You are on page 1of 502

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-Strae 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 ber 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)
Jol 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. Jrgen Zuchowski GDMB Gesellschaft fr 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,
Queens 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. Jrg 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. Jrgen Zuchowski GDMB Gesellschaft fr Bergbau, Metallurgie, Rohstoff-
und Umwelttechnik e. V.
(Copper2010 Organizing Committee Chairman)
Mareike Hahn GDMB Gesellschaft fr Bergbau, Metallurgie, Rohstoff-
und Umwelttechnik e. V.
Thomas Marbach GDMB Gesellschaft fr Bergbau, Metallurgie, Rohstoff-
und Umwelttechnik e. V.
Dipl.-Ing. Jens Harre GDMB Informationsgesellschaft mbH
Mareike Mller 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 7
th
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 conferences 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 worlds 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 industrys
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. Daz, 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 SO
2
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 21
st
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 worlds 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. Daz
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 Wdka
Mine to Heap in Mantoverde Anglo American Division 2693
Manuel Daz, Cristian Salgado, Carlos Prez, Cristian Alvayai,
Leonardo Herrera, Gabriel Zrate
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 Gonzlez S., Alejandra Mejas J., Claudia Rodrguez 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. zdeniz

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 Kudeko, Ph.D. Jacek Pyra, Ph.D. Eng. Jerzy Sobociski
Control of Bubble Size in a Laboratory Flotation Column 2829
Miguel Maldonado, Dr. Andr Desbiens, Dr. ric Poulin, Dr. Ren del Villar,
Alberto Riquelme
Flow Process in the Aerator of the Flotation Machine 2845
Preliminary Simulations
Adam Maka, Adam Fic, Andrzej Sachajdak, Ireneusz Szczygie
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 FeCl
3
Solutions with FeCl
2
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 Gutirrez, Dipl.-Ing. Jorge Ferreiro
Removal of Arsenic from Copper Smelter Wastewaters by 2988
Airlift Electrocoagulation
Dr. Henrik K. Hansen, Dipl.-Ing. Claudia Gutirrez, M.Sc. Patricio Nuez
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

Proceedings of Copper 2010 2542



Proceedings of Copper 2010 2543
Copper Sulphide Smelting:
Past Achievements and Current Challenges
Carlos M. Daz
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 SO
2
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.

Daz
Proceedings of Copper 2010 2544
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 Kelloggs 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 Queneaus 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 worlds 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.
Copper Sulphide Smelting: Past Achievements and Current Challenges
Proceedings of Copper 2010 2545
2 The Oxygen Age
The commercialization of Incos 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 SO
2
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. Furukawas 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 Outokumpus 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, SO
2
strong process off-gas streams amenable to sulphur fixa-
tion in acid or sulphur dioxide liquefaction plants.
Daz
Proceedings of Copper 2010 2546
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). Outokumpus 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
Copper Sulphide Smelting: Past Achievements and Current Challenges
Proceedings of Copper 2010 2547
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]. ISASMELT
TM
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,
Daz
Proceedings of Copper 2010 2548
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 SO
2
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 O
2
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 % O
2

enriched air for preheated low-oxygen enriched air (21-30 % O
2
, 1000 C), 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 % O
2
. The three ISASMELT
smelters that participated in the survey reported oxygen enrichment of the combustion gas in the 50
Copper Sulphide Smelting: Past Achievements and Current Challenges
Proceedings of Copper 2010 2549
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 % O
2
tonnage oxygen [24]. The Mitsubishi process operating numbers corre-
spond to the Gresik smelter that processes a 30.9 % Cu concentrate, using 99.5 % O
2
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
[O
2
%]
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
Daz
Proceedings of Copper 2010 2550
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: Nm
3
O
2
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 % O
2
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 [O
2
%] Nm
3
O
2
consumed/h/m
3
of reaction shaft
1949mid 1970s (Av 4 fces) Preheated air 55
Mid 1970smid 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
Copper Sulphide Smelting: Past Achievements and Current Challenges
Proceedings of Copper 2010 2551
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 countrys infrastructure. Commissioning
of modern copper smelters, and guaranteeing concentrate feed for these plants, became part of the
countrys economic development program. Years later, through a different path, Indias economy
also began expanding at high annual rates. As in the case of China, access to basic materials became
a key element of Indias 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.


Daz
Proceedings of Copper 2010 2552
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 19872008, 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.
Copper Sulphide Smelting: Past Achievements and Current Challenges
Proceedings of Copper 2010 2553
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 Chiles 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. Cochilcos and Marsdens
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
Cochilcos and Marsdens 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.
Daz
Proceedings of Copper 2010 2554
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
Cochilcos 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. Marsdens numbers are based on Miami ISASMELT Smelter operating data. Pirets 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 Pirets energy requirement for
flash smelting-PS converting given in Table 7 and the calculated energy consumption for Isasmelt
Copper Sulphide Smelting: Past Achievements and Current Challenges
Proceedings of Copper 2010 2555
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. Cochilcos number is the average for the three Chilean copper refi-
neries [37, 38]. Cochilcos 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
Daz
Proceedings of Copper 2010 2556
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) CO
2
emis-
sions are highly dependant on the type of fuel used [40, 43].
Table 10: Direct (heating fuel) and indirect (electric energy generation) CO
2
emissions for various
types of fuel.
Type of fuel Direct CO
2
emission [t/GJ] Indirect CO
2
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 CO
2
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 CO
2
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 CO
2
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 CO
2
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.
Copper Sulphide Smelting: Past Achievements and Current Challenges
Proceedings of Copper 2010 2557
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 SO
2
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.
Daz
Proceedings of Copper 2010 2558
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 countrys 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 authors 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-
thors 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
Gpfert and Carlos Landolt were instrumental in uncovering comprehensive studies conducted by
Cochilcos researchers on energy consumption and greenhouse emissions in Chiles 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] DAZ, 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. 27
th
-Sept. 1
st
1973, p. II.
[2] INCO Staff (QUENEAU, P.E. & SPROULE, W.K.) (1955): Oxygen Flash Smelting. JOM 7, 7:
742-750.
Copper Sulphide Smelting: Past Achievements and Current Challenges
Proceedings of Copper 2010 2559
[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] DAZ, 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. & DAZ, C. (2003): Development of the El Teniente Converter
technology. COPPER 2003-COBRE 2003, VOL. IV Pyrometallurgy of Copper (Book 2),
Ed. C. DAZ 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 Vedantas 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 Shans Houma smelter, Peoples 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. DAZ et al., MetSoc of CIM: 3-106.
[15] SOFRA, J. & MATUSEWICZ, R. (2003): Ausmelt technology copper production technology for
the 21
st
century, Ibid.: 157-172.
[16] ARTHUR, P.S. & PARTINGTON, P.J. (2007): Latest developments with copper ISASMELT
TM
.
Cu2007, Vol. III (Book 2), Ed. A.E.M. WARNER et al., MetSoc of CIM: 3-15.
Daz
Proceedings of Copper 2010 2560
[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. & DAZ, 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. DAZ 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. DAZ 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. & DAZ, 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.
Copper Sulphide Smelting: Past Achievements and Current Challenges
Proceedings of Copper 2010 2561
[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 Energa y Emisiones de gases de Efecto Invernadero de la Minera del Cobre de
Chile. Ao 2008 (Energy Consumption and Greenhouse Gas Emissions in Chiles 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 Todays 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 climtico
(The Chilean copper industry addresses climate change). Seminario Internacional Minera
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] DAZ, 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.

Proceedings of Copper 2010 2562


Proceedings of Copper 2010 2563
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 worlds 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 industrys
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
Proceedings of Copper 2010 2564
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
Chinas smelting capacity has surpassed that of Japan. Putting aside the discussion of a subsidized
growth, Chinas 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-
Sustainable Growth Strategy for Japanese Copper Business
Proceedings of Copper 2010 2565
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
Proceedings of Copper 2010 2566
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. MMCs 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
1990s, 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
J
a
p
a
n
E
U
A
s
i
a
N

A
m
e
r
i
c
a
S

A
m
e
r
i
c
a
E
n
e
r
g
y
-
E
f
f
i
c
i
e
n
c
y

Sustainable Growth Strategy for Japanese Copper Business
Proceedings of Copper 2010 2567
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
4
33
11
Copper Conc.
Coal Burner
ASR / SR
T
o

B
o
i
l
e
r
Matte
Slag
Cu conc.
Cu conc.
ASR / SR
4
33
11
Copper Conc.
Coal Burner
ASR / SR
T
o

B
o
i
l
e
r
T
o

B
o
i
l
e
r
Matte
Slag
Cu conc.
Cu conc.
ASR / SR


Kato
Proceedings of Copper 2010 2568
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 Naoshimas 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 Onahamas reverberatory furnace now the
worlds 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-slag
Lance air
(50-60 % O
2
)
Lance air
(30-40 % O
2
)
Lift
tanks
S furnace
CL furnace
C furnace
Anode furnace
Lime, sludge
& coolant (C-slag)
Pressurized
tanks
Matte,
slag
Matte
Scraps
Anode scrap
Casting machine
Blister copper
O gas Electrodes
Granulated slag Granulated C-slag
Return to S, C furnace Sale

Sustainable Growth Strategy for Japanese Copper Business
Proceedings of Copper 2010 2569
Figure 5: Schematic flow sheet of the incinerating and melting plant
Figure 6: Schematic flow of MMCs recycling business
Slag and metal Track scale
Quencher
Rotary 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
Smokestack Induced
draft Fan
Onerous recycling
materials
Automobile
shredder residue
Washing treatment plant
for fly ash from incinerators
Steam
Power plant of
Naoshima
Smelter&Refinery
Caustic soda
Secondary
combustion
chamber
Waste heat boiler
Copper bearing
sludge
Printed circuit
board scrap
Household
appliance residue

Cement
Plants
Copper
Smelters
Cement
Construction
materials and
concerete
Rolled copper
products
(Electric wire,
pipes and lead
frames)
Ingots
Creating
infrastructure
Thermal
Power plants
Automobiles Appliances Lifestyles
Industrial
wastes
(Coal ash, sludge,
and excavated soil)
Shredder dust,
battery and
copper scrap
Discarded
appliances
Waste
tires
Appliance
recycling
plants
Waste
plastics
Waste
plastics
Metal-bearing
residue
Metal-bearing
residue
No!
Landfill
100%
Recycle
Cement
Plants
Copper
Smelters
Cement
Construction
materials and
concerete
Rolled copper
products
(Electric wire,
pipes and lead
frames)
Ingots
Cement
Construction
materials and
concerete
Rolled copper
products
(Electric wire,
pipes and lead
frames)
Ingots
Creating
infrastructure
Thermal
Power plants
Automobiles Appliances Lifestyles
Industrial
wastes
(Coal ash, sludge,
and excavated soil)
Shredder dust,
battery and
copper scrap
Discarded
appliances
Waste
tires
Appliance
recycling
plants
Waste
plastics
Waste
plastics
Metal-bearing
residue
Metal-bearing
residue
Industrial
wastes
(Coal ash, sludge,
and excavated soil)
Shredder dust,
battery and
copper scrap
Discarded
appliances
Waste
tires
Appliance
recycling
plants
Waste
plastics
Waste
plastics
Metal-bearing
residue
Metal-bearing
residue
No!
Landfill
100%
Recycle
No!
Landfill
100%
Recycle

Kato
Proceedings of Copper 2010 2570
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. MSCs 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 MCIs various technologies and its solid customer base with MMCs oxygen-free copper and
alloy technologies, resulting in an overall strengthening of the MMC Groups 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. MMCs
integrated supply chain is shown in Figure 7.


Sustainable Growth Strategy for Japanese Copper Business
Proceedings of Copper 2010 2571
Figure 7: MMC groups 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 business
Copper mine Smelting & rening Fabrication & alloys Users
Smelting & rening 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
rened
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 plating Wire rod
Onahama
rened
230 kt
75 % of
260 kt
Scraps
60 kt
Rened sales
to others
90 kt

Shower
Bath
Kitchen
CO2 Heat Pump Unit
Water Tank Unit
Heat Heat
Exchanger Exchanger
Air
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
Exchanger Exchanger
Air
Water
High Copper Alloy Tubes are used for
heat exchanger in the heat pump unit
CO2
Cycle

Kato
Proceedings of Copper 2010 2572
It reduces energy consumption by one-third and reduces CO
2
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 industrys 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 todays 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 todays 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.

Proceedings of Copper 2010 2573












Mineral Processing

Proceedings of Copper 2010 2574



Proceedings of Copper 2010 2575
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
Proceedings of Copper 2010 2576
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 2
4
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 2
SiO

2 3
Al O
. . LOI 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

Optimization of Copper Concentrate Bioleaching
Proceedings of Copper 2010 2577
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 4
2
NH SO
4 2
.7 MgSO H O
2 4
K HPO KCl
( )
3 2
2
. 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 2.410
8
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 H
2
SO
4
.
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.
Ahmadi, Ranjbar, Schaffie, Manafi
Proceedings of Copper 2010 2578
2.4 Design of experiments
In the present research, a two-level full factorial 2
4
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 2
4
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
Optimization of Copper Concentrate Bioleaching
Proceedings of Copper 2010 2579
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.
Ahmadi, Ranjbar, Schaffie, Manafi
Proceedings of Copper 2010 2580


(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
Fe
R T
ORP or E E
n F Fe
+
+


= +


Optimization of Copper Concentrate Bioleaching
Proceedings of Copper 2010 2581
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 6
3 2 6 ( ) ( ) 8 Fe X HSO H O XFe 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
N
u
m
b
e
r

o
f

B
a
c
t
e
r
i
a

(
C
e
l
l
s
/
m
L

1
*
1
0
^
7
)

A
B
Ahmadi, Ranjbar, Schaffie, Manafi
Proceedings of Copper 2010 2582
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 H
2
SO
4
.
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 4
3 2
Sulphur oxidizing microorganisms
S 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.

Optimization of Copper Concentrate Bioleaching
Proceedings of Copper 2010 2583









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
N
u
m
b
e
r

o
f

B
a
c
t
e
r
i
a

(
C
e
l
l
s
/
m
L

1
*
1
0
^
7
)

A
B
Ahmadi, Ranjbar, Schaffie, Manafi
Proceedings of Copper 2010 2584
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 Ag
2
S 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 Ag
2
S (Equation 10). In bioleaching systems, bacteria oxidize sulphure and ferrous ions pro-
duced in Equation 10 and accelerate the rate of reaction [12].
S Ag Fe Cu Ag CuFeS
2
2 2
2
2 4 + + +
+ + +
(9)
+ + +
+ + +
2 0 3
2
2 2 2 Fe S Ag Fe S Ag
(10)
It was also found that Ag
2
S 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
N
u
m
b
e
r

o
f

B
a
c
t
e
r
i
a

(
C
e
l
l
s
/
m
L

1
*
1
0
^
7
)

A
B

Optimization of Copper Concentrate Bioleaching
Proceedings of Copper 2010 2585
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, 249257.
Ahmadi, Ranjbar, Schaffie, Manafi
Proceedings of Copper 2010 2586
[5] NORRIS, P.R., BURTON, N.P. & FOULIS, N.A.M. (2000): Acidophiles in bioreactor min-
eral processing- Extremophiles, 4, 7176.
[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, 4756.
[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, 19361943.
[8] OLSON, G.J. & CLARK, T.R. (2004): Fundamentals of metal sulfide biooxidation. Mining
Engineering, 56, 4046.
[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, 100111.
[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 68C. 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, 77101, 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] LPEZ-JUREZ, A., GUTIRREZ-ARENAS, N. & RIVERA-SANTILLN, R.E. (2006):
Electrochemical behavior of massive chalcopyrite bioleached electrodes in presence of silver
at 35 C. Hydrometallurgy, 83, 6368.
[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, 3138.
Optimization of Copper Concentrate Bioleaching
Proceedings of Copper 2010 2587
[18] HIROYOSHI N., MIKI, H., HIRAJIMA, T. & TSUNEKAWA, M. (2001): Enhancement of
chalcopyrite leaching by ferrous ions in acidic ferric sulfate solutions. Hydrometallurgy, 60,
185197.
[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, 17751781.
[23] DAS, T., AYYAPPAN, S. & CHAUDHURY, G.R. (1999): Factors affecting bioleaching ki-
netics of sulfide ores using acidophilic micro-organisms. BioMetals, 12, 110.

Proceedings of Copper 2010 2588



Proceedings of Copper 2010 2589
Industrial NSC Pressure Oxidation of Combined
Copper and Molybdenum Concentrates
Corby G. Anderson
a
, Todd S. Fayram
b
, Larry G. Twidwell
c

a
Allihies Engineering Incorporated
b
Continental Metallurgical Services
c
Montana 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
Proceedings of Copper 2010 2590
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)
+ 2HNO
3

(aq)
+ 3H
2
SO
4

(aq)
3MeSO
4
+ 3S
(s)
+ 2NO
(g)
+ 4H
2
O (1)
However, it has been postulated and confirmed that the actual reaction species is NO
+
and not NO
3
-

(Anderson 1992, Anderson 1996, Baldwin 1996, Gok 2009). The addition of or presence of NO
2
-

instead of NO
3

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.
NaNO
2

(aq)
+ H
+
HNO
2

(aq)
+ Na
+
(2)
Nitrous acid further reacts to form NO
+
.
HNO
2

(aq)
+ H
+
NO
+

(aq)
+ H
2
O (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, H
2
ref.)
Fe
3+
Fe
3+
+ e
-
Fe
2+
0.770 V
HNO
3
NO
3
- + 4H
+
+3e
-
NO + 2H
2
O 0.957 V
HNO
2
NO
2
-
+ 2H
+
+ e
-
NO + H
2
O 1.202 V
O
2

(g)
O
2
+ 4H
+
+ 4e
-
2H
2
O 1.230 V
Cl
2

(g)
Cl
2 (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
Proceedings of Copper 2010 2591
NO
(g)
+ O
2

(g)
2NO
2

(g)
(5)
2NO
2

(g)
2NO
2

(aq)
(6)
2NO
2

(aq)
+ 2NO
(aq)
+ 4H
+
4NO
+

(aq)
+ 2H
2
O (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
+
+ O
2

(g)
2Me
2+

(aq)
+ 2S + 2H
2
O (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.
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2592
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.
Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2593
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.
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2594
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 Software
Original Scale
Cu Recovery, %
97.85
22.6
X1 = E: Reaction Time
X2 = D: Temp
Actual Factors
A: Grind Time = 10
B: Solids = 10
C: 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
e
c
o
v
e
r
y
,

%


E: Reaction Time D: Temp

Design-Expert Software
Original Scale
Mo Recovery, %
84.25
10.24
X1 = A: Grind Time
X2 = D: Temp
Actual Factors
B: Solids = 10
C: Initial Acidity = 25.00
E: 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
e
c
o
v
e
r
y
,

%


A: Grind Time
D: Temp

Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2595
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 Software
Original Scale
Re Recovery, %
87.92
0
X1 = E: Reaction Time
X2 = B: Solids
Actual Factors
A: Grind Time = 10.00
C: Initial Acidity = 75.00
D: 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
e
c
o
v
e
r
y
,

%


E: Reaction Time B: Solids

Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2596
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 [10
6
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 [10
6
t] Grade
Cu % Mo %
North 18.166 0.271 0.005
South 25.593 0.278 No Data
Total 41.759 0.275 ---


Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2597

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
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2598

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
Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2599
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.
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2600
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.
Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2601
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 H
2
SO
4
, 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/m
2
. The actual meas-
ured temperature of the electrolyte and current density were 31.1 C and 313 A/m
2
. 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 H
2
SO
4
O:A Ratio- Extract 1.0-1.4:1 Strip 1.5: 1
Stage efficiency Extract 95 % Strip 95 %

Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2602
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 Kennecotts 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.
Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2603
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/m
2
/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
Thickener U/F density 50 % solids
Flux rate t/m
2
/h 0.52
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
Thickener U/F density 50 % solids
Flux rate t/m
2
/h 1.0
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.
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2604
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 P
80
of 325 mesh)
Flux rate t/m
2
/h 0.487
NSC POX Material (includes any vertical mill ground material to include leach residues)
(Assumes 65 % solids, product has a P
80
of 10 micron)
Flux rate t/m
2
/h 1.0
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

Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2605
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-
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2606
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 P
65
of 74 micrometer for the
downstream processes of leaching and/or flotation.
Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2607
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 m
3
/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 m
3
/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.
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2608
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 m
3
/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 m
2
, and the cathodes will be pulled on a seven-day stripping cycle. An average current density of
258 (maximum 275) A/m
2
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 m
2
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 P
80
of
10 m. 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
Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2609
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 m
2
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-
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2610
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
Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2611
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
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2612
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.
Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2613
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.
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2614
[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.
Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2615
[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.
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2616
[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.
Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2617
[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.
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2618
[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.
Industrial NSC Pressure Oxidation of Combined Copper and Molybdenum Concentrates
Proceedings of Copper 2010 2619
[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 Companys 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.
Anderson, Fayram, Twidwell
Proceedings of Copper 2010 2620
[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.

Proceedings of Copper 2010 2621
HPGRs in Copper Ore Comminution
A Technology Broke Barriers
E. Burchardt, N. Patzelt, J. Knecht, R. Klymowsky
Polysius AG
Graf-Galen-Strae 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 Freeports 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 Platinums Mogalakwena, March 2008,
600 ktpm; Assmangs BKM Iron Ore Project, July 2008; Northam Platinums 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
Proceedings of Copper 2010 2622
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 80
th
, 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,
todays 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
Proceedings of Copper 2010 2623
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 90
th
. 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 Freeports 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 90
th
. 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








Burchardt, Patzelt, Knecht, Klymowsky
Proceedings of Copper 2010 2624
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


















HPGRs in Copper Ore Comminution
Proceedings of Copper 2010 2625
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.








Burchardt, Patzelt, Knecht, Klymowsky
Proceedings of Copper 2010 2626
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.

HPGRs in Copper Ore Comminution
Proceedings of Copper 2010 2627
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.
Burchardt, Patzelt, Knecht, Klymowsky
Proceedings of Copper 2010 2628
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
HPGRs in Copper Ore Comminution
Proceedings of Copper 2010 2629
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.










Burchardt, Patzelt, Knecht, Klymowsky
Proceedings of Copper 2010 2630
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.







HPGRs in Copper Ore Comminution
Proceedings of Copper 2010 2631
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
tomorrows comminution circuits is obvious. This is a result of lower grade and harder ore bodies to
be treated in the future. Restriction of CO
2
emissions to avoid global warming has become an im-








Burchardt, Patzelt, Knecht, Klymowsky
Proceedings of Copper 2010 2632
portant issue on the world wide political agenda and may result in a limitation of available energy
(CO
2
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, todays 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 crushers 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-
HPGRs in Copper Ore Comminution
Proceedings of Copper 2010 2633
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.

Burchardt, Patzelt, Knecht, Klymowsky
Proceedings of Copper 2010 2634
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.















HPGRs in Copper Ore Comminution
Proceedings of Copper 2010 2635
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 m. 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 m 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 todays crushing and milling circuits will be
an on-going development in the near future. A revolution would be to apply HPGRs for finish
grinding.
Burchardt, Patzelt, Knecht, Klymowsky
Proceedings of Copper 2010 2636
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; 7
th
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 Indonesias 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

Proceedings of Copper 2010 2637
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 Concepcin Chagres Smelter, Anglo American
Edmundo Larenas 285 Camino Troncal s/n
Concepcin, Chile Catemu, Chile
J. Vargas, M. Artigas
Paipote Smelter, ENAMI
Camino Pblico s/n
Copiap, 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
Proceedings of Copper 2010 2638
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 Minera) is situated near Copiap 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
Proceedings of Copper 2010 2639
Figure 1: Current flow sheet of the Hernn 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
Inyeccin
Conc. Seco
Metal
Blanco
Anodos
Inyeccin
Mezcla CPS
Plantas de Acido
Sulfrico
Secador Secador
Escorias a Botadero Escorias a Botadero
CT
CPS CPS
RAF RAF
Gases
Gases
Blister
POX POX
HELE HELE HELE
Precipitador
Electrosttico
Gases
Metal
Blanco
Circulantes Circulantes
Precipitador
Electrosttico
Gases
Recepcin Recepcin
Mezcla RAM Mezcla RAM
Gases
Polvos HE
CNU - slice

Cardona, Hernandez, Araneda, Parra, Bahamondes, Parada, Vargas, Artigas
Proceedings of Copper 2010 2640
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 Nm
3
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. Radiacin R. Conveccin R. Lurgi
EJE ESCORIA CONVERSION
Bomba Moller
Cobre
Blister
Gases Combustin, Quemador
y Reduccin
ANODOS
BLISTER
Camara de Mezcla
VTI Conversin
VTI Flash
C Rad
4
CPS 4
CPS 3
CPS 2
Lurgi
Horno Flash
C Rad
3
C Rad
2
HLE 2
HR 2
HR 1
Rueda
de
Moldeo
TOLVA
400
Bronces
Botadero 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

Evaluation of Copper Losses in Slag Cleaning Circuits from Two Chilean Smelters
Proceedings of Copper 2010 2641
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 (1
st
and 3
rd
ladles) were taken before
the charge to the EAF and two samples (1
st
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 cm
3
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 SiO
2
and LECO titration for S. The magnetite contents were obtained by
Cardona, Hernandez, Araneda, Parra, Bahamondes, Parada, Vargas, Artigas
Proceedings of Copper 2010 2642
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 (Cu
1-x
S-FeS) on matte inclusions.
Mixed sulfides: Pyrrotite (FeS)-Bornite solid solution (Cu
1-x
S-FeS) on mixed inclusions
(This phase appears on solid matrix and grows during the cooling).
Fe-silicates
"Fayalite"
Fe-silicates (Fe
2
SiO
4
) 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.
Evaluation of Copper Losses in Slag Cleaning Circuits from Two Chilean Smelters
Proceedings of Copper 2010 2643
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 1
st
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
SiO
2
24.19 23.72 31.36 24.36 24.36 26.18 26.74 27.17 28.12 28.43 28.17
Fe
3
O
4
22.85 20.2 19.7 22.15 23.75 6.10 6.00 3.90 4.10 5.00 5.40
Al
2
O
3
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.
Cardona, Hernandez, Araneda, Parra, Bahamondes, Parada, Vargas, Artigas
Proceedings of Copper 2010 2644
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 m. 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 Ferets diameters bigger than 10 m are
shown in Figures 4 and 5.

Evaluation of Copper Losses in Slag Cleaning Circuits from Two Chilean Smelters
Proceedings of Copper 2010 2645
Table 5: TC-EAF slags phases distribution
Slag Courrent
TC tapping Slag Final Slag
Tapping 1
st
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
3
9
7
9
1
2
0
1
6
0
2
0
0
2
4
0
2
8
0
3
2
0
3
6
0
4
0
0
4
4
0
4
8
0
5
2
0
5
6
0
6
0
0
6
4
0
6
8
0
7
2
0
7
6
0
8
0
0
8
4
0
8
8
0
9
2
0
9
6
0
9
8
0
R
e
l
a
t
i
v
e

f
r
e
q
u
e
n
c
y
,

%
Diameter, m
TCS-1 TCS-2 TCS-3 TCS-4 TCS-6

Cardona, Hernandez, Araneda, Parra, Bahamondes, Parada, Vargas, Artigas
Proceedings of Copper 2010 2646
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
Fe
3
O
4
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
3
9
7
9
1
2
0
1
6
0
2
0
0
2
4
0
2
8
0
3
2
0
3
6
0
4
0
0
4
4
0
4
8
0
5
2
0
5
6
0
6
0
0
6
4
0
6
8
0
7
2
0
7
6
0
8
0
0
8
6
0
9
0
0
9
4
0
9
8
0
R
e
l
a
t
i
v
e

f
r
e
q
u
e
n
c
y
,

%
Diameter, m
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

Evaluation of Copper Losses in Slag Cleaning Circuits from Two Chilean Smelters
Proceedings of Copper 2010 2647
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
% SiO
2
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
% Fe
3
O
4
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
% Al
2
O
3
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
C
u
,
w
t
%
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
F
e
3
O
4
,

w
t
%
Time (min)
h=127cm h=106cm h=85cm h=64cm h=43cm

Cardona, Hernandez, Araneda, Parra, Bahamondes, Parada, Vargas, Artigas
Proceedings of Copper 2010 2648
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-spinels grains and f) matte inclusion, both in EF-2 sample. (Sp=
Spinel phase; Fy: Fayalite phase; Sm=solid matrix; Mte=Matte).
The slag phases distributions and sizes distribution of the Cu-sulfide inclusions (matte droplets)
were obtained using Qemscan

-Tescan system in mode PMA with 2000 m of field phase, 25 kV


voltage acceleration. The slag phases distributions at 106 cm height at the end of the settling are
showed in the Table 7.

Evaluation of Copper Losses in Slag Cleaning Circuits from Two Chilean Smelters
Proceedings of Copper 2010 2649
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 m. 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.
Cardona, Hernandez, Araneda, Parra, Bahamondes, Parada, Vargas, Artigas
Proceedings of Copper 2010 2650
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 m. Sulfides and copper in matte
inclusions at different heights: h=106 cm, b) 86 cm, c) 43 cm.
a
0
0
1
10
100
200 195 185 175 165 155 145 135 125 115 105 95 85 75 65 55 45 35 24 14
R
e
l
a
t
i
v
e

f
r
e
q
u
e
n
c
y

%
Diameter, m
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
Evaluation of Copper Losses in Slag Cleaning Circuits from Two Chilean Smelters
Proceedings of Copper 2010 2651
The sizes distribution obtained in the slag samples of the upper zone of the bath are almost similar:
All the particles are below 75 m. There are 1 % of 100-190 m 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 Fe
3
O
4
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
Fe
3+
content in the slag and the cooling rate of the slag sample. The classical report of the Fe
3
O
4

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 Fe
3+
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 - d
u=
12
(1)
Where: u=settling rate [m/s],
m
=density of matte drops [kg/m
3
],
e
= density of slag [kg/m
3
],
d=diameter of matte drops [m],
e
=slag viscosity [Pas].
The viscosities were calculated by the Kv method [8] based in the chemical composition of slags
(Table 4 and 6). The values 0.18 Pas and 0.12 Pas were used for the SCF and EAF, respectively.
Cardona, Hernandez, Araneda, Parra, Bahamondes, Parada, Vargas, Artigas
Proceedings of Copper 2010 2652
Considering 60 minutes of settling in the SCF (see Table 2), can be recovered all the particles with
a diameter higher than 135 m. The normal operation is 40 min of settling, for this condition, the
calculated critical diameter is 145 m. The high SiO
2
and Al
2
O
3
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 m 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 Fe
3
O
4
reduction for slag formation:
FeS (matte) + 3Fe
3
O
4
(slag) + 5SiO
2
(slag) = 5(2FeOSiO
2
)+SO
2
(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 PSO
2
for the SCF can be determined in function of the equilibrium constant K
I
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(Fe
3
O
4
) could be consid-
ered near unity.
( )
3 4
2
2 2
3
5
2
( )
( / )
I Fe O FeS
SO FeS
FeOSiO SiO
K a a
p K T a
a a
= =
(3)

Supposing a constant temperature (1250 C), total pressure (1 atm), the a(FeS)
matte
were calculated
by FactSage
tm
for the range of copper contents in matte droplets (60-76 % Cu) and the values are
between 0.149-0.027. The ratio a(2FeOSiO
2
)/a(SiO
2
)1.93 (activities were calculated by Fact-
Sage
TM
). The pSO
2
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].
Evaluation of Copper Losses in Slag Cleaning Circuits from Two Chilean Smelters
Proceedings of Copper 2010 2653
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 pSO
2
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
Al
2
O
3
content increase the slag basicity), an optimizing of the Fe/SiO
2
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; its 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., ECHEVERRA, P., DAZ, C.M. (2002). Factors governing
Slag Cleaning in an Electric Furnace. Canadian Metallurgical Quarterly. Vol. 41, No 4 . pp.
465-474.
Cardona, Hernandez, Araneda, Parra, Bahamondes, Parada, Vargas, Artigas
Proceedings of Copper 2010 2654
[6] PARADA, R., CHACANA, P.(2008). Chagres Evolution after Debottnecking project.
12
th
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., GOI, C., CASTELL L., BRANDALEZE E. VERDEJA L. F. Y PARRA, R.
(2006). Estimacin de la viscosidad de escorias fayalticas utilizando el mtodo kv y el
mtodo experimental del plano inclinado. Revista de Metalurgia, 42(2). Pg. 84-90.
[9] SLAG ATLAS (1995). Edited by Verein Deutscher Eisenhttenleute(VDEh). Germany.
[10] HERREROS, O., QUIROZ R., MANZANO, E. BOU C., VIALS, J. (1998). Copper extrac-
tion from reverberatory and flash furnace slags by chlorine leaching. Hydrometallurgy 49. Pp
87-101.

Proceedings of Copper 2010 2655
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
Proceedings of Copper 2010 2656
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 90s of 20
th
century and in
2004 [7, 8]
Ore component
Rudna Polkowice Sieroszowice Lubin
90s 2004 90s 2004 90s 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
C
org
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 90s of 20
th
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
Proceedings of Copper 2010 2657
is over 45 %. For organic carbon (C
org
) 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 C
org
, %
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 (C
org
)
content and metal concentration. The correlation between metal and C
org
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.
Chmielewski, Luszczkiewicz
Proceedings of Copper 2010 2658

y = 0.0523x
2
+ 0.2947x + 0.988
R
2
= 0.929
0
5
10
15
20
25
0 2 4 6 8 10 12 14 16 18
Content of organic carbon, %
C
o
n
t
e
n
t

o
f


C
u
,

%




.


y = 46.379x
0.6661
R
2
= 0.6279
0
50
100
150
200
250
300
350
400
450
500
0 2 4 6 8 10 12 14 16
Content of organic carbon, %
C
o
n
t
e
n
t

o
f

A
g

,

%




.

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
Content of organic carbon, %
C
o
n
t
e
n
t

o
f

N
i
,


g
/
t


.


0
50
100
150
200
250
300
0 1 2 3 4 5
Content of organic carbon, %
C
o
n
t
e
n
t

o
f

C
o
,


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,
Leaching of Gangue Matter in Technological Flotation Circuit of Polish Copper Ores
Proceedings of Copper 2010 2659
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 1
st
cleaning flotation [8]. According to these figures, almost 70 % of the organ-
ic carbon and some 30 % of copper circulate between 1
st
cleaning flotation and additional milling
operations. The shale by-product (middlings) from 1
st
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 1
st
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, %
C
org
, % C
org
recovery,%
FINAL CONCENTRATE
LUBIN CONCENTRATOR
FINAL TAILINGS
FEED
(Run of Mine ore)

Chmielewski, Luszczkiewicz
Proceedings of Copper 2010 2660
Figure 6: Simplified flowsheet of first technological line of Lubin Concentrator.
Figure 7: Flowsheet and mass balance of copper and organic carbon for 1
st
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
M
2nd 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, %
C
org
, % C
org
recovery,%
Cu recovery, %
C
org
recovery,%
Yield, %
1st CLEANING FLOTATION
Feed: concentrate from
rougher flotation
Concentrate
to 2nd cleaning flotation
Middlings to rougher
flotation after cycloning
and regrinding

Leaching of Gangue Matter in Technological Flotation Circuit of Polish Copper Ores
Proceedings of Copper 2010 2661
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.
COPPER
BASE AND
NOBLE METALS
FLOTATION
TAILINGS
Technological process of copper recovery
at the KGHM Polish Copper
Shale-enriched
by-product
Non-oxidative leaching
with H
2
SO
4
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

Chmielewski, Luszczkiewicz
Proceedings of Copper 2010 2662
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 1
st
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, %
R
e
c
o
v
e
r
y

o
f

o
r
g
a
n
i
c

c
a
r
b
o
n

i
n

c
o
n
c
e
n
t
r
a
t
e
,

%



.
lack of floatabilit y differences
St andard Xant .
Fuel oil 100g/Mg
Fuel oil 200g/Mg
Fuel oil 400g/Mg
Fuel oil 600g/Mg

Leaching of Gangue Matter in Technological Flotation Circuit of Polish Copper Ores
Proceedings of Copper 2010 2663
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 1
st
cleaning)
in comparison with a processing feed (run-of-mine ore) and final concentrate produced in
2004
Material Cu, % Ag, g/t C
org
, % Ni, g/t Co, g/t Pb, % Zn, g/t Fe, % S
c,
% 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 1
st
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:
CaCO
3
+ H
2
SO
4
+ H
2
O = CaSO
4
2H
2
O + CO
2
(1)
MgCO
3
+ H
2
SO
4
= MgSO
4
+ CO
2
+ H
2
O (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 H
2
SO
4
is very rapid and can be performed at
ambient temperatures in reactors with mechanical stirring of a simple construction.
The amount of H
2
SO
4
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
Chmielewski, Luszczkiewicz
Proceedings of Copper 2010 2664
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.
p
H
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 H
2
SO
4
required for carbonates leaching is the maxi-
mum demand for acid (
max
SO H
4 2
z ), which is the mass of pure H
2
SO
4
necessary for a total decomposi-
tion of carbonates in 1 kg of dry solid feed. The
max
SO H
4 2
z parameter should be determined analytically
from laboratory tests. H
2
SO
4
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 H
2
SO
4
/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 CO
2
.
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.
Leaching of Gangue Matter in Technological Flotation Circuit of Polish Copper Ores
Proceedings of Copper 2010 2665
The samples of middlings being the tailings from the 1
st
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 H
2
SO
4
) was also carried out. For the flotation experiments a Mek-
hanobr sub-aeration type laboratory flotation machine equipped with a 1 dm
3
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

Chmielewski, Luszczkiewicz
Proceedings of Copper 2010 2666
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, %
R
e
c
o
v
e
r
y

o
f

b
a
r
r
e
n

p
a
r
t

i
n

t
a
i
l
i
n
g
,

%



.
Lack of upgrading
Standard flot at ion
50% of decomposition
70% of decomposition
90% of decomposition
Cu


0
10
20
30
40
50
50 60 70 80 90 100
Recovery of copper in concentrate, %
R
e
c
o
v
e
r
y

o
f

b
a
r
r
e
n

p
a
r
t

o
f

t
a
i
l
i
n
g
,

%

.
Lack of upgrading
Standard flot at ion
50% of decomposition
70% of decomposition
90% of decomposition
Cu


0
10
20
30
40
50
50 60 70 80 90 100
Recovery of silver in concentrat e, %
R
e
c
o
v
e
r
y

o
f

b
a
r
r
e
n

p
a
r
t

o
f

t
a
i
l
i
n
g
,

%


.
Lack of upgrading
Standard flotation
50% of decomposit ion
70% of decomposit ion
90% of decomposit ion
Ag


0
10
20
30
40
50
50 60 70 80 90 100
Recovery of component in concent rate, %
R
e
c
o
v
e
r
y

o
f

b
a
r
r
e
n

p
a
r
t

o
f

t
a
i
l
i
n
g
,

%

.

Lack of upgrading
Cu
Ag
Co
Zn
Fe
Corg
70% of carbonate
decomposition

Leaching of Gangue Matter in Technological Flotation Circuit of Polish Copper Ores
Proceedings of Copper 2010 2667
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
o
C) 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
o
C) 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 time, min.

C
u

c
o
n
c
e
n
t
r
a
t
i
o
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, min.
C
u

c
o
n
c
e
n
t
r
a
t
i
o
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 H
2
SO
4
concentration was kept at the level of 50 g/dm
3
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.
Chmielewski, Luszczkiewicz
Proceedings of Copper 2010 2668
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 (Cu
2
S)
and bornite (Cu
5
FeS
4
) can be leached very rapidly by oxygenated H
2
SO
4
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
o
C 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.
C
o

c
o
n
c
e
n
t
r
a
t
i
o
n
,

m
g
/
l




.





0
7
15
30



0
5
10
15
20
25
30
0 60 120 180 240 300
Leaching time, min.
N
i

c
o
n
c
e
n
t
r
a
t
i
o
n
,

m
g
/
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 d
80
exceeded 100 m even after non-oxidative leaching
with H
2
SO
4
(carbonates decomposition R
w
= 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 m particle size fraction and its additional
regrinding seems to be very desirable to facilitate further leaching.
Leaching of Gangue Matter in Technological Flotation Circuit of Polish Copper Ores
Proceedings of Copper 2010 2669


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 1
st
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
Chmielewski, Luszczkiewicz
Proceedings of Copper 2010 2670
3. Flotation experiments demonstrated that upgrading the middlings was nearly impossible without
a chemical pretreatment. Non-oxidative leaching of Lubin middlings with H
2
SO
4
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 C
org
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-Zieliska J. 1990, Petrography and genesis of copper-bearing shales from
Foresudetic Monocline. Rudy i Metale Nieelazne. R.35, Nr 5-6, 128-133, in Polish
[2] Rydzewski A., 1996, Lithology of the deposit rocks. In: Monografia KGHM Polska Mied
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
Leaching of Gangue Matter in Technological Flotation Circuit of Polish Copper Ores
Proceedings of Copper 2010 2671
[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

Proceedings of Copper 2010 2672



Proceedings of Copper 2010 2673
Pressure Leaching of Shale Middlings from Lubin
Concentrator in Oxygenated Sulphuric Acid
Tomasz Chmielewski, Jerzy Wdka
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 1
st
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
o
C. 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/dm
3
. 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
Proceedings of Copper 2010 2674
- high rate of leaching reactions,
- elimination of SO
2
, 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 1
st
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

Proceedings of Copper 2010 2675
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 H
2
SO
4
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 1
st
cleaning from Lubin Concentrator. Pressure
leaching examinations were performed in the temperature range of 100-180
o
C 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 1
st
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 H
2
SO
4
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
Proceedings of Copper 2010 2676
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 Gogw 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
Pressure Leaching of Shale Middlings in Oxygenated Sulphuric Acid

Proceedings of Copper 2010 2677
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 1
st
sampling campaign in 2006 and
2
nd
sampling campaign in 2007. Middlings from 1
st
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
(1
st
and 2
nd
sampling campaigns 2006 and 2007).
CONTENT
Cu, % Fe, % Ni, g/t Co, g/t Pb, % As, %
1
st
campaign 2.72 1.76 374 572 1.51 0.090
2
nd
campaign 2.60 1.89 328 613 - 0.085
CONTENT
Ag, g/t Zn
,
g/t S
c,
% Sso
4
, % C
total,
% C
org
, %
1
st
campaign 190 1200 2.95 1.45 14.30 8.96
2
nd
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 (Cu
2
S) and bornite (Cu
5
FeS
4
), easiest to
leach copper sulfides, are dominating in the material. The content of chalcopyrite (CuFeS
2
), 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
Proceedings of Copper 2010 2678
that beneficial mineralogical segregation of minerals takes place during the 1
st
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 d
80
exceeded 100 m even after non-oxidative leaching
with H
2
SO
4
(carbonates decomposition, R
w
= 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, R
w
(within 30-90 %).

20
30
40
50
60
70
80
90
100
10 100 1000
Particle size, m
C
u
m
u
l
a
t
i
v
e

r
e
c
o
v
e
r
y
,

%



.
0
Rw - 30%
Rw - 50%
Rw - 70%
Rw - 90%

Pressure Leaching of Shale Middlings in Oxygenated Sulphuric Acid

Proceedings of Copper 2010 2679
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 m 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
o
C 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
Proceedings of Copper 2010 2680
The experiments were performed in 2.0 dm
3
autoclave having the stirrer and sampling pipe made of
teflon. The following experimental procedure was used: A teflon beaker containing 1.0 dm
3
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
H
2
SO
4
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
o
C 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 Results and discussion
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
o
C
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
o
C required about 120 min. of activation
time. This was most likely due to the formation of H
2
S observed at the initial stage of the process.
Remarkable acceleration of the process was subsequently detected after 120 min. of activation.
Unexpectedly, at 120
o
C 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
o
C the leaching appeared to be very rapid, although some decrease
of Cu concentration was observed at 160 and 180
o
C versus 140
o
C. 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
o
C, then iron
concentration slightly drops as the result of precipitation of iron(III) oxide (Fe
2
O
3
) or goethite
(FeOOH). The appearance of soluble iron in leaching solution results from leaching of Cu-Fe sul-
phides, mainly bornite Cu
5
FeS
4
and chalcopyrite CuFeS
2
.
Pressure Leaching of Shale Middlings in Oxygenated Sulphuric Acid

Proceedings of Copper 2010 2681

0
500
1000
1500
2000
2500
3000
0 30 60 90 120 150 180 210
Leaching time, min
C
u

c
o
n
c
e
n
t
r
a
t
i
o
n
,

m
g
/
d
m
3




.
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, min
F
e

c
o
n
c
e
n
t
r
a
t
i
o
n
,

m
g
/
d
m
3


.

100
120
140
160
180
Fe

0
10
20
30
40
50
0 30 60 90 120 150 180 210
Leaching time, min
C
o

c
o
n
c
e
n
t
r
a
t
i
o
n
,

m
g
/
d
m 3
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 H
2
SO
4
: 5 %,
solid to liquid ratio, s/l: 1:10.

Pressure leaching of cobalt was practically not observed at temperatures 100 and 120
o
C with detected
concentration of cobalt about 5 mg/dm
3
. Remarkable acceleration of leaching of Co was recorded at
temperatures above 140
o
C 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
o
C the highest
leaching recovery was observed with the final Ni concentration of about 180 mg /dm
3
. At tempera-
tures exceeding 140
o
C the leaching rate of nickel decreases, which are hard to explain at this stage
Chmielewski, Wodka
Proceedings of Copper 2010 2682
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
o
C when
detected concentration of As was about 10 mg/dm
3
. Significant increase in leaching rate of As was
observed at temperatures above 140
o
C 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
o
C. At 140
o
C
observed were the highest Cu, Co and As recoveries. It was also detected that at temperatures ex-
ceeding 180
o
C iron control became possible as a result of precipitation of FeOOH or Fe
2
O
3
. 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 FeAsO
4
2H
2
O at
temperatures above 180
o
C 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/dm
3
. 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
o
C and above
90 % for Fe at 160
o
C. 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/dm
3
at temperature of 140
o
C 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 doesnt leach before introduction of oxygen and course of copper leaching resembles
those observed during investigations of the effect of temperature.
Pressure Leaching of Shale Middlings in Oxygenated Sulphuric Acid

Proceedings of Copper 2010 2683
0
500
1000
1500
2000
2500
3000
0 30 60 90 120 150 180 210
Leaching time, min. .
C
o
p
p
e
r

c
o
n
c
e
n
t
r
a
t
i
o
n
,

m
g
/
d
m
3



.
2%
3%
4%
5%

0
10
20
30
40
50
60
0 30 60 90 120 150 180 210
Leaching time, min.
C
o
b
a
l
t

c
o
n
c
e
n
t
r
a
t
i
o
n
,

m
g
/
d
m
3



.
2%
3%
4%
5%

Figure 5: Effect of sulphuric acid concentration on copper and cobalt pressure leaching.
(Temperature: 140
o
C, 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/dm
3
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
L
e
a
c
h
i
n
g

r
e
c
o
v
e
r
y
,


%



.
H
2
SO
4
concentration, %
Cu
Fe
Ni
Co
As

Chmielewski, Wodka
Proceedings of Copper 2010 2684
For 20 g/dm
3
of H
2
SO
4
the concentration of ferric ions in solutions was 680 mg/dm
3
and for
50 g/dm
3
of H
2
SO
4
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/dm
3
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/dm
3

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/dm
3
. Kinetic curves observed
for Cu and Co leaching at acid concentrations above 30 g/dm
3
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 H
2
SO
4
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/dm
3
of H
2
SO
4
. 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 H
2
SO
4
concentration (2 %) the precipitation of Fe(III) from
the solution was observed, most likely as FeOOH and Fe
2
O
3
and the concentration of Fe in solution
decreased to about 1.3 g/dm
3
. 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
o
C), 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].
Pressure Leaching of Shale Middlings in Oxygenated Sulphuric Acid

Proceedings of Copper 2010 2685
0
500
1000
1500
2000
2500
3000
3500
0 30 60 90 120 150 180 210
Leaching time, min.
C
u

c
o
n
c
e
n
t
r
a
t
i
o
n
,

m
g
/
d
m
3



.
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, min
C
o

c
o
n
c
e
n
t
r
a
t
i
o
n
,

m
g
/
d
m
3




.
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
o
C, concentration of H
2
SO
4
: 5 %, liquid/solid ratio: 10:1.
Figure 8: Effect of oxygen partial pressure on metals recovery from Lubin middlings.
Temperature: 140
o
C, H
2
SO
4
: 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
L
e
a
c
h
i
n
g

r
e
c
o
v
e
r
y
,


%



.
Oxygen partial pressure, atm .
Cu
Fe
Ni
Co
As

Chmielewski, Wodka
Proceedings of Copper 2010 2686
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,
o
C

oxygen par-
tial pressure,
atm
H
2
SO
4
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, CuFeS
2
) 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.

Pressure Leaching of Shale Middlings in Oxygenated Sulphuric Acid

Proceedings of Copper 2010 2687
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 1
st
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 H
2
SO
4
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
Proceedings of Copper 2010 2688
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
o
C and under oxygen pres-
sure exceeding 5 atm. From pressure leaching experiments performed at temperatures 120-180
o
C,
under oxygen pressure 2.5-10 atm, at H
2
SO
4
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
o
C 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 H
2
SO
4
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.
Pressure Leaching of Shale Middlings in Oxygenated Sulphuric Acid

Proceedings of Copper 2010 2689
[5] NICE R.W., (2004): The Western Australian nickel laterites project What we have learned?,
Pressure Hydrometallurgy 2004, Proc. 34
th
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
Proceedings of Copper 2010 2690
[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. 34
th
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. 6
th
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. 6
th
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, ICNOP07, 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, BIOPROCOP06, Lubin 2006,
KGHM Cuprum, Wroclaw 2006, 125-145 (in Polish).
[22] DHUGUES P., NORRIS P.R., HALLBERG K.B., SNCHEZ 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. 111120.
[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).
Pressure Leaching of Shale Middlings in Oxygenated Sulphuric Acid

Proceedings of Copper 2010 2691
[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.

Proceedings of Copper 2010 2692


Proceedings of Copper 2010 2693
Mine to Heap in Mantoverde Anglo American
Division
Manuel Daz, Cristian Salgado, Carlos Prez, Cristian Alvayai, Leonardo Herrera
Gabriel Zrate
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 Chaaral 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.
Daz, Salgado, Prez, Alvayai, Herrera, Zrate
Proceedings of Copper 2010 2694
1 Introduction
The Mantoverde mine is located in the Chaaral 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
Proceedings of Copper 2010 2695
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 CaCO
3

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)
C
u
s
o
l

R
e
c
o
v
e
r
y

(
%
)

Daz, Salgado, Prez, Alvayai, Herrera, Zrate
Proceedings of Copper 2010 2696
It was also found that heap height and CaCO
3
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 CaCO
3
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
C
u
s
o
l

R
e
c
o
v
e
r
y

(
%
)
Col 31
Col 33

Mine to Heap in Mantoverde Anglo American Division
Proceedings of Copper 2010 2697
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)
C
u
s
o
l

R
e
c
o
v
e
r
y

(
%
)
Col 31
Col 33

Daz, Salgado, Prez, Alvayai, Herrera, Zrate
Proceedings of Copper 2010 2698
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
C
u
s
o
l

R
e
c
o
v
e
r
y

(
%
)
Predicted
Actual

Mine to Heap in Mantoverde Anglo American Division
Proceedings of Copper 2010 2699
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
N
e
t

A
c
i
d

C
o
n
s
u
m
p
t
i
o
n

(
k
g
/
t

o
r
e
)
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
N
e
t

A
c
i
d

C
o
n
s
u
m
p
t
i
o
n

(
k
g
/
t

o
r
e
)
0
0,2
0,4
0,6
0,8
1
1,2
1,4
1,6
Predicted
Actual
CaCO3

Daz, Salgado, Prez, Alvayai, Herrera, Zrate
Proceedings of Copper 2010 2700
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, CaCO
3
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
P
L
S

v
i
s
c
o
s
i
t
y

(
c
p
)

Mine to Heap in Mantoverde Anglo American Division
Proceedings of Copper 2010 2701
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. Zrate. Anlisis Operacin Pilas Mantoverde. Julio 2007. Interim Report.
[2] C. Salgado, C. Prez, C. Alvayai, G. Zrate, Acid management in heap leaching. Are we doing
right? HydroCopper 2009, E. Domic, J. Casas Eds. Gecamin, Santiago, Chile. pp. 47-56.

Proceedings of Copper 2010 2702


Proceedings of Copper 2010 2703
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
Proceedings of Copper 2010 2704
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
hrp vmp
vmp
hrp vmp
est lib
r
r r
r
r r
P

= (1)
P
lib.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
*
* *
*
* *
*
* *
. vmp hrp
hrp
vmp hro
hrp
hrp vmp
vmp
hrp vmp
est Locked Partially
r r for
r
r r
r
r r
r
r r
P >

(2)
) ... ))]...( exp( 1 )( exp( 1 [
* *
*
* *
*
* *
. vmp hrp
hrp
hrp vmp
vmp
hrp vmp
est Locked Partially
r r for
r
r r
r
r r
P <

(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
Proceedings of Copper 2010 2705
Correspondingly, the probability of fully-locked valuable mineral particles is [2]:
) ... ]...( ) ))][( exp( 1 )( exp( 1 [
* * 3
*
* *
*
* *
*
* *
. vmp hrp
hrp
vmp hro
hrp
hrp vmp
vmp
hrp vmp
est Locked Fully
r r for
r
r r
r
r r
r
r r
P >

(4)
) ... ...( 0
* *
. vmp hrp est Locked Fully
r r for P < =

(5)
1.2 Liberated Particle Leaching
Liberated particle leaching can be described mathematically by the expression:
] ) 1 ( 1 [ ) (
y
lib vmp lib lib vmp
r k r t = (6)
k
lib
is a reaction constant (sec/cm), 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
, , 3
2
2
,
, PD lib PD lib
eff x
vmp lib PD
PD lib vmp
D C
r k
r t = (7)
C
x
is the concentration of reactant, D
eff
is the the effective diffusivity, and k
PD,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
lib vmp lib lib vmp
r k r t

(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 4/(4/3)
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
, , 3
2
2
,
, PD lock PD lock
eff x
hrp lock PD
PD lib vmp
D C
r k
r t = (9)
Free, Jurovitzki
Proceedings of Copper 2010 2706
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 (H
2
SO
4
, 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(SO
4
)
3
, nH
2
O 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 (HNO
3
, 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
Predicting the Effect of Locked, Partially Locked, and Liberated Minerals
Proceedings of Copper 2010 2707
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.
3 Results and discussion
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
Proceedings of Copper 2010 2708
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,
k
lib
= 1.0710
4
sec/cm, k
PD
,
lock
/D
eff
= 4.610
9
cmsec/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,
k
lib
= 5.7610
5
sec/cm, k
PD
,
lock
/D
eff
= 3.210
9
cmsec/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
F
r
a
c
t
i
o
n

R
e
a
c
t
e
d
Time (sec)
Model Measured

0
0.02
0.04
0.06
0.08
0.1
0.12
0 500000 1000000 1500000
F
r
a
c
t
i
o
n

R
e
a
c
t
e
d
Time (sec)
Model
Measured

Predicting the Effect of Locked, Partially Locked, and Liberated Minerals
Proceedings of Copper 2010 2709
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, k
lib
= 1.9210
11
sec/cm, k
PD
,
lock
/D
eff
= 6.410
9
cmsec/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
F
r
a
c
t
i
o
n

R
e
a
c
t
e
d
Time (sec)
Model Measured

Free, Jurovitzki
Proceedings of Copper 2010 2710
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
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, 7
th
International Copper 2010 Cobre 2010 Conference at Hamburg,
Germany, Proceedings of Copper 2010, Volume 7, pp. 2711-2719.

Proceedings of Copper 2010 2711
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]:
CuSiO
3
2H
2
O + 12H
+
= 6Cu
2+
+ 8H
4
SiO
4
(1)
CuO + 2H
+
= Cu
2+
+ H
2
O (2)
CuSO
4
3Cu(OH)
2
+ 6H
+
= 4Cu
2+
+ 6H
2
O (3)
KAlSi
3
O
8
+ 4H
+
= K
+
+ Al
3+
+ 3H
4
SiO
4
(4)
(H,K)
2
(Mg,Fe)
2
Al
2
(SiO
4
)
3
+ 10 H
+
= 2K
+
+ 2Al
3+
+ 3H
4
SiO
4
+ 2(Fe,Mg)
2+
(5)
Fe
2
O
3
+ 6H
+
= 2Fe
3+
+ 3H
2
O (6)
FeO(OH) + 3H
+
= Fe
3+
+ H
2
O (7)
Free
Proceedings of Copper 2010 2712
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]:
3Fe
3+
+ K
+
+2SO
4
2-
+ 6H
2
O = KFe
3
(SO
4
)
2
(OH)
6
+ 6 H
+
(8)
3H
2
O + K
+
+ 2SO
4
2-
+ 3Al
3+
= KAl
3
(OH)
6
(SO
4
)
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
r
r P

=
*
exp 1 ) ( (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
Proceedings of Copper 2010 2713
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
hrp vmp
vmp
hrp vmp
est lib
r
r r
r
r r
P

= (11)
P
lib.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- P
lib.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
*
* *
*
* *
*
* *
. vmp hrp
hrp
vmp hro
hrp
hrp vmp
vmp
hrp vmp
est Locked Partially
r r for
r
r r
r
r r
r
r r
P >

(12)
) ... ))]...( exp( 1 )( exp( 1 [
* *
*
* *
*
* *
. vmp hrp
hrp
hrp vmp
vmp
hrp vmp
est Locked Partially
r r for
r
r r
r
r r
P <

(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
*
* *
*
* *
*
* *
. vmp hrp
hrp
vmp hro
hrp
hrp vmp
vmp
hrp vmp
est Locked Fully
r r for
r
r r
r
r r
r
r r
P >

(14)
) ... ...( 0
* *
. vmp hrp est Locked Fully
r r for P < =

(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
Proceedings of Copper 2010 2714
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
lib vmp lib lib vmp
r k r t = (16)
k
lib
is a reaction constant (sec/cm), 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
, , 3
2
2
,
, PD lib PD lib
eff x
vmp lib PD
PD lib vmp
D C
r k
r t = (17)
C
x
is the concentration of reactant, D
eff
is the the effective diffusivity, and k
PD,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 30
V
a
l
u
a
b
l
e

M
i
n
e
r
a
l

E
x
p
o
s
u
r
e

(
%
)
Size (mm)
Simplified Model
Miller et al. (2003)

Predicting Leaching Solution Acid Consumption as a Function of pH
Proceedings of Copper 2010 2715
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 4/(4/3)
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
lib vmp lib lib vmp
r k r t

(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
, , 3
2
2
,
, PD lock PD lock
eff x
hrp lock PD
PD lib vmp
D C
r k
r t = (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
Proceedings of Copper 2010 2716
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 Results and discussion
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, k
lib
= 3.8410
6
sec/cm,
k
PD
,
lock
/D
eff
= 3.210
9
cmsec/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
F
r
a
c
t
i
o
n

R
e
a
c
t
e
d
Time (sec)
Model pH 1.5
Measured pH 1.5
Model pH 2.5
Measured pH 2.5

Predicting Leaching Solution Acid Consumption as a Function of pH
Proceedings of Copper 2010 2717
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 (m). The ore particles were 14 + 20 Mesh.
The model parameters were r
*
vmp
= 0.004 cm, k
lib
= 1.3510
6
sec/cm,
k
PD
,
lock
/D
eff
= 110
9
cmsec/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
F
r
a
c
t
i
o
n

R
e
a
c
t
e
d
Time (sec)
Model 820 microns
Measured 820 microns
Model 1420 microns
Measured 1420 microns

Free
Proceedings of Copper 2010 2718
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
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
Predicting Leaching Solution Acid Consumption as a Function of pH
Proceedings of Copper 2010 2719
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. 331340.

Proceedings of Copper 2010 2720


Proceedings of Copper 2010 2721
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 SO
2
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 Kupferhtte 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 SO
2
for sulfuric acid manufacture. In
the sixteenth century, pyrite was heap leached in the Harz mountains in Germany and in Ro Tinto
mines in Spain to recover its copper content and in the nineteenth century pyrometallurgical
processes were developed.
Habashi
Proceedings of Copper 2010 2722

Figure 1: Naturally-occurring pyrite cubes Figure 2: French translation of Johann
Friedrich Henckels 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 SO
2
needed for sulfuric acid manufacture:
2 FeS
2
+
11
/
2
O
2
4 SO
2
+ Fe
2
O
3
(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
Copper from Pyrite A Short History
Proceedings of Copper 2010 2723
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 Kupferhtte 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
Proceedings of Copper 2010 2724
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 Kupferhtte 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).


Copper from Pyrite A Short History
Proceedings of Copper 2010 2725
Figure 10: Importing pyrite from worldwide suppliers to Duisburger Kupferhtte 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
Proceedings of Copper 2010 2726

Figure 11: Salt roasting of pyrite cinder Figure 12: Leaching plant for purple ore
Cu + CuCl
2
Cu
2
Cl
2(ppt)
(2)
Cu
2+
+ Fe Cu + Fe
2+
(3)
Copper for the first step was obtained from the second step. It is interesting to note that the blue
CuCl
2
solution turns pink after reduction and precipitation of Cu
2
Cl
2
due to the presence of Co
2+
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:
Cu
2
Cl
2
+ Ca(OH)
2
Cu
2
O + CaCl
2
+ H
2
O (4)
2Cu
2
O + C 4Cu + CO
2
(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 Na
2
SO
4
10H
2
O (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 CuSO
4

solution obtained was electrolyzed to produce copper cathodes.
Copper from Pyrite A Short History
Proceedings of Copper 2010 2727
Figure 13: Cementation plant

Habashi
Proceedings of Copper 2010 2728
Figure 14: Sodium sulfate plant

Copper from Pyrite A Short History
Proceedings of Copper 2010 2729
The residue, called purple ore, now a high-grade iron ore (6163 % 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 Kupferhtte 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
Proceedings of Copper 2010 2730
Figure 16: Blast furnace plant producing pig iron and lead-silver alloy

Copper from Pyrite A Short History
Proceedings of Copper 2010 2731


Figure 17: General flowsheet of operations at Duisburger Kupferhtte
4 Pyrite and the steel industry
Noting the success of Duisburger Kupferhtte, 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
Proceedings of Copper 2010 2732
slurry containing 68 % solids. When the solids are filtered off, the solution contains 2025 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 SO
2
. In the middle part of
the furnace, the reduction zone, SO
2
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: FeS
2
FeS +
1
/
2
S
2
(6)
Oxidation zone: FeS
+
3
/
2
O
2
FeO + SO
2
(7)
Middle zone: SO
2
+ C CO
2
+
1
/
2
S
2
(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.

Copper from Pyrite A Short History
Proceedings of Copper 2010 2733


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 SO
2
. 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 SO
2
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
Proceedings of Copper 2010 2734



Figure 20: Herman Frasch
(1851-1914)
Figure 21: Extracting sulfur
by Frasch process
Figure 22: Storage of
Frasch sulfur
H
2
S-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 SO
2
. Hydrogen sulfide must first be separated from the gases by an absorption
desorption process (then oxidized by a controlled amount of oxygen at 400
o
C on aluminum oxide
or bauxite bed using Claus reaction discovered in 1883 by the German chemist Carl Friedrich Claus:
H
2
S + O
2
S + H
2
O (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 Kupferhtte, 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, Mtallurgie Extractive Qubec,
Qubec City 2009. Distributed by Laval University Bookstore Zone
Copper from Pyrite A Short History
Proceedings of Copper 2010 2735
[5] F. Habashi, The Recovery of Elemental Sulfur from Sulfide Ores. Proceedings of XXXVI
Congrs 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 Fools Gold, Eng. & Ming. J. 170 (12), 5964
(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-
ferhtte, 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 Schlmme und Fllprodukte im Schachto-
fen, Z. Erzbergbau u. Metallhttenw. 22(1), 22 28 (1969)
[11] C.Nuez 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 Congrs International de
Minralurgie, volume 3 (1985)
[12] W. Teworte, Duisburger Kupferhtte, company brochure published in Duisburg, W. Germany
1957
[13] E. Wiklund, Die Nutzung von Pyrit als Grundlage fr die Produktions von Schwefelsure
und Eisenoxyd, Aufbreitungs- Technik (6),285-289 (1977)

Proceedings of Copper 2010 2736


Proceedings of Copper 2010 2737
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-Strae 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
Proceedings of Copper 2010 2738
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
Proceedings of Copper 2010 2739
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

Hartmann, Horbach, Kramer
Proceedings of Copper 2010 2740
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

Copper Crud Treatment with 3 Phase Centrifuge
Proceedings of Copper 2010 2741
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.

Hartmann, Horbach, Kramer
Proceedings of Copper 2010 2742
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.

Copper Crud Treatment with 3 Phase Centrifuge
Proceedings of Copper 2010 2743
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.

Hartmann, Horbach, Kramer
Proceedings of Copper 2010 2744
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.


Copper Crud Treatment with 3 Phase Centrifuge
Proceedings of Copper 2010 2745
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-Flssig 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


Proceedings of Copper 2010 2746


Proceedings of Copper 2010 2747
A Specific Electrode for On-line pH
Measurement in SAG Cleaner Flotation Circuits
Christian Hecker C., Alejandro Gonzlez S., Alejandra Mejas J. Claudia Rodrguez F.
Hecker Electroqumica Industrial S.A Codelco Chile
Sargento Bernardo Cuervas 093 Divisin 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
dont 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, Gonzlez, Mejas, Rodrguez
Proceedings of Copper 2010 2748
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
Proceedings of Copper 2010 2749
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.
3 Results and discussion
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, Gonzlez, Mejas, Rodrguez
Proceedings of Copper 2010 2750
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.
R
2
= 0,89
-500
-450
-400
-350
-300
-250
8 11 13 16 18 21 23 26 28 31 33
E

/
m
V
/
(
A
g
/
A
g
C
l
)
Volumetric Titration / ml

A Specific Electrode for On-line pH Measurement in SAG Cleaner Flotation Circuits
Proceedings of Copper 2010 2751
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.; ALARCN E.; ZAPATA V.
Patente de Invencin. Presentada en el INAPI Chile. Solicitud Nmero 2035-2009.
[2] BUSTAMANTE A., GONZALEZ A. (2009): Desarrollo de un sensor de pH para circuitos de
limpieza en medio fuertemente alcalino, en flotacin colectiva.


Proceedings of Copper 2010 2752


Proceedings of Copper 2010 2753
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 Fe
3+
and Fe
2+
was supplied to the top of the column and the effect
of redox potential on copper extraction was investigated.
When 0.04 M Fe
3+
/ 0.06 M Fe
2+
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 Fe
3+
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, CuFeS
2
, 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 km
2
in area and several m in
height) using dilute sulfuric acid at ambient temperature. Operating conditions of heap leaching are
Hiroyoshi, Kuwazawa, Takehara, Tsunekawa
Proceedings of Copper 2010 2754
relatively moderate: temperature inside the ore pile increases due to the exothermal leaching reac-
tion but it is not over 70
o
C; 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 Fe
3+
is generated in-
situ by microbiological oxidation of Fe
2+
extracted from the ore, but the Fe
3+
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 Fe
3+
to Fe
2+
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 Fe
3+
/Fe
2+

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 m) was prepared from a massive specimen
(Copper Queen Mine, Arizona, USA). Silica (SiO
2
), sphalerite (ZnS), and molybdenite (MoS
2
) 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 H
2
SO
4
, Fe(SO
4
)7H
2
O, Fe
2
(SO
4
)
3
nH
2
O, and CuSO
4
5H
2
O 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
H
2
SO
4
and 0.1 kmol m
-3
Fe
3+
. A 500 cm
3
Erlenmeyer flask containing 2 g of the
Redox Potential Control in Column Leaching of Chalcopyrite
Proceedings of Copper 2010 2755
ground chalcopyrite and 200 cm
3
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 cm
3
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 Fe
3+
and
Fe
2+
were calculated from the total Fe concentration and the redox potential value E, assuming the
following material balance equation and Nernst equation for the Fe
3+
/Fe
2+
redox pair.
[Total Fe] = [Fe
3+
] + [Fe
2+
] (1)
E = 0.67 + 0.059 log ([Fe
3+
]/[Fe
2+
]) (2)
2.3 Column leaching experiments
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 cm
3
of pure water as a binder. The aggregate was charged
in 0.1 m diameter x 1 m height cylindrical PVC column. The H
2
SO
4
solution (pH 1.5), containing
known concentrations of Fe
3+
, Fe
2+
, and Cu
2+
,was stored in a reservoir tank and was supplied to the
column top at a flow rate of 2.0 dm
3
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 cm
3
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.
Hiroyoshi, Kuwazawa, Takehara, Tsunekawa
Proceedings of Copper 2010 2756
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, Fe
3+
in the solution is consumed and Fe
2+
is formed by the following reac-
tion.
CuFeS
2
+ 4 Fe
3+
= Cu
2+
+ 5Fe
2+
+ 2S (3)
Therefore, the Fe
3+
concentration decreases and the Fe
2+
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
P
u
m
p
Drainage tank
Solution reservoir
Column with sampling port
0.1m interval

Redox Potential Control in Column Leaching of Chalcopyrite
Proceedings of Copper 2010 2757
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 Cu
2+
concentration
changes [5-6, 12-13]. The authors have proposed a normalized redox potential, E
normal
, as an indica-
tor to compensate for the effect of Cu
2+
concentration [5-6, 12-13]. The normalized redox potential,
E
normal
, is calculated from the concentrations of Cu
2+
, Fe
3+
, and Fe
2+
using the following equation.
(4)
As shown in Figure 3, the leaching rates are higher at E
normal
between 0 ~ 1, and the maximum rate
is achieved at E
normal
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
R
e
d
o
x

p
o
t
e
n
t
i
a
l
,

E
/

V

v
s
.

S
H
E
Time / h
C
u

e
x
t
r
a
c
t
e
d
/
k
m
o
l

m
-
3

Hiroyoshi, Kuwazawa, Takehara, Tsunekawa
Proceedings of Copper 2010 2758
Figure 3: Cu extraction rate as a function of redox potential and normalized redox potential.
3.2 Column leaching experiments
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, E
normal
. 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, Fe
3+
/ total Fe ratio ([Fe
3+
]/[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
Fe
3+
(without Fe
2+
) in Experi-
ment 1, and 0.04 kmol m
-3
Fe
3+
and 0.06 kmol m
-3
Fe
2+
in Experiment 2.
The fraction of extracted Cu was much higher in Experiment 2 ([Fe
3+
]/[T.Fe] = 0.4) than in Ex-
periment1 ([Fe
3+
]/[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 Fe
3+
/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
C
u

e
x
t
r
a
c
t
i
o
n

r
a
t
e
,

K
/

m
o
l

h
-
1
m
-
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
C
u

e
x
t
r
a
c
t
i
o
n

r
a
t
e
,

K
/

m
o
l

h
-
1
m
-
2
Normalized redox potential, E
normal
/ -

Redox Potential Control in Column Leaching of Chalcopyrite
Proceedings of Copper 2010 2759
The values of the redox potentials, E and E
normal
, were much higher in Experiment 1 (E > 0.7 V,
E
normal
> 1.5) than in Experiment 2 (E = 0.60~ 0.65 V, E
normal
= 0.5 ~ 1.0). Referring to Figures 3
and 4, the values of E and E
normal
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 Fe
3+
/total Fe ratio of 0.4 (sulfuric acid solution containing
0.04 kmol m
-3
Fe
3+
and 0.06 kmol m
-3
Fe
2+
) was continued until 90 days. The results are shown in
Figures 5 and 6.
To evaluate the effect of Cu
2+
addition on the leaching, after day 27, 0.02 kmol m
-3
Cu
2+
was added
in the solution supplied to the column top.
Figure 4: The effect of the Fe
3+
/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 Fe
3+
/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 E
normal
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 Cu
2+
.
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 Cu
2+
addition started at day 27 did not affect the Cu
extraction rate. On the other hand, Fe extraction rate decreased clearly after the 27
th
day, implying
that Cu
2+
addition suppresses Fe extraction. As a result, the Fe
2+
concentration at the bottom of the
column began to decrease at the 27
th
day and it caused increases in the redox potential E and the
normalized redox potential E
normal
at the bottom. At day 27, the values of E and E
normal
were near the
optimum values for chalcopyrite leaching (E = 0.6 V, E
normal
= 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
F
r
a
c
t
i
o
n

o
f

C
u

e
x
t
r
a
c
t
e
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
R
e
d
o
x

p
o
t
e
n
t
i
a
l
,

E

/

V

v
s
.

S
H
E
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
N
o
r
m
a
l
i
z
e
d

r
e
d
o
x

p
o
t
e
n
t
i
a
l
,

E
n
o
r
m
a
l
/

-
Time / day
Exp.1 (column bottom)
Exp.2 (column bottom)

Hiroyoshi, Kuwazawa, Takehara, Tsunekawa
Proceedings of Copper 2010 2760
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
o
C), 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
Fe
3+
and 0.06 kmol m
-3
Fe
2+
.
The concentration of Cu
2+
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 27
th
day when Cu
2+
addition was started.
Figure 6 shows the vertical distribution of Cu concentration, redox potential, E, and normalized
redox potential, E
normal
, in the leaching column at day 22 and day 72. Because Cu
2+
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 Cu
2+
addition in the solution. At both day 22
th
and 72
th
, 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, E
normal
, 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
N
o
r
m
a
l
i
z
e
d

r
e
d
o
x

p
o
t
e
n
t
i
a
l
,

E
n
o
r
m
a
l

/

-
Time / day
Column bottom
0.55
0.60
0.65
0.70
0 50 100
R
e
d
o
x

p
o
t
e
n
t
i
a
l
,

E
/

V

v
s
.

S
H
E
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
F
r
a
c
t
i
o
n

o
f

C
u

e
x
t
r
a
c
t
e
d

/

-
Time / day
Cu
Fe
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0 50 100
F
e
2
+
c
o
n
c
e
n
t
r
a
t
i
o
n

/

k
m
o
l

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
F
e
3
+
c
o
n
c
e
n
t
r
a
t
i
o
n

/

k
m
o
l

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
p
H
Time / day
Column top
Column bottom

Redox Potential Control in Column Leaching of Chalcopyrite
Proceedings of Copper 2010 2761
optimum values for chalcopyrite leaching (E = 0.6 V, E
normal
= 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
Fe
3+
and 0.06 kmol m
-3
Fe
2+
.
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 Fe
3+
/Fe
2+
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
Fe
3+
/ 0.06 kmol m
-3
Fe
2+
solution in compared
with the Cu extraction rate obtained using 0.1 kmol m
-3
Fe
3+
. 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
Fe
3+
/ 0.06 kmol m
-3
Fe
2+
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
H
e
i
g
h
t

/

m
Normalized redox potential, E
normal
/-
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
H
e
i
g
h
t

/

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
H
e
i
g
h
t

/

m
Cu concentration / kmol m
-3
Day 22th
Day 72th

Hiroyoshi, Kuwazawa, Takehara, Tsunekawa
Proceedings of Copper 2010 2762
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.
Redox Potential Control in Column Leaching of Chalcopyrite
Proceedings of Copper 2010 2763
[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.

Proceedings of Copper 2010 2764


Proceedings of Copper 2010 2765
Separation Characteristics of Chalcopyrite and
Pyrite via Bench Scale Flotation Investigations
S. Kelebek, S. Reeves Z. El Jundi
Queens 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. zdeniz
Seluk 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, zdeniz
Proceedings of Copper 2010 2766
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
Proceedings of Copper 2010 2767
2 Materials and Methods
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 m) 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
F
80
of 135 m down to 75 m.
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.


Kelebek, El Jundi, Reeves, zdeniz
Proceedings of Copper 2010 2768
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 Results and Discussion
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 F
80
of 123 m and pH 10 nearly doubles at a
60 % recovery when the particle size (F
80
) is reduced to 106 m. 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 F
80
of 89 m, 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).


Separation Characteristics of Chalcopyrite and Pyrite Using Flotation
Proceedings of Copper 2010 2769
Figure 1: Grade-Recovery performance of the sample with 1.88 % Cu under various conditions of
particle size (F
80
89-123 m), 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
C
u

G
r
a
d
e


(
%
)
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
C
h
a
l
c
o
p
y
r
i
t
e


r
e
c
o
v
e
r
y


(
%
)
Pyrite recovery (%)
F80 123 um, pH 10
F80 106, pH 10
F80 89 um, pH 11
No selectivity line

Kelebek, El Jundi, Reeves, zdeniz
Proceedings of Copper 2010 2770
However, it is clear that best selectivity is observed at the highest pH of 11 and at the finest grind
size (F
80
) of 89 m, 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

Separation Characteristics of Chalcopyrite and Pyrite Using Flotation
Proceedings of Copper 2010 2771
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 F
80
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 F
80
of around 76 m 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
P
y
r
i
t
e


R
e
c
o
v
e
r
y


(
%
)
Time (min)
F80 123 um, pH 10
F80 106, pH 10
F80 89 um, pH 11

Kelebek, El Jundi, Reeves, zdeniz
Proceedings of Copper 2010 2772
Figure 5: Grade-Recovery performance of the sample with 5.0 % Cu under various conditions of
particle size (F
80
76-135 m), 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 F
80
between 76 and 135 m. 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 F
80
of 135 m.
However, when the grind size is reduced to an F
80
of 76 m, 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
C
u
m
u
l
a
t
i
v
e

g
r
a
d
e
,

C
u


(
%
)
Cu recovery (%)
F80 134 um
F80 104 um
F80 76 um

Separation Characteristics of Chalcopyrite and Pyrite Using Flotation
Proceedings of Copper 2010 2773
Figure 6: Cu recovery as a function of time with the 5.0 % Cu sample under various conditions of
particle size (F
80
76-135 m), pH (11) and PEX (20 g/tonne), MIBC (20 g/tonne)
Figure 7: Pyrite recovery as a function of time with the 5.0 % Cu sample under various conditions
of particle size (F
80
76-135 m), 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
C
u

R
e
c
o
v
e
r
y

(
%
)
Time (Minutes)
F80 104 um
F80 134 um
F80 76 um


Kelebek, El Jundi, Reeves, zdeniz
Proceedings of Copper 2010 2774
The finer the grind size (F
80
), 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: Grade-Recovery performance of the sample with 3.35 % Cu under various conditions of
particle size (F
80
68-139 m), pH (11) and PEX (18 g/tonne), MIBC (18 g/tonne)
Figure 8 shows the copper grade-recovery response at three grind sizes (F
80
), namely, 68, 94 and
139 m while the pH and potassium ethyl xanthate levels are fixed at about 11 and 18 g/tonne,

Separation Characteristics of Chalcopyrite and Pyrite Using Flotation
Proceedings of Copper 2010 2775
respectively. Although the F
80
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 F
80
values from 90 m to 95 m.
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].
Kelebek, El Jundi, Reeves, zdeniz
Proceedings of Copper 2010 2776
Figure 9: Impact of head grade on grade-recovery performance of the samples (Grind size, F
80
90-
95 m; 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 (CuFeS
2
), 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.


Separation Characteristics of Chalcopyrite and Pyrite Using Flotation
Proceedings of Copper 2010 2777
Figure 10: Upgrading ratio of copper as a function of recovery (Grind size, F
80
90-95 m; 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 F
80

range of 135-76 m) 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
C
u

u
p
g
r
a
d
i
n
g

r
a
t
i
o
Cu Recovery (%)
5.0%Cu, F80 ~95 um
1.88%Cu, F80 ~90 um
3.35%, F80 ~93 um
3.36%Cu calc

Kelebek, El Jundi, Reeves, zdeniz
Proceedings of Copper 2010 2778
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 F
80
values ranging from 76 m to 135 m, 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
C
u


r
e
c
o
v
e
r
y


(
%
)
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

Separation Characteristics of Chalcopyrite and Pyrite Using Flotation
Proceedings of Copper 2010 2779
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, Queens 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: 6980.

Proceedings of Copper 2010 2780




Proceedings of Copper 2010 2781
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 m, Cu 1.2 %, Fe 20.4 %, S 2.4 %, SiO
2
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
Proceedings of Copper 2010 2782
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
Proceedings of Copper 2010 2783
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 (Cu
2
S/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 m) 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

Konishi, Saitoh, Shuto, Ogi, Kawakita, Kamiya
Proceedings of Copper 2010 2784
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 m. 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 cm
3
capacity [6-8]. Reactor contents were mixed at 250 rpm by a paddle impeller and air was
sparged continuously at 1000 cm
3
/min. In leaching experiments with A. brierleyi at 65 C and at a
solution pH of 1.8, the initial cell concentration was 1.010
14
cells/m
3
-suspension and the initial
mineral-liquid loading ratio was 5 kg/m
3
. In some experiments, Fe
2
(SO
4
)
3
was added so that the
initial concentration of ferric iron in the leach solution was 0.5 or 1.0 kg/m
3
. A solution sample of
2 cm
3
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, , was determined from the copper content in the solution at spe-
cific times divided by the initial copper content in the mineral samples.
Bioleaching of Crude Chalcopyrite Ores by the Thermophilic Archaean Acidianus Brierleyi
Proceedings of Copper 2010 2785
3 Results and Discussion
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 W
0
/V of 5 kg/m
3
and an initial pH of 1.8. The results of
these bioleaching experiments are shown in Figure 3, where the fraction of chalcopyrite leached, ,
and the concentration of the free cells in the leach solution, X
L
, 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
X
T0
= (1.2-1.5)10
14
cells/m
3
: () natural crude ore, an initial chalcopyrite-liquid load-
ing ratio of W
0
/V = 5 kg/m
3
; () concentrate, an initial chalcopyrite-liquid loading ratio
of W
0
/V = 0.17 kg/m
3
; () artificial crude ore (mixture of 0.17 kg/m
3
concentrate and
4.83 glass beads), an initial ore-liquid loading ratio of W
0
/V = 5 kg/m
3
.
0
0.2
0.4
0.6
0.8
1.0
C
o
p
p
e
r

l
e
a
c
h
e
d
,

[
-
]
0 2 4 6 8 10
0
2
4
6
8
C
o
n
c
.

o
f

f
r
e
e

c
e
l
l
s
,
X
L

1
0
-
1
4
[
c
e
l
l
s
/
m
3
]
Time, t [day]
0
0.2
0.4
0.6
0.8
1.0
C
o
p
p
e
r

l
e
a
c
h
e
d
,

[
-
]
0
0.2
0.4
0.6
0.8
1.0
C
o
p
p
e
r

l
e
a
c
h
e
d
,

[
-
]
0 2 4 6 8 10
0
2
4
6
8
C
o
n
c
.

o
f

f
r
e
e

c
e
l
l
s
,
X
L

1
0
-
1
4
[
c
e
l
l
s
/
m
3
]
Time, t [day]
0 2 4 6 8 10
0
2
4
6
8
C
o
n
c
.

o
f

f
r
e
e

c
e
l
l
s
,
X
L

1
0
-
1
4
[
c
e
l
l
s
/
m
3
]
Time, t [day]


Konishi, Saitoh, Shuto, Ogi, Kawakita, Kamiya
Proceedings of Copper 2010 2786
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 X
L
. The rates of bacterial growth and chalcopyrite oxidation for the crude ore
(CuFeS
2
3.3 %, Fe
3
O
4
26.7 %, and SiO
2
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 W
0
/V = 5 kg/m
3
, an initial pH of 1.8 and 65 C: () copper leaching
with A. brierleyi at an initial cell concentration of X
T0
= 1.210
14
cells/m
3
; () iron
leaching with A. brierleyi at an initial cell concentration of X
T0
= 1.210
14
cells/m
3
;
() copper leaching without A. brierleyi (sterile control); () iron leaching without
A. brierleyi (sterile control).
0 2 4 6 8 10
0
0.2
0.4
0.6
0.8
1.0
Time, t [day]
F
r
a
c
t
i
o
n

o
f

m
e
t
a
l
s

l
e
a
c
h
e
d
,


[
-
]
0 2 4 6 8 10
0
0.2
0.4
0.6
0.8
1.0
Time, t [day]
F
r
a
c
t
i
o
n

o
f

m
e
t
a
l
s

l
e
a
c
h
e
d
,


[
-
]


Bioleaching of Crude Chalcopyrite Ores by the Thermophilic Archaean Acidianus Brierleyi
Proceedings of Copper 2010 2787
Bioleaching experiments with A. brierleyi were done to examine the effect of initial ore-liquid load-
ing ratio W
0
/V. The results of these experiments are shown in Figure 5 where the concentration of
free cells in the leach solution, X
L
, and the fraction of chalcopyrite leached, , are plotted as a func-
tion of operating time. When the initial amount of the crude chalcopyrite ore, W
0
/V, was increased
from 5 to 20 kg/m
3
, the free cell concentration, X
L
, increased markedly but the leaching fraction of
copper, , changed slightly.
Figure 5: Effect of initial mineral-liquid loading ratio W
0
/V on the rate of chalcopyrite leaching
with A. brierleyi at an initial pH of 1.8 and an initial cell concentration of
X
T0
= (0.8-1.2)10
14
cells/m
3
: () W
0
/V = 5 kg/m
3
; () W
0
/V = 10 kg/m
3
;
() W
0
/V = 20 kg/m
3
; () W
0
/V = 50 kg/m
3
.

Konishi, Saitoh, Shuto, Ogi, Kawakita, Kamiya
Proceedings of Copper 2010 2788
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/m
3
. 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 W
0
/V = 5 kg/m
3
and an initial cell concentration of
X
T0
= (1.2-1.8)10
14
cells/m
3
: () pH 1.2; () pH 1.5; () pH 1.8; () pH 2.0.

Bioleaching of Crude Chalcopyrite Ores by the Thermophilic Archaean Acidianus Brierleyi
Proceedings of Copper 2010 2789
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]:
4CuFeS
2
+ 17O
2
+ 2H
2
SO
4
4CuSO
4
+ 2Fe
2
(SO
4
)
3
+ 2H
2
O (1)
When ferric ions are present in the leach solution, chalcopyrite is chemically oxidized with ferric
ion:
CuFeS
2
+ 2Fe
2
(SO
4
)
3
CuSO
4
+ 5FeSO
4
+ 2S
0
(2)
In the presence of A. brierleyi, ferrous ions and elemental sulfur are microbially oxidized:
4FeSO
4
+ O
2
+ 2H
2
SO
4
2Fe
2
(SO
4
)
3
+ 2H
2
O (3)
2S
0
+ 3O
2
+ 2H
2
O 2H
2
SO
4
(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 W
0
/V of 5 kg/m
3
(Figure 7). Chalcopyrite leaching was stimulated in the presence of A.
brierleyi. Chalcopyrite bioleaching at [Fe(III)]
0
= 0 kg/m
3
resulted in an increase in the liquid-phase
ferric ion concentration to 0.04 kg/m
3
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/m
3
, the ferric ion concentration
in the solution increased to 1.04 kg/m
3
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.
Konishi, Saitoh, Shuto, Ogi, Kawakita, Kamiya
Proceedings of Copper 2010 2790
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 W
0
/V = 5 kg/m
3
, an initial
pH of 1.2 and an initial cell concentration of X
T0
= (1.2-1.5)10
14
cells/m
3
:
() [Fe(III)]
0
= 1.0 kg/m
3
; () [Fe(III)]
0
= 0.5 kg/m
3
; () [Fe(III)]
0
= 0.0 kg/m
3
;
() [Fe(III)]
0
= 1.0 kg/m
3
; () [Fe(III)]
0
= 0.0 kg/m
3
.
0 2 4 6 8 10
0
0.2
0.4
0.6
0.8
1.0
C
o
p
p
e
r

l
e
a
c
h
e
d
,


[
-
]
Time, t [day]
0 2 4 6 8 10
0
1
2
3
4
5
6
C
o
n
c
.

o
f

f
r
e
e

c
e
l
l
s
,
X
L


1
0
-
1
4

[
c
e
l
l
s
/
m
3
]
Time, t [day]


Bioleaching of Crude Chalcopyrite Ores by the Thermophilic Archaean Acidianus Brierleyi
Proceedings of Copper 2010 2791
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 m; Cu 1.2 %, Fe 20.4 %, S 2.4 %, SiO
2
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.
Konishi, Saitoh, Shuto, Ogi, Kawakita, Kamiya
Proceedings of Copper 2010 2792
[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, 4
th
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.

Proceedings of Copper 2010 2793
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)
Schuhstrae 60
Erlangen, Germany

Keywords: Hybrid flotation, Los Pelambres
Abstract
This paper presents a study to estimate the effect of Siemens SIMINE
CIS
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 SIMINE
CIS
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
Proceedings of Copper 2010 2794
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 AMSAs 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 P
c
, the probability of attachment of particles to the bubbles P
a

and the probability of detachment of the particles from the bubble P
d
[3, 5].
( )
d a c
P P P P = 1 (1)
In literature, a generalized formula can be found for the probability of the particle collision with gas
bubbles P
c
depending on the particle size d
p
and the bubble d
b
size [5]:
n
b
p
c c
d
d
k P

= (2)
The Parameters k
c
and n are dependent on the Reynolds number Re i.e. the hydrodynamic milieu
during the flotation process.
Hybrid Flotation
Proceedings of Copper 2010 2795
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
d
d
d
P (3)
where d
p,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 SIMINE
CIS
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.
Krieglstein, Grossmann
Proceedings of Copper 2010 2796
Figure 1: SIMINE
CIS
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
SIMINE
CIS
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 m/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
Proceedings of Copper 2010 2797
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
SIMINE
CIS
HybridFlot flotation cell was implemented in the existing flotation circuit as a pre-
rougher cell. The concentrate from the SIMINE
CIS
HybridFlot flotation cell was led to a pneumatic
cell which was operated as cleaner and produced final concentrate. The tailings from the SIMINE
CIS

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 SIMINE
CIS
HybridFlot flota-
tion cell. During the in-plant testing of the SIMINE
CIS
HybridFlot flotation cell at Los Pelambres,
samples were collected from the feed, concentrate and tailings flows. Primarily, the SIMINE
CIS
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 SIMINE
CIS
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
SIMINE
CIS
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-

Krieglstein, Grossmann
Proceedings of Copper 2010 2798
ventional mechanical rougher cells and concentrate from the SIMINE
CIS
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 Q
s
= 120 t/h.

Hybrid Flotation
Proceedings of Copper 2010 2799
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 SIMINE
CIS
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 SIMINE
CIS
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 SIMINE
CIS
HybridFlot flotation cell
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 SIMINE
CIS
HybridFlot flotation cell is integrated into the flotation cir-
cuit. The highest molybdenum grade in the concentrate was achieved at low solid mass flow
Q
s
= 100 t/h in the investigated range.
By integrating the SIMINE
CIS
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 SIMINE
CIS
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
100
R
e
c
o
v
e
r
y

i
n

%
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)

Krieglstein, Grossmann
Proceedings of Copper 2010 2800
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 m ON 97.5 % 5.9 %
OFF 95.2% 6.6 %
10 m < d < 44 m ON 97 % 6.9 %
OFF 96 % 6.9 %
44 m < d < 74 m ON 97.8 % 38.1 %
OFF 96.3 % 29.9 %
74 m < d ON 96 % 37 %
OFF 90.2 % 29.7 %

100 120 140
0
5
10
15
20
25
30
35
C
o
n
t
e
n
t

i
n

%
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
Proceedings of Copper 2010 2801
The highest increase in recovery is achieved for the particle fractions with a mean particle diameter
d > 74 m and d < 10 m. 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 SIMINE
CIS
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 SIMINE
CIS
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 (d
50
= 18 m) is fine.
Figure 6: Cumulative undersize of the feed measured by a sieve test
The cumulative undersize of the concentrate of the SIMINE
CIS
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 m. This underlies the assumption that

Krieglstein, Grossmann
Proceedings of Copper 2010 2802
the SIMINE
CIS
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 SIMINE
CIS
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 SIMINE
CIS
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 SIMINE
CIS
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 SIMINE
CIS
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 SIMINE
CIS
HybridFlot flotation cell
All results presented so far show that an implementation of a single SIMINE
CIS
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
Proceedings of Copper 2010 2803
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 SIMINE
CIS

HybridFlot flotation cell are summarized in Table 2, classified by 2 different particle size fractions.
Table 2: Molybdenum recovery and grade with SIMINE
CIS
HybridFlot flotation cell
Particle Size Recovery Grade Enrichment factor approx.
d < 44 m 18.5 % 15.9 % -
d > 44 m 37.0 % 50.0 % -
Overall 22.1 % 24.0 % 8 10
The overall recovery of the SIMINE
CIS
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 SIMINE
CIS
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 SIMINE
CIS
HybridFlot flotation cell.
6 Conclusions
The test with the first SIMINE
CIS
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 SIMINE
CIS
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 m), but also in the coarser particle size fractions (d > 74 m). 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 SIMINE
CIS
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 SIMINE
CIS
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 SIMINE
CIS
HybridFlot flotation cell into the process as pre-rougher
stage.
Krieglstein, Grossmann
Proceedings of Copper 2010 2804
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 SIMINE
CIS
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 SIMINE
CIS
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 SIMINE
CIS
HybridFlot flotation cells if applicable in different sizes.
Prospectively, Siemens is also striving to transfer its SIMINE
CIS
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
Proceedings of Copper 2010 2805
References
[1] WILLS, B. & NAPIER-MUNN, T. (2006): Wills 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 ber Flotationsgrundlagen unter besonderer Bercksichtigung der
Korngre, 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

Proceedings of Copper 2010 2806


Proceedings of Copper 2010 2807
Perspectives of Copper Mining Industry
Development in Poland
Ph.D. Eng. Jan Kudeko, Ph.D. Jacek Pyra, Ph.D. Eng. Jerzy Sobociski
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.
Kudeko, Pyra, Sobociski
Proceedings of Copper 2010 2808
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
Proceedings of Copper 2010 2809
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 D
1
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 D
2
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).
Kudeko, Pyra, Sobociski
Proceedings of Copper 2010 2810

Table 1: Reserves, supposed economic and not supposed economic resources of copper KGHM
Polska Miedz S. A. deposit as for 31.12.2005

Perspectives of Copper Mining Industry Development in Poland
Proceedings of Copper 2010 2811
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 - -
C
1
6856 108.3 2293 31.2
C
2
28,244 348 927 10.8
subtotal 35,613 463.8 3220 42.0
In protection
pillars
A+B 1,740 29.2 - -
C
1
8223 155.4 7 0.1
C
2
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 C
1
20,153 271 7965 88
In protection pillars C
1
17,650 327 1396 15
Total C
1
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
[km
2
]
Thickness of
copper
bearing
inteval
[m]
Average
content of
Cu
e

[%]
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
Kudeko, Pyra, Sobociski
Proceedings of Copper 2010 2812
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.
Perspectives of Copper Mining Industry Development in Poland
Proceedings of Copper 2010 2813
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).
Kudeko, Pyra, Sobociski
Proceedings of Copper 2010 2814
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).


Perspectives of Copper Mining Industry Development in Poland
Proceedings of Copper 2010 2815
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.


Kudeko, Pyra, Sobociski
Proceedings of Copper 2010 2816
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 km
2
. 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.
Perspectives of Copper Mining Industry Development in Poland
Proceedings of Copper 2010 2817
Figure 5: Output in KGHM Polska Miedz S. A. mines
Figure 6: Concentrate production in KGHM Polska Miedz S. A.


Kudeko, Pyra, Sobociski
Proceedings of Copper 2010 2818
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.


Perspectives of Copper Mining Industry Development in Poland
Proceedings of Copper 2010 2819
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]:



Kudeko, Pyra, Sobociski
Proceedings of Copper 2010 2820
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 flotation processing 10 USD/Mg,
unit costs of metallurgical processing Cu 870 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 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).
Perspectives of Copper Mining Industry Development in Poland
Proceedings of Copper 2010 2821
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


Kudeko, Pyra, Sobociski
Proceedings of Copper 2010 2822
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 flotation processing 10 USD/Mg,
unit costs of metallurgical processing Cu 870 USD/Mg Cu,
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.
Perspectives of Copper Mining Industry Development in Poland
Proceedings of Copper 2010 2823
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



Kudeko, Pyra, Sobociski
Proceedings of Copper 2010 2824
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.

Perspectives of Copper Mining Industry Development in Poland
Proceedings of Copper 2010 2825
Figure 16: Predicted production of metallic Ag from own KGHM resources
Figure 17: Predicted net profit from electrolytic Cu production from own KGHM resources


Kudeko, Pyra, Sobociski
Proceedings of Copper 2010 2826
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.

Perspectives of Copper Mining Industry Development in Poland
Proceedings of Copper 2010 2827
References
[1] KGHM Polska Miedz S.A.: Annual Report 1998, Lubin.
[2] KGHM Polska Miedz S.A.: Annual Report 2005, Lubin.
[3] KGHM Polska Miedz S.A.: Annual Report 2006, Lubin.
[4] KGHM Polska Miedz S.A.: Annual Report 2007, Lubin.
[5] KGHM Polska Miedz S.A.: Annual Report 2008, 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.

Proceedings of Copper 2010 2828


Proceedings of Copper 2010 2829
Control of Bubble Size in a Laboratory Flotation
Column
Miguel Maldonado, Dr. Andr Desbiens, Dr. ric Poulin Alberto Riquelme
Dr. Ren del Villar
Universit Laval Universidad de Concepcin
Department of Electrical and Computer Engineering Department of Electrical Engineering
Department of Mining, Materials and Metallurgical Engineering
LOOP (Laboratoire dobservation et doptimisation des procds)
Qubec (QC), Canada Concepcin, 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 S
b
) 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 S
b
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
Proceedings of Copper 2010 2830
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
6
g
b
J
S
d
= (1)
where J
g
is the superficial gas velocity and d
32
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
Proceedings of Copper 2010 2831
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
1
32
2
1
N
i
i
N
i
i
d
d
d
=
=
=

(2)
where d
i
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
o
C 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
o
ref
g g
a
T C
J J
P cmH O
( ( +
=
( (
+

(3)
where P
a
is the absolute pressure measured in cm H
2
O, T is the temperature in degrees Celsius and
J
g
ref
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:
(%) 100
g
P
L


= (4)
where P is the pressure differential in cm H
2
O 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
Maldonado, Desbiens, Poulin, del Villar, Riquelme
Proceedings of Copper 2010 2832
modified version of the McGills bubble viewer [11] was implemented to measure bubble size. The
image processing system was performed by a dedicated computer using MatLabs 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.

Control of Bubble Size in a Laboratory Flotation Column
Proceedings of Copper 2010 2833
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 (J
ls
), i.e. the water flow rate through the gap divided by the cross sectional area of the
gap:
ls
ls
gap
Q
J
A
= (5)
where Q
ls
is the volumetric water flow rate of liquid in the sparger and A
gap
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
j
(x) called component densities as follows:
Porous
ring
gap
sleeve
Top
View
Maldonado, Desbiens, Poulin, del Villar, Riquelme
Proceedings of Copper 2010 2834
1
( ) ( )
M
j j
j
f x w x
=
=

(6)
1
1, 0 1, 1, ,
M
j j
j
w w j M
=
= =

(7)
( ) 1, 1, ,
j
x dx j M
+

= =

(8)
where M is the model order, i.e. the number of components considered. The coefficients w
j
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
j
(x) are chosen to be fixed Gaussian functions.
( )
2
2
1 1
( ) exp
2 2
j
j
x
x


=
`


)
(9)
where
j
, j =1,,M , while M and are fixed parameters.
Minimizing the log-likelihood of the samples and applying a stochastic gradient descent approach
the following recursive equation is obtained [17]:
( 1) ( ) ( / ( 1)) ( )
j j j
w k w k P j x k w k ( + = + +

(10)
When is selected to be constant, it introduces an exponentially decaying envelop (1- )
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 j
x k w
P j x k
f x k

= (11)
The Sauter mean bubble diameter can be estimated on-line from the Gaussian mixture parameters as
follows [19]:
( )
( )
3 2 2
1
32
2 2
2
1
, ( ) ( 3 )

( )
( )( )
,
M
j j j
j
M
j j
j
x f x k dx w k
d k
w k
x f x k dx


+
=

+
=

+
= =
+

(12)
Control of Bubble Size in a Laboratory Flotation Column
Proceedings of Copper 2010 2835
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 ( = 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.


Maldonado, Desbiens, Poulin, del Villar, Riquelme
Proceedings of Copper 2010 2836
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.


Control of Bubble Size in a Laboratory Flotation Column
Proceedings of Copper 2010 2837
Figure 7 shows the static gain (d
32
/J
ls
) 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 0.226
ls
d J

= (13)
where
32
d 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
s
e
G s
s

=
+

(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

Maldonado, Desbiens, Poulin, del Villar, Riquelme
Proceedings of Copper 2010 2838
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 H
2
norm of the error e for a given input v:
2
( ) ( ) 2
min min 1 ( ) ( ) ( )
Q s Q s
e G s Q s v s ( =


(15)
For a step input, i.e. v(s) = 1/s and a stable system factored into
( ) ( ) ( )
M A
G s G s G s =

where
( )
A
G s

is an
all-pass transfer function, containing right half plane zeros and delays, and
( )
M
G s
a minimum phase
transfer function, the optimal H
2
controller is:
1
( ) ( )
M
Q s G s

=

(16)
Then, the controller is augmented by a filter F(s):
( ) ( ) ( ) Q s Q s F s =

(17)
where the filter F(s) is usually selected to be:
( )
1
( )
1
r
F s
s
=
+
(18)
0 2000 4000 6000 8000 10000 12000 14000 16000 18000
0
50
100
150
200
250
300
350
time (s)
i
n
t
e
r
m
e
d
i
a
t
e

v
a
r
i
a
b
l
e
,

v
(
t
)
v(t)
estimated
v(t)

Control of Bubble Size in a Laboratory Flotation Column
Proceedings of Copper 2010 2839
and r is selected to obtain a realizable controller. The complementary sensitivity function is given
by:
( , ) ( )
( , )
1 ( , ) ( ) ( )
Q s G s
T s
Q s G s G s

=
( +

(19)
The maximum peak (Mp) tuning [20] is used to find the value of IMC filter 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
s
Q 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.

Maldonado, Desbiens, Poulin, del Villar, Riquelme
Proceedings of Copper 2010 2840
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 J
ls
= 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.

Control of Bubble Size in a Laboratory Flotation Column
Proceedings of Copper 2010 2841
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

Maldonado, Desbiens, Poulin, del Villar, Riquelme
Proceedings of Copper 2010 2842
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 Universit Laval.
Control of Bubble Size in a Laboratory Flotation Column
Proceedings of Copper 2010 2843
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, 5
th
International Mineral Processing
Seminar, Santiago, Chile, pp.215220.
[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, 5
th
International Mineral Processing Seminar, Santiago, Chile, pp. 315325.
[5] MOILANEN, J., REMES, A. (2008): Control of the flotation process. Procemin 2008, 5
th
In-
ternational Mineral Processing Seminar, Santiago, Chile, pp. 305313.
[6] LIU, J.J., MACGREGOR, J.F. (2008): Froth-based modeling and control of flotation proc-
esses. Minerals Engineering, Vol.21, pp.642651.
[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 DURO, 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 GONALVES, 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] OKEEFE, C., VIEGA, J., and FERNALD, M. (2007): Application of passive sonar technolo-
gy to mineral processing and oil sands applications. Proceedings of the 39
th
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. 48
th
Annual Conference of
Metallurgists, Sudbury, Canada, pp.35-146.
Maldonado, Desbiens, Poulin, del Villar, Riquelme
Proceedings of Copper 2010 2844
[14] NESSET, J.E., HERNANDEZ-AGUILAR, J., ACUA, 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] TRVN, 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, 5
th
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.

Proceedings of Copper 2010 2845
Flow Process in the Aerator of the
Flotation Machine Preliminary Simulations
Adam Maka, Adam Fic, Andrzej Sachajdak, Ireneusz Szczygie
Silesian University of Technology
Institute of Non-ferrous Metals, Sowiskiego 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.
Maka, Fic, Sachajdak, Szczygie
Proceedings of Copper 2010 2846
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
Proceedings of Copper 2010 2847
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.

Maka, Fic, Sachajdak, Szczygie
Proceedings of Copper 2010 2848
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 = +

t
p
Dt
D
=

w
w)
2
(

Flow Process in the Aerator of the Flotation Machine Preliminary Simulations


Proceedings of Copper 2010 2849
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.

Maka, Fic, Sachajdak, Szczygie
Proceedings of Copper 2010 2850
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


Flow Process in the Aerator of the Flotation Machine Preliminary Simulations
Proceedings of Copper 2010 2851
Figure 6: Velocity distribution at the outlet of traditional stator
Figure 7: Velocity distribution at the outlet of modified stator


Maka, Fic, Sachajdak, Szczygie
Proceedings of Copper 2010 2852
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


Flow Process in the Aerator of the Flotation Machine Preliminary Simulations
Proceedings of Copper 2010 2853
Figure 10: Velocity vectors at the outlet of the stator
Figure 11: Velocity vectors at the outlet of the rotor


Maka, Fic, Sachajdak, Szczygie
Proceedings of Copper 2010 2854
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


Flow Process in the Aerator of the Flotation Machine Preliminary Simulations
Proceedings of Copper 2010 2855
Figure 14: Flotation machine at work
Figure 15: Outlet of rotor


Maka, Fic, Sachajdak, Szczygie
Proceedings of Copper 2010 2856
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.

Flow Process in the Aerator of the Flotation Machine Preliminary Simulations
Proceedings of Copper 2010 2857
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 Users 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., GRNSTRAND S., Numerical modeling of an Outokumpu flota-
tion device. 3
th
Int. Conf. on CFD In The Minerals and Process Industries, Melbourne,
Australia, 2003.

Proceedings of Copper 2010 2858


Proceedings of Copper 2010 2859
Analysis of Fine Particles Behaviour in Flotation
of Polish Copper Ores
Ph.D. Eng. A. Potulska
KGHM CUPRUM Ltd Research & Development Centre
Street Gen. Wadysawa Sikorskiego 2-8
53-659 Wrocaw, 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
Proceedings of Copper 2010 2860
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
Proceedings of Copper 2010 2861
2 Characteristic of Polish copper ores
The copper deposit mined by KGHM Polska Mied 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 Mied 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
Proceedings of Copper 2010 2862
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 Mied 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 C
org
, 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.
Analysis of Fine Particles Behaviour in Flotation of Polish Copper Ores
Proceedings of Copper 2010 2863
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 Lubins 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 Mied 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
Proceedings of Copper 2010 2864
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
Analysis of Fine Particles Behaviour in Flotation of Polish Copper Ores
Proceedings of Copper 2010 2865
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 dm
3
/h of air, impeller rotati-
on 670 rpm, collector consumption (potassium xanthate) 100 g/Mg and frother consumption (
terpineol) 100 g/Mg of ore. The achieved results were recalculated using Fuerstenau`s curves for the
same amount of gangue recovery in tailings
r
= 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 =87.4 % =7.8 % =23.2 %
80 % <0.025 mm =88.8 % =7.9 % =23.3 %
90 % <0.025 mm =89.8 % =8.0 % =23.3 %
copper recovery in concentrate,
r
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, %
R
e
c
o
v
e
r
y

o
f

g
a
n
g
u
e

m
i
n
e
r
a
l
s

i
n

t
a
i
l
i
n
g
s
,

%
brak wzbogacania
65% <0,025 mm
80% <0,025 mm
90% <0,025 mm


0
20
40
60
80
100
85 90 95 100
Cu recovery in concentrate, %
R
e
c
o
v
e
r
y

o
f

g
a
n
g
u
e

m
i
n
e
r
a
l
s

i
n

t
a
i
l
i
n
g
s
,

%
brak wzbogacania
60
80
110

Potulska
Proceedings of Copper 2010 2866
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

r
= 80 % the following concentration results were obtained:
60 dm
3
/h =86.2 % =7.8 % =23.1 %
80 dm
3
/h =89.8 % =8.0 % =23.3 %
110 dm
3
/h =88.8 % =7.9 % =23.4 %
copper recovery in concentrate,
r
gangue recovery in tailings.
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

r
= 80 %, the following concentration indexes were obtained:
590 rpm =86.1 % =7.7 % =23.3 %
670 rpm =89.8 % =8.0 % =23.3 %
790 rpm =88.8 % =7.9 % =23.2 %
copper recovery in concentrate,
r
gangue recovery in tailings.
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 dm
3
/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
Analysis of Fine Particles Behaviour in Flotation of Polish Copper Ores
Proceedings of Copper 2010 2867
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 dm
3
/h, impeller rotation 670 rpm
The results were recalculated basing on Fuerstenau`s curves for the same value of gangue recovery
in tailings
r
= 80 % (see Figure 3). The following concentration results were obtained:
70 g/Mg =89.1 % =8.0 % =23.3 %
100 g/Mg =89.8 % =8.0 % =23.3 %
150 g/Mg =87.8 % =7.9 % =23.2 %
copper recovery in concentrate,
r
gangue recovery in tailings.
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
Cu recovery in concentrate, %
R
e
c
o
v
e
r
y

o
f

g
a
n
g
u
e

m
i
n
e
r
a
l
s

i
n

t
a
i
l
i
n
g
s
,

%
brak wzbogacania
70 g/Mg
100 g/ Mg
150 g/ Mg

no concentration
Potulska
Proceedings of Copper 2010 2868
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
C
u

r
e
c
o
v
e
r
y
,

%

Analysis of Fine Particles Behaviour in Flotation of Polish Copper Ores
Proceedings of Copper 2010 2869
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.

>0.025 0.015-0.025 0.010-0.015 <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
Proceedings of Copper 2010 2870
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 plants 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 plants 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.
Analysis of Fine Particles Behaviour in Flotation of Polish Copper Ores
Proceedings of Copper 2010 2871
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, 117
[3] PEASE J. D., CURRY D., YOUNG M. F., (2006): Designing flotation circuits for high fines
recovery. Minerals Engeneering, 19: 831840
[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), 18
[5] GAUDIN A. M., (1963): Flotacja. Tum. z II wydania oryginau: J. Olszewski, T. Piaseczny.
Wyd. lsk, 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, 151166
[7] TRAHAR W. J., WARREN L. J., (1976): The flotability of very fine particles a review.
Inter-national Journal of Mineral Processing, 3: 103131
[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): Wpyw drobnego mielenia na flotacj rud cynkowo-oowiowych.
Prace Naukowe Instytutu Gornictwa Politechniki Wrocawskiej, nr 116. Seria: Konferencje ;
nr 47, Wrocaw
[10] BORTEL R, (1967): Wpyw mineraw ilastych na flotowalno siarczkw metali niee-
laznych. Praca Doktorska, Politechnika lska, Gliwice
[11] WIENIEWSKI A., (1994): Problemy zwikszenia selektywnoci wzbogacania flotacyjnego
rud metali nieelaznych. Konferencja Komitetu Grnictwa PAN Aktualne zadania nauki w
grnictwie, Ustro
[12] GROTOWSKI A., (1997): Koncepcja rozwiza przerbki z zastosowaniem nowych tech-
nologii flotacji w aspekcie kompleksowego zagospodarowania grniczego wynikajcego z
etapu I, sprawozdanie CBPM Cuprum, Wrocaw
[13] USZCZKIEWICZ A., WIENIEWSKI A., (2006): Kierunki rozwoju technologii
wzbogacania rud w krajowym przemyle miedziowym. Grnictwo i Geoinynieria, Rok 30,
Zeszyt 3/1, AGH Uczelniane Wydawnictwa Naukowo-Dydaktyczne, ISSN: 1732-6702, 181
196
Potulska
Proceedings of Copper 2010 2872
[14] GROTOWSKI A., PAKULSKA B., MIZERA W., (2004): Wstpne okrelenie moliwoci
zastosowania mynw 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 Mied S.A. cz II, Praca zbiorowa pod redakcj Piestrzyski A., Wyd. KGHM Cuprum
CBR, Spka z o. o., Wrocaw, Lubin, 463472
[16] SPALISKA B., STEC R., SZTABA K., (2007): Miejsce i rola przerbki rudy w kompleksie
technologicznym KGHM Polska Mied S.A.. In: Monografia KGHM Polska Mied S.A..
cz II, Praca zbiorowa pod redakcj Piestrzyski A., Wyd. KGHM Cuprum CBR, Spka z
o. o., Wrocaw, Lubin, 463472
[17] TUMIDAJSKI T. et al., (2007-2008): Report of Investigation, Project No. KGHM-ZW-U-
0011
[18] GROTOWSKI A., DBKOWSKI R., B., HENZEL S., 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): Wpyw drobnego mielenia na flotacj krajowych rud miedzi. Roz-
prawa Doktorska, Wrocaw
[20] DRZYMAA 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, 5565
[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, 198210

Proceedings of Copper 2010 2873
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
Proceedings of Copper 2010 2874
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 1980s) 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 1990s.
The cement industry was the first to accept this new technology in the late 1980s 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 McMoRans 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 HPGRs 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
Proceedings of Copper 2010 2875
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 HPGRs 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:

Rana, Chandrasekaran, Wood
Proceedings of Copper 2010 2876
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 authors 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 versus SAG Milling Technology in Hard-Rock Mining Review and Analysis
Proceedings of Copper 2010 2877
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),
Kppern,
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 Kppern 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 Kppern 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.
Rana, Chandrasekaran, Wood
Proceedings of Copper 2010 2878
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.
HPGR versus SAG Milling Technology in Hard-Rock Mining Review and Analysis
Proceedings of Copper 2010 2879
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].
Rana, Chandrasekaran, Wood
Proceedings of Copper 2010 2880
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
HPGR versus SAG Milling Technology in Hard-Rock Mining Review and Analysis
Proceedings of Copper 2010 2881
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 %
Rana, Chandrasekaran, Wood
Proceedings of Copper 2010 2882
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

Proceedings of Copper 2010 2883
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
11 TegataGakuen cho
Akita City 0108502, Japan

Keywords: Selective leaching, arsenic, copper resources, enargite, sodium hydrosulfide
Abstract
Conventional smelting processes cannot treat Cuores/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 (Cu
3
AsS
4
) contained in Cuores/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 Cu
3
AsS
4
was transformed to Cu
2
S.
1 Introduction
Enargite (Cu
3
AsS
4
) and tennantite (Cu
12
As
4
S
13
) are the main minerals with high As content that
associate with copper sulfides such as covellite (CuS), chalcocite (Cu
2
S), chalcopyrite (CuFeS
2
),
Bornite (Cu
5
FeS
4
), 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
Proceedings of Copper 2010 2884
options will be sought. Reviews on Curesources containing enargite are covered widely in litera-
ture by many researchers [1-8].
A method to remove As from the Cuores/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 Cu
2
S by selective dis-
solution of As into solution with sodium hypochlorite (NaClO) or sodium sulfide (Na
2
S) 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 (Cu
3
AsS
4
) to Cu
2
S in NaHSNaOH 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 (Na
2
S)
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 Na
2
S (~13 %) since
leaching of As from enargite is a function of sulfur [S
2-
] 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 m 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
Proceedings of Copper 2010 2885
Table 1: Main chemical and phase compositions in Cuores/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 CuFeS
2
, FeS
2

(2) Cu conc. 0.95 22.60 18.17 CuFeS
2
, FeS
2

(3) Cu conc. 1.67 11.49 9.71 CuFeS
2
, FeS
(4) Cu conc. 3.11 33.46 14.34 FeS
2
, Cu
2
FeSiS
4

(5) Enargite ore 8.32 33.71 16.34 Cu
3
AsS
4
, CuFeS
2

(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 Cu
3
AsS
4
, FeS
2

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 (50200) 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.
Tongamp, Takasaki, Shibayama
Proceedings of Copper 2010 2886
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, RINT2200 / PC system with
a CuK irradiation source ( = 1.5405 ) 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 (ICPAES/OES) atomic/optical emission spectrometer, SPS3000 (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 m 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 +
H
2
O
Sample preparation
Na
3
AsS
3

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
Selective Leaching of Arsenic from Copper Ores and Concentrates
Proceedings of Copper 2010 2887
Figure 2: Arsenic dissolution behavior as a function of leaching time for a 1.0 g enargite ore (8)
Leaching of -75+32 m 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 80C 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
S
2-
/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


A
r
s
e
n
i
c

r
e
m
o
v
a
l

(
%
)
50 C


60 C



70 C
80 C


95 C


Tongamp, Takasaki, Shibayama
Proceedings of Copper 2010 2888
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
(S
2-
/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 (S
2-
/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 (S
2-
/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
9
5
.
4
0

%
8
7
.
9
1

%
4
7
.
1
7

%
1
0
.
0
5

%


95
o
C 80
o
C 60
o
C 30
o
C
A
r
s
e
n
i
c

d
i
s
s
o
l
u
t
i
o
n

(
%
)
Leaching temperature, T /
o
C
Selective Leaching of Arsenic from Copper Ores and Concentrates
Proceedings of Copper 2010 2889
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 (S
2-
/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 Cuconcentrate samples mixed with enargite ore within 6 h.

0 1 2 3 4
0
20
40
60
80
100


(c): S/As = 6.1
(b): S/As = 12.3
(a): S/As = 12.3
A
r
s
e
n
i
c

d
i
s
s
o
l
u
t
i
o
n

(
%
)
Leaching time, t / hours
Tongamp, Takasaki, Shibayama
Proceedings of Copper 2010 2890
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 Cu
3
AsS
4
from leach residues
obtained after 1 h. On the other hand, characteristic peaks of Cu
2
S 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)
A
r
s
e
n
i
c

d
i
s
s
o
l
u
t
i
o
n

(
%
)
Leaching time, t / hours
100 g/L (S/As = 61)
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9 10
A
r
s
e
n
i
c

d
i
s
s
o
l
u
t
i
o
n

(
%
)
200 g/L NaHS (S/As = 6.7)
Leaching time, t / hours

100 g/L NaHS (S/As = 3.3)
150 g/L NaHS (S/As = 5.1)
Selective Leaching of Arsenic from Copper Ores and Concentrates
Proceedings of Copper 2010 2891
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 (Cu
3
AsS
4
) to Cu
2
S 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 S
2-
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 S
2-
ions to HS
-
, and maintains a high concentration of S
2-
ions
to dissolve As from Cu
3
AsS
4
.
2Cu
3
AsS
4
+ 3NaHS + 3NaOH = 2Na
3
AsS
4
+ 3Cu
2
S + 3H
2
O (1)
NaHS Na
+
+ HS
-
(2)
HS
-
+ OH
-


S
2-
+ H
2
O (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
Cu
3
AsS
4
to Cu
2
S as represented by Reaction (1) in NaHS media is similar that for Na
2
S. 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
2 /degree
0h
1h
3h

Cu
3
AsS
4
Cu
2
S
FeS
2
SiO
2 6h

Tongamp, Takasaki, Shibayama


Proceedings of Copper 2010 2892
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-)
1/3
and 1-(1-)
2/3
(interface/surfacechemical reaction control models) and diffusion controlled mechanisms;
1-3 (1-)
2/3
+ 2 (1-) and (1-2/3)-(1-)
2/3
.
Selective Leaching of Arsenic from Copper Ores and Concentrates
Proceedings of Copper 2010 2893
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:
13(1)
2/3
+2(1) = kt (4)
In the expression, k is the linear kinetic constant; t is the reaction time and , represents fraction
arsenic removed from enargite. A linear relationship from the plot of reaction time (t) versus
1-3 (1-)
2/3
+2 (1) 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 NaHSNaOH 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 m) and (-75+32 m) 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
313K
298K
1

3
(
1

)
2
/
3
+
2
(
1

)
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
RootMSE (SD) = 0.29009
l
n

k
1/T
E
a
= 70.26 4.74 kJ/mol
Tongamp, Takasaki, Shibayama
Proceedings of Copper 2010 2894
To investigate effect of NaHS concentration, the (-75+32 m) 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 S
2-
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 m) 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 NaHSNaOH 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 (Cu
3
AsS
4
) 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 Cuores/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 (Cu
3
AsS
4
) was transformed to Cu
2
S as a result of reduction of As
5+
in enargite being
reduced to As
3+
(Na
3
AsS
4
) by S
2-
offered by NaHS. Xray 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 (Cu
3
AsS
4
) and subsequent formation of Cu
2
S in
NaHSNaOH system occurs according to the reaction:
2Cu
3
AsS
4
+ 3NaHS + 3NaOH = 2Na
3
AsS
4
+ 3Cu
2
S + 3H
2
O.
4. The kinetics of As leaching from enargite in NaHSNaOH solution was well represented by the
diffusion controlled model; kt = 1-3 (1-)
2/3
+2 (1-) and the activation energy was calculated to
be 70.26 4.74 kJ/mol.
Selective Leaching of Arsenic from Copper Ores and Concentrates
Proceedings of Copper 2010 2895
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, EarthScience Reviews, 2008, 86, pp. 6288.
[2] FILIPPOU D., STGERMAIN P., GRAMMATIKOPOULOS.: Recovery of metal values
from CopperArsenic minerals and other related resources, Mineral Processing & Extractive
Metall. Rev., 2007, 28, pp. 247298.
[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] BAL P. and ACHIMOVIOV M.: Selective leaching of antimony and arsenic from me-
chanically activated tetrahedrite, jamesonite and enargite, Int. J. Miner. Process, 2006, 81, pp.
4450.
[5] GUO H. and YEN W.T.: Selective flotation of enargite from chalcopyrite by electrochemical
control. Miner. Engin., 2005, 18, pp. 605612.
[6] VIALS J., ROCA A., HERNNDEZ M.C., BENAVENTE O.: Topochemical transforma-
tion of enargite into copper oxide by hypochlorite leaching, Hydrometallurgy, 2003, 68, pp.
183193.
[7] WELHAM N.J.: Mechanochemical processing of enargite (Cu
3
AsS
4
), Hydrometallurgy, 2001,
62, pp. 165173.
[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. 603612.
[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. 1526.
[10] KANTAR C.: Solution and flotation chemistry of enargite, Colloids and Surfaces A: Physical
and Engineering Aspects, 2002, 210, pp. 2331.
[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. 109119.
Tongamp, Takasaki, Shibayama
Proceedings of Copper 2010 2896
[12] PAUPORT 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. 119.
[13] MENACHO J.M., ALIAGA, W., VALENUELA R., RAMOS V., OLIVARES L.: Selective
flotation of enargite and chalcopyrite, Minerals, 1993, 48 (203), pp. 3339.
[14] BAL P. and ACHIMOVIOV M.: Mechanochemical leaching in hydrometallurgy of
complex sulphides, Hydrometallurgy, 2006, 84, pp. 6068.
[15] CURRELI L., GHIANI M., SURRACCO M., ORR G.: Beneficiation of a gold bearing
enargite ore by flotation and As leaching with Nahypochlorite. Minerals Engineering, 2005,
18, pp. 849854.
[16] DELFINI M., FERRINI M., MANNI A., MASSACCI P., PIGA L.: Arsenic leaching by Na
2
S
to decontaminate tailings coming from colemanite processing, Miner. Engin., 2003, 16, pp.
4550.
[17] BAL P., ACHIMOVIOV M., BASTL Z., OHTANI T., SNCHEZ M.: Influence of
mechanical activation on the alkaline leaching of enargite concentrate, Hydrometallurgy,
2000, 54, pp. 205216.
[18] HERREROS R.O., QUIROZ HERNNDEZ M.C., VIALS J.: Dissolution kinetics of enar-
gite in dilute Cl
2
/Cl media, Hydrometallurgy, 2002, 64, pp. 153160.
[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. 469476.
[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.

Proceedings of Copper 2010 2897
Copper Leaching from Molybdenite in Acidic
FeCl
3
Solutions with FeCl
2

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 FeCl
3
-FeCl
2
-HCl. The results showed
that leaching temperature, solid to liquid ratio and the concentration of FeCl
3
and FeCl
2
(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 FeCl
2
. 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 (MoS
2
) 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 MoS
2
, 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 MoS
2
.
Copper occurs mainly as covellite (CuS), chalcopyrite (CuFeS
2
), chalcocite (Cu
2
S) 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], FeCl
3
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
Proceedings of Copper 2010 2898
Cu
2+
, Fe
3+
, O
2
, Cl
2
and so on [5]. Among them, Fe
3+
is more efficient and can be recovered more
easily than others. When it comes to acid, HCl and H
2
SO
4
are commonly used in copper leaching and
HCl made the leaching more easily than H
2
SO
4
[6, 7].
FeCl
3
and HCl were selected as the oxidizing agents and acid respectively in the leaching solutions
for this work. Moreover, FeCl
2
was also added to the solution. On one side, FeCl
2
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 [Fe
3+
] / [Fe
2+
] ratio, decreases with FeCl
2
addition.
Mineralogical analysis showed that CuS and CuFeS
2
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 Fe
3+
and Fe
2+
. 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 MoS
2
, and the
impurities were CuS, CuFeS
2
and SiO
2
.
Copper Leaching from Molydenite in Acidic FeCl
3
Solutions with FeCl
2

Proceedings of Copper 2010 2899
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-SiO
2
, 3-MoS
2
, 4-CuFeS
2
, 5-MoS
2
)
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, 04 h; temperature,
70100 C; solid / liquid ratio, 0.1-0.35 kg/L; the concentration of FeCl
3
, 100-200 g/L; the
concentration of FeCl
2
, 200-300 g/L; stirring speed, 300 rpm; and pH=0. The leaching solution was
filtrated and analyzed by atomic absorption spectrophotometer (AAS).

Zhang, Tumen-Ulzii, Li
Proceedings of Copper 2010 2900
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 Results and discussion
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-H
2
O and Fe-Cl-H
2
O systems. From
Figure 3 it can be seen that both CuS and CuFeS
2
dissolve at pH lower than 4. A comparison of
Figure 4 and Figure 3 shows the redox potential of Fe
3+
is much higher than that of CuFeS
2
and
CuS, and can dissolve them at pH lower than 0. So FeCl
3
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 FeCl
3
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.
S FeCl CuCl FeCl CuFeS 2 5 4
2 2 3 2
+ + = + (1)
S FeCl CuCl FeCl CuS + + = +
2 2 3
2 2 (2)
Copper Leaching from Molydenite in Acidic FeCl
3
Solutions with FeCl
2

Proceedings of Copper 2010 2901
Anode:
0 2
2 S Cu e CuS + =
+

(3)
0 2 2
2
4 S Fe Cu e CuFeS + + =
+ +

(4)
Cathode:
+ +
= +
2 3
Fe e Fe (5)
Figure 3: Eh-pH diagram of the Cu-Fe-S-H
2
O system at 25 C
14 12 10 8 6 4 2 0
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 i mi t s
CuFeS2
CuS
Cu(OH)2
Cu(+a)
Cu(+2a)

Zhang, Tumen-Ulzii, Li
Proceedings of Copper 2010 2902
Figure 4: Eh-pH diagram of the Fe-Cl-H
2
O 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 FeCl
3
and FeCl
2
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.
14 12 10 8 6 4 2 0 -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 imit s
Fe
Fe2O3
Fe3O4
Fe(OH)2
Fe(+3a)
Fe(+2a)

Copper Leaching from Molydenite in Acidic FeCl
3
Solutions with FeCl
2

Proceedings of Copper 2010 2903
Figure 5: Effect of temperature on the copper extraction in the ferric chloride leaching of
molybdenite concentrate (pH = 0, S: L = 1:10, FeCl
3
concentration: 100 g/L,
FeCl
2
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 FeCl
3
concentration 200 g/L, FeCl
2
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


C
o
p
p
e
r

E
x
t
r
a
c
t
i
o
n

(
%
)
Leaching Time (min)
70
0
C
80
0
C
90
0
C
100
0
C

Zhang, Tumen-Ulzii, Li
Proceedings of Copper 2010 2904
Figure 6: Effect of S: L on the copper extraction in the ferric chloride leaching of molybdenite
concentrate (pH=0, 100 C, FeCl
3
concentration: 200g/L, FeCl
2
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 FeCl
3
FeCl
2
leaching
system, the initial redox potential of the leaching solution can be controlled by FeCl
3
/ FeCl
2
ratio,
according to the Nernst equation (Equation 6),
) (
) (
log 059 . 0 67 . 0
2
3
+
+
+ =
Fe a
Fe a
E

(6)
As shown in Figure 7, at a constant concentration of FeCl
2
(300 g/L), copper extraction increased
with increasing FeCl
3
concentration. And it can be seen in Figure 8 that copper extraction decreased
with increasing FeCl
2
concentration at a constant FeCl
3
concentration (100 g/L).
0 50 100 150 200 250
0
10
20
30
40
50
60
70
80


C
o
p
p
e
r

E
x
t
r
a
c
t
i
o
n

(
%
)
Leaching Time (min)
S:L=1:10
S:L=1:5
S:L=1:2.5

Copper Leaching from Molydenite in Acidic FeCl
3
Solutions with FeCl
2

Proceedings of Copper 2010 2905
Figure 7: Influence of FeCl
3
concentration on the copper extraction in the ferric chloride leaching
of molybdenite concentrate (pH=0, S:L=1:10, 100 C, FeCl
2
concentration: 300 g/L)
Figure 8: Influence of FeCl
2
concentration on the copper extraction in the ferric chloride leaching
of molybdenite concentrate (pH=0, S:L=1:10, 100 C, FeCl
3
concentration: 100 g/L)
0 50 100 150 200 250
0
10
20
30
40
50
60
70
80
200 g/L FeCl
3
- 300 g/L FeCl
2

100 g/L FeCl
3
- 300 g/L FeCl
2



C
o
p
p
e
r

E
x
t
r
a
c
t
i
o
n

(
%
)
Leaching Time (min)

0 50 100 150 200 250
0
10
20
30
40
50
60
70
80


C
o
p
p
e
r

E
x
t
r
a
c
t
i
o
n

(
%
)
Leaching Time (min)
100g/L FeCl
3
- 300g/L FeCl
2
100g/L FeCl
3
- 200g/L FeCl
2

Zhang, Tumen-Ulzii, Li
Proceedings of Copper 2010 2906
3.5 Synergetic effect of FeCl
2

The synergetic effect of FeCl
2
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 FeCl
2
. The value of a
H
+
enhanced more
than two times duo to the addition of 3 mol/ kg FeCl
2
. Then the efficiency of the leaching can be
enhanced when FeCl
2
is added into the leaching solution. Moreover, the addition of FeCl
2
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 FeCl
2
, but it increases to more than 107 C when 3 mol/ kg of FeCl
2
has been
added.
Figure 9: Influence of concentration of FeCl
2
on a
H
+
in a solution of HCl (0.1 mol/ kg) +
FeCl
3
(1 mol/ kg) + FeCl
2
(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 3.0
0.08
0.10
0.12
0.14
0.16
0.18


a
H
+

(
m
o
l
/

k
g
)
Concentration of FeCl
2
(mol/ kg)

Copper Leaching from Molydenite in Acidic FeCl
3
Solutions with FeCl
2

Proceedings of Copper 2010 2907
Figure 10: Influence of concentration of FeCl
2
on the boiling point of the solution of HCl
(0.1 mol/kg) + FeCl
3
(1 mol/kg) + FeCl
2
(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],
t k X
s
=
3 / 1
) 1 ( 1

(7)
RT Ea
Ae k
/
= (8)
RT E A k
a
/ ln ln = (9)
where X is the fraction of leached copper amount at time t (min), and k
s
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 3.0
101
102
103
104
105
106
107
108


B
o
i
l
i
n
g

p
o
i
n
t

(

C
)
Concentration of FeCl
2
(mol/ kg)

Zhang, Tumen-Ulzii, Li
Proceedings of Copper 2010 2908
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

-
L
n
k

(
s
-
1
)
T
-1
/10
-3
(K
-1
)

Copper Leaching from Molydenite in Acidic FeCl
3
Solutions with FeCl
2

Proceedings of Copper 2010 2909
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 FeCl
3
/FeCl
2
ratio. The addition of FeCl
2
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 - H
2
SO
4
-O
2
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.
Zhang, Tumen-Ulzii, Li
Proceedings of Copper 2010 2910
[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. 18920.
[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
conferenceEMC 2007, June 11-14, 2007, Dusseldorf, Germany.
[10] Mohammad A. H., Sam K., Adnan A. H.: Ferric chloride leaching of chalcopyrite: synergetic
effect of CuCl
2
, Hydrometallurgy, 2008, 91, pp. 8997.
[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.

Proceedings of Copper 2010 2911












Recycling

Proceedings of Copper 2010 2912



Proceedings of Copper 2010 2913
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 subproducts 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
Proceedings of Copper 2010 2914
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
Proceedings of Copper 2010 2915
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 2
nd
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
Proceedings of Copper 2010 2916
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.
Shifting Core Business Vision: From Copper to Polymetallics
Proceedings of Copper 2010 2917
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, its possible to think that new breakthroughs could follow the breakdown of the core-business
paradigm regarding new visions of what mining industry its 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
Proceedings of Copper 2010 2918
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.
Shifting Core Business Vision: From Copper to Polymetallics
Proceedings of Copper 2010 2919
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
Proceedings of Copper 2010 2920
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.
Shifting Core Business Vision: From Copper to Polymetallics
Proceedings of Copper 2010 2921
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,
N60, 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, N2.
[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): Intuicin y Razn, p 152; Buenos Aires.

Proceedings of Copper 2010 2922


Proceedings of Copper 2010 2923
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 m 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 Agencys 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
Proceedings of Copper 2010 2924
1 Introduction
Arsenic contamination is one of the worlds major environmental issues. Many methods have been
reported for the removal of arsenic from water environments [114]. Most reported methods involve
adsorption of arsenic on iron compounds and this phenomenon has long been a subject of interest
[1517].
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 As
2
O
3
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 (FeAsO
4
2H
2
O) has been developed and
reported [1821]. 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 FeAsO
4
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 FeAsO
4
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
Proceedings of Copper 2010 2925
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 H
3
AsO
4
; content: 58-62 %; mean: 60 %). Analytical-
grade iron (II) sulfate 7-hydrate (FeSO
4
7H
2
O) 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 FeSO
4
7H
2
O
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.
Fujita, Shibata, Nakamura
Proceedings of Copper 2010 2926
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 (ICPAES) (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
Agencys 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 m 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)SO
4
.7H
2
O
Fe/As molar ratio: 1.5
Temp.: 95
o
C
Stirring rate: 800 rpm
O
2
gas flow: 4 L/min
Reaction time: 7 h
Reaction
Sampling
Pressure filtration
Precipitate
Washing
Scorodite powder Solution

Solubility of Scorodite Synthesized by Oxidation of Ferrous Ions
Proceedings of Copper 2010 2927
weeks. After a pre-determined time, the sample bottle was removed from the shaker, the mixture
was filtered through a 0.2 m PTFE membrane filter, and the filtrate was analyzed by inductively
coupled plasma mass spectrometry (ICPMS) (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 m 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 CH
3
COOH solution) 3.00
TCLP 4.93 (0.1mol/L CH
3
COOH 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 H
2
SO
4
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 HNO
3
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 H
2
SO
4
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 H
2
SO
4
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 H
2
SO
4
5.47, 2.96
Fujita, Shibata, Nakamura
Proceedings of Copper 2010 2928
3 Results and discussion
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

30m 6m

Solubility of Scorodite Synthesized by Oxidation of Ferrous Ions
Proceedings of Copper 2010 2929
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 CH
3
COOH 0.1 mol/L CH
3
COOH 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 KH
2
PO
4
and NaH
2
PO
4
0.01 mol/L Na
2
B
4
O
7

pH
As Fe S
pH
As Fe S
mg/L mg/L mg/L mg/L mg/L mg/L
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 H
2
SO
4
was used to adjust the pH.
Fujita, Shibata, Nakamura
Proceedings of Copper 2010 2930
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, H
2
SO
4
or HNO
3
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 HNO
3
, H
2
SO
4
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
mg/L mg/L mg/L mg/L mg/L mg/L
Water 5.74 4.61 0.01 0.56 <10 H
2
SO
4
3.96 3.99 0.05 0.91 <10
TCLP 2.88 3.00 3.02 0.11 1.48 <10 H
2
SO
4
2.97 3.03 0.21 1.37 20
TCLP 4.93 4.88 4.89 0.16 0.21 <10 H
2
SO
4
1.97 1.99 1.83 2.87 210
CaO 12.52 12.17 2.55 0.09 <10 H
2
SO
4
1.22 1.28 11.00 9.84 1630
CaO 12.30 12.10 2.20 0.11 <10 H
2
SO
4
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 HNO
3
3.98 3.99 0.05 0.98 <10
Mg(OH)
2
9.00 7.58 1.27 0.60 <10 HNO
3
2.99 3.01 0.08 1.72 <10
NaOH 12.50 9.46 99.20 60.50 30 HNO
3
1.99 2.04 0.78 2.90 <10
NaOH 11.06 8.28 21.20 5.43 <10 HNO
3
1.12 1.12 3.89 5.93 <10
NaOH 9.70 5.83 0.09 0.05 <10 HNO
3
0.12 0.16 29.34 26.11 <10

Solubility of Scorodite Synthesized by Oxidation of Ferrous Ions
Proceedings of Copper 2010 2931
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.
*: H
2
SO
4
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;
FeAsO
4
2H
2
O + H
2
O = Fe(OH)
3
+ HAsO
4
2-
+ 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
Fujita, Shibata, Nakamura
Proceedings of Copper 2010 2932
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 (FeAsO
4
2H
2
O) 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, H
2
SO
4
or HNO
3
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 HNO
3
, H
2
SO
4
, 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.
Solubility of Scorodite Synthesized by Oxidation of Ferrous Ions
Proceedings of Copper 2010 2933
References
[1] WILKIE J.A., HERING J.G. (1996): Colloids Surf., A 107: 97110
[2] MANNING B.A., GOLDBERG S. (1996): Soil Sci. Soc. Am. J., 60: 121131
[3] JAIN A., RAVEN K.P., LOEPPERT R.H. (1999): Environ. Sci. Technol., 33: 11791184
[4] MENG X., BANG S., KORFIATIS G.P. (2000): Water Res., 34: 12551261
[5] GAO Y., MUCCI A. (2001): Geochimica et Cosmochimica Acta, 65: 23612378
[6] DIXIT S., HERING J.G. (2003): Environ. Sci. Technol., 37: 41824189
[7] KATSOYIANNIS I., ZOUBOULIS A., ALTHOFF H., BARTEL H. (2002): Chemosphere, 47: 325332
[8] AHN J.S., CHON C.M., MOON H.S., KIM K.W. (2003): Water Res., 37:24782488
[9] FUKUSHI K., SATO T., YANASE N. (2003): Environ. Sci. Technol., 37: 35813586
[10] MUNOZ J.A., GONZALO A., VALIENTE M. (2002): Environ. Sci. Technol., 36: 34053411
[11] GIASUDDIN A.B.M., KANEL S.R., CHOI H. (2007): Environ. Sci. Technol., 41: 20222027
[12] JANG J.H., DEMPSEY B.A. (2008): Environ. Sci. Technol., 42: 28932898
[13] MERCER K.L., TOBIASON J.E. (2008): Environ. Sci. Technol., 42: 37973802
[14] SARKAR A., BLANEY L.M., GUPTA A., GHOSH D., SENGUPTA A.K. (2008): Environ. Sci.
Technol., 424: 2684273
[15] HINGSTON F.J., ATKINSON R.J., POSNER A.M., QUIRK J.P. (1968): 9th Int. Congr. Soil
Science Transactions (Adelaide, Australia), Vol. I: 669678
[16] HINGSTON F.J., POSNER A.M., QUIRK J.P. (1971): Discussions of the Faraday Society,
No.52: 334342
[17] PIERCE M.L., MOORE C.B. (1982): Water Res., 16: 12471253
[18] FUJITA T., TAGUCHI R., ABUMIYA M., MATSUMOTO M., SHIBATA E., NAKAMURA T.
(2008): Hydrometallurgy, 90: 92102
[19] FUJITA T., TAGUCHI R., ABUMIYA M., MATSUMOTO M., SHIBATA E., NAKAMURA T.
(2008): Hydrometallurgy, 90: 8591
[20] FUJITA T., TAGUCHI R., ABUMIYA M., MATSUMOTO M., SHIBATA E., NAKAMURA T.
(2008): Hydrometallurgy, 93: 3038
[21] FUJITA T., TAGUCHI R., ABUMIYA M., MATSUMOTO M., SHIBATA E., NAKAMURA T.
(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: 1926
Fujita, Shibata, Nakamura
Proceedings of Copper 2010 2934
[24] NISHIMURA T., TOZAWA K. (1984): J. Min. Metall. Inst. Japan, 100: 11381144
[25] ROBINS R.G. (1980): 4
th
Joint Meeting of MMIJ-AIME, Technical Session D-1, (Tokyo, Ja-
pan), 2544
[26] ROBINS R.G. (1981): Metall. Trans. B 12B: 103109
[27] ROBINS R.G. (1983): 3
rd
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), 291310
[28] ROBINS R.G. (1987): Am. Mineralogist, 72: 842844
[29] ROBINS R.G. (1990): EPD Congress 90, Ed. D. R. Gaskell, (The Minerals, Metals & Mate-
rials Society, Anaheim, California), 93104
[30] KRAUSE E., ETTEL V.A. (1985): Proc. 15
th
Annual CIM Hydrometallurgy Meeting, (CIM,
Vancouver, Canada), 5/15/20
[31] KRAUSE E., ETTEL V.A. (1988): Am. Mineralogist, 73: 850854
[32] KRAUSE E., ETTEL V.A. (1989): Hydrometallurgy, 22: 311337
[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), 469488
[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), 545559
[35] EMETT M.T., KHOE G.H. (1994): EPD Congress 1994, Ed. G. W. Warren, (The Minerals,
Metals & Materials Society, San Francisco, California), 153166
[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), 281286
[37] LANGMUIR D., MAHONEY J., MACDONALD A., ROWSON J. (1999): Geochimica et Cosmo-
chimica Acta, 63: 33793394
[38] LANGMUIR D., MAHONEY J., ROWSON J. (2006): Geochimica et Cosmochimica Acta, 70:
29422956
[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), 439451
[40] BLUTEAU M.C., DEMOPOULOS G.P. (2007): Hydrometallurgy, 87: 163177


Proceedings of Copper 2010 2935
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 Chtelet, 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
Proceedings of Copper 2010 2936
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 PCBs (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 Materials and Methods
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
Proceedings of Copper 2010 2937
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 K
2
Cr
2
O
7
, 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 (d
50
) as well as d
10
and d
90
as given by the image
analysis system are derived as 1.27 mm, 0.43 and 2.03 mm respectively.
Gaydardzhiev, Bastin, Bareel, Goffinet
Proceedings of Copper 2010 2938
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 m. 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.

Biosolubilization of Copper from Waste Electric Cables
Proceedings of Copper 2010 2939
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

Gaydardzhiev, Bastin, Bareel, Goffinet
Proceedings of Copper 2010 2940
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
Biosolubilization of Copper from Waste Electric Cables
Proceedings of Copper 2010 2941
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 H
2
SO
4
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)
C
o
n
c
e
n
t
r
a
t
i
o
n

o
f

C
u

(
g
/
L
)
0
2
4
6
8
10
12
14
16
18
20
C
o
n
c
e
n
t
r
a
t
i
o
n


o
f

Z
n

a
n
d

P
b

(
m
g
/
L
)
Cu
Zn
Pb

Gaydardzhiev, Bastin, Bareel, Goffinet
Proceedings of Copper 2010 2942
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

Biosolubilization of Copper from Waste Electric Cables
Proceedings of Copper 2010 2943
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

Gaydardzhiev, Bastin, Bareel, Goffinet
Proceedings of Copper 2010 2944
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.

Biosolubilization of Copper from Waste Electric Cables
Proceedings of Copper 2010 2945
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, 642647

Proceedings of Copper 2010 2946


Proceedings of Copper 2010 2947
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 (FeAsO
4
2H
2
O) 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 plants 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 m 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
Proceedings of Copper 2010 2948
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 arsenics 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 (FeAsO
4
2H
2
O) by reaction with iron(II) ion and O
2
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


Proceedings of Copper 2010 2949
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.

Kubo, Abumiya, Matsumoto
Proceedings of Copper 2010 2950
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).
As
2
S
3
+ 3CuSO
4
+ 4H
2
O 3CuS + 2HAsO
2
+ 3H
2
SO
4
(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)
O
Leaching
Solution Residue
Oxidation Copper Smelter
Oxidized Solution
Crystallization
SCORODITE Effluent
Stockpile Solution Treatment
Oxidizing agent
FeSO7HO
O-gas

Dowa Mining Scorodite Process

Application to Copper Hydrometallurgy


Proceedings of Copper 2010 2951
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, H
2
O
2
(concentration: 33~35 %) is used as the oxidizing agent. The reaction in this process is
expressed by Equations (2) and (3).
HAsO
2
+ H
2
O
2
H
3
AsO
4
(2)
HAsO
2
+ H
2
O
2
H
2
AsO
4
-
+ 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).
2H
3
AsO
4
+ 2FeSO
4
+ O
2
+ 3H
2
O 2FeAsO
4
2H
2
O + 2H
2
SO
4
(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 plants 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.

Kubo, Abumiya, Matsumoto
Proceedings of Copper 2010 2952
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)
A
r
s
e
n
i
c

M
T

/

M
o
n
t
h

0%
20%
40%
60%
80%
100%
9 10 11 12 1 2 3 4 5
2008 2009
Period (year - month)
E
x
t
r
a
c
t
i
o
n

(
%
)

Dowa Mining Scorodite Process

Application to Copper Hydrometallurgy


Proceedings of Copper 2010 2953
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.
L
e
a
c
h
a
t
e

A
s

(
m
g
/
L
)
Regulatory threshold: 0.3mg/L
Average: 0.07mg/L

Kubo, Abumiya, Matsumoto
Proceedings of Copper 2010 2954
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
Method
Leachate 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.

Dowa Mining Scorodite Process

Application to Copper Hydrometallurgy


Proceedings of Copper 2010 2955
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 m. Particles of diameter less than 5 m 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
2(deg) / (Cu-K)
I
n
t
e
n
s
i
t
y

a
.
u
.

Scorodite
ICDD-PDF: 00-37-0468
Lab. test sample
Plant made sample

Kubo, Abumiya, Matsumoto
Proceedings of Copper 2010 2956
Figure 7: Comparison of SEM images of scorodite
Figure 8: Particle size distributions of scorodite







Lab. test sample Plant made sample
20m 20m
0%
3%
6%
9%
12%
15%
1 10 100 1000
Diameter (m)
F
r
e
q
u
e
n
c
y

(
%
)
Plant made sample
Lab. test sample

Dowa Mining Scorodite Process

Application to Copper Hydrometallurgy


Proceedings of Copper 2010 2957
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 (m)
Specific surface area (m
2
/g)
Moisture (Wet basis %)
Lab. test sample
23.29
0.13
8
Plant made sample
22.48
0.2
11
Kubo, Abumiya, Matsumoto
Proceedings of Copper 2010 2958
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. 5
th
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 (24), 92102.
[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 (24), 8591.
[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.
[10] US-EPA. (2004): Test method for the evaluation of solid waste, physical/chemical methods,
Method 1310B, Extraction procedure toxicity test method and structural integrity test, U.S.
Environmental Protection Agency, Washington, DC.
[11] US-EPA. (1986): Test method for the evaluation of solid waste, physical/chemical methods,
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)

Proceedings of Copper 2010 2959
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
Proceedings of Copper 2010 2960
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
Proceedings of Copper 2010 2961
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
Proceedings of Copper 2010 2962
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

Recycling of Electric Home Appliances in Minamata
Proceedings of Copper 2010 2963
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
Proceedings of Copper 2010 2964
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


Recycling of Electric Home Appliances in Minamata
Proceedings of Copper 2010 2965
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
Proceedings of Copper 2010 2966
Figure 8: Treatment and sorting process in Act-B

Recycling of Electric Home Appliances in Minamata
Proceedings of Copper 2010 2967
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
Proceedings of Copper 2010 2968
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

Recycling of Electric Home Appliances in Minamata
Proceedings of Copper 2010 2969
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

Proceedings of Copper 2010 2970


Proceedings of Copper 2010 2971
New Standards in Environmental Protection for
Copper Recycling
Dr. Franz-Josef Westhoff, Dr. Claus Meyer-Wulf
Aurubis AG
Kupferstrae 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
fr 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
Proceedings of Copper 2010 2972
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
Proceedings of Copper 2010 2973
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.

Westhoff, Meyer-Wulf
Proceedings of Copper 2010 2974
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 [g per m and day]
TA Luft
1986
TA Luft
2002
Average

Arsenic and its inorganic compounds
calculated as Arsenic
g/(md) no value 4 Year
Lead and its inorganic compounds
calculated as Lead
g/(md) 250 100 Year
Cadmium and its inorganic compounds
calculated as Cadmium
g/(md) 5 2 Year
Nickel and its inorganic compounds
calculated as Nickel
g/(md) no value 15 Year
Mercury and its inorganic compounds
calculated as Mercury
g/(md) no value 1 Year
Thallium and its inorganic compounds
calculated as Thallium
g/(md) 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
New Standards in Environmental Protection for Copper Recycling
Proceedings of Copper 2010 2975
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 fr 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 wassergefhrdenden Stoffen (Concrete construction when handling
substances being harardous to waters) of the DAfStb (Deutscher Ausschuss fr Stahlbeton =
German committee for reinforced concrete) and consequently also the requirements of the VAwS
(Verordnung ber Anlagen zum Umgang mit wassergefhrdenden 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.
Westhoff, Meyer-Wulf
Proceedings of Copper 2010 2976
Figure 2: Construction of new outside storage are