The Extractive Metallurgy of Lead (R. J. Sinclair)

The Extractive Metallurgy of Lead By

Roderick J Sinclair Consultant in extractive metallurgy and formerly with The Electrolytic Zinc Company of Australasia and Pasminco Limited.

The Australasian Institute of Mining and Metallurgy Spectrum Series Volume Number 15 2009

Los Metalurgistas

Published by: THE AUSTRALASIAN INSTITUTE OF MINING AND METALLURGY Level 3, 15 - 31 Pelham Street, Carlton Victoria 3053 Australia

i

© The Australasian Institute of Mining and Metallurgy 2009 First Edition, June 2009

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means without permission in writing from the publishers.

The AusIMM is not responsible as a body for the facts and opinions advanced in any of its publications.

ISBN 978 1 921522 02 4

Desktop published by: Angie Spry and Kristy Pocock for The Australasian Institute of Mining and Metallurgy

Compiled on CD ROM by: Visual Image Processing Pty Ltd PO Box 3180 Doncaster East Victoria 3109

ii

Preface The history of lead is as old as the recorded history of mankind. Its use as a valuable material in society has been equally long and varied. In more recent times awareness of lead’s toxicity has restricted its widespread use and many older applications have been replaced by newer materials or have been phased out. Today the use of lead is dominated by the automotive lead-acid battery, and a key feature of this application is the ability to achieve a high level of recovery and recycle of scrap batteries. This attribute now makes lead the most recycled metal in use and approaching 60 per cent of the world’s supply of lead is provided by recycled metal. Secondary processing and smelting is consequently as important a part of the extractive metallurgical industry as primary extraction from ores and concentrates. The growing market share of batteries and the corresponding growing availability of secondary lead has meant that the demand growth has largely been met by secondary lead and the production of primary lead has been static or in decline for many decades. Coupled with increased regulation and controls of the environmental and occupational health aspects of the industry, there has been little incentive to change other than to meet higher regulatory standards. There has been a steady decline in the number of operating primary smelters and technology change has been slow. Nevertheless the industry plays a vital role in the supply of materials to society and there is a need for awareness of processing options so that the most efficient and cost-effective methods of lead extraction and refining can be applied. There is a singular deficiency in the technical literature of a comprehensive text covering the extractive metallurgy of lead. The purpose of this text is to hopefully fill that gap and to summarise the main processes in use for lead extraction and refining, the reasons why they are used, and the key features of their design and operation. It is primarily written for those in the industry, as an introduction to the issues involved, and to provide a means of developing a broader perspective of the extractive lead industry, and the ramifications of actions within any one sector of the lead production chain. It is by no means an exhaustive exposé of all aspects of individual processing steps in the extractive metallurgy of lead, but it is hoped that it has covered most key aspects and can serve as a reference and guide to stimulate further enquiry as required. This work follows the completion of a similar text on the extractive metallurgy of zinc, written with the same purpose in mind. The two metals are so closely associated in terms of mineral occurrence and extraction, that it seemed necessary to develop a complementary text on lead and have companion reference volumes covering each metal. Some of the details in this text repeat to some extent sections in the earlier zinc text, such as the coverage of slag fuming, but this has been done to allow each to stand alone, with sufficient information for those only interested in lead. As with the zinc text, the material is drawn from both the technical literature and from a long term association with the industry and many experienced and competent technical experts over many years. Much appreciation is expressed to my colleagues for comments, in particular Jim Happ and Denby Ward, and for the support and encouragement from Dr Rod Grant. Roderick J Sinclair

iii

Part A – General Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Chapter 1 – Industry Perspective and Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 Introduction, Properties and Uses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 World Supply and Demand. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 The Lead Smelting Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7 Primary Smelting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9 Secondary Lead Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13 References and Further Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15

Chapter 2 – Historical Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Lead Production in Early Times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17 The Lead Blast Furnace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21 Preparation of Blast Furnace Feed. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 Blast Furnace Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25 Lead Refining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26 Silver Recovery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26 Direct Smelting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27 Secondary Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28 Historical Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29 References and Further Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29

Chapter 3 – Raw Materials. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Lead Mineralogy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31 Separation and Concentration Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Commercial Lead Concentrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36 Commercial Terms for the Purchase of Standard Lead Concentrates . . . . . . . . 38 Commercial Terms for the Purchase of Bulk Concentrates . . . . . . . . . . . . . . . . . 40 Commercial Terms for the Sale of Lead Bullion . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Secondary Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41 References and Further Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42

Part B – Primary Smelting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

Chapter 4 – Sintering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Process Chemistry and Thermodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 The Sintering Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .46

xiii

The Structure of Sinter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49 Process Operating Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50 Updraught Sintering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55 Sinter Machine Capacity and Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Gas Handling and Cleaning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .60 Sulfuric Acid Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62 References and Further Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64

Chapter 5 – The Blast Furnace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65 Chemical Principles and Thermodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Furnace Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .67 Slag Characteristics and Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Furnace Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 Furnace Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80 Environmental Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86

Chapter 6 – The Imperial Smelting Furnace ( ISF) . . . . . . . . . . . . . . . . . . . . . . . . . 89 General Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 Process Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 Slag Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 Evolution of Furnace Design and Operation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Coke Use and Furnace Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96 References and Further Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97

Chapter 7 – Direct Smelting Processes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99 The Boliden Lead Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .105 The Kaldo Process (Top Blown Rotary Converter – TBRC) . . . . . . . . . . . . . . . 106 The Kivcet Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109 The Queneau-Schuhmann-Lurgi (QSL) Process . . . . . . . . . . . . . . . . . . . . . . . . 116 Top Submerged Lance (TSL) – Slag Bath Processes . . . . . . . . . . . . . . . . . . . . 119 The Isasmelt Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .120 The Ausmelt Lead Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .123 The Outokumpu Lead Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .126

Chapter 8 – Smelter By-Products and Treatment Processes. . . . . . . . . . . . . . . . 129 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .129

xiv

Slag Cleaning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .129 Zinc Recovery from Slags. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .130 The Conventional Slag Fuming Furnace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Top Submerged Lance Slag Fuming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 High Intensity Fuming Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .144 Fume Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .144 Electric Arc Fuming Furnace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146 Treatment of Lead Smelter Mattes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .147 Sinter Plant and Smelter Dusts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .147 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .148

Chapter 9 – Electrochemical Reduction Processes . . . . . . . . . . . . . . . . . . . . . . . 151 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .151 Processes Based on Molten Salt Electrolysis . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Processes Based on Aqueous Electrolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163

Part C – Secondary Smelting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

Chapter 10 – Secondary Materials and Pretreatment . . . . . . . . . . . . . . . . . . . . . 167 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167 Lead-Acid Battery Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .168 Battery Breaking and Separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .170 Paste Desulfurisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .172 Processing of Secondary Residues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .174

Chapter 11 – Secondary Smelting Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .175 Reverberatory Furnace. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .175 The Blast Furnace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178 The Electric Arc Furnace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .180 Rotary Furnace Smelting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .181 Top Blown Rotary Converter (TBRC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188 Top Lance Slag Bath Reactors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .188 Electrowinning Processes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .188 Refining of Secondary Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .193 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .194

xv

Part D – Refining of Lead Bullion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

Chapter 12 – Thermal Refining of Primary Lead Bullion. . . . . . . . . . . . . . . . . . . . 197 Methods and Equipment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .199 Copper Removal or Copper Drossing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 Softening for Arsenic, Antimony and Tin Removal . . . . . . . . . . . . . . . . . . . . . . . 205 Removal of Silver and Other Precious Metals . . . . . . . . . . . . . . . . . . . . . . . . . . 210 Separation of Thallium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .217 Separation of Zinc from Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .217 Separation of Bismuth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .219 Final Caustic Refining. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .220 Refining of Secondary Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .221 Summary of Common Impurities, Their Control and Recovery . . . . . . . . . . . . . 221 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .224

Chapter 13 – Electrolytic Refining of Lead. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227 Process Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .227 Practical Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .230 Current Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .236 Periodic Current Reversal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .237 Bipolar Electrode Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .238 Final Refining of Cathode Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .238 Anode Slimes Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .238 Other Electrolytic Refining Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .241

Chapter 14 – Alloying and Casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 Handling Molten Lead and Alloying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .243 Casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .243

Part E – Environmental and Economic Issues . . . . . . . . . . . . . . . . . . . . . . . . . . 247

Chapter 15 – Health and Environment Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .249 Lead in the Environment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .249 The Toxicology of Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .249 Exposure Pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .251 Occupational Standards and Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 External Environmental Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .254

xvi

Product Controls and Life Cycle Management . . . . . . . . . . . . . . . . . . . . . . . . . . 256 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .257

Chapter 16 – Energy Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 Purpose and Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .259 Energy Consumption for the Sinter Plant–Blast Furnace. . . . . . . . . . . . . . . . . . 259 Thermal Refining of Lead Bullion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .261 Electrolytic Lead Refining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .261 Direct Smelting Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .262 Electrochemical Lead Extraction Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . 264 Comparison of Extraction Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266 Energy Consumption in Supply of Lead Concentrates. . . . . . . . . . . . . . . . . . . . 266 Energy Consumption for Secondary Lead Production . . . . . . . . . . . . . . . . . . . . 267

Chapter 17 – Costs and Economics of Lead Production . . . . . . . . . . . . . . . . . . . 269 Purpose and Basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .269 Smelting by the Sinter Plant–Blast Furnace . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 Smelting by the Kivcet Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .274 Smelting by the Isasmelt Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .277 Comparison of Smelting Technologies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 Lead Refining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .280 Metal Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .284 By-Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .285 Overall Economics for Refined Lead Production . . . . . . . . . . . . . . . . . . . . . . . . 286 Economics of Secondary Lead Production. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289

Appendix 1 – Properties of Lead and Associated Compounds. . . . . . . . . . . . . . . 293 Lead Metal Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .293 Binary Lead Rich Eutectics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .294 Properties of Lead Oxides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .294 Vapour Pressures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .295 Silver Metal Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .295 Thermodynamic Properties of Compounds Involved in Lead Extraction . . . . . . 296 Heat Capacities at Constant Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299

xvii

PART A GENERAL CONTEXT This part of the text covers the general structure of the lead smelting industry, including its scope, its history and details of raw material supplies used for the recovery of lead metal. Chapter 1 Chapter 2 Chapter 3

Industry Perspective and Introduction Historical Background Raw Materials

CHAPTER 1 Industry Perspective and Introduction INTRODUCTION, PROPERTIES AND USES Lead is a metal of wide historical significance. It is now a very mature commodity and as such exhibits declining intensity of use, with broad replacement in many of its traditional uses. Much of the replacement results from acute awareness of the effects of lead on human health and the environment. Lead was widely used in ancient times, dating back over 7000 years. It was often mined and produced as a co-product of silver, which was highly prized for ornamentation and jewellery and later for coinage. Lead served as a collector for silver and gold and often smelting was conducted primarily for this purpose. Lead was separated from the precious metals by oxidation in the ‘cupellation process’. The Phoenicians and later the Romans mined silver and lead in Spain. Lead was mined at Laurium in ancient Greece and on the islands of Rhodes and Cyprus. The Romans also produced lead in Britain and in ancient Gaul. In the Middle Ages, silver and lead mining and production flourished at Rammelsburg, in the Hartz region of Germany, and in the Erzgebirge, and in Upper Silesia. Large deposits were later found and developed in the New World – in the USA, Mexico and Canada, as well as in Australia and these deposits represent major supplies of lead today. In Roman times, lead was used for making water piping, for lining water tanks and baths, as a roofing material and as a seal for weatherproofing buildings. It was used in soldered lead sheets by the Assyrians in the Hanging Gardens of Babylon. The Latin word ‘plumbum’ for lead has been synonymous with the working of lead metal for handling water, hence the trade of ‘plumbing’. Lead’s low melting point and softness enabled it to be used to seal bronze and iron connectors into building stone, and this can still be seen in many ancient buildings and ruins today. It was used for the construction of large windows from smaller fragments of glass at a time before large-sheet glass production was possible. Stained glass windows still remain as a prominent example of this art. Because of its high density and ease of moulding, lead was used as a projectile in warfare, initially for slingshots and catapults, but following the invention of gunpowder and firearms, was primarily used for the manufacture of ammunition. The production of lead shot using a high tower to form small spherical shapes was a significant industry up until the 19th century. Also due to the ease of moulding, as well as the hardness of its alloy with antimony, lead was used by Gutenberg in the first printing process for the fabrication of moveable type, and is still the basis of large-scale printing type setting where this is still used. However, the new technologies of offset and electronic printing are rapidly replacing this use. Lead’s oxides as red and white lead were used as paint pigments dating from ancient Egyptian times until the mid 20th century. They provide good pigment coverage and relatively stable colour, but have been phased out of use in recent times for health and environmental reasons. The unique electrochemical properties of lead in combination with its oxides and sulfates provided a means of constructing high capacity and high powered electrical storage batteries. This low cost application has developed into the major use for lead today, principally for starting, lighting and ignition (SLI) power supply in the automotive industry. Traction batteries for fork-lift trucks, buses and other heavy vehicles are commonly lead-acid. Large installations of lead-acid batteries are used for standby and uninterruptible power supplies, and for electrical energy storage from renewable

The Extractive Metallurgy of Lead

Spectrum Series Volume 15

3

CHAPTER 1 – Industry Perspective and Introduction

energy sources. Because lead in this application remains concentrated and is not dispersed, it can be readily recovered, making lead the most recycled metal, at around 60 per cent of total world supply. This has given rise to a major part of the lead smelting industry being structured around secondary sources of feed. The use of lead-acid batteries for portable power sources such as power tools has not been high due to lead’s high density, and alternatives such as nickel-cadmium, nickel-metal-hydride and lithium-based batteries have predominated, with much higher energy-to-weight ratios. In this regard, the future trends in electric vehicles are not yet clear, but lead-acid batteries have the advantage of relatively low cost, with high power delivery. As a pure metal, lead is soft and malleable with low mechanical strength. This is an advantage in some applications such as weatherproofing, but one consequence is that under stress the metal will easily deform to relieve that stress, or ‘creep’, and this can take place over long periods of time. Indeed, lead can creep under its own weight, and to avoid this effect the safe tensile stress is 1.7 MN/m2 and in compression, 2.75 MN/m2. Lead can be alloyed to improve its strength properties, and antimony was commonly used as a hardener. Pure lead is in fact rarely used. The corrosion resistance of lead is due to the formation of dense coherent surface films such as oxide, carbonate or sulfate. This, coupled with its ability to be alloyed and rolled into sheet, has enabled lead to be used as a construction material in the chemical industry, particularly in sulfuric, phosphoric or chromic acid environments. For these applications it was often used as a protective coating on steel, applied by melting and wiping, or ‘burning’, the lead onto the steel surface. The high density of lead, and the fact that its oxides will dissolve in glass without causing colouration, have enabled its use to increase refractive index and form decorative ‘crystal’ glass products. High quality crystal can contain up to 70 per cent lead and was first introduced in the 17th century. Many lead compounds have unique properties with corresponding useful applications. The organo-metallic compound tetraethyl lead has been important as an additive to automotive fuel to control pre-ignition in the internal combustion engine. It is effective in very small amounts, and the petrol-driven internal combustion engine and, indeed, the automobile itself, owed much of their early development to this use. Tetraethyl lead represented a large use of lead in the mid 20th century, but health and environmental concerns have seen this largely eliminated in the early 21st century. Lead compounds are also used for a range of plastic stabilisers to overcome the degradation of the plastic by heat and UV radiation. This is particularly applied to polyvinyl chloride (PVC), where it is used for construction applications such as house siding, window frames and rainwater products. Degradation causes decomposition and loss of HCl from the polymer structure, in turn causing discolouration and brittleness. A number of base metal salts, particularly lead, zinc, tin and cadmium, are effective in HCl bonding and preventing free HCl formation. The lead salts are usually tri-basic lead sulfate, phosphate or stearate. There are some legislative requirements that products of this nature must be recycled because of their lead content. Other significant uses of lead are for the sheathing of electrical and communication cables, and for protection against high energy radiation. Its high coefficient of absorption of X-rays and gamma rays at 0.48 cm-1, combined with the ability to dissolve lead oxides in glass, provided for the construction of cathode ray tubes for television and computer monitor applications. This has been an important use for lead, although there is a significant trend to replacement by more compact alternative display technologies, such as LCD and plasma screens. Apart from addition to glass, radiation shielding in many forms in the nuclear industry relies on the use of lead.

4




A generalised list of the major uses of lead may be given as follows:

• batteries: • automotive SLI, and • energy storage, • sheet for the building industry, • sheathing of power and telecommunication cables, • plastic stabiliser chemicals, • radiation shielding: • cathode ray tubes, and • general applications, • ammunition, • corrosion protection: • chemical lead applications, • glass additive for production of crystal, • glazes for ceramics, • colouring pigments for plastics, • paints: • both pigment and preservative uses, • weights, • sound insulation barriers, and • automotive fuel additives. Many old uses of lead such as for ammunition remain, but many have been phased out with the availability of new materials, and because of the recognition of the health hazards associated with some of those old uses. Health concerns have also seen lead removed from paints and petrol. Lead poisoning and its effect on mental health has been known about for many years, and regulations covering permissible maximum lead levels in the blood of those working with lead have been introduced and progressively tightened. Children are more susceptible to lead poisoning, and the effect on childhood mental development has been a significant issue in the formulation of environmental controls. The ancient practices of using lead in cosmetics and sweetening wine by storing it in lead containers or adding lead acetate have long been eliminated. These practices have even been given as a cause for the downfall of the Roman Empire, for the madness of King George III of England and his subsequent loss of the American colonies. Lead shot in cartridges for hunting waterfowl has largely been replaced with iron shot because of concerns about the poisoning of birds from shot ingestion collected with feed from the bottom of waterways. With these health and environmental pressures, the pattern of lead use has shifted markedly, as illustrated in Table 1.1 which shows lead end uses for 1960 and for 2005. Due to environmental and health concerns, the clear general trend is to replace lead in the dispersive uses and to concentrate its application to uses where it can be recycled. This trend will necessarily see an increase in the proportion of lead supply derived from secondary processing, and minimal growth or a decline in the future demand for primary lead.



5


TABLE 1.1 End uses for lead in 1960 and 2005 (source: Lead Development Association). End use

1960

2005

Batteries

28%

75%

Pigments and chemicals

10%

8%

Rolled extrusions

16%

6%

Alloys and ammunition

15%

5%

Cable sheathing

18%

2%

Miscellaneous (including tetraethyl lead)

13%

4%

WORLD SUPPLY AND DEMAND World consumption of lead totalled close to 7 800 000 t in 2005, of which about 3 400 000 t was derived from mine and primary smelter production. The balance came from secondary production from recycled scrap products – predominantly batteries. Total world lead consumption and mine production since 1970 are illustrated in Figure 1.1 and 2005 production figures are given in Table 1.2. 8000 7000

Lead ’000 (t)

6000 5000 4000 3000 2000 1000 0 1970

Consumption

1975

1980

1985

1990

Mine Production

1995

2000

2005

Year

FIG 1.1 - Global lead consumption and mine production.

TABLE 1.2 Lead metal and lead mine production for 2005 (source: International Lead Zinc Study Group). Region

Metal (t)

Europe

1 702 000

256 000

Africa

130 000

130 000

America

2 043 000

1 013 000

Asia

3 486 000

1 322 000

Oceania World total

6

Mine (t)

276 000

715 000

7 636 000

3 436 000




Figure 1.1 shows a slight decline in mine production, although it has been relatively steady over the past two decades. Mined lead will correspond closely with primary metal production at around 3 300 000 t/a. The gap between mine production and total consumption closely matched the production of lead from secondary sources, which was 4 200 000 t/a in 2005 and growing significantly. This increase in secondary lead production matches the increasing proportion of lead being used for lead-acid batteries. Primary lead is produced largely from the smelting of lead sulfide (galena). This is often mined in conjunction with zinc sulfides where both metals are sought. However, the growth in demand for zinc has outstripped the growth in demand for primary lead and there has been a relative decline in the mined lead to zinc ratio, to match smelter requirements. The ratio of lead to zinc mined was 0.7 in 1960, declining to 0.5 in 1983 and to 0.32 in 2005.

THE LEAD SMELTING INDUSTRY The lead smelting industry is divided broadly into primary and secondary smelters, producing a crude lead bullion, and refineries, removing impurities from the crude bullion to achieve the market grade of refined lead as set by the London Metal Exchange (LME) or the customer. Refineries may be directly associated with the smelting operations or may be separate independent operations, taking crude bullion from the smelters. There are, for instance, large independent refining operations in Japan and in the UK. Japan, for example, has a surplus of stand-alone refining capacity and has traditionally purchased bullion. This can be an efficient approach where a primary smelter is located at a mine site and the refinery is located close to final refined lead markets. Refining of primary bullion is more complex than for secondary bullion, and most independent refineries have the capability for handling primary bullion. Secondary refining can be relatively simple with few impurities to remove, and is usually part of the secondary smelter. A broad schematic of the structure of the industry is shown in Figure 1.2. Primary lead production is based on the smelting of lead sulfide concentrates. There is a large disparity between regional mining and smelting operations and a significant world trade in lead concentrates. Table 1.3 shows the major lead mining countries and Table 1.4 the major smelting capacities in 2004. Primary smelters are often associated with major mining operations, but are usually centrally located in major industrial centres. Because of the environmental issues associated with lead smelting sites in the past, it will be very difficult in the future to obtain licences for the construction of new TABLE 1.3 Major lead mining countries in 2004 (source: Lead Development Association). Country

Mined lead production (tonnes of contained lead)

Australia

654 000

China

618 000

USA

464 000

Peru

308 000

Canada

200 000

Mexico

152 000

Others

804 000

Total

3 200 000



7


TABLE 1.4 Major lead smelting production in 2004. Country

Total smelter production (t/a)

Production from mined lead (t/a)

Secondary lead production (t/a)

China

1 533 000

1 350 000†

183 000

USA

1 338 000

290 000

1 048 000

Australia

390 000

365 000

25 000

Japan

375 000

175 000

200 000

Canada

370 000

250 000

120 000

Germany

350 000

75 000

275 000

Kazakhstan

330 000

300 000

30 000

Italy

283 000

130 000

153 000

UK

176 000

0

176 000

Others

1 685 000

135 000

1 540 000

Total

6 820 000

3 070 000

3 750 000

†

†

Estimated.

Scrap Collection

Mining

Ore

Breaking and Separation

Mineral Processing Zinc and Copper Concentrates Tailings

Waste

Plastics

Lead Concentrate

Residues

Primary Lead Smelting

Secondary Smelting Slag Fuming

Slag

Sulfuric Acid

Waste Slag

Lead Bullion

Zinc Oxide Lead Bullion

Lead Refining

Copper Matte

Silver and Gold Antimonial Lead

Refined Lead

FIG 1.2 - Lead extraction industry schematic.

8




primary lead smelting sites. As a result there have been very few new greenfield primary smelters constructed in the past 20 to 30 years. This is reinforced by the fact that primary smelting of lead has been static over that period, so there has been no need for additional capacity. This approach is likely to continue, and existing primary smelters will tend to be upgraded or replaced with improved technology at existing sites. Secondary lead smelting tends to be localised around major population centres and the supply of waste batteries, due to the relatively high cost associated with the transport of used batteries. The secondary smelting technologies used are also suited to relatively small-scale operations in comparison with primary smelters, which can benefit significantly from the economies of scale. The balance of lead metal flows for the total lead industry is illustrated in Figure 1.3, with the horizontal width of the bars representing the annual tonnage of metal produced and used. This illustrates the relatively high level of production by secondary lead recycling in comparison with new lead from mine production. New lead essentially reports to a growing inventory of lead-acid batteries and other metal uses, and to losses from the system as dispersive uses. The inventory effect reflects both the growing demand for batteries and the life of the battery before it is scrapped and recycled.

Lead Residues from Zinc / Copper Smelting Mine Production

Scrap Batteries Secondary Smelting

Recycle Residues

Primary Smelting Lead Metal Produced

Batteries RecycleRecycle Scrap Batteries

Metal

Other Uses

Inventory Growth Losses

Losses

Recycle Metal Scrap

FIG 1.3 - Lead metal cycle.

PRIMARY SMELTING Primary lead smelting is largely based on the treatment of lead sulfide (galena) concentrates. A number of processes are used but the traditional sinter plant–blast furnace technology (as illustrated in Figure 1.4) has predominated. The sinter plant eliminates sulfur and produces an agglomerated material with lead feed and fluxes present as oxides, which is subsequently reduced to lead metal in the blast furnace using metallurgical coke. Crude lead bullion is refined either by the thermal process, which individually separates impurities, or by the electrorefining process, to give a refined lead with less than 0.01 per cent total impurities.



9


Lead concentrates

Recycles

Fluxes – silica, lime, ironstone

Gas Cleaning

Sinter Plant

Dusts

Sulfuric Acid Production

Sulfuric acid

Gas Cleaning

Blast Furnace

Dusts Slag to waste or fumer Lead bullion to refinery

FIG 1.4 - Sinter plant–blast furnace flow sheet.

More recent process developments have been applied to direct smelting in which sulfur elimination and oxide reduction take place in the one unit, enabling the heat of sulfide oxidation to be utilised, and thus improving the overall thermal efficiency of the process. Direct smelting processes avoid the use of metallurgical coke as a relatively high cost fuel and reductant. The incentive to change from the sinter plant–blast furnace technology has also been driven by environmental issues, since these operations are difficult to contain, and can contribute significant emissions of lead particulates to the atmosphere. The major primary lead smelting processes in use are:

• the sinter plant–blast furnace combination (see Chapters 4 and 5), • the Imperial Smelting Process (also a sinter plant with closed top blast furnace for co-production of zinc) (see Chapter 6), and

• direct smelting processes (see Chapter 7): • • • • •

the Kivcet process, the QSL process, the ISASMELT and Ausmelt processes, the Boliden process, and

the Kaldo TBRC process. As indicated in Table 1.4, production of lead from primary sources is of the order of 3 100 000 t/a. However, the capacity of primary smelters is significantly in excess of this figure, since most primary smelters also accept varying proportions of secondary materials as part of their feed. These additional feeds are commonly in the form of lead residues, containing oxide lead as well as sulfates. These residues may arise from scrap processing or from other metal extraction such as zinc and copper,

10




which give rise to fumes and leach residue containing high levels of lead. In the latter case, the lead produced from such materials is still primary lead, but is not accounted for in the statistics for mine lead production. These materials can represent around ten per cent of total lead output from the sinter plant, and lead bullion production from the blast furnace can be significantly greater due to direct feeds to the furnace, particularly if those materials contain metallic lead. It is therefore difficult to arrive at a figure which truly represents primary lead production capacity; however, it will be of the order of 4 000 000 t/a (2005). The number of primary smelters listed in 2004 totals 53, and Table 1.5 shows the distribution of the world’s primary lead smelters by the process used. TABLE 1.5 Distribution of primary smelter capacity by process type in 2004. Process type

Capacity (t/a)

Percentage of total

Number of smelters

Sinter plant–blast furnace

2 470 000

70%

34

Imperial Smelting Furnace

280 000

8%

8

Kivcet

360 000

10%

3

QSL

270 000

8%

3

Other processes Total

160 000

4%

5

3 540 000

100%

53

Cumulative Capacity Above ’000 (t/a)

The sinter plant–blast furnace technology represented over 90 per cent of total primary lead capacity in 1980, so there has been a significant replacement of that technology. There has been virtually no additional primary capacity in that period, hence there has been a net closure of blast furnaces and the remaining plants are relatively old. It is likely that there will need to be progressive closures of sintering–blast furnace operations and replacement by direct smelting technologies in the future. The capacity distribution of primary smelting capacity is shown in Figure 1.5. The vertical axis represents cumulative capacity of plants above a given plant size, as given by the horizontal axis. Figure 1.5 indicates that almost half of the primary production capacity (or 1 750 000 t), is attributed to plants above 100 000 t/a, and 80 per cent of the primary production capacity (or 2 800 000 t) is attributed to plants above 50 000 t/a capacity, of which there are only 25 presently operating, as detailed in Table 1.6. Of those 25, three use the Kivcet process, three use the QSL 4000 3500 3000 2500 2000 1500 1000 500 0 0

50

100

150

200

250

Smelter Capacity ’000 (t/a)

FIG 1.5 - Distribution of world primary lead smelter capacity.



11


process, one uses Kaldo and two Ausmelt technology. The remaining 16 use sinter plant–blast furnace. The average primary smelter capacity is 67 000 t/a of lead bullion, whereas the median capacity is 50 000 t/a. TABLE 1.6 Major primary lead smelters in 2004 (capacity over 50 000 t/a). Region, country and company name

Location

Process

Annual capacity (t)

Europe Belgium Umicore

Hoboken

S-BF

125 000

France Metaleurop

Noyelles Godault

S-BF

110 000

Germany Berzelius Stolberg Metaleurop Weser Blei

Binsfeldhammer Nordenham

QSL Ausmelt

100 000 90 000

Porto Vesme

Kivcet

100 000

Ust Kamenogorsk Chimkent

Kivcet S-BF

140 000 160 000

Kosovska Mitrovica

S-BF

125 000

Ronnskar

Kaldo

55 000

Herculaneum Glover

S-BF S-BF

205 000 95 000

Trail, BC Belledune

Kivcet S-BF

120 000 108 000

Mexico Met Mex Penoles

Torreon

S-BF

180 000

Peru Centromin

La Oroya

S-BF

93 000

Zhouzhou Baiyin (Gansu) Fankou (Guandong) Shenyang (Liaoning)

S-BF QSL S-BF S-BF

100 000 52 000 60 000 70 000

India Hindustan Zinc

Chanderiya (Rajasthan)

Ausmelt

50 000

Japan Toho Zinc Co

Chigirishima

S-BF

90 000

South Korea Korea Zinc Co

Italy Eniresorse Kazakhstan Kazpolymetal Serbia Trepca Sweden Boliden Mineral AB Americas USA Doe Run Doe Run Canada Teck-Cominco Brunswick M&S Co

Asia and Oceania China Zhouzhou Smelter Baiyin Northwest Smelter Fankou Mine Shenyang Smelter

Onsan

QSL

120 000

North Korea Korea Metals and Chemicals

Mumpyong

S-BF

90 000

Australia Nyrstar Xstrata Zinc

Port Pirie Mount Isa

S-BF S-BF

220 000 150 000

12




Refining is a significant and separate part of primary lead smelting, and two different approaches are used involving pyrometallurgical separation processes or electrorefining. Pyrometallurgical methods involve the oxidation of selected impurities from molten lead bullion for collection as a slag or dross, or the precipitation of impurities to form a dross or crust by the addition of reagents and/or by changes in temperature. A number of steps are usually applied for removal of copper, arsenic and antimony, silver and precious metals, zinc, then bismuth, and finally, residual minor impurities and drossing reagents by treatment with caustic soda. Operations are usually conducted in externally heated crucibles or ‘kettles’, holding between 100 and 300 t of molten lead. A significant part of the refining operation involves the recovery of by-products, particularly the precious metals silver and gold. Electrorefining involves the transfer of lead from an impure anode sheet, through an electrolyte to a high purity lead cathode. Crude bullion, after copper, arsenic and antimony removal, is cast into anodes, which are placed in tank cells. The electrolyte commonly used in the Betts Process is a solution of lead fluorosilicate and free fluorosilicic acid. Lead is deposited on lead starter sheets, which are removed from the cells and melted to high purity refined lead. Impurities are contained in the anode slimes and are collected and processed by pyrometallurgical methods for recovery of precious metals, bismuth and copper.

SECONDARY LEAD PRODUCTION Secondary lead is primarily sourced from scrap lead-acid batteries but also processed scrap metallics such as sheet and pipe. Secondary operations are characterised by relatively small plants in comparison with primary smelters, and are sized to handle scrap availability within a local area. This is determined by the economics of scrap battery collection and transport to the secondary operation, and it follows that the largest secondary plants are located in the high vehicle density areas of the USA. The first step in secondary lead processing is the breaking and separation of scrap batteries. In this step, batteries are shredded or disintegrated, then the battery components are separated by physical methods into metallic components, pastes containing lead oxides and sulfate, plastics from battery cases and plate separators, and waste battery acid, which is usually neutralised with lime to form gypsum. Polypropylene recovered from cases is a valuable material and can be recycled for reuse. The metallic components may be simply melted to recover lead, largely contaminated with antimony, but the battery pastes are treated in a smelting process in which they are reduced, using a carbon-based fuel, to lead bullion and a waste slag. Processes used for secondary smelting include the following:

• blast furnace, • reverberatory furnace, • short rotary furnace, • rotary kiln, • submerged lance slag bath reactor (eg Isasmelt or Ausmelt Processes), • electric furnaces, and • leaching and electrowinning processes. Some refining of secondary bullion is required for the removal of antimony, arsenic, copper and tin. This is usually done in kettles using standard pyrometallurgical refining techniques, but is far less extensive than for primary lead.



13


Cumulative Capacity Above ’000 (t/a)

Secondary lead is recovered either as ‘soft lead’ or as ‘hard’ or antimonial lead. The metallic components of automobile batteries such as plate grids and posts may be made from antimonial lead alloys containing up to ten per cent antimony, but usually less than three per cent. This provides the source of antimony in secondary lead, but it can be controlled to some extent by separately processing metallics and non-metallic scrap. There is a trend to the use of calcium lead alloys in place of antimony for sealed batteries, which significantly reduces the quantity of antimonial lead produced by secondary smelters. 4000 3500 3000 2500 2000 1500 1000 500 0 0

20

40

60

80

100

120

140

Capacity ’000 (t/a)

FIG 1.6 - Distribution of world secondary lead smelter capacity.

100

Primary

90

Secondary

Number of Plants

80 70 60 50 40 30 20 10 0 0 to 20

20 to 40

40 to 60

60 to 80

80 to 100 to 120 to 140 to 160 to 180 to 200 to 100 120 140 160 180 200 220

Capacity Range ’000 (t/a)

FIG 1.7 - World lead smelters – size distribution.

14




There are around 150 secondary smelters worldwide with a median capacity of 15 000 t/a of lead, although there are many small plants and the first quartile size is 6000 t/a. Figure 1.6 shows the distribution of world secondary capacity as the cumulative capacity above a given plant size, and compares with Figure 1.5 covering primary smelters. Clearly secondary smelters are much smaller than primary smelters. The number of primary and secondary smelters within a given size range is illustrated in Figure 1.7, which shows the significant difference in numbers and in plant capacities. The scale of these plants also has an impact on the technologies used for secondary smelting in comparison with primary smelting, and the most common approach is the use of the short rotary furnace.

REFERENCES AND FURTHER READING Henstock, M E, 1996. The Recycling of Non-Ferrous Metals, 342 p (International Council on Metals and the Environment: Ottawa). International Lead Association website, . International Lead Association Europe website, . Lead Development Association International website, . Siegmund, A H J, 2000. Primary lead production – A survey of existing smelters and refineries, in Proceedings Lead-Zinc 2000, pp 55-116 (The Minerals, Metals and Materials Society: Warrendale).



15

CHAPTER 2 Historical Background As already discussed, lead has been used by humans for over 7000 years. Lead oxides were used as pigments in ancient Egypt and also as glazes for pottery, and objects made of the metal have been found dating from 3800 BC. The Chinese used lead coinage dating back to 2000 BC. Lead sheet was reportedly used in the construction of the Hanging Gardens of Babylon, and it is known that lead was used at that time for embedding bronze and iron connection brackets into stone blocks used in construction. It was extensively used by the Romans in building construction, for water pipes, for coinage and in warfare. The Romans were also familiar with lead-tin alloys for use as solder. Mining of lead ores is recorded at Mount Laurion in Greece in the fifth century BC. It was mined by the Phoenicians in Spain and later by the Romans in the Rio Tinto region, as well as in Derbyshire in Great Britain and widely throughout Europe, but particularly in Silesia, Bohemia and the Hartz Mountain area of Germany. The history of lead is also inextricably linked with the mining and recovery of silver, which was produced for its value as a currency of trade, as well as a precious metal for the manufacture of jewellery and artefacts. Because of lead’s association with silver and its potential use for degrading silver coinage, lead mining and smelting operations were often closely controlled by the application of strict laws.

LEAD PRODUCTION IN EARLY TIMES Lead can be reduced from its oxide at relatively low temperatures compared with other metals, and the use of a wood fire is sufficient to produce lead metal. Early lead smelting methods used a stack of wood and ore piled in a hollow or ‘bole’ on the side of a hill crest to utilise strong winds to intensify the fire. In Britain these smelting sites were known as ‘bolehill’ or ‘bloomery’ sites, and were common for metal smelting in general. A small retaining wall could be built around the base to retain a bed of coals and provide a reducing zone. Channels allowed molten lead to run out from the furnace. The next development was the application of hand, or foot, operated bellows to provide an air blast to the hearth, which enabled the smelting site to be more conveniently located near ore supplies. The furnace was constructed as a short, square shaft with a bottom opening for the bellows and to allow metal and slag to run out. The shaft was packed with charcoal and ore. An example of this type of furnace is the Catalan Forge, introduced in Spain about 700 AD. These furnaces were primarily used for iron production and evolved into the blast furnace in time. The furnaces in use for general smelting applications in the 1500s, including lead and silver, have been described by Agricola, in the first detailed descriptions of smelting practices up to that time. (Agricola, 1950). The furnace was typically a rectangular shaft 370 mm wide by 460 mm deep and 1500 mm high, equipped with a single tuyere through the rear wall close to the hearth, and operated by bellows. It was constructed of stone for the rear and side walls, with brick for the front wall. The hearth was made of rammed clay mixed with powdered charcoal. A number of furnaces were constructed against a large stone wall, behind which were located a series of bellows – one for each furnace – operated by a shaft linked to a water wheel. Figure 2.1 gives an illustration taken from De Re Metallica (Agricola, 1950).



17

CHAPTER 2 – Historical Background

FIG 2.1 - Medieval smelting furnaces (from Agricola, 1950).

The medieval furnaces operated on charcoal, but wood was also used when smelting lead ores, which was the simplest of the various smelting operations for which the furnaces were employed. For the smelting of silver and gold ores, lead was also added to the furnace and to the forehearth, as a solvent for the precious metals. Production of slags and matte were also common with precious metal smelting and there was considerable recycling and reworking of these materials. The furnace generally operated only for a few days and was then cleared of accretions and the walls were replastered with ‘lute’, a paste of clay and fine charcoal. The skill of the furnace operator was most important in regulating the air blast from the bellows and in the placement of the ore charge towards the front of the shaft so as to avoid the formation of a sintered mass or ‘sow’. Natural fluxing materials such as fluorspar were often also added to the charge depending on the nature of the ore. In the 16th century, lead smelting tended to develop towards the use of a shallower hearth akin to the blacksmith’s forge, with rear fixed tuyeres blown by bellows that were driven by water power, and a mix of selected lump lead ore or concentrate and charcoal was piled on the hearth and hand rabbled. Crushing and simple gravity or hand-sorting of ores was becoming more common at this time. Hearths of this type were the Scotch hearth, as shown in Figure 2.2, and the Moffat ore hearth. Early hearth dimensions were 0.6 to 0.9 m2, with a central depression of around 100 mm deep to retain lead.

18




To Chimney

Charge Door

Charge and Fuel Bed

Tuyere Lead Pool in Hearth

Lead Pot

Air Blast FIG 2.2 - Schematic of the Scotch hearth furnace.

The hearth was started with a charcoal fire, onto which lead ore was added with more fuel. The charge was worked by hand-stirring with iron tools. Lump material was removed onto the front working stone, was broken to allow oxidation and then pushed back onto the heap. The basin in the hearth filled with molten lead, which then overflowed into a cast iron collection pot located at the front of the hearth. Lime was added at around one to 1.5 per cent of the ore charge to cover the molten lead and enhance the formation of a crumbly slag, which allowed good blast penetration and sulfur removal. The temperature was kept as low as practical to maintain this slag regime and to minimise lead fuming. A lumpy slag was removed periodically and generally reported around 20 per cent lead. The high lead slag was stockpiled and, with improved smelting techniques, many of these old hearth slags were reworked to recover additional lead. The hearth furnace required lump material of high lead grade to avoid excessive dusting and fuming, and to minimise slag formation and loss of lead in that slag. It was consequently favoured by early lead smelting operations in the Mississippi Valley with clean high-grade galena ore. At these sites a water-jacketed version was developed, constructed of water-cooled cast iron panels in a U-configuration on the long axis, and termed the ‘American Water Backed Hearth’. A later development favoured by the Missouri lead producers was the Newman Hearth, a mechanically rabbled version of this technique. An eight foot (2400 mm) long by 20 inch (203 mm) wide hearth containing an eight inch deep bed could produce three tons of lead in eight hours from high-grade concentrates. The mechanical version relieved the smelterman of the laborious task of constantly hand-rabbling the charge to break up accretions and ‘sows’, with exposure to heat and fumes.



19


The copious fume emissions from these furnaces and the need to process finer ores eventually favoured the use of the reverberatory style hearth furnace, which did not require the use of bellows and is illustrated in cross-section in Figure 2.3. These furnaces probably evolved from the open oven style hearth furnaces used at Carni in Austria, and the Saxon furnaces, which resembled baking ovens.

Filling port Firebox Flue

Hearth

Tap hole FIG 2.3 - Reverberatory lead furnace.

For the standard reverberatory furnace, a batch of galena was added to the furnace hearth and was roasted with hand-rabbling for about two hours, in which time part of the lead sulfide was directly oxidised to lead sulfate. The resulting mixture of lead sulfate and unreacted lead sulfide was thoroughly mixed and the temperature of the furnace was increased. This allowed the ‘roast reaction’, as given in Equation 2.1, to take place, with copious emission of sulfur dioxide: PbSO4 + 2PbS = 3Pb + 2SO2

(2.1)

Any silica in the concentrate tended to react with lead oxide (PbO) to form lead silicate, and in the final stage of the process lime was added to the furnace charge and mixed in with the slag and unreacted ore, for the purpose of decomposing the lead silicate in accordance with Equation 2.2: 2PbSiO3 + 2CaO + C = 2CaSiO3 + CO2 + 2Pb

(2.2)

Following this step, molten lead bullion was tapped from the base of the furnace hearth as a crude impure or ‘hard’ lead. In some operations, particularly for those processing lump feed, the roast-reduction cycle was repeated a number of times, with the temperature raised for each cycle. In this situation silver tended to concentrate in the first run lead bullion, and could be four times the silver content of the final lead run. This was a useful approach to handling high silver ores, so as to reduce the effort required in silver recovery by the Pattinson Process or by cupellation. The earliest reverberatory hearths were introduced around 1720 in Silesia and England. Early types were the Corinthian furnace, the English or Flintshire furnace and the Silesian furnace. Silesian furnaces used in Germany were up to 25 foot (7600 mm) long by 8 foot (2400 mm) wide, with five working doors in each side. Labour requirements were 15 to 20 man hours per tonne of lead produced, with coal consumption close to 0.8 tonnes per tonne of lead produced.

20




In later practice during the 1800s two hearth furnaces were commonly used: one for calcining with gases going to a chamber-type sulfuric acid plant, and the second or ‘flowing’ furnace operating at a higher temperature, in which calcined material and coke were melted with the addition of lime and fluorspar to produce a fluid slag. Often pig iron or scrap iron was also added as a reductant. As well as lead bullion and a low lead slag, matte (or ‘regulus’) was also produced and could contain significant amounts of lead. The matte was reworked by calcining and returned to the reduction furnace, thus producing a second matte enriched in copper. Matte could be recycled a number of times. The above smelting operation still resulted in a significant amount of lead being volatilised, forming a fume in the exit gases from the furnace, probably rich in toxic elements such as arsenic, as well as sulfur dioxide. This caused the destruction of vegetation around the smelting operation, and the poisoning of farm animals and cattle feeding nearby. To reduce this effect, long horizontal flues up to 1.5 km in length were constructed from the furnace to the final vent stack, allowing the bulk of the fume to settle out onto the walls and base of the flue. Collected fume contained of the order of 33 per cent lead and was recycled. Collection methods were improved by the addition of drop-out or condensing chambers immediately following the smelting furnace, which were introduced in England around 1780. The crude bullion from the smelting furnace was allowed to oxidise in shallow open pans and the dross skimmed from the top removed arsenic and antimony, ‘softening’ the resulting lead metal. The use of iron metal additions to reduce galena or lead oxide was first noted in India in the 14th century. In this method, iron was combined with lead sulfide to form metallic lead and an iron matte as in Equation 2.3, thus limiting the formation and emission of SO 2 (Dube, 2006): PbS + Fe = FeS + Pb

(2.3)

Lead ore, charcoal and iron were placed in crucibles within a furnace and later removed to separate and recover the lead. In the late 18th century in Europe, iron reduction was applied with the addition of high-grade lump ore, charcoal and iron to a small shaft furnace. Iron use was around 12 to 15 per cent of the ore charge. This practice first appeared in Claustal in the Upper Hartz region and later at Tarnowitz in Silesia and Przibram in Bohemia. The use of iron for lead reduction also occurred in Japan, but using an open pan hearth filled with burning charcoal into which lead ore and pig iron were charged. Iron use was two to three times higher than reported in European practice.

THE LEAD BLAST FURNACE The hearth furnace processes described above were only efficient for the smelting of lump high-grade ores with low levels of associated ‘earthy materials’ such as silica and iron minerals, as well as zinc. To handle lower grade materials it was preferable to use a higher temperature process and produce a molten slag containing the gangue minerals, and hence the blast furnace was applied. The blast furnace for the production of iron had evolved from the early charcoal, fuelled hearth and shaft furnaces blown by water-powered bellows. Coke was first used by Abraham Derby at Coalbrook Dale, in England in 1713, and with the invention of the steam engine and its use for operating blowing engines, together with the development of the hot blast by Neilson in Glasgow in 1830, the iron blast furnace was well established by the mid 19th century. Furnaces were originally constructed from massive sandstone blocks bound with iron straps. Later, thin steel shell construction was used lined with refractory brick.



21


Early shaft furnaces for lead were used in Freiberg and in the Harz region in Germany. These evolved from the earlier hearths, using a rectangular cross-section with one tuyere at the back of the furnace facing a tapping access at the front. This developed into a horseshoe shape with a number of tuyeres and finally into a circular shaft. The Castilian furnace from Spain was an example of the early circular shaft furnaces for lead. It was constructed of sandstone blocks and is shown in Figure 2.4. Flue Charging Ports

Sandstone Block Shaft

Tuyeres (5)

Lead Pot Slag Car

Rammed Hearth

FIG 2.4 - Castilian blast furnace.

Later furnaces, such as the Lower Harz furnace, the Claustal furnace and the Przibram furnace (small versions of an iron blast furnace), were of brick construction. In 1863 the Pilz furnace in Germany introduced water-cooled cast iron plates at the base of the furnace in the tuyere zone. This was followed in 1891 by the American Water Jacketed furnace, at Great Falls in Montana, and the Globe smelter furnace at Denver, Colorado. Water jackets contributed significant benefits by enabling rapid repairs and much longer operating campaigns, due to the reduction in accretion formation in critical narrow areas of the furnace. With the adoption of water jackets, furnace design tended to move from circular to rectangular cross-section for construction simplicity, also allowing significant increases in productivity from each furnace. The Rochette furnace used at the Atenau smelter near Claustal in the Upper Hartz region introduced the concept of a long rectangular hearth, with rows of tuyeres on each of the long sides and tap holes at each end. The main driver for this change was the realisation that blast penetration from the tuyeres was a limiting factor; the only way to get higher production was therefore to retain optimum width for this purpose and to make the furnace longer. Around the end of the 19th century large furnaces were 36 inches wide (914 mm) × 108 inches long (2740 mm) at the tuyere level (2.5 m 2 hearth area), and around 16 ft high (4870 mm).

22




With the use of higher blast pressures at the Port Pirie smelter, blast penetration at the base of the furnace was increased to an optimum width of around 1200 mm. Capacity was further increased by extending the length of the furnace, but this reached limits at around 7.5 m, set by the ability to tap slag from one end. By tapping slag at both ends or from the centre, the length could be extended to around 11 m, giving a hearth area of around 13 m 2, which is similar to the typical lead blast furnace today. Although the width at the base of the furnace was limited to achieve blast penetration, narrow shafts permitted shaft accretions to readily bridge across and block the furnace. This was corrected by expanding the width of the furnace above the tuyeres using a short sloping section or ‘bosh’, with either a tapered or straight upper shaft. In the early 1900s the upper limit to the furnace width was around 2000 mm. In 1935 Port Pirie further extended the width at the tuyeres to 1524 mm and added a ‘chair jacket’ and a second upper row of tuyeres with a width of 3048 mm. Further, in 1940 the formation of accretions was minimised by extending the water jackets to the full height of the furnace. This design evolved to the largest lead blast furnace currently in operation. The Port Pirie blast furnace is shown in Figure 5.6 (Chapter 5). The use of oxygen enrichment of blast air has also enabled the capacity of the blast furnace to be further increased, and is applied in most operations. Details of blast furnace performance and operation are covered in Chapter 5. In 1960 the first standard commercial scale Imperial Smelting Furnace (ISF) was constructed at Swansea in the UK, as an adaptation of the lead blast furnace, to simultaneously produce zinc and lead. The furnace operated with a hot top to retain zinc in the vapour phase. The top was sealed and gases passed through a lead splash condenser to strip zinc from the gas phase into a lead-zinc bullion which could be cooled for separation of crude zinc and lead metals. The ratio of zinc to lead production from these units is generally more than 2:1, and lead production from the standard unit is close to 40 000 t/a. Thirteen plants were constructed around the world but due to unfavourable economics a number of these have now closed. Details are given in Chapter 6.

PREPARATION OF BLAST FURNACE FEED Originally, lump sulfide ores were fed to the blast furnace, but this tended to produce large quantities of matte, requiring appropriate levels of iron flux to form the matte. Silica fluxing was also important to displace lead from sulfates and to form a slag. Limits on the input of sulfur due to excessive matte formation were generally around 15 per cent in the feed material. Speiss (an iron-arsenic-antimony intermetallic) was often also present as a separate phase. For lower grade ores with higher sulfur contents, and for ores with high arsenic content, it was necessary to eliminate these elements by roasting prior to feeding to the blast furnace. Early roasting was done in open heaps over fuel beds with a tunnel beneath to supply combustion air. This was applied at Broken Hill, Australia for the sintering of lead slimes, which were formed into bricks and dried, and then stacked in an open heap over a fuel bed and burned for ten to 15 days. The resulting material was sintered and quite suitable as blast furnace feed. Open compartmented pads or stalls were also used, often venting to a central collection flue for combustion gases. However, roasting at low temperatures produced fine calcines, which were not a practical feed for the blast furnace, and higher temperature roasting to melt or sinter the material was necessary, so that lump material could be produced. To effectively achieve this, roasting furnaces were developed in the form of reverberatory hearths, and were introduced around 1790. Typical roasting furnaces were up to 5 m wide and 20 m long, with a fire at one end and exhaust flue at the other end. A series of doors along the side of the furnace allowed access for hand-raking and



23


movement of material along the length of the furnace from the cold feed end. Lead concentrate and ironstone, lime and silica fluxes were fed at the cold end. The hot end of the hearth, near the fire, contained a depression or sump called the ‘fuse box’ where the calcine melted and from where it was manually scrapped out into slag pots. The slag was cooled and solidified, and then broken into lump material suitable for blast furnace feed. A typical roaster of the size indicated would process 5 t/d of lead feed using three men per shift and consuming 3 t/d of coal. At the Pontgibaud smelter in France, a reverberatory hearth was used for batch calcining, followed by elevation in furnace temperature to cause surface melting and sintering of the calcine into an agglomerated mass, rather than complete melting of the charge. This was withdrawn from the furnace, cooled and broken into lump for blast furnace feed. In order to improve the intensity and efficiency of the roasting process, the Huntington-Heberlein process was introduced in the 1890s. Partly roasted material from the hearth roaster was moistened and placed in a pot fitted with a lower grate (or converter), with a layer of hot material on the grate, and was subjected to an air blast. The construction of the roasting pot or converter is shown in Figure 2.5.

Hood

Trunnions

Grate

Air Blast FIG 2.5 - Sintering pot (or converter).

These converters could be considered the forerunners of today’s updraft sintering machines. However, they had serious deficiencies: they were batch operations, they generated much fume, and the work of manually handling and breaking the sintered material from the pots was laborious and unhealthy. This approach significantly improved the productivity of both the roasting operation and the blast furnace, and lead smelters in the early 1900s had large numbers of converters producing blast furnace feed.

24




Alternative processes at the time were the Bradford-Carmichael process, in which lead ore was mixed with dehydrated gypsum (plaster) as a binder, was formed into lumps and then processed in the converter, the Savelsberg Process, which fired ore and limestone over a fuel bed in a converter to produce a sintered material as blast furnace feed. A circular rotating furnace using a downward blast through a bed of ore and limestone, covered with a surface layer of wood chips as a starting fuel, was developed by the Cerro de Pasco Corporation in Peru to provide a suitable sintered blast furnace feed. The major advance to overcome the disadvantages of the pot roasting methods came with Dwight-Lloyd sintering machine. Originally this consisted of a series of boxes with an open grate base, running on rails over a suction box. The boxes or ‘pallets’ were firstly pushed through a small reverberatory furnace to ignite the top surface of the charge in each pallet box, and then continued over the suction box until combustion was completed. The pallets were inverted to empty the contents and were returned to the beginning of the process. The principle was extended to the development of the standard downdraft sinter machine around 1910 and later to the updraft machine in 1955. Details of current sintering processes are covered in Chapter 4.

BLAST FURNACE PRODUCTS Early blast furnace operations processing lead ores or concentrates from hand-sorting or gravity separation methods had to contend with much higher levels of impurity metals than later operations processing flotation concentrates. In particular, blast furnace feed contained high levels of sulfur, iron, arsenic, copper and zinc. As well as bullion and slag, the blast furnace produced significant quantities of matte and speiss, and the presence of zinc created significant problems with furnace accretions and high slag viscosities. As well as containing iron at around 40 to 45 per cent, matte contained about 12 per cent lead, most of the copper from blast furnace feed, about half the zinc and a substantial proportion of the silver. Matte was initially roasted in open heaps or stalls to remove sulfur and was then recycled to the blast furnace, but later reverberatory roasting furnaces were used. A shaft kiln was used at the Harz smelter in Germany. The recycle of roasted matte to the blast furnace resulted in the progressive enrichment of the copper content of matte, until it reached a grade where it could be processed to blister copper in a converter. Table 2.1 shows the composition of matte produced after five recycling stages from a smelting operation around the 1890s (Hoffman, 1899). TABLE 2.1 Progressive matte enrichment. Matte production cycle Lead content (%)

1

2

3

4

5

13.5

8.3

10.0

9.0

9.0

Iron content (%)

48

44

31

20

12

Copper content (%)

5.7

12.8

27.8

42.9

50.9

Sulfur content (%)

25

20

21

18

21

Zinc proved to be a particularly troublesome problem and many techniques were developed to remove zinc from blast furnace feed. Mineral separation techniques using gravity had limited effectiveness for some ores, and in particular cases it was necessary to leach zinc from roasted ores



25


either using water after low temperature sulfation roasting, using dilute sulfuric acid to extract zinc sulfate, or water and SO2 to remove zinc as a sulfite. The development of practical froth flotation from around 1913 substantially improved the separation of zinc and lead sulfides and presented cleaner lead concentrates to the lead smelters. This also increased the fineness of concentrates and necessitated the introduction of sintering methods for blast furnace feed preparation. Speiss was the other significant blast furnace product from earlier smelting operations. It contained significant amounts of entrained particulate lead as well as silver and gold. Speiss was often roasted in heaps or a calcining furnace and recycled to the blast furnace. At the Trail smelter in Canada it was treated in a bottom blown converter with the addition of molten lead. The lead captured most of the silver and gold and no doubt significant amounts of arsenic were volatilised into the gas stream.

LEAD REFINING Simple cooling of the lead bullion from the smelting furnace initially allowed the separation of a black dross containing most of the dissolved zinc, iron, tin and oxygen in hot furnace bullion. This dross could be skimmed off and worked up if the tin content was high enough. With further cooling, much of the copper and sulfur content came out of solution forming copper-rich crusts. The process was termed ‘copper drossing’. In early lead-refining practices, further purification firstly involved the oxidation of bullion in shallow open pans. The dross formed contained antimony and arsenic and was skimmed off until the lead was ‘softened’. Oxidation softening in a reverberatory furnace was practised in the mid 1800s and was developed into a continuous operation in the early 1900s. The alternative Harris Process, for removal of arsenic, antimony and tin by the addition of caustic soda and sodium nitrate, was introduced in 1920. Separation of low levels of silver was not practised until the development of the Pattinson Process in 1829 as detailed in the next section. The Parkes Process for silver and gold removal by zinc addition was introduced in 1872. Remaining copper and other impurities were originally removed by the addition of zinc, and the removal of zinc was by chlorine to form a zinc chloride dross or by drossing with caustic soda. Vacuum dezincing of lead was developed as a practical technique in 1946. ‘Fine’ copper removal by sulfur drossing was developed in 1923. Bismuth became a significant issue as the uses of lead became more demanding and was the prime reason for development of the Betts Electrolytic refining process in 1902. The alternative KrollBetterton process involving the use of calcium and magnesium metal additions was introduced in 1936. Details of thermal refining practices are covered in Chapter 12 and electrolytic refining in Chapter 13.

SILVER RECOVERY From early times silver was an important source of wealth, but particularly so during the Middle Ages in Europe. Many early lead smelting operations were for the prime purpose of recovering silver, and lead could be regarded as a collector for silver and as a by-product. Extensive mining of silver with co-product copper and lead occurred throughout central Europe, notably in Austria, Saxony and the Harz district of northern Germany. Most silver ores are sulfides and contain argentite or silver glance (Ag2S), although there are also many complex mixed sulfides with antimony, arsenic, copper and lead. In general these minerals occur with lead sulfide or galena and with copper sulfide ores.

26




Up to the Middle Ages silver rich ores were smelted with charcoal in hearth or shaft furnaces as previously described, usually with the addition of lead and the use of a lead pool in the forehearth to collect a lead-silver alloy. Silver was then separated from the lead by cupellation. The cupellation process involved blowing air over a crucible or pan containing molten alloy on a bed of bone ashes, so as to oxidise the lead to molten litharge. The bone ashes absorb the molten litharge, but surface tension effects cause rejection of the residual silver, which remains as a pool of metal on the surface. This procedure was probably known as far back as 2500 BC, and is mentioned in a number of places in the Old Testament. It is also well accounted for in descriptions of the Mount Laurion Mines in Attica, Greece, which provided much of the wealth of the city state of Athens around 500 BC. Mount Laurion ores contained 40 to 90 ounces (or 1.2 to 2.8 kg) of silver per tonne of lead. The litharge produced by cupellation was known as ‘spuma argenti’ and could be re-smelted back to lead metal. Although normal lead bullion usually contained varying but relatively small amounts of silver, it was not until the development of a process for the desilverising of lead by Pattinson in 1829 that silver could be produced from ores containing relatively small quantities, such as the ores from Great Britain. The separation of silver by the Pattinson process used fractional crystallisation, in which molten lead was cooled and partly solidified in a pan while being briskly stirred. The solid lead crystals were relatively pure, leaving silver in the remaining liquid. A row of about nine pans were used, each heated by a fire from below. Crude bullion from the smelter was fed to the middle pan, from which solid crystals were transferred to the first pan on one side and the remaining liquid to the first pan on the other side. This process was repeated from one pan to the next up and down the line to give a purified lead with low silver content at one end and a residual liquid of around 9 kg of silver per tonne at the other end. The silver rich lead was subjected to cupellation to recover a silver bullion. Today, precious metals are separated from lead bullion using the Parkes Process, following the removal of copper, arsenic and antimony. In this process, zinc is added and the lead bullion is cooled to precipitate a zinc-silver alloy, which is removed and separately treated.

DIRECT SMELTING Up until the 1980s, the sinter plant–blast furnace technology was almost exclusively used for the production of primary lead. The early exception was the Boliden electric furnace process from the 1950s, and the Boliden Kaldo or top-blown rotary converter in the 1970s. Major effort was concentrated in the l970s and 1980s on alternative smelting processes to achieve sulfide oxidation directly to lead bullion without full oxidation, to lead oxide, followed by reduction of the lead oxide to metal. The aim was not only to reduce overall energy consumption and the use of coke as a costly reducing agent, but also to address growing attention to occupational health and environmental issues associated with the older technologies. This could be achieved by full enclosure of processes and reduction in the volume of gases produced by more intensive smelting using high levels of oxygen enrichment. As a result, the Kivcet flash smelting process, the QSL process and the Sirosmelt top-submerged lance slag bath process (as Isasmelt and Ausmelt processes) were all developed at this time and are detailed in Chapter 7. With direct smelting from sulfide to lead metal, involving partial oxidation in a single stage, it is not possible to achieve good lead recovery with high sulfur elimination and low lead levels in slag. Hence, most processes involve a two-stage sequence of oxidation, followed by reduction of the lead oxide so formed to lead metal. Conducting these steps within the same piece of equipment can enable the excess heat generated by oxidation to be used in the reduction stage, and hence greatly improve the thermal efficiency of the smelting operation.



27


Copper smelting in the Balkans Neolithic Age

Lead metal produced in Egypt Egyptian Kingdoms

Bronze Age

Hellenistic Period 0

Iron Age

Roman Empire

Silver recovered by cupellation

Bolehill lead smelters Ore concentration by water washing

Catalan Forge for lead smelting Iron blast furnace Stamp mills for ore crushing

Medieval World Renaissance Europe

Hearth smelters for lead

1713 Use of coke for iron smelting 1720 Reverberatory hearth smelter for lead

Industrial Revolution

1778 Fume collection from smelter gases 1790 Roasting of lead ores in a hearth furnace 1829 Pattinson silver removal process

Industrial Age

1863 Water jacketed blast furnace 1872 Parkes process for silver removal 1895 Sintering pots in use

1900 AD

1902 Betts electrolytic lead refining 1910 Dwight Lloyd sintering machine 1920 Harris process for arsenic /antimony removal

World War I

1923 Copper removal by sulfur addition 1935 Full water jacketed blast furnace 1936 Kroll Bretterton process for bismuth removal

World War 2

1946 Vacuum dezincing process 1953 Boliden electric furnace lead process 1955 Updraft sintering machine 1976 Kaldo lead smelting process 1978 Kivcet process development 1980 Isasmelt lead process 1984 QSL process Electronic Age

1990s Electrowinning processes

FIG 2.6 - Key technology milestones.

SECONDARY LEAD Until the widespread use of the lead-acid battery, metallic lead scrap was originally simply melted for reuse by smelters or by consumers and foundries. Although the battery was invented by Gaston Plante in 1859, it was not until the automobile became ubiquitous that specialised secondary smelters came

28




to prominence. When the quantity of available batteries was relatively small they could be added to a primary blast furnace. However, as the quantity grew, this created problems with the sulfate load and the introduction of chloride from the plastic materials contained in the batteries. Initially the older reverberatory furnaces were used for processing battery scrap, with slag treatment in a small blast furnace. This practice has persisted in North America, but in Europe the reverberatory furnace evolved from the stationary form to the short rotary furnace. This provided much improved mixing and operational efficiency, and separate equipment for slag treatment was not required. The use of soda slags in the short rotary furnace also captured sulfur from the battery paste, and provided an efficient simple operation. However, problems with the disposal of soda slags resulted in the development of pre-treatment techniques to remove sulfur prior to smelting, and this has been coupled with the separation of battery components and the recovery of polypropylene as a valuable by-product. Presently there is a range of process configurations used for the recovery of secondary lead. These are detailed in Chapters 10 and 11. The secondary lead industry now represents almost 70 per cent of total lead supply and hence exceeds the scale of the primary smelting industry.

HISTORICAL SUMMARY A timeline summary of important milestones in the development of lead smelting technology is shown in Figure 2.6. Major changes in smelting technology have been few over the past 250 years and may be limited to:

• the reverberatory hearth process; • the blast furnace and application of water jackets; • the Dwight-Lloyd sintering machine; • direct smelting technologies – Kivcet, QSL and Top submerged lance processes; and • direct leach – electrowin processes – yet to be commercialised. In relation to lead refining major innovations have mainly occurred in the 70-year period from the late 1800s to the early 1900s, largely due to the quality demands for lead as the industrial age developed.

REFERENCES AND FURTHER READING Agricola, G, 1950. De Re Metallica (1556), translation by H C and L H Hoover (Dover Publications Inc: New York). Collins, H F, 1910. The Metallurgy of Lead (Charles Griffin and Co: London). Dube, R K, 2006. The extraction of lead from its ores by the iron-reduction process: A historical perspective, Journal of Metals, October, pp 18-23. Eissler, M, 1891. Metallurgy of Argentiferous Lead: A Practical Treatise on the Smelting of Silver – Lead Ores (Crosby-Lockwood and Son: London). Hoffman, H O, 1899. The Metallurgy of Lead (The Scientific Publishing Company: New York). Also revised edition, 1918 (McGraw Hill: New York). Percy, J, 1870. The Metallurgy of Lead (John Murray: London).



29

CHAPTER 3 Raw Materials Lead sulfide or galena is the main source mineral for lead. It was earlier mined as massive rich veins and was smelted directly. Today it is usually extracted as a co-product with zinc mineralisation and is separated and concentrated using flotation.

LEAD MINERALOGY The common lead minerals are given in Table 3.1. TABLE 3.1 Common lead minerals. Mineral Galena

Formula

Lead content (%)

PbS

86.6%

Anglesite

PbSO4

68.3%

Cerussite

PbCO3

77.5%

Leadhillite

PbSO4.2PbCO3.Pb(OH)2

76.8%

Jamesonite

Pb4FeSb6S14

50.8%

Pyromorphite

3Pb3P2O8.PbCl2

76.4%

Bournonite

3(Pb,Cu2)S.Sb2S3

24.7%

Mimetite

(PbCl)Pb4(AsO4)3

69.7%

As the predominant source of lead, galena deposits were commonly formed by hydrochemical processes through the cooling of mineral-rich solutions associated with magmatic eruptions or intrusions, and occur in beds or veins. Galena can occur as a replacement or metasomatic deposit in limestone associated with dolomitisation, and is typified by the Mississippi zinc–lead deposits. Associated minerals are frequently zinc sulfide or sphalerite, chalcopyrite, pyrite and minor mixed sulfide minerals often containing silver. Weathering and oxidation of the primary sulfides leads to the formation of deposits of anglesite and cerussite, which were the early source minerals in Britain. Most lead ores also contain silver, antimony, arsenic and bismuth. Silver is an economically important constituent of lead concentrates and is often necessary to allow the smelting operation to be profitable. Silver may be present as argentite (Ag2S) but is probably more commonly present in association with antimony, copper and arsenic minerals such as pyrargyrite (Ag3SbS3), proustite (Ag3AsS3), freieslebenite ((Pb,Ag)8Sb5S12), polybasite ((Ag,Cu)16(Sb,As)2S11), and tetrahedrite ((Cu,Fe)12Sb4S13) in which silver can partly replace copper. Small amounts of many of these minerals can be present in solid solution in galena. The presence of gold is also common and can be of value as it is recovered with the silver. Bismuth is an unwanted element in refined lead and is difficult to remove in the refining process. Hence, its presence in lead concentrates attracts a significant cost penalty. In general there is a complex series of sulfide minerals of the form MS:(As,Sb,Bi)2S3 in which M may be Pb, Fe, Cu, Ag or Tl. There is a wide range of ratios between the two components and mix or replacement of the various elements. This gives rise to the common presence of these elements to varying degrees in lead concentrates.



31

CHAPTER 3 – Raw Materials

SEPARATION AND CONCENTRATION METHODS Early lead smelting used directly mined massive sulfide vein ores rich in galena, as well as surface deposits of carbonates (cerussite). Any concentration may originally have been by hand-sorting of lump ore. However, because of their high density, it was relatively simple to separate and concentrate coarse-grained lead ores by gravity concentration methods. Relative densities of commonly occurring minerals are shown in Table 3.2. TABLE 3.2 Density of common minerals. Mineral

Specific gravity

Hardness (Mohs’ Scale)

Galena (PbS)

7.4 - 7.6

2.5 - 3.0

Anglesite (PbSO4)

6.1 - 6.4

2.8 - 3.0

Cerussite (PbCO3)

6.5 - 6.6

3.0 - 3.5

Sphalerite (ZnS)

3.9 - 4.1

3.5 - 4.0

Smithsonite (ZnCO3)

4.3 - 4.5

5.0

Willemite (Zn2SiO4)

3.9 - 4.2

5.5

Pyrite (FeS2)

5.0

6.0 - 6.5

Pyrrhotite (FeS)

4.6

3.5 - 4.6

Siderite (FeCO3) Haematite (Fe2O3)

3.9

3.5 - 4.0

4.9 - 5.3

5.5 - 6.5

As shown in Table 3.2, there are significant differences between the specific gravity of common lead minerals and associated minerals. These differences enabled gravity classification methods such as elutriation classifiers, spirals, shaking tables, jigs and vanners to be used to produce a concentrate, and gravity concentration was the sole method used prior to the development of flotation in the early 1900s. One issue was the relative softness of the lead minerals, which results in the formation of slimes in crushing and grinding circuits. After size classification using equipment such as the Dorr rake classifier, the fines were filtered to form a lead-rich concentrate of around 50 - 60 per cent lead. The classifier sands were then subjected to gravity separation on shaking tables or similar devices to reject gangue minerals and yield a concentrate of 60 - 70 per cent lead content. To use these methods it was necessary to limit the degree of size reduction and to separate the minerals at the earliest opportunity, and hence it was a successful approach only for relatively coarse-grained ores. The development of flotation separation methods was originally for the purpose of separating zinc minerals from mixed lead–zinc ores and initially from gravity separation tailings. Once this great breakthrough was achieved, attention was turned to the separation and concentration of lead minerals and thereafter flotation became the primary method of concentrating sulfide ores. In the past, lead ores were mined for the production of lead and silver. Today this is uncommon and most lead concentrates are produced in conjunction with zinc concentrates from lead–zinc ores. As discussed in Chapter 1, the ratio of lead to zinc mined has fallen significantly to around 0.3 in 2005. It is therefore common for ores to contain around five per cent lead or less. Generally the aim is to produce a marketable lead concentrate around 60 per cent lead or greater, although concentrates of 50 per cent lead are produced and treated. There is reasonable flexibility in the ability of smelting processes to handle a wide range of concentrate grades (50 - 75 per cent lead) and the principal effect of lower grade is the reduction in capacity of the initial or sulfur elimination stages of the smelting process.

32




Separation and concentration involves two main steps: 1. size reduction of the ore so that the individual mineral grains are liberated and separated from one another, and 2. selective physiochemical separation of the individual mineral grains by froth flotation to form separate metal concentrates. Size reduction for mineral grain liberation involves crushing and grinding to an 80 per cent passing size of less than 50 µm and for particularly complex fine grained ores to less than 10 µm. Fine grinding is a high cost in both equipment and energy and there have been developments to improve the efficiency of this operation, such as tower and stirred mills. Fine grinding is a particular problem for lead sulfide recovery because of the softness of galena and its tendency to fragment into ultra fine particles or ‘slimes’, which are difficult to capture and recover during flotation. Because of this problem it is preferable to float lead minerals before extensive fine grinding. Fine grinding can then be applied to the tailings of galena flotation and is primarily targeted at improving zinc recovery and concentrate grade. Primary crushing may reduce ‘as mined ore’ to less than 200 mm and then, in the conventional approach, it is further reduced in secondary and tertiary crushers followed by rod and ball mill grinding to the required particle size. Final sizing classification is usually achieved by operating the ball mill in closed circuit with hydrocyclones. If the ore is suitable, autogenous grinding may be used, in which large lump ore is added directly to a large tumbling mill. This replaces conventional secondary and tertiary crushing, rod and possibly ball mills, saving on the cost of steel grinding media, which can be consumed at a rate of the order of 1 - 2 kg/t of ore. The low density of lump ore in comparison with steel balls requires a much larger diameter mill to achieve the same grinding forces. Full autogenous grinding is often difficult to achieve if the ore tends to fracture and break down readily on a macro scale; however, in this case some steel balls can be added to give semi-autogenous operation and greatly extend the application of this approach, to the point where this has largely replaced the conventional multi-stage size reduction. For fine grinding, additional ball mill capacity is necessary. For very fine grinding tower or stirred mills have been developed to significantly reduce energy costs. In this type of equipment, grinding of the mineral pulp occurs within an agitated bed of coarse gravel. The bed may be agitated by circulation using a screw or by slow-moving paddle mixers. These mills have significantly lower specific energy consumption, which is critical for size reductions to 10 µm or less. To avoid ‘sliming’, fine grinding can be applied partway through the separation process rather than on total ore feed and after separation of galena and a large part of the gangue minerals. This minimises ‘sliming’ and reduces the quantity of material subjected to fine grinding, and hence minimises overall grinding costs. Physical separation of minerals generally relies on the use of froth flotation. Flotation is based on the following principles:

• sulfide minerals may be conditioned by the addition of surface active chemicals to cause them to become water repellent (hydrophobic),

• the collisions between mineral particles with hydrophobic surfaces and bubbles of air within the mineral pulp will result in the attachment of the mineral particles to the air bubbles,

• mineral particles which remain wetted (hydrophilic) will not attach to the air bubbles, and • reagents are used to produce a froth of reasonable stability to allow its collection and separation from the surface of the mineral pulp.



33


Flotation equipment consists of a series of agitated tanks through which the mineral pulp flows and into which air is dispersed as a stream of fine bubbles. A froth is formed on the surface, containing the hydrophobic mineral particles, and is skimmed off into a trough where it collapses and flows into a collection tank. It is a relatively simple matter to bulk float all sulfide minerals by the addition of surface active chemicals known as collectors. This type of material is usually an organic molecule containing a sulfur-bearing group at the polarised end which can bond with the sulfide mineral, forming an attached organic and hydrophilic surface layer. The most common collectors are: Dithiophosphates such as:

• sodium diethyl dithiophosphate

Na+(PS2(OC2H5)2)-

Xanthates such as:

• sodium isopropyl xanthate

Na+(CS2-O-C3H7)-

Differential flotation can be achieved when only specific minerals are floated. This is done by the use of additives to the pulp to depress or promote the collection of particular mineral surfaces. The particular response of minerals contained in an ore will differ widely, and hence there is a considerable variation in flotation practices from one plant to another. The general approach, however, is to first float copper and depress the other base metal sulfides, then float lead and finally zinc. The aim is generally to depress pyrite, but this can be difficult and it is often the major diluent in lead and zinc concentrates. The pH of the pulp is an important depressant, showing selectivity for individual minerals and depression in the following order as pH is raised:

• sphalerite (ZnS), • pyrrhotite (FeS), • galena (PbS), • pyrite (FeS2), • chalcopyrite (CuFeS2), • sphalerite activated with copper sulfate, and • tetrahedrite ((CuFe)12Sb4S13) for silver mineralisation. Other depressants that have been used are zinc sulfate for sphalerite and cyanide for sphalerite and pyrite, although CMCs (carboxy methyl cellulose – dextrose or starch derivatives) tend to be more commonly used at the present time. Following depression of most of the sulfide mineralisation, copper can be initially floated, followed by the flotation of lead after pH adjustment. Copper sulfate is generally then added to promote the flotation of sphalerite. In simplistic terms, copper ions react with the zinc sulfide surface to form a copper sulfide layer which will readily respond to the collector. The action of cyanide as a depressant is partly due to its ability to form a complex with copper ions in solution and prevent any such activation. The kinetics of the flotation process are an important factor and some of the relative reaction rates may be quite different, allowing some opportunity for differentiation by this means, and placing emphasis on the importance of process residence times. Other factors of importance in control of the process are the pulp density, pulp temperature, aeration rates and bubble size. A sample copper/lead/zinc separation flow sheet is shown in Figure 3.1.

34




Ore Input

Primary Grind

Copper Rougher Float concentrate

Float tails

Copper Cleaner

Copper Scavenger Concentrate

Tailing

Copper concentrate

Tail

Lead Rougher Tailing

Concentrate

Lead Scavenger

Lead Cleaner Tailing

Concentrate

Re-grind

Tailing

Lead concentrate

Zinc Rougher Concentrate

Tailing Zinc Scavenger

Zinc Cleaner Tailing

Concentrate

Re-grind Zinc concentrate

Final Tailing

FIG 3.1 - Typical flotation flow sheet for copper/lead/zinc ore.

There are many different flow sheets used for the handling of particular ores; however, a typical flow sheet usually involves an initial flotation stage or ‘rougher circuit’, the crude concentrate from which is refloated in a ‘cleaner circuit’ for upgrading. The tailings from the ‘rougher circuit’ are also refloated in a ‘scavenger circuit’ to recover any residual mineral values. A multiplicity of such stages and the recycling of intermediate streams can lead to a highly complex flow sheet with a high level of internal recycle and difficult control problems. The basic aim is to optimise the grade of the concentrate produced and the recovery of valuable metals into their respective concentrates. This may be expressed as a grade-recovery curve for a particular ore, as illustrated in Figure 3.2. The position of the curve is to some extent a function of the flow sheet used, but is critically dependent on the grind size. Finer grinding will improve minerals liberation, but the recovery of very fine particles or slimes can be reduced, due in large part to their inability to collide with and attach to air bubbles. Specialised flotation cells to generate fine bubbles can improve this situation and significantly improve lead recovery from the ore in such situations.



35


Recovery % Decreasing Grind Size

0

Ore Grade

Pure Mineral Grade Concentrate Grade % Zn

FIG 3.2 - Concentrate grade–recovery relationship.

Conditioning of the pulp with reagents prior to flotation is also of great importance, particularly when processing very finely ground material. This involves the degree of agitation and residence time used, as well as the concentration of added reagents in the pulp. The particular position on the grade-recovery curve for flotation plant operation is dictated by competitive pressures for sale of the concentrate and overall economics of the mining operation, including transportation issues. It is a complex balance, and many variables, together with the total mineral value recovered from the ore, must be considered in order to select the optimum set of operating conditions. This involves consideration of all primary and by-product concentrates produced. In some instances it is impractical to grind fine enough to achieve satisfactory separation, or excessive sliming causes loss of recovery. For these types of ores, production of a bulk or middling concentrate containing high levels of two or more valuable metals at relatively high recovery may be the only feasible method for metal concentration. This can be a suitable feed to an Imperial Smelting Furnace. In such circumstances it is possible to produce part of the zinc and lead content of the ore as high-grade concentrates, and part as a bulk or middling concentrate to maximise recovery. The economics of the mining operation are largely determined by the head grade of the ore and the metal recovery achieved into commercially acceptable concentrates.

COMMERCIAL LEAD CONCENTRATES Lead sulfide concentrates produced by the above described separation processes have individual characteristics which need to be carefully assessed by the smelter, although they are largely treated as a commodity, with compositions falling within common and accepted ranges and limits. In general, lead concentrates vary over a much wider range of compositions than zinc concentrates. This is partly due to the greater tolerance of the standard smelting processes to accept gangue minerals which form smelter slags, but it also due to the wider variation of lead recovery as a function of concentrate grade and the wider range of optimum grades for a particular mining operation than applies for zinc concentrates. Of key importance is the lead to sulfur ratio of the concentrate, since low levels (corresponding with low lead grades) will markedly restrict the lead throughput of the sulfur elimination stage of the smelting process (see Chapter 4, Table 4.2).

36




Also of major importance is the content of precious metals, particularly silver, which contributes significant value, and the presence of the critical impurity elements arsenic, antimony and bismuth, which attract cost penalties. Smelters often have a limited capacity to remove penalty elements within the lead bullion refining operation, and hence need to balance the intake of impurities so as to remain within various capacity limitations. This can often require careful blending of a range of feed concentrates to obtain the optimum feed mix. The average composition range for traded lead concentrates is given in Table 3.3, which also shows the maximum limits commonly preferred for critical impurity elements. TABLE 3.3 Commercial lead concentrate specifications. Element

Normal range

Lead

55 - 75%

Zinc

3 - 15%

Iron

2 - 12%

Sulfur

14 - 25%

Silica Calcium Aluminium

Preferred maximum limits 10%

2 - 10% 0.05 - 1.5% 0.2%

Arsenic

0.02 - 0.5%

0.2%

Antimony

0.01 - 0.3%

0.2%

Silver Gold

100 - 2000 g/t 0 - 5 g/t

Copper

0.005 - 1.0%

Cadmium

0.005 - 0.20%

Nickel

5 - 50 g/t

Cobalt

0 - 20 g/t

Mercury

5 - 50 g/t

Selenium

0 - 30 g/t

Tin Bismuth

10 g/t

2 - 50 g/t 5 - 1000 g/t

10 g/t

Fluorine

5 - 500 g/t

100 g/t

Chlorine

10 - 1000 g/t

500 g/t

To form slags in the smelting process, fluxes need to be added; these generally consist of CaO (as limestone), SiO2 (as sand or ground quartz), and in some situations Fe2O3 (as ironstone). Certain dilution of concentrates by flux additions is necessary, particularly for sinter plant–blast furnace operations, since typical sinter composition is in the range of 45 - 50 per cent lead with defined ratios of FeO:SiO2:CaO. Generally the amounts of CaO and SiO2 in concentrate are below requirements for fluxing, hence the presence of these components in the concentrate is not of concern to the smelter unless they are in excessive amounts. Other gangue impurities such as MgO, Al2O3, Mn, Na and K will also report to the slag with minimal impact, and hence their presence in minor amounts is not of significance.



37


There are limits to the amount of zinc contained in the lead concentrate, as determined by the capacity of the slag to hold the zinc in solution. The limits of zinc in slag are around 18 per cent, and for a common slag to lead bullion ratio of 1:1 might suggest an upper limit of zinc in a 60 per cent lead concentrate of ten per cent. Higher levels may be accepted if blended down in smelter feed with concentrates containing lower levels of zinc. The presence of some zinc is beneficial in sinter–blast furnace operations due to its effect on the microstructure and quality of sinter, and at low levels economic recovery of zinc from slag by fuming will not be possible (see Chapter 8). A range of impurity elements need to be separated in the lead refining process. If the concentrate is relatively clean with few impurities, it may be possible to produce a primary bullion with minimal impurity separation procedures to produce an acceptable refined lead. If impurities are in excess of the minimal levels, then cost penalties can be applied to cover the additional operations and associated costs of removing those impurities. In particular, this applies to arsenic, antimony and bismuth. Alternatively, the material may be unacceptable or will only be accepted in limited amounts as part of a feed blend, depending on the capabilities of the particular smelter and associated refinery. For feed to the Imperial Smelting Process, there can be penalties on silica, as well as arsenic, bismuth and antimony.

COMMERCIAL TERMS FOR THE PURCHASE OF STANDARD LEAD CONCENTRATES The structure of the terms formula on which the price of lead concentrates is determined is essentially based on payment for 95 per cent of the lead content of the concentrate (subject to a minimum deduction of three percentage points) at the prevailing market price for refined lead, such as the London Metal Exchange (LME) settlement price, less a treatment charge, which is a fixed amount per tonne of concentrate. Examples of lead payments at different concentrate grades are given in Table 3.4. At lower grades the price formulation to some degree reflects that the losses in metal from the smelter are governed by the quantity of slag produced. TABLE 3.4 Lead concentrate metal payments. Concentrate lead content (%)

Lead in concentrate paid for (%)

Lead content paid for (%)

75

71.25

95

65

61.75

95

60

57

95

55

52

94.5

50

47

94

45

42

93.3

In addition to lead payments, silver and gold are also paid for above minimum levels. Silver payments are commonly made for 90 - 95 per cent of the silver content above 50 g/t, and gold for 85 per cent of the gold content above 1.0 - 1.5 g/t. The treatment charge is negotiable and can escalate with the lead price, but is typically of the order of US$250 per tonne of concentrates at an LME lead price of US$1000 per tonne. The escalation is typically 15 per cent of the increase in the lead price, but is a negotiable item. In times of surplus

38




concentrate supply the negotiated treatment charge will tend to be high, favouring the purchaser, whereas in times of concentrate shortage the treatment charge will tend to be low, favouring the seller. The treatment charge normally covers the production of refined lead and so includes charges for both the smelting and bullion refining operations. Refining can be separately charged and the treatment charge for refining is around US$150 per tonne of bullion, which represents about US$80 per tonne of concentrates or around one-third of the total treatment charge. In addition to the basic lead treatment charge, charges are levied for the refining of silver and gold and for the removal of key penalty elements. Silver refining charges are typically US$12 per kg of silver paid for and gold refining at US$200 per kg of gold paid for in the concentrate. Penalties for arsenic, antimony and bismuth are:

• arsenic

US$4 per 0.1 per cent of contained arsenic above 0.2 or 0.3 per cent,

• antimony

US$3 per 0.1 per cent of contained antimony above 0.2 or 0.3 per cent, and

• bismuth

US$1 per 0.01 per cent of contained bismuth above 0.05 per cent.

For a typical lead concentrate, the following gives a calculation of standard commercial terms:

• concentrate grade:

60 per cent lead (Pb) 1000 g/t silver (Ag) 3 g/t gold (Au) 0.5 per cent arsenic (As) 0.3 per cent antimony (Sb) 0.06 per cent bismuth (Bi)

• metal prices:

lead US$1100 per tonne silver US$10 per troy ounce gold US$600 per troy ounce

• metal payments:

lead 60 per cent × 95 per cent × US$1100 = US$627 per tonne silver (1000 - 50) × 90 per cent = 855 g/t = 27.49 troy oz/t = 27.49 × US$10 = US$274.9 per tonne gold (3 - 1) × 85 per cent = 1.7 g/t = 0.0546 troy oz/t = 0.0546 × US$600 = US$32.8 per tonne Total metal payments = US$934.7 per tonne of concentrates

• treatment charges:

lead TC = US$265 per tonne silver refining = 0.855 kg × US$12 = US$10.26/t gold refining = 0.0017 kg × US$200 = US$0.34/t Total charges = US$275.6 per tonne of concentrates

• penalties:

arsenic = (0. 5 - 0.3)/0.1 × US$4 = US$8/t antimony = (0.3 - 0.3) = 0 bismuth = (0.06 - 0.05)/0.01 × US$1 = US$1/t Total penalties = US$9 per tonne of concentrates Total value of concentrate = (934.7 - 275.6 - 9.0) = US$650.1 per tonne



39


The smelter primarily makes a return from the treatment charge, plus any free metals derived from recoveries achieved for lead and precious metals above the recoveries implied in the terms. For instance it may be possible to achieve 98 per cent lead recovery, giving three per cent free metal. Silver and gold recoveries may also significantly exceed the terms values, depending on process efficiencies. Smelter revenues are also boosted by an ability to recover and sell by-products such as sulfuric acid and copper, as well as some minor elements such as antimony in the form of antimonial lead alloys, mercury and cadmium. In some instances zinc can be recovered from smelter slags by fuming. Pricing of concentrates for delivery on a particular date is based on average daily LME metal prices over a defined period (‘The Quotational Period’) and usually is set as the month following the month of delivery. This is intended to provide a close match with the timing of consumption of the concentrate and delivery of product metal. Otherwise there can be serious discrepancies and variations in pricing of the metal contained in the concentrates used, and the pricing of the metal recovered and sold from those concentrates.

COMMERCIAL TERMS FOR THE PURCHASE OF BULK CONCENTRATES The Imperial Smelting Process produces both zinc and lead, and can blend standard zinc concentrates with lead concentrates or, preferably, can use mixed bulk concentrates. Since the terms for low grade and mixed bulk concentrates are more favourable to the smelter due to the greater percentage of ‘free metal’, it is in the interests of an ISF smelter to source as much of its raw material as possible as bulk concentrates. The metal output of the plant will be reduced as more lower grade materials are processed due to the increased formation of slag, the associated losses of zinc in that slag, and the fuel utilisation to heat and melt the slag. Hence there is a balance between lower cost raw materials and lower zinc production, which leads to an optimum grade of mixed feed. The best option is possibly the total use of bulk feed of relatively high grade in terms of total zinc plus lead content. The terms for mixed bulk concentrates represent a mix between standard zinc and lead concentrate terms and the following is a typical example:

• zinc payment for the balance of the zinc content after deduction of 7.5 units at the LME zinc price for SHG zinc metal,

• lead payment for the balance of the lead content after deduction of three units at the LME lead price, • silver payment for 90 per cent of the balance after deduction of 90 g per tonne of concentrate, and • gold payment for 60 per cent of the balance after deduction of 2 g per tonne of concentrate. Penalties and refining charges are similar to lead concentrate terms and often there can be additional penalties, such as for silica above a minimum level. This critically depends on the mix of feed stocks normally received by the smelter. The treatment charge is commonly US$10 to US$15 above the ruling treatment charge for standard zinc concentrates, at around US$220 per tonne, and with the price escalator three per cent lower than standard zinc concentrates at around nine per cent.

COMMERCIAL TERMS FOR THE SALE OF LEAD BULLION The production of refined lead is commonly broken into two discrete stages:

• the smelting stage, with sulfide concentrate feed producing a crude lead bullion; and • the refining stage, in which a range of impurities are removed from the bullion to produce refined lead.

40




These two stages may be within the one smelting site or may be two separate businesses, with crude lead bullion being the traded intermediary. Traded lead bullion is commonly derived from the treatment of primary concentrates, but can also be lead bullion produced from secondary sources. Secondary bullion generally contains much fewer impurities. As an example, ISF smelters do not usually include lead refining operations and consequently sell their crude lead bullion to a refinery. The commercial terms for the sale of lead bullion are typically as follows:

• lead payment

98 per cent of the lead content at the prevailing LME price for lead

• silver payment

98 per cent of the silver content with a minimum deduction of 50 g/t at the specified silver price

• gold payment

95 per cent of the gold content with a minimum deduction of 1 g/t at the specified gold price

• copper payment

80 per cent of the copper content with a minimum deduction of two per cent at the LME price for copper

• lead refining charge

US$150 per tonne of bullion treated

• silver refining charge US$12 per kg of silver paid for • gold refining charge

US$200 per kg of gold paid for

• bismuth penalty

US$15 per 0.1 per cent contained bismuth

Penalties will depend on the refinery and the process used. For instance, for the electrolytic refineries the cost structures for impurity removal and treatment are quite different to the pyrometallurgical refineries, where the costs of removal of bismuth, for instance, are quite high.

SECONDARY MATERIALS Metallic scrap is one significant source and can be purchased by the smelter or refinery at prices reflecting a nominal discount to the prevailing LME price for refined lead. However, the bulk of secondary lead is derived from the processing of recycled scrap lead-acid batteries. The trade is very localised with no general standard terms and the cost to the secondary smelter often simply reflects the cost of collection of scrap batteries. The other source of secondary lead can be low-grade residues such as oxides or sulfate leach residues, often sourced from electrolytic zinc plants as secondary leach residue and containing between 20 and 40 per cent lead. Often these residues contain substantial amounts of silver, which can economically justify treatment. Similar residues can be sourced from copper smelting operations, as well as miscellaneous dusts and fumes containing high levels of lead and precious metals. Many of these metallurgical fumes, such as copper converter dusts, can contain up to 30 per cent lead but also usually contain high levels of minor impurities such as bismuth, arsenic, antimony, cadmium, selenium and tin, which add significantly to the bullion refining load. Terms are similarly structured to sulfide concentrates, with the exception that treatment charges will be negotiable given the individual smelter’s capacity and ability to take additional secondary materials. Treatment charges are generally lower per tonne of material treated, but up to 50 per cent higher when expressed as per tonne of contained lead. There may also be additional penalties where high levels of particular impurities are present.



41


REFERENCES AND FURTHER READING Henley, K J, Radke, F and Tilyard, P A, 1998. Determining the deportment of minor trace elements in base metal concentrates with examples from Broken Hill, in Proceedings The Mining Cycle – AusIMM Annual Conference, pp 381-388 (The Australasian Institute of Mining and Metallurgy: Melbourne). Sutherland, K L and Wark, I W, 1955. Principles of Flotation (The Australasian Institute of Mining and Metallurgy: Melbourne). Wilson, P C and Chanroux, C, 1993. Lead, in Cost Estimation Handbook for the Australian Mining Industry, pp 346-348 (The Australasian Institute of Mining and Metallurgy: Melbourne).

42



PART B PRIMARY SMELTING This part of the text covers smelting processes used for the extraction of crude lead bullion from ores and mineral concentrates. Chapter 4 Chapter 5 Chapter 6 Chapter 7 Chapter 8 Chapter 9

Sintering The Blast Furnace The Imperial Smelting Furnace (ISF) Direct Smelting Processes Smelter By-Products and Treatment Processes Electrochemical Reduction Processes

CHAPTER 4 Sintering The purposes of sintering are:

• to roast lead sulfide concentrates so as to remove sulfur, and • to achieve sufficient temperature to cause partial melting in order to form a porous cake of sufficient cohesion and strength to be suitable as feed to the blast furnace.

PROCESS CHEMISTRY AND THERMODYNAMICS The primary roasting reaction required is shown in Equation 4.1: PbS + 1.5O2 = PbO + SO2

(4.1)

Other side reactions that can occur are shown in Equations 4.2 to 4.6: PbS + 2O2 = PbSO4

(4.2)

2PbS + 3.5O2 = PbO.PbSO4 + SO2

(4.3)

PbS + O2 = Pb + SO2

(4.4)

PbS + PbSO4 = Pb + 2SO2

(4.5)

PbS + 2PbO = 3Pb + SO 2

(4.6)

Relevant thermodynamic data for the above reactions are given in Table 4.1. TABLE 4.1 Thermodynamic data for principal reactions. Heat of reaction at 25 C (kJ/g mole PbS)

Heat of reaction at 1000 C (kJ/g mole PbS)

Gibbs Free Energy change at 25 C (kJ/g mole PbS)

Gibbs Free Energy change at 1000 C (kJ/g mole PbS)

4.1

-419.4

-390.8

-392.3

-318.6

4.2

-822.5

-831.6

-716.6

-369.1

4.3

-628.8

-613.4

-573.2

-358.9

4.4

-198.7

-183.7

-203.4

-220.8

4.5

+212.5

+232.0

+154.9

-72.6

4.6

+255.4

+245.3

+182.4

-41.0

Reaction number

Most reactions are highly exothermic and, with a highly negative Gibbs Free Energy change, have the potential to readily proceed. The ‘roast-reduction reactions’ given as Equations 4.5 and 4.6 are endothermic and have a negative Gibbs Free Energy change only at high temperature. Reaction equilibria for these reactions are controlled by the partial pressures of oxygen and sulfur dioxide, and can be simply explained by reference to the phase diagram for the Pb-S-O system as



45

CHAPTER 4 – Sintering

shown in Figure 4.1. This shows the equilibrium boundaries between various phases at two different temperatures of 1000 and 1200°C, in terms of the log10 of the equilibrium partial pressures (in atmospheres) of sulfur dioxide and oxygen. Since the system is at atmospheric pressure, the log (partial pressure) cannot exceed zero. To maximise PbO formation and ensure that sulfates are avoided, the temperature must be high and the SO2 levels low, particularly if excess oxygen is present. This is achieved in the sinter bed by sweeping generated SO2 away from the reaction zone with a high gas flow. If oxygen levels are low, (moving to the left of the diagram in Figure 4.1), there is an opportunity to form metallic lead. This effect is more probable at higher temperatures. PbSO4 0 PbO.PbSO 4

PbS

PbO.PbSO 4 -5

Log(SO 2 partial pressure - atm)

Pb

PbO

-10 1200 oC

1000 oC

-15 -15

-10

-5

0

Log(O 2 partial pressure - atm)

FIG 4.1 - Phase diagram PbO-S-O system.

Figure 4.1 assumes solid state conditions for the non-gaseous components. In reality some phases will be molten at these temperatures and will have mutual solubility, which will alter the activities from unity. In particular the solubility of sulfur both in lead metal and PbO will mean that elimination of sulfur is not as complete as implied by Figure 4.1.

THE SINTERING PROCESS Sintering was first developed as a batch pot roasting technique using a crucible fitted with a grid base through which air could be blown. Coal was first ignited on the base then sulfide materials were progressively added and burned until the crucible was full. It was then emptied and the process repeated. The Huntington-Heberlein Pot was a widely used form of the batch sintering method. This approach was soon mechanised with the development of the Dwight Lloyd sintering machine based on the principles of the chain grate stoker used for coal-fired boilers, as shown in Figure 4.2. It consisted of an endless looped chain of steel pallets, each made from a frame fitted with steel bars to form a grate. Material to be sintered was placed on the grate which then moved under a gas or oil fired

46




Feed Ignition Stove Feed Hopper Gas Air

Sinter Bed

Moving Grate

Product Sinter To Crusher Suction Boxes Gas Containing SO2

Fan

FIG 4.2 - Schematic of a Dwight Lloyd downdraught sintering machine.

ignition stove to heat and ignite the top layer. Air was drawn through the charge using a series of suction windboxes located beneath the pallet chain and sealed by side plates attached to each pallet. As the grate slowly moved forward, the combustion zone moved downward through the material until combustion was completed. At the end of the machine, air passing through the sintered material served to cool the sinter before it was discharged. The pallets moved over guides at the discharge end of the machine and returned beneath the windboxes to the feed end. The time required for sintering a bed of up to 400 mm deep is usually of the order of 30 minutes, which represents the speed of one pallet traverse from the position of ignition to the discharge end or ‘tip end’ of the machine. The process may be considered in terms of a number of zones moving through the charge bed as illustrated in Figure 4.3. Air

Ignition Stove

Wet Charge Zone

Charge Drying Gas Cooling Zone

Charge Heating to Ignition Zone

Reaction Zone

Air Heating Sinter Cooling Zone

FIG 4.3 - Separate zones in the sinter charge bed – downdraught conditions.



47


In the top zone, sinter already formed is cooled by incoming air and the air is heated. Hot air then passes through the reaction zone where combustion of the fuel occurs with some degree of melting. In the third zone hot combustion gases heat the unburned charge up to ignition temperature and the gases are progressively cooled and pass to the fourth zone where drying of the material takes place and gases are further cooled and humidified. As the reaction front moves through the sinter bed the temperature at any one point will rise to a maximum level and then decline as the bed is cooled by incoming air. The temperature profile of a point within the bed as a function of time (or position along the machine) is illustrated in Figure 4.4. The ‘peak bed temperature’ is the key parameter to follow and control. 1400

Temperature (degrees Celsius)

1200

1000

800

600

400

200

0 0

5

10

15

20

25

30

Time (minutes)

FIG 4.4 - Typical sinter bed temperature profile.

Gas permeability of the bed during the sintering process is a critical issue for maintaining productivity and sinter quality. Permeability can be reduced by the effect of moisture evaporation from the charge and recondensation in the colder downstream layers of the bed. More importantly, with dry feeds permeability can be reduced by entrainment of fine particles on drying causing blockage of the bed by collection and concentration of these fines in the downstream layers, aggravated by the collection of excess moisture. To avoid this some form of binding of fine concentrate particles to the return sinter is needed and often the addition of a small amount of CaO will serve this purpose. In addition, excessive melting and formation of liquid phases can also cause a reduction in bed permeability. However, a sinter feed mix which generates excessive amounts of low melting point phases is not likely to be of acceptable quality, and needs to be avoided in any event. Sinter quality is of key importance for efficient operation of the blast furnace, but is difficult to define and measure in quantitative terms, as the ultimate measure is essentially optimum performance of the blast furnace. Sinter must be strong enough to withstand handling without significant degradation and breakage. It should be reasonably porous to allow gas-solid reduction reactions to take place to the maximum extent, and it should have a suitably high melting temperature or softening range in order to

48




maintain its structure and integrity for the maximum possible time during the reduction process. Strength and a high softening temperature are most important and can be compromised by a structure which is too open and porous, even though this is also desirable for access of reduction gases. Clearly the reaction temperature and peak bed temperature as discussed above are important and key parameters for the formation of sinter with the required composition and properties. The uniformity of attainment of this condition throughout the sinter bed is critical. In order that this uniformity is achieved there must be constant conditions for the combustion reactions. This in turn requires close attention to the sulfur or fuel content of the charge, and hence, uniformity in the blending of charge components. Sizing and packing uniformity of the bed is also important to allow for even airflow and oxygen access. A number of operating parameters need to be closely controlled to achieve this, as discussed below.

THE STRUCTURE OF SINTER In order to perform in a blast furnace, sinter must retain its physical structure for as long as possible before melting and forming slag. Lead oxide reduction rates are highest from gas-solid reactions and lowest from the liquid slag. If the sinter softens and becomes a sticky mass it can restrict access of reduction gases to PbO surfaces and can, in extreme cases, plug the furnace and restrict the flow of both the charge and gas. The key parameter is therefore the melting temperature or the softening temperature range of the sinter. This is very much a function of the mineralogical composition and structure of the sinter. The structure may simplistically be considered as a matrix of crystalline phases cemented together by a glassy phase usually composed of lead silicate glass (CaO-PbO-SiO2). The crystalline phases are primarily the high melting point zinc containing minerals melilite (a complex mixture of calcium and aluminium silicates containing zinc, iron and magnesium) of which hardystonite is one particular form, and ferrites such as franklinite (ZnFe2O4), which contains the bulk of the trivalent iron and can also contain magnesium and aluminium as a replacement for zinc. This structure develops only at high temperature and it is important that peak bed temperature is above 1200°C for this to occur. Poorly performing sinters exhibit excessive lead silicate phases without the tight matrix of intergrown crystal phases to provide structural support. This allows the sinter to soften and slump as the temperature is raised. The relative amount and melting temperature of the glassy phase or phases are also critical in terms of determining the softening of the sinter structure. For instance, lead silicates have low melting points in the range of 600 to 800°C depending on composition. Melting points can be raised by the addition of CaO to increase the CaO:SiO2 ratio. The addition of CaO usually raises the temperature at which softening commences as well as reducing the temperature range between initial softening and the molten state. This effect is due to the decrease in the volume fraction of lead silicate phases with a wide softening range. For a typical sinter containing 46 per cent Pb, five per cent Zn, 11 per cent Fe, nine per cent CaO, and 11 per cent SiO2, the glassy matrix will represent close to 40 per cent of the volume of the sinter, the melilite phase is of the order of 30 per cent and the ferrites of the order of 20 per cent. The microstructure of sinter can be quite complex where low-grade feed concentrates are processed and there is a significant level of gangue minerals present. However, consideration of the details generally points to the need for high peak bed temperatures during sintering to promote crystallisation, and to the need for correct fluxing in order to achieve the appropriate matrix



49


composition as well as the ultimate process slag composition. In evaluating the needs for achieving a given slag composition with respect to SiO2 in particular, the contribution from the ash derived from coke used in the blast furnace must be taken into account. The residual level of sulfur in sinter is also important in achieving the most suitable sinter structure as well as meeting maximum limits for the blast furnace, and should generally be less than two per cent. Common levels are around 1.7 per cent but can be as low as 0.7 per cent.

PROCESS OPERATING PARAMETERS Sinter charge The basic composition of sinter is determined by the target composition of blast furnace slag and a lead content generally in the range of 45 to 50 per cent. Limits on the lead content of sinter at around 50 per cent are imposed to provide for sufficient slag fall within the blast furnace and to restrict the formation of metallic lead within the sintering process. Good quality sinter usually contains some metallic lead, but excessive amounts can result in drainage to the base of the sinter bed where it can freeze and cause blockages of the airflow. It can also drain into the grate and windboxes, again causing blockages as well as corrosion of the grate by PbO. This problem is more severe with downdraft rather than updraft machines. The sinter plant feed consists primarily of lead sulfide concentrates diluted with recycle sinter, flue dusts, reverts and fluxing materials. Fluxes consist of silica sand, lime or limestone, and haematite if additional iron units are required. The flux additions are determined by the target composition of blast furnace slag with preset ratios of various components such as CaO:SiO2, FeO:SiO2 and Fe:Zn ratios (details are given in Chapter 6 – Slag Composition). This determines the target composition of sinter and hence determines the feed additions required to the sinter plant. In addition it is common practice to also process secondary materials such as lead-rich leach residues arising from electrolytic zinc operations, reclaimed battery pastes and miscellaneous recycle fumes and metallurgical residues rich in lead and precious metals. Lead is predominantly present as lead sulfate in these materials, and they also commonly contain zinc and iron. Lead concentrates contain a range of impurity elements, primarily zinc and copper but also arsenic and antimony as well as variable amounts of silver and gold. The composition of blast furnace coke with respect to its lime, silica and alumina contents also needs to be taken into account when calculating the required sinter composition and corresponding flux additives to give the required blast furnace slag composition. Most sinter plants involve significant spillage from conveyors, from chute and equipment blockages and from equipment maintenance activities. The quantity of spillage is highly variable, but can be in the range of one to four per cent of sinter production. Spillage is collected into a storage bin from where it is returned at a controlled rate to the sinter plant charge, together with dusts and fumes collected from gas cleaning and from ventilation systems.

Fuel content of the sinter charge In practice, for the sintering of lead sulfide concentrates, a sulfide sulfur content of the charge of around six to seven per cent is required to achieve the necessary peak bed temperature. Since lead concentrates normally range from 15 per cent to 25 per cent sulfide sulfur they must either be diluted or partially roasted before sintering. Some dilution with return material occurs, but dilution with recycled sinter is the preferred approach because of the high degree of control over the sinter feed mix

50




particle sizing that can be achieved. Depending on concentrate grade, fluxing needs and other inert recycle materials added to the sinter charge, the ratio of recycle sinter to net product sinter will be between 1:1 and 3:1. The benefit of using crushed recycled sinter (termed ‘return sinter’) is to provide a relatively coarse inert particle, which acts to form an open permeable structure allowing good airflow and controlled energy release (from combustion) per unit volume, so as to prevent excessive fusion and collapse of the bed. Another key benefit of return sinter is that fine concentrates in the charge will coat onto the coarse sinter particles, improving bed permeability and thus providing maximum oxygen access for rapid combustion. Reduced levels of return sinter will cause a decrease in bed permeability, reducing possible gas flow and hence, sinter machine capacity. On the other hand gas permeability is enhanced by increased levels of return sinter, but fuel value of the charge is reduced and peak bed temperature is lowered. This leads to poorer quality sinter with increased fines, which report to ‘returns’ and reduce sinter bed permeability, thus reducing gas flow and machine capacity. The product sinter in this case is also weak and readily breaks up in the shaft of the blast furnace, reducing permeability and blast furnace throughput. Hence, there is an optimum level of recycle sinter and of fuel to maximise sinter machine capacity. This normally corresponds to a sulfide sulfur content of sinter feed of six to seven per cent and a minimum recycle ratio of 1:1 (recycle: net sinter production), or 50 per cent of total sinter production. If the minimum return sinter setting, or a heavy load of residues results in inadequate fuelling, then a fuel supplement in the form of coke fines can also be used. One kilogram of carbon is roughly equivalent to 2.2 kg of sulfide sulfur. Coke can be used in the case of a heavy charge of residues containing sulfates, which have highly endothermic decomposition reactions; it also assists by reducing the sulfates. However, it needs to be noted that coke tends to burn preferentially to lead sulfide and if fine, can burn too rapidly. It generally broadens the temperature profile and must be carefully sized to raise the peak temperature. Coke also promotes the formation of metallic lead within the product sinter and its use has been attributed to an increase in lead volatility, and therefore an increase in the production of fume in sinter gases. However, a partially prereduced sinter can bring benefits in the blast furnace smelting stage by raising the sinter softening temperature.

Charge mixing and moisture control Because of the high degree of compositional uniformity required, controlled proportioning and mixing of all feed components is critical. This includes weigh feeding and controlled blending of a number of concentrate feeds as well as fluxing materials, recycle fumes and residues, and return sinter. For this purpose the plant usually includes a feed blending and proportioning facility made up of a series of storage silos for each component, each fitted with weigh feeders and discharging onto a common collection conveyor which feeds the primary mixer. Another method used for blending of feed materials is the use of ‘bedding’ or the building of a heap made up of layers of component materials on the floor of a storage building. The material is then reclaimed by fully excavating the face of the heap, thus producing a blend in the same proportions as the heap was constructed. This method is inflexible in that the blend cannot be varied once the heap has been constructed, unless trimming additions of mix components can be made after reclamation. Many lead sinter plants incorporate secondary materials such as lead residues containing lead sulfate in their charge. The decomposition of sulfate is endothermic and can significantly affect peak bed temperature and the quality of the sinter produced. Consequently, there are limits to the amount of



51


residues and secondary materials added to the charge. Some compensation can be achieved by the addition of coke fines to the charge, but this can burn too quickly and must be carefully sized to provide the desired effect. As a general rule the total quantity of secondary materials should not exceed 25 per cent of the charge, but will depend on the total fuel value of the concentrates used as determined by the relative amounts of minerals such as galena and pyrite present in the concentrates. In order to have sufficient uniform flow of air through the sinter bed there needs to be a uniformly sized porous bed, with a reasonably coarse average particle size. Since concentrates are usually quite fine, the sizing of the charge will largely be determined by the sizing of the return sinter representing the bulk of the charge, and the degree of agglomeration achieved in feed preparation. The ideal situation is to have uniformly sized and relatively coarse particles of return sinter coated with a layer of fine concentrates, thus forming an open structure with the concentrates exposed to the airflow to the maximum extent. To meet this condition, return sinter is crushed using a combination of ribbed and flat crushing rolls to a particle size of less than 5 mm. Moisture is added to the charge to assist in binding the concentrate particles to the surface of the sinter. The total charge is mixed in a pug mill and tumbled in a drum or table mixer to achieve the coating action and some degree of pelletisation and conditioning. Disc pelletisers have been used in place of a drum, but generally the residence time is too short for efficient final mixing and product uniformity, especially in terms of size distribution and penetration of moisture evenly throughout the charge. Some water is added in the form of filter cake from returned dusts and fumes collected from gas cleaning and ventilation systems. The level of moisture is quite critical for the mixing and preparation operation, and is usually between six and seven per cent. Too much moisture will physically reduce porosity and permeability of the bed, whereas too little will not give the degree of uniformity of coating of the sinter particles and hence, uniformity of the particle sizing of the bed. Inadequate moisture addition will also cause a reduction in bed permeability. This is due to the lack of attachment of fines to the coarser particles and their tendency to move and block gas passage through the bed. Excessive moisture can also give rise to condensation in colder regions of the bed, resulting in blockage and reduced airflows. This effect can result from the addition of residues containing chemically bound water such as water of hydration or hydroxides and will place limits on the acceptable proportion of these materials in the charge mix. Measurement of the moisture content of sinter feed with feedback to direct water additions at various mixing points is thus an essential part of good sinter machine control. Various devices are used for online moisture measurement, including infrared devices and electrical conductivity probes. The latter are most common and can simply be two trailing electrodes in contact with the feed material on a conveyor belt. These devices are not highly reliable and need careful and regular calibration since they are affected by the presence of soluble salts, which can change the electrical conductivity of the free interstitial water in the sinter bed.

Machine feeding Even laying of the charge across the width of the sinter machine grate is critical in avoiding variable permeability and uneven airflow through the bed. A reciprocating conveyor or swing chute are commonly used to evenly fill an open hopper located above the grate. The gap between the grate and the base of the hopper sets the depth of the bed and can be adjusted using a sliding plate. Bed depths range from 200 to 500 mm. The width of the gap (or bed depth) in relation to the width of the hopper (or throat width) can also influence the packing density of the bed and needs to be carefully evaluated. It generally should be greater than 1:1. Variations in the depth of material held in the hopper can also impact to some degree

52




on the packing density and it is preferable to maintain the hopper at a constant level. This can be done by varying the machine speed for a given feed rate or by varying the feed rate to the hopper for a set machine speed. The latter control is probably preferable but is more complex since feedback control to all components of the feed preparation chain will be necessary. However, since this represents a fine balancing control it is more usual to set a feed rate and vary machine speed as a balancing control.

Gas distribution

SO2 concentration (%)

In general the combustion of most of the sulfur tends to take place towards the feed end of the sinter machine, and the concentration of SO2 in gas is much lower towards the ‘tip end’ of the machine. A typical gas concentration profile is shown in Figure 4.5 for a machine with ten windboxes. Average SO2 levels as shown are about half the maximum and need to be above five per cent for sulfuric acid production by the standard contact plant technology. 10 9 8 7 6 5 4 3

Point Average

2 1 0 0

2

4

6

8

10

Windbox No. FIG 4.5 - Typical sinter gas composition profile.

Total airflow through the sinter machine far exceeds the requirements for sulfur combustion, in part due to cooling requirements at the ‘tip end’ of the machine and leakage around the grate pallets. High air leakage will be accentuated by deeper beds and hence, windbox pressure. This will dilute SO2 levels in sinter plant gas and create difficulties with an associated sulfuric acid plant. If windbox pressure is reduced to reduce leakage, then airflow will be lower and the machine speed will need to be reduced, thus affecting machine capacity. Consequently, there are compromises between the operating variables of bed depth, machine speed, windbox pressure and gas composition to give optimum productivity. Since there is considerable excess oxygen in the gas stream it is common practice to recirculate gas from the tip end of the machine back into the front end in order to reduce the volume of sinter gas and raise the average SO2 concentration from 4.5 per cent to around six or seven per cent so that it is suitable for sulfuric acid production. Otherwise it has been the practice to collect only part of the gas from the feed end for acid production and discharge the tail end gas to atmosphere. However, that practice is less acceptable today, and tip end gas recirculation is now an important requirement for environmental control.



53


Ventilation and hygiene The handling of high lead oxide materials is hazardous and strict controls of lead particulates in air in the workplace is required. Consequently, in addition to combustion gases, there is a large volume of ventilation air for the purpose of collecting dust generated at many points within the sinter train. This covers conveyors and transfer points, and particularly the sinter size reduction, screening and handling equipment. Ventilation air is usually filtered in a bag house before discharge to atmosphere. Temperatures are generally not high at around 100 to 150°C and polyester bags are commonly used, fitted with reverse pulse cleaning mechanisms. Bag filter areas required for these duties are of the order of 30 m2 per 1000 m3/h of air filtered. Collected dusts are commonly pulped in water for ease of handling, and are filtered and the filter cake combined with dusts collected from sinter plant combustion gases and blast furnace fumes for return to the sinter feed mixing facility. The total load of collected dusts can be up to 2.5 per cent of total sinter production, and the volume of ventilation gases for the sintering operation can be of the order of 1500 Nm 3/h/m2 of sinter plant grate area.

Sinter handling Sinter breaks away and falls from the tip end of the machine in large slabs, which fall through breaker bars and a set of spiked rolls to break the sinter into lumps, which are separated using a 25 mm screen. Oversize represents product sinter suitable for feed to the blast furnace. Undersize is cooled through a drum or cooler conveyor and is then crushed using two sets of rolls in series. The final rolls are normally set at around 5 to 6 mm. The resulting crushed sinter constitutes return material and is sent via a surge bin to the feed mixing facility. Control of the gap in the final roll’s crusher is quite critical to the sizing of the return sinter and to the performance of the process and the production of quality sinter. These rolls tend to wear unevenly and give rise to a variable gap width and hence, a wider size distribution of crushed sinter than is desirable. Hence, regular maintenance and re-machining of the rolls’ surfaces is necessary. Coarse product sinter from the primary screen is sent directly to surge bins from where it is proportioned and mixed with lump coke to provide feed to the blast furnace. Provision is made to crush and return product sinter to make up the required recycle ratio if necessary. Under steady state operation of the sinter machine, control of the amount of product sinter returned is normally through the level of material in the returns sinter surge bin, which is drawn off at a steady rate to satisfy feed blend requirements. Severe imbalances and instability of the process can result from the production of poor quality sinter. Friable sinter of low strength gives excessive fines, which can exceed the requirements for recycle. In such a situation recycle sizing can be reduced, sending more fine material to product, which can have a detrimental effect on blast furnace operation. Finer sizing of the returns sinter can also affect sinter machine performance by restricting the permeability of the sinter bed, and can possibly perpetuate the problem. This situation can particularly arise on machine start-up and it is often preferable to discard poor quality material to stockpile for later slow reclamation as returns. The problem can also arise with departure from critical controls such as moisture content of the feed, sizing and feed blending controls, and can lead to process instability and difficulty in restoring product sinter quality. A disturbance will recycle with the returns cycle time, which can be measured in hours. It may dampen with time (requiring many hours), but in extreme cases can increase out of control, necessitating rejection from circuit to restore stable operation. In this regard the sintering process is inherently unstable and is a particularly difficult process with a high degree of intolerance to departure from stable operating conditions.

54




Sinter tends to degrade and disintegrate with time and hence stockpiling for any length of time can significantly reduce its quality with adverse effect on the blast furnace. This property of sinter varies greatly from one operation to another depending on the nature of the concentrates and other feed materials used. A high calcium content can aggravate this problem.

UPDRAUGHT SINTERING The initial development of sintering was based on the principles described above, wherein air passes down through the sinter bed to windboxes located beneath the sinter machine grate and under suction – so termed ‘downdraught sintering’. The downdraught action tends to compress the bed against the grate and reduce permeability as well as tending to lodge particles in the grate openings, causing blockages. It was also found that where significant levels of metallic lead are present in the sinter it tends to more readily drain down onto the grate where it can freeze, causing blockages, but more particularly causes severe corrosion of the grate bars. ‘Lead fall’ into the windboxes themselves could also be a major operational burden. In many cases sinter gases were also too low in SO2 content for sulfuric acid production and were discharged to atmosphere. In the downdraught operation any molten oxide phases formed tended to flow downwards into unreacted material and could dissolve some of the sulfides. Hence, it was difficult to reduce sulfur to low levels in product sinter and often grinding and re-roasting was necessary (Willis, 1980). These problems can be overcome to a large extent by blowing air upwards from pressurised windboxes and collecting sinter gas from a hood covering the machine. The first updraft sinter machine was developed at the Port Pirie smelter in the early 1950s (Burrow, Ridley and Adams, 1956). This concept significantly improved the productivity and efficiency of the sintering operation and is now the most commonly used design. A typical schematic of an updraft sinter machine is given in Figure 4.6. Ignition of the charge is still by downdraught on a shallow layer of around 30 mm laid on the grate ahead of the ignition stove. After ignition of that layer the full charge depth of 300 to 400 mm is placed on top of the ignited layer and airflow is reversed to updraught for the remainder of the machine. The combustion zone moves upwards through the charge and breaks through the upper surface at around 80 per cent of the machine length. Feeding the machine is then in two parts – one position for the ignition layer and one for the main layer. In some machines entirely separate feed systems are used; in others the feed is split to the ignition hopper and the main hopper using an intermittent diversion chute or conveyor. The disadvantage of the latter system is that the level in the hoppers varies and as indicated above can vary the packing density of the sinter charge. It also makes smooth control of the machine more difficult if controls are linked to the level of material in the main feed hopper. Machine speed is generally in the range of 1.0 to 1.5 m/min, and machines are generally wider at around 3 m to minimise the edge effects which result in air bypassing the bed. Initially updraught sinters performed poorly in the blast furnace due to relatively low softening temperatures, which was raised by restricting the combined lead content (as PbO) to less than 35 per cent (Grant and Cunningham, 1971). This was achieved by the addition of coke fines to reduce some of the combined lead to metallic lead. Such a practice was not considered for downdraught sintering because of the significant problems with ‘lead fall’ into the windboxes which already existed.



55


To Acid Plant

Gas Cleaning

Raw Material Hoppers Dust Separation

ESP Scrubber

Conditioning Drum

Return Sinter

Dust Collection and Treatment

Mixer Splitter Main Feed Hopper

Ignition Hopper and Stove

Hood

Air Ignition Gas to Gas Cleaning

Air Spiked Rolls

Tip-end Gas Recycle

Screen

Crushing Rolls

Cooler Crushing Rolls Product Sinter

Return Sinter

FIG 4.6 - Flow sheet of an updraught sintering machine.

SINTER MACHINE CAPACITY AND PERFORMANCE In general the sintering process involves a large number of variables and the formation of a material with a complex microstructure. It is a difficult process to control and is inherently unstable when subject to significant disturbances. Good performance thus tends to rely on a skilled and experienced workforce. The generation of poor quality sinter can have significant impact on the performance of the reduction processes which follow, particularly the blast furnace, and can greatly restrict its throughput. Sinter machine capacity is normally expressed in terms of the sulfur burning capacity and is in the range of 1 - 2 t/m2/d of grate area, with a typical figure of 1.8 t/m2/d. For a typical lead concentrate containing 60 per cent Pb and 20 per cent S producing a sinter containing 45 per cent Pb and 1.7 per cent S, this would correspond with a theoretical net sinter production of around 13.5 t/m2/d of grate area or 6.1 t of Pb throughput per m 2 per day.

56




Sulfur burning capacities are relatively uniform amongst different plants, but sinter or lead throughput varies significantly depending on the grade of the concentrate and the grade of the sinter produced. This is illustrated in Table 4.2 and shows considerable variation in machine capacity. Clearly capacity in terms of sinter output is significantly higher for primary lead smelters treating high-grade lead concentrates with relatively low sulfur contents in comparison with ISF plants. TABLE 4.2 Commercial sinter machine capacities. Plant (see Table 1.6)

Grate area (m2)

Sulfur burning rate 2 (t/m /d)

Sinter rate 2 (t/m /d)

Lead capacity 2 (t/m /d)

Primary lead smelters Belledune (Canada)

99

1.7

9.7

3.6

Chigirishima (Japan)

33

1.8

16.0

6.9

East Helena (USA)

83.3

1.2

10.1

2.6

Glover (USA)

70.8

1.1

13.2

5.2

Herculeneum (USA)

90

1.5

16.0

7.8

Hoboken (Belgium)

56

1.6

17.6

6.2

Mount Isa (Australia)

93

2.3

11.4

5.5

Port Pirie (Australia)

83.6

1.9

15.5

7.8

Torreon (Mexico)

160

1.3

9.4

3.1

Chanderiya (India)

120

1.6

6.5

Cockle Creek (Australia)

95

1.8

6.8

Hachinohe (Japan)

90

1.78

7.4

Harima (Japan)

70

1.63

7.8

Porto Vesme (Italy)

70

2.2

8.6

ISF smelters

Sinter machines vary largely in size, ranging from small machines of 12 m2 grate area up to 110 m2 with a grate width of up to 3 m. Details of a number of commercial machines are given in Table 4.3. Comparison between the operating performances of different sinter machines needs to recognise that the sinter plant is coupled with a blast furnace and the blast furnace could be the primary restriction, thus limiting the output of the sinter plant. It would appear to be a common feature that smelters processing high-grade lead concentrates are usually limited by blast furnace capacity rather than the sinter plant. The opposite is also the case, and smelters treating low-grade lead concentrates, or a high level of secondary materials are often restricted by sinter plant capacity. ISF smelters demonstrate that sinter production capacity per unit grate area is much lower than for primary lead smelters due to the much higher sulfur content of concentrates at around 32 per cent, compared with around 18 to 20 per cent for lead concentrates. Otherwise sulfur burning capacity, as the rate determining parameter, is much the same. The effect of lead concentrate grade on sinter machine capacity is most significant and may be illustrated in Table 4.4 for a basic sulfur burning capacity of 1.8 t/m2/d, a target sinter feed of 6.5 per cent S and a target sinter composition of 45 per cent Pb and 1.7 per cent residual S. In all cases the net sinter gas production at five per cent SO2 will be 1050 Nm3/m2 of sinter grate area. Because of the varying Pb:S ratio the volume of sinter gas per unit of lead will also increase markedly as the concentrate lead grade is reduced.



57

Plant

Belledune Binsfeld- Chigirishima hammer

Lead production (t/a)

108 000

92 000

East Helena

Glover

Herculeneum Hoboken Mount Isa Nordenham Port Pirie Torreon

90 000

70 000

120 000

225 000

125 000

156 000

102 300

215 000

180 000

Concentrates (t/a)

275 000

141 000

235 000

160 000

320 000

315 000

335 000

474 200

% Pb

45 to 72

62

to 75

76.5

72

40

51

65

43.5

%S

10 to 35

18

13 to 30

13.5

16.1

23

23.1

18

19.8

5.5

to 30

10.5

% Fe

2 to 25

Hearth length (m) × width (m)

3.0 × 33

1.5 × 25

2 × 15.2

Hearth area (m )

99

37.5

33

Bed depth (mm)

355

Speed (m/min)

1.1 to 1.4

2


Returns ratio

83.3

3

1.2

3.3

10

11.6

6.6

2.44 × 29

3 × 30

2 × 28

3.05 × 30.5

3.05 × 27

70.8

90

56

267

254

1.0 to 1.1

0.75 to 1.15

93

36

83.6

440

300

0.65 to 2.5

1.2

1.3


1

1.25

1.6 to 1.8

1.7

1.81

1.21

1.11

1.54

1.64

2.29

1.7

1.94

3.6

7.5

6.9

2.6

5.2

7.8

6.2

5.5

8.1

7.8

3.1

9.7

18.6

16.0

10.1

13.2

16.0

17.6

11.4

19.3

15.5

9.4

% Pb

43.5

42

41.9

28 to 36

45.1

30 to 40

44.1

43

48

41.5

%S

1.6

2

1.8

1.3 to 1.9

1.7

2 to 2.5

1.7

1.6 to 2.1

1.7

% Fe

17.5

17

11.9

10 to 18

15.5

6 to 12

10.2

10

11.5

11.3

(Nm /h)

120 000

91 000

21 000

59 000

230 000

15 800

80 000

45 000

75 000

91 700

% SO2

4.2

5 to 6.5

7

3.3

4

5.5

2.8

4 to 5.0

5.5

5

2

Sulfur capacity (tS/m /d) 2

Lead throughput (tPb/m /d) 2

Net sinter (t/m /d)

3.2

160

1.5 1.29

Sinter 1.3

Gas 3


58

TABLE 4.3 Data for commercial lead sinter plants.


TABLE 4.4 Effect of lead concentrate grade on sinter plant capacity. Lead concentrate grade (% Pb) Sulfur content (%)

50

60

65

70

75

23

20

18

16.5

15

Pb:S ratio

2.17

3.00

3.61

4.24

5.00

Sinter recycle ratio

3.10

2.11

1.67

1.28

1.07

Sinter production (t/m2/d)

9.4

13.5

16.6

20.8

24.5

Lead throughput (t/m2/d)

4.26

6.10

7.52

9.43

11.07

Sinter gas volume (Nm3/t Pb)

5910

4140

3350

2670

2270

Sulfur recovery to gas (%)

92.0

88.7

86.5

83.6

81.2

Theoretical capacities in terms of sinter production given in Table 4.4 are illustrated in Figure 4.7 together with actual plant data. The fact that actual sinter output can be higher than theoretical is due to the addition of oxidised lead materials, such as residues, to the sinter feed as diluent. Where capacity falls well short of theoretical, it is often due to the restrictive capacity of the blast furnace, which as indicated above is more common with the processing of higher grade concentrates. There are also variations in the sulfur burning rate of 1.8 t/m2/d due to characteristics of the sinter bed, such as porosity and reaction characteristics of the particular concentrates, which will cause significant variations. 30

Sinter production (t/m2/d)

25

20 Plant data

15

Calculated

10

5

0 30.00%

40.00%

50.00%

60.00%

70.00%

80.00%

% Pb in concentrates

FIG 4.7 - Calculated and actual sinter plant capacities’ dependence on lead concentrate grade.

The use of oxygen enrichment of sinter plant air feed has been evaluated as a means of increasing machine capacity in terms of the sulfur elimination rate. It is debatable whether there is any benefit, since excessive bed temperatures must be avoided, requiring a reduction in the fuel content of sinter feed and hence, lead sulfide feed. However, even if there were some benefit, oxygen is used in considerable excess and inefficiently for sulfur combustion. On this basis alone enrichment is generally not economically justified.



59


One means of increasing the lead throughput of a sinter machine is to use an oxidised lead feed material as a sinter returns replacement. For this purpose the material needs to be correctly sized and some form of agglomeration may be necessary, but it must be sufficiently strong to retain its form and size. Clinkered Waelz oxide is a suitable material for this purpose and is used in ISF sinter plants. The use of lead-rich slag produced by slag bath oxidation of lead concentrates, such as the Isasmelt process, has been proposed as a method of adding high lead inert material to sinter feed and increasing lead throughput. In general terms sinter machine installations are complex with many moving parts. The operating environment is hot and dusty and consequently can result in relatively high maintenance costs. There is a complex array of parameters to be balanced and an inherent tendency towards instability with a high reliance for optimum performance on the use of experienced operators. The high level of solids recycle, and hence, the total material handling loads per unit of output, significantly adds to the costs of the process. For these reasons the sinter plant–blast furnace combination is no longer cost competitive in comparison with direct smelting processes for new greenfields installations.

GAS HANDLING AND CLEANING The net gas output from the sinter machine is at a temperature of 200 to 500°C and contains a significant load of dusts, fumes and volatile materials, including compounds of lead, cadmium, mercury, chlorine and fluorine. The SO2 content can be variable from 2.5 per cent to 6.5 per cent, depending on measures taken to raise its concentration such as tip end gas recirculation. Gas may also be separated from the front end and tip end, giving relatively high SO2 concentrations from the front end and low concentrations from the tip end. For the production of sulfuric acid by the contact process gas strength should be above five per cent and preferably closer to seven per cent. This is clearly difficult from lead sinter plants with a single gas pass and either only front gas is used or tip end gas recirculation is applied. Where only front end gas is used for sulfuric acid production, overall sulfur recovery to acid is of the order of 60 per cent; the remainder is discharged to atmosphere. Gas that is not processed in an acid plant needs to be cooled to below 200°C using an open humidification tower before being cleaned using bag filters to remove dusts and fumes before discharge to atmosphere. In this process it is important to ensure that the gas temperature is not lowered below its dewpoint as this can lead to serious corrosion problems. Although total or partial discharge of sinter gas was commonly practiced in the past it is becoming less acceptable. For acid production sinter gas is usually cleaned by drop out of dust in an open chamber or flue followed by hot electrostatic precipitators of the plate and wire type with two or three fields in series. This is followed by wet gas scrubbing and wet electrostatic precipitators to remove mist from the scrubbing operation. A mercury removal scrubber may be included, if mercury levels in the clean gas are high enough to give product acid above 1 ppm mercury. Mercury removal is not always necessary as it is with zinc concentrate roasting. Total collected dusts generally amount to between one and two per cent of the input sinter charge and contain relatively high levels of lead, zinc, arsenic, antimony and cadmium, and sometimes other volatile elements such as selenium and mercury. Collected dusts are usually directly recycled to the sinter plant feed as a filter cake, but can be separately treated by leaching prior to recirculation in order to remove highly volatile impurities such as cadmium and prevent high circulating loads developing. Many devices are used for wet gas scrubbing such as spray towers, packed towers, plate scrubbers such as the Peabody scrubber or Venturi scrubber. Scrubbing liquor is circulated through the scrubbing system and becomes acidic due to the absorption of SO3 contained in sinter plant gas. Often the

60




scrubbing system consists of two stages, with the first as a quench or humidifying stage to reduce gas temperature from 300°C to less than 100°C, followed by a second scrubbing–cooling stage, which lowers the temperature to less than 38°C. The final temperature is important since the gas will be saturated with water vapour and the temperature determines the amount of water going forward to acid production, which must be less than that required to combine with the SO3 formed. Hence, there is a direct relationship between the SO2 content of the gas and the maximum temperature of that gas leaving the scrubbing system, as shown in Table 4.5. TABLE 4.5 Temperature of wet gas as a function of SO2 content. SO2 content of sinter gas (%)

Maximum temperature of wet gas to acid production (°C)

8

40.1

7

38.1

6

35.2

5

32.2

4

28.6

The ‘weak acid’ scrubbing liquor containing sulfuric acid as well as chlorides and fluorides can be very corrosive and appropriate materials of construction must be used. One particular area of high corrosion potential is the boundary region where hot gas first meets scrubbing liquor. This is usually a troublesome and high maintenance area for the containment equipment. In general these problems appear to be less severe in the case of lead sintering than for zinc sintering or roasting operations. The weak acid scrubbing liquor is bled from the system to maintain either a preset maximum acidity or a maximum fluorine concentration. Otherwise the bleed will merely be set by the water balance control. This solution also contains lead, zinc and cadmium dissolved from sinter dusts as well as mercury and selenium. Mercury and selenium react to form mercuric selenide, which precipitates and can foul heat transfer surfaces. The presence of selenium can be beneficial in increasing the capture of mercury by the gas scrubbing system. In some instances the scrubbing system includes a final stage of sacrificial silica packing to ensure full capture of fluorine and to protect the acid conversion catalyst. Following scrubbing and cooling, suspended droplets of scrubbing solution or mist are removed in wet electrostatic precipitators. The cleaned gas may still contain unacceptable levels of mercury vapour, which must be removed before transfer of gas to the acid conversion plant. There are a number of mercury removal techniques but the most commonly used is the Boliden–Norzink mercury removal process, which uses mercuric chloride solution to scrub the gas and absorb mercury according to Equation 4.7 (Dyvik, 1985): Hgo + HgCl2 = 2 HgCl

(4.7)

Mercurous chloride or calomel is insoluble and precipitates from solution. It can be separated in a thickener and removed from the system. Part of the calomel is treated with chlorine to regenerate the mercuric chloride scrubbing solution. The system is capable of achieving less than 0.5 ppm mercury in product acid or lower if reduced gas temperatures are used. It is important for efficient operation that the gas coming forward from the electrostatic precipitators is free from suspended mist and is classed as ‘optically clear’, otherwise contamination of the mercury scrubbing circuit will quickly destroy mercury removal efficiency.



61


Following mercury removal the gas is suitable as feed to a conventional wet gas sulfuric acid plant, of which there are many designs and configurations. This clean gas should be capable of production of high-quality sulfuric acid. The ‘weak acid’ or scrubbing solution bleed represents a disposal issue and is usually sent to an effluent treatment facility where the acid is neutralised with lime to form gypsum and the dissolved metals are precipitated. The filter cake can be returned to the sinter plant feed where the CaO content will be useful as a flux addition. The filtrate solution containing chlorides can be discarded, provided it meets relevant environmental standards. After cleaning by scrubbing, the gas may be passed to a standard contact sulfuric plant.

SULFURIC ACID PRODUCTION Sulfur dioxide in cleaned sinter gas is oxidised to sulfur trioxide at temperatures of around 400 to 500°C over a vanadium pentoxide catalyst. Resulting SO3 is absorbed in strong sulfuric acid and reacts with contained water to form H2SO4. The first stage of the acid plant involves drying the gas and any added air in a packed tower (drying tower), where it is contacted with 93 to 96 per cent sulfuric acid at a temperature of around 65°C. Acid is circulated between the drying and absorption towers to maintain the acid strength in the drying tower and feed water into the absorption tower circuit. This provides most of the water required for the formation of sulfuric acid by reaction with SO3. Hence, there are limits on the moisture content of input gas as detailed in Table 4.5. Dry cold gas usually then passes through the gas blower, is heated to 420°C and passes to the first catalyst pass. The oxidation of SO2 to SO3 is exothermic and gas exiting the first pass may be of the order of 600°C. Heat is extracted by heat exchange with incoming cold gas before transfer to the second catalyst pass. There are many different arrangements of heat exchange between incoming cold gas and the various gas streams to and from the catalyst beds, dependent on the particular plant design. At least three, but more commonly four catalyst beds are used, and it is generally aimed to operate at temperatures between 420 and 450°C. Following the third catalyst pass, gas is treated in the interpass absorption tower to absorb SO3. Exit gas from the interpass tower is reheated by heat exchange and passes to the fourth catalyst bed for conversion of residual SO2, and then to the final absorption tower. With this double absorption system, SO2 conversion can exceed 99.5 per cent. A flow sheet of a typical sulfuric acid installation is given in Figure 4.8. Single absorption is also used but achieves lower conversion at 98.0 to 98.5 per cent, and emissions of SO2 in tail gas are much higher (five times). Unless tail gas scrubbing is applied, single absorption plants will generally not meet normal environmental emission standards. Controlled water additions are made to the final absorption circuit to control product and circulating acid strength at 98.0 to 98.5 per cent H2SO4. Insufficient addition will result in oleum formation with high corrosion potential, as well as the potential for fume generation. Acid circulating over the absorption and drying towers must be cooled by heat exchange with cooling water and represents a significant thermal load of low-grade heat. Plate exchangers are most commonly used for this duty but shell and tube exchangers are also common. It is important to closely control the temperature of circulating acid for optimum performance of the towers and to prevent the generation of fine mist or SO3 fume, which can carry over into tail gas. Absorption tower acid is generally controlled within the range of 80 to 90°C and high efficiency candle mist eliminators are used to minimise carryover.

62




1st Pass Heat Exchangers 3&4

Heat Exchangers 1&2

Blower

Converter

Tail Gas

Air

Air Clean Gas Input Acid Pump Tank

Cooler Acid Pump Tank

Cooler

Water Additions Product Acid Drying Tower

Stripping Tower

Interpass Absorber

Final Absorber

FIG 4.8 - Typical sulfuric acid plant flow sheet.

Product acid is drawn from the final absorption tower circuit and is cooled to 40°C before storage in mild steel tanks. Sulfuric acid produced from sinter gas is black in colour due to the presence of fine carbon. This is derived from organic compounds such as flotation reagents present in lead concentrates, which distil into the sinter gas on initial heating of the sinter charge. These compounds are captured in drying tower acid and are decomposed by the strong acid to form finely divided carbon. Because of this the acid has limited application – such as use for fertiliser production; otherwise it can be treated with hydrogen peroxide to remove the coloration and form clean ‘white’ acid for more general application.



63


REFERENCES AND FURTHER READING Burrow, W R, Ridley, K L and Adams, F C, 1956. Up-draught sintering at Port Pirie, The AusIMM Proceedings, 180:179-206. Dyvik, F, 1985. Application of the Boliden Norzink mercury removal process to sulfuric acid production, in Proceedings Extractive Metallurgy ’85 Symposium, p 189 (Institution of Mining and Metallurgy: London). Grant, R M and Cunningham, B C, 1971. The relationship between sintering practice and lead blast furnace performance at Port Pirie, paper A7 1-1 (The Minerals, Metals and Materials Society – American Institute of Mining, Metallurgical and Petroleum Engineers: Warrendale). Siegmund, A H J, 2000, Primary lead production – A survey of existing smelters and refineries, in Proceedings Lead-Zinc 2000, pp 55-116 (The Minerals, Metals and Materials Society: Warrendale). The AusIMM, 1958. Sintering Symposium (The Australasian Institute of Mining and Metallurgy: Melbourne). Willis, M, 1980. The physical chemistry of lead extraction, in Proceedings Lead-Zinc-Tin ’80 Symposium, pp 437-476 (The Minerals, Metals and Materials Society – American Institute of Mining, Metallurgical and Petroleum Engineers: Warrendale).

64



CHAPTER 5 The Blast Furnace INTRODUCTION The blast furnace consists of a vertical shaft in which lead oxide contained in sinter is reduced to lead metal by a counter-current flow of gas rich in carbon monoxide. Elevated temperatures are required and the necessary heat and CO are generated by the combustion of coke with an air blast at the base of the shaft. Coke is used to provide a non-fusible support for the charge and a free open structure to facilitate uniform gas flow up through the shaft. Reduction of free PbO in the sinter by gaseous CO occurs from the solid state in the upper shaft. The sinter melts in the lower part of the shaft and the bulk of the contained lead, in the form of glassy silicates, is reduced to lead metal as it flows down over the hot coke bed. Further reduction can take place in the slag pool contained in the hearth of the furnace. The combustion zone at the base of the shaft is intense and is generated by an air blast through a set of nozzles or tuyeres, impinging on the remaining coke charge. CO2 is generated and then reacts with hot coke to form CO. Molten lead and slag collect in the furnace hearth with lead forming a lower layer, which can be tapped either separately or with slag in a continuous tapper. With continuous tapping a forehearth can then be used to separate slag and lead bullion. The capacity of the blast furnace may be simply determined by the ability to burn carbon, which in turn relates to the oxygen supply or blast rate. The latter is controlled by the permeability of the shaft charge and hence depends critically on the nature and structure of the sinter and coke feed as well as the presence of accretions, which can reduce the cross-sectional area of the shaft. Comparative capacity can be defined by the carbon burning rate per unit of shaft cross-section. The carbon is consumed to supply heat and CO, which in turn is consumed to reduce PbO, as well as Fe2O3, CuO, ZnO and other minor metal oxides. Since most sinters have a reasonably consistent composition of around 40 to 45 per cent Pb, it follows that the sinter treatment rate and bullion production rate per unit of shaft cross-section also give meaningful comparative measures. The ratio of CO:CO2 in the exit gas from the shaft is also an important parameter to be considered in determining operational efficiency and is usually of the order of 0.4 for an open top furnace, but can vary considerably. Under hot top conditions, which may represent excessive coke in the charge, the ratio can increase to 1.0. The net heat generated in the furnace from a given quantity of coke is greatly reduced as the CO:CO2 ratio increases. However, the distribution of that heat generation between the lower and upper sections of the shaft is of considerable importance.

CHEMICAL PRINCIPLES AND THERMODYNAMICS The basic reactions taking place in the furnace shaft are given in Equations 5.1 to 5.11: C + O2 = CO2

(5.1)

C + CO2 = 2CO

(5.2)

PbO + CO = Pbl + CO2

(5.3)

3Fe2O3 + CO = 2Fe3O4 + CO2

(5.4)



65

CHAPTER 5 – The Blast Furnace

Fe3O4 + CO = 3FeO + CO2

(5.5)

FeO + CO = Fe + CO2

(5.6)

ZnO + CO = Znv + CO2

(5.7)

Znv + PbO = Pb + ZnO

(5.8)

Znv + PbS = ZnS + Pb

(5.9)

PbSO4 + CO = PbO + SO2 + CO2

(5.10)

PbS + 2PbO = 3Pbl + SO2

(5.11)

The thermodynamic parameters for each reaction are given in Table 5.1. TABLE 5.1 Thermodynamic factors for shaft reactions. Reaction

Heat of reaction at 25°C (kJ/gmole)

Heat of reaction at 1100°C (kJ/gmole)

Free energy at 25°C (kJ/gmole)

Free energy at 1100°C (kJ/gmole)

5.1

-393.51

-392.48

-394.38

-403.21

5.2

172.43

169.54

119.83

-78.03

5.3

-58.57

-71.05

-68.39

-79.36

5.4

-50.75

-40.98

-62.53

-114.85

5.5

23.05

-17.55

24.05

-0.76

5.6

-12.60

-14.04

-12.75

-9.48

5.7

195.85

183.27

156.02

26.24

5.8

-254.42

-254.32

-224.41

-105.60

5.9

-232.65

-221.73

-193.74

-71.21

5.10

121.0

169.38

67.19

-143.18

5.11

255.93

246.84

182.37

-63.57

It is of note that except for ZnO, the principal reduction reactions of metal oxides with CO are exothermic at operating temperature. However, the reaction of CO2 with coke regenerating CO for reduction is highly endothermic, as well as the decomposition of lead sulfate (Equation 5.10) and the ‘roast reaction’ as given by Equation 5.11. The equilibrium conditions for each reaction may be expressed in terms of the partial pressure ratios of CO to CO2 and are illustrated in Figure 5.1. As a simplification this assumes solid–gas reactions with unit activity of the solid oxide reactants. If in the molten phase, such as PbO dissolved in slag, the activity will be much lower and the equilibrium ratios for CO:CO2 will be correspondingly higher. For partial pressure ratios of one or above, covering the bulk of the reactions zones of the shaft, Figure 5.1 indicates that PbO reduction should proceed readily, ZnO can be reduced to zinc vapour above 800°C, and iron oxides will be reduced primarily to FeO. In Figure 5.1 the ZnO reduction equilibrium is shown for a zinc vapour partial pressure of 0.01 atmospheres or one per cent in the gas stream. Zinc partial pressure will vary widely; however, this serves only to illustrate that zinc vapour will be present at partial pressures of this order.

66




6

4

Log(CO:CO2)

2

PbO C Fe2O3

0

Fe3O4 FeO

-2

ZnO

-4

-6 600

700

800

900

1000

1100

1200

1300

Temperature °C

FIG 5.1 - Equilibrium CO:CO2 ratios for reduction reactions (ZnO represents a Znv partial pressure of 0.01 atm).

FURNACE PERFORMANCE The internal workings of the blast furnace are complex, with a range of reactions as above, often with non-uniform gas flow due to charge variability and to accretion formation. The influence of volatile components such as zinc metal and lead sulfide is significant and they can cause significant shifts in heat generation from the base to higher levels in the shaft. Feed composition and furnace configuration can have a major influence on performance and no two furnaces will operate in exactly the same way. This makes operation more an art than a science, with high reliance on the experience of the operators. It is also difficult to generalise on the performance characteristics and the following description must be taken as one possible explanation of the mode of operation of the lead blast furnace. There have been a number of attempts to model the lead blast furnace, notably Lumsden (1971), Madelin, Sanchez and Rist (1990), as well as descriptions of the process chemistry by Willis (1980) and Oldwright and Miller (1936).

Reaction zones In simple terms, the furnace shaft may be divided into four zones.

Zone 1 – tuyere zone The tuyere zone is from the hearth upwards to a point just above the tuyeres. This is the high temperature zone in which carbon is oxidised by the air blast to CO 2 at temperatures above 1500°C. Generally the lead blast furnace does not operate at high blast pressures compared with the iron blast furnace or the ISF, and ‘raceways’ are not an important feature of the tuyere zone. Consequently there can be differences in the oxygen content of gas across the furnace, with the chance of some gas with relatively high oxygen levels being deflected by the charge and passing upwards in close proximity to the wall. Towards the centre oxygen will be depleted and CO levels will tend to increase and the temperature fall.



67


Liquid phases pass down through this zone over the hot coke bed and into the slag bath. Passage of molten lead over the coke bed should counter any potential oxidation on passage through the flame region. Overall the gases leaving this zone have an oxygen potential (or CO:CO2 ratio) in thermodynamic equilibrium with the slag and bullion formed and at the tapping temperature. There will be local variations from tuyere to tuyere along the furnace and across its width, which may result in departure from average equilibrium conditions. However, in general the outcome in terms of slag composition and metal balance can be controlled by the carbon to air mass ratio. The height of Zone 1 has been indicated by Madelin, Sanchez and Rist (1990) at 400 mm above the tuyere level, but is likely to vary significantly and this should only be taken as an order of magnitude guide.

Zone 2 In this zone, immediately above the tuyere level, reduction reactions predominate, particularly the reaction of CO2 with coke to form CO at temperatures above 1000°C. Sinter commences to melt in this zone and lead oxide in the form of glassy lead silicates is reduced by CO to metallic lead. Zinc oxide is also reduced to form zinc metal vapour, which rises with the gas stream. Any lead sulfide present can be volatilised and may react with zinc vapour in colder adjacent regions such as the furnace walls or centre of the furnace, to form zinc sulfide and lead metal. This can give rise to the formation of accretions. The reaction of CO2 with coke to form CO is endothermic and will reduce the temperature in this zone, but will be counteracted to some extent by the reduction of lead oxides, which is exothermic.

Zone 3 Above Zone 2 the charge is essentially in the solid state and below 900°C. Heat transfer from rising gases to the descending charge occurs within this zone, and free PbO within the sinter is reduced to lead metal by both CO and by zinc vapour in accordance with Equation 5.8. Reaction of zinc vapour with CO2 to form ZnO occurs as the temperature falls below 800°C and generates heat. This tends to form accretions on the colder furnace walls. However, zinc oxide formed in the upper part of the furnace will descend with the charge and will again be reduced to zinc vapour at lower levels in the shaft, thus forming a circulating load within the charge. This effect reduces temperatures in the lower shaft by the endothermic reduction reaction and raises temperatures in the upper shaft by the exothermic oxidation reaction. Although the furnace input of zinc can be low, the development of this circulating load can mean that zinc still has a significant bearing on the performance of the furnace in relation to the distribution of heat within the shaft.

Zone 4 The upper zone of the shaft essentially serves to transfer heat from the rising gas to the descending charge with minimal reaction, and to remove any moisture contained in the coke. This zone can vary widely from one furnace to another and exit gas temperature can vary upwards from around 200°C. In general the gas flow is non-uniform. It is higher close to the furnace walls and lower in the centre of the shaft. Gases close to the wall can still contain oxygen, particularly at lower blast rates and consequently reduction is higher in the centre and temperatures are lower. This can lead to the

68




formation of a rigid core of material in the centre of the furnace. The effect can be more pronounced at lower blast rates where more gas channels up the walls and gives rise to higher temperatures further up the shaft, which in turn can cause melting and aggravate channelling. Gas composition at the exit of the shaft is a key indicator of the overall mass balance of the blast furnace and can be used as a controlling parameter. Normally this is a CO:CO2 ratio of close to 0.4:1. The gas composition profile will show a low CO:CO2 ratio in Zone 1 at the tuyere level rising in Zone 2 as CO2 reacts with hot coke and falling as CO is used for reduction in Zones 2 and 3. This basic picture will be affected by the volatilisation of zinc and its reaction with CO2 as the temperature falls to regenerate CO. Bearing in mind that a circulating load of zinc can build up between Zones 1 and 3, this can have a significant effect on the gas composition profile and also on the zonal heat balance. Bypassing of air up the walls of the furnace or through channels can result in the presence of some oxygen in exit gas. This can be up to four per cent, suggesting utilisation of only 80 per cent of the oxygen in blast air. A schematic of a vertical cross-section through a typical blast furnace illustrating the above zones and characteristics is shown in Figure 5.2. Charge feed

Gas offtake

Furnace charge Zone 4 Wall accretions Zone 3

Zone 2

Central accretion or ‘deadman’ Tuyeres

Zone 1

Blast air Molten slag Molten lead

Hearth

FIG 5.2 - Typical vertical cross-section of a lead blast furnace.

The above picture of the performance of the furnace is idealised and is complicated by non-uniformity across the shaft section, giving rise to significant gas compositional changes from the walls to the centre of the shaft with higher oxidation potential at the walls. Conditions at each tuyere



69


are also likely to vary in terms of access to coke and molten slag, as well as variations in gas flow. Because of these multi-dimensional variations and the random occurrence of accretions, which profoundly affect gas flow patterns in the shaft, it has been particularly difficult to adequately model the operation of the blast furnace. Performance of a furnace is best with a new clean furnace and charge, without accretions and when gas flow patterns are reasonably uniform. Performance generally deteriorates from the start and the maintenance of a reasonable ongoing performance is the art of an experienced operating crew.

Accretions As illustrated in Figure 5.2 accretions are a key feature of the lead blast furnace and must be adequately managed to maintain operation. Accretion formation is complex and as indicated above, results from reactions of volatile components such as zinc and lead sulfide as well as fusion of the charge and solidification in cooler regions of the furnace. Analysis has been reported by Oldwright and Miller (1936), Ruddle (1957) and Polyvyannyi et al (1971). Wall accretions can be of two types, those near the hearth comprising essentially of frozen slag forming materials, and those higher up consisting of charge material bonded by condensed volatile material or reaction products therefrom, such as zinc oxide lead sulfide and zinc sulfide. This latter form of accretion can be sintered by an increase in temperature brought about by a rise in the level of high temperature zone boundaries. Since these high temperature zones can rise and fall with furnace variations, the accretions can exhibit a layered structure with varying degrees of sintering. Lead present in wall accretions is largely as metal derived from the reaction of PbS with zinc. The sulfide content of wall accretions tends to increase at lower levels of the furnace. Hearth accretions are usually sulfide rich, extremely hard and refractory and can be a major cause for shutdown for cleaning when they interfere with the flow of bullion or slag to the tapper. The central accretion or ‘dead man’ or ‘sow’ is comprised of loosely bonded semiplastic fine charge material formed as a result of attrition of larger material, displaced to the centre by the main gas flow from the tuyeres, and bonded in much the same way as wall accretions. Since the ‘dead man’ is not removed and normally acquires a reasonably consistent shape and size, and further since it does not normally impede the flow of lead bullion to the tapper, it is likely that the binding material has a melting point below the temperature of the slag bath, and that the base of the column of the central accretion is a reasonable open bed of coke.

Coke consumption reactions and heat balance There must be careful balancing of operating parameters to achieve the necessary temperatures and heat balance in various sections of the furnace shaft. Zone 1 supplies the heat and CO to drive the furnace and must receive the required amount of coke to achieve the required temperature and to generate reduction gases at the base of the shaft. For a given sinter feed and blast rate, the operating parameters relating to fuel which affect furnace performance are:

• coke supply rate, • coke reactivity, and • coke sizing.

70




Coke supply rate If the coke rate is increased by increasing the relative proportion of coke in the furnace feed without change to the blast air, the CO:CO2 ratio in smelter gas will rise, which means that less heat is developed and the bottom temperature falls. In the extreme this has the potential to cause a freeze and virtually stop the operation of the furnace and the phenomena is known as ‘over-coking’. An overall mass balance of the furnace indicates that a five per cent change in the coke rate will change the CO:CO2 ratio in exit gas from 0.4:1 to 0.58:1 or by almost 50 per cent. This provides a very sensitive indicator of the furnace condition. If the furnace is ‘under-coked’ with the coke supply reduced to give a low CO:CO2 ratio of say 0.2, then the slag temperature will rise and can theoretically exceed 1300°C. The oxidation conditions would increase and the loss of lead in slag would rise. Clearly, in practice this situation could be rapidly controlled by reducing blast air, but this change will reduce lead production. However, an increase in coke rate will also correct the situation without reducing lead production, which tends to be counter-intuitive as a means of reducing bottom temperature. Factors that affect the supply of coke to the oxidation zone will have a most significant impact on the performance of the furnace, and are critical for adequate control of the furnace and for achievement of optimum performance. It should also be noted that vapour phase reactions can shift heat generation from the base to higher levels and reduce the amount of coke reaching lower levels of the furnace. This will arise with high levels of zinc or sulfur in the furnace charge.

Coke reactivity Since all the coke is consumed the overall heat generation and mass balance will not change due to coke reactivity. A more reactive coke will promote reaction with CO2 to form CO in Zone 2. This reaction is endothermic and will tend to lower the temperature at this point. The lower temperature will promote the oxidation of zinc vapour by CO2, which is exothermic and will to some extent compensate the effect of coke reactivity by heating the solids entering Zone 2. In the extreme this could lead to partial melting of solids and then refreezing in the lower zone due to the formation of CO (Oldwright and Miller 1936; Hopkins and Haney, 1954). It is unlikely that coke reactivity will affect combustion rates and hence combustion efficiency in Zone 1 where high temperatures are reached. Consequently, reactivity will only affect the relative position of Zone 2 and is likely to have only minor impact on the performance of the furnace.

Coke size The prime role of coke as a fuel and reductant as distinct from other fuels is to provide structure and support for the charge, to allow uniform flow of gas and to separate lumps of sinter to prevent charge fusion in the plastic temperature range. For this purpose the coke needs to be of similar size to the sinter and is preferably of a uniform sizing to maintain the maximum voidage. With sinter at +50 -75 mm, coke should be similarly sized, although it is likely there will be some finer material present from handling and attrition. For a given coke supply the size of the coke will affect reactivity in the furnace. Coarser material will react less in the upper shaft and get through to the oxidation zone. In some operations there is a practice of feeding both coarse and fine coke and to vary the proportions to allow consumption of the fines in the upper zones, enhancing the retention of the size of the coarser coke into the lower levels of the furnace. This provides an additional degree of control for optimisation of the furnace, but it is generally best to minimise excessive amounts of both very coarse and very fine coke to achieve a relatively narrow size distribution.



71


SLAG CHARACTERISTICS AND COMPOSITION Lead blast furnace slags are basically represented by the CaO:FeO:SiO2 system, but with significant additions of ZnO, Al2O3, Fe2O3 and MgO. The basic ternary phase diagram is given in Figure 5.3, and shows that the commonly used slag region is within the low liquidus temperature zone of the eutectic trough and roughly along the line joining Ca2SiO4 and Fe2SiO4 (Winterhager and Kammel, 1961).

FIG 5.3 - Phase diagram CaO-FeO-SiO2 system.

Slag viscosity is a most important parameter for the performance of the furnace, with low viscosity preferred to allow separation of slag and bullion, and to allow tapping of the furnace at a practical rate. As well as lowering viscosity by raising temperature, slag viscosity at a given temperature will increase with increases in the content of silica and alumina, and will decrease with increases in the content of CaO, FeO and MgO, as well as PbO. Silica and alumina tend to form polymers or network bonding in the melt, increasing viscosity, whereas those bonds can be broken by the addition of CaO, FeO and MgO. However, changes in composition can also raise melting points of the slag above practical levels, such as by MgO addition, and can also increase the tendency to form solid phases in suspension, which can effectively raise viscosity. The latter effect can be promoted by increases in alumina, and by the iron content, which can promote the formation of solid magnetite (Fe3O4). Increasing the oxygen potential of the slag will tend to increase the magnetite content, but this is offset by the increase in PbO content, which acts as a powerful fluxing agent.

72




To work within a reasonable and practical melting temperature range, at or below 1200°C, it is necessary to operate within a restricted region within the following boundaries of an idealised ternary system:

• 30 to 40 per cent by weight

SiO2


FeO


CaO

This will give average ratios as follows:

• FeO:SiO2

1.3

• CaO:SiO2

0.7

The presence of ZnO and Al2O3 will of course change the characteristics of the molten slag to some extent, but it has been common practice and far less complex to initially consider the basic ternary system as a means of defining the operating slag regime. Table 5.2 gives typical slag compositions and composition ratios for most of the major lead blast furnace operations. TABLE 5.2 Actual slag compositions and ratios. Component FeO

Range

Average

25 - 36%

30%

Fe3O4

2 - 5%

3%

SiO2

19 - 25%

22%

CaO

12 - 20%

16%

Al2O3

1 - 11%

6%

MgO

1 - 3%

2%

ZnO

8 - 22%

13%

PbO

1 - 4%

2.5%

CaO:SiO2

0.65 - 1.05

0.8

FeO:SiO2

1.14 - 1.76

1.43

Basicity index

0.9 - 1.3

1.05

Fluidity index

1.5 - 2.3

1.75

The basicity index = molar ratio of (CaO + MgO + FeO + ZnO + PbO) to (2SiO2 + 3Al2O3). The fluidity index = mass ratio of (CaO + MgO + FeO) to (SiO2 + Al2O3).

The concept of ‘slag basicity’ is analogous to pH in aqueous systems and is useful in characterising general behaviour of the slag as it affects the following:

• slag melting point and ‘shortness’ or softening – melting range; • slag viscosity and therefore the ability to separate fine lead droplets; • distribution of impurity elements between slag and molten lead in contact with the slag; • electrical conductivity; • oxygen and sulfur capacities; and



73


• interfacial tension between slag and molten lead, as affecting lead droplet size and a tendency of the slag to foam. Most simply, basicity can be expressed as the CaO:SiO2 ratio or by the more complete basicity index as shown in Table 5.2. The fluidity index also shown in Table 5.2 is clearly a closely related empirical guide, demonstrating that viscosity is influenced by basicity. A more basic slag will have a higher fluidity or lower viscosity. Of most importance is the effect of slag basicity on the activity of dissolved species such as PbO and ZnO. The activity coefficient increases with increasing basicity and its effect can be illustrated by Equations 5.12 and 5.13: Pb + ½ O2 = PbOslag K=

a PbO PO 2

0. 5

=

(5.12)

γ PbO . N PbO PO 2

0. 5

(5.13)

where: K = the equilibrium constant for Equation 5.12 aPbO = the activity of PbO in the slag PO2 = the partial pressure of oxygen γPbO = the activity coefficient of PbO in the slag NPbO = the mole fraction of PbO in the slag From Equation 5.13, if the activity coefficient is increased by making the slag more basic, then for a given fixed oxygen partial pressure, the mole fraction of PbO in the slag will fall. That is, the PbO content of a slag in equilibrium with lead metal is reduced at a given oxygen potential by an increase in basicity. Basic slags have a similar effect on the ZnO activity coefficient and consequently will assist in the fuming of zinc from the slag. High CaO:SiO2 ratios tend to promote the formation of melilite structures in sinter, increasing its strength and raising the softening temperature. Industry correlations suggest that at a lime:silica ratio of 0.6 the softening temperature of typical slags is around 1025°C, whereas at a lime:silica ratio of 1.2 the softening temperature rises to 1110°C. Higher softening temperature means that the sinter maintains its strength further down the blast furnace shaft, improving shaft permeability and generally allowing for increased treatment rates. Slag fluidity is also increased to improve furnace performance. Extra lime addition will increase the slag make and lead losses in slag, but this can be compensated by the lower equilibrium concentration of lead in the more basic slag. To some extent CaO can be replaced by ZnO in the slag and high zinc slags will generally correspond with lower lime to silica ratios. One downside to a high calcium slag regime is that the high CaO content of sinter is known to cause its rapid degradation by absorption of water. This adversely affects the ability to hold stockpiles of sinter for any length of time. The Mount Isa lead smelter is a prime example of an operation with high lime to silica ratios at around 1.2. Basicity can theoretically be increased by raising the iron content, but high iron slags can have the potential risk of forming magnetite (Fe3O4) in the lower oxidation zone of the shaft and in the slag bath.

74




This forms as a solid phase in suspension in the slag and considerably raises its viscosity, making tapping difficult as well as the separation of slag and bullion. Simple tests are available for measuring magnetite and can be applied to grab samples of slow cooled or granulated slag. This problem is not generally encountered with the lead blast furnace, but can occur in the ISF. In the case of high zinc inputs in concentrates and sinter, the quantity of slag will be dictated by the limiting content of zinc in slag to around 17 per cent (or 22 per cent ZnO). Higher levels of zinc will dramatically raise the slag viscosity and 17 per cent is generally regarded as a practical upper limit, although for a typical operation no more than 15 per cent is preferred because of effects on accretion formation and higher upper shaft temperatures. High zinc inputs in sinter plant feed will therefore require additional fluxing to maintain these levels in blast furnace slag. The lead content of slag is the sum of oxidised/dissolved lead and suspended droplets or prills, and it is often difficult to distinguish between the two. Dissolved lead is affected by the oxygen potential of the slag and entrained lead by slag viscosity. The total content of lead in typical blast furnace slag at 2.5 per cent is no longer acceptable for the dumping of slag as land fill, or for use as a construction material, since it will fail the USEPA Toxicity Leach Procedure (TCLP), which is used as a benchmark for the suitability of material for landfill. Some form of additional slag cleaning is therefore required before dumping or utilisation. This may take the form of an electric furnace for settling and further reduction, or a slag fuming furnace. The slag fuming furnace will primarily be aimed at recovering zinc but is most efficient in also recovering most of the residual lead. Details of these subsidiary operations are covered in Chapter 8.

FURNACE CONSTRUCTION The lead blast furnace consists of a rectangular shaft with sidewalls made up from a series of watercooled hollow steel jackets. Refractory lining is unnecessary since an accretion layer forms on the inner surface to protect the steel from attack. The shaft narrows at the ‘bosh’ to the lower section containing the tuyeres. This allows for the reducing volume of the charge as coke is consumed and the sinter melts. In older furnaces the top shaft can be refractory brick within a steel shell, but the lower section around the tuyeres will be of waterjacketed construction. Furnaces constructed in this way tend to burn out the joint between the upper refractory and the lower water jackets and the full steel furnace was a significant improvement. It is also easier to remove accretions from the steel surface compared to refractories and this is a key advantage of a water jacketed furnace. The tuyeres are fitted through the jackets and are constructed of copper, often fitted with stainless steel tips to reduce corrosion and erosion. The cold air blast is sufficient to keep the tuyere bodies cool. Jackets at the tuyere level can be refractory lined on initial installation, but this is gradually replaced by accretions. The influence of the tuyeres can cause burn out of jackets, particularly with oxygen enrichment of blast air, but this can be overcome by the projection of the tuyere further into the furnace. The lower section of the furnace sits on a hearth made up of several layers of refractory brick and a castable refractory crucible, all held within a steel frame. The construction of the hearth and selection of the bricks and refractories used is critical to avoid leakage of molten lead and expansion of the brickwork, which can limit the operating life of the furnace before reconstruction. If properly constructed this is not a problem and hearth life can be determined more by the build up of accretions, which gradually restrict the flow of bullion to the tapping point.



75


The integrity of the jackets and any lining refractory is another key element in determining furnace life, which is in the range of two to seven years, before extensive re-bricking and reconstruction is required. An advantage of water-jacketed construction is that failed water jackets can be easily and quickly replaced compared with difficult repairs for a refractory lined furnace shaft, although with maintenance and good cooling water flow and the use of corrosion inhibitors in the cooling water, jacket failure is not a common problem. A schematic of a typical lead blast furnace is given in Figure 5.4, and shows an open topped furnace with feeding by rail-mounted dump cars and fitted with a central gas offtake hood.

FIG 5.4 - Cross-section of a typical lead blast furnace.

76




The width of the furnace at the tuyeres or lower section is set by the penetration of the air blast and is in the range of 1.2 to 2.0 m, but more commonly 1.5 to 1.8 m. The tuyeres are set at a height of around 450 mm above the furnace tap hole. The upper shaft width can be expanded to 2 to 3 m and the overall shaft height is around 5 to 6 m. Some experimentation with tuyere angle of inclination and distance above the hearth may be needed for each furnace to achieve the desired crucible temperature and lead content in slag. The length of the furnace cross-section is variable depending on the required capacity, but is limited by the distance slag is required to flow to the tapping hole. The limiting distance for slag flow is of the order of 6 to 7 m, which dictates the length of a furnace with end tapping. A central tapping hole will allow for increased length and in practice furnaces range from 5 to 10 m. Tuyere spacing is also variable from 310 to 480 mm, but averages around 450 mm, giving ten to 20 tuyeres per side. Blast air volume ranges between 1000 and 2000 Nm3/h per m2 of hearth area. Tuyeres have a jet diameter ranging from 50 to 80 mm and carry 0.12 to 0.18 Nm3/s of air, indicating a superficial velocity of 40 to 60 Nm/s. Blast pressure is in the range of 15 to 25 kPa, quite low compared with the iron blast furnace, but necessary because of the continuous tapping systems used. Tuyeres are constructed of copper and can have stainless steel inserts at the tip to limit corrosion and erosion. A sight glass is fitted at the external end to allow the operator to view into the furnace. Airflow to individual tuyeres can be controlled by dampers and they can be isolated to allow cleaning by punching through with a bar from the sight glass port. Blast air is usually delivered by Rootes type blowers and is normally not preheated, but can be enriched with oxygen. Furnace cooling is by water circulating though the furnace wall jackets. The water is usually pumped from a head tank to provide security supply, through the jackets and then through a heat exchanger, removing heat to an evaporative water cooling system. It is important for the circulating water to be treated and softened as internal corrosion or blockage of jackets can be a major cause of failure. The temperature of water leaving each jacket is usually measured and recorded, since loss of cooling will lead to jacket failure and leakage of water into the furnace. This is difficult to detect, but can cause solidification of the charge adjacent to the leak and eventual blockage of the shaft and shutdown. The furnace design at the Port Pirie smelter differs from the standard approach, having a second row of tuyeres in the upper section and 990 mm above the lower row. There are fewer tuyeres in the upper row and the airflow is less than for the lower tuyeres. This particular arrangement was designed to allow for an increase in the upper shaft width to 3.0 m to overcome problems associated with accretions bridging across the narrower shaft, particularly when operating with relatively high zinc concentrations in slag, at 18 per cent (Green, 1977). Details of the lower section and hearth construction of the Port Pirie furnace, and a plan showing the centrally located tapping arrangement are given in Figure 5.5.

Furnace top and furnace feeding A measured mix of sinter and coke is delivered into a weighed transfer hopper from separate feed bins and the contents are dumped into the top of the furnace. Some furnaces can be fitted with fixed feed hoppers and others use rail-mounted bottom dump transfer cars. Most furnaces have a top open to the atmosphere with a central off-gas draft hood. The hood is under suction and considerable dilution air is sucked in down through the top of the furnace charge. The off-gas volume is consequently three to four times the blast air volume, depending on the hood suction applied and in some cases can be significantly higher. The purpose of the open top is to provide access for the regular cleaning of accretions from the furnace walls. It also facilitates the



77


FIG 5.5 - Port Pirie blast furnace design (Fern and Jones, 1980).

78




charging of large and irregular plant recycle materials, but has the major disadvantage that the development of gas channels within the charge due to accretion build-up and uneven gas distribution can locally overload the off-gas draughting and cause a blow-out of gas from the surface of the charge directly to atmosphere. This gas carries a heavy load of lead and zinc oxide fume and is a significant hazard and a serious environmental issue. Enclosure of the top of the furnace is possible, but impedes access to remedy or prevent an impending or existing gas blow. A sealed top furnace such as the ISF fully prevents this situation. However, the propensity to form accretions is far greater in the lead furnace where zinc vapour is re-oxidised in the upper section, than is the case in the ISF. Doors are often installed over the top and opened only during charging, but this does not provide a seal or adequately control a major furnace blow. Some installations use multiple furnaces with one on standby to allow a shutdown for accretion removal. However, this is a highly inefficient and costly practice. This undesirable feature of the standard lead blast furnace is perhaps one of the reasons why its use will be phased out, and it is unlikely to be applied in any new lead smelting operation.

Tapping Tapping of lead bullion is usually continuous through an inverted lead siphon located at one end or on the side of the furnace. This maintains a fixed level of molten lead in the crucible, which is important to preserve the integrity of the base of the hearth and prevent attack from the higher temperature slag. Slag may be separately tapped, intermittently or continuously, through an underflow-overflow weir arrangement. For batch tapping the furnace is fitted with a water-cooled tapping breast containing the tapping hole. The tapping hole is plugged with fireclay and is opened manually using a bar or an oxygen lance and slag is directed into a settling forehearth before overflow to either a ladle or granulator. Bullion is collected and siphoned from the base of the settler and fed to the bullion collection system. For continuous tapping, the Asarco design of continuous tapper is commonly used and provides for both bullion and slag tapping at the one point (Roy and Stone, 1963). A water-cooled steel tapping breast contains an underflow weir to provide a furnace seal and is attached to a refractory lined channel connected to a forehearth in which bullion and slag are separated by settling. Lining is commonly chrome magnesite brick. The forehearth is a steel shell lined with castable refractory. It contains an overflow and underflow weir for separation of lead and slag into two separate streams, and a refractory lined cover fitted with gas burners to maintain a fluid slag surface. Since refractory life is limited in this duty, the forehearth is usually removable and can be mounted on rails for rapid changeover, which may take from two to eight hours. With continuous tapping of slag it is necessary to balance the head of slag in the tapping device with the weir height and gas pressure in the furnace. Factors such as erosion of the weir bricks, or a rise in gas pressure due to changes in furnace conditions, or the nature of the charge can cause a blow-out of hot gas from the tapper, and ventilation must be provided to accommodate this situation. This can be corrected rapidly by reducing the blast rate. Fine adjustments to the weir height can be made by inserting or removing thin (10 - 20 mm) slices of high quality brick into the ‘V-notch’ of the weir. Lead bullion flows from the tapping facility via a Y launder into either of two steel ladles lined with castable refractory and located in a pit adjacent to the furnace. The ladles are usually around 10 t in capacity and are handled by overhead gantry crane, transporting the ladles to the copper drossing operation. Similarly, molten slag can be run into cast steel ladles or directly into a granulator. The usual form of granulator is a launder carrying a large flow of water into which the slag stream is directed. The slag stream is broken up by the water flow and chilled to form particles similar to coarse sand. The stream



79


flows into a pit where the slag settles out and can be recovered by scraper conveyor and elevated to a hopper for drainage of retained water, and from where it can be trucked to disposal. Normal slags can be granulated without difficulty, but in situations where matte, speiss or bullion is present, explosions can occur in the granulation launder due to encasement and superheating of water. The ratio of the mass flow of granulation water to the mass flow of slag is typically 30:1.

Gas handling system Off-gas from the top of the furnace may be three to four times the blast air input and in addition an equivalent volume of ventilation air can be drawn from around the furnace, covering tapping points, slag granulation, transfer of bullion to ladles and drossing of bullion. Total ventilation air and off-gas is usually cooled by water injection in a spray tower to around 150°C and is then filtered in a bag house and the cleaned gas discharged to atmosphere. It is important to maintain the gas above its dewpoint during filtration to prevent corrosion and bag failure. Solids loading of the gas to the bag house can range from 2 to 10 g/Nm3. Using a modern bag house with reverse pulse cleaning, a filter area of 30 m2 per 1000 Nm3/h is suitable. Collected dusts are pulped in water, filtered and the filter cake added to sinter plant feed. If the dusts contain substantial amounts of impurities such as cadmium or thallium, the filtrate may be processed to recover these metals. A feature of the traditional blast furnace and sinter plant is the huge volume of dilute process and ventilation gases loaded with fumes and dusts which need to be collected and cleaned. Off-gas will contain some sulfur dioxide and, depending on its concentration, removal by scrubbing could be necessary, although this is generally not the case (see section below on sulfur balance).

FURNACE OPERATION Feed preparation and feeding methods Both sinter and coke need to be correctly and closely sized for optimum performance of the blast furnace. Sinter is delivered as predominantly -150 mm +50 mm lump material from the sinter plant with undersize forming recycle within the sinter plant (‘return sinter’). Purchased coke normally contains an excessive amount of fines as received and needs to be screened before use in the blast furnace. The coke fines (‘breeze’) can be used as part of the sinter plant charge and for dross conditioning during the refining operation. Furnace coke is sized similarly to sinter at +50 mm, but fine coke at +10 mm may also be added to control coarse coke size retention into the lower shaft. Coke and sinter may be added separately to the furnace in alternating layers or may be mixed together prior to charging. Accurate weighing of the components as batches is the normal practice, often using transportable hoppers on a weigh scale. The hopper is then moved to the top of the furnace and the contents dumped in. Alternatively the charge may be dumped into a subsidiary hopper for controlled feeding into the furnace. Placement of the charge, in particular the rilling of coarse lump material towards the wall of the furnace can cause segregation and a more open structure against the wall, thus affecting gas distribution and encouraging accretion formation on jackets and wall sections. This will also aggravate the formation of a central dead zone within the furnace as well as raising oxygen levels close to the wall, and is an important consideration in the design of the charging system. Most charging systems therefore add feed at the furnace walls, allowing the coarse material to rill towards the centre of the shaft rather than the walls.

80




Furnace accretion removal The formation of accretions within the charge and particularly on the walls of the furnace is a major issue and can restrict gas flow and hence furnace throughput, as well as causing gas channelling and surface eruptions or gas blows. Accretion formation is discussed above and the main areas of concern are the wall accretions, the central accretion or ‘deadman’ or ‘sow’ and the hearth accretion. One factor contributing to the formation of the central accretion or ‘deadman’ is the lack of adequate penetration of the blast. This can be corrected by increasing blast velocity through reduction in the diameter of the tuyeres and also by reducing the width of the furnace at the tuyere level. The Naoshima Smelter in Japan reported an effective removal of the central accretion by these measures, including reduction of furnace width from 1.66 to 1.42 m, and significantly increased furnace capacity (Moriya, 1989). The presence and location of accretions can be identified by temperature distributions and gas flow patterns. Removal of wall accretions can simply be done by barring down with access through the top of the furnace, either manually or by using mechanical hammering devices. For difficult accretions, explosive charges are employed. In this case a pipe is driven down into the accretion and an electrically detonated charge is dropped down the pipe and exploded. Damage to the jackets is always a risk with this approach as well as the hazard to operators with the handling of explosives. In the extreme, the furnace can be ‘burnt down’ by discontinuing feed, allowing the charge to drop down to just above the tuyere level. Ready access to the walls is then possible for the removal of any remaining deposits as necessary, and this approach will effectively remove the central ‘dead man’ or ‘sow’. As indicated above, the hard hearth accretion forms relatively slowly and eventually causes interruption to the tapping of the furnace. It requires a complete shutdown for removal.

Furnace capacity Furnace capacity is ultimately the lead production rate as tonnes per day per square metre of hearth area, but is basically controlled by the carbon burning capacity in tonnes of carbon burned per day per square metre of hearth area. This in turn relates to the air blast or oxygen throughput that can be achieved, as influenced by the permeability of the charge. Hence the importance of uniform sizing of coke and sinter, and the retention of sinter structure without degradation. Data of plant performance for a range of smelters are given in Table 5.3. Carbon use is for reduction, but is mainly for the provision of process heat and appears in the heat content of molten slag, lead bullion, exit gases and heat lost from the furnace walls in cooling water. In this sense carbon demand is more closely related to sinter throughput than other factors. From Table 5.3 fuel use is generally within the range of 90 to 100 kg of carbon per tonne of sinter processed, whereas the carbon used per tonne of bullion varies over a much wider range of 170 to 370 kg/tonne. This emphasises the advantage of maximising the lead content of sinter. Actual fuelling requirements can be established from detailed heat and mass balances for a particular furnace. However, for a typical lead blast furnace, a fuelling level of 90 kg/t of sinter would appear to be a reasonable figure for efficient operation. If the top gas CO:CO2 ratio is 0.4 then 2.286 kg of oxygen are required per kilogram of carbon burned. This is supplied by reduction of sinter and by the air blast. For a sinter of 45 per cent Pb and ten per cent Fe the reduction oxygen supplied will be close to 4.9 per cent, hence one tonne of sinter will supply 49 kg of oxygen and consume 21.43 kg of carbon. The remaining (90 - 21.4) = 68.6 kg of carbon will require 156.8 kg of oxygen to be supplied by the air blast, representing 523 Nm3 of blast air.



81

Belledune Binsfeld- Chigirishima East Helena hammer

Plant Number of operating furnaces

Glover

Herculeneum Hoboken Mount Isa Nordenham Port Pirie

Torreon

1

1

1

2

1

2

2

1

1

1

3


108 000

92 000

90 000

70 000

120 000

225 000

125 000

156 000

102 300

215 000

180 000

Sinter feed rate (t/h)

40

29

22

35

39

30

41

44

29

54

62.5

Coke feed rate (t/h)

5.1

2.8

3.15

4.6

2.66

5.8

9.73

Furnace dimensions (m × m) 2 Cross-section area (m )

Number of tuyeres

1.68 × 6.4 10.75

10.7

2.05 × 7.93

1.9 × 7.74

1.76 × 8.5

1.3 × 7.5

1.83 × 7.02

1.2 × 7.6

1.5 × 10.7

1.52 × 6.4

9.35

16.2

14.8

14.96

9.75

12.85

9.12

16.2

9.73

32

42

40

46 + 32

21

8400

16 500

13 600

25 500

26 000

27 000

20 000

21

25 - 27

23.5

23

21

36

3


1.7 × 5.5

Blast volume (Nm /h)

20 500

Oxygen in blast air (%)

26

Coke cal value (kJ/kg)

23 7000

Coke fixed carbon (%)

90

Bullion (t/d)

360

280

S content (%)

1.3

<1

228

220

25

22.5

23.5

25

7600

6200

7000

6700

93

93

82

365

700

347

<2

0.09

510

83 290

650

0.15

0.4 2.2

Cu content (%)

2.5

0.03

0.02

Matte produced (t/d)

15

0

30

0

31

0

Slag produced (t/d)

430

85 - 110

280

330

760

350

650

Pb content (%)

3.5

1.25

3.6

1.8

2.2

2

2.5

500 0.06

41

40

150 - 200

690

600

2.6

1.75

2.6

1.3 31.4


FeO content (%)

36

33

28

25.8

30.8

33

32.5

22.8

32.5

26.3

CaO content (%)

15.5

17.5

20

19.6

15

12

16.3

24.5

20.5

15.1

20.5

SiO2 content (%)

20.5

20.5

21

22.3

27

22

24.9

21

19.5

21.6

19.9

255 000

40 000

19 200

36 000

170 000

25 000

92 500

40 000

0.8

0.8

2.5

Off gas volume (Nm3/h) Off gas dust load (t/h)

1.25

0.4

To 60 000 0.5

3

Performance ratios 2 Carbon use (t/m .day) 2

Sinter input (t/m .day) 2

Bullion output (t/m .day) 3

2

5.2

4.2

4.7

7.0

7.0

7.1

89.3

10.2 40.9

56.5

51.9

63.2

48.1

50.5

82.2

76.3

80

33

33

24

14

25

23

18

40

32

40

17

898

1019

919

1705

2023

1667

770

100

66

98

85

Blast rate (Nm /h.m )

1907

Carbon use (kg/t of sinter)

114

92

89

51.4


82

TABLE 5.3 Data for commercial lead blast furnaces.


Common tuyere capacity is 700 Nm3/h and blast air supply from two opposing tuyeres is 1400 Nm3/h for a furnace length of around 450 mm (tuyere spacing), or a furnace cross-sectional area of 0.765 m2 for a 1.7 m wide furnace. This represents 1830 Nm3/h of blast air per square metre of hearth area. Hence 1 m2 of hearth can handle 1830/523 = 3.5 t/h of sinter (84 t/m2.day), and lead bullion production will be around 1.5 t/h.m 2 (36 t/m2.day). In this case the carbon burning rate is 84 × 90/1000 = 7.5 t of carbon/m 2.day. For the above illustration the overall heat balance for the blast furnace will be as shown in Table 5.4. TABLE 5.4 Blast furnace overall heat balance. Heat supplied by coke combustion to CO2

2510 MJ/t of sinter

100%

Heat absorbed by metal reduction reactions

748 MJ/t of sinter

29.8%

Heat content of slag at 1200°C

778 MJ/t

31.0%

Heat content of bullion at 1200°C

82 MJ/t

3.3%

Heat content of exit gas at 220°C

133 MJ/t

5.3%

Heat absorbed by reduction of CO2 to CO

516 MJ/t

20.5%

Heat lost in cooling water

253 MJ/t

10.1%

Table 5.4 illustrates the general distribution of heat generated by the full combustion of carbon to CO2 and the significant loss of potential heat by the formation of CO and its content in the exit gas. Reduction in the CO:CO2 ratio in exit gas from 0.4 as assumed above to 0.2 would reduce coke consumption for the same furnace operation and heat balance conditions from 90 kg/t of sinter to 81 kg/t of sinter. By raising the lead content of sinter from 45 per cent to 48 per cent for the same assumed furnace conditions, lead bullion production can be raised by seven per cent for the same coke consumption. Although the above calculations are simplistic they can be used to clearly illustrate the impact of key process parameters on furnace performance, such as the CO:CO2 ratio in exit gas, the Pb content of sinter, the blast rate and the oxygen content of blast air. There are other practical limits on these parameters to maintain practical operating temperatures at various regions in the furnace shaft, and to minimise the formation of accretions and other obstructions.

Oxygen enrichment of blast air Experience with oxygen enrichment of blast air is illustrated in Figure 5.6 (Fern and Jones, 1980), which shows that furnace capacity is increased by five times the oxygen percentage increase. A two per cent enrichment from 21 per cent to 23 per cent will raise furnace lead production by ten per cent. Blast furnace production is thus directly proportional to the oxygen flow rate given a constant tuyere volume and other furnace parameters. There is some indication that oxygen enrichment can increase the level of lead in slag, but according to Fern and Jones this can be counteracted by raising the CaO content of the slag. Most blast furnaces today run with some level of oxygen enrichment of blast air to maximise furnace capacity and this usually ranges up to 25 per cent oxygen content in blast air. Using the above mass balance calculation as an example, by raising the level of oxygen in blast air from 21 per cent to 23 per cent the blast air requirement per tonne of sinter would reduce from



83


45 40

% Change in Production

35

East Helena 30

Bunker Hill Chimkent

25

Trail Hoboken

20

Ust Kamenogorsk Trend

15

Linear (Trend) 10 5 0 20

22

24

26

28

30

% Oxygen in Blast Air

FIG 5.6 - Effect of oxygen enrichment on furnace capacity at various lead plants (Fern and Jones, 1980).

523 Nm3 to 478 Nm3 and the quantity of sinter handled per square metre of hearth area would rise from 3.5 t/h to 3.82 t/h or an increase of close to ten per cent for the same gas flow rate. This is in agreement with observed practice.

Blast preheating Blast preheating theoretically has the potential to significantly reduce the requirements for internal heat generation and hence carbon consumption, and is commonly used in other blast furnace operations. Blast air is usually preheated by heat exchange with hot off-gases directly, or after combustion of the CO content to further raise their temperature. However, in the case of the lead blast furnace, the open top arrangement leads to excessive dilution of the off-gas and leaves little opportunity to use flue gases as the heat exchange medium to achieve significant blast air preheat temperatures. The level of CO is also too low for use as a fuel for preheating as in other blast furnace operations. The counter-current operation and low top temperatures also means that there is effective recovery of heat from the furnace gases to the descending charge and hence it achieves charge preheating rather than blast air preheating, which is more necessary with hot top operation such as with the Imperial Smelting Furnace. There is also the view that preheating of blast air and maintenance of bottom furnace temperature will mean a reduction in the coke rate and hence the volume of gas flowing up through the furnace. This will reduce the heat available at the top of the furnace to meet requirements of charge heating and in the limit would necessitate charge preheating – as required for the ISF.

General operating practices In general the blast furnace operates best at its maximum design blast rate. Any attempts to run at reduced blast usually results in poor gas flow distribution with attendant problems of accretion formation or central dead zone enlargement due to low blast penetration into the charge. Sinter quality is of paramount importance in maintaining a high blast rate and good gas distribution and it is often

84




preferable to reject poor quality sinter for recycle in the sinter plant rather than attempt to treat it through the blast furnace, since any upsets resulting from poor gas distribution and resulting in accretion formation can take considerable time to correct. Conditions at the tuyere level are also critically dependent on the coke supply rate as well as the air supply rate, and if this is deficient temperatures will fall, reducing the smelting rate, even though the blast rate has been maintained. Blast furnace condition at the tuyere level can be indicated to the operator by ‘clear and bright’ tuyeres as observed through the sight glasses. Among other measures, this can also indicate coking conditions or whether the furnace is receiving sufficient or excessive coke supply at the tuyere level. Coking requirements are discussed above in the section on Coke Consumption Reactions and Heat Balance. A number of temperature measurements of the charge at various levels within the furnace is beneficial in indicating imbalances and poor gas distribution, or in indicating inappropriate fuelling conditions. However, this is seldom practised. The condition of the slag is also a key indicator to the operator. A change in slag viscosity can result from changes in slag temperature and from sinter compositional changes or from the formation of magnetite as a result of excessive oxygen potential at the base of the furnace. High slag viscosities can cause major difficulties with tapping and the separation of bullion from slag. It can cause a buildup in slag levels in the furnace and interference with the operation of the tuyeres or even blockage of the tuyeres. With continuous tapping, a reduction in slag viscosity can also result in excessive lowering of the slag level and a blow-out of gas through the tap hole. Stable slag conditions are therefore of major importance.

ENVIRONMENTAL ISSUES Particulate lead emissions The main environmental concern with lead blast furnace operation relates to the open top and the emissions therefrom. As discussed above under ‘Furnace Top and Furnace Feeding’ ‘gas blows’ can cause bypass of the venting system leading to emissions directly to atmosphere. The emitted gas contains a heavy load of lead and zinc oxides. Although infrequent, this is becoming unacceptable and means of capture or avoidance are necessary. Enclosure of the top of the furnace can contain emissions, but will interfere with access for the management of accretions, particularly when a blow is occurring. The removal of accretions is a constant need with the lead blast furnace and is generally regarded as incompatible with an enclosed top. Alternatively, extensive hooding and a high level of ventilation is necessary to capture ‘blows’ in most situations, but carries the cost of a high volume of gas to be moved and filtered. Apart from the furnace top there are a number of points of fume emission such as from tapping operations as well as high lead dust from raw materials handling operations, which all require extensive ventilation to capture such emissions.

Sulfur in off-gas and sulfur balance As indicated in Chapter 4, sinter contains between 1.0 and 2.0 per cent sulfur and typically 1.7 per cent for a lead content of around 45 per cent. This is the major input of sulfur to the blast furnace, although coke can contribute of the order of five per cent of the total sulfur supply.



85


The sulfur content of blast furnace bullion is around 0.5 per cent and slag around 1.5 per cent. The remaining sulfur reports to furnace off-gases. Some may be captured as a basic lead sulfate in fume, which is recycled to the sinter plant, otherwise the remainder reports as SO2 in filtered off-gases and is normally discharged to atmosphere. Equilibrium conditions at the base of the furnace would suggest that the SO2 partial pressure should be of the order of 10-7 to 10-6 atmospheres. Consequently the sulfur content of the gas stream from the base of the furnace should be negligible and sulfur will be captured in the bullion and in slag. It follows that the presence of SO2 in furnace exit gas results from reaction higher in the shaft, probably the decomposition of sulfate according to Equation 5.10 and the reaction between volatilised PbS and PbO according to Equation 5.11. A typical balance may be as follows per tonne of product lead:

• sinter – 2.27 tonnes at 45 per cent Pb and 1.7 per cent S

38 kg of S per tonne of lead

• coke – 0.2 tonnes at one per cent S

2 kg/t

• bullion – 1.0 tonnes at 0.5 per cent S

5 kg/t

• slag – 1.2 tonnes at 1.5 per cent S

18 kg/t

• fumes – 35 kg at six per cent S

2 kg/t

• balance of sulfur reporting to blast furnace gas

15 kg/t

In this example 40 per cent of the sulfur reports to the blast furnace gas, but this proportion will greatly depend on sinter composition. Typical gas volumes emanating from the blast furnace operation are as follows:

• blast furnace gases

1200 Nm3/ t of lead

• furnace top dilution air

3500 Nm3

• ventilation air

5000 Nm3

Ventilation airflows in particular vary greatly from one plant to another depending on particular configurations and design. If the total gas flows are handled collectively for fume separation the total gas flow will be 9700 Nm3 per tonne of blast furnace lead and will contain 15 kg of sulfur or 30 kg of SO2. This represents a concentration of 3.1 g of SO2 per Nm3 or approximately 0.11 per cent by volume. Without dilution by ventilation air the concentration would rise to 6.4 g of SO2 per Nm3 or 0.22 per cent by volume. Emission standards for SO2 vary widely, but a common concentration limit is 1.25 g of SO2 per Nm3, which has to be met by sulfuric plant emissions. On this basis, sulfur emissions from the blast furnace commonly exceed normal standards by a substantial margin. This has indeed resulted in the closure of some European lead blast furnaces or the conversion to alternative technologies. The alternative is to scrub the gas after fume collection with a lime slurry, which is capable of reducing the SO2 concentration to 0.01 per cent by volume or 0.3 g per Nm3. Otherwise a combination of total plant ventilation gases from sinter plant and refining operations with blast furnaces gases may be able to meet overall emission concentration limits for sulfur, but at the expense of large gas volume movements per tonne of lead produced (at around 12 000 Nm3/tonne of lead).

REFERENCES Collins, H F, 1910. The Metallurgy of Lead (Charles Griffin and Co: London).

86




Fern, J H and Jones, R W, 1975. The development and operation of a continuous tapping lead blast furnace at Port Pirie, in Proceedings AusIMM Annual Conference, pp 155-165 (The Australasian Institute of Mining and Metallurgy: Melbourne). Fern, J H and Jones, R W, 1980. Lead blast furnace smelting with oxygen enrichment at the BHAS plant, Port Pirie, South Australia, in Proceedings Lead-Zinc-Tin 80 – World Symposium and 109th Annual Meeting of the AIME, pp 321-332 (Metallurgical Society of American Institute of Mining, Metallurgical and Petroleum Engineers: Littleton). Grant, R M and Cunningham, B C, 1971. The relationship between sintering practice and lead blast furnace performance at Port Pirie, TMS/AIME paper selection A71-1. Green, F A, 1977. The Port Pirie Smelters (The Broken Hill Associated Smelters Proprietary Limited: Melbourne). Hoffman, H O, 1899. The Metallurgy of Lead, fifth edition (The Scientific Publishing Company: New York). Hopkins, R J and Haney, L B, 1954. Thoughts on lead blast furnace smelting, Journal of Metals, March, pp 1209-1213. Iles, W M, 1902. Lead Smelting (John Wiley: New York). Lumsden, J, 1971. The physical chemistry of the lead blast furnace, in Proceedings Metallurgical Chemistry Symposium (ed: O Kubaschewski), pp 533-548 (National Physical Laboratory, HMSO: London). Madelin, B, Sanchez, G and Rist, A, 1990. Investigation and modelling of the non-ferrous blast furnaces of Metaleurop, in Proceedings Lead and Zinc 90 Symposium, pp 571-596 (The Minerals, Metals and Materials Society: Warrendale and American Institute of Mining, Metallurgical and Petroleum Engineers: Littleton). Moriya, K, 1989. Achievement in lead smelting during a quarter century at Mitsubishi-Cominco’s Naoshima smelter, in Proceedings Primary and Secondary Lead Processing Symposium, Halifax, August, pp 71-86. Oldwright, G L and Miller, V, 1936. Smelting in the lead blast furnace, Trans AIME, 121:82-105. Polyvyannyi, I R, Elyakov, I I, Gaivaronskii, A G, Ivakina, L P and Anan’ev, N I, 1971. Investigation into lead blast furnace smelting with preheated oxygen rich blast by freezing the furnace, Trudy Inst Met i Obogaschch Akad Nauk Kaz SSR, 43:54-84. Rist, A and Meyson, N, 1967. A dual graphic representation of blast furnace heat and mass balances, Journal of Metals, 19:50-59. Roy, J T and Stone, J R, 1963. Lead blast furnace continuous tapping, Journal of Mining, 15:827-829. Ruddle, R W, 1957. Difficulties Encountered in Smelting in the Lead Blast Furnace, 55 p (Institution of Mining and Metallurgy: London). Willis, G M, 1980. The physical chemistry of lead extraction, in Proceedings Lead-Zinc-Tin ’80 Symposium, pp 457-476 (The Minerals, Metals and Materials Society: Warrendale, and American Institute of Mining, Metallurgical and Petroleum Engineers: Littleton). Winterhager, H and Kammel, R, 1961. Viscosity measurements in process slags from lead and copper recovery processes, Erzmetall, 14:319-328.



87

CHAPTER 6 The Imperial Smelting Furnace (ISF) GENERAL INTRODUCTION The Imperial Smelting Furnace (ISF) was primarily developed as a blast furnace for the production of zinc. It relies on the concept of reducing zinc oxide in a shaft furnace to produce zinc vapour in the furnace gases. By maintaining high temperatures at the top of the shaft reversion to zinc oxide can be prevented. The hot furnace gases are ducted from the furnace to a condenser where they are chilled using a spray of molten lead at around 500°C, which condenses and absorbs the zinc from the gas stream. This concept requires close control of furnace gas composition and a sealed top furnace. A key advantage of the ISF is that it can simultaneously produce lead bullion as well as zinc and can therefore treat mixed lead and zinc concentrates. Lead production is normally 40 to 50 per cent of the zinc production, and the fuel required in addition to that used for zinc production alone is relatively small, within the normal operating range. Volatilisation of lead is minimised by lowering the residual sulfur content of sinter. The combination of lower sulfur in sinter and a more strongly reducing environment in the crucible of an ISF than a lead blast furnace means that bullion is lower in sulfur than bullion produced from the standard lead blast furnace. The aim is also to produce a slag substantially lower in zinc than the standard lead blast furnace at around six to seven per cent Zn, and requiring more reducing conditions at the base of the furnace and greater interaction of coke and blast air with the slag pool in the furnace hearth. The ISF was developed from the standard lead blast furnace, but with evolutionary change to a more intensive operation, more akin to the iron blast furnace conditions at the tuyere level. The standard ISF design had an upper shaft cross-section of 17.2 m2 and a cross-section at tuyere level or hearth area of 13.2 m2. Data for the standard ISF is compared with the lead blast furnace in Table 6.1 and demonstrates the higher intensity operation. TABLE 6.1 Comparison of the Imperial Smelting Furnace and lead blast furnace. Furnace type

Lead blast furnace

Imperial Smelting Furnace

Air blast (Nm3/h per m2 hearth area)

785

2500

Carbon use (t/d per m2 hearth area)

7.4

15.5 8.3

2

Lead output (t/d per m hearth area)

36

Zinc + lead (t/d per m2 hearth area)

36

27

Carbon used kg/t of sinter

90

375

ISF plants operating in 2000 are listed in Table 6.2. A number of these plants have subsequently closed.

PROCESS DESCRIPTION A general process flow sheet is shown in Figure 6.1. Furnace feed is prepared from sulfide concentrates, oxide feeds and residues as sinter using a standard updraught sintering machine with attached sulfuric acid plant. As for the lead blast furnace, high-quality sinter in terms of uniform coarse sizing and strength is essential for good performance and efficient operation of the blast furnace. Details of sintering operations are given in Chapter 4.



89

CHAPTER 6 – The Imperial Smelting Furnace (ISF)

TABLE 6.2 Imperial Smelting Furnace smelters in operation in 2000. Location

Country

Shaft area (m2)

Year started

Max zinc output (t/yr)

Max lead output (t/yr)

Avonmouth

UK

27.2

1967

105 700

51 200

Chanderiya

India

21.5

1991

61 400

31 000

Cockle Creek

Australia

24.2

1961

97 300

40 700

Copsa Mica

Romania

17.2

1966

34 100

17 100

Duisburg

Germany

19.3

1966

97 400

45 100

Hachinohe

Japan

27.3

1969

114 400

52 100

Harima

Japan

19.4

1966

88 900

28 600

Miasteczko

Poland

19.0

1972

84 600

36 000

Noyelles Godault

France

24.6

1962

115 700

48 300

Porto Vesme

Italy

19.0

1972

84 600

36 000

Shaoguan No 1

China

18.7

1975

81 500

34 800

China

17.2

1996

70 900

31 900

Macedonia

17.2

1973

59 700

30 500

Shaoguan No 2 Veles

Coke

Sinter

Sinter Hopper

Coke Preheater

Coke Hopper

Charge Bucket Gas Scrubber Gas LCV Gas

Bell Seals

Air Preheater

Condenser

Furnace Lead Pumps Cooling Launder

Separation Bath

Tuyeres

Blue Powder / Liquor ISF Zinc

Forehearth

Slag

Lead Bullion

FIG 6.1 - Schematic layout of the imperial smelting process.

90




The standard furnace is a rectangular shaft constructed of a steel casing cooled externally by water shower and lined with refractory brick. Original dimensions were 6.3 m × 3.2 m at the outer steel casing and 17.1 m2 upper shaft cross-section within the brick lining. The furnace width narrows progressively to the hearth. Above the hearth are two opposing rows of tuyeres (one row on each side of the furnace), inclined towards the hearth and with around eight tuyeres in each side. Tuyere nozzles are made of copper with annular water-cooling channels. Air injected through the tuyeres is generally preheated to between 800 and 1000°C using Cowper stove recuperators and heated by combustion of the CO from furnace off-gases. Preheat is an important means of increasing furnace capacity and it has been reported that 100°C of preheat is equivalent to a three to five per cent increase in furnace capacity (Lee, 2000). The top of the furnace is sealed and gases are taken through a side off-take to the condenser chamber. The standard furnace has one condenser chamber whereas the larger Avonmouth furnace design incorporated two condenser chambers located at opposite sides of the furnace. Furnace feed consists primarily of sinter of the appropriate composition to give the required slag formulation, and coke preheated to around 700°C. Other feed materials such as briquettes made from zinc oxides, and metallic scrap can also be added to the furnace feed. Feed is charged to the furnace using weighed buckets, which are dumped into a double ‘bell seal’ hopper arrangement located in the furnace roof. The bell seals are water-cooled steel construction and efficient sealing is very important. At the tuyere zone of the furnace, carbon combustion takes place predominantly forming CO and generating most of the heat required to operate the furnace. Temperatures up to 1600°C are generated in the blast zone or tuyere raceways. As gases rise through the shaft, zinc and lead oxides are reduced by CO, and carbon is consumed by the CO2 so generated to reform CO. These reactions are endothermic and temperatures decline towards the top of the shaft. When the sinter melts it trickles into the slag pool contained in the furnace hearth from where further zinc reduction takes place but at a lower rate. A higher temperature assists and most ISFs operate with a slag melting temperature of above 1200°C and a lower oxygen potential than the lead blast furnace. Increasing the lime content of the slag can be used to raise the melting point of the slag and increase the activity of dissolved ZnO. Sinter composition ratios are commonly targeted at:

• FeO: SiO2

2.7:1

• CaO: SiO2

1.4:1

• FeO: (CaO+SiO2+Al2O3)

1.1:1

Coke will usually add significant quantities of silica (up to 50 per cent of sinter feed amounts) and will alter these ratios for blast furnace slag, which is the ultimate target. Slag and bullion are tapped from the furnace hearth into a forehearth where they are separated. Slag flows from the forehearth into a granulation launder and lead bullion flows to a lead ‘kettle’ of around 100 tonnes capacity either directly or via batch ladles. The bullion is allowed to cool in the kettle to separate a copper-lead dross, reducing the copper content to around 0.2 per cent. The resulting lead bullion is then cast into blocks of up to four tonnes each for transfer or sale to a lead refinery. Lead reduction tends to occur in the top zone of the furnace by reaction of lead oxides with zinc vapour and with CO. The reactions are mildly exothermic. To ensure that the temperature of the gas stream leaving the top of the furnace does not fall below a critical level where re-oxidation of zinc occurs, some air is introduced to burn CO and generate additional heat (‘top air’). Generally the critical temperature to prevent re-oxidation is around 950°C.



91


Although the action of air injection does generate higher CO2 concentrations and increases the potential for zinc re-oxidation, this is offset by the effect of the temperature increase achieved. Top air can represent of the order of 12 per cent of the air volume injected through the tuyeres. Gas leaving the furnace generally has the composition of five to seven per cent Zn vapour, 16 to 18 per cent CO, ten to 13 per cent CO 2 and should have a temperature of close to 1000°C. The lead splash condenser consists of a chamber approximately 8.5 m long × 5.5 m wide × 1.15 m high, and contains a pool of lead 0.4 m deep. Vertical baffles divide the chamber into four sections. Each section is fitted with two rotors located at the surface of the lead pool and driven by a vertical shaft projecting up through the roof to the electric motor drives. The inclined blades of the rotor throw up a spray of lead droplets into the gas stream, scrubbing out zinc and cooling the gas from around 1000°C at the inlet to around 500°C at the outlet. Molten lead is continuously circulated through the condenser, exiting at a temperature of 550°C and containing approximately 2.5 per cent zinc. The hot lead flows through a cooling launder where the temperature is reduced to 450°C, at which point the solubility of zinc is reduced to 2.1 per cent (see Table 11.3). The lead flows to a settling bath where excess zinc floats to the surface and is separated using an overflow/underflow weir arrangement. Separated liquid zinc at 98.6 per cent zinc content is pumped to a holding furnace prior to refining, while the cooled lead is circulated back to the condenser. Gases exiting the condenser pass through a scrubbing system where solids containing zinc escaping the condenser are collected as ‘blue powder’. The gas with its high CO content is used as fuel for the preheating of blast air in Cowper stoves and for the preheating of coke.

SLAG COMPOSITION There are limits to the level of zinc in final slag, because as the oxygen potential is lowered to reduce more zinc, the tendency to reduce FeO and form iron is increased. At slag temperatures of 1350°C the presence of metallic iron can cause considerable difficulties in the tapping process. Generally zinc levels are not reduced much below six per cent in final slag for this reason. Apart from the zinc content, slag composition is important for control of the appropriate melting point and viscosity of the slag, to ensure suitable operation of the hearth, and the ability to tap the furnace and achieve good separation of slag and bullion. Composition adjustment is made by the addition of fluxing components to the sinter feed, but can also dictate the concentrate feed mix and place limits on the intake of particular feed materials. Ratios of major components are targeted in sinter as indicated above, but additional amounts of some components such as silica will be sourced from the coke and shift these ratios for slag. Slag compositions will vary depending on available feeds to the smelter, but are typically as shown in Table 6.3 with a liquidus temperature of around 1220°C. The chemistry of slags, the various phases which can be present, and their interactions is complex and minor variations of some components can have significant effects. For instance Al2O3 at high levels results in the precipitation of Hercynite (FeO.Al2O3), which raises the liquidus temperature of the slag but lowers the potential to form metallic iron. At low concentrations of Al2O3 there is increased potential to form metallic iron with associated problems. Such subtle effects can have a significant impact on the hearth and tapping operations.

EVOLUTION OF FURNACE DESIGN AND OPERATION As indicated above, there has been progressive improvement in the design and capability of the standard blast furnace and considerable expansion in its capacity.

92




TABLE 6.3 Typical Imperial Smelting Furnace slag composition. CaO

18.5%

SiO2

19.5%

FeO

41.5%

ZnO

9.5%

PbO

1%

Al2O3

8%

MgO

0.5%

Other

1.5%

Component ratio CaO: SiO2

Typical value

Range in values

0.95

0.8 to 1.4

FeO: SiO2

2.1

1.8 to 2.9

FeO: (CaO+SiO2+Al2O3)

0.9

0.7 to 1.2

The original standard furnace had an upper shaft cross-section of 17.1 m2 constructed of a refractory lined steel shell (6.3 m × 3.2 m), which was cooled externally by means of a water shower. The lower shaft narrowed at the tuyere level and was constructed of two rows of water jackets. Twenty-six tuyeres as two opposing rows (each of 13) were used at relatively low intensity or blast rate, in keeping with standard lead blast furnace designs. Figure 6.2 illustrates the general vertical cross-section of the original furnaces.

FIG 6.2 - Original imperial smelting process furnace configuration water jackets in lower shaft.

The formation of accretions in the shaft was a particularly serious problem with the early design, aggravated by water leaks from the jackets forming the lower shaft. This problem became more significant as attempts were made to increase the blast intensity and productivity of the furnace. It was



93


recognised that the required characteristics of the furnace were more akin to the higher intensity iron blast furnace than to the lead blast furnace. These characteristics are:

• the higher carbon requirements per unit of metal produced, • the highly reducing conditions required, • the use of blast air preheat for fuel efficiency, and • the higher blast rates required for furnace productivity. It was recognised that the narrow hearth configuration as shown in Figure 6.2 was unsuitable for high intensity operation and for maximum retention of slag in the hearth in order to minimise the final zinc content of slag. This latter requirement followed from the recognition that a significant level of ZnO reduction occurred from the slag pool held in the hearth. As a consequence of these issues the hearth of the furnace was widened, water jackets in the lower shaft were replaced with a showercooled steel casing, the number of tuyeres was increased and the blast rate was increased. Raceways developed at the tuyere level as in the iron blast furnace, creating greater mobility of coke at the tuyere zone and greatly increasing the interaction with the slag pool. The general cross-section of the furnace changed to that shown in Figure 6.3.

FIG 6.3 - Later configuration of the lower shaft.

The formation of a zone of solid fused material in the centre of the furnace just above the tuyere zone (or ‘deadman’) provides support for the charge column above, but if excessive restricts the flow of material and gases, thus limiting the capacity of the furnace. The extent of this formation is controlled by limiting the separation between opposing tuyeres, as dictated by the blast velocity and hence, blast volume and the number of tuyeres installed. The lower sections of the ‘deadman’ may be composed of relatively high melting point iron and speiss, as well as unreacted coke. As the blast rate is increased, the raceway size increases and greater separation is required to prevent excessive overlap of adjacent raceways. Hence, for a given furnace the number of tuyeres must be reduced. To limit excessive overlap of opposing tuyere raceways, it is necessary to reduce

94




velocity by increasing the diameter of the tuyeres. However, it is necessary to ensure that there is a minimum degree of overlap under all operating conditions. There are thus a number of complex interactions in furnace design and a balance between the blast rate, tuyere diameter, the number of tuyeres, and the furnace width, which determine capacity and the efficiency of zinc recovery from the feed material. Assessment of raceway coverage is an important tool in design and can be done using an equation developed by Wlodarczyk, as published by Harris (1981). The standard furnace now has typically 16 tuyeres as single rows of eight on each side. The angle of the tuyeres to the horizontal is also a parameter of some importance, although it varies between individual plants. High angles will give greater interaction with the slag pool, but will induce high wear rates on the furnace hearth refractories. Initial furnace designs had a tuyere angle of 10° but this has increased to between 15 and 20°.

Accretions Accretions form in the ISF as zinc oxide deposits by reoxidation of zinc vapour in colder areas of the furnace, which essentially means the furnace walls. However, critical areas are above the upper surface of the charge, and include the furnace roof and condenser entry. Accretions grow and progressively reduce the cross-sectional area and capacity of the furnace. Extensive growth can also generate significant hoop stresses on the furnace and can result in splitting of the casing and lifting of upper sections of the shaft. Movements in the upper shaft can affect the attachment to the condenser and if movement is transferred to the condenser can severely impact on its performance by affecting the level of the lead pool in the condenser. Regular shutdown time for the cleaning of accretions is consequently necessary, and is usually on a three to six week cycle. Explosives may be used to remove accretions from the furnace shaft, but in other areas manual cleaning using jackhammers is necessary. As an alternative to blasting, the furnace may be burnt down to tuyere level with a high coke charge and the offtake temperature is allowed to rise so as to remove accretions. This action tends to marginally reduce the life of shaft refractories and may risk damage to the furnace casing.

Slag and bullion tapping The original ISF design was based on batch tapping with a variable level of slag and bullion in the base of the furnace, and most operating furnaces remain with this configuration. However, as indicated above, the interaction of the slag pool with the tuyere raceways is important in maximising the reduction and elimination of zinc from the slag. Hence, it can be argued that control of the slag level has important implications and steady control can only be achieved with continuous tapping. Varying levels of slag caused by intermittent tapping will also cause varying back pressure on the blast air supply, thus varying the blast rate. There are also implications for the separation of lead from slag. When tapping is at a high rate and flow through the forehearth is high, losses of lead, copper and precious metals in slag will be higher. There are thus significant advantages in adopting continuous tapping. However, experience has shown that although the above advantages can be achieved, a different flow regime and phase distribution occurs within the furnace, and this can lead to difficulties with the discharge of speiss from the furnace with slag. Iron speiss can be solid within the hearth and form a layer between the lead and slag at the end of the hearth opposite the tapping position. To avoid this it is necessary to achieve a high level of mixing and hence, strong interaction of the tuyere raceways with the slag pool, in turn requiring increased inclination of the tuyeres (Holliday et al, 1987).



95


The higher pressure operation of the ISF compared with the lead blast furnace means that weir heights in the forehearth are greater. The combination of taller weirs, lower bullion and slag flows, and higher heat losses around the weir in the ISF, results in more difficulty in maintaining the tapping hole open compared with a lead blast furnace.

COKE USE AND FURNACE CAPACITY Coke quality and uniformity is of importance and needs to be sufficiently strong to avoid degradation during handling and passage through the preheaters and furnace shaft. Volatile content should be low to avoid the presence of hydrogen, and reactivity should be relatively low, within the range of cokes normally available. The presence of hydrogen results in the formation of water vapour, which can react with zinc vapour in the upper shaft as the temperature falls to reform zinc oxide, and hence, volatiles should be low. As with the lead blast furnace a highly reactive coke causes the reduction zone to move to higher levels in the furnace and reduces the amount of carbon reaching the tuyere zone, thus lowering the temperature at the base of the furnace. This lowers the reduction of zinc from the slag pool and raises the loss of zinc in slag. The lower temperatures can also create difficulties with slag viscosity and the tapping of slag from the furnace. These effects can be overcome by increasing the coke consumption, but it is preferable to use a low reactivity coke. As indicated, coke is preheated to 700 or 800°C in a shaft in countercurrent flow to combustion products from the burning of condenser exit gas. Since the exit gas conditions for the furnace are relatively well defined in terms of temperature, CO and CO2 contents, the carbon consumption of the furnace will be directly related to oxygen input in blast air and top air. The furnace can in fact be defined closely in terms of its coke burning capacity. This can be increased per unit of oxygen supplied by preheating of the blast air and coke, and by oxygen enrichment of blast air, but a typical figure given average conditions as indicated above, is 1.2 kg of carbon per 1000 Nm3/h of oxygen input in blast air. The maximum blast air input volume for the standard furnace of 17.2 m2 upper shaft cross-section is close to 34 000 Nm3/h. The typical furnace blast maximum intensity is close to 2000 Nm3/h/m2 of shaft area. On this basis such a standard furnace has the capacity to burn: 34 000 × 24 × 21 per cent × 1.2 × 10 -3 = 205 tonnes per day of carbon For coke with a fixed carbon content of 85 per cent this represents close to 240 tonnes per day of coke as the maximum burning rate. Practical rates will be below this figure due to plant intensity and availability issues. The manner in which carbon is utilised for zinc production is affected by heat balance issues as well as chemical requirements for metal reduction. The heat balance is influenced by the amount of slag produced, the blast preheat and the preheat achieved in coke and sinter. Apart from the quantity of zinc oxide reduced, the chemical factors include the lead and iron contents of sinter and the moisture entering the furnace in blast air and in sinter or additives. Based on typical operating conditions as indicated above and a lead production of close to 50 per cent of the zinc production, the empirical formula used to relate carbon consumption to zinc production is given by Equation 6.1: Carbon consumed = 0.655 × zinc produced + 0.152 × slag produced

96


(6.1)



For a typical slag production of 0.9 tonne per tonne of zinc produced, Equation 6.1 indicates a carbon to zinc ratio of 0.8 or a zinc to carbon ratio of 1.25. Hence, the maximum zinc production from a standard furnace would be 205 × 1.25 = 256 tonnes per day. This represents the zinc which can be produced in furnace gases to the condenser and is reduced in practice by inefficiencies, loss of plant availability for cleaning and breakdown, and by reduced intensity of operations for various reasons. For 92 per cent recovery of vaporised zinc to product metal and 90 per cent plant utilisation this will equate to an annual capacity of 77 400 tonnes. Fuel can be injected through the tuyeres either in the form of coke fines, or as pulverised coal, fuel oil or gas. This basically replaces lump coke and is done as a cost saving measure, rather than affecting production capacity. However, there can be offsets with the use of hydrogen containing fuels such as oil, natural gas and coal, which give rise to problems during condensation due to the reaction of zinc and water vapour, which is much faster than the oxidation of zinc vapour by CO2. As an adjunct to this, the presence of water in blast air due to natural humidity can be a significant problem, firstly in consuming carbon in the lower shaft and later in causing condensation problems due to the presence of water vapour. Dehumidification of blast air can be justified in some circumstances. The impacts of a number of these factors were reviewed and modelled by Kellogg (1990). Oxygen enrichment of blast air gives a marginal reduction in carbon use per tonne of zinc produced but a significant increase in capacity of the furnace of the order of four per cent extra zinc production per one per cent oxygen enrichment in blast air.

REFERENCES AND FURTHER READING Harris, C F, 1981. Imperial Smelting Furnace hearth and raceway considerations, in Proceedings International Blast Furnace Hearth and Raceway Symposium, pp 13/1-13/8 (The Australasian Institute of Mining and Metallurgy: Melbourne). Holliday, R J, Fitzgibbons, D P, Arthur, A F and Bath, R A, 1987. Development of a continuous tapping system for the Imperial Smelting Furnace, in Proceedings Pyrometallurgy ’87 Symposium, pp 521-535 (Institution of Mining and Metallurgy and Institute of Metals: London). Kellogg, H H, 1990. A practical model of the imperial smelting zinc-lead blast furnace, in Proceedings Lead-Zinc ’90, pp 549-569 (The Minerals, Metals and Materials Society: Warrendale). Lee, R W, 2000. The continuous evolution of the imperial smelting process, in Proceedings Lead-Zinc 2000, pp 455-466 (The Minerals, Metals and Materials Society: Warrendale). Temple, D, 1980. Zinc-lead blast furnace – the key developments, Metallurgical Transactions B, 11(3):343-352.



97

CHAPTER 7 Direct Smelting Processes PRINCIPLES The sinter plant–blast furnace technology for lead production is highly capital intensive with a large amount of mechanical equipment, and generates large volumes of process gas and ventilation air to capture dusts and fumes generated from many points in the process. It is not simple to control or maintain at maximum efficiency and places a high reliance on experienced operators. In fact the sinter plant–blast furnace technology has not been economically viable for a new ‘greenfield’ smelter for many decades and alternative technologies have been sought to significantly reduce both capital and operating costs. Apart from excessive capital costs, the particular operating cost item of concern has been the use of metallurgical coke. Control of environmental emissions and occupational health matters are also significant issues and the meeting of strict environmental standards can be extremely difficult and costly for sinter plant and blast furnace technology. The sinter plant–blast furnace process involves two stages, firstly fully oxidising sulfides to oxides, followed by reduction of lead oxide to metal, as shown in Equations 7.1 and 7.2. The same end point can be reached by partial oxidation only, as indicated by Equation 7.3. PbS + 1.5O2 = PbO + SO2

(7.1)

PbO + CO = Pb + CO2

(7.2)

PbS + O2 = Pb + SO2

(7.3)


(7.4)

The reaction given by Equation 7.3 may in fact proceed by first oxidation of PbS to PbO (or PbSO4 at lower temperatures), followed by reaction of PbO (or PbSO4 ) with PbS, which is termed the ‘roast reaction’ as given by Equation 7.4. This approach was used in the earlier hearth processes for the smelting of high-grade lead material, but was never complete. Significant amounts of PbO were formed and lost in flue dusts, slags and residues, particularly if lower grade concentrates were processed. Complexities also arose with the presence of zinc, which formed zinc oxide and had to be removed in slag. Slag formation required the addition of silica as well as lime and iron. Silica tended to react with any oxidised lead to form stable silicates, further reducing the efficiency of the process. For reasons of poor metal recovery as well as serious emission problems, the hearth process was replaced by the sinter plant–blast furnace process. However, there are significant inherent energy savings in the use of this approach and attempts have been made to utilise the ‘roast reaction’ in the development of new smelting processes treating concentrates directly to produce lead metal. In the separate sinter plant–blast furnace process the heat of combustion of sulfides in the sinter plant is only utilised for drying and preheating feed before combustion, but is otherwise lost. If it could be effectively captured and transferred into the reduction stage then a substantial decrease in carbon fuel use could be achieved. A typical heat balance is shown in Table 7.1 for the sinter plant and blast furnace, together with a hypothetical balance for a single stage direct smelting process with the same end point slag and bullion conditions from an identical concentrate feed.



99

CHAPTER 7 – Direct Smelting Processes

TABLE 7.1 Heat balances for lead smelting (megajoules per tonne of lead produced). Component

Sinter plant

Blast furnace

Overall sinter (BF)

Direct smelting

Heat in feed

35

Heat in air/oxygen

25

240

35

35

80

105

6

4570

2850

3040

3040

581

3360

7750

3471

Heat lost in slag

1670

1670

Heat lost in bullion

150

150

150

Inputs

Heat of reaction Sulfides

4570

Fuels Total input

4630

Outputs Heat lost from sinter

2210

1970 1670

Heat lost in gases

2190

690

2880

1351

Equipment heat loss

230

850

1080

300

Total output

4630

3360

7750

3471

Note: Heat balances are based on processing a concentrate containing 60 per cent Pb, 20 per cent S, seven per cent Fe and six per cent Zn.

As shown in the illustration in Table 7.1, heat requirements from fuel can be substantially reduced if the separate stages can be combined. The characteristics of direct smelting can be simply demonstrated by reference to the equilibrium phase diagram for the Pb-S-O system as shown in Figure 7.1 for a temperature of 1200°C. This is the practical temperature required to achieve molten and free flowing reduced slags. (Oxidised slags with a high lead content are still fluid at temperatures below 1100°C.) As well as the equilibrium lead phases the diagram shows the equilibrium partial pressure of SO2 and indicates that even in a pure SO2 atmosphere at 1 atmosphere (atm), metallic lead can exist. An oxygen-based smelting process is required to produce SO2 at 1 atm partial pressure, whereas an air-based system will produce a partial pressure of SO2 closer to 0.1 atm. Some dissolution of sulfide in bullion will occur with a merging of the boundary between lead metal and lead sulfide. Lead sulfide contains 13 per cent S as the extreme, and two other positions of three per cent S and 0.3 per cent S in bullion are shown, as influenced by the partial pressure of free sulfur alone. As temperature is reduced the Pb to PbO and PbSO4 boundary moves down and to the left of the diagram to lower O2 and S2 equilibrium partial pressures, thus restricting the region where metallic lead can be formed and favouring the formation of lead sulfate or basic lead sulfates. In reality and under high temperature direct smelting conditions, sulfate phases shown in Figure 7.1 do not exist, and only two immiscible phases are present – namely sulfur containing bullion and a sulfur containing oxide slag (Willis, 1980; Ward, 1985). A lower oxygen potential will favour lower levels of PbO in slag, but for a given SO2 partial pressure such a change will result in higher levels of sulfur in bullion. Similarly higher oxygen potential will produce higher lead slags and lower sulfur levels in bullion. The normal regions for direct smelting as well as sintering and the blast furnace are also shown in Figure 7.1.

100




FIG 7.1 - Equilibrium diagram for the Pb-S-O system at 1200°C (Yazawa and Nakazawa, 1998; Matyas and Mackey, 1976).

Pure direct smelting utilising Equation 7.3 alone thus tends to produce high lead slags and a high sulfur content bullion in comparison with the sintering–blast furnace approach. The key issues for any direct smelting process are then the adequate elimination of sulfur and the production of a residual slag of low lead content. The principal reaction to be considered is Equation 7.4 for which the equilibrium constant is given by Equation 7.5: K=

a Pb . p SO 2

(7.5)

a PbO . a PbS

where: aPb = the activity of lead metal p SO 2 = the partial pressure of SO2 in equilibrium with the melt aPbO = the activity of PbO in the slag aPbS = the activity of PbS in the slag Assuming aPb = 1.0, then Equation 7.5 can be rewritten as: K. aPbO =

p SO 2 a PbS

indicating that the activity of PbO in slag is inversely proportional to the activity of PbS, or sulfur in the bullion, for a fixed temperature and partial pressure of SO2. Figure 7.2 illustrates a form of this relationship between the lead content of slag and the sulfur content of bullion, taken from a range of actual smelting plant data (Zaitsev and Margulis, 1985).



101


45% 40% 35%

Lead in slag

30% 25% 20% 15% 10% 5% 0% 0.00%

0.50%

1.00%

1.50%

2.00%

2.50%

3.00%

3.50%

Sulfur in bullion

FIG 7.2 - Relationship between S in bullion and Pb in slag for single stage direct smelting (after Zaitsev and Margulis, 1985).

It should be noted that the equilibrium constant K in Equation 7.5 is temperature dependent, as are the activity coefficients of PbO and PbS, which has ramifications for the yield of volatiles from the direct smelting process. For a slag of given PbO and PbS contents, the activity of PbO will increase slightly with temperature while that of PbS will decrease significantly – as demanded by changes in the equilibrium constant and activity coefficients with temperature. Calculated activities for a slag containing 33.3 per cent Pb are illustrated in Figure 7.3. This also shows that a slag of this composition cannot exist below 1070°C where the calculated activity of PbS is greater than 1.0, an actual impossibility since a separate liquid PbS phase would have formed below this temperature. 1.6 1.4 1.2

Activity

1 Act. PbO

0.8

Act. PbS

0.6 0.4 0.2 0 1000

1050

1100

1150

1200

1250

1300

Temperature (°C)

FIG 7.3 - Slag activities at 33.3 per cent Pb in slag (Ward, 2007).

102




The volatilisation of a particular species (Pb, PbO, or PbS) is determined by the vapour pressure of the species and its activity as a function of temperature. The vapour pressure of individual pure components are given in Table 7.2 and the actual calculated vapour pressures of components above a slag containing 33.3 per cent Pb, as illustrated in Figure 7.3, are given in Table 7.3 for temperatures above the limiting minimum for a single slag phase. TABLE 7.2 Vapour pressures of lead compounds (vapour pressures in atm). Temperature ( C)

900

1000

1100

1200

1300

0.0036

0.0223

0.106

0.4087

1.327

PbO

0.00060

0.00357

0.01632

0.0608

0.1915

Pbl

0.00048

0.0019

0.0066

0.0192

0.0488

PbS

TABLE 7.3 Vapour pressures of lead compounds above a 33.3 per cent Pb slag (vapour pressure in atm). Temperature C)

1100

1200

1300

PbS

0.072

0.098

0.173

PbO

0.0016

0.0064

0.0208

Although the activity of PbS in the slag decreases with temperature there is still an increase in vapour pressure and there will be a marked increase in volatilisation with an increase in temperature from 1200 to 1300°C leading to the generation of fumes and dusts. This will favour smelting at the lowest practical temperature, conducive to the formation of fluid slag and above the critical point of separation of a separate PbS matte phase as indicated above. Minimisation of gas volumes will reduce losses of lead from the smelting bath by volatilisation, and hence oxygen based smelting or high levels of oxygen enrichment are desirable. It also follows that low-grade concentrates with substantially higher S to Pb ratios will yield correspondingly higher SO2 generation and gas flows. Hence volatilisation of lead will be proportionally higher and in some direct smelting processes may result in excessive losses or circulating loads. Low-grade concentrates will also generate more slag resulting in greater losses of lead as a result of high lead slags.

Direct smelting process configuration Meeting the requirements of high sulfur elimination, low lead slags and minimal volatilisation in a direct smelting operation utilising the ‘roast reaction’ alone in a single stage is relatively impractical, and involves unacceptable compromises between these requirements. The only practical approach is the use of a two stage process, in which extensive oxidation of sulfide occurs in the first stage (oxidation stage) to eliminate sulfur, followed by reducing conditions in the second stage (reducing stage) to lower the level of lead in slag. Carbon must be used for promoting reducing conditions in the second stage if the level of sulfur in bullion is to be low. The use of oxygen can also minimise the amount of lead volatilisation. The oxidation stage can be structured to produce a lead-rich slag and some lead bullion by utilising the roast reaction to some extent. The lead-rich slag can be subsequently reduced to recover the remaining lead as bullion. With integration of the two stages, significant fuel savings can be achieved. These principles form the basis of a number of direct smelting processes.



103


The Boliden lead process was the first of the direct smelting processes in which the first stage attempted to utilise the roast reaction to the maximum extent and was carried out in an electric furnace as an alternative heat input device to minimise gas volumes and lead volatilisation. Sulfur elimination was poor, giving high levels of sulfur in bullion, which was subsequently oxidised in a converter to lower the sulfur content. The slag remained high in lead as well as zinc and was treated in a slag fumer. Lead volatilisation was high and the process was effective only for the smelting of high-grade concentrates. Subsequent direct smelting developments have aimed for higher sulfur elimination in the first stage to give high lead first stage slags and include the Kivcet process, the Queneau-SchuhmannLurgi (QSL) process, the Kaldo process and top submerged lancing (TSL) used in the Isasmelt and Ausmelt processes, and are detailed in the following sections of this chapter.

The role of iron in refractory protection The equilibrium composition of iron in the slag is also of importance and is controlled in bath smelting processes to ensure some precipitation of magnetite (or zinc ferrite) onto the refractories in order to ameliorate the effects of highly aggressive litharge slags, produced under the oxidising conditions of the first direct smelting stage. To provide effective protection, the refractory surface temperature should be below the magnetite liquidus temperature. Although special grade refractories such as fused chrome magnesite are used for these applications, they are not completely inert to litharge attack. The chrome magnesite spinel itself is virtually immune to litharge attack, but any residual magnesite or traces of silica will accumulate at grain boundaries and will be vulnerable, leading to disintegration of the refractory structure. The phase diagram for the Fe-O-S system is shown in Figure 7.4, corresponding with Figure 7.1, and indicates that the direct smelting region does have the potential to produce magnetite. Figure 7.4 is, however, a simplification since it does not account for the solubility of Fe3+ in slags and the presence of solid species with highly variable Fe2+ and Fe3+ contents (Jak, Zhao and Hayes, 2000).

FIG 7.4 - Equilibrium diagram for the Fe-O-S system at 1200°C.

104




At the required oxygen potential for the formation of magnetite to occur and provide refractory protection there will generally need to be some compromise between the PbO content of slag and the residual sulfur level in bullion.

THE BOLIDEN LEAD PROCESS The Boliden lead process applied the ‘roast reaction’ in a liquid slag phase in an electric furnace. It was part of a copper and lead smelting complex at the Ronnskar smelter in northern Sweden. It was also integrated with a slag fumer to recover zinc and lead and later to some degree with the Kaldo top blown rotary converter (TBRC) discussed below, which was initially used to treat dust and fumes containing lead that could not be handled by the electric furnace (Elvander, 1965). The process was developed to smelt local high-grade concentrates and was not competitive for the smelting of low-grade materials, as indicated above for application of the roast reaction. Design concentrate feed for this process was 70 - 75 per cent Pb, 14 per cent S, one to three per cent Zn, two per cent Fe, five to six per cent SiO2. The smelter was closed when supply of these high-grade concentrates was exhausted and smelting continued using the Kaldo TBRC alone. Figure 7.5 illustrates the original configuration of the integrated lead smelting operations at the Ronnskar smelter. The electric furnace has subsequently been closed with lead operations confined to the TBRC. Lead concentrate Fluxes

Dryer Air/oxygen Electric Furnace Air

Gas to Sulfuric Acid

Gas Cleaning

Bullion

Converters

Slag Fumer

Gas Cleaning

Clinkering Kiln Lead Bullion

Dusts

Slag to waste Lead fume

Zinc oxide to Zinc Smelter

Gas to Sulfuric Acid

Lead scrap Copper plant dusts Slag Kaldo TBRC

Gas Cleaning

Dusts Dusts

FIG 7.5 - Original lead smelting at Boliden – Ronnskar smelter.



105


The electric furnace was a 6 MW unit with four in line Soderberg electrodes. High-grade concentrates, limestone, iron cinders and recycle dusts and fumes were predried to two per cent moisture and fed into the electric furnace via two flash burners located between electrodes one and two and between electrodes two and three. The solids entered through a central vertical pipe surrounded by tangential jets of oxygen enriched air creating a vortex and flame at around 1150°C. Oxygen input was less than stoichiometric and around 70 per cent of sulfur was oxidised in the burners. Reactions in the flame formed lead bullion and a slag/matte, which fell into the slag bath contained in the bottom of the electric furnace where they separated into an upper slag and lower bullion layer. Further oxidation of sulfur continued in the slag bath according to Equations 7.6 and 7.7: PbS + 2PbO = 3Pb + SO 2

(7.6)

PbS + PbSO4 = 2Pb + 2SO2

(7.7)

Coke was also added to the furnace as fines at one per cent of the concentrate input to assist in reducing the lead level of slag, which contained around 4.1 per cent Pb as well as zinc. Carbon was also added from the electrodes and represented around 1.3 per cent of the concentrate input. Gas from the furnace contained copious fumes and dusts at around 40 per cent of concentrate feed, and around eight per cent SO2. Dust and fume collected from gas cleaning was returned to the electric furnace and represented a significant recycling load. Bullion was tapped continuously via a siphon from the centre of the furnace and slag was tapped intermittently from one end. Lead bullion contained around three per cent S, which was transferred to either of two small batch operated Pierce Smith converters, blown with air to remove sulfur. Slag from the converters and dusts collected from converter gases were also returned to the electric furnace. A copper matte was periodically recovered from the converters and the bullion was sent to the lead refinery. Slag from the electric furnace was too high in lead (and zinc) for disposal and was further processed in a conventional slag fuming furnace, where zinc and lead were fumed off by the addition of coal. The fume produced was unsuitable as feed for a zinc smelter and was treated in a clinkering kiln to fume off halides and lead and leave a zinc oxide clinker now suitable as feed to an electrolytic zinc plant. The fume produced from the clinkering kiln contained over 50 per cent Pb, but was unsuitable for recycling to the electric furnace due to the high concentration of impurities. This material was processed in the Kaldo process utilising a batch operated top blown rotary converter (TBRC), and was later installed at the Ronnskar smelter. This plant also treated lead-rich converter dusts from the copper smelter. Capacity of the electric furnace was around 55 000 t/a of lead bullion and electrical energy consumption was around 900 kWh per tonne of lead bullion produced.

THE KALDO PROCESS (TOP BLOWN ROTARY CONVERTER – TBRC) The Kaldo process was also an innovation developed at the Boliden Mineral AB – Ronnskar smelter for the processing of difficult fumes and dusts from copper smelting and from the lead slag fuming operation as shown in Figure 7.5. The process proved to be highly flexible and capable of handling a wide variety of lead bearing materials including sulfate residues and was later adapted to the smelting of low-grade concentrates. It is now predominantly used for the processing of secondary materials.

106




Gas to Acid Plant

Lead concentrate or lead rich dusts

Filter

Fluxes

Venturi Scrubber

Rotary Dryer

Oxygen Screen

Rotary Converter

Air Pnuematic Conveyor

FIG 7.6 - Layout of Boliden Kaldo process plant.

A layout of the Kaldo plant is given in Figure 7.6 (Nystedt, 1980; Pazour, 1982). The Kaldo furnace is a batch operated rotary converter vessel with an inside diameter of 2500 mm and a length of 6500 mm, with a working volume of 12.5 m3. It is lined with refractory brick and the working volume increases as the brickwork wears. The initial charge is around 90 tonnes. The brick lining consists of a 120 mm backing layer plus a 450 mm wear layer of chrome magnesite brick (15 18 per cent Cr2O3), which reduces back to 50 mm before replacement. Thickness is regularly measured using a laser beam and rebricking frequency is every two to three months, representing lost time of around five per cent. The relatively low lost time is due to the use of changeover converter shells to allow re-bricking off-line. Oxygen or enriched air is blown through a lance and impinges on a pool of material in the vessel. The main lance is water-cooled and designed with multiple inner tubes to inject solid feed material with air as well as 98 per cent oxygen. A flame is formed which impinges on the slag pool within the vessel. Smaller lances are used to inject flue dusts or fuel oil. For concentrate smelting the concentrates are pneumatically fed through the inner pipe of the main lance with air and oxygen fed through the outer pipe. The converter spins around its longitudinal axis at a variable speed of between 0.5 and 30 rev/min, but normal speeds are in the range of five to 15 rev/min. The vessel is held within a frame which can be tilted for filling and emptying. The whole assembly is enclosed within a box-like enclosure fitted with top and front access doors. Material can be charged into the vessel in a vertical position either from a batch hopper or skip by overhead crane. Fluxes, coke, return slag and scrap are added this way. The lower part of the enclosure houses a number of flat top rail cars carrying ladles for either slag or bullion. The total enclosure is vented through a bag house at around 80 000 Nm3/h and is maintained at a negative pressure. Process gas at around 20 000 Nm3/h is cooled and scrubbed in a venturi before transfer to a sulfuric acid plant. The processing of an intermittent gas stream, both in terms of volume and SO2 content is a major problem for an acid plant, but is uniquely handled at the Ronnskar smelter by the use of a cold sea water absorption system, as shown in Figure 7.7.



107


To acid plant

To Stack Bypass during smelting Bypass for heating stage

Wet electrostatic precipitator

Absorption column

Stripping column

SO2 – water storage Hot air or air + steam

Lime

Neutralisation Cold water supply

FIG 7.7 - Gas handling system for discontinuous SO2 production.

This system absorbs SO2 from the variable gas stream into cold water, which is stored and then continuously stripped by heating to 60°C before return to the sea. The system is only practical because of the unlimited supply of cold low-salinity sea water available to the smelter. Alternative processes are available using amine solutions for the absorption of SO2, which can be recovered by steam stripping. However, the inventory cost of the amine reagent can be quite high. In other situations the handling of intermittent gas flow is a major impediment for the Kaldo Process. The furnace operating sequence involves charging the furnace with pelletised feed, recycled slag and lime, and firing with an oil–oxygen burner. The charge is heated slowly under slow speed rotation. The melting stage is completed when the SO2 content of the off-gases declines. The reduction stage follows with fine coke injected through the lance and with the oil burner continuing to operate. The reduction stage is completed when the lead content of slag falls below one per cent. The temperature is then raised to 1150°C and most of the slag is emptied from the furnace while retaining the lead bullion. A new charge is then added and the above sequence repeated. After discharge of the second slag load, the bullion is partially refined by the addition of iron turnings to combine with arsenic and copper to form a speiss. Oil burning is continued to retain the temperature above 1000°C. This stage is relatively rapid and is completed when suitably low levels of arsenic and copper are reached in the bullion. The speiss containing 30 per cent As, 50 per cent Fe, 12 per cent Cu, 3.5 per cent Pb and 1.0 per cent Sn is separately poured from the furnace. The lead bullion is then poured into ladles and is allowed to cool to 500°C, at which point any residual speiss settles to the top of the molten lead as a crust. The underlying lead bullion is poured into a second ladle for transfer to the lead refinery. The composition of lead bullion at this stage is <0.1 per cent As, <0.1 per cent Cu, 0.1 - 0.5 per cent Bi and 0.2 per cent S. Close to 50 per cent of the arsenic input and 90 per cent of the copper input reports to speiss. Of the arsenic input 35 per cent reports to smelter gas, and 15 per cent to slag.

108




Final slag is below one per cent Pb and contains between seven and 15 per cent Zn and around 22 - 29 per cent Fe, 23 - 25 per cent SiO2 and 20 per cent CaO, and is sent to a slag fuming furnace to recover zinc. The total smelting cycle described above takes 12 hours with the order of two hours for each stage of smelting (two stages), reduction (two stages) and refining (one stage). The remaining time is used for various transfer operations. Dusts produced contain of the order of 65 per cent Pb and six to ten per cent Zn, and can contain around 20 per cent of the lead input. Capacity of the standard furnace has been given as close to 55 000 t/a of lead bullion from average lead concentrates at 62 per cent Pb, with typical fuel and consumption data as:

• oxygen use

(m3/t of concentrate feed)

120

• coke use

(kg/t of concentrate feed)

30

• oil use

(L/t of concentrate feed)

6.5

• power use

(kWh/t of concentrate feed)

70

The TBRC is a highly flexible process with the capability of handling a wide range of primary and secondary materials. It is also possible to use the converter for fuming zinc from lead slags. The disadvantages are the high level of mechanical equipment with attendant maintenance issues, and the intermittent nature of SO2 rich gas production, which significantly adds to capital equipment requirements.

THE KIVCET PROCESS The Kivcet Process (an acronym in Russian for oxygen flash cyclone electrothermal process), was developed by the Vniitsvetmet Institute in Kazakhstan in 1967 for the treatment of both copper and lead concentrates (Sychev et al, 1985). A 25 t/d pilot plant was built and operated for a number of years followed by a plant to treat 500 t/d of mixed copper-zinc concentrates at Glubokoe in 1970. That plant produced a copper matte and zinc rich slag for slag fuming in the electric furnace to recover zinc oxide. Initial developments were based on the use of a cyclone smelting chamber mounted on the top of the electric furnace in which the principal reactions take place in a thin slag film running down the walls of the chamber. Because of the aggressive nature of lead rich slags towards refractories and the much slower burning rates of lead concentrates compared with copper, attention was shifted to the use of the flash burner shaft. A low sulfur bullion was achieved by chilling the furnace hearth to lower the bullion temperature to about 500°C, giving rise to the formation of a lead–copper matte phase as well as slag and bullion. The first full size commercial lead furnace was constructed for Kazzink at Ust Kamenogorsk in 1985, treating 450 t/d of lead concentrates. Apart from a small furnace built in Bolivia but never operated, two other lead units have been built, one at the Porto Vesme smelter complex in Sardinia in 1987 (Perillo et al, 1990) and one at Teck-Cominco’s Trail smelter in BC Canada, completed in 1997 (Walker, 1998). The Kivcet furnace design is a combination of an oxygen based flash smelting shaft and an electric furnace for slag cleaning and settling. Significant features are the partition wall between the smelter shaft and the electric furnace, and the gas off-take shaft and waste heat boiler, which are designed to handle high loads of sticky dusts. A schematic of the furnace is shown in Figure 7.8.



109


Gas Off-take Shaft

Boiler Convective Section

Feed Blend

Coal or Coke

To Gas Cleaning Oxygen

Coke Gas to ZnO Collection

Flue Dusts for Recycle

Smelter Shaft

Coke ‘Checker’

Afterburner

Electric Furnace

Slag layer Bullion layer Partition Wall

Slag /Matte

Lead Bullion

FIG 7.8 - Schematic of the Kivcet furnace.

Feed consists of lead concentrates, lead sulfate or oxide residues, fluxes and some coal or coke. There are no constraints on the proportion of sulfides and oxides in the feed mix and fuel value can be adjusted by the addition of coal or coke. High levels of oxidised lead in feed requires a substantial increase in the fuel load. It is found necessary to prepare feed materials by wet mixing followed by drying in a rotary kiln dryer. This produces micro agglomerates of the feed components in close physical contact, which promotes reactions between PbS and oxidised components of the feed mix such as PbO and PbSO4. In so doing it minimises the volatilisation of PbS and the generation of fine dusts as it is oxidised in the furnace atmosphere. Generally the target moisture level of the feed is around one to two per cent. Where a high proportion of residues are used in the feed, carbon addition is needed for reduction, particularly of ferric iron. This should be included in the wet feed mix prior to drying, whereas additional coal as fuel should be added separately to the burner. Coke is only added to the electric furnace if zinc fuming is required. The gas stream would then require afterburning, cooling and filtration to recover zinc oxide.

Smelting shaft The feed is dropped into a cyclone flash burner into which industrial grade oxygen (95 per cent) is injected tangentially, surrounding the feed. A flame is produced projecting downwards into the smelting shaft, with a temperature between 1350 and 1420°C, controlled by varying oxygen or feed flow rates. Within the flame and reaction shaft, carbon is partly burnt, sulfides are oxidised fully, sulfates are decomposed in accordance with Equations 7.8 to 7.14, and the oxides are formed into a molten slag phase.

110




PbS + 1.5O2 = PbO + SO2

(7.8)

ZnS + 1.5O2 = ZnO + SO2

(7.9)

FeS + 1.5O2 = FeO + SO2

(7.10)

PbSO4 = PbO + SO2 + 0.5O2

(7.11)

ZnSO4 = ZnO + SO2 + 0.5O2

(7.12)

C + O2 = CO2

(7.13)

CO2 + C = 2CO

(7.14)

There is a view from the process developers that lead metal is initially formed on the galena surfaces and is then oxidised to PbO, although other mechanisms can result in the same outcome. The composition of particles in the burner shaft as a function of residence time or height in the shaft would then be expected to show a progressive decrease in the PbS content, and initial rise then decline in the metallic lead content and a progressive rise in the PbO content as oxidation is completed. Figure 7.9a illustrates measured composition as a function of shaft height. Figure 7.9b shows the corresponding temperature (Lyamina and Shumskii, 2006). This demonstrates that lead is mainly oxidised to PbO in the shaft. Depending on the form of lead-rich residues in the feed there may be varying amounts of complex materials such as ferrites and jarosite from zinc plant leach residues, which also need to be decomposed, and as indicated above coal as a reductant to the feed mix is required in addition to coal as fuel. The rate of decomposition is generally lower than the sulfide oxidation reactions and longer residence times are required in the smelting shaft with the treatment of these materials. Residence times are set by the gas velocity, hence cross-section of the shaft and the length of the shaft. The original Kazzink lead furnace was extensively modified to handle a high proportion of residues (75 per cent), and the shaft length extended from 3 to 5 m. In general the ratio of feed rate to smelting shaft cross-section or ‘specific feed rate’ is relatively fixed at between 45 and 50 t/d per square metre of shaft cross-section. There can be up to four burners at the top of the shaft and the feed rate per burner is of the order of 300 to 350 t/d.

Gas offtake system The molten phase produced in the smelting shaft falls to the bottom and separates from the gas phase, which passes under a dividing wall and into the vertical off-take shaft. The off-take shaft is lined with a membrane boiler wall, heated by radiation from the gas. Gas is normally cooled to around 700°C, but preferably less than 800°C before entering the convection section of the boiler. Although the oxygen potential in the smelting shaft is sufficiently high to minimise excessive lead sulfide volatilisation, there is still a significant level of volatilisation of lead in various forms with the high flame temperature. As a result, as the gas is cooled a significant dust and fume load is developed. It is estimated that 75 per cent of the lead in dusts arises from volatilisation in comparison with mechanical entrainment of feed material. As the gas cools, molten PbO will be condensed in the shaft and will be sulfated by surplus oxygen in the flue gas (or injected into the uptake shaft) forming sticky material, which will readily adhere to the colder boiler walls. The uptake shaft is designed to handle



111


PbS

A

Pb

PbO

40 35

Composition (%)

30 25 20 15 10 5 0 0

0.5

1

1.5

2

2.5

2

2.5

3

Shaft height (m)

Temperature (°K)

B

1800 1700 1600 1500 1400 1300 1200 1100 1000 0

0.5

1

1.5

3

Shaft height (m)

FIG 7.9 - Lead deportment in shaft material (A) and temperature (B) as a function of shaft height.

this by the use of spring hammers to dislodge deposits, causing them to fall back into the slag bath at the base of the shaft. Gas below 800°C is no longer able to form sticky accretions and can be handled in the convection section of the boiler where it is cooled to around 350°C. The high dust loadings at the base of the uptake shaft act as a scrubber for volatile species (such as arsenic and tin) in the flue gas, which can drop out and enter the bath as involatile species (eg lead arsenate) and thus report to lead bullion. A dust drop-out section is also included between the uptake shaft and the convection section. It is important that there are no horizontal flat sections where fume can collect and especially where the temperature is high enough so that the PbO content can fuse the fume and dusts into inaccessible accretions. Total dusts passing through the boiler and collected in the gas cleaning train typically represent between five and 15 per cent of the feed and contain over 50 per cent Pb, around ten per cent Zn depending on feed composition, eight per cent S and relatively high cadmium. Recirculation of all the dust would cause cadmium levels to build up and hence part of the dusts are separated and leached to remove cadmium before return to the furnace.

112




The SO2 content of the gas will vary widely depending on the proportion of lead concentrates in the feed compared with residues. With a concentrate feed the gas may be 45 per cent SO2, whereas with predominantly oxide residues as feed the level will fall to 15 per cent. Steam production is around 0.4 t/t of feed. Gases leaving the boiler are passed through an electrostatic precipitator before scrubbing prior to transfer to a sulfuric acid plant. The electrostatic precipitator usually contains four fields in series, with two parallel lines.

Coke layer or ‘coke checker’ A key feature of the Kivcet furnace is the layer of incandescent coke floating on the slag bath at the base of the smelting shaft. The molten phases from the flash burner impinge on the coke layer and trickle through into the slag bath beneath. Reduction occurs within this layer and the main reactions involved are shown in Equations 7.15 to 7.18. Also shown below is the heat of reaction at 1200°C. PbO

+ CO = CO2 + Pb

Fe2O3 + CO = CO2 + 2FeO ZnO

+ CO = CO2 + Znv

C + CO2 = 2CO

-72.7 kJ/mole

(7.15)

-35.7 kJ/mole

(7.16)

+182.7 kJ/mole

(7.17)

+170.9 kJ/mole

(7.18)

It has been demonstrated that the reaction mechanism is basically via gaseous reduction of the PbO slag with CO to form CO2, which in turn reacts with the coke to form more CO. The overall reduction is mildly endothermic even at high CO2:CO ratios in the gas phase. Hence the maintenance of the coke temperature must come from a combination of excess heat in the melt and coke arriving on the surface of the layer, from combustion of CO immediately above the layer by means of excess oxygen in the furnace atmosphere or by deliberate oxygen injection, and from the underlying slag bath temperature by circulation from the electric furnace. The coke layer is maintained at about 100 to 150 mm thick with a temperature of 1100 to 1200°C. The thickness can be measured on a regular basis using a steel bar probe. Extra carbon is added to the coke layer as unburned coke from the burner to make up for consumption by the reduction reactions. Coal or coke at a particle size in excess of 5 mm is usually added for this purpose, and is carbonised in the shaft but is not burned to any significant extent. Since the shallow coke layer has minimal strength requirements, as distinct from a shaft furnace, it is not necessary to conform to normal rigid coking coal specifications. On the other hand it is still necessary for the carbon to arrive at the coke layer in granular form. If it decrepitates to a fine powder under the thermal shock of entering the burner flame, it is more likely to be carried out in the gas stream than enter the coke layer. It would appear that any low volatile coal with reasonable coking properties is a suitable reductant for the Kivcet process. Lead oxide reduction is relatively rapid and efficient, whereas zinc oxide is much slower and only a small proportion forms zinc vapour to report in dusts and fumes. Ferric oxide reduction is more endothermic than PbO reduction, and if present in significant amounts such as from zinc plant leach residues, will lower the temperature of the coke layer and reduce the efficiency of PbO reduction. To



113


compensate, additional finely divided coal can be added to the feed to both raise the flame temperature and achieve some further reduction before the coke layer. Also, oxygen can be injected just above the coke to burn emitted CO to CO2 (Slobodkin and Kluev, 1998).

Electric furnace Metallic lead formed in the coke layer rapidly falls through the slag to form a bullion layer on the hearth of the furnace. The slag and bullion flow under a water-cooled copper partition or curtain wall immersed into the slag bath, then into the electric furnace. Electrical energy input maintains the slag temperature at around 1200 to 1300°C and power input is controlled by the depth of immersion of the electrodes into the slag layer. Slag circulates from the electric furnace into the base of the burner shaft and thus also provides heat to the coke layer, helping to maintain its operating temperature. The original Kazzink furnace had six in line electrodes and it was intended that a significant level of zinc fuming would be carried out. Since this was not particularly efficient in comparison with a separate slag fuming furnace, subsequent designs have substantially reduced electric furnace bath volume with three electrodes in line. Small quantities of coke are added to complete lead reduction, some zinc fuming can occur and a zinc oxide fume is collected from electric furnace off-gases. Power input is of the order of 150 kWh per tonne of furnace feed through prebaked graphite electrodes; coke addition can be up to 5 kg/t of furnace feed and electrode carbon used is between 1 and 2 kg/t of furnace feed. The total bath depth in the electric furnace is between 1500 and 2000 mm, with the lower bullion layer at around 900 mm. The bullion is kept at a significantly lower temperature than the slag to protect the hearth, and is tapped through an underflow siphon at the end of the furnace. Slag is tapped intermittently. In some operating conditions a copper matte can be formed and can be tapped separately as in Russian plant practice. This requires the temperature of the bullion to be lowered to around 500°C, but enough copper remains in the bullion to still require a copper drossing operation and there are significant losses of copper in slag. In addition the tapping of matte can be difficult. If the recovery of copper is a significant economic issue then it is preferable to operate at higher bullion temperatures to avoid matte formation and retain copper in the bullion for subsequent recovery by copper drossing. As an example, the Teck-Cominco plant operates at bullion temperatures at 850 to 950°C in order to prevent matte formation (Ashman et al, 2000). Electric furnace gases generally pass through an after-burner to burn CO and to oxidise any zinc to zinc oxide fume, are then cooled and cleaned by passing through a scrubber or bag house filter. Although it is possible to recover the zinc content of the slag by fuming in the electric furnace with coke additions, this requires a much larger furnace volume and is not generally regarded as competitive with separate conventional slag fuming.

Comparative plant details Table 7.4 provides details of a number of Kivcet furnaces and operating conditions. From Table 7.4 bullion with less than 0.5 per cent sulfur can be achieved as well as slags containing around one per cent lead. However, where slag fuming facilities are available for zinc recovery, low lead levels can be sacrificed for furnace throughput as in the case of the Teck-Cominco owned Trail plant. The lead content of slag will be almost fully recovered with the zinc oxide fume and will be recycled in the form of zinc plant leach residues containing lead as lead sulfate.

114




TABLE 7.4 Kivcet furnace details. Plant

Kazzink (original)

Kazzink (upgraded)

Porto Vesme

Trail (Teck-Cominco)

86

57

88

120

Dimensions Hearth area (m2) Length (m)

19.2

13

20

24

Width (m)

4.5

4.5

4.5

5

Smelting shaft area (m )

12

12

14

30

Shaft height (m)

3

5

6.2

7.2

2

2

2

4

2

Number of burners 2

Uptake shaft area (m )

10

10

11

25

Electric furnace area (m2)

62

34

54

55

Number of electrodes

6

3

3

3

600

600

800

900

Feed rate (t/d)

480

600

720

1400

Specific feed rate (t/d.m2 of shaft area)

40

50

51

47

72%

25%

73%

27%

Oxygen (Nm3/t of feed)

185

268

165

208

Coke use (kg/t of feed)

17.5

25

45

110 (coal)

Electrode power (kWh/t)

275

160

170

110

Electrode diameter (mm) Performance

Feed (% concentrates)

Other power (kWh/t feed)

60

Electrode carbon (kg/t)

2.1

1.0

Product streams Lead bullion (t/d)

280

%S %Cu

0.8

2.1

Matte (t/d)

300

350

0.4

1.0

0.6

2.0

10

9

%Pb

12

11.4

40

31.4

%Cu

23

35

25

45.8

20

5

0.33

%Fe %S

21

20

Slag (t/d) %Pb

0.4 - 2.0

1.2

12

15.8

120

650

4

5

%FeO

30

26

28

%SiO2

22

22

21

20

12.7

9

17.8

400

360

22 - 25

12 - 15

%CaO %Zn Smelter gas volume (Nm3/t of feed) %SO2


14 12 - 18

11.4

200 - 250 40 - 50

15 - 17


115


Furnace construction The furnace is built within a steel frame to support the side walls. Structural support and expansion forces are transmitted to the frame through metal springs allowing for movement from thermal expansion. The hearth is an inverted arch constructed on a steel plate base and composed of a lower graphite brick layer and an upper chrome magnesite brick layer, separated by a stainless steel sheet. The lower support plate is cooled by a series of channels supplied with air from variable speed fans. Air cooling is used rather than water which would present a significant danger. The lower refractories are cooled to below the melting point of lead so that leakage cannot occur. The side walls of the furnace are made up of water-cooled copper jackets, lined with chrome magnesite brick. Each jacket is individually cooled and outlet water is kept below 50°C. The furnace roof is a suspended arch made from chrome magnesite brick. The internal separation walls are similarly constructed of water-cooled copper elements lined with brick. The gap between the slag layer and the bottom of the wall separating the smelting and gas off-take shafts is critical, and accretion build-up at this point is a significant issue. The Kivcet process is robust and highly flexible in the range of possible feed materials, from high-grade concentrates through to secondary leach residues. It produces a low volume of smelter gas rich in SO2 and suitable for sulfuric acid production. It is also well contained with minimal opportunity for environmental emissions, particularly in comparison with sinter plant–blast furnace technology.

THE QUENEAU-SCHUHMANN-LURGI (QSL) PROCESS As discussed in ‘Principles of Direct Smelting’, a significant problem with direct oxygen smelting using the roast reaction (Equation 7.3) has been the volatility of lead sulfide at reaction temperature and the associated lead content of the slag produced. One means of overcoming these difficulties was proposed by Queneau and Schuhmann (1974), in which the reactions are conducted within a molten slag bath overlying a layer of molten lead. Oxygen is injected into the base of the reactor in much the same way as in a bottom blown converter. Oxygen rapidly reacts with the molten lead to form PbO, which transfers to the slag layer and reacts with added lead sulfide to form SO2 and lead metal as in Equation 7.19: PbO + PbS = 2Pb + SO2

(7.19)

The slag formed is further treated by carbothermic reduction in an effective second stage to reduce its lead content before discharge from the reactor. A long cylindrical reactor was proposed with feed and oxygen injection at one end and coal injection at the other end. Slag flowed from the oxidation end to the reduction end and out of the vessel, whereas lead bullion flowed in the reverse direction to exit at the oxidation end. The process was developed by Lurgi Chemie und Huttentechnik and hence the name QSL Process. A demonstration unit was constructed in Duisburg, Germany in 1981, followed by the construction of commercial plants for Berzelius Metalhutten at Stolberg, Germany, for Korea Zinc at Onsan in South Korea and at Baiyin in China. A fourth unit was constructed for Cominco at Trail, British Columbia but was unsuccessful and was replaced by a Kivcet unit. The failure of the Cominco plant has been attributed to the attempted use of natural gas for slag reduction rather than coal as used for the other plants. A schematic of the QSL reactor is given in Figure 7.10.

116




Boiler

Feed hopper Gas to Acid plant

Air or enriched air Gas Off-take Shaft

Slag layer

Bullion layer Oxygen

Bullion output

Oxidation Section

Coal and oxygen

Slag

Reduction Section

FIG 7.10 - The Queneau-Schuhmann-Lurgi (QSL) reactor.

Metallurgy and reactor design In the oxidation section of the reactor, feed materials including concentrates, residues, fluxes such as silica and lime, and coal as an auxiliary fuel are dropped into the slag bath. Predrying of these materials is unnecessary. Oxygen is injected into the base of the furnace through tuyeres, which have a surrounding shroud of nitrogen to protect the tuyere from the corrosive effects of the PbO formed. Oxygen passes into the molten lead layer, rapidly forming PbO and heat. The PbO rises into the slag layer where it reacts with sulfides and sulfates from the feed to produce SO2 and metallic lead, which falls back into the bullion layer. The lead bullion under highly oxidising conditions is relatively low in sulfur and can be directly tapped from the vessel at around 1100°C through an inverted siphon. Sulfur levels in bullion of less than 0.5 per cent can be achieved. The slag at this point is high in PbO at around 40 per cent and flows to the reduction section of the reactor. The oxidation section temperature is controlled by varying the input of oxygen and auxiliary fuel as required. The PbO content of the slag at this point is adjusted by the feed to oxygen ratio. Oxygen, or oxygen enriched air is added to the gases leaving the oxidation section to remove any volatilised sulfides. The gas then passes through an uptake shaft lined with a membrane boiler wall, which serves to cool the gases to below 700°C before entry to the convective section of the waste heat boiler. Sulfated sticky lead dusts are collected in the offtake shaft and removed by rapping equipment, causing the collected dusts to dislodge and fall back into the reactor. Following the boiler, gases at around 350°C are cleaned in an electrostatic precipitator with four fields in series, followed by further cooling and scrubbing prior to supply to a sulfuric acid plant. Gas composition is normally in the range of 12 to 20 per cent SO2, but with high levels of CO2 at around 15 per cent and high levels of moisture from the concentrates also at around 15 per cent. Oxygen can be of a similar level to SO2 and the balance is nitrogen.



117


Dusts from the boiler and gas cleaning system are recycled back to the reactor, but depending on the cadmium content part may be removed and leached to extract cadmium before recycling to prevent development of a large circulating load. Slag from the oxidation section flows over a weir separating the two sections to the reduction stage where coal and oxygen are injected to reduce the lead content. Metallic lead formed by reduction flows back in the reverse direction to the oxidation section. Slag flows out of the reactor via an overflow weir and is targeted to contain less than five per cent Pb. It was initially proposed that the reduction section with five coal injectors would form a series of well mixed cells giving a gradation in the lead content of slag along the reactor to a low final value. To some extent this could be assisted by baffles; however, it has been found that diffusion rates at the operating temperature of around 1250°C are so high that it is difficult to achieve any gradation in slag composition and the achievement of low lead slags has been difficult. However, levels of 2.5 per cent lead in slag have been achieved. To assist in settling entrained metallic lead in slag, an area without injectors is provided at the slag discharge end of the reactor to allow metallic lead to separate before the slag is discharged. Slag composition as set by flux additions is targeted at close to blast furnace slag composition with a typical slag from the Stolberg smelter reporting: two to three per cent Pb, 25 to 28 per cent FeO, 21 to 23 per cent SiO2, 21 to 23 per cent CaO and seven to eight per cent Zn (Pullenberg and Rohkohl, 2000). Part of the zinc content of slag is also reduced and reports as zinc vapour to the reduction gases together with some lead. Air or oxygen enriched air is added to oxidise the zinc to an oxide fume and the gases are then cooled and finally cleaned in a bag filter to recover the zinc fume. Although oxidation and reduction gases can be separated, this is not necessary provided mix gas strength in terms of the SO2 content remains suitable for acid production. It should be borne in mind that a significant quantity of nitrogen is added as shroud gas around each injector and represents around one quarter of the oxygen volumetric flow. The length of the reduction section of the reactor is at least twice the length of the oxidation section but is of reduced diameter. The relative slag volumes and residence times are the critical parameters and are determined by the relative amounts of sulfides and oxide lead materials in the feed mix, particularly the amount of lead to be reduced. The lead pool maintained in the reduction zone is minimal and merely provides a channel for the lead to flow back to the oxidation section. The cylindrical reactor can be rotated through 90° when the process is interrupted to lift the injectors clear of the slag bath. This also allows access to the injectors for maintenance or replacement. Injector life is relatively short and has been quoted for the Stolberg unit at 800 hours for the oxidation section and 1300 hours for the reduction section. Overall plant availability was given as 85 per cent. The reactor is constructed in much the same way as a rotary kiln but with limited rotation. It is generally lined with chrome magnesite brick. The experience with refractory wear has been a problem, with particular areas opposite the injectors and at the slag/gas interface line showing high wear rates. Improved cooling of refractories and profiling of the refractory surface has progressively improved and prolonged refractory life.

Comparative plant design data Table 7.4 lists comparative design data for the four commercial plant designs to date. Note that the Cominco plant did not achieve design expectations and was replaced.

118




TABLE 7.5 Design data for Queneau-Schuhmann-Lurgi (QSL) smelters (Mager and Schulte, 1989). Plant

Stolberg

Korea Zinc

Bayin

Cominco

80 000

60 000

52 000

120 000

Oxidation (m)

3.5

4.5

3.5

4.5

Reduction (m)

3.0

4.0

3.0

4.0

Reactor length (m)

33

41

30

41

Oxidation section length (m)

11

13

10

13

Reduction section length (m)

22

28

20

28 47

Lead production (t/a) Reactor diameter

3

Slag volume in oxidation (m )

24

47

21

Slag volume in reduction (m3)

30

74

26

74

% concentrates in raw material

63

53

100

53

% lead in raw material

45

35

66

34

27.34

34.5

16.0

81.7

Raw materials (t/h)

20.8

22.7

11.0

52.0

Silica (t/h)

0.04

0.4

1.0

Limestone (t/h)

0.3

2.7

0.7

6.3

Recycle oxidation fume (t/h)

4.3

5.0

2.2

15.4

Total feed (t/h dry)

Recycle leach residue (t/h)

1.3

Recycle fumed slag (t/h) Coal fines (t/h)

1.5 1.2

1.9

2.8

0.5

5.5

4700

7300

2250

13 000

Coal to reduction (t/h)

0.9

1.4

0.7

Natural gas to reduction (Nm3/h)

400

Oxygen to oxidation (Nm3/h)

1,750

3

Off gases (Nm /h) Oxidation

22 400

Reduction

31 000

14 000

65 000

44 000 25 300

Slag production (t/h)

7.1

8.8

4.0

28.4

Bullion production (t/h)

9.6

7.9

7.2

33.8

Steam production (t/h)

14.3

19.0

6.0

33.0

Oxidation slag (%Pb)

50

40

40

35

Reduction slag (%Pb)

2.5

2.0

2.5

5.0

The QSL process has significant advantages over sinter plant–blast furnace technologies. It is a single step smelting process, is well contained with low emissions and can accept a wide range of feed materials. It does not require dry feed as with the Kivcet process; however, in its early stages it did not prove to be as robust or as flexible as the Kivcet process, particularly with attempts to use natural gas for reduction and in its ability to handle quantities of zinc plant residues.

TOP SUBMERGED LANCE (TSL) – SLAG BATH PROCESSES The top submerged lance–slag bath reactor for injection of gases and fuel into a slag bath was developed in Australia by the Commonwealth Scientific and Research Organisation (CSIRO), and was termed the SIROSMELT Process. The original aim was to avoid the refractory problems



119


associated with submerged injection tuyeres by using a submerged lance separated from the refractory lining of the reactor and which could be withdrawn and readily replaced in the event of maintenance problems. The key development was the design of the lance, which relied on the formation of a frozen slag layer on its outer surface to protect the steel from attack by aggressive slags, such as those high in PbO content or molten metals. The lance consisted of an outer steel tube containing one or more inner tubes for the injection of fuel or feed materials. Cold air is blown down the outer tube which is fitted with helical baffles to impart a swirl to the airflow and increase the heat transfer coefficient to the tube wall. Superficial gas velocity is around 100 m/s and the outer tube is cooled sufficiently to freeze a layer of slag on its outer surface. The lance is in fact initially slowly lowered into the slag bath to sufficient depth to cause enough splashing to form the frozen slag coating before it is fully immersed. The reactor can be a simple vertical cylindrical vessel lined with refractory brick. The slag bath is highly turbulent as a result of the gas injection by the lance, requiring the vessel to be relatively high in comparison with the slag bath depth, due to the high degree of splashing and the need to avoid carryover of spray into the gas offtake, where it can freeze on the walls and cause accretions and eventual blockage. For this reason top burners can be used to ensure that freeboard gas temperature is high enough to melt accretions and prevent build-up. Initial application of this process to lead (and copper) was developed by Mount Isa Mines and was termed the Isasmelt process. Other applications and later also tin, copper and lead smelting and zinc fuming were developed by Ausmelt Limited, both as licensees of the Sirosmelt technology.

THE ISASMELT PROCESS The Isasmelt lead process was developed as a two-stage process for the smelting of lead concentrates. The first stage is a continuous oxidation stage to eliminate sulfur as SO2 and form a lead-rich slag, and in some cases lead bullion. The second stage involves the reduction of the lead rich slag by the addition of carbon to form lead bullion and a slag with a low lead content. Gases from each stage are separately handled through waste heat boilers and gas cleaning equipment. The oxidation stage gas can be passed to a sulfuric acid plant. A small demonstration unit was constructed at Mount Isa but was only operated for a short time and then only the oxidation section appeared to have been routinely operated. In that case the lead-rich slag was solidified and crushed for use as a replacement for return sinter in a standard sinter plant. This increased the lead throughput of the sinter plant accordingly. Product sinter was reduced in a blast furnace. No applications of the full Isasmelt lead process to the smelting of lead concentrates have been implemented, although it has been applied to the smelting of secondary materials, but using a batch operating mode rather than a continuous process as originally envisaged (Matthew et al, 1990). Figure 7.11 shows a schematic of the two-stage Isasmelt lead process.

Oxidation stage Concentrates and fluxes are mixed in a paddle mixer and fed via a weigh feeder directly into the slag bath reactor. Wet feed is satisfactory, eliminating any need for drying, but high moisture will consume more fuel and increase the volume of smelter gases. Flux additions are set to provide a slag with an FeO:SiO2 ratio of 2:1 and an FeO:CaO ratio of 5:1. Higher levels of CaO will increase the activity of PbO and hence will aid subsequent reduction, but this will be offset by the increased volume of slag and greater lead loss at the same final lead in slag value.

120




Concentrates Silica Recycled dusts

Lime

Boiler To acid plant Electrostatic precipitator To stack Fuel

Mixer Air + oxygen

Boiler Gas off-take shaft

Coal

Fuel

Bag filter

Air

Cooling tower Oxidation stage

Reduction stage

Bullion

Slag

FIG 7.11 - Isasmelt lead smelting process.

Air enriched to around 30 per cent oxygen is injected through the lance, and fuel in the form of pulverised coal, oil or natural gas is also injected as required to maintain the bath temperature at around 1150 to 1200°C. Off-gas containing eight to 12 per cent SO2 passes through a vertical membrane cooled offtake shaft to a waste heat boiler, electrostatic precipitator, gas scrubber and sulfuric acid plant. Lead sulfide volatilisation is largely suppressed by rapid incorporation of the feed into the molten slag bath under oxidising and highly turbulent conditions. Nevertheless approximately ten to 15 per cent of the lead in feed can be volatilised and appear in sulfated form in dusts collected from the boiler and gas cleaning plant. Oxidation rates are quite rapid and as a guide, smelting rates equivalent to about 1.3 t/h of sulfur as sulfides in feed can be processed per cubic metre of slag bath volume. The lead content of oxidation stage slag is in the range of 40 to 55 per cent Pb and the balance of lead from feed can form bullion. The amount of bullion formed will depend on feed composition, but can be controlled to some extent by the stoichiometric excess of oxygen supplied for the smelting reactions. At 105 per cent stoichiometric oxygen supply a maximum amount of bullion will form, whereas at 125 per cent bullion formation can be eliminated, producing a high lead slag only.



121


Slag and any bullion formed can continuously overflow from the reactor into the second reduction reactor, or it may flow into a settler to separate bullion and slag, if bullion separation at this point is required. Bullion from the oxidation stage can contain higher levels of precious metals and bismuth, and lower levels of sulfur, arsenic and antimony than bullion from the reduction stage; copper levels are similar. For this reason there may be advantages in separating the two bullions. Conditions in the oxidation reactor favour the formation of magnetite and zinc ferrite, which tend to precipitate and coat the surface of the refractories (chrome magnesite brick) and the lance. This provides an effective protection layer, leading to relatively long refractory life for this stage. The reduction stage, however, does not have this protection and consequently has a much shorter refractory and lance life. To enhance this effect the operating temperature can be held just below the liquidus temperature of the slag.

Reduction stage Slag reduction takes place in a similar vessel to the oxidation reactor, but with the injection of fine coal and air. Lump coal can also be directly added but reduction rates appear to be favoured by the injection of fine coal through the lance. The operating temperature of the reduction is higher than the oxidation stage since the liquidus temperature of the slag rises as the level of PbO is lowered. Typical operating temperatures are 1200 to 1250°C. There have been no long-term commercial operations of a continuous reduction stage and details of a preferred mode of operation and performance are quite unclear. However, it would appear that this operation has been problematic due to the difficulty with attainment of low final lead levels and excessive fuming of lead. Tail slags with less than two per cent lead have been demonstrated in pilot plant operations, but at long residence times and practical operating levels for typical slags may be well above five per cent. Lead fuming, expressed as a proportion of input lead, is higher for low lead content slags; gas volumes are also high compared with other direct smelting processes, aggravating this problem. Zinc fuming also increases as the lead level in slag is lowered and can be of the order of 50 per cent. It is substantially increased with the operating temperature of the reduction stage. Fuming during reduction is understood to be a significant issue for this process, and can develop a large circulating load. Gases from the reduction reactor pass through an uptake shaft, waste heat boiler, quench cooler and bag house filter to recover fume. The fume may need to be partly leached to separate cadmium and zinc before recycle. Slag overflows from the reactor into a forehearth to separate bullion, and good settling is essential for the achievement of low final lead levels in discard slag. It would appear that there could be advantages in operating the slag reduction stage in a batch mode rather than continuously. This approach may allow a progressive change in reducing conditions to minimise fuming, but few details of this practice are available. In general the oxidation stage of the Isasmelt process appears to be robust and practical but there are clearly difficulties with the reduction stage, which has yet to be satisfactorily demonstrated. Indeed the application of the Isasmelt process to secondary lead processing at the Britannia Lead refinery in the United Kingdom smelts battery paste materials to produce bullion and a high lead slag at around 50 per cent Pb, which is separated and reduced in a rotary furnace rather than in the Isasmelt furnace with its substantially higher gas flows. An Isasmelt furnace is also used at Metal Reclamation Industries in Malaysia with a capacity of around 50 000 t/a of lead bullion.

122




The ISA-YMG smelter in Qujing, China uses an Isasmelt first stage furnace to smelt concentrates with partial recovery of lead as bullion. The high lead slag is cast into biscuit moulds producing a lump slag that is then smelted with coke in a specially designed blast furnace. The blast furnace has modifications in the crucible area to allow intensive tuyere blast interaction with the pool of slag and coke in the hearth so that target levels of lead in final slag can be achieved. This smelter configuration replaces the sinter plant and is being proposed as a future direction for primary lead smelting technology.

THE AUSMELT LEAD PROCESS The Ausmelt lead process is similar in principle to the Isasmelt process above but differs in detailed design (Mounsey and Piret, 2000). Ausmelt supply a broad range of top submerged lance technologies and have constructed four primary lead smelters using this approach, including a 140 000 t/a of bullion plant for Metaleurop at Nordenham in Germany, a plant for Goldfields of South Africa at Tsumeb in Namibia, an 80 000 t/a of bullion unit for Korea Zinc at Onsan, South Korea and a 60 000 t/a smelter for Hindustan Zinc at Chanderiya, India. Ausmelt offers a highly flexible arrangement with one, two or three furnaces in series, operating in either batch or continuous mode or in any combination thereof. In the smelting of sulfide concentrates it is usually preferable to have the first oxidation stage operating continuously in order to supply SO2 to any associated sulfuric acid plants, which are unable to handle large variations in throughput. Slag reduction can be operated in either batch or continuous mode and a third stage can be added if it is required to recover zinc by fuming. A batch operated oxidation stage may be suitable for processing secondary materials, oxides or leach residues. In fact all the above installations operate with only one furnace each, which is used for differing duties on a campaign basis.

Smelting or oxidation stage Of prime importance in this stage is the deportment of lead to bullion, slag and fume. Ausmelt data has substantially defined this relationship as highly dependent on the lead concentration in the feed as well as the stoichiometric oxygen:concentrate ratio. Table 7.6 summarises the typical lead distribution for high, medium and low-grade feed materials. TABLE 7.6 Deportment of lead in the smelting stage. Feed grade (% Pb)

Smelting temperature ( C)

Smelter slag grade (% Pb)

Deportment of Deportment of Deportment of lead to bullion (%) lead to slag (%) lead to fume (%)

60 - 80

1000 - 1100

40 - 60

75 - 85

5 - 15

5 - 15

45 - 60

1100

25 - 40

36 - 60

25 - 45

15 - 20

1150 - 1250

20 - 30

20 - 50

30 - 50

20 - 30

<45

Note: refer to Mounsey and Piret (2000).

The variation of these distributions with oxygen supply for one particular feed is shown in Figure 7.12. The deportment of lead to bullion can be increased by increasing the lead in feed as in Table 7.6, by increasing the oxygen enrichment of the air, by decreasing the temperature and by decreasing the sulfur in feed (which strongly corresponds with higher lead grade concentrates).



123


100 90

Lead deportment (%)

80 70 60

To bullion

50

To slag To fume

40 30 20 10 0 60

70

80

90

100

110

120

130

140

Oxygen Stoichiometry (%)

FIG 7.12 - Effect of oxygen stoichiometry on lead deportment.

Reaction rates in the oxidation stage are very high and largely limited by the capacity of the gas handling system.

Slag reduction stage Slag from the smelting or oxidation stage is carried out in a separate second stage by the addition of coal and air. As the PbO content falls the liquidus temperature of the slag rises and this stage generally operates above 1200°C. The higher temperature will produce increased quantities of fume and from this viewpoint batch operation can be beneficial in progressively raising the temperature as the slag is reduced. Final slags containing less than three per cent lead are feasible. However, the deportment of lead to fume or bullion has not been defined in a generalised way as for the oxidation stage above, but is a significant issue and based on operating plants does not appear to be a particularly attractive option. The Metaleurop smelter at Nordenham uses a single furnace and was initially designed to campaign oxidation and reduction stages, granulating and stockpiling high lead slags from the oxidation stage for later campaign reduction. However, the unit operates only as the smelting stage and lead-rich slags are sold to other smelters (as a sinter feed supplement). The Tsumeb smelter was similarly a single reactor with campaign operation of smelting and reduction stages but ceased operation in 1998 due to closure of the mine complex. In the situation where the zinc content of feed is relatively high, slags can contain up to 18 per cent zinc, and recovery can be considered. Although some zinc will fume during slag reduction down to two per cent lead, the bulk will remain and can be fumed in a third stage with coal addition and operating at higher temperatures of between 1250 and 1300°C. This stage may be operated continuously in a third furnace or on a batch campaign basis using the reduction furnace. If extensive zinc fuming is practised, allowance must be made for secondary air input above the bath to oxidise zinc vapour to an oxide fume, and for the high level of heat so generated. Gas cleaning facilities must also be capable of collecting and handling the zinc fume generated. Ausmelt reactors operating in this mode are in use at the Onsan plant of Korea Zinc.

124




Reactor design The conventional design has been a steel shell lined with refractory brick and cooled by shower sprays on the outer surface. The upper section on reduction reactors where afterburning occurs can operate at quite high gas temperatures and are subject to severe refractory conditions, requiring additional shell cooling. More recent designs for the Korea Zinc fuming reactors have used a water jacketed section for the top of the reactor and the roof as part of the boiler system. This allows a protective frozen slag layer to form.

THE OUTOKUMPU LEAD PROCESS The Outokumpu flash smelting process for lead is an extension of the successful flash smelting technology developed by Outokumpu for copper and nickel and applied to lead concentrates. It was developed and piloted at a 5 t/h treatment rate in the late 1980s but has not been commercially applied. Lead concentrates are dried and fed to a flash burner with 95 per cent oxygen gas, located at the top of a smelting shaft. A high lead slag and some bullion are produced and collect in the base of the furnace. Gases rich in SO2 pass up through a separate off-take shaft to cool and allow volatile lead compounds to sulfate and be separated as a fume for subsequent collection in the waste heat boiler and electrostatic precipitator. In common with other direct smelting processes the smelting stage aims to achieve high sulfur elimination and hence requires oxidation conditions, which will necessarily result in the formation of a high lead slag of between 20 and 60 per cent lead. The higher the grade of the concentrate feed, the higher will be the proportion of lead in concentrates reporting to bullion. For low-grade concentrates all the lead can report to slag with no or minimal bullion formation in the smelting stage. The slag and bullion are continuously tapped from the smelting furnace into a separate electric furnace where the PbO in slag is reduced to lead bullion by injection of coal with nitrogen through lances. Pilot tests by Outokumpu showed that the rate of reduction of PbO by a floating coke layer was slow and could be increased five times by injection of coal into the slag bath. Off-gas from the electric furnace can contain significant amounts of lead and zinc, which are oxidised to a fume by the addition of air in an after burner chamber. If high zinc recovery is required from the slag the electric furnace temperature must be raised, but this will cause a substantial increase in the amount of lead to fume. Bullion and slag are separately and continuously tapped from the electric furnace. The efficiency of the electric furnace for reduction and the degree of fuming of lead are the key areas of concern with this process, but no commercial scale data is available to demonstrate performance. Indicative consumption figures for a high-grade lead concentrate of 76 per cent Pb have been given as:

• oxygen use

120 Nm3/t of concentrate

• propane fuel use

5 kg/t of concentrate

• coal for reduction

18 kg/t of concentrate

• power use

100 kWh/t of concentrate

(Weenink, de Puy and Duyvesteyn, 1990).



125


REFERENCES Ashman, D W, Goosen, D W, Reynolds, D G and Webb, D J, 2000. Cominco’s new lead smelter at Trail operations, in Proceedings Lead-Zinc 2000, TMS Symposium, pp 171-185. Elvander, H L, 1965. The Boliden lead process, in Proceedings Metallurgical Society AIME, Pittsburg, pp 225-243. Jak, E, Zhao, B and Hayes, P, 2000. Phase equilibria and thermodynamics of zinc fuming slags, in Proceedings Minprex 2000, pp 479-484 (The Australasian Institute of Mining and Metallurgy: Melbourne). Lyamina, M A and Shumskii, V A, 2006. Theoretical questions in the treatment of lead-bearing raw materials by the Kivcet process (translated by D H Ward) (The Eastern Mining and Metallurgical Research Institute for Non-Ferrous Metals (Vniitsvetmet), Ministry of Industry and Trade: Kazakhstan Republic). Mager, K and Schulte, A, 1989. Process and technological aspects of the first four QSL plants, in Proceedings Canadian Institute of Mining and Metallurgy (CIM) Symposium on Primary and Secondary Lead Processing, pp 3-14 (Canadian Institute of Mining and Metallurgy: Montreal). Matthew, S P, McKean, G R, Player, R L and Ramus, K E, 1990, The continuous Isasmelt process, in Proceedings Lead-Zinc 90, The Minerals, Metals and Materials Society (TMS) Symposium, Anaheim, pp 889-901. Matyas, A G and Mackey, P J, 1976. Metallurgy of the direct smelting of lead, Journal of Metals, 28(11):10-15. Mounsey, E N and Piret, N L, 2000. A review of Ausmelt technology for lead smelting, in Proceedings Lead-Zinc 2000, The Minerals, Metals and Materials Society (TMS) Symposium, Pittsburg, pp 149-169. Nystedt, P, 1980. Operation of the Boliden lead Kaldo plant for treatment of lead bearing dusts, Canadian Institute of Mining and Metallurgy (CIM) Bulletin, May. Pazour, D A, 1982. Boliden tests TBRC for lead concentrate smelting, World Mining, September, pp 63-65. Perillo, A, Carminati, A, Schuermann, P and Berger, N, 1990. The Kivcet furnace construction at Porto Vesme in Proceedings Lead-Zinc 90, The Minerals, Metals and Materials Society (TMS) Symposium, Anaheim, pp 903-917. Petersson, S and Erickson, S, 1977. Autogenous smelting of lead concentrates in TBRC, in AIME Symposium, A77-11 (The Minerals, Metals and Materials Society (TMS): Warrendale). Pullenberg, R and Rohkohl, A, 2000. Modern lead smelting at the QSL plant, Berzelius Metall in Stolberg, Germany, in Proceedings Lead-Zinc 2000, The Minerals, Metals and Materials Society (TMS) Symposium, pp 127-148. Queneau, P E and Schuhmann, R, 1974. The QS process, Journal of Mining, 26(8):14-16. Slobodkin, L V and Kluev, G F, 1998. Kivcet process – A universal technology for treating lead containing materials, in Proceedings Lead-Zinc Processing Symposium, pp 715-722 (Canadian Institute of Mining and Metallurgy (CIM)). Sychev, A P, Kuenov, A S, Sannikov, Y I, Slobodkin, L V, Lysenko, V A and Grinin, Y A, 1985. Completion of an important stage in the implementation of the Kivcet-TSS process at the Ust Kamenogorsk lead-zinc combine, Tsvetnye Metally, 29(1):11-15. Walker, M, 1998. Kivcet smelter on-stream at Trail, Mining Magazine, 178(4):256-263. Ward, D H, 1985. Scientific and technical developments in lead smelting and refining, in Proceedings Scientific and Technical Developments in Extractive Metallurgy, G K Williams Memorial Volume, pp 25-32 (The Australasian Institute of Mining and Metallurgy: Melbourne). Ward, D H, 2007. Personal communication. Weenink, E M, de Puy, E R and Duyvesteyn, W P C, 1990. Computer modelling of mass and energy balances for the Outokumpu lead flash smelting process, in Proceedings Lead-Zinc 90, The Minerals, Metals and Materials Society (TMS) Symposium, Anaheim, pp 597-606.

126




Willis, G M, 1980. The physical chemistry of lead extraction, in Proceedings Lead-Zinc-Tin ’80, The Minerals, Metals and Materials Society (TMS) – AIME Symposium, pp 457-476. Yazawa, A and Nakazawa, S, 1998. Comparison between copper, lead and nickel smelting processes from thermodynamic viewpoints, in Proceedings Sulfide Smelting Symposium ’98, pp 39-48 (The Minerals, Metals and Materials Society (TMS): Warrendale). Zaitsev, V Y and Margulis, E V, 1985. The Metallurgy of Lead and Zinc (translation by D H Ward, 1992).



127

CHAPTER 8 Smelter By-Products and Treatment Processes INTRODUCTION Apart from lead bullion the smelting process produces slag, which often contains residual lead and recoverable quantities of zinc. Flue dusts are also produced in varying quantities and, depending on the operating conditions of the smelter, a matte and speiss can also be produced. Matte can be produced as a means of separating copper, but can also capture a significant portion of the precious metals, silver and gold. The following details processes used to treat these streams and recover valuable by-product metals.

SLAG CLEANING Reduction of slag in the blast furnace to below 1.5 to 2.0 per cent Pb is uncommon and will raise coke consumption. In the extreme at lead levels at or below one per cent in slag there is potential for the formation of metallic iron and the production of speiss phases, which cause accretions and tapping difficulties. On the other hand, with more oxidised slags solid phases can form, raising slag viscosity and impeding tapping and lead separation from the slag in the forehearth. Such solid phases may be magnetite, or melilite arising from a high zinc content. Operating the smelter at moderate residual lead levels in slag to avoid these problems will incur lead and precious metal losses and thus a separate slag cleaning stage to effect further reduction and lead separation can be justified. Lead is present in slag both as fine droplets, which have not settled out, and as dissolved PbO. Apart from using a separate slag cleaning stage for additional reduction of PbO, simply raising the temperature and providing extended settling time can then be a significant benefit. Such slag cleaning is best done using an electric furnace where high slag temperatures can be achieved without large gas volumes and consequent heat losses. With residence times of the order of eight hours and the addition of coke, lead in slag can be reduced down to 0.5 per cent. As well as bullion, an alloy phase can be produced rich in iron and copper, depending on the copper levels in smelter feed. Often the higher temperature alone will reduce slag viscosity and allow fine droplets of lead to settle, significantly reducing the lead content of the final slag. An installation of this type was installed at the Umicor Hoboken smelter in Belgium (Van Negen, Maes and Cocklebergs, 1990). In that operation 20 per cent of the charge was cold slag and reported operating data is as follows:

• blast furnace slag charge

2.45 per cent Pb, 1.28 per cent Cu, 7.2 per cent Zn, 118 g/t Ag

• cleaned slag

0.55 per cent Pb, 0.37 per cent Cu, 5.77 per cent Zn, 10 g/t Ag

• flue dust (3.4 per cent of charge)

18.1 per cent Pb, 54.1 per cent Zn

• bullion (0.4 per cent of charge)

88.6 per cent Pb, 5.21 per cent Cu, 4243 g/t Ag

• alloy (3.6 per cent of charge)

26.3 per cent Pb, 25.5 per cent Cu, 1.99 per cent Zn, 19.2 per cent Fe, 2460 g/t Ag

• coke consumption

two per cent of charge



129

CHAPTER 8 – Smelter By-Products and Treatment Processes

• power use

280 kWh/t of charge

• electrode carbon use

12 kg/MWh

ZINC RECOVERY FROM SLAGS Process chemistry and thermodynamics Depending on the zinc content of raw material feeds to the smelter, slag can contain up to 17 per cent zinc. This is a practical solubility limit beyond which solid zinc containing melilite phases tend to precipitate raising the viscosity of the slag. Higher zinc levels also contribute to significant accretion problems in the blast furnace. Use is made of the relatively high vapour pressure of zinc to recover it from slags by reduction and vaporisation into a gas stream, usually followed by oxidation of the vapour to form zinc oxide as a fine particulate fume. It is theoretically possible to condense zinc metal from the gas stream, but in practice this is difficult due to impurities present and to re-oxidation of zinc vapour by CO2 and water vapour. The presence of sulfur in lead smelter slags arising from the need to retain some sulfur in bullion, also means the PbS will volatilise during fuming operations. This will react with zinc vapour as the gas is cooled to form ZnS and metallic lead. Because of these complications the concept of direct condensation of zinc from slag fuming gases has not been applied commercially to the treatment of lead smelter slags, although some process developments incorporate this concept – see electric furnace slag fuming, below. The most common approach uses the injection of coal with air and oxygen into a molten slag bath at around 1200°C. Zinc oxide is reduced to zinc vapour according to Equations 8.1 to 8.3: ZnO + CO = Zng + CO2

(8.1)

CO2 + C = 2CO

(8.2)

Overall reaction ZnO + C = Zng + CO

(8.3)

The vapour pressure of zinc above a slag bath relates to the CO/CO2 partial pressure ratio in the gas and the activity of ZnO in the slag according to Equation 8.4: pZn =

K1 . p CO . a ZnO p CO 2

(8.4)

where: aZnO is the activity of ZnO dissolved in the slag K1 is the equilibrium constant of Equation 8.1 at the particular temperature The activity of ZnO can be related to the mole fraction of ZnO in the slag by the activity coefficient as in Equation 8.5: aZnO = γZnO . NZnO

(8.5)

where: γZnO = activity coefficient of ZnO NZnO = the mole fraction of ZnO in the slag

130




A review and assessment of ZnO activity coefficients on this basis was given by Grimsey et al (1993), and is illustrated in Figure 8.1. This data is for a slag of 8.5 per cent Al2O3 content at 1300°C and the slag composition has been condensed so that ZnO + CaO + SiO2 + FeO equals 100 per cent. Isobars have been constructed for constant ZnO activity coefficients, and are shown as dotted lines. It is clear that the activity coefficient largely depends on the CaO/SiO2 ratio and increases as the ratio increases.

60

10

50

20

30

40

wt% CaO

wt% SiO 2

3.0 2.5

40

50

2.0

60 wt% ZnO+FeO

1.5

70

80

FIG 8.1 - Activity coefficient for zinc oxide in slag at 1300°C.

At a fixed CaO/SiO2 ratio the activity coefficient also increases as the Al2O3 content is increased. For example, a typical lead blast furnace slag may have an activity coefficient of 2.0 at 8.5 per cent Al2O3 and this will increase to around 2.5 at 22.5 per cent Al2O3. This is a significant simplification of a complex system and more detailed thermodynamic modelling of the ZnO-PbO-FeO-Fe2O3CaO-SiO2 system with slag liquidus phenomena, component activities and the effect of precipitating phases on viscosity under various conditions has been given by Jak and Hayes (2002). Liquid zinc cannot be formed in the slag unless the vapour pressure is above the saturation vapour pressure of liquid zinc at the operating temperature. The saturation vapour pressure is given by Equation 8.6, and under atmospheric pressure conditions at around 1200 to 1300°C the saturation vapour pressure well exceeds the vapour pressure derived from reduction. Hence liquid zinc cannot form in the slag bath. Log pZn = 5.227 −

6163 T

(8.6)

where: pZn is in atmospheres T is in K



131


The equilibrium constant for zinc oxide reduction according to Equation 8.4 is given as a function of temperature in Table 8.1, together with the partial pressure of zinc vapour in equilibrium with a slag containing four per cent ZnO, as the end-point slag fumer composition, and a reduction gas ratio of CO:CO2 of 2:1. Also shown is the saturation vapour pressure of liquid zinc metal according to Equation 8.6. TABLE 8.1 Equilibrium constant for ZnO reduction. Temperature (K)

1100

1200

1300

1400

1500

Equilibrium constant (atm)

0.00167

0.00901

0.0305

0.130

0.374

Zn partial pressure (atm)

0.0002

0.001

0.0046

0.0156

0.045

0.42

1.233

3.064

6.68

13.13

Zn saturation vapour pressure (atm)

There has been conjecture about the principal mechanism for reduction from a molten phase and whether the process is in chemical equilibrium or is controlled by reaction kinetics and mass transfer. Initial studies of the reduction reactions in slag fuming (Bell, Turner and Peters, 1955; Kellogg, 1957) assumed that reaction rates were sufficiently high at operating temperatures above 1000°C for equilibrium conditions to prevail. Kellogg (1967) developed a mathematical model of the slag fuming process on this basis, and this was subsequently further developed and applied successfully at Port Pirie (Grant and Barnett, 1975) and at Boliden (Bygden et al, 1985). Empirical models were also developed for fuming processes, indicating an approach to a first order reaction with respect to zinc concentration in the slag as indicated by Equation 8.4 (Izbakhanov et al, 1990). Application of the models developed for one operation did not necessarily apply to another and slag properties such as viscosity clearly had a significant influence. Some operations more closely followed a zero order reaction in which the zinc reduction rate was independent of the zinc concentration in the slag, at least above a level of around four per cent Zn in slag (Richards, Brimacombe and Toop, 1985). This work indicated that solid carbon within the melt was significant, and the kinetics of the process may be controlled by the injection of coal particles into the slag bath as opposed to retention as a suspension in the gas phase. The kinetics can therefore be enhanced by high velocity injection. In practice neither temperature nor oxygen potential (or CO:CO2 ratios) remain constant and hence there can be significant departure from first order with respect to ZnO activity with rates to some degree dependent on heat and mass balance issues. The rate of reduction of zinc oxide in slag is also influenced by the FeO (or the ferrous to ferric iron ratio) and the sulfur contents of the slag. Rates are progressively increased as the FeO content increases. This is attributed to the reduction of zinc oxide by ferrous iron and the subsequent reduction of the ferric iron formed by CO. Rates may also be reduced by the presence of sulfur in the slag, thought to be caused by surface blockage of the gas–slag interface due to the surface active nature of sulfur in the slag, but may also be due to the impact on ZnO activity (Dal, Jahanshahi and Rankin, 1997). There are clearly a number of parallel reduction reactions and it would appear that their predominance depends on physical conditions and the operating conditions of the furnace. Some reactions may have low resistance and are not rate limiting, and others such as those involving solid carbon in the melt are kinetically controlled and can be enhanced by giving attention to carbon particle sizing and injection methods. Slag properties, particularly viscosity can have a bearing on the kinetics and can also change as the zinc concentration is depleted. However, it is now accepted that with an adequately deep slag bath limiting the opportunity for gas bypassing, equilibrium is essentially achieved so that the Kellogg model can be applied (Ward, 2000).

132




In most practical slag reduction operations coal is used as the source of carbon. Hydrocarbons are present and hydrogen can be formed giving rise to water vapour in reduction gases, which can result in re-oxidation of zinc vapour as the temperature is lowered. This will make condensation of metallic zinc from reduction gases difficult, but is not a disadvantage for fuming operations producing ZnO. The use of natural gas should therefore be a suitable fuel, but it was found that reduction rates fell significantly when coal was replaced by natural gas. Work at the Chimkent Lead Smelter in Russia (Izbakhanov et al, 1990) attributed this to the slow decomposition of the hydrocarbons, which then tended to pass through the molten slag and burn above it. Oxygen potential in the injected gas was therefore high and tended to oxidise iron in the slag to magnetite. This precipitated as a solid phase and raised the viscosity of the slag, further reducing reduction rates, or in the extreme caused the bath to freeze. Cracking and precombustion of the natural gas to CO and hydrogen in external chambers before injection into the slag bath overcame most of these difficulties (Polyvyanni, 1976). Processes utilising molten phase reduction usually employ in situ combustion of fuel to maintain operating temperature and to provide the necessary heat to balance endothermic reduction reactions and heat loss through the furnace walls. The gas phase is thus further diluted with nitrogen as well as having high levels of CO2 for maximum fuel utilisation efficiency. The thermal efficiency of the process can be improved by preheating input air to reduce the combustion fuel required and by the use of oxygen enrichment to reduce the volume of gas to be heated. However, the flow of gas does serve to enhance the removal of zinc vapour from the bath and there are limits to the positive impact of oxygen enrichment. An issue in some of the processes used for the reduction of zinc oxide from slags operating at extreme reducing conditions is the control or avoidance of the formation of metallic iron. Iron has a wide melting range depending on the carbon content. The minimum liquidus temperature of 1130°C corresponds with a eutectic at 4.3 per cent carbon. This rises to 1535°C for pure iron with no carbon content. For carbon contents above 4.3 per cent the liquidus temperature again rises. From usual reduction processes it would appear that the carbon content of iron formed can be around three per cent, melting at 1250°C. If the iron is decarburised by moving to reaction zones of higher oxygen potential, it can freeze and cause severe accretion problems. Therefore, operating conditions are often selected to avoid or minimise the formation of metallic iron. Iron formation results from the reaction given in Equation 8.7: FeO + CO = Fe + CO2

(8.7)

The equilibrium constant for the reaction 8.7 can be expressed in Equation 8.8: K=

p CO 2 × a Fe p CO × a FeO

= exp( −

∆G ) R× T

(8.8)

where: aFe is the activity of the iron formed taken as 1.0 aFeO is the activity of FeO dissolved in the slag The activity of FeO in ternary system slags is shown in Figure 8.2, from which it can be deduced that typical slags of around 35 per cent FeO will have an activity of around 0.4. This need not necessarily apply to a more complex multi-component system, but illustrates the possible variation and range of FeO activity coefficients.



133


SiO 2

0.1

0.2 0.3

0.4 0.5 0.6 0.7 0.8 0.9

FeO

CaO

FIG 8.2 - Activities of FeO in system FeO-CaO-SiO2 at 1600°C (Elliot, Gleiser and Ramakrishna, 1963).

The competing reduction of ZnO and FeO can be illustrated by the equilibrium ratios of CO:CO2 as a function of the temperature at which reduction occurs. These equilibrium curves are shown in Figure 8.3 for zinc at levels in a typical slag of 2.5 per cent and seven per cent, and for iron reduction at FeO levels in slag of 15 per cent and 45 per cent. Reduction will occur for CO:CO2 partial pressure ratios in regions above the equilibrium curves. For regions above all curves, both ZnO and FeO will be reduced and iron will be formed. For regions at high temperatures above the zinc equilibrium curves and below the iron curves, ZnO will be reduced without the formation of metallic iron. Raising the zinc content or lowering the iron content will extend this region of no iron formation to lower operating temperatures. At low temperatures FeO will be reduced in preference to ZnO. 3.5 Zn 2.5%

3

Zn 7%

Log(Pco /P co2 )

2.5

FeO 15%

2

FeO 45%

1.5 1 0.5 0 -0.5800

1000

1200

1400

1600

-1 Temperature ( C)

FIG 8.3 - Equilibrium curves for ZnO and FeO reduction.

134




For typical situations with slags containing around 35 per cent FeO, zinc can be fumed to levels of 2.5 per cent in final slag without iron formation if the temperature is above about 1250°C. For higher iron slags, such as the typical ISF slag, this temperature must be above 1300°C, or if the slag contains seven per cent zinc, above 1200°C. The tendency to form metallic iron during fuming is significantly greater for ISF slags than for normal blast furnace slags, where few such related problems are normally encountered. As well as removing zinc, fuming will also remove most of the lead content of the slag and if this is recycled by return of zinc smelter residues to the lead smelting stage, then almost complete lead recovery can be achieved in an integrated operation. This ability of lead to fume is illustrated by the vapour pressure of lead compounds at 1250°C:

• PbS vapour pressure at 1250°C

0.987 atm

• PbO (liquid)

0.154 atm

• Pb (liquid metal)

0.040 atm

THE CONVENTIONAL SLAG FUMING FURNACE The conventional slag fuming furnace was developed for the injection of coal into hot blast furnace slag, with secondary air injection above the slag bath to oxidise zinc, and included equipment for gas cooling and fume collection. Slags which typically contain between ten and 18 per cent Zn and around 2.5 per cent Pb, 30 per cent FeO, 25 per cent SiO2 and ten per cent CaO are suitable. Zinc recovery is economic only at high levels of zinc in slag, and hence slag fuming furnaces are now only applied where the zinc content of slag is at the upper end of the range at around 16 per cent or more. In fact when such a facility is in place it benefits the lead smelter to source concentrates containing zinc in order to optimise the levels in slag. No payment is required for such zinc content in lead concentrates and it is ‘free metal’ to the lead smelter. The first commercial installation of the standard slag fuming furnace was at the Asarco, East Helena lead smelter in Montana in 1927, and this furnace design has subsequently been applied at a number of other major lead smelters. Copper smelter slags rich in zinc are also processed in conventional slag fumers such as at Boliden’s Rönnskär smelter in Sweden, and at the Flin Flon smelter of Hudson Bay Mining and Smelting Company. The standard furnace is constructed as a long narrow rectangular box of steel water-cooled panels. A row of tuyeres is fitted to each long side just above the furnace floor level so that a suspension of fine coal in air is blown into the molten slag through the tuyeres. A schematic of the slag fuming furnace is shown in Figure 8.4. A typical furnace has cross-sectional dimensions of 2.4 m wide by 6.4 m long and 6.4 m high. The walls are made of rows of water jackets, with three rows each of seven jackets making up each long side for the typical size. The floor is also made up of water-cooled panels. There are three tuyeres in each lower jacket, giving 21 tuyeres per side. Furnaces may vary in size by varying the number of side wall panels, but the width is usually constant and dictated by the depth of penetration of the blast from the tuyeres into the melt (Hancock, Hart and Pelton, 1970; Ashman et al, 2000). The water-cooled panels or jackets freeze a protective layer of slag on their inner surface, which can be studded to assist in adherence of slag. This form of containment rather than refractories has been chosen because of the aggressive conditions and the high turbulence and abrasive nature of the bath, and has been highly successful with jacket life extending well over ten years and up to 20 years.



135


Water Tube Boiler Section

Membrane Wall Shaft

Gas to Cooling and Fume Collection

Feed Slag

Dust Collection Fuming Furnace

Tap Hole

Slag Level

Tuyeres

Spent Slag to Granulation

FIG 8.4 - Conventional slag fuming furnace.

The price of this form of containment is the additional fuel required to compensate for the relatively high heat loss through the jackets to maintain the protective slag layer. Water is circulated through the jackets and is flashed to low pressure steam. Furnaces are normally operated in a batch mode and the typical size with a 15.4 m2 cross-section will hold approximately 45 to 50 tonnes of slag. Slag is charged as molten blast furnace slag using a ladle and also as cold slag or as residues containing zinc. In the first part of the batch cycle, air and pulverised coal are blown into the furnace to fully melt the bath and to raise the temperature of the contents to around 1250°C. The air rate for this part of the cycle is set to provide sufficient oxygen for complete combustion of coal and maximum heat generation. Some fuming of zinc occurs, but at a rate which is about one third of the rate under reducing conditions. The time taken for the heating phase depends on the proportion of cold material charged and the temperature of the molten slag feed. The second phase of the cycle is the fuming phase in which the coal rate is increased, but the air rate is reduced to ensure that the oxygen supply is sufficient for 60 to 70 per cent of complete combustion. Fuming is completed when the zinc content of the slag is reduced to between two to three per cent and the furnace is then tapped to complete the cycle. Bath temperature falls during fuming but rises at the end of the cycle and can be purposely raised at the end of the batch by raising the air rate to assist in furnace tapping by lowering slag viscosity and hence reduce the time required for tapping. Alternatively lowering the coal addition rate for a fixed air rate at the end of the cycle will also allow complete combustion to CO2 and will raise the slag temperature in preparation for tapping. During the initial heating-oxidation stage of a batch slag fuming process the ferrous:ferric iron ratio decreases and if this stage is extended too long it can result in magnetite formation, which can precipitate from the slag. Reduction of magnetite will occur first in the fuming phase in preference to

136




ZnO reduction, acting as a buffer that limits the initial rate of zinc reduction. Because of this effect close control of the heating cycle of a batch fuming process to minimise magnetite formation is an important issue. Overall cycle times are generally of the order of two to three hours, giving an effective fuming rate within the range of 200 to 280 kg of zinc per hour per square metre of furnace cross-section. A typical metal elimination curve is given in Figure 8.5 showing the zinc concentration in the slag bath with time. This relates to the fuming phase of a cycle only, starting from an initial slag composition of 17 per cent Zn. If an initial heating phase is used then some fuming will occur during this part of the cycle at lower rates. Fuming rates may be taken from Figure 8.5 and expressed as per cent decline in zinc content per hour. This is shown in Figure 8.6. 18

Zinc Content %

16 14 12 10 8 6 4 2 0 0

50

100

150

200

Time - minutes FIG 8.5 - Typical zinc in slag reduction during fuming.

8

Fuming Rate %/h

7 6 5 4 3 2 1 0 0

5

10

15

20

Zinc Content % FIG 8.6 - Typical zinc fuming rate versus slag composition.



137


Fuming rate - tonnes Zn/hour

For a furnace with a 50 tonne initial charge the corresponding fuming rates in tonnes of zinc per hour are shown in Figure 8.7 as a function of furnace slag composition. 5 4.5 4 3.5 3 2.5 2 1.5

Instant Average

1 0.5 0 0

5

10

15

20

Zinc in Slag % FIG 8.7 - Zinc fuming rate versus slag composition (50 tonne capacity furnace).

Two curves are given in Figure 8.7: the lower curve representing the instantaneous fuming rate at the particular slag composition during the cycle, and the upper curve representing the overall average fuming rate to that particular tail end slag composition. As illustrated, the rate declines significantly as fuming proceeds and the zinc content of slag is reduced. For a total cycle, including time allowances for charging and heating and for tapping of the furnace, the tonnes of zinc fumed will be averaged over an extended time period, only part of which will cover high fuming rates. By way of example, for a total cycle, including 30 minutes for furnace charging and 15 minutes for tapping, the overall average zinc fuming rate as a function of final tail slag composition is shown in Figure 8.8. 3.00

Av. Zinc Fuming rate t/h

2.50 2.00 1.50 1.00 0.50 0.00 0

2

4

6

8

10

12

14

16

Residual Zn in slag %

FIG 8.8 - Average overall fuming rate versus tail slag composition (50 tonne capacity furnace – batch operation).

138




The curve in Figure 8.8 is relatively flat with little difference in overall furnace capacity when operating between three and nine per cent residual zinc in slag from a feed of 17 per cent zinc. However, zinc recovery is maximised at the lower residual zinc level and it is common practice to operate at a tail slag composition of around 2.5 to three per cent zinc. Below this there will be a significant fall-off in furnace capacity. If the fuming furnace can be charged with hot slag, then the average fuming rates over the total cycle can be increased by eliminating the heating phase. An electric slag holding furnace may be used for this purpose and indeed Teck-Cominco’s Trail smelter feeds hot slag directly at 1340°C from an electric furnace associated with its Kivcet lead smelter (Ashman et al, 2000). In the situation with hot slag feed, the optimum tail slag composition can be lower and Teck-Cominco operate to a tail slag of close to 2.0 per cent. Zinc is vaporised from the slag bath as zinc metal but is oxidised above the bath by the introduction of secondary air. Secondary air also burns CO to CO2 and any fine coal particles in suspension in gas emitted from the bath, and is controlled to achieve an oxygen level of five per cent in furnace flue gas. As a result the temperature of furnace gases above the bath can rise to 1600°C. The furnace gases pass up through a shaft composed of a water-cooled membrane wall to a water tube boiler where the temperature is reduced to around 650°C. The membrane wall is integrated within the boiler system. Accumulation of dust on the boiler tubes and walls is a significant problem and soot blowing is extensively required, usually using saturated steam. Gases from the boiler are further cooled to below 200°C and are filtered using bag filters to capture the zinc oxide fume. Gas cooling is commonly by air coolers and by humidification in water spray towers. Air cooling by exchange with furnace combustion air to achieve a preheat of up to 600°C is a useful way to improve thermal efficiency and reduce coal consumption. However, in the total gas handling system there are major issues with the accumulation of sticky dust deposits and the need for constant cleaning of heat exchange surfaces and ducts to prevent flow restriction. Typical plant availability is 88 per cent and is largely dictated by the need for cleaning parts of the gas train rather than the fuming section. Downtime can vary significantly and is critically dependent on the design of the gas handling system. Coal is usually pulverised to 90 per cent minus 200 mesh and is transported in a fluidised state using cold air to the tuyeres, where it is injected with blast air into the slag bath. Coal consumption is of the order of 16 to 25 per cent of the slag charge where the slag contains around 17 per cent Zn, and represents between 1.1 and 1.7 tonnes of coal per tonne of zinc recovered. The lower end of this range requires full preheating of blast air, but variations also depend on the extent of charge heating required and the proportion of cold feed, as well as the length of the fuming cycle. The presence of hydrocarbons should theoretically increase zinc fuming rates. However, in practice there have been mixed messages. In some cases softer high volatile coals are beneficial, but at the Trail smelter coal consumption was found to be proportional to the fixed carbon content with little effect from the volatile content. This is said to be due to slower reaction rates of volatiles, which tend to burn above the slag bath (Yurko, 1970), but may be an artefact of mass balance control and departure from equilibrium conditions. Ash from coal dissolves in the slag and dilutes the zinc content. This tends to reduce zinc recovery and hence low ash coals generally give higher zinc recovery. The use of high ash coal at the expense of zinc recovery may, however, be justified by the lower price of that coal. As with zinc, lead will fume efficiently, possibly assisted if some sulfur is present in the slag. A slag containing 2.5 per cent Pb will be reduced to less than 0.05 per cent with a lead recovery in excess



139


of 98 per cent. Lead is fumed ahead of zinc during the first part of the cycle. Cadmium, halides and arsenic will also be fumed with high recovery. Germanium, indium, thallium, antimony and tin are fumed to varying degrees. Operation of slag fuming furnaces in a continuous rather than batch mode has been practised, but is not common. A notable exception was at the Non-ferrous Metallurgical Works at Plovdiv in Bulgaria. If operated continuously, the composition of the slag in the furnace must be that of the tail slag. Since fuming rates reduce as the zinc content falls, the fuming rate for continuous operation will be at the minimum. The reduced fuming rate needs to be offset by the time eliminated from the cycle for charging, heating and batch tapping. The initial experience with the Plovdiv plant has shown reduced capacity with continuous operation, but each case needs to be evaluated for its particular circumstances. Details of operating conventional slag fuming furnaces are given in Table 8.2. TABLE 8.2 Representative conventional slag fuming furnaces (Grant, 2004). Plant Company

Chihuahua Asarco

Chimkent

Flin Flon

Kazakhstan Hudson Bay

Plovdiv

Port Pirie

Ronnskar

Trail

Dimiter Blagoev

Nyrstar

Boliden

Teck Cominco 1998

Year commissioned

1952

1974

1952

1970

1966

1964

Number of furnaces

1

1

2

1

2

1

1

Batch

Batch

Batch

Continuous

Batch

Batch

Batch

2.3 × 2.54

4.57 × 2.44

8.1 × 2.4

6.4 × 2.44

6.25

9.00

Operation Length × width (m)

6.4 × 2.44

8.48 × 2.41

6.4 × 2.44

Height (m)

7.32

7.2

7.32

Hearth area (m2)

15.61

20.45

15.61

5.85

11.15

19.4

Tuyeres per side

14

14

21

11

15

26

21

Slag charge (t)

39

120

72 - 75

25 @ 13 t/h

40 - 45

105

70 - 85 16 - 18%

Slag – Zn%

15.61

11.7%

11 - 17%

8 - 10%

12.5 - 14%

16 - 18%

9.5%

Slag – CaO/SiO2

0.88

0.7 - 0.8

0.047

0.61

0.7

1.33

0.61

Slag depth (m)

1.02

1.5

1.2 - 1.3

1.1

0.97 - 1.09

1.44

1.13 - 1.36

Tail slag (Zn%)

1.9%

1.9%

0.7 - 1.0%

2.0 - 2.5%

2.2%

1.0%

2.5%

Cycle time (min)

80

130 - 150

180

110 - 200

120

150 - 180

Slag rate (t/day)

646

1100

560

312

380

1200

650

Fuel type

Coal

Nat gas

Coal

Oil

Coal

Coal

Coal

2.1 - 2.3

1.56

1.06 - 1.10

2.8

4.05

6.50

2.94

5.86

5.04

5.25

Fuel use (t/t zinc) Productivity (tZn per m2 day)

6.04

There are opportunities to increase fuming rates for a continuously operated furnace by dividing it into a series of compartments with progressively lower levels of zinc in each. However, the general conclusion appears to have been that the best capacity is achieved by batch operation.

TOP SUBMERGED LANCE SLAG FUMING As an alternative to the conventional slag fuming furnace, a top submerged lance slag bath reactor may be used for zinc fuming. In this process fuel as gas, oil or pulverised coal is injected with oxygen

140




enriched air through a vertical submerged lance into a molten slag bath contained in a refractory lined vertical cylindrical vessel. Lump coal can also be added as a reductant. The high degree of turbulence generated by the injected gas promotes reaction rates and heat transfer. The principles of the submerged lance rely on the cooling effects of the injected air passing down through the lance to freeze a protective layer of slag on its outer surface, otherwise the lance would be rapidly consumed at the operating temperature of around 1300°C. These principles were developed by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) in Australia with early application to tin fuming in 1978. Further development of application to zinc fuming was by Ausmelt with commercial application at the Onsan Smelter of Korea Zinc and by Mitsui at the Hachinohe ISF smelter. The process has also been applied to copper smelting and to lead smelting by Mount Isa Mines (Isasmelt process), as discussed in Chapter 7. One key part of these processes is the design of the lance so as to achieve stable operation and long life (Floyd and Swayn, 1998). A schematic of the submerged lance slag bath process is shown in Figure 8.9. To Atmosphere Gas Cooler

Fuel

Bag Filter

Primary Air Secondary Air

Fan

Feed Boiler Lance

Fume

Slag

FIG 8.9 - Schematic of submerged lance slag bath reactor.

Feed materials together with fluxes and lump coal are directly fed into the bath through a port in the reactor lid. Fuel is injected with air enriched with up to 45 per cent oxygen into the bath through a lance, which is supported on a hoist so that it can be lowered into the bath and raised free of the reactor for cleaning or replacement. The need to have an adequate volume of gas flowing through the lance to provide cooling limits the degree of oxygen enrichment possible. Secondary or after-burning air is also supplied through the lance assembly but exits from the lance above the bath level. Additional secondary air enters through the feed and lance ports, but these are normally as constricted as possible to limit this uncontrolled input. The vessel is maintained under slight suction from the gas handling system. The slag bath is highly turbulent and the reactor vessel needs to be tall (at about three times the vessel diameter) in



141


order to prevent blockages in the gas outlet duct due to splashing. Since the lance needs to be fully withdrawn from the reactor, a relatively high structure is required to accommodate the plant. Secondary combustion can generate elevated temperatures in the gas above the bath, and adding to the required cooling load before the fume is separated. As shown in Figure 8.9, gas is cooled firstly in a waste heat boiler consisting of a membrane wall shaft to drop the temperature to below the point where fume is sticky. It then enters the convective water tube section of the boiler. Additional cooling is by humidification using water sprays to drop the temperature below 200°C so that fume can then be collected using bag filters. For a small plant the cost of a boiler may not be justified and cooling by humidification alone can be used. Collection of fume using an electrostatic precipitator is also practical and will allow collection at a higher temperature than a bag filter (400 - 500°C), thus limiting the degree of cooling required. If an electrostatic precipitator is used it should contain at least three fields in series to ensure efficient removal of fume to low levels. To ensure that fume is removed and exhaust gas meets environmental standards, the use of a bag filter often has been preferred. The gas handling and fume collection facilities are essentially the same as for the conventional slag fuming furnace, although in this case it is possible to significantly reduce the volume of furnace gases by a higher degree of oxygen enrichment. Gas volumes are also reduced by fewer opportunities for gas leakage and air ingress.

Top-submerged lancing (TSL) – slag bath reactor operating principles and construction The chemical principles are the same as those described for the conventional slag fuming furnace and fuming rates are of a similar order. However, it is usually preferable to operate the TSL slag bath reactor continuously rather than in batch mode. This can avoid the need for feed slag holding facilities and larger gas handling volume capacity, which is the primary cost of the plant. The cost of the reactor is relatively small in comparison with the conventional fuming furnace and extended residence time can be more readily accommodated. Zinc is fumed from molten slag as metal vapour and is oxidised to ZnO fume above the bath by secondary air. Considerable heat is generated by secondary combustion, and can be partly captured by the bath as molten slag is projected into the gas space by the turbulence generated by the lance. This heat capture by the bath reduces the actual fuel needs within the bath and improves fuel efficiency. The high level of turbulence is in the centre of the reactor around the lance and well away from the reactor wall, unlike the conventional slag fuming furnace where injection tuyeres are located in the furnace wall and gas can rise along the wall. In the case of the conventional furnace, refractory lining of the furnace walls is not possible and the water jacket containment design is necessary. For the submerged lance reactor refractory lining is possible, but even so there is sufficient turbulence to severely impact on refractory life. In some installations refractory life was less than three months with plant availability under 80 per cent. However, this has generally improved with experience, and refractory life exceeding two years can be achieved using chrome magnesite fused grain re-bonded bricks for typical iron silica slags. Backing the refractory layer with water-cooled panels can also assist in increasing refractory life. The use of a water jacketed system alone places a limit on the operating temperature, which cannot be much above the liquidus temperature of the slag. This limitation can be relaxed with refractory lining provided critical temperatures for degradation of the bricks are not exceeded. With the use of water jackets, the formation of metallic iron during fuming must be avoided. This will limit the reducing conditions which can be applied, and the final level of zinc in tail slag should be above

142




two per cent. This is not so critical for the refractory lined reactor and final slags of one per cent Zn or below are practical. The selection of the final zinc content of slag will be determined largely by fuming rate considerations, but increased temperature can assist in comparison with the conventional furnace. Although the risk to the integrity of the furnace associated with iron formation is reduced with refractory lining, it may need to be avoided over long periods of operation with particular slags, such as ISF slag, since it can give rise to severe accretion formation and blockage of the furnace. Figure 8.10 illustrates the effect of reducing conditions as indicated by the final level of zinc in slag, on the formation of metallic iron. This data relates to the fuming of ISF slag and shows the combustion stoichiometry required to avoid metallic iron formation.

Zinc and Iron in Fumer slag %

6

Metallic Iron

5

Zinc 4 3 2 1 0 20

30

40

50

60

70

Combustion stoichiometry %

FIG 8.10 - Metallic iron and zinc content of slags versus reducing conditions (Sekiguchi and Azuma, 1998).

Due to the heat developed by secondary combustion of zinc vapour and CO above the bath, high gas temperatures can be developed in the upper section of the reactor. This can also affect refractories by exceeding their critical breakdown temperature. In this situation the use of a water-cooled wall has benefits, particularly as loss of heat at this point assists in gas cooling. Following these principles, Korea Zinc has proposed the use of cooling water circulating through copper sets within the refractory in the lower part of the reactor and in contact with the slag bath, and a water-cooled membrane wall with frozen slag layer protection for the upper section (Kim and Lee, 2000; Mounsey and Piret, 2000). These questions of containment and refractory life are the primary concerns in the application of this process and are still in an evolutionary phase at the time of writing. Otherwise the approach has the advantage of relative simplicity. Operation of the slag bath reactor in a continuous mode is feasible and in this case it is practical to use multiple stages in series with slag flowing from one reactor to another. This can increase the overall fuming rate for a given reactor volume but is also useful to separate gas streams if the feed material contains sulfur. When operated in batch mode, fuming rates for the submerged lance slag bath reactor are comparable with average rates achieved in the batch operated conventional fuming furnace as detailed above. This gives an indicative average decline in zinc content of six per cent per hour between 17 and three per cent zinc content. From Figure 8.6 the continuous fuming rate at a slag composition of two per cent zinc is equivalent to around four per cent decline in zinc content per hour.



143


For a reactor containing 1 m slag depth, the slag contained is approximately 3 t/m2 of cross-section and the zinc depletion rate is thus of the order of 0.12 t of zinc fumed per hour per square metre of furnace cross-section. This represents a crude indication of reactor capacity only, as gas flow rates are also a major determining parameter and can differ significantly from the conventional slag fumer in relation to bath cross-sectional area.

HIGH INTENSITY FUMING PROCESSES These processes generally involve the injection of zinc containing material into a high temperature reducing flame. Zinc oxide is reduced to zinc metal vapour reporting to the gas stream. Slag as a suspension in the gas stream is separated either in a settling or impingement chamber or in a cyclone. Secondary air is added to the gas stream after slag separation to burn CO to CO2 and zinc vapour is oxidised to form zinc oxide fume, which is separated after cooling by gas filtration. Principles are the same as described above for other fuming processes, with the exception that the reduction step is conducted at high temperatures, with very high reaction rates and hence small equipment volumes. High levels of oxygen enrichment are also used so that gas volumes and subsequent gas handling equipment size is considerably reduced. A number of variants of this approach have been proposed and evaluated but there are no full-scale commercial operations solely for the purpose of zinc recovery. Two processes closest to commercial development for zinc applications are the Horsehead Flame Reactor and the Contop Flame Cyclone Reactor.

FUME TREATMENT Crude zinc oxide fume produced by the fuming process contains a range of other volatile impurities, particularly lead and halides, but also contains fine particulate slag and some coal ash carried over from the fuming furnace with the gas stream. The carry-over depends to a large extent on the process used but is generally less than one per cent of slag produced for the conventional fuming furnace and between two and five per cent for submerged lance furnaces. The presence of other volatile elements can be significant, depending on the application of the fume and the process used for conversion to zinc metal. If the fume is to be used as feed to a retorting process or to a blast furnace (ISF) then the requirements are not so critical, other than the need for densification or briquetting, and either a secondary clinkering kiln or a hot briquetting process is commonly used. If the fume is to serve as feed to an electrolytic zinc process there is a need to remove halides, which are only tolerated in limited amounts in sulfate electrolytes, due to the detrimental effects on electrodes during electrolysis, as well as the occupational health risk associated with chlorine generation from open electrolytic cells. It is therefore usual to treat fume to remove halides prior to feeding to an electrolytic zinc process. Raw fume may be washed with water to remove halides. To prevent loss of zinc and other metals into the wash solution it is usual to use sodium carbonate solution for fume washing. This converts soluble zinc and other metal chlorides and fluorides to sodium chloride or fluoride and insoluble metal carbonates which are retained with the fume. Wash solutions are generally maintained at a pH of 9.0 to 9.2. Alternatively the fume can be first washed with water and the resulting solution treated with sodium carbonate or caustic soda to precipitate zinc (Chabot and James, 2000). Halides may be removed by heating the fume in a gas or oil fired rotary kiln to around 1200°C with a residence time of more than 12 hours. This can remove over 90 per cent of the chlorine content and

144




95 per cent of the fluorine content of raw fume. Secondary kiln treatment can also be used to remove significant amounts of lead, but higher temperatures are required. Removal is achieved in part by volatilisation of species such as PbCl2, PbF2 and ZnCl2. Complex fluoro-chlorides, PbClF and ZnClF are also known to be present. The hydrolysis reaction given in Equation 8.9 also takes place, with the water vapour being derived from the combustion of fuel and may be the main route for removal of halogens from the fume. ZnF2 + H2O = ZnO + 2HF

(8.9)

In general halogen removal improves with higher firing and exit temperatures. Rotary kilns used for de-halogenation require a capacity of the order of 0.6 m3 per tonne of raw fume processed per day, and typical kilns are of the order of 2.0 to 2.5 m OD and 20 to 30 m in length. Kiln slope is around 4° and refractory lining is typically 150 to 250 mm thick. Examples of commercial operations are given in Table 8.3. TABLE 8.3 Dehalogenation kiln installations (Grant, 2004). Plant year

Company

No kilns

Kiln Length (m)

O diameter (m)

Rotation (rev/min)

Chihuahua 1952

Asarco

1

22.86

2.13

1.43

El Paso

Asarco

2

18.29

2.44

1.3

Kellogg 1943

Firing end (°C)

Exit gas (°C)

Feed rate (t/d)

1400

650 - 740

100 - 112

1384 - 1440 705 - 755

130 - 140

Bunker Hill

1

22.86

2.13

Ronnskar 1964

Boliden

1

39.6

2.38

Port Pirie 1966

Nyrstar

2

27.5

2.44

To 2.5

Asarco

1

22.86

2.13

1.3

1260

International smelting and refining

1

21.34

2.13

1.1

1260

Selby 1953 Tooele

0.94 - 1.05 1290 - 1370 600 - 760 700 1150 - 1200 400 - 650

68 - 77 127 200 90

704 - 815

73

Kiln gases pass through cyclones to remove dusts and are then scrubbed with a sodium carbonate solution before discharge to atmosphere. Scrubbing liquor can be around 20 g/L Na2CO3 at a pH of 8.5. The carbonate scrubbing liquor precipitates zinc, lead and cadmium carbonates and leaves the halides in solution. The collected and precipitated solids from the scrubber are separated by thickening and filtration and are returned to the kiln feed. Fuel consumption is of the order of 3000 MJ/tonne of fume processed. Product fume from a dehalogenation kiln is usually cooled in a rotary drum cooler. The raw fume will greatly increase its bulk density and particle size, and will require grinding before feeding to a leaching operation. The presence of some impurities in the fume can influence the densification, for instance the presence of silica derived from dust carry-over from the fuming operation can be of concern if above 0.5 per cent in raw fume, since it tends to form a hard clinker, which is difficult to grind. Generally kiln processing will be more efficient in removing halides to low levels, but is more expensive than washing. The dehalogenated zinc oxide fume is then sold to and processed in a zinc smelter for reduction to zinc metal.



145


ELECTRIC ARC FUMING FURNACE If the oxidation of zinc vapour from a fuming process can be avoided, and zinc can be recovered directly by condensation, then considerable costs can be avoided and the viability of the zinc recovery process can be significantly improved. This requires the use of a sealed system to maintain a low oxidation potential and minimum volume of gas. The conventional slag fuming furnace is not suitable for this purpose and a sealed electric arc furnace is most appropriate. The concept has been applied at the Duisberger Kupferhutte smelter in Germany (Dietrich, 1962), at the Toho Zinc smelter at Chigirishima, Japan (Itoh, Yamakita and Yoneoka, 1980) and at the Glubokoe smelter in Kazakhstan, but more recently developed by Mintek in South Africa as the Enviroplas Process, using a high temperature DC plasma arc furnace rather than an AC submerged arc, operating at 1400°C and connected to a conventional lead splash condenser from the Imperial Smelting Furnace. This arrangement is shown in Figure 8.11. Slag + Coke Feed

Electrodes Splash Condenser LCV gas to scrubber Zn vapour

Cooler

Discard slag Pb-Fe alloy

Zinc metal FIG 8.11 - Electric furnace slag fuming and direct zinc recovery.

Slag and coke fines are fed into the furnace through a seal, and residual slag and an iron rich alloy are tapped continuously at a temperature of around 1350°C. This approach has been applied to lead blast furnace slag from the Belledune smelter of the Brunswick Mining and Smelting Company, containing 14 to 20 per cent ZnO, with a CaO:SiO2 ratio of 0.8 and a FeO:SiO2 ratio of 1.0 to 1.5. The single pass zinc recovery achieved was generally above 70 per cent (LeClair, Hancock and Hickey, 1998). The zinc condenser will produce zinc metal, but also a small quantity of matte and ‘Blue Powder’ as a mixture of surface-oxidised fine zinc particles as well as zinc/lead oxide dross. Condenser efficiency to metal was around 70 to 80 per cent. Excessive production of Blue Powder and condenser drosses is often a significant problem with these systems. Zinc fuming rates in the electric furnace were directly proportional to the power input, with 180 kg of zinc per hour per square metre of furnace area obtained at a power density of 1000 kW/m2 of furnace area. The zinc level in residual slag could be reduced to less than three per cent and lead to less than 0.2 per cent. Zinc fuming rates for other previous fuming operations using AC submerged arc resistance furnaces range from 10 to 70 kg of zinc per hour per square metre of furnace area.

146




TREATMENT OF LEAD SMELTER MATTES Normally copper in lead smelter feed is sufficiently low to fully dissolve in bullion and be recovered by copper drossing as part of the refining stage. However, in the case of high copper feeds (above five per cent of the lead content), a matte can be formed during smelting. This will require retention of sufficient sulfur to form the matte, which tends to be a mixture of lead, iron and zinc sulfides as well as copper sulfide. With direct smelting processes such as Kivcet, matte can be tapped directly from the furnace at the appropriate level, separate from bullion and slag. Where matte is separately produced from the blast furnace it is usual to separate slag first and then separate matte from bullion in a separate forehearth or settling pot where it can rise to the surface and freeze. Speiss can also form as a separate phase at this stage, as a frozen layer from cooled bullion. Speiss is usually a waste product, but depending on the smelter feed composition can be a point of concentration of nickel. Smelter mattes may contain:

• 20 to 35 per cent

Cu

• ten to 25 per cent

Pb


Fe

• five to nine per cent

Zn


S

The main method of processing matte is by oxidation with oxygen in a converter. Initially zinc and iron sulfides are oxidised first to ZnO and FeO respectively, followed by PbS oxidation and then Cu2S to copper metal and SO2. The oxidation of PbS can be to metallic lead as well as PbO, and the lead bullion so formed can dissolve in the copper up to four per cent. Excess lead metal can also dissolve copper. Alternatively lead can be separated from the matte by deliberate fuming or by adding a flux such as silica to the converter to enhance lead slagging. To minimise these complications it is preferable to raise the level of copper in matte to 50 per cent or more before converting. This can be done by processing low-grade matte back through the blast furnace, either as a recycle stream or as a separate campaign, as was the practice at the Hoboken smelter, recovering lead bullion and an upgraded copper matte. Alternatively with smaller quantities, a rotary furnace can be used, particularly with the addition of iron turnings to displace lead from the matte and form lead bullion. Any lead bullion recovered from matte treatment will be rich in copper and can form a significant circulating load through the lead bullion refining stage (see Chapter 12 – Copper Removal or Copper Drossing). Converting of copper matte will produce ‘blister copper’, which can be cast into anodes for subsequent electrolytic refining. Converter dusts are predominantly lead oxides and basic sulfates containing zinc and can be recycled to the lead smelting operation. In general the control of these operations and the handling of a range of side and recycle streams can be difficult and a separate matte is rarely produced.

SINTER PLANT AND SMELTER DUSTS Flue dusts and fumes collected from sinter plant and smelter gases can contain concentrations of volatile metals as well as lead and zinc. Elements which tend to concentrate in sinter plant dusts are



147


cadmium, thallium and selenium. Usually these dusts are simply recycled and find an outlet in slag, but if a high circulating load develops it can be necessary to provide a bleed by leaching these elements from the dusts before recycling. Apart from lead, blast furnace dusts are often enriched in cadmium, zinc and indium, and can be a source for recovery. Again separation before recycle may be necessary to provide an outlet and prevent the development of heavy circulating loads. In direct lead smelting with no sintering operation, some of the volatile elements will be retained and captured by lead bullion and may need special removal procedures. In particular this applies to thallium.

REFERENCES Ashman, D W, Goosen, D W, Reynolds, D G and Webb, D J, 2000. Cominco’s new lead smelter at Trail operations, in Proceedings Lead-Zinc 2000, pp 171-185 (The Minerals, Metals and Materials Society (TMS): Warrendale). Bell, R C, Turner, G H and Peters, E, 1955. Fuming of zinc from lead blast furnace slags, J of Metals, March, p 472. Blanks, R F and Ward, D W, 1968. Development of a cyclone furnace for slag fuming, in Advances in Extractive Metallurgy, pp 234-244 (Institution of Mining and Metallurgy: London). Bygden, J, Seetharaman, S, Srinivasan, N S and Täpp, J E, 1985. Application of Kellogg’s model to the slag fuming practice in Sweden, in Proceedings Zinc ’85, pp 171-183 (Min and Met Society of Japan: Tokyo). Chabot, S S and James, S E, 2000. Treatment of secondary zinc oxides for use in an electrolytic zinc plant, in Proceedings Lead-Zinc 2000, pp 739-750 (The Minerals, Metals and Materials Society (TMS): Warrendale). Dal, I, Jahanshahi, S and Rankin, W J, 1997. Effects of iron oxide and sulfur control on the rate of reduction of ZnO from slags by CO-CO2 gas mixtures, in Proceedings Fifth International Conference on Molten Slags, Fluxes and Salts ’97, pp 123-133 (Iron and Steel Society: Warrendale). Dietrich, A, 1962. The development of electrothermic zinc production at Duisberg copper works, Erzmetall, 15(4):181-189. Elliot, J F, Gleiser, M and Ramakrishna, V, 1963. Thermochemistry for Steelmaking, vol III (Adison Wesley: New York). Floyd, J M and Swayn, G P, 1998. An update of Ausmelt technology for zinc and lead processing, in Proceedings Zinc and Lead Processing, pp 861-874 (Canadian Institute of Mining and Metallurgy (CIM): Montreal). Grant, R M, 2004. Information provided in November. Grant, R M and Barnett, L J, 1975. Development and application of the computer model of the slag fuming process at Port Pirie, in Proceedings South Australia Conference 1975, pp 247-265 (The Australasian Institute of Mining and Metallurgy: Melbourne). Grimsey, E J, Li, H, Hayes, P and Hae, G L, 1993. The thermodynamics of zinc oxide in iron silicate slags, in Proceedings World Zinc ’93, pp 431-437 (The Australasian Institute of Mining and Metallurgy: Melbourne). Hancock, G C, Hart, D H and Pelton, L A H, 1970. Lead smelting and refining and slag fuming at the Broken Hill Associated Smelters Pty Ltd, in Proceedings World Symposium on Mining and Metallurgy of Lead and Zinc, vol 2, pp 790-823 (American Institute of Mining, Metallurgical and Petroleum Engineers (AIME): Littleton). Itoh, U, Yamakita, T and Yoneoka, Y, 1980. The recovery of PW zinc from lead blast furnace slags by electro-thermic process at Chigirishima smelter of Toho Zinc Co Ltd, in Proceedings Australia/Japan Extractive Metallurgy Symposium, pp 313-319 (The Australasian Institute of Mining and Metallurgy: Melbourne).

148




Izbakhanov, K S, Naimanbeav, M A, Tabynbaev, N B and Kolosov, B V, 1990. Use of natural gas and evaporative cooling for slag fuming at the Chimkent lead plant, USSR, in Proceedings Lead-Zinc ’90, pp 375-388 (The Minerals, Metals and Materials Society (TMS): Warrendale). Jak, E and Hayes, P, 2002. Phase equilibria and thermodynamics of zinc fuming slags, Canadian Metallurgical Quarterly, 41(2):163-174. Kellogg, H H, 1957. A new look at zinc fuming, Eng Min J, 158(3):90. Kellogg, H H, 1967. A computer model of the slag fuming process, Transactions of the Metallurgical Society of AIME, 239:1439-1449. Kim, M B and Lee, W S, 2000. The QSL slag fuming process using an Ausmelt furnace, in Proceedings Lead-Zinc 2000, pp 331-343 (The Minerals, Metals and Materials Society (TMS): Warrendale). Leclair, R, Hancock, P and Hickey, T, 1998. Lead smelting operations at Noranda’s Belledune plant: From vertical integration to custom smelting, in Zinc and Lead Processing, pp 109-124 (Canadian Institute of Mining and Metallurgy (CIM): Montreal). Mounsey, E N and Piret, N L, 2000. A review of Ausmelt Technology for lead smelting, in Proceedings Lead-Zinc 2000, pp 149-169 (The Minerals, Metals and Materials Society (TMS): Warrendale). Polyvyanni, I R, 1976. Oxygen and natural gas in the metallurgy of lead, Iza Nauka Kazakhakoi SSR, Alma Ata, Chapter 5. Richards, G G, Brimacombe, J K and Toop, G W, 1985. Kinetics of the zinc fuming process, Metallurgical and Materials Transactions B, 16(3):513-527. Sekiguchi, T and Azuma, S, 1998. Slag fuming at the Hachinohe smelter, in Zinc and Lead Processing, pp 299-311 (Canadian Institute of Mining and Metallurgy (CIM): Montreal). Van Negen, P, Maes, R and Cocklebergs, C, 1990. The Hoboken complex lead-copper smelter: A survey of main technological developments during the last decade, in Proceedings Lead-Zinc ’90, pp 933-951 (The Minerals, Metals and Materials Society (TMS): Warrendale). Ward, D H, 2000. Rate limitations in the slag fuming process, in Proceedings Minprex 2000, pp 471-478 (The Australasian Institute of Mining and Metallurgy: Melbourne). Yurko, G A, 1970. Slag fuming process at the Cominco smelter, Trail, British Columbia, in Proceedings World Symposium on Mining and Metallurgy of Lead and Zinc, vol II, pp 330-347 (American Institute of Mining, Metallurgical and Petroleum Engineers (AIME): Littleton).



149

CHAPTER 9 Electrochemical Reduction Processes BACKGROUND The possibility of electrowinning lead from primary concentrates has been investigated for well over a century without the development of a robust and sustainable commercial operation. There are attractions in reducing the large number of separate stages involved in conventional smelting and refining operations, but there are considerable difficulties in handling a wide range of impurities, extreme variations in the levels of minor impurities, and in the corrosive environment often encountered with appropriate systems, such as in chlorine metallurgy. Most of the early development attempts centred on the use of chloride systems that relied on the reaction of lead sulfide with chlorine to produce lead chloride and elemental sulfur, followed by electrolysis of the lead chloride to produce lead metal and chlorine for recycle. PbS + Cl2 = PbCl2 + S0

(9.1)

Cathode Pb2+ + 2e = Pb

(9.2)

Anode

2Cl2- = Cl2 + 2e

(9.3)

The simplicity of the concept was attractive and early work was undertaken by Ashcroft at Cockle Creek, Australia in the 1890s amongst others, on a dry system with electrolysis of molten lead chloride. Lead concentrates were dry chlorinated with chlorine gas and the resulting molten lead chloride was electrolysed in a heated cell using graphite electrodes. The build-up of impurities in the molten salt electrolyte was a major issue. Subsequent developments focused on multi-metal complex concentrates, producing mixed chlorides of zinc, lead and copper. Copper and lead metal were recovered by displacement from the melt by the stage addition of molten zinc, leaving zinc chloride for electrolysis in a molten salt cell. This approach may have produced a reasonable quality lead metal but the impurity problems remained and were perhaps more difficult to handle with zinc electrowinning from a molten salt (Ashcroft, 1933). The alternative line of investigation was the use of an aqueous chloride system rather than molten salts. Early work was based on a chloridising roast of the sulfide concentrates with the addition of salt. The resulting calcine was leached with hot brine to give a lead chloride–sodium chloride solution. This was purified by cementation with lead powder to remove more noble impurities and electrolysed in a diaphragm cell. The cathode deposit tends to be crystalline and dendritic or even spongy, and must be constantly removed by scraping or it will grow across to the anode, causing a short circuit. The cathode lead collects in the base of the cell, is removed, compacted to minimise oxidation, and is then melted and cast into lead ingot. Some early plants used rotating cathodes to effect continuous removal of the lead deposit. Chlorine gas is recovered from the anode compartment and in early plants with a chloridising roast, was used to make calcium hypochlorite (bleaching powder). Later operations recycled chlorine to the leaching operation for direct extraction of lead from the sulfide concentrate – eliminating the roasting step and producing elemental sulfur. In the period 1918 to 1925 five commercial plants using these methods were constructed in the USA, with one each in Canada, Australia and England. The



151

CHAPTER 9 – Electrochemical Reduction Processes

extreme corrosion problems encountered with the generation and handling of chlorine was a serious complication, and this, together with problems associated with impurities, saw the demise of these operations within a short period of time. In more recent development efforts, and to avoid difficulties associated with handling chlorine, an alternative has been the use of ferric chloride for direct oxidation of lead sulfide to lead chloride and elemental sulfur. The resulting leach solution can be electrolysed in a diaphragm cell to deposit lead at the cathode and re-oxidise ferrous iron back to ferric iron at the anode. Indeed ferric chloride is a most effective leachant for PbS with almost complete attack in less than 15 minutes at 100°C, to form lead chloride and elemental sulfur. Other hydrometallurgical systems are limited due to the need for soluble lead salts. The normal sulfate system used for other base metals is not applicable since lead sulfate is insoluble. Nitrate is unacceptable because nitrate ions are reduced at the cathode to give NO2, NO and NH3. Suitable electrolytes with high lead solubility and high conductance have been well developed for electrorefining of lead and include:

• Fluosilicic acid H2SiF6 • Fluoboric acid

HBF4

• Sulfamic acid

HNH2SO3

• Dithionic acid

H2S2O6

Dithionic acid tends to decompose at normal cell operating temperatures to H2SO4 and SO2. The sulfate will then precipitate lead and SO2 will be reduced at the cathode to H2S, which in turn will precipitate PbS. Sulfamic acid is also unstable at higher current densities and tends to break down to form ammonium sulfate, in turn precipitating lead sulfate. Hence, the most suitable practical alternative electrolytes are fluosilicic and fluoboric acid systems. One significant issue for these alternative systems, where the anion does not take place in the anode reaction, is the tendency to form PbO2 at the anode in competition with the formation of oxygen in accordance with Equations 9.4 and 9.5: Pb2+ + 2 H2O = PbO2 + 4H+ + 2e 2H2O = 4H+ + O2 + 4e

(Eo = -1.46 volts)

(Eo = -1.23 volts)

(9.4) (9.5)

Although Equation 9.5 has the lower equilibrium potential there is a significant overpotential of at least 0.5 volts depending on the nature of the anode. This then favours the formation of PbO2 over oxygen. In the case of chloride the standard potential for chlorine formation is -1.36 volts with little overpotential and hence this reaction is favoured. If ferrous iron is present then oxidation at the anode in accordance with Equation 9.6 has a standard electrode potential of -0.7 volts and this reaction will predominate. Fe2+ = Fe3+ + e (Eo = -0.70 volts)

(9.6)

The formation of PbO2 in significant amounts reduces the recovery of metallic lead from the process and may require recycle of PbO2 by reduction and re-solution. However, it has been found that various additives can suppress the formation of PbO2 on graphite anodes, such as the presence of 1.5 g/L of phosphorus in solution as H3PO4. A disadvantage is that phosphorus can result in an increase in the impurities in cathode lead. Arsenic at around 75 mg/L added as sodium arsenate will also suppress PbO2 formation on graphite anodes. The use of different anode materials can also be effective.

152




The use of iron to effect the direct leaching of lead from sulfide concentrates and its part in the anode reaction is attractive in both providing an effective leaching reagent and in minimising the voltage of the electrowinning cell, at the same time avoiding the formation of PbO 2.

PROCESSES BASED ON MOLTEN SALT ELECTROLYSIS Electrolyte composition The attractiveness of molten salt systems is the ability to produce high purity metal in molten form ready for direct casting into ingots. The cells should therefore operate at temperatures above the melting point of lead (327°C). Lead chloride melts at 501°C and hence is suitable, but it also has a high vapour pressure at practical temperatures – 2.8 mm Hg at 550°C. This leads to significant fuming from the cells, and creates problems by condensing on cold surfaces in the chlorine gas handling system. Lead metal has some solubility in molten lead chloride and hence cathode lead will dissolve to some degree into the electrolyte and will react with chlorine at the anode to reform lead chloride. This effectively reduces the current efficiency of the cell to around 85 per cent. The addition of alkaline earth metal chlorides to the molten salt mix can substantially lower the melting point and operating temperature as shown in Table 9.1, as well as the solubility of lead metal. TABLE 9.1 Liquidus temperatures for various salt mixes (Wong and Haver, 1977). Mole % in mixture

Liquidus temperature (°C)

100% PbCl2

501

28% NaCl + 72% PbCl2

410

36% LiCl + 64% PbCl2

400

49% KCl + 51% PbCl2

410

34% LiCl + 39% KCl + 27 PbCl2

320

Magnesium is generally avoided since it causes severe attack of graphite anodes. PbCl2 forms complexes in the molten salt mix with KCl, NaCl and LiCl in decreasing order of strength, and hence KCl will have the greatest effect on reducing fuming. In a tertiary salt mix, LiCl will have a major impact on lowering the melting point, allowing lower temperature operation, which also reduces fuming. LiCl also increases the conductivity of the melt to a greater extent than KCl due to its smaller cation size. Work by the US Bureau of Mines (USBM) determined an optimum salt mix of LiCl, KCl and PbCl2 as shown in the last row of Table 9.1. The operating temperature selected for an electrolytic cell was 450°C for low cell voltage, low lead solubility and relatively low fuming (Wong and Haver, 1977). Other developers have used different salt mixes with NaCl replacing LiCl or even KCl because of the cost associated with recovery from electrolyte bleed to control impurities.

Cell design A number of different cell designs have been proposed, but these are based on the use of a brick-lined cell with graphite electrodes. Major differences arise from the configuration of the electrodes. It is possible to use vertical electrodes or horizontal electrodes or even slanted electrodes. Vertical



153


electrodes have a higher risk of contact between the falling lead and rising chlorine gas, particularly with narrow electrode spacing. This will result in re-formation of lead chloride and loss of current efficiency. Horizontal electrodes allow better separation of the two product streams, and tilting of the electrodes will facilitate escape and prevent blanketing of the anode with chlorine gas. The simplest form consists of a flat cathode located just above and electrically connected to a pool of molten lead, which collects in the base of the cell. This in turn is connected through a steel bar to the power supply. The anodes are graphite plates held above the cathode using ceramic spacers and provide for the escape of chlorine gas to collect in the upper sealed space of the cell. The cell is fed with dry lead chloride, which is controlled by the density of the electrolyte, and lead metal is periodically tapped from the cell via a siphon. A schematic of the USBM cell is shown in Figure 9.1.

FIG 9.1 - Molten salt cell for lead chloride (US Bureau of Mines).

Provision is made to supply alternating current to the cell to maintain a molten electrolyte whenever electrolysis was discontinued. With an electrode spacing of 19.1 mm, cell voltage is 4.7 at a current density of 6800 amps/m2. Current efficiency is 93 per cent and energy consumption 1.32 kWh/kg of lead. With the use of meshing saw tooth surfaces on the electrodes to facilitate the removal of chlorine gas, spacing can be reduced to 12.7 mm, cell voltage to 2.5 and current efficiency can be raised to 98.7 per cent, giving an energy consumption of 0.66 kWh/kg of product lead. A more complex design is the use of a stack of horizontal bipolar electrodes in which the lower surface is the anode and the upper surface is the cathode. Only electrical connections are required for the top and bottom electrodes. The problem with this system is that current can bypass around the edges of the electrodes from top to bottom of the cell, and hence narrow gaps between the edges of the electrodes and the walls of the cell are required. Sufficient room is necessary to provide for the flow of lead to the base of the cell and chlorine gas to the top of the cell. Current bypass is reflected in a loss of current efficiency. A cell using this principle was developed by Alcoa for the production of aluminium from aluminium chloride, and has been applied to lead chloride, as shown in Figure 9.2. Electrode spacing was 9.5 mm and inter-electrode voltage 2.8. Current efficiency was 95 per cent, giving an energy consumption of 1.1 kWh/kg of lead.

154




FIG 9.2 - The Alcoa bipolar cell – designed for aluminium chloride.

Clearly electrode spacing is a critical parameter in achieving low energy use, but the design must allow for ready separation of lead and chlorine to ensure high current efficiencies. There have been a number of different designs proposed with this in mind, but it would seem that monopolar design is favoured because of relative simplicity. In general, current densities of between 1000 and 1500 amps/m2 are practical, cell voltage of 2.5 to 3.0 are achievable and current efficiencies of 95 per cent are attainable. The build-up of impurities in the electrolyte bath is a significant problem in maintaining this performance, and an accumulation of iron for instance can cause current efficiency to decline to 50 per cent. Similarly, copper can cause a major decline in current efficiency. Electrode life is unclear because of the lack of long-term operating experience. However, attack of graphite anodes can be a significant problem if magnesium is present and erosion can result from the addition of moisture with the lead chloride feed and the generation of oxygen at the anode. US Bureau of Mines work indicated that only copper, silver and zinc (to a minor extent) appeared to co-deposit with lead. Copper and silver could be removed by prior treatment of the melt with lead before electrolysis. Elements less noble than lead can build up in the electrolyte, such as calcium, magnesium, sodium, iron and zinc. These effects necessitate the feeding of relatively pure lead chloride as well as the ability to bleed or purge the electrolyte periodically to maintain impurities at a low level. These aspects of prolonged operation of a molten lead chloride cell have yet to be fully resolved.

Preparation of lead chloride feed The essential lesson from past attempts to develop processes for the production of lead by molten chloride electrolysis was that the electrolyte must be of high purity. Chlorination of lead concentrates and feeding of the resulting lead chloride directly to an electrolytic cell was unsuccessful.



155


Consequently, the first stage of any process must be the preparation of high purity lead chloride. This will inevitably require the preparation of an aqueous solution and separation of a high purity lead chloride therefrom. Fortunately, this can be facilitated by crystallisation utilising the high temperature coefficient of solubility for lead chloride. Lead chloride can be prepared by direct chlorination of lead concentrates either as dry chlorination in a suitable reactor such as a fluidised bed, or by wet chlorination in a suspension in water as part of a leaching operation. Alternatively lead concentrates can be leached with a ferric chloride solution according to Equation 9.7: PbS + 2FeCl3 = PbCl2 + 2FeCl2 + So

(9.7)

As indicated above, this reaction proceeds rapidly and efficiently and can be relatively selective with lower extraction efficiency for zinc and copper sulfides. However, there may be concerns with low extractions for silver and gold, which are key economic by-products for lead extraction. For this reason an aggressive leach is preferred, which then necessitates comprehensive purification procedures. Where ferric chloride is used for leaching and converted to ferrous iron, it can be regenerated by reaction of the solution (after lead chloride removal) with chlorine from the electrolytic cell. The can be done using a packed absorption tower irrigated with depleted solution in exchange with an upflow of gas from the anode compartment of the cells. Whichever method is used for preparation, a hot aqueous leach solution is the end point, from which a pure lead chloride must be produced. Cementation with lead powder can be used to separate and recover copper, silver, gold and bismuth as well as removing arsenic and antimony. Separation of lead chloride by crystallisation can then be used, relying on the high temperature dependence of lead solubility as shown in Table 9.2. TABLE 9.2 Solubility of lead in chloride solutions. Temperature (°C)

PbCl2 concentration in water (g/L)

PbCl2 concentration in a FeCl2/NaCl solution (g/L)

20

9.9

23

60

18

49

95

31.7

98

100

33.4

110

The solubility of lead in brine solutions is much higher than in water alone, due to the formation of strong chloro complexes. Commonly, sodium chloride concentrations of the order of 250 g/L are used, together with iron concentrations for ferric iron leaching of the order of 25 g/L. This also applies to the leaching of silver, which can be readily solubilised in brine solutions, whereas AgCl is quite insoluble in water alone. Table 9.3 shows the equilibrium chloride concentrations at which particular complexes form. There is clearly benefit in using brine solutions with a high chloride concentration where solubility differentials of the order of 50 g/L of lead can be obtained. Many process developments have therefore proposed to use crystallisation simply by cooling the leach solution as a means of obtaining pure lead chloride. The crystals must then be thoroughly washed and dried in preparation for feeding to a fused salt electrolytic cell.

156




TABLE 9.3 Chloro complex formation (equilibrium chloride concentration for complex formation at 25°C). Complex

Total Cl concentration (g/L)

PbCl+

2

PbCl3-

14

PbCl42-

140

AgCl32-

8

3-

AgCl4

263

Processes using these techniques are the US Bureau of Mines process (Wong, Haver and Sandburg, 1980), the UOP process (Stauter, Tolley and Um, 1978), and the St Joe Minerals Corporation developments (Bounds, 1980). However, there has been insufficient long-term experience on a commercial scale to know whether there are any significant issues with the transfer of particular impurities with the lead chloride. Indications from pilot testing are that cathode lead of 99.99+ per cent quality can be produced. A typical flow sheet is shown in Figure 9.3 representing the USBM process. Lead Sulfide Concentrate NaCl / FeCl 3 solution

Make up reagents

Leaching

Bleed Stream Treatment

Crystallisation Leach residue and sulfur

To by-product recovery

Filtration and Washing Lead chloride

Drying

Molten Salt Electrolysis

Chlorination

Chlorine

Molten lead

Casting

Product Lead

FIG 9.3 - Flow sheet for aqueous ferric chloride leaching – molten salt electrolysis (US Bureau of Mines).



157


PROCESSES BASED ON AQUEOUS ELECTROLYSIS Chloride systems Rather than use molten salt chloride electrolysis, it is possible to directly electrolyse an aqueous brine solution of lead chloride. As indicated in the Background section above, the cathodic deposition of lead from a chloride solution will be crystalline and dendritic, and needs to be continuously removed from the cathode by scraping or shaking. The lead deposit will collect in the base of the cell and can be continuously removed using a conveyor system such as a screw. Because the finely divided lead will oxidise readily it needs to be compacted by briquetting or extruding through high pressure rolls before feeding to the melting pan. If a lead chloride–brine solution is electrolysed, then chlorine will be evolved at the anode and can be collected and used for leaching. An acidic brine solution is preferred for high conductivity and cell voltages of around 2.5 to 3.0 are possible with current efficiencies for lead deposition of 95 per cent. The PLACID process, developed by Tecnicas Reunidas in Spain, has been applied to electrowinning lead from lead carbonate derived from secondary sources, and uses an electrolyte of lead chloride and sodium chloride only. In this case the cell has a diaphragm permeable only to hydrogen ions and produces an acidic spent electrolyte (hydrochloric acid) used for leaching lead carbonate. The anolyte is an acidic solution free of chloride so that oxygen is produced rather than chlorine. The dendritic or sponge deposit of lead is shaken from the cathodes and collected by conveyor belt. The lead is pressed to remove electrolyte and form platelets, which are fed into a melting kettle. Energy consumption is reported as 0.9 kWh/kg of lead (Frias, Garcia and Diaz, 2000). A schematic of the PLACID flow sheet is shown in Figure 9.4. It is also possible to electrolyse a solution containing iron in the ferrous state in a compartmented– diaphragm cell. Ferrous iron will not interfere with the cathode reactions but will be oxidised to ferric iron in the anode compartment. To achieve this, feed solution enters the cathode compartment and electrolyte passes through the diaphragm, constructed of an inert fabric, to the anode compartment and exits the cell. The anolyte can then be reused for leaching. Any escape or return of ferric iron to the cathode compartment will result in its reduction to ferrous iron in preference to lead deposition, thus reducing process current efficiency for lead recovery. In the case where leach solution is used for cell feed, purification of the solution by cementation with lead powder is necessary, as well as the treatment of a bleed to prevent the build-up of other impurities such as zinc, calcium and magnesium. It is also important that the iron present in feed solution is fully in the ferrous state to maintain high current efficiency by avoiding the reduction of ferric iron at the cathode. To ensure this, it is preferable to have a two-stage counter-current leach in which excess sulfides are in contact with the final leach solution. Some excess ferrous iron (10 to 15 g/L) should also be maintained in the anolyte to impede the formation of chlorine at the anode. Solution compositions are generally around 80 g/L Pb in feed solution, reducing to 20 g/L in spent electrolyte, which allows the cell to operate at a temperature of around 40°C. Iron levels are around 40 - 50 g/L and sodium chloride at 250 g/L. Cell voltages will be high due to the resistance of the diaphragm and extra spacing involved at 3.5 to 4.5 volts depending on design. Current efficiencies of 95 per cent can be achieved with current densities of 200 amps/m2. Electrodes can be stainless steel for the cathode and graphite for the anode, although there are reported difficulties with graphite attack. The alternative anode is titanium or ruthenium coated titanium. In the case of titanium, the use of expanded metal against the back of the diaphragm to provide an internal channel for the anolyte is a favoured approach, particularly if current densities are relatively low.

158




Oxide Lead Feed NaCl / HCl solution

Leaching Lime

Sulfate Removal Leach residue Gypsum Cement to by-product recovery

Purification

Oxygen Water

Cathode

Anode

H+ Cathode lead powder Briquetting Melting & Casting

Product Lead

FIG 9.4 - Aqueous chloride electrolysis for lead (PLACID process).

The Minemet process follows these principles but also incorporates an additional purification stage following cementation. This uses ion exchange with IMACTI-GT73 resin with good selectivity for copper and silver to remove the last traces prior to lead electrowinning. The resin is stripped with HCl and the strip solution is recycled back to first stage purification (Demarthe and Georgeaux, 1980). A schematic flow sheet is shown in Figure 9.5. The INTEC process is another alternative chloride-based system proposed for the direct recovery of lead from sulfide concentrates (Everett and Moyes, 1992). The process was primarily developed for the processing of copper concentrates using the cupric/ cuprous ion couple for oxidation; however, later developments used a halogen complex for oxidation and this is more suited to the application to lead extraction where copper must be absent during electrodeposition of lead. The halogen complex or ‘Halex’ is BrCl2, and is generated at the anode as an alternative to chlorine gas. It is quite soluble and avoids the complexities and difficulties associated with chlorine handling. The electrolyte is basically a brine solution containing around 250 g/L NaCl and 25 g/L NaBr together with lead. Lead is leached from the sulfide concentrate using a multistage counter-current operation, in essence according to Equation 9.8: 3PbS + 2BrCl2 = 3Pb2+ + 4Cl- + 2Br- + 3So



(9.8)

159


PbS concentrate

Stage 1 Leach Lead powder

Stage 1 Purification Stage 2 Leach

HCl

Cementate to by-product recovery

Stage 2 Purification Leach residue + sulfur

IX eluate

NaCl / FeCl 3 solution

Anode

Cathode

Cathode lead powder Bleed stream processing

Briquetting Melting & Casting

Product Lead

FIG 9.5 - Aqueous ferric chloride leaching and electrolysis (Minemet process).

Any iron present in the leach solution is removed by precipitation as goethite (FeOOH). The solution is then purified by cementation with lead powder to remove copper, silver, bismuth, arsenic and antimony, leaving impurities such as zinc, which do not co-deposit with lead. The purified solution is electrolysed in a diaphragm cell, which uses a coated copper cathode and a titanium mesh anode coated with ruthenium and iridium oxides. An ion permeable membrane separates the cathode and anode compartments. The cathode is fabricated from a dimpled copper sheet coated with an inert adhesive sheet between the dimples, leaving numerous sites of high current density to promote dendritic growth of the lead deposit. The crystalline lead falls from the cathode and is collected in the base of the cell. Halex is regenerated in the anode compartment according to Equation 9.9: Br- + 2Cl- = BrCl2 + 3e

(9.9)

Spent electrolyte can then be treated to remove impurities such as zinc or removed as a bleed stream before recycle to the leaching stage. The INTEC process follows similar principles to other aqueous chloride electrowinning systems, but uses ‘Halex’ in place of chlorine gas, avoiding the difficulties of handling chlorine. Halex is a strong oxidant and can attack many minerals including pyrite. Consequently, there is less selectivity

160




in the leaching step than from the use of ferric chloride, giving rise to an increased load of impurities. It would appear that methods of handling impurity control and removal from the electrolyte are far from resolved for application of this technology to the extraction of lead.

Aqueous versus molten salt electrolysis The use of aqueous chloride electrolysis in comparison with molten salt systems has the disadvantage of producing a powdered lead cathode deposit in comparison with molten lead. The production of chlorine can be common to both systems but can be avoided in the aqueous system if iron leach solutions are used as the electrolyte, or proton permeable membranes are used to allow for a separate anolyte solution composition. No clear preference has emerged to date from the many process options examined.

Fluosilicate systems Electrolysis of lead fluosilicate solutions is well established for lead refining by the Betts process, as detailed in Chapter 13. The principal advantage is the ability to produce a dense cathode deposit rather than a powder deposit. The process has largely been examined for the treatment of secondary lead materials, which are converted either into lead carbonate or PbO from mixed sulfate–oxide residues, and are then leached in fluosilicic acid. The US Bureau of Mines developed an approach along these lines as an extension of the Betts electrorefining process to electrowinning (Cole, Lee and Paulson, 1981). This approach was also developed and applied on a commercial scale by RSR Corporation in the USA (Prengaman and McDonald, 1990). The lead fluosilicate solution is electrolysed in open tank cells using lead starter sheet cathodes and graphite anodes coated with PbO2. Two possible competing anode reactions can occur, as shown by Equations 9.10 forming oxygen and Equation 9.11 forming PbO 2. H2O = 2H+ + ½ O2 + 2e

(9.10)

2H2O + Pb2+ = PbO2 + 4H+ + 2e

(9.11)

If PbO2 forms it must be recovered from the anode and be recycled. It does not reduce effective current efficiency for cathode lead, but creates a troublesome recovery and recycle operation, and hence must be suppressed. As discussed above, various additives will suppress the formation of PbO2, such as phosphoric acid developed by the USBM or sodium arsenate. The RSR process uses arsenic at 0.5 to 1.0 g/L. Effectively H2SiF6 is formed at the anode and is recycled to the leaching stage, and in the RSR process the leach solution contains 150 g/L Pb, and the recirculating electrolyte is held at 60 g/L Pb with H2SiF6 at 63 g/L. Antimony can build up in the electrolyte to 2 g/L, after which it is no longer extracted during leaching. Arsenic, antimony, bismuth or zinc in solution will not co-deposit with lead, but copper will deposit and needs to be removed. In the RSR operation this is removed from the melted cathode deposit by conventional drossing techniques rather than by leach solution purification. Additives such as glue and lignin sulfonate are needed to ensure smooth, dense cathode deposits. Current densities up to 200 amps/m2 are practical, cell voltages are in the range of 2.6 to 3.0 and operating temperatures around 35 to 40°C.



161


Where an anode producing oxygen is in contact with the fluosilicate electrolyte, graphite has been found to deteriorate; however, this can be corrected either by coating the graphite with PbO2, as used by RSR Corporation, or the alternative use of titanium also coated with PbO2, as proposed by the USBM (Smith, Sandberg and Cole, 1979). For processes aimed at direct leaching of lead sulfide concentrates the most practical application is to make use of the ferric/ferrous iron couple. As with the chloride system, ferric fluosilicate can be used as an effective leaching reagent to produce a solution of lead according to Equation 9.12: Fe2(SiF6)3 + PbS = 2FeSiF6 + PbSiF6 + So

(9.12)

The leaching rates are highly temperature dependent and are quite high at temperatures above 60°C. In contrast zinc sulfide leaching rates are far slower giving rise to the possibility of selective leaching. However, this does not consider the extraction of valued by-products from lead concentrates such as silver and copper. The leach solution can be purified by cementation with lead powder and can then be electrolysed to produce high purity lead at the cathode. Ferrous iron is oxidised at the anode of a compartmented cell in the same way as for the chloride system above. The advantage of this reaction is that electrode potentials are sufficiently low to avoid competing reactions, which either form oxygen or PbO2. This will mean that ordinary graphite can be used with long life.

Fluoborate systems Fluoborate systems are generally similar to fluosilicate, but it has not been widely used for lead electrochemical processes due to its higher cost. However, fluoborate does have distinct advantages over fluosilicate:

• solution conductivity is higher, • smooth deposits of lead can be produced at much higher current densities, • the vapour pressure of HF above a hot solution is an order of magnitude lower, and • the fluoborate ion is more stable. The use of fluoborate for lead electrowinning from both secondary and primary sources has been developed by Engitec SpA in Italy and is known as the FLUOBOR process (Olper, 1998). The FLUOBOR process uses the ferric/ferrous couple with oxidation of ferrous iron in the anode compartment of a diaphragm cell. Oxygen and PbO2 formation are completely avoided and a simple graphite anode can be used. Primary lead sulfide concentrates are leached with ferric fluoborate solution in much the same way as for ferric chloride detailed above. The solubility of lead fluoborate is much greater than the solubility of lead chloride and the kinetics of the leaching reaction are faster. This gives more possibility of selective leaching for lead than is possible with ferric chloride leaching, but again selectivity to lead may not be beneficial for processing sulfide concentrates where by-product values from silver in particular can be of major importance to venture economics. Leaching conditions are a pH of less than one and a temperature of 50 to 90°C. The leaching reaction is shown in Equation 9.13: 2Fe(BF4)3 + PbS = 2Fe(BF4)2 + Pb(BF4)2 + So

162


(9.13)



Leach residue and elemental sulfur are separated by filtration and the clarified solution is treated with lead powder to cement out impurity elements Cu, Ag, Sb, As, Se, Te and Bi. The purified solution is then electrolysed in a diaphragm cell with a 316 stainless steel cathode and graphite anode. Lead is deposited as a dense deposit, which is periodically stripped from the stainless steel cathode. Solution flows through the diaphragm to the anode where ferrous iron is oxidised to ferric iron for return to the leaching stage. Cathode current density is around 300 amps/m2 and cell voltage is about two, giving an energy consumption of around 550 kWh/tonne of recovered lead. In order to remove impurities such as iron and zinc, which will build up in the electrolyte, a bleed stream is processed to remove these elements by a crystallisation procedure. Elemental sulfur can be extracted from the leach residue using perchlorethylene as a solvent at 60°C and is crystallised from the solvent by cooling to 20°C. Otherwise, flotation techniques for sulfur recovery followed by melting and filtration could be used. The FLUOBOR process has been extended to the treatment of secondary materials or pastes from lead–acid batteries, and to the refining of impure lead bullion. For refining, crude smelter bullion is granulated and then leached using a ferric fluoborate solution in a rotating stainless steel drum. Impurities with a higher electrochemical potential than lead are not dissolved and remain in the leach residue so long as they are in contact with residual lead. Other impurities will dissolve but will generally not be deposited on the cathode, except possibly for tin. However, ferric iron can oxidise Sn2+ to Sn4+, which will precipitate as metastannic acid, H4SnO4, which can be removed by filtration before electrolysis. For the treatment of battery pastes, which are a mixture of PbSO4, PbO2, PbO and metallic lead, a sulfidisation process has been proposed, converting all lead components into PbS, which can then be subjected to ferric fluoborate leaching as above. Sodium sulfide can be used for this step but involves a complex procedure for its recovery. Alternatively, a biological sulfate reduction process developed by PAQUES Biosystems of the Netherlands has been applied (Olper et al, 2000).

REFERENCES Ashcroft, E A, 1933. Chlorine smelting with chloride electrolysis, Trans IMM, 43:151-255. Bounds, C O, 1980. Chloride based lead extraction – Possibly an economic solution to the environmental restraints on lead smelting, in Proceedings Canadian Institute of Mining and Metallurgy (CIM) 82nd Annual Meeting (Canadian Institute of Mining and Metallurgy (CIM): Montreal). Cole, E R, Lee, A Y and Paulson, D L, 1981. Electrolytic method of recovery of lead from scrap batteries, Research investigation no 8602 (US Bureau of Mines). Demarthe, J M and Georgeaux, A, 1980. Hydrometallurgical treatment of lead concentrate, in Proceedings Lead-Zinc-Tin ’80, pp 426-444 (The Minerals, Metals and Materials Society (TMS): Warrendale and American Institute of Mining, Metallurgical and Petroleum Engineers (AIME): Littleton). Everett, P K and Moyes, A J, 1992. The INTEC copper process, in Proceedings International Conference on Extractive Metallurgy of Gold and Base Metals, pp 287-292 (The Australasian Institute of Mining and Metallurgy: Melbourne). Frias, C, Garcia, M and Diaz, G, 2000. New clean technologies to improve lead acid battery recycling, in Proceedings Lead-Zinc 2000, pp 791-801 (The Minerals, Metals and Materials Society (TMS): Littleton). Liddell, D M (ed), 1962. Chlorine metallurgical processes, in Handbook of Non-Ferrous Metallurgy, vol II, pp 1151-1188 (McGraw-Hill Book Company: New York). Olper, M, 1998. Fluoborate technology – A new challenging way for primary and secondary lead processing, in Proceedings Lead and Zinc Processing Symposium, pp 185-198 (Metallurgical Society and Canadian Institute of Mining and Metallurgy (CIM): Montreal).



163


Olper, M, Maccagni, M, Buisman, C J N and Schultz, C E, 2000. Electrowinning of lead battery paste with the production of lead and elemental sulphur using bioprocess technologies, in Proceedings Lead-Zinc Symposium, pp 803-813 (The Minerals, Metals and Materials Society (TMS): Warrendale). Prengaman, R D and McDonald, H, 1990. RSR’s full scale plant to electrowin lead from battery scrap, in Proceedings Lead-Zinc ’90, pp 1045-1056 (The Minerals, Metals and Materials Society (TMS): Warrendale). Smith, L L, Sandberg, K G and Cole, E R, 1979. Method of producing lead dioxide coated anode, US Patent 4,159,231, June 16. Stauter, J C, Tolley, W R and Um, R T, 1978. The recovery of lead from sulphide concentrates using a chlorination/brine leach/electrolysis process, in Proceedings ACS Meeting, Anaheim, March. Wong, M M and Haver, F P, 1977. Fused salt electrolysis for the production of lead and zinc metals, in Proceedings International Symposium on Molten Salt Electrolysis in Metal Production, pp 21-29 (Institution of Mining and Metallurgy (IMM): London). Wong, M M, Haver, F P and Sandberg, K G, 1980. Ferric chloride leach – Electrolysis process for the production of lead, in Proceedings Lead-Zinc ’80 Conference, pp 445-454 (The Minerals, Metals and Materials Society (TMS): Warrendale and American Institute of Mining, Metallurgical and Petroleum Engineers (AIME): Littleton).

164



PART C SECONDARY SMELTING This part of the text covers the preparation of secondary feed materials and the smelting processes used for lead extraction. This predominantly covers the treatment of scrap lead-acid batteries. Chapter 10 Chapter 11

Secondary Materials and Pretreatment Secondary Smelting Methods

CHAPTER 10 Secondary Materials and Pretreatment INTRODUCTION As indicated in Chapter 1, global lead consumption from secondary sources approached four million tonnes per year in 2005, or 60 per cent of total world consumption. Around 20 per cent of world consumption is for uses where recycling is difficult, such as for plastics stabilisers, for TV tube glass, for shot and ammunition. Of the remainder, ten per cent is used for rolled or extruded alloys and cable sheathing, which have long-term applications, and 70 per cent is used for batteries. Recyclable lead therefore is predominantly from used automotive batteries, with some from reclaimed sheet, cable sheathing and other metallic scrap. In addition there are various residues, drosses and flue dusts containing lead. Secondary residues are often handled by primary smelters as a supplement to concentrate feeds. Secondary smelters accept metallic scrap, but are primarily oriented to the processing of scrap lead acid batteries, which represent more than 85 to 90 per cent of secondary smelter feed. Up to the 1970s the secondary industry was often small scale and localised in major population centres for ease of collection of used automotive batteries. It was also oriented to the return of the lead produced to the battery manufacturers. This was not difficult since most batteries used antimonial lead alloy for the grid material and hence the secondary smelter could recover both antimonial lead and soft lead, which could be suitably blended for return to the battery manufacturers. Since that time there have been significant changes, as follows:

• Increased environmental regulations affecting the gas and fugitive dust emissions from the smelter, as well as constraints on the wastes produced, such as smelter slags.

• Stricter occupational health regulations, which have impacted on handling methods and the generation of dust and general exposure to lead in the workplace.

• Changes in the design of lead acid batteries, particularly the grid alloys for the purpose of producing maintenance free batteries, such as sealed or valve regulated, recombinant batteries. Alloys have changed from antimony lead to calcium-tin-lead with many proprietary minor additives, and requiring much higher purity lead supplies.

• Changes in the non-metallic components of the battery such as the case and separator materials. Polypropylene has virtually completely replaced hard rubber or ebonite casings, and has significant recyclable value in itself. These changes have increased the operating costs of secondary smelting operations and have increased the need for some refining of secondary lead rather than simply blending for suitable antimony contents. Separation of battery components is more important because of the value of recovered plastic components, in particular polypropylene. Economies of scale have become significant and secondary operations have become far more automated and capital intensive in order to meet these changed requirements. As a result smelters have become part of major metal producing corporations rather than small local enterprises. There has also been a general trend to restrict the disposal of lead acid batteries in waste landfill, mandating their collection and recycling. Many countries introduced schemes for battery recycling, which included the levying of charges on new batteries to cover the costs of collection and recycling,



167

CHAPTER 10 – Secondary Materials and Pretreatment

or a deposit system to provide incentive for return of an old battery when it is replaced. This has led to a significant increase in the availability of secondary feed and secondary lead production. Originally smelting involved the processing of whole batteries, usually after draining of the electrolyte, either in a reverberatory hearth furnace or in a blast furnace. However, differences in the composition of grid lead, which is alloyed, and paste lead, which is high-grade lead, gave a significant advantage to separation before smelting. Metallic lead also only needed to be melted without consuming high-grade fuel such as coke. Problems with chlorine from the separators producing volatile lead chloride, and the recoverable value of polypropylene from battery cases all contributed to the benefits from separation of the battery into its various components and the separate handling of each component as appropriate. Few operations now process whole batteries and breaking and separation prior to smelting is the common approach.

LEAD-ACID BATTERY COMPOSITION To understand the requirements for processing scrap lead-acid batteries, as the predominant component of lead scrap, it is useful to review the operation of a battery and its construction. The lead-acid battery uses a positive electrode or plate composed of PbO2 as the active material, and a negative electrode or plate, composed of sponge lead as the active material. The electrolyte is dilute sulfuric acid and upon discharge both active materials are converted into lead sulfate according to the reactions shown in Equations 10.1 and 10.2. This explanation of the action of the battery is referred to as the ‘double sulfate theory’. It may be a simplistic view, but broadly explains the essential principles of the operation of the battery. Positive plate

PbO2 + SO42- + 4H+ + 2e = PbSO4 + 2H2O

Negative plate Pb + SO42- = PbSO4 + 2e

(10.1) (10.2)

Reading Equations 10.1 and 10.2 from left to right is the battery discharge reaction, and from right to left is the charging reaction. Clearly, as the battery discharges the concentration of sulfuric acid in the electrolyte decreases and hence the density of the electrolyte will fall, which is used as a measure of the charge condition. The active materials are held within a lattice or grid made of lead metal to act as the current conductor and into which the materials are pasted. Many different techniques are used for the manufacture and preparation of battery plates, but will not be covered in this text. The active paste material changes in volume as the above reactions take place and must be designed as a porous structure to accommodate these changes over many charge and discharge cycles, without degradation and spalling of material from the plates. Positive and negative plates should be close to reduce the electrical path through the electrolyte, but must avoid the possibility of contact, which causes an internal short circuit. Hence separators are used as inert porous plates, usually constructed of PVC or microporous polyethylene. The general construction of the lead-acid automotive or starting lighting ignition (SLI) battery using the above principles is shown in Figure 10.1. Grids need to be corrosion resistant and have reasonable strength and dimensional stability. Soft lead is unsuitable and antimony was successfully used for many years as an alloying element for this purpose. However, antimony tends to migrate into the negative active material and lowers the hydrogen overvoltage on that material, increasing the generation of hydrogen during charging and the

168




1. Lead alloy grid 6. Negative electrode set for one cell 2. Separator - porous 7. Full electrode and separator assembly insulating plate for one cell 3. Positive plate 8. Battery case 4. Separator 9. Positive terminal 5. Negative plate 10. Positive electrode set for one cell FIG 10.1 - Construction of the automotive lead-acid battery.

associated decomposition of water and loss from the battery. In these circumstances the battery has to be topped up with water from time to time, requiring ‘maintenance’. The presence of antimony is also one of the main causes of self-discharge of the battery on standing. This led to the development of lead calcium alloys to minimise grid corrosion and gassing, giving rise to the ‘maintenance free’ battery. With the ability to avoid water decomposition and gassing, the electrolyte can be immobilised by the use of glass microfibre mats or by gelation with silica, giving rise to the sealed battery. Antimonial lead alloys used typically contain 4.5 to 6.0 per cent Sb, 0.1 to 0.4 per cent As, 0.2 to 0.5 per cent Sn and 0.02 to 0.06 per cent Cu. There are also a range of antimonial alloys containing up to 4.5 per cent Cd. Calcium lead alloys contain 0.05 to 0.08 per cent Ca, 0.1 to 2.0 per cent Sn and around 0.05 per cent Ag. Overall, the lead-acid automotive battery is close to 60 per cent lead and is typically made up of the following components:

• sulfuric acid electrolyte, 24 per cent by weight (around 18 per cent H 2SO4); • grids – metallic lead, 22 per cent by weight; • top lead – connector posts, four per cent by weight; • pastes, 38 per cent by weight; • case – polypropylene, seven per cent by weight; and • separators – five per cent by weight. An average analysis of separated pastes from scrap batteries is shown in Table 10.1. Separators are primarily PVC and contain around 21 per cent chlorine.



169


TABLE 10.1 Average battery paste analysis. Lead sulfate

PbSO4

50 - 60%

Lead dioxide

PbO2

15 - 30%

Litharge

PbO

5 - 10%

Metallic lead

Pb

2 - 5%

Antimony

Sb

0 - 0.5%

Carbon

C

2 - 4%

Silica

SiO2

0.5 - 1.0%

Moisture

H2O

8 - 15%

Total lead content

69 - 75%

Based on the above average figures, if the components are separated and only the pastes need to be processed in the smelting operation, the input tonnage to be processed is reduced to around 50 per cent in comparison with whole battery smelting.

BATTERY BREAKING AND SEPARATION A number of proprietary battery breaking systems are available, in particular those offered by MA Industries and the CX System from Engitec–Tonolli. The sequence of operations for battery breaking and separation is shown in Figure 10.2. Primary breaking is either by hammer mill or by spiked rolls in combination with a knife shredder. Batteries are handled by front end loader into a hopper with apron feeder discharging onto a conveyor and into the primary breaker. Breakage releases the dilute sulfuric acid electrolyte and equipment must be constructed of stainless steel and be fully enclosed to contain acid spray and fine particulates. If crushing rolls are used the discharged material is wet screened to separate fines and electrolyte, and the coarse material is passed under a magnet to remove any tramp steel and is then processed in the knife shredder. The material crushed to around -50 mm is passed over a vibrating screen with a flow of water to wash out fine paste material. The screen underflow passes to a settler where the pastes are settled out and the overflow is recirculated over the screen or discarded to either acid reclaim or to neutralisation and disposal. The primary screen oversize consists of grid metal, separator plastics and case plastics. It is fed to a water filled tank where the polypropylene case material floats and is collected and washed. The remainder of the material is extracted from the bottom of the tanks by drag or screw conveyor and is passed to a hydrodynamic separator or elutriating column where an upward flow of water separates out plastic material as overflow from metallic lead, which sinks to the base of the column and is extracted by drag or screw conveyor. Coarse materials are dewatered and washed on fine mesh screens and stockpiled for further processing. Metallics may simply be melted in a pot or standard lead kettle or can be melted in a short rotary furnace, with some collection of a litharge slag to separate antimony and arsenic and produce a soft lead. Separator plastics are usually disposed of in landfill but polypropylene is thoroughly washed, and extruded as pellets for sale to battery manufacturers for new case production.

170




Whole Batteries

Alternative Route

Roll Crusher

Hammer Mill

Oversize Wet Screen

Magnetics

Undersize Shredder Settler

Acid Discard Oversize

Wet Screen

Magnetics Undersize

Float Separator

Polypropylene

Acid discard Elutriation Separator

Pastes

Separator plastics Metallic Lead

FIG 10.2 - Generalised flow sheet of battery breaking and separation.

Battery pastes may be directly sent to smelting or can be further processed for desulfurisation prior to smelting. In some cases the final separation of metallics and plastics from battery separators may not be included and the mix is sent to the secondary smelter. The plastics can be used as fuel for reduction in smelting but have the disadvantage of introducing chlorine and associated problems with volatile chlorides in the smelter gas handling system, as well as causing the formation of noxious compounds in exhaust gases such as dioxins.

Sulfuric acid reclamation Sulfuric acid may be reclaimed for reuse in batteries by filtration of discard from the primary paste collection settler, or it may be neutralised with lime and the resulting gypsum filtered off and discarded.



171


Because of the low cost of sulfuric acid and the difficulties of handling and storing as a dilute solution (around 20 per cent), there is generally no justification for acid reclamation and reuse. However, in situations where the battery recycling facility is part of a larger battery manufacturing operation on the same site, then acid reclamation would be sensible. Also, if it is part of a primary smelting operation with a sulfuric acid plant then the dilute recovered acid can be used to provide make up water for acid production and hence can be recovered. The East Penn Manufacturing Co Inc in Pennsylvania, USA as a major battery manufacturer with secondary operations has developed an acid recycling facility in which filtered reclaimed acid is subjected to a solvent extraction process to remove metal impurities – principally iron. Iron is reduced from a level of around 15 mg/L to less than 1.5 mg/L by extraction into an organic solvent. It is electrolytically reduced to the ferrous state and is stripped from the solvent using a low volume of more concentrated sulfuric acid (Leiby, 1993).

PASTE DESULFURISATION In the smelting of battery pastes the sulfur from lead sulfate is either captured in the slag as a matte or reports to the smelter gases and must be removed before discharge to atmosphere by scrubbing. Many secondary processes were designed to operate with soda-iron slags, which captured sulfur as Na2S or FeS matte as the bulk of the smelter slag. Depending on performance of the operation some sulfur still reported to smelter gas, but at the time was sufficiently low to permit discharge to atmosphere. However, regulations have tightened and in some instances scrubbing is now necessary in addition to the use of soda-iron slags. A further complication is the stability of soda-iron slags in landfill dumps. FeS in wet acidic conditions can oxidise to Fe2O3 and elemental S, releasing considerable heat, which can be sufficient to raise the temperature and cause ignition of the sulfur and the remainder of the matte. This process causes expansion and disintegration of the slag and exposes the material to leaching of heavy metal components. Consequently the slag is treated as a hazardous material with associated high disposal costs. As a result there has been a move towards the use of more conventional inert silica slags and the elimination of sulfur from the secondary smelter feed by using a paste desulfurisation process. Desulfurisation depends on the fact that lead carbonate has a much lower solubility than lead sulfate. Hence lead sulfate can be converted to the carbonate by treatment with a solution of sodium or ammonium carbonate. With the use of ammonium carbonate the reaction is shown in Equation 10.3. The paste is agitated with a strong carbonate solution for one or two hours until the S content is reduced from around five per cent to less than 0.5 per cent. PbSO4 + (NH4)2CO3 = (NH4)2SO4 + PbCO3

(10.3)

The ammonium carbonate solution can be regenerated by first reacting the ammonium sulfate solution with lime, as in Equation 10.4, to form gypsum, which is separated by filtration to leave an ammonium hydroxide solution. This is then treated with CO2 by absorption in a packed tower to reform ammonium carbonate solution as in Equation 10.5. CO2 can be generated from kiln gases from the burning of limestone to produce the lime required for the first stage.

172

(NH4)2SO4 + Ca(OH)2 + 2H2O = CaSO4.2H2O + 2NH4OH

(10.4)

2NH4OH + CO2 = (NH4)2CO3 + H2O

(10.5)




With the use of sodium carbonate, the solution pH is higher and there is a tendency to produce a basic carbonate (hydrocerrusite), according to Equation 10.6: 3PbSO4 + 2Na2CO3 + 2H2O = Pb3(CO3)2(OH)2 + 2Na2SO4 + H2SO4

(10.6)

With treatment at high pH or temperatures above 55°C there is a tendency for sodium to enter the solid phase as NaPb2(CO3)2(OH), and if sodium carbonate is present in excess, the double salt Na2CO3.2PbCO3 can form. These effects will incorporate sodium into the desulfurised paste, as can the presence of NaCl in the solution, which is usually the case for industrial grade soda ash. Hence one to two per cent Na is usually always present in paste desulfurised by this method. Conversion of lead to carbonate is also lower at around 85 to 90 per cent and the sodium must be discarded in smelter slag (Queneau et al, 1998; Stout, 1993). An alternative to carbonates for desulfurisation is sodium hydroxide producing hydrated lead oxide according to Equation 10.7: PbSO4 + 2NaOH = Na2SO4 + PbO.nH2O + (1-n)H2O

(10.7)

This results in a cleaner separation with no sodium present in the desulfurised paste (Reynolds, Hudson and Olper, 1990). The sodium sulfate solution resulting from past desulfurisation can be filtered, evaporated and crystallised to produce sodium sulfate for sale. It is used in water treatment and for detergent manufacture as the quality from this source is sufficiently free of metals for this purpose. In the case where sodium carbonate is used the residual solution is usually neutralised with waste acid to remove any residual carbonate before evaporation and crystallisation. Alternatively it is possible to regenerate sodium hydroxide by reaction with lime to remove sulfate as gypsum, according to Equation 10.8: Na2SO4 + Ca(OH)2 + 2H2O = CaSO4.2H2O + 2NaOH

(10.8)

Engitec have proposed the electrolysis of the sodium sulfate solution in a diaphragm cell to regenerate sodium hydroxide at the cathode and sulfuric acid at the anode to give a battery grade sulfuric acid as the anolyte.

Paste sulfate reduction A novel approach in treating paste in order to provide feed for a paste lead electrowinning process is proposed by Engitec, in which all lead containing salts are converted into PbS by using anaerobic bacteria. This follows the successful commercial application of sulfate reducing bacteria to treat smelter waste waters for the separation and recovery of heavy metals as sulfides. Reactions are as in Equations 10.9 to 10.11: PbSO4 + 4H2 = PbS + 4H2O

(10.9)

PbO + H2SO4 + 4H2 = PbS + 5H2O

(10.10)

PbO2 + H2SO4 + 5H2 = PbS + 6H2O

(10.11)



173


The bacterial leaching step requires the injection of H2 and CO2, which can be produced by cracking natural gas. The carbon dioxide is used for bacterial growth and the hydrogen as an electron donor for the reduction reactions as above. The lead sulfide is leached with ferric fluoborate solution to dissolve lead and produce elemental sulfur. The resulting lead–ferrous fluoborate solution is electrolysed in a diaphragm cell to produce a pure lead cathode and a ferric fluoborate solution at the anode for recycle to leaching. This process is also applicable to primary lead production from lead sulfide concentrates and thus allows a high degree of integration of secondary and primary lead smelting operations. In any event the conversion of pastes into lead sulfide allows integration with most primary smelting operations (Olper et al, 2000).

PROCESSING OF SECONDARY RESIDUES There are many lead containing residues from other metallurgical operations including flue dusts, drosses and leach residues, often containing lead as lead sulfate and often containing a wide range of other impurity elements. Large tonnages of such residues are available from the electrolytic production of zinc as sulfate leach residue and from copper smelting as flue dusts. In general these materials contain substantial quantities of gangue elements such as silica, lime and iron oxides, but often contain substantial amounts of precious metals which represent the bulk of the value. Smelting will produce much slag as well as a relatively impure bullion requiring extensive refining. This is quite incompatible with normal secondary smelting, oriented to scrap battery processing, and hence the treatment of such secondary residues is usually handled by primary smelters as an addition to concentrate feeds. Depending on the smelting process used there can be limits to the acceptable proportion of feed represented by such residues. For the sinter plant–blast furnace combination the input of sulfate residues is commonly limited to around 25 per cent of net new feed. The decomposition of lead sulfate is endothermic and can limit the attainment of peak bed temperatures in the sinter plant, affecting the quality of sinter produced, which is critical to blast furnace performance. This is discussed in detail in Chapter 4. Many of the direct smelting processes covered in Chapter 7 are more able to cope with residue feeds, particularly the Kivcet and Kaldo processes.

REFERENCES Leiby, R A, 1993. Secondary lead smelting at East Penn Manufacturing Co Inc, in Proceedings EPD Congress 1993, pp 943-957. Olper, M, Maccagni, M, Buisman, C J N and Schultz, C E, 2000. Electrowinning of lead battery paste with the production of lead and elemental sulfur using bioprocess technologies, in Proceedings Lead-Zinc 2000, TMS Symposium, Pittsburg, pp 803-813 (The Minerals, Metals and Materials Society (TMS): Warrendale). Queneau, P B, James, S E, Downey, J P and Livelli, G M, 1998. Recycling lead and zinc in the United States, in Proceedings Zinc and Lead Processing – CIM Symposium, Calgary, pp 127-153. Reynolds, R M, Hudson, E K and Olper, M, 1990. Advances in lead acid battery recycling: Engitec’s automated CX Breaker system, in Proceedings Lead-Zinc ’90 Symposium, Anaheim, pp 1001-1022 (The Minerals, Metals and Materials Society (TMS): Warrendale). Stout, M E, 1993. Secondary lead recovery from spent batteries, in Proceedings EPD Congress 1993, pp 967-979.

174



CHAPTER 11 Secondary Smelting Methods GENERAL Secondary lead smelting is predominantly based on the treatment of scrap lead-acid batteries, as detailed in Chapter 10. A wide range of smelting methods are used for the treatment of secondary lead, originally based on traditional smelting methods such as the hearth and blast furnace, feeding whole batteries, but are changing under stricter environmental controls to the use of prior separation and specialised reduction in rotary furnaces or newer electrochemical techniques. Processes are essentially classified by the type of equipment used, as follows:

• reverberatory furnace based processes, coupled with additional slag reduction equipment such as the blast furnace;

• blast furnace; • rotary kiln; • short rotary furnace; • submerged lance slag bath reactor; • electric furnace; and • electrochemical treatment methods. In many operations a combination of two processes is used; the first for bulk smelting of total lead containing scrap and the second for reduction of lead-rich slags produced in the first stage. Often impurities can be concentrated into the first stage slag, reducing the quantity of lead requiring refining. Figure 11.1 shows the more common combinations and interrelationships between the primary and secondary smelting processes. The primary smelting stage feed varies from whole drained batteries to disintegrated whole batteries, mixed paste and metallics after separation of plastics, pastes only after separation of metallics and plastics, and pastes after desulfurisation. Some processes are unsuited to different types of feeds. The feed for the secondary smelting stage in a combination is the slag from the primary process. Bullion from each stage may be acceptable for reuse in battery manufacture or it may require some refining. Where slag recycle is shown in Figure 11.1 it is usually done on a separate campaign basis to enrich an alloying element into a separate bullion stream. For example it is common to accumulate first stage blast furnace slags for recycle through the furnace in a campaign. This serves to recover extra lead as well as concentrate alloying elements such as antimony, which tend to preferentially report to the slag.

REVERBERATORY FURNACE The reverberatory furnace is possibly an extended use of the old hearth smelting process. It has generally been replaced by other methods, but is still widely used in the USA. The furnace is rectangular in shape with a shallow hearth and arched roof. The floor and roof slope down to the firing end, and the opposite end contains a feeding port where the feed mix is continuously or intermittently pushed into the furnace. A natural gas or oil burner flame extends to the



175

CHAPTER 11 – Secondary Smelting Methods

Whole Drained Batteries

Slag recycle

Battery Breaking

Blast Furnace

Short Rotary Furnace Plastics

Separation Stage 1

Rotary Kiln Slags for secondary smelting

Paste + Metallics Metallics

Separation Stage 2

Top Lance Slag Bath Reactor

Blast Furnace

Reverberatory Furnace

Electric Arc Furnace

Kettle Melting Pastes

Short Rotary Furnace

Slag recycle

Paste Desulfurisation

Short Rotary Furnace

Leaching & Electrowinning

FIG 11.1 - Secondary smelting process options.

centre of the furnace and melts the charge, which flows into a pool on the hearth, settling as a lower layer of lead bullion and an upper layer of slag. Slag continuously overflows, whereas bullion can be tapped intermittently or continuously through an underflow weir and lead well. It is often the practice to adjust the level of bullion on the hearth to maintain a relatively shallow slag layer, without causing bullion to be lost with the slag. Furnace sizes vary widely, with hearth areas from 2 to 70 m2. Construction commonly uses chrome magnesite brick to the slag level and high alumina brick above. Suspended arch roofs are used in the larger furnaces. A schematic of the reverberatory furnace is shown in Figure 11.2. The charge consists of broken battery components premixed with coal or coke fines at about five per cent by weight. The charge is fed by conveyor onto a mechanical stoker which pushes it into the furnace. Initially the charge is dried, the metallic lead then melts followed by PbO, allowing the reactions shown in Equations 11.1 to 11.6 to proceed. Because the battery pastes are closely mixed with the grid metal in the charge, lead metal acts as a reducing agent to convert PbSO4 and PbO2 to PbO. PbS is an intermediary which can be oxidised according to the ‘roast reaction’ – Equation 11.3 or reaction with PbO as in Equation 11.4. PbO is also reduced to lead metal by carbon as in Equation 11.5.

176




Exhaust ports

Arched Roof

Exhaust gas Drying feed

Feed Burner

Feed port

Working ports

Bullion

Hearth

Slag

FIG 11.2 - Reverberatory furnace schematic.

Because the furnace conditions are not strongly reducing, a PbO-rich slag will be present and antimony, arsenic and tin tend to be oxidised and remain in the slag as per Equation 11.6. The lead bullion produced is thus relatively free of alloying components, although any copper, bismuth and silver present in the feed mix is captured by the bullion. Pb + PbO2 = 2PbO

(11.1)

4Pb + PbSO4 = 4PbO + PbS

(11.2)

PbS + PbSO4 = 2Pb + 2SO2

(11.3)


(11.4)

2PbO + C = 2Pb + CO2

(11.5)

2Sb + 3PbO = 3Pb + Sb 2O3

(11.6)

Furnace gases are cooled either by direct water injection or by indirect heat exchange to below 200°C and are then cleaned in a bag filter. Collected dusts are recycled to the feed mix. Although most secondary smelters using this technology in the past discharged filtered gas directly to atmosphere, the SO2 content no longer makes this permissible in most locations. Hence gas scrubbing to remove SO2 using lime or ammonia solutions is necessary for the application of this technology. Slag production is roughly one tonne for every four tonnes of bullion produced and typical slags contain around 75 per cent Pb and close to 90 per cent of the antimony in the feed material, with bullion commonly reporting around 0.1 per cent Sb. The reverberatory furnace is not efficient for reduction and the slag is normally treated in separate equipment with a higher reduction potential. The blast furnace has commonly been used, but also the short rotary furnace and electric furnace are used for this duty. The bullion produced from reduction of reverberatory slag is enriched in alloying elements such as antimony and tin and requires refining or careful blending to produce suitable alloys. In some instances first run reverberatory slag can be re-run through the furnace on a campaign basis. The slag is mixed with coal at around five per cent of feed and more than half the lead content



177


can be recovered as bullion. The resulting second run slag is enriched in impurity elements and reduced in volume for separate treatment. However, this does leave an intermediate bullion stream with somewhat higher impurity levels and will possibly need to be refined.

THE BLAST FURNACE The blast furnace is used in two modes, one for the smelting of whole batteries and the second for the reduction of reverberatory slags. A key feature of the blast furnace is its requirement for bulk feed to maintain an open structure in the shaft so that gas flow is not impeded. As a result it is not able to process fine material and application to separated battery paste reduction is impractical. Secondary blast furnaces are much smaller than primary smelter furnaces and are often circular in cross-section rather than rectangular as with larger furnaces. Furnace shaft diameter at the tuyeres ranges from 0.7 to 1.5 m, and working height is from 2.5 to 3.8 m. Larger furnaces, mainly used for whole battery smelting, tend to be of rectangular cross-section similar to the primary smelter units (see Chapter 5).

Whole battery smelting This is typified by the Varta process implemented by Varta Batterie AG at Hannover, Germany and as used by Boliden-Bergsoe AB at Landskrona, Sweden. The batteries are drained and charged with fluxes (lime, silica and shredded scrap iron), coke and an amount of recycle lump slag. The slag has a melting point above 1000°C and maintains an open structure in the shaft until it nears the tuyere zone with a temperature of around 1150°C. Coke use is around 180 kg per tonne of crude lead bullion. Larger furnaces of rectangular cross-section are generally required in this case and the Varta furnace is 4 m2 in hearth area and 7 m high. The furnace used at Britannia Lead in the United Kingdom (UK) was similar at 1.1 m wide by 4 m long at the tuyere level with eight tuyeres per side and 4.7 m high (Koch and Niklas, 1989; Taylor and Moore, 1980). Water is evaporated in the upper part of the shaft, followed by decomposition of the plastic components as the temperature rises to 500°C. Lead also melts and reacts with PbO2 to form PbO. This is reduced by CO to lead further down the shaft, and sulfates are reduced to PbS. A feature of the Varta process is the use of metallic iron to react with PbS and form lead and FeS as a matte, thus capturing the majority of the sulfur in feed. The crucible at the base of the furnace contains lead bullion and a mixed slag–matte phase. Lead bullion is usually tapped continuously via the lead well into a ladle, and slag may be tapped intermittently into slag pots where the matte separates. The solidified slag and matte can be broken up and mechanically separated. The lump slag is crushed to an appropriate size and part is returned to the furnace. The matte phase contains excess iron and is represented as (FeS + 0.35Fe). It contains 25 to 26 per cent sulfur and six to nine per cent Pb as entrained metal. Matte provides the main outlet for sulfur and is usually disposed of in landfill. Furnace gases pass through a fuel fired afterburner to raise gas temperature to 1000°C in order to combust organics distilled from decomposing plastics in the upper shaft. After burning the gases pass through a drop-out chamber and are then cooled and filtered through a bag house before discharge to atmosphere. Although most of the sulfur is captured as matte this will not be complete, and the advent of tighter emission standards may make scrubbing of final gas necessary. The destruction of organics by afterburning requires added fuel in addition to the CO content of the gas and is a significant inefficiency in the energy requirements of this process.

178




Chlorine from PVC separators forms volatile lead chloride, which vaporises in the lower shaft and condenses in the upper shaft, forming a recycle loop as well as wall accretions. Some lead chloride reports to flue dusts, again causing accretions in the gas handling system, and collected dusts must be leached for removal of chlorine before recycling to the furnace feed. Another difficulty with the processing of whole batteries is the tendency of the cases to melt in the upper part of the shaft to form a sticky mass bridging across and blocking the furnace. The blast furnace is also a high energy consumer for various reasons, not the least being the requirement to melt a high circulating load of slag. Furnace heat is supplied by the burning of coke, which is a costly fuel. Together with the complications of consuming plastic materials, the bulk of which can be a valuable by-product, and the environmental issues with disposal of lead-rich mattes, these problems have rendered the blast furnace smelting of whole batteries an unattractive approach.

Blast furnace smelting of reverberatory slags The furnaces used in this case are relatively small and commonly circular in cross-section. A schematic of a typical small-scale furnace is shown in Figure 11.3. Construction of the shaft is water jacketed mild steel, sitting on a brick crucible fitted with lead well and siphon. The top is sealed using a cone seal through which the charge is periodically dropped. The furnace feed is crushed reverberatory furnace slag and it is fuelled with coke at around ten per cent of the charge. Lime, silica and shredded scrap iron are added as fluxes, making up about 20 per cent of the charge. Feed

Charge Hopper Cone seal

Gas Offtake

Water jacketed shaft

Shaft

Dam Tuyere Bullion spout Slag spout Crucible

FIG 11.3 - Typical secondary blast furnace cross-section.



179


In some cases the bullion produced can be rich in antimony and to achieve high antimony recoveries under these circumstances it is necessary to avoid speiss formation, and hence keep the iron input relatively low. In turn this means that slags are lower in iron than normal primary smelter slags with FeO:SiO2 ratios at 0.8:1 and CaO:SiO2 at 0.9:1. This compares with typical primary smelter slag ratios of 1.3 and 0.7 respectively. The lead content of slag is typically between one and three per cent. Where lower antimony slags are processed and recovery is not a critical issue, then more typical lead blast furnace slags can be used. Where feed has a significant sulfur content, iron can be added to capture the sulfur as a matte, and in this case relatively high iron slags can be produced with FeO:SiO2 ratios above 2.0 and CaO:SiO2 ratios below 0.5. Blast furnace gases pass through an afterburner to convert CO to CO2, and are then cooled by water injection and cleaned in a bag filter before discharge to atmosphere. Collected dusts are recycled to the reverberatory furnace. It is usual to operate the blast furnace on a campaign basis, stockpiling feed until sufficient has been accumulated for a reasonable campaign run of from one to four weeks. It is also possible to re-run blast furnace slags under different conditions, particularly if it is desired to concentrate antimony or other alloying elements into a small quantity of final bullion. This mode of operation provides considerable flexibility and allows the furnace to be cleaned between runs, minimising operational difficulties due to accretion build-up and maximising the efficiency of the furnace (Pike, 1990).

THE ELECTRIC ARC FURNACE The electric arc furnace has been applied as a replacement for the blast furnace by RSR Corporation for the purpose of treatment of reverberatory furnace slags (Eby, 1990). RSR recycle first run slags through the reverberatory furnaces to produce a second run slag enriched in alloying elements such as antimony. The second run slag reports 35 to 50 per cent Pb, five to ten per cent Sb, one to three per cent As and one to three per cent Sn and is fed to the electric arc furnace mixed with coke fines at five to seven per cent, limestone at four to seven per cent and iron turnings at zero to five per cent of the charge weight respectively. Mixed feed is dried in a rotary dryer using furnace off gases and air to burn CO and is dropped through a port in the furnace roof. Lead oxide is reduced by carbon to metal along with antimony and tin oxides to produce a bullion containing these elements. A residual slag floats on the surface of the bullion and is maintained at a minimum depth of at least 150 mm to provide a resistance path for the electrodes. Any shallower depth has the danger of electrical shorting through the underlying bullion. During operation slag depth will increase to 450 mm. Slag temperature is relatively high and high iron levels are not required to maintain a fluid slag as is the case in the blast furnace. Hence FeO:SiO2 ratios can be in the range 0.15 to 0.4, with CaO:SiO2 ratios around 1.0. This will avoid the formation of a speiss phase. Sulfur in the charge will form a matte phase usually composed of CaS and FeS. The lead content of slag is between 0.5 and two per cent and the total of lead, antimony, arsenic and tin is generally less than 4.5 per cent, compared with six to 12 per cent for the blast furnace. The slag is disposed of in landfill. The RSR furnace is 4.8 m internal diameter by 4 m high and is fitted with three 355 mm diameter graphite electrodes, rated at 4.0 MVA. Electrode consumption is around 7 kg/t of slag smelted. Construction is a water-cooled steel shell lined with chrome magnesite brick, with a suspended domed roof. Electrodes are sealed into the roof.

180




The furnace is operated by slowly adding feed to the furnace and building up a layer of solid charge to a set level. Feed is then stopped and power is increased to fully melt the charge, and at completion the slag is tapped and run down to the minimum level. During the above cycle of about eight hours from one slag tap to another, bullion is also periodically tapped. Energy consumption per cycle is 12 000 to 14 000 kWh. Lead production per cycle is understood to be of the order of 30 tonnes.

ROTARY FURNACE SMELTING Two broad types of rotary furnace are used – the rotary kiln and the short rotary furnace. Rotary kilns are characterised by large length to diameter ratios of the order of 10 to 15:1, whereas for the short rotary furnace the length to diameter ratio rarely exceeds 1.5:1. One important aspect of the operation of these furnaces is the slag regime used. Traditionally, a high sodium slag has been used since it had the characteristics of a relatively low melting point and low viscosity, allowing for good separation of slag and bullion. It also provided a means of capturing sulfur into the slag rather than reporting to the smelter gas and the need for costly gas scrubbing. However, there are significant costs in the use of soda ash and there are environmental problems with the disposal of soda slags (see below). As a result of these issues the lime–iron–silica slags used in primary smelting have become more favoured.

Slag regimes The use of soda slags is primarily a means of capturing sulfur from the smelting operation and the slag is essentially a sulfide or matte – usually a mixture of FeS and Na2S. The normal fluxes used are shredded iron scrap as cast iron chips or steel turnings and soda ash (sodium carbonate). Small amounts of lime and/or silica might be added depending on the composition of feed material, but is generally not needed with pure battery feeds. For the smelting of battery pastes the reactions given by Equations 11.7 to 11.14 take place: PbO + CO = Pb + CO2

(11.7)

CO2 + C = 2CO

(11.8)

PbSO4 + 4CO = PbS + 4CO2

(11.9)

PbS + CO + Na2CO3 = Pb + Na2S + 2CO2

(11.10)

PbS + Fe = FeS + Pb

(11.11)

PbO2 + Pb = 2PbO

(11.12)

PbO2 + 2CO = 2CO2 + Pb

(11.13)

The gas phase mainly contains CO and CO2, but some SO2 can be present due to limited reaction according to Equation 11.14 – ‘the roast reaction’: PbS + PbSO4 = 2Pb + SO2



(11.14)

181


It should be noted that CO2 generated from sodium carbonate will react according to Equation 11.8 and thus will lead to increased fuel consumption. For this reason and because of the cost of soda ash, the balance between the addition of iron and soda ash should favour a greater proportion of iron. The slag resulting from the above is basically Na2S and FeS with a small amount of PbO and possibly CaO and SiO2. A phase diagram for the Na-Fe-S system is given in Figure 11.4 and shows that between 25 and 75 per cent FeS the melting point is relatively low. The lowest melting point eutectic is 640°C at 36 per cent FeS and the highest for a liquid slag is 760°C at 60 per cent FeS. 1200

Temperature °C

1000

Liquid

800

3 Na 2 S.4FeS 760°C

5Na 2 S .2FeS (730°)

695°C 650°C

640°C

600

400 0

20

60

40

80

100

Composition %

Na2S

FeS

FIG 11.4 - Na-Fe-S system phase diagram.

A high FeS content in the matte tends to increase the lead content and reduce lead recovery to bullion by reducing the reaction given by Equation 11.11. The equilibrium constant (k) for Equation 11.11 is given by Equation 11.15: Log k = -1610/T + 2.388

(11.15)

where: T is the temperature in K At 1000°C, k = 13.3, so that the activity of PbS in the matte is the activity of FeS divided by 13.3. A high FeS content of the matte also has more tendency to precipitate Fe or magnetite (Fe3O4) at lower reduction potentials. These are solid phases at normal operating temperatures and will raise the viscosity of the matte or slag mix, tending to increase the loss of lead as entrained metal. An optimum Na2S:FeS ratio in matte is regarded as 1:1.86. If significant slag forming components are present, such as in residue smelting, a separate slag phase will form in equilibrium with the matte. The Na2O activity coefficient in slag is low compared

182




with matte and will cause sodium to transfer from the matte to the slag until activities are equal. The sodium composition of slag will thus be higher than matte and the sodium contained in slag is essentially wasted as a means of capturing sulfur. Consequently, for cases where a significant amount of slag is produced the aim should be for a low Na2S:FeS ratio in the matte of little more than 1:4. A reasonable level of sodium is needed to maintain low slag viscosity and high levels unnecessarily consume soda ash as a costly reagent. Some Na2S will also dissolve in slag and an optimum level of sodium in slag is regarded as 25 per cent as (Na 2S + Na2O). An iron–soda matte–slag mix tends to be emulsified and is difficult to separate (Queneau, Cregar and Mickey, 1989). Although soda slags provide a ready means of capturing sulfur and provide a low viscosity slag/ matte at relatively low temperature for the efficient separation from bullion, there are significant associated problems, which may be listed as follows:

• Soda ash is a costly reagent. • Soda slags are very aggressive to chrome magnesite brickwork and to steel if exposure occurs. • Soda slags are hygroscopic and can expand and break down in landfill. Under these conditions the FeS can be oxidised to Fe2O3 and So, generating heat and raising the temperature sufficiently to ignite the sulfur and the remaining sulfides.

• The slags contain water soluble heavy metal compounds, which can be leached in landfill situations, particularly when breakdown occurs as above. Because of the above, disposal of soda slags are classified as hazardous wastes with significant disposal difficulties and are unacceptable in many locations, hence there has been significant incentive to change to a more conventional silicate slag regime as used in primary smelting (see Chapter 5). This will require an increase in operating temperature to 1150 to 1200oC, which can be assisted by the use of oxygen enrichment. A matte can still be generated using metallic iron fluxing, but sulfur has more potential in this case to report to the gas phase. The solubility of sulfur in the silica slag is also limited, and in general the sulfur should be less than one per cent in the feed material. This will essentially require desulfurisation of the battery pastes prior to smelting with the use of silica slag. Target silica slags are generally of the olivine structure with melting points in the range of 1100 to 1200°C and composition as follows:

• ten to 15 per cent

CaO

• 40 per cent

SiO2


FeO

Fluxing materials required are silica, shredded iron scrap and lime, with coal as the reductant.

The rotary kiln It is possible to process whole batteries in the rotary kiln, but more often it is used to smelt leadbearing components after separation of plastic materials. Thus the rotary kiln can perform similar functions to the blast furnace but with a higher degree of flexibility due to the ability to handle fine material, and it is free of many of the problems associated with the blast furnace. The primary disadvantage is the relatively short life of refractories. Kilns are constructed as a steel shell fitted with two or three outer rings as tyres, which sit on trunnion rollers. The kiln is inclined at between two and four per cent. A large ring gear attached to



183


and encircling the outside of the kiln is driven by a pinion gear, which in turn is driven by an electric motor. Rotational speed is usually variable from around 0.3 to 1.5 rev/min depending on kiln diameter, but usually the nominal design speed represents a peripheral speed of 10 m/min. The kiln is lined with high alumina brick at the feed end and chrome magnesite brick in the reaction zones and discharge end. This brickwork overlays a full backing of insulating firebrick. The discharge end is narrowed to form a dam for the collection of molten lead and slag. Figure 11.5 shows a schematic of a smelting kiln installation. Stack

After Burner Drop-out Chamber Battery Scrap Feed

Coke

Bag Filter

Soda Ash

Iron

Dusts

Screw conveyor

Burner Kiln

Slag Bullion

FIG 11.5 - Rotary kiln – secondary smelter.

Feed is composed of battery materials, other lead-bearing residues and scrap, fine coke or coal, and fluxes. The fluxes used depend on the type of slag and can be soda ash (sodium carbonate) and shredded iron scrap for the production of soda slags or limestone, silica and iron for the production of silica slags. Separate bins and weighfeeders for each component are used to continuously blend the feed, which is fed by conveyor into the kiln. The kiln is fired at the discharge end with a natural gas or oil burner and the flame is directed at the surface of the slag pool, since its main use is control of the viscosity of the slag to ensure good separation of lead bullion. The burner is used for initial heating on start-up, but the oxidation of coal and the exothermic heat of reduction of PbO are sufficient to maintain the operating temperature of the kiln. Gases leaving the feed end are cooled and filtered in a bag house before discharge to atmosphere. Dusts collected in the flues and bag house are recycled to the kiln feed. Slag continuously overflows the dam at the discharge end and is collected in slag pots where it solidifies and can be crushed for campaign reprocessing if required or for disposal. Bullion is intermittently tapped through a tap hole by stopping the kiln, opening the tap hole and rotating the kiln so that the tap hole is at the bottom. Lead is collected via a chute into a ladle. Rotary kiln smelters are operated at the Harz-Metall plant at Goslar, Germany and at the Pedricktown plant of NL Industries in New Jersey, USA. The Harz-Metall kiln of 3.1 m internal diameter by 40 m long had an average capacity of between 130 and 300 t/d of lead bullion depending on the nature of the feed material processed. Performance details are reported as in Table 11.1 on the basis of the production of a soda slag.

184




TABLE 11.1 Performance data for the rotary kiln smelting process (Schenker, 1990). Feed material

Metallics

Oxides

Refinery drosses

Lead residues

Feed rate (t/d)

350

270

400

200

Cl rich flue dusts 240

Lead bullion output (t/d)

300

195

300

130

150

Number of tappings per day

12

8

12

5-6

6

Iron chips flux (kg/t feed)

70

160

160

160

120

Soda ash flux (kg/t feed)

40

90

90

90

320

Reduction coal (kg/t feed)

40

70

70

70

100

Slag make (kg/t lead)

210

350

340

460

960

Fuel gas used (kWh/t lead)

500

890

690

990

860

The composition of rotary kiln slag varied depending on the compounds to be eliminated in the slag, which are principally S, SiO2 and CaSO4, and is within the range shown in Table 11.2. TABLE 11.2 Composition of rotary kiln slag. FeO

Na2O

S

SiO2

CaO

C

Pb + Sb

25 - 45%

13 - 27%

10 - 15%

4 - 8%

1 - 2.5%

2 - 6%

3 - 5%

In one variant of the rotary kiln by the ‘Oliforno process’ operated by Accumulatoren – Fabrik, Oerliken, Switzerland, the kiln is used to incinerate whole batteries to produce a molten slag and lead bullion, which are simply granulated together on discharge from the kiln. The kiln is designed with a relatively steep incline to allow molten lead to drain rapidly and not be held up in the kiln and possibly be oxidised. The mixed granulated material is then processed in a short rotary furnace with coal or coke for reduction to lead bullion. The plastic components of the battery provide the bulk of the fuel for the kiln, but the problems of handling chloride rich dusts and sulfur in kiln gases remain.

Short rotary furnaces These furnaces are the workhorses of the secondary smelting industry. They have similarities with the reverberatory furnace, but are operated on a batch basis, providing considerable flexibility in being able to adjust the fluxing regime. The thorough mixing action greatly enhances its ability to conduct reduction reactions, as well as allowing optimum reaction time for drossing. The rotary furnace is used on a campaign basis to process separated battery components, such as melting the metallic lead fraction and separately reducing pastes. It can be used for controlled oxidation of bullion for drossing out impurities such as arsenic and antimony, as well as reduction of oxide drosses to concentrate impurities such as antimony. The construction of the short rotary furnace is relatively simple. It consists of a drum fitted with two steel ring tyres resting on trunnion rollers and driven through a ring gear. Diameters range from 2 to 5 m with lengths from 2.5 to 7 m, and usually have a burner fitted in one end with the gas exit flue



185


located above the burner at the same end. The charging door is fitted in the other end as well as the tapping port, with access for both charging equipment and tapping ladles. A schematic of a typical installation is shown in Figure 11.6.

Stack Mixed Feed

After-burner & Quench chamber

Charge Hopper Ventilation Air

Drop-out Chamber

Charge

Slag Pots

Burner

Bag Filter

Dusts recycled

Rotary Furnace Bullion Holding Pot

FIG 11.6 - Schematic of a short rotary furnace installation.

Furnace lining is usually chrome magnesite brick and the life of the refractories is relatively short. Most installations have more than one furnace to allow for downtime for re-bricking. The most common furnace size is 3.5 m diameter by 5 m long with an effective working volume of 7 m3. Maximum rotational speed is 1 rev/min. Burner capacity is of the order of 2.5 MW and furnace gas volume 10 000 Nm3/h. This volume may double with afterburning and dilution prior to gas cleaning. Ventilation air covering the furnace surrounds and ancillary operations can total up to 50 000 Nm3/h for such a furnace installation. Operation is on a batch basis and components are commonly premixed with coal and fluxes before feeding. The furnace is fed by means of a retractable conveyor through the charging door and loaded by front end loader, or more commonly by a ‘spoon’ hopper fitted to the front of a fork-lift truck, which is pushed into the furnace and rotated through 180 degrees to dump the contents. After completion of charging the furnace is intermittently rotated to turn over the material for drying, which takes about two to three hours. When drying is complete the furnace is rotated continuously under maximum burner operation. On completion of meltdown a second charge is added followed by a second drying and melting cycle. A third charge can sometimes be added. Following final meltdown the furnace is run at operating temperature for a set time of around two hours and is then ready for tapping. Temperature is raised prior to tapping to facilitate the separation of slag and bullion. The tap hole is opened and the furnace rotated to first tap slag and matte, followed by bullion into separate ladles. Furnace gases usually pass through an afterburner, followed by a drop-out chamber, a cooler and bag house before discharge to atmosphere. Bag house areas used are typically 100 m2 per m3/s of gas passing.

186




Collected dusts are recycled to feed mixing. Dusts produced from smelting battery pastes can be up to ten per cent of the bullion production, but this can be reduced by more than 50 per cent if the feed is crudely pelletised in a drum or disc pelletiser. Short rotary furnaces have been used for processing whole batteries as well as mixed components and separated pastes and metallics. The furnace has been widely used for processing pastes with the production of soda slags. A typical sulfur balance for a short rotary furnace is given in Table 11.3. TABLE 11.3 Sulfur distribution in a rotary furnace with soda slags. Output stream

S deportment

Slag

80%

Bullion

7%

Dusts

11%

Gases

2%

The problems associated with soda slags detailed above have led to a shift to desulfurisation of pastes and the reduction of desulfurised pastes with the production of lime silica slags. This has necessitated furnace operation at higher temperature and the use of oxygen enrichment of the gas burner. It is preferable to separately process pastes and metallics because of the different impurity or alloying elements present. Metallic components are simply melted in the furnace, but may be oxidised to some extent to separate calcium, antimony, arsenic and tin into a small volume of PbO-rich slag. That slag can be accumulated and separately treated by addition of coke or coal at the end of the cycle to produce an impure bullion, which can be sent to a refinery. In past practice the main impurity was antimony and the resulting lead–antimony alloy could simply be used for the production of specific antimonial lead alloys. Currently there is a much larger range of alloying elements for battery grids, particularly calcium and tin. Calcium will be oxidised readily and report to the final slag, whereas tin will report to the impure bullion with antimony and will need to separated by standard refining procedures. Paste materials are separately processed with the addition of coke fines and fluxes as detailed above, and generally produce a high-grade lead bullion. Table 11.4 illustrates performance of a short rotary furnace of 3.5 m diameter for various feeds and with the production of a soda slag. TABLE 11.4 Performance data for a typical short rotary furnace (Schenker, 1990). Feed materials

Whole batteries

Metallics

Oxides (pastes)

Refinery drosses

Feed rate (t/d)

90

130

60

100

Lead bullion production (t/d)

70

110

45

80

Reduction coal (kg/t feed)

40

80

120

Slag (kg/t lead) Gas use (kWh/ t lead) Cycle time (hours)


150 - 200

60

750

50

410

170

870

470

10.5 (Three loads)

12 (Four loads)

7

4.5


187


TOP BLOWN ROTARY CONVERTER (TBRC) The TBRC is represented by the Kaldo process, developed and operated by Boliden in Sweden and detailed in Chapter 7. This process is suitable for both primary and secondary smelting and can accept whole batteries if required. The Kaldo system at the Ronnskar smelter has adequate facilities for the capture of sulfur and paste desulfurisation is unnecessary.

TOP LANCE SLAG BATH REACTORS Two proprietary designs are available as the Isasmelt process and the Ausmelt process, both following developments of the CSIRO in Australia for both primary and secondary lead smelting and as detailed for primary lead smelting in Chapter 7. The Isasmelt was first adapted for secondary smelting at the Britannia Lead smelter at Northfleet, England in 1991, followed by a unit at the Metal Reclamation Industries’ secondary smelter in Malaysia in 2000 with a capacity of 50 000 t/a of lead. In this case batteries are fully separated and the metallics are separately melted by campaign through the Isasmelt or using rotary furnaces. Desulfurised pastes containing around one per cent sulfur are processed in the Isasmelt furnace on a batch–semi-continuous basis. Pastes are added continuously together with coal, and lead bullion is regularly tapped until the slag volume builds up to a preset level, which takes around 36 hours. Slag composition at this point is between 55 and 65 per cent PbO. Bath temperature is slowly raised during this time from 900 to 1000°C to maintain slag fluidity as impurities build up. Bullion is relatively pure at over 99.9 per cent Pb and impurities such as arsenic, antimony and tin accumulate in the slag. Paste addition is stopped at this time and slag is tapped into a short rotary furnace for reduction with coal to form an impure antimonial lead and a final silica slag containing less than 0.5 per cent lead. It is possible to continue slag reduction with coal and flux additions in the Isasmelt reactor at an increased temperature of 1200°C, but this practice has not been as efficient as the use of a separate slag reduction stage in the rotary furnace. The total cycle is 40 hours including 36 hours of continuous paste feeding and two hours of fluxing and slag reduction. Lead production and the lead balance during that time is given as (Brew, Fountain and Pritchard, 1991):

• 175 tonnes of lead in paste feed, • 170 tonnes of soft lead bullion, • five tonnes of alloyed impure lead, • 0.03 tonnes of lead in final slag, and • ten tonnes of lead recycled in flue dusts. The advantages of this approach are the use of silica slags low in residual lead, and high thermal efficiency. The Ausmelt plant at the Metalleurop Nordenham smelter in Germany is designed to process both primary and secondary feeds and is detailed in Chapter 7.

ELECTROWINNING PROCESSES To address the problems associated with slags, environmental emissions and occupational health issues from secondary smelting operations, a number of hydrometallurgical processes involving the electrowinning of lead from battery pastes have been developed. One or two only have been operated

188




on a small commercial scale, but without any significant acceptance for new or expanded secondary smelting operations. Despite this the technologies appear to adequately address the issues and are mostly founded on well established principles used in electrolytic lead refining. The processes also closely relate to technology developed for primary hydrometallurgical extraction, as covered in Chapter 9. In this case batteries are fully separated, metallic components are simply melted and drosses from that operation together with desulfurised battery pastes are subjected to hydrometallurgical extraction of lead followed by electrowinning from the leach solution. As detailed in Chapter 9, solutions used for lead electrolytes are fluosilicates as used in the Betts refining process, fluoborates and chlorides. After desulfurisation, battery paste material contains PbO, Pb(OH)2, PbCO3, PbO2 and some fine metallic Pb. A common problem with hydrometallurgical extraction is that both metallic lead and PbO2 are not soluble in the leaching solutions. A common approach is to make use of the reaction shown in Equation 11.16, by the addition of lead powder as required under acidic conditions. Pb + PbO2 = 2PbO

(11.16)

The fine metallic lead contained in the paste material will contribute to this reaction and will in turn be dissolved. As shown in Chapter 10 – Table 10.1 ‘Average battery paste analysis’, lead will be distributed in battery paste as follows:

• sulfates converted to soluble salts

51 to 64 per cent of contained lead

• PbO – soluble

eight to 11 per cent of contained lead

• PbO2 – insoluble

24 to 32 per cent of contained lead

• Pb – insoluble

four to six per cent of contained lead

Based on this analysis, the amount of powdered lead required to reduce PbO2 is an equivalent amount to the lead as PbO2 less the amount of free metallic lead already present, that is 20 to 26 per cent of the contained lead in the paste material. This could be regarded as the required amount of cathode lead recovered which needs to be recycled to the leaching stage, and is a significant amount. An alternative proposed for the RSR process is the use of carbon or organics present to reduce PbO2 during drying, in which the paste is heated to 290°C. The low temperature operation will avoid problems of lead fuming and is sufficient to achieve reduction. In line with the above composition, carbon equivalent to 1.2 to 1.5 per cent of the lead content would be required to achieve reduction of PbO2. Another option with the RSR process is the use of SO2 during paste desulfurisation leaching with sodium carbonate; however, this will generate additional sulfate at around 50 per cent more than generated from the removal of the lead sulfate content alone. In the Engitec process, based on a fluoboric acid electrolyte, hydrogen peroxide is used according to Equation 11.17 during leaching with HBF 4. PbO2 + H2O2 + 2HBF4 = Pb(BF4)2 + 2H2O + O2

(11.17)

Residual H2O2 in solution is useful in this case in preventing the formation of PbO2 at the anode during electrolysis.



189


The other significant issue for hydrometallurgical electrowinning processes is the formation of PbO2 at the anode. As discussed in Chapter 9 (Equations 9.10 and 9.11), competing reactions at the anode result in the formation of oxygen or PbO2. PbO2 deposition is unwanted since the aim is to recover all lead as metal by deposition on the cathode. Various techniques have been developed to prevent the deposition of PbO2, such as the presence of phosphorus in solution as developed for the US Bureau of Mines process and the presence of arsenic in solution as used in the RSR process (see below). In the proposed Ginatta electrowinning process the addition of cobalt to the electrolyte at around 200 mg/L is used to reduce the oxygen overvoltage, thus promoting oxygen evolution and reducing any formation of PbO2. It is also claimed that the use of cobalt prevents the degradation of graphite and enables it to be used without PbO2 coatings, or the use of titanium anodes (Ginatta, 1984). Figure 11.7 gives a generalised flow sheet for the electrowinning processes with various options for handling PbO2 shown as dotted lines. Separated Battery Pastes

Optional SO2 addition to reduce PbO2 Paste Desufurisation

Optional C addition to reduce PbO2 Drying Optional lead powder recycle for PbO 2 reduction Optional reagents to reduce or facilitate the reduction of PbO2 Leaching Reagents for deposit levelling and suppression of PbO2 formation Filtration

Electrolysis

Cathode lead

FIG 11.7 - Generalised flow sheet for electrowinning lead from battery pastes.

The US Bureau of Mines process This process was developed in the 1980s to address environmental problems associated with secondary lead smelting at that time. It extended the Betts electrorefining technology using fluosilicate electrolytes to process desulfurised battery pastes and produce a dense cathode deposit of pure lead metal.

190




Desulfurised paste is leached with fluosilicic acid as spent electrolyte to dissolve lead, as in Equation 11.18: PbO + H2SiF6 = PbSiF6 + H2O

(11.18)

Leaching time is around one hour at 50°C and can give a solution of up to 150 g/L Pb. Metallic lead powder is added to the leach to reduce PbO2, as discussed above. After filtration the solution is fed to the electrolytic cells, maintaining an electrolyte composition of the order of 70 g/L Pb and 90 g/L H2SiF6. Lower levels of lead in solution will cause the cathode deposit to lose density and become spongy. The solution contains animal glue at 0.05 g/L and lignin sulfonate at 4 g/L as grain refiners to aid the production of a smooth and level cathode deposit. Lead starting sheets are used for the cathode. Initially graphite anodes were used but with oxygen evolution they were attacked and deteriorated rapidly. Hence titanium anodes coated with PbO2 were ultimately selected. To prevent the formation of PbO2 the USBM process uses the presence of phosphorus in solution at around 1 to 2 g/L, which is added as phosphoric acid (H3PO4 at 3.1 to 6.3 g/L). 1.5 g/L P in solution will correspond with one per cent lead deposition on the anode as PbO2, whereas at 0.5 g/L the lead deposition on the anode will be closer to 20 per cent. Phosphoric acid may be lost in the leaching step and needs to be continuously added to the electrolyte. Current densities proposed are 170 amps/m2 of cathode area in line with electrorefining practice (see Chapter 13), and current efficiencies are 95 to 97 per cent. Cell voltages are around 2.9 volts, giving energy consumption of 0.78 kWh per kilogram of lead recovered (Cole, Lee and Paulson, 1985).

The RSR process As described above, the preferred method of preparation of desulfurised battery pastes is low temperature drying to decompose PbO2, and it has been found that sufficient organic materials are present to provide carbon for the reduction reaction. The dried paste is leached in spent electrolyte to dissolve lead in accordance with Equation 11.18. Lead concentration is raised to around 150 g/L or up to 200 g/L but must maintain a free fluosilicic acid concentration of at least 50 g/L to prevent hydrolysis and precipitation of lead compounds. Lead recovery to the leach solution is about 95 per cent and the leach residue containing 30 per cent lead and two to three per cent sulfur can be recycled to the desulfurisation step, or part discarded after conversion to a slag in an electric furnace. The electrowinning electrolyte is maintained at around 60 g/L lead by recirculation and contains glue and lignin sulfonate for deposit levelling. Each cell contains 50 anodes and 51 cathodes of 2 m2. Cathodes are starter sheets fabricated from lead sheet. Anodes are graphite but are prepared by electroplating a smooth coating of PbO2 over a non-conducting inert mesh material. Deposition of the PbO2 is continued until the mesh material is completely covered, and this technique provides a reinforced conductive coating that is coherent and is not subject to breakage or spalling (Prengaman and McDonald, 1980). The formation of PbO2 on the anode during normal electrolysis is prevented by the addition of arsenic to the electrolyte at a level of 0.5 to 1.0 g/L. Boric acid is also added to the electrolyte to prevent the decomposition of fluosilicate and the release of HF vapour from the cells. Any copper in the electrolyte will deposit with lead, but arsenic, antimony and bismuth will remain in the electrolyte. It has been noted that antimony builds up to a maximum of around 2 g/L in



191


the electrolyte and then is no longer extracted during leaching. High purity lead is deposited at better than 99.9 per cent and any minor impurities in the cathode lead, such as copper, can be removed by conventional refining approaches after melting in preparation for ingot casting. Purification of the electrolyte by lead powder addition could possibly be used to avoid this (Prengaman and McDonald, 1990).

The Engitec process The Engitec process for battery paste processing is based on the use of a fluoborate electrolyte. Desulfurised pastes are leached in spent electrolyte with the addition of H2O2 to decompose PbO2 and produce a solution of up to 200 g/L Pb. After filtration the leach solution is added to a circulating electrolyte to maintain lead at around 100 g/L and 100 g/L of free HBF4. Glue and lignin sulfonate are used for cathode deposit levelling, and electrolyte temperature is held at 35°C. Stainless steel cathode sheets are used, from which the lead deposit can be readily stripped. The cathode current density is around 330 amps/m2 and cycle time 48 hours between stripping operations. Cathode purity is in excess of 99.99 per cent. The anodes have been specially developed by Engitec and consist of a composite wire of copper with an outer layer of tantalum on which PbO2 has been deposited to lower the oxygen overpotential. Each anode consists of a grid of wires held in a frame and operates at a high current density of around 1500 amps/m2 of anode surface. The high current density and the presence of residual H2O2 from the conversion of PbO2 in the leaching step prevent the formation of PbO 2 at the anode. Energy use is reported to be around 0.9 kWh/kg of cathode lead (Reynolds, Hudson and Olper, 1990). Subsequent to the development of the above process for battery paste processing, Engitec have extended their fluoborate technology to the extraction of lead from sulfides by leaching with ferric fluoborate solution, followed by electrowinning in a diaphragm cell to recover cathode lead and re-oxidise ferrous iron in solution back to the ferric state for recycle to the leaching stage. This is detailed in Chapter 9 – ‘Fluoborate Systems’. Although this involves the use of a diaphragm electrolytic cell with increased voltage drop through the diaphragm, the anode potential for ferrous iron oxidation is much lower than for oxygen evolution and there is no possibility of PbO2 deposition on the anode. As a result standard graphite anodes can be used. Application of this approach to battery scrap processing has been proposed by conversion of pastes into lead sulfide by means of a biological sulfate reduction process, which effectively follows the reactions given by Equations 11.19 to 11.21 (Olper et al, 2000): PbSO4 + 4H2 = PbS + 4H2O

(11.19)

PbO + 4H2 + H2SO4 = PbS + 5H2O

(11.20)

PbO2 + 5H2 + H2SO4 = PbS + 6H2O

(11.21)

Lead sulfide so produced together with metallic lead from battery grids can then both be leached with ferric fluoborate according to Equations 11.22 and 11.23, allowing a common system for handling all battery components, and producing high quality refined lead. PbS + 2Fe(BF4)3 = Pb(BF4)2 + 2Fe(BF4)2 + So

192


(11.22)



Pb + 2Fe(BF4)3 = Pb(BF4)2 + 2Fe(BF4)2

(11.23)

The Ginatta process The Ginatta process also uses a fluorborate electrolyte and is based on the leaching of battery pastes which have been desulfurised with ammonium carbonate. The electrowinning process is licensed to MA Industries, who promote it in conjunction with their battery breaking technology (Ginatta, 1984). In the leaching stage PbO2 is reduced by lead metal, as in Equation 11.16, but this is promoted by the use of a ‘redox pair’ such as Fe3+/Fe2+ in the solution in small amounts, as shown by Equations 11.24 and 11.25: Pb + 2Fe3+ = Pb2+ + 2Fe2+

(11.24)

PbO2 + 2Fe2+ + 4H+ = Pb2+ + 2Fe3+ + 2H2

(11.25)

In any event it is most likely the reaction given by Equation 11.16 proceeds in this way during acid leaching, whether by iron or another ‘redox pair’. Some scrap metallic lead may need to be added to the leaching stage to allow for complete conversion of PbO2. The leach solution contains 100 g/L lead and the spent electrolyte 40 g/L. As mentioned above, cobalt at 200 mg/L is used to promote oxygen formation at the anode and prevent the deposition of PbO2. This also enables the use of ordinary uncoated graphite for the anodes with minimal degradation. Triton x-100 and phenolphthalein are used as levelling agents and are said to enable the use of higher cathode current densities at 400 amps/m 2.

The PLACID process The PLACID process uses a chloride system with a hydrochloric acid leach of direct battery pastes without prior desulfurisation. Details are given in Chapter 9 – ‘Processes Based on Aqueous Electrolysis – Chloride Systems’ and a flow sheet is shown in Figure 9.4. Lead powder is used during the leaching stage to reduce PbO2, constituting a significant circulating load. Sulfate is removed from the leach solution by precipitation of gypsum with lime. The disadvantage of the use of a chloride system is the production of lead powder or sponge at the cathode rather than a dense deposit as with fluosilicate and fluoborate electrolytes. In this case such an output would facilitate recycle to the leaching stage, but the bulk of the cathode deposit must be compressed by briquetting or roll extrusion prior to melting to minimise oxidation.

REFINING OF SECONDARY LEAD Much of the secondary lead bullion produced, particularly from paste materials, will be suitable for direct recycle for battery manufacture, whereas bullion derived from battery grid materials or from general scrap metal will contain a range of impurities which may need to be removed. In some smelting options these impurity elements can be concentrated into first run slags and then into bullion produced from the reduction of those slags in the secondary smelting stage. This can significantly reduce the volume of lead to be refined. The most common impurities are antimony, tin and copper, with some arsenic, zinc, cadmium and iron. Bismuth and silver are rarely significant impurities.



193


The refining method used is usually a simplified thermal refining process following the procedures applied for primary lead refining as detailed in Chapter 12. Copper is first removed by sulfur drossing in a pan or kettle, which will also remove any nickel present. It may be difficult to reach low copper levels with low tin and silver contents, but this can be overcome by the addition of pyrite with sulfur. Following this, a simplified Harris process can be applied by treating the bullion with sodium hydroxide and sodium nitrate. This will remove antimony and tin in particular, but also traces of arsenic, zinc and cadmium. It is possible to recover antimony and tin from refining drosses if this is required using methods outlined in Chapter 12.

REFERENCES Brew, R B M, Fountain, C R and Pritchard, J, 1991. Isasmelt for secondary lead smelting, in Proceedings Lead-Zinc ’90 Tenth International Lead Conference, Nice, pp 170-181 (Lead Development Association: London). Cole, E R, Lee, A Y and Paulson, D L, 1985. Electrolytic method of recovering lead from scrap batteries, Journal of Metals, 37(2):79-83. Eby, D, 1990. Electric furnace smelting at RSR Corporation, in Proceedings Lead-Zinc ’90 Symposium, Anaheim, pp 825-839 (The Minerals, Metals and Materials Society (TMS): Warrendale). Ginatta, M V, 1984. US Patent 4 451 340, May 29. Koch, M and Niklas, H, 1989. Processing of lead-acid-battery scrap: The Varta process, in Proceedings Minerals, Metals and Materials Society (TMS) Symposium on Productivity and Technology in the Metallurgical Industries, Cologne, September, pp 495-500 (The Minerals, Metals and Materials Society (TMS): Warrendale). Olper, M, Maccagni, M, Buisman, C J N and Schultz, C E, 2000. Electrowinning of lead battery paste with the production of lead and elemental sulphur using bioprocess technologies, in Proceedings Lead-Zinc 2000 Symposium, Pittsburg, pp 803-813 (The Minerals, Metals and Materials Society (TMS): Warrendale). Pike, K N, 1990. Secondary lead blast furnace smelting at East Penn Manufacturing Co Inc, in Proceedings Lead-Zinc ’90 Symposium, Anaheim, pp 955-969 (The Minerals, Metals and Materials Society (TMS): Warrendale). Prengaman, R D and McDonald, H B, 1980. US Patent 4 236 978, December 2. Prengaman, R D and McDonald, H B, 1990. RSR’s full scale plant to electrowin lead from battery scrap, in Proceedings Lead-Zinc ’90 Symposium, Anaheim, pp 1045-1056 (The Minerals, Metals and Materials Society (TMS): Warrendale). Queneau, P B, Cregar, D E and Mickey, D K, 1989. Optimizing matte and slag composition in rotary furnace smelting of leady residues, in Proceedings Primary and Secondary Lead Processing Symposium, Halifax, pp 145-178 (Canadian Institute of Mining and Metallurgy (CIM): Montreal). Reynolds, R M, Hudson, E K and Olper, M, 1990. Advances in lead acid battery recycling: Engitec’s automated CX Breaker system, in Proceedings Lead-Zinc ’90 Symposium, Anaheim, pp 1001-1022 (The Minerals, Metals and Materials Society (TMS): Warrendale). Schenker, G, 1990. Lead recycling from battery scrap and other raw materials in Metaleurop’s lead smelting plant in Oker, in Proceedings Lead-Zinc ’90, pp 979-999 (The Minerals, Metals and Materials Society (TMS): Warrendale). Taylor, J D and Moore, P J, 1980. Secondary lead smelting at Britannia Lead Company Ltd, in Proceedings Lead-Zinc-Tin ’80 Symposium, pp 1003-1022.

194



PART D REFINING OF LEAD BULLION This part of the text covers refining methods used for the removal of a range of impurity elements from crude lead bullion produced by smelting operations. It also covers the recovery of valuable by-product metals such as silver. Chapter 12 Chapter 13 Chapter 14

Thermal Refining of Primary Lead Bullion Electrolytic Refining of Lead Alloying and Casting

CHAPTER 12 Thermal Refining of Primary Lead Bullion Primary lead bullion is an efficient solvent for a range of metallic elements during smelting operations and these elements need to be removed in order to produce pure lead metal. Impurities generally impart hardness to the metal, particularly arsenic and antimony, and lead is softened as these impurities are removed. Hence in the past, pure lead was referred to as ‘soft lead’ and early refining techniques, particularly for removal of arsenic and antimony, were called softening. Removal of impurities or refining is mainly carried out by thermal refining methods, which are commonly conducted by batch processing in open pots or ‘kettles’. Impurities are generally removed individually or in groups in a series of steps. The usual refining steps are made in the following order:

• copper removal or copper drossing (in two stages); • arsenic, tin and antimony removal or ‘softening’; • silver and other precious metal separation; • removal of zinc (added for silver recovery); • bismuth removal; and • caustic refining to remove a range of minor residual elements. Figure 12.1 illustrates a generalised flow sheet of the thermal refining sequence. In some cases special removal steps are added, such as tellurium removal, where these elements are present in significant amounts in the ores treated. Also, particular smelters may have less than the full range of separation stages depending on the impurities present and the smelter’s source of lead concentrates; for instance bismuth removal can often be omitted. However, in all cases copper drossing and caustic refining are required. Some of the steps in the refining process can be conducted continuously, but this is not common practice and the associated plant is only justified for large lead refineries over say 120 000 t/a of lead production. Batch processing can be highly flexible and is more easily controlled to final product specifications. Because of the high density of lead one batch is commonly around 200 tonnes and the number of batches required per day is quite limited. A significant part of any refinery can be the various processes required to treat drosses and residues produced from each of the separate refining steps for the purpose of recovering by-products, recovering lead for recycle, or rendering the material suitable for disposal. Such processes include the conversion of copper dross into a marketable form for copper recovery, the production of silver and gold, the recovery of zinc for recycle, the recovery of antimony to produce antimonial lead alloy and the conversion of arsenic into a stable material suitable for disposal. The alternative approach for lead refining is the electrolytic refining process, which is covered in Chapter 13, but this must be preceded by copper removal and usually by some softening. It also requires a final caustic refining stage and is not a full replacement of thermal refining.



197

CHAPTER 12 – Thermal Refining of Primary Lead Bullion

Hot Lead Bullion

Lead concentrate Copper Drossing

Iron Copper dross Melting Lead

Sulfur

Sulfur Drossing

Arsenic speiss Dross to smelter Copper matte to copper smelter

Oxygen or air

Sod. Hydroxide

Harris Process Alternative

Oxygen Softening

Sod. Nitrate

Carbon Slag

Iron Slag Reduction Rotary Furnace

Hydromet. Slag Treatment

Lime

Zinc

Sod. Antimonate Calcium Stannate

Arsenic speiss

Calcium Arsenate

Desilvering Zinc crusts

Antimonial Lead Alloy

Liquation Vacuum Dezincing

Lead return

‘Triple Alloy’ Ag/Zn/Pb

Zinc return

Vacuum Distllation

Calcium + Magnesium Cupellation

Kroll Betterton Bismuth Removal

Silver dore

Drosses

Electrolytic Refining

Sod. Hydroxide Ca /Mg slag

Anode slimes

Melting Bi /Pb alloy

Sod, Hydroxide + sod. nitrate Air

PbO slag to smelter

Caustic Refining

Refined Silver

Oxidation

Gold recovery

Gold Slag to smelter

PbO slag to smelter

Refined Lead to casting

Bismuth metal

FIG 12.1 - Thermal lead refining sequence.

198




METHODS AND EQUIPMENT Most refining is carried out in a batch mode and techniques are generally based on the addition of reagents to molten lead to form a separate solid or molten phase, which separates from the lead and floats to the surface as a dross, crust or slag to be removed by scraping or skimming. The principal equipment used is the open top pan or kettle holding between 100 and 300 tonnes of molten lead. Figure 12.2 shows a cross-section of a typical refining kettle of 3.4 m diameter, but diameters vary from 2.5 to 4 m. The upper cylindrical part of a 3.4 m diameter 200 tonne kettle is around 1.5 m long, and 10 mm of depth represents one tonne of lead. Steel thickness is 30 mm and unalloyed steel is used (ASTM -A284-55T). The kettle sits within a brick furnace fired by natural gas or oil and with provision for air cooling. The level of the operating floor of the refinery is usually one metre below the rim of the kettle so that the furnace is located and accessed from a lower floor or basement. Support Plate

Working Floor Vent Flue

1.7 m Cooling Air Injection

Cooling Air

Gas Burner

Furnace Walls

FIG 12.2 - Typical refining kettle.

An agitator fitted in a removable frame can be placed on the top of the kettle and is handled by an overhead crane. Similarly a submersible centrifugal pump can be fitted for transfer of molten lead from one kettle to another. The kettle is also fitted with a ventilation hood to capture emitted fumes, and this also needs to be removable to allow access for placement of other equipment. Ventilation requirements range between 800 and 1200 Nm 3/h per m2 of kettle surface area. Drosses are removed by skimming from the surface either manually using a perforated spoon ladle, or by a crane-operated scoop, or by mechanical equipment such as a backhoe, pneumatic grabs, or using proprietary skimming equipment based on drag or screw conveyors such as are manufactured by Worthwick. The refining kettles need to be serviced by an overhead gantry crane to lift covers, pumps, agitators and skimming devices at various stages of the batch cycle.

COPPER REMOVAL OR COPPER DROSSING The copper content of primary bullion depends on the composition of the concentrates treated and can be up to six per cent, but more commonly between one and three per cent. Copper needs to be removed to less than 0.01 per cent in this stage and is achieved by cooling the bullion to reduce the solubility of copper as well as by final treatment with sulfur to precipitate copper sulfide (Cu2S). It may be considered a two-stage process of first cooling bullion from around 1000°C to around 400°C to



199


produce a ‘copper dross’, and then a sulfur addition at close to the freezing point or ‘sulfur drossing’ stage. Sulfur drossing was also known as the Colcord or Hulst process. The cooling of hot impure bullion initially precipitates a PbS-Cu2S matte using the dissolved sulfur, or a mix of matte and solid alpha copper. As sulfur is depleted, copper may be precipitated as an arsenide (Cu3As) or as metallic copper (Davey, 1963). As the bullion cools the nature of the precipitated material will change as the copper and sulfur contents of the bullion are depleted. Thereafter, solid Cu2S and PbS may be precipitated at certain Cu:S ratios in the bullion. The composition and texture of the materials rejected from solution are very sensitive to the composition of the smelter bullion treated. Figure 12.3 shows the lead corner of the equilibrium phase diagram for the Pb-Cu-S system and indicates that on cooling some PbS will form as well as Cu2S.

FIG 12.3 - Pb-Cu-S system – lead corner (Davey, Jensen and Seguit, 1980).

The copper to sulfur ratio is important and with bullions low in sulfur content, such as from the Imperial Smelting Furnace, copper dross has a high content of metallic copper. The precipitated copper compounds are less dense than lead and float to the surface as a voluminous dross and are skimmed off. A significant amount of lead is entrained in the dross and further processing is useful to separate and recycle the entrained lead and yield a copper matte suitable for sale to a copper smelter. The second stage or sulfur drossing stage merely adds excess sulfur in elemental form to ensure complete precipitation of copper as Cu2S, and can reduce the level of copper in bullion to less than 0.005 per cent. Figure 12.3 suggests that the copper level in lead cannot be reduced below the equilibrium solubility of around 0.06 per cent by sulfur addition (or down to 0.02 per cent if arsenic or antimony are present), whereas in practice it is possible to go an order of magnitude below this to 0.001 to 0.005 per cent copper. It has been found that these levels cannot be reached in the absence of silver or tin, which act as promoters of the decoppering reactions. A percentage of 0.002 silver is required to achieve copper levels of around 0.001 per cent (Davey and Happ, 1971). It has also been found that continued stirring of the dross with decoppered bullion can cause the copper to revert and redissolve into the bullion. The explanation for this phenomenon lies in the fact that the reactions are controlled by kinetics and not thermodynamic equilibrium.

200




The solid products of reaction float to the surface and are removed from the body of the molten metal. Reaction 12.1 below is slow, particularly in the presence of silver or tin, whereas Reactions 12.2 and 12.3 are very fast. Similarly, reactions between metallic lead and copper sulfides (Reaction 12.4) are slow and equilibrium is slow to establish (Davey, 2000). However, reversion will occur if stirring is continued for long enough after the bullion has reached levels of around 0.002 per cent Cu. Pb + S = PbS

(12.1)

Cu + S = CuS

(12.2)

Cu + CuS = Cu2S

(12.3)

Pb + Cu2S = PbS + 2Cu (in bullion)

(12.4)

Pyrite (FeS2) can be added as a source of sulfur, usually as a supplement together with elemental sulfur, and can enable effective decopperising without the presence of silver or tin.

Batch copper drossing methods Usually two or four kettles are required for batch processing, one or two for each stage. One may be filling or emptying while the other is carrying out the drossing operation. It is common practice to transfer bullion from the smelter to the refining operation in ten to 20 tonne ladles. A number of these can be in use and can provide a buffer storage as well as being used to allow the bullion to cool to around 600°C before transfer to the refining kettles. It is common for a crust to form on the ladles as the lead cools and this crust can consist of a mix of ‘tin dross’ (zinc, tin and iron spinels), frozen matte and copper speiss, in some cases with a concentration of nickel if this is present in smelter feed. In fact this can be used as a preliminary purification stage. The kettles are cooled by blowing air around the outer surface of the kettle, within the combustion furnace. Water coils can also be used and in some cases direct water addition to the surface of the lead is practised, but has the considerable risk of causing an explosion by water entrapment in frozen lead. The molten lead is agitated while cooling takes place and coke breeze or sawdust is usually added to the surface of the lead to aid the formation of a dry friable dross that can be easily removed. There is a misconception that drying out the dross with such methods avoids entrainment of a large amount of lead, but the reality is that the additives or stirring merely encourage surface oxidation of the dross with its entrained metallics, creating the impression of ‘drying’. Typically the dross will contain 40 to 60 per cent metallic lead. Lead is cooled to 380 to 400°C in the first stage and copper is reduced to between 0.1 to 0.5 per cent. Batch time may be of the order of eight to 24 hours. The first stage also removes a portion of the arsenic, tin and antimony as Cu3As, Cu3Sn and Cu2Sb, in the form of speiss, or as separately precipitated metallides. Arsenic removal may be more than 50 per cent depending on relative levels of copper, arsenic and sulfur, but antimony is significantly less due to the preferential formation of arsenides rather than antimonides. In the second ‘sulfur drossing’ stage, the lead is cooled to close to the bullion’s freezing point of about 320°C. Operators can sense the freezing point by watching for the formation of frozen lead rims on the edge of the kettle. Once the bullion is cooled, elemental sulfur is stirred into the lead bullion by addition into the agitator vortex. Additional water or air cooling may be applied to counter the heat release from sulfide formation, since best results are obtained at the lowest temperatures. The solubility of sulfur at this temperature is around 0.7 per cent and this will react with residual copper to



201


form CuS initially; the CuS then rapidly reacts with further copper in bullion to form Cu2S, which is kept in suspension during stirring. Sawdust is commonly added to the surface of the lead when the drossing is complete to kill the reversion reaction by coating the Cu2S particles with oxide and produce a dry manageable dross. The sulfur dross so produced is low in copper and can contain high levels of PbS as well as Cu2S. It is removed by scraping from the surface of the molten lead after raising the temperature to about 360°C and is commonly recycled to the smelting operation or the sinter plant. Sulfur use is between one and two kilograms per tonne of lead, but it is also a common practice to use pyrite in addition to elemental sulfur for copper precipitation, although this degrades the final matte produced with iron. Batch times for sulfur drossing are generally less than four hours and final copper levels in lead can be as low as 0.002 per cent. The drossing reactions are very fast (five to ten minutes), so most of the time involved is with cooling the batch and then later dross removal. Any copper not removed at this stage will be extracted during desilverising with the zinc-silver alloy, although this is not desirable as it complicates the recovery of silver. Final dross grades produced can range from ten up to 50 per cent copper. Copper dross from the complete drossing operation will also contain much of the tin present in primary bullion and virtually all of the iron, tellurium and indium. If it is desired to separately recover these metals, particularly tin, then the first stage can be controlled to remove predominantly copper and the second stage can be delayed until after softening by the Harris process which will extract these metals as well as antimony and arsenic. This is practised at the Hoboken smelter, in which copper is removed to 0.15 per cent in the first stage prior to the Harris process and then the remaining copper is removed by sulfur drossing to less than 0.02 before desilverising. It is important to reduce copper to low levels prior to the desilverising operation. The presence of silver and tin in bullion is beneficial to the sulfur drossing operation, possibly due to a change in surface tension of the bullion at the molten sulfur/lead interface, suppressing PbS formation.

Continuous copper drossing Continuous copper drossing was developed and installed by the Port Pirie smelter in 1962. The process uses a gas fired reverberatory furnace and uses the concept of mixing cold lead at 360°C with incoming hot bullion to cause copper precipitation within the mixing zone. This avoids the need for cooling through heat transfer surfaces, and the problem of adherence of copper precipitates to the cooling surfaces, building up accretions and blocking the furnace. The copper dross produced floats to the top of the lead where it is heated by gas burners, and melts to form a molten matte that is maintained at about 1150°C. The matte can be tapped from the furnace substantially free of metallic lead and usually contains above 50 per cent copper. A cross-section of the Port Pirie continuous copper drossing furnace (CDF) is shown in Figure 12.4 (Fern and Shaw, 1983). Feed bullion enters at around 950°C at one end and displaces an equivalent volume of cold lead from the furnace outlet at the side of the furnace. Bullion at around 380°C is pumped from the far end through a circulation launder where it is cooled to 360°C by means of immersed water-cooled plates. It is then returned to a separated compartment at the base of the furnace from where it flows upwards to mix with the incoming hot bullion. The furnace operates with a number of layers increasing in temperature from the bottom to the top surface.

202




Bullion Feed

Tap Hole Slag Vent Flue

Gas Burner Matte Bullion

Bullion discharge Outlet

Bullion Feed Pump Tank

Return Bullion Inlet

Bullion Circulation Outlet

Bullion Circulation Pump Tank Tapping Spout

Cooling Plates

Burner

Product Bullion Launder

Circulation Launder

FIG 12.4 - Continuous copper drossing unit (Port Pirie).

A thin slag layer is maintained over the matte to provide protection from the furnace atmosphere. This can be formed by the addition of silica to form lead silicate, but also contains zinc, iron and tin extracted from the bullion. The latter elements can render the slag viscous and unworkable although this effect can be compensated by increasing the PbO formation and by silica addition. Because of the interference of silica on subsequent treatment of matte, Port Pirie converted to a slag based on the PbO-As2O3-Sb2O3 system by adding some ‘softener slag’ as the sole flux. Experience with the CDF has shown that accretions can form on the furnace side walls at the mixing zone level, but this is manageable if the copper to sulfur ratio is below a critical value of 3:1. At this ratio a matte of 48 to 50 per cent Cu, 34 to 36 per cent Pb and 16 per cent S can be produced. The stoichiometric ratio of Cu:S is 3.96:1, so the critical ratio represents an excess of around 30 per cent sulfur above stoichiometric requirements. Sulfur can be added to supplement the level in primary bullion and is added into a vortex developed in the return of circulating lead from the cooling launder back into the furnace (Wiltshire, Miller and Bauer, 1989). Bullion from the CDF contains of the order of 0.06 per cent copper and must be further removed by sulfur drossing. This is also carried out continuously at the Port Pirie smelter using a continuous sulfur drossing unit (CSD) and lowers the copper content of lead to less than 0.005 per cent. This unit consists of two stirred pots in series followed by a third unstirred pot fitted with a scraper to remove the dross formed, as shown in Figure 12.5. The first pot is fitted with cooling coils and is for cooling only, although sawdust as a conditioner is added to the surface of the lead. The second pot is fitted with cooling coils and sulfur is continuously added from a feeder into the agitator vortex. Dross is continuously removed from the third pot in the series and bullion flows under a weir to exit the system. Dross collected from the CSD is high in lead and can be recycled to the smelting operation.



203


Sulfur Feeder

Dross Granulator

Stirred Pans

Product Launder

Feed Launder

Sawdust Feeder

Cooling coils

Dross Scraper

FIG 12.5 - Schematic of continuous sulfur drossing unit (Port Pirie).

Treatment of copper dross The composition of copper drosses (or mattes) covers a wide range depending on other impurities present and the efficiency of the methods used, but typically the copper content is in the range of 30 to 45 per cent. Sulfur may be 15 to 20 per cent and lead in the range of 20 to 50 per cent. The dross can be upgraded by melting with separation into a copper matte, a speiss and metallic lead for recycle. This operation can be conducted in a range of different furnaces, including a shaft furnace (campaigned through a blast furnace), a reverberatory furnace, a top submerged lance furnace and an electric furnace or a rotary furnace. The addition of lead sulfide concentrate during melting can convert any metallic copper present into sulfide matte and form additional metallic lead. Iron turnings can also be added to bind arsenic as a speiss. Bullion is separated from the mix of slags from the melting furnace during tapping and the slag mix is cast into pans and allowed to settle and solidify. It separates into three layers of slag, matte and speiss and is broken up and mechanically separated into the three phases. The matte can be sold to a copper smelter or can be converted to blister copper; however, returns are generally poor due to impurity penalties for elements such as lead and arsenic. The lead recovered from melting operations is relatively high in copper, which represents a significant circulating load on the copper drossing operations. There are many variations to these operations depending on the available outlets by way of copper smelting operations, and the nature of the feed they can accept. At the Port Pirie smelter, copper matte from the continuous drossing furnace was initially sold to copper smelters, but at a poor return. Subsequently plant was installed to extract the copper as cathode copper and could be justified by the available tonnage of copper recovered. This involved grinding the matte and leaching using a sulfuric acid-sodium chloride solution at a temperature of around 85 to 90°C. The leach residue is predominantly lead sulfate and sulfur and is returned to the sinter plant. The leach solution is subjected to solvent extraction using highly selective LIX reagents to remove copper and the raffinate is recycled to leaching or partly to a bleed to control soluble impurities. Loaded solvent is carefully cleaned of entrained aqueous solution to minimise chloride carryover to the cell house, and is stripped using spent electrolyte (dilute sulfuric acid solution) and fed to the electrowinning tank-house for deposition of cathode copper. Antimonial lead anodes and stainless steel cathodes are used for electrolysis (Lal and McNichol, 1987).

Copper dross treatment using sodium metal (the Glover process) Crude copper dross generally contains a substantial amount of entrained lead metal and PbS as well as Cu2S. As above, standard practice is to melt the dross to release entrained lead and leave a mixed

204




copper lead matte. Another means of upgrading developed at the Glover smelter is the treatment of the copper dross in a separate small kettle with sodium metal. The sodium is first dissolved in a small batch of lead bullion to which the dross is added. Sodium reacts with the PbS to form Na2S and metallic lead. Sodium sulfide has a low melting point and produces a fluid matte at a kettle temperature of around 650°C, and the matte is upgraded to around 30 per cent copper and depleted in lead to about three per cent. The matte can be skimmed from the surface and solidified. It is much more acceptable as a saleable product to copper smelters, justifying the cost of the sodium. This concept can be extended to the direct application of sodium metal to the copper drossing kettles where the addition of sodium redistributes the sulfides and arsenides and enhances dross fluidity to facilitate separation of matte from lead bullion. A separate speiss phase also tends to form but requires the addition of sodium at least equivalent to 100 per cent of the stoichiometric amount corresponding with sulfur in excess of Cu2S. The reactions with sodium form Na2S, Na3As and Na3Sb in that order. This can be a useful means of arsenic and antimony separation if present in only small amounts.

Removal of minor amounts of copper Where high purity concentrates are processed and only minor amounts of copper have to be removed, one method used at the Naoshima smelter in Japan is to add aluminium to form the intermetallic compound CuAl2. Aluminium is added as a 40 per cent alloy with zinc because of its low melting point at 560°C compared with aluminium at 660°C. Zinc is removed from the lead later either in the dezincing process or by caustic soda treatment. This method produces a dry powdery dross, which is easily handled and can reduce copper levels to less than one gram per tonne (Moriya, 1989).

SOFTENING FOR ARSENIC, ANTIMONY AND TIN REMOVAL Arsenic and antimony levels in primary bullion are generally in the range of 0.2 to two per cent for each element, but antimony is commonly higher than arsenic. The amount of tin present varies widely but rarely exceeds 0.1 per cent. Two techniques are generally used for separation of these elements:

• oxygen softening, and • caustic extraction or the Harris process. Softening relies on the fact that arsenic, antimony and tin are oxidised preferentially to lead and if an oxide slag is produced it will extract and concentrate these metals as well as any zinc present. If antimony levels are less than 0.5 per cent these operations can be eliminated at this stage of the refining operation and reliance placed on final caustic refining.

Oxygen softening In this case oxygen or air or mixtures thereof are blown into molten lead at temperatures in the usual range of 500 to 800°C, to form a PbO-rich slag. Arsenic, antimony and tin are oxidised preferentially to lead and are concentrated in the slag. Arsenic is most efficiently removed by this approach (98 per cent extraction), followed by antimony (90 per cent extraction) and then tin. It is likely that PbO is initially formed by oxidation followed by reaction with the impurities to form various arsenites (nPbO.As2O3), antimonites (nPbO.Sb2O3) and stannates (nPbO.SnO2), where n is 1, 3 or 5. A liquid slag, as a solution of impurities dissolved in PbO, can contain varying amounts of impurities, depending on the final target level of impurities in softened bullion; the lower the target the lower the level of impurities in the slag.



205


The softening process may be conducted continuously or by batch, and both kettles and reverberatory furnaces are used. Originally reverberatory furnaces were used as a continuous process with compressed air injection into the molten lead through air-cooled lances. In more recent forms of the process, oxygen-enriched air is blown into lead bullion in a kettle through two or three lances. Provided the temperature is high enough the slag formed can be sufficiently fluid to be continuously tapped and run over a chilled plate to solidify. The slag may be displaced from the kettle by the progressive addition of lead bullion to the kettle. With high levels of impurities, or if a high-grade slag is desired to work up to antimonial lead, two stages in series may be used. Generally the process is autogenous as the heat generation from oxidation will maintain the temperature with little fuel input. However, this is influenced by the grade of oxygen or air enrichment used as well as the operating temperature selected. Batch times also vary widely within the range of eight to 20 hours and depend on the temperature used, the initial level of impurities and the degree of oxygen enrichment of injected air. High rates are achieved with pure oxygen injection, but there is an economic balance due to efficiency of use. A fluid PbO-rich slag is produced at temperatures of 700 to 800°C, but lower temperature oxidation may produce a solid dross. Lower temperatures can be used in the presence of a solvent melt such as NaOH or NaCl to improve selectivity and reduce the amount of lead oxidised. Final levels of antimony in bullion are the control indicator for the softening process and are commonly around 0.05 per cent, but can be as low as 0.01 per cent with two softening stages.

Treatment of softener slags There are various approaches to the handling of softener slags. Usually they are reduced with coke fines or char in a short rotary furnace to produce an antimonial lead and an arsenic-rich slag from which arsenic can be extracted by leaching with caustic soda. Arsenic is then precipitated as arsenic trioxide for sale or by lime to form calcium arsenite for disposal. Alternatively the addition of iron turnings to the rotary furnace will produce a speiss containing around 40 per cent arsenic for disposal as well as an antimonial lead metal. The antimonial lead is usually cast into ingots for later use in producing lead alloys. The antimony content of the alloy can range up to 30 per cent.

The Harris process The Harris process removes arsenic, antimony and tin by oxidation with sodium nitrate and reaction with sodium hydroxide to form sodium arsenite, sodium antimonite and sodium stannate, which form a molten salt solution in excess sodium hydroxide. Sodium chloride can be added to extend the solvent (Paschen and Winterhagen, 1968). The overall reactions are given in Equations 12.5 to 12.7: 2As + 2NaNO3 + 4NaOH = 2Na3AsO4 + N2 + 2H2O

(12.5)

2Sb + 2NaNO3 + 4NaOH = 2Na3SbO4 + N2 + 2H2O

(12.6)

5Sn + 4NaNO3 + 6NaOH = 5Na2SnO3 + 2N2 + 3H2O

(12.7)

Arsenic will react with sodium nitrate alone according to Equation 12.8: 2As + 6NaNO3 = 2Na3AsO4 + 3N2 + 5O2

206


(12.8)



Lead will also react with sodium nitrate as in Equation 12.9: 5Pb + 2NaNO3 = 4PbO + Na2PbO2 + N2

(12.9)

The ease of oxidation is greatest for arsenic and decreases in order – tin, antimony and lead. Full oxidation of antimony does not occur until all the arsenic and tin are oxidised and extensive oxidation of lead does not occur until all the antimony is oxidised. The process reactions are diffusion controlled and require good agitation and dispersion of the molten salt phase into the molten lead through a vortex induced by the mixer impeller. An alternative method (as used by the Hoboken smelter) is to pump bullion at high rates into a cylindrical vessel mounted over the kettle, with sodium nitrate added from a feeder into the cylinder. A temperature above 450°C is required and a residence time of the order of 12 hours to remove impurities to acceptable levels, represented by a recorded antimony level of less than 0.0005 per cent. This is significantly better than for oxygen softening, but the process has a higher cost in terms of reagents used and the extended residence time required. Standard kettles are used and the process is conducted on a batch basis. A guide to reagent consumption for the individual impurities, expressed as kilogram per kilogram of impurity, is shown in Table 12.1 (Zaitsev and Margulis, 1985). TABLE 12.1 Reagent use for the Harris process (kilogram of reagent per kilogram of impurity). Reagent NaOH

Arsenic

Tin

Antimony

2.9

1.8

1.5

NaNO3

1.0

0.6

0.6

NaCl

1.0

0.6

0.5

The caustic slag contains arsenic, tin and antimony as sodium salts as indicated above. Of these, only sodium arsenite is partly soluble whereas the rest are in suspension, and sufficient NaOH must be used to maintain a reasonably fluid slag.

Treatment of caustic slags from the Harris process The slags are tapped from the kettle, granulated, and then subjected to separation by hydrometallurgical methods. The initial step is leaching with water, which will readily extract excess sodium hydroxide and sodium chloride leaving a residue of the salts of arsenic antimony and tin. The solubility of these salts is illustrated in Figure 12.6, with tin the most soluble, followed by arsenic and with antimony relatively insoluble. The leach slurry is usually cycloned to separate metallic lead granules, which are entrained in the slag during separation from the surface of the molten lead in the refining kettle. Figure 12.6 shows that if the concentration of NaOH in the initial leach is above 30 to 35 per cent, then all salts are insoluble and are retained in the leach residue. Sodium stannate is hydrated under these conditions to form Na2SnO3.3H2O. The excess sodium hydroxide-sodium chloride solution is evaporated to recover the reagents for recycle. A dilute water leach of the residue will extract tin and arsenic and leave antimony as insoluble sodium antimonite. This residue will also contain any tellurium, indium and selenium contained in the bullion and if these elements are present in recoverable amounts then a two-stage Harris process can be employed (see below).



207


18 16

Concentration (% by mass)

14 12 10

Sn As

8

Sb

6 4 2 0 -2 0

5

10

15

20

25

30

35

40

NaOH concentration %

FIG 12.6 - Solubility of impurity elements in caustic solution.

The leach solution containing arsenic and tin is treated with a controlled addition of lime (CaO) to firstly precipitate tin as calcium stannate (CaSnO3.3H2O), and then by further lime addition after dilution to a NaOH concentration of less than 80 g/L to precipitate arsenic as calcium arsenite (Ca3(AsO4)2). In some instances calcium carbonate is used for tin precipitation followed by lime addition, firstly to remove carbonate as calcium carbonate and then to precipitate arsenic. This approach may give a cleaner separation of tin if required. Following precipitation of arsenic, the residual solution can be evaporated to recover sodium hydroxide for recycle to bullion treatment.

Two-stage Harris process The two-stage process uses the fact that arsenic and tin are more easily separated from bullion than antimony. In the first stage kettle, sufficient sodium nitrate and sodium hydroxide are added to separate all the arsenic and tin, leaving most of the antimony to be removed in the second stage with excess reagent. The slags from the two stages are separately treated by the same hydrometallurgical routine as above, but in the case of the first stage slag, the initial leach residue or ‘black’ antimony cake concentrates the minor elements tellurium, indium and selenium. These can be separately recovered by treatment of the leach residue. The initial leach residue from the second stage slag is relatively pure sodium antimonite. Leach solutions may be combined for recovery of arsenic and tin by calcium precipitation. A generalised process flow sheet for the treatment of Harris process slag is shown in Figure 12.7. The two-stage approach is used at the Hoboken lead smelter in Belgium. In that case the ‘black’ antimony residue is acid leached to separate indium and tellurium leaving a lead-antimony residue. Tellurium is precipitated from solution using sulfur dioxide followed by indium precipitation as a hydroxide. The indium hydroxide is re-dissolved in acid and cemented from the resulting solution as indium metal by the addition of zinc dust (DeKeyser and Jespers, 1981).

208




Second Stage Harris Slag

First Stage Harris Slag

NaOH Leach

NaOH Leach

Wet Cyclone

Lead to remelt

Wet Cyclone

Lead particles to remelt

Filter

Filter

Solution Water Water Leach

Water Leach

Filter

Filter

As + Sn solution

CaO

Tin Precipitation ‘Black’ sod. antimonate Sb + (Se, In, Te) To further processing

‘White’ sod. antimonate 48% Sb Filter

CaO Arsenic Precipitation

Calcium Stannate 40 - 42% Sn

Filter Calcium Arsenate 15 - 25% As Evaporation NaOH (+NaCl) recycle to the Harris Process

FIG 12.7 - Process flow sheet for treatment of Harris process slags.



209


Separation of tellurium If tellurium is present in significant amounts it can be removed by an initial treatment of lead bullion with a small amount of sodium hydroxide at around 420°C. Tellurium rapidly reacts according to Equations 12.10 and 12.11. Although tellurium is far more reactive than arsenic, tin or antimony, the resulting slag will contain some of these elements. 3Te + 6NaOH = 2Na2Te + Na2TeO3 + 3H2O

(12.10)

xTe + Na2Te = Na2Te(x+1)

(12.11)

When the slag is leached in water the soluble sodium polytelluride can be oxidised with air to precipitate metallic tellurium as in Equation 12.12: Na2Te(x+1) + H2O + ½ O2 = 2NaOH + (x+1)Te

(12.12)

Sulfur dioxide can also be used to precipitate tellurium metal from sodium polytelluride solution. A Russian approach uses the addition of sodium metal to produce a high melting point intermetallic compound Na2Te, which can be collected in a caustic slag and skimmed from the surface of the molten lead. This material can contain up to 25 per cent Te.

REMOVAL OF SILVER AND OTHER PRECIOUS METALS Prior to the development of procedures to remove silver it was necessary to cupel all the lead bullion to oxidise the total lead content to PbO, leaving a silver doré. The PbO-rich slag was then reduced back to metallic lead free of silver. This was of course practical and economical only for lead bullion with a relatively high silver content.

The Pattinson process In 1839 a refining process for silver removal was developed by H L Pattinson in England. This process used the fact that lead and silver form a eutectic at around two per cent silver with a melting point of 305°C compared with the melting point of lead at 327°C. The lead corner of the lead-silver phase diagram is shown in Figure 12.8. For any lead bullion containing less than two per cent silver, freezing will first produce crystals of pure lead, leaving silver dissolved in the residual lead melt. A typical operation used a series of small pans or kettles (say nine). Bullion was fed to the middle pan and was allowed to cool so that relatively pure lead crystallised from the melt. The crystals were skimmed from the pan using a perforated ladle and were transferred to the pan on the right, whereas the remaining liquid was transferred to the pan on the left. This process was repeated for each pan in the series ending with the pan on the extreme right containing lead almost free of silver and the pan on the extreme left containing lead significantly enriched in silver at say 30 times the input silver content or up to one per cent. The enriched lead was then cupelled to yield a silver doré and a litharge slag, but considerably less lead has to processed in this way compared with the old system of full cupellation. An improvement to this approach was developed by Luce-Rozan in Marseilles in which dry steam was injected into the molten lead while cold water was sprayed onto the surface. The steam blast kept the temperature much more uniform and the enriched alloy approached closer to the eutectic composition of two per cent silver. Generally only two pans were required for this variant of the process.

210




420

400

Liquid

380 Temperature

360

Solid Silver + Liquid

340

320

Solid Lead + Liquid 300 0

2

4

6

8

Per cent Silver

Solid Lead + Eutectic

Solid Silver + Eutectic

FIG 12.8 - Lead-silver binary phase diagram.

The Pattinson process was replaced around 1880 by the Parkes process, which used zinc to extract silver, and is now obsolete.

The Parkes process The precious metals, particularly silver and gold, are removed by the Parkes process, which uses the ability to form high melting point intermetallic compounds with zinc of the form AgZnm. The technique was developed by Parkes in Birmingham in 1852, and made practical by Balbach in New Jersey with a procedure for treatment of the zinc crusts in 1872. The process has not been supplanted since that time. Gold has a greater affinity to react with zinc than silver and is thus readily separated with silver, which is the predominant precious metal associated with the production of lead. Removal of silver using zinc must follow copper removal and ‘softening’. Copper forms intermetallic compounds with zinc, as does arsenic and tellurium. The presence of antimony and tin also affects the performance of the desilverising operation, in particular separation of the silver-zinc alloy phase. As shown in the zinc-silver phase diagram in Figure 12.9, there are five identified solid phases. The extremes α and η are solids solutions and there are three intermetallic compounds; β, γ and ε. In practice the addition of molten zinc to lead bullion followed by cooling will result in the formation of these intermetallics as solid precipitates, which are mainly of the ε form but also contain η towards the



211


end of the process. The process also relies on the fact that above 420°C zinc and lead form two separate liquid phases with a range of compositions and merge into a single liquid phase above 798°C, as shown in the zinc-lead phase diagram in Figure 12.10. There are thus no solid phases of zinc and lead to interfere with the separation of solid phase zinc-silver intermetallic compounds. Temp. o C 900 Liquid Phase 800 700

600 500

400

300 200 100 0

10

20

30

Ag

40

50

60

90

80

70

100

Zn

Zn content % by mass

FIG 12.9 - Zinc-silver phase diagram. Temp. o C

Vapour Phase

900 One Liquid Phase 800

700 Two Liquid Phases 600

500 417 400 318 300

200 0

Zn

10

20

30

40

50

60

70

Lead % by mass

80

90

Pb

FIG 12.10 - Zinc-lead phase diagram.

212




Generally the Parkes process is conducted as a two-stage batch process using standard lead kettles or pans. In the first stage, recycled zinc crusts from the second stage are added and dissolved into hot lead coming forward from arsenic and antimony separation at around 600°C. The lead is then cooled to 480°C to allow solid phases to form and float to the surface as a crust. This requires the zinc content at the start to be above the saturation level at 480°C, which is around two per cent. The crust is skimmed from the surface of the molten lead and typically contains 60 to 70 per cent lead as entrained metal, 25 to 30 per cent zinc and up to five to seven per cent silver, as well as up to one per cent copper and some arsenic and antimony. It is important to avoid oxidation of the crust at this stage, to ease downstream handling and to increase zinc recovery. The residual lead from the first stage may contain 0.01 to 0.03 per cent silver and is subjected to an addition of around 10 to 12 kg of zinc per tonne of lead bullion. The temperature is initially raised to dissolve the zinc, and is then lowered to close to the freezing point or 320°C. After a suitable time the surface crust is skimmed off and is recycled to the first stage. The silver content of the lead at this point is generally reduced to less than 5 g/t (0.0005 per cent), but the residual dissolved zinc content is 0.5 to 0.7 per cent. In the case where large amounts of silver are contained in lead bullion, a three-stage Parkes process can be used, such as the case at the Torreon Smelter in Mexico, which has an input silver content of 1.4 per cent and uses around 18 kg of zinc per tonne of bullion. Using this technique some selectivity between silver and gold can be obtained, since gold is concentrated in the first crusts.

Davey desilverising procedure The Davey desilverising method consists of a two-stage counter current system using one kettle and has been applied at the Noyelles Godault refinery in France. The kettle is fitted with a floating shallow annular trough, fabricated from mild steel, which is filled with water during the cooling cycle to leave a solid crust at the surface of the kettle. The sequence starts with fresh bullion from softening being pumped into a kettle containing solid crusts retained from the second stage. Ingot lead from the liquation of silver crusts is also added at this point. When filled, the temperature of the kettle is raised to around 470°C with agitation to dissolve the retained crust and ingot lead. Agitation is then stopped and the kettle is allowed to settle and zinc silver crystals to collect at the surface. The crystals are then scooped off the surface of the lead and are transferred to a liquation kettle for melting under a CaCl2-NaCl covering flux. A 28 per cent silver-lead alloy is produced and a lower lead layer for recycling. After removal of the zinc-silver crystals, fresh zinc is added to the kettle with additions calculated according to the silver and copper content of the incoming lead from softening, as given in Equation 12.13: % Zn added = 0.634 + 1.676 × (Ag%) + 1.0 × (Cu% - 0.02%) (12.13) After completion of zinc addition the cooling tray is placed on the bath and water is continually added to maintain a level in the tray. The lead is gently stirred to promote heat transfer and when the kettle temperature has fallen to 320°C the cooling tray is removed, leaving a bridging solidified crust at the surface. The liquid lead in the base of the kettle is then pumped out as desilvered lead and the above sequence repeated. Cycle time is around 13.5 hours with four hours for stage 1, two hours for zinc addition, 6.5 hours for cooling and one hour for desilvered lead transfer. This procedure is most suited to mid-range silver contents of softened lead at around 0.3 per cent.



213


Continuous desilverising For many years the Port Pirie smelter operated a continuous desilverising kettle, developed by G K Williams in 1932. This process uses a tall cylindrical cast iron vessel 3.05 m diameter × 7.05 m tall. Molten lead bullion at 600 to 650°C flows into the top and through a layer of molten zinc 1 m deep and held at 530 to 580°C, where it becomes saturated with zinc at that temperature. It then flows down through the kettle and is progressively cooled to 320°C, close to its freezing point at the base. Silver-zinc alloy precipitates and floats up to the surface and is incorporated into the zinc layer. Depleted lead flows back up through a central siphon to overflow at the top, and is reheated in the process. The zinc layer becomes enriched in silver and its melting point rises. It is periodically ladled off before precipitation of solid phases begins to occur, and fresh zinc is added. The alloy contains ten to 15 per cent silver, 60 per cent zinc and around 25 per cent lead, and 0.5 per cent copper. This composition is significantly higher in zinc and lower in lead than crusts from the batch Parkes process. Figure 12.11 gives a diagram of the Williams desilverising kettle.

FIG 12.11 - Continuous desilverising kettle at Port Pirie (Dawson, 1980).

Zinc is aggressively corrosive to the cast iron kettle and the top section in contact with the zinc alloy layer is lined with chrome magnesite brick. However, lower sections serve as heat transfer surfaces in contact with bullion and corrosion is significant, especially with the two upper castings. Consequently the life of the kettles is only of the order of two years. This imposes a significant cost penalty and adverse economics have caused the Port Pirie operation to replace this elegant approach with the more conventional batch system, using standard kettles.

Treatment of silver-zinc alloy The silver-zinc crusts contain substantial quantities of entrained lead, depending on the efficiency of the skimming procedure. Hence the first processing step generally involves the removal of excess free

214




lead by melting and liquation, or in some cases by pressing the hot crusts to squeeze out excess molten lead using a Howard press. The operation of the Howard press is illustrated in Figure 12.12. Compressed air

Crust flows into press Pressed crust Molten lead return to kettle Crust

Molten lead

FIG 12.12 - Schematic of the operation of the Howard press.

For liquation, the crusts are melted in a crucible or induction furnace at around 650°C using a salt flux cover to minimise crust oxidation. The upper silver-zinc layer (or ‘triple alloy’) is poured off and cast into bars, and the lower lead layer is returned to the main refinery stream ahead of the desilvering operation. Drosses separated from this melting operation are usually returned to the smelter, but can contain significant amounts of silver and thus represent a circulating load. Typical performance is shown in Table 12.2. TABLE 12.2 Typical performance of silver separation stages. Stream

Silver

Zinc

Lead

10%

20%

65%

Triple alloy – after liquation

30%

65%

3%

Impure silver (retort bullion) – after distillation

80%

5%

10%

Doré – after cupellation

99+%

Parkes crust

Copper

0.2%

The silver-zinc alloy is further separated by removal of the zinc by distillation to leave a residual impure silver. Distillation was originally conducted using the Faber Du Faur process in silicon carbide retorts fitted with an air-cooled zinc condenser. The retorts were located within a gas-fired furnace. Coke breeze or charcoal were usually added to the charge to prevent oxidation as well as calcium borate to aid in the prevention of ‘Blue Powder’ formation in the zinc condenser. This process required high temperature operation at around 1250°C and high fuel costs as well as a high cost in retort replacement. The retorting technique has largely been replaced by the use of an electrically heated vacuum furnace, also fitted with a zinc condenser. There are a number of different devices for this purpose with heating by induction, by electrical resistance through a slag layer or by radiant heating bars as in the Leferrer furnace. The sealed electrical induction furnace under vacuum with an attached zinc condenser appears to be the most favoured approach – the vacuum induction retort. Operating costs are much lower using this technique and zinc separation is much higher and above 90 per cent.



215


Furnaces are constructed using an induction heated crucible of graphite or silicon carbide, contained within a refractory lined vacuum vessel. The zinc condenser may be external, in the vacuum train, or may be contained within the lid of the containment vessel. Seals are by rubber O rings in water-cooled flanges and the ability to maintain vacuum is a critical aspect of efficient operation. Target operating pressure is around 80 mbar absolute. The final temperature of the retort bullion (impure silver) formed within the retort at the end of a batch is in the range of 1100 to 1200°C. The detection of the end point of the batch distillation is also an important aspect of operation and usually is determined by the vacuum pressure and the temperature of the residual metal. Typical performance is shown in Table 12.2. The distilled zinc product is recycled to the desilverising stage and the retort bullion is treated by cupellation, in which air or oxygen is blown into the molten metal to oxidise the lead and residual zinc to form a slag and leave a molten silver doré containing gold and some copper. Cupellation was traditionally conducted in a hearth within a reverberatory furnace with lance injection of air or oxygen-enriched air. However, reactors such as a bottom blown converter, top blown rotary converter or top submerged lance reactor (Ausmelt or Isasmelt) are now more commonly used, with much higher efficiency. Again the operations are batch and can have batch times of the order of eight hours. Sodium nitrate may be added to assist oxidation of some impurities. The bottom blown converter has the advantage of direct injection of oxygen into the bullion layer, allowing a much deeper litharge slag layer to minimise the entrainment of metal. Oxygen use efficiency also approaches 100 per cent (Barrett and Knight, 1989). Lead-rich slag from the cupellation stage is returned to the lead smelter, and the silver doré is prepared by casting into anode plates for electrolytic refining or is granulated for ‘parting’.

Electrolytic silver refining The most common approach for the production of high purity silver is electrolytic refining. Anodes of silver doré are electrolysed using a silver nitrate electrolyte and pure silver is deposited on stainless steel cathodes. Silver deposits on the cathode in a dendritic crystalline form, which readily separates from the cathode by scraping and collects in the bottom of the cell. The anode is separated by a cloth membrane and retains the anode slimes. The slimes contain gold, platinum and palladium and are further processed to recover these metals. The electrolyte is neutral and can contain up to 250 g/L of silver nitrate. Copper tends to build up in the electrolyte and is controlled by a bleed to maintain a level below 40 g/L. The bleed solution is treated with zinc dust to precipitate a silver cement which is returned to cupellation. Make-up silver nitrate is produced by the digestion of doré in nitric acid. The cathode deposit of silver crystals is washed and filtered and then melted in a graphite crucible and cast into 23 kg (1000 troy ounce) bars. The cells used are commonly the Balbach-Thum cell design, which uses a single anode within an upper diaphragm compartment and a single horizontal cathode sheet on the base of the cell. The cell is constructed from PVC. Silver deposits in dendritic form and needs to be periodically scraped from the cathode to prevent short circuiting. It is then removed from the cell. Anode current densities used are in the range of 450 to 650 amps/m2, and for a cell with anode area of around 0.25 m2 the theoretical production is around 0.5 kg/h of silver per cell (Mantel, 1960). The Moebius cell uses a more conventional and compact cell design with vertical electrodes. The anodes are enclosed in cloth bags and the cathodes are equipped with vertical mechanical scrapers to continuously remove the silver deposit, which drops to the bottom of the cell.

216




Maintenance of a consistent and uniform electrolyte composition by recirculation is important for the achievement of optimum cathode purity and product silver of better than 99.999 per cent can be obtained. The anode slime containing all the gold and any platinum and palladium is first treated with sulfuric acid to extract residual silver and copper, and is then melted and cast into crude gold bar or into anode plates for electrolytic refining, depending on the presence of the other precious metals. The Wohlwill process is commonly used for electrolytic refining. It uses an electrolyte of ten per cent HCl and seven to eight per cent Au and rolled gold cathodes. Current densities are high at 1000 amps/m2. The electrolyte is periodically withdrawn and treated with SO2 to precipitate gold which is recycled, and then with NH4Cl to precipitate platinum as (NH4)2PtCl6, leaving palladium in solution which can be subsequently precipitated with HNO3.

Silver parting Separation of silver and gold by parting involves dissolving silver from the doré using hot concentrated sulfuric acid or nitric acid, leaving gold as the residue. The residual gold is simply melted and cast into bars or may be further refined by chlorination. For effective dissolution of silver the silver to gold ratio must be greater than 2:1, otherwise the silver will not dissolve. Ferrous sulfate is added to the clarified silver solution to reduce silver ions and precipitate metallic silver, which is filtered from the solution, melted and cast into ingot. For dorés rich in gold that cannot be extracted by acid leaching, the doré is melted and chlorine is injected. A silver chloride slag will be formed leaving pure gold. The silver chloride slag includes traces of other metals, but some metal chlorides such as zinc, bismuth and arsenic will be volatilised.

SEPARATION OF THALLIUM Normally thallium, like mercury, is effectively removed by volatilisation in the sintering process. For direct smelting processes there may not be sufficient volatilisation with direct capture into the lead bullion, and a direct method of separation may be required, as is the case at the Trail smelter with the introduction of the Kivcet process. In this case thallium is removed by the addition of zinc chloride to molten bullion, displacing thallium as a thallium chloride dross, and leaving zinc in the bullion for later removal.

SEPARATION OF ZINC FROM LEAD Metallic zinc is added for desilverising and leaves around 0.5 to 0.7 per cent residual zinc in the lead. The most common method of residual zinc removal is by vacuum distillation, although methods involving oxidation into a slag are available. Batch vacuum distillation of zinc from lead was developed by the St Joseph Lead Company at Herculeneum, Missouri in the 1940s using a bell placed in a standard kettle as in Figure 12.13b (Isbell, 1949). Vacuum distillation utilises the widening differences in vapour pressure between zinc and lead at lower temperatures, as shown in Table 12.3. The rate of mass transfer is also increased at reduced total pressure. Batch operation in a kettle of standard size is the most common approach. In one form of the equipment, the kettle is covered with a lid containing a condenser as a flat water-cooled plate and sealed onto the rim of the kettle by the use of rubber seals set in a water-cooled flange (Figure 12.13a).



217


TABLE 12.3 Vapour pressures of zinc and lead. Temperature (°C)

600

700

800

900 714.2

Zinc vapour pressure (mm Hg)

11.2

59.4

231.3

Lead vapour pressure (mm Hg)

0.00035

0.0051

0.045

0.27

Vapour pressure ratios (Zn/Pb)

31 540

11 625

5160

2631

Condenser surface

Condenser surface

Lead flow

Seal

Vacuum Lead feed

Feed launder

Vacuum Agitator

Vacuum Lead exit seal pot A. Batch Equipment

B. Batch Equipment

C. Continuous Equipment

FIG 12.13 - Schematic of vacuum dezincing equipment. (A) Standard kettle developed by Britannia Refined Metals. (B) Equipment design to avoid seals. (C) Continuous equipment developed at the Port Pirie smelter.

The lid is placed on the kettle and a relatively high vacuum is applied at around 0.05 mbar. The lead is agitated and zinc distils over from the lead to collect as a solid crystalline material on the cold condenser surface. Batch time is of the order of eight hours at an operating temperature of 600°C. At completion the vacuum is released, the lid is lifted and the zinc deposit is removed by hammering and scraping. The recovered zinc is recycled to the desilverising stage. An alternative device, shown in Figure 12.13b, uses an inverted open cylinder or ‘bell’ with a condenser surface at the top as in Figure 12.13a. This is immersed into the kettle to form a hydraulic seal with the lead rising up into the cylinder as the vacuum is applied. No seals are required but access for zinc removal from the condenser plate is more difficult. A ‘mud’ slurry can be applied to the condenser surface to facilitate condensed zinc removal. Continuous dezincing equipment uses a vacuum vessel in which lead is fed to a peripheral weir to overflow as a film down the walls to the bottom of the vessel from where it is withdrawn through a barometric leg seal. The condenser consists of a central water-cooled projection from the lid in the form of a truncated cone. The lid sits on the rim of the vessel and is sealed by a rubber ring located in a water-cooled flange. Zinc collects as a crystalline solid deposit on the surface of the condenser. The process is periodically interrupted for about ten minutes every eight to 12 hours to remove the condenser and replace it with a cleaned unit. Zinc is removed by hammering and scraping as with the other forms of dezincing equipment. The equipment is illustrated in Figure 12.13c (Davey and Williams, 1956). These processes generally remove in excess of 90 per cent of the zinc content of the lead, down to a level of less than 0.05 per cent.

218




Zinc can also be removed by oxidation with air or oxygen and captured in a sodium hydroxide slag. However, this approach is not efficient or selective for zinc removal but is useful as a final cleaning stage. In a similar approach chlorine injection into molten lead will preferentially react with zinc to form a molten zinc chloride slag, which may be separated by skimming from the surface of the lead. The handling of chlorine gas in open kettles has inherent hazards and this approach has not often been used on a large scale, although there is one application at the Trepca smelter in Serbia, where a slag containing 92 to 95 per cent ZnCl2 is produced. The major disadvantage is that metallic zinc is not recovered for recycle to the desilverising step.

SEPARATION OF BISMUTH Because of its adverse effect on the electrochemical and corrosion properties of lead in particular situations (notably the lead-acid battery), the specification for lead of 99.99 per cent quality requires bismuth to be below 0.005 per cent. In many applications this is not critical and levels up to 0.03 per cent can be acceptable. Bismuth is also chemically so close to lead that its separation is difficult. However, depending on the source of lead concentrates, a suitable level can often be achieved without separate removal, but it can also be present in many ores and in that case, a separate removal stage is required. Separation relies on the formation of high melting point intermetallic compounds with calcium and magnesium – Ca3Bi2 with a melting point of 1350°C and Mg3Bi2 with a melting point of 823°C, as well as the ternary CaMg2Bi2. This technique is known as the Kroll-Betterton process. In their initial work Kroll and Betterton found that separation of bismuth was poor with the use of calcium or magnesium alone, but they later discovered the presence of both metals dissolved in lead provided synergy and greatly improved separation. As well as forming individual intermetallics, the ternary compound is also formed, which may be expressed as a combination of the two separate compounds – Ca3Bi2.2Mg3Bi2 (or 3CaMg2Bi2) (Betterton and Lebdeff, 1936). In practice the mass ratio of magnesium to calcium used is between 2:1 and 4:1, and depends on the relative cost of the reagent metals. The process is promoted by the addition of antimony, which can also form intermetallic compounds as a replacement for bismuth, and is necessary to achieve low levels of residual bismuth. It is suggested that the antimony compound is isomorphous with the bismuth compound and a solution of the two causes a lowering of the activity of the bismuth complex (Moodie, 1976). The concentrations of the metals in lead at its melting point (327°C) have been given by the solubility product of the ternary compound according to Equation 12.14 (Davey, 1979): Log10([Ca][Mg]2[Bi]2) = -7.37

(12.14)

where the concentration of each metal is expressed in weight per cent. The effect of antimony on the above solubility product has not been defined; however, bismuth can be removed to less than 0.005 per cent by this process. In practice calcium and magnesium need to be dissolved into the lead bullion as a first step and this encounters significant problems due to the ready oxidation of the reagents at operating temperature, and due to the low density relative to lead which causes them to float to the surface of the lead. Magnesium is also prone to burn when exposed to air at kettle temperatures. Hence the metals are handled in ingot form and are usually placed in a wire cage, immersed into the molten lead in a



219


standard refining kettle, with circulation of lead by pump or stirrer over the alloying metals to aid dissolution. It is also important that the surface disturbance of the lead bath is minimal to avoid oxidation – particularly of the calcium content. Even so, oxidation of 15 per cent of the calcium input and ten per cent of the magnesium input is to be expected. Additions of calcium plus magnesium are generally in the range of 2 - 5 kg/t of lead. Batch temperatures begin at 450 to 500°C for dissolution of the reagents, and are reduced to close to 330°C at the end of the batch. Batch times range between ten and 20 hours. Depending on the bismuth content, two batch stages can be used with antimony additions into the second stage. Such additions are typically 0.4 kg per kilogram of bismuth and are usually less than 1 kg/t of lead. It is also possible to dissolve the calcium and magnesium in a small part of the lead stream with addition of that alloy mix to the main debismuthising kettle. In this case better control of potential oxidation can be achieved using an enclosed kettle for dissolution either purged with nitrogen or with the surface covered with a layer of charcoal. Calcium-magnesium alloys at either 70 per cent Mg at 30 per cent Ca or 65 per cent Mg at 35 per cent Ca are also available for this application, and can be added without the wire cage and with reduced oxidation loss. The bismuth dross formed is removed from the surface of the lead on completion of the batch. Drosses normally contain between three and ten per cent Bi, and commonly around six per cent. Calcium plus magnesium total around two per cent and the remainder is lead. Drosses can be upgraded by pressing or centrifuging to remove entrained lead. Bismuth dross is treated by melting at 500 to 600°C with the addition of sodium hydroxide and sodium nitrate. Calcium, magnesium and antimony are oxidised and removed in a caustic slag, leaving a bismuth lead alloy with around six to ten per cent bismuth. With upgrading of the dross this can be raised to around 15 to 25 per cent. The bismuth lead alloy can be further processed to bismuth metal using a similar series of procedures as used above for lead refining and removing the bulk of the lead as a litharge slag. The Kroll Betterton process is relatively expensive due to the cost of calcium and magnesium and the fact that they cannot be recovered. There is some flexibility in the relative amounts of calcium and magnesium used, providing some opportunity for cost optimisation.

FINAL CAUSTIC REFINING The final stage in most refining operations is treatment with sodium hydroxide, often also with the addition of sodium nitrate as an oxidising agent. This stage particularly removes remaining amounts of zinc and antimony, but also removes residual calcium and magnesium from the bismuth removal stage. Sodium nitrate tends to oxidise lead to PbO according to Equation 12.15: 5Pb + 2 NaNO3 = 5PbO + Na2O + N2

(12.15)

PbO then oxidises the various impurity elements as in Equation 12.16: Zn + PbO = ZnO + Pb

(12.16)

Arsenic and antimony are also removed in accordance with the principles of the Harris process as detailed above, and this is particularly applicable if oxygen softening has been used, where higher residual levels of arsenic and antimony remain than from application of the Harris process.

220




Caustic refining is conducted in a standard refining kettle at a temperature of 450 to 580°C with a batch time in the range of six to 12 hours. Reagents are in flake or powder form and are mixed into the molten lead by addition to the vortex created by the agitator. Reagent use varies depending on the impurity load, but is generally less than 2 kg/t of lead and sodium nitrate use is usually less than half the amount of sodium hydroxide. The oxides are absorbed into the caustic slag layer as a dross, which can be skimmed from the surface of the melt. At this point the resulting lead will be fully refined and can be cast into ingots.

REFINING OF SECONDARY LEAD The impurities contained in secondary lead bullion are generally confined to copper, tin, calcium, antimony and arsenic, and the full range of refining is not required. Copper is generally quite low and can be controlled by simple sulfur drossing in a kettle. All other impurities can be removed by ‘softening’, usually using a simplified Harris process by treatment with sodium hydroxide and sodium nitrate, or by oxygen injection, depending on the amounts involved.

SUMMARY OF COMMON IMPURITIES, THEIR CONTROL AND RECOVERY Molten lead is a particularly good solvent for many base metals present during smelting unless they can be readily oxidised under lead smelting conditions and report to slag. Alternatively the presence of a matte phase during smelting can also capture many impurities as well as precious metals. Otherwise the impurities present in lead bullion need to be progressively removed, as described above. The following provides a brief summary reference for the common impurities encountered in extractive lead metallurgy.

Antimony Antimony present in smelter feed reports primarily to bullion, with smaller amounts in slag, and in any matte or speiss formed. It is the major impurity in secondary lead where it is sourced from antimonial lead alloys used in lead-acid batteries. Antimony is oxidised preferentially to lead, and hence is removed by oxidation with air or oxygen in the ‘softening’ process or by sodium nitrate used with sodium hydroxide in the Harris process. The litharge slag from ‘softening’ can be reduced to form an antimonial lead alloy, and the caustic slag from the Harris process can be treated to produce sodium antimonate.

Arsenic Arsenic present in smelter feed reports to lead bullion as well as slag and matte, and can also form a separate speiss phase containing up to 30 per cent, depending on the relative amounts present. It can also report in significant amounts to smelter fumes. Bullion commonly contains up to two per cent arsenic. Arsenic is preferentially oxidised in the softening or Harris processes, but more readily than antimony. Consequently when softener slag is reduced to form an antimonial lead alloy, arsenic can be retained in the residual slag from where it can be extracted by leaching and precipitation as arsenic trioxide or as calcium arsenite. Alternatively it can be extracted from caustic slags from the Harris process by leaching and precipitation with lime as calcium arsenite, which can contain around 20 per cent As.



221


Bismuth Being chemically similar to lead, bismuth reports readily to bullion during smelting and is difficult to separate. It is usually removed by precipitation from molten lead as a high melting point intermetallic compound with calcium and magnesium using the Kroll Betterton process. Treatment of the bismuth dross with sodium nitrate and sodium hydroxide yields a bismuth-lead alloy from which lead can be removed by selective oxidation and further caustic treatment to give a refined bismuth. Bismuth is retained in anode slimes from the electrolytic lead refining process and can be recovered using similar caustic treatment procedures after recovery of silver.

Cadmium Cadmium is normally volatilised during sintering and smelting and reports to fumes from those processes. Otherwise some may report to smelter slag, similar to zinc.

Calcium During smelting, calcium reports exclusively to slag, but is subsequently introduced into lead bullion as calcium metal additions during removal of bismuth by the Kroll Betterton process. The residual calcium is readily removed by oxidation in the final caustic refining step with the addition of sodium hydroxide and sodium nitrate.

Copper During smelting copper predominantly reports to lead bullion, although it will preferentially report to any matte phase formed during smelting. Copper is commonly present at up to three per cent in primary lead bullion. It has a low solubility in lead close to its freezing point, but most importantly readily forms Cu2S with any sulfur present at low solubility. Consequently copper is removed by cooling smelter bullion and by the addition of sulfur or pyrite to form a matte. This is usually the first step in the lead refining sequence and can remove copper to levels below 0.002 per cent. Copper may be recovered by treatment of the copper drosses or matte in a conventional copper smelting operation and can be a significant by-product for primary lead smelting operations.

Gold Gold follows silver during smelting and in the refining operations. Following cupellation, which separates residual lead from precious metals, silver is recovered by electrolytic refining, leaving a crude gold residue containing any platinum group metal present. The crude gold is melted, cast into anodes and electrolytically refined.

Indium Indium tends to be distributed between smelter fumes, slag and lead bullion. It is separated during the softening process or in caustic slags from the Harris process. If present in significant amounts indium is recovered with tellurium from the first stage Harris process slags by acid leaching to yield a solution of tellurium and indium from which indium is precipitated as a hydroxide.

222




Iron Iron predominantly reports to smelter slags and any matte formed. The content of iron in bullion is not significant, and any present is removed as matte during copper drossing.

Magnesium Similar to calcium, magnesium is not present in primary smelter bullion, but is added as magnesium metal in the removal of bismuth by the Kroll Betterton process. The residual magnesium is readily removed by oxidation in the final caustic refining step with the addition of sodium hydroxide and sodium nitrate.

Mercury Mercury is efficiently separated by volatilisation during sintering or direct smelting, reporting to the SO2 rich gas stream. It will contaminate sulfuric acid produced from the gas stream, and usually needs to be removed during wet scrubbing prior to gas drying and acid production.

Nickel During smelting any nickel present in concentrates will be distributed to both slag and bullion, but will report to a greater extent to any matte or speiss formed on cooling of the bullion. Nickel present in bullion is readily removed with copper during sulfur drossing.

Platinum group metals (PGMs) PGMs follow silver and remain as the final residues following gold refining.

Selenium Selenium is normally separated by volatilisation during sintering or smelting operations and predominantly reports to fumes. Any present in lead bullion can be separated in caustic slags from the Harris process, but is relatively minor and of no real consequence.

Silver Silver normally reports to lead bullion during smelting but will also report to any matte formed. It is the most valuable by-product metal contained in primary lead bullion and efficient recovery is of key importance. Silver is normally separated from lead bullion following copper removal and softening, using zinc addition to form a high melting point silver zinc alloy (the Parkes process). Zinc is recovered from the alloy by melting and vacuum distillation, leaving a lead-silver mixture from which lead is removed by oxidation or ‘cupellation’. The resulting silver doré is electrolytically refined to recover pure silver cathode and leave an anode residue of gold plus any PGMs present.

Tellurium Tellurium primarily reports to lead bullion during smelting. It follows arsenic, antimony and tin into softener slags or caustic slag formed in the Harris process. It will concentrate in the first stage Harris



223


slag and can be extracted by a water leach followed by precipitation as tellurium metal from the leach solution by oxygen or by SO 2.

Thallium Thallium is normally separated by volatilisation in the sintering operation, but in direct smelting of lead concentrates can be at relatively high levels in bullion. It can be removed by the addition of zinc chloride to form a thallium chloride dross, leaving zinc in the bullion for subsequent removal.

Tin Tin tends to follow arsenic and antimony and is removed from lead bullion in softener slags or caustic slag from the Harris process. It is the most soluble of the three metals in caustic slags and can be selectively leached with water and precipitated from the resulting solution by the addition of lime to form calcium stannate containing around 40 per cent Sn.

Zinc During smelting zinc reports to slag, although it can also be present in any matte formed at up to eight per cent. It is only present in trace amounts in lead bullion. However it is added in excess in the Parkes process for the separation of silver, but is subsequently removed by vacuum distillation and residual traces by the final caustic refining stage. Treatment of lead bullion with sodium hydroxide and with oxidation by oxygen or sodium nitrate will efficiently remove zinc into a caustic slag.

REFERENCES Barrett, K G and Knight, R P, 1989. Operation of the bottom blown oxygen cupel, in Proceedings American Institute of Mining, Metallurgical and Petroleum Engineers (AIME) Conference, Las Vegas, February (American Institute of Mining, Metallurgical and Petroleum Engineers (AIME): Littleton) Betterton, J O and Lebdeff, Y, 1936. Debismuthisation of lead with alkaline earth metals, including magnesium, Transactions of American Institute of Mining, Metallurgical and Petroleum Engineers (AIME), 21:205-225. Davey, T R A, 1963. Phase systems concerned with the copper drossing of lead, Transactions of the IMM, 72:533-562. Davey, T R A, 1979. The physical chemistry of lead refining, in Proceedings Lead-Zinc-Tin ’80 Symposium, pp 477-507 (Metallurgical Society and American Institute of Mining, Metallurgical and Petroleum Engineers (AIME): Littleton). Davey, T R A, 2000. Equilibrium versus kinetics in lead refining, in Proceedings Lead-Zinc 2000 Symposium, pp 617-636 (The Minerals, Metals and Materials Society (TMS): Warrendale). Davey, T R A and Happ, J V, 1971. The decoppering of lead, tin and bismuth by stirring with elemental sulfur, Proceedings of the Australasian Institute of Mining and Metallurgy, 237:63-70. Davey, T R A, Jensen, G and Seguit, E R, 1980. Decoppering lead and hard lead by stirring with sulfur, in Proceedings Australia/Japan Extractive Metallurgy Symposium, pp 301-312 (The Australasian Institute of Mining and Metallurgy: Melbourne). Davey, T R A and Williams, K C, 1956. Continuous vacuum dezincing plant at the BHAS Pty Ltd, Port Pirie, Proceedings Australasian Institute of Mining and Metallurgy, 180:1-11. Dawson, R, 1980. Silver recovery at the Broken Hill Associated Smelters Pty Ltd, Port Pirie, SA, in Mining and Metallurgical Practices in Australia, pp 10:249 (The Australasian Institute of Mining and Metallurgy: Melbourne).

224




DeKeyser, J and Jespers, W, 1981. The Harris refinery of MHO, in Proceedings 110th American Institute of Mining, Metallurgical and Petroleum Engineers (AIME) Meeting, Chicago, February (American Institute of Mining, Metallurgical and Petroleum Engineers (AIME): Littleton). Fern, J H and Shaw, R W, 1983. Copper separation and recovery at the BHAS lead smelter, Port Pirie, South Australia, in Proceedings MMIJ – The Australasian Institute of Mining and Metallurgy Joint Symposium, Sendai. Isbell, W T, 1949. Development of a process for vacuum dezincing of lead, Transactions of the American Institute of Mining, Metallurgical and Petroleum Engineers (AIME), 182:186. Lal, R and McNichol, J H, 1987. The BHAS copper leach plant, The Minerals, Metals and Materials Society (TMS) technical paper A87-1. Mantel, C L, 1960. Electrochemical Engineering, fourth edition, pp 166-171 (McGraw Hill). Moodie, J, 1976. Debismuthising lead, PhD thesis (unpublished), The University of Melbourne. Moriya, K, 1989. Achievement in lead smelting during a quarter century at Mitsubishi-Comino’s Naoshima smelter, in Proceedings Primary and Secondary Lead Processing Symposium, pp 71-86 (Canadian Institute of Mining and Metallurgy (CIM): Montreal). Paschen, P and Winterhagen, H, 1968. Die raffination von blei mit atznatron, Erzmetall, 21:14-20. Wiltshire, P G, Miller, M J and Bauer, J C, 1989. Improved decoppering of lead bullion at Port Pirie, in Proceedings Non-Ferrous Smelting Symposium – 100 Years of Lead Smelting and Refining at Port Pirie, pp 65-70 (The Australasian Institute of Mining and Metallurgy: Melbourne). Zaitsev, V Y and Margulis, E V, 1985. The metallurgy of zinc and lead, Metallurgiya, p 96.



225

CHAPTER 13 Electrolytic Refining of Lead Electrolytic refining involves electrochemically dissolving lead from an impure anode and depositing lead free of impurities onto a lead cathode. The impurities predominantly remain at the anode as a residue or anode slime, from which valuable by-product metals can be recovered. The electrolytic process for the refining of lead is based on the use of a fluosilicate electrolyte and was developed by Betts in 1902. At that time the electrolytic refining method was the only technique available for the efficient handling of lead bullion produced from high bismuth concentrates, until development of the Kroll-Betterton process. Although it is ideally an alternative to multistage thermal refining, it does not fully replace all steps, and copper and tin removal by thermal methods is generally required as well as final caustic refining. For electrolytic refining, an electrolyte is required that has a reasonable lead solubility, is stable, has a high electrical conductivity and will yield a smooth compact deposit of lead. Various organic acids have good lead solubility and conductivity but tend to be unstable. It was found during the early development of the process that fluosilicic acid, fluoboric acid and sulfamic acid were most suitable and fluosilicic acid was the least costly. Sulfamic acid systems were also used, but showed instability at high current densities. Consequently, most electrolytic refining operations are based on a fluosilicate electrolyte.

PROCESS PRINCIPLES The chemical processes involved are simply the dissolution of the lead anode according to Equation 13.1 to produce lead ions in solution, and the discharge of lead ions and the deposition of metallic lead at the cathode according to Equation 13.2: Pb0 = Pb2+ + 2e

Anode Cathode

Pb2+ + 2e = Pb0

(13.1) (13.2)

The above reactions are the reverse of each other and the reversible electrode potential under standard conditions (E0) is -0.13 volts. For non-standard conditions the electrode potential is given by Equation 13.3: E = E0

RT × ln(a) nF

(13.3)

where: E0 is the electrode potential under standard conditions R is the universal gas constant (8.314 joules K-1mole-1) T is the temperature in K a is the ionic activity of the metal ion in solution = molar concentration × mean ion activity coefficient (which is dependent on solution composition and the ionic strength of the solution) F is Faraday constant = 96 500 coulombs per mole n is the number of electrons involved per mole



227

CHAPTER 13 – Electrolytic Refining of Lead

E0 values are given for various metals at 25°C relative to the standard hydrogen electrode in Table 13.1. TABLE 13.1

Standard electrode potentials (relative to the H + electrode at 25°C). 0

E volts

Electrode reaction 2+

Au

0

/ Au

+1.42

-

+1.36

Cl2 / Cl

Hg2+ / Hg0

+0.85

Ag+ / Ag0

+0.80

Fe3+ / Fe2+

+0.77

5+

Sb

3+

/ Sb

+0.75

SeO32- / Se0

+0.74

Te4+ / Te0

+0.63

As5+ / As3+

+0.58

-

O2 / OH

+0.40

Cu2+ / Cu0

+0.34

As3+ / As0

+0.30

Bi3+ / Bi0

+0.20

3+

0

+0.10

Pb2+ / Pb0

-0.13

Sn2+ / Sn0

-0.14

Ni2+ / Ni0

-0.25

Sb

/ Sb

H+ / H0

2+

Co

0

0

/ Co

-0.28

In3+ / In0

-0.34

Cd2+ / Cd0

-0.41

Fe2+ / Fe0

-0.44

Ga3+ / Ga0

-0.56

2+

0

Zn / Zn

-0.76

Mn2+ / Mn0

-1.10

Al3+ / Al0

-1.66

Mg+ / Mg0

-2.39

In the practical situation electrodes depart from equilibrium and for the electrode process to proceed at a practical rate, a higher potential than the equilibrium value is needed to overcome the energy loss in activating ions taking part in the reaction and in transporting ions to the electrode surface. The actual observed electrode potential for an operating cell is then: Ea = E + ηa +

(13.4)

c

where: E is the reversible potential as defined above ηa is the activation overpotential (or chemical polarisation) is the concentration overpotential (or concentration polarisation) c

228




The activation overpotential may be collectively made up of three broad processes: 1. the energy required to decompose complex ions in solution, 2. the energy barrier to be overcome to discharge the ion, and 3. the energy required to arrange the metal on the electrode surface in a particular crystalline state. The concentration overpotential normally has little impact until high currents are reached and depends on boundary layer effects, either fluid surface film effects or the build up of a slime layer, eventually limiting the current to a level determined by ionic diffusion. For electrorefining the standard potential of lead at each electrode will be the same and the equilibrium voltage (at zero current) between the two lead electrodes is therefore zero. In practice it is desired to transfer lead from the anode to the cathode and a current will flow, giving a potential drop across the solution path between the two electrodes as well as overpotentials at each electrode. This will define the starting voltage for a refining cell. As the anode lead is dissolved the residual impurities will form a slime layer. This will impede the flow of lead ions and hence the concentration overpotential will rise and the anode potential will increase at a given current. Hence the cell voltage will progressively increase as the anodes are consumed. Removal of the slime layer can be used to reduce the cell voltage. Cell voltages can vary from 0.45 volts for new anodes up to 0.6 volts for old anodes and a typical cycle range is shown in Table 13.2. TABLE 13.2 Refining cell voltage range over cycle. Voltage drop due to:

Start

Finish

Cathode contact

0.07

0.04

Cathode sheet resistance

0.04

0.02

Electrolyte resistance

0.35

0.35

0

0.15

Anode sheet resistance

0.01

0.01

Anode contact resistance

0.01

0.02

Total

0.48

0.59

Anode slime resistance

Theoretically, elements with electrode potentials lower than lead in the electrochemical series will also dissolve at the anode, and from Table 13.1 this includes tin, nickel, iron, cadmium and zinc. All the elements with higher electrode potentials, including antimony, arsenic, bismuth, silver and gold, selenium and tellurium will remain at the anode and will not transfer to solution as long as lead is present. However, as refining proceeds the potential drop across the slime layer will increase and potential at the outer surface can reach a critical level for the dissolution of some impurities. Bismuth is the most critical for refining performance and it is generally acknowledged that the critical anodic overpotential above which bismuth starts to dissolve rapidly is 200 mV. Only elements with electrode potentials higher than lead will co-deposit with lead at the cathode. Hence, those elements that dissolve with lead from the anode are theoretically not able to deposit at the cathode and will remain and build up in solution. Ideally this automatically allows only the predominant metal to transfer from the anode to the cathode, but does require treatment of the electrolyte to remove those impurities which tend to accumulate. However, departure from equilibrium caused by high transfer currents and different electrode overpotentials for dissolution and deposition can broaden the range of elements that can be co-dissolved at the anode and co-deposited



229


at the cathode. This consideration impacts on tin in particular with an electrode potential so close to lead, and also antimony to some extent, although it is largely insoluble in the electrolyte. Hydrogen will not form at the cathode because of its relatively high activation overpotential. If an element is insoluble in the electrolyte then its activity in solution is very low and the deposition potential in accordance with Equation 13.3 can be reduced below that of lead so that it cannot co-deposit on the cathode. This applies in the case of iron and to some extent antimony. The capacity of the system is governed by the rate of transfer of lead ions from the anode to the cathode, which is expressed by Faraday’s Law: m=

ItM nF

(13.5)

where: m is the mass deposited in grams I is the current passed in amps t is the time of passage of the current in seconds M is the molecular weight of the material deposited n is the number of electrons involved per mole F is the Faraday constant = 96 500 coulombs per mole From the above expression, and for the deposition of 1 kg of lead, 258.7 amp hours (It) are theoretically required. The efficiency with which electric current is used for the dissolution and deposition of lead is an important parameter and can differ slightly for the anode and the cathode. If the anode current efficiency is higher, then more lead will dissolve than will deposit on the cathode and the lead concentration of the electrolyte will tend to rise, and vice versa. This can be controlled by the lead content of the anode, which is generally around 98.5 per cent Pb for balance, but it can also be controlled by adding or removing lead salts from the electrolyte. Generally the cathode current efficiency is slightly lower than the anode current efficiency, but is usually between 93 and 98 per cent.

PRACTICAL OPERATIONS A generalised flow sheet of lead refining using the Betts process is shown in Figure 13.1 and details of various operations are given in Table 13.3.

Electrolytic cells Cells used are open topped rectangular tanks constructed of concrete and lined with PVC. Cell dimensions are dictated by the size of the electrodes, and in particular the anode dimensions. Anodes are generally in the range of 200 to 300 kg starting weight with thicknesses of around 30 mm. Immersed depth is commonly about 1.3 times the electrode width, hence the following is typical: • Total anode length 1120 mm Range 950 to 1250 mm • Immersed depth

1020 mm

850 to 1150 mm

• Width

780 mm

570 to 980 mm

• Thickness

30 mm

25 to 35 mm

• Immersed surface area • Weight

230

2

1.59 m 300 kg

0.97 to 2.2 m2 190 to 400 kg




Hot Smelter Bullion

Copper Drossing

Copper Dross

Sulfur

Sulfur Drossing

Sulfur Dross to Smelter

Oxygen Softening

Softener Slag

Crude Lead

Anode Casting Cathode Starter Sheets

Anodes

Electrolytic Cells

Anode slimes

Cathode lead

Washing

Melting and Caustic Refining Carbon

Slag to smelter

Starter Sheet Preparation

Melting and Reduction

Ingot Casting

Alloy of Pb, Ag, Bi, Cu

As/Sb Fume Pb/Sb Slag

Refined Lead

Converter Stage 1

As/Sb Fume Oxygen

Coal Sb/Cu/Bi Slag Sulfur

Rotary Furnace Reduction

Converter Stage 2 Silver Dore

Oxygen

Electrolytic Refining

Copper Matte

PB/Bi Alloy for Further Refining

Gold Refining Refined Silver Refined Gold

FIG 13.1 - Flow sheet for electrolytic lead refining (Betts process).



231

Operation

Takehara (Japan)

Chigirishima (Japan)

Harima (Japan)

Kamiok (Japan)

Shenyang (China)

Teck Cominco (Canada)

La Oroya (Peru)

Onsan (Korea)

Electrolyte – Pb (g/L)

236 - 270

75

90 - 100

75

74

170 - 180

75

90 - 110

60

65 - 75

500

1000

– Free H2SiF6

52 - 61

Additives – Lignin sulfonate (g/t)

110 - 115

75

38 - 58

85 - 95

1000

180

300 - 450

275

600

200 - 300

Glue (g/t)

210 - Aloes

Other (g/t) Electrolyte temperature (°C)

500 550 Beta naphthol

28 - 43

35 - 38

40 - 43

35 - 45

32 - 45

38 - 42

40

35 - 38

Current density – A/m cathode

120 - 140

147

185

135

154 - 172

178 - 192

156

145 - 170

Cell voltage

0.55

0.55

0.46

0.45 - 0.5

0.5 - 0.6

0.5 - 0.7

165

125

120

143

175

1.0

0.68

1.3

1.8

0.8

2


0.5 - 0.6

0.47

Electrolytic energy (kWh/t)

156

143

Anodes – Sb %

0.98

1.25 0.25

Bi %

0.50

As %

0.01

Cu %

0.02

Sn %

0.02

Size – Length × height × thickness (mm) Weight (kg)

0.13

0.3

0.40

1.5

0.25

0.05

0.10

0.26

0.15

0.02

0.01

<0.05

0.10

0.10

0.073

0.05

0.10

0.044

1150 × 1000 × 32 1200 × 800 × 24 970 × 740 × 35 1140 × 990 × 20 920 × 620 × 23 410

Anode spacing (mm) The Extractive Metallurgy of Lead

Life (days)

0.03

0.5

4-5

Scrap (%)

0.004

0.01 880 × 670 × 30

940 × 690 × 25 1250 × 800 × 25

250

280

250

140

200

150

270

100

110

110

95

100

100

100

7

8

6

2.5

6

4

6

30

26.5

45

20 - 25

40

49

38

Anode slime (as % of anode wt)

2.4 - 3.6

2.9

1.3

1.4

1.1 - 1.4

4

4

1.2

Cathode starting thickness (mm)

1.0

0.6 - 1.0

0.8

0.7

0.8 - 1.0

0.5

0.6

0.8

Cathode life (days)

4

7

5

2.5

6

4

6

Number cathodes per cell

43

29

44

33

25

41


232

TABLE 13.3 Electrolytic lead refining operations by the Betts process (Gonzalez-Dominguez, Peters and Dreisenger, 1991; Siegmund, 2000).


Cathodes are 20 mm greater in dimensions than the anode and are usually 0.7 to 1.0 mm thick as starting sheets, growing to between 12 and 23 mm (120 or 220 kg respectively) depending on the cycle time in comparison with the anode. The number of anodes per cell is commonly around 28, but there are some larger installations with over 40 anodes per cell. Cathodes total one more than the number of anodes and are located at the ends of the cell and between each anode. Electrodes are handled by overhead gantry crane and are lifted from the cell at the end of their nominated life. Centre line anode spacing is commonly 100 mm, resulting in a gap between new anode and cathode surfaces of 34.5 mm with a 1 mm thick cathode starting sheet. The cathode thickness will increase, but the anode thickness will not decrease to the same extent due to the residual slime layer. Hence the gap will decrease with time. If the gap is too small there is an increasing risk of short circuiting between the two electrodes. Anode head-bars rest on a busbar along one side of the cell and cathode head-bars rest on a busbar on the opposite side. Cells may be placed side by side, in which case the anode busbar for one cell serves as the cathode busbar for the adjacent cell. This means that cells are electrically connected in series with connections to the DC power supply only at the end of each cell bank. Cathode current densities used generally range between 140 and 190 amps/m2. For a typical electrode immersed surface area of 1.66 m2 from the above typical dimensions, the current flow per cathode is up to 320 amps and for a cell with 28 anodes is 8960 amps. This current will transport around 1.18 kg/h of lead per cathode, requiring 180 hours (7.5 days) to deposit a full load of 210 kg. Anode life is typically six to eight days and cathode life may be the same or half the anode life. In the case where the cathode life is half the anode life it is often the practice to lift both electrodes from the cells, but to scrub half life anodes to remove anode slimes and return them to the cell for the remaining half life. This practice will reduce the anode potential and the average cell voltage and there must be a balance between the energy cost savings and the extra costs of handling and possibly providing extra cathode starting sheets. With a theoretical 258.7 amp hours per kilogram of lead and a current efficiency of 95 per cent the actual current required is 272.3 amp hours per kilogram. For an average cell voltage of 0.5 over the anode cycle life the energy use is 136.2 DC watt hours per kg (or DC kilowatt hour per tonne of lead). With busbar and conversion losses of the order of ten per cent the AC energy consumption is close to 150 AC kWh per tonne of refined lead, which is a relatively low cost item.

Anodes Anodes are cast from smelter bullion after copper separation and partial softening if needed, either in flat open moulds on a casting wheel as a measured amount, or into closed ‘book moulds’. One critical aspect of the performance of the anode is that it should dissolve at a uniform rate and this can be influenced by the microstructure and uniformity of the anode as well as by the impurity content. In this regard the casting method and the cooling rate are most important, and the use of book moulds is preferable in achieving better control and uniformity of the surface structure. With an open mould as in wheel cast anodes, the upper and lower surfaces are quite different in terms of crystal structure, oxidation and flatness. The other key issue with lead anodes is the nature of the anode slime deposit. It is desirable that this adheres firmly to the anode and does not separate and fall to the bottom of the cell. If it separates it can cause a short circuit between the electrodes. It can then also react differently with the electrolyte, since various impurities, which would otherwise be in electrical contact with lead and



233


remain in the slimes, could dissolve in the electrolyte and co-deposit with lead, thus reducing the purity of the cathode product. It is also important that the slime deposit adhering to the anode is reasonably porous to allow access by the electrolyte and the diffusion of lead away from the anode. A dense deposit will block the anode dissolution reaction in the extreme, but will raise the cell voltage to a point that may cause impurities in the outer layers of the deposit to dissolve. This effect will be aggravated by increased cell temperatures. Key factors affecting slime adherence are the microstructure of the anode and the presence of certain impurity elements. The microstructure determines the distribution of impurity containing phases at the grain boundaries, which can form a residual honeycomb structure in the slime layer after the lead has been dissolved, and is influenced by the cooling rate of the anode during casting. Apart from microstructure, certain impurity elements significantly affect the adherence of the slime layer, in particular bismuth and antimony, and to some extent arsenic and silver (Tanaka, 1977). Bismuth has considerable effect but is not controllable and would not be purposely added. Antimony is the key controllable element and is adjusted to achieve a suitable slimes deposit. Generally, if antimony is much above 1.5 per cent, the slimes become dense and blocking, and if below 0.5 per cent the slime layer will not adhere. Copper will give rise to a dense slime deposit and will cause uneven dissolution of the lead and levels generally should be below 0.1 per cent or 0.05 per cent. Given these constraints it is normally necessary to remove copper from smelting bullion using the decopperising techniques outlined in thermal refining – Chapter 12. Since copper is often a major impurity in crude lead bullion, prior removal can also significantly reduce the amount of anode slimes which have to be treated. As indicated above, tin is one impurity with an electrode potential close to lead that will dissolve and co-deposit on the cathode. Hence it can be necessary to remove tin before casting anodes, although smaller amounts can be removed from cathode lead in the final caustic drossing stage. As outlined under thermal refining (see Chapter 12), tin is removed together with arsenic and antimony in the softening process and depending on the levels of tin in smelter bullion as well as antimony, some degree of softening will probably also be required. This step may leave less antimony than is desired for control of the slime deposit and it can be necessary to also add antimony to the bullion before anode casting. Given the above requirements for control of anode composition it is therefore often necessary to subject smelter bullion to decopperising and softening prior to electrolytic refining and the casting of anodes. Anodes are lifted from the cells at the end of their nominated life, which may be between six and eight days. The slime deposit is removed by brushing using steel brushes and the residual ‘scrap’ lead is recycled back to anode casting. Residual lead consists of that part of the anode above solution line as well as the remaining sheet below the solution line, which is needed to structurally hold the slime deposit. The amount of unrefined lead recycled in this way is about 25 to 40 per cent of input lead. As indicated above, anodes may be lifted at half their nominal life, cleaned and returned to prevent irregular dissolution and to reduce cell voltage. As the anode is dissolved and the slime layer increases, it is noted above that the voltage increases for a given cell current and there is a tendency for increased dissolution of unwanted impurities and contamination of the lead cathode. This effect is shown by the fact that the impurity level, exemplified by the bismuth content of refined lead, increases as the anode scrap ratio falls, as shown in Figure 13.2. This clearly illustrates that there are limits to the extent of recovery of lead from each anode and the life of the anode.

234




14

Bi content of refined lead (ppm)

12

10

8

6

4

2

0 25

30

35

40

45

Scrap ratio (%)

FIG 13.2 - Effect of anode scrap ratio on Bi content (Kubota et al, 1995).

Cathodes The standard approach has been to use thin lead starter sheets for the cathode and to melt the total cathode after removal from the cell. The alternative approach is to use stainless steel cathodes and strip the deposited lead from the starting cathode. The latter has been applied more recently in secondary lead refining using fluoroborate electrolytes (Olper, 1998). There must be a balance between the requirements for producing starter sheets and equipment required for cathode stripping. Lead sheet for production of starting sheets is usually produced using a water-cooled cast iron drum partly immersed in a bath of molten lead and rotating such that lead solidifies in a thin layer on the surface of the drum and is peeled off the top as a continuous sheet. Thickness of the sheet is controlled to between 0.8 and 1 mm by the temperature of the bath and the depth of immersion, as well as the speed of rotation and flow of coolant inside the drum. Cathode starting sheets are cut from the continuous sheet and are attached to a copper head bar by wrapping around and riveting. The cathode starter sheet is straightened in a press before being loaded into the cells. Weight is generally around 10 kg. The lack of rigidity of the starting sheet is a significant factor in limiting the size of the electrodes. Lead has a tendency to form dendritic deposits and any such formations will bridge across from cathode to anode causing an electrical short circuit. It is therefore important to ensure that the deposit is dense and smooth and this is achieved partly by the selection of the electrolyte, but also by the addition of smoothing agents such as animal glue or various lignin sulfonates, aloes, or other organic agents. Beta naphthol can sometimes also be added. Use of these types of reagents is approximately 1 kg/t of lead in total. When cathodes are lifted from the cells the head bars are removed for reuse and the lead cathodes are washed, then melted in a standard refining kettle.

Electrolyte The electrolyte is a solution of lead fluosilicate and free fluosilicic acid. The lead concentration is normally in the range of 75 to 120 g/L and the free fluosilicic acid 65 to 100 g/L. The total equivalent fluosilicic acid is close to twice the free acid amount.



235


Electrolyte temperature is normally in the range of 36 to 48°C, and little control by way of heating or cooling is needed. At high operating temperatures fluosilicic acid has a significant vapour pressure due to its decomposition, according to Equation 13.6. Both HF and SiF4 are toxic and need to be removed by adequate ventilation of the cell house. H2SiF6 = 2HFv + SiF4v

(13.6)

The lead content of the electrolyte may rise or fall depending on the relative current efficiencies of the anode and cathode reactions and can be controlled by precipitating lead as sulfate from a bleed stream or by passing the electrolyte over a column of granulated lead (or adding PbO) to increase the level. Generally the anode current efficiency is higher than the cathode efficiency by up to 0.5 per cent and lead will tend to build up and must be reduced from time to time. Cominco have proposed a control by simply passing a bleed stream through an electrowinning cell using graphite anodes, and see this as preferable to lead sulfate precipitation. With electrowinning, PbO2 tends to deposit at the anode, but can be suppressed by the addition of 75 mg/L of arsenic to the electrolyte as sodium arsenate. However, anode life is very poor at around three months (Gonzalez-Dominguez, Kirby and Heim, 1995). Minor impurities can be controlled by a bleed of electrolyte and often the bleed resulting from entrainment in anode slimes is sufficient for this purpose. As indicated above, elements that will dissolve are mainly tin, nickel, iron, cadmium and zinc. Tin should be removed prior to anode casting if it is significant, otherwise if present in small amounts the electrolyte can be passed through a column of granulated lead for removal by cementation. This is suitable for controlling the bismuth concentration in electrolyte to less than 25 mg/L, which is equivalent to around 10 ppm in cathode lead. Arsenic and antimony are virtually insoluble in the electrolyte and remain in the slimes. Zinc and cadmium are present in very small amounts and will not co-deposit with lead, but can be controlled by a small bleed of electrolyte. Iron normally does not dissolve from the anode to any appreciable extent due to its presence in crude lead bullion as a sulfide or matte rather than metallic iron. Nickel is generally not present in sufficient quantities in lead bullion for there to be a concern with accumulation in the electrolyte. As a result of the above, electrolyte purification requirements are minimal.

Typical installation performance Based on the above details, the characteristics and performance of a typical Betts electrolytic refining installation can be described as in Table 13.4.

CURRENT MODULATION In order to maintain the overpotential at the slime layer surface, and consequently the total potential drop across the slime layer below a critical value, it is possible to commence an anode cycle at high current density and progressively reduce through the life cycle as the slime layer builds. This will enable cell productivity to be raised, but does add major complications in current control to individual cells, since cells will generally be operating on a range of anode life cycles as not all anodes are lifted at the one time. Consequently, current modulation is not normally practised.

236




TABLE 13.4 Betts process characteristics and performance. Individual weight Anodes

Cathodes

Total cell

Total installation

270 kg

Number per cell

28

Total starting anode load

7560 kg

Immersed area

1.59 m2

Life

Six days or 144 hours

Starter weight

10 kg

Number per cell

29

Immersed area

1.66 m2

Current density

190 amps/m

2

Current efficiency

95%

Current supply

8831 amps

Cell voltage – average

0.5 volts

Lead transfer rate

32.43 kg/h

Lead transfer per anode cycle

4670 kg

Slime produced

150 kg

Residual anode

2740 kg

Scrap ratio

36% of anode weight

Refined lead production

100 000 t/a (285 t/d)

Number of cells required

366 operating (370)

Total series cell voltage drop

183 V

Contacts and busbar voltage drop

20 V

DC supply voltage

203 volts

DC power input

1793 kW

DC energy per tonne of lead

151 kWh/tonne

PERIODIC CURRENT REVERSAL Short circuiting between anode and cathode by dendritic growth of cathode lead can be a significant problem and requires constant cell voltage monitoring to detect short circuits from this source as well as from the deformation of electrodes. The practice of periodic current reversal (PCR) has been used to assist in smoothing the cathode deposit and minimising dendritic growth. Consequently, the number of short circuits can be significantly reduced. The effect of the reverse current is to dissolve the deposited lead and this is concentrated at high points or dendrites, thus smoothing the deposit. The ratio of forward to reverse current is around 100:1, with a cycle time of less than one minute. It is also found that the free acid level in the slimes layer is greater under this practice, reducing the voltage drop through the slime layer and giving a lower cell voltage. This effect on energy consumption is greater than the impact of current reversal, so that the overall energy consumption is slightly reduced at normal current densities. Electrical switching facilities have been a complication in the past and this practice has rarely been applied, but the application of thyristoformers can allow the technique to be applied more easily.



237


BIPOLAR ELECTRODE CELLS In a bipolar system only the end electrodes in each cell are connected to the electrical supply. Each electrode plate in the cell between the end electrodes will become an anode on one side and a cathode on the other. Current can bypass around the edges of the electrodes to flow directly through the electrolyte to the end electrodes and represents an energy efficiency loss. This can be minimised by fitting the electrodes close to the sides and baffles in the bottom of the cell. The concept is to fill the cell with cast lead anodes, which dissolve only on the one side, and pure cathode lead is deposited on the other side. The electrodes are lifted from the cell at the end of a cycle and the freshly deposited lead is separated from the residual anode. The advantages of the system may be as follows:

• Starter sheets are not required. • Larger electrodes are possible with stronger electrodes rather than using thin starter sheets, which are easily bent.

• There are no electrical contacts for every electrode – only the two end electrodes per cell. This reduces cell voltage to a minor extent and can allow higher current densities.

• Since all electrodes are lifted at the end of a cycle, mechanical handling is simpler. • It is more practical to cover the cells to contain vapour emissions. Cominco developed a bipolar cell using 4 m2 electrodes, each weighing around 700 kg and operating between 200 and 265 amps/m2, with spacing of 70 mm. At the higher than normal current density, periodic current reversal (PCR) and the application of smoothing agents were necessary to minimise dendritic growth. PCR operated on 18 cycles per minute, with current reversal for 150 ms per cycle. Current efficiency was 90 per cent and cell voltage 2.6. Voltage was programmed to reduce during the operating cycle to maintain the voltage drop through the slime layer at less than 200 mV. Cathode lead was stripped from the residual anode by flexing the electrode in a press to break the bond between the two and allow them to fall apart.

FINAL REFINING OF CATHODE LEAD Melted cathode lead is usually subjected to treatment with sodium hydroxide and some oxidation with air or oxygen in the melting kettle to remove traces of antimony and tin before casting into ingots. The product is particularly high grade and can achieve 99.999+ per cent lead content. Drosses from final refining can be treated with softener slags and slime treatment slags in a short rotary furnace or in a blast furnace campaign to recover antimonial lead, which can be used for alloying and for control of the antimony content of anodes.

ANODE SLIMES TREATMENT The composition of anode slimes will be highly variable but typical ranges are shown in Table 13.5. Selenium and tellurium can sometimes be present in recoverable amounts. Treatment of the slimes is primarily for the purpose of recovering the precious metals silver and gold, but also to recover lead and antimony for recycle.

238




TABLE 13.5 Typical anode slime composition ranges. Element

Range

Antimony

40 - 50%

Silver

5 - 25%

Bismuth

4 - 15%

Lead

10 - 25%

Gold

100 - 300 g/t

Copper

1 - 4%

Arsenic

1 - 6%

The recovered slimes are first washed and filtered and then are melted at 850 to 900°C under reducing conditions, usually with the addition of coal or coke fines at about two to three per cent of the weight of material processed. A short rotary furnace can be used for this operation. Arsenic and some antimony tend to fume or collect in a lead antimonite slag (PbO.Sb2O3). The bulk of the lead, silver, bismuth, copper and gold form a molten metal phase, which is transferred to a converter where it is blown with air at around 900°C. Residual arsenic and antimony will fume and are removed in the gas stream. Lead, some antimony, copper and bismuth form a slag. Slag is tapped at an intermediate stage and oxidation is continued at 1000°C until all antimony is removed. Final slag contains predominantly copper and bismuth, and the metal is essentially a silver doré. The silver doré is separated and refined following the same procedures outlined in Chapter 12 for thermal refining. The bismuth-rich slags from the converter operation and from doré cupellation can be reduced in a short rotary furnace with coal and some sulfur addition to give a bismuth–lead alloy and a copper matte. The bismuth–lead alloy can be refined by a similar sequence as used for the thermal refining of lead, as follows:

• oxidation to remove arsenic, antimony and lead; • silver removal by zinc addition; • zinc removal by chlorine, which also removes any residual lead as a chloride slag; and • final caustic treatment to yield a refined bismuth metal. OTHER ELECTROLYTIC REFINING SYSTEMS The San Gavino lead refinery in Sardinia operated a system based on a sulfamic acid electrolyte from 1955 until the early 1960s. The choice was influenced by the lack of suitable materials to produce a fluosilicate electrolyte and the fact that most of the major impurities – bismuth, arsenic, antimony and silver – were virtually insoluble, tin formed an unstable salt and the polarisation potential for copper was very high while that for lead was quite low. This gave a performance similar to that of a fluosilicate electrolyte although with higher power costs. Tin remains in the anode slimes due to its low solubility, giving an advantage for the use of sulfamic acid electrolyte for bullion containing high levels of tin (Stracchi et al, 1967). Electrolyte composition was 80 to 85 g/L lead and 45 to 50 g/L free sulfamic acid (HNH2SO3). One to two grams per litre of tannic acid (tannin) was also used as a smoothing agent. Current densities used were around 90 amps/m2, and it was found that as the current density increased to 100 amps/m2 or above there was an increasing degree of decomposition of sulfamic acid.



239


Decomposition resulted in the formation of ammonium sulfate and hence caused the precipitation of lead sulfate from solution. The losses amounted to 4 kg of sulfamic acid and 2.5 kg of PbO per tonne of refined lead at 90 amps/m2, but increased rapidly with the current density. Cell voltage was also 0.7 at these conditions with an energy consumption of 210 kWh/t of lead. This is significantly higher than for the fluosilicate system, which has a much higher solution conductivity. As a consequence San Gavino changed to the fluosilicate electrolyte and was able to double the productivity of the refinery at significantly lower cost (Freni, 1965). Fluoboric acid (HBF4) is a very similar alternative to fluosilicic acid (H2SiF6) as a basis for the electrolytic refining of lead, but is a significantly more expensive reagent. It was used at the Norddeutsche Affinerie refinery in Germany in the 1950s but has been replaced with the more conventional fluosilicate system. Fluoborate has the advantage of higher solution conductivity and can produce smooth dense cathode deposits at much higher current densities than are possible with fluosilicates. It also has an HF vapour pressure an order of magnitude lower than fluosilicate solution and can operate at higher temperatures. However, cost has dictated against the use of fluoborates in the conventional Betts refining process. A range of electrolytic technologies using fluoborates have been developed by Engitec SpA in Italy, basically as an electrowinning technology (see Chapter 9), but this has also been adapted as a novel approach to lead electrorefining (Olper, 1998). This technology uses a ferric/ferrous couple to oxidise and dissolve lead from granulated bullion. The dissolved lead is then electrodeposited on a cathode in a compartmented cell, the depleted solution flows into an anode compartment where iron is re-oxidised to the ferric state, and the solution returns to the leaching stage. The reactions involved are shown in Equations 13.7 to 13.9: Leaching reaction: Cathode reaction: Anode reaction:

Pb + 2Fe(BF4)3 = Pb(BF4)2 + 2Fe(BF4)2 Pb(BF4)2 + 2e + 2H+ = Pb + 2HBF4

2Fe(BF4)2 + 2HBF4 = 2Fe(BF4)3 + 2H+ + 2e

(13.7) (13.8) (13.9)

The cell uses a graphite anode within a fabric diaphragm compartment from which the spent electrolyte is drawn. Filtered leach solution as feed electrolyte is fed to the open topped cathode compartment containing a stainless steel cathode from which the deposited lead is stripped. Current densities are of the order of 300 amps/m2, which are 60 per cent higher than commonly used in the Betts process. Virtually all impurities in the lead bullion are retained in a leach residue, comparable to the anode slimes in the Betts process, but the leaching process can be continuous with the addition of fresh granulated lead and the separation of residue by filtration of the leach solution. There is no recycle of unconsumed lead. Ferric iron will oxidise tin to Sn4+, which will precipitate in the leaching stage as metastannic acid (H4SnO4) and is hence separated with the leach residue. It is therefore possible to process smelter bullion directly without softening, although decopperising will still be beneficial to provide a separate copper-rich stream, to reduce the quantity of mixed leach residue for further processing and to minimise interference with subsequent silver recovery. The disadvantage with this approach is the higher cell voltages resulting from the use of a diaphragm and the oxidation of iron at the anode with an E0 of +0.77 volts and an equilibrium cell voltage of 0.91. The resultant energy consumption is around 500 kWh/t of refined lead compared with around 150 kWh/t for the Betts process.

240




REFERENCES Betts, A G, 1908. Lead Refining by Electrolysis (John Wiley and Sons: Hoboken). Freni, E R, 1965. Electrolytic lead refining in Sardinia, Journal of Metals, November, pp 1206-1214. Gonzalez-Dominguez, J A, Kirby, R C and Heim, H, 1995. An evaluation of lead electrowinning in H2SiF6 – PbSiF6 electrolytes, in Proceedings Zinc and Lead ’95 Symposium, Sendai, pp 713-721. Gonzalez-Dominiguez, J A, Peters, E and Dreisinger, D B, 1991. The refining of lead by the Betts process, J of Applied Electrochemistry, 21:189-202. Kubota, H, Kusakabe, T, Takei, K and Takewaki, M, 1995. Current operations of Sumitomo Metal Mining’s Betts lead electrorefining, in Proceedings Zinc and Lead ’95 Symposium, Sendai, pp 353-366. Mantel, C L, 1960. Electrochemical Engineering, fourth edition, pp 185-192 (McGraw Hill). Olper, M, 1998. Fluoborate technology – A new challenging way for primary and secondary lead processing, in Proceedings Zinc and Lead Processing Symposium, pp 185-198 (The Metallurgical Society and Canadian Institute of Mining and Metallurgy: Montreal). Siegmund, A H J, 2000. Primary lead production – A survey of existing smelters and refineries, in Proceedings Lead-Zinc 2000 Symposium, pp 55-116 (The Minerals, Metals and Materials Society (TMS): Warrendale). Stracchi, P M, Peruzzi, R, Rozzoli, A, Sinigaglia, D and Vicentini, B, 1967. The refining of lead by the Betts process, Electrochim Met, 2:95. Tanaka, T, 1977. Adhesion of anode slime on anode surface in electrolytic refining of lead, Metallurgical Transactions B, 8(4):651-660.



241

CHAPTER 14 Alloying and Casting HANDLING MOLTEN LEAD AND ALLOYING Remelting, alloying and holding molten lead in preparation for casting operations is usually done in standard refining kettles as described in Chapter 12. Kettle capacity varies depending on alloy lot sizes but can range from 10 - 300 t, with larger kettles used for straight casting of pure refined lead. Alloys are prepared in batches by adding weighed components to an agitated kettle. In the case of antimony, antimonial lead of highly variable composition is often available from the reduction of antimony rich slags, and is used for the preparation of antimonial alloys. Composition is checked by spectrographic analysis and is adjusted and re-checked before casting. Molten lead is transferred from the kettle to the casting installation by a submerged centrifugal pump through steel piping. The piping is lagged and often electrically or open flame heated to prevent freezing at low flow rates. The kettle is drossed periodically by manual skimming and the dross can be recycled to a short rotary furnace for reduction and conversion back to metallic lead. Depending on the alloying components it may be necessary to recycle the lead bullion so produced back to the refinery feed.

SPECIFICATIONS Most primary lead is sold as ‘soft lead’ or ‘pig lead’, with a purity of at least 99.90 per cent. High quality lead is specified as ‘four nines’ or 99.99 per cent plus. Specifications for impurity contents are given in various Standards as shown in Table 14.1. Otherwise a wide range of alloys is produced to customer specifications, particularly antimonial and calcium–tin alloys and have not been standardised. Each battery manufacturer tends to have its own particular requirements to suit its own manufacturing needs and battery specifications.

CASTING Lead is commonly traded in the form of 25 kg ingots (pig lead), or as blocks of 500 kg, one tonne or two tonne weight. Typical dimensions are shown in Figure 14.1. Impure bullion as feed to an independent refinery is also commonly traded and is often cast and handled as four tonne blocks. The 25 kg ingots are often assembled into bundles of approximately one tonne. Such a bundle is made up of 42 ingots as seven layers, each of six ingots arranged in alternate direction per layer to give a bundle 560 mm square by 490 mm high. Four steel straps are used to hold the bundle together, with two straps in each direction. The bundle is arranged such that the tines of a fork-lift can get under the bottom layer for easy handling. Two tonne blocks vary significantly but may be 850 mm square at the top surface by 300 mm high. Similarly the four tonne blocks used for unrefined bullion are close to 1 m square at the top surface and about 450 mm high.



243

CHAPTER 14 – Alloying and Casting

TABLE 14.1 Composition of standard grades of lead. Element

United States ASTM pig lead specification BS 29 – 1955 Corroding Chemical

Ag max (%)

0.0015

Ag min (%) Cu max (%)

Pb Acid Common Pure Pure Smelter Pure Chemical Pb lead copper type A 99.99 99.94 Pb copper de-silvered Pb lead 99.99 99.985 0.002

0.002

0.001

0.001

0.001

0.0025

0.002

0.001 0.001

0.080

0.080

0.0025

0.001

0.001

0.001

0.08

0.003

0.001 0.001

0.040

0.040

Trace

0.001 0.001

0.002 0.0015

Cu min (%) Ag + Cu max (%)

0.020

India United Kingdom IS 27 – 1965 BS 334 – 1934

Germany DIN 1719 – 1951

0.04

0.0025

As max (%)

0.001

0.001

0.001

0.001

Sb max (%)

0.001

0.002

0.002

0.002

0.002

0.001 0.002

Sn max (%)

0.001

0.001

0.001

0.001

Trace

0.001 0.001

0.002

0.001 0.001

As + Sb + Sn max (%)

0.002

0.002

0.002

0.005

Zn max (%)

0.001

0.001

0.001

0.002

0.001

0.001

0.001

0.001

Fe max (%)

0.002

0.002

0.002

0.002

0.001

0.001

0.001

0.001

0.003

0.001 0.001

Bi max (%)

0.050

0.005

0.025

0.150

0.005

0.01

0.05

0.01

0.005

0.005

Cd max (%)

Trace

Co max (%)

0.001

Ni max (%)

0.001

Lead min (%)

99.94

99.90

99.90

99.95

99.99 99.985

99.94

99.90

99.99

99.99 99.94

600 mm

70 mm

80 mm

0.05

600 mm

300 mm

560 mm

25 kg Ingot

400 mm

400 mm

One Tonne Block

FIG 14.1 - Typical ingot dimensions.

Ingot (or pig) casting Two general types of machine are used for ingot casting – the rotary wheel caster and the straight line chain mould machine. The casting wheel uses a rotary table with water-cooled moulds arranged around the outer periphery of the wheel. Molten metal is poured into the moulds at a fixed station. The mould

244




then moves around with the wheel until close to the return position when the mould is rotated through 180 degrees and the ingots released onto a collection conveyor. The mould is then returned to its upright position ready for refilling. These machines were the first of the mechanised casting lines and have largely been superseded. They are not highly productive for their size and the relatively large floor area they occupy. Steady movement of the mould is difficult to achieve, and as a result the ingots may have ripple marks on the surface, which are less attractive to customers than a smooth surface finish. The straight line machines use the principle of a continuous line of moulds attached to two parallel chains driven by a variable speed tractor or caterpillar type drive. The line is inclined upwards from the pouring end and passes through a hood where the top surface is initially heated until full freezing of lead in the moulds has occurred, followed by water sprays to finally cool the ingots. Lead ingots are released as the chain turns over at the upper end to return. At the lower end lead is poured into the moulds at a controlled rate, usually set by the speed of the pump in the relevant supply kettle with fine adjustment of flow using a needle valve or equivalent device. The machine speed can also be set and adjusted to give the required production rate. There are many different pouring devices aimed at interrupting the flow of molten lead between one mould filling and the next arriving at the filling position. These devices are in the form of wheels and tilting (or oscillating) ladles. Their performance is largely judged by trouble-free operation and the consistency of weight of each ingot. To minimise dross formation, gentle flow of metal into the mould is important, without generating turbulence or air entrapment, such as would occur by pouring from a significant height above the mould. Submerged pouring systems are gaining favour, since they can minimise splash and dross formation. Following the pouring position there is a skimming station where dross is skimmed from the surface of the ingot. Complete removal of dross is important for the production of high quality ingots with a smooth surface appearance. Many automated skimming devices have been designed and trialled, but few have been able to perform with complete satisfaction and this commonly remains a manual task in most plants. Improved robotics technology mimicking manual operations has the promise of fully automating this boring and tiring (but essential) task. Ingots released from the casting machine are sorted and aligned on a conveyor and then can be automatically stacked into bundles. Otherwise the bundle stacking task is performed manually. Straight line machines can achieve casting rates of around 20 tonnes per hour.

Block casting Lead blocks are usually cast in water-cooled steel moulds on a wheel casting machine. In this case the moulds are filled at one position to set level and the time of rotation of the wheel allows for solidification and adequate cooling before the mould returns to the filling station. At some position before this point, the blocks are lifted from the moulds using a hoist or crane, which either attaches to two steel lifting eyes cast into the block or uses a vacuum lifting pad which attaches to the top surface of the block. The steel lifting eyes are normally used for four tonne bullion blocks for subsequent refining, whereas other means are used for pure lead and alloys, usually as one or two tonne blocks. The smaller blocks are then handled by fork-lift truck. The cooling time for large blocks of four tonnes is of the order of one hour. Blocks are usually stamped with relevant product information such lot numbers, grade and weight, and edge flashings are manually trimmed before despatch.



245


Sheet lead casting Soft lead is produced as sheet by direct casting using a rotating water-cooled drum immersed in a bath of molten lead. The lead solidifies on the surface of the drum and is peeled off from the top surface as the drum rotates, as shown in Figure 14.2. Finishing Rolls

Water Cooled Drum

Lead Sheet

Coil of Finished Lead Sheet

Molten Lead Bath

FIG 14.2 - Schematic of lead sheet casting.

Sheet thickness is controlled by the speed of the drum, the depth of immersion, the temperature of the molten lead bath and the intensity of water cooling applied to the inside of the drum. The sheet is usually run through a set of finishing rolls to provide a smooth surface and a precise gauge. Standard gauges are expressed as weight per unit area and are shown in Table 14.2. Width of the sheet is normally 900 to 1000 mm and is usually produced as a roll on a mandrel containing 1500 to 2000 kg of lead. Alloys may also be produced in this way, for example for the production of battery grids by punching and expanding. TABLE 14.2 Standard lead sheet gauges. 2

246

Weight ( kg/m )

Thickness (mm)

5

0.441

8

0.705

10

0.882

15

1.323

20

1.764



PART E ENVIRONMENTAL AND ECONOMIC ISSUES This part of the text covers the critical issues associated with the environmental impact of lead, its effect on human health and the necessary controls. Energy consumption for key processes is reviewed as well as processing costs and economics. Chapter 15 Chapter 16 Chapter 17

Health and Environmental Issues Energy Consumption Costs and Economics of Lead Production

CHAPTER 15 Health and Environment Issues INTRODUCTION The effects of lead on human health and its impact on the environment have been a significant concern to society for a long time, and have been investigated perhaps more thoroughly than for most other metals. This has led to the development of regulations controlling the lead industry in most developed countries and greatly influences the design and operation of lead processing plants. Of prime importance is the limitation and control of exposure of personnel working in the industry to lead in a harmful form, and the requirement to monitor the health of those personnel as well as emissions to the external environment. Because of the importance of this aspect of lead processing, and the influence it has on processes and methods of operation, it is useful to provide some understanding of the underlying science and reasoning behind the controls and regulations imposed on the industry.

LEAD IN THE ENVIRONMENT Although lead is a ubiquitous heavy metal it is relatively rare, constituting only 13 parts per million (ppm) of the earth’s crust compared with 50 and 70 ppm for copper and zinc respectively. It is more concentrated in granites at around 49 ppm due to its radiogenic origins. It has been concentrated as sulfide minerals largely from precipitation from hot brines. Apart from the weathering of sulfides its concentration in surface soils tends to relate to the presence of human activity, and is derived from its use in paints, pesticides, combustion of coal and most importantly the use of lead in automobile fuel. Lead is strongly bound within soils and is quite insoluble, limiting bioavailability and limiting its dispersion over time. Many of the uses of lead, which contributed to environmental contamination, such as automotive fuel and paints, have been phased out leaving most contamination as an historic legacy. Urban dust containing lead from these sources has been a major source of wider human exposure, but has declined markedly in recent decades as a result of ‘safe use’ and ‘abatement’ programs which flowed from a better understanding of the toxicity profile of lead. Most urban soils contain lead at between 15 and 40 ppm, and a soil is regarded as contaminated if it has over 500 ppm total lead content. Sea water contains 0.3 mg/L lead with fresh waters substantially lower. Drinking water standards are 15 µg/L in the USA and vary between 10 and 25 µg/L. Effluent standards vary from less than 0.5 mg/L to below 0.1 mg/L. Marine organisms tend to concentrate heavy metals, particularly shellfish, but lead has substantially lower concentration factors than other metals at 2600:1 compared with 32 500 for zinc, 10 500 for copper and 4500 for cadmium. There are often greater concerns for environmental contamination from other metals associated with lead smelting such as arsenic, antimony, cadmium, selenium and thallium. The impact of lead on the environment per se is not of prime importance and it is its entry to the food chain and potential for human exposure that is the primary determinant of emission and contamination standards.

THE TOXICOLOGY OF LEAD Ingestion is the main means of intake of lead into the blood stream. Absorption through the skin or in the lungs is minor, and inhaled lead in dust is usually expelled from the lungs in mucous and is



249

CHAPTER 15 – Health and Environment Issues

absorbed in the stomach. Adult humans absorb between five and 15 per cent of ingested lead and generally retain less than five per cent of the amount absorbed (Goyer, 1996). Due to metabolic differences young children absorb more ingested lead at around 30 to 40 per cent. Lead competes with several essential elements (iron, calcium and zinc) and dietary deficiencies in these elements will increase the absorption of lead. This is a particularly important issue for children and proper nutrition is essential to minimising risk. Once absorbed, lead is deposited in mineral tissue such as bones and teeth, in some soft tissues such as the liver, kidneys and brain, and some is retained in the blood stream. The distribution of total body lead to the bones is about 70 per cent, increasing to 90 per cent with age. Lead in blood is excreted and under particular physiological and pathological conditions, such as osteoporosis, chronic disease, pregnancy and lactation, lead can be released from the bones into the blood stream. This coupled with the sensitivity of young children places women of childbearing age at a high level of susceptibility. The measurement of bone lead is the most definitive indicator but is quite difficult to assess and interpret. Consequently the blood lead level (BLL) is used as the principal indicator of exposure. It is usually expressed in micrograms per decilitre of blood and can be measured from finger prick samples as a first order screen or from drawn blood samples. The impacts of high levels of lead exposure can be severe, with massive nervous system breakdown (encephalopathy) and kidney failure. Table 15.1 lists health effects at high levels of lead in blood. TABLE 15.1 Lead health effects for high level exposure (Boreiko, 2000; Wilson, 2000). Lethality Encephalopathy

Children: BLL > 125 µg/dL Adults: BLL > 300 - 400 µg/dL Children: BLL > 80 µg/dL Adults: BLL > 100 µg/dL

Colic Kidney toxicity

BLL > 60 - 100 µg/dL Prolonged exposure > 60 µg/dL

Anaemia

BLL > 50 µg/dL

Neurobehavioural performance Male reproduction Female reproduction

40 - 50 µg/dL Subtle changes above 40 µg/dL Miscarriage at high exposure Subtle effects possible at 15 - 30 µg/dL

The chief impacts of lead on human health at lower levels of exposure are neurotoxicity and some kidney effects. Nerve conduction is reversibly slowed in peripheral nerves at blood lead levels of 30 µg/dL, and significant effects can occur above 60 µg/dL. Kidney disease can be caused but requires prolonged exposure at relatively high levels. At BLL up to 60 - 80 µg/dL there can be biological changes in the kidney but they are largely reversible. Lead contained in the bones can affect skeletal growth in children and can affect vitamin D metabolism. Although there have been suggestions of causality, there is no definitive evidence of the carcinogenic potential of lead, and lead exposure is not considered to be of concern from this viewpoint. High levels of lead have induced tumours in animals and for this reason alone it has been classified as a ‘possible’ human carcinogen. However, most studies have found no relationship with cancer but those which have been inconclusive, have concurrent exposure to confounding substances such as arsenic and to smoking.

250




Studies into elevated blood pressure show no significant relationship or are inconclusive; however, suggestions have been made that a 1 mm of mercury increase in blood pressure could be associated with a doubling of the blood lead level. For an individual this is of little consequence in comparison with other effects. The effect on childhood development has been a particular concern to society and there have been many epidemiological studies relating to lead exposure effects on early childhood functions such as psychomotor, cognitive and behavioural problems. The effect on IQ development has been given as one to three IQ points decrease for a 10 µg/dL increase above 10 - 15 µg/dL. The general conclusions are that while low levels of lead exposure can cause such a small IQ deficit, there are many confounding factors with larger impacts which have not been adequately controlled. These are nutrition and key mineral deficiencies, general health history, parental IQ, parental attention and child rearing practices and general socio-economic status, which may in turn influence the above factors. There is also the effect of reverse causality, where children of lower IQ may have a greater tendency to a behaviour which increases exposure to lead (Pocock, Smith and Baghurst, 1994).

EXPOSURE PATHWAYS Exposure of the general public to lead is usually through eating or smoking in the presence of lead contamination, or by the inhalation of dusts and fumes in an occupational environment, particularly lead materials processing. Dietary intake in food and drinking water is relatively minor and has been estimated at around 20 µg per day for the average USA population. Of the other sources, hygiene is most important in limiting intake, and smoking will greatly increase risk. Occupational exposure involves the greatest risk and strict controls are necessary for those involved in the lead smelting industry.

Blood lead levels (BLL) As a result of various studies and the conclusion of the toxicity effect levels as indicated above, public health standards have been promulgated by most national health authorities, such as the Centers for Disease Control and Prevention (CDC) in the USA. These standards stipulate values above which BLL is regarded as elevated, or values above which medical intervention is recommended. BLL for the general population, based on the effects on children has been progressively lowered from 55 µg/dL pre-1970 to 10 µg/dL in 1991. Unfortunately these levels are often regarded as ‘poisoning limits’, whereas symptomatic lead poisoning is well above as indicated in Table 15.1. The CDC action guidelines for intervention are shown in Table 15.2. A number of surveys of children in US cities were conducted from 1976 to 1994 under the National Health and Nutrition Examination Survey (NHANES). Results are shown in Table 15.3 and indicate a dramatic drop in BLL as a result of various controls and abatement programs, but most particularly due to the removal of lead from automotive fuel.

OCCUPATIONAL STANDARDS AND CONTROLS An extensive body of legislation has been developed in most countries for the protection of workers involved in the processing of lead. Controls are primarily targeted at limiting the lead content of air to which workers are exposed. Cleanliness standards are also applied and the monitoring of all exposed workers is mandatory, involving regular blood lead determinations and record evaluation to determine trends.



251


TABLE 15.2 Centers for Disease Control and Prevention’s recommended actions for children. BLL µg/dL

Recommended action

<10

• Reassess or rescreen in one year. No additional action necessary unless exposure sources change.

10 - 14

• Provide family lead education. • Provide follow up testing. • Refer to social services, if necessary.

15 - 19

• • • •

Provide family lead education. Provide follow up testing. Refer to social services, if necessary. If BLLs persist (ie two venous BLLs in this range at least three months apart) or worsen, proceed according to actions for BLLs 20 - 44.

20 - 44

• • • •

Provide coordination of care (case management). Provide clinical management. Provide environmental investigation. Provide lead hazard control.

45 - 69

• Within 48 hours, begin coordination of care (case management), clinical management, environmental investigation and lead hazard control.

>70

• Hospitalise child and begin medical treatment immediately. Begin coordination of care (case management), clinical management, environmental investigation and lead hazard control immediately.

TABLE 15.3 Decline in blood lead level for children aged one to five years from 1976 to 1994. Mean BLL (µg/dL)

Prevalence of BLL >10 µg/dL

Prevalence of BLL >20 µg/dL

NHANES II (1976 - 1980)

15

88%

24.7%

NHANES III Phase 1 (1988 - 1991)

3.6

8.9%

1.1%

NHANES III Phase 2 (1991 - 1994)

2.7

4.4%

0.4%

Blood sampling may be capillary sampling from a finger prick or by directly drawing blood from a vein. The finger prick method may be a useful a first screen, but is prone to contamination from particles of lead on the skin, and thorough washing prior to sampling is necessary. Analysis requires the use of a certified laboratory and atomic absorption methods are usually employed. Maximum blood lead levels are usually stipulated above which a worker must be removed from further exposure. These levels are generally in the range of 50 to 70 µg/dL for men and 20 - 40 µg/dL for women. In many places women of childbearing age are not employed where lead exposure can occur. It is also common in the industry to have warning levels and also a requirement that a worker cannot return to the workplace where exposure has occurred until the BLL drops by at least 10 µg/dL below the removal limit. Table 15.4 shows BLL removal limits in a wide range of countries. BLLs are not strongly correlated with exposure such as lead in air, and depend to a large extent on the individual’s habits and hygiene, and particularly smoking history. Smoking in the workplace is commonly banned and smokers are often unsuitable for employment in high exposure situations due to their impaired ability to remove dust from the lungs. Conditions for employment often also require workers to be clean shaven, since beards will retain dust and interfere with the proper sealing of respirators.

252




TABLE 15.4 Maximum blood lead levels for occupational exposure for men (Wilson, 2000). Maximum lead level (µg/dL)

Country

80

India, Namibia, South Africa

70

Belgium, Denmark, European Economic Community (EEC), France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Spain, Thailand

60

Israel, Japan, Morocco, Peru, United Kingdom

50

Australia, Canada, Finland, Norway, Sweden, USA

Exposure controls Lead in air is the primary control required and maximum levels for routine occupational exposure without adverse health effects are normally in the range of 50 to 150 µg/m3, expressed as an eight hour time weighted average. In relation to lead smelting operations, exposure to associated metals can also be significant and relevant threshold limit values (TLVs) listed by the US Occupational Safety and Health Administration (OSHA) and the American Conferences of Governmental Industrial Hygienists (ACGIH) are given in Table 15.5. TABLE 15.5 Threshold limit values for occupational exposure. 3

Element Lead Antimony Arsenic Bismuth Cadmium Copper Selenium Tellurium Thallium Zinc

Maximum lead in air (µg/m ) 50 500 10 500 10 200 200 100 100 5000

Air sampling may be by static samplers or by personal samplers worn by individuals. The latter give a more accurate picture of likely levels of exposure to plant operators, but can be subjected to tampering. Samplers consist of a micropore filter capable of capturing particles of 0.2 µm, and a pump to draw air through the filter. For personal samplers the air rate is of the order of 2 L/min, whereas for static samplers the air rate is substantially greater at around 2 L/s. Emissions of lead dusts are controlled by ventilation at all point sources of dust and fume. Ventilation air is filtered in a bag house before discharge to atmosphere. In some cases two bag filters in series may be used to ensure no escape of dust with failure of the primary filter. Wet scrubbers may also be used where moist gases are to be processed or where acid mists are present. Because of the very fine nature of lead fumes, other collection devices such as cyclones or electrostatic precipitators are not suitably efficient and are not generally used for ventilation duties. In situations where exposure cannot be adequately controlled by ventilation, such as in equipment maintenance or clean-up operations, respirators must be worn. These should be tight fitting to provide



253


an adequate seal to the face and should be fitted with replaceable filter cartridges which are usually changed every day. An alternative is a full head and face shield, incorporated into a safety helmet with clear visor, supplied with a flow of fresh filtered air from a belt mounted pump and filter unit. This provides far more comfort, particularly in a hot environment, but is relatively expensive and needs battery charging between use. Protective clothing in the form of boiler suits, safety helmets, safety glasses, gloves and boots are normally supplied to all smelter workers and clothing is washed on a daily basis. Change rooms are designed in two parts as ‘clean’ and ‘dirty’ rooms with showering facilities between the two. Entry to the workplace is through the ‘dirty’ room and at the end of a shift workers will enter the ‘dirty’ room, remove work clothing and either store in a locker or place in a bin for washing. They will then pass through a shower to the ‘clean’ room fitted with lockers for normal clothing. Contaminated work clothing, towels, etc should be laundered on site where there is provision to handle contaminated wash water in a suitable effluent treatment facility. Eating should not be permitted in a lead work environment and it is usual practice to provide a canteen on the ‘clean’ side of the change facility, so that workers must remove work clothes and wash before eating. Within the workplace facilities need to be provided for washing such as by means of a foot-operated washing sink, together with suitable drinking water fountains, also preferably foot operated. Good housekeeping and workplace cleanliness is an essential part of controlling exposure. Design for ease of cleaning is an important aspect of any new facility, with the provision of smooth floors, avoidance of ledges and horizontal surfaces, which can collect dust, and the provision of vacuum cleaning facilities. Cleaning should include attention to regular road sweeping and washing down to prevent wind blown dust. Conveyor systems need to be fully enclosed and all transfer points need to be fully ventilated. Housekeeping measures of this nature will make a major difference to both workplace air and to emissions to the external environment.

EXTERNAL ENVIRONMENTAL CONTROLS Apart from the internal workplace environment, any lead processing facility must control its emissions of lead in air and water and the quality of the surrounding external environment. Air quality is usually monitored by the use of high volume samplers located at strategic points on, or outside the boundary of the processing plant, and such measurements are usually supervised by the regulating authority. Standards for ambient lead in air concentrations apply and are issued under ‘Clean Air’ regulations and range from 0.5 to 2.0 µg/m3. Some particular standards are shown in Table 15.6. TABLE 15.6 Ambient lead in air standards. Country

Lead in air (µg/m3)

Australia

0.5 averaged over 12 months

European Union

2.0 averaged over 24 hours

United Kingdom

0.5 averaged over 12 months 0.25 after December 2008

USA

1.5 averaged over three months

WHO recommendation

254

0.5 averaged over 12 months




Liquid effluents are similarly required to meet local standards and in this case due to the relative insolubility of lead it is often associated metals that require closest attention. Liquid effluents are commonly processed by lime neutralisation to precipitate metals, aiming for pH levels of the order of nine to ten. Typical effluent discharge standards are given in Table 15.7. TABLE 15.7 Typical discharge water standards. Element

Typical limit (mg/L)

Lead

0.2

Zinc

1.5

Copper

0.3

Cadmium

0.03

Manganese

0.5

Arsenic

0.25

Mercury

0.01

Selenium

0.02

Thallium

0.01

Chlorine

250

Fluorine

2.0

Ambient water quality maximum concentration of lead in drinking water has typically been 50 µg/L, but in the USA has been changed to an ‘action limit’ of 15 µg/L. Standards to protect aquatic life are given as 65 µg/L for freshwater systems and 210 µg/L for saltwater systems.

Effluent treatment There are many effluent treatment strategies depending on particular circumstances, but if it can be generalised, effluents may be grouped into two broad categories for separate handling approaches: Category 1: dilute streams which do not contain deleterious impurities and can be recycled for reuse in the smelting operation. Category 2: effluents containing impurity elements which represent outlets from the plant circuit and should not be returned. Category 1 effluents may be separately collected in ponds and can be used as process water after neutralisation of any acidity. The total proportion of effluents in this category should be maximised so as to minimise the requirements for effluent processing and generation of associated solid wastes. Category 2 effluents are predominantly weak acid bleed solution from smelter gas cleaning operations. This requires neutralisation and processing to remove all regulated elements – mainly lead, arsenic, antimony, zinc, cadmium, manganese, mercury, selenium and fluoride. It is generally the case that alkali metals, chloride and some sulfate can be tolerated in discharged effluent within set limits, and can provide an outlet for these elements from the smelting operation. The usual treatment procedure involves lime neutralisation to a pH of around ten often coupled with a sulfide precipitation polishing stage to scavenge base metals to low levels. The concentration of lead under these conditions is around 0.1 mg/L and is not a limiting factor. Lime neutralisation may be conducted in two stages; first to a pH of up to four to neutralise free acid and to produce a relatively clean gypsum for sale as a by-product, followed by a second high pH stage and sodium sulfide



255


addition to produce a contaminated gypsum. This approach significantly reduces the quantity of final effluent treatment solids for disposal. However, if the effluent treatment solids are recycled to the smelting operation the calcium from lime can be used as a flux to form slag, and the minimisation of final solids in this way is not necessary. In this situation the first stage neutralisation can use limestone rather than lime with a significant saving in reagent costs. A generalised flow sheet for a typical lime neutralisation effluent treatment procedure is given in Figure 15.1. Critical operating features of the lime neutralisation plant are the settling and filtration characteristics of the final solids, which require careful pH control and the use of flocculants. Combined Solution for Treatment Limestone

First Stage Neutralisation pH 4

Lime

Wash Water

Second Stage Neutralisation pH 10

NaHS

Filter

By-product Gypsum

Effluent Discard

Solids to Disposal

FIG 15.1 - Generalised effluent treatment – using lime neutralisation.

Disposal of the final effluent treatment solids follows many avenues, but commonly is recycled to the lead smelting operation, where the contained calcium can be used as a flux. Otherwise it can be placed in separate sealed ponds or sent to toxic landfill.

PRODUCT CONTROLS AND LIFE CYCLE MANAGEMENT In order to limit the input of lead into the environment many regulatory controls for the use of lead products have been introduced. Restrictions on the use of lead in paint and in automotive fuel are widespread. Other measures include bans on the use of lead shot for sporting ammunition and restrictions on the lead content of waste materials. Most of these regulatory restrictions are specific to individual countries, but one international treaty under the administration of the United Nations Environmental Program is the Basel Convention to control the cross border transfer of industrial

256




waste. This treaty prohibits the transfer of waste from developed countries to undeveloped countries with poor or non-existent environmental regulations. It has promoted a high degree of recycling within developed countries, and in particular the recycling of lead acid batteries. Another initiative to broaden the scope of lead acid battery recycling in developing countries and to promote a sustainable life cycle on a global basis is the ‘Green Lead Initiative’. Member corporations or organisations within ‘Green LeadTM’ will be appraised and certified as complying with its aims and must only trade with other accredited members. Thus mines will sell to accredited smelters who in turn will sell to accredited battery manufacturers who sell only to accredited retailers. The retailers must have used battery collection facilities in place and supply accredited secondary smelting operations. Consumers will be encouraged to purchase batteries only from certified battery retailers and to return used batteries. These types of controls and voluntary initiatives are important in minimising the impact of lead on the natural environment and in countering extreme calls for bans on the use of lead.

REFERENCES Boreiko, C J, 2000. Lead and zinc: A study of technological contrasts and shared regulatory concerns, in Proceedings Lead-Zinc 2000 Symposium, pp 39-52 (The Minerals, Metals and Materials Society (TMS): Warrendale). Goyer, R A, 1996. Casarett and Doull’s Toxic Effects of Metals in Toxicology – The Basic Science of Poisons, fifth edition (ed: C D Klaassen) (McGraw Hill: New York). Piomelli, S, Rosen, J F, Chisolm, J J and Graef, J W, 1984. Management of childhood lead poisoning, J of Pediatrics, 105:523. Pocock, S J, Smith, M and Baghurst, P, 1994. Environmental lead and children’s intelligence: A systematic review of epidemiological evidence, British Medical Journal, 309:1189-1196. Wilson, D N, 2000. Health and hygiene in the modern lead and zinc industry, in Proceedings Lead-Zinc 2000 Symposium, pp 289-306 (The Minerals, Metals and Materials Society (TMS): Warrendale).



257

CHAPTER 16 Energy Consumption PURPOSE AND SCOPE The consumption of energy for lead production has not been anywhere near as high as for other metals such as zinc or aluminium, and hence technology change has not focused on this aspect to the same extent. Focus for technology change has been more on the elimination of costly fuels such as coke and replacement with coal, and the confinement of the process to reduce environmental emissions. Nevertheless energy use for the smelting and production of metals in general is an issue of growing significance as the environmental consequences of energy production and use receive closer attention and scrutiny. The purpose of this chapter is to quantify, in a general way, the total energy use for the production of lead by the major process options. The production of lead metal can be divided into two broad areas – the basic smelting operation to produce a crude lead bullion, and the refining operation. Since the sinter plant–blast furnace method of smelting coupled with thermal refining has been the workhorse of the industry; they are examined in greater detail herein as a benchmark for other processes. Smelting processes are assumed to terminate at the production of crude bullion, excluding the copper drossing operation which has been included in the refining operations. In the evaluation of energy inputs to the various smelting processes the following heat values have been used for individual fuels:

• coal

30.7 GJ/t

• coke

28.5 GJ/t

• natural gas

37 MJ/Nm3

• electric power

10.91 MJ/kWh heat energy equivalent The electric power heat energy equivalent is the typical fuel heat energy required to generate 1 kWh of electrical energy.

ENERGY CONSUMPTION FOR THE SINTER PLANT–BLAST FURNACE The following provides generalised quantification of the various components of the energy balance for the sinter plant–blast furnace process and draws on generalised mass balance data shown in Table 16.1 on the basis of one tonne of product lead. TABLE 16.1 Mass balance information for the sinter plant–blast furnace. Component Concentrate feed Fluxes added Sinter produced Sulfur burned Sulfuric acid produced Coke used in blast furnace Slag produced


Quantity tonnes per tonne of product lead 1.7 0.7 2.3 0.3 0.8 0.22 1.1


259

CHAPTER 16 – Energy Consumption

Concentrate handling The handling of incoming concentrates, their storage, blending with flux materials and feeding to the sinter plant requires the use of mobile equipment, feed bins and conveyor systems. Electric power used is around 5 kWh/t of materials handled or 12 kWh/t of product lead. Fuels used for mobile equipment represents 75 MJ/t of material handled or 180 MJ/t of product lead.

Sinter plant Fuel is supplied as natural gas to the ignition stove of the sinter plant at around 300 MJ/t of product lead. Coke breeze is often added to the charge particularly when residues are processed and may represent around 8 kg/t of sinter charge or 18 kg/t of product lead providing a heat value input of 510 MJ/t of product lead. Heat generated by the roasting reactions is lost from the hood of the machine and in sinter gases from which heat is not recovered as steam. Otherwise energy input is in the form of electric power to a wide range of fans to supply combustion air and handle sinter plant gases, as well as miscellaneous pumps, conveyors, the drive motors for the sinter strand and product sinter crushers, and feed conditioning equipment. Total electrical power input is related to the sulfur burned, which defines the gas flow and size of the sinter hearth area and is 300 kWh/t of sulfur burned or 90 kWh/t of product lead.

Gas cleaning and sulfuric acid production It is assumed that a conventional wet gas scrubbing system and double absorption acid plant are included. Power is used for fans and blowers as well as acid circulation pumps and cooling towers. Typical electric power consumption for plants based on six to seven per cent SO2 in acid plant feed gas, is 120 kWh/t of acid produced. In most lead sinter plants the gas is lower in SO2 at around five per cent, which will raise power consumption to 150 kWh/t of acid produced, which equates to 120 kWh/t of product lead. Fuel for preheating and start-up is more frequent with sinter gas feed and is taken as 25 Nm3 of gas per tonne of acid produced, equivalent to 930 MJ/t of acid or 750 MJ/t of product lead.

Blast furnace The primary energy input is from the combustion of coke at 0.22 t/t of lead, representing an input of 6270 MJ/t of product lead. Electric power is used for blast air blowers and gas handling fans as well as fans for ventilation air. Total power use is estimated at 80 kWh/t of bullion, and natural gas use for miscellaneous heating uses 5 Nm3/t of lead, equivalent to 185 MJ/t of product lead. Most furnaces operate with some oxygen enrichment and a nominal figure of 20 kg/t of lead has been assumed equivalent to a power input of 10 kWh/t of product lead. No heat is recovered from blast furnace operations.

Summary Table 16.2 summarises the energy inputs to the smelting stage and production of crude lead bullion.

260




TABLE 16.2 Summary of energy inputs for the sinter plant–blast furnace. Plant section

Electrical energy (kWh/t)

Electrical fuel equivalent (MJ/t)

Concentrate handling

12

130

Sinter plant

90

980

Gas cleaning/acid

120

1310

Blast furnace

90

980

Ventilation

40

450

Total

312

3850

Coke or coal (MJ/t)

Fuel (MJ/t)

Total fuel equivalent (MJ/t)

75

205

510

300

1790

750

2060

6270

185

7435

6780

1310

11 940

450

THERMAL REFINING OF LEAD BULLION Refining operations are assumed to cover all crude bullion impurity separation procedures including the first copper drossing operation, which is often regarded as part of primary smelting as distinct from refining operations. All these operations are normally carried out in standard refining kettles or kettles modified for the specific operation. Energy input primarily involves electrical energy input for agitation, lead pumping and ventilation duties, plus heating of the kettles by natural gas to maintain temperature or to run through temperature cycles as required for the particular separation. The natural gas used in each case depends on the temperature cycles used and the batch time required for each operation. Estimates for each of the refining operations are given in Table 16.3. TABLE 16.3 Energy inputs for thermal refining. Plant section



Coke or coal (MJ/t)

Fuels (MJ/t)


4.5

50

150

1500

1700

750

780

40

180

Copper drossing Softening

3

30

2.5

25

Softener slag treatment

2

20

Desilverising

14

150

690

840

Vacuum dezincing

6

65

310

375

Debismuthising

14

150

1450

1600

3

35

250

285

100

155

Oxygen for softening

25 120

Zinc for desilvering

420

Magnesium/calcium Caustic refining

900

Casting and despatch

5

55

Ventilation system

24

250

Total

78

830

250 270

5090

7510

ELECTROLYTIC LEAD REFINING As an alternative to thermal refining, lead may be refined by the Betts electrolytic process. However, this still requires a number of steps of thermal refining, such as copper drossing, softening and final caustic refining.



261


As detailed in Chapter 13 the electrical energy required for electrolysis is around 150 kWh/t of product lead. Some additional power for solution handling and fuels is required for the anode casting operation and for melting cathode lead in preparation for casting and starter sheet preparation. Estimates are provided in Table 16.4. TABLE 16.4 Energy inputs for electrolytic refining. Plant section



Coke or coal (MJ/t)

Fuels (MJ/t)


4.5

50

150

1500

1700

750

780

40

180

Copper drossing Softening Oxygen for softening Softener slag treatment Anode casting

3

30

2.5

25

2

20

25 120

2

10

152

1660

Cathode melting

1

10

150

160

Caustic refining

2

20

150

170


5

55

100

155

Ventilation system

16

175

Total

190

2055

2770

5095

Electrolytic refining

80

90 1660

175 270

The energy input for electrolytic refining is significantly less than for thermal refining and the difference is close to the energy input for bismuth separation. If in a particular situation bismuth separation is not required, there would no energy advantage for conventional electrolytic refining.

DIRECT SMELTING PROCESSES Direct smelting processes provide alternatives to the sinter plant–blast furnace above, up to the stage of crude bullion production, but still require the addition of refining operations to produce high-grade lead. Three alternatives have been evaluated – the Kivcet process, the QSL process and the Isasmelt process – as typical of the top submerged lance slag bath reactor. Evaluations are based on comparable feeds, predominantly lead concentrates with low residue inputs. The extensive use of oxygen is common in these cases and associated electrical energy input for its production has been taken as 500 kWh per tonne of oxygen. The recovery of heat is also possible with these processes in contrast to the sinter plant and blast furnace. Recovered energy is taken as the heat content of recovered steam, taken as 2.65 MJ/kg of steam. Smelter gases are relatively rich in SO2, significantly reducing the gas cleaning and acid production requirements in comparison with a sinter plant.

The Kivcet process Operating parameters for the Kivcet process have been assumed as follows per tonne of product lead:

• oxygen requirements

365 Nm3 (0.52 tonnes)

• coal input

100 kg

• electrode carbon use

2.3 kg

262




• electrode power

300 kWh

• other power

125 kWh

• natural gas general use

20 Nm3

• natural gas for feed drying

50 Nm3

• steam recovered

1.4 tonnes

Estimates of energy inputs are given in Table 16.5. TABLE 16.5 Summary of energy inputs for the Kivcet process. Plant section Concentrate handling Concentrate drying Flash furnace Oxygen use Gas cleaning/acid Ventilation Total Steam recovered Net total

Electrical energy (kWh/t) 12 2 425 260 70 15 784

Electrical fuel equivalent (MJ/t) 130 20 4740 2845 760 165 8660

Coke or coal (MJ/t)

3135

Fuels (MJ/t) 75 1850 740 250

3135

2915

Total fuel equivalent (MJ/t) 205 1870 8615 2845 1010 165 14 710 -3710 11 000

On this basis the Kivcet process has very similar net energy consumption to the sinter plant–blast furnace. This may be surprising for an intensive process, but results from the high use of electric power for both the electric furnace and for oxygen production and the low efficiency of conversion of heat energy to electrical energy in comparison with its direct use.

The Queneau-Schuhmann-Lurgi (QSL) process Operating parameters for the QSL process have been assumed as follows per tonne of product lead:


450 Nm3 (0.64 tonnes)

• coal input

175 kg

• electric power

280 kWh

• natural gas general use

20 Nm3

• steam recovered

1.0 tonnes

Estimates of energy inputs are given in Table 16.6.

The Isasmelt process It is assumed that the Isasmelt process operates as two stages with two reactors as discussed in Chapter 7. The first is an oxidation stage and the second reduces first stage slag. Process oxygen requirements are provided as enriched air at 35 per cent rather than pure oxygen in the above alternatives. This will reduce the energy required for oxygen production but will increase the energy required for air supply and gas handling.



263


TABLE 16.6 Summary of energy inputs for the Queneau-Schuhmann-Lurgi (QSL) process. Plant section




12

130

QSL furnace

280

3055

Oxygen use

320

3490

Gas cleaning/acid

70

760

Ventilation

20

220

Total

702

7655

Coke or coal (MJ/t)

Fuels (MJ/t)


75

205

5370

740

9165

250

1010

1065

14 090

3490 220 5370

Steam recovered

-2650

Net total

11 440

Operating parameters for the Isasmelt process have been assumed as follows per tonne of product lead:


285 Nm3 (0.41 tonnes) to Stage 1

• coke fines to Stage 1

6 kg

• coal input to Stage 2

150 kg

• natural gas to Stage 1

3.5 Nm3

• natural gas to Stage 2

25 Nm3

• electric power

320 kWh (200 to Stage 1, 120 to Stage 2)

• steam recovered

1.0 tonne

Estimates of energy inputs are given in Table 16.7. TABLE 16.7 Summary of energy inputs for the Isasmelt process. Plant section




12

130

Stage 1 furnace

200

2200

Oxygen use

205

2240

Stage 2 furnace

120

1310

Gas cleaning/acid

90

1000

Ventilation

30

330

Total

657

7210

Coke or coal (MJ/t)

Fuels (MJ/t)


75

205

170

140

2510

4605

925

6840

450

1450

1590

13 575

2240

330 4775

Steam recovered

-2850

Net total

10 725

ELECTROCHEMICAL LEAD EXTRACTION PROCESSES A number of processes have been reviewed in Chapter 9, but none of these have yet been commercialised and no actual data is available on energy consumption. These processes produce relatively high purity lead and do not require a full refining procedure other than a final caustic treatment of melted cathodes to remove final traces of some impurities such as arsenic and antimony.

264




The following provides theoretical estimates for direct comparison with the other smelting processes plus refining reviewed above. The basis used for this estimate is a full hydrometallurgical system using ferric iron to effect the oxidation of PbS to produce soluble lead and elemental sulfur, followed by solution purification with lead powder and electrolysis of the purified solution in a diaphragm electrolytic cell to deposit lead on the cathode and reoxidise iron to ferric iron at the anode for recycle to the leaching stage. Possible systems using this approach are fluoborates and fluosilicates as well as chlorides, and are possibly the most likely to be developed in the future. There are differences in the conductivity of the different systems with significant effect on cell voltages and hence electrolytic energy use. Assumed cell voltages for electrowinning lead from the different systems and the corresponding energy use for 95 per cent current efficiency in all cases have been taken as follows:

• chloride system

3.5 volts and 950 kWh/t of cathode lead

• fluosilicate system


• fluoborate system


There are thus distinct advantages in the use of fluoborates and this has been used as the basis for energy input estimates. These systems will recover relatively pure cathode lead and valuable by-products such as silver and copper must be recovered from purification cements. An allowance is made for the pyrometallurgical processing of these materials to achieve separate by-product recoveries and is taken as the energy required for normal thermal refining for the lead used but excluding debismuthising (4590 MJ/t for five per cent of cathode lead or 230 MJ/t of product lead plus the full quantity of zinc for desilvering at 420 MJ/t of product lead). Assumed operating parameters per tonne of product lead are as follows:

• electrical energy for leaching and solution handling

180 kWh/t

• electrical energy for electrowinning

600 kWh/t

• lead powder for purification as five per cent

30 kWh/t

Estimates of energy inputs are given in Table 16.8. TABLE 16.8 Summary of energy inputs for electrowinning. Plant section



Coke or coal (MJ/t)

Fuel (MJ/t) 75



12

130

Leaching/solution handling

180

1960

1960

Electrowinning

600

6550

6550

Purification Pb powder

30

325

325

By-product recovery

3

31

14

180

Zinc for desilvering

205

225 420

Cathode melting

1

10

150

Caustic refining

2

20

150

170


5

55

100

155

655

10 390

Ventilation

20

220

Total

853

9301



160

220 14

265


For a process based on the chloride system rather than fluoborates, electrolytic power will increase from 600 to 950 kWh/t and the total energy input will increase from 10 390 MJ/t of lead to 14 400 MJ/t.

COMPARISON OF EXTRACTION PROCESSES Table 16.9 compares the total energy inputs required for production of refined lead from sulfide concentrates. TABLE 16.9 Comparison of smelting and refining processes’ energy inputs. Smelting process

Refining process

Energy input smelting (MJ/t of Pb)

Energy recovery (MJ/t of Pb)

Energy input refining (MJ/t of Pb)

Total net energy input (MJ/t of Pb)

Sinter – BF

Thermal

11 940

7510

19 450

Sinter – BF

Electrolytic

11 940

5095

17 035

Kivcet

Thermal

14 710

-3710

7510

18 510

Kivcet

Electrolytic

14 710

-3710

5095

16 095

QSL

Thermal

14 090

-2650

7510

18 950

QSL

Electrolytic

14 090

-2650

5095

16 535

Isasmelt

Thermal

13 575

-2850

7510

18 235

Isasmelt

Electrolytic

13 575

-2850

5095

15 820

Leach – electrowinning

(Fluoborate)

10 390

10 390

Leach – electrowinning

(Chloride)

14 400

14 400

Table 16.9 suggests that there is little difference in the overall net energy input for most of the alternative smelting processes, but there is potentially a significant saving with the use of leaching electrowinning techniques. Electrolytic refining offers the lowest energy use for full refining, but as noted above if bismuth removal is not required there is little difference between the two approaches. In general the energy input required for smelting sulfide concentrates to crude lead bullion is of the order of 11 to 12 GJ per tonne of lead, and the energy required to produce refined lead is 17 to 19 GJ per tonne. The adoption of direct smelting methods is not dictated by energy savings but by other issues such as the reduction in the vast volumes of gas to be handled by the sinter–blast furnace process and the consequent reduction in plant capital cost. Elimination of the use of coke as a high cost reductant can be a key consideration. Environmental containment and occupational health issues are also a major factor in adoption of direct smelting methods.

ENERGY CONSUMPTION IN SUPPLY OF LEAD CONCENTRATES In order to determine the total energy requirements for the production of lead metal it is necessary to include the energy inputs for mining and mineral separation operations. Lead is most commonly now mined in combination with other metals, particularly zinc and sometimes copper. In this situation the energy consumed in mining and smelting operations can be assigned to the total product metals recovered and is typically as follows:

• mining operations

5000 MJ/t of metal recovered

• ore crushing


266




• grinding and flotation


• total mine site energy input


Transport of concentrate from the mine site to the smelter is highly variable depending both on distances and the mode of transport, such as road, rail or sea. A typical figure for transport may be of the order of 400 MJ/t of concentrate or 700 MJ/t of contained lead. In this case the energy consumed in supply of lead in concentrate form to the smelter is 10 100 MJ/t of lead. Adding the energy required for concentrate supply to the energy inputs for smelting gives the total energy consumption for the production of primary lead by the standard blast furnace–thermal refining approach as close to 30 000 MJ/t of refined lead.

ENERGY CONSUMPTION FOR SECONDARY LEAD PRODUCTION As detailed in Chapter 11 secondary lead smelting is basically concerned with the processing of scrap lead–acid batteries, with metallic scrap representing a minor part. There are many processes in use, but the following analysis of energy consumption examines the most common approach involving battery breaking and separation, followed by separate processing of metallics and battery pastes in a short rotary furnace. The assumed performance parameters have been taken as follows:

• battery waste lead content

60 per cent

• metallic lead portion of total lead

40 per cent

• paste lead proportion of total lead

60 per cent

• battery breaking: • power consumption • fuel use for paste desulfurisation

45 kWh/t of scrap 270 MJ/t of scrap

• melting metallic lead components: • power consumption • fuel consumption

60 kWh/t of lead 610 MJ/t of lead

• smelting battery pastes and reduction to bullion: • power consumption • coal addition • fuel consumption

200 kWh/t of lead 80 kg/t of lead 2500 MJ/t of lead

In addition the product lead may require some refining, particularly to remove antimony and tin, and it is assumed that caustic refining will be sufficient for this purpose. Table 16.10 summarises the energy inputs required for the production of high-grade secondary lead. On the above basis the energy required for the recovery of secondary lead is of the order of 35 per cent of the energy required for primary lead production from sulfide concentrates. If the energy required for mining, mineral separation and supply of lead concentrates to the primary smelters is considered, the total energy for secondary lead recovery is less than 25 per cent of the total for primary lead.



267


TABLE 16.10 Summary of energy inputs for secondary lead processing (data per tonne of product lead). Plant section



Coke or coal (MJ/t)

Fuels (MJ/t)


Material handling

5

55

80

135

Battery breaking

75

820

450

1270

Metallics melting

24

260

Paste smelting

120

1310

2

Caustic refining

245

505

1500

4285

20

150

170

100

155

2525

6740


5

55

Ventilation

20

220

Total

251

2740

268

1475

220 1475



CHAPTER 17 Costs and Economics of Lead Production PURPOSE AND BASIS The objective of this chapter is to provide a general outline of the costs associated with the production of lead by different processes and to give an indication of the important factors affecting economic performance. The basis for comparison of different processes is the sinter plant–blast furnace, followed by thermal refining as the principal method of production of primary lead. No new smelters using this technology have been constructed for many decades. Consequently there is a lack of data on actual construction costs and the figures provided herein must be regarded as broad estimates at best. Hopefully the following chapter provides a comparative framework and cost analysis, which can be updated from time to time. The framework is aimed at identifying the major cost elements that influence the conduct of the industry, as well as giving an indication of the relative importance of each cost element. Information is drawn from personal sources within the industry and its suppliers, and from a range of published cost data. This data covers a long period of time and it has been necessary to update costs to the present time frame. All data has been expressed in US dollars as at June 2005, and may be adjusted by CPI rates from that date.

SMELTING BY THE SINTER PLANT–BLAST FURNACE The design of a lead smelter depends to a significant extent on the nature of the feed materials processed, particularly the grade of the concentrates. In simple terms this is due to the large possible variation in the sulfur to lead ratio in feed materials and hence the size of the sinter plant required, which is dictated by the sulfur burning capacity. Usually lead sinter has a relatively common lead composition at around 45 per cent and hence the lead blast furnace sizing is not so critically dependent on the nature of the feed. For this reason the cost estimates provided are based on a standard concentrate feed of 60 per cent Pb and 20 per cent S content. Capacity is standardised at 100 000 tonnes per annum (t/a) lead production, representing the median capacity smelter. By-products and impurity elements are also highly significant to the economics of lead smelting operations, but in terms of costs, are associated more with the refining operation than the smelting operation. The assumptions used for construction of cost estimates and economic parameters are shown in Table 17.1. For a sinter plant producing 100 000 t/a of lead on the above assumed inputs, and with an operating time of 320 days per year, derived performance data is shown in Table 17.2.

Capital costs The construction costs for a new lead smelter have been developed for construction under USA or equivalent conditions. Construction costs in other localities will depend on the cost of construction labour, the use of local construction materials, and the extent of inclusion of locally manufactured



269

CHAPTER 17 – Costs and Economics of Lead Production

TABLE 17.1 Assumptions used for lead smelter estimates. Parameter

Value


100 000

Concentrate grade Pb

60%

S

20%

Fe

6%

CaO

0.5%

SiO2

3%

Ag

1000 g/t

Au

3 g/t

Cu

0.5%

Sulfur recovery to acid

85%

SO2 content of sinter gas

5%

Pb content of sinter

45%

FeO content of slag

30%

CaO content of slag

16%

SiO2 content of slag

22%

Pb content of slag

1.5%

TABLE 17.2 Derived performance data for the sinter plant–blast furnace (100 000 t/a of lead). Parameter

Derived value

Lead recovery from concentrate

98.3%

Concentrate required (t/a)

169 500

Sinter gas volume Nm3/h

52 500

Acid produced (t/a)

86 500

Acid produced (t/d)

270

Fluxes – total input (t/a)

93 000

Net sinter output (t/a)

222 000

Slag produced (t/a)

115 000

Coke burned in BF (t/a)

20 000

equipment such as drives, filters, electrical supply equipment, etc. It needs to be recognised that for construction in lesser-developed countries, the cost per worker may be substantially lower, but due to the lower labour productivity the overall labour cost component may not be greatly reduced. Construction costs for remote locations can also be elevated due to the need to locate labour for the construction period. This may be minimised by off-site prefabrication of as much equipment as possible. The assumed configuration of the sample plant is illustrated in Figure 17.1 and covers current practice for a plant treating average sulfide concentrates by sintering and processing of sinter in a blast furnace to produce crude lead bullion. Refining of the crude lead bullion is considered in a separate section.

270




Lead Concentrates

Fluxes

Raw Materials Storage

Feed Blending

Ventilation

Sinter Plant

Gas Cleaning

Blast Furnace

Gas Cooling & Cleaning

Acid Production

Coke

Slag Granulation

Bullion Handling

Slag to Disposal

Bullion to Refinery

Sulfuric Acid To Sale

FIG 17.1 - Generalised flow sheet and plant subdivision.

The broad subdivision of costs as illustrated in Figure 17.1 may be defined as follows: Raw materials handling and feed blending

Concentrate and flux materials receival, storage and controlled blending facilities to provide sinter feed.

Sinter plant

Conventional up-draught sinter plant including feed mixing with return sinter, feed conditioning and moisture control, sinter strand, product sinter crushing and screening, sinter storage.

Gas cleaning

Dust removal by drop-out chambers and electrostatic precipitators, cooling and wet scrubbing, and wet gas mist removal by electrostatic precipitators.

Acid plant

Conventional double absorption acid plant based on a five per cent SO2 feed gas.

Blast furnace

Covers furnace charging system with sinter and coke, blast air supply, shaft furnace, furnace cooling facilities, tapping facilities and forehearth. Also includes blast furnace gas cooling and dust removal system (bag-house).

Ventilation

Total system for collection of dusts from all areas of the smelter and associated dust removal equipment.

Slag disposal

Includes slag granulation, separation and disposal facilities.

Bullion handling

Handling equipment such as cranes, transfer cars, ladles, etc for transfer of crude bullion from the smelter to the refinery.

Effluent treatment

Covers a lime neutralisation plant for processing weak acid effluent from gas cleaning and general effluents that cannot be recycled as process water.



271


Utilities and services

Includes allowances to cover the connection of power and water, drainage works, maintenance facilities, laboratories and offices. This is estimated at 12 per cent of direct plant costs for the above items.

Site works

Site earthworks, roads, access, fencing and drainage. This is estimated at six per cent of direct plant costs.

Costs given for each section are fully constructed costs but exclude project engineering and supervision, which is estimated separately as 15 per cent of the project cost. An allowance for licence fees for propriety equipment and processes has been included at two per cent of the project cost. Capital costs for the sinter plant–blast furnace smelter are given in Table 17.3 for a smelter with an output of 100 000 t/a of lead, and corresponding with the parameters outlined in Tables 17.1 and 17.2. TABLE 17.3 Lead smelter capital costs – sinter plant–blast furnace (cost basis – June 2005). Plant section

US$ million

US$/t of lead pa

Raw material handling

14

140

Sinter plant

38

380

Gas cleaning

21

210

Acid plant

11

110

Blast furnace

55

550

Ventilation

22

220

Slag disposal

8

80

Bullion handling

8

80

Effluent treatment

5

50

Sub total

180

1800

Utilities and services at 12%

22

220

Site works at 6%

11

110

213

2130

Land and access

Sub total – direct plant cost

12

120

Licence fees

4

40

Engineering at 15% of direct costs

32

320

Sub total – indirect costs

48

480

261

2610

Total plant capital cost

Clearly circumstances and location will significantly change these figures, but Table 17.3 should provide a guide to the major items to be considered and the general order of cost for new facilities under USA conditions.

Direct operating costs Direct operating costs may be broadly subdivided into labour (for operating and maintenance), maintenance materials, electric power, reduction carbon (coke), other fuels, and operating materials and supplies. Typical costs for a standard sinter plant–blast furnace smelter as at June 2005 are given in Table 17.4 for a plant of 100 000 t/a of lead annual capacity. The subdivision of plant sections corresponds with capital cost items given in Table 17.3.

272




TABLE 17.4 Direct operating costs for a 100 000 t/a lead smelter. Sinter plant–blast furnace (costs in US$ as at June 2000). Plant section

Number employed

Man hours Labour cost per t $000’s


16

27.2

880

Power kWh/t

Electric power $000’s

Maint materials $000’s

Materials and supplies $000’s

12

72

336

1517

Sinter plant

45

76.5

2475

90

540

921

338

Gas cleaning

10

17

550

45

270

523

75

Acid plant

14

23.8

770

75

450

261

105

Coke $000’s

286

Other fuels $000’s

Total $000’s

Total $ t Pb

30

2836

28.4

120 300

1886

18.9

85

2750

90

540

1529

750

9163

91.6

12

20.4

660

40

240

548

270

1718

17.2

Slag disposal

8

13.6

440

5

30

209

60

739

7.4

Bullion handling

8

13.6

440

2

12

200

60

712

7.1

Effluent treatment

10

17

550

3

18

114

225

907

9.1

Administration

35

58.8

1903

6.5

39

0

1730

3672

36.7

Totals

208

353

11 418

2211

4640

5131

3806

524

27 730

277.3

114.2

22.1

46.4

51.3

38.1

5.2

277.3

1.

Power is costed at 6 US cents per kWh.

2.

Coke is costed at US$160 per tonne with a thermal value of 28.5 GJ/t.

3.

Natural gas fuel is costed at US$4 per GJ.

273



50

Ventilation

Notes:

74

46.8 14.2

Blast furnace

Total per tonne of lead

3520

4680 1418


Labour costs are estimated from the number of employees required in each section covering both operating and maintenance, and an average annual employment cost of US$55 000 per employee. Employment cost should be the total including direct wages and all associated expenses such as taxes, pension scheme costs, insurances and benefits. Maintenance materials are simply estimated as 2.5 per cent of the direct construction capital cost for each section. This cost will be lower for a new plant but should rise to the 2.5 per cent level for a well-established plant. Power is estimated from typical consumption figures for each area and a power cost of six cents per kilowatt hour in this example. Similarly coke and other fuels are estimated on the basis of typical energy consumption data provided in Chapter 16. Materials and supplies are again typical in dollar terms and represent a range of reagents and consumables as well as external services provided to each section. Oxygen is included in the cost of materials and supplies for the blast furnace at around $1.50/t of lead bullion for marginal blast air enrichment by two per cent. The cost of administration reflects higher salaried support and management staff, as well as services such as insurance and site taxes. These costs will vary widely depending on the location of the plant and whether it is a stand-alone operation or part of an industrial complex. As for the capital cost data, these figures are designed to provide a guide to the major operating cost items to be considered in any smelter evaluation. Total smelting costs are close to US$277/t of lead produced from the smelting operation. Of this cost 41 per cent represents labour costs, 24 per cent is for energy and 35 per cent is for materials (including maintenance materials at 16 per cent).

SMELTING BY THE KIVCET PROCESS The Kivcet process as a means of primary lead smelting replaces the sinter plant and blast furnace with a single unit, with significantly reduced requirements for labour and for environmental containment. Oxygen is used in place of air with greatly reduced gas volumes, significantly reducing the cost of gas handling equipment, but at the expense of added energy for oxygen production. Lead levels in slag are higher in this case at 3.5 per cent rather than 1.5 per cent assumed for the blast furnace. This will lower overall lead recovery unless slag fuming facilities are included. Cost estimates are provided on the same general basis of concentrate feed composition as used for the sinter plant–blast furnace estimates above and using energy data given in Chapter 16.

Capital costs Table 17.5 gives the capital cost estimate for the production of 100 000 t/a of lead in bullion by the Kivcet process. Total capital costs for the Kivcet plant at US$2015 per annual tonne of lead in bullion is lower by 23 per cent than the comparable capital for a sinter plant and blast furnace at US$2610 per annual tonne.

Direct operating costs Direct operating costs for a Kivcet smelter as at June 2005 are given in Table 17.6 for a plant of 100 000 t/a of lead annual capacity. The subdivision of plant sections corresponds with capital cost items given in Table 17.5.

274




TABLE 17.5 Lead smelter capital costs – Kivcet process (cost basis – June 2005). Plant section

US$ million

US$/t of lead pa


14

140

Kivcet plant

77

773

Gas cleaning

8

83

Acid plant

11

108

Ventilation

8

77

Slag disposal

8

83

Bullion handling

8

80

Effluent treatment

4

36

Sub total

138

1385

16

160

Utilities and services at 12% Site works at 6%

8

80

162

1625

Land and access

12

120

Licence fees

3

30


24

240

39

390

201

2015


Sub total – indirect costs Total plant capital cost

Labour costs are estimated from the number of employees required in each section covering both operating and maintenance, and an average annual employment cost of US$55000 per employee. Employment cost should be the total including direct wages and all associated expenses such as taxes, pension scheme costs, insurances and benefits. Maintenance materials are simply estimated as 2.5 per cent of the direct construction capital cost for each Section. This cost will be lower for a new plant but should rise to the 2.5 per cent level for a well established plant. Coal is used in the Kivcet furnace at a cost of US$80/t in place of coke used for the blast furnace at US$160/t. Power is estimated from typical consumption figures for each area and a power cost of six cents per kilowatt hour in this example. Similarly coal and other fuels are estimated on the basis of typical energy consumption data provided in Chapter 16. Gas fuels are higher in this case due to its use for Kivcet furnace feed drying. In this case waste heat steam can be used to generate power and give a credit of 337 kWh per tonne of lead. Materials and supplies are estimates of a range of reagents and consumables as well as external services provided to each section. The cost of administration reflects higher salaried support and management staff, as well as services such as insurance and site taxes. These costs will vary widely depending on the location of the plant and whether it is a stand-alone operation or part of an industrial complex. As for the capital cost data, these figures are designed to provide a guide to the major operating cost items to be considered in any smelter evaluation. Total smelting costs are close to US$195/t of lead produced from the smelting operation. Of this cost 38 per cent represents labour costs, 24 per cent is for energy and 38 per cent is for materials (including maintenance materials at 18 per cent). Without waste heat energy recovery total costs would rise to $211/t of lead with energy representing 32 per cent.



275

Plant section

Number employed


Power kWh/t





16

27.2

880

12

72

341

1509

Kivcet plant

36

Oxygen supply

2

61.2

1980

425

2550

1932

540

3.4

110

260

1560

Gas cleaning Acid plant

10

17

550

11

90

207

75

14

23.8

770

55

330

269

105

Ventilation

5

8.5

275

15

90

193

Slag disposal

8

13.6

440

5

30

Bullion handling

8

13.6

440

2

12

Effluent treatment

10

17

550

2

12

Administration

22

37.1

1199

4.2

Waste heat credit

5

8.5

275

-337

136

231

7469 74.7

Totals


Total per tonne of lead Notes:


1.


2.

Coal is costed at US$80 per tonne with a thermal value of 30.7 GJ/t.

3.


Coal $000’s

800


Total $000’s

Total $ t Pb

30

2831

28.3

1036

8837

88.4

1670

16.7

922

9.2

1574

15.7

113

671

6.7

209

60

739

7.4

200

60

712

7.1

90

225

877

8.8

26

0

1090

2315

23.1

-2022

70

38

-1640

-16.4

2750

3510

3814

800

1166

19 508

195.1

27.5

35.1

38.1

8.0

11.7

195.1

100


276

TABLE 17.6 Direct operating costs for a 100 000 t/a Kivcet lead smelter (costs in US$ as at June 2000).


SMELTING BY THE ISASMELT PROCESS The Isasmelt process as a means of primary lead smelting replaces the sinter plant and blast furnace with a two top submerged lance reactors. Oxygen enriched air is used, reducing gas volumes and the cost of gas handling equipment, but at the expense of added energy for oxygen production. Lead levels in slag are taken as relatively high in this case at five per cent rather than 1.5 per cent assumed for the blast furnace. Lower levels can be achieved but at high residence times in the reduction stage and higher equipment costs. This will lower overall lead recovery unless slag fuming facilities are included. Cost estimates are provided on the same general basis of concentrate feed composition as used for the sinter plant–blast furnace estimates above and using energy data given in Chapter 16.

Capital costs Table 17.7 gives the capital cost estimate for the production of 100 000 t/a of lead in bullion by the Isasmelt process. TABLE 17.7 Lead smelter capital costs – Isasmelt process (cost basis – June 2005). Plant section

US$ million

US$/t of lead pa


14

140

Isasmelt plant

62

618

Gas cleaning

8

83

Acid plant

11

108

Ventilation

12

119

Slag disposal

8

83

Bullion handling

8

80

Effluent treatment

4

36

Sub total

127

1270


15

150

Site works at 6%

7

75

149

1495

12

120

Sub total – direct plant cost Land and access Licence fees

3

30


22

225

Sub total – indirect costs Total plant capital cost

37

375

187

1870

Total capital costs for the Isasmelt plant at US$1870 per annual tonne of lead in bullion is lower by 28 per cent than the comparable capital for a sinter plant and blast furnace at US$2610 per annual tonne.

Direct operating costs Direct operating costs for an Isasmelt lead smelter as at June 2005 are given in Table 17.8 for a plant of 100 000 t/a of lead annual capacity. The subdivision of plant sections corresponds with capital cost items given in Table 17.7.



277

Plant section

Number employed


16

Isasmelt plant Oxygen supply


Power kWh/t





27.2

880

12

72

341

1509

50

85

2750

320

1920

1897

750

2

3.4

110

205

1230

Gas cleaning

10

17

550

35

210

207

75

Acid plant

14

23.8

770

55

330

284

105

Ventilation

5

8.5

275

30

180

299

Slag disposal

8

13.6

440

5

34

209

Bullion handling

8

13.6

440

2

12

Effluent treatment

10

17

550

4

28

Administration

25

41.8

1353

4.8

Waste heat credit

5

8.5

275

-260

153

259

8393 83.9

Totals Total per tonne of lead Notes:


1.


2.

Coal is costed at US$80 per tonne with a thermal value of 30.7 GJ/t.

3.


Coal $000’s

1240


Total $000’s

Total $ t Pb

30

2828

28.3

426

8983

89.8

1340

13.4

1042

10.4

1669

16.7

113

866

8.7

60

743

7.4

200

60

712

7.1

90

225

892

8.9

29

0

1230

2612

26.1

-1560

70

38

-1178

-11.8

2485

3600

4157

1240

636

20 510

205.1

24.9

36.0

27.8

12.4

6.4

205.1

180


278

TABLE 17.8 Direct operating costs for a 100 000 t/a Isasmelt lead smelter (costs in US$ as at June 2000).


Labour costs are estimated from the number of employees required in each section covering both operating and maintenance, and an average annual employment cost of US$55 000 per employee. Employment cost should be the total including direct wages and all associated expenses such as taxes, pension scheme costs, insurances and benefits. Maintenance materials are simply estimated as 2.5 per cent of the direct construction capital cost for each section. This cost will be lower for a new plant but should rise to the 2.5 per cent level for a well established plant. Coal is used in the Isasmelt furnace at a cost of US$80/t in place of coke used for the blast furnace at US$160/t. Power is estimated from typical consumption figures for each area and a power cost of six cents per kilowatt in this example. Similarly coal and other fuels are estimated on the basis of typical energy consumption data provided in Chapter 16. In this case waste heat steam can be used to generate power and give a credit of 195 kWh per tonne of lead. Materials and supplies are estimates of a range of reagents and consumables as well as external services provided to each section. The cost of administration reflects higher salaried support and management staff, as well as services such as insurance and site taxes. These costs will vary widely depending on the location of the plant and whether it is a stand-alone operation or part of an industrial complex. As for the capital cost data, these figures are designed to provide a guide to the major operating cost items to be considered in any smelter evaluation. Total smelting costs are estimated at US$205/t of lead produced from the smelting operation, essentially the same as for the Kivcet process. Of this cost 41 per cent represents labour costs, 21 per cent is for energy and 38 per cent is for materials (including maintenance materials at 18 per cent).

COMPARISON OF SMELTING TECHNOLOGIES Table 17.9 gives a comparison of estimated capital and operating cost for the standard sinter plant– blast furnace technology with the Kivcet and Isasmelt processes as two examples of direct smelting technologies. TABLE 17.9 Comparison of smelting process costs (costs in US$ per tonne of lead). Process

Sinter plant–blast furnace

Kivcet

Isasmelt

2610

2015

1870

Labour

114.2

74.7

83.9

Energy

65.4

47.2

43.6

Capital ($/t/a) Direct operating costs

Materials

97.7

73.2

77.6

Total

277.3

195.1

205.1

98.3%

96.2%

94.6%

Lead recovery to bullion

The two direct smelting technologies show similar overall direct operating costs, although the Isasmelt process may have lower capital cost. However, lead losses in slag are higher and recovery correspondingly lower for the Isasmelt process, which will raise raw material costs. This is covered in more detail below in the section on overall economics for refined lead production.



279


Competitive cost curve The competitive position of a smelter in comparison with others is shown by use of the cumulative cost curve in which the direct cash cost for each of the Western world’s smelters is plotted against the cumulative tonnage capacity, arranged in ascending cost order. The general form of the curve is shown in Figure 17.2 for Western world lead smelters, predominantly sinter plant–blast furnace smelters, but also including later direct smelting operations.

Direct cash cost US$/ tonne Pb

600

500

400

300

200

100

0 0

500

1000

1500

2000

2500

Cumulative production 000's tonnes

FIG 17.2 - World lead smelter direct cash costs (excluding refining).

This data shows that 85 per cent of production is from plants with direct operating costs within the range of $160 to $350/t of lead produced. The first quartile boundary is $195/t, and the third quartile upper boundary is $320/t. The above estimate for the sinter plant–blast furnace technology is around the average mid range of the curve, whereas the direct smelting technologies fall close to the first quartile boundary. The curve covers a broad range and is not particularly flat, which is usually the case for a mature commodity. However in the case of lead, economics can be dictated to a large extent by the recovery of by-products such as silver or low raw material costs, and relatively high costs expressed per tonne of lead can be tolerated enabling such smelters to survive. Secondary residues or lead residues from zinc or other smelting operations can be used to significantly reduce raw material costs.

LEAD REFINING Although lead refining operations are often part of a primary smelting complex they can be independent and can be separated from a costing viewpoint. There are two main refining technologies – thermal and electrolytic, as discussed in Chapters 12 and 13 respectively. In this case costs are provided for removal of the full range of impurities and represent the maximum cost position. Depending on the nature of the raw materials processed by a smelter, this may not be necessary and some impurity removal steps can either be simplified or eliminated. In this sense it is difficult to compare the costs of individual refining operations.

280




Capital costs of thermal refining As discussed in Chapter 12, thermal refining involves six principal impurity removal stages, plus ingot casting and support services such as ventilation. The six impurity removal stages are:

• copper removal or copper drossing; • softening or arsenic, antimony and tin removal; • silver and precious metal removal; • zinc removal; • bismuth removal; and • caustic refining. The basis for cost estimates is the treatment of 100 000 tonnes per year of product lead. Generalised capital costs for a thermal refinery are shown in Table 17.10. Costs for each stage can vary widely and those given are for full refining at stage using standard refining kettles for each stage. TABLE 17.10 Thermal lead refinery capital costs (cost basis – June 2005). Plant section

US$ million

US$/t of lead pa

Copper drossing

6

60

Softening – oxygen pans

4

40

Desilverising

4

40

Dezincing

3

30

Debismuthising

4

40

Caustic refining

3

30

Holding kettles

1.5

15

Ingot casting

10

100

Ventilation

15

150

50.5

505


6

60

Site works at 6%

3

30

59.5

595

Land and access

2

20

Licence fees

1

10


9

90

12

120

72

715

Sub total


Sub total – indirect costs Total refinery capital cost

Capital costs for electrolytic refining As detailed in Chapter 13, electrolytic refining retains some of the thermal refining steps but replaces the removal of silver (and consequently zinc) and bismuth by the electrolytic refining stage. It generally produces a higher grade lead which can command a premium sale price.



281


Table 17.11 gives a generalised estimate of the capital costs of the electrolytic lead refinery. This is based on general performance data given in Table 13.3 and Table 13.4 and assumes the electrolytic plant will contain 370 cells each holding 28 anodes and consuming 1850 kW AC for the refining of 100 000 t/a of lead. Apart from the cells the electrolytic plant includes anode casting and cathode starting sheet fabrication facilities. TABLE 17.11 Electrolytic lead refinery capital costs (cost basis – June 2005). Plant section

US$ million

US$/t of lead pa

Copper drossing

6

60

Softening – oxygen pans

4

40

Anode casting

2

20

Cathode starter sheet fabrication

2

20

Electrolytic plant

40

400

Caustic refining

3

30

Holding kettles

1.5

15

Ingot casting

10

100

Ventilation Sub total Utilities and services at 12% Site works at 6%

7

70

75.5

755

9

90

4.5

45

89

890

Land and access

2

20

Licence fees

1

10


13

130


Sub total – indirect costs Total refinery capital cost

16

160

105

1050

On the basis of the above estimates the capital costs of electrolytic refining are substantially greater than for thermal refining by approximately 47 per cent.

Operating costs Estimates for operating costs for both thermal and electrolytic refining are given in Table 17.12 on the same basis as used for smelting operating cost estimates given above.

Comparison of refining processes Costs of the two refining technologies are compared in Table 17.13. The electrolytic refinery has substantially higher capital costs than the thermal refining process which relies on a series of standard refining kettles. Operating costs are also higher by US$11/t of product lead, although energy costs are marginally lower for electrolytic refining. It is possible that product premiums from the electrolytic refinery could cover the additional direct cost, but would not cover the return required for the additional capital cost.

282




TABLE 17.12 Direct operating costs for a 100 000 t/a lead refinery (costs in US$ as at June 2000). Plant section

Number employed

Man hours per t

Copper drossing

14

23.8

Softening

10

Desilverising

8

Dezincing

10

Debismuthising Caustic refining

Labour cost Power kWh/t $000’s


Maint Materials materials and supplies $000’s $000’s


Total $000’s

Total $ t Pb

600

1732

17.3

Thermal refinery 770

4.5

17

550

7.5

13.6

440

14

17

550

6

14

23.8

770

8

13.6

440

27

150

185

45

100

139

316

1150

11.5

84

100

68

276

968

9.7

36

75

75

124

860

8.6

14

84

100

129

580

1663

16.6

3

18

75

60

100

693

6.9

4

6.8

220

1

6

38

30

20

314

3.1

Casting

24

40.8

1320

4

24

250

180

20

1794

17.9

Ventilation

12

20.4

660

24

144

375

90

1269

12.7

Administration

21

35.4

1144

4

24

75

1040

2283

22.8

Totals

125

212

6864

492

1338

1996

2036

12 725

127.3

68.6

4.9

13.4

20.0

20.4

127.3

Total per tonne of lead Electrolytic refinery Copper drossing

14

23.8

770

4.5

27

150

185

600

1732

Softening

10

17

550

7.5

45

100

139

316

1150

11.5

Electrolytic plant

46

78.2

2530

155.4

932

900

345

150

4857

48.6

Caustic refining

8

13.6

440

3

18

75

60

100

693

6.9

Holding kettle

4

6.8

220

1

6

38

30

20

314

3.1

Casting

24

40.8

1320

4

24

250

180

20

1794

17.9

Ventilation

8

13.6

440

12

72

235

50

797

8.0

Administration

23

38.8

1254

4

24

78

1140

2496

25.0

Totals

137

233

7524

1148

1826

2129

1206

13 833

138.3

75.2

11.5

18.3

21.3

12.1

138.3

Total per tonne of lead

283

Note: Power is costed at 6 US cents per kWh, and natural gas fuel is costed at US$4 per GJ.

17.3



Holding kettle


TABLE 17.13 Comparison of refining process costs (costs in US$ per tonne of lead). Process

Thermal refinery

Electrolytic refinery

715

1050

Labour

68.6

75.2

Energy

25.3

23.6

Materials

33.4

39.6

Total

127.3

138.3

Capital ($/t/a) Direct operating costs

Most electrolytic refineries were built before the successful application of the Kroll-Betterton process for bismuth removal as the only reasonable approach for bismuth removal. Since that time no new electrolytic refineries have been installed and the adverse economics are clearly indicated by these figures.

METAL PRICING Generally lead product prices are set by international metal exchanges, principally the London Metal Exchange (LME). Product price is therefore usually outside the control of the producer. The demand for metals as commodities is determined by the international trade cycle. Supply and demand are rarely in equilibrium, and for this reason there is a need for an international terminal market as a gathering of traders to purchase and sell metal, establishing a net supply-demand balance and setting prices accordingly. The bulk of metal traded on world markets is contracted directly between producer and consumer, but is priced on the basis of the international terminal market quotation. The LME has three functions in relation to metal trading: 1. To register prices as set by supply and demand trends. This is done on a daily basis. 2. To receive and deliver physical metal which is held by traders in LME warehouses. A warrant is issued giving title to metal held in a number of warehouses worldwide. 3. To provide facilities for hedging contracts of which both forward selling and buying contracts are available. The market trades seven base metals including lead, and there are four opportunities to trade each day. Daily prices are quoted at the end of each morning trade as representative of the contracts made. Both a cash and three month forward delivery price are quoted. The three month price may be above the cash price (termed a ‘Contango’) or below the cash price (termed a ‘Backwardation’). The LME market, in large part, reflects future expectations as well as current market conditions. Pricing is generally determined by market dynamics, considering factors such as:

• consumption demand and growth projections; • supply chain projections for both mines and smelters; and • stock holdings in smelters, LME warehouses and with end consumers – or market liquidity. The timing of changes in these market fundamentals can give rise to significant differences between the present and future metal prices. The involvement of option traders and hedge funds speculating on these balances is another important issue, and this activity can significantly influence pricing, deviating the market away from its underlying fundamentals in the short term.

284




To some extent forward pricing is responsive to metal stocks, and will generally fall if stocks move well above normal levels, and vice versa. The alternative to price setting by a terminal market is the use of a ‘Producer Price’, set by a cartel of producers, or by one major producer which others follow. This scheme was used up until the mid 1980s when both Producer Price and terminal market pricing schemes operated. Generally LME pricing is regarded as more independent, reflecting the market supply and demand balance. It can be manipulated by traders, but only over short periods of time. Metal can always be sold at negotiated prices between producer and consumer, but since the smelter is often separate from the mine supplying concentrate, a common metal pricing system is needed for both mine and smelter products, and the Producer Pricing approach is not particularly practical or equitable. Metal held in LME warehouses must conform to set quality standards, and for lead there are a number of standards as given in Table 14.1 in Chapter 14.

BY-PRODUCTS For the typical lead smelter the principal by-products will be silver and gold, copper dross, sulfuric acid and antimony metal, usually in the form of antimonial lead alloy. Other possibilities are arsenic compounds and zinc oxide if slag fuming facilities are installed.

Silver and gold Silver is a key by-product and can return realised metal values of up to 50 per cent or more of the primary lead value. It is, however, highly variable and depends on the raw materials used. Silver is efficiently recovered into the primary lead bullion from most sulfide concentrates. It is separated during refining as a zinc–silver alloy which is further processed by distillation and cupellation to give a silver doré, which is electrolytically refined to pure silver. Gold is recovered with silver and separated during the final silver refining stage to yield a gold doré which can be sold for final refining. Both metals are cast into bars for sale at metal exchange prices with minimal realisation costs. The example used as the basis for cost estimates in this Chapter assumes a lead concentrate containing 60 per cent lead, 1000 g/t silver and 3 g/t gold. For recoveries of 96 per cent and 93 per cent for silver and gold respectively, and for prices of US$10 per troy ounce for silver and US$600 per troy ounce for gold, the returns are US$523/t of product lead for silver and US$91/t of product lead for gold.

Copper matte Copper drossing as the first stage in refining can yield a copper sulfide material for sale to a copper smelter. Often such drosses only contain ten to 15 per cent copper and receive a poor return. However, this can be upgraded to a rich copper matte with recovery of entrained lead by using a short rotary or reverberatory furnace. With continuous copper drossing, matte grades can be up to 50 per cent copper, but are typically between 30 and 45 per cent. Apart from sulfur the other main constituent of copper matte is lead, which is not favoured by copper smelters in large amounts due to the effects on refractories. Hence prices paid for this material can reflect standard copper concentrate terms but with penalties for the lead content. A decision on whether to install dross upgrading or continuous copper drossing equipment will depend on the level of copper normally encountered in smelter feed concentrates.



285


In the example used for the basis of cost estimates in this chapter, the copper content of concentrates is 0.5 per cent and for a blast furnace slag of 0.15 per cent copper, recovery into matte would be around 80 per cent, which would yield 0.017 tonnes of copper matte containing 40 per cent copper per tonne of lead produced. Taking smelting charges and penalties at US$1000/t of contained copper and the LME copper prices at US$7500/t, the return for copper matte is US$44/t of lead produced.

Sulfuric acid The sulfur to lead ratio in lead concentrates is highly dependent on the concentrate grade and hence the quantity of sulfuric acid produced will vary accordingly. For a typical lead smelting operation as detailed above, with a concentrate feed of 60 per cent Pb, the sulfuric acid production will be close to 0.865 tonnes per tonne of lead produced. Acid produced from the sintering operation will be ‘black’ acid, discoloured by organic carbon distilled from entrained flotation reagents and kerogen material in the sinter feed, whereas acid produced from direct smelting processes will be ‘white’ acid due to the higher temperatures which destroy the organic species present and oxidises the carbon to CO 2. Black acid has limited application, such as the manufacture of fertilisers, otherwise it must be treated to remove the colouration by oxidation of the carbon with hydrogen peroxide. In either case the return from ‘black’ acid will be at a discount to the return from ‘white’ acid, and for example could be taken as around US$10/t. The net sale price for ‘white’ sulfuric acid will vary greatly depending on smelter location and proximity to acid markets, but by way of example may be taken as US$20/t ex works. This represents a credit of 0.865 × 20 = $17.3/t of lead produced, whereas as ‘black’ acid the return will be around $8.65/t of lead produced. In addition, the basic price of sulfuric acid as a traded commodity depends on the world market price of raw sulfur or brimstone. The supply-demand balance for brimstone can vary widely and is dependent on the general availability of smelter acid as well as other factors such as its production rate and the stock position. Given also that freight costs for shipping sulfuric acid from the smelter can be significant, particularly for smelters in remote locations, there are many situations where returns from acid sales can be negative, representing a disposal cost. Regular output of acid is essential to maintain smelter operation, as storage is limited, and often there can be no option but to accept poor or negative returns.

Other by-products Antimonial-lead alloys are the main additional by-product, but returns depend on local demand and the particular alloys required, and are difficult to quantify in a general way. With the popularity of calcium–lead alloys for sealed, maintenance free batteries, the price of antimonial alloys declined, but has resurged due to increased demand for specialised batteries. Clearly this market is quite volatile. In some instances other by-products such as tellurium, tin, bismuth, germanium and selenium can provide valuable returns to the smelter, but this is unusual.

OVERALL ECONOMICS FOR REFINED LEAD PRODUCTION Revenues are obtained from the sale of lead metal and by-products. Generally lead is sold at a premium to the LME price. The premium is in part established by the cost of accessing metal from alternative sources such as a LME warehouse or another supplier, and

286




will vary according to the location of the customer. Premiums are also received for upgraded metal such as alloys. For the purposes of example the overall metal price premium may be taken as an average level of US$50/t of lead sold. Costs borne by the smelter and refinery in producing lead metal are:

• The cost of concentrates. In accordance with Chapter 3 ‘Commercial Terms for Purchase of Standard Lead Concentrates’ concentrate feed has been costed on the basis of a 60 per cent lead content, 1000 g/t of silver and 3 g/t of gold, and a treatment charge of US$265/t of concentrate at a lead price of US$1100/t, giving a net cost of US$650/t of concentrate.

• The direct cash costs for the smelter operation. These have been summarised in Table 17.9 for three different smelter technologies.

• The direct cash costs for lead bullion refining. These have been summarised in Table 17.13 for the two different refining technologies.

• Annual capital expenditure required to sustain the smelter operation. This represents the ongoing annual expenditure on equipment replacement and modification. It may be estimated as 50 per cent of the cost of new plant spread over 30 years. This recognises an average plant life of 30 years and assumes that part of the installation, such as structures, would not need replacement.

• The cost of delivery of product metal to the customer including handling, freight, insurance and sales organisation expenses. This will vary greatly from one plant to another and will depend on relative locations of the smelter and its customers. These costs will be low for plants located in Europe and Japan, but will be high for plants located in Canada and Australia for example. For assessment purposes an average delivery cost is taken as US$40/t of lead. Lead price is the main variable and uncertainty in any assessment of smelter economics, and is outside the control of the smelter operator. A new venture or investment in lead extraction facilities, whether mine or smelter, must be able to withstand wide movements in the lead price and at least remain cash positive at all points in the expected price cycle. Projection of future pricing is perhaps the most difficult issue and the greatest uncertainty facing any new investment.

Basis used for evaluation of economics of lead production Assumptions used in evaluating raw material costs and revenues are as follows: Metal prices (LME)

Concentrate grades


Lead

US$1100 per tonne

Copper

US$7500 per tonne

Silver

US$10 per troy ounce

Gold

US$600 per troy ounce

Lead

60 per cent

Copper

0.5 per cent

Silver

1000 g/t

Gold

3 g/t


287


Recoveries

Lead

98.3 per cent for blast furnace 96.2 per cent for Kivcet 94.6 per cent for Isasmelt

Copper

80 per cent

Silver

96 per cent

Gold

93 per cent

Economic evaluation of lead production Using the above assumptions Table 17.14 details the costs and returns of lead production using the sinter plant–blast furnace process and the Kivcet process as an example of direct smelting technology, both followed by thermal refining. TABLE 17.14 Economics of lead production. Item

Sinter plant–blast furnace process US$/tonne of lead

Kivcet process US$/tonne of lead

Smelter capital

2610

2015

Refinery capital

715

715

Total capital

3325

2730

Revenues Lead sales at LME price

1100

1100

Premium

50

50

Silver

523

535

Gold

91

93

Copper matte

44

45

Sulfuric acid

8.6

18.7

Total revenue

1817

1842

Costs Raw materials

1102

1127

Smelter operating

277.3

195.1

Refinery operating

127.3

127.3

Sustaining capital

55.4

45.5

Delivery costs Total costs

40

40

1602

1535

Cash margin

215

307

As return on investment

6.5%

11.2%

The returns on investment indicated would not normally be regarded as sufficiently attractive to justify investment, and smelter costs need to be substantially reduced, or returns by way of treatment charges need to be substantially increased. New direct smelting technologies offer improved economics but returns have been insufficient to justify new greenfields smelters at prevailing metal prices.

288




This particular situation results from the lack of a need for new smelting facilities over many years and the fully depreciated state of most of the primary smelters. Through competitive pressures, smelting charges have been pushed down to levels where direct costs are covered with reduced margin to cover capital charges. However, coupled with the environmental benefits, direct smelting is the only option for primary smelting plant replacement or expansion. The impact of lead price on the margin, and return on investment for the sinter plant–blast furnace process is given in Table 17.15. From this it can be seen that the margin increases by around $29/tonne of lead produced for every $100/tonne increase in the lead price. This reflects the distribution of lead price gains between the concentrate supplier and the smelter and is a function of the concentrate purchase terms – particularly the escalation of the treatment charge with the lead price. TABLE 17.15 Effect of lead price on smelter economics. LME lead price (US$/t)

800

900

1000

1100

1200

1300

1400

1500

Revenues (US$/t Pb)

1517

1617

1717

1817

1917

2017

2117

2217

Concentrate cost (US$/t Pb)

888

960

1031

1102

1173

1244

1316

1387

Other costs (US$/t Pb)

500

500

500

500

500

500

500

500

Margin (US$/t Pb)

129

157

186

215

244

273

301

330

Return on investment (%)

3.9

4.7

5.6

6.5

7.3

8.2

9.0

9.9

Although the above illustrates the factors involved in lead smelter economics there are wide variations in the costs and returns for lead smelters. It is therefore not possible to define costs in a general way, as in the above examples, and a detailed assessment of each individual situation is necessary.

ECONOMICS OF SECONDARY LEAD PRODUCTION As an example, indicative costs are provided for the treatment of scrap lead-acid automotive batteries to produce refined lead using battery breaking and paste desulfurisation, followed by melting and reduction in a short rotary furnace, as described in Chapters 10 and 11. A typical small scale unit handling 35 000 tonnes per year of batteries is considered with a lead production of around 19 000 tonnes per year. Relevant production data is as follows: 35 000 t/a • batteries processed

• refined lead produced

19 000 t/a

• wastes for disposal (slag, gypsum, plastics)

6500 t/a

• polypropylene recovered

2100 t/a

A plant of this size will have a battery breaker and shredder, separation plant to recover metallic lead, pastes, polypropylene and other plastic and a single rotary furnace of 7 - 8 m3 working volume. Capital costs will be of the order of US$15 million for battery handling, breaking and separation, and US$10 million for the smelting furnace and ancillaries. The total cost of US$25 million represents US$715/t of batteries per annum or US$1315/t of lead per annum. Operating costs are detailed in Table 17.16. If the cost of collection and transport of scrap batteries is taken as $100/t (or $194/t of lead), then the total cost of refined lead produced is US$633/t.



289


TABLE 17.16 Operating cost for secondary lead production. Item

Quantity per annum

Unit cost

Annual cost $000’s

Cost per t Pb

Labour

26

$55 000 pa

1430

75.2

Power

4.8 m kWh

$0.08

384

20.0

Fuels

11 000 GJ

$4.5

50

2.6

Soda ash

2950 t

$250

738

38.8

Lime

980 t

$80

78

4.1

195

10.2

Battery breaking

Other supplies Maintenance materials

2% capital

Sub total

300

15.8

3175

167.1 49.2

Smelting operation Labour

17

$55 000 pa

935

Power

2.7 m kWh

$0.08

216

11.4

Fuels

66 000 GJ

$4.5

297

15.6

Fluxes

1800

$50

90

4.7

Oxygen

4000 t

$160

640

33.6

Other supplies Maintenance materials

2.5% capital

Caustic refining costs Waste disposal

6500 t

$60

Sub total

143

7.5

250

13.2

380

20

390

20.5

3341

175.8

Administration Labour

3

75 000

Supplies and services Sub total Sustaining capital cost

5% capital

Grand total

225

11.8

350

18.4

575

30.2

1250

65.8

8341

438.9

Credit may be obtained from the recovery of polypropylene at 110 kg/t of lead produced. With a current value of US$1.30/kg the credit is worth $143/t of lead or 33 per cent of the direct processing costs, and will reduce the overall direct cost of refined lead to $490/t. This is clearly a most significant by-product. With an LME lead price of US$1100/t the profit margin is $467/t of lead or 35 per cent of fixed capital. Production of lead by this approach is therefore quite profitable in comparison with primary lead production at the assumed lead price, and may be a factor in depressing the returns for primary lead smelters in a situation where the growth in new lead consumption has been negligible over many years. However, it should be noted that the returns from secondary smelting are far more sensitive to lead price since there is no compensating reduction in raw material costs as the lead price falls, which is the case for primary smelters. The sensitivity of the operating cash margin to lead price is illustrated in Table 17.17. The cash break-even lead price in the above example would be US$624/t, below which the secondary smelter would operate at a cash loss.

290




TABLE 17.17 Price sensitivity of secondary smelters. LME lead price (US$/tonne)

700

800

900

1000

1100

Return (US$/t Pb)

750

850

950

1050

1150

Scrap collection

194

194

194

194

194

Smelting

440

440

440

440

440

Costs (US$/t Pb)

Realisation

40

40

40

40

40

Total costs

674

674

674

674

674

Margin (US$/t Pb)

76

176

276

376

476

Return on investment (%)

5.8

13.3

21.0

28.6

36.2

The relative sensitivities of cash margin to LME lead price is shown in Figure 17.3. For primary smelters a fall in lead price will cause a proportional fall in raw material costs and a decline in allowable treatment charge, but the decline in treatment charge will be only a fraction of the lead price decline and in some contracts may in fact have a floor below which it cannot fall. This has the effect of significantly reducing metal price sensitivity. Primary

Secondary

1000

Cash m argin US$/t Pb

800

600

400

200

0

-200 400

600

800

1000

1200

1400

1600

Lead Price US$/t

FIG 17.3 - Sensitivity of operating cash margin to lead price for primary and secondary smelters.

In the past, the high sensitivity of secondary lead production to lead price has acted to limit any decline in the lead price, since it reaches a point where some secondary operation is not viable and production is reduced, decreasing supply and hence forcing an increase in price or at least a halt in any price decline. Primary smelters will remain cash positive at the point where secondary smelters become cash negative. The reverse situation does not apply to secondary lead, since scrap availability is relatively inflexible, and any significant growth in lead demand must be met from increased primary lead production. This will result in increased concentrate demand and cause an increase in lead price.



291


The above may present a somewhat simplistic view, but does illustrate the interplay of primary and secondary operations in influencing lead price movements. The other major factor is of course mine supply.

292



Sponsors The Australasian Institute of Mining and Metallurgy would like to thank the following sponsors for their generous support of this volume.

v

APPENDIX 1 Properties of Lead and Associated Compounds LEAD METAL PROPERTIES Atomic number

82

Atomic weight

207.19

Stable isotopes, relative abundance Mass number

204

206

207

208

Per cent abundance

1.48

23.6

22.6

52.3

Valence

Usually 2, sometimes 4 or 1

Melting point

327.4°C

Boiling point (at 760 mm Hg)

1751°C

Vapour pressure Temperature °C

987

1167

1417

1611

1751

Vap pressure mm Hg

1.0

10.0

100.0

400.0

760

Density

Solid at 20°C

11.34

Solid at 327.4°C

11.005

Liquid at 327.4°C

10.686

Coefficient of linear thermal expansion (polycrystalline)

29.3 × 10-6 °C-1

Coefficient of volumetric thermal expansion

87.9 × 10 °C

Heat capacity of solid

J.mole .°C

-6

23.6 + 0.00962 . T

-1

-1

-1

-1

-1

-1

-1

(T in °K )

Heat capacity of liquid

32.4 - 0.00305 . T

J.mole .°C

Heat capacity of vapour

20.79

J.mole .°C

Heat of fusion

5425.38

J.mole at 327.4°C

Heat of vaporisation

175068.1

J.mole at 1525°C

Thermal conductivity of solid

(25°C)

0.347

J.sec-1.cm-1.°C-1

Thermal conductivity of solid

(327.4°C)

0.310

J.sec .cm .°C

Thermal conductivity of liquid

(327.4°C)

0.159

J.sec .cm .°C

Liquid surface tension

4.70 Nm.cm


-1 -1

-1

-1

-1

-1

-1

-1

-1


293

APPENDIX 1 – Properties of Lead and Associated Compounds

-6

-3

Electrical resistivity at 20°C (polycrystalline solid)

20.648. 10 ohm.cm

Temperature coefficient per °C

0.00336

Tensile strength

126.55 - 175.77 kg cm -2

Modulus of elasticity

0.155 × 106 kg cm-2

BINARY LEAD RICH EUTECTICS Metal

% Pb by weight

Liquidus temp °C

95.3

304

Ag

Lead rich compound

Ca

90.0

630

Cu

99.94

326

CaPb3

K

98.1

277

KPb4

Na

97.3

307

NaPb3

Li

99.3

235

LiPb

Mg

97.8

253

Mg2Pb

Cd

82.6

248

Zn

99.5

318.2

Sn

38.1

183

Sb

88.9

252

As

97.2

288

Te

99.975

326.3

PbTe

Se

79.5

860

PbSe

PROPERTIES OF LEAD OXIDES Oxide Molecular weight

PbO

PbO2

Pb3O4

223.21

239.19

685.57

α Red β Yellow

Dark brown or black

Orange to red

Structure

α Tetragonal β Orthorhombic

Orthorhombic or tetragonal

Spinel

Density

α 9.2 - 9,5 β 9.5 - 9.9

9.165 - 9.40

9.1

888°C

Decomposes at 290°C

830°C in oxygen Decomposes at 500°C

Heat of formation (kJ/mole)

α -219.28 β -217.86

-276.65

-734.29

Free energy of formation (kJ/mole)

α -188.82 β -188.07

-212.42

-617.14

Entropy (J/°C.mole)

α 65.27 β 67.36

76.48

211.29

Heat capacity (J/°C.mole)

α 45.77 β 45.85

62.22

142.26

Colour

Transition temperature Melting point

294

488.5°C




VAPOUR PRESSURES Vapour pressures (P) expressed in mm of mercury and as a function of temperature may be expressed by the following equations, where T is the temperature in °K. Lead metal

log(P) = 7.518 - 9386/T

Lead monoxide (PbO)

log(P) = 9.457 - 11476/T

Lead sulfide (PbS)

log(P) = 10.196 - 11368/T

SILVER METAL PROPERTIES Atomic number

47

Atomic weight

107.868

Stable isotopes, relative abundance Mass number

107

109

Per cent

51.82

48.18

Melting point

960.8°C

Boiling point (at 760 mm Hg)

2212°C

Density

(Solid at 25°C) 10.5

Heat capacity of solid

25.54

J.mole-1°C-1 (T in °K)

Heat capacity of liquid

30.54


Heat capacity of vapour

20.8


Heat of fusion

11 945

J.mole-1 (at 960.8°C)

Heat of vaporisation

254 052

J.mole-1 (at 2212°C)



295


THERMODYNAMIC PROPERTIES OF COMPOUNDS INVOLVED IN LEAD EXTRACTION Source: Handbook of Chemistry and Physics – CRC. Compound

Mol wt

Density (298°K) g/cm3

Pb (liquid)

207.19

11.344

PbS

239.27

PbO (yellow) PbO2

Heat of formation (298°K) kJ/mole

Free energy of formation (298°K) kJ/mole

7.5

-94.31

-92.67

223.21

9.53

-217.86

-188.49

239.21

9.375

-276.65

-218.99

PbSO4

303.27

6.2

-918.39

-811.24

PbCO3

267.22

6.60

-700.00

-626.34

PbSiO3

283.27

6.49

-1082.82

-999.97

PbCl2

279.12

5.80

-359.20

-313.97

FeS2

119.98

5.0

-177.90

-166.69

FeS

87.91

4.6

-95.06

-97.57

FeO

71.85

5.7

-266.52

-244.35

Fe2O3

159.69

5.24

-822.16

-740.99

SiO2

60.07

2.65

-859.39

-805.00

SO2

64.06

-296.81

-299.91

H2SO4

98.06

1.834

-810.40

-733.92

CaCO3

100.09

2.711

-1211.13

-1133.03

CaO

56.08

3.32

-634.71

-603.75

CaSO4.2H2O

172.14

2.32

-2005.52

-1780.17

ZnS – sphalerite

97.44

4.102

-202.92

-198.32

ZnS – wurtzite

97.44

4.087

-189.53

-184.93

ZnO

81.38

5.606

-348.38

-318.19

ZnSO4

161.45

3.74

-978.55

-871.56

ZnCO3

125.39

4.42

-812.53

-731.36

Zn2SiO4

222.85

4.05

Zn (g)

65.37

0.0029

130.44

94.93

CO (g)

28.01

0.00125

-110.54

-137.28

CO2 (g)

44.01

0.00197

-393.51

-394.38

CH4 (g)

16.04

0.000717

H2O (g)

18.01

296


-74.85

-50.79

-241.84

-228.59



HEAT CAPACITIES AT CONSTANT PRESSURE (Joule. mole-1. °K-1 as a function of temperature T°K) Cp = a + b T + c T2 + d T-2 Suitable for elevated temperatures around 1000°C. Compound

a

Pb (liquid)

28.451

b

c

d

PbO

43.221

0.01331

PbO2

53.137

0.03263

PbCl2

66.442

0.03493

PbS

44.476

0.01678

PbSO4

110.458

FeS2

74.768

0.005565

FeS

50.417

0.01142

FeO

52.802

0.006242

-318 821

Fe2O3

103.428

0.06711

-1 771 506

Fe3O4

172.255

0.07874

-4 098 228

Fe (solid)

25.606

0.01406

Fe (liquid)

34.10

-1 274 028

SiO2

45.815

CaCO3

82.341

0.04975

-1 286 998

CaO

41.84

0.02025

-451 872

CaSO4

77.487

0.09192

-656 051

Zn (g)

20.786

ZnO

49.003

0.005104

-912 279

ZnS

49.246

0.005272

-485 344

ZnSO4

117.152

0.02301

SO2 (g)

47.697

0.005916

-855 628

SO3 (g)

58.158

0.02552

-1 347 248

C

17.154

0.004268

-878 640

CO (g)

26.861

0.006966

-8.2 E-7

CO2 (g)

25.999

0.04350

-1.48 E-5

CH4 (g)

14.146

0.07550

-1.8 E-5

H2 (g)

29.066

-0.00084

2.013 E-6

H2O (g)

30.359

0.009615

1.184 E-6



297

INDEX

Index Terms

Links

A Accretions in boiler systems in the blast furnace Agricola

112 70 17

Alcoa bipolar electrolytic cell Alloying methods

155 243

Aluminium for copper separation American water jacketed furnace

205 22

Anode slimes in lead refining

238

Anodes for electrolytic refining Antimony

233 221

recovery from softener slag

206

recovery in Harris process

208

removal from lead

205

Arsenic

221

removal from lead

205

Ausmelt furnace See Slag bath reactor Ausmelt process

123

for secondary lead

188

reactor design

125

Autogenous grinding

33

This page has been reformatted by Knovel to provide easier navigation.

Index Terms

Links

B Balbach-Thum cell

216

Battery – lead acid breaking and separation

170

composition

168

Battery paste desulfurisation

172

Battery paste analysis

170

Betts process

227

performance details Bismuth

230

236 222

recovery

239

removal from lead

219

Blast furnace

65

accretions

81

capacity determinants

81

coke requirements

70

construction

75

development and evolution

21

energy consumption

259

energy use

260

feeding

80

feeding practice

77

for secondary lead

178

gas handling

80

heat balance

83

oxygen enrichment

83

performance data

81

Port Pirie design

78

reaction zones

67

slag composition

72

tapping requirements

79

treatment of reverb slags

73

179


Index Terms Blood lead levels Blue Powder

Links 250

251

92

Boliden lead process

105

Bulk concentrates commercial terms

40

By-products revenues By-products from zinc production

285 286

C Cadmium

222

Calcium

222

use for bismuth separation Castilian blast furnace Casting machines

219 22 244

Cathode starting sheets for lead refining

233

Caustic soda refining

220

Cementation for solution purification Chloride complexes of lead and silver

160 156

Chlorine for zinc separation

219

from electrolytic process

154

removal from fume

145

Chrome magnesite refractories

104

Coke reactivity

70

reactivity in ISF

96

sizing for blast furnace

70

use in blast furnace

70

use in Kivcet process use in the ISF Coke layer in the Kivcet process

113 96 113


Index Terms

Links

Condenser lead splash

92

Converting lead bullion Copper

106 222

removal from lead

199

removal with aluminium

205

Copper dross treatment with sodium

204

Copper dross treatment

204

Copper drossing

201

continuous method

202

Copper matte as a by-product Cost curve for world plants

285 280

Costs comparison of refining processes

282

for Isasmelt process

277

for Kivcet process

274

for refining operations

280

of sinter plant blast furnace

269

of smelting and refining operations

269

smelting process comparisons

279

Cupellation

27

216

Current modulation in electrolytic refining

236

D Davey desilvering process

213

Dehalogenation See Chlorine removal of zinc fume Desulfurisation of battery paste

145 172


Index Terms

Links

Direct smelting

27

energy use

262

fuel requirements

100

principles

99

processes

99

Dithionic acid electrolyte Dwight Lloyd sintering machine

152 25

46

E Economics of refined lead production

286

Economics of secondary lead production

289

Effluent standards

255

Effluent treatment methods

255

Electric furnace for lead smelting

105

for slag cleaning

129

in Kivcet process

114

treatment of secondary slags

180

use for slag fuming

146

Electrode potentials, standard

228

Electrolysis of lead chloride

151

Electrolytic cells for molten salt electrolysis

153

Electrolytic refining anode slimes treatment

238

anodes

233

cathodes

235

cell voltage

229

cells

230

chemical principles

227


Index Terms

Links

Electrolytic refining (Cont.) costs

281

electrolyte control

235

fluoborate electrolyte

240

performance

236

starter sheets

235

sulfamic acid electrolyte

239

Electrowinning for secondary lead

188

of primary lead

151

Electrowinning processes energy use

264

Energy consumption in smelting and refining

259

Energy consumption comparison for extraction

266

processes for concentrate production

266

for secondary lead

267

Engitec Fluobor process

162

Engitec secondary lead process

192

Environmental impact of lead

249

Environmental lead controls

254

F Faber Du Faur furnace

215

Ferric chloride

156

Flash smelting See Kivcet process Flash smelting shaft Kivcet

110

Flotation See Froth flotation Flotation process See Froth flotation


Index Terms

Links

Flue dusts handling and treatment

147

Fluoborate electrolysis systems

162

electrolyte

152

Fluorine in sinter gas removal from fume

61 145

Fluosilicate electrolysis systems

161

electrolyte

152

Fluosilicic acid

235

Froth flotation

32

collectors

34

depressants

34

flow sheets

35

33

Fume treatment prior to zinc extraction

144

iron formation during

143

Fuming

G Ginatta process

193

Gold

222 by-product returns

285

chlorine refining

217

refining

217

separation from silver

217

Graphite anodes Gravity concentration

158 32


Index Terms

Links

H Halex

159

Harris process

206

treatment of caustic slags

207

two-stage configuration

208

Health effects of lead Hearth furnace

249 19

Heat balances smelting processes

99

Heat capacities of compounds

297

Howard press

215

Huntington-Herbelein process

24

46

I Imperial Smelting Furnace See ISF process Impurities in lead

221

Indium

222


208

Ingot dimensions

244

Intec process

159

Iron

223 formation during fuming

143

formation during ZnO reduction

133

Isasmelt process

120

costs

277

energy use

263

for secondary lead

188

ISF blast air preheat

91

blast air rates

96

coke use

96

furnace accretions

95

furnace capacity

96


Index Terms

Links

ISF (Cont.) furnace design

92

furnace top temperature control

92

process

89

slag and bullion tapping

95

slag formation and composition

92

tuyere characteristics

94

K Kaldo process

106

Kinetics of ZnO reduction from slags

132

188

Kivcet furnace construction

116

gas offtake shaft

111

Kivcet process

109

costs

274

energy use

262

performance

114

Kroll Betterton process

219

L Lance processes using

119

Lance reactors for secondary lead

188

casting methods

243

Lead

consumption

6

health exposure limits

251

in air standards

254

in slag from direct smelting

101

metal cycle metal quality specifications

8 243


Index Terms

Links

Lead (Cont.) mine production presence in the environment secondary production

6 249 7

standard ingot dimensions

243

toxicity

249

uses

5

Lead blast furnace slags fuming

135

Lead block casting

245

Lead bullion commercial terms

41

from the ISF

95

Lead carbonate

172

Lead chloride aqueous electrolysis

158

molten salt electrolysis

153

preparation from sulfide

156

solubility

156

Lead concentrates commercial specifications

36

commercial terms

38

grade versus recovery

36

37

Lead dioxide formation at anode

152

in hydrometallurgical processing

189

Lead metal properties Lead minerals

161

191

293 31

Lead oxides properties

294

Lead sheet production equipment

246


192

Index Terms

Links

Lead smelting capacity of smelters historical developments

8 29

industry structure

7

primary smelters

9

secondary capacity

13

Lead, sulfur, oxygen system

101

Lead-copper-sulfur system

200

Lead-silver phase diagram

210

Lead-zinc phase diagram

212

Leferrer furnace

215

Lime in waste water treatment

255

London Metal Exchange

284

metal prices

38

M Magnesium

223

use for bismuth separation Magnetite formation Matte

219 104 25

treatment of Medieval furnaces

147 17

Melilite importance in sinter structure Mercury

49 223

removal from sinter gas Metal payment in concentrates Minemet process Mineral separation Moebius cell

61 39 159 32 216

Molten salt electrolysis impurities

155


Index Terms

Links

N Nickel

223

O Occupational exposure controls

253

Occupational health standards

251

Oliforno process

185

Ore size reduction Outokumpu lead process

33 125

Oxygen enrichment in ISF blast air

96

in blast furnace

83

in Isasmelt process

121

in QSL process

116

use for softening lead

205

Oxygen stoichiometry in Ausmelt process

121

P Parkes process

211

Pattinson process

27

Peak bed temperature – sinter

48

210

Penalty elements in lead concentrates

39

PLACID process

158

Platinum Group Metals

223

193

Polypropylene recovery from batteries Port Pirie blast furnace Pricing of metals

170 23 284


Index Terms

Links

Q QSL process

116

energy use

263

performance

118

R Reaction zones in blast furnace

67

Refining by electrolytic methods

227

by thermal methods

197

energy consumption

259

energy use in electrolytic process

261

energy use in thermal process

261

equipment

199

kettles

199

secondary lead

193

221

Refractory protection by magnetite Retorting of silver crusts

104 215

Reverberatory furnace for secondary lead

175

Reverberatory hearth

20

Roasting

23

Rotary furnaces

181

for secondary smelting

185

performance with soda slags

187

Rotary kiln for secondary smelting

183

secondary smelting performance RSR electrolytic process for secondary lead

184 161 191


Index Terms

Links

S Scotch hearth

18

Secondary lead economics

289

effect of lead price on economics

290

energy consumption for recovery

267

Secondary lead residues

174

Secondary raw materials

41

pretreatment

167

Secondary smelting

175

processes

13

Selenium

223

Short rotary furnace See Rotary furnace Silver

223

by-product returns

285

continuous removal process

214

cupellation

216

electrolytic refining

216

parting

217

properties

295

recovery efficiency

215

recovery methods

26

removal from lead

210

treatment of zinc alloy

214

zinc phase diagram

211

Silver minerals

31

Sinter composition for blast furnace

49

composition for ISF

91

handling problems

54

physical structure

49

quality

48

Sinter charge preparation

50


Index Terms

Links

Sinter gas handling

60

Sinter gas scrubbing liquor

62

Sinter machine capacity and performance

56

downdraft

47

effect of concentrate grade

59

returns control

54

Sintering

45

coke additions

51

early processes

25

feed moisture control

51

fuel requirements

50

gas distribution and recycle

53

machine feeding

52

recycle ratios

51

updraught versus downdraught

55

Sirosmelt reactors

119

Skimmings from ingot casting

245

Slag composition for blast furnace

73

granulation

79

ISF composition

92

Slag bath reactor design and operation

142

zinc fuming process

141

Slag cleaning

129

Slag fuming

130

conventional furnace batch operation

136

design

135

zinc recovery

138

furnace boiler

139

furnace design

135

furnace operation

136


Index Terms

Links

Slag fuming (Cont.) high intensity processes

144

using slag bath reactor

141

zinc elimination rates

137

Slag fuming conventional furnace coal use

139

Slag viscosity

72

Sliming of lead minerals

33

Smelter by-products

40

285

Smelting energy consumption

259

Soda slags disposal issues

183

in secondary smelting

181

Sodium hydroxide use for final lead refining Sodium nitrate

206 220 206

220

Sodium sulfate from waste battery processing Sodium-iron-sulfur system

173 182

Softener slag treatment Softening

206 205

Speiss

26

St Joe Minerals process

95

157

Submerged lance reactor See Slag bath reactor Sulfamic acid electrolyte Sulfate reducing bacteria

152 174

Sulfur in blast furnace

85

in bullion from direct smelting

101

removal from battery paste

172


Index Terms

Links

Sulfur dioxide content of gas for acid production discontinuous production

61 108

efficiency of conversion to sulfuric acid

62

Sulfur drossing for copper removal

201

203

Sulfuric acid black acid

63

by-product

286

production equipment

62

recovery from batteries

171

T TBRC for lead smelting

106

secondary smelting

188

Tecnicas Reunidas

158

Tellurium

223


208

separation from lead and recovery

210

Temperature critical ISF exit gas Thallium

91 224

removal from lead Thermal refining costs

217 197 281

Thermodynamic properties of compounds

296

Tin

224 recovery in Harris process

208

removal from lead

205

Titanium anodes

158


Index Terms

Links

Top submerged lance processes for direct smelting

119

for slag fuming

140

secondary smelting

188

Toxicology of lead

249

Treatment charge

39

Tuyeres design for blast furnace

91

in the blast furnace

78

raceways

94

94

U UOP process

157

US Bureau of Mines Molten salt electrolytic cell

154

US Bureau of Mines process

157

for secondary lead

190

V Vacuum dezincing

217

Vapour pressure lead compounds

103

of lead compounds

295

Vapour pressure of zinc

131

Volatilisation of lead in direct smelting

103

W Waste heat recovery in slag fuming

139

Water quality standards for lead

255

Water jackets slag fuming furnace

135


Index Terms

Links

Williams process

214

Wohwill process

217

Z Zinc

224 in blast furnace feed

26

limits in lead concentrates

38

quality specifications

285

removal from lead

217

separation by chlorine

219

use for desilverising

211

vacuum distillation equipment

217

Zinc fuming from lead slags

130

kinetics

132

reduction equilibria

132

Zinc oxide See Fume activity in slag

131

Zinc vapour pressure

131

Zinc-silver phase diagram

211


The Extractive Metallurgy of Lead (R. J. Sinclair)

Recommend Documents