity is observed at the highest pH of 11 and at the finest grind

size (F80) of 89 µm, where 90-95 % chalcopyrite is obtainable at about 10-12 % pyrite recovery.

The pyrite data was obtained using mineralogical stoichiometry and results of chemical assays as-

suming that chalcopyrite and pyrite are the only copper bearing sulphide and iron bearing sulphide,

respectively. In order to check the extent of the validity of this approach, the amount of iron of the

products calculated from chalcopyrite and pyrite was combined the get the total and compared with

the total amount of iron obtained by assaying. Results of this comparison for this sample are shown

in Figure 3. The amount of iron obtained by assaying has turned out to be somewhat smaller than

that found by calculation. This might suggest a co-presence of an iron deficient copper sulphide

such as bornite and chalcocite in trace amount. Nevertheless, as can be noted from the graph, the

agreement between the two sets of data is reasonably good. Thus, the method used for estimation of

the amount of pyrite in these samples may be regarded as a valid approach for practical purposes.

Figure 3: Comparison of the amount of iron by calculation and of iron by actual assaying of the

flotation products



Figure 4: Flotation recoveries of pyrite as a function of time (with the sample having 1.88 % Cu)

Flotation recoveries of pyrite as a function of flotation time are shown in Figure 4, which indicates a

substantial variation from 20 % to 95 % at the end of 11 minutes depending on the experimental

conditions. While the pyrite flotation is highly sensitive to grind size and pH level, the kinetic trends

of chalcopyrite under same conditions are quite similar (see examples given later). It is this charac-

teristic difference that makes the production of high grade copper concentrates possible with these

ore samples.

3.2 Behaviour of ore samples with 5.0 % Cu

According to calculations, this high grade sample was estimated to have a chalcopyrite/pyrite ratio

of 1.3 (assuming that its sulphide mineralogy is similar to the lower grade one tested). The effect of

particle size on flotation performance of this sample was also assessed. Figure 5 shows the copper

grade-recovery performances obtained for all three tests carried out at various F80 values. Several

points can be made here. First of all, upgrading is significantly better when compared to the pre-

vious case with the lower grade sample. The concentrate grade decreases systematically as the feed

becomes coarser. Best results are obtained at the finest grind size with an F80 of around 76 µm and

at a high pH of around 11. Under these conditions, the overall concentrate grade is 32 % at 95 %

copper recovery. This is a highly remarkable response for a primary stage flotation based on the

presence of chalcopyrite as the only copper bearing mineral.

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10 12

Pyri

te R

eco

very

(%

)

Time (min)

F80 123 um, pH 10

F80 106, pH 10

F80 89 um, pH 11




particle size (F80 76-135 µm), pH (11) and PEX (20 g/tonne), MIBC (20 g/tonne)

X-ray diffraction analysis of the combined concentrate from this test did not indicate presence of

any other copper mineral. However, presence of some other copper bearing minerals such as bornite

and chalcocite at a level below the detection limit of XRF is still likely. It is also notable that the

concentrate grade-recovery curves converge at about 55 % copper recovery. This suggests that the

copper bearing mineral(s) in this sample are liberated well and this liberation is not sensitive to var-

iations of F80 between 76 and 135 µm. In this case, it can be conservatively stated that about 50 %

of copper bearing minerals are highly liberated at a grind size corresponding to an F80 of 135 µm.

However, when the grind size is reduced to an F80 of 76 µm, the cumulative mass of particles libe-

rated (hence recovered) increases to nearly 95 %.

Copper flotation recoveries corresponding to these three tests are shown as a function of time in

Figure 6. It is immediately noticeable that all recovery data are represented practically by a single

line. This observation is important in that it supports the well-liberated nature of chalcopyrite in this

ore discussed earlier.

Figure 7 shows the corresponding data on flotation kinetics for pyrite. As can be noted from the

figure, in contrast to the case with chalcopyrite, pyrite flotation is highly dependent on experimental

conditions of the tests especially in relation to particle size.

10

15

20

25

30

35

20 30 40 50 60 70 80 90 100

Cu

mu

lati

ve g

rad

e, C

u (%

)

Cu recovery (%)

F80 134 um

F80 104 um

F80 76 um



Figure 6: Cu recovery as a function of time with the 5.0 % Cu sample under various conditions of


Figure 7: Pyrite recovery as a function of time with the 5.0 % Cu sample under various conditions

of particle size (F80 76-135 µm), pH (11) and PEX (20 g/tonne), MIBC (20 g/tonne)

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10 12

Cu

Reco

very

(%

)

Time (Minutes)

F80 104 um

F80 134 um

F80 76 um



The finer the grind size (F80), the greater the degree of pyrite depression is in the presence of lime. It

is clear that the poorest grade of the copper concentrate in Figure 5 results from co-flotation of py-

rite at its highest rate. In general, the tests with this high grade sample (at 5 % Cu) have demonstrat-

ed very good separation characteristics as it is amenable to liberation of its chalcopyrite and pyrite

as the sulphide gangue at relatively coarse grind size. However, flotation of pyrite requires chemical

control by lime to maintain a high pH (> 10) and proper dosage of xanthate (e.g., starvation level).

Otherwise, it can ruin the quality of copper concentrates to be produced.

Additional tests were carried out using a mixture of the high grade and low grade samples to see if

there are any interactions that may influence the flotation separation characteristics between chalco-

pyrite and pyrite.

3.3 Behaviour of blended samples

For the purpose of these tests, the blending ratio was kept nearly at 1 to 1. Flotation charges were

prepared by repeated blending and and finally by riffling. As a result, a head grade at about 3.35 %

Cu was obtained. The testing involved essentially the same type of reagents as before with some

variation in their dosages. The grind size was also slightly different.



Figure 8 shows the copper grade-recovery response at three grind sizes (F80), namely, 68, 94 and

139 µm while the pH and potassium ethyl xanthate levels are fixed at about 11 and 18 g/tonne,



respectively. Although the F80 values were not identical to the previous case in Figure 5, the grade-

recovery performance shows expected trends with respect to grind size differences. It can be seen

that despite a lower head grade, high grade copper concentrates reaching 30 % Cu at 90 % recovery

are still obtainable at primary stage of separation as long as the pH level is sufficiently high and the

fineness of grind is sufficiently low.

Additional tests with these samples were carried out at lower pH values. The results indicated little

difference in upgrading behaviour of copper at pH 10 compared to 11 when the pH regulator was

lime. Use of soda ash for adjustment to pH 10 resulted in premature loss of selectivity. When the pH

level was further reduced (with less lime) to about 9, there was a significant deterioration of the

copper grade due to co-flotation of additional pyrite [10], which is expected on the basis of Barsky

relation [8].

3.4 Impact of head grade on flotation performance

In general, it is known that the grade-recovery performance is dependent on the head grade of the

ore tested. A higher grade feed typically yields a higher grade concentrate at an equivalent recovery.

This relationship has been explored with the present ores since some samples were available at dif-

ferent head grades. This also gave opportunity to examine the behaviour of the mid-grade blend in

comparison with the sum of individual behaviour of the high-grade and low grade samples based on

the same blend ratio. The grade-recovery curves obtained are shown in Figure 9. A separate screen

analysis performed for each case indicated some differences in the F80 values from 90 µm to 95 µm.

It is assumed that such differences do not interfere with the head grade effect significantly. As ex-

pected, the higher concentrate grades are associated with the higher grade feeds. The copper concen-

trate grade of 30.5 % obtained at an equivalent recovery of 95 % drops to 22 % Cu when the feed

grade drops from 5.0 % Cu to 1.88 % Cu. The chalcopyrite/pyrite ratio is believed to have played a

role in such a difference in performance (see later). It is interesting that the grade-recovery perfor-

mance of the blend at about 3.35 % Cu head grade can be predicted based on individual behaviour

of the high and low grade samples. This suggests that there are no significant interactions between

high grade and low grade components when procesed together. These components in the blend ap-

pear to act independently, typical of their individual flotation behaviour. It is speculated that a high

pH of 11 may have contributed to this independent behaviour. In some cases, especially when the

slurry pH has a lower value, an iron sulphide component in a blend has a much greater flotation

activity than its individual behaviour. Some of these phenomena are connected to inadvertent acti-

vation and oxidation effects that may be induced by galvanic interactions, which can result in signif-

icant losses of concentrate grade-recovery for the valuable metal [11].



Figure 9: Impact of head grade on grade-recovery performance of the samples (Grind size, F80 90-

95 µm; pH 11; and PEX & MIBC (at 18 g/tonne except for 1.88 % Cu with 12 g/tonne)

The grade-recovery performance of the lowest grade sample (1.88 % Cu) has produced the lowest

grade concentrate overall. However, a concentrate grade of 22 % Cu at 95 % recovery is still an im-

pressive performance for a rougher-scavenger stage considering its relatively low head grade. The

cumulative concentrate grade when divided by the head grade of a sample yields a so-called called

“upgrading ratio”, which is sometimes used in performance analysis of flotation circuits. The up-

grading ratios are shown for the current data in Figure 10. As can be noted, this way of plotting in-

dicates a superior performance for the low grade sample, which has switched its position with the

high grade sample. There may be several reasons for the superior upgrading of this sample including

one related to its mineralogical characteristics. Another important reason is likely related to use of

less xanthate (12 g/tonne) compared to other cases (18 g/tonne). An important mineralogical differ-

ence is that this low grade sample has a much greater pyrite/chalcopyrite ratio. Flotation selectivity

for these series of tests can also be examined by plotting copper recoveries as a function of iron re-

coveries. Although it involves the iron from chalcopyrite (CuFeS2), this is a direct approach com-

pared to the method plotting calculated pyrite recoveries (e.g., Figure 7). The results presented in

Figure 11 show that flotation selectivity for all three tests is significantly in favour of copper (com-

pared to non-selectivity line as a reference). However, the flotation selectivity is distinctly better for

the lower grade sample than the other two samples. It should be noted that the lower grade sample

has the highest pyrite (Py)/chalcopyrite (Cp) ratio of 3.8 versus 1.3 for the high grade sample.



Figure 10: Upgrading ratio of copper as a function of recovery (Grind size, F80 90-95 µm; pH 11;

and PEX & MIBC (at 18 g/tonne except for 1.88 % Cu with 12 g/tonne)

Thus, the amount of ethyl xanthate used per unit weight of pyrite present in this sample is even less

than the amount stated earlier. This xanthate level was apparently just sufficient for flotation of

chalcopyrite, but not sufficient to trigger flotation of a significant amount of pyrite. In a more de-

tailed study of an intermediate grade copper ore sample, the ethyl xanthate dosage (used at a range

of 10-20 g/tonne) was found to be more significant as a variable than the grind size (used in an F80

range of 135-76 µm) and pH level within the range of 9-11 [11]. It may also be noted from this fig-

ure that the selectivity between copper and iron for the blended sample can be mathematically pre-

dicted (denoted by triangles) based on individual flotation behaviour of the high and low grade

samples.

4 Summary

Flotation characteristics of a series of pyrite-rich copper ore samples were investigated in connec-

tion with selective separation chalcopyrite. The ore grade ranged from 1.88 % Cu to 5.0 % Cu with

the pyrite/chalcopyrite ratios varying from 1.3 to 3.8. The ore samples had excellent liberation cha-

racteristics. Three main parameters assessed through testing were:

0

2

4

6

8

10

12

14

0 10 20 30 40 50 60 70 80 90 100

Cu

up

gra

din

g r

ati

o

Cu Recovery (%)

5.0%Cu, F80 ~95 um

1.88%Cu, F80 ~90 um

3.35%, F80 ~93 um

3.36%Cu calc



Figure 11: Impact of head grade on flotation selectivity between copper and iron (for legend refer

to Figure 10)

1. fineness of grind with the F80 values ranging from 76 µm to 135 µm, 2. pH values mainly 10 to

11, and 3. Dosage of ethyl xanthate ranging from 10 g/tonne to 20 g/tonne. The dosage of frother

(MIBC) was kept close to that of xanthate. Copper recoveries obtained in a typical rougher-scavenger

flotation were high, typically above 95 %. The overall concentrate grades range from 22 % Cu to 32 %

Cu at 95 % copper recovery, depending on the feed grade. Pyrite was found to be highly floatable, and

difficult to control once exposed to pulp conditions that would trigger its flotation. The samples were

generally found to be “good behaving” in flotation since their metallurgical response to the levels of

these main variables indicated expected variations. In all cases, the chalcopyrite-pyrite separation

was improved by a finer grind size, a higher pH and a lower xanthate dosage. It has been found that

the recovery of chalcopyrite was not sensitive to differences in these variables. This is in sharp con-

trast with the behaviour of pyrite which indicated substantial variations in flotation kinetics depend-

ing on the experimental conditions. Separation of the two minerals has been achieved by taking ad-

vantage of this difference, i.e., by proper control of the variables affecting pyrite flotation the most.

The level of potassium ethyl xanthate has been found to be the most important aspect in the chemi-

cal control of flotation. The low grade sample had a superior upgrading ratio in the process. An in-

crease in the copper content of the ore or an increase in the pyrite/chalcopyrite ratio did not seem to

have any negative impact on separation characteristics of chalcopyrite from pyrite.

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50

Cu

re

co

very

(

%)

Fe recovery (%)

5.0%Cu,Py/Cp: 1.3

1.88%Cu, Py/Cp: 3.8

3.35%, Py/Cp: 2

3.36%Cu calc

non-selectivity line



It can be concluded that the chalcopyrite-pyrite ore samples tested respond to conventional treat-

ment using lime and ethyl-xanthate very well. The method allows production of high grade copper

concentrates even at the roughers-scavenger stage.

Acknowledgements

The authors would like to thank Mr. Eric Beck for supplying the ore samples and Mrs. M. Bailey for

help provided in analytical procedures and sample preparation during the course of these bench

scale investigations.

References

[1] MISRA, K.C. (1999): Understanding Mineral Deposits, Kluwer Academics Publishers, 845 p

[2] HEYES, G.W. and TRAHAR, W.J. (1977): The natural floatability of chalcopyrite, Int. J.

Mineral Processing, 4: 317- 344

[3] GARDNER, J. and WOODS, R. (1979): An electrochemical investigation of the natural floa-

tability of chalcopyrite”, Int. J. Mineral Processing, 6: 1-16

[4] YOON, R.H. (1981): Collectorless flotation of chalcopyrite and sphalerite ores by using

sodium sulfide, Int. J. Mineral Processing, 8: 31-48

[5] KELEBEK, S. and HULS, B. J. (1991): Collectorless flotation of chalcopyrite in the nickel-

copper ores from Sudbury Basin, COPPER-91, Mineral Processing and Process Control, Vol.

II, G.S. Dobby, S.A. Argyropoulos and S.R. Rao Eds., Pergamon Press, 171-186

[6] TRAHAR, W. (1984): Pulp potential in sulphide flotation, in: Principles of mineral flotation–

The Wark symposium, M.H. Jones and J.T. Woodcock, Eds. The Australian Institute of Min-

ing and Metallurgy, Melbourne, 117-135

[7] BULATOVIC, S.M. (2007): Handbook of Flotation Reagents: Chemistry, Theory and Prac-

tice, Flotation of Sulfide ores, Vol. 1, Elsevier, 446 p

[8] SUTHERLAND, K.L. and WARK, I.W. (1955): Principles of Flotation, Australasian Ints.

Min. Met., Melbourne, Australia

[9] GAUDIN, A.M. (1957): Flotation, 2nd edn. McGraw-Hill, New York.

[10] KELEBEK, S. (2009): An investigation of options for selective recovery of copper from a

pyrite-rich sulphide ore, Queen’s University, Mining Engineering, Kingston, ON, Canada.

[11] KELEBEK, S. and NANTHAKUMAR, B. (2007): Characterization of stockpile oxidation of

pentlandite and pyrrhotite through kinetic analysis of their flotation, Int. J. Mineral

Processing, 84: 69–80.



Bioleaching of Crude Chalcopyrite Ores by the

Thermophilic Archaean Acidianus brierleyi in a

Batch Reactor

Y. Konishi, N. Saitoh, M. Shuto, T. Ogi

K. Kawakita, T. Kamiya

Osaka Prefecture University Japan Oil, Gas and Metals National Corporation

Department of Chemical Engineering Metals Mining Technology Department

1-1 Gakuen-cho, Naka-ku 1310 Omiya-cho, Saiwai-ku

Sakai, Osaka 599-8531, Japan Kawasaki 212-8554, Japan

Keywords: Copper recovery, chalcopyrite, low-grade ores, acidophilic thermophile

Abstract

This paper describes the bioleaching of chalcopyrite by the thermophilic archaean Acidianus brier-

leyi at 65 °C and at pH 1.8-2.0 in a batch reactor. The thermophile A. brierleyi was able to solubilize

chalcopyrite much faster than the commonly used leaching mesophile Acidithiobacillus ferrooxi-

dans at 30 °C. About 65 % of the copper was leached from the crude chalcopyrite ore (size range

< 25 µm, Cu 1.2 %, Fe 20.4 %, S 2.4 %, SiO2 46.3 %), and the copper extraction rate was almost

the same as that from a chalcopyrite concentrate (Cu 22.5 %, Fe 27.8 %, S 31.4 %), suggesting that

the effect of gangue minerals such as magnetite and silica in the crude ore on the leaching was in-

significant. About 10 % iron was leached from the crude ore in 10 days in the presence or absence

of A. brierleyi. In other words, A. brierleyi selectively leached chalcopyrite while magnetite leach-

ing by A. brierleyi was negligible. We thus conclude that thermophilic bioleaching with A. brierleyi

is attractive as an economical and environmentally friendly process for good copper extraction from

low-grade primary sulfides.

1 Introduction

Most primary copper production utilizes terrestrial sulfide ores containing 0.5-2 % Cu. Typically,

copper sulfide concentrate obtained by froth flotation of the ores contains 20-30 % Cu and this is

used as a starting material in pyrometallurgical smelting. Due to the increase in copper consumption

high-grade terrestrial mineral resources have been depleted, and copper production from low-grade

and complex ores have become important. As a method to produce copper from low-grade and

complex ores, biohydrometallurgical processes have great potential.

Konishi, Saitoh, Shuto, Ogi, Kawakita, Kamiya


The use of microorganisms to facilitate the extraction of precious and base metals from mineral

resources is referred to as “bioleaching”. The bioleaching of terrestrial sulfide minerals has devel-

oped into a successful and expanding area of research in biotechnology [1, 2]. Bioleaching proc-

esses consume a little energy (i.e. are economically attractive) and are environmentally safe for the

recovery of copper from low grade sulfide ores. Figure 1 is a simplified schematic representation for

the role of biotechnology in the copper industry [3]. Engineering options for bioleaching have

evolved from heap leaching to stirred-tank leaching. Heap bioleaching is typically used for low-

grade sulfide ores containing 0.5 % or less copper and tank bioleaching is used mostly for the ex-

traction of copper from high-grade sulfides.

Figure 1: A simplified schematic representation of the role of bioleaching in the copper industry

(based on a figure presented by M.E. Clark et al. [3]).

Bioleaching of Crude Chalcopyrite Ores by the Thermophilic Archaean Acidianus Brierleyi


Figure 2: A simple flowchart of a copper heap bioleaching process.

Common microbes used in sulfide leaching are the mesophilic bacteria Acidithiobacillus ferrooxi-

dans and Leptospirillum ferrooxidans, which grow optimally at 20-30 °C and at pH 1-3. The ther-

mophilic archaea, Acidianus brierleyi and Sulfolobus acidocaldarius, which have pH optima of 1-2

and temperature optima for activity of 60-70 °C are also candidate microbes [4-6]. In the copper

industry, bioleaching processes have been used commercially for the recovery of copper from low-

grade secondary sulfides (Cu2S/CuS). Figure 2 shows a simple flowchart of a copper heap bioleach-

ing process. Copper in the ore is extracted as ions in the solution, recovered as metal by solvent ex-

traction and electrowinning. However, low recovery from primary copper sulfides, particularly from

chalcopyrite, and protracted leach cycles, has delayed the full commercialization of bioheap leach-

ing for low-grade primary sulfide ores.

Economically, chalcopyrite is the most important copper resource. Thermophilic archaea have been

shown to solubilize chalcopyrite concentrates much faster than the mesophiles. The use of A. brier-

leyi at 65 °C resulted in a greater than 90 % leaching of copper concentrate (particle size range of

38-53 µm) in 10 days [7, 8]. These results show that thermophile performance is more effective than

mesophile operating for the bioleaching of chalcopyrite concentrates. However, further work is re-

quired for the development of copper bioleaching from low-grade chalcopyrite ores because the



depletion of terrestrial high-grade sulfide ores will require that copper be recovered from low-grade

ores in the near future. This paper describes the bioleaching of crude chalcopyrite ores by the ther-

mophilic archaean Acidianus brierleyi at 65 °C and at pH 1.8-2.0 in a batch reactor.

2 Experimental Section

2.1 Mineral substrates and microorganisms

The mineral substrates used in this study were a natural chalcopyrite concentrate and crude ore, ob-

tained from the Atacama Mine, Chile. The natural minerals were ground and sieved to obtain a size

range of < 25 µm. The chemical composition of the concentrate was 23.5 wt. % copper, 28.6 wt. %

iron and 31.5 wt. % sulfur, respectively. The composition of the crude ore was 1.15 wt. % copper,

20.36 wt. % iron, 2.63 wt. % sulfur and 46.27 wt. % silica, respectively. The mineral composition of

these ores was determined by X-ray diffraction (XRD). The XRD scan of the crude ore indicates

that chalcopyrite is present as a major component of the sulfide mineral and a large amount of silica

as well as magnetite is also present.

The acidophilic thermophile used was A. brierleyi DSM 1651, obtained from the German Collection

of Microorganisms and Cell Cultures (DSMZ). The original DSM 1651 strain was adapted by a

multiple transfer technique to the medium containing the chalcopyrite concentrate as the sole energy

source. The basal salts medium used was that described by Silvermann and Lundgren [9]. The

adapted strain was subcultured aerobically at 65 °C in the modified medium at pH 1.2, supple-

mented with 1 % w/v chalcopyrite and 0.05 % w/v yeast extract. Three-day old cells that were

grown in the medium were used in the subsequent experiments.

2.2 Apparatus and procedure

The batch reactor used to perform the bioleaching tests was an air-sparged stirred vessel of

1000 cm3 capacity [6-8]. Reactor contents were mixed at 250 rpm by a paddle impeller and air was

sparged continuously at 1000 cm3/min. In leaching experiments with A. brierleyi at 65 °C and at a

solution pH of 1.8, the initial cell concentration was 1.0×1014

cells/m3-suspension and the initial

mineral-liquid loading ratio was 5 kg/m3. In some experiments, Fe2(SO4)3 was added so that the

initial concentration of ferric iron in the leach solution was 0.5 or 1.0 kg/m3. A solution sample of

2 cm3 was periodically withdrawn from the reactor. The number of free cells in the liquid samples

was counted using a Petroff-Hausser counting chamber. The solution samples were also analysed

for copper and iron by ICP-AES (inductively coupled plasma-atomic emission spectroscopy). The

fraction of chalcopyrite leached, α, was determined from the copper content in the solution at spe-

cific times divided by the initial copper content in the mineral samples.




3.1 Copper bioleaching for concentrate and crude ore

We compared the leaching ability of A. brierleyi between the chalcopyrite concentrate and the crude

ore at an initial mineral-liquid loading ratio W0/V of 5 kg/m3 and an initial pH of 1.8. The results of

these bioleaching experiments are shown in Figure 3, where the fraction of chalcopyrite leached, α,

and the concentration of the free cells in the leach solution, XL, are plotted as a function of operating

time.

Figure 3: Comparison of copper bioleaching between crude chalcopyrite ore and chalcopyrite

concentrate at an initial pH of 1.8 and an initial cell concentration of

XT0 = (1.2-1.5)×1014

cells/m3: (■) natural crude ore, an initial chalcopyrite-liquid load-

ing ratio of W0/V = 5 kg/m3; (▲) concentrate, an initial chalcopyrite-liquid loading ratio

of W0/V = 0.17 kg/m3; (●) artificial crude ore (mixture of 0.17 kg/m

3 concentrate and

4.83 glass beads), an initial ore-liquid loading ratio of W0/V = 5 kg/m3.

0

0.2

0.4

0.6

0.8

1.0

Co

pp

er l

each

ed, α

[-]

0 2 4 6 8 100

2

4

6

8

Co

nc.

of

free

cel

ls,

XL×

10

-14

[cel

ls/m

3]

Time, t [day]

0

0.2

0.4

0.6

0.8

1.0

Co

pp

er l

each

ed, α

[-]

0

0.2

0.4

0.6

0.8

1.0

Co

pp

er l

each

ed, α

[-]

0 2 4 6 8 100

2

4

6

8

Co

nc.

of

free

cel

ls,

XL×

10

-14

[cel

ls/m

3]

Time, t [day]0 2 4 6 8 10

0

2

4

6

8

Co

nc.

of

free

cel

ls,

XL×

10

-14

[cel

ls/m

3]

Time, t [day]



About 65 % of the copper in the crude chalcopyrite ore was leached in 10 days in a batch reactor.

The bioleaching of copper in the crude chalcopyrite ore resulted in a significant increase in the free

cell concentration XL. The rates of bacterial growth and chalcopyrite oxidation for the crude ore

(CuFeS2 3.3 %, Fe3O4 26.7 %, and SiO2 46.3 %) were almost the same as those for the chalcopyrite

concentrate (Cu 22.5 %, Fe 27.8 %, S 31.4 %). We conclude that gangue minerals in the crude ore,

such as magnetite and silica, contribute little to chalcopyrite bioleaching with A. briereleyi.

3.2 Bioleaching of crude chalcopyrite ore by A. brierleyi

Figure 4 shows the bioleaching behavior of crude chalcopyrite ore with A. brierleyi at 65 °C and at

an initial pH of 1.8. Rapid leaching of chalcopyrite with A. brierleyi was obtained for the crude ore

and 65 % of the copper was leached in 10 days. By comparison, the bioleaching of iron in the crude

chalcopyrite ore was very slow and only about 10 % of the iron was leached in 10 days in the pres-

ence or absence of A. brierleyi. In other words, the leaching of copper took place selectively when

A. brierleyi was used on chalcopyrite because magnetite leaching with A. brierleyi was negligible.

These results demonstrate that thermophilic bioleaching with A. brierleyi is attractive as an eco-

nomical and environmentally friendly process for good copper extraction from a crude chalcopyrite

ore.

Figure 4: Rate data on the batch bioleaching of crude chalcopyrite ore at an initial mineral-liquid

loading ratio of W0/V = 5 kg/m3, an initial pH of 1.8 and 65 °C: (●) copper leaching

with A. brierleyi at an initial cell concentration of XT0 = 1.2×1014

cells/m3; (▲) iron

leaching with A. brierleyi at an initial cell concentration of XT0 = 1.2×1014

cells/m3;

(○) copper leaching without A. brierleyi (sterile control); (△) iron leaching without

A. brierleyi (sterile control).

0 2 4 6 8 100

0.2

0.4

0.6

0.8

1.0

Time, t [day]

Fra

cti

on

of

meta

ls l

ea

ch

ed

, αα αα

[-]

0 2 4 6 8 100

0.2

0.4

0.6

0.8

1.0

Time, t [day]

Fra

cti

on

of

meta

ls l

ea

ch

ed

, αα αα

[-]



Bioleaching experiments with A. brierleyi were done to examine the effect of initial ore-liquid load-

ing ratio W0/V. The results of these experiments are shown in Figure 5 where the concentration of

free cells in the leach solution, XL, and the fraction of chalcopyrite leached, α, are plotted as a func-

tion of operating time. When the initial amount of the crude chalcopyrite ore, W0/V, was increased

from 5 to 20 kg/m3, the free cell concentration, XL, increased markedly but the leaching fraction of

copper, α, changed slightly.

Figure 5: Effect of initial mineral-liquid loading ratio W0/V on the rate of chalcopyrite leaching

with A. brierleyi at an initial pH of 1.8 and an initial cell concentration of

XT0 = (0.8-1.2)×1014

cells/m3: (●) W0/V = 5 kg/m

3; (○) W0/V = 10 kg/m

3;

(◎) W0/V = 20 kg/m3; (▲) W0/V = 50 kg/m

3.



Figure 6 shows the effect of initial pH on the bioleaching of the crude chalcopyrite ore with A. bri-

erleyi at an initial ore-liquid loading ratio of 5 kg/m3. In the range of the initial pH between 1.2 to

2.0 the effects of pH on the bacterial growth and chalcopyrite leaching were slightly.

Figure 6: Effect of initial pH on the rate of chalcopyrite leaching with A. brierleyi at an initial

mineral-liquid loading ratio of W0/V = 5 kg/m3 and an initial cell concentration of

XT0 = (1.2-1.8)×1014

cells/m3: (■) pH 1.2; (▲) pH 1.5; (●) pH 1.8; (◆) pH 2.0.



3.3 Bioleaching mechanism of copper in crude ore with A. brierleyi

The mechanisms of microbial action in sulfide mineral oxidation are usually discussed in terms of a

direct microbial attack on the sulfide and an indirect attack via ferric iron. For the bioleaching of

chalcopyrite with A. brierleyi, both the oxidizable metal moiety (Fe(II)) and the sulfide moiety are

considered to be simultaneously attacked by separate enzymes [10]:

4CuFeS2 + 17O2 + 2H2SO4 → 4CuSO4 + 2Fe2(SO4)3 + 2H2O (1)

When ferric ions are present in the leach solution, chalcopyrite is chemically oxidized with ferric

ion:

CuFeS2 + 2Fe2(SO4)3 → CuSO4 + 5FeSO4 + 2S0 (2)

In the presence of A. brierleyi, ferrous ions and elemental sulfur are microbially oxidized:

4FeSO4 + O2 + 2H2SO4 → 2Fe2(SO4)3 + 2H2O (3)

2S0 + 3O2 + 2H2O → 2H2SO4 (4)

Because ferric ions are automatically supplied by the solubilization of chalcopyrite, the bioleaching

of chalcopyrite may occur by both direct microbial action and indirect chemical action.

To examine the mechanism of chalcopyrite leaching with A. brierleyi, bioleaching experiments

were carried out at different initial ferric ion concentrations, [Fe(III)]0, at an initial ore-liquid load-

ing ratio W0/V of 5 kg/m3 (Figure 7). Chalcopyrite leaching was stimulated in the presence of A.

brierleyi. Chalcopyrite bioleaching at [Fe(III)]0 = 0 kg/m3 resulted in an increase in the liquid-phase

ferric ion concentration to 0.04 kg/m3 after 10 days but no accumulation of ferrous iron in the leach

solution was found. At an initial ferric iron concentration of 1.00 kg/m3, the ferric ion concentration

in the solution increased to 1.04 kg/m3 after 10 days. There was a marked difference in the ferric ion

concentration during the bioleaching. Nevertheless, the initial addition of ferric ions to the A. brier-

leyi cultures only had a slight effect on the rate of chalcopyrite leaching. A previous study [11] re-

ported that the reaction kinetics of chemical leaching exhibits a half order dependence on the ferric

ion concentration. If chemical leaching with ferric ion contributes to chalcopyrite leaching, the ini-

tial concentration of ferric iron, [Fe(III)]0, would have a significant effect on the bioleaching rate.

Figure 7 shows that the bioleaching rates are independent of the ferric ion concentration. This dem-

onstrates that chemical leaching by ferric ion has a negligible contribution to bioleaching with A.

brierleyi. Thus, we conclude that chalcopyrite leaching by A. brierleyi is predominantly due to the

direct microbial attack by A. brierleyi cells.



Figure 7: Effect of initial ferric iron concentration [Fe(III)]0 on the rate of chalcopyrite leaching

with A. brierleyi at an initial mineral-liquid loading ratio of W0/V = 5 kg/m3, an initial

pH of 1.2 and an initial cell concentration of XT0 = (1.2-1.5)×1014

cells/m3:

(▲) [Fe(III)]0 = 1.0 kg/m3; (■) [Fe(III)]0 = 0.5 kg/m

3; (●) [Fe(III)]0 = 0.0 kg/m

3;

(△) [Fe(III)]0 = 1.0 kg/m3; (○) [Fe(III)]0 = 0.0 kg/m

3.

0 2 4 6 8 100

0.2

0.4

0.6

0.8

1.0

Co

pp

er

leach

ed, α

[-]

Time, t [day]

0 2 4 6 8 100

1

2

3

4

5

6

Co

nc. o

f fr

ee c

ell

s,

XL ×

10

-14 [

cell

s/m

3]

Time, t [day]



4 Conclusion

We investigated the mechanism and kinetics of crude chalcopyrite ore leaching by the thermophilic

archaean Acidianus brierleyi at 65 °C and at pH 1.8-2.0 in a batch reactor. About 65 % of the cop-

per in the crude chalcopyrite ore (size range < 25 µm; Cu 1.2 %, Fe 20.4 %, S 2.4 %, SiO2 46.3 %)

was leached in 10 days in the presence of Acidianus brierleyi. The copper leaching rate for the crude

ore was almost the same as that for a chalcopyrite concentrate (Cu 22.5 %, Fe 27.8 %, S 31.4 %),

suggesting that the effects of the gangue minerals such as magnetite and silica is little. The bioleach-

ing of iron was very slow and only about 10 % iron was leached in 10 days in the presence or ab-

sence of A. brierleyi. In other words, chalcopyrite leaching selectively took place and magnetite

leaching was negligible. These results lead us to conclude that thermophilic bioleaching with A.

brierleyi is attractive as an economical and environmentally friendly process for good copper extrac-

tion from low-grade primary sulfides.

References

[1] Rawlings, D. E., and Johson, D.B., 2007, Biomining, Springer, Heidelberg.

[2] Donnati, E. R., and Sand, W., 2007, Microbial Processing of Metal Sulfides, Springer, Hei-

delberg.

[3] Clark, M. E., Batty, J. D., van Buuren, C. B., Dew, D. W., and Eamon, M. A., 2006, “Bio-

technology in Minerals Processing: Technological Breakthroughs Creating Value,” Hydro-

metallurgy, 83, pp. 3-9.

[4] J. A. Brierley, Acidophilic thermophilic archaebacteria: potential application for metal recov-

ery, FEMS Microbiol. Rev., 75 (1990) 287-292.

[5] J. A. Brierley, C. L. Brierley, Microbial mining using thermophilic microorganisms, in T.D.

Brock (ed.), Thermophiles: General, Molecular, and Applied Microbiology, John Wiley &

Sons, New York, 1986, pp. 279-305.

[6] Y. Konishi, S. Yoshida, S. Asai, Bioleaching of pyrite by acidophilic thermophile Acidianus

brierleyi, Biotechnol. Bioeng., 48 (1995) 592-600.

[7] Y. Konishi, S. Asai, M. Tokushige, T. Suzuki, Kinetics of the bioleaching of chalcopyrite

concentrate by acidophilic thermophile Acidianus brierleyi, Biotechnol. Prog., 15 (1999) 681-

688.

[8] Y. Konishi, M. Tokushige, S. Asai and T. Suzuki, ”Copper recovery from chalcopyrite con-

centrate by Acidophilic Thermophile Acidianus brierleyi in batch and continuous-flow stirred

tank reactors”, Hydrometallurgy, Vol. 59, 2001, 271-282.



[9] M. P. Silvermann, D. G. Lundgren, Studies on chemoautotrophic iron bacterium Ferrobacil-

lus ferrooxidans. I. An improved medium and a harvesting procedure for securining high cell

yields, J. Bacteriol., 77 (1959) 642-647.

[10] H. L. Ehrlich, Geomicrobiology, 4th

Ed., Marcell Dekker Inc., New York, NY, USA, 2002.

[11] P. B. Munoz, J. D. Miller, M. E. Wadsworth, Reaction mechanism for the acid ferric sulfate

leaching of chalcopyrite, Metall. Trans. B, 10B (1979) 149-158.


Hybrid Flotation – Newly Developed Flotation

Technology for Increased Recovery –

Especially in the Finest Particle Fractions

W. Krieglstein, L. Grossmann

Siemens AG, Industry Sector, Industry Solutions Division, Metals Technology, Mining

(I IS MT MI)

Schuhstraße 60

Erlangen, Germany

Keywords: Hybrid flotation, Los Pelambres

Abstract

This paper presents a study to estimate the effect of Siemens’ SIMINECIS

HybridFlot flotation tech-

nology on the processing of copper-molybdenum sulphide minerals. A pilot flotation cell was tested

in the Antofagasta Minerals S. A. (AMSA) operation Minera Los Pelambres (MLP), Chile, in 2007.

In Hybrid Flotation, the so-called pneumatic spray-in principle is combined with the column

method. Using SIMINECIS

HybridFlot flotation cells, an increase of molybdenum recovery in the

selective process is achieved particularly with regard to finest and coarsest particles.

1 Introduction

Froth flotation is an important method for the separation and enrichment of mineral raw materials.

This separation method is based on the interfacial phenomenon in the three-phase system liquid-

solid-gas [1]. Among others, flotation reagents play a decisive role in the separation process [2]. By

using these reagents, the flotation or non-flotation of particular minerals can be regulated. Ahead of

froth flotation, grinding is required for the liberation of the valuable minerals from the gangue to

achieve a good concentrate grade and recovery. Fine and ultra fine particle fractions generated by

the grinding process may hinder the downstream separation process especially in terms of recovery

and turn the flotation into a complicated process. Fine particles have a small mass and relatively

high surface area.

These properties can affect and result in the following phenomena: a higher dissolution rate in fluid,

a rigidity of froth, a high pulp viscosity, a heterocoagulation, a low probability of collision between

Krieglstein, Grossmann


particles and gas bubbles, a low particle momentum, and a gangue particle entrainment into the con-

centrate product [3].

This may also lead to a higher consumption of flotation reagents and to undesirable coating of gas

bubbles and valuable particles with ultrafine slimes [3]. It is also believed that overcoming the en-

ergy barrier between particle-particle and particle-bubble becomes more difficult for fine parti-

cles [3]. A continuously increasing appearance of ores with falling head grades and their benefici-

ation effect the flotation of fine particles to become more and more relevant in the future [4].

The Mining segment at the Siemens Metals Technologies Unit is currently developing a portfolio

for flotation cells, aiming to be able to provide complete flotation process solutions in future. In this

regard, Siemens Mining Technologies has developed and built a pilot flotation cell, commissioned

and tested in AMSA’s Minera Los Pelambres (MLP), Chile. In this paper, the test results using the

Hybrid Flotation method in the molybdenum beneficiation process will be shown.

2 Flotation

One of the main challenges for flotation equipment providers is the development of flotation cells

which disperse a sufficient amount of fine air bubbles into the pulp for the recovery of fine grinded

valuable particles among coarser particle fractions.

The flotation process takes place under following conditions:

- A particle-gas bubble contact is given,

- Particles attach to bubbles and form aggregates,

- The particle-bubble-aggregates do not detach during floating from the pulp zone into the froth

zone of the machine and are transported from there into the concentrate collecting devices.

The probability P for a particle being recovered during the flotation process depends on the prob-

ability of the particle-bubble collision Pc, the probability of attachment of particles to the bubbles Pa

and the probability of detachment of the particles from the bubble Pd [3, 5].

( )dac PPPP −⋅⋅= 1 (1)

In literature, a generalized formula can be found for the probability of the particle collision with gas

bubbles Pc depending on the particle size dp and the bubble db size [5]:

n

b

p

ccd

dkP

⋅= (2)

The Parameters kc and n are dependent on the Reynolds number Re i.e. the hydrodynamic milieu

during the flotation process.

Hybrid Flotation


The probability of the attachment depends, among others, on the hydrophobicity of the parti-

cles [3, 5]. The most important mechanism of the flotation is the attachment of particles to the gas

bubbles and it can be influenced by the addition of flotation reagents [3].

The probability of the detachment depends, among others, on the collision with other particles and

on the turbulence of the flotation cell [3] and can be defined as follows [5]:

5.1

max,

=

p

p

dd

dP (3)

where dp,max is the particle size of the major floatable particle.

Furthermore, bubble and particle size distributions and their relation play a decisive role in the flota-

tion process [2, 3, 5].

As a consequence, research and development activities are focused on the hydrodynamic character-

istics of the flotation cells beside the areas of specific energy consumption, maintainability and

others.

3 Hybrid Flotation

In the design of Siemens’ SIMINECIS

HybridFlot flotation cell, a so-called pneumatic spray-in prin-

ciple is combined with the column method (Figure 1). The cell operates without an agitator, because

the ore slurry is sprayed into the cell by high-pressure nozzles. The gas is added to the ore slurry in

mixing chambers before the 3-phase mixture enters into the cell. The special design of the nozzles

ensures the generation of finest gas bubbles. This considerably improves the frequency of contact

between the gas bubbles and very fine particles (Equation 2), and also improves the ability of the

particles to stick to these bubbles. The resulting mixture is sprayed into the flotation cell in a tan-

gential arrangement.



Figure 1: SIMINECIS

HybridFlot flotation cell [6]

Additional gas is added to an internal column in the second stage (Figure 1), which ensures that the

ore particles not “captured” in the first stage can collide with the gas bubbles of the second stage

and be transported to the surface. The froth concentrate produced during the process is taken to the

edge of the flotation tank and drained off. Short retention times of the pulp in the machine and new

patented additional drain gutters reduce the risk of loosing already captured particles. The

SIMINECIS

HybridFlot flotation cell as tested at Minera Los Pelambres can be operated with pulp

feed having up to 50 percent solid content and has a nominal capacity of up to 400 m³/h.

4 Flotation Process in Minera Los Pelambres

In Minera Los Pelambres, the porphyry copper minerals like chalcopyrite, chalcocite, bornite, covel-

lite and molybdenite are processed with typical porphyry ores head grades for copper and molybde-

num. The recovery of molybdenum is carried out as a by-process of the flotation of copper sulphide

minerals. Concentration and separation takes place in two steps, carried out in a bulk flotation plant

and a molybdenum plant. In the bulk flotation process, rougher, scavenger and cleaner flotation are

Hybrid Flotation


implemented with the associated concentrate and regrind stages [2]. Bulk copper-molybdenum con-

centrate goes to a thickener prior to its treatment in the molybdenum plant (Figure 2). The

SIMINECIS

HybridFlot flotation cell was implemented in the existing flotation circuit as a pre-

rougher cell. The concentrate from the SIMINECIS

HybridFlot flotation cell was led to a pneumatic

cell which was operated as cleaner and produced final concentrate. The tailings from the SIMINECIS

HybridFlot flotation cell were fed to the rougher flotation bank of the selective process.

Figure 2: Flow sheet of the molybdenum beneficiation process in Mine Los Pelambres

5 Test Results

This paper presents the results of the in-plant equipment testing of the SIMINECIS

HybridFlot flota-

tion cell. During the in-plant testing of the SIMINECIS

HybridFlot flotation cell at Los Pelambres,

samples were collected from the feed, concentrate and tailings flows. Primarily, the SIMINECIS

Hy-

bridFlot flotation cell was integrated into the existing flotation circuit and was operated for 4 hours

and respectively 3 samples were taken from the flows. The concentrates from the regular process

and the concentrate of the SIMINECIS

Hybrid Flot flotation cell were united, see Figure 3. After-

wards, the cell was disconnected from the flotation process. The plant was operated then without the

SIMINECIS

HybridFlot flotation cell until a stabilisation of the process (after approx. 2 hours) was

achieved. The sampling for characterising the existing flotation circuit began after the stabilisation.

As a consequence from this test set up, the concentrate mix consisting of concentrate from the con-



ventional mechanical rougher cells and concentrate from the SIMINECIS

HybridFlot flotation cell

was compared with concentrate coming only from the conventional mechanical rougher cells. The

tests were realized at different feed mass flow levels, ranging from 100 to 120 and 140 tons per hour

and the mass concentration was uniform. At the molybdenum concentrate sampling point (Figure 3),

the different concentrate samples were taken and analysed.

Figure 3: Test set up and Sampling points

The results of the analysis of the samples taken are illustrated in Figure 4. The metallurgical perform-

ance of the existing molybdenum flotation circuit could be enhanced by an increasement of molybde-

num recovery at per average 1.2 percentage points. This improvement of the metallurgical perform-

ance is associated with a higher selectivity of the flotation process. The recovery of copper and iron

into the concentrate is dropped by an average of 1.7 and 1 percent respectively. For the recovery of

molybdenum, the optimum operation range was found at the solid mass flow Qs= 120 t/h.

Hybrid Flotation


Figure 4: Recovery of molybdenum, copper and iron in the flotation process

Figure 5 demonstrates the selectivity of the existing molybdenum flotation circuit with and without

the SIMINECIS

Hybrid Flot flotation cell in operation. In Figure 5, the content of valuable mineral

(molybdenum) and the unwanted minerals in the concentrate depending on the solid mass flow are

illustrated. The flotation is seen to be improved by integration of the SIMINECIS

HybridFlot flota-

tion cell and the molybdenum grade is noted to be increased by 0.6 percent on average. It can be

concluded that the selectivity of the process is higher when the SIMINECIS


is integrated in the flotation circuit. The copper and iron grades in the concentrate drop by an aver-

age of 1.1 points when the SIMINECIS

HybridFlot flotation cell is integrated into the flotation cir-

cuit. The highest molybdenum grade in the concentrate was achieved at low solid mass flow

Qs = 100 t/h in the investigated range.

By integrating the SIMINECIS

HybridFlot flotation cell into the existing flotation circuit, an increase

of molybdenum recovery and grade in the concentrate can be achieved with simultaneously increas-

ing of the selectivity of the flotation process.

With the SIMINECIS

HybridFlot flotation cell, the valuable particles can be removed from the pulp

quickly. The residence time in the cell was between 3.8 and 5.3 minutes. In the existing flotation

bank at Los Pelambres, the pulp has a typical residence time of approx. 20 to 30 minutes, other se-

lective rougher processes do show even much higher residence times.

100 120 140

0

10

20

30

40

50

60

70

80

90

100R

eco

ve

ry in

%

Solid Mass Flow in t/h

Molybdenum (with Hybrid Flotation Cell)

Molybdenum (without Hybrid Flotation Cell)

Copper (with Hybrid Flotation Cell)

Copper (without Hybrid Flotation Cell)

Iron (with Hybrid Flotation Cell)

Iron (without Hybrid Flotation Cell)



Figure 5: Molybdenum, copper and iron grades in the concentrates

Recovery and molybdenum grade in the concentrate for the molybdenum rougher flotation process

are summarized in Table 1, classified by different particle size fractions. Analysis of particle sizes

was carried out by sieve tests.

Table 1: Molybdenum recoveries and grades in the concentrates of the rougher flotation process

Particle Size Hybrid Flotation Recovery Grade

d < 10 µm ON 97.5 % 5.9 %

OFF 95.2% 6.6 %

10 µm < d < 44 µm ON 97 % 6.9 %

OFF 96 % 6.9 %

44 µm < d < 74 µm ON 97.8 % 38.1 %

OFF 96.3 % 29.9 %

74 µm < d ON 96 % 37 %

OFF 90.2 % 29.7 %

100 120 140

0

5

10

15

20

25

30

35C

on

tent in

%

Solid Mass Flow in t/h

Molybdenum (with Hybrid Flotation Cell)

Molybdenum (without Hybrid Flotation Cell)

Copper (with Hybrid Flotation Cell)

Copper (without Hybrid Flotation Cell)

Iron (with Hybrid Flotation Cell)

Iron (witout Hybrid Flotation Cell)

Hybrid Flotation


The highest increase in recovery is achieved for the particle fractions with a mean particle diameter

d > 74 µm and d < 10 µm. It can be assumed that the improvement of the metallurgical performance

of the molybdenum rougher flotation process is attributable to the previous recovery of fine particles

by the SIMINECIS

HybridFlot flotation cell. A plausible explanation of the data shown in table 1 can

be given by the assumption that two major effects are responsible for the overall metallurgical per-

formance improvement by implementing the SIMINECIS

HybridFlot flotation cell: Firstly, the addi-

tionally recovered fine particles fractions per se lead to a higher overall recovery without having a

negative effect on the grade. Secondly, the removal of these fine particle fractions might positively

influence the ability to recover coarser particle fractions in the down stream mechanical rougher

cells by preventing undesired effects like coating of valuable coarser mineral particles and gas bub-

bles by fine particles.

The findings and assumption as described above lead to the necessity of a more detailled analysis of

the influence of particle sizes. The cumulative undersize of the feed is shown in Figure 6. It can be

seen that the major part of the particles in the feed (d50 = 18 µm) is fine.

Figure 6: Cumulative undersize of the feed measured by a sieve test

The cumulative undersize of the concentrate of the SIMINECIS

Hybrid Flot flotation cell, measured

by Cyclosizer, is illustrated in Figure 7. It can be noted that the majority of the particles in the con-

centrate (approx. 58 %) has a particle size smaller than 10 µm. This underlies the assumption that



the SIMINECIS

HybridFlot flotation cell has a high recovery efficiency especially for fine and ultra

fine particle fractions.

Samples were drawn form the inner cylinder section (column method) and the peripheral section

(spray-in principle) of the SIMINECIS

HybridFlot flotation cell, see Figure 1. It was assumed that the

majority of the fine particle fractions are recovered in the first flotation stage in the upper section of

the SIMINECIS

HybridFlot flotation cell, i.e. by the spray-in principle. On the contrary, particles in

the in-cylinder are slightly finer than those in the peripheral part of the SIMINECIS

HybridFlot flota-

tion cell. This minimal effect can be explained by the sampling. The froth from in-cylinder is led to

the peripheral part of the cell (Figure 1). Therefore, both different froth products generated by the

two different principles in the SIMINECIS

HybridFlot flotation cell were mixed, before the entity of

both froth products was led out of the machine.

It is for this reason that a comparison of the different froth products was only possible by comparing

a froth mix against the product of the second stage, e.g. in-cylinder principle. A significant differ-

ence in terms of recoveries and grades of molybdenum of both froth products generated by the two

principles in the machine could therefore not be proven.

Figure 7: Cumulative undersize of the concentrate of SIMINECIS


All results presented so far show that an implementation of a single SIMINECIS

HybridFlot flotation

cell as an additional pre-rougher stage into an existing copper-molybdenum selective flotation pro-

cess can improve recovery and selectivity of the overall process, attributable to the improved re-

Hybrid Flotation


covery of fine particle fractions, the consequential positive impact on the downstream mechanical

rougher cells and the simple provision of additional process volume.

To also address the question to which extent a single Siemens cell may replace an entire bank of

mechanical rougher cells, the results of the metallurgical performance test of the single SIMINECIS

HybridFlot flotation cell are summarized in Table 2, classified by 2 different particle size fractions.

Table 2: Molybdenum recovery and grade with SIMINECIS


Particle Size Recovery Grade Enrichment factor approx.

d < 44 µm 18.5 % 15.9 % -

d > 44 µm 37.0 % 50.0 % -

Overall 22.1 % 24.0 % 8 … 10

The overall recovery of the SIMINECIS

HybridFlot flotation cell was 22.1 % with an overall 24 %

molybdenum grade in the concentrate, resulting in a range of 8 to 10 for the enrichment factor. It

can be concluded that, as expectable, a single SIMINECIS

HybridFlot flotation cell has a lower re-

covery than usually a full size mechanical rougher flotation bank has. On the other hand, the achiev-

able molybdenum grade for the concentrate respectively the ratio of concentration is very high with

the SIMINECIS

HybridFlot flotation cell.

6 Conclusions

The test with the first SIMINECIS

HybridFlot flotation pilot cell in the selective copper-molybdenum

process showed an increasement of molybdenum recovery by more than 1 percentage point per av-

erage of the overall rougher process. Depending of the feed load of the process, affecting the opti-

mum operation range of the SIMINECIS

HybridFlot flotation cell, overall recovery increasement by

more than 2 percent could be reached. The recovery increase is shown especially in the fine particle

fractions (d < 10 µm), but also in the coarser particle size fractions (d > 74 µm). It can be assumed

that this improved overall recovery of the coarser particles is due to a reduction of an undesired

coating of the valuable particles and gas bubbles with ultrafine slimes by recovering fine particle

fractions, operating the SIMINECIS

HybridFlot flotation cell as a pre-flotation stage of the process.

Furthermore, the ability to create fine gas bubble sizes, very short retention times and a very quick

removal of the froth products lead to an increased recovery of fine particle fractions. The molybde-

num concentration could be enriched from a typical concentration of less than 3 % in the feed pulp

up to approximately 24 % percent on average and up to nearly 40 % as maximum in the concentrate

product of the SIMINECIS

HybridFlot flotation cell using only one cell. The fraction of fine particles

and coarse particles which are lost in the existing flotation process can be recovered with good effi-

ciency by integration of the SIMINECIS

HybridFlot flotation cell into the process as pre-rougher

stage.



7 Outlook

Based on the achieved results, ongoing research and development works are currently done to de-

scribe the hydrodynamic characteristics of both the two basical principles in the SIMINECIS

Hybrid-

Flot flotation cell separately as well as an integrated, interacting system. To be able to address these

fields of research work, a laboratory device is set up and operated accordingly. Furthermore, a

downscaled mobile test unit of the SIMINECIS

HybridFlot flotation cell is available to be able to

conduct further test work in operational industrial sites.

Once detailled hydrodynamic characterizations and kinetic modelling are available in more details,

partial process solutions and in a further step fully integrated flotation processes can be elaborated

by combining and cascading SIMINECIS

HybridFlot flotation cells – if applicable in different sizes.

Prospectively, Siemens is also striving to transfer its SIMINECIS

HybridFlot flotation technology

from copper-molybdenum selective processes into applications in copper bulk and other minerals

flotation processes.

Acknowledgments

Siemens AG would like to acknowledge the contributions of the personnel at the Minera Los Pe-

lambres operation of Antofagasta Minerals S.A., Chile. This project would not have been possible

without Mr. Dalibor Dragisevic, Mr. Nestor Villarroel, Mr. Jorge Cortinez and Mr. Andres Soto.

The authors wish to express their gratitude to them because of their invaluable openness, assistance

and support during the project. Their commitment in terms of time and openness to use new tech-

nology is gratefully acknowledged.

Hybrid Flotation


References

[1] WILLS, B. & NAPIER-MUNN, T. (2006): Will’s Mineral Processing Technology: An Intro-

duction to the Practical Aspects of Ore Treatment and Mineral Recovery, Oxford (Butterworth-

Heinemann) – ISBN 0-8247-9264-5

[2] BULATOVIC, S. M. (2007): Handbook of Flotation Reagents. Chemistry, Theory and Practice,

Amsterdam (Elsevier) – ISBN 978-0-444-53029-5

[3] MATIS, K. A. (1995): Flotation science and engineering, New York (Marcel Dekker) – ISBN

978-0-750-64450-1

[4] CROZIER, R. D. (1992): Flotation: theory, reagents and ore testing, Oxford (Pergamon) –

ISBN 0-08-041864-3

[5] LI, R. (1996): Untersuchung über Flotationsgrundlagen unter besonderer Berücksichtigung der

Korngröße, Aachen (Shaker) – ISBN 3-8265-1635-4

[6] VIDUYETSKY, M. G., GARIFULIN, I. G. & MALZEW, V. A. (2006): Patent WO 069995,

Pneumatic flotation column comprising a foam collecting container



Perspectives of Copper Mining Industry

Development in Poland

Ph.D. Eng. Jan Kudełko, Ph.D. Jacek Pyra, Ph.D. Eng. Jerzy Sobociński

KGHM CUPRUM Ltd Research and Development Centre

ul. Gen. Sikorskiego 2-8

53-659 Wroclaw, Poland

Keywords: Mining, copper, production, profit

Abstract

To determine mining profits possibility, copper ore and accompanying minerals were analyzed on

the basis of the copper ore resources. Future technical and economic developments of the copper

industry in Poland for the next ten years were evaluated considering actual state of copper ore stock

balance. Presented production-economic results obtained in the last 12 years by KGHM Polska

Miedz S. A. led to estimate company effectiveness's forecast in aspect of macroeconomic conditions

and estimated production values and costs.

1 Introduction

Development policy of domestic copper mining industry is a very important issue for the Polish

economy especially the regional one. Defining it at the very early phase of investment cycle, permits

not only for selection of adequate development strategy but also contribute to prepare the most op-

timum procedure of already chosen policy. Type of mineral, quality of deposit, presence of base and

accompanying minerals together with development of winning and processing technology, deter-

mine to the great extend, the development trends of KGHM Polska Miedz S. A. However, the

changes in word market situation, marked metals demand or even employment opportunities in the

region of possible deposit occurrence have a substantial impact on changes of previously deter-

mined development policy. Adapting to the market situation, basing on business analyzes of already

confirmed metals deposits at established assumption, it is possible to determine the project profit-

ability conditions and choose the most cost effective of deposit extraction as well and time of its

duration. Expanding the owned resource base by adding new deposit or widening the boundaries of

already mined deposit makes the possible mining period longer. Thus, very important are the explo-

ration works on perspective areas, where depending on economic conditions, further activities may

be taken.

Kudełko, Pyra, Sobociński


Using new technologies in mining of copper ore results in more cost effective production, what has

the direct relation with copper production profitability, even when mining the ore having worse

parameters. However utilization of these technologies gives the measurable results in longer time

perspective, due to the substantial costs of their winning and implementation.

The impact of world market metal prices fluctuation on the worthwhile mining of the poly-metallic

deposits is presented in the paper. This was the base to analyze the KGHM Polska Miedz S. A.

development trends in the future.

2 Reserves of copper deposits in Poland

Polish copper deposits are of sedimentary type and occur in the Zechstein (Permian) formation on

the Foresudetic Monocline, as well as in Northsudetic Basin. Presence of 14 copper deposits con-

firmed: 11 on the Foresudetic Monocline and 3 in the Northsudetic Basin.

At present 6 deposits on the Foresudetic Monocline: “Lubin-Malomice” “Polkowice”,

“Sieroszowice”, “Radwanice Wschod”, “Rudna” and “Glogow Gleboki – Przemyslowy” (GG-P) are

mined. Other deposits such as “Bytom Odrzanski”, “Gaworzyce”, “Glogow”, “Radwanice Zachod”

and “Retkow” are now not developer are the future resource base.

On the Northsudetic Basin 3 deposits: “Niecka Grodziecka", “Wartowice" and “Nowy Kosciol" are

located.

Deposits of copper ore mined by KGHM are located on Lower Silesia between Lubin and Glogow.

The deposit area confirmed by exploration extends 40 km along the longness and 20 km along the

dip on the depth of from 600 to 1380 meters. The orebody occurs within the Zechstein sedimentary

formation, which is inclined in the form of monocline towards north-east.

Economic reserve base of KGHM copper deposit as for 31.12.2005 are 922 million Mg of ore,

21.3 million Mg of copper and 58.5 thousand Mg of silver (Table 1).

The biggest amount of reserves bound in pillars is in “Lubin-Malomice” and “Rudna” deposit –

about 150 million Mg each, mainly in protection pillar of Lubin town and protection pillar of “Ze-

lazny Most” tailings management facility.

Getting the reserves of “Glogow Gleboki-Przemyslowy” deposit will let to restore the base of re-

sources supposed economic from years 1998-1999 and base of economic reserve base from early

nineties of previous century (Table 1).

In further time perspective there will be a possibility to enlarge the resources of copper mining in-

dustry by the resource supposed economic of copper ore occurring on the following reserve areas –

“Gaworzyce” (44.8 million Mg) and "Radwanice Zachod" (18.6 million Mg) located west from

“Sieroszowice” and “Radwanice Wschod” deposits as well as deposits located along the inclination,

towards north-east and north from “Glogow Gleboki deposit – Przemyslowy” i.e. “Bytom Odrzan-

Perspectives of Copper Mining Industry Development in Poland


ski” (31.5 million Mg) and “Retkow” (135.8 million Mg). Due to the absence of present economic

criteria for copper deposits below 1250 m, reserve area “Glogow” does not have the confirmed sup-

posed economic resources of copper ore.

In so called Old Miedz Basin the deposit with potential mining capability is joint depost Niecka

Grodziecka and Wartowice (Table 2, 3). Total supposed economic resources are there 84.2 million

Mg of ore, and contain 1.3 million Mg of copper and 4.6 thousand Mg of silver.

Presence of perspective supposed economic resources of copper ore should be bound with both the

foresudetic area (Zechstein deposits), and sedimentary-magmatic rock of north periphery of Upper

Silesian Coal Basin (GZW). However extracting the foresudetic deposits is very impedimental due

to the unfavorable geological conditions, while north periphery of GZW in not enough explored yet.

Probable reserves in Zechstein formation on the Foresudetic monocline amount 29.7 billion Mg of

ore (165 million Mg Cu). Explored reserves of D1 category (west from Wartowice deposit and Gaw-

rony – Scinawa area) are estimated to be about 3.9 billion Mg of ore (about 17 million Mg of Cu).

These are areas with relatively high level of exploration, but mostly with not enough high metals

concentration within two meters interval. Probable reserves of copper ore explored in D2 category

are estimated to be 25 billion Mg of ore (150 million Mg of Cu). Exploratory boreholes network is

very irregular since most of them were drilled by petroleum industry and their spacing is adapted to

gas and oil prospection. The situation is also made more complicated by the fact that they may be

located totally within the area of gas symptoms, making difficult the future mining (Table 4).



Table 1: Reserves, supposed economic and not supposed economic resources of copper KGHM

Polska Miedz S. A. deposit – as for 31.12.2005



Table 2: Niecka Grodziecka copper deposit-reserves specification

Type of re-

sources

Exploration

stage

Supposed economic resources Not supposed economic resources

Ore

[thousand Mg]

Cu

[thousand Mg]

Ore

[thousand Mg]

Cu

[thousand Mg]

Out of pillars A+B 513 7.5 - -

C1 6856 108.3 2293 31.2

C2 28,244 348 927 10.8

subtotal 35,613 463.8 3220 42.0

In protection

pillars

A+B 1,740 29.2 - -

C1 8223 155.4 7 0.1

C2 809 17.1 - -

Subtotal 10,772 201.8 7 0.1

Total reserves 46,385 665.6 3227 42.1

Table 3: Copper resources in Wartowice deposit to the depth of 1250 m

Type of resources Exploration

stage

supposed economic resources notsupposed economic re-

source

Ore

[thousand Mg]

Cu

[thousand Mg]

Ore

[thousand Mg]

Cu

[thousand Mg]

Out of pillars C1 20,153 271 7965 88

In protection pillars C1 17,650 327 1396 15

Total C1 37,803 598 9361 103

Table 4: Level of probable resources of copper ore on the depth 1250-2000 m (cat. E)

(Rydzewski et al., 1996)

Item Calculative field

Interval of

depth

[m]

Field

area

[km2]

Thickness of

copper

bearing

inteval

[m]

Average

content of

Cue

[%]

Perspective resources

Ore

[million Mg]

Cu

[million Mg]

1

Foresudetic

Monocline – area

III

(15 boreholes)

1250-1500

1500-2000

50

400 5.0 0.64 5600 35

2

Foresudetic

Monocline - area

IV

(Borzecin –

Sulmierzyce)

(27 boreholes)

1250-1500

1500-2000

350

2030 3.08 0.55 170,000 86



3 Economic premises for “Wartowice – Niecka Grodziecka”

deposits management

Using the results of „Analyze of grounds for Niecka Grodziecka deposits (Konrad, Wartowice)

management” study, the analyze of cost-effectiveness of Wartowice – Niecka and predicted macro

economic conditions.

Essential parameters decisive in determining the investment project profitability in poly-metallic

ores mining and processing industry are as follows:

• deposit resources,

• basic and accompanying elements content,

• annual production figures,

• time period of investment phase,

• investment costs of deposit development and processing plant construction

• costs of mining and processing,

• prices and possibilities of produced metals marketing.

Among the above parameters, characterizing the project, the most difficult to estimate, especially in

longer time perspective, are the macroeconomic values such as copper price, silver price and ex-

change rates. Therefore predicting the values of those parameters the trends in last years were taken

into consideration (Figure 1 and 2).

Carrying the economic analyze in order to define the value of profit from extraction and processing

the ores from Wartowice – Niecka Grodziecka deposits for the predicted schedule of production

based on reserves, the twenty years period of mine and concentrator operation was assumed.

The calculations were carried out using the method of drawn calculus for the technological circuit,

at the assumption, that the Company (mine, concentrator and metallurgical plant) final product is a

electrolytic copper and metallic silver.

Calculation of profit from selling the produced electrolytic copper and silver obtained as a result of

ore extraction and processing from Wartowice – Niecka Grodziecka deposits consisted in defining

the following values:

• costs of extraction,

• costs of flotation processing,

• costs of metallurgical processing Cu,

• costs of metallurgical processing Ag,

• investment cost for mine and concentrator construction,

• depreciation of the above investments.



The sum of calculated costs gave the production costs.

The production income is composed of:

• income from electrolytic Cu sale calculated as product of copper amount and price of 1 Mg of Cu

in USD and American dollar exchange rate,

• income metallic Ag sale calculated as product of silver and price of 1 kg of Ag in USD and

American dollar exchange rate.

Gross profit was computed as a difference between the income and costs of production.

The following values describing the deposit and investment project were used in calculations [7]:

• reserves of deposits – 60,000,000 Mg,

• content of Cu in the ore – 1.23 %,

• content of Ag in the ore – 46.0 g/Mg,

• production – 3,000,000 Mg/year,

• period of investment– 11 years,

• discounted investment costs – 1100 million USD,

• unit costs of extraction – 33 USD/Mg,

• unit costs of flotation processing – 10 USD/Mg,

• unit costs of metallurgical processing Cu – 900 USD/Mg Cu,

• unit costs of metallurgical processing Ag – 13.7 USD/kg Ag.

Additionally the following values of yields obtained in flotation and metallurgical processing were

assumed:

• flotation yield of Cu – 90 %,

• metallurgical yield of Cu – 97 %,

• flotation yield of Ag – 86 %,

• metallurgical yield of Ag – 95 %.

The following values of macroeconomic parameters were used:

• electrolytic copper price – 6000 USD/Mg,

• metallic silver price – 386 USD/Mg (12 USD/troz).



Figure 1: Average prices of electrolytic copper on LME in years 1996-2008

Figure 2: Average prices of metallic silver on LBM in years 1996-2008

Using those parameters the gross profit for example year of mine operation was calculated. More-

over the sensitivity analysis of annual gross profit value and cumulated gross profit for the whole

operation period of project on the copper price (Figure 3 and 4).



Figure 3: Sensitivity of annual gross profit on copper price change (Wartowice-Niecka

Grodziecka project)

Figure 4: Sensitivity of cumulated gross profit on price change of electrolytic copper (Wartowice-

Niecka Grodziecka project)

Analyzing the presented economic figures calculated for “Wartowice – Niecka Grodziecka” deposit

management project it is evident that profit from flowing production will be maintained only if the

average copper price within the next three decades is not lower than 6000 USD per Mg.



Thus the perspectives of Polish copper industry development at present conditions may be are prom-

ising only for KGHM Polska Miedz S. A.

4 Conditions for copper industry development in Poland in

KGHM Polska Miedz S. A.

KGHM Polska Miedz S. A. it is a company having almost 50 years long tradition, applying the most

modern methods of extraction and processing of copper ore, being the world leader of high quality

copper and silver as well as other raw materials. Copper ore is extracted from the Europe biggest

and one of the world biggest copper deposits, which is located between Glogow, Lubin and

Sieroszowice. The deposit has the area of about 550 km2. The company extracts the copper ore in

three mines: Lubin, Polkowice - Sieroszowice and Rudna. Annual output is nearly 30 million Mg of

ore while copper production exceeds 500,000 Mg of electrolytic copper and over 1200 Mg of metal-

lic silver due to processing the ore from own resources and purchased concentrates. Moreover small

amounts of nickel (nickel sulfate), crude lead, metallic selenium, and platinum-palladium sludge are

produced. In recent years increased also the metallic gold production what was the result of process-

ing the purchased concentrates with higher, than in domestic concentrates, Au content.

Granted in 2004 by Minister of Environment the concession for mining the copper ore from

“Glogow Gleboki Przemyslowy” (GG-P) deposit, what increased the reserves by 26 %. KGHM Pol-

ska Miedz S. A. currently extracts 6 deposits: “Lubin-Malomice” “Polkowice”, “Sieroszowice",

“Radwanice Wschod”, “Rudna” and “Glogow Gleboki – Przemyslowy”.

Production figures of the Company in years 1996-2007 are shown on drawings: 5 – output extrac-

tion, 6 – concentrate production, 7 – electrolytic copper production, 8 – production of metallic sil-

ver. Additionally changes of copper content in the ore and copper in concentrate changes are shown

in Figures 9 and 10.



Figure 5: Output in KGHM Polska Miedz S. A. mines

Figure 6: Concentrate production in KGHM Polska Miedz S. A.



Figure 7: Production of electrolytic copper in KGHM Polska Miedz S. A.

Figure 8: Production of metallic silver in KGHM Polska Miedz S. A.

The presented results show that the production in KGHM mines since 2002 had been increasing

gradually up to the maximum level reached in 2006. Production of copper during last years in in-

creased first of all due to enlarged purchase of concentrates, while silver production since 2003 has

dropped as a result of lower metals (both copper and silver) content in own resources, what was also

noticed in copper content reduction in concentrate.



Figure 9: Copper content in output mined in KGHM Polska Miedz S. A. mines

Figure 10: Copper content in KGHM Polska Miedz S. A. concentrate

Carrying the economic analyze in order to define the value of profit from extraction and processing

the ores from GG-P deposit for the predicted schedule of production based on reserves, the period

from 2008 to 2040 of mine and concentrator operation was assumed. The calculations were carried

out using the method of drawn calculus for the technological circuit, at the assumption, that the

Company (mine, concentrator and metallurgical plant) final product is a electrolytic copper and me-

tallic silver.

In this analyze the following values describing the deposit and investment project were used [6]:



• deposits reserves – 204,000,000 Mg,

• content of Cu in the ore – 2.08 %,

• content of Ag in the ore – 64.7 g Mg,

• annual average mining production – 6,000,000 Mg/year,

• investment period (till getting the average mining level) – 15 years,

• investment cost of region development and preparing for extraction – 660 million USD,

• unit costs of extraction – 50 USD/Mg,



• unit costs of metallurgical processing Ag – 13.7 USD/kg Ag.

Additionally the following values of yields obtained in flotation and metallurgical processing were

assumed:

• flotation yield of Cu – 90 %,

• metallurgical yield Cu – 97 %,

• flotation yield of Ag – 86 %,

• metallurgical yield of Ag – 95 %.

The following values of macroeconomic parameters were used:

• electrolytic copper price – 4500 USD/Mg,

• metallic silver price – 386 USD/Mg (12 USD/troz).

Using those parameters the gross profit for example year of mine operation from GG-P was calcu-

lated. Moreover the sensitivity analysis of annual gross profit value and cumulated gross profit for

the whole operation period of project on the copper price (Figures 11 and 12).



Figure 11: Sensitivity of annual gross profit on electrolytic copper price for production in GG-P

mine

Figure 12: Sensitivity of cumulated gross profit value on electrolytic copper price for production in

GG-P mine

5 Prognosis of production value and KGHM Polska

Miedz S. A. efficiency

Basing on production program elaborated for „Appendixes to Copper Deposit Management Design

of “Lubin-Malomice, “Polkowice II”, ”Sieroszowice I”, “Rudna I”, “Rudna II”, “Radwanice

Wschod” and “Glogow Gleboki-Przemyslowy mines”, the prognosis of extraction level as well as



copper and silver production from own reserves for years 2008-2040 was made. Studying the cur-

rent trends in extraction and processing cost changes and predicting the values of macroeconomic

parameters during next decades, the economic analyze was made with defining the predicted net

profit from selling the copper and silver from own reserves.

Forecast of production volume, of production costs and of net profit was made using in calculation

the following values:

• unit costs of extraction – 47.7 USD/Mg,



• unit costs of metallurgical processing Ag – 13.7 USD/kg Ag,

• Cu flotation yield – 90 %,

• Cu metallurgical yield – 97 %,

• Ag flotation yield – 86 %,

• Ag metallurgical yield – 95 %.

The following macroeconomic parameters were used:

• electrolytic copper price:

2009 – 6500 USD/Mg,

2010 – 6000 USD/Mg,

2011-2040 – 4500 USD/Mg,

• metallic silver price:

2009 – 14 USD/troz,

2010 – 13 USD/troz,

2011-2040 – 12 USD/troz.



Figure 13: Copper production costs vs. copper prices

In Figure 13 the comparison of copper price changes and production costs of electrolytic copper in

years 1996-2008 is presented. The charts show the substantial increase of production costs since

2004 when metal prices started to substantially increase.

Figure 14: Predicted mining production in KGHM mines



Figure 15: Predicted production of electrolytic Cu from own KGHM reserves

In Figures 14, 15 and 16 the predicted volume of extraction in KGHM mines and predicted Cu and

Ag production from own reserves are presented. Extraction level should be stable until 2030 (28-

30 million Mg), only in last decade the drop of production to the level of 20-25 million Mg is ex-

pected. Prognosis of copper and silver production during the next three years is on the level of

400 thousand Mg of Cu and 1000 Mg of Ag due to carrying out the mining operations in areas with

lower metals content in the deposit. Starting the extraction in “Glogow Gleboki” region will let to

increase Cu production and especially Ag production due to better ore parameters in this area (aver-

age Cu content of 2.08 % and Ag content of 65 g/Mg). In Figure 17 the predicted net profit is pre-

sented, which at the copper price of 4500 USD/Mg Cu is on the level of 160-210 million USD for

the year.

In Figure 18 the predicted unit cost of electrolytic copper production taking into consideration the

revenue from silver sale, is shown.



Figure 16: Predicted production of metallic Ag from own KGHM resources

Figure 17: Predicted net profit from electrolytic Cu production from own KGHM resources



Figure 18: Predicted unit costs of electrolytic copper production from own KGHM reserves

6 Conclusions

1. Among currently prospected resources of copper deposits only the deposits extracted KGHM

Polska Miedz S. A. have the economic importance and only the development of those deposits

extraction (“Glogow Gleboki – Przemyslowy”) and if necessary development of “Gaworzyce”

and “Radwanice Zachod” deposits will be cost-effective.

2. Mine construction and mining of “Wartowice – Niecka Grodziecka” deposit will be profitable if

copper price, during next three decades, is not lower than 6000 USD/Mg.

3. Decrease of mineralization in copper deposit extracted by KGHM Polska Miedz S. A. in current

five-year period does not have the substantial impact on the Company economic results.

4. Development and mining of “Glogow Gleboki – Przemyslowy” (GG-P) deposit allow to main-

tain the production on the current level during next two decades.



References

[1] KGHM Polska Miedz S.A.: Annual Report 1998, Lubin.





[6] SADECKI Z. et al., (2006): Appendix no. 1 to Copper Deposits Management Design for

Glogow Gleboki-Przemyslowy”. KGHM CUPRUM sp. z o.o. CBR, Wroclaw.

[7] SOBOCINSKI J. et al., (2007): Study on reasons of Grodziec Basin deposit (Konrad, War-

towice) management. KGHM CUPRUM sp. z o.o. CBR, Wroclaw.



Control of Bubble Size in a Laboratory Flotation

Column

Miguel Maldonado, Dr. André Desbiens, Dr. Éric Poulin Alberto Riquelme

Dr. René del Villar

Université Laval Universidad de Concepción

Department of Electrical and Computer Engineering Department of Electrical Engineering

Department of Mining, Materials and Metallurgical Engineering

LOOP (Laboratoire d’observation et d’optimisation des procédés)

Québec (QC), Canada Concepción, Chile

Keywords: Flotation column control, frit-and-sleeve sparger, Wiener model, IMC control,

bubble size control, bubble size distribution

Abstract

Gas dispersion properties have proven to be key variables of the flotation process. Among these

properties, bubble surface area flux (BSAF or Sb) has been reported to linearly correlate with the

flotation rate constant; therefore, it is a potential manipulated variable to achieve a desired metallur-

gical performance. BSAF can be represented as a combination of two other gas dispersion proper-

ties: the superficial gas velocity and the Sauter bubble mean diameter; thus, controlling Sb implies

controlling bubble size and superficial gas velocity. This work focuses on the control of the Sauter

mean bubble diameter. To improve BSAF controllability, a so-called frit-and-sleeve sparger device

was implemented in a laboratory flotation column to regulate bubble size independently from super-

ficial gas velocity. For control purposes, the Sauter bubble mean diameter was indirectly calculated

from the bubble size distribution, estimated by using a Gaussian mixture model. An IMC controller

based on a Wiener model, was implemented in the laboratory flotation column set-up. Tracking

performance and rejection of gas velocity and unmeasured frother concentration variations were

then evaluated.

1 Introduction

Flotation is a commonly used method for separating valuable minerals (metal containing) from use-

less mineral (gangue). Its performance is determined by the valuable-mineral concentrate grade and

recovery. Whereas the first of these two variables can be measured on-line using an X-ray on-stream

analyzer (OSA), the latter must be estimated from steady-state material balance, which strongly lim-

Maldonado, Desbiens, Poulin, del Villar, Riquelme


its its use for regulatory control purposes. Moreover, the long sampling times of these OSA devices,

usually multiplexed, favour the use of a hierarchical control where secondary variables are con-

trolled to reject the frequent disturbances occurring in this type of process.

Recent studies [1] have shown that the performance of a flotation device basically depends on three

factors: the particle floatability, the froth recovery and the “bubble surface area flux”, a combination

of two other gas-dispersion properties, superficial gas velocity and bubble size [2]. This finding

suggests that for a well conditioned pulp, a given metallurgical performance can be achieved by

modifying froth recovery and bubble surface area flux. Consequently, two control approaches have

been proposed. The first focuses on controlling some froth characteristics, such as froth speed,

which have been widely applied [3-5], or froth appearance [6]. The second approach aims at con-

trolling some gas dispersion properties in the collection zone, such as gas hold-up and superficial

gas velocity [7-10]. This approach has been motivated by the recent availability of industrial gas-

dispersion sensors [11, 12]. This article extends previous works on control of some gas dispersion

properties, to the control of bubble size, represented by the Sauter mean bubble diameter, a key step

towards controlling the bubble surface area flux.

Bubble surface area flux can be mathematically expressed as follows:

32

6g

b

JS

d= (1)

where Jg is the superficial gas velocity and d32 is the Sauter mean bubble diameter. Since there is no

uncertainty associated to Equation (1), controlling BSAF implies controlling superficial gas velocity

and the Sauter mean bubble diameter.

In flotation column operation, bubble size is affected by frother type and concentration, gas rate and

sparging system. Frother dosage regulation is usually implemented using a ratio feed-forward con-

trol based on the actual tonnage processed. Nevertheless, because of the limitations of feed-forward

control in the presence of modelling uncertainties and unknown disturbances, such as frother persis-

tency, evaporation rate and most importantly, the effect of reprocessed water (still containing some

residual frother), frother concentration in a given flotation machine is hard to assess. Consequently,

and despite the fact that the authors have already proposed a method to evaluate on-line the frother

concentration [13], in this work this variable is considered as an unmeasured disturbance. Since

superficial gas velocity also modifies bubble size [2, 14], it influences BSFA directly through the

numerator of Equation (1) and indirectly by affecting the bubble size (denominator). Whenever fea-

sible, the sparger system adds another control degree of freedom to modify bubble size in flotation

columns. To take advantage of this fact, in this work a ‘frit-and-sleeve’ sparger [15] was imple-

mented, which allows the modification of the bubble size independently from gas velocity, thus

improving BSAF controllability.

Control of Bubble Size in a Laboratory Flotation Column


To obtain the benefits of feedback control, bubble size must be measured on-line. For diagnosis

purposes, Sauter mean bubble diameter is usually calculated after a set of collected images have

been processed by using the following equation:

3

132

2

1

N

i

i

N

i

i

d

d

d

=

=

=

∑

∑ (2)

where di is the equivalent diameter of bubble i and N is the number of counted bubbles. This calcu-

lation assumes a steady state condition, i.e. the data points (bubble sizes in this case) are sampled

from a steady bubble size density function, and as such, is not suitable for control purposes. In this

work an on-line estimation of the Sauter mean bubble diameter based on the bubble size density

estimation was implemented.

2 Flotation column set-up

The flotation column used in this work is composed of three sections made of polycarbonate tubes

for a total height of 5 m; the internal diameter of the bottom, intermediate and upper sections are

respectively 15.24 cm, 10.16 cm, and 5.08 cm. A frit-and-sleeve sparger is mounted in the bottom

part of the column as shown in Figure 1. Gas flow rate is measured through a mass flow sen-

sor/controller (Aalborg model GFC17), which also provides an estimate of the volumetric flow

based on a reference condition (21.1 oC and 101.3 kPa). Therefore, its readings must be converted to

the actual tests conditions of temperature and pressure, measured by the sensors shown in Figure 1

using the following equation:

2

1033.23 ( ) 273.15

1033.23 ( ) 294.16

oref

g g

a

T CJ J

P cmH O

+=

+ (3)

where Pa is the absolute pressure measured in cm H2O, T is the temperature in degrees Celsius and

Jgref

is the gas velocity calculated from the air mass flow meter readings at reference conditions. A

differential pressure transmitter DP (model ABB 264DS) was tapped between 250 cm and 320 cm

above the sparger to measure gas hold-up. For a two-phase system (air-water), this latter can be

measured using the following relationship:

(%) 100g

P

Lε

∆= ⋅ (4)

where ∆P is the pressure differential in cm H2O and L is the distance between taps, here 70 cm.

Data acquisition was performed by a HMI/SCADA software (iFIX®

) working under a Windows

XP®

operating system. An Opto 22 I/O system was used to centralize sensor and actuator signals. A



modified version of the McGill’s bubble viewer [11] was implemented to measure bubble size. The

image processing system was performed by a dedicated computer using MatLab’s Image Acquisi-

tion toolbox®

and ‘Image J’ free license software which was used to process the collected images to

obtain bubble sizes. A communication link between MatLab applications running on the control

station and the image processing computer was implemented using MatLab's Instrumentation tool-

box®

under Ethernet UDP protocol.

Figure 1: Column set-up.



3 Frit-and-Sleeve Sparger

The frit-and-sleeve sparger, depicted in Figure 2, consists of a porous ring surrounded by a sleeve

forming a gap through which water is passed to produce shear and consequently modifying bubble

size. This sparger provides another control degree-of-freedom to modify bubble size, the superficial

water velocity (Jls), i.e. the water flow rate through the gap divided by the cross sectional area of the

gap:

lsls

gap

QJ

A= (5)

where Qls is the volumetric water flow rate of liquid in the sparger and Agap is the cross sectional

area of the gap.

Figure 2: Frit-and-sleeve sparger.

4 Bubble size density estimation

To obtain the benefits of feedback control, it is necessary to continuously (on-line) measure the Sau-

ter mean diameter with a sampling time much shorter than its fastest dynamic. In this work, the Sau-

ter mean diameter was calculated from the estimation of whole bubble size distribution, represented

by its probability density function (PDF). The problem of estimating the density function from data

points is known as “density estimation”. Depending on the used model structure, these estimations

are classified in parametric, nonparametric and semiparametric methods [16]. Mixture models are

semi-parametric models, where the density function f(x) is represented as a linear combination of

functions φj(x) called component densities as follows:

Porous

ring

gap

sleeve

Top

View



1

( ) ( )M

j j

j

f x w xφ

=

= ⋅∑ (6)

1

1, 0 1, 1, ,M

j j

j

w w j M=

= ≤ ≤ =∑ … (7)

( ) 1, 1, ,j

x dx j Mφ

+∞

−∞

= =∫ … (8)

where M is the model order, i.e. the number of components considered. The coefficients wj are

called mixing parameters and represent the prior probability of a data point having been generated

from component j of the mixture model.

The component densities could be any density function or even a combination of density functions

with different functional forms. Nevertheless, depending on their properties and structure, the pa-

rameter identification might become very difficult. An interesting case is obtained when the compo-

nents φj (x) are chosen to be fixed Gaussian functions.

( )2

2

1 1( ) exp

22

j

j

xx

µ

φ

σπ σ

−

= − ⋅

(9)

where µj , j =1,…,M , while M and σ are fixed parameters.

Minimizing the log-likelihood of the samples and applying a stochastic gradient descent approach

the following recursive equation is obtained [17]:

ˆ ˆ ˆ( 1) ( ) ( / ( 1)) ( )j j j

w k w k P j x k w kη + = + + − (10)

When η ��is selected to be constant, it introduces an exponentially decaying envelop η (1- η)k-i

applied to the influence of the data point x(k-i) [18]. P(j|x(k)) is the posterior probability determined

using Bayes’ theorem as follows:

( ( ))( | ( ))

( ( ))

j jx k w

P j x kf x k

φ ⋅

= (11)

The Sauter mean bubble diameter can be estimated on-line from the Gaussian mixture parameters as

follows [19]:

( )

( )

3 2 2

1

322 22

1

ˆ, ( ) ( 3 )

ˆ ( )

ˆ ( )( ),

M

j j j

j

M

j j

j

x f x k dx w k

d k

w kx f x k dx

µ µ σ

µ σ

+∞

=−∞

+∞

=−∞

⋅ +

= =

+⋅

∑∫

∑∫ (12)



Figure 3 shows an example of bubble size density estimation using the previously described

algorithm.

Figure 3: Example of a bubble size density estimation using a mixture of thirty equally-spaced

fixed Gaussian distributions (σ = 0.005).

5 The Wiener model

A Wiener model is used to represent the dynamic relationship between superficial water velocity

and Sauter mean diameter. Wiener models consist of a linear system in series with a memory-less

(static) nonlinear element, as shown in Figure 4. When this model is used for control purposes, the

nonlinear element must be selected in such a way that the nonlinearity function is invertible.

Figure 4: Wiener model structure.

The Wiener model was identified in two steps; first, the static nonlinear element was determined by

using steady-state information, then, the linear dynamic element was identified by using step re-

sponse information. The technique used in either case is described hereafter.



5.1 Static nonlinear element

To explore the static relationship between superficial water velocity in the sparger and Sauter mean

bubble diameter, several experimental tests were conducted in the laboratory flotation column. Fig-

ure 5 contains three subplots, each for a given superficial gas velocity, i.e. 0.5, 0.8 and 1.1 cm/s. For

a given superficial gas velocity, each subplot shows a monotone decreasing steady state relationship

between superficial water velocity in the sparger and the Sauter mean bubble diameter for three dif-

ferent frother concentrations. In general, it can be observed that the effect of frother concentration is

to shift this decreasing relationship without significantly modifying its shape.

Figure 5: Static relationship between Sauter mean diameter and the superficial water velocity in

the sparger for different frother concentration and superficial gas velocities.

Figure 6 shows the same pattern; in this case, it can be observed that superficial gas velocity acts by

shifting the nonlinear relationship between superficial water velocity and Sauter mean diameter.

Figure 6: Static relationship between Sauter mean diameter and the superficial water velocity

through the sparger for different frother concentration and superficial gas velocities.



Figure 7 shows the static gain (∆d32/∆Jls) corresponding to the nonlinear relationships shown in

Figures 5 and 6 as a function of the superficial water velocity in the sparger. The selected nominal

gain shown in Figure 7 corresponds to the following nonlinear relationship between Sauter mean

diameter and superficial water velocity:

0.256

32 3.706 0.226ls

d J−

= ⋅ − (13)

where 32d and ls

J are respectively the steady-state Sauter mean diameter and superficial water

velocity.

Figure 7: Static gain of the system vs superficial water velocity in the sparger.

5.2 Linear dynamic element

A simple unitary gain, first-order lag system with time delay was used to represent the dynamic be-

haviour between superficial water velocity in the sparger and the Sauter mean diameter. Notice that

the static gain is captured by the nonlinear static block. The following nominal model was identified

using step response information:

90

( )90 1

se

G ss

−

=

+

ɶ (14)

Figure 8 shows the Wiener intermediate variable v(t) (see Figure 4), in this case, a filtered version of

the superficial water velocity obtained by inverting the nonlinearity function (Equation 13) and the



simulated response of the identified linear system (Equation 14). It can be seen that the Wiener

model is able to capture reasonably well the nonlinear dynamic behaviour of the process.

Figure 8: Wiener model validation.

6 Wiener model based control

Assuming there is no uncertainty in the non-linear block, the static non-linearity can then be com-

pletely removed by performing the inverse of the non-linearity (see Figure 9). Consequently, the

controller synthesis can be performed by the classical internal model control (IMC) framework, i.e,

by minimizing the H2 norm of the error e for a given input v:

2( ) ( ) 2min min 1 ( ) ( ) ( )Q s Q s

e G s Q s v s = − ⋅ ɶ ɶ

ɶ ɶ (15)

For a step input, i.e. v(s) = 1/s and a stable system factored into ( ) ( ) ( )M AG s G s G s=ɶ ɶ ɶ where ( )

AG sɶ is an

all-pass transfer function, containing right half plane zeros and delays, and ( )MG sɶ a minimum phase

transfer function, the optimal H2 controller is:

1( ) ( )M

Q s G s−

=ɶ ɶ (16)

Then, the controller is augmented by a filter F(s):

( ) ( ) ( )Q s Q s F s=ɶ (17)

where the filter F(s) is usually selected to be:

( )

1( )

1r

F ssλ

=

+ (18)

0 2000 4000 6000 8000 10000 12000 14000 16000 180000

50

100

150

200

250

300

350

time (s)

inte

rmed

iate

vari

ab

le,

v(t

)

v(t)estimated

v(t)



and r is selected to obtain a realizable controller. The complementary sensitivity function is given

by:

( , ) ( )( , )

1 ( , ) ( ) ( )

Q s G sT s

Q s G s G s

λλ

λ

⋅=

+ ⋅ − ɶ

(19)

The maximum peak (Mp) tuning [20] is used to find the value of IMC filter λ for parametric uncer-

tainty in the process model. The objective is to find the smallest IMC filter time-constant that as-

sures that no closed-loop frequency response (complimentary sensitivity function) will have more

than specified Mp (i.e., the specified maximum peak).

Considering a 20 % uncertainty in time delay and time constant and a 5 % uncertainty in the static

gain of the nominal process model (Equation 14), the following IMC controller was obtained:

90 1( )

82.6 1

sQ s

s

+=

+

(20)

For implementation purposes, this controller was discretized with a sampling time of 5 seconds.

Figure 9: IMC based Wiener model.

7 Experimental results

To evaluate performance of the control system, the laboratory flotation column was first filled with

a solution of 10 ppm frother concentration Dowfroth 250. Then, changes on superficial gas velocity

and frother concentration were implemented to simulate disturbances affecting the bubble size con-

trol loop.



Figure 10 shows the controlled variable, i.e, Sauter mean diameter, and the manipulated variable,

the superficial water velocity in the sparger. A good tracking performance can be observed when the

pump speed is not in a saturated condition. Approximately at 12,000 s, a gas velocity step change

from 0.5 to 0.8 cm/s was implemented. This originated an increase in bubble size, which was rap-

idly compensated by increasing superficial water velocity in the sparger. Later on, at about 13,000 s,

the bubble size set-point was decreased to 0.6 mm, but this new set-point was not achievable since

the pump speed saturated (at Jls = 340 cm/s). Finally, frother concentration was changed from

10 ppm to 20 ppm at around 16,000 s. It can be seen that the bubble size was consequently affected

(reduced), but the control system immediately reacted by decreasing the superficial water velocity,

which allowed the bubble size to return to its previous value (set-point).

Figure 11 shows the time evolution of both gas hold-up and bubble surface area flux for test condi-

tions shown in Figure 10. A significant correlation can be seen between gas hold-up and bubble

surface flux, initially suggesting that both variables carry similar information and consequently ei-

ther variable could be used for control purposes.

Figure 10: IMC based Wiener model: Tracking and regulation performance.



Figure 11: IMC based Wiener model: relation between Sauter mean diameter, gas hold-up and

BSAF.

8 Discussion

It has been shown that for controlling BSAF, reference values for superficial gas velocity and Sauter

mean diameter must be provided. Different combinations of gas velocity and Sauter mean diameter

could generate the same BSAF, thus an optimal combination of these variables producing a desired

BSAF must be determined based on the desired metallurgical performance. Something similar oc-

curs with gas hold-up, since different conditions of gas velocity, frother concentration and bubble

size can generate the same gas hold-up value. As it is widely accepted that bubble size plays a key

role on flotation performance, one possible control approach is to regulate bubble size by modifying

the superficial water velocity in the sparger and to regulate bubble surface area flux (or gas hold-up)

by manipulating gas velocity.

One potential problem with controlling bubble size via Sauter mean diameter is that all the available

information regarding the shape of the bubble size distribution such as multi-modal and tailing



behaviour is completely lost in this averaging exercise. In fact, it is possible to generate very differ-

ent bubble size distributions having the same Sauter mean diameter. Therefore, using this latter for

control purposes will be appropriate only for unimodal narrow bubble size distributions.

In this study, frother concentration was considered to be an unknown disturbance. However, since it

has a strong effect on bubble size, its use as a manipulated variable deserves to be explored. More-

over, for mechanical cells where no spargers are used, it becomes a more relevant manipulated vari-

able to modify bubble size.

9 Conclusions

This work explored the control of Sauter mean diameter as a first step towards the control of bubble

surface area flux. A frit-and-sleeve sparger was implemented to allow the modification of the bub-

ble size by injecting water through a gap.

The Sauter mean diameter was calculated on-line from the measured (image processing) bubble size

distribution. A Gaussian mixture model was proposed and its parameters were determined by mini-

mizing the log-likelihood of the data points and using a gradient descent method.

A Wiener model-based IMC was proposed to control the Sauter mean diameter by modifying the

superficial water velocity through the frit-and-sleeve sparger. Good tracking performance (bubble

size set-point) and rejection of unmeasured changes in superficial gas velocity and frother concen-

tration were obtained in a laboratory flotation column.

Acknowledgements

This work was conducted under the project NSERC (National Science and Engineering Research

Council of Canada) with the following industrial partners: COREM, Xstrata-Ni Strathcona Mill,

Agnico-Eagle Mines Division Laronde.

Authors would like to acknowledge the McGill mineral processing group for providing a frit-and-

sleeve sparger prototype. M. Maldonado would also like to thank the Chilean Council for Science

and Technology (Conicyt) for partially financing his Ph.D. studies at Université Laval.



References

[1] GORAIN, B.K., FRANZIDIS, J-P., MANLAPIG, E.V. (1997): Studies on impeller type,

impeller speed and air flow rate in an industrial scale flotation cell – Part4: Effect of bubble

surface area flux on flotation performance. Minerals Engineering, Vol.10, No.4, pp.367-379.

[2] FINCH, J.A. AND DOBBY, G.S. (1990): Column Flotation, Pergamon Press, Oxford (UK).

[3] BARRIA, A., VALDEBENITO, M. (2008): Implementation of rougher flotation control sys-

tem at Codelco Chile, Andina Division. Procemin 2008, 5th

International Mineral Processing

Seminar, Santiago, Chile, pp.215–220.

[4] CORTES, G., VERDUGO, M., FUENZALIDA, R., CERDA, J., CUBILLOS, E. (2008):

Rougher flotation multivariable predictive control; Concentrator A-1 division Codelco Norte.

Procemin 2008, 5th

International Mineral Processing Seminar, Santiago, Chile, pp. 315–325.

[5] MOILANEN, J., REMES, A. (2008): Control of the flotation process. Procemin 2008, 5th

In-

ternational Mineral Processing Seminar, Santiago, Chile, pp. 305–313.

[6] LIU, J.J., MACGREGOR, J.F. (2008): Froth-based modeling and control of flotation proc-

esses. Minerals Engineering, Vol.21, pp.642–651.

[7] BERGH, L.G. and YIANATOS, J.B. (1993): Control alternatives for flotation columns. Min-

erals Engineering, Vol. 6, No. 6, pp. 631-642.

[8] CARVALHO, T. and DURÃO, F. (2002): Control of a flotation column using fuzzy logic

inference. Fuzzy Sets and Systems, Vol.125, pp.121-133.

[9] PERSECHINI, M.A.M, JOTA, F.G., PERES, A.E.C. and GONÇALVES, F (2004): Control

strategy for a column flotation process. Control Engineering Practice, Vol.12, pp.963-976.

[10] MALDONADO, M., DESBIENS, A. and DEL VILLAR, R. (2009): Potential use of model

predictive control in optimizing the column flotation process. International Journal of Mineral

Processing, Vol.93, pp.26-33.

[11] GOMEZ, C.O. and FINCH, J.A. (2007): Gas dispersion measurements in flotation cells. In-

ternational Journal of Mineral Processing, Vol.84, pp.51-58.

[12] O’KEEFE, C., VIEGA, J., and FERNALD, M. (2007): Application of passive sonar technolo-

gy to mineral processing and oil sands applications. Proceedings of the 39th

Annual Meeting

of the Canadian Mineral Processors, CIM, Ottawa (Canada), pp.429-457.

[13] MALDONADO, M., DESBIENS, A, DEL VILLAR, R. and AGUILERA, R. (2009): On-line

estimation of frother concentration in flotation processes. 48th

Annual Conference of

Metallurgists, Sudbury, Canada, pp.35-146.



[14] NESSET, J.E., HERNANDEZ-AGUILAR, J., ACUÑA, C., GOMEZ, C.O. and FINCH, J.A.

(2006): Some gas dispersion characteristics of mechanical flotation machines. Minerals Engi-

neering, Vol.19, pp.807-815.

[15] KRACHT, W., GOMEZ, C.O and FINCH, J.A. (2008): Controlling bubble size using a frit-

and-sleeve sparger. Minerals Engineering, Vol.21, pp.660-663.

[16] BISHOP, C.M. (1996): Neural network for pattern recognition, Oxford University Press, 1st

edition.

[17] TRÅVÉN, H. (1991): A neural network approach to statistical pattern classification by semi-

parametric estimation of probability density functions. IEEE Transactions on Neural Net-

works, Vol.2, No.3.

[18] ZIVKOVIC, Z. and VAN DER HEIJDEN, F. (2004): Recursive unsupervised learning of fini-

te mixture models. IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol.26,

No.5, pp.651-656.

[19] MALDONADO, M., DESBIENS, A., DEL VILLAR, R., GIRGIN, E. and GOMEZ, C.

(2008): On-line estimation of bubble size distributions using Gaussian mixture models.

Procemin 2008, 5th

International Mineral Processing Seminar, Santiago, Chile, pp.389-398.

[20] STRYCZEK, K, LAISECA, M, BROSILOW, C, LEITMAN, M. (2000): Tuning and design

of single input, single output control systems for parametric uncertainty. AIChE Journal, Vol.

46, No.8, pp.1616-1631.


Flow Process in the Aerator of the

Flotation Machine – Preliminary Simulations

Adam Mańka, Adam Fic, Andrzej Sachajdak, Ireneusz Szczygieł

Silesian University of Technology

Institute of Non-ferrous Metals, Sowińskiego 5

Institute of Thermal Technology, Konarskiego 22

44-100 Gliwice, Poland

Keywords: Flotation, technology, computation, CFD modeling

Abstract

In this paper, the numerical aspect of the flow phenomena in the flotation machine is presented. The

efficiency of the flotation process strongly depends on the fluid flow in the main part of the ma-

chine: in the aerator. This element is responsible for mixing suspended solids and dispersing air.

The paper presents preliminary results of mathematical modeling. The process of flotation consists

of a number of phenomena which provide serious numerical difficulties. One can enumerate rota-

tion, two phase flow, foam formation etc. To the knowledge of authors there is no complete numeri-

cal model available for the flotation machine. The long term task of the project is to create a com-

plete model of the machine. Such a model would be very helpful in the process of constructing and

modernization of the flotation machine. As it was mentioned, due to the difficulties connected with

flotation phenomena modeling, only some of them were taken under consideration. First step, de-

scribed in this paper, shows how to handle one phase flow in the aerator. The commercial package

Ansys Fluent was utilized for the analysis. The results were compared with the measurements per-

formed on the real machine. Obtained results are satisfying and encouraging for further develop-

ment.

1 Introduction

Flotation is the industrial process of particle separation by dissolving air in the pulp under pressure

and then releasing the air at atmospheric pressure into a flotation tank. The released air forms tiny

bubbles which adhere to the suspended matter causing it to float to the surface of the water where it

may then be removed by a skimming device.

Mańka, Fic, Sachajdak, Szczygieł


The process is realized in the flotation machine. An example is shown in Figure 1. In the machine

complicated, multiphase (gas, liquid and solid) fluid flows take place. In the near past, modeling of

such a process was impossible due to the lack of sufficient computer capacity and the processor ef-

ficiency. So, construction of flotation machines based mainly on empirical relationships and on the

experience of the constructors. Nowadays, the numerical simulations of flotation phenomena be-

came possible due to powerful enough machines as well as due to the sophisticated CFD packages

like Fluent or CFX. Recently first attempts of modeling flotation processes using CFD packages can

be noticed [1, 3, 4]. There are even some thematic conferences devoted to numerical modeling of

flotation [4]. Unfortunately, these first attempts provide only simplified models of partial phenome-

na taking place during flotation. So, there is still a lot of work to do in this field.

The flotation machine consist of tank, rotor and stator (aerator), air feed mechanism as well as pulp

delivery and discharge mechanism. Air bubbles are provided to the pulp by the aerator which con-

sists of rotor and stator. The aerator is shown in the Figure 2. Details of the aerator construction

strongly influence the quality of the final products, as well as operational parameters of the ma-

chine: air and power consumptions. The Computational Fluid Dynamics (CFD) tool describing the

behaviour of the fluids in the flotation machine, especially in the aerator surroundings would be

very helpful in the optimization process of the flotation machine. The partial model of such machine

was presented in [1].

Figure 1: Flotation machine

Flow Process in the Aerator of the Flotation Machine – Preliminary Simulations


The fluid flows in the flotation machine show complex physical dynamics which have a reflection

in complicated mathematical expressions. The features of the flow are:

• three dimensional flows,

• turbulent flows,

• multiphase flows,

• unsteady state.

Figure 2: Aerator

The phenomena in the flotation machine are described by:

• continuity equation,

• momentum equation,

• energy equation,

• turbulence model,

• rotation model,

• material transport,

• foam formation model,

• phase model.



At this stage of computation, the energy balance, multiphase flow and foam formation are omitted

what makes the system of partial differential equations simpler, but what does not mean it is simple.

With the mentioned simplifications the governing equations can be presented as follows:

• continuity equation:

(1)

• momentum equation:

(2)

• turbulence model.

For the solution of the system of equations the commercial package Fluent was employed. Fluent

utilizes the controlled volume method to convert the governing differential equations into algebraic

ones. Unfortunately, the set of governing equations is nonlinear, so an iterative procedure of solu-

tion is necessary. Additionally, the swirling elements of the aerator are the source of serious troubles

during the creation of a numerical model. There are several possibilities of rotation consideration:

Mixing Plane Model, Multiple Reference Frame Model and Sliding Mesh Model. All of them were

taken under consideration, but as the investigations showed, not all of them can be applied. Mixing

plane model can not be applied due to the multiphase flow. The distance between the rotor and sta-

bilizer (stator) blades is low (about 30 mm), which is the source of the transient character of the

phenomena taking place in the aerator. Due to that the only model of rotations which can be used is

“Sliding Mesh”. The rotating and stationary parts of the aerator should be divided by a so-called

interface. The mesh in the rotating region slides over the stationary cells along this interface. The

negative effect of such an attempt realizes in the long term of computations. The blade transfer be-

tween the appropriate stator blades should be stretched to the several dozens of time steps, which by

the practical rotational speeds of the rotor makes this time step extremely short.

The presented system of governing equations has to be accomplished with the well-matched pack of

initial and boundary conditions. Improper selection of boundary conditions can result in enlarging of

computational time, or, what is worse, in loss of convergence.

The following boundary conditions were set:

• wall: no slip condition at the walls of the tank

• rotor: rotation 30 rad/s

• free surface at the top of the tank

• in the first stage of simulation no air inlet were defined

As the initial condition a motionless rotor and motionless fluid was assumed.

( ) 0=⋅⋅∇+

∂

∂wρ

ρ

t

pDt

D∇−⋅∇⋅=

⋅w

w) 2(µ

ρ



1.1 Grid construction

Mesh definition is a significant part of CFD modeling. Although each commercial CFD package is

equipped with mesh generator, this task should be realized very carefully. In fact proper definition

of mesh is one of the most consuming parts of the whole modeling process. The quality of the mesh

size significantly influences the quality of the solution.

In the presented model, the structure of the mesh was constructed with tetrahedron elements. As it

was mentioned previously, rotating and motionless parts of aerator were divided by an interface.

The total number of elements is 3 millions. One exemplary part of mesh is presented in the Figure 3.

Figure 3: Mesh generated in the rotor domain

2 Results of computations

Two types of stators were investigated – the typical one, which is shown in the Figure 4 and the

modified one shown in Figure 5. For these two cases the full CFD procedure was performed. Within

that, the fields of velocity, pressure, turbulence intensity were calculated. The results of computation

are demonstrated in the Figures 6-13.



The results of numerical analysis was qualitatively compared with the results of the real numbers,

observed at the machine. In the authors opinion, the obtained results encourage for further investiga-

tions.

Figure 4: Aerator with traditional stator

Figure 5: Aerator with modified stator



Figure 6: Velocity distribution at the outlet of traditional stator

Figure 7: Velocity distribution at the outlet of modified stator



Figure 8: Velocity distribution at the horizontal cross section of the traditional stator

Figure 9: Velocity distribution at the horizontal cross section of modified stator



Figure 10: Velocity vectors at the outlet of the stator

Figure 11: Velocity vectors at the outlet of the rotor



Figure 12: Path lines (flow disturbances) in the tank for the traditional stator

Figure 13: Path lines (flow disturbances) in the tank for the modified stator



Figure 14: Flotation machine at work

Figure 15: Outlet of rotor



Figure 16: Outlet of stator

3 Conclusion

CFD is a helpful but still not complete tool for visualizations and checking design assumptions in

flotation machines. Result of the first view on this topic is the development of new, more efficient

numerical models for multiphase flows, planned to introduce in CFD packages. As was mentioned

above the results of numerical analysis was qualitatively compared with the results from the real

machines. Finally the CFD-aided design tool for flotation processes are planned to be prepared.

Presented simulations proved that the modified stator introduces less flow disturbance in the flota-

tion process.



References

[1] DEGLON D. A., Meyer C. J., CFD modelling of stirred tanks: Numerical considerations.

Minerals Engineering, 19, 2006, 1059-1068.

[2] Fluent 6.2 User’s Guide. Fluent Inc., 2005.

[3] KOH P. T. L., SCHWARZ M. P., CFD model of self-aerating flotation cell. Int. J. Mineral

Process, 85, 2007, 16-24.

[4] TIITINEN J., VAARNO J., GRÖNSTRAND S., Numerical modeling of an Outokumpu flota-

tion device. 3th

Int. Conf. on CFD In The Minerals and Process Industries, Melbourne,

Australia, 2003.



Analysis of Fine Particles Behaviour in Flotation

of Polish Copper Ores

Ph.D. Eng. A. Potulska

KGHM CUPRUM Ltd Research & Development Centre

Street Gen. Władysława Sikorskiego 2-8

53-659 Wrocław, Poland

Keywords: Flotation, copper ores, liberation

Abstract

One of the most important reasons for lowering the nonferrous metal ores concentration indexes is

the finer and finer mineralization and in consequence the problems with liberation of the sulphide

minerals from the ore. This problem applies to both Polish and world non-ferrous metal ores. In

order to define the possibility for improving the concentration of sedimentary type ores – occurring

in LGOM deposit – laboratory tests were carried out aimed at definition of the influence of different

factors, such as rotational impeller speed, xanthate consumption and air flow rate on flotation effec-

tiveness of fine particles of Polkowice ore. Moreover, the influence of ore liberation on the fine par-

ticles flotation results was investigated. The best results were obtained for ore liberated in 95 %,

which means that 90 % of the particles were below 0.025 mm. In order to accurately define the be-

havior of fine particles, the mineralogical and chemical analyses were made for narrow size classes:

0.010; 0.010-0015; 0.015-0.025; 0.025-0.040 and +0.040 mm for all concentration products. The

results of improved concentration tests showed, that effective flotation of the feed with 90 % of the

particles <0.025 mm is possible.

1 Introduction

For cost and technical effective production of metallic copper, on the concentration stage it is neces-

sary to obtain a flotation concentrate, with the required Cu content, and as high as possible recovery

rates. Achieving such concentration indexes affords in all cases a high level of Cu minerals libera-

tion from the gangue components. It is carried out through grinding, and is the more difficult the

finer is the ore mineralization. Thus, the concentrator trying to liberate the copper bearing minerals,

produces certain amounts of undesired fine and ultra-fine particles, which float worse than the rest

of the feed in current circuit. Due to the grinding circuits constrains it is impossible to prevent effec-

tively their generation. From the other side, due to increasing non-ferrous metals demand, it is more

Potulska


and more often necessary to extract the deposits, where metals are bound in the finest particle

classes. Therefore, effective production and flotation of fine particles is a common problem, and

increasing of their recovery becomes one of the biggest research challenges in the mineral

processing and a condition for improvement of competitiveness for many concentrators.

Despite of identification of this problem almost 100 years ago, and substantial amount of tests, only

a limited number of ways for fine flotation, have been developed until now. In general, two methods

of fine flotation are proposed. In the first one all particles are floated together since it is considered

that fine particles are taken to the froth product together with coarse particle. The second solution

considers separate flotation of the particles having various sizes, which enables optimization of the

process conditions differently for fine and coarse particles.

The method for separate flotation of the individual size classes was described already in 1927 by

Taggart [1] who pointed out its advantages, and functionality of application. During tests on flota-

tion of very fine grained lead-zinc ore from Mt Isa – Australia, it was shown [2, 3] that application

of very fine regrinding [4] and floating very narrow size classes gives a Pb recovery increase by

5 %, and by 10 % for Zn; moreover, a rise of Pb content in concentrate by 5 % and Zn by 2 %, as

well as improvements of process stability, were also stated. Such results were possible to obtain due

to high level of sulfide minerals liberation during regrinding in an Isa Mill. The lead and zinc

rougher concentrate was reground to 0.012 mm, and zinc cleaner tailings to 0.007 mm.

All papers, starting from the first half of the twentieth century, such as already quoted classical pa-

pers of Taggart [1] and Gaudin [5], through later ones [6-8] – more detailed, elaborated in the se-

venties stress first of all the influence of particle size on flotation and in the smaller extend a fine

grinding. Most of the utility investigations were made for already mentioned fine-mineralized lead-

zinc ore deposits in Australia. Lately, similar experiments were also made in the Zn-Pb ore concen-

trator in Ireland. However, they covered narrow scope and only partial results were published [9].

It must be pointed out that those studies dealt with the so called „primary” type of ores, generated

during magmatic processes. In the word literature there is no wide range of investigations, concern-

ing the specific, sedimentary type ores – which are mined for example in Poland. Attempts to find

out the reason of worse floatability of the Polish copper ores, and to find the necessary measures

were being undertaken many times [10, 11]. However, they have not given any satisfactory results.

In the case of the Polish copper concentrators, the issue of fine particle concentration improvement

is considered as one of the most important problems [12, 13]. The possibility of IsaMill application

to improve the flotation of „difficult-to-float” middlings from Polkowice concentrator, which con-

tain lots of not liberated fine valuable minerals, was also studied [14], but due to very tough condi-

tions to concentrate this material, and limited scope of investigations, the improvements were in-

considerable.

Analysis of Fine Particles Behaviour in Flotation of Polish Copper Ores


2 Characteristic of Polish copper ores

The copper deposit mined by KGHM Polska Miedź S.A. substantially differs from typical, world-

wide extracted porphyry deposits. Main differences are: presence of silver, which is a source of ad-

ditional profits, and a specific mineralization, that causes serious problems encountered especially

during mineral processing stage. These unorthodox types of ore make many world experiences with

copper ore processing technologies applicable only in very narrow extend with reference to Polish

copper ores.

One of the more important features of the ore processed by KGHM Polska Miedź is a presence of

three types of rocks with different kinds of mineralization, i.e. sandstone, shale and limestone. In

1985-1991, average compositions of ore processed in the concentrators were as follows [12]:

Sandstone Shale Carbonates

Lubin concentrator 54.0 % 8.1 % 37.9 %

Polkowice concentrator

Rudna concentrator

9.9 %

43.3 %

5.9 %

5.5 %

84.2 %

51.2 %

In spite of changes, which have been occurred since then, the processed ores maintained their basic

nature and the feeds from individual concentrators are basically still:

• sandstone ore – Lubin

• carbonate ore – Polkowice

• ore with equal amounts of sandstone and carbonate content – Rudna [15].

Sandstone ore has a form of light-grey, fine grained, compact sandstones, containing mainly quartz

and small amounts of feldspars and other minerals, bound by carbonate or clayey binder. Metal

bearing mineral particles are mostly not bigger than 0.200 mm, and in general, within the range be-

tween 0.050 and 0.200 mm. Carbonate ore occurs in the form of lime dolomites and less often of

dolomite lime stones. Minerals of gangue are mainly dolomite, calcite, anhydrite and clay minerals.

Metal bearing minerals mostly lay in the range from 0.030 to 0.200 mm. Shale ore contains about

85 % of clay minerals and carbonates, about 7 % of. organic matter and small amounts of quartz.

Copper minerals are predominantly in the size from 0.005 to 0.040 mm [16].

It must be stressed, that in all ore types, significant part of copper bearing minerals have particle

sizes even smaller than the finest of the above mentioned limits. That is why improvement of flota-

tion effectiveness of these fine particles is now a crucial issue in raising the total Cu recovery. Pres-

ence of three lithological components in the ore, having different and variable crushing and grinding

properties and different floatability, as well as variability of their contents, create specific challenges

both at the grinding and the flotation stage. Particular problems cause shales during flotation or even

dewatering.

Potulska


Copper ores processed in all concentrators occur mainly in the form of chalcocite, bornite or chal-

copyrite as well as other minor copper bearing minerals. Average content of the most important

copper bearing sulphide minerals in 2007-2008 is presented in Table 1 [17].

Table 1: Average content of the most important copper bearing sulphide minerals in the

processing plants of KGHM Polska Miedź in 2007-2008 [17]

Bornite Chalcocite,

digenite

Chalco-

pyrite

Pyrite,

markazite Coveline Sphalerite

Tennan

tite Galena

Lubin

concentrator 33.1 % 14.5 % 26.1 % 17.4 % 3.5 % 1.8 % 2.0 % 1.9 %

Polkowice

concentrator 13.6 % 55.5 % 12.0 % 10.9 % 3.7 % 1.2 % 1.5 % 1.6 %

Rudna

concentrator 24.8 % 40.4 % 7.2 % 15.0 % 5.8 % 3.1 % 1.5 % 2.3 %

One may identify the following reasons for rating the Polish copper ores among the so-called diffi-

cult ones:

- presence and variability (quantity, quality) of three ore types, which require different grinding

and flotation conditions,

- small sized sulphide mineral particles and thus resulting the necessity of fine grinding,

- relatively fast transport of some gangue minerals to concentrate, especially shales, which worsens

or makes impossible obtaining a good quality of concentrates,

- slow kinetics of sulphide mineral particles that result in relatively long flotation times,

- poor separation of impurities during flotation process, especially Corg, As and Pb,

- relatively long time required to thicken final concentrates - their moisture content amounts to

11-13 %, even after filtration in a filter press, and that is the reason why the filter cakes require

additional thermal drying.

Such problems generally are not encountered in such extent in the case of world wide mined por-

phyry copper ores, and in other sedimentary types of ores – at least, they are not described in the

literature. The detailed analysis showed that an increase of the fine copper bearing minerals libera-

tion and improvements of their flotation should help to eliminate or reduce the most important

problems.



3 Characteristic of fine size classes concentration in the

industrial scale

Despite of the differences between the feeds processed in the individual concentrators, the concen-

tration results are quite similar due to proper selection of technology (see Table 2).

Table 2: Copper ore processing results in Polkowice and Rudna concentrators in 2008 [18]

Concentrator Metal content Concentration

ratio

Recovery,

% Feed Concentrate Tailings

Copper, %

Polkowice 1.74 24.98 0.21 14.0 87.9

Rudna 1.86 26.19 0.23 14.1 88.4

Silver, ppm

Polkowice 31 424 5 13.7 85.9

Rudna 46 605 6 10.3 81.8

In Table 2 results of only two concentrators were presented. In the last one – Lubin concentrator –

the concentrate contains only 15-16 % Cu, but the Ag content is the highest among all KGHM con-

centrators and amounts to 750-820 g/t. Authors [18] explained the considerably low copper content

in Lubin’s concentrate as an effect of specific properties of waste rock – mainly the presence of or-

ganic matter. Besides mineral and petrographic composition of the ore being processed in the Lubin

concentrator, also the grain size distribution of ore minerals and their associations are of significant

importance with regard to concentration rates [18]. These last two factors also strongly determine

the possibility to obtain concentration indexes in two other KGHM concentrators. That is the reason

why one of the proposed solutions can offer the essential improvement of the concentration results,

namely the finest grinding.

Tests carried out in the KGHM Polska Miedź concentrators indicate that the amount of metals lost

in the finest particle classes of tailings comes from the carbonates flotation line and exceeds some-

times even 60 % of all losses. Table 3 presents the example data for Lubin concentrator, including

particle size distribution of tailings comming from flotation line of sands (sandstone tailings) and

flotation line of carbonates (carbonate tailings) from Lubin, copper content and its distribution in

individual particle classes.

Potulska


Table 3: Particle size distribution of flotation tailings, copper content and its distribution in size

classes for sands and carbonate tailings in Lubin concentrator [18]

Class, mm Recovery, % Cum. % pass. Cu content, % Cu distribution, %

Sands tailings

0.150 8.7 91.3 0.33 22.2

0.100 34.1 57.2 0.10 26.5

0.075 8.7 48.5 0.09 6.1

0.044 9.2 39.3 0.12 8.6

-0.044 39.3 - 0.12 36.6

N 100.0 100.0 0.13 100.0

Carbonates tailings

0.150 2.6 97.4 0.60 8.8

0.100 11.4 86.0 0.20 12.8

0.075 8.5 77.5 0.19 9.0

0.044 13.6 63.9 0.16 12.2

-0.044 63.9 - 0.16 57.2

N 100.0 100.0 0.18 100.0

Presented results show the differences between sandstone and carbonate ores processed in KGHM

concentrators and their tailings. Sandstone and carbonate tailings have different particle size distri-

bution (carbonate tailings are finer) and copper content (carbonate tailings contain more copper).

The exceptionally high participation of 0.10-0.15 mm particle class is specific for sandstone tailings

and reaches even 34.4 %. It is caused by accumulation of the quartz particles, difficult to grind.

Copper content in the finest particle class (<0.044 mm) for all types of tailings is indeed usually the

lowest one but in this class, the highest copper mass is accumulated. A similar type of Cu distribu-

tion is observed in the case of the Polkowice concentrator (carbonate tailings) and the Rudna con-

centrator (sandstone and carbonate tailings).

4 Results

In order to determine the influence of particle sizes on flotation results for particles below

0.044 mm, especially below 0.020-0.015 mm, and to propose the remedial measures, detailed labor-

atory tests on grinding and flotation of the Polkowice ore were carried out [19]. The subject of these

tests was fine-mineralized ore from the Polkowice concentrator having the following lithological



composition: carbonate ore (51 %), sandstone ore (13 %) and shale ore (36 %). Among many para-

meters deciding about the flotation results, the ones whose change in industrial practice is relatively

simple were selected:

° particle size distribution of feed,

° rotation speed of impeller,

° aeration level of flotation slurry,

° amount of added collector.

To determine the influence of particle size on the flotation results, the laboratory flotation tests were

carried out with feed having the following particle size distribution: 65, 80 and 90 % below 0.025 mm.

For all samples the same flotation conditions were used, i.e. aeration – 80 dm3/h of air, impeller rotati-

on – 670 rpm, collector consumption (potassium xanthate) – 100 g/Mg and frother consumption (α–

terpineol) – 100 g/Mg of ore. The achieved results were recalculated using Fuerstenau`s curves for the

same amount of gangue recovery in tailings εr = 80 % (see Figure 1). That type of curves enables a

proper comparison of flotation experiments because they are insensible to changes of mineral content

in feed [20]. The following concentration results were obtained:

65 % <0.025 mm ε =87.4 % β =7.8 % γ =23.2 %

80 % <0.025 mm ε =88.8 % β =7.9 % γ =23.3 %

90 % <0.025 mm ε =89.8 % β =8.0 % γ =23.3 %

ε – copper recovery in concentrate, εr – gangue recovery in tailings.

Figure 1: Polkowice ore flotation vs. Figure 2: Polkowice ore ground to 90 %

particle class <0.025mm content <0.025mm flotation vs. consump-

(in Fuerstenau system) tion air (in Fuerstenau system)

0

20

40

60

80

100

87 89 91 93 95 97 99

Cu recovery in concentrate, %

Re

cove

ry o

f g

an

gu

e m

ine

rals

in t

ailin

gs, %

brak wzbogacania

65% <0,025 mm

80% <0,025 mm

90% <0,025 mm

0

20

40

60

80

100

85 90 95 100


Recovery

of

gangue m

inera

ls in t

aili

ngs,

%

brak wzbogacania

60

80

110

Potulska


With the increase of feed fineness, evaluated by content of particle class below 0.025 mm, copper

recovery progressively, increases. The analysis of the sulfide minerals liberation was made in order

to find reasons of such behaviour and the results are given in Table 4.

Table 4: Liberation of sulfide minerals in tested ore samples

Size class,

mm

65 % <0.025 mm 80 % <0.025 mm 90 % <0.025 mm

Liberated

minerals, %

Distribution,

%

Liberated

minerals, %

Distribution,

%

Liberated

minerals, %

Distribution,

%

>0.040 62.0 27.6 68.5 16.3 71.0 4.3

0.025-0.040 67.0 10.70 85.0 22.8 87.7 11.4

<0.025 84.0 61.7 89.5 60.9 93.9 84.3

100.0 100.0 100.0

Evident correlation is visible: the finest grinding, the highest liberation level. It explains the bad

floatability of the most coarses (65 % below 0.025 mm) feeds, caused by its insufficient liberation.

Figure 2 presents the relation between flotation results and the aeration of flotation slurry. After

recalculating the results using Fuerstenau`s curves for the same value of gangue recovery in tailings

εr = 80 % the following concentration results were obtained:

60 dm3/h ε =86.2 % β =7.8 % γ =23.1 %

80 dm3/h ε =89.8 % β =8.0 % γ =23.3 %

110 dm3/h ε =88.8 % β =7.9 % γ =23.4 %


Similarly, the influence of the rotational impeller speed on flotation results was investigated. After

recalculating the results, using Fuerstenau`s curves for the same gangue recovery in tailings

εr = 80 %, the following concentration indexes were obtained:

590 rpm ε =86.1 % β =7.7 % γ =23.3 %

670 rpm ε =89.8 % β =8.0 % γ =23.3 %

790 rpm ε =88.8 % β =7.9 % γ =23.2 %


Studies on the influence of aeration and rotational impeller speed showed that at given particle size

distributions of ore (90 % <0.025 mm) and in tested flow sheet [19], the most advantageous condi-

tions for flotation occurred at an air consumption amounting to 80 dm3/h (1.33 m

3/h/m

3 of slurry), a

collector consumption of 100 g/Mg and at a rotational impeller speed of 670 rpm.

Except of grain size, hydrodynamic conditions and aeration, the flotation results depend also on the

type and amount of the flotation reagents added, especially frothers and collectors, which determine



Eh and pH levels in floated slurry. Their control during flotation helps to affect on flotation results,

since the copper sulfide minerals float efficiently only at specific values of Eh and pH [21]. The

influence of the potassium xanthate consumption, within the range from 70 to 150 g/Mg, was tested

only during this study. Tested ore was ground to 90 % <0.025 mm. Frother consumption was the

same in all experiments of this series and amounted to 100 g/Mg. The results are presented on

Figure 3 in the form of Fuerstenau`s curves.

Figure 3: Results of flotation of feed ground to 90 % <0.025 mm for different consumption of

potassium ethylo-xanthate: 70, 100 and 150 g/Mg. (curves in Fuerstenau pattern). Flota-

tion conditions: aeration – 80 dm3/h, impeller rotation – 670 rpm

The results were recalculated basing on Fuerstenau`s curves for the same value of gangue recovery

in tailings εr = 80 % (see Figure 3). The following concentration results were obtained:

70 g/Mg ε =89.1 % β =8.0 % γ =23.3 %

100 g/Mg ε =89.8 % β =8.0 % γ =23.3 %

150 g/Mg ε =87.8 % β =7.9 % γ =23.2 %


The best results were obtained for xanthate consumption using up to 100 g/Mg, which is a bit higher

than the standard consumption. Further increase of xanthate consumption, up to 150 g/Mg, gave

slightly worse results, which can often be observed in the case of xanthates.

In order to determine the flotation performance of individual particle classes of ore, the products of

standard ore flotation having 90 % particles below 0.025 mm were screened on 0.005, 0.010, 0.015

and 0.025 mm by micro-sieves and Cu content as well as types of minerals presented in the obtained

0

20

40

60

80

100

85 90 95 100


Re

cove

ry o

f g

an

gu

e m

inera

ls in

ta

ilin

gs, %

brak wzbogacania

70 g/Mg

100 g/Mg

150 g/Mg

no concentration

Potulska


fractions were determined. This was the basis to define the liberation of sulfide minerals (Figure 4)

as well as their floatability (see Figure 5) for all particle classes.

In the finest classes below 0.010 and 0.010-0.015 mm, liberation of sulfide minerals is practically

the same for all flotation products. Only in the most coarse classes (0.025 and above 0.040 mm) one

can see the evident difference between liberation level for concentrate and for tailings. Into concen-

trate, independently from particle class size, report almost solely liberated sulfide mineral grains.

The highest Cu recoveries are achieved for 0.010-0.015 and 0.015-0.025 mm size classes. While

drop of recovery for sizes above 0.025 mm can be explained by insufficient liberation of sulfide

minerals, for sizes below 0.010 mm, the recovery drop must be caused by specific features of the

fine particles flotation, because as show Figure 4, their liberation is adequate.

Figure 4: Liberation levels of sulfide minerals Figure 5: Copper distribution in all size

in flotation products for feed 90 % of classes of flotation products

particles <0.025 mm. (relative values) (concentrate + middlings)

Among possible reasons for worse flotation results of the finest size classes, the energetic condi-

tions of fine particles surface, improper hydrodynamic conditions during flotation, insufficient con-

sumption of collector, too short flotation time, too wide range of particle size classes, etc. must be

considered.

Based on the achieved results, the chart presenting the floatability of individual sulfide minerals vs.

their particle size distributions is shown in (Figure 6).

70

80

90

100

0 10 20 30 40 50

Particle size, mm

Cu r

ecovery

, %



Figure 6: Mineral recoveries for the individual size classes of concentrate. Particle size distributi-

on of feed 90 % <0.025 mm, air consumption – 80 dm

3/h, impeller rotation – 670 rpm

and collector consumption – 70 g/Mg

Graphs of floatability and recoveries for the main sulfide minerals indicate, that the best floatability

for all size classes of ore being tested, have galena and chalcocite with digenite. Flotation recoveries

of those minerals practically do not depend (galena) or only slightly depend (chalcocite with dige-

nite) on class size. Visibly worse flotability have bornite and chalcopyrite, and definitely the worst –

coveline and pyrite with markazite.

>0.0250.015-0.0250.010-0.015<0.010

Particle size, mm

0

10

20

30

40

50

60

70

80

90

100

Bornite

Chalcocite with digenite

Chalcopirite

Pirite with markazite

Galena Coveline

Potulska


5 Conclusions

Research results enable to formulate the following conclusions:

1. Grinding fine mineralized ore from Polkowice mine to 90 % <0.025 mm warrants better condi-

tions of flotation than in the case of grinding to 60 and 80 % <0.025 mm. At liberation of sulfide

minerals amounting 95 % (90 % <0.025 mm), maximum recovery of copper bearing minerals in

froth product – equal to plant’s concentrate, amounting to 97-98 %, is observed for particles

within the range of 0.015-0.025 mm. Particles bigger than 0.040 mm are usually poorly libe-

rated.

2. Maximum copper recovery in flotation concentrate containing 25 % Cu, equal to plant’s concen-

trate, is observed for 0.010-0.025 mm size class. For <0.010 mm size class, under the experi-

mental conditions, a drop of recovery is visible, according to published data, concerning the

floatability of sulfide minerals.

3. The increase of rotational impeller speed in the flotation cell from 590 to 670 rpm, gives a cer-

tain improvement of floatability for all tested particle size distribution ranges. The increase of

slurry aeration also has an advantageous influence on flotation of all particle classes.

4. Concentration of the main sulfide minerals during flotation of Polkowice-Sieroszowice ore,

ground to 90 % < 0.025 mm, may be ranked in the following decreasing order: galena, chalco-

cite with digenite, chalcopyrite, bornite and coveline with pyrite and markasite. Floatability of

chalcocite and galena for all tested size classes, from above 0.025 to below 0.010 mm, changes

in a relatively low degree. However, coveline and pyrite with markasite show the worst floata-

bility of all tested ranges of particle size distributions, especially for the finest class – below

0.010 mm.

Acknowledgments

During this study, as one of the analytical methods, the optical analysis was used. Hereby, I would

like to thank Ph.D. Antoni Muszer for carrying out the time-consuming mineralogical investigations

and for his help during the completion of this study.

This study – as a research project – was financed by the Polish Ministry of Science and Higher Edu-

cation, from its resources for science investigations for 2005 – 2007. It was also supported by funds

of First Scholarship Program of Integrated Operational Program for Regional Development, where

75 % of founds were covered by the European Social Fund and 25 % by the Polish State budget.



References

[1] TAGGART A. F., (1956): Handbook of Mineral Dressing, Ores and Industrial Minerals. John

Wiley and Sons, New York, Section 12: 92, 97

[2] PEASE J. D., YOUNG M. F., CURRY D., JOHNSON N. W., (2004): Improving fines recov-

ery by grinding finer. MetPlant, Centenary of Flotation Symposium, Brisbane, Australia, 1–17

[3] PEASE J. D., CURRY D., YOUNG M. F., (2006): Designing flotation circuits for high fines

recovery. Minerals Engeneering, 19: 831–840

[4] HARBORT G., MURPHY A., VARGAS A., YOUNG M., (1999): The introduction of the

Isa-Mill for ultrafine grinding in the Mt Isa Lead/Zinc concentrator. For presentation at Ex-

temin 99, Arequipa, Peru, Sept., (Xstrata Technology), 1–8

[5] GAUDIN A. M., (1963): Flotacja. Tłum. z II wydania oryginału: J. Olszewski, T. Piaseczny.

Wyd. Śląsk, Katowice

[6] TRAHAR W. J., (1976): The selective flotation of galena from sphalerite with special reffer-

ence to the effect of particle size. International Journal of Mineral Processing, 151–166

[7] TRAHAR W. J., WARREN L. J., (1976): The flotability of very fine particles – a review.

Inter-national Journal of Mineral Processing, 3: 103–131

[8] LYNCH A. J., JOHNSON N. W., MANLAPIG E. V., THORNE C. G., (1981): Mineral and

coal flotation circuits. W: Fuerstenau, D.W. Ed., Dev. Miner. Process, Elsevier Science Pub-

lisher B. V., Vol. 3, Amsterdam

[9] POTULSKA A. (2006): Wpływ drobnego mielenia na flotację rud cynkowo-ołowiowych.

Prace Naukowe Instytutu Gornictwa Politechniki Wrocławskiej, nr 116. Seria: Konferencje ;

nr 47, Wrocław

[10] BORTEL R, (1967): Wpływ minerałów ilastych na flotowalność siarczków metali nieże-

laznych. Praca Doktorska, Politechnika Śląska, Gliwice

[11] WIENIEWSKI A., (1994): Problemy zwiększenia selektywności wzbogacania flotacyjnego

rud metali nieżelaznych. Konferencja Komitetu Górnictwa PAN „Aktualne zadania nauki w

górnictwie”, Ustroń

[12] GROTOWSKI A., (1997): Koncepcja rozwiązań przeróbki z zastosowaniem nowych tech-

nologii flotacji w aspekcie kompleksowego zagospodarowania górniczego wynikającego z

etapu I, sprawozdanie CBPM Cuprum, Wrocław

[13] ŁUSZCZKIEWICZ A., WIENIEWSKI A., (2006): Kierunki rozwoju technologii

wzbogacania rud w krajowym przemyśle miedziowym. Górnictwo i Geoinżynieria, Rok 30,

Zeszyt 3/1, AGH Uczelniane Wydawnictwa Naukowo-Dydaktyczne, ISSN: 1732-6702, 181–

196

Potulska


[14] GROTOWSKI A., PAKULSKA B., MIZERA W., (2004): Wstępne określenie możliwości

zastosowania młynów IsaMill do domielenia rudy miedzi w KGHM Polska Copper S.A.,

sprawozdanie CBR Cuprum sp. z o.o. sprawozdanie nr U-004/KU/04

[15] KIJEWSKI P., JAROSZ J., (2007): Odmiany litologiczne rudy. In: Monografia KGHM Pol-

ska Miedź S.A. część II, Praca zbiorowa pod redakcją Piestrzyński A., Wyd. KGHM Cuprum

CBR, Spółka z o. o., Wrocław, Lubin, 463–472

[16] SPALIŃSKA B., STEC R., SZTABA K., (2007): Miejsce i rola przeróbki rudy w kompleksie

technologicznym KGHM Polska Miedź S.A.. In: Monografia KGHM Polska Miedź S.A..

część II, Praca zbiorowa pod redakcją Piestrzyński A., Wyd. KGHM Cuprum CBR, Spółka z

o. o., Wrocław, Lubin, 463–472

[17] TUMIDAJSKI T. et al., (2007-2008): Report of Investigation, Project No. KGHM-ZW-U-

0011

[18] GROTOWSKI A., DĘBKOWSKI R., B., HENZEL SŁ., MIZERA A., MIZERA W.,

PAKULSKA B., SADECKI Z., SZAFRAN A., (2006): D2.2 Evaluation of the existing tech-

nologies used for mining and processing of black shale ore in KGHM Polish Copper. Propos-

als of new technologies for selective separation of shale material from Lubin and Talavivara

shale deposit, sprawozdanie KGHM Cuprum sp. z o.o.

[19] POTULSKA A. (2008): Wpływ drobnego mielenia na flotację krajowych rud miedzi. Roz-

prawa Doktorska, Wrocław

[20] DRZYMAŁA J., AHMED H. A. M., (2005): Mathematical equations for aproximation of

sepa-ration results using the Fuesrtenau upgraiding curves. International Journal of Mineral

Processing, Vol. 76, No. 1-2, April 4, 55–65

[21] RICHARDSON P. E, WALKER G. W., (1985): The flotation of chalcocite, bornite, chal-

copy-rite, and pyrite in an electrochemical – flotation cell. In: Proceed 5th Int. Min. Process.

Congr., Cannes, 198–210


HPGR versus SAG Milling Technology in

Hard-Rock Mining – Review and Analysis

Irshad Rana, Kris Chandrasekaran Ken Wood

Fluor Enterprises, Mining & Metals San Francisco Office Fluor Canada Limited

4140 Dublin Blvd., Suite 300 1075 West Georgia St., Suite 700

Dublin, California, USA Vancouver, Canada

Keywords: HPGR, SAG milling, high pressure grinding rolls, hardrock mining, cement, diamond,

kimberlite, copper

Abstract

High Pressure Grinding Roll (HPGR) technology has been developed based on the pioneering stu-

dies carried out by Professor Klaus Schoenert in Germany on single and packed bed particle fracture

under compression. This work demonstrated significantly lower specific energies for comminution

than those obtained from grinding ball mills. The machine consists of a pair of counter rotating rolls

with one fixed and the other floating. Gravity fed between the rolls, ore crushing takes place by the

mechanism of inter-particle breakage by compression exerted by the hydro-pneumatic spring on the

floating roll. The cement industry became the early adopter of HPGR technology in treating rela-

tively softer material (klinker). Other applications have since followed with kimberlite secondary

and tertiary crushing in the diamond industry, autogenous mill pebble crushing in iron ore

processing, limestone crushing, concentrate fine grinding and a number of others, such as phos-

phate, gypsum, titanium slag, gold ores and coal. Cyprus Sierrita operations in Arizona, USA, were

the first to attempt a large scale HPGR application in 1996 on harder and more abrasive copper ore.

Although the trial outcome did not result in the retrofit of HPGRs to the Sierrita fine crushing cir-

cuit, Phelps Dodge later purchased Cyprus Sierrita and, possibly capitalizing on this experiment,

became the first mining company to introduce HPGRs on a world class copper project at Cerro

Verde (Peru). Phelps Dodge was later purchased by Freeport McMoRan and the company now op-

erates four HPGR units (2.4 m x 1.7 m) instead of a traditional SAG mill circuit at Cerro Verde and

also operates two HPGRs (2.0 m x 1.8 m) at their Grasberg Property in Indonesia. Boddington and

Bendigo gold projects in Australia have also included HPGR units in their plants. Although the

technology is offering attractive possibilities when compared with SAG milling, site specific factors

must be carefully considered in decision making, such as ore wetness, stickiness, abrasion index and

work index etc. Based on various characteristics including plant layout, the capital costs for HPGR

installations are 10 to 25 % higher than the SAG milling process. This higher CAPEX is tempered

Rana, Chandrasekaran, Wood


by 15 to 25 % reduction in energy cost with potential overall operating cost savings in the range of

10-20 %. As inter-particle crushing promotes micro-cracking, an additional benefit of improved

recoveries from downstream processing may also be realized. This paper will also present an over-

view of the HPGR application in heap leaching technology where micro-cracking could have a posi-

tive influence on metal recoveries.

1 Introduction

Since the application of HPGRs at Cerro Verde (Peru) in 2006 instead of a traditional SAG milling

circuit, the technology has been receiving close scrutiny for projects in the hard-rock mining indus-

try (copper & gold ores). Although Professor Klaus Schoenert at Karlsruhe & Clausthal Universities

in Germany (early 1980’s) demonstrated significantly lower specific energies for crushing in a

packed particle bed under compression than those obtained in the tumbling ball mills, the technolo-

gy was not adopted by the mining industry until in the late 1990’s.

The cement industry was the first to accept this new technology in the late 1980’s for the softer

clinker crushing and several hundred machines are now operating for this sector worldwide. Most of

the applications are in pre-treating feed for the ball mills and this has significantly improved the

downstream grinding capacity and the total energy consumption has been reduced as a result.

In the traditional mineral processing industry, diamond mining in South Africa began utilizing

HPGRs to promote controlled micro-cracking in ores to liberate diamonds without damaging the

product quality. This application has now been adopted in Australia and Canada as well and the

technology has become a standard in most new diamond plants. In iron ores, HPGRs are used large-

ly for pre-treatment of pellet feed to increase plant capacities and for pebble crushing. This has been

now widely accepted in Brazil, India, Sweden, Russia and the USA.

Freeport McMoRan’s Sierrita operations were the first in copper mining to perform plant tests of

the HPGR technology in 1996 in an effort to replace the existing tertiary crushers. Sierrita flowsheet

is three stage crushing followed by ball milling. The feed material was abrasive quartz diorite with a

nominal particle size of less than 50 mm. However, due to the excessive wear problems on the then

available roll surface protection materials, the pilot test was abandoned. The Sierrita pilot test work

and other efforts in the hard-rock mining led Freeport to review the HPGR technology for the Feasi-

bility Study of Cerro Verde Project in Peru instead of the traditional SAG milling flowsheet. The

study by Fluor Canada in Vancouver showed improved economics with HPGR’s and the commis-

sioning of Cerro Verde HPGRs in 2006 marked the culmination of the efforts by many in the indus-

try over the past number of years to have the technology accepted as a legitimate alternative to the

conventional grinding approach for hard-rock mineral processing. Other projects that have recently

adopted HPGR technology include Freeport (Indonesia), Boddington (Australia) and Amplats Pot-

gietersus (South Africa).

HPGR versus SAG Milling Technology in Hard-Rock Mining – Review and Analysis


Freeport Sierrita and Bagdad Operations in Arizona, USA, studied the use of HPGRs again in 2008.

The application was evaluated at the feasibility level (Fluor San Francisco Office located in Dublin,

California, USA) for quaternary crushing mode to increase the milling capacity at Sierrita. At Bag-

dad the expansion was studied to utilize HPGR instead of a conventional SAG milling flowsheet

with two stage crushing followed by HPGR’s and a ball mill. After the completion of feasibility

studies, both projects were postponed due to the world economic crisis at that time.

2 HPGR Technology

The machine consists of two counter rotating rolls mounted in a robust frame with each roll sup-

ported by large bearing blocks. One of the bearing blocks is fixed while the other is allowed to float

on rails. Pressure is applied on the floating roll against the fixed by an oil hydraulic system. A stop

block prevents the rolls from contacting each other. Figure 1 provides a typical schematic cross sec-

tion of a HPGR.

Figure 1: Schematic Representation of HPGR (permission from Anguelov et al. [3])

The HPGR machinery is described in detail in numerous papers and vendor brochures along with

the definitions of terminology and the ways to carry out testing/sizing procedures. This article on the

other hand will present the main distinguishing features of the technology:



• Ore is choke fed, by gravity, between the rolls and the crushing takes place by a mechanism of

autogenous, compressive breakage in the particle bed. This is in contrast to the tumbling SAG or

ball mills where particles are broken in a random hit and miss manner.

• Crushing is accomplished largely by the pressure exerted on the floating roll by the system hy-

draulics and this pressure could vary (50-250 MPa) depending upon the ore characteristics, roll

diameter and nature of the roll surface protection. The compressive forces involved render the

ore density between the rolls in the range of 70-85 % solid by volume. The hydraulic system al-

lows the rolls to open to pass hard metal tramp material, protecting the rolls.

• Both rolls are identical and interchangeable and with variable speed drive the roll speeds are ad-

justable to obtain optimum grinding conditions. Pressure on the floating roll and the roll speed

are the most important operating parameters and can be controlled by an operator.

• Depending upon the ore type the HPGR product could be compressed into a cake, flake or bri-

quette. The downstream processing may require de-agglomeration in a scrubber etc. Material

handling systems must account for the sheet/flake-like material characteristics.

• HPGR product has a greater percentage of fines than can be found in a conventional crusher and

the coarser particles contain extensive micro-cracking which reduces the grinding work needed

in downstream processing to liberate minerals/metals from the ore [1].

The author’s experience analysing testwork indicates that for the samples tested, HPGR product

contains a similar amount of fines to SAG product but the HPGR final product has less extreme

fines (slimes) and less material in the coarse fractions. It is felt that this has advantages for sub-

sequent flotation.

• Micro-cracking is shown to be beneficial for heap leaching processes where the solution must

penetrate deep into the particles during leaching processes to dissolve metal and enhance kinetics

to generate the pregnant leach solution (PLS).

• Diameters of rolls typically range from 0.5 m to 2.8 m and the roll widths vary from 0.2 m to

1.8 m. For comparison, Cerro Verde HPGRs are: 2.4 m diameter and 1.65 m wide processing

2500 tonnes per hour through each unit with a power rating of 5000 kW.

• As the ore dwell time between the rolls is short, the machine settings can be easily altered by the

control system, permitting the ease of quick process control and compensation for the fluctuating

feed ore properties.



• HPGR technology is characterized by:

• Minerals are liberated generally along the grain boundaries,

• High and steady throughput rates if the ore is dry and non-sticky,

• Reduced noise levels,

• Low and manageable dust loading,

• The penetration of the HPGR technology in the hard-rock mining application was limited by

the wear resistance of the roll surfaces. However, recent advances in wear-resistant materials

and surfaces have improved roll service lives in the range of 2000 to 6000 hours for the abra-

sive hard rocks. The roll service lives can exceed 15,000 hours when treating softer materials.

• The following three manufacturers are currently supplying HPGR machines:

• Polysius (a Thyssen Krupp Company),

• Köppern,

• KHD Humbolt Wedag AG.

Except for minor differences, the three suppliers use largely similar technology.

Polysius employs a roll design with high aspect ratio (bigger diameter with smaller width) whereas

KHD and Köppern utilize a lower aspect ratio. Larger diameter rolls are expensive but offer a wider

operating gap and the roll surfaces are exposed to a smaller portion of the material processed. This

machine characteristic helps improve the roll wear life.

Tungsten carbide studs are being used to generate an autogenous wear layer on the roll surface un-

der a patent by KHD. Polysius and Köppern are also offering HPGRs with this innovation under a

license from KHD. Different roll surfaces are available, such as: pattern welded, studded, Hexadur

lined, and chevron welded. Equipment manufacturers provide detailed brochures with the available

roll surfaces based on the applicable ore hardness.

• The HPGR has a lower unit capacity than SAG mills and as such is more readily scalable for

increasing plant capacity, although conveying systems may prove to be limitations to expansion.

3 Screening Parameters for HPGR vs. SAG Milling

3.1 Impact crushability work Index

The Bond Crushing Work Index of 13 and higher indicates that the ore to be crushed is a hard ore.

An impact work index of 20 and higher indicates that the ore is a very hard ore. The hard competent

rock is a candidate for the application of HPGR.



The compressive strength of the material to be crushed determines the amount of energy that can be

absorbed by the material. This in turn determines the HPGR power requirement.

3.2 Abrasion Index

The abrasion index of the material to be crushed will determine the metal wear of the HPGR liner

material. Unless the feed ore has a very high abrasion index, HPGR is suitable for size reduction

such that single stage milling in a ball mill subsequent to the HPGR can be adopted for the grinding

circuit. With a high abrasion index, SAG mills will also experience high media consumption and

liner wear.

3.3 Ore Characteristics

Ore containing significant amounts of clay and talc minerals along with excessive moisture can

cause problems in maintaining a stable autogenous layer between the studs. The recommended feed

moisture is up to 4 %. Feed moisture significantly in excess of this will cause operational material

handling problems with the HPGR and associated bins, feeders, upstream crushers, conveyor trans-

fers, etc. Higher feed moisture also increases metal wear. These are conditions for which Autogen-

ous and SAG mills were originally applied.

3.4 Plant Layout

The layout of the HPGR facilities influences the economics due to the physical site conditions. The

use of HPGR in lieu of SAG mill needs the following additional facilities.

• The primary crushed ore secondary crushing and screening facility. The secondary crusher oper-

ates in closed circuit with the screens.

• HPGR feed bin and HPGRs.

• Fine screens operating in closed circuit with HPGRs.

• Conveyors for material transfer between facilities.

• Dust control systems.

All the above facilities will occupy a relatively large area. In hilly terrain, especially where different

process facilities are required to be located at different elevations, the plant area comes at a pre-

mium. In addition, the terrain could involve a large quantity of earthwork for site preparation and

this quantity could be overwhelming. Even though the HPGR technology may be suitable to adopt,

the project economics may not support HPGR application.



4 Down-stream Processing Considerations

HPGR technology offers certain unique down-stream considerations in hard-rock mining applica-

tions due to the micro-cracking induced in the crushed particles, as described below.

4.1 Reduced Work Index

Laboratory tests have been carried out on various ores using the standard Bond grinding index pro-

cedures and have indicated reductions in Bond ball mill work index (Wi) on the HPGR product

compared with the original feed. This would, of course, vary from ore to ore and is dependent on the

grind size tested, but there have been reported reductions in the range of 0-25 %. Some recent cop-

per project tests have shown Wi reductions of 6-21 % [1]. HPGR suppliers have created grindability

test methodologies that start with a coarser feed than the Bond ball mill work index test and as a

result take into account the effect of micro-cracking on the coarser size fractions as well as the in-

creased production of fines from the HPGR compared to conventional crushers. These tests show

reductions in ball mill energy requirements more in the range of 10-40 % for the same samples as

the Wi values noted above, compared to conventionally crushed feed [1]. In comparison to SAG

mill product, only the Wi reduction would be expected to be evident since both SAG and HPGR

products have similar fine content. Such reduction in ball mill grinding energy would result in either

smaller ball mill grinding equipment, higher downstream capacity for grinding ore or a better grind

for a similar capacity. The mill operators therefore could opt for either higher capacity or better

metal recovery from the grinding fineness.

4.2 Flotation Recovery

Flotation losses tend to be greatest in the extreme fines and the coarse fractions. SAG product has

more material in both these fractions compared with HPGR product and can be assumed to be more

susceptible to recovery losses. In addition, HPGR product tends to break along grain boundaries,

leading to better liberation at coarser sizes. This may allow a coarser grind for the same recovery or

higher recovery at a given grind compared to SAG product.

4.3 Leach Recovery

Comparison of gold leach recoveries on a size by size basis on both South African and a refractory

Nevada ore showed improved recovery for ore that had been treated in an HPGR circuit compared

to conventional. Gold leach recoveries of 5-20 % higher were experienced with the HPGR

product [2].



4.4 Ball Mill Optimization

The relatively tightly sized HPGR product after screening allows optimization of ball mill media

size for further improvement in ball mill efficiency and capacity in a close circuit configuration. The

reduced coarse material content in the slurry can also reduce wear on ancillary equipment such as

pumps, piping and cyclones compared to SAG circuits.

4.5 HPGR in Heap Leaching

Heap leaching of copper and gold ores is a major consideration for HPGR technology. The main

advantages for HPGR units versus the conventional crushing circuits are:

• Micro-cracking of particles would permit better penetration of lixiviant,

• Better mineral liberation and exposure to leaching,

• Increased oxygen penetration.

Due to the factors above, the following observations have been reported in heap leaching regimes:

• HPGR crushed ores exhibit better solution percolation compared with the conventional

crushing [1].

• Approximately 60 % of the coarse ore particles (+100 mesh) are micro-fractured and hence the

diffusion paths to metal minerals are significantly better. This results in better metal recoveries.

• The overall reaction kinetics are significantly improved for the ores with slow penetration rates.

• Improved lixiviant accessibility through micro-cracks leads to reduction in leach reagent con-

sumption rate.

5 HPGR vs. SAG Milling Cost Considerations

In light of various trade off studies carried out recently, HPGR capital costs are 8-12 % higher than

the comparable SAG mill costs. This project capital gap appears to be narrowing as the HPGR tech-

nology is improving and the applications are better understood and analyzed. The increased capital

is mainly due to the requirement for a secondary crushing circuit and the associated conveyors and

screens to handle the processing. The material handling required with the HPGR mandates a larger

layout footprint and the associated civil earthworks.

HPGR energy utilization is more efficient as crushing takes place by a mechanism of autogenous

compressive breakage in a particle bed in contrast to the tumbling SAG mill where particles are

broken in a random hit and miss manner. Thus the comparatively higher HPGR capital cost is tem-

pered by 15-25 % reduction in energy cost with potential overall operating cost savings in the range



of 10-20 %. Next to the energy cost, absence of grinding media cost (ball consumption) is a major

factor in reducing the HPGR operating cost.

Anguelov et al. [3] presented the results of trade off studies, in a recent paper, between the HPGR

and the SAG milling technologies for various hard-rock mining projects. Without naming the

projects involved, given below are the results for capital and operating costs from the referenced

publication:

Table 1: Overall Mill Operating Cost

Table 2: Overall Capital Cost

Fluor San Francisco Office has carried out studies recently between HPGR and SAG milling circuits

and produced similar results (HPGR capital + 10-12 %).

Projects SAG Mill HPGR

Project A 5.30 $/t 4.56 $/t

Project B 2.24 $/t 1.63 $/t

Project C 4.98 $/t 3.62 $/t

Project D 2.66 $/t 2.03 $/t

Projects HPGR vs. SAG

Project A + 6.4 %

Project B - 14.3 %

Project C + 8.2 %

Project D + 9.6 %



6 Conclusions

Based on the review and analysis of HPGR versus SAG Milling above, the following conclusions

are presented:

• HPGR technology should be evaluated in light of the site specific factors for greenfield and

brownfield projects. This includes the available layout space, topography of the area, energy cost,

labor cost and the remoteness of the project site. In a brownfield project, the evaluation must

consider the “fitting in” of the HPGR technology within the current mix of the process flowsheet

equipment.

• Wet, sticky feed and ores containing excessive clays should be evaluated critically for the materi-

al flow bottlenecks in HPGR type flowsheets. These are probably better suited to SAG milling.

• Extensive laboratory and pilot plant testwork should be carried out before deciding between the

HPGR and SAG milling circuit. This would shed light on the energy consumption, ore hardness

and ore abrasiveness indices.

• HPGR capital costs are expected to be 8-12 % higher than the SAG mill circuit, under the current

state of technology.

• Overall operating costs are expected to be 10-20 % lower than the SAG mill circuit due to the

lower energy cost with the HPGR circuit and the absence of grinding media cost (roll liners con-

sidered).

• Energy costs are expected to be 15-25 % lower for the HPGR circuit compared with the SAG

mill.

Acknowledgments

The authors wish to thank Fluor Enterprises and the management of Fluor San Francisco Office

(Mining & Metals) for granting permission to publish and allowing resources to develop the paper.

References

[1] Unpublished test data on copper/gold ores from North and South America

[2] PATZELT, N., KNECHT, H., BAUM, W. (1995): Case made for High Pressure Roll Grinding

in Gold Plants – Mining Engineering; June 1995, 524-529

[3] ANGUELOV, R., GHAFFARI, H., ALEXANDER, J.: High Pressure Grinding Rolls (HPGR),

an Alternative Technology versus SAG Milling, Conference: Communition of Ores, Common

Wealth of Cornwall, UK. 2008 & China Mining Conference, 2008


Selective Leaching of Arsenic from Copper Ores

and Concentrates Containing Enargite in

NaHS Media

William Tongamp, Yasushi Takasaki, Atsushi Shibayama

Akita University

Faculty of Engineering and Resource Science

1–1 Tegata–Gakuen cho

Akita City 010–8502, Japan

Keywords: Selective leaching, arsenic, copper resources, enargite, sodium hydrosulfide

Abstract

Conventional smelting processes cannot treat Cu–ores/concentrates containing significant amount

of toxic arsenic, and development of pretreatment methods to remove arsenic (As) is needed. This

work reports selective leaching of As from enargite (Cu3AsS4) contained in Cu–ores/concentrates

using NaHS media. Samples were obtained from various sources with As content ranging from 0.9

to 14 wt. % and Cu content from 20 to 40 wt. %. Leaching experiments were conducted at 30-95 °C

under atmospheric pressure with pulp densities from 100 to 1000 g/L. Concentration of NaHS was

varied from 50 to 200 g/L, and 50 to 100 g/L of NaOH was added. The kinetic analysis of the data

found that the reaction is controlled by a product layer diffusion process with an activation energy of

70.3 ± 4.7 kJ/mol. Chemical analysis and X-ray diffraction analysis of the leached samples showed

that As content in the samples were reduced to less than 0.5 wt. % within 3 to 6 h even at high pulp

density of 1000 g/L, and that Cu3AsS4 was transformed to Cu2S.

1 Introduction

Enargite (Cu3AsS4) and tennantite (Cu12As4S13) are the main minerals with high As content that

associate with copper sulfides such as covellite (CuS), chalcocite (Cu2S), chalcopyrite (CuFeS2),

Bornite (Cu5FeS4), and are normally recovered by flotation and offered as Cu-resource to smelters

for Cu recovery. The high content of As effectively reduces value of the Cu-resource due to envi-

ronmental implications of toxic As that require additional processing steps for containment of As.

As demand for Cu continues to grow, the need to source Cu from all resources including those that

contain high As is necessary and as such more economical and environmentally friendly processing

Tongamp, Takasaki, Shibayama


options will be sought. Reviews on Cu–resources containing enargite are covered widely in litera-

ture by many researchers [1-8].

A method to remove As from the Cu–ores/concentrates containing enargite is flotation [2, 5, 9-13];

by employing slurry potential and pH control with depressants and activators, a clean concentrate

with As content < 0.5 wt. % are recovered. Conversion of enargite to CuO or Cu2S by selective dis-

solution of As into solution with sodium hypochlorite (NaClO) or sodium sulfide (Na2S) in alkaline

solutions with NaOH is also effective to remove As from the Cu-ores/concentrates [1-5, 6, 14-17].

Other hydrometallurgical treatment options through chloride and sulfates leaching, or in ammonical

solutions and at elevated temperatures and pressures is reported in many literature listed above and

elsewhere [18, 19].

In a recent study, the same authors have introduced sodium hydrosulfide (NaHS) as a leaching me-

dia for the conversion of enargite (Cu3AsS4) to Cu2S in NaHS–NaOH solution [20], and have

shown that near complete removal of As (> 99 %) in enargite can be achieved.

The behavior of enargite conversion in NaHS media is very similar to that in sodium sulfide (Na2S)

media [2, 4, 6, 14, 17], however, selection of NaHS was based on its application in copper flotation

and also its higher sulfur content (~43 %) per unit weight NaHS as compared to Na2S (~13 %) since

leaching of As from enargite is a function of sulfur [S2-

] concentration.

In the present study, the effects of parameters such as particle size, slurry stirring rate, NaHS con-

centration and temperature were investigated to study the reaction kinetics of enargite in NaHS–

NaOH media.

2 Experimental

2.1 Sample preparation

Enargite ores and Cu-concentrates samples used in this work have been obtained from various

sources with differing amounts of As ranging from 0.8 to 14 wt. % and Cu from 10 to 40 wt. %. Ore

samples were first prepared by crushing and grinding followed by sizing of both ores and concen-

trates to obtain a size fraction of -75+35 µm for all leaching experiments. Samples with As content

< 1.0 wt. % were also blended with enargite ore with As content >8.0 wt. % to obtain a composite

with As content at 3.0 wt. %. Sample information and chemical analysis of the original samples is

given in Table 1.

Selective Leaching of Arsenic from Copper Ores and Concentrates


Table 1: Main chemical and phase compositions in Cu–ores/concentrate samples with As impu-

rity from various sources.

Sample identity As (%) Cu (%) Fe (%) main composition

(1) Cu conc. 0.81 26.62 25.40 CuFeS2, FeS2

(2) Cu conc. 0.95 22.60 18.17 CuFeS2, FeS2

(3) Cu conc. 1.67 11.49 9.71 CuFeS2, FeS

(4) Cu conc. 3.11 33.46 14.34 FeS2, Cu2FeSiS4

(5) Enargite ore 8.32 33.71 16.34 Cu3AsS4, CuFeS2

(6) mixture (1) + (5) 3.00 25.84 21.04 -

(7) mixture (2) + (5) 3.00 25.69 21.08 -

(8) Enargite ore 3.65 12.32 23.37 Cu3AsS4, FeS2

2.2 Leaching

All leaching experiments were conducted in a 200 ml Teflon beaker immersed in a water bath and

total solution volume was kept constant at 100 ml, fully enclosed with a thermometer fixed for

monitoring temperature and cooling system to prevent solution evaporation. NaHS as leaching me-

dia and NaOH to maintain high pH and prevent sulfur hydrolysis were obtained as pure chemical

reagents from Wako Chemicals – Japan and added at (50–200) g/L.

After adjusting temperature to set point, samples at varying pulp densities ranging between 100 and

1000 g/L were weighed and introduced into the solution containing NaHS. Slurry temperature dur-

ing leaching was varied between 30 to 95 °C to evaluate the effect of temperature; most tests were

conducted at 80 °C in this work and leaching time varied between 5 min to 10 h. At the end of each

experimental run, the solids were separated from solution by filtration to obtain filtrate and residue

for analyses.



Figure 1: Schematic illustration of experimental flowsheet for enargite leaching and main condi-

tions.

2.3 Characterization and analyses

X-ray diffraction (XRD) powder analysis was carried out using Rigaku, RINT–2200 / PC system with

a CuKα irradiation source (λ = 1.5405 Å) for starting samples and residues after leaching to observe

changes in phase formations. Solution potential and pH both at the beginning and end of each leaching

experiment was measured using laboratory Eh/pH meters to observe changes and relate to relationship

of Eh and pH conditions in enargite transformation. Analyses of filtrates were carried out using an

Inductively Coupled Plasma (ICP–AES/OES) atomic/optical emission spectrometer, SPS–3000 (Seiko

Instruments Inc.) to monitor dissolution of As, Cu and Fe during leaching.

3 Experimental results

3.1 Effect of temperature

The influence of temperature on As dissolution was investigated by conducting experiments from

30 to 95 °C using the -75+32 µm fraction of enargite ore (8) while keeping reagent concentrations

fixed at 100 g/L (1.35 M) NaHS + 50 g/L (1.25 M) NaOH and the results obtained are shown in

Figure 2.

(crush/grind/sizing/blending)

NaSH + NaOH +

H2O

Sample preparation

Na3AsS3

solution

Selective leaching

Solid/Liquid separation

Arsenic treatment Clean Cu conc

Experimental conditions

Pulp density: 100 – 1000 g/L

NaHS addition: 50 – 200 g/L

Leaching time : 10 min – 12 h

Temperature: 30 – 95 °C

Stirrer speed: 150 – 750 rpm

NaOH addition: 50 – 100 g/L

Enargite



Figure 2: Arsenic dissolution behavior as a function of leaching time for a 1.0 g enargite ore (8)

Leaching of -75+32 µm at different temperatures in a 100 g/L (1.35 M) NaHS and

50 g/L (1.25 M) NaOH solution. Stirring rate was fixed at 550 rpm.

These data shows that temperature exerts significant effect on As leaching rate. At 30 °C, As disso-

lution reached only 13 % in 30 min and shows no improvement after that. Increasing temperature to

40 °C, As dissolution reaches 20 % in the first 30 min and reaches 36 % in 120 min. Further in-

crease in temperature, As dissolution significantly increases reaching over 99 % in 120 min for sam-

ple leached at 80°C and over 99 % in 90 min for sample leached at 95 °C.

In a second case, another sample; Cu-concentrate (3) containing 1.67 wt. % As was leached for 3 h

at temperatures ranging between 30 and 95 °C but at a much higher pulp density of 500 g/L is

shown in Figure 3. NaHS and NaOH additions were fixed at 100 g/L each and these corresponds to

S2-

/As ratio of 16.20 (mol/mol ratio). At 30 °C As dissolution reached only 10 %, but with increase

in temperature, As dissolution reached over 47 % at 60 °C, 87 % at 80 °C, and over 95 % at 95 °C.

To reduce As content from 1.67 wt. % to <0.50 wt. %, up to 70 % dissolution of As in the sample is

required and this result indicates that within 3 h and at 80 °C As content in original sample was re-

duced to about 0.22 wt. %. Based on this result, all proceeding experiments were conducted at

80 °C and leaching times varied from 1 to 6 h.

0

20

40

60

80

100

0 20 40 60 80 100 120

30 °C

Leaching time, t / min

Ars

enic

rem

ov

al (

%)

50 °C

60 °C

70 °C

80 °C

95 °C



Figure 3: Dissolution behavior of As from Cu-concentrate (3) at different temperatures. Leaching

time was fixed at 3 h, pulp density at 500 g/L with NaHS and NaOH additions

at 100 g/L.

3.2 Effect of NaHS and NaOH concentration

For subsequent leaching experiments to investigate effect of NaHS and NaOH concentrations on As

leaching, temperature was kept constant at 80 °C and pulp density fixed at 500 g/L. Cu-concentrate

(3) with As content at 1.67 wt. % was leached for 1 to 4 h and varying NaHS/NaOH concentrations

and results obtained is shown in Figure 4. At 100 g/L NaHS and NaOH additions respectively

(S2-

/As = 16.20), As dissolution rapidly increases to 70 % in the first 1 h and gradually increases to

above 80 % within 3 h and almost complete dissolution in 4 h. This shows As content in this sample

could be reduced to below 0.5 wt. % within 1 to 4 h. However, when NaHS addition is reduced to

50 g/L (S2-

/As = 8.10), overall dissolution does not exceed 40 %. When addition of NaOH is re-

duced to 50 g/L and maintaining NaHS at 100 g/L, dissolution rate of As is slightly decreased and

only reaches 70 % dissolution after 3 h for the same sample. This result indicates that at 500 g/L

pulp density and leaching temperature of 80 °C, concentration of both NaHS and NaOH of at least

100 g/L (S2-

/As = 16.20) is required to reduce As content in the sample to <0.50 wt. %

within 2 to 4 h.

0

20

40

60

80

100

95.4

0 %

87.9

1 %

47

.17 %

10

.05 %

95 oC80

oC60

oC30

oC

Ars

enic

dis

solu

tio

n (

%)

Leaching temperature, T / oC



Figure 4: Effect of NaHS and NaOH concentrations on As dissolution as a function of leaching

time at 80 °C and pulp density of 500 g/L. (a) Both NaHS/NaOH addition at 100 g/L;

(b) NaOH addition at 50 g/L; (c) NaHS addition at 50 g/L.

3.3 Effect of pulp density on leaching

Varying pulp densities at 100 to 600 g/L was investigated to determine optimum pulp density at

which As content in the samples could be reduced to <0.50 wt. % within 6h and at low NaHS/NaOH

concentration and leaching temperature fixed at 80 °C. As discussed in previous sections above,

results at 500 and 600 g/L show that such an objective could be established as shown in Figure 5. At

600 g/L (S2-

/As = 13.50), leaching rate is slightly lower than 500 g/L and As dissolution reaches

70 % after 2 h, however it is also clear that As content can be reduced to <0.50 wt. % within 3 to

6 h. For a pulp density of 100 g/L, complete removal of As was achieved within 3 h under similar

conditions.

Varying addition of NaHS and NaOH to observe As dissolution of sample (7) with a pulp density of

1000 g/L was performed and the results are shown in Figure 6. At 200 g/L NaHS addition

(S/As = 6.7) over 80 % of As is dissolved in 1 h and reaches over 95 % in 6 h and total dissolution

achieved in 10 h. Decreasing NaHS addition to 150 g/L (S/As = 5.1) and 100 g/L (S/As = 3.4), the

As dissolution is reduced accordingly and reaches 70 to 80 % in 6 h, after which it remains consis-

tent at prolonged leaching up to 10 h. This result suggests that a S/As molar ratio of at least 6 is re-

quired from a stoichiometric requirement of S/As = 1.5 to reduce As content from 3.0 wt. % to be-

low 0.5 wt. % for the Cu–concentrate samples mixed with enargite ore within 6 h.

0 1 2 3 40

20

40

60

80

100

(c): S/As = 6.1

(b): S/As = 12.3

(a): S/As = 12.3A

rsen

ic d

isso

luti

on (

%)

Leaching time, t / hours



Figure 5: Dissolution behavior of As from Cu-concentrate (3) at different pulp densities as a func-

tion of leaching times at 80 °C. NaHS/NaOH concentrations fixed at 100 g/L.

Figure 6: Arsenic leaching behavior of blended sample (7) with arsenic content at 3.0 wt. % at a

very high pulp density of 1000 g/L. Effect of NaHS and NaOH addition on arsenic dis-

solution as a function of leaching time at 80 °C was investigated and aggressive stirring

at 650 to 750 rpm was performed to keep sample evenly dispersed in slurry.

X-ray diffraction analysis of starting enargite ore (5) with As content of 8.32 wt. % and compared to

leach residues obtained after leaching for different times; 1, 3 and 6 h is shown in Figure 7. The

result shows near complete disappearance of characteristic peaks of Cu3AsS4 from leach residues

obtained after 1 h. On the other hand, characteristic peaks of Cu2S clearly appear from residues ob-

tained after 1 h leaching and their intensity increases significantly as leaching progresses.

0

20

40

60

80

100

0 1 2 3 4 5 6

600 g/L (S/As = 10.2) 500 g/L (S/As = 12.3)

Ars

enic

dis

solu

tio

n (

%)


100 g/L (S/As = 61)

0

20

40

60

80

100

0 1 2 3 4 5 6 7 8 9 10

Ars

enic

dis

solu

tio

n (

%)

200 g/L NaHS (S/As = 6.7)


100 g/L NaHS (S/As = 3.3) 150 g/L NaHS (S/As = 5.1)



Figure 7: X-ray diffraction pattern of original enargite ore (5) and the leach residues obtained after

leaching for different times at 80 °C. Pulp density, NaHS and NaOH additions were kept

constant at 500 g/L and 100 g/L respectively.

3.4 The leaching mechanism

The transformation or leaching mechanism of enargite (Cu3AsS4) to Cu2S in NaHS and NaOH me-

dia can be represented by Reaction (1) below. The dissociation of NaHS added as a leaching me-

dium can be represented by Reaction (2), which further dissociates to provide S2-

as shown by Reac-

tion (3) in the presence of OH- (NaOH). Therefore, maintaining high pH in high alkaline region

(> 12.5) prevents the hydrolysis of the S2-

ions to HS-, and maintains a high concentration of S

2- ions

to dissolve As from Cu3AsS4.

2Cu3AsS4 + 3NaHS + 3NaOH = 2Na3AsS4 + 3Cu2S + 3H2O (1)

NaHS → Na+ + HS

- (2)

HS- + OH

- ⇌

S

2- + H2O (3)

The stoichiometric requirements according to Reaction (1), shows that one mole of As would re-

quire 1.5 moles of sulfur (S/As = 1.5) for dissolution of As from enargite. The transformation of

Cu3AsS4 to Cu2S as represented by Reaction (1) in NaHS media is similar that for Na2S. NaHS was

chosen as reagent for this work due to its application in Cu flotation and higher sulfur content (2 to

3 times higher).

10 20 30 40 50 60 70

2θ /degree

0h

1h

3h

∆∆

Cu3AsS

4

Cu2S

FeS2

SiO26h

∆



Target As dissolutions for each of the samples used in this work to reduce As content to

< 0.50 wt. % as computed and actual As dissolution results obtained within 4 to 6 h of leaching are

shown in Table 2. Samples (1) and (2) due to low As content would require between 40 to 50 %

dissolution for As to be reduced to < 0.5 wt. %, however, both samples were not treated separately

but blended with sample (5) to prepare samples (6) and (7). The results show that over 98 % of As

in sample (3) was dissolved within 4 h and As dissolution for samples (4) and (5) reaches 90 to

98 % from 4 to 6 h. Analysis of leach residues by ICP obtained after leaching for As content corre-

spond well with filtrate or leach solution analysis discussed in sections above with As content in all

leach residues remain below 0.5 wt. %.

Table 2: Distribution of As and Cu in starting sample and leached residues. Leaching conditions;

pulp density at 500 g/L, NaHS and NaOH additions at 100 g/L respectively, temperature

and time at 80 °C and leaching conducted for 4 h.

Sample identity As content (wt. %) Cu content (wt. %)

original sample leach residue original sample leach residue

(1) Cu conc. 0.81 - 26.62 -

(2) Cu conc. 0.95 - 22.60 -

(3) Cu conc. 1.67 < 0.35 11.49 (13 ~ 15)

(4) Cu conc. 3.11 < 0.40 33.46 (38 ~ 42)

(5) Enargite ore 8.32 < 0.50 33.71 (38 ~ 44)

(6) mixture (1) + (5) 3.00 < 0.30 25.84 (28 ~ 33)

(7) mixture (2) + (5) 3.00 < 0.30 25.69 (28 ~ 33)

(-) no seperate leaching experiments conducted

In all leaching experiments, Cu and Fe remain undissolved and Cu content slightly increases due

leaching of As into solution.

3.5 Enargite leaching kinetics

The rate of reaction between solid particles and leaching reagent could be determined by a non-

catalytic heterogeneous model such as diffusion through the fluid film, diffusion through product

layer, or one involving surface chemical reaction. Experimental results obtained at different tem-

peratures in this work have been evaluated using several kinetic models; 1-(1-α)1/3

and 1-(1-α)2/3

(interface/surface–chemical reaction control models) and diffusion controlled mechanisms;

1-3 (1-α)2/3

+ 2 (1-α) and (1-2/3α)-(1-α)2/3

.



Figure 8: (a) Linearization of the experimental data for As leaching at different temperatures and

(b) Arrhenius plot using linear kinetic constant (k) values for determination of activation

energy.

The diffusion controlled models were observed to produce good linearity and data from current

work was analyzed using the shrinking core model or rate expression represented by Equation (4)

below:

1–3(1–α)2/3

+2(1–α) = kt (4)

In the expression, k is the linear kinetic constant; t is the reaction time and α, represents fraction

arsenic removed from enargite. A linear relationship from the plot of reaction time (t) versus

1-3 (1-α)2/3

+2 (1–α) could confirm this model and Figure 8(a) clearly supports the kinetic model

chosen for linearization of the experimental results, giving correlation coefficients for the linear

plots drawn through the origin consistently over 0.9 and nearing 1.0 for experiments above 40 °C,

confirming that leaching of As from enargite in NaHS–NaOH solution is controlled by the diffusion

mechanism as controlling step. Kinetic constant values from Figure 8(a) were used to construct an

Arrhenius plot for determination of activation energy as shown in Figure 8(b), giving an excellent

linear dependency of the kinetic constant with temperature. From the slope of the straight line in this

figure, the activation energy was determined as 70.26 ± 4.74 kJ/mol.

The effect of particle size, NaHS concentration and stirring rates on As leaching were also investi-

gated as a function of leaching time at 80 °C. High As leaching of over 99 % was achieved within

1 h and 2 h for the finer fractions (-32 µm) and (-75+32 µm) respectively. For the coarser fractions,

As leaching remained consistently below 60 % indicating that higher As removal can be achieved at

finer fraction, attributed to by increased surface area of the sample.

0.0

0.2

0.4

0.6

0.8

1.0

0 20 40 60 80 100 120

363K

353K

343K

333K

313K298K

1−3(1

−α

)2/3+

2(1

−α

)

Leaching time, t / min

0.00270 0.00285 0.00300 0.00315 0.00330 0.00345

-14

-13

-12

-11

-10

-9

-8

R = -0.99103

Root−MSE (SD) = 0.29009

ln k

1/T

Ea = 70.26 ±4.74 kJ/mol



To investigate effect of NaHS concentration, the (-75+32 µm) fraction was leached in 50 (0.68 M),

75 (1.01 M), 100 g/L (1.35 M) + 50 g/L (1.25 M) NaOH for 120 min. Slurry temperature and stir-

ring rate were fixed at 80 °C and 550 rpm respectively. At 100 g/L (1.35 M) NaHS addition, over

99 % As is removed within 2 h, however, the leaching rate decreases when NaHS concentration is

reduced to 75 (1.01 M) and 50 g/L (0.68 M), reaching up to 85 % and 55 % for the two NaHS con-

centrations respectively in 2 h. Low As removal at 50 and 75 g/L NaHS addition could be attributed

to by hydrolysis of S2-

ions to HS- ions as in both cases pH values decreased from 13.7 for 100 g/L

NaHS addition to below pH 12.7 and pH 12.3 for 75 and 50 g/L NaHS respectively. For tests at

100 g/L NaHS addition, the slurry pH and potential (Eh vs. SHE) remained consistently over 13.5

and Eh values ranging between -500 to -550 mV.

To investigate effect of stirring rate, the (-75+32 µm) fraction was leached in

100 g/L (1.35 M) NaHS + 50 g/L (1.25 M) NaOH at 80 °C under three different stirring rates. At

550 rpm, over 99 % of arsenic removal was achieved within 2 h; however, on decrease in stirring

speed arsenic removal proceeds slower and reached 89 and 82 % in 2 h for 360 and 200 rpm,

respectively. These results suggest that the leaching of enargite in NaHS–NaOH media is of diffu-

sion control, since a chemical reaction controlled process would show not much effect of stirring

speed on the leaching rate.

4 Conclusion

Selective leaching of As impurity in Cu-ores/concentrates containing enargite (Cu3AsS4) by leaching

in NaHS and NaOH media was investigated in this work. The results obtained and discussed can be

summarized as follows:

1. As contents in Cu–ores/concentrates can be reduced to <0.50 wt. % within 1 to 3 h for samples

with As content <4.0 wt. % and within 3 to 6 h for samples with As content <10.0 wt. % at tem-

peratures between 80 to 90 °C. Such result was also obtained for pulp density at 1000 g/L.

2. Enargite (Cu3AsS4) was transformed to Cu2S as a result of reduction of As5+

in enargite being

reduced to As3+

(Na3AsS4) by S2-

offered by NaHS. X–ray diffraction analysis of leach residues

and solution pH (> 12.50) and potential (< -500 mV (vs. SHE)) measurements correspond well

with this result.

3. Selective leaching of As from enargite (Cu3AsS4) and subsequent formation of Cu2S in

NaHS–NaOH system occurs according to the reaction:

2Cu3AsS4 + 3NaHS + 3NaOH = 2Na3AsS4 + 3Cu2S + 3H2O.

4. The kinetics of As leaching from enargite in NaHS–NaOH solution was well represented by the

diffusion controlled model; kt = 1-3 (1-α)2/3

+2 (1-α) and the activation energy was calculated to

be 70.26 ± 4.74 kJ/mol.



As impurity in Cu-ores/concentrates can be successfully removed be leaching in NaHS media as

shown by the results in this work. Treatment of solution from the leaching operation for removal of

As is a next step for complete process development.

References

[1] LATTANZI P., DA PELO S., MUSU E., ATZEI D., ELSENER B., FANTAUZZI M., ROSSI

A.: Enargite oxidation: A Review, Earth–Science Reviews, 2008, 86, pp. 62–88.

[2] FILIPPOU D., ST–GERMAIN P., GRAMMATIKOPOULOS.: Recovery of metal values

from Copper–Arsenic minerals and other related resources, Mineral Processing & Extractive

Metall. Rev., 2007, 28, pp. 247–298.

[3] MIHAJLOVIC I., STRBAC N., ZIVKOVIC Z., KOVACEVIC R., STEHERNIK M.: A po-

tential method for arsenic removal from copper concentrates, Miner. Engin., 2007, 20, pp. 26–

33.

[4] BALĂŽ P. and ACHIMOVIČOVĂ M.: Selective leaching of antimony and arsenic from me-

chanically activated tetrahedrite, jamesonite and enargite, Int. J. Miner. Process, 2006, 81, pp.

44–50.

[5] GUO H. and YEN W.T.: Selective flotation of enargite from chalcopyrite by electrochemical

control. Miner. Engin., 2005, 18, pp. 605–612.

[6] VIŇALS J., ROCA A., HERNĂNDEZ M.C., BENAVENTE O.: Topochemical transforma-

tion of enargite into copper oxide by hypochlorite leaching, Hydrometallurgy, 2003, 68, pp.

183–193.

[7] WELHAM N.J.: Mechanochemical processing of enargite (Cu3AsS4), Hydrometallurgy, 2001,

62, pp. 165–173.

[8] PADILA R., FAN Y., SANCHEZ M., WILKOMIRSKY I.: Processing high arsenic copper

concentrates. In: Sanchez, M.A., Vergara, F., Castro, S.H., (Eds.), 1998, Environment, Inova-

tion in Mining and Mineral Technology vol. 2. University of Conception, Chile, pp. 603–612.

[9] SENIOR G.D., GUY P.J., BRUCKARD W.J.: The selective flotation of enargite from other

copper minerals — a single mineral study into relation to beneficiation of the Tampakan de-

posit in the Philippines, Int. J. Miner. Process, 2006, 81, pp. 15–26.

[10] KANTAR C.: Solution and flotation chemistry of enargite, Colloids and Surfaces A: Physical

and Engineering Aspects, 2002, 210, pp. 23–31.

[11] FORNASIERO D., FULLSTON D., LI C., RALSTON J.: Separation of enargite and tennan-

tite from non-arsenic copper sulfide minerals by selective oxidation or dissolution, Int. J.

Miner. Process, 2001, 61, pp. 109–119.



[12] PAUPORTÊ TH., SCHUHMANN D.: An electrochemical study of natural enargite under

conditions relating to those used in flotation of sulfide minerals, Colloids and Surfaces A:

Physicochemical and Engineering Aspects, 1996, 111, pp. 1–19.

[13] MENACHO J.M., ALIAGA, W., VALENUELA R., RAMOS V., OLIVARES L.: Selective

flotation of enargite and chalcopyrite, Minerals, 1993, 48 (203), pp. 33–39.

[14] BALĂŽ P. and ACHIMOVIČOVĂ M.: Mechano–chemical leaching in hydrometallurgy of

complex sulphides, Hydrometallurgy, 2006, 84, pp. 60–68.

[15] CURRELI L., GHIANI M., SURRACCO M., ORRŬ G.: Beneficiation of a gold bearing

enargite ore by flotation and As leaching with Na–hypochlorite. Minerals Engineering, 2005,

18, pp. 849–854.

[16] DELFINI M., FERRINI M., MANNI A., MASSACCI P., PIGA L.: Arsenic leaching by Na2S

to decontaminate tailings coming from colemanite processing, Miner. Engin., 2003, 16, pp.

45–50.

[17] BALĂŽ P., ACHIMOVIČOVĂ M., BASTL Z., OHTANI T., SĂNCHEZ M.: Influence of

mechanical activation on the alkaline leaching of enargite concentrate, Hydrometallurgy,

2000, 54, pp. 205–216.

[18] HERREROS R.O., QUIROZ HERNĂNDEZ M.C., VIŇALS J.: Dissolution kinetics of enar-

gite in dilute Cl2/Cl¯ media, Hydrometallurgy, 2002, 64, pp. 153–160.

[19] DUTRIZAC J.E. and MACDONALD R.J.C.: The kinetics of dissolution of enargite in acidi-

fied ferric sulfate solution, Can. Metall. Q., 1971, 11(3), pp. 469–476.

[20] TONGAMP W., TAKASAKI Y., SHIBAYAMA A.: Removal of arsenic from copper

ores/concentrates by alkaline leaching in NaHS media, Hydrometallurgy, 2009, 98, pp. 213–

218.


Copper Leaching from Molybdenite in Acidic

FeCl3 Solutions with FeCl2 Yan Zhang, Narangarav Tumen-Ulzii, Zhibao Li

Chinese Academy of Sciences

Institute of Process Engineering

National Engineering Laboratory for Hydrometallurgical Cleaner Production Technology

Beijing 100190, China

Keywords: Molybdenite, copper, leaching

Abstract

This work aims to remove copper from molybdenite ore by leaching. The leaching process of

covellite and chalcopyrite was investigated in the media of FeCl3-FeCl2-HCl. The results showed

that leaching temperature, solid to liquid ratio and the concentration of FeCl3 and FeCl2 (or redox

potential) affected on the leaching significantly. Numerical simulation showed that the hydrogen

ion activity and the boiling point of the solution increases with the addition of FeCl2. Data analysis

using a shrinking core model suggested that the leaching process is controlled by chemical reaction

with activation energy of 79 kJ/mol.

1 Introduction

Molybdenum disulfide (MoS2) with the purity of more than 98 % is a good solid lubricant, and it is

used in automobile, aerospace and military field. Molybdenite concentrate, as usually marketed,

contains the mineral MoS2, along with silicious material, iron, copper, and other impurities [1].

Copper is an undesired impurity in molybdenite concentrate. Selective leaching of copper is the

premise to get high purity MoS2.

Copper occurs mainly as covellite (CuS), chalcopyrite (CuFeS2), chalcocite (Cu2S) and cuprite

(CuO) in molybdenite ore. Many approaches have been reported for the leaching of copper from

molybdenite concentrate, including leaching with dichromate-sulphuric acid mixture [2], HF & HCl

[3], FeCl3 in acidic solutions [4] and other chemicals. Usually, both oxidizing agent and acid are

included in these methods. The oxidizing agents commonly used for the leaching of copper include

Zhang, Tumen-Ulzii, Li


Cu2+

, Fe3+

, O2, Cl2 and so on [5]. Among them, Fe3+

is more efficient and can be recovered more

easily than others. When it comes to acid, HCl and H2SO4 are commonly used in copper leaching and

HCl made the leaching more easily than H2SO4 [6, 7].

FeCl3 and HCl were selected as the oxidizing agents and acid respectively in the leaching solutions

for this work. Moreover, FeCl2 was also added to the solution. On one side, FeCl2 can increase the

activity of H+, on the other, it can increase the boiling point of the solution [8, 9]. Furthermore, the

redox potential, determined by [Fe3+

] / [Fe2+

] ratio, decreases with FeCl2 addition.

Mineralogical analysis showed that CuS and CuFeS2 were the main impurities in the molybdenite

concentrate used in this work. The objective of the present work is to remove these copper

sulphides from molybdenite concentrate. Experiments were carried out to investigate the effects of

temperature, solid to liquid (S/L) ratio, and concentration of Fe3+

and Fe2+

. The kinetic of the copper

extraction were also studied.

2 Experimental

2.1 Materials

The molybdenite concentrate used in this work was obtained from the Erdenet Mine of Mongolia.

The particle size of the concentrate was from 200 to 500 mesh. The chemical composition

determined by X-Ray Fluorescence (XRF) is shown in Table 1. The phase identification and image

analysis were achieved by polarizing microscope and scanning electron microscope-energy

dispersive spectrometry (SEM-EDS). Polarizing microscope micrograph in Figure 1 combined with

SEM-EDS analysis in Table 2 showed that the main phases of the concentrate were MoS2, and the

impurities were CuS, CuFeS2 and SiO2.

Copper Leaching from Molydenite in Acidic FeCl3 Solutions with FeCl2


Table 1: The elements content of the molybdenite concentrate by XRF

Chemical species Content (% w/w) Chemical species Content (% w/w)

Mo 50.09 Ca 1.78

S 37.01 Mn 0.06

Fe 4.30 Al 0.09

Cu 2.71 K 0.34

Si 1.96 Mg 0.34

Pb 0.07 Cl 0.41

Figure 1: The phases of molybdenite concentrate investigated by polarizing microscope

(1-CuS, 2-SiO2, 3-MoS2, 4-CuFeS2, 5-MoS2)

2.2 Methods

Leaching experiments were carried out in a round bottom glass reactor of 500 ml. The leaching

experiments were done under the following conditions: leaching time, 0～4 h; temperature,

70～100 °C; solid / liquid ratio, 0.1-0.35 kg/L; the concentration of FeCl3, 100-200 g/L; the

concentration of FeCl2, 200-300 g/L; stirring speed, 300 rpm; and pH=0. The leaching solution was

filtrated and analyzed by atomic absorption spectrophotometer (AAS).



Table 2: The SEM-EDS analysis of the position 1-5 in Figure 1

Elements

%

Position

O Si S Fe Cu Al Mo

1 1.14 0.36 33.05 0.36 65.13 - -

2 57.87 39.55 2.22 0.35 - - -

3 - - 46.73 - - 0.57 52.7

4 2.81 0.37 36.79 28.73 31.04 0.26 -

5 5.36 - 45.8 0.56 - 0.44 47.84


3.1 Eh-pH equilibrium diagrams and electrochemical mechanism

Figure 3 and Figure 4 present Eh-pH equilibrium diagrams generated by chemical reaction and

equilibrium software HSC chemistry at 25 °C for Cu-Fe-S-H2O and Fe-Cl-H2O systems. From

Figure 3 it can be seen that both CuS and CuFeS2 dissolve at pH lower than 4. A comparison of

Figure 4 and Figure 3 shows the redox potential of Fe3+

is much higher than that of CuFeS2 and

CuS, and can dissolve them at pH lower than 0. So FeCl3 solutions with pH=0 were prepared as the

leaching solution.

Equation (1) is reported to represent the dissolution of chalcopyrite in ferric chloride [10]. Covellite

dissolution in acidic FeCl3 solutions can be represented by Equation (2) [11]. Consequently, the

oxidation proceeds according to an electrochemical mechanism (Equations (3), (4) and (5)) occurs

at the anode parts of the mineral, and iron reduction occurs at the cathode parts.

SFeClCuClFeClCuFeS 254 2232 ++=+ (1)

SFeClCuClFeClCuS ++=+ 223 22 (2)



Anode:

022 SCueCuS +=−+−

(3)

022

2 4 SFeCueCuFeS ++=−++−

(4)

Cathode:

+−+

=+23

FeeFe (5)

Figure 3: Eh-pH diagram of the Cu-Fe-S-H2O system at 25 °C

14121086420

2.0

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5

-2.0

Cu - Fe - S - H2O - System at 25.00 C

C:\HSC5\EpH\CuFeS25.iep pH

Eh (Volts)

H 2 O L im it s

CuFeS2

CuS

Cu(OH)2

Cu(+a)

Cu(+2a)



Figure 4: Eh-pH diagram of the Fe-Cl-H2O system at 25 °C

3.2 Effect of leaching temperature

The effect of temperature on the copper extraction was studied over the range between 70-100 °C,

at pH 0 and S: L = 1: 10. The concentrations of FeCl3 and FeCl2 in the leaching solutions are 100

and 300 g/L respectively. The leaching curves in Figure 5 indicate a marked effect of temperature

on the copper extraction. It can be observed that copper extraction increased from 31.9 % to 63.1 %

as the temperature increased from 70 °C to 100 °C after leaching for 4 hours.

14121086420-2

2.0

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5

-2.0

Fe - Cl - H2O - System at 25.00 C

C:\HSC5\EpH\FeCl25.iep pH

Eh (Volts)

H 2 O L im it s

Fe

Fe2O3

Fe3O4Fe(OH)2

Fe(+3a)

Fe(+2a)



Figure 5: Effect of temperature on the copper extraction in the ferric chloride leaching of

molybdenite concentrate (pH = 0, S: L = 1:10, FeCl3 concentration: 100 g/L,

FeCl2 concentration: 300 g/L)

3.3 Effect of solid: liquid ratio

The effect of solid: liquid (mass/ volume) ratio from 1:2.5 to 1:10 on copper extraction was studied

using an initial FeCl3 concentration 200 g/L, FeCl2 concentration 300 g/L, and temperature 100 °C.

It can be seen from Figure 6 that the leaching rate gradually increases in copper leaching process

with the decrease of solid: liquid ratio, and the copper extraction of leaching with S:L = 1:10 is

almost 20 % higher than leaching with S:L = 1:2.5 after leaching of 4 hours. This may be because

of that the decrease in solid: liquid ratio enhances medium diffusion and makes the reaction more

sufficient [12].

0 50 100 150 200 250

0

10

20

30

40

50

60

70

Co

pp

er

Extr

ac

tio

n (

%)

Leaching Time (min)

700C

800C

900C

1000C



Figure 6: Effect of S: L on the copper extraction in the ferric chloride leaching of molybdenite

concentrate (pH=0, 100 °C, FeCl3 concentration: 200g/L, FeCl2 concentration: 300g/L)

3.4 Effect of concentration of ferric chloride and ferrous chloride

The leaching of copper is an oxidation process for both chalcopyrite and covellite. Several

researchers have reported that the chalcopyrite dissolution rate depends on the redox potential of the

solution, and that the best result is achieved under moderately oxidizing conditions [13]. The

leaching of covelite would be also affected by the redox potential. In FeCl3 – FeCl2 leaching

system, the initial redox potential of the leaching solution can be controlled by FeCl3/ FeCl2 ratio,

according to the Nernst equation (Equation 6),

)(

)(log059.067.0

2

3

+

+

+=

Fea

FeaE

(6)

As shown in Figure 7, at a constant concentration of FeCl2 (300 g/L), copper extraction increased

with increasing FeCl3 concentration. And it can be seen in Figure 8 that copper extraction decreased

with increasing FeCl2 concentration at a constant FeCl3 concentration (100 g/L).

0 50 100 150 200 250

0

10

20

30

40

50

60

70

80

Co

pp

er

Extr

ac

tio

n (

%)

Leaching Time (min)

S:L=1:10

S:L=1:5

S:L=1:2.5



Figure 7: Influence of FeCl3 concentration on the copper extraction in the ferric chloride leaching

of molybdenite concentrate (pH=0, S:L=1:10, 100 °C, FeCl2 concentration: 300 g/L)

Figure 8: Influence of FeCl2 concentration on the copper extraction in the ferric chloride leaching

of molybdenite concentrate (pH=0, S:L=1:10, 100 °C, FeCl3 concentration: 100 g/L)

0 50 100 150 200 250

0

10

20

30

40

50

60

70

80

200 g/L FeCl3- 300 g/L FeCl

2

100 g/L FeCl3- 300 g/L FeCl

2

Co

pp

er

Extr

acti

on

(%

)

Leaching Time (min)

0 50 100 150 200 250

0

10

20

30

40

50

60

70

80

Co

pp

er

Extr

ac

tio

n (

%)

Leaching Time (min)

100g/L FeCl3- 300g/L FeCl

2

100g/L FeCl3 - 200g/L FeCl

2



3.5 Synergetic effect of FeCl2

The synergetic effect of FeCl2 was studied by numerical simulation using OLI Systems’ Stream

Analyzer software package [9]. As can be seen in Figure 9, the hydrogen ion activity (a H+)

increases rapidly with the increase of the concentration of FeCl2. The value of a H+ enhanced more

than two times duo to the addition of 3 mol/ kg FeCl2. Then the efficiency of the leaching can be

enhanced when FeCl2 is added into the leaching solution. Moreover, the addition of FeCl2 can

increase the boiling point as shown in Figure 10. The curves in Figure 5 indicate that higher copper

extraction occurs at higher temperature. So the optimum leaching temperature is near the boiling

point of the solution, and increase in the boiling point of the solution can improve the leaching

efficiency of copper. As can be seen in Figure 10, the boiling point of the leaching solution is lower

than 102 °C without FeCl2, but it increases to more than 107 °C when 3 mol/ kg of FeCl2 has been

added.

Figure 9: Influence of concentration of FeCl2 on a H+ in a solution of HCl (0.1 mol/ kg) +

FeCl3 (1 mol/ kg) + FeCl2 (0-3 mol/ kg) at 25 °C with the aid of OLI Systems’ Stream

Analyzer software

0.0 0.5 1.0 1.5 2.0 2.5 3.00.08

0.10

0.12

0.14

0.16

0.18

aH

+ (

mo

l/ k

g)

Concentration of FeCl2 (mol/ kg)



Figure 10: Influence of concentration of FeCl2 on the boiling point of the solution of HCl

(0.1 mol/kg) + FeCl3 (1 mol/kg) + FeCl2 (0-3 mol/kg) with the aid of OLI Systems’

Stream Analyzer software

3.6 Kinetic studies

The leaching kinetic was studied based on a shrinking core model using the experimental data

shown in Figure 5. As shown in Figure 11, the dissolution of copper was found to follow Equation

7 [10, 14],

tkX s=−−3/1)1(1

(7)

RTEaAek

/−

= (8)

RTEAk a /lnln −= (9)

where X is the fraction of leached copper amount at time t (min), and ks is the apparent rate constant

(1/min). This result implies that the leaching process is controlled by surface chemical reaction.

Based on Arrhenius equations (Equation 8 and Equation 9), an activation energy was evaluated as

79 kJ/mol from the Arrhenius plot (Figure 12).

0.0 0.5 1.0 1.5 2.0 2.5 3.0101

102

103

104

105

106

107

108

Bo

ilin

g p

oin

t ( °° °°

C)

Concentration of FeCl2 (mol/ kg)



Figure 11: Variation of 1-(1-X)1/3

over time at various temperatures

Figure 12: Arrhenius plot for the dissolution of copper from molybdenite concentrate based on the

apparent rate constants calculated from Figure 5

40 60 80 100 120 140 160 180 200 220 240 260

0.104

0.112

0.120

0.128

0.136

0.144

0.152

0.160

0.168

0.176

0.184

70 °C

80 °C

90 °C

1-(

1-x

)1/3

Time (min)

2.74 2.76 2.78 2.80 2.82 2.84 2.86 2.88 2.90 2.92

8.0

8.2

8.4

8.6

8.8

9.0

9.2

9.4

9.6

9.8

-Ln

k (

s-1

)

T-1

/10-3

(K-1

)



4 Conclusions

In this work, the dissolution process of covellite and chalcopyrite in molybdenite has been

investigated. It was found that dissolution rate of them increased with an increase in temperature,

and the optimum leaching temperature was near the boiling point of the solution. The S: L ratio

affect on the leaching since relatively less solid made the diffusion more easily. The initial redox

potential of the leaching solution was another important factor for the leaching, which can be

controlled by FeCl3/FeCl2 ratio. The addition of FeCl2 increased the hydrogen ion activity of the

solution and the boiling point of the solution. The kinetic study using a shrinking core model

suggested that the copper extraction rate is controlled by the surface reaction with activation energy

of 79 kJ/ mol.

Acknowledgements

The support of the National Nature Science Foundation of China (Grant No.20876161) and the

National Basic Research Program of China (973 Program, 2007CB613501) are gratefully

acknowledged.

References

[1] Byung-Su K., Manis K.J., Jinki J., Jae-chun L.: Leaching of impurities for the up-gradation of

molybdenum oxide and cementation of copper by scrap iron, Int. J. Miner. Process., 2008, 88,

pp. 7-12.

[2] Ruiz, M.C., Padilla, R.: Copper removal from molybdenite concentrate by sodium dichromate

leaching, Hydrometallurgy, 1998, 48 (3), pp. 313-325.

[3] Mankhand, T.R., Prasad P.M.: Lime enhanced hydrogen reduction of molybdenite.

Metallurgical Transactions, 1982, 13, pp. 275-282.

[4] Ikumapayi. F. K.: Purification of molybdenite concentrates, Master Thesis of Lulea

University of Technology, 2008.

[5] Herreros O., Vinals J.: Leaching of sulfide copper ore in a NaCl - H2SO4-O2 media with acid

pre-treatment, Hydrometallurgy, 2007, 89, pp. 260-68.

[6] Lu Z.Y., Jeffrey M.I., Lawson F., An electrochemical study of the effect of chloride ions on

the dissolution of chalcopyrite in acidic solutions, Hydrometallurgy, 2000, 56, pp. 145-155.



[7] Lu Z.Y., Jeffrey M.I., Lawson F., The effect of chloride ions on the dissolution of

chalcopyrite in acidic solutions, Hydrometallurgy, 2000, 56, pp. 189–20.

[8] Demopoulos G. P., Zhibao L., Becze L., Moldoveanu G., Cheng T., Harris G. B.: New

technologies for HCl regeneration in chloride hydrometallurgy. ERZMETALL, 2008, 61, pp.

89-98.

[9] Demopoulos G. P., Zhibao L., Becze L., Moldoveanu G., Cheng T., Harris G.B.: New

technologies for HCl regeneration in chloride hydrometallurgy european metallurgical

conference（EMC） 2007, June 11-14, 2007, Dusseldorf, Germany.

[10] Mohammad A. H., Sam K., Adnan A. H.: Ferric chloride leaching of chalcopyrite: synergetic

effect of CuCl2, Hydrometallurgy, 2008, 91, pp. 89–97.

[11] Antonijevic M.M., Dimitrijevic M.D., Stevanovic Z.O., Serbula S.M., Bogdanovic G.D.:

Investigation of the possibility of copper recovery from the flotation tailings by acid leaching,

Journal of Hazardous Materials, 2008, 158, pp. 23-3.

[12] Zeng L., Yongcheng C.: A literature review of the recovery of molybdenum and vanadium

from spent hydrodesulphurization catalysts Part II: Separation and purification,

Hydrometallurgy, 2009, 98, pp. 10-20

[13] Cordob E.M., Munoz J.A., Blazque M.L., Gonzale F., Ballester A.: Leaching of chalcopyrite

with ferric ion. Part II: Effect of redox potential, Hydrometallurgy, 2008, 93, pp. 88-96.

[14] Cordob E.M., Munoz J.A., Blazque M.L., Gonzale F., Ballester A.: Leaching of chalcopyrite

with ferric ion. Part III: Effect of redox potential on the silver-catalyzed process,

Hydrometallurgy, 2008, 93, pp. 97-105.


Recycling



Shifting Core Business Vision: From Copper to

Polymetallics – A Recycling Point of View

Juan Ignacio Barturen Zabala

Codelco Chile

Once Norte 1291. Edificio Corporativo Codelco Norte

Calama, Chile

Keywords: Recycling, polymetallics

Abstract

It is everybody knowledge that mining business is a time and site restricted business due to ore

grade and technology availability to extend to a maximum its profits. This ore grade constant de-

crease is somehow balanced by both technologies breakthrough that allow recovering targeted ele-

ments (i.e. copper) from lower grade ores, and the possibility to recover others minerals that may

slow down cash costs growing through subproduct’s commercializing.

Chuquicamata mine is not an exception, and given its near 100 years of production life, lower

grades are no minor subject. Although almost every copper related recovery technology is being

used, or had being used, and several arsenic confining technologies had being developed to keep

producing copper in spite of this impurity; higher costs, human resource problems, and lower pro-

duction rates are unavoidable in the short term, and will keep increasing its effects. Something must

be made to keep Chuquicamata a profitable business.

In the last 10 years many social transformation had occur and still surprise us with new forms of

communication, science develops, etc., but there is one relevant change (to mine business) that

should shift how do we think the whole mining business: Recycling.

But recycling not just as a secondary business for better respect the environment and avoid danger-

ous waste, but a main crucial shift in how we evaluate our ores and products and tails.

How about to achieve commercialize every last atom of every element founded, whichever these

are, wherever flow they concentrated in the cooper process? How about to avoid waste?

Some by-products recovery had point in this direction, Molybdenum mainly, and also Rhenium,

Silver and Gold from anodic slug, Sulphuric acid and so on, but because those products are in such

high concentrations that no to do so was unthinkable, a very bad business.

Barturen


But, what about the rest of the elements contained in tails, slugs, intermediate process flows and the

like? What if there is more to recover (and profit)? To answer this is both a commercial and tech-

nological matter in the first place, is there a market for those elements? Is possible to recover those

elements without interfering the core business?

Nevertheless those questions are crucial for the success of such a challenge, first we have to face

one main, primary, foremost question: is this endeavor relevant for business? Is there any interest in

undertake this strategy? Some guidelines for these questions should be revealed.

1 Introduction

In this first XXI century decade many changes had happened, as to the start of this new millennium,

the complete deploy of the ‘new economy’ of internet [1, 2], the arise of Chine and India as a

worldwide player [3], the loss of confidence in the economy thesis that build the subprime crisis [4],

the ever growing agreement in the Climate warming and its consequences, the ecology concept get

into our lives, among others changes had had a big impact in how we see and understand our world.

This changes had occur in every aspect of our lives, how we communicate between each other, how

companies treat their employees, buy theirs supplies, distribute their products. The supply chain

management have developed to a proper business by its own [5]; marketing had evolved from sell-

ing to bonding firm to the customer, and so on. This implies that companies had included many as-

pects into their strategy not evaluated before, bringing the theory of the stakeholders to front be-

cause profit and costs aren't the only aspects to analyse firms and the conclusions of such analysis

aren't enough to understand and withstand environmental changes and challenges.

But is the ecologic view of the world an aspect that is crucial now-a-days to the mining business,

and is going to increase its importance. From Lean production [6], to the Factor 4 [7] and other en-

vironmental thesis there is one point of agreement: recycling is not an option. Is a must in a world

deep concerned in maximize the output of the production factors: land, capital, labour, technology;

without any environmental cost or ambient damage.

Then, to better profit our resources is not just an enterprise policy, but a main goal for our society

due to the scarcity of them (especially of the minerals that allow built the world as we know it) and

to avoid the impact on the environment that waste produce.

2 Environment factors

In 2009, the high prices cycle that started on 2004 come to an end, and in some cases, a very painful

end. Many conclusions have been written these months regarding this, but also so many agree that if

Chine (and India) keeps its growing momentum, the world mining industry would be safe, or less

Shifting Core Business Vision: From Copper to Polymetallics


damaged. And this have been validated with facts, as most metal industry had not suffer any deep

damage, and most had recover their dynamism from 2nd

Q of 2009.

But this might disguise the real problem or lessen the effects over an industry that had got into

structural problems everywhere. Mining operation faces all around the world supply problems,

mainly supply of water, energy and qualified workers. Although water & energy problems aren't

easy to resolve and may block the mine development and extraction, is the labour factor – once wa-

ter and energy are solved – the biggest challenge to keep mining business running. One main reason

for this is the location for these operations as they are remote located, far away from industry end

users, big cities and several of them are in deserts, jungles or the like. So, there are two ways to ad-

dress this problem: bring enough well prepared workers to mine sites, or prepare from the locals the

needed working force. Both alternatives faces the same problem in dealing with people so mining

facilities more and more depends on technology to diminish worker's lack of preparation.

Moreover, this highly specialized work force, as they became aware of their importance, put another

problem to the struggling mining industry: the strike menace. This fact, plus the ever lowering ore's

grade put forward one simple question, how can industry maximize the output of their labour fac-

tor?

On one hand, there is the possibility of, through education, increase the productivity of workers,

giving them tools for innovations in their work place that could led to patenting and knowledge, and

preparing them well enough in order to avoid work accidents.

On the other hand, technology allow to overcame this by developing process control, robots, intelli-

gent vehicles, energy efficient equipment, etc, that allow companies to lower demands of workers

unions and to keep controlled the payroll. In this sense, the core-business philosophy had an impor-

tant role in disaggregate large companies into single-focus, less integrate firms, and consequently,

less complex to manage the organization.

The industry struggles to keep running its operations despite these problems and restrictions, and

have found some relief focusing on their main business, but as environmental factor keep getting

stronger, and prices for every metal stays pushing, there must be a solution for these supply prob-

lems and a turnaround over the core-business vision.

3 The core-business/integration dilemma

Delocalization is one of the main effects over world economy of following core-business strategy as

the idea of focusing only in ‘what make us different’, by allowing other industries to emerge (ser-

vices mainly) in any country, as long as this choice bring cost and quality advantages.

To mining industry this also have been true, as companies define their business and got rid of any

operation not directly related to their core-business, and in some others respects it didn't, for one

reason mainly: you can not delocate mines and processing facilities. But mining and mineral proc-

Barturen


essing firms had deploy the core-business philosophy (with some exceptions) focusing in their main

products, which results in copper companies only producing copper, steel companies only concern-

ing in iron and so on. This had act as a blindfold to strategic vision.

From early 90's, the core business concept drove all industries to a narrow view of their relation to

the customer and the rest of the world. This concept have had a big impact in how mining compa-

nies think themselves, as many start to sell their secondary business to better focus on their core-

business. For sub product's development these have been disastrous as many plants had to close, and

no more integration projects have been developed. For porphyry copper mines, which characterized

for having many metals contained among the ore, these means that tails gets lost with no value add-

ed but water recovering (if mine is in a dry zone). Tailing dams are now even an environment prob-

lem due to the foreclose actions needed and soil remediation.

Many questions may arise from this fact: is not possible to lessen this problem? Is there any alterna-

tive? What else can be done from these materials?

The last question is very significant because the answer may lead to the conclusion that something

else can be done in terms of better profit from these wasted resources, which for certain have some

value due to its metal contents. It may be argued that the Fe, Si and other elements contained in tails

and the like have not the value that copper has, or that there is no technology available for recover

these elements economically, but it's also true that need brought always new forms to adjust industry

to new paradigms.

Figure 1 shows a typical copper processing operation, on the left hand, the sulphur processing line,

and on the right side, the oxide processing line. Is possible to state that -for a given amount of re-

sources located on mine site- there is a limited production capacity for the targeted element, as well

as limited technology availability for this element to be recovered. So, for the question of what else

can be done, it's imperative to look down to the “waste”, “garbage” flows produced by recovering

copper in this case. Several by-products are being recovered in some facilities from anodic Slug,

and by concentrating Sulphuric acid, for example, but once again arise the question of what else can

be done.



Figure 1: Copper processing description

There are two kinds of waste generated, one related to different flows and materials discarded by

mineral processing operations; and stocks, related to large amount of materials dispose as garbage.

So different strategies could be follow in recovering valuable elements from these wastes. The for-

mer waste, named discarded flow, could lead to discrete modules that will extract other sub-

products – depending on which element is available – without interfering core normal operation.

The latter, named stock consists in large deposits of tails and graves, which historically are being

dumped with no further use, but nevertheless contain valuable elements waiting to be extracted.

Then, it’s possible to think that new breakthroughs could follow the breakdown of the core-business

paradigm regarding new visions of what mining industry it’s all about, and where to find new value

in mining disregarded wastes.

The steel industry have seen these lasts years how big companies as ArcelorMittal start to integrate

new operations to protect themselves from price volatility, and supply shortcuts for their energy

(mainly coal) and special metals needs for steel alloys. This is the opposite direction followed by the

same company (both Arcelor and Mittal, previous the merge) that in previous years had sell some

coal mines, iron-rich sands, and minor business (as related to Steel production) following the core-

business drive, but facing the same problem of loosing operation flexibility, business stability, and

mainly, loosing strategic business opportunities related to supply or commercial ventures that could

allow to bypass – as this year made clear- or lessen the effects of a worldwide economic crisis.

Furthermore, imagine the possibilities for mining industry in developing new technologies to re-

cover more minerals form the same ore, the impact of such an endeavour may lead to a complete

revolution not only in the know-how and knowledge management assets of the firm, but in getting

more from the labour factor, and then, increasing many times the financial output (i.e., stock value)

by getting into business accountability resources now labelled as “waste”.

Barturen


4 The no-waste business policy

As had been stated, is no longer possible to keep wasting so many resources both from an economic

and environmental point of view. So a major shift in the business vision is required. The proposition

is to a no waste policy that will (almost automatically) lead to a polymetallic production to better

profit the resources available in situ on the ore and mines worldwide.

The propose policy tends to unite both environmental and economic objectives into one main enter-

prise goal of increase value for the company and avoid environmental damages to earth.

For certain there are several restrictions in what can be done to follow this policy, among all is the

technology availability what came in first place, as is imperative this technology to achieve the ob-

jective of a polymetallic production evolution from a single-metal facility. Yes, economic reasons

had determined why some elements are recovered and others don't, but these economic reasons are

mainly defined by the technology that makes it possible and its operating costs.

Some guidelines for doing this may be followed:

• Broad scope for metal recovery.

• Analyze and quantified metal content in ore, tails, tails dam, graves and any waste deposit.

• Screen metal occurrence for better focus on subproduct development.

• Search for technology availability.

• Design a proper business model for processing these products.

• Adjust financial aspects for capital investing and technology development.

• Encourage technology innovations.

Figure 2 describes the proposed no-waste policy process, the organizational responsible and the ex-

pected outcome for each step.



Figure 2: The No-Waste Policy

The target for a less volume of waste to dispose is achieve by a selective process intended to deter-

mine the economic potential to be recover by processing wasted materials, and the technical feasi-

bility to make this possible. Two main products could be defined, a less volume of waste to dispose,

which means a less costly disposal operation; and the innovation acquired by this search for new

metallurgic and chemical processes.

These guidelines are not exhaustive, and are yet to go through management approval, but the prima-

ry idea is to design a business strategy that, in the long term, allow the industry shift to a polymetal-

lic production intended to avoid any waste, given that every atom of element that goes through min-

eral processing had consume energy, water, chemicals and other resources, which in the long term

would be unacceptable to loss.

Maybe is 'common sense' that mineral processing only focused on what is profitable, disregarding

the rest, but is also true that 'common sense' isn't static, but is gradually increased by science and

technology [8]. Then, although now it seems that is not an option to recovers every element from

Barturen


the ore, this might easily change if some technology breakthrough happened, or the economic value

of this wasted elements increase to a certain level enough for allowing their recovery without eco-

nomic penalty.

This policy is meant to prepare the organization for seek and develop actions that would lead to a

polymetallic mineral processing, but is not a recipe to undertake this change, because this shall be

restricted by technology advances yet to come.

Maybe to declare that polymetallic is the future of the mining industry is a bold statement, but as

long as prices keep rocketing, ore's grades keep lowering and metal's world demand stay as strong as

consumer technology envisions new uses and aids to our lives, to dump any metal once crushed,

grind, floated, lixiviated or whatever process undertake will be a luxury that none would be able to

afford.

5 Conclusion

Our society will face major challenges in the near future regarding environmental problems, eco-

nomic paradigms changes, social stress and technology development. Most changes point in one

direction: to better use the resources available.

This means that not only climate global warming affect us as a distant not-so-clear menace, not only

the economy is changing its principles to avoid structural weakness, not only workers will become

more aware of their rights and not only technology will keep its trend of awesome advances; but

reassures that the scarcity concept contained in any extracting industry as mining and mineral

processing will be more and more relevant.

Organizations must be prepared for this change in how we value our resources, and mostly, how we

value waste; in order to keep its operations running in the long term. To do so, imply a major shift

in mining industry, that should drove them to maximize their productivity and resource recovery,

transforming -wherever possible- actual facilities into polymetallic processing operations that will

lead not only to a business economic burst, but also to lessen environmental impact.



References

[1] KELLY, K. (1999): New Rules for the New Economy, 16-18, New York.

[2] BLINDER, A.S. (2000): The Internet and the New Economy, Brookings Policy Brief, June,

N°60, Washington.

[3] ENGARDIO, P. (2007): Chindia: How Chine and India are revolutionizing global business,

McGraw-Hill, New York.

[4] SOROS, G. (2008): The New Paradigm for Financial Markets: The Credit Crisis of 2008 and

what it means, Public Affairs, New York.

[5] MENTZER, J.T.; DeWitt, W; Keebler, J.S.; Min, S.; Nix, N; Smith, C.D.; Zacharia, Z.,(2001):

Defining Supply Chain Management, Journal of Business Logistics, Vol 22, N°2.

[6] WOMACK, J.P.; JONES, D.T. (1996): Lean Thinking, Simon&Schuster, New York.

[7] Von WEIZSACKER, E.; LOVINS, A.B.; LOVINS, L.H., (1997): Factor Four: Doubling

Wealth – Halving Resources, Allen&Unwin.

[8] BUNGE, M. (1996): Intuición y Razón, p 152; Buenos Aires.



Solubility of Scorodite Synthesized by Oxidation

of Ferrous Ions

Tetsuo Fujita Etsuro Shibata, Takashi Nakamura

Dowa Metals & Mining Co., Ltd. Tohoku University, Institute of Multidisciplinary

Research for Advanced Materials

217-9 Shimo-kawabata Furumichi Iijima 2-1-1 Aoba-ku Katahira

Akita, Japan Sendai, Japan

Keywords: Scorodite, solubility, NaOH, NaCl, CaO

Abstract

An atmospheric scorodite synthesis process was developed, in which ferrous ions were oxidized by

oxygen gas in the presence of pentavalent arsenic ions. The synthesized scorodite was well-

crystallized in a short retention time of 1 to 7 h. The resulting scorodite particles were as large as

15 µm in diameter, had a moisture content of less than 10 % were readily washable, and featured

excellent packing properties. This scorodite synthesis can be incorporated into a hydrometallurgy

process.

The solubility of scorodite synthesized via this novel atmospheric process was investigated. Envi-

ronmental batch leaching tests were performed according to the United States Environmental Pro-

tection Agency’s Toxicity Characteristic Leaching Procedure (TCLP) test (method 1311), and the

Japanese Ministry of the Environment Notification No. 13 method, with minor modifications. In the

TCLP test, scorodite particles released a very low, almost negligible, concentration of arsenic in the

pH range 3 to 5, suggesting long-term stability. In leach tests with various pH solutions, it was

shown that the scorodite solubility was significantly affected by outside environmental factors such

as pH, rather than by long-term stability. In addition, the results showed that scorodite released high

concentrations of arsenic under specific leaching conditions. In particular, the presence of a combi-

nation of CaO and NaCl in alkaline leaching solutions had a significant effect on arsenic release.

Development of scorodite storage methods based on accurate evaluation of the conditions that cause

arsenic leaching, which will minimize the risk of environmental arsenic contamination, is crucial.

Fujita, Shibata, Nakamura


1 Introduction

Arsenic contamination is one of the world’s major environmental issues. Many methods have been

reported for the removal of arsenic from water environments [1–14]. Most reported methods involve

adsorption of arsenic on iron compounds and this phenomenon has long been a subject of interest

[15–17].

Arsenic is present in ore impurities and occurs in the by-products of metallurgical processes. In the

past, arsenic compounds were used in pesticides, herbicides, fungicides, and wood preservers. How-

ever, because of the environmental impact of As2O3 toxicity, many countries have banned the use of

arsenic-containing agricultural products and arsenic consumption has therefore greatly declined.

Thus, the production volume of arsenic far exceeds industrial demand, and the development of ef-

fective technologies for elimination, fixation, and storage of arsenic are highly sought after.

A novel atmospheric process for the synthesis of scorodite (FeAsO4⋅2H2O) has been developed and

reported [18–21]. The essential feature of this process involves precipitation of scorodite by oxidiz-

ing ferrous ions contained in a solution containing a high concentration arsenic (V). Scorodite crys-

tals release only very low or negligible levels of arsenic when tested according to the Japanese Min-

istry of the Environment Notification No. 13 method. However, this test result should not be

construed as the sole basis for judgment of scorodite stability. A wide range of experimental condi-

tions must be used for comprehensive analysis of scorodite leaching behavior.

Investigations into the leaching of arsenic have been widely reported in the literature. Nishimura and

Tozawa [23, 24] calculated the solubility products for FeAsO4 and found that heat treatment en-

hanced its immobilization. Robins [25-29] conducted a series of studies on the solubility and stabili-

ty of scorodite, comparing the thermodynamic properties of FeAsO4 and scorodite. Krause and Ettel

[30-32] examined the solubilities of ferric arsenate at various Fe/As molar ratios. Harris and Mo-

nette [33, 34] reported the solubilities of smelter/refinery-associated arsenic compounds, including

ferric arsenate. Emett and Khoe [35, 36] explored the effect of co-impurities such as cadmium and

lead on the solubility of arsenic. The recent work by Langmuir et al. [37, 38] provides new insights

into the thermodynamic behavior of scorodite. Bluteau and Demopoulos [39, 40] presented an ex-

tensive discussion of the mechanism of arsenic leaching from scorodite.

Although a large number of studies have been reported to date, they seem to fall short of establish-

ing the safest and most convincing methods for scorodite storage. This is primarily because of a lack

of evidence for long-term scorodite stability. The solubility is affected by time-dependent changes in

the stored scorodite environment, therefore research aimed at eliminating the risk of arsenic dissolu-

tion will be unlikely to yield robust conclusions regarding storage or disposal.

In this study, the solubility of crystalline scorodite synthesized by our novel atmospheric process

was investigated. Environmental leaching tests were performed according to batch leaching proce-

dures. The objectives of the present study were (1) to measure and compare the levels of arsenic

Solubility of Scorodite Synthesized by Oxidation of Ferrous Ions


leached from synthesized scorodite and (2) to examine the concentrations of arsenic leached from

scorodite into various aqueous solutions at various pH values to understand the scorodite solubili-

ties in various external environments.

2 Experimental

2.1 Materials

We used an arsenic (V) acid solution (labeled H3AsO4; content: 58-62 %; mean: 60 %). Analytical-

grade iron (II) sulfate 7-hydrate (FeSO4·7H2O) was used as the iron source. All chemicals used in

the tests were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). Compressed

oxygen gas (purity: 99.9 %, Akisan Kogyo K. K., Akita, Japan) was used as the oxidizing agent.

2.2 Synthesis of scorodite

The experimental procedure for the production of scorodite is shown in Figure 1. 4 L of arsenic test

solution (50 g/L) was measured and poured into a beaker, and a sufficient amount of FeSO4·7H2O

was weighed and added to the solution to give an Fe/As molar ratio of 1.5. The total volume of so-

lution was 4 L. The beaker was placed on a heater with a titanium stirrer and the reaction mixture

was agitated at 200 rpm until it reached a pre-determined temperature. When the required tempera-

ture was attained, oxygen gas (4 L/min) was introduced into the mixture via a glass tube to begin

reaction, while stirring at 800 rpm. After the reaction had progressed for a pre-defined period of

time, part of the slurry was collected, cooled to 60 °C, and its pH and oxidation reduction potentials

(ORP) were measured. The pH value was determined using manual temperature compensation

(MTC), set at 30 °C. ORP was measured with an Ag/AgCl reference electrode, with a 3.3 mol/L

KCl filling solution. The remaining reaction slurry was pressure-filtered through a 1 mm pore-size

polytetrafluoroethylene (PTFE) membrane filter at 0.4 MPa to separate the solid from the liquid.

The filtrate was subjected to inductively coupled plasma (ICP) measurements.

The solid filter cake was re-suspended and washed three to five times with distilled water until the

pH of the wash water exceeded 3.9. The weight ratio of solid to distilled water was 1 to 10. After

each washing, the solid was separated from the liquid by pressure filtration. Finally, the solid was

dried at 60 °C for 18 h and carefully pulverized in a mortar. The obtained solids were subjected to

chemical composition and powder characteristics analyses.



Figure 1: Procedure for scorodite synthesis

2.3 Analytical methods

ICP analyses were performed on the solid and on the post-reaction filtrate using inductively coupled

plasma atomic emission spectroscopy (ICP–AES) (ICAP577, Nippon Jarrell-Ash Co. Ltd., Kyoto,

Japan) to determine their chemical compositions. Scanning electron microscopy (SEM) was con-

ducted on solids using a Hitachi FE-SEM S-4500 (Hitachi Ltd., Tokyo, Japan).

2.4 Leach tests

2.4.1 Long-term leach tests

Long-term leach tests were performed according to the United States Environmental Protection

Agency’s Toxicity Characteristic Leaching Procedure (TCLP) test (method 1311), with minor mod-

ifications. The TCLP leach test is usually carried out as follows. A 100 g of sample and 2 L of acetic

acid buffer solution (pH 2:88 or 4.95) are mixed in a polyethylene bottle. Extraction is performed by

rotary shaking for 18 h and solid-liquid separation is carried out using a 0.6-0.8 µm pore-size glass

fiber filter (GFF). In this study, a 5 g (dry weight) sample was added to 100 g of extraction fluid in a

high-density polyethylene sample bottle; the bottle was placed horizontally, and the mixture was

shaken for a period of from 6 h to 35 days. The maximum duration of 35 days was based on data

reported by Bluteau and Demopoulos [39, 40], which showed the attainment of equilibrium in 5

Prepared solution 4 Liter

As content: 50 g/L

Fe(II)SO4.7H2O

Fe/As molar ratio: 1.5

Temp.: 95 oC

Stirring rate: 800 rpm

O2 gas flow: 4 L/min

Reaction time: 7 h

Reaction

Sampling

Pressure filtration

Precipitate

Washing

Scorodite powder Solution



weeks. After a pre-determined time, the sample bottle was removed from the shaker, the mixture

was filtered through a 0.2 µm PTFE membrane filter, and the filtrate was analyzed by inductively

coupled plasma mass spectrometry (ICP–MS) (SPQ9000, Seiko Instruments Inc., Chiba, Japan).

In these tests, the following buffers were used for the extraction fluid: 0.1 mol/L aqueous acetic acid

(pH 2.88), a mixture of 0.1 mol/L acetic acid solution and 0.1 mol/L sodium hydroxide solution

(pH 4.93), a mixture of 0.025 mol/L potassium dihydrogen phosphate and 0.025 mol/L sodium di-

hydrogen phosphate (pH 6.86), and 0.01 mol/L sodium tetraborate (pH 9.18). The samples were

sealed in the bottle under a normal air atmosphere.

2.4.2 Solvent-specific leach tests

Leach tests using different extraction fluids were carried out according to the method stipulated by

the Japanese Ministry of the Environment, with several modifications. After a 10 g sample and

100 g of extraction fluid were weighed and mixed in a high-density polyethylene bottle, the bottle

was sealed, placed horizontally and shaken for 6 h. Upon completion, the mixture was filtered

through a 0.20 µm PTFE membrane filter, and the filtrate was analyzed by ICP-MS. These experi-

ments employed the extraction fluids listed in Table 1.

Table 1: Solvent conditions used in the 6 h leach test

Condition of solvent pH

Distilled water 5.74

TCLP 2.88 (0.1mol/L CH3COOH solution) 3.00

TCLP 4.93 (0.1mol/L CH3COOH and NaOH solution) 4.88

Alkaline water conditioned by CaO 12.52, 12.30, 10.75, 10.14, 9.00

Alkaline water conditioned by Mg(OH)2 10.48, 10.41, 10.64, 9.00

Alkaline water conditioned by NaOH 12.50, 11.06, 9.70

Acid water conditioned by H2SO4 3.96, 2.97, 1.97, 1.22, 0.35

Acid water conditioned by HCl 3.99, 2.99, 1.98, 1.11, 0.13

Acid water conditioned by HNO3 3.98, 2.99, 1.99, 1.12, 0.12

8.2 g/L NaCl solution conditioned by CaO 12.50, 12.00, 11.06

8.2 g/L NaCl solution conditioned by H2SO4 5.80, 2.99

16.5 g/L NaCl solution conditioned by CaO 12.51, 12.00, 11.01

16.5 g/L NaCl solution conditioned by H2SO4 5.55, 2.99

33 g/L NaCl solution conditioned by CaO 12.51, 12.00, 11.05

33 g/L NaCl solution conditioned by H2SO4 5.47, 2.96




3.1 Synthesis and characteristics of scorodite particles

Figure 2 presents SEM images of the scorodite precipitate. Large scorodite crystals with low mois-

ture content (11.2 %) were formed, yielding a high arsenic precipitation percentage. The final pH

and ORP of the synthesis solution were -0.22 and 474 mV (vs Ag/AgCl), respectively.

Tables 2 and 3 summarize the results of scorodite leach tests with the different extraction buffers.

The pH values given in the tables were taken immediately after shaking was complete (before solid/

liquid filtration and separation). These results show that the scorodite crystals were quite stable,

with arsenic leaching less than 0.5 mg/L in the pH range 3 to 5, though arsenic leached at a concen-

tration of 3 mg/L at pH 6.9, and at a concentration of 300 mg/L at pH 9. The value of 300 mg/L in-

dicates that about 2 % of the arsenic was released from the scorodite.

Figure 2: SEM images of precipitated samples

30µm 6µm



Table 2: Long-term scorodite leach tests under a normal air atmosphere using buffer solutions

with pH values of 2.88 and 4.93

Stirring

days

0.1 mol/L CH3COOH 0.1 mol/L CH3COOH and NaOH

pH

As Fe S

pH

As Fe S

mg/L mg/L mg/L mg/L mg/L mg/L

0.25 2.88 0.19 0.70 <10 4.95 0.16 0.06 <10

1 2.90 0.24 0.90 <10 4.95 0.16 0.06 <10

2 2.90 0.28 0.93 <10 4.96 0.25 0.07 <10

3 2.90 0.38 1.09 <10 4.95 0.26 0.06 <10

5 2.91 0.32 1.04 <10 4.95 0.20 0.07 <10

7 2.88 0.31 1.03 <10 4.94 0.23 0.07 <10

14 2.90 0.34 1.14 <10 4.96 0.19 0.09 <10

21 2.91 0.50 1.33 <10 4.96 0.18 0.07 <10

28 2.90 0.43 1.39 <10 4.95 0.20 0.12 <10

35 2.89 0.40 1.23 <10 4.95 0.18 0.09 <10

Table 3: Long-term scorodite leach tests under a normal air atmosphere using buffer solutions

with pH values of 6.86 and 9.18

Stirring

days

0.025 mol/L KH2PO4 and NaH2PO4 0.01 mol/L Na2B4O7

pH As Fe S

pH As Fe S


0.25 6.92 1.05 <0.04 <10 9.26 4.18 0.60 <10

1 6.92 1.50 0.06 <10 9.20 32.3 9.11 <10

2 6.91 1.53 0.06 <10 9.19 40.3 13.3 <10

3 6.91 1.88 0.07 <10 9.15 71.3 28.1 <10

5 6.91 2.13 0.12 <10 9.04 137.0 54.9 <10

7 6.89 2.07 0.09 <10 9.08 117.0 42.5 <10

14 6.91 2.49 0.13 <10 8.84 253.0 15.6 10

21 6.91 3.00 0.11 <10 8.75 321.7 24.4 <10

28 6.90 2.77 0.17 <10 8.69 312.0 26.6 10

35 6.89 3.14 0.15 <10 8.66 351.7 26.3 <10

Tables 4 and 5 show the results of 6-hour scorodite leach tests conducted using various solvents.

The leach tests using buffer solutions with pH values of 2.88 and 4.93 were performed again in this

series of experiments and the results differ from the corresponding data reported above. Table 5

gives the results of leach tests that involved the presence of a given amount of NaCl. In this experi-

ment, CaO or H2SO4 was used to adjust the pH.



When Mg(OH)2 or CaO was used, the leached arsenic concentration was as low as approximately

3 mg/L under strongly basic conditions (pH 10 or higher), whereas arsenic leached at the higher

concentration of 10 mg/L in the pH range 8-10. The arsenic leaching concentration under alkaline

conditions using NaOH was as high as 99.2 mg/L. The liquid-solid ratio was 10:1 in this leach test

and it was calculated that about 0.3 % of the arsenic was released from the scorodite.

When HCl, H2SO4 or HNO3 was employed, the arsenic leaching concentration was less than 1 mg/L

at pH 3, and 2 mg/L at pH 2. However, when the pH was reduced from 1 to 0, the concentration of

leached arsenic increased considerably. The sequence of acids arranged in decreasing order of arsen-

ic leaching was HNO3, H2SO4 and HCl.

For leach tests conducted in the presence of NaCl, the arsenic leaching concentration was 0.3 mg/L

or less under acidic to neutral pH conditions, though it exceeded 1000 mg/L (dissolved As ratio:

approximately 3 %) under basic conditions.

Table 4: 6-hour scorodite leach tests with various solvents

Solvent Initial

pH

Final

pH

Leachate

Solvent Initial

pH

Final

pH

Leachate

As Fe S As Fe S


Water 5.74 4.61 0.01 0.56 <10 H2SO4 3.96 3.99 0.05 0.91 <10

TCLP 2.88 3.00 3.02 0.11 1.48 <10 H2SO4 2.97 3.03 0.21 1.37 20

TCLP 4.93 4.88 4.89 0.16 0.21 <10 H2SO4 1.97 1.99 1.83 2.87 210

CaO 12.52 12.17 2.55 0.09 <10 H2SO4 1.22 1.28 11.00 9.84 1630

CaO 12.30 12.10 2.20 0.11 <10 H2SO4 0.35 0.42 94.74 78.04 15130

CaO 10.75 8.36 3.91 0.54 <10 HCl 3.99 3.98 0.04 0.97 <10

CaO 10.14 6.83 0.51 0.27 <10 HCl 2.99 3.04 0.13 1.68 <10

CaO 9.00 5.14 <0.01 0.24 <10 HCl 1.98 2.03 1.06 2.94 <10

Mg(OH)2 10.48 10.03 1.66 0.05 <10 HCl 1.11 1.12 6.70 7.73 <10

Mg(OH)2 10.41 9.88 8.59 <0.05 <10 HCl 0.13 0.15 202 166 <10

Mg(OH)2 10.64 9.26 6.66 0.12 <10 HNO3 3.98 3.99 0.05 0.98 <10

Mg(OH)2 9.00 7.58 1.27 0.60 <10 HNO3 2.99 3.01 0.08 1.72 <10

NaOH 12.50 9.46 99.20 60.50 30 HNO3 1.99 2.04 0.78 2.90 <10

NaOH 11.06 8.28 21.20 5.43 <10 HNO3 1.12 1.12 3.89 5.93 <10

NaOH 9.70 5.83 0.09 0.05 <10 HNO3 0.12 0.16 29.34 26.11 <10



Table 5: 6-hour scorodite leach tests with various solvents in the presence of NaCl

Solvent Initial pH Final pH

Leachate

As Fe S

mg/L mg/L mg/L

NaCl=8.2g/L † 12.51 10.47 927 29.9 30

NaCl=8.2g/L † 12.00 9.55 258 26 <10

NaCl=8.2g/L † 11.06 8.46 25.80 0.07 <10

NaCl=8.2g/L * 5.80 4.61 0.04 0.58 <10

NaCl=8.2g/L * 2.99 3.05 0.18 1.54 20

NaCl=16.5g/L † 12.51 10.52 1041 26.7 40

NaCl=16.5g/L † 12.00 9.71 278 2.76 10

NaCl=16.5g/L † 11.01 8.32 22.05 <0.05 <10

NaCl=16.5g/L * 5.55 4.49 <0.01 0.66 <10

NaCl=16.5g/L * 2.99 3.05 0.10 1.55 20

NaCl=33g/L † 12.51 10.54 1164 21.9 50

NaCl=33g/L † 12.00 9.85 309 2.54 10

NaCl=33g/L † 11.05 8.36 26.63 <0.05 <10

NaCl=33g/L * 5.47 4.42 <0.01 0.68 <10

NaCl=33g/L * 2.96 3.03 0.30 1.61 20

†: CaO was used for adjustment of pH.

*: H2SO4 was used for adjustment of pH.

The results of leach tests in Tables 2-5 indicate that the solubility of scorodite depends strongly on

pH. The experimental results indicate that high pH values increase the concentration of arsenic ac-

cording to the following reaction;

FeAsO4·2H2O + H2O = Fe(OH)3 + HAsO42-

+ 2H+ (1)

The above results (Tables 4 and 5) indicate that scorodite leached at only very low concentrations in

the pH range 3-6. The data also show that scorodite leached at higher concentrations at pH values

below 3, and that it leached at high concentrations when the solvent was made alkaline by adding

NaOH. Considering that under neutral to acidic pH conditions only low concentrations of arsenic

leached into solvents with high NaCl concentrations, we may conclude that alkaline pH conditions,

not Na+ ions, contributed to the enhanced arsenic leaching.

Low concentrations of scorodite leached into solvents containing calcium and magnesium ions. This

finding is consistent with the solubility products reported by Nishimura and Tozawa [28] for cal-

cium arsenate and magnesium arsenate. One point worth noting is that the above study by Nishimu-

ra and Tozawa found that no calcium arsenate leached at pH 8-10, resulting in an apparent discre-

pancy with the results reported here. The discrepancy arises because the leach test in their study

measured the leaching of scorodite into a calcium-containing aqueous solution. It is presumed that



in the test for leaching into the calcium-containing alkaline solution reported in this article, scoro-

dite leached a low level of arsenate, which then reacted with calcium ions, present in abundance to

form calcium arsenate, resulting in a very low arsenic concentration. Similar reaction schemes prob-

ably dominated in the case of magnesium.

It was observed that addition of NaCl to the Ca-containing alkaline solution promoted leaching of

arsenic from scorodite at a level similar to or higher than that of the NaOH-containing solution. This

observation indicates that the effect of lime on the suppression of arsenic leaching is cancelled out

by the presence of NaCl. Sodium ions may react with arsenic ions to form sodium biarsenate, how-

ever the sodium biarsenate will dissolve easily in an alkaline solution. The sodium ions probably

enhanced the dissolution of arsenic from scorodite, with the chlorine ions present also being able to

react with the scorodite ferric ions.

The combination of NaCl and CaO may constitute the composition of fly ash from blast furnaces

recently developed at industrial waste treatment sites. The results of leach testing in this research

demonstrate the absolute necessity of separating the storage locations of fly ashes from blast fur-

naces for scorodite precipitates because these materials, when dumped together, constitute a high

risk of arsenic leaching.

4 Conclusions

Stable scorodite (FeAsO4·2H2O) particles were produced by introducing oxidizing gas into a reac-

tion mixture containing ferrous sulfate and a high concentration of arsenic (V) to convert ferrous

ions to ferric ions. The solubility of scorodite synthesized by this novel atmospheric process was

investigated. The long-term solubility of the new scorodite product was the same as previously re-

ported results. In addition, it was shown that scorodite solubility was significantly influenced by

surrounding environmental factors such as pH rather than by long-term stability.

The results can be summarized as follows:

(1) When Mg(OH)2 or CaO was used, the leached arsenic concentration was as low as approxi-

mately 3 mg/L under strongly basic conditions (pH 10 or higher), whereas arsenic leached at the

higher concentration of 10 mg/L in the pH range 8-10.

(2) The arsenic leaching concentration under alkaline conditions using NaOH was as high as

99.2 mg/L.

(3) When HCl, H2SO4 or HNO3 was employed, the arsenic leaching concentration was less than

1 mg/L at pH 3, and less than 2 mg/L at pH 2. However, when the pH was reduced from 1 to 0,

the concentration of leached arsenic increased considerably. The sequence of acids, arranged in

order of decreasing effect on arsenic leaching, was HNO3, H2SO4, and HCl.

(4) For leach tests conducted in the presence of NaCl, the arsenic leaching concentration was

0.3 mg/L or less under acidic to neutral pH conditions, though it exceeded 1000 mg/L under

basic conditions.



References

[1] WILKIE J.A., HERING J.G. (1996): Colloids Surf., A 107: 97–110

[2] MANNING B.A., GOLDBERG S. (1996): Soil Sci. Soc. Am. J., 60: 121–131

[3] JAIN A., RAVEN K.P., LOEPPERT R.H. (1999): Environ. Sci. Technol., 33: 1179–1184

[4] MENG X., BANG S., KORFIATIS G.P. (2000): Water Res., 34: 1255–1261

[5] GAO Y., MUCCI A. (2001): Geochimica et Cosmochimica Acta, 65: 2361–2378

[6] DIXIT S., HERING J.G. (2003): Environ. Sci. Technol., 37: 4182–4189

[7] KATSOYIANNIS I., ZOUBOULIS A., ALTHOFF H., BARTEL H. (2002): Chemosphere, 47: 325–332

[8] AHN J.S., CHON C.M., MOON H.S., KIM K.W. (2003): Water Res., 37:2478–2488

[9] FUKUSHI K., SATO T., YANASE N. (2003): Environ. Sci. Technol., 37: 3581–3586

[10] MUNOZ J.A., GONZALO A., VALIENTE M. (2002): Environ. Sci. Technol., 36: 3405–3411

[11] GIASUDDIN A.B.M., KANEL S.R., CHOI H. (2007): Environ. Sci. Technol., 41: 2022–2027

[12] JANG J.H., DEMPSEY B.A. (2008): Environ. Sci. Technol., 42: 2893–2898

[13] MERCER K.L., TOBIASON J.E. (2008): Environ. Sci. Technol., 42: 3797–3802

[14] SARKAR A., BLANEY L.M., GUPTA A., GHOSH D., SENGUPTA A.K. (2008): Environ. Sci.

Technol., 424: 268–4273

[15] HINGSTON F.J., ATKINSON R.J., POSNER A.M., QUIRK J.P. (1968): 9th Int. Congr. Soil

Science Transactions (Adelaide, Australia), Vol. I: 669–678

[16] HINGSTON F.J., POSNER A.M., QUIRK J.P. (1971): Discussions of the Faraday Society,

No.52: 334–342

[17] PIERCE M.L., MOORE C.B. (1982): Water Res., 16: 1247–1253

[18] FUJITA T., TAGUCHI R., ABUMIYA M., MATSUMOTO M., SHIBATA E., NAKAMURA T.

(2008): Hydrometallurgy, 90: 92–102






(2009): Hydrometallurgy, 96: 189-198

[22] FUJITA T., TAGUCHI R., MATSUMOTO M., SHIBATA E., NAKAMURA T. (2009): Hydrometal-

lurgy, 96: 300-312

[23] NISHIMURA T., TOZAWA K. (1978): Reprint from the Bull. The Research Institute of Mineral

Dressing and Metallurgy, Tohoku University, 34: 19–26



[24] NISHIMURA T., TOZAWA K. (1984): J. Min. Metall. Inst. Japan, 100: 1138–1144

[25] ROBINS R.G. (1980): 4th

Joint Meeting of MMIJ-AIME, Technical Session D-1, (Tokyo, Ja-

pan), 25–44

[26] ROBINS R.G. (1981): Metall. Trans. B 12B: 103–109

[27] ROBINS R.G. (1983): 3rd

Int. Symp. on Hydrometallurgy, Hydrometallurgy Research, Devel-

opment and Plant Practice, Eds. K. Osseo-Asare and J. D. Miller, (The Metallurgical Society

of AIME, Georgia, Atlanta), 291–310

[28] ROBINS R.G. (1987): Am. Mineralogist, 72: 842–844

[29] ROBINS R.G. (1990): EPD Congress ’90, Ed. D. R. Gaskell, (The Minerals, Metals & Mate-

rials Society, Anaheim, California), 93–104

[30] KRAUSE E., ETTEL V.A. (1985): Proc. 15th

Annual CIM Hydrometallurgy Meeting, (CIM,

Vancouver, Canada), 5/1–5/20

[31] KRAUSE E., ETTEL V.A. (1988): Am. Mineralogist, 73: 850–854

[32] KRAUSE E., ETTEL V.A. (1989): Hydrometallurgy, 22: 311–337

[33] HARRIS G.B., MONETTE S. (1989): Arsenic Metallurgy Fundamentals and Applications, Eds.

R. G. Reddy, J. L. Hendrix and P. B. Queneau, (The Metallurgical Society of AIME, Phoenix,

Arizona), 469–488

[34] HARRIS G.B., MONETTE S. (1989): Productivity and Technology in the Metallurgical Indus-

tries, Eds. M. Koch and J. C. Taylor, (The Minerals, Metals & Materials Society, Cologne,

Germany), 545–559

[35] EMETT M.T., KHOE G.H. (1994): EPD Congress 1994, Ed. G. W. Warren, (The Minerals,

Metals & Materials Society, San Francisco, California), 153–166

[36] KHOE G., CARTER M., EMETT M., VANCE E.R., ZAW M. (1994): 6th AusIMM Extractive Metal-

lurgy Conference, Ed. G. W. Warren, (AusIMM publications, Brisbane, Queensland), 281–286

[37] LANGMUIR D., MAHONEY J., MACDONALD A., ROWSON J. (1999): Geochimica et Cosmo-

chimica Acta, 63: 3379–3394

[38] LANGMUIR D., MAHONEY J., ROWSON J. (2006): Geochimica et Cosmochimica Acta, 70:

2942–2956

[39] BLUTEAU M.C., DEMOPOULOS G.P. (2004): Waste Processing and Recycling in Mineral and

Metallurgical Industries V Fifth Int. Symp., Hamilton, Ontario, Eds. S. R. Rao, F. W. Harri-

son, J. A. Kozinski, L. M. Amaratunga, T. C. Cheng and G. G. Richards, (CIM, Montreal,

Canada), 439–451

[40] BLUTEAU M.C., DEMOPOULOS G.P. (2007): Hydrometallurgy, 87: 163–177


Biosolubilization of Copper from

Waste Electric Cables

Asst. Prof. S. Gaydardzhiev, D. Bastin, Eng. F. Goffinet Dr. P. F. Bareel

University of Liege Comet Traitements S. A.

Mineral Processing and Recycling – GeMME

B 52, Sart Tilman Rivage de Boubier 25

4000 Liege, Belgium 6200 Châtelet, Belgium

Keywords: Copper, recycling, technogenic products, bio leaching

Abstract

The paper reports on results from a laboratory tests for bacterial leaching of copper from scrap ca-

bles. The studied material is a reject fraction obtained after dismantling and separation of electric

cables during recycling of end-of-life vehicles (ELV). The copper has been met predominately in

pure metallic form as tiny irregular shaped wires often coated with tin and well liberated from the

plastic isolations. For bringing copper in solution, a bacterially assisted agitative leaching with

mixed consortium of mesophylic microorganisms has been chosen. Continuous bacterial adaptation

of the cultures to the substrate has been envisaged in order to provide an efficient way for ferrous

iron regeneration during the leaching. It has been established that under optimal conditions of pH,

density and temperature it is possible to recover nearly the total copper within short leach duration.

The obtained pregnant leach solution could be subjected to subsequent copper recovery via solvent

extraction, while the solid leached residue could be considered as non-metallic material containing

plastics suitable for recycling.

1 Introduction

Technogenic wastes derived from recycling of End-off-Life Vehicles (ELV) are viewed as growing

stream for use as a secondary resource base. Recent statistic estimates the world automotive park as

nearly 900 million cars, a figure which grows annually at constant rate of 2.25 % [1]. Moreover, due

to the forthcoming implementation of the new EU regulation in 2015 stipulating automotive indus-

try to recycle cars up to 95 % by weight, there is considerable interest in developing post-shredder

technologies capable to increase the recycling rate for both nonmetallic and metallic vehicle com-

ponents. During ELV recycling, cars are dismantled and depolluted and delivered to shredding and

further to post-shredder treatment where various material streams are formed. The separation flow

Gaydardzhiev, Bastin, Bareel, Goffinet


sheets used often involve combination of wind sichters, screens, magnetic and eddy-current separa-

tors. The dismantled electric and electronic equipment like board computers, CD players, navigation

units, electric cables etc., are processed independently in order to concentrate targeted components

like plastics, non-ferrous metals, and glass. The material stream in which electric cable scraps are

predominately concentrated as a rule is not suitable for land-filling or as an additive for the civil

engineering sector due to the elevated content of metals. Therefore, quite often a dedicated mechan-

ical post-processing involving grinding by knife mills and gravity separation (shaking tables, zigzag

sichters, pneumatic separators, etc.) is envisaged. The purpose is to obtain enriched in copper frac-

tions acceptable by the refineries as by-product to usual feeds. However, the heterogeneity of the

material and the irregular shape of wires and plastics render the gravimetric separation difficult

which reflects in non-negligible losses of metal values with the “tailing” fractions. This was the case

of the present study dealing with material presenting a mixture between the light fraction from spiral

separation and reject fraction after separation in pneumatic table of ground scrap cables. Hydrome-

tallurgical treatment could recover the remaining copper from this material, provided the metal sur-

face is accessible to aqueous leach solutions. Although not completely separated from their plastic

isolation, the cables are relatively finely ground and the isolation loosened during fragmentation,

therefore the wire surfaces could be deemed accessible to lixiviants.

The use of microorganisms to bring metals from wastes into solution could offer a low-cost alterna-

tive to the classical hydrometallurgical processes. Bio assisted solubilization can be utilized to oxi-

dize zero valent copper through leaching by Fe(III) and sulphuric acid, where Fe(II) oxidation is

carried out by the microorganisms. Several studies have been performed recently mainly for recov-

ery metals from PCB’s (printed circuit boards) with various technological aspects being discussed

by Brandl et al. [2], Ilyas et al. [3], Yang et al. [4], Pham et al. [5]. Bioleaching has been also inves-

tigated by Misra et al. [6] for recovery of lithium from spent secondary batteries. Biotechnological

recovery of non-ferrous metals from scrap cables has not been yet reported. Therefore, the present

work has an objective to evaluate the feasibility of this option and to investigate certain process re-

levant parameters such as pulp density, pH and effect of culture adaptation on the material upon the

degree of metals leaching. The bacterial adaptation to the substrate as suggested by Elzeky & Atia

[7] is an important parameter during bioleaching of mineral ores and it is expected this factor to play

an essential role due to the quite specific nature of the cable scraps.


2.1 Culture

A mixed consortium of acidophilic bacteria from the genus Acidithiobacillus ferrooxidans, Acidi-

thiobacillus thiooxidans, and Leptospirillum ferrooxydans supplied by the Department of

Geoecololgy, University of Mining and Geology in Sofia, Bulgaria has been used. Bacteria were

incubated under non-sterile conditions at 35 °C in 250 mL Erlenmayer flasks on orbital shaker using

Biosolubilization of Copper from Waste Electric Cables


an iron containing 9K medium prepared as described in [8]. Standard shake flask serial transfer was

used for adapting the cultures to the substrate. After three serial transfers, inoculation of a 2-liter

fermentor has taken place by transferring the cultures having reached their logarithmic growth

phase.

2.2 Analysis

Copper, lead, zinc and iron in the leach solutions have been analyzed by atomic adsorption spec-

trometer (AAS Perkin Elmer) following filtration of the samples. It has been tried to determine fer-

rous iron by titration with K2Cr2O7, but the strong interference effect from the bivalent copper pre-

sent in the solution has rendered this method non applicable. Sulphuric acid and sodium hydroxide

of analytical grade (Merck AG, Darmstadt) have been used to maintain the desired pH during leach-

ing. Particle size analysis of the sample has been realised by two alternative methods: by using a

Zephyr image analysis system (Ochio, Belgium) and through dry sieve analysis.

2.3 Scrap cables

The as received from the car recycling company sample of scrap cables was subjected to characteri-

sation and leaching without pre-treatment. Its particle size distribution obtained through both image

analysis and dry sieving is presented in Figure 1. A quite good coincidence of particle size distribu-

tion curves obtained by the two methods could be noted. The image analysis system determines the

particle size while capturing objects in motion and the implemented algorithm yields the weighted

sieve diameter. The mean diameter of the material (d50) as well as d10 and d90 as given by the image

analysis system are derived as 1.27 mm, 0.43 and 2.03 mm respectively.



Figure 1: Particle size distribution of the waste cables as given by sieve and image analysis

The content of the principal metals of interest in the scrap cables and their repartition in the differ-

ent granulometric classes is given in Table 1. It should be noted that the principal metal of interest is

copper which is concentrated in majority in the mesh size < 500 µm. However, the mass yield of

this class has been found less than 10 %. For quantitative determination of other elements which

could not be detected by the AAS, the sample has been subjected to XRF analysis. The analysis

gave the following elements in respective concentrations [%]: Ca – 6.58; Sn – 0.03; Al – 1.09;

Cl – 2.47; F – 1.09; Mg – 0.21; O – 8.07; P – 0.02; Ni – 0.04.

Figure 2 presents a general view of the material under study. The immediate impression is that the

predominant part of the wires is well liberated and metallic copper could be well distinguished es-

pecially on the background of the black foamy-look composites. Visible are several pieces of wires

coated with tin and pieces of brass as well. It should be noted that PVC and polyethylene are the

plastics most often met in the automotive cables. Other type of plastics such as EFTE, PET and PU

are less common. The issue of eventual dissolution of plastics during the non-ferrous metals leach-

ing should be addressed in view assessing the level of bacterial tolerance towards chlorinated

products.



Table 1: Concentration of main metals of interest and their granulometric repartition

Screen opening [mm] Cu [%] Fe [%] Zn [%] Pb [%]

+ 2

- 2 + 1

- 1 + 0.5

- 0.5 + 0.3

- 0.3 + 0.15

- 0.15

0.87

2.15

9.96

43.18

60.10

37.10

0.11

0.11

0.19

0.36

0.43

0.97

0.03

0.08

0.25

0.11

0.09

0.24

0.20

0.14

0.15

0.18

0.17

0.15

Input material 9.83 0.17 0.12 0.15

Figure 2: View of the as received scrap cables



2.4 Leaching procedure

Two different modes of agitative leaching have been tested. The first one, aimed for evaluating the

material aptitude to leaching and for adapting the bacterial culture to the substrate, has taken place

inside 250 mL Erlenmayer flasks on an orbital shaker. The second mode leaching performed at in-

creased pulp density, has been realised in a 1-liter double wall reactor equipped with a glass stirrer

and agitator run at 800 rpm (Janke & Kunkel RW20). In both cases temperature has been kept

around 35 °C, either by placing the shaker inside “hot” room or by heating the reactor with thermo-

static bath. The leach solution has been produced by fermentor operating in batch mode and inocu-

lated with the culture under consideration. The required volume of leach solution has been periodi-

cally drained and the fermentor refilled with a fresh 9K nutrient medium. At the beginning of the

leaching experiment, 1 litre solution produced by the fermentor has been placed inside the leaching

reactor and predetermined amount of cables added. The solution containing the biomass was charac-

terized by redox potential of minimum 690 mV and concentration of ferric iron about 8 g/L, sug-

gesting that bacteria have reached their logarithmic phase of growth and that almost complete oxi-

dation of the ferrous to ferric iron by the bacteria has taken place. The pH set value for the leaching

was 1.9, chosen in view optimal bacterial growth and for prevention of excessive built up of ferrous

hydroxides which could lead to copper losses. The adjustment of the pH to the set value has been

realised either through periodical check and correction or via continuous pH control. At the former

case, pH has been measured in a given interval and when values above 1.9 registered, sulphuric acid

(10N) added to adjust pH back to 1.9. At the latter case, a pH electrode placed in the agitation vessel

and micro dosing pump have been connected to a control system (Consort R305) operating in pH-

stat mode. The Eh value (vs. SHE) has been measured as well and registered on-line by a PC to-

gether with the pH. Bacterial leaching has been monitored through periodical sampling of the pulp

and every time sample was taken, an equivalent amount of distilled water has been added. After

bioleaching, the pulp has been subjected to solid liquid separation using a filter paper and the re-

maining solid analysed for mass balance calculation.

3 Results and discussions

3.1 Leaching in Erlenmayer flasks at low pulp density with culture

non-adapted to the scrap cables

The initial bioleaching tests have been done with the aim to adapt the bacterial cultures to the sub-

strate and to evaluate the acid consumption required for keeping pH below 1.9. 10 grams of cable

scrap material have been contacted with 150 ml bacterial solution in Erlenmayer flasks. It should be

noted that for these tests, the bacterial culture has been already adapted during previous studies to

copper concentration of 4 g/L coming from metal bearing substrate different than the cables. The

results regarding pH and Eh evolution as well as the acid consumption for keeping pH below 1.9 are



presented at Table 2, while Figure 3 shows the leaching kinetics for copper, zinc and lead. The

leaching has been terminated after 48 hours based on visual observation of pulp coloration and on

disappearance of copper wires. Iron concentration in the solution has not been measured since it has

been assumed that due to pH variation part of the iron has precipitated.

Table 2: Variation in pH and Eh during leaching and amount of sulphuric acid (10N) totally

consumed

Time [h] Eh [mV] pH measured pH adjusted H2SO4 added cumulatively [mL]

0 690 2.04 - -

1 512 2.30 1.84 0.47

3 414 2.40 1.88 0.87

6 386 2.68 1.80 1.27

27 430 2.36 1.78 1.67

28 440 2.35 1.77 2.07

30 458 2.09 1.84 2.30

48 558 2.18 1.78 2.50

It could be seeen that after approximately 40 hours, following an innitial drop, the redox potential

potential begins to rise which well corraborates with the end of copper leaching and indicates the

emerging of regenerated from the bateria ferric iron. As shown in Table 2, the pH has to be adjusted

back all along during the leaching.This indicates that the material is alkaline in nature and consumes

acid.

Figure 3: Leaching rate of copper, zinc and lead. Conditions: 6 % (w/v) pulp density, Erlenmayer

flasks, 35 °C, Cu in g/L, Pb and Zn in mg/L

0

1

2

3

4

5

6

7

0 10 20 30 40 50 60

Leaching time (h)

Co

nce

ntr

ati

on

of

Cu

(g

/L

)

0

2

4

6

8

10

12

14

16

18

20

Co

nce

ntr

ati

on

o

f Z

n a

nd

Pb

(m

g/

L)

Cu

Zn

Pb



The results shown in Figure 3 suggest that it took less than 48 hours to bring copper in solution with

recovery of about 99 %. Zinc is mobilized faster than copper, while lead is not leached at all and

remains in the solid phase. The concentration of zinc in the pregnant leach solution is very low

which is logicall owing to its low innitial content. On comparison of the results from Figure 3 and

Table 2 it could be noticed that copper leaching does not follow the redox potential trend. Therefore

the sulphuric acid added for keeping the pH below 1.9 (totally 2.5 mL) eventually contributes also

to leach copper. The rise in the redox potential at the end of the test shows that the bacterial culture

even not adapted to the same substrat has been able to regenarate ferrous to ferric iron. The plastic

materials ussually contain calcium carbonate as filler (in our case about 16 % as given by the XRF

analysis), which explains the alkaline nature of the cables. Consequently it has been important to

delineate the amount of acid consumed by the plastic and the amount of acid responsible for

mobilizing copper. To evaluate the degree of acid consumption, the material has been brought in

contact with sulfuric acid under agitative mode with pH being kept between 1.8 and 2. The test was

terminated after 50 hours when pH has stabilized and sulphuric acid total consumption has been

calculated as 0.12 g/g material. During this leaching test, characterized by absence of

microorganisms and ferric iron, only 2 % of copper has been brought in solution. This fact idicates

that the presence of bacteria and ferric iron do enable and do accelerate mobilization of copper. The

XRD analysis of the leached residue is presented in Figure 4. The characteristics pics are

suggesting that the predominate compound of the precipitate is synthetic gypsum, resulting from the

reaction between the sulphuric acid and the calcium carbonate.

Figure 4: XRD spectra of the residue after acid leaching



3.2 Leaching at controlled pH with adapted bacterial culture

The test realised in agitative vessel at density of 12 % (w/v) has been performed to study the effect

of the continuous pH control on the degree of copper mobilization. At this instance, the leach sus-

pension has contained an adapted to the cables bacterial culture. Figure 5 portrays the leaching ki-

netics for copper, zinc and lead and the total iron concentration, while Figure 6 gives the variations

in pH and Eh during the leaching.

Figure 5: Leaching rate of copper, zinc and lead. Conditions: 12 % (w/v) pulp density, agitative

reactor, 35 °C, Cu in g/L, Pb and Zn in mg/L

From the results it could be depicted that likewise in the Erlenmayer tests, zinc is leached rapidly at

the very beginning. The total iron concentration is stable at the first 20 hours after which begins to

decrease. The reduction of iron in the process solution from 7.3 to 6.2 g/L after the first 20 hours up

to the end of copper leaching reaction (hour 80) accounted for around 75 % copper dissolution. Fe

concentration decreased slowly and steadily as the leaching proceeded and iron concentration

reached about 5 g/L at the end. Possibly, regardless the continuous pH control, some part of the iron

is possibly precipitated as jarosite. The Eh curve in Figure 6 indicates a rapid sink of the redox po-

tential at the beginning following from a prolonged period of stabilization at around 400 mV. Al-

though copper solubilization comes to the end after 80 hours, the potential remains stable at about

400 mV until the hour 100, after which it rises rapidly. The dissolution of copper has been increas-

ing steadily indicating that bacteria are well adapted to the light jarosite media and continue to rege-

nerate the iron regardless its slightly diminishing concentration. The amount of consumed acid per

gram material has been the same as the one estimated in the Erlenmayer leaching tests. The slight



decrease in zinc and lead concentration at the end of the leaching cycle could be due to sorption

effects caused by the formed iron precipitates.

Figure 6: Evolution of Eh and pH during leaching. Conditions: 12 % (w/v) pulp density, agitative

reactor, 35 °C

4 Conclusions

The study although preliminary, allows drawing certain conclusions regarding the approach feasibil-

ity. The experiments of the leaching of scrap cables have indicated that metallic copper could be

recovered effectively by bio-hydrometallurgical process.

It has taken about 40 hours to leach the total copper from a suspension with 6 % density (w/v) using

a bacterial culture adapted to other substrate. Copper concentration has reached 6.3 g/L. The

adapted to the material bacterial culture has recovered about 100 % of the copper in 80 hours

(3.5 days) from a suspension at 12 w/v density and have brought copper concentration of 12.8 g/L in

the PLS. Hence increasing twice the pulp density has doubled the leaching time required to reach

same recovery, with copper concentration in PLS increasing also twice. This correlation shows that

there are no inhibition effects from the cable material on the bacterial culture, which is already resis-

tive following its initial adaptation to other metal bearing substrate. Slight variation in pulp pH in

certain margins does not influence leaching efficiency.

Neither the plastics nor the presence of tin have affected the bacterial activity and their ability to

regenerate iron. Ferric iron seems the main factor for copper mobilization, but iron speciation

method should be found in this particular case.

The final copper concentration makes the PLS suitable for further treatment via solvent extraction

for copper extraction with the additional advantage that the solution does not contain zinc.



References

[1] WINFIELD, P., HUTCHINSON, A. & PEMBERTON, M., (2007): Whole life vehicle waste

streams, Oxford Brookes University.

[2] ILYAS, S. et al., (2007): Bioleaching of Metals from Electronic Scrap by Moderately Thermo-

philic Acidophilic Bacteria, Hydrometallugy, 88, 180-188

[3] BRANDL, H., BOSHARD, R. & WEGMANN, M. (2001): Computer-munching Microbes:

Metal Leaching from Electronic Scrap by Bacteria and Fungi, Hydrometallugy, 59, 319-326

[4] PHAM, V. & TING, Y. (2009): Gold Bioleaching of Electronic Waste by Cyanogenic Bacteria

and its Enhancement with Bio-oxidation, Advanced material research, 71-73, 661-664

[5] YANG, T. et al., (2009): Factors Influencing Bioleaching Copper from Waste Printed Circuit

Boards by Acidithiobacillus ferrooxidans, Hydrometallugy, 97, 29-32

[6] MISHRA, D. et al., (2008): Bioleaching of Metals from Spent Lithium Ion secondary Batteries

using Acidithiobacillus ferrooxidans, Waste Management, 28, 333-338

[7] ELZEKY, M. & ATTIA, Y., (1995): Effect of Bacterial Adaptation on Kinetics and Mechanism

of Bioleaching Ferrous Sulfides, Chemical Engineering Journal, 56, B115-B124

[8] SILVERMAN, M. & LUNDGREN, D. (1959): Studies on the chemoautotrophic iron bacterium

Ferrobacillus ferrooxidans. I. An improved Medium and a Harvesting Procedure for securing

High Cell Yields, Journal of Bacteriololgy, 77, 642–647



Dowa Mining Scorodite Process® – Application

to Copper Hydrometallurgy

H. Kubo, M. Abumiya, M. Matsumoto

Dowa Metals & Mining Co. Ltd.

217-9 Shimo-kawabata Furumichi Iijima

Akita, Japan

Keywords: Scorodite, arsenic, ferrous oxidation, ferric arsenate, crystalline

Abstract

Dowa Metals & Mining Co. Ltd. has developed a new method for arsenic immobilization. Specifi-

cally, arsenic fixation is achieved by synthesizing crystalline scorodite (FeAsO4·2H2O) under at-

mospheric conditions.

Since June 2008, a plant handling 30 MT/month of arsenic has been operating at the Kosaka

Smelter to produce crystalline scorodite from non-ferrous intermediate materials. The results from

this plant’s operation are reported here.

The process utilized by this plant consists primarily of three steps: 1) the process of leaching arsenic

from metallurgical intermediate materials containing arsenic; 2) the oxidation of As(III) to As(V) in

the leaching solution; 3) the crystallization of scorodite from the solution at 95 °C under atmos-

pheric conditions.

This plant can use two types of intermediate products as the starting material. The first is arsenic

sulfide precipitated from the wastewater from sulfuric acid production. The other is liberation slime

from copper electrolysis.

The scorodite formed in this plant has good crystallinity, large particle size (>20 µm on average),

high X-ray diffraction intensities, and filtering and handling attributes similar to those of laboratory-

synthesized samples. The concentration of arsenic with respect to the solubility of scorodite is low

enough to satisfy the environmental regulations of most countries.

This achievement may contribute to the arsenic immobilization methods used in the non-ferrous

smelting industry.

Kubo, Abumiya, Matsumoto


1 Introduction

Today, non-ferrous smelters are facing serious problems in handling low-grade ores and concen-

trates, which contain large amounts of various impurities. This is particularly evident in copper

smelting due to the continuously increasing trends seen in the amounts of arsenic present in concen-

trates. Hence, copper smelters are especially prone to operational difficulties. Therefore, establish-

ment of a reliable method for arsenic management is considered to be urgent.

Two types of widely known intermediate products in the copper smelting process contain large

amounts of arsenic. One is arsenic sulfide, which is precipitated from the wastewater solutions in

sulfuric acid plants, and the other is liberation slime created from copper electrolysis. These prod-

ucts, which also contain large amounts of copper, are usually recycled into the smelting process for

copper recovery. However, increasing arsenic charge from raw materials also increases the volume

of reiterative arsenic remaining in the process, and makes the operation of smelters gradually diffi-

cult. Then, the recycling of these products cannot be continued, and this finally results in an unde-

sired temporary stockpile of these intermediate products in an unstable form within the smelters’

facilities. An appropriate and reliable method of treatment for arsenic extraction and fixation from

such intermediate products is urgently needed to cope with arsenic’s rising risks and problems such

as its outflow, exposure, and space for the safe storage of such arsenic-bearing products.

Studies on the treatment of arsenic-containing non-ferrous intermediate products have been per-

formed and reviewed [1-3]. However, a process which resolves this issue satisfactorily has yet to be

developed, and many unresolved problems remain.

Dowa Metals & Mining Co. Ltd. has developed a process for treating arsenic sulfide and liberation

slime for them to function as starting materials. This process involves simultaneous leaching of ar-

senic at a high extraction ratio and copper residue recovery. The copper residue is used as a resource

for copper smelting, and the dissolved arsenic in the solution is transformed into crystalline scoro-

dite (FeAsO4·2H2O) by reaction with iron(II) ion and O2 gas in the solution [4-6]. The scorodite is

considered to be the most stable arsenic compound ever known.

Based on the achievement process described above, DOWA began experiments to verify the indus-

trial capability of a plant with a capacity of 30 MT/month of arsenic since June 2008, and has suc-

ceeded in proving that it is possible to produce crystalline scorodite industrially on a commercial

scale.

This process was named and trademarked as DMSP®

, and the achievement verified through experi-

mental operation of the DMSP®

plant is reported in this paper.

Dowa Mining Scorodite Process®

– Application to Copper Hydrometallurgy


2 About the DMSP® Plant

2.1 Plant design concept and features

Figure 1 shows the appearance of the DMSP®

plant and its equipment.

The DMSP®

plant was designed according to the following concepts and features:

• a simple process consisting of only three steps - leaching, oxidation and crystallization,

• material flexibility – the capability to treat various products with a high arsenic content, for ex-

ample sulfide, metal, and oxides, as starting materials,

• complementation – a good affinity with existing smelting processes, installable as a pre-

treatment process attachable to existing smelting facilities without any modifications, and with a

variety of possible combinations.

Figure 1: Photographs of the DMSP®

plant: external view (left) and an internal view (right)

2.2 Process overview

The DMSP®

consists of three processes. The first process is arsenic leaching and recovery of copper

residues. The second process is oxidation of As(III) in the leached solution to As(V). The third

process is crystallization, immobilizing the arsenic dissolved in the oxidized solution obtained from

the oxidation process, by forming scorodite in a solution of sulfuric acid under atmospheric condi-

tions. Figure 2 shows a flow diagram of this process.



Figure 2: DMSP®

flow diagram

2.3 Leaching process

In this process, arsenic contained in the materials is leached into solution, while copper is not

leached and remains as a residue. Two kinds of arsenic-containing products are treated as target

materials. One is arsenic sulfide obtained from sulfide treatment of the wastewater solutions from

sulfuric acid production, and the other is liberation slime created in copper electrolysis.

This process is completed by the application and development of the Kowa-process [7].

The essential reaction is represented by Equation (1).

As2S3 + 3CuSO4 + 4H2O → 3CuS + 2HAsO2 + 3H2SO4 (1)

The sulfuric acid generated by this reaction is consumed to leach copper from liberation slime.

Solid-liquid separation is carried out by pressure filtration. The obtained solution moves on to the

next oxidation process. The residue is supplied to the copper smelting process as a material.

High Arsenic Bearing Materials

(Arsenic sulfide, Liberation slime)

O₂

Leaching

Solution Residue

Oxidation Copper Smelter

Oxidized Solution

Crystallization

SCORODITE Effluent

Stockpile Solution Treatment

Oxidizing agent

FeSO₄・7H₂O

O₂-gas




2.4 Oxidation process

Before supplying arsenic to the crystallization process, the As(III) in the solution obtained through

arsenic leaching must be oxidized to As(V) by utilizing the oxidizing agent in this process, because

the arsenic must exist in the form of As(V) in order to synthesize well-formed scorodite. In this

plant, H2O2 (concentration: 33~35 %) is used as the oxidizing agent. The reaction in this process is

expressed by Equations (2) and (3).

HAsO2 + H2O2 → H3AsO4 (2)

HAsO2 + H2O2 → H2AsO4- + H

+ (3)

2.5 Crystallization process

In this process, oxygen gas is introduced into a solution containing ferrous sulfate and As(V) under

atmospheric conditions at 95 °C, thereby immobilizing the arsenic by scorodite formation. This

process is expressed by Equation (4).

2H3AsO4 + 2FeSO4 + ½O2 + 3H2O → 2FeAsO4·2H2O + 2H2SO4 (4)

After completion of this reaction, the mixture undergoes solid-liquid separation by pressure filtra-

tion and washing, and immobilized scorodite is finally obtained.

The solution after the crystallization process, which still contains some arsenic, is appropriately

treated.

3 Operation outline

3.1 Record of the amount of fixed arsenic

The DMSP®

plant was completed in June 2008.

After three months of day-time test runs with equipment checks, an experimental process in which

the successive operations were carried out with gradual increases in the quantity treated. It was con-

firmed that the plant’s operational capability was 30 MT/month of arsenic, which was the amount

initially estimated. Figure 3 shows the operation record in terms of the fixed arsenic amount from

the initial month of the operation. Dowa succeeded in satisfactorily starting up this plant from test

runs to regular operation.



Figure 3: Record of DMSP®

plant – total amount of fixed arsenic

3.2 Arsenic extraction efficiency

Figure 4 shows a summary of the record of arsenic extraction efficiency through the leaching proc-

ess. Although operational extraction efficiency became slightly lower during the initial stage, it in-

creased to around 93 % and maintained this level consistently after a leaching adjustment.

Figure 4: Summary of arsenic extraction efficiency in the DMSP®

plant

0

5

10

15

20

25

30

35

40

6 7 8 9 10 11 12 1 2 3 4 5

2008 2009

Period (year - month)

Ars

enic

MT

/ M

on

th

0%

20%

40%

60%

80%

100%

9 10 11 12 1 2 3 4 5

2008 2009

Period (year - month)

Ex

trac

tio

n (

%)




3.3 Solubility of Scorodite

Using the procedure specified by the Ministry of Environment, Notification No. 13 [8], solubility

tests were carried out on the scorodite obtained in the plant to investigate the leaching behavior of

major hazardous elements.

Although there was some fluctuation among the samples as shown in Figure 5, the scorodite was

found to be very stable, with an arsenic solubility of less than 0.3 mg/L, this meets the Japanese

standards for classification as non-hazardous waste. Table 1 shows the results of leaching tests us-

ing procedures applied in other countries [9-13]. Table 2 shows the results of leaching tests on typi-

cal samples for hazardous heavy metals other than arsenic.

Figure 5: Arsenic solubility trends for scorodite produced at the DMSP®

plant

(MOE Notification No. 13)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0 10 20 30 40 50 60

Sample No.

Lea

chat

e A

s (m

g/L

)

Regulatory threshold: 0.3mg/L

Average: 0.07mg/L



Table 1: Arsenic solubility of scorodite produced at the DMSP®

plant (other testing methods)

Sample a Sample b Sample c

Notification No. 13 (Japan) 0.06 0.07 0.03

TCLP (U.S. EPA Method 1311) 0.053 0.038 0.034

Availability Test (NEN 7341) 0.070 0.088 0.031

EP (U.S. EPA Method 1310B) 0.045 0.040 0.028

MEP (1) (U.S. EPA Method 1320) 0.31 0.24 0.24

MEP (2) 0.26 0.32 0.37

MEP (3) 0.46 0.48 0.54

MEP (4) 0.31 0.31 0.36

MEP (5) 0.37 0.33 0.37

MEP (6) 0.23 0.29 0.32

MEP (7) 0.23 0.29 0.31

MEP (8) 0.25 0.26 0.32

MEP (9) 0.28 0.51 0.33

MethodLeachate As (mg/L)

Table 2: Result of leaching test for hazardous heavy metals in scorodite produced at the DMSP®

plant

Hg Pb Cd Se Cr(VI)

0.005 0.3 0.3 0.3 1.5

<0.0005 <0.01 <0.03 <0.01 <0.02

Threshhold limit (mg/L)

Test result (mg/L)

3.4 Characterization of Scorodite

The precipitates produced at the DMSP®

plant have almost the same properties as those of scorodite

synthesized in laboratory experimentations. A chemical analysis by inductively coupled plasma

(ICP), X-ray diffraction (XRD), scanning electron microscopy (SEM), particle size analyzer, and

specific surface area (SSA) was conducted in order to determine the scorodite properties.

Table 3 shows a comparison of the iron and arsenic contents of precipitates produced at the DMSP®

plant and those of laboratory-produced samples. The XRD patterns are shown in Figure 6.

The results proved that the precipitates produced in the plant are crystalline scorodite, and that they

have the same properties as crystalline scorodite synthesized in a laboratory. The XRD patterns

were typical of those for scorodite, which indicated the absence of any other iron compounds such

as jarosite. In addition, the Fe/As molar ratio of the scorodite was almost equal to its stoichiometric

ratio.




Table 3: Chemical composition of the scorodite

As Fe Fe/As

wt % wt % molar ratio

30.9 24.1 1.05

31.5 25.3 1.08

Plant made sample

Lab. test sample

Figure 6: Comparison of XRD patterns of the scorodite

Figure 7 shows SEM images of the scorodites produced at the plant and scorodite synthesized in a

laboratory. Data of the particle size distributions of the two scorodite samples were shown in

Figure 8. Other data on the particle characteristics are summarized in Table 4. SEM images show

that scorodite produced at the plant had large, homogeneous single crystals with a diameter of about

20 µm. Particles of diameter less than 5 µm diameter were not observed, and the particles had a

typical distribution. The plant-produced scorodite had excellent filtration properties, and was con-

venient to handle on an industrial scale. The final moisture content after pressure filtration was

around 10 %.

10 15 20 25 30 35 40 45 50 55 60

2θ(deg) / (Cu-Kα)

Inte

nsi

ty （a

.u. ）

○：Scorodite

ICDD-PDF: 00-37-0468

Lab. test sample

Plant made sample



Figure 7: Comparison of SEM images of scorodite

Figure 8: Particle size distributions of scorodite

Lab. test sample Plant made sample 20µm 20µm

0%

3%

6%

9%

12%

15%

1 10 100 1000

Diameter (µm)

Fre

qu

ency

(%

)

Plant made sample

Lab. test sample




Table 4: Powder characteristics of scorodite

4 Conclusion

The DMSP®

has been developed for arsenic immobilization and copper recovery in copper smelters.

Practical experiments have been subsequently carried out at a new plant with satisfactory results,

verifying that it is capable of handling 30 As-MT/month of arsenic. The plant start-up was achieved

without difficulty, which showed the great potential of the DMSP®

and the significant technological

expertise of the DOWA group. The plant operation record was quite satisfactory. The high arsenic

extraction efficiency in the leaching process and the low solubility of arsenic from scorodite made at

the plant also proved the superiority of the DMSP®

as a solution to the above-mentioned problems.

This plant enabled the utilization of intermediate products in copper smelting, which were arsenic

sulfide and liberation slime. In the leaching process, only arsenic was leached from them and copper

residues were recovered as a recycled resource. Leached arsenic could be immobilized as crystalline

scorodite. The scorodite produced in the plant, like that made in a laboratory, was proved to be crys-

talline scorodite that was single, homogeneous and of a large particle size. Solubility of the scoro-

dite was very low, and the leached arsenic concentration completely met environmental standards.

This development of an appropriate industrial procedure for arsenic treatment and copper recovery

is of significant importance in the immobilization of hazardous substances and the efficient use of

limited natural resources. The DMSP®

, considered to be an industry-changing arsenic treatment

method with a good affinity for copper smelting processes, is expected to be further improved and

developed in the future.

Acknowledgements

For their remarkable contributions to the success in developing this process and practice experi-

ments at this plant, thanks to all of the members of KOSAKA SMELTING, DOWA HOLDINGS,

DOWA ECOSYSTEMS and DOWA METALS & MINING, who also supported the work to com-

plete this report.

Lastly, we dedicate this paper to the late Mr. Hiroshi Inoue, former President of Kosaka Smelting

Co., Ltd., who passed away on June 13, 2009.

Average particle diameter (µm)

Specific surface area (m2/g)

Moisture (Wet basis %)

Lab. test sample

23.29

0.13

8

Plant made sample

22.48

0.2

11



References

[1] WELHAM, N. J., MALATT, K. A., VUKCEVIC, S. (2000): The stability of iron phases presently used

for disposal from metallurgical systems - A review. Minerals Engineering 13 (8-9), 911-931.

[2] RIVEROS, P.A., DUTRIZAC, J.E., SPENCER, P., (2001): Arsenic disposal practices in the metal-

lurgical industry. Canadian Metallurgical Quarterly 40 (4), 395-420.

[3] HARRIS, B. (2003): The removal of arsenic from process solutions: theory and industrial prac-

tice. In: YOUNG, C., ALFANTAZI, A., ANDERSON, C., JAMES, A., DREISINGER, D., HARRIS,

B. (Eds.), Hydrometallurgy 2003 Proc. 5th

Intl. Symposium. TMS, Warrendale, pp. 1889–

1902.

[4] FUJITA, T., TAGUCHI, R., ABUMIYA, M., MATSUMOTO, M., SHIBATA, E., NAKAMURA, T.

(2008a): Novel atmospheric scorodite synthesis by oxidation of ferrous sulfate solution. Part I.

Hydrometallurgy 90 (2–4), 92–102.

[5] FUJITA, T., TAGUCHI, R., ABUMIYA, M., MATSUMOTO, M., SHIBATA, E., NAKAMURA, T.

(2008b): Novel atmospheric scorodite synthesis by oxidation of ferrous sulfate solution. Part

II Effect of temperature and air. Hydrometallurgy 90 (2–4), 85–91.

[6] FUJITA, T., TAGUCHI, R., KUBO, H., SHIBATA, E., NAKAMURA, T. (2009): Immobilization of

arsenic from novel synthesized scorodite - Analysis on solubility and stability. Materials

Transactions 50 (2), 321-331.

[7] KONDO, Y. (1980): Recovery and fixation of arsenic from metallurgical intermediates. In:

Proceedings of Fourth Joint Meeting MMIJ-AIME, Tokyo. pp. 45-58.

[8] Ministry of the Environment, (1973): Determination of metals and other substances contained

in industrial waste. Ministry of the Environment Notification No. 13, Ministry of the Envi-

ronment, Tokyo, February 17, 1973. (in Japanese)

[9] US-EPA. (1992): Test method for the evaluation of solid waste, physical/chemical methods,

Method 1311, Toxicity characteristic leaching procedure. U.S. Environmental Protection

Agency, Washington, DC.


Method 1310B, Extraction procedure toxicity test method and structural integrity test, U.S.

Environmental Protection Agency, Washington, DC.


Method 1320, Multiple extraction procedure, U.S. Environmental Protection Agency, Washington,

DC.

[12] NEN 7341. (1995): Leaching characteristics of solid earthy and stony building and waste ma-

terials. Leaching tests. Determination of the availability of inorganic components for leaching,

Nederlands Normalisatie Instituut, Delft, The Netherlands. (in German)

[13] NEN 7371. (2004): Leaching characteristics of granular building and waste materials. The

determination of the availability of inorganic components for leaching. The maximum avail-

ability leaching test, Nederlands Normalisatie Instituut, Delft, The Netherlands. (in German)


Recycling of Electric Home Appliances

in Minamata

Yoshihiro Watanabe

Act-B Recycling Co., Ltd

278-6, Shiohama-cho, Minamata City

Kumamoto, Japan

Keywords: Electric home appliances, recycling, treatment and sorting, CRT

Abstract

Targeting the creation of a circular economy and society through the effective use of resources and

the reduction of waste production discarded at landfill sites, the Home Electrical Appliance Recycle

Law was enforced in April 2001 in Japan. The law requires manufacturers and importers to collect

and recycle their own appliances. Products addressed by the law include air-conditioners, televi-

sions, refrigerators and washing machines.

The industry has formed two consortia (Group A, Group B) to meet this responsibility. Each group

has the obligation to establish regional consolidation centre and to ensure the transport of collected

products from these centres to recycling facilities.

Group B has 16 recycling plants. Act-B Recycling Co. was founded in 1999 as one of these facilities

and started its operation in 2001. Since then we have been recycling home appliances collected from

the regional consolidation centres scattered in middle and southern part of Kyushu.

1 Introduction

The Home Appliance Recycling Law was enacted in Japan in 2001. According to this law, manufac-

turers and home appliance importers are required to collect and recycle their own appliances and

consumers must bear the collection and recycling costs (see Figure 1). Items mentioned in the law

include air-conditioners, TV sets, refrigerators and washing machines.

Watanabe


Figure 1: Flow of Home Appliance Recycling

In order to accept and recycle used home appliances, manufacturers have established Group A and

Group B recycling plants throughout the country. Each group has established regional collection

sites for collecting used home appliances and carried to recycling plants where they are dismantled,

classified, and then recycled (see Figure 2).

The target used home appliances for each group are listed as follows (manufacturers for imported

home appliances are omitted because of small numbers):

Group A: Panasonic, Toshiba and others;

Group B: Hitachi Appliance, Mitsubishi Electric, Sharp, Sony, Fujitsu General, Sanyo Electric and

others;

Group B consists of 16 recycling plants.

Recycling of Electric Home Appliances in Minamata


Figure 2: Location of Recycle Plants in Japan

Act-B Recycling (Act-B) started its operation as a member of Group B in 2001. Our target area is

the whole of Kyushu except for Northern Kyushu (see Figure 3).

Watanabe


Figure 3: Areas where Act-B collects

1.1 Act-B Company Profile

Capital: 200 M yen (1 US$ = 100 yen)

Shareholders: DOWA Eco System Co., Ltd (55 %), 2 Financial Institutions (15 %)

Home Appliance Manufacturers (Hitachi Appliance, Mitsubishi Electric, Sharp, Sony, Fujitsu Gen-

eral, Sanyo Electric) (total 30 %)

Employees: 104

Business Summary: Dismantlement and Classification of used home appliances (TV, refrigerator,

washing machine, air-conditioner), PCs and OA appliances



1.2 Company History

December 1999: Founded

April 2001: Operation started

October 2002: Obtained ISO14001 certification

February 2006: Dowa Mining Co. obtained 55 % shares of Act-B

August 2009: The number of incoming household appliances exceeds 3 million

2 Actual Performance of Collection of Used Home Appliance

The annual total number of collected home appliances in Japan since the home appliance recycling

project started is shown in Figure 4. The domestic yearly yield for FY 2008 is estimated to be ap-

proximately 24 million of which 12.9 million were actually collected (7.8 million for Group B,

5.1 million for Group A) and the rest has been used both domestically and abroad as reused items.

The numbers collected for Act-B are shown in Figure 5.

Figure 4: Change in used home appliance collection throughout Japan

(Grand total of Groups A & B)

Watanabe


Figure 5: Change in used home appliances which Act-B received

Figure 6 shows the discharged amount per household that was received by Act-B, however, this

only shows the number of collected home appliances for Group B. Considering the number reused

and collected for Group A, a number which is three times more should be considered as the actual

discharged number per household.

Figure 6: Discharged number of Group B home appliance collected by Act-B per household

(Reuse and Group A home appliance are not included)

※ Considering the amount reused and collected for Group A, the actual number will three times higher



3 Recycling Rate

The recycling rate of Act-B of FY2008 and the legal standard are shown in Figure 7.

Our recycling rate is far better than the standard substantially and also other plants in both group A

and B show similar recycling rates.

Figure 7: Recycling rate by items

4 Treatment and Sorting Process and Amount of Treated Items

The treatment and sorting process of used home appliances at Act-B are shown in Figure 8.

Regarding the dismantling process, first high-value-added components are removed manually, and

classified and subdivided to add more value. After manual dismantling, empty bodies are crushed

and classified into iron, nonferrous metals and plastics through the classification process.

Chlorofluorocarbons (CFCs) are recovered and stored in different cylinders according to type.

Watanabe


Figure 8: Treatment and sorting process in Act-B



5 Circular Situation of Resources

The classification, collection rate and destination of our treated resources by items in FY2008 are

shown in Table 1.

Table 1: Circular situation of resources at Act-B

Watanabe


6 Future Challenges

The recycling rate including heat recovery has reached an average of 93 %. In order to increase the

recycling rate, we need to improve the effective utilization of wastes such as CFCs, dust, glass

waste etc.

CRT glass currently used as raw material for CRT-base televisions will become difficult to use as

recycled material because of the transition to flat-screen televisions.

Due to its lead content, waste funnel glass cannot be taken directly to landfill sites. Therefore, pos-

sible alternative utilization methods need to be considered as soon as possible.

7 Local Harmony

In Act-B Recycling, we value our friendship with people in the community and we actively partici-

pate in collaborative work with welfare institutions.

Cooperative work with welfare sectors:

1. Group Eco: People with mild disabilities are dismantling CBs from TVs and air-conditioners

for their vocational training to be involved in society again.

2. Wakuwork: People with intellectual disabilities are dismantling PCs.

3. Hottohausu: Fetal and infantile Minamata disease patients and people with disabilities are mak-

ing chopstick rests using glass from CRT-based TVs. Act-B is cooperating with

the material handling and techniques.

Figure 9: Recycling activities



References

[1] Association Electric Home Appliances (2008): The annual report of the home appliances recy-

clingn

[2] Kagoshima Prefecture, Homepage,

http://www.pref.kagoshima.jp/tokei/bunya/jinko/suikei/home2.h21.6.1.html

[3] Kumamoto Prefecture, Homepage,

http://www.pref.kumamoto.jp/site/statistics/h21jinko-m.html

[4] Miyazaki Prefecture, Homepage,

http://www.pref.miyazaki.lg.jp/contents/org/honbu/toukei/toukeimiya/index.html

[5] Saga Prefecture, Homepage, http://www.pref.saga.lg.jp/web/_26298.html

[6] Nagasaki Prefecture, Homepage,

http://www.pref.nagasaki.jp/toukei/new_date/nen_geppou/tuki/kokuseichousa/suikei/suikei.htm

[7] Oita Prefecture, Homepage,

http://www.pref.oita.jp/10800/chosakekka/jinko/nenpo/h20/jinko.html

[8] Okinawa Prefecture, Homepage, http://www.pref.okinawa.jp/toukeika/estimates/estidata.html

[9] Nakatsu City, Homepage, http://www.city-nakatsu.jp/contents/toukei/HP-data/sub2.html



New Standards in Environmental Protection for

Copper Recycling

Dr. Franz-Josef Westhoff, Dr. Claus Meyer-Wulf

Aurubis AG

Kupferstraße 23

D-44149 Lunen, Germany

Keywords: Kayser Recycling System (KRS), fugitive emissions, material storage and pre-treatment

Abstract

After replacing the shaft furnaces and converters by the KRS (Kayser Recycling System), numerous

measures have been implemented at the Aurubis recycling centre in Lunen in recent years to reduce

emissions further, in particular diffuse emissions. These included many technically demanding

investment projects. On the one hand, new standards have been set by the construction of storage

halls and the design of outside storage facilities. On the other hand, diffuse emissions from smelting

facilities have been minimised by the construction and installation of appropriate collection systems

and hoods, etc. These have already had a noticeable effect by lowering emission rates.

Deposition values of arsenic, lead, cadmium and nickel measured by the LANUV NRW (Landesamt

für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen = Northrhine Westphalia Office for

Nature, Environment and Consumer Protection) in the vicinity of the works already reflect the first

significant improvements.

In addition, the suspended dust PM 10 immission values measured in Lunen since 2008, including

the immediate vicinity of the works, show that the limit values of the TA-Luft (Technical

Instruction on Air Quality Control) are not exceeded, both as regards quantity and heavy metal

substances. Consequently, no dust and metal values laid down by law to protect human health are

exceeded in the town of Lunen, despite the Aurubis recycling activities and various other industrial

production sites.

1 Introduction

Modern recycling forms an integral part of a sustainable raw material supply. The more we recycle

metals like copper, the less we need to use our natural resources.

Westhoff, Meyer-Wulf


Lunen is the recycling centre of the Aurubis Group, where more than 300,000 tonnes of secondary

raw materials are processed into copper cathodes each year. The by-metals contained in those

materials, such as tin, lead, nickel and precious metals, are also recycled or enriched for subsequent

winning processes.

Products at the Lunen recycling centre include high-purity copper cathodes, tin/lead alloy, zinc-

bearing KRS oxide, nickel sulphate and copper sulphate, precious metal-bearing anode slimes and

iron silicate sand.

The main production facilities are the KRS, the anode furnace and the tankhouse. The KRS Plus

Project, including the installation of a new TBRC furnace, is currently being carried out to optimise

the Lunen site further and will come into operation in 2011.

This will enable the value added in multi-metal recycling to be further increased in the next few

years. Aurubis AG has invested more than € 100 million in technology and environmental

protection at the Lunen recycling centre in the last ten years, of which € 59 million alone has

focused on environmental protection.

2 Improvements due to the KRS

The KRS was installed to replace three shaft furnaces and two converters at the beginning of 2002.

In addition, the lead/tin alloy production was technologically modified and modernised. As a result,

the specific energy consumption has been reduced by 53 % and specific emissions of carbon dioxide

by 64 %.

The stack emission of particles has been substantially reduced by more than the factor 10 by

focusing those emissions on the former converter filter plants, which were renewed or updated in

the last decade, and closing down old filter plants. In addition, diffuse emissions were significantly

reduced by improved collection.

New Standards in Environmental Protection for Copper Recycling


Figure 1: Reduction of energy consumption and carbon dioxide emissions by the KRS

3 Requirements of TA Luft 2002

3.1 More stringent deposition values

Nonetheless the improvements achieved with the KRS were not sufficient to meet the more

stringent deposition values of the new TA Luft of 2002 in the vicinity of the Lunen works. It should

be noted here that the old TA Luft of 1986 values were previously maintained despite comparably

high shaft furnace and converter emissions.

The new TA Luft resulted in the following changes:

1. The deposition values for lead and cadmium were reduced by 60 %.

2. New stringent values were introduced for arsenic and nickel.

3. New or more stringent values were introduced for mercury and thallium; these do not pose a

problem in Lunen.

4. The measuring and assessment strategy was changed from being area-related (i.e. average of four

single readings taken in a square kilometre measuring network) to spot-related.



Thus, the LANUV NRW installed various new measuring points in Lunen as of 2003, in some

instances in the immediate works’ vicinity. This inevitably resulted in limit values being exceeded,

as was the case at other German sites for metal production.

Table 1: TA Luft Deposition Values [µg per m² and day]

TA Luft

1986

TA Luft

2002

Average

Arsenic and its inorganic compounds

calculated as Arsenic

µg/(m²·d) no value 4 Year

Lead and its inorganic compounds

calculated as Lead

µg/(m²·d) 250 100 Year

Cadmium and its inorganic compounds

calculated as Cadmium

µg/(m²·d) 5 2 Year

Nickel and its inorganic compounds

calculated as Nickel


Mercury and its inorganic compounds

calculated as Mercury


Thallium and its inorganic compounds

calculated as Thallium

µg/(m²·d) 10 2 Year

3.2 Emission reduction programme

In the course of the implementation of the amended TA Luft 2002, an extensive emission reduction

programme was developed since 2005 in cooperation with the responsible authorities (at that time

Staatliches Umweltamt = State Environmental Agency Lippstadt and Bezirksregierung = Regional

Government Arnsberg) involving capital expenditure in the amount of € 10 million. The detailed

programme was defined in the course of 2006 and became legally binding in 2007 as part of a KRS

modification permit.

The emission reduction programme included the analysis of all emission sources at the Lunen works

to identify possible improvements and a series of different measures were consequently agreed. The

aim was in particular to reduce fugitive emissions in the smelters and during storage and handling of

feed materials.

The capital expenditure of the emission reduction programme was initially estimated at

€ 10 million. However, during implementation it was seen that both the originally estimated

expenditure and the agreed measures were not sufficient to achieve an optimal reduction of the

emissions. Consequently, further measures were resolved on a voluntary basis that required



additional capital expenditure of € 25 million, raising the total capital expenditure volume to € 35

million for the programme. On top of this, a further € 17.5 million will be invested in the coming

KRS Plus Project environmental protection measures, bringing the total capital expenditure by 2011

on environmental protection at the Aurubis Lunen site to more than € 52.5 million.

4 Description of the Main Measures

4.1 Storage and Handling

4.1.1 Outside Storage Places

Scrap and lumpy non-dusting feed materials are generally stored in outside storage areas. It cannot

be totally ruled out that small amounts of substances hazardous to waters adhere to the materials.

Therefore, in collaboration with the FEhS-Institut für Baustoff-Forschung e.V. (Research Institute

for Building Materials), Duisburg, we have developed an innovative sealing system for the concrete

surfaces of our storage areas. This consists of inward water stop metal strips, which are arranged in

a square grid pattern in advance of the concrete layer construction and then filled and covered up

with a water-proof concrete mass. This new system fulfils the requirements of the guideline

“Betonbau beim Umgang mit wassergefährdenden Stoffen“ (Concrete construction when handling

substances being harardous to waters) of the DAfStb (Deutscher Ausschuss für Stahlbeton =

German committee for reinforced concrete) and consequently also the requirements of the VAwS

(Verordnung über Anlagen zum Umgang mit wassergefährdenden Stoffen = Ordinance on

installations for handling substances being hazardous to waters).

As a further improvement, walls have been constructed round the outside storage areas made of

concrete bricks which are placed on top of each other with the lowest layer erected either directly on

the storage area or on simple small foundation strips. The walls protect the material from wind

blown dispersal and thus reduce fugitive emissions.



Figure 2: Construction of new outside storage areas

Figure 3: Storage Area with Concrete Brick Walls



4.1.2 New Storage Hall

Dusting nonferrous metal-bearing materials are to be stored in halls in accordance with the relevant

environmental regulations. Since the amount of such feed materials in Lunen has risen in the recent

years, it was necessary to build a significantly larger storage hall to complement the already existing

ones.

This was completed and commissioned in October 2009 after a construction time of exactly one

year.

Characteristic data are:

• Length 150 m

• Width 70 m, self-supporting

• Ridge height 21.6 m

• Floor space 10,600 m²

• Building volume 210,000 m³

• Storage capacity 12,000 t

• Capital expenditure € 7.5 million

The surface area is constructed to be water-proof just like the a.m. outside storage areas since the

stored materials are sprinkled with water as required. Furthermore, the hall contains an automatic

fire-extinguishing system for detecting and preventing fires. A tyre washing facility is in place for

the vehicles leaving the area.



Figure 4: New storage hall

Figure 5: Inside view of the new storage hall



4.2 Smelter

4.2.1 KRS Smelter

Already the initial KRS smelter design included extensive devices to minimise emissions. These

have been modified and improved on several occasions since the furnace was commissioned. One of

these improvement measures was to fit the lance port with a movable hood, which enables a better

capture of emitted fumes.

Figure 6: Movable hood at lance port

Nevertheless, it became clear that the improvements to the suction hood in the KRS sector and the

tin/lead alloy plant were not sufficient to collect the fugitive emissions satisfactorily, since the

suction performance for the entire system totalling 300,000 m³/h was too low. Consequently, a

further Filter No. 5 was erected which provides an additional suction volume of 120,000 m³/h.



Figure 7: New KRS filter 5

4.2.2 Tin/Lead Alloy Plant

The tin/lead alloy plant consists of a rotary furnace, a pretreatment of the crude tin/lead alloy, a

refining facility and a casting plant for ingots.

The rotary furnace has hoods on both sides to collect the emissions arising during charging with

liquid and solid input materials and when tapping the products. The originally existing hoods were

extensively improved; the current hood form enables almost quantitative collection.

The tin/lead alloy treatment is performed in a ladle with various additives. To collect the released

metal fumes, the treatment is performed in a closed housing, which is aspirated. The off-gases are

cleaned in a filter plant.

As part of the emission reduction programme, the housing was enlarged and modified to handle

copper fractions which are by-products of this process and were handled outside before.



4.2.3 Anode Plant

The anode plant consists of two anode furnaces, of which the tiltable anode furnace 5 which was

built in 1995 is mainly in operation, and the anode casting plant. The anode plant was improved in

several steps with the aim of reducing emissions.

In the first step, several hoods were improved, in particular at the charging doors, the slag tapping

and the casting launder between the furnaces and the casting wheels.

In a further step, the cast mould vapours that have been discharged directly into the atmosphere

before were captured, mixed with hot off-gases and dedusted in one of three former shaft furnace

filters being reactivated for this purpose.

Despite visible improvements, the remaining emissions from the ridge turrets were still

considerable. Consequently the ridge turrets have been closed in the anode furnace sector and

likewise connected to the former shaft furnace filter plant by means of a newly installed suction

duct.

Figure 8: Suction from the anode furnace



5 Emissions and Immissions

5.1 Historical trend of Emissions

As a result of the reduction programme, the emissions from the Lunen recycling centre have been

continually reduced. For example, the specific copper emissions per tonne of produced copper have

been reduced by 62 % from 2004 to 2009. Similar reduction rates were obtained for arsenic, lead,

cadmium as well as for zinc, tin and total dust. The final parts of the emission reduction programme

implemented in 2009, such as the new storage hall and the additional KRS filter plant will result in

further improvements in the coming years.

5.2 Immissions of suspended dust

Immissions of suspended dust (PM 10) are measured in Lunen at three measuring points by the

LANUV NRW, with one of the measuring points in the immediate vicinity of the Aurubis Lunen

works. The results of 2008 show that the actual TA-Luft limit values as well as the future targets of

the European legislation are not exceeded in regard to dust quantity and heavy metal concentrations.

5.3 Immissions of Dust and Metals Deposition

The deposition values in the vicinity of the Aurubis Lunen works have continuously improved since

2006 after implementation of the emission reduction programme and the numerous voluntary

measures. For example, the average value of the 10 LANUV measuring points from 2006 to 2008

shows reductions of up to 45 %.

In detail average reductions of all ten Lunen LANUV measuring points:

• Dust quantity -17 %

• Contents, such as lead -25 %

cadmium -15 %

arsenic -45 %

The only metal, which was not reduced is nickel. Reasons have still to be clarified.

Despite the achieved improvements, the limit values at some measuring points were still exceeded,

at least until 2008. The effect of the measures taken in 2009 to reduce these further (storage hall,

filter KRS) will only be apparent in the coming years.

At this point it should be mentioned that according to the LANUV publication of results deposed

dust consists of coarse dust which cannot be breathed in. Increased deposition values are not related

to the current emissions of an industrial site, but can arise from the remobilisation of dust from

surfaces, such as roads, areas or the ground, etc.



5.4 Outlook

The emission reduction programme will result in further improvements of the ambient air situation,

especially with regard to deposition values in the immediate vicinity of the works. Consequently it

can be seen as an appropriate basis for new technical projects including capacity increase, as the

KRS-Plus project.

Predicted immission values considering on the one hand the final effectiveness of all measures of

the emission reduction programme and on the other hand the relatively low additional emissions

will meet nearly all limit and target values. Exceptions exist only with nickel and copper deposition

values as well as a single lead value at Lunen municipal harbour. Regarding copper, it must be

mentioned that the deposition value is not a TA Luft value, but only derived from

Bundesbodenschutzverordnung BBodSchV (Federal Soil Protection Ordinance) for soil protection,

taking into account the most sensitive use, children’s playgrounds. If other uses like agriculture etc.

are taken into account, the predicted values do not incorporate any hazard.



Posters

Poster of Copper 2010 2986

Recovery of Copper from Copper Smelter

Wastewater by Electrodialysis

Dr. Henrik K. Hansen, Dipl.-Ing. Claudia Gutiérrez, Dipl.-Ing. Jorge Ferreiro

Universidad Técnica Federico Santa Maria

Departamento de Ingeniería Química y Ambiental

Avenida España 1680

Valparaíso, Chile

Copper smelters – and the sulfuric acid plants that are connected to them – generate large amount of

acidic wastewater containing a series of inorganic contaminants, where arsenic and copper are key

contaminants. Heavy metals are typically precipitated as hydroxides but arsenic remains in the

nearly pH-neutral wastewater. Arsenic is precipitated with a Fe(III) treatment, and the combined

solids are dried, calcined and disposed off as a hazardous waste. Valuable elements such as copper

are therefore lost in these deposits of waste materials.

This work analyses the possibility to recover copper before a precipitation of arsenic and other

heavy metals. The process analyzed is electrodialysis, where primarily the separation of copper from

arsenic is studied. Copper is found as Cu2+

in the acidic wastewater – in a concentration of about

1100 mg L-1

– and therefore expected to migrate towards the cathode. On the other hand, arsenic –

in a concentration of about 5400 mg L-1

– is mostly present in trivalent state, and therefore expected

to migrate to the anode as arsenite species.

The results showed that raw copper smelter wastewater was difficult to process, mainly due to the

very low pH (around 0). Only small amounts of copper were removed towards the cathode. At this

pH most of the current probably was carried by H+ and SO4

2-. On the other hand, after a preliminary

Ca(OH)2 addition in order to reach a pH of 2, copper could be removed completely towards the

cathode, whereas arsenic stayed in the original compartment or moved towards the anode. Other

cations such as lead also migrated but too much lesser extend to the cathode compartment and a

separation of cationic species was possible, too. These results are very promising for a possible

recovery of copper from smelter wastewater by electrodialysis.

RReeccoovveerryy ooff ccooppppeerr ff rroomm ccooppppeerr ssmmeell tteerr wwaasstteewwaatteerr bbyy

eelleecctt rrooddiiaall yyss iiss..

Henrik K. Hansen, Patricio Núñez, Jorge Ferreiro and Claudia Gutierrez

Departamento de Ingeniería Química y Ambiental, Universidad Técnica Federico Santa María, Casilla 110-V, Chile E-mail: [email protected]

Introduction At the Codelco El Teniente copper smelter, the wastewater treatment is generally as shown in Figure 1. The final sludge is dried and disposed off as a hazardous waste product. The waste product contains arsenic, copper and other heavy metals. Lately, the copper smelter has shown interest in developing processes to recover the copper before the arsenic precipitation steps. A possible solution could be electrodialysis (ED). Figure 2 shows a typical ED setup. In the acidic smelter wastewater, copper is expected to be present as cationic species (Cu2+), whereas arsenic probably exists as uncharged or anionic species (arsenite or arsenate). This would make an ED separation of Cu and As possible. The purpose of this work is to evaluate if it is possible to separate copper from arsenic in acidic copper smelter wastewater by ED. The efficiency of the process will be tested on raw wastewater and wastewater added Ca(OH)2 in order to from raise pH to 1, 2 and 3, respectively, and to precipitate initially some amounts of Ca(SO4). Parameters that would be analysed is the treatment time and the current density. The feasibility of the process is evaluated from the copper and arsenic displacement within the different compartments of the ED cell.

Experimental The ED technique was tested experimentally on copper smelter wastewater – either as raw wastewater or with adjusted pH. Practically no copper and arsenic were precipitated during the pH adjustment. The raw wastewater was sampled at the El Teniente copper mine in VI Region of Chile. Table 1 gives the experimental data.

Figure 1. Actual wastewater treatment.

Figure 2. Experimental ED scheme.

Table 1. Experimental details.

Sample type CAs mg L-1

CCu mg L-1

pH

Current density A m-2

ED Time hr

Raw wastewater 2600 ± 80 580 ± 30 0.6 ± 0.2 150 - 300 0 - 5 pH 1 2570 ± 30 530 ± 20 1.0 ± 0.1 150 - 300 0 - 5 pH 2 2630 ± 40 540 ± 30 2.0 ± 0.1 150 - 300 0 - 5 pH 3 2510 ± 40 500 ± 20 3.0 ± 0.1 150 - 300 0 - 5

Results

From the experiments it was found that copper migrated with the electric field to compartment 2. No copper was found in compartment 4 in all experiments. On the other hand, arsenic migrated mainly towards compartment 3. some small amount of arsenic was found in compartment 2 but the experiments confirmed that it could be possible to separate copper from arsenic by ED. Figure 3 shows an example of the distribution of As and Cu during ED with time.

Compartment 2

0

100

200

300

400

500

600

0 1 2 3 4 5 6ED time (hr)

Co

nce

ntra

tion

(m

g L

-1)

Cu As

Compartment 3

0

500

1000

1500

2000

2500

3000

0 1 2 3 4 5 6ED time (hr)

Con

cent

ratio

n (m

g L

-1)

Cu As

Compartment 4

0

100

200

300

400

500

600

700

800

0 1 2 3 4 5 6ED time (hr)

Con

cent

ratio

n (m

g L

-1)

Cu As

Figure 3. As and Cu distribution during ED. pH = 2, current density = 225 A m -2.

Figure 4 shows the copper distribution as a function of ED time for 4 different wastewaters. It is clear to see that copper is removed faster from solutions adjusted to a higher pH. This is also expected since the competition between Cu2+ and H+ would favour Cu2+ at a higher pH. On the other hand, it can also be seen that it is necessary to adjust the pH since copper in raw wastewater does not seem to be affected by the electric field – only after 5 hours there is a slight tendency that copper is starting to move towards compartment 2.

Compartment 2

0

100

200

300

400

500

600

0 1 2 3 4 5 6

ED time (hr)

Con

cent

ratio

n (m

g L

-1)

Raw w astew ater pH 1 pH 2 pH 3

Compartment 3

0

100

200

300

400

500

600

700

0 1 2 3 4 5 6ED time (hr)

Con

cent

ratio

n (m

g L

-1)

Raw w astew ater pH 1 pH 2 pH 3

Figure 4. Cu distribution during ED for different pH. Current density = 300 A m -2. Table 2 shows the current efficiency with respect to copper. It can be seen from the table that still at pH 3 other species than copper are mainly carrying the current. Anyway, the main purpose of this work is fulfilled – copper and arsenic can be separated by ED. The next step would be to follow the other species during ED. From the experiments it could be estimated that in order to remove 1 mol Cu, 32.7 MC is needed at pH 1, 11.2 MC at pH 2, and 7.5 MC at pH 3.

Table 2. Current efficiency with respect to copper after 2 hour ED.

Wastewater Current density (A m-2)

Charge (C)

Current efficiency (%)

pH 1 150 225 300

1.18 0.88 0.59

0.7 0.4 0.8

pH 2 150 225 300

1.18 0.88 0.59

1.5 1.9 1.8

pH 3 150 225 300

1.18 0.88 0.59

2.5 2.6 2.7

Conclusions

The results showed that in the raw copper smelter wastewater, it was difficult to separate copper from arsenic with ED, mainly due to the very low pH. Most of the current was used in transporting H+.

On the other hand, after a preliminary Ca(OH)2 to adjust pH to 1, 2 or 3, copper was efficiently removed towards the cathode. At the same time, arsenic remained in the wastewater or was removed towards the anode.

ED could be an interesting tool in order to recover Cu from smelter wastewater and decreasing the amount of waste generated.

The authors wish to thank FONDECYT Project Nº 1100440 for economical support

a) Oxidized raw wastewater from copper smelter gas cleaning. pH: ~1 As: 5000 – 15000 mg L-1

Ca(OH)2

Floculant

b) Wastewater after 1st precipitation pH: ~ 10 As: 100 – 1000 mg L-1

1st Precipitation

2nd Precipitation

Fe3(SO4)2/ FeCl3 HCl/H2SO4

Floculant

Filtration

Drying and calcination

c) Treated wastewater pH: ~ 7 As: 0.1 – 5 mg L-1

Sludge

Sludge

CaCO3

Filtrate

Stabilized solid residue with arsenic and heavy metals

Wet solids

Wastewater

A: Anion exchange membrane C: Cation exchange membrane 1: Cathode compartment 2: Cation concentration compartment 3: Wastewater compartment 4: Anion concentration compartment 5: Anode compartment


Removal of Arsenic from Copper Smelter

Wastewaters by Airlift Electrocoagulation

Dr. Henrik K. Hansen, Dipl.-Ing. Claudia Gutiérrez, M.Sc. Patricio Nuñez

Universidad Técnica Federico Santa Maria

Departamento de Ingeniería Química y Ambiental

Avenida España 1680

Valparaíso, Chile

Pyrometallurgic copper processing generates large amounts of arsenic that vaporize as arsenic

trioxide. This compound is absorbed from the gas flow leading into the sulphuric acid plant together

with a variety of heavy metals, creating a highly acidic contaminated wastewater. Wastewater from

copper smelters contains typically considerable amounts of copper, lead, cadmium, zinc, arsenic and

mercury. Heavy metals can precipitated as hydroxides with a lime treatment but arsenic remains in

the nearly pH-neutral wastewater. Combined Ca(OH)2 and FeCl3 precipitation can deal with most of

the arsenic but since the arsenic concentration in the gas phase changes due to the batch wise

operation of the smelter, it is difficult to predict and control the chemical dosage for the

precipitation of the arsenic compounds.

Electrocoagulation (EC) has shown its potential for arsenic removal due to formation of ferric

hydroxide-arsenate precipitates. This work evaluates the feasibility of EC as a treatment process at

various stages during conventional copper smelter wastewater treatment – with focus on arsenic.

The reactor used is a batch airlift electrocoagulator.

The results showed that raw copper smelter wastewater was difficult to treat for arsenic and heavy

metals with EC, mainly due to the very low pH. On the other hand, after a preliminary Ca(OH)2

treatment for sulphate and heavy metal removal, arsenic could be removed totally by EC. In addi-

tion, EC could also be applied as a final remediation control tool for arsenic since the national

threshold value for wastewater discharge could rapidly be reached when the conventional method

did not clean the wastewater sufficiently.

AArrsseenniicc rreemmoovvaall ff rroomm ccooppppeerr ssmmeell tteerr wwaasstteewwaatteerr bbyy

eelleecctt rrooccooaagguullaatt iioonn iinn aann aaii rr ll ii ff tt rreeaaccttoorr ..

Henrik K. Hansen, Patricio Núñez and Claudia Gutierrez

Departamento de Ingeniería Química y Ambiental, Universidad Técnica Federico Santa María, Casilla 110-V, Chile E-mail: [email protected]

Introduction

At the Codelco El Teniente copper smelter, the wastewater treatment is generally as shown in Figure 1. Metals such as a copper and zinc are mainly precipitated as hydroxides together with calcium sulphate in the first step but a large amount of arsenic remains in the slightly alkaline wastewater. Combined Fe3(SO4)2 and FeCl3 addition in acid deals with the arsenic but since the arsenic concentration in the gas phase changes due to the batch wise operation of the smelter, it is difficult to predict and control the chemical dosage for the precipitation of the arsenic compounds. This work focuses on the treatment of copper smelter wastewater with an airlift batch electrocoagulation (EC) reactor. The efficiency of the process will be tested for the raw wastewater and wastewater from different stages in the actual wastewater treatment plant. Parameters that would be analysed is the treatment time and the applied electric current. The results are evaluated in terms of arsenic removal efficiency.

Experimental

The EC technique was tested experimentally on copper smelter wastewaters from three different stages during the actual wastewater treatment: a) raw wastewater (after oxidation) - WW1, b) wastewater after Ca(OH)2 precipitation – WW2, and c) treated wastewater – WW3. Figure 2 shows the experimental setup and Table 1 gives experimental details.

Figure 1. Actual wastewater treatment. Figure 2. Experimental setup.

Table 1. Experimental details.

Sample type Sample Code CAs mg L-1

Time Current

a) Raw wastewater WW1 5250 ± 150 60 – 300 min 2 – 5 A b) After 1st precipitation WW2 910 ± 40 15 – 120 min 2 – 5 A c) Treated wastewater WW3 4.38 ± 0.22 2 – 20 min 0.8 – 5A

Results

Tables 2-4 show the EC results for arsenic removal from WW1, WW2 and WW3 respectively. When first analysing the raw wastewater (WW1), it can be noted that the arsenic and heavy metal removal seems very slow. Only for the experiments carried out at 5A and longer than 240 min duration some indication of removal was observed. In fact, only in these two experiments some amounts of orange/brown solids were formed during the EC process. In these experiments the final pH reached 1.4 and 1.8, respectively. Probably the pH was too low in the other experiments to form the necessary ferric hydroxides - or hydrated ferric oxides (HFO), which are known to adsorb arsenic and heavy metals efficiently. Therefore, the EC process should have been carried for longer time in order to first neutralize the wastewater to a certain pH level before any arsenic removal could be an option.

Table 2. EC results for WW1.

EC Time (min)

I (A)

Final pH CAs,final

(mg L-1) Arsenic removal

(%) 60 2 0.7 5100 0

120 2 0.7 5150 0 300 2 0.9 5300 0 60 3 0.7 5050 0

120 3 0.8 5400 0 300 3 1.1 5200 0 60 5 0.7 5400 0

120 5 0.8 5320 0 180 5 1.0 5300 0 240 5 1.4 4990 6 300 5 1.8 4750 10

Another option could be the implementation of the EC process after the actual Ca(OH)2 addition – at point b) on Figure 1. This specific treatment has several objectives: 1) to remove sulphates, b) to increase pH to around 10 in order to precipitate heavy metal cations, and c) to precipitate some arsenic as Ca3(AsO4)2. The clarified solution from the sedimentation was sampled and called WW2 in this work. From Table 3 it can be observed that arsenic could be removed to less than the detection limit after 240 minutes with 2A when treating WW2 with EC. Increasing the current to 5A, complete removal was obtained after 180 minutes.


EC Time (min)

I (A)

Final pH CAs,final


(%) 30 2 9.9 600 34.1 60 2 9.9 320 64.8 90 2 10.3 170 81.3 120 2 10.3 70 92.3 180 2 10.4 17 98.1 240 2 10.5 < 0.02 > 99.9 30 5 10.0 330 63.7 60 5 10.0 150 83.5 90 5 10.4 49 94.6 120 5 10.6 9.7 98.9 180 5 10.7 < 0.02 > 99.9 240 5 10.9 < 0.02 > 99.9

Arsenic is removed rapidly from the treated wastewater WW3 (Table 4). For a 2 min treatment, using the lowest current (0.8 A), already 98% of the arsenic has been removed. For higher currents, at 2 min the arsenic concentration is below detection limit. The same is the case for longer treatment at all applied current densities. For all experiments, the arsenic concentration after treatment was below the Chilean norm for discharge (0.5 mg L-1). This means that the EC process could be a very efficient control option for the treated wastewater. Therefore, the Fe(OH)3 dosage could be controlled with the applied current.


EC Time (min)

I (A)

Final pH CAs,final


(%) 2 0.8 7 0.08 98.2 5 0.8 7 < 0.02 > 99.5

20 0.8 7 < 0.02 > 99.5 2 2 7 < 0.02 > 99.5 5 2 7 < 0.02 > 99.5

20 2 7 < 0.02 > 99.5 2 5 7 < 0.02 > 99.5 5 5 7 < 0.02 > 99.5

20 5 7 < 0.02 > 99.5

Conclusions The results showed that raw copper smelter wastewater was difficult to treat for arsenic and heavy metals with EC, mainly due to the very low pH. On the other hand, after a preliminary Ca(OH)2 treatment for sulphate and heavy metal removal, arsenic could be removed totally by EC. In addition, EC could also be applied as a final control tool for arsenic since the national threshold value for wastewater discharge could rapidly be reached when the conventional method did not clean the wastewater sufficiently.

The authors wish to thank FONDECYT Project Nº 1085118 for economical support.

Iron electrodes

O2

a) Oxidized raw wastewater from copper smelter gas cleaning. pH: ~1 As: 5000 – 15000 mg L-1

Ca(OH)2

Floculant

b) Wastewater after 1st precipitation pH: ~ 10 As: 100 – 1000 mg L-1

1st Precipitation

2nd Precipitation

Fe3(SO4)2/ FeCl3 HCl/H2SO4

Floculant

Filtration

Drying and calcination

c) Treated wastewater pH: ~ 7 As: 0.1 – 5 mg L-1

Sludge

Sludge

CaCO3

Filtrate

Stabilized solid residue with arsenic and heavy metals

Wet solids


Quantitative Mineralogy: X-ray Analytics of

Copper Ores and Concentrates

Dr. Karsten Knorr

Bruker AXS GmbH

Östliche Rheinbrückenstraße 49

76187 Karlsruhe, Germany

X-ray diffraction (XRD) is the only direct method for evaluating the mineralogy of a sample. Each

mineral (phase) is identified by its unique XRD-fingerprint. Amounts of crystalline and amorphous

phases are obtained in a sub-% to 100 % range. This is of large relevance to processing of geo-

materials since properties such as the rock-type, hardness, ore concentrations, floatability, cleavage,

acid consumption, … solely depend on the type of minerals present in the specimen and not the

chemistry.

Modern Silicon-strip detector technology makes the measurement of full XRD patterns fast and

therefore, applicable to industry use. The entire process from the sample preparation to the final data

evaluation takes about 15 min and can completely be automated. The poster presents the example of

a Copper concentrate analyzed with the Rietveld method.

The results of the XRD support mine planning and operation. Furthermore they may help deciding

where to mine, what mill and milling times to use, how to further process the material and how to op-

timize the processing conditions. In that sense XRD may contribute to improved mine profitability.

Benefits

X-ray fluorescence (XRF) is an established technology for the determination of the chemical composition of materials. It delivers concentrations of elements from Be to U in a concentration range from sub-ppm to 100%.

Properties of relevance to processing of geo-materials (rock-type, hardness, ore concentrations, floatability, cleavage, acid consumption, …) solely depend on the type of minerals present in the specimen and not the chemistry.

X-ray diffraction (XRD) is the only direct method for evaluating the mineralogy of a sample. Each mineral (phase) is identified by its unique XRD-fingerprint. Amounts of crystalline and amorphous phases are obtained in a sub-% to 100% range.

1

1

2

3

Secondary list of mineral classes

Results reconciliation

Primary minerals grouped into ore and gangue

Karsten KnorrBruker AXS GmbH,76187 Karlsruhe, Germany

The specimen presented to the X-ray diffractometer (Fig. 1) is typically a pellet of powder having about 1 to 10 µm grain size. The same sample can be used for XRF analysis.

A typical measurement is shown in figure 2, together with the TOPAS data evaluation. The full diffraction pattern is calculated based on the individual contributions of the minerals identified in the specimen (Rietveld method). The quantification requires no standard samples. It is based on the well-known tabulated crystal structures of minerals.

Modern Silicon-strip detector technology makes the measurement fast and therefore, applicable to industry use. The entire process from the sample preparation to the final data evaluation takes about 15 min and can completely be automated.

A typical result is given in figure 2. It shows the concentration of minerals (primary minerals table) found in a Copper concentrate.

Different presentations of this result are given in figure 3. Minerals of the primary results table are grouped into ore and gangue, or into several mineral groups.

The mineral concentrations can be compared to results from QUEMSCAN/MLA minerals liberation analysis.

Finally, from the minerals the chemical composition may be obtained and compared to, or combined with XRF, IR, absorption and other chemical analysis.

The quantitative mineralogy is directly obtained from XRD data. These results support

• Geology,

• Mine planning,

• Mine operation,

• Metallurgical testing, and

• Leaching and leach modeling.

The ore characterization lab may anticipate the ore response to the processing conditions and improve the general performance of the operation.

Fig. 1 D8 ADVANCE diffractometer with fast LYNXEYE Silicon strip detector and 90 position sample changer.

Fig. 2 TOPAS data evaluation of a Copper concentrate.

Fig. 3 Different presentations obtained from the primary minerals concentration (see Fig. 2).

Quantitative X-ray mineralogy improves mine profitability

• Evaluation of ore grades

• Target areas for selective mining

• Higher mill throughput

• Consume less energy at the mill

• Increased recovery

• Reduced acid consumption duringleaching

• Less waste material.

Objectives

Methods

Summary

Results

Quantitative Mineralogy: X-ray Analytics of Copper Ores and Concentrates

Copper 2010, Hamburg

XRD / XRF

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Total Silicate 21.95

Total Ca Sulfate 1.93

Total Carbonate 5.34

Total Fe Oxide 1.28

Total Pb-Sulfide 0.32

Total Mo Sulfide 0

Total Zn Sulfide 3.04

Total Fe Sulfide 12.42

Total Cu Sulfide 53.73

53.73

XRD_secondary

3 0

10

20

30

40

50

60

70

80

Ore

wt%

Goethite

Hematite

Magnetite

Galena

Molybdenite

Sphalerite

Arsenopyrite

Pyrhhotite

Marcassite

Pyrite

Digenite, Chalcocite

Bornite

Covellite

Chalcopyrite

0

5

10

15

20

25

30

35

Gangue

wt%

Lizardite

Chlorite

Talc

Amphibole

Olivine, Pyroxene

Andalusite

Garnet

Clay

Swelling Clay

Kaolinite

Mica

Plagioclase

K-Feldspar

Quartz

Bassanite

Gypsum

Anhydrite

Ankerite

Magnesite

Siderite

Dolomite

Calcite

2

Copper concentrate

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Total Silicate 21.95 20.03

Total Ca Sulfate 1.93 1.22

Total Carbonate 5.34 3.68

Total Fe Oxide 1.28 1.87

Total Pb-Sulfide 0.32 0.32

Total Mo Sulfide 0 0

Total Zn Sulfide 3.04 3.04

Total Fe Sulfide 12.42 11.76

Total Cu Sulfide 53.73 59.38

XRD_secondary QUEMSCAN_secondary


Continuously Cast Copper Alloys in

Thin Dimensions and Their Applications

Dr. rer. nat. Eberhard E. Schmid

Berkenhoff GmbH

Berkenhoffstraße 14

D-35452 Heuchelheim, Germany

Under its umbrella brand bedra®

, the Berkenhoff GmbH produces and distributes innovative non-

ferrous wires used for a number of special applications. The Berkenhoff GmbH, a medium-sized

enterprise founded by Karl Berkenhoff in 1889, has two production sites in central Germany and by

means of its agencies and sales companies it operates in more than 80 countries worldwide.

The Berkenhoff GmbH is considered to be the only manufacturer of precision wires made of copper

and copper alloys that generally has everything at its disposal in-house. After melting and alloying

the raw material specifically to its function and after the following continuous casting, the product

“wire” is manufactured to customer specifications by cold rolling and drawing. Between the latter

steps of the manufacturing process, the material has to be annealed at a certain temperature by

applying the necessary heat treatment. Depending on the requirements, the wire surface may be

coated with galvanic layers. Furthermore, the packaging is customer-oriented and in line with

demand, thus rounding off the wire product. In this way, the highest levels of quality “Made in

Germany“ and in-house processing stability are guaranteed for all steps in the production process.

More than one hundred different copper alloys are produced, among them are alloys including

different brasses, phosphor bronzes or special bronzes, copper-nickel or nickel silver; among them

are also several special alloys and precipitation-hardenable alloys. These alloys are used for wires

employed in electrical discharge machining, in electronics for connector pins or heating conductors,

in the automotive industry as filler material for welding and brazing using lasers or arc beams,

furthermore for spectacle frames, as anchoring wires in tooth brushes and in costume jewellery.

Therefore the alloys are continuously cast in thin dimensions, i.e. up to 20 mm in diameter, and are

intended for the small-scale production of fine wires. Up to 12 strands can be drawn simultaneously

from a casting furnace. This ensures a uniform and constant quality in a tight production process

line. By means of six different casting units alloys with the most various compositions may be cast

efficiently. Moreover, alloys can be changed quickly and flexibly using moderately sized melting

and casting furnaces. The melting and alloying process is closely connected to a correct alloy

adjustment. The close cooperation between the foundry and the chemical analysis department,

which is equipped with state of the art analysis technology, makes it possible that even alloys with

compositions limitations exceeding below international standards can be produced as well.

Great flexibility due to several casting units with different layouts:

7 melting furnaces: chamber capacity up to 2000kg, output up to 800 kg/h

6 casting furnaces: chamber capacity up to 1200kg, performance up to 7t/day

Efficient change of alloys through moderately-sized casting furnaces

Short reaction time

Production monitoring through a process control system

Independence due to in-house foundry and furnace construction

Continuously cast copper alloys in thin dimensions and their applications

The only in-house manufacturer of precision wires made of copper alloysCasting, rolling, annealing, drawing and packagingSurface refinement through galvanic electroplatingConstant quality and processing stability in the overall process at the highest levelCooperation with partners from industry and research institutions

Tin bronzes (phosphor bronzes)17 alloy variants up to CuSn12eg Bl5, B65, CuSn6, CuSn8

Brasses11 alloy variants from CuZn2 to CuZn36eg Ms80, Ms70, Ms64

Special bronzessilicon bronzes with 5 alloy variants up to CuSi3 aluminium bronzes with alloy variants up to AlBz8manganese bronzes with 3 alloy variants up to CuMn12Ni2

Special alloyseg CuAg1, CuAg0,5Zn1

Precipitation-hardenable alloyseg CuNi3Si, NIBRODAL116, CA725

Nickel & nickel silver alloys10 alloy variants from Ns10 up to Ns25eg CuNi6, Ns18

More than 100 different alloys Fields of application

Features of the casting processCasting processSchematic layout of a continuous casting unit; the flow of the heat transfer is also marked (S. Riedel, H. Ricken, TU München UTG)

Transfer from the melting furnace to the casting furnace

Chemical analysis departmentProduction of alloys with composition limitations run below international standards

Specific alloy adjustments as soon as the metal is molten

Adjusting and securing special alloy tolerances through numerous in-house copper-based calibration samples

Purity of alloys guaranteed by

• Use of pure metals (LME qualities)

• Admixture of own homogenous production scrap

• No addition of toxic materials such as Pb, Cd, Hg, Be, Cr, Sb, As

• No purchase of external scrap

Mechanical engineering:Fine wire electrodes in electrical dischargemachining (EDM)

Electronics:Connector pins for contactsheating conductor wires

Joining technology:Filler electrodes for brazing and welding (WIG, MIG, laser, plasma)

Medical engineering:Anchoring wires in tooth brushesscalpel, catheters

Optics and visual effects:Pre-material for spectacle framescostume jewellery, decorative applications

Contact detailsBerkenhoff GmbH

• Dr. Eberhard Schmid• Head of Casting Processes & Metallurgy• [email protected]• Berkenhoffstr. 14, D-35452 Heuchelheim• www.bedra.com

Berkenhoff GmbH • Pascale Papadakis• Plant Manager• [email protected]• Berkenhoffstr. 14, D-35452 Heuchelheim• www.bedra.com



Authors Index



Authors Index

Abisheva ................................................... 893

Abumiya ................................................. 2947

Adham ...................................................... 587

Adji ........................................................ 1143

Aguayo ..................................................... 285

Ahan ....................................................... 1437

Ahmadi ................................................... 1725

Ahmadi ................................................... 2575

Ahokainen .............................................. 1003

Al Shakarji ............................................. 1237

Alfaro ....................................................... 301

Almaraz ............................................ 919, 931

Altenberger ................................................... 3

Alvani ....................................................... 247

Alvayai ................................................... 2693

Alvear ....................................................... 615

Ammon .................................................. 2341

Anderson ................................................ 2589

Anming ........................................... 823, 1313

Ante .......................................................... 601

Antrekowitsch, H. .................................... 863

Anyszkiewicz ......................................... 1803

Anzinger ................................................. 1713

Aracena .................................................. 2443

Araneda .................................................. 2637

Arbizu .................................................... 1599

Arellano .................................................... 931

Aromaa ................................................... 1959

Arthur ....................................................... 615

Artigas .................................................... 2637

Aslin ....................................................... 1253

Asselin ................................................... 1909

Atsumi ........................................................ 89

Bahamondes ........................................... 2637

Balladares ............................. 465, 1259, 1273

Baranek ...................................................... 13

Bareel ..................................................... 2935

Barrón ...................................................... 931

Barturen ................................................. 2913

Bassa ...................................................... 2189

Basson ................................ 1737, 1753, 1771

Bastin ..................................................... 2935

Basu ................................... 2307, 2381, 2499

Baxter ..................................................... 1783

Bayanmunkh .......................................... 1495

Beas ........................................................ 1355

Bednarski ............................................... 2059

Begazo ................................................... 1327

Behrens .................................................. 2267

Beltran .................................................... 2071

Bender .................................................... 1271

Benke ..................................................... 1803

Bernhard ................................................. 2341

Bhatt ....................................................... 2307

Björklund ............................................... 1155

Blackett .................................................. 1281

Authors Index


Bombach ...................................... 1675, 1687

Botín ................................................. 309, 319

Brantes ..................................................... 333

Brees ....................................................... 1393

Bronk ...................................................... 2341

Brouwer .................................................. 2159

Bruch ...................................................... 2397

Bruening ................................................. 1941

Bryant ....................................................... 971

Buchholz .................................................. 587

Burchardt ................................................ 2621

Bustos ..................................................... 1355

Byszyński ................................................. 631

Byun ......................................................... 831

Caballero ............................ 1013, 1095, 1273

Campbell .................................................. 351

Capanema ............................................... 2453

Cardona .................................................. 2637

Cartagena ............................................... 2283

Cekel ........................................................ 517

Chandrasekaran ...................................... 2873

Chauhan ....................................... 2381, 2499

Chen, S. .................................................. 1183

Chen, T. T. ............................................. 1815

Chisakuta ................................................ 1585

Chmielarz ......................................... 13, 1803

Chmielewski ................................ 2655, 2673

Cho ......................................................... 1437

Chugh ................................. 2307, 2381, 2499

Ciura ........................................................ 255

Clayton ................................................... 1635

Cocalia ................................................... 2059

Coleman ............................... 685, 1079, 2517

Collao ..................................................... 1871

Cook ....................................................... 1617

Coombs .................................................. 2101

Coursol ..................................................... 649

Cronje .................................................... 1999

Cuenca ..................................................... 919

Cwolek ..................................................... 255

Czernecki ................................................. 669

Daelman ................................................. 2159

Dale ........................................................ 1941

Deaulmerie ............................................. 2295

Degel ...................................................... 1213

Delbeke .................................................... 365

Delgado .................................................. 1871

Demuth .................................................. 1051

Deneys .................................................... 2327

Denoël .................................................... 2295

Desai ...................................................... 2307

Desbiens ................................................. 2829

Devia ...................................................... 1293

Devos ....................................................... 379

Dianbang ................................................ 2527

Díaz, C. M. ..................................... 649, 2543

Diaz, G. .................................................. 1899

Díaz, M. ................................................. 2693

Authors Index


Dixon ............................................ 1845, 1909

Djurov .................................................... 1307

Dominguez ............................................. 1635

Dreisinger ............................................... 2087

Dryga ...................................................... 1063

Dutrizac .................................................. 1815

Dyussekenov .......................................... 2483

Eberle ......................................................... 65

Ebin ............................................................ 21

Eckenbach .................................................. 31

Edens ........................................................ 413

Eghbalnia ............................................... 1845

Eklund .................................................... 1003

Ekman, E. ............................................... 1545

Ekman, S. ............................................... 1857

El Jundi .................................................. 2765

Ellis ........................................................ 1593

Emmerich ............................................... 1585

Enriquez ................................................. 2327

Eriksson .................................................. 1253

Estelle ............................... 185, 191, 207, 217

Evans ...................................................... 2473

Fagerlund ................................................. 699

Fayram .................................................... 2589

Feather .................................................... 1585

Fernandez ............................................... 1327

Ferreiro ................................................... 2986

Fic ........................................................... 2845

Filzwieser, A. ..................... 1233, 1247, 1713

Filzwieser, I. ................................ 1247, 1713

Fleury ..................................................... 1871

Font .............................................. 1013, 1095

Ford ........................................................ 1035

Forsén ..................................................... 1959

Free .................................... 1649, 2703, 2711

Frias ....................................................... 1899

Friedrich, B. ........................................... 1495

Fuentes ................................................... 1095

Fujita ...................................................... 2923

Fukushima .............................................. 1531

Gamweger ................................................ 713

Garycki ..................................................... 631

Gaydardzhiev ......................................... 2935

Gebski .................................................... 1123

Gehrckens .............................................. 2117

Gencer ........................................................ 21

George-Kennedy ........ 811, 1123, 1297, 2415

Gernerth-Mautner Markhof ................... 2341

Gerth ........................................................ 533

Geveci .................................................... 1115

Ghahremaninezhad ................................ 1909

Girouard ................................................... 587

Gizicki ...................................................... 669

Goffinet .................................................. 2935

Gómez .................................................... 2127

González A. ........................................... 2747

González, P. ........................................... 2127

González, S. ........................................... 2127

Authors Index


Gostyński ................................................. 631

Grasser ................................................. 49, 65

Green ........................................................ 247

Gregory .................................................. 1237

Griez ....................................................... 2295

Grossmann ............................................. 2793

Gunnewiek ............................................. 2357

Gupta ...................................................... 1035

Gurmen ....................................................... 21

Gutiérrez ....................................... 2986, 2988

Habashi ................................................... 2721

Hannemann .............................................. 387

Hansen .......................................... 2986, 2988

Hanusch .................................................... 721

Hao ............................................................. 65

Hapçı ........................................................ 399

Hartingh ................................................... 351

Hartmann ................................................ 2737

Hashikawa .............................................. 1521

Hashimoto .............................................. 1345

Häuser .................................................... 1687

Hayes .............................. 731, 761, 811, 1297

He ........................................................... 1237

Hebner ...................................................... 533

Hecker .......................................... 1355, 2747

Heferen ................................................... 1253

Hein ........................................................ 1925

Henao ............................................... 731, 761

Hernández, B. ........................................... 919

Hernández, C. .......................................... 465

Hernandez, L. ......................................... 2637

Herrándiz ................................................. 309

Herrera, E. ................................................ 749

Herrera, L. .............................................. 2693

Hessling ................................................... 561

Hidayat ..................................................... 761

Hinrichs-Petersen ............................. 517, 561

Hirai ......................................................... 779

Hiroyoshi ............................................... 2753

Hisahi ......................................................... 89

Hiskey .......................................... 1367, 1463

Hoffmeister .............................................. 413

Hojda .......................................................... 81

Holleis .................................................... 1051

Hong ....................................................... 1437

Horbach .................................................. 2737

Hoshi ........................................................ 779

Houbart .................................................... 379

House ..................................................... 2137

Hufschmidt ............................................ 2397

Hughe ....................................................... 961

Hundrieser .............................................. 2369

Hyde ......................................................... 685

Iida ......................................................... 1199

Ishmurzin ................................................... 65

Ito ........................................................... 1403

Iwahori ................................................... 1403

Izatt, N. E. .............................................. 1941

Authors Index


Izatt, S. R. ............................................... 1941

Jacobsen ..................................................... 99

Jåfs ........................................................... 699

Jak .......................... 731, 761, 793, 811, 1297

Jara D. ...................................................... 439

Jara, H. ................................................... 1013

Johnston ................................................. 2039

Jones ....................................................... 1079

Joshi ....................................................... 2381

Joy .......................................................... 1379

Jun .................................................. 823, 1313

Jung .......................................................... 831

Jurovitzki ................................................ 2703

Juszczyk ................................................... 255

Kalisch ................................................... 1213

Kamenetzky ............................................ 2059

Kamiya ................................................... 2781

Kapusta ..................................................... 839

Kashefipour ............................................ 2341

Kassebaum ............................................. 2397

Kato ........................................................ 2563

Kaur ........................................................ 2415

Kawakita ................................................ 2781

Kawanaka ............................................... 1025

Kawecki ........................................... 137, 155

Kelebek .................................................. 2765

Khandelwal .................................. 2381, 2499

Kiani ....................................................... 2433

Kim ........................................................ 1393

King ......................................................... 971

Kiyotani ................................................. 1199

Klymowsky ............................................ 2621

Knecht .................................................... 2621

Knorr ...................................................... 2990

Knych ....................................... 119, 137, 155

Kobayashi .............................................. 1531

Köhler ........................................................ 81

Kojima ....................................................... 89

Kokourine ................................................ 587

Komori ................................................... 1403

Koncik .................................................... 1233

Kondoh ...................................................... 89

Konetschnik ............................................. 863

Konishi ................................................... 2781

Kopyto ....................................................... 13

Korpi ...................................................... 1569

Kosaka ....................................................... 89

Köster ....................................................... 879

Kozhakhmetov ......................................... 893

Kramer ................................................... 2737

Krawiec .............................................. 13, 669

Krieglstein .............................................. 2793

Krystkowiakv ........................................... 413

Kubo ...................................................... 2947

Kück ......................................................... 561

Kudełko .................................................. 2807

Kumar .................................................... 1413

Kumaresan ............................................. 2499

Authors Index


Kunze ..................................................... 1213

Kuwazawa .............................................. 2753

Kvyatkovskiy ........................................... 893

Kwaśniewski .................................... 137, 155

Lankinen ................................................... 907

Larinkari ................................................. 1423

Lee, S. G. ................................................ 1437

Lee, S. M. ............................................... 1437

Leon ....................................................... 2357

Leszczyńska-Sejda ................................. 1803

Leuprecht ................................................ 1663

Li, S. ........................................................... 89

Li, Z. ....................................................... 2897

Lillo ........................................................ 1449

Lin ............................................................ 961

Lindgren ................................................... 699

Litwinionek ............................................ 1803

Llorca ..................................................... 2071

López, C. .......................................... 919, 931

Łoś ............................................................ 425

Lu ............................................................. 587

Lübbe ....................................................... 387

Ludwig ................................................. 49, 65

Łukomska ................................................. 425

Lundström .............................................. 1959

Luszczkiewicz ........................................ 2655

Lyu ......................................................... 1437

Maccagni ................................................ 1973

Mackay ................................................... 1095

Mackey ............................................. 649, 945

Maldonado ............................................. 2829

Malec ....................................................... 255

Malek ..................................................... 2397

Mamala .................................... 119, 137, 155

Manafi .......................................... 1725, 2575

Mańka .................................................... 2845

Mansikkaviita ........................................ 1183

Mariscal ................................................... 749

Martin .................................................... 1899

Martínez, M. .......................................... 1355

Martínez, E. ............................................. 931

Mathur .................................................... 2499

Matsumoto ............................................. 2947

Matusewicz .............................................. 961

Mayhew .................................................. 1983

McKenna .......................................... 971, 987

Meichi ............................................ 823, 1313

Mejias .................................................... 1899

Mejías J. ................................................. 2747

Mena ...................................................... 2283

Menge ...................................................... 173

Meskers .................................................. 2159

Meyer-Wulf ............................................ 2971

Michels ............................ 185, 191, 207, 217

Miczkowski .............................................. 669

Miettinen ................................................ 1003

Miki .............................................. 1737, 1753

Minnaar .................................................. 1999

Authors Index


Mishra .................................................... 2307

Moats ........................ 1463, 1367, 1379, 1483

Möller ..................................................... 1495

Money .................................................... 2517

Montes .................................................... 2017

Morales ................................................... 1617

Moran ............................... 185, 191, 207, 217

Morgenstern ............................................. 533

Morimitsu ............................................... 1511

Motomura ................................................. 779

Moyano ........................................ 1013, 1095

Muller ..................................................... 1771

Müller ........................................................... 3

Murray ...................................................... 547

Muthuraman ........................................... 1063

Nagai ............................................ 1025, 1521

Nagaraj ................................................... 2453

Nakamura ............................................... 2923

Nakano ................................................... 1531

Narita ...................................................... 1345

Naumann .................................................. 231

Nees ........................................................ 2137

Nematollahi ............................................ 2433

Newman ........................................... 971, 987

Nexhip ........................ 811, 1297, 2327, 2415

Nicol ..... 1281, 1559, 1699, 1737, 1753, 1771

Nieminen ................................................ 1545

Nikoloski ................................................ 1559

Nikus ...................................................... 1569

Nisbett .................................................... 1585

Nuñez ..................................................... 2988

Ogi ......................................................... 2781

Okada ..................................................... 1403

Olguín .................................................... 2189

Orhan ....................................................... 399

Orr .......................................................... 2137

Oshiumi .................................................. 1511

Ospanov ................................................... 893

Oterdoom ............................................... 1213

Oue ......................................................... 1531

Özdeniz .................................................. 2765

Padilla .......................................... 2017, 2443

Palacios, J. ............................................. 2029

Palacios, M. ................................... 319, 1173

Palmu ..................................................... 1545

Pankewitz ............................................... 2267

Parada, F. ...................................... 465, 1273,

Parada, R. ............................................... 2637

Parhar ..................................................... 1983

Park ........................................................ 2483

Parra ............................................. 1273, 2637

Partington ................................................. 615

Patzelt .................................................... 2621

Paulitsch ................................................... 863

Peacey .................................................... 1035

Peart ....................................................... 2453

Peippo ...................................................... 907

Pérez, O. ......................................... 439, 2171

Authors Index


Pérez, C. ................................................. 2693

Perkins .................................................... 1379

Pesl ........................................................... 863

Ping .......................................................... 823

Plascencia ......................................... 919, 931

Plewka ...................................................... 425

Polfliet .................................................... 2295

Ponce .............................................. 465, 2189

Poplar ..................................................... 2327

Potesser .................................................. 1051

Potulska .................................................. 2859

Poulin ..................................................... 2829

Prayoga ................................................... 1143

Prengaman .............................................. 1593

Pyra ........................................................ 2807

Raabe ...................................................... 1687

Raaber .................................................... 1247

Rameshni .................................................. 477

Ramírez .................................................. 1599

Rana ....................................................... 2873

Ranasinghe ............................................. 1063

Ranjbar ......................................... 1725, 2575

Rantala, A. .............................................. 1423

Rantala, J. ............................................... 1569

Ravishankar ............................................ 2453

Reed ....................................................... 1079

Reeves .................................................... 2765

Reghezza ................................................ 1273

Renner .................................................... 2283

Reuter, M. A. ........................................... 961

Reyes ........................................................ 919

Riedle ................................................... 49, 65

Ríos ........................................................ 1599

Riquelme ................................................ 2829

Risopatron .............................................. 2215

Riveros ................................................... 1213

Robinson ............................ 1379, 1617, 1635

Robles .................................................... 1999

Robles-Vega ........................................... 2473

Rodríguez ............................................... 2747

Rojas, F. ....................................... 1013, 1095

Rojas, N. ................................................ 1259

Ronsyn ................................................... 2295

Rooyen, van ........................................... 1999

Rosales ......................................... 1013, 1095

Roth .......................................................... 379

Roux ....................................................... 1999

Ruhs ....................................................... 1233

Ruimin ................................................... 2527

Ruiz .............................................. 2017, 2443

Rüşen ..................................................... 1115

Russell, M. ............................................... 497

Russell, R. .............................................. 1063

Sachajdak ............................................... 2845

Sadri ....................................................... 1123

Safe .......................................................... 497

Saitoh ..................................................... 2781

Sakala ..................................................... 1413

Authors Index


Salgado ................................................... 2693

Salomon de Friedberg, S. ......................... 685

Salomon-de-Friedberg, H. ...................... 1983

Samei ........................................................ 247

Sánchez ................................................... 2029

Sánchez-Corrales ................................... 2473

Sandoval ....................................... 1617, 1635

Santo ........................................................ 477

Sarswat ................................................... 1649

Sarvinis .................................................... 587

Sasai ....................................................... 1025

Sato ........................................................ 1143

Schaffie ........................................ 1725, 2575

Schillinger .................................................. 49

Schlitt ..................................................... 2039

Schlutzkus ................................................ 517

Schmid ................................................... 2992

Schneider .................................................. 271

Schuhmacher .......................................... 2397

Schwarz .................................................... 533

Scriba ..................................................... 1783

Shah .............................................. 2381, 2499

Shameli .................................................. 1123

Shibata .................................................... 2923

Shibayama .............................................. 2883

Shimokawa ............................................. 1345

Shin .......................................................... 831

Shuto ...................................................... 2781

Śmieszek .......................................... 255, 669

Smith, A. ................................................ 2267

Smith, D. .................................................. 465

Smyrak ..................................................... 119

Snowdon ................................................ 1079

Sobociński .............................................. 2807

Soderstrom ............................................. 2059

Sohn ....................................................... 2483

Soto ........................................................ 2059

Spoljaric ................................................. 1051

Staley ..................................................... 1379

Stantke ................................................... 1663

Stelter, M. .................................... 1675, 1687

Stodulski .................................................. 631

Stoilov ...................................................... 247

Ströder .................................................... 1173

Stuart ...................................................... 1559

Subiabre ................................................... 301

Sue Yek .................................................. 1253

Sun ........................................................... 961

Sunyer .................................................... 2071

Suontaka ................................................ 1155

SuPing .................................................... 1167

Szczygieł ................................................ 2845

Takahashi ................................................. 779

Takasaki ................................................. 2883

Takehara ................................................. 2753

Talja ....................................................... 1183

Tanaka .................................................... 1199

Thakre ................................ 2307, 2381, 2499

Authors Index


Tiroi ........................................................ 1143

Tjandrawan ............................................. 1699

Toda ......................................................... 779

Tohn ......................................................... 587

Tomioka ................................................. 1531

Tongamp ................................................ 2883

Topkaya .................................................. 1115

Torres ..................................................... 2071

Torsner ................................................... 1857

Tsunekawa ............................................. 2753

Tsymbulov ............................................... 793

Tumen-Ulzii ........................................... 2897

Tuuppa ................................................... 1545

Twidwell ................................................ 2589

Uhrie ....................................................... 1379

Ulloa ......................................................... 465

Umeda ........................................................ 89

Unger ...................................................... 1635

Urbanowski .............................................. 631

Utigard ..................................................... 931

Vargas .................................................... 2637

Veenstra ........................................... 971, 987

Vega ....................................................... 1783

Velásquez-Yévenes ...................... 1737, 1753

Vera ........................................................ 1355

Vermeiren ............................................... 2295

Villafañe ................................................... 547

Villar, del ............................................... 2829

Villavicencio .......................................... 1355

Viñals ..................................................... 2071

Virtanen ................................................. 1545

Voermann ......................................... 971, 987

Wada ...................................................... 1511

Walck ....................................................... 351

Walker .................................................... 2517

Walkowicz ............................................... 119

Wang ...................................................... 1393

Warczok ................................................. 1213

Watanabe ............................................... 2959

Wei ......................................................... 1167

Weiqun ................................................... 2527

Wenzl ................................. 1233, 1247, 1713

Westhoff ................................................ 2971

Wierzbicki (Marchewka) ......................... 255

Wilbrand .................................................. 271

Wilkomirsky ................................ 1259, 1273

Willbrandt ................................................ 561

Wilson .................................................... 2415

Wódka .................................................... 2673

Wood ...................................................... 2873

Woodling ................................................. 577

Wraith ...................................................... 945

Wu .............................................................. 65

Xie .......................................................... 2087

Yamaguchi, K. ....................................... 1287

Yamaguchi, Y. ....................................... 1521

Yamamoto .............................................. 1025

Yañez ..................................................... 2059

Authors Index


Yunxiao .................................................. 1313

Zagorodnyaya ........................................... 893

Zambika ................................................. 1413

Zamorano ............................................... 2357

Zárate ..................................................... 2693

Zauner .................................................... 1051

Zauter ........................................................... 3

Zavala ..................................................... 2235

Zeidabadi ................................................ 2433

Zekovic .................................................. 2249

Zeng ......................................................... 577

Zhang ..................................................... 2897

Zhao ............................................... 811, 1297

Zhi .......................................................... 2527

Zhixiang ................................................. 2527

Zhuo ............................................... 823, 1313

Zimba ..................................................... 1413

Zúñiga .................................................... 1273



Keywords Index



Keywords Index

3D-model ..................................... 1233, 2357

3D-drawing ............................................ 1233

abandonment ............................................ 285

AC + DC ................................................ 1355

acid consumption ......................... 2693, 2711

acid mist ................... 1237, 1271, 1617, 2137

acid treatment ........................................... 601

acidophilic thermophile .......................... 2781

acoustic emission ................................... 1123

acousto ultrasonic-echo (AU-E) ............. 1123

additives ................................................. 1687

African Copperbelt ................................. 1999

Ag-Se-O chemical potential diagram ..... 1293

alkaline media ........................................ 2747

alunite ..................................................... 2071

ammonium perrhenate ............................ 1803

amorphous coating ................................. 1511

annular jet ............................................... 2483

anode .................................. 1559, 1617, 1635

anode behavior ....................................... 1495

anode furnace ........................................... 183

anode passivation ............... 1367, 1463, 1495

anode quality .......................................... 1327

anode refining ........................................ 2327

anode slime ............................................ 1495

anode slimes ................................. 1393, 2159

anode spacing ......................................... 1521

anodes ........................................... 1423, 1699

anodic deposition ................................... 1511

anodic polarization ................................. 1725

antimicrobial .................... 185, 191, 207, 217

antimony ...................................... 1495, 1599

arsen ....................................................... 1599

arsenic ..................... 1035, 1495, 1783, 2071,

.............................................. 2883, 2947, 517

arsenic removal ........................................ 669

As ........................................................... 2443

ASR treatment ....................................... 2563

Ausmelt Top Submerged Lance (TSL) .... 961

automated production ............................ 2295

baffle design ........................................... 2499

base metals ............................................. 2453

basic theory ............................................ 2527

bath slopping ............................................ 839

bath smelting ............................................ 945

BB05xi ....................................................... 81

BB95 .......................................................... 81

bendability ................................................. 81

beneficiation ............................................. 285

bio leaching ............................................ 2935

bioleaching ............................................. 2575

bismuth .......................... 89, 255, 1495, 1599

bismuth removal .................................... 1941

black shales ............................................ 2655

bleed treatment ....................................... 1599

boiler tube ................................................ 907

Keywords Index


bottle roll tests ........................................ 2039

bronze ................................................. 81, 255

bronzes ......................................................... 3

bubble size control ................................. 2829

bubble size distribution .......................... 2829

bubble size measurement ....................... 1237

buoyancy power ....................................... 839

business optimization ............................. 2127

by-product treatment .............................. 2159

C19400 ....................................................... 81

C52400 ....................................................... 81

cables audio-video .................................... 119

calcium ferrite slag ......................... 811, 1297

campaign life .................................... 699, 987

campaign life ............................................ 987

CaO ........................................................ 2923

capital ..................................................... 2101

cast cooling element ................................. 987

castability ................................................. 255

cast-in pipe coils ....................................... 987

casting ................................................ 65, 721

cathode ................................................... 1253

cathode production capacity ................... 1143

cathode quality ................... 1271, 1413, 1635

cathodes .................................................. 1423

cement .................................................... 2873

centrifuge ............................................... 2737

CESL ...................................................... 1983

CFD ...................................... 931, 1003, 2499

CFD design ............................................ 2357

CFD modeling ........................................ 2845

CFD simulation ...................................... 1313

CFD-simulation ..................................... 2397

chalcopyrite ............. 1725, 1737 ,1753, 1771,

.................................. 1815, 1845,1909, 1959,

.................................. 2017, 2753, 2765, 2781

channel induction furnace ........................ 173

Chile ............................................... 439, 1213

Chilean copper industry ......................... 2171

chloride .................... 1737, 1753, 1771, 1857

chloride leaching .................................... 1959

chloride media ........................................ 1649

chlorination ............................................ 1403

chronopotentiometry .............................. 1367

Clyde ...................................................... 2517

CO2 .......................................................... 713

CO2 emissions ........................................ 2117

cobalt ...................................................... 1593

cobalt refinery ........................................ 1585

CoJet process ......................................... 2327

colemanite .............................................. 1115

column tests ........................................... 2039

combustion accelerated .......................... 1025

community engagement ........................... 351

compressed air ......................................... 271

computation ........................................... 2845

concentrate ................................... 2017, 2517

concentrate blending .............................. 2381

concentrate burner .................................. 1025

Keywords Index


condition monitoring .............................. 1123

connector .................................................... 81

consumption ............................................. 301

continous casting ........................................ 49

continuous converting .............................. 793

contirod .................................................... 119

Contirod®

................................................. 137

convergent – divergent nozzle ................. 931

converter slag ........................................... 863

converters ................................................. 839

converting ....................................... 649, 2543

conveying ............................................... 2517

cooling ...................................................... 907

cooling media ......................................... 1247

cooling system .......................................... 699

cooling technology ................................. 1247

cooling water ............................................ 271

copper ........ 31, 271, 285, 301, 365, 379, 425,

.......................... 439, 517, 533, 561, 615, 631,

.......................... 649, 721, 749, 761, 793, 863,

........................ 1063, 1259, 1281, 1327, 1345,

........................ 1423, 1483, 1531, 1545, 1559,

........................ 1569, 1599, 1649, 1675, 1713,

........................ 1783, 1845, 1871, 1899, 1925,

........................ 1983, 2039, 2071, 2433, 2543,

.................................. 2589, 2673, 2693, 2737,

................................... 2807, 2873, 2897, 2935

copper alloys .............. 81, 185, 191, 207, 217

copper business ...................................... 2563

copper cathode ....................................... 1585

copper concentrate ............ 1095, 1183, 1725,

....................................................... 2575, 2747

copper concentrates .............. 893, 1013, 1273

copper concentrator ................................ 2235

copper converting ........................... 811, 1297

copper cooler ............................................ 971

copper demand ....................................... 2215

copper electrorefining ....... 1307, 1367, 1437,

...................................................... 1663, 1687

copper electrowinning .................. 1379, 1585

copper extraction technology ................. 1167

copper fabrication .................................. 2563

copper global production ....................... 2215

copper industry ....................................... 1167

copper infiltration .................................... 247

copper losses .......................................... 2637

copper losses to slag .............................. 1115

copper market ........................................ 2215

copper material properties ....................... 231

copper matte ................................... 931, 1287

copper matte smelting ............................ 1115

copper mine projects .............................. 2215

copper nanoparticles .................................. 21

copper ores ......................... 2453, 2655, 2859

copper production .................................. 1511

copper recovery .......... 399, 1095, 2029, 2781

copper refining ................... 1521, 1941, 2397

copper refining electrolysis .................... 1495

copper resources .................................... 2883

copper rod .................................................. 99

copper rod plasticity examination by AR test .. 13

Keywords Index


copper sector .......................................... 2249

copper selenide ....................................... 1815

copper slag ............................................. 2283

copper smelter ........................................ 2415

copper smelting ........................................ 731

copper solvent extraction ....................... 2059

copper telluride ...................................... 2473

copper trade ............................................ 2215

copper white metal ................................. 2283

copper-cobalt concentrate ........................ 587

corrosion ............................... 699, 1593, 1857

cost effective .......................................... 1143

cost escalation ........................................ 2101

cost function ........................................... 2171

cost reduction ............................... 1051, 2369

cost saving .............................................. 1585

costs ........................................................ 2101

crane operation ....................................... 1569

critical blast flow rate ............................... 839

CRT ........................................................ 2959

crucible melting furnace ........................... 173

crud treatment ........................................ 2737

crystalline ............................................... 2947

Cu-40Zn brass ............................................ 89

CuAg0.1 ..................................................... 99

CuAg0.10 ................................................. 137

Cu-ETP1 .................................................... 99

CuFe2P ....................................................... 81

CuMg0.2 ..................................................... 99

CuMn20Ni20 ............................................... 3

Cu-OFE ...................................................... 99

cupric ion leaching ................................. 1815

current ...... 1281, 1355, 1449, 1521, 1687, 1713

current efficiency ........ 1437, 1649, 1687, 1713

CuSn10 ...................................................... 81

cyanide ................................................... 2087

cyclone reactor ......................................... 879

data interoperability ............................... 2127

decopperized anode slime smelting ....... 1293

development ................. 387, 721, 1233, 1247

diamond ................................................. 2873

direct current .......................................... 1437

direct emissions ..................................... 2117

direct to blister furnace .......................... 1063

discourse communities ............................. 351

distribution ............................................. 1281

distribution coefficient ........................... 1495

distribution ratio of minor elements ...... 2415

doré matte refining ................................. 1293

doré metal refining ................................. 1293

dryer ....................................................... 2517

dust ........................................................... 893

dust emissions .......................................... 517

dusting rate ............................................. 1025

dynamic modeling .................................. 2235

ecobronze ................................................... 81

economic ................................................ 1999

Ecuprex .................................................. 1973

Keywords Index


EE-core ..................................................... 173

efficiency .................................................. 333

EIS .......................................................... 1909

electric furnace ......................................... 961

electric home appliances ........................ 2959

electro-chemical ....................................... 425

electrochemical dissolution .................... 1909

electrochemical potential ....................... 2747

electrochemistry ..................................... 1845

electrolysis ................................................ 399

electrolyte ............................................... 1307

electrolyte composition .......................... 1463

electrolyte temperature ........................... 1675

electrolyte treatment ............................... 1663

electrolytic .............................................. 1599

electrolytic copper ...................................... 13

electrolytic copper powder ....................... 399

electrorefining .......... 1355, 1531, 1675, 1713

electrorefining process ........................... 1327

electrowinning ......... 1237, 1253, 1281, 1355,

.................................. 1449, 1483, 1545, 1559,

.................................. 1617, 1635, 1649, 1699,

................................... 1713, 1973, 2137, 2189

emission ................................................... 713

emissions .......................................... 477, 649

enargite ..................... 1035, 1783, 1983, 2883

energy ..... 301, 561, 649, 713, 1617, 1635, 2543

energy efficiency .................... 185, 271, 2117

energy optimization .................................. 497

energy saving ...... 779, 831, 1449, 1511, 2369

energy saving ........................................... 831

engineering design ................................... 547

environment ... 285, 533, 561, 577, 721, 2543

environment protection .......................... 1167

environmental engineering ....................... 547

environmental impact .............................. 547

environmental improvement .................... 413

environmental preservation .................... 2563

environmental protection ....................... 2369

EPMA ............................................ 811, 1297

ethyl xanthate ......................................... 2765

EU regulations ......................................... 365

evaporation ............................................. 1663

excellence ................................................. 309

exhaust gas system ................................... 413

expansion ............................................... 1143

experimental .......................................... 1271

experimental estimation ........................... 919

extraction ............................. 893, 1259, 2721

ferric arsenate ......................................... 2947

ferric sulphate leaching .......................... 1815

ferrous oxidation .................................... 2947

Finite Element Analysis (FEA) ................ 587

fire refining .............................................. 649

flameless combustion ............................. 1051

flash dryer ................................................ 779

flash furnace ........................................... 1079

flash smelting ..... 669, 699, 1313, 2381, 2415

flash smelting furnace ............ 823, 879, 1155

Keywords Index


flash smelting process ............................ 2499

flotation ................... 2189, 2433, 2589, 2655,

.............................................. 2765, 2845, 2859

flotation column control ......................... 2829

flotation reagents optimization ............... 2453

flow conditions ....................................... 1675

fluid bed ................................................... 587

fluid flow .................................................. 823

fluidization ............................................. 1025

flux ........................................................... 685

fluxes ........................................................ 893

forging ...................................................... 155

freezing line ............................................ 2307

frit-and-sleeve sparger ............................ 2829

froth flotation ......................................... 2453

fugitive emission ...................................... 685

fugitive emissions .......................... 517, 2971

furnace campaign life ............................... 971

furnace inspection .................................. 1025

furnace integrity ............................... 971, 987

gas cleaning .............................................. 497

gas flow .......................................... 879, 1003

gas sparging ............................................ 1545

gas-gas cooler ......................................... 1173

gas-solid ................................................. 2483

glue ......................................................... 1687

gold ........................... 1259, 1393, 1403, 1783

granulation ............................................. 2283

graphite particles ........................................ 89

grinding ........................................ 2189, 2621

gypsum ..................................................... 465

hardrock mining ..................................... 2873

harvesting plan ....................................... 1569

Hatch furnace technology ........................ 971

head grade .............................................. 2171

health ........................................................ 561

heap leach .............................................. 2711

heap leach modeling .............................. 2703

heap leaching ......................................... 2693

hearth accretion ...................................... 2327

heat and mass balance ............................ 1063

heat loss ................................................. 2307

heat recovery .................................. 497, 1003

heat transfer ............................................ 2307

heat treatment ........................................... 155

heavy metals ............................................. 517

heavy oil consumption ............................. 779

high current density ...................... 1545, 1675

high pressure grinding rolls ................... 2873

high speed flow ...................................... 1079

high speed pressure measurement .......... 1079

high-arsenic ............................................ 1273

historic R+D ............................................. 945

history .................................................... 2721

hospital superbugs .................................... 217

hospital-acquired infection .............. 207, 217

HPGR ........................................... 2621, 2873

Hybrid flotation ...................................... 2793

Keywords Index


HydroCopper®

........................................ 1959

hydrogen reduction ..................................... 21

hydrogen sulfide ..................................... 1973

hydrometallurgical ....................... 1403, 1871

hydrometallurgy ...... 1393, 1783, 1845, 2017,

................................... 2039, 2189, 2693, 2737

image analysis ........................................ 1237

IMC control ............................................ 2829

impact ....................................................... 387

implementation ...................................... 2499

improvement ........ 387, 699, 713, 1213, 1413

improvements ......................................... 1379

impure anodes ........................................ 1495

impurities .......................... 1327, 1463, 1663,

impurities in cathode copper ...................... 13

impurity management ............................. 2159

indirect emissions ................................... 2117

indoor air quality ...................................... 185

induction casting ........................................ 31

induction melting ....................................... 31

inductive heating ...................................... 173

industrial test .......................................... 1095

inertization ............................................. 2071

infection control ....................................... 207

information system ................................. 1423

infrared ................................................... 1123

injection .................................................. 2483

innovation ................................................ 577

intensified smelting process ................... 1313

investment .............................................. 1413

ion exchange .......................................... 1803

ionic liquids ........................................... 1247

iron ......................................................... 1379

iron-silicate slag ..................................... 1287

ISA 2000 ................................................ 1307

ISA process ............................................ 1521

ISASMELTTM

................ 615, 749, 907, 1327

ITER ............................................................. 3

jacket ...................................................... 2307

Kayser Recycling System (KRS) ........... 2971

kimberlite ............................................... 2873

Konkola Copper Mines .......................... 1413

k-ε turbulence model .................... 2397, 2499

La Coipa mine ........................................ 1035

laboratory practices testing ...................... 191

labour productivity ................................... 439

lance ....................................................... 2483

leaching ..... 893, 1585, 1725, 1737, 1753, 1771,

............................... 1783, 1973, 1983, 1999, 2029,

................................. 2189, 2655, 2721, 2753, 2897

lead ......................................... 631, 863, 1699

lead anodes ............................................. 1593

lead free brass .......................................... 255

lead removal ............................................. 669

lead-free ..................................................... 89

liberation ...................................... 2703, 2859

lime ................................................ 465, 2765

limestone .................................................. 685

liquidus .................................................... 731

Keywords Index


liquidus temperature ............... 761, 811, 1297

LIX 7950 ................................................ 2087

LOI measurement ................................... 2433

Los Pelambres ........................................ 2793

low-grade ores ........................................ 2781

machinability .............................................. 89

macrosegregation ................................. 49, 65

maintenance optimization ...................... 1173

management ............................ 309, 319, 387,

management decisions ........................... 2127

management system ............................... 1423

manganese .................................... 1593, 1699

manganese dioxide ................................. 1483

markets ................................................... 2543

material cycle ............................................. 81

material flow management ..................... 2341

material storage and pre-treatment ......... 2971

matte converting ....................................... 945

matte making .......................................... 2527

measurement .................................. 517, 1079

mechanical properties .............................. 255

metal ....................................................... 2087

metal ions ............................................... 1857

metal recovery ................................ 863, 1213

metallurgy ................................................ 319

metals ....................................................... 425

metastable Cu1+xSe phase .................... 1815

METTOP-BRX-Technology .................. 1713

micro alloys ................................................ 99

microstructure .......................................... 255

mineralogy from elements ..................... 2381

minerals applications ............................. 2621

minerals references ................................ 2621

miniaturization ........................................... 81

mining ............................ 285, 319, 333, 2807

mining industry ................................ 533, 577

mining projects ........................................ 547

minor element deportment ..................... 2415

mist suppression agents ......................... 1271

Mitsubishi process ..... 685, 1143, 1199, 2563

Mitsubishi smelting process .................. 2483

model ..................................................... 1649

model application ................................... 2381

modeling .................... 919, 1003, 1063, 2711

moderate thermophile ............................ 2575

modernization .......................................... 961

modified phosphinic acid formulation ... 2059

Molecular Recognition Technology (MRT)

................................................................ 1941

molybdenum ................................ 2059, 2589

molybdenum recovery ............................ 2029

MRSA ...................................................... 217

multicomponent ........................................... 3

multi-frequency current ......................... 1355

NaCl ....................................................... 2923

NaOH ..................................................... 2923

neutralizing .............................................. 465

new technology ...................................... 1871

Keywords Index


nickel ...................................... 793, 863, 1663

nitrate solution ....................................... 1687

nodulation .............................................. 1599

Non-Destructive Testing (NDT) ............ 1123

NSC ........................................................ 2589

numerical simulation ................................ 823

off-gas ........................ 601, 1003, 1173, 2357

off-gas cleaning ........................................ 601

off-gas duct .................................. 1173, 1233

off-gas flow ............................................ 2499

off-gas line ............................................. 1003

off-gas recirculation ............................... 1051

on-line particle size analysis .................. 2267

operating ................................................. 2101

operation ....................................... 1143, 1345

operational cost ........................................ 497

operational stability .................................. 477

operations ................................................. 533

optimization ....... 721, 879, 1003, 1569, 2575

optimizing operations ............................. 1379

OPUS ..................................................... 2267

ORP ........................................................ 1393

oscillation ............................................... 1367

O-SR process .......................................... 1199

oxidation .............................. 793, 2443, 2753

oxyfuel ................................................... 1051

oxygen .......................................... 1495, 2543

oxygen burner ......................................... 1051

oxygen evolution anode ......................... 1511

oxygen free copper ................................... 119

oxygen free silver copper ......................... 137

oxygen-enriched bottom blowing .......... 2527

oxygenfree copper ...................................... 31

panel data ................................................. 439

parallel electrolyte flow ......................... 1713

passivation ............................................. 1559

passivation mechanism .......................... 1463

PCR ........................................................ 1675

penetration depth .................................... 2483

permanent cathode ................................. 1521

permanent cathodes ................................ 2295

phase diagram .......................................... 731

phase diagrams ......................................... 761

pilot molten layer reactor ....................... 1273

pilot tests ................................................ 2039

pilot trials ................................................. 945

planning ................................................. 2295

plant continuity ...................................... 2127

plant structuring ..................................... 2137

pneumatic ............................................... 2517

polarization ............................................ 1531

polymer additive .................................... 1531

polymetallics .......................................... 2913

potential ............................. 1737, 1753, 1771

practice production ................................ 2527

precious metal ........................................ 1287

precious metals ........................................ 631

precipitation ................................. 1599, 1973

Keywords Index


pressure leaching .......................... 2017, 2673

pressure oxidation .................................. 2589

prices ...................................................... 2101

primary materials ................................... 1899

process advisor ....................................... 2381

process advisor software ........................ 1155

process control ............................. 1063, 1155

process heat balance ................................. 631

process modernization .............................. 631

process optimization .................... 2369, 2397

production ...................................... 477, 2807

production management system ............. 2341

productivity .......................... 713, 1449, 2543

profit ....................................................... 2807

project ..................................................... 1871

protection ............................................... 1559

P-S Converter ......................... 919, 931, 1183

public health registration .......................... 191

pyrite ................................. 1845, 2721, 2765,

pyritic ores ................................................ 945

pyrometallurgy .............. 379, 615, 749, 1115,

............................................. 1183, 2443, 2637

Radiant cooling chamber (RCC) ............ 2357

rate-controlling step ............................... 1959

reaction shaft .......................................... 1313

reagents .................................................. 1925

real time transparency ............................ 2341

recovery .......................................... 301, 1259

recyclabilityy .............................................. 81

recycle ...................................................... 425

recycling ..... 379, 2117, 2159, 2913, 2935, 2959

redox potential ............................. 1725, 2753

reduction ...................... 477, 793, 2397, 2753

refinery ....................... 721, 1345, 2295, 1253

refractory ................................................ 2307

refractory engineering .................. 1233, 1247

refractory wear ....................................... 1123

residence time distribution ....................... 919

residue .................................................... 1259

residue disposal ...................................... 1035

resistance .................................................. 699

resource utilization ................................. 1167

resources .................................................. 333

rhenium .................................................. 1803

rhenium recovery ..................................... 893

risk and uncertainty ................................ 2249

risk assessment ......................................... 365

roasting ........... 1035, 1259, 1783, 2029, 2721

robotic .................................................... 1253

rod mill ..................................................... 271

rolled lead calcium tin alloys ................. 1593

rotary furnace ........................................... 721

rotating cylinder electrode ....................... 399

RotoFeed®

...................................... 685, 1079

SAG milling ........................................... 2873

Sb ........................................................... 2443

scorodite ............................. 1983, 2923, 2947

scrap melting optimization .................... 2327

Keywords Index


secondary materials ...................... 1013, 1899

selective leaching ................................... 2883

sensitivity to process variables ............... 2765

settler ........................................................ 823

shale ore ................................................. 2673

short circuits ........................................... 1379

sick building syndrome ............................ 185

signal conductor ......................................... 99

silicate ...................................................... 761

silver ....................................................... 1259

simulation ..................................... 1281, 2295

simulation of solidification ........................ 49

slag ......................................... 379, 731, 2159

slag chemistry ........................................... 749

slag cleaning ............... 649, 1173, 1213, 2637

slag coating thickness ............................. 2307

slag treatment ........................................... 863

slags ................................................ 761, 2029

slide gate .................................................. 713

slime ....................................................... 1463

slurry analysis ......................................... 2267

smelter .................................................... 1167

smelter modernization ............................ 1327

smelting ............ 649, 1013, 1273, 2381, 2543

smelting capacity .................................... 1095

SO2 ................................................... 477, 601

SO2 treatment ........................................... 413

SO3 ........................................................... 601

social commitment ................................... 561

sodium hydrosulfide .............................. 2883

sodium tellurate ..................................... 2473

sodium tellurite ...................................... 2473

software development ............................ 1155

solar copper strips .................................... 231

solar copper tubes .................................... 231

solar flatpanel absorbers .......................... 231

solar heat .................................................. 231

solidification .............................................. 65

solidus .................................................... 2307

solubility ...................................... 1599, 2923

solvent extraction .... 1545, 1585, 1899, 1925,

.................................. 1999, 2087, 2137, 2189

southern Peru ................................. 615, 1327

space-time yield ..................................... 1545

special coated solar strip .......................... 231

spray forming ............................................... 3

stabilization ............................................ 2071

stainless steel cathodes ........................... 1307

standards .................................................. 285

steam dryer ............................................. 1183

steam generation ...................................... 831

steam selling ............................................ 831

steel ........................................................ 1857

stirring reactor ........................................ 1213

strategy ..................................................... 533

stripping ................................................. 1253

sulfide .................................................... 2453

sulfide ore ................................................ 587

Keywords Index


sulfuric acid .............................................. 601

sulphatation ............................................ 2499

sulphating roasting ................................... 587

sulphur ...................................................... 477

sulphuric acid ......................................... 1857

surface disinfection .................................. 207

surface tension ........................................ 1271

sustainability ..... 309, 319, 333, 387, 561, 2453

sustainable development ...... 351, 2137, 2249

tails leach ................................................ 2589

tankhouse ........................... 1307, 1423, 1569

tankhouse current efficiency .................. 1379

tankhouse design .................................... 1449

tankhouse purity control ......................... 1941

taphole ...................................................... 713

taphole acoustic monitoring ..................... 971

technogenic products .............................. 2935

technology ...................................... 425, 2845

Tellurium ................................................ 2473

temperature distribution ........................... 823

Teniente Converter ....................... 1013, 1095

ternary ........................................................ 65

thermal conductivity ............................... 2307

thermodynamics ......................................... 65

thermodynamics of precious metals ....... 1293

thin wires .................................................. 119

thiourea ......................................... 1531, 1687

tin ............................................................. 863

tin bronze ................................................... 49

tin bronzes .............................................. 3, 65

tin-coated ................................................... 81

titanium anode ....................................... 1483

Tongling, Jinchang ................................... 961

Top Submerged Lance (TSL) .................. 615

Top Submerged Lance (TSL) furnace ...... 907

traction equipment ................................... 155

translog ................................................... 2171

treatment and sorting ............................. 2959

trolley wire ............................................... 137

tube decoiling procedure .......................... 231

Tungsten skeleton .................................... 247

tuyere submergence .................................. 839

ultrasonic extinction ............................... 2267

Ultrasonic Spray Pyrolysis (USP) .............. 21

Upcast®

............................................ 119, 137

utilities ................................................... 2369

validation ..................................... 2381, 2499

Vaniukov converter .................................. 793

Vanyukov furnace .................................... 893

Variable Current .................................... 1437

vat leaching ............................................ 2039

volatilization .......................................... 2443

W/Cu compacts ........................................ 247

waffle cooler ............................................ 987

waste .............................................. 425, 2071

waste acid treatment ................................. 465

waste gas .................................................. 271

waste heat boiler ............................ 879, 2499

Keywords Index


waste heat recovery .................................. 779

water ................................ 301, 333, 533, 577,

water-cooling ......................................... 1247

wet chlorination ...................................... 1393

white metal ......................... 1095, 1259, 1273

Wiener model ......................................... 2829

wire rod ............................................ 119, 271

wireless measurement ............................ 2747

zinc ......................................................... 1899

zincex ..................................................... 1899


Copper Volume 7.pdf

Documents