CSRP08 Proceedings Lowres.pdf#Page=49

CSRP’08Delivering sustainable solutions to the minerals and metals industries

Produced by

CENTRE FOR SUSTAINABLE RESOURCE PROCESSING

26 Dick Perry Avenue, Kensington WA 6152 Australia

2nd annual conference

18 – 19 November 2008Brisbane, Queensland, Australia

PROGRAM

TUESDAY 18 NOVEMBER 2008 (The Long Room)8:30 Welcome, OHS, Introduction and Summary of CSRP Highlights

Stevan Green (CEO, CSRP) and Erica Smyth (Independent Director, CSRP)

8:45 Keynote Speaker

Dr John Cole (General Manager, Queensland Office of Clean Energy)

09:15 [1] ZERO WASTE & MINOR ELEMENTS

Chair: Warren Bruckard (CSIRO) (Q&A to follow each session 1 presentation)

Developing SUSOP® – a structured, methodical mechanism for incorporating sustainability principles

into the design and operation of minerals processing plants

Ben McLellan (Sustainable Minerals Institute, University of Queensland)

Overview of the Early Removal Process – Small Scale Development Work

Warren Bruckard (CSIRO Minerals)

Thermodynamics of Selective Roasting of Arsenic from Tennantite Containing Copper Concentrates

Steven Wright (CSIRO Minerals)

Techno-Economic Assessment of Early Arsenic Removal from Copper Ores

Nawshad Haque (CSIRO Minerals)

10:30 morning tea

11:00 [2] CO2 BREAKTHROUGH IN METAL PRODUCTION

Chair: Sharif Jahanshahi (CSIRO)

Water Conservation in the Mining Industry – case study of gold mining and refining

Goen Ho (Murdoch University)

Overview of the CO2 Breakthrough Program and Linkage to IISI

John Mathieson (BlueScope Steel Research)

Survey of Sustainable Biomass Resources for Iron and Steel Industry

Nawshad Haque (CSIRO Minerals)

Demonstration of Recarburisation of Liquid Steel Using Renewable Carbon at OneSteel

Michael Somerville (CSIRO Minerals)

Piloting the Integrated Dry Granulation and Heat Recovery Process at CSIRO

Dongsheng Xie (CSIRO Minerals)

Value of the Projects – Industry Perspective

Phillip Ridgeway (OneSteel)

Panel Discussion (All session 2 speakers available for Q&A)

12:40 networking lunch

1:30 [3a] ENERGY EFFICIENT LIBERATION & COMMINUTION

Chair: Malcolm Powell (University of Queensland)

Minerals, Metals and Sustainability – a text book for earth resources graduates of the future

John Rankin (CSIRO Minerals)

Demand-Supply Interaction on Future Mining Resource Production: The Coal Model

Geoffrey Evans (University of Newcastle)

Overview of Eco Efficient Liberation

Malcolm Powell (JKMRC, University of Queensland)

Energy Efficient Circuits and Progressive Liberation

Zeljka Pokrajcic (PhD Student, JKMRC, University of Queensland)

Fair Measures of Energy Usage

Fiesal Musa (PhD Student, JKMRC, University of Queensland)

Improving Grinding Efficiency with the IsaMill™

Rob Morrison (JKMRC, University of Queensland)

2:45 afternoon tea

3:00 [3b] ENERGY EFFICIENT LIBERATION & COMMINUTION

Chair: Malcolm Powell (University of Queensland)

High Pressure Grinding Rolls Multiple Pass

Marko Hilden (JKMRC, University of Queensland)

Update on the JK Rotary Breakage Tester

Toni Kojovic (JKMRC, University of Queensland)

Improvement of Energy Efficiency of Rock Comminution through Reduction of Thermal Losses

Nenad Djordjevic (JKMRC, University of Queensland)

Towards a Virtual Comminution Machine

Rob Morrison (JKMRC, University of Queensland)

The New Energy Logging from the Discrete Element Method

Nirmal Weerasekara (JKMRC, University of Queensland)

Progress on the Unified Comminution Model

Malcolm Powell (JKMRC, University of Queensland)


4:30 end presentation sessions

4:35 [4] PARTICIPANTS’ FORUM (Lady Thiess Room)

Chair: Philip Bangerter (Hatch)

The purpose of the Participants’ Forum is to provide a forum for consultation among the CSRP Participants.

7:00 Pre-dinner drinks and canapés served on the Riverside Terrace, Customs House

7:30 Conference Dinner in The Long Room, Customs House

WEDNESDAY 19 NOVEMBER 2008 (The Long Room)

08:30 Welcome and Information on CSRP2

Stevan Green (CEO, CSRP)

09:00 [5] INDUSTRY PANEL SESSION

Chair: Erica Smyth (Director, CSRP)

This is an opportunity for our industry participants to tell us what we can do to better serve their needs and toprovide an opportunity to respond to day 1 of the conference.

10:00 morning tea

10:30 [6] GEOPOLYMERS

Chair: Arie van Riessen (Curtin University of Technology) (Q&A to follow each session 6 presentation)

Australian Life Cycle Initiative (AusLCI) and CSRP database: Australian data

Damien Giurco (Institute for Sustainable Futures, University of Technology Sydney)

Highly Sensitive Quantitative Phase Analysis of Fly Ashes for Use as Geopolymer Source Materials

Ross Williams (Centre for Materials Research, Curtin University)

Synthesis Paths and Performance Evaluation of Geopolymer Binder Systems Derived from Major

Mineral Processing and Mining Waste Feedstock Materials

Kwesi Sagoe-Crentsil (CSIRO Materials Science & Engineering)

Managing Coal-Fired Power Station Solid By-Products

Terry Gourley (Geopolymer Alliance)

Thermal Character of Geopolymers Synthesised from Class F Fly Ash Containing High

Concentrations of Iron and -Quartz

William Rickard (Centre for Materials Research, Curtin University)

Rational Utilisation of Fly Ash for Geopolymer Processing

Temuujin Jadambaa (Centre for Materials Research, Curtin University)

12:15 networking lunch

1:15 [7] EDUCATION

Dan Churach (CSRP) and Jim Avraamides (CSRP)

1:45 [8] BAUXITE RESIDUE

Chair: Evan Jamieson (Alcoa)

Development of a Sustainability Roadmap for the Kwinana Industrial Area (WA)

Karin Schianetz (Centre of Excellence in Cleaner Production, Curtin University)

Production and Application of Red Sand™ as ReSand®

The Role of CSRP in Facilitation of a Sustainability Project

Evan Jamieson (Alcoa)

Understanding Bauxite Residue from the Community’s Perspective – A review of a community

consultation in the Peel-Harvey Catchment

Catherine Pattenden (Sustainable Minerals Institute, University of Queensland)

Using LCA to Compare the Environmental Performance of Bauxite Residue, Lime and Bauxsol™ in

the Treatment of Acid Mine Drainage at Mt Morgan Mine

Daniel Tuazon (University of Queensland)


3:00 Conference Close

Stevan Green (CEO, CSRP)

CONTENTS

Developing SUSOP® – A structured, methodical mechanism for incorporating sustainability principles into

the design and operation of minerals processing plants......................................................................................... 6

Thermodynamics of Selective Roasting of Arsenic from Tennantite Containing Copper Concentrates ........... 8

Techno-Economic Evaluation of Early Arsenic Removal from Copper Ores ................................................... 11

Water Conservation in the Mining Industry – Case study of gold mining and refining.................................... 13

Overview of the CO2 Breakthrough Program and Linkage to IISI .................................................................... 17

Survey of Sustainable Biomass Resources for the Iron and Steel Industry........................................................ 21

Production of Charcoal for Recarburisation Plant Trials..................................................................................... 23

Piloting the Integrated Dry Granulation and Heat Recovery Process at CSIRO ............................................... 26

Minerals, Metals and Sustainability – A book for the future .............................................................................. 28

Demand-Supply Interaction on Future Mining Resource Production: The coal model .................................... 30

Applying DEM Outputs to the Unified Comminution Model ............................................................................ 32

Energy Efficient Comminution Circuits – A modified grinding strategy and the selection of a target product

size........................................................................................................................................................................... 34

Improving Grinding Efficiency with the IsaMill™ ............................................................................................. 38

Multiple-pass High Pressure Grinding Rolls Circuits ......................................................................................... 40

Update on the JKRBT (JKMRC Rotary Breakage Tester).................................................................................. 42

Improvement of Energy Efficiency of Rock Comminution through Reduction of Thermal Losses ................ 46

Towards a Virtual Comminution Machine ........................................................................................................... 48

The New Energy Logging from the Discrete Element Method .......................................................................... 50

Australian Life Cycle Initiative (AusLCI) and CSRP Database: Australian Data............................................. 52

Quantitative Phase Analysis of Fly Ashes for Use as Geopolymer Source Materials....................................... 54

Synthesis Paths and Performance Evaluation of Geopolymer Binder Systems Derived from Major Mineral

Processing and Mining Waste Feedstock Materials............................................................................................. 56

Managing Coal-Fired Power Station Solid By-Products ..................................................................................... 58

Thermal Character of Geopolymers Synthesised from Class F Fly Ash Containing High Concentrations of

Iron and -Quartz ................................................................................................................................................... 60

Rational Utilisation of Fly Ash for Geopolymer Processing............................................................................... 62

Sustainability Roadmap for the Kwinana Industrial Area / Extended Abstract................................................. 65

Production and Application of Red SandTM

......................................................................................................... 67

Using Life Cycle Assessment to Compare the Environmental Performance of Bauxite Residue, Lime and

Bauxsol™ in the Treatment of Acid Mine Drainage at Mount Morgan Mine................................................... 70

Developing SUSOP® – A structured, methodical mechanism for incorporating

sustainability principles into the design and operation of minerals processing plants

B. McLellan and G. Corder

University of Queensland, Sustainable Minerals Institute, Centre for Social Responsibility in Mining, QLD

Introduction, Background and Objectives

The current key activity of the CSRP’s

Sustainable Development (SD) Program is the

development of the SUSOP® (SUStainable

OPerations) mechanism for incorporating SD

principles into the design and operation of

industrial processing plants. SUSOP® will fill a

major gap in the current industry approach to SD

integration in design and operation, by providing a

comprehensive and rigorous mechanism to:

• Generate feasible plant design and operating

options according to SD principles.

• Evaluate the sustainability benefits and

impacts of these options.

• Assess each option using an SD-based

decision-support process.

• Support the engineering, project management

and subsequent operations of projects.

SUSOP® aims to:

• Guide the users to the tools they should apply

to evaluate the sustainability impact of their

operations.

• Demonstrate the systematic links between

activities across the project or production

cycle.

• Provide a framework of SD principles to

guide possible project and production options.

• Generate an SD Balance Sheet that illustrates

the benefits and impacts of each option on a

chosen sustainability framework.

• Use the SD Balance Sheet to determine the

impact on over-arching SD principles.

• Incorporate a dynamic element to predict the

sustainability impacts of each option under

future scenarios.

Methodology

The development of SUSOP® draws on three

strands of research:

• Examination of literature on the fundamentals

and available SD tools to clarify the state of

play and identify gaps.

• Case study development with industrial sites

and consulting companies.

• Development of in-house databases and

technology assessment tools to support

SUSOP®.

Key Results

The key result of the work so far has been the

development of the SUSOP® concept, based on

the collaborative efforts of the CSRP SD program

working group. The acknowledgement that the

processing industries include a vast array of

activities with flow-on effects from pre-feasibility

and planning through to remediation and

restoration have led to the first key element of

SUSOP® - entry points across the project and

production cycles. Each industry user can locate

their activity within one of the boxes in Figure 1.

From that starting point, users are guided through

the following process (Figure 2), and will be

directed to key tools and relevant information at

each step.

Figure 1: Multiple entry points on the project / production

cycle (after Allen et al., 1997).

The SD Opportunity Assessment Process for

identification and evaluation of sustainable

development (SD) opportunities is the most

intensive part of SUSOP® and requires the most

input from the project proponent or operator.

Figure 2: SUSOP® concept diagram.

The initial step, Familiarisation, will ensure the

users and relevant stakeholders are aware of the

meaning of sustainable development and

sustainability in the context of their operation.

Achieving this level of understanding may involve

workshops or training in SUSOP® and SD

concepts.

Following the Familiarisation step, SUSOP®

will

guide users through Goal Scoping in an SD and

specific operational sense (i.e. What is the goal of

the project operationally, and what are the key SD

concerns?). Identifying the goals and context

enables the SUSOP® study to be developed

according to an agreed scope.

The Identification step takes the goals and

context, and examines the operation in an effort to

develop options for improving SD performance.

Screening tools and workshops are some of the

elements that can be applied in this step.

Evaluation then guides the user to a selection of

tools or approaches with which to quantify and

rank these opportunities. Examples include life

cycle assessment, ecological footprint, social

assessment, industry ecology, green engineering

etc. A range of approaches will be required for a

comprehensive analysis. An important part of the

research is to identify additional tools that don’t

currently exist. Development of these tools will

form part of the research agenda for CSRP 2.

The outputs from this Identification and

Evaluation process then feed information back

into the selected Sustainable Development

Framework. Frameworks such as the 5 Capitals

Model, corporate frameworks or Triple Bottom

Line are all equally valid for use in SUSOP®.

These results will be used to generate an SD

Balance Sheet indicating the benefits and impacts

of the operation or selected alternatives.

The information from the SD Balance Sheet,

combined with Conventional Business Case

data, can then be used in the Decision Support

using tools such as Multi-Criteria Analysis, to

produce a Decision based on the incorporation of

agreed SD principles.

Monitoring and Evaluation is the ongoing

review of the selected options, once implemented.

This step ensures that the expected SD benefits

are achieved and any negative impacts are

minimised. SUSOP® will be able to be applied

periodically or at each stage of development in

order to identify emerging sustainability

opportunities.

Highlights / Benefits

• Development of the SUSOP® concept that

incorporates a multiple-entry points approach

across the project and production cycles and

utilises a known or customised SD framework

which is then the basis for generating an SD-

based decision.

Conclusions and Future Direction

The SUSOP® concept is complete. Industry-based

case studies are and will be conducted leading to

the publication of SUSOP® Mk I, a formal

documentation of the concept and process, which

is due for completion in 2010.

Acknowledgements

This work was carried out under the auspices of

the Centre for Sustainable Resource Processing,

which is established and supported under the

Australian Government’s Cooperative Research

Centres Program. The authors also acknowledge

the key contributions of the SD Program working

group.

References

1. B.C. McLellan and G.D. Corder “SUSOP® Concept Document”. CSRP Report. July 2008.

2. Allen, D., F. Consoli, et al. (1997). Public policy

applications of Life-Cycle Assessment. SETAC.

Thermodynamics of Selective Roasting of Arsenic from Tennantite Containing

Copper Concentrates

S. Wright, W. Bruckard and F. Jorgensen

CSIRO, Minerals Down Under National Research Flagship, VIC

Background

The CSRP project 2D8 (Early removal and safe

disposal of arsenic and other minor elements

during base metal processing) examined the

selective removal of arsenic from a copper

concentrate and its safe disposal through

alternative processes. While the concept was

developed several years ago, it was only exposed

and evaluated recently (Jahanshahi et al., 2006).

The proposed approach is applicable to ores

where the mineralogy allows separation by

flotation into high and low arsenic (As) streams.

A sample of high As ore from an NSW mine was

chosen as the raw material to demonstrate the

overall treatment at small scale. Differential

flotation produced two copper concentrates

(Bruckard and Davey, 2007). The high As

fraction required further treatment to produce a

material suitable for blending with the low As

fraction for sale. Roasting has potential to remove

the As from the high As stream and produce a

low-volume fume for disposal (Jorgensen, et al.,

2007). This is a major benefit as dispersion of As

into several streams during smelting is minimised.

Roasting can be performed at the

mine/concentrator, away from urban centres (see

Jahanshahi et al., 2006 for more details).

Encapsulation of the fume in geopolymers, (Brew

and Vance, 2007) or other materials such as slags

and concrete with high resistance to leaching

provide an outlet for the fume in the mine as an

engineering material as well a means of safe

disposal. This paper investigates the conditions

needed to selectively remove As from concentrate

through a roasting step. The approach taken

involves use of thermodynamic free energy

minimisation to simulate the high temperature

reactions taking place during the “selective

roasting” of high As copper concentrate.

Roasting Simulations

The simulations were performed with the CSIRO

Thermochemistry system (Turnbull and Wadsley,

1988). Composition of the simulated concentrate

is given in Table 1. The thermodynamic data of

all relevant As species was included in the system

definition, as well as the potential reaction

products of the concentrate. The thermodynamic

data for Tennatite was taken from the literature

(Seal et al., 1990).

Table 1: Input species used in simulating roasting in a

neutral atmosphere.

Species Mass (g) Species Mass (g)

N2 gas 100.0

Cu10Fe2As4S13 11.2 MgO 1.2

Cu5FeS4 20.2 SiO2 39.0

CuFeS2 16.4 Al2O3 8.9

FeS2 0.27 CaO 2.4

The effects of various initial conditions were

evaluated at temperatures between 600 to 800°C

with the focus on the As deportment and the

stability of the As containing phases. The results

of various scenarios follow.

Sulphur Dioxide and Temperature

The distribution of As to the gas phase is shown in

Figure 1 at temperatures between 600°C and

800°C for gas atmospheres containing up to 4%

sulphur dioxide. In a neutral atmosphere (SO2

free), the removal of As approaches 100% at

600°C. However as the SO2 pressure increases,

complex behaviour with respect to the As

distribution is observed. This can be attributed to

the relative stability of solid calcium arsenate

(Ca3As2O8), calcium sulphide (CaS) and calcium

sulphate (CaSO4). Removal of As from the

concentrate approaches 100% in up to 2% SO2

atmospheres at 650, 700 and 750°C.

0

20

40

60

80

100

0 0.01 0.02 0.03 0.04

SO2 partial pressure (atm)

As

dis

trib

ute

d t

o G

as

(%

)

600

650

700

750

800

Figure 1: Effect of SO2 partial pressure on the selective

removal of As from Cu concentrate.

The stability of calcium phases at 800°C is shown

in Figure 2. At low SO2 partial pressures, CaS is

the dominant phase, but its proportion is reduced

as the SO2 partial pressure increases, and the

proportion of Ca3As2O8 increases. Consequently

less As distributes to the gas phase. Above 4%

SO2 in the gas, CaSO4 becomes stable relative to

CaS and Ca3As2O8 and the As deporting to the gas

increases. This trend is also repeated at

temperatures below 800°C, e.g. Figure 3 shows

the stability range at 650°C. As the temperature

decreases, the pSO2 where CaSO4 becomes stable

decreases, and is less than 1% at 650°C.

0

20

40

60

80

100

0 0.01 0.02 0.03 0.04 0.05

pSO2 (atm)

Ca

ph

as

e d

istr

ibu

tio

n (

%)

Ca3As2O5

CaS

800°C

CaSO4

Figure 2: Distribution of calcium phases at 800°C as a

function of SO2 partial pressure.

0

20

40

60

80

100

0 0.002 0.004 0.006 0.008

pSO2 (atm)

Ca

ph

as

e d

istr

ibu

tio

n (

%)

Ca3As2O5

CaS

650°C CaSO4

Figure 3: Distribution of calcium phases at 650°C as a

function of the SO2 partial pressure.

Partial Roasting with Air

Addition of oxygen (O2) to partially roast the

concentrate yielded different results than

equilibrating a N2-SO2 gas with the concentrate.

In Figure 4 sufficient O2 was added to N2 to

produce SO2 partial pressures matching those

shown in Figure 1. At the same partial pressures

of SO2 the arsenic removal from the concentrate

was lower with the N2/O2 input than in the N2/SO2

input. The equilibrium sulphur pressure (pS2) is

lower in the case of addition of N2 and O2, and

consequently the equilibrium vapour pressure of

As in the gas was lower. If too much air added so

that the pSO2 is sufficiently high, then magnetite

(Fe3O4) and ferric arsentate (Fe3As2O8) can be

formed.

0

20

40

60

80

100

0 0.005 0.01 0.015 0.02

SO2 partial pressure (atm)

As

dis

trib

ute

d t

o G

as

(%

)

700°C

N2 + SO2

N2 + O2

Figure 4: Effect of oxygen addition at 700°C (plotted as a

function of pSO2) on the separation of As.

Addition of Pyrite

Addition of a sulphidant to increase pS2 increased

As separation. Under conditions where the As

recovery would normally be low (700°C,

pSO2=0.011 and 800°C, pSO2 = 0.0029) addition

of pyrite (FeS2) increased the removal of arsenic

(Fig. 5).

0

10

20

30

40

50

60

70

80

90

100

0 5 10 15

FeS2 added (g/100g)

Ga

se

ou

s A

s d

ep

ortm

en

t (%

)

700°C

800°C

Figure 5: Effect of FeS2 added to concentrate on the

recovery of As.

Increasing the sulphur partial pressure increased

the stability of CaS. At 700°C maximum removal

of As was achieved by adding 11 g of FeS2 per

100 g of dry concentrate. At 800°C, addition of

25 g of FeS2 increases the Arsenic deportment to

the gas phase to 100%.

Conclusions

The proportion of arsenic reporting to the gas

phase is related to the temperature and the

thermodynamic stability of Ca3As2O8, CaS and

CaSO4. Roasting conditions exist at temperatures

between 600 and 700°C where arsenic can be

removed from an arsenic bearing concentrate by

using an a neutral (N2) atmosphere or an

atmosphere with a low concentration of SO2.

Partial oxidation of the concentrate is not advised,

as the simulations show that the sulphur potential

of the system is lower than the equivalent N2/SO2

gas mixture and that magnetite (Fe3O4) can be

formed. If too much air is added, ferric arsenate

(Fe3As2O8) becomes stable and arsenic is retained

in the concentrate.

Addition of FeS2 to the concentrate at

temperatures above 700°C increases the sulphur

partial pressure and enhances the stability of CaS

compared to that of Ca3As2O8. This increases As

removal.

There are several potential thermodynamic

operating windows to remove As from an As-rich

copper concentrate. The most appropriate will

depend on the rate at which As can be volatilised

from the concentrate, and the conditions where the

structure of the concentrate and the fuel value of

the concentrate are not compromised.

Acknowledgements

This work was carried out under the auspice of the

Centre for Sustainable Resource Processing,



Centres Program. This project was financially

supported by CSIRO and CSRP.

References

1. Bruckard W.J. and Davey K.J., 2007. Arsenic

Separation in Copper circuits Using Flotation, First CSRP Annual Conference, Melbourne, 43-44.

2. Brew D.R.M. and Vance E.R., 2007. Arsenic

Remediation Using Geopolymers, First CSRP Annual Conference, Melbourne, 47-48.

3. Jahanshahi S., Bruckard W.J., Chen C. and Jorgensen

F.R.A., 2006. Management of Minor Elements in the

Production of Base Metals, Green Processing 2006, Newcastle, 25-30, AusIMM, Parkville.

4. Jorgensen F.R.A., Hall T.P. and Sanetsis S., 2007.

Selective Removal of Arsenic from a Copper

Concentrate, First CSRP Annual Conference, Melbourne, 45-46.

5. Seal, R.R., Essene, E.J., Kelley, W.C., 1990. Canadian Mineralogist, Vol 28, pp. 725-738.

6. Turnbull, A,G. and Wadsley, M.W., 1988, The CSIRO

Thermochemistry System, CSIRO Division of Mineral Products.

Techno-Economic Evaluation of Early Arsenic Removal from Copper Ores

N. Haque and T. Norgate

CSIRO Minerals, VIC

Introduction

There are significant amount of undeveloped

copper ore deposits containing arsenic minerals

(e.g. enargite and tennantite) which can produce

copper concentrates but these are high in arsenic

(0.7%) (Smith and Bruckard, 2007). Arsenic

contents above 0.5% are generally not acceptable

to smelters. Recently, a new process has been

proposed at CSIRO Minerals to separate high

arsenic copper minerals from other copper

minerals in flotation (Jahanshahi et al., 2006;

Senior et al., 2006) based on exploiting

differences in pulp potential and pH. This

technology involves a two-stage flotation process

accompanied by an additional roasting stage.

There are less environmental impacts with this

new process because of reduced arsenic, sulphur

and SO2 emissions to the land and air. However,

the new proposed process involves additional

capital (roaster and additional flotation units) and

operating costs. The aim of this techno-economic

evaluation study is to provide first estimates of

any additional benefits from the new flowsheet

compared with conventional copper processing

technology and to determine whether these

additional benefits justify the additional costs of

the process.

Methodology / Technique

Process modelling of the conventional (base case)

flowsheet and the new proposed flowsheet was

undertaken using the METSIM process simulation

software. The mass and energy balances obtained

from the METSIM simulations of the respective

flowsheets were used to estimate the operating

costs of both processes. Several assumptions

were made in estimating these costs as outlined

below.

The roaster is assumed to operate at 700°C in a

nitrogen atmosphere and the reaction time in the

roaster is 6 min (Padilla et al., 2001). The roaster

fuel energy requirement was estimated to be 0.27

GJ/t of concentrate, while the electrical energy

consumption was calculated to be 1.5 kW based

on the estimated size (5.2 m long, 1.7 m diameter)

of the rotary kiln for the required feed rate.

Key Results / Findings

The estimated operating cost of the proposed

process was approximately 17% higher ($76/t

conc. cf. $65/t conc.) than the conventional

process. The contributions to the operating cost

of both processes are compared in Figure 1. The

cost of labour was the major contributor followed

by electricity and reagents.

0

10

20

30

40

50

60

70

80

90

Rea

gent

s

Ele

ctric

ity

Wate

r

Labo

ur

Cons

& m

ain

t

Fue

l

As stab

ilisa

tion

Tot

alO

pera

tin

g c

ost

($/t

co

ncen

trate

)

Base case

New process

Figure 1: Comparison of operating cost contributions for

base case and new process.

Estimated revenue for both processes is shown in

Table 1. The price received for the final copper

concentrates from both flowsheets was based on

the copper content (96.7% payable – Mining

Journal, December 2006), and an average copper

price over 2006/2007 of US$7000/t (Mining

Journal, November 2007).

Table 1: Revenue from copper concentrate for both flowsheets.

Parameter Base case New process

Ore (t/h) 1000 1000

Concentrate (t/h) 32.8 34.1

Concentrate grade

- % Cu

- % As

26.4

0.53

25.4

0.08

Cu revenue

- $/t conc

- $/t ore

2234

73.20

2149

73.33

As penalty

- $/t conc

- $/t ore

12.53

0.41

0.00

0.00

Net revenue ($/t ore) 72.79 73.33

Economic benefits of proposed process

Based on the economic results (and underlying

assumptions), the net revenue derived from the

new proposed process over the base case

conventional process was estimated to be $0.54/t

ore. The increased operating cost of the new

process over the conventional process was

estimated to be $0.46/t ore, giving an overall

increase in revenue of $0.08/t ore. However, this

analysis does not take into account the increased

capital cost (additional flotation cells, roaster and

associated equipment). For the assumed base

case nominal flow rate of 1000 t/h ore federate

(approximately 8.0 Mtpa at 92% plant

availability), the additional revenue received after

taking the increased operating costs into account

corresponds to about $0.6 M/y.

Effect of roaster arsenic removal efficiency

The As in the final concentrate product increased

by a factor of about three (2388 ppm cf. 777 ppm)

at 60% efficiency compared with 90% efficiency.

Thus a roaster efficiency of only 60% would see

an arsenic penalty imposed on the final

concentrate of $1.46/t concentrate (or $0.05/t ore),

which reduces the net revenue from $0.54/t ore to

$0.49/t ore, still slightly above the increase in

operating cost ($0.46/t ore).


Benefits of the proposed new flowsheet include

the following:

• Potential utilisation of underdeveloped copper

ore deposits that contain otherwise

unacceptable levels of arsenic.

• Processing of this ore using the proposed

process is likely to be economically viable.

• The penalty for arsenic content (i.e. US$3 per

0.1% As content above 0.2%, averaged from

various sources) in the copper concentrate

product is significantly reduced due to the

reduction of arsenic using the proposed

process.

• If the smelter cannot accept copper

concentrates with arsenic content above 5000

ppm, the economics of the new process

improves significantly since the proposed

process is the only option for utilisation of this

ore.

• Removal of arsenic from the flotation circuit

during processing, stabilisation and safe

storage or utilisation would yield additional

benefits.

• Reduction of sulphur and SO2 emission from

the smelter.

• It should also be noted that smelter limits for

As are projected to become more stringent in

the future which will further enhance the

benefits of this new approach.


This evaluation showed that the yield of

concentrate will be slightly higher with the

proposed process compared with the conventional

(base case) process. This resulted in an increase

in net revenue (copper revenue less arsenic

penalty) of $0.54/t ore with the new process.

However, it should be emphasised that this

evaluation was based on the premise that the

conventional process was capable of producing a

final concentrate with an arsenic content low

enough to make it acceptable to smelters. If no

acceptable concentrate can be produced from the

ore by conventional processing, meaning a high

arsenic ore body cannot be mined and processed,

then the economic benefits of the new proposed

flowsheet increase dramatically, with potential

environmental benefits as a bonus. These first

estimates of the economic benefits of the new

proposed process are sufficiently encouraging to

warrant further development of the flowsheet.

Acknowledgements

This work was carried out with financial support

from CSIRO and the Centre for Sustainable

Resource Processing, which is established and

supported under the Australian Government’s

Cooperative Research Centres Program. The

authors gratefully acknowledge Warren Bruckard,

Kevin Davey and Frank Jorgensen from CSIRO

Minerals for their assistance during this study.

References

1. S. Jahanshahi, W.J. Bruckard, C. Chen and F.R.A.

Jorgensen “Management of minor elements in the

production of base metals”. In: Green Processing 2006 –

3rd International Conference on the Sustainable

Processing of Minerals, 5-6 June, Newcastle, Australia,

AusIMM & CSIRO. pp 25-30. 2006.

2. R. Padilla, Y. Fan and I. Wilkomirsky “Decomposition

of enargite in nitrogen atmosphere”. Canadian Metallurgical Quarterly 2001, 40(3): 335-342.

3. G.D. Senior, P.J. Guy and W.J. Bruckard “The selective

floatation of enargite from other copper minerals – a

single mineral study in relation to beneficiation of the

Tampakan deposit in the Philippines”. International

Journal of Mineral Processing 2006, 81:15-26.

4. L.K. Smith, and W.J. Bruckard “The separation of

arsenic from copper in a Northparkes copper-gold ore

using controlled-potential flotation”. International

Journal of Mineral Processing 2007, 84:15-24.

Water Conservation in the Mining Industry – Case study of gold mining and refining

R. Cocks , G. Ho1, S. Dallas

1, M. Anda

1and Aidan Giblett

2(industry supervisor)

1Murdoch University, Environmental Technology Centre, WA

2Newmont, WA

Introduction

Water is an essential input to mining and mineral

processing and demand is increasing due to

growing activities. Climate change has resulted in

reduced water availability and this is the case in

arid parts of Australia, where many mining and

mineral processing operations take place.

Governments in Australia now require major

water users to prepare a Water (Efficiency)

Management Plan to achieve reduction in water

consumption. Water auditing is now a well

established tool for conducting a systematic,

structured and rigorous water use performance

assessment that includes costed options for water

conservation. There are benefits resulting from an

industry conducting a water audit that is not only

due to the savings in water costs, but also the

heat/energy and materials dissolved or suspended

in the water. Reduction in heated process water

results in reduction of not only water volume, but

also the cost of heating the water. Reduction in

water volume may also mean reduction in the cost

of pumping the water. Water conservation

includes reduction in the amount of pollutants in

water streams or load of pollutants. Final disposal

of the water will impact less on the environment.

If the pollutants are nutrients (Nitrogen and

phosphorus) the water will be suitable for

landscaping. Minimising final disposal of

wastewater is an objective of water auditing and

water conservation. Zero discharge of stormwater

is a requirement in many cases at mining sites,

and zero emission of contaminated water may be

feasible and an objective to be considered as part

of achieving sustainable development.

Water Auditing and Water Conservation

The water audit process commences with a review

of the audit scope and objectives. Once the

domain of the audit has been agreed, water flows

into and out of the domain are determined and

quantity and quality are measured. Water auditing

is different to environmental auditing ISO 14000

in that quantifying and qualifying water use for

the purposes of auditing water is only covered in

part by ISO 14010 and 14011, and the emphasis

of environmental auditing is compliance with

objectives. The water auditing process will, on the

other hand, provide the objectives for

environmental auditing by recommending or

specifying objectives for achieving water

conservation, that is the water conservation

strategy (Sturman et al., 2004).

In the water audit process water and material

balances are fully carried out in a systematic

manner. Once the scope and objectives of the

water audit have been established a schedule is

drawn up and may include a selection list of the

people that would contribute to, and or be part of

a water audit team. Resources are also tabulated in

the context of availability, and resources that may

be required and would need to be introduced into

the audit process, and for example the need for a

clamp-on Acoustic Doppler Velocimeter for water

flow measurement and meter calibration. A

proposed water audit procedure includes many

facets and starts with a flow diagram of the water

audit domain that identifies the number of unit

operations. Once a site-wide water balance model

has been produced, recommended future changes

to operations and processes are weighted against

the financial capacity for implementing change

and the economic justification to do so within the

scope of water conservation. The collation of data

from an initial site visit underpins a

methodological approach towards conducting a

water audit and achieving a foreseeable outcome

in terms of audit closure. Closure refers to the

difference between input and output flows or

material fluxes, and usually a maximum 10 %

difference is an objective. This also integrates

with the resources on hand and the ability of the

mining company to provide additional support for

ongoing measurement and assessment of the water

balance account.

Application to Gold Mining and Refining

Newmont The Gold Company™ owns and

operates the Waihi Goldmine in New Zealand’s

North Island, the Tanami Granites Operation in

the Northern Territory and the Jundee-Nimary

Mine in the Little Sandy Desert north of

Kalgoorlie, WA. Joint venture mines include the

Fimiston-Gidji operation run by Kalgoorlie

Consolidated Gold Mine (KCGM) in a 50-50

share arrangement with Barrick Gold Australia

and the Boddington Gold Mine where 66.67 % is

owned by Newmont and 33.33 % by AngloGold

Ashanti. An approach to sustainable sourcing, use

and reuse of water was met through the use of the

water auditing tool.

Water balance models have been designed for

each case study and in turn set a framework for

water auditing exercises to be carried out in

ongoing research. Historically, Newmont has

audited water use in the refining operation at its

respective mine sites. However, a general audit of

a site-wide water account has yet to be

maintained. A possible exception is the Waihi

Gold Mine’s compliance with New Zealand’s

Resource Management Act 1991(RMA) that sets

out a series of amendments pertaining to mines

using land, beds of lakes and rivers, the use of

water and water quality on discharge or seepage

from mining operations. To comply with water

quality conditions set by The RMA’s Resource

Consents and Water Rights Bill, comprehensive

routine monitoring of flow rates and water quality

is carried out. With regard to process water

monitoring, the Tanami Granites Operation is by

far the most efficient water-user over its gold mill

and personnel accommodation domain.

Furthermore, several conclusive water audit

exercises have been conducted in 2001 and 2003

at the Granites. The recorded results bear

testament to the validity of conserving water and

accounting for every kilolitre used in gold-ore

processing. Therefore, the question is raised as to

how much raw water will be conserved by

accounting for all water uses and establishing

closure in a site-wide water audit at any of

Newmont’s gold mine sites. An overall objective

of carrying out a water audit is to arrive at a water

management strategy for each mine site.

Newmont’s water data gathering is

comprehensive both with meteoric water flows

that are described as non-process water stocks and

process water flows used in all aspects of gold

extraction from pits and declines through to

crushing/milling of gold ore and final gold

recovery. At each mine site water data were

collated to form an overall picture of various

water sources, flow rates and end uses.

Information gathering included contacting staff to

retrieve data and opening archival IT systems

where historical data could be obtained. As a

general rule the metallurgical departments were

the greatest source of information followed by

process operations staff and bore/dewatering

personnel. Data acquisition workbooks displayed

an annual consumption rate of raw and recycled

water stocks and excel spreadsheets produced

daily and weekly water consumption rates. From

this information water balance flow diagrams

were produced to inform a methodological

approach and further research by examining site-

wide water flows via a formal water audit

exercise. Examples of both non-process and

process water schematics are illustrated in Figure

1 and Figure 2 respectively:

Figure 1: Pre-audit site-wide water schematic of process

water sources and uses.

Figure 2: Pre-audit site-wide water schematic of non-

process water sources and flows.

Within a water management strategy, continuous

improvement of maintenance procedures support

site-operating systems and a structured

methodology monitors key performance indicators

for ongoing full compliance of, for example;

Waihi Gold Mine’s ‘Consents’ permit. Each mine

site was examined in terms of ‘fit for purpose’

water use in various operations and for example;

types of water used in dust suppression,

dewatering, site services and elution/gold

recovery. Furthermore, in non-disturbed areas of a

mine site’s domain, non-process or environmental

water flows are examined. This methodology aims

to build in water quality to the pre-audit

conditions of achieving audit closure and the

inclusion of a hierarchical table of water quality

and uses (Figure 4).

Results

Figure 3 shows the amount of raw water used per

ounce of gold produced at each mine site. The

figures are a reflection of raw water used for

crushing and milling ore and the amount of gold

produced in any one of the gold recovery

operations. Factors affecting the amount of raw

water used include the amount of gold produced

per tonne of processed ore, where water use

increases with a lower ratio of recoverable gold to

ore tonnage. Other factors not tabled include the

geochemistry of ore relating to rock types and

processes required to recover gold from differing

geological matrices. An example of the difficulty

of gold recovery from a certain ore type is

experienced at KCGM where the high sulphide

content in the Fimiston Pit ore requires further

processing at the Gidji Roaster plant and using

more water to recover the gold from an extra

processing cycle. Therefore, the comparisons are

only an indication of water use efficiencies that

will assist in benchmarking and building an over

all framework of a mine site water account and

ultimate water audit closure. Even though KCGM

uses more water than the other mine sites, a total

water audit of the KCGM domain would lead to

further efficiencies in its gold operations and

water uses.

Figure 3: Comparison of water efficiencies.

The future direction with NZ’s RMA ’91 is that

management of non-process water diversion,

collection and treatment is central to the effective

control of environmental conditions over Waihi

Gold Mine’s leased area. Typically, sites with an

effective water management strategy maintain a

water balance model as an effective tool for

informed risk based decision making and this

applies to all Newmont mine sites.

The National Water Initiative underpins the

implementation of water use strategies at state

levels. Western Australia’s State Water Allocation

Strategy advocates the use of water metering at all

stages and classifications of water used in the

operation of, for example; any mining venture.

Within the strategy, a new metering program is

proposed and may be rolled out on a priority basis

with a view that comprehensive metering will

inform government agencies on the draw of water

from the ‘consumptive pool’ of a region.

To achieve reliable, cost-effective and efficient

performance in water extraction and use, mine

sites will establish a rigorous basis for water

resource management and trading in order to

provide the best available monitoring and

measurement information on water availability

and use. This follows on from the State’s Rights

in Water and Irrigation (Approved Meters) Order

of 2003 and is relevant to the standards,

technologies and operating practices under review

by the National Water Commission. User-

compliance

with water access entitlement will be monitored

where responsibility lies with user (mine sites)

water meter reporting. The water classification

table sets the basis for a generic application of a

hierarchical approach to a water management

strategy and water audit process.

References

1. Sturman, J., Ho, G., Mathew, K. 2004. Water Auditing and Water Conservation. London: IWA Publishing.

Water source TDS Mine usage Approx. usage Classification

Scheme < 1,500 mg/L Accom/ops < 1% Potable

Bore potable < 1,500 mg/L Accom/ops

RO – elution

< 5% Fresh

Bore brackish 1,500-15,000

mg/L

SAG/Ball mills < 20% Raw

Bore saline 15,000-35,000

mg/L

SAG/Ball mills < 20% Raw

Paleo-channel

Hyper-saline

> 35,000 mg/L Dust suppression Variable Raw

Tailings decant 15,000-35,000

mg/L

Sag/Ball mills 60% of raw water

inflow

Recycled

Treated wastewater 1,500-15,000

mg/L

Revegetation

Tailings dam

< 1% Recycled

Hydrocyclones Barren

elate

15,000-35,000

mg/L

Sag/Ball mills Up to 6 cycles of

re-use

Re-used

Meteoric < 1,500 mg/L Tailings dam

Water storage

Variable Raw

Figure 4: Water Classification Table.

0

500

1000

1500

2000

2500

Copper

Nicke

l

Lead

Zinc

Alum

inium

Ste

el

Glo

bal G

HG

em

issio

ns (

Mt

CO

2e / y

)

0

10

20

30

40

50

Au

str

alian

GH

G e

mis

sio

ns (

Mt

CO

2e / y

)

Global Australia

Figure 1: Australian and global annual greenhouse gasemissions for various primary metals (Norgate et al).

9

Towards Carbon Neutral Metal & Cement Production

Reduce GHG &

Limestone

System innovation builds on industrial ecology, regional synergi es,

and collaboration across normal business boundaries

VALUE

IMPACT

Wheat Power Iron & Steel Cement

Bio -

mass

CharFarming Pyrolysis

Tree

Planting

Soil Quality

Zero Waste

-Smelting

Planting

Soil Quality

(Salinity Control)

GHG NeutralRenewable

Energy

Cement

Making

SlagDry

Granulation

Power

Reduce Emission

& Water Usage

Slag

Hot

- Reducing GHG Emissions, Fresh Water Usage and Salinity

Figure 2: Conceptual flowsheet for low GHG steel and

cement production using biomass derived charcoal and drygranulation of slags.

Overview of the CO2 Breakthrough Program and Linkage to IISI

S. Jahanshahi1, J.G. Mathieson

2 and P. Ridgeway

3

1 CSIRO, Minerals Down Under National Research Flagship, VIC 2 BlueScope Steel Research, NSW 3 OneSteel, NSW

Introduction

Currently the steel industry is a moderate

contributor to global GHGs, but this ranking may

change given the initiatives on reducing GHGs by

the power generation and transport industries and

the rapid growth in demand for steel. Virgin

production typically produces around 2 t-CO2e/t-

steel it is thus imperative to identify major

opportunties for reducing the industry’s energy

consumption and GHG emissions and develop

sustainable solutions/technologies in response to

the challenges ahead.

Major Opportunities in Reducing GHG

Emissions

CSIRO life cycle assessment studies have

indicated that extraction and refining of metals

from concentrates are the most energy intensive

steps in the mineral cycle(i,ii)

, and are thus

responsible for the major portion of GHG

emissions. Taking into account the tonnage of

different metals being produced in Australia and

globally, it then becomes clear that major

opportunities for reducing GHG emissions in the

sector are in the production of steel from iron ore

and aluminium from bauxite (see Figure 1).

Through collaboration with the steel industry,

CSIRO and CSRP are involved in development

and demonstration of technologies for step-change

reductions in GHG emissions in metal extraction

and refining processes. These include:

• use of renewable carbon as fuel and reductant

• integrated heat recovery and dry granulation

of slags.

These technologies could be linked in the value

chain, as shown in Figure 2, to result in a

reduction of over 500 million tonnes per annum of

GHGs globally. It is interesting to note that the

potential reduction in GHGs resulting from

implementation of these technologies is similar to

the current level of total GHG emission by

Australia.

Renewable Fuels and Reductants

The steel industry is almost entirely dependent on

fossil carbon for fuel and reductant and

contributes significantly to GHG emissions in

Australia and globally. Processed biomass (e.g.

char) is a potential source of renewable energy

and carbon for metallurgical reactors, providing

an essentially GHG-neutral form of fuel/reductant.

By-products (e.g. bio-diesel or electricity) from

production of char from biomass can also

contribute to reduction of GHGs from industry.

The supply of biomass should (and can) be

managed in sustainable or even restorative ways

to avoid adding stress to agricultural and forestry

eco-systems. Our studies have shown that partial

replacement of fossil carbon by bio-char over the

next decade or two will create a demand for

biomass supply of up to 10 million tonnes per

year. Such volumes of biomass could

theoretically be supplied in a sustainable way

through use of large volumes of forestry and saw

mill residues currently being produced and/or

plantation of short rotation-cycle woody biomass,

such as the oil mallee that could meet the

metallurgical carbon requirements of Australia’s

steel industry. The innovative challenge is to find

viable solutions that capture the inherent

sustainability opportunities associated with the

use of biomass char and other forms of renewable,

recycled or waste carbon. Furthermore, by using

deep-rooted biomass that has been grown in

salinity-prone areas, we will also help to improve

soil quality in those areas. It is worth noting that a

plantation program of short rotation trees (oil

mallee) has already commenced in the WA wheat

belt to lower the water table and hence address the

salinity issue in this region. Full rehabilitation of

salinity-affected farmland will produce 25 million

tonnes per annum of biomass. The local and

overseas steel industries could potentially provide

an excellent market opportunity for such volumes

of renewable carbon.

In addition to the environmental benefits

associated with substitution of biomass-derived

charcoal for coal, there are also some economic

benefits. Studies by Mathieson(i)

have shown the

value-in-use of pyrolysied woody biomass could

be up to 80 per cent higher than a reference coal

used for pulverised coal injection into a modern

blast furnace. This study shows that conditions

used for pyrolysis of biomass could result in

production of a range of chars with varying value-

in-use, thus it may be possible to design and

produce semi-charcoals that optimise blast

furnace heat and mass balance. This concept of

“designer chars” is being pursued through the

International Iron & Steel Institute’s CO2

Breakthrough Program, and is focusing on the

production of chars with tuned properties for

different applications in the iron and steel making

processes. These chars are aimed at replacing a

significant portion of the fossil fuel and reductant

used by the industry. In particular, we aim to

develop technologies to replace 100 per cent of

tuyere-injected fuels, i.e. about 30 per cent of

blast furnace fuel (excluding stoves), 50 per cent

of solid fuel used in sintering of iron ore, 2 to 10

per cent of coking coal blend and 100 per cent of

carbon used for recarburiser and slag foaming

agent in steel making over the next 5 to 15 years.

Integrated Heat Recovery and Dry

Granulation of Slags

In most modern smelters, a significant portion of

high-grade waste heat is recovered from processes

through the off-gas handling systems. However,

the thermal energy that resides in the slag is lost to

the atmosphere. Capture and use of such high-

grade waste heat is technically feasible and

economically attractive. Studies have shown

large potential benefits associated with the

application of suitable technology in integrated

metallurgical plants. The challenge has been in

development of a versatile technology platform

that maximises use of by-product energy and

material. This is being addressed through

development of an integrated process for heat

recovery and conversion of the slag into a glassy

by-product suitable for cement production.

An integrated steel plant producing one million

tonnes of steel a year, also produces 300 000

tonnes of slag. The Australian steel industry

produces more than two million tonnes of slag

each year. Blast furnace slags are generally either

water-granulated to produce glassy particles that

can be used in cement production or air-cooled in

large pits for land-fill or low-value applications

such as road base materials. Steelmaking slags are

generally air-cooled and land filled or used in

road-based applications.

Since 2002, CSIRO has been involved in

development of a new technology for waste heat

recovery and dry granulation of slags. The

CSIRO process involves feeding molten slag on to

a disc rotating at high speed. This breaks the slag

into small droplets that solidify rapidly, producing

glassy granules with similar properties to those

produced by water granulation. The granules can

be used as supplementary cementitious material in

the production of cement, which is a key

constituent of concrete. Globally, about 2.3 billion

tonnes of cement is produced every year.

Producing one tonne of Portland cement

consumes about 3000 mega-joules of electrical

and thermal energy and emits about 800

kilograms of carbon dioxide. Granulated slag can

substitute for up to 70 per cent of Portland

cement, leading to significant energy and GHG

reductions.

The conventional wet granulation method

involves significant capital costs, consumes large

amounts of water, generates acid mists and does

not recover any of the valuable heat. With lower

capital cost and benefits in heat recovery and

reduced pollution, dry granulation is an attractive

alternative to conventional wet granulation. A

recent economic assessment of CSIRO’s new

integrated process identified potential annual

savings in fuel costs alone of two to three dollars

per tonne of steel, which translates to about $20

million annually for the Australian industry and

over $3 billion for the global steel industry.

Successful development and commercialisation of

the technology in Australia will result in

reductions of GHG emissions by millions of

tonnes, reduction of water consumption by

thousands of millions of litres and conversion of

millions of tonnes of by-product slag into cement.

CSIRO’s pilot-plant work has provided insights

into process design and control, and how slag

atomisation can be optimised to produce fine

granulates. The results have demonstrated how

enhanced fast cooling can improve granulate

handling and have assisted in the design of a

compact unit to reduce cost and take advantage of

heat recovery. Work is now progressing on

optimising the scale-up of the integrated process

using a combination of computational fluid

dynamics modelling and pilot-plant measurement,

product evaluation and plant measurements (e.g.

current process slag temperature off-takes) at

OneSteel and BlueScope Steel sites.

Value of the Projects – Industry

Perspective

Industry engagement is a critical success factor in

development and implementation of the emerging

technologies.

BlueScope Steel and OneSteel are the leading

firms in the Australian iron and steel industry,

employing over 20,000 people across several

hundred sites, contributing over $1.6 billion per

annum in exports, and servicing customers in the

manufacturing, infrastructure, agriculture and

building & construction sectors. Steel

manufacture accounts for approximately 3% of

Australia's annual greenhouse gas emissions. The

iron and steel manufacturing industry, which is

emissions-intensive and trade-exposed, is one of

the industry sectors that will be most affected by

the proposed Australian Carbon Pollution

Reduction Scheme (CPRS).

BlueScope Steel and OneSteel believe reducing

GHG emissions is a global problem that requires a

global response. Our companies recognise the role

of the Australian iron and steel industry, along

with other sectors of the Australian economy, and

the global industry, in transitioning to a low

emissions economy in a sustainable and

economically responsible way.

Whilst viable incremental energy and greenhouse

efficiency improvements will continue to be

sought, realistically any breakthrough iron and

steelmaking technology to radically reduce the

emission of CO2 in the iron reduction process is

likely to be medium-term (5-10 years) to long-

term (greater than 10 years) in terms of

timeframes required in new technology

development, commercial proving, and take-up in

capital lead times within companies in an industry

where assets are long-lived between major

upgrades.

Both BlueScope Steel and OneSteel are actively

supporting the dry slag granulation and heat

recovery and biomass R&D projects given their

theoretical potential to deliver some significant

GHG savings in the medium term, with some

potential within the short-term. In supporting the

technical work of CSIRO Minerals, we as a team

contribute on strategic direction, technical

application advice, industry R&D and in-kind site

trial and sample support, apart from cash funding

contributions.

In regard to the biomass project, if 10% of the

carbon consumption of the world iron and steel

industry was able to be replaced with biomass

(charcoal), this would be approximately

equivalent to abating the GHG of Australia's

national GHG inventory per annum. The

CSIRO/BlueScope/ OneSteel biomass project

reported on at this conference is a valuable set of

work where the product chain from biomass

material sourcing, transport and logistics,

pyrolysis techniques, and application is being

researched. Just as 'oils ain't oil's’, charcoals are

not all the same. Charcoals must be designed in

properties against particular applications with the

steel industry from steel recarburisation, to coke-

making additions, to sintering, to blast furnace

injection. This then required differing pyrolysis

condition, and hence a pyrolysis unit that can dial

up different charcoal properties (% moisture, %

volatiles etc). To make use of forestry and other

biomass source waste residues, this pyrolysis

unit then must consider feed variation

requirements. There are linkages, opportunities

and barriers that must be investigated between

these process chain steps in consort.

The work also seeks to bridge the 'chicken and

egg' syndrome of steelmakers waiting for biomass

availability, and potential biomass producers

waiting for steelmakers to generate a stable

market. The project seeks to fast-track R&D in

biomass uses that have potential for early

application, even if not involving large GHG

abatement, such as use as an alternative for steel

recarburiser. In this way both the steel industry

and external parties down the supply chain will be

shown practical demonstration of biomass use to

stimulate interest and support in other larger areas

of potential biomass application. It also provides

information key to other areas of work.

The first plant trial on use of bio-char in steel

making for recarburiser at OneSteel’s Sydney

Steel Mill is currently in preparation to be

undertaken with the next few months. Parallel

activities on sustainable supply of biomass and

techno-economics of the technologies and supply

chain have been commissioned through

collaborative projects involving a number of

stakeholders representing mineral and agri-

industries, local councils, state and local

governments, investors, other technology

providers and steelmaking companies.

The project on dry slag granulation and heat

recovery has also attracted interest from the

Australian steel industry, cement producers and

engineering firms. The project team at CSIRO is

currently evaluating the conceptual integrated

process through a pilot plant campaign at CSIRO

before plant trials at one of the iron making plants

in Australia. There is already very strong interest

expressed to CSIRO from parties around the

world in the CSIRO prototype work, and it

appears to have significant commercial value

if the research and development work reaches it

goals and becomes commercially viable and

proven. This project has several environmental

attributes as discussed in the conference report,

and in particular for OneSteel where traditional

wet slag granulation has not been viable in

utilising its generated blast furnace slag for

cement application, a breakthrough in this

technology is of interest. Particularly that it is dry

granulation rather than wet, is a very dry region of

Australia relying on the Murray River.

The International Iron & Steel Institute’s CO2

Breakthrough Program has recently acknowledged

the transformational nature of the biomass and dry

granulation projects and enrolled these R&D

projects as part of its worldwide portfolio of

projects. This allows engagement of a broader

range of collaborators and sponsors, and possibly

quicker technology uptake by overseas iron and

steel producers in the future.

Acknowledgements

This work was carried out under the auspices and

with the financial support of the Centre for

Sustainable Resource Processing, which is

established and supported under the Australian

Government’s Cooperative Research Centres

Program. Financial support from CSIRO, through

the Minerals Down Under National Research

Flagship, OneSteel and BlueScope Steel is also

acknowledged.

References

1. T. Norgate, S. Jahanshahi and W.J. Rankin, Assessing

the environmental impact of metal production processes, Journal of Cleaner Production, 15, pp.838-848, 2007

2. T. Norgate and S. Jahanshahi , Opportunities for

reducing energy consumption and greenhouse gas

emissions in mineral processing and metal production;

In Proceedings of the Chemeca 2007 conference, Melbourne, 2007.

3. J. G. Mathieson, The Value-in-Use of Some Biomass-

Derived Blast Furnace Injectants, BlueScope Steel

unrestricted report, BSR/N/2007/071, December 2007.

Survey of Sustainable Biomass Resources for the Iron and Steel Industry

N. Haque1, M. Somerville

1, S. Jahanshahi

1, J.G. Mathieson

2 and P. Ridgeway

3

1CSIRO Minerals, VIC 2BlueScope Steel Research, NSW 3OneSteel, NSW


A collaborative project between BlueScope Steel,

OneSteel and CSIRO through CSRP commenced

in late 2006 to identify, evaluate and demonstrate

specific opportunities where biomass or

sustainable and renewable carbon can be used in

iron making and steelmaking processes. As one

of the activities of this project, a survey of

currently available biomass resources from

several regions of Australia was carried out and is

reported here. Residues have been targeted

because of their current low value, waste disposal

issues and being readily available in some parts of

Australia. Potential waste biomass categories are:

woody biomass from forestry and wood

processing, residues from agriculture and

horticulture and biomass from woody weeds.

Figure 1: Map showing the regions for the survey of

biomass.

Methodology

The data collection was based on interviews and

meetings with personnel from the state forest

agencies and wood processing industries. Other

data were collected from the open literature,

including forest management plans and plantation

inventory data. The amount of non-forestry

resources was estimated from the production of

the agricultural and horticultural industries.

The green wood volume has been converted to an

oven-dry biomass equivalent, based on reasonable

conversion factors for harvest recovery in the

forest, greenmill recovery, drymill recovery in the

sawmill (Ximenes et al., 2005; Ximenes and

Gardner, 2006), wood density (Bootle, 2005;

Kininmonth and Whitehouse, 1991) and moisture

content (MC). Finally, the estimated amount of

residues has been reported as equivalent oven-dry

tonnes unless otherwise stated. Dry weight has

been taken as the basis because biomass achieves

this stable reference condition after drying at

105±2°C for 24 hours. Prices are currently paid

for the bone-dry weight of wood chips by the fibre

based mills and the well-established export

market.


Total dry residues from all regions were estimated

to be around 7.5 Mt per year. Forest residues (e.g.

wood chips, reject logs, out-of-specification logs,

bark, stump, branches, foliage (leaves and twigs)

and other biomass in the tree crown) are 45% of

total residues. Wood processing residues (e.g.

chips, sawdust, shavings, off-cuts) are 30% of

total residues. Non-forestry residues (e.g.

biomass from grain crops such as wheat and

maize, olive pomace, grape skin, almond waste,

cut flower waste and sugarcane based residues

(e.g. bagasse, infield cane crop residues), woody

weeds such as Camphor Laurel and waste from

macadamia nut processing) account for 25% of

the total biomass resources. Sample residues are

shown in Figure 2.

Figure 2: Photos of sample residues in the forest and during

processing.

The estimated weights of biomass residues are

shown in Table 1. Residues generated in the

forest during harvesting are largely uncommitted

Northern NSW

Central NSW

South East NSW

Southern NSWSE SA, SW Vic

Southern Vic

by the plantation owners. In NSW, the majority

of the plantations are owned by state forests,

whereas in Victoria and the GT region, plantations

are owned by both state forests and also private

companies. The electrical power generation

companies are the sector most interested in these

residues. Recently, there is some interest from the

bio-fuel companies for making bio-crude or

transport fuel.

Table 1: Consolidated annual production of biomass

residues from all regions studied.

Resource category Thousand dry tonnes

per annum

Forestry harvesting residues 3,385

Wood processing residues 2,255

Non-forestry residues 1,905

Total 7,546

Figure 3 shows the estimated requirements for

charcoal by the steel industry in Australia for

various categories of potential applications

(Langberg et al., 2007; Mathieson, 2008). In the

medium term (5-10 years), the theoretical

requirement will consume about 24% of the

charcoal making potential (2.3 Mt/yr) from the

sustainable biomass available from the regions

investigated. However, based on 50% availability

of total biomass (i.e. from 7.6 Mt/yr to 3.7 Mt/yr)

for making charcoal, 48% of the total estimated

charcoal can supply all the iron and steel

industry’s requirements.

Figure 3: Estimated charcoal requirement for the Australian

steel industry (*assumes R&D success, and commercial

viability).

This assumes that the Australian iron and steel

industry requirements are represented by the two

main steel companies, BlueScope Steel and

OneSteel. This will also depend on the success of

R&D into the supply chain, pyrolysis, and

biomass application in terms of technical,

environmental and economic parameters.


• Potential utilisation of, and value-adding by,

undervalued residues from the forestry,

agricultural and horticultural industries.

• Likely large net reduction of CO2 emission

from the iron and steel industry.


There appears to be sufficient biomass residues to

supply the estimated theoretical charcoal

requirements that may potentially be utilised in

the Australian iron and steel industry for

metallurgical processing in the short to medium

term.

It is envisaged that this study will be expanded

into a more detailed and focussed evaluation of

biomass sources concurrent with the other

research and development works on biomass

processing, transport, pyrolysis and steel industry

application.

The necessity for high quality biomass for

charcoal making (low contamination as evidenced

by ash content and low variability of products)

also requires future study on the issues related to

supply of high quality biomass from existing

sources.

Acknowledgements

This work was carried out by CSIRO under the

auspice of the Centre for Sustainable Resource

Processing, which is established and supported

under the Australian Government’s Cooperative

Research Centres Program, with financial support

from BlueScope Steel, CSIRO, OneSteel and

CSRP.

References

1. K.R. Bootle “Wood in Australia: Types, Properties and

Uses”. 2nd Edition. McGraw-Hill Book Company, Sydney. 2005.

2. J.A. Kininmonth and L.J. Whitehouse “Properties and

Uses of New Zealand Radiata Pine”. Volume 1: Wood

Properties, Forest Research Institute, Rotorua, New Zealand. 1991.

3. D.E. Langberg, M.A. Somerville and T. Norgate “The

use of charcoal in the iron and steel industry”. DMR 3278, CSIRO Minerals. 2007.

4. J.G. Mathieson “Private Communication”, Clayton, Oct

2008.

5. F.A. Ximenes, W.D. Gardner and J.F. Marchant

“Carbon flow following the harvest of blackbutt trees

and their conversion into sawn productions”. Research

Report No 41, DPI NSW. 2005.

6. F.A. Ximenes and W.D. Gardner “Biomass allocation

following radiata pine thinning operations in southern

NSW forests”. CRC for Greenhouse Accounting Draft Report by DPI NSW. 2006.

Production of Charcoal for Recarburisation Plant Trials

M. Somerville1, J. Mathieson

2, P. Ridgeway

3, and M. Davies

3

1CSIRO Minerals, VIC 2BlueScope Steel, NSW 3OneSteel, NSW


The recarburisation of molten steel has been

identified as a suitable first step in the utilisation

of renewable and sustainable carbon in the iron

and steel industry1. Recarburisation is the

addition of extra carbon to molten refined steel to

produce particular steel grades. In general

recarburisers need to have a high carbon content

(>95 %), with low ash, volatiles and moisture.

Traditional recarburisers include materials such as

calcined anthracite or other high grade carbon

sources but could also include charcoal. Previous

kilogram-scale work conducted at CSIRO2

showed that:

• Charcoal readily dissolved into molten steel at

high rates of recovery

• Carbon recovery to steel was directly related

to the volatile content of the recarburiser

• Alkali components in the recarburiser ash can

be reduced and report to the gas stream.

In light of the success of this preliminary work,

the decision was made to proceed to plant-scale

recarburisation test work using charcoal at

OneSteel’s Sydney EAF Steel Mill (SSM) at

Rooty Hill during 2008-09.

The major part of the planning for these trials was

the preparation of a large amount of charcoal (3-4

tonnes). The key specification of the charcoal

was a volatile and moisture content of less than 1

wt %. Previous experience indicated that a

pyrolysis temperature of 1000 °C would be

required to produce charcoal of this quality.

The Illawarra Coke Company (ICC) agreed to

support the project by allowing one of their coke

ovens at their Corrimal plant to be used in a

preliminary test and a further five ovens to be

used to produce the main batch of charcoal. In

normal operations 20-30 tonnes of fine locally

sourced coal is loaded into the ovens through a

port in the oven roof. Heat is generated through

the combustion of volatile components above the

coal bed. At the end of a production cycle (2-3

days), the coke is pushed from the ovens using a

ram. The coke is then quenched, crushed and

sized The coke ovens at Corrimal were suitable

for this charcoal making exercise because they

can provide a stable high temperature of about

1000 °C and the ovens were available with

convenient access. Heat was available in the

empty coke ovens from hot walls and from

operating adjacent ovens.

This presentation will cover some experimental

results generated during the preliminary and main

charcoal making process at Corrimal.

Preliminary Trial

Methodology / Experimental Technique

The aim of the preliminary trial was to determine

the time required to produce low volatile charcoal

from supplied E. Rossie wood (or scribbly gum)

using the Corrimal coke ovens operating at about

1000 °C.

Seasoned wood was cut into lengths of about 1

metre and had a diameter of between 3 and 10 cm.

The wood logs were loaded into a standard 200

litre fuel drum. Thermocouples fitted to the

drums measured the wood and furnace

temperatures during the trial.

Three drums of wood were added to a recently

emptied oven. One drum was removed after 2, 4

and 6 hours. Once removed, each drum was

quenched with water to prevent burning of the

charcoal. Representative samples of the charcoal

were collected and analysed at HRL

Technologies. Figure 1 shows a photograph of

the drums of wood inside the Corrimal coke

ovens.

Figure 1: Photograph of drums of wood inside the Corrimal

coke ovens.


Figure 2 shows the temperature within each drum

and the temperature of the coke oven during the

pyrolysis trial. These plots show the oven

temperature reached a maximum of about 1130

°C, probably due to combustion of the evolved

volatiles, after about 1 hour, before decreasing to

about 960 °C after 6 hours. The wood reached the

oven temperature, about 1000 °C, after 2 hours,

before decreasing slightly over the next 4 hours to

about 980 °C.

Table 1 shows the proximate analysis of the initial

wood, the three samples of charcoal (2, 4 and 6

hours) and a typical Recarburiser used at SSM.

The high moisture content of the charcoal

samples at about 70 % is due to the water

quenching of the hot charcoal. The 4 and 6 hour

charcoal samples showed increasing ash contents

and a corresponding decrease in fixed carbon

content. This is most likely due to extra burning

of the charcoal during the extended stay in the

oven.

0

200

400

600

800

1000

1200

7:00 8:00 9:00 10:00 11:00 12:00 13:00

Time

Tem

pera

ture

(°

C)

drum 2

drum 1

drum 3

oven

Figure 2: Wood and oven temperatures during pyrolysis

trial.

The composition of the 2 hour charcoal sample

compares very favourable with the recarburiser:

The ash content is lower, the volatile content is

about the same, while the fixed carbon content is

slightly higher. These results indicate that 2 hours

residence time of wood in a coke oven at about

1000 °C should be adequate to achieve a low

volatile content of the charcoal while minimising

burning and hence the ash content.

Table 1: Proximate analysis of the original wood, charcoal

samples and SMM recarburiser.

Sample Moistur

e (%)

Ash

(db%)

Volatile

(db%)

FC

(db%)

Wood 15.7 0.2 78.8 21.0

Char 2 hr 67.6 1.5 0.7 97.8

Char 4 hr 69.7 3.1 0.2 96.7

Char 6 hr 69.3 6.9 0.4 92.7

SSM

recarb.

0.6 5.3 0.7 94.0

Charcoal Making for Plant Trials


Approximately 20 tonnes of wood (E. Rossie) was

supplied from a property near Colinton NSW.

The wood was delivered as debarked logs 1.8

metres in length and from about 60 mm to about

200 mm in diameter. Approximately 40 logs were

strapped together to form bundles using

conventional steel strapping. Each bundle

weighed about 1 tonne. Figure 3 shows a

photograph of the wood bundles.

During the charcoal making four bundles were fed

into each oven using a site fork lift. The charcoal

was removed from the oven using the coke

discharge ram which pushed the charcoal into the

waiting buckle of a larger front end loader (shown

in Figure 4). The charcoal was dumped into a

large steel bin where it was quenched with water

to minimise burning due to exposure of the hot

charcoal to air

Figure 3: Photograph of wood bundles used to produce

charcoal for recarburisation trials.

The aim was to allow the wood to reach 1000 °C

before pushing and subsequent quenching of the

charcoal. The results of the previous drum test

work indicated that this should occur by about 2

hours.

The temperature of the fourth (last charged)

bundle of wood from each oven was monitored

during the charcoal making run. Grab samples of

the quenched charcoal from each oven were

collected and analysed by HRL Technologies.

Figure 4: Photograph of hot charcoal being removed from

the Corrimal coke ovens.


Figure 5 shows the temperatures recorded for each

oven during the charcoal making. The x-axis

shows the time from the final bundle addition for

each oven. The typical behaviour evident from

Figure 5 is a rapid increase in temperature to

about 400 °C, followed by a plateau of about 3

hours and finally an increase in temperature to

approach the oven temperature of about 1000 °C.

The temperature of one oven (33) remained

stubbornly low, but still displayed the temperature

plateau albeit at a lower temperature than the

other ovens. The reason for this lower

temperature is not known but may be due to the

position of the thermocouple in the wood bundle.

The temperature of wood/charcoal increased

much faster in the preliminary trial than in the

main charcoal making exercise. This is likely to

be due to a more open wood pile in the trial which

increased gas flows and hence convective heat

transfer during pyrolysis. The collapse of the

wood bundles into a bed of smouldering charcoal

would have resulted in relatively slow heat

transfer and hence a slower rate of pyrolysis.

0

200

400

600

800

1000

1200

0:00 1:00 2:00 3:00 4:00 5:00 6:00

Time from final charging (hr:min)

Te

mp

era

ture

°C

oven 39

oven 37

oven 35

oven 33

oven 31

Figure 5: Temperature of the wood in each oven during

charcoal making.

Proximate analysis results are shown in Table 2.

These results show that the target volatile content

of the charcoal (less than the 1 %) was reached.

The ash content was generally less than 1 % but

the charcoal from oven 37 contained 1.6 % ash.

The average ash content was 0.85 %. Table 1

shows that the ash content of the initial wood was

about 0.2 %. This indicated that the recovery of

material to charcoal is about 23 % and suggests

that the charcoal yield from the original 20 tonnes

of wood may be around 4.6 tonnes. However

some of the ash on the charcoal is likely to have

been washed away during water quenching.

Hence, the charcoal yield was probably less.

Table 2: Proximate analysis of charcoal samples collected from each coke oven.

Sample Moisture

(%)

Volatiles

(db%)

Ash

(db%)

FC

(db%)

Oven 31 58.3 0.83 0.83 98.3

Oven 33 63.5 0.80 0.73 98.5

Oven 35 50.4 0.29 0.63 99.1

Oven 37 59.1 <0.01 1.59 98.4

Oven 39 51.6 <0.01 0.48 99.5


This project work was successful in producing a

significant quantity of charcoal of the required

volatile content for recarburising plant trials. The

next stage in processing the charcoal is drying

followed by crushing and sizing to a suitable

product for SSM plant operations. Issues with the

preparation of a final charcoal product of less than

1 % water need to be finalised before the

recarburising trials can be completed.

Acknowledgements





Centres Program. Financial support for this work

from CSIRO Minerals Down Under Flagship,

BlueScope Steel and One Steel are also

acknowledged. The cooperation of the

management and staff of ICC prior and during the

trials at Corrimal is gratefully recognised.

References

1. D.E. Langberg, M.A. Somerville and T. Norgate, “The

Use of Charcoal in the Iron and Steel Industry”, CSRP Report DMR-3278, October 2007.

2. D. Langberg, M. Somerville and B. Washington,

“Recarburisation of Molten Steel Using Charcoal”. CSRP Report DMR-3277, September 2007

Piloting the Integrated Dry Granulation and Heat Recovery Process at CSIRO

D. Xie, J. Donnelley, R. Flann, Y. Pan, S. Sanetsis and B. Washington

CSIRO Minerals, VIC

Introduction

The Australian steel industry produces more than

2 million tonnes of molten slags at around 1500°C

each year, which contain a large amount of

sensible heat. Currently, these slags are either air

cooled (Fig.1) or water granulated with some

adverse environmental impact such as air

pollution and excessive water usage. The large

amount of waste heat is not recovered.

Figure 1: Slags were poured and air cooled in a slag pit at

OneSteel Whyalla Steelworks.

Dry slag granulation1 is emerging as an attractive

alternative to water granulation to provide an

environmental friendly solution. Some major

design difficulties however have been experienced

in previous studies. Recent work at CSIRO since

2002 has resulted in major progresses in the

process design to overcome some of these

difficulties.

To build on the success of CSIRO’s work, a four

year CSRP project has commenced in August

2006 to extend the dry granulation work to

incorporating heat recovery.2 The project is also

jointed sponsored by Australian steel industry

(BlueScope Steel and OneSteel). The key

objectives were:

• To dry granulate molten blast furnace slags

and produce glassy slag products for cement

manufacture. Thus, reducing GHG emission

from cement industry.

• To recover high grade waste heat from molten

slags, and utilise the heat in steel plants to

reduce energy cost and hence GHG emission.

• To reduce water consumption and emission of

sulphur to atmosphere.


Dry slag granulation could involve pouring

molten slag onto a spinning disc or cup to produce

fine slag droplets using centrifugal forces as

shown in Figure 2. In comparison with alternative

approaches such as air blast and rotary drums, the

spinning disc dry granulation is more efficient in

producing fine slag droplets for fast quenching

and efficient heat recovery.

Figure 2: Dry granulation of molten slags at a spinning disc.

The dynamic process involves fast droplet

breakup, rapid cooling and solidification. There

are several major technical challenges: including

(1) optimal disc design and process control to

provide smooth and well-controlled granulation;

(2) fast quenching and solidification of droplets to

produce glassy granulates for cement

manufacture; (3) compact reactor design for

recovering high grade heat and reducing the

capital and operation cost. A conceptual process

based on a two-step dry granulation/heat recovery

operation has been developed. This is shown

schematically in Figure 3.

Hot air >6 00°C

Granules <5 0°C

Hot air >6 00°C

Air 25 °C

Air 25 °C

Dry granulation (Spinning disc)

Heat exchanger(Packed bed)

Slag

1500°C

Slag

1500°C

Drying

Preheating

Steam

Power

Desalination

Cement

Solid 900°C

Figure 3: Conceptual flowsheet for the new integrated dry

granulation and heat recovery process.

In this process, the molten slag is first atomised in

a dry granulator and slag droplets generated are

fast quenched and solidified to produce glassy

granules. The hot granules are then removed to a

packed-bed heat exchanger to extract the

remaining heat contained in the slag. The waste

heat is recovered in the form of hot air, which

could be used on-site for drying, preheating and

other purposes.

A new pilot plant was constructed (Figure 4) at

CSIRO to prove the conceptual design. A series of

tests using industrial blast furnace slags were

carried out to establish key operating parameters

and further optimise the process design.

Figure 4: New pilot plant at CSIRO.

Some technical and operational challenges were

experienced, including efficient air delivery and

smooth handling of hot granulate collected. These

challenges have been resolved through design

modification and process optimisation. The pilot

plant could now be run smoothly to produce fine

granulates as shown in Figure 5.

Figure 5: Slag products from dry slag granulation.

The slag products collected were subjected to size

analysis. The results indicated that more than 90%

of the products were less than 1.5 mm in diameter

as shown in Figure 6. These slag samples are

being further assessed for their suitability for

cement applications.

0

10

20

30

40

50

60

70

80

90

100

0 1000 2000 3000 4000 5000

Aperture (_m)

Cum

ula

tive %

passin

g

T15_R7

T15_R8

T16_R3

T16_R4

T17_R1

T18_R1

Figure 6: Accumulative size distribution of dry granulated

slags.


A new integrated dry slag granulation pilot plant

has been designed, constructed and commissioned

at CSIRO. The pilot plant is being used to prove

the concept of integrated dry slag granulation and

heat recovery. Test results with industrial blast

furnace slags have confirmed the feasibility of the

integrated process.


The new integrated slag treatment process will

allow for full value recovery (slag products and

heat recovery) from molten slags with a potential

reduction of 1.8 million tonnes GHG emission

each year for Australia.


• Pilot scale tests have shown that the integrated

dry granulation and heat recovery is feasible.

Slag products are being assessed for

suitability for cement application.

• CFD modelling and plant measurements are

being carried out for anticipated scale up and

plant trials in 2009.

Acknowledgements





Centres Program. The work is also jointly

sponsored by CSIRO, OneSteel and BlueScope

Steel.

References

1. Featherstone, W B: "Slag Treatment Improvement by

Dry Granulation," Iron and Steel Engineer, 1998, July, 42.

2. D Xie, Y Pan, R Flann, B Washington, S Sanetsis, J

Donnelley, T. Norgate, S Jahanshahi, “Heat Recovery

from Slag through Dry Granulation”, The 1st CSRP

Annual Conference, 21-22 November 2007, Melbourne, Australia.

Minerals, Metals and Sustainability – A book for the future

W.J. Rankin

CSIRO Minerals, VIC


A project has recently been established within the

Centre for Sustainable Resource Processing

(CSRP) with the objective of writing and

publishing a book which describes, within a

sustainability framework, the nature of mineral

resources, the role that minerals and their products

play in society, and the science and technology of

the minerals industry.

While there are many books on ores and

mineralogy, mining, and processing of metals –

from elementary, descriptive books through to

advanced technical books – these are usually

written from a relatively narrow discipline

perspective. The proposed book will have a

multi-disciplinary approach which integrates the

physical and earth sciences with the social

sciences, ecology and economics.

The intended readership of the book will be wide

and the book will be written in a manner to be

easily understood by readers with diverse

academic levels, backgrounds and interests.

However, it will be aimed particularly at:

• students of engineering and applied science

(mining, metallurgical, civil, chemical,

mechanical, environmental) and geology, as

an integrated overview of the minerals

industry and/or as an introduction to specific

aspects of the industry;

• practising engineers, geologists and scientists

who want a modern, and broader perspective

of the minerals industry;

• students of economics, social sciences and

related disciplines who need an overview of

the minerals industry;

• professionals in government service in areas

such as resources, environment and

sustainability; and

• non-technical professionals working in the

minerals industry or in sectors servicing the

minerals industry.

The planned book will bring together recent

knowledge and understandings of the role of

minerals in modern society and the implications

of sustainable development on the use of minerals

and mineral products and of the minerals industry.

It will draw on a wide body of knowledge

including the work of CSRP.

It is hoped the book, when published, will have an

impact on the approach and content of relevant

undergraduate and graduate courses and that it

will result in an increase in the knowledge and

understanding by graduates and professionals in

industry and government service.

Methodology

The working title for the book is: Minerals,

Metals and Sustainability. The text will be

written around thirteen chapters arranged in a

logical sequence. Chapters 1 to 4 will introduce

readers to the concept of the earth as the source of

all materials, how materials are utilised in society

(with particular focus on inorganic materials), and

the issue of sustainability in relation to finite

resources. Chapters 5 to 8 will examine two

important classes of materials, metals and cement,

in detail. Chapters 9 to 13 will address the

sustainability challenges.

The proposed chapter titles are:

Chapter 1 Introduction

Chapter 2 Mineral resources

Chapter 3 Minerals and their role in society

Chapter 4 The Sustainability challenge and

opportunity

Chapter 5 Why are metals special?

Chapter 6 How metals are produced

Chapter 7 Cement – Glue of industrial society

Chapter 8 Alternatives to metals and cement

Chapter 9 Minerals and the environment

Chapter 10 Natural limits – What are they?

Chapter 11 Towards zero waste

Chapter 12 What does the future hold?

Chapter 13 A sustainability strategy for minerals

and metals

In general, no prior specialist knowledge will be

assumed other than an elementary knowledge of

chemistry and physics.

To make the book accessible, Chapters 1 to 4, 5 to

8 and 9 to 13 will be able to be read as

independent, free-standing blocks if desired and,

where necessary, cross-references will be

provided. For the same reason, more specialist

material, which could be omitted without loss of

continuity, will be placed in the later sections of

chapters. In this manner, the entire book, or

sections of it, will be capable of being studied at

various depths depending on the reader’s needs

and background.

The author, John Rankin, will draw on expertise

from a wide range of researchers and industry

personnel as needed through group and one-on-

one discussions, and will seek peer review of

chapter drafts as they are prepared by “key”

researchers and an “Editorial Panel”. Much of the

knowledge and understanding required for the

project is available within CSRP and its

Participants.

Progress

The project was formally established as a CSRP

project in mid-August 2008. In the 12-month

period leading up to this, the scope of the book

was developed and the proposal was discussed

with stakeholders. There was generally

widespread support for the project and this led to

commitment to support a two-year project within

CSRP. The project is fully resourced, with

funding being provided by CSRP, CSIRO

(through the Minerals Down Under Flagship) and

seven industry sponsors (BHP Billiton, Rio Tinto,

Alcoa, Xstrata, Hatch, BlueScope Steel,

OneSteel). CSIRO Publishing and the AusIMM

have indicated a strong interest in publishing the

book. Formal writing of the text commenced in

mid-August.

Demand-Supply Interaction on Future Mining Resource Production: The coal model

S. Mohr and G. Evans

University of Newcastle, Chemical Engineering, NSW

Introduction

Future supply of mineral and energy resources is

commonly predicted by applying Hubbert’s bell

curve analysis to historical production [1,2,3]. The

approach has a number of limitations. Firstly, it is

not theoretically immediate why production

should follow a bell curve distribution, and indeed

there are a number of cases where production is

asymmetric. Secondly, there is no mechanism for

accounting for changes in production rates by

external influences such as wars, economic

recessions, international crises etc. In this study an

alternative model is described, which includes

demand-supply interaction and production

schedule inputs for production output. The

Oceania coal industry will be used to demonstrate

this model.

Modelling Approach

The model is based on a market approach,

whereby the supply of a resource, such as oil or

coal, is influenced by: demand for it, production

capacity, and the amount of reserves available to

supply that market. The market is defined

depending on the nature of the resource. For

example, oil is shipped globally so it would be

appropriate to consider the global market.

Conversely, the global transport of natural gas is

relatively minor, so it would be more appropriate

to consider a regional or continental market.

The modelling approach for mineral production

from mining operations is described below:

1. Identify what the market is, regional, global,

etc, and define the intrinsic demand for the

resource. Here, intrinsic means the demand

that the supply is aiming for. Market

expectation (or demand) is a function of

supply availability; e.g. if supply cannot

satisfy demand then the market must adjust to

reduce demand.

2. Define the Ultimately Recoverable Resources

(URR) available to the market, and these are

obtained from literature values.

3. The URR are used to calculate the number of

mines with a given production capacity.

Demand-Supply interactions regulate the

scheduling of individual mine outputs,

including start-up and shut-down durations,

and steady-state production rates based on

historical data.

4. An iterative procedure is applied, whereby

intrinsic demand and supply try to equalise.

When sufficient capacity is available the

production is equal to the intrinsic demand,

brought about by a combination of production

from existing operating mines and by bringing

new mines on-stream. Increased production

from existing mines can also be implemented

in an effort to achieve the intrinsic demand. A

point is reached, however, where production

is not sufficient to meet existing intrinsic

demand, so the intrinsic demand must be

reduced.

Figure 1: Schematic of Mineral Model.

The model is illustrated in Figure 1. The approach

is implemented iteratively, so at a given point in

time, t, we know the following: the intrinsic

demand of the market MD[t], the total supply to

the market MS[t], the number of mines online M[t]

and the production, Pj[t], and scheduling for each

mine.

From time, t, information, intrinsic demand and

supply for time t+1, are obtained as follows:

The intrinsic demand MD[t+1] is given by:

][][]1[

tk

DDDetMtM =+ , (1)

where kD[t] is the rate variable, which is a function

of the equilibrium value, kD0, and the difference

between demand and supply at time, t, such that:

=][

][][][ 10

tM

tMtMkktk

S

DS

DD , (2)

where k1 is a proportionality constant.

The supply, MS[t+1], in year, t+1, is the sum of

the production from each mine, Pj[t+1], i.e.:

[ 1]

1[ 1] [ 1]

M t j

S jM t P t

+

=+ = + . (3)

In eq. (3), M[t+1] is the number of mines online at

time t+1, which is given by:

( )[ ]

[ ]

[ 1] [ ]S

S

M tk t

URRT TM t M M M t e+ = , (4)

where MT is the total number of mines, based on

the URR and the production scheduling applied to

all mines, and kS[t] is the supply rate variable and

is given by:

+=][

][][][ 20

tM

tMtMkktk

S

DS

SS , (5)

where kS0 is the equilibrium supply rate variable

and k2 is a constant.

Production for a given mine is assumed to take 4

years to linearly ramp up to the maximum

production, MP, and similarly 4 years to ramp

down to zero production at the end of its useful

life. There is scope for an individual mine to

upgrade production in response to demand to a

level of 2MP. The number of mines that upgrade

in the tth

year, MU[t+1], is equal to:

=+ ][][

][][]1[ 3 tMk

tM

tMtMktM U

S

DS

U , (6)

where k3 is a proportionality constant, and kU is

the minimum gap needed between intrinsic

demand and supply before mines start upgrading.

Wars, financial slowdowns, etc, can reduce

intrinsic demand, which can be manually inputted

into the model, causing mines to be taken offline.

As described above, the model incorporates an

iterative approach, whereby production from

individual sources is optimised so that the demand

is met. When this can no longer be achieved by

the resources available, then the demand is revised

downward. The result is a demand-supply

interaction reflecting the reality of finite resources

and the market responding accordingly.

Example: Oceania Coal Demand-Supply

The model output is illustrated for the demand-

supply of coal in the Oceania market. Sources of

coal are Australian bituminous, sub-bituminous,

lignite, and other sources (e.g. New Zealand). The

corresponding URR values are given in Table 1.

The other inputs to the model are the constants

described in eqs. (1)-(6). These constants affect

the intrinsic demand and supply, and were

approximated by fitting the model to the European

coal market which has already peaked.

Table 1: Oceania Coal URR and peak year.

Type URR (Gt) Peak year

Aust. Bituminous 51 2058

Aust. Sub-bituminous 3 2033

Aust. Lignite 40 2088

Other 1 2056

Total 95 2066

The predicted supply and demand for coal types in

Oceania are shown in Figure 2. Also shown

(dotted line) is demand based on the current

growth rate.

Figure 2: Demand-Supply Model for Oceania.

It can be seen that future supply curves for each of

the different coal types is not a symmetric bell

curve, as assumed by Hubbert. This is due to the

intrinsic demand-supply interaction being based

on the historical data for the European market,

which itself is not symmetric. While a different

demand-supply interaction would change the

shapes of the curves, the peak years do not vary

greatly (less than 10 year) to those given in Table

2. Current projections show supply cannot meet

intrinsic demand after 2030, which is well below

the current business-as-usual scenario. The model

predicts a peak in total supply of coal occurring at

around 2066, and beyond that is a rapid decline in

production as the URR is quickly exhausted.

Acknowledgements

The Authors acknowledge the CSRP for their

financial support.

References

1. Zittel W., Schindler J., Energy Watch Group,

http://www.energywatchgroup.org/fileadmin/global/pdf/EWG_Report_Coal_10-07-2007ms.pdf.

2. Laherrere J., Groningen Annual Energy Conv. 2006, http://www.oilcrisis.com/Laherrere/groningen.pdf

3. Hubbert M. K., World Wildlife Fund’s Conference, The

Fragile Earth: Towards Strategies for Survival, San Francisco, 1976.

Applying DEM Outputs to the Unified Comminution Model

M.S. Powell1, N. Weerasekara

1, I. Govender

2 and M. Khanal

1

1University of Queensland, JKMRC, QLD 2University of Cape Town, Centre for Minerals Processing, South Africa

Introduction

The concept of a unified comminution model

(UCM) was formulated with the vision of

bringing all comminution models onto a common

base. The structure of the model relies on

knowledge of the mechanical environment in the

comminution devices - the Discrete Element

Method (DEM) has been used as the tool to

simulate this.

The model structure tackles the fundamental

causes of rock breakage, considering them to be

independent of the comminution equipment. The

equipment is considered to be the mechanism that

applies a given comminution environment to the

ore particles. Thus, once an understanding of the

processes of ore breakage is gained, the

mechanical environment from any type of

equipment can be overlaid to provide a model that

predicts the production of broken material.

Model Definition

The Unified Comminution Model (UCM) traces

the mechanical collision environment experienced

by particles in a comminution device and

calculates the resultant damage to the particles in

order to predict the progeny of the device.

The structure and key inputs to the model are

illustrated in Figure . The central block outlines

the model principle, and the peripheral blocks

outline the inputs that are required for the model.

A full description of the model is available in

Powell et al 2008.

DEM Outputs

The energy spectrum provided by DEM

simulations is divided into the tangential and

normal specific collision energies (Figure 2) per

rock size. This figure is rich with information,

showing the difference in tangential versus normal

energies; the increase in specific energy as the

rock size decreases; The reduction in the

maximum specific energy experienced by larger

rocks; and the dramatic increase in total number

of impacts for the smaller rocks (there are many

more of them compared to large rocks). These

individual energy spectra per size are used as the

input to the UCM breakage model.

Figure 1: UCM model structure.

The following observations can be drawn from the

plot of the energy spectra:

• A tiny fraction of impacts have sufficient

energy to break rocks in one impact (Ecrit

0.28 kWh/t)

• a few percent of collisions can break rocks

through multiple impacts (Eo 3.7 E-3 kWh/t).

• the vast majority of the impact energy is

insufficient to cause bulk body breakage

• the smaller particles receive considerably

more impacts

• the smaller particles are subjected to

considerably higher specific energies of

impact.

This simulation output emphasises the importance

of the low energy surface impact abrasion

damage, and of the ball mill mode of breakage.

Therefore rounding and abrasion are likely to be

the dominant comminution mechanisms for

particles larger than 63 mm, and incremental

breakage important for particles in the size range

of 5.6 to 63 mm.

Applying these rates to the contents of a mill

allows the rate of breakage of each particle size to

be calculated. Adding the breakage size

distributions, that are a function of the impact

energy, to this information yields the overall size

distribution of the progeny. This is being applied

to the modelling of a pilot SAG mill in the first

instance.

Model principle

Breakage process

+

Mechanical environment

Breakage product

+

transport function +

Time stepping

Product

DEM input

Mechanical environment

Contact detection energy, angle, force

Mode of contact

Probability distribution

mode & partner

Distribution of mode

PB

Framework

Particles +

Breakage

then

Convert to mass

PB by mass

Tomography, MLA

Breakage testing

Mode of contact

Tests to reproduce these

Input - force, energy

Output – size dist, liberation

Transport function

Mechanistic flow relationship Later

SPH/DEM

Tangential energy distributions

1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

1.E+05

1.E+06

1.E+07

1.E-10 1.E-09 1.E-08 1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01

Specific energy, kwh/t

cu

mu

lativ

e n

o o

f im

pa

cts

180mm

125mm

90mm

63mm

45mm

31mm

22mm

16mm

11mm

8mm

5.6

Normal energy distributions

1.E+00

1.E+01

1.E+02

1.E+03

1.E+04

1.E+05

1.E+06

1.E+07

1.E-08 1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01

Specific energy, kwh/t

cu

mu

lativ

e n

o o

f im

pa

cts

180mm

125mm

90mm

63mm

45mm

31mm

22mm

16mm

11mm

8mm

5.6

Figure 2: Specific energy distributions per particle size

class.

Figure 3: Fitted function to energy distributions.

The curves given in Figure 2 have been fitted to

generic functions, that allow interpolation

between sizes and a simple functional form to be

used in the UCM energy distribution model,

Figure 3.

We have also shifted from the traditional use of

the spring and dashpot energy to using the input

kinetic energy to analyse the breakage. This

energy corresponds to the energy that is measured

in the breakage tests, and is very different to the

dashpot energy, as illustrated in Figure 4.

Figure 4: Comparison of Dashpot and kinetic energies at

collision.

Conclusions

The principle of the application of the UCM has

been demonstrated, and the DME modelling

capability is considerably advanced. However, the

ore testing criteria have delayed proper model

validation, which is currently underway. This

clearly indicates the need for advanced pilot and

ore breakage characterisation work, which will

allow accurate validation and calibration of

advanced mill models, and models of other

comminution and classification equipment.

There are a number of areas to be addressed in the

capability of DEM:

• Apportioning of energy between particles

• Energy available for breakage during a

collision

• Left-over energy- input power not accounted

for in particle damage

• Sub-DEM size material

• Solids and slurry transport

• Solids discharge function.

References

1. Powell, M.S., Govender, I., and McBride, A.T., (2008).

Applying DEM outputs to the unified comminution

model – the SAG mill. Minerals Engineering, Special edition – DEM07. DOI: 10.1016/j.mineng.2008.06.010

E Ecrit

Energy Efficient Comminution Circuits – A modified grinding strategy and the

selection of a target product size

Z. Pokrajcic

University of Queensland, JKMRC, QLD

Introduction

This paper describes two techniques that can be

used in the design of comminution circuits for

improved efficiency and a reduction in total

energy consumption. They are:

• A modified grinding strategy where efficient

grinding equipment is used and techniques

employed to decrease grinding media

consumption

• A more reasoned and thorough approach to

the selection of target product size for a

comminution circuit involving a better

understanding and interpretation mineral

liberation data.

Background

There has been much discussion about the energy

intensive nature of comminution processes. It has

been quoted several times that comminution

processes account for 3%-4% of the global

electrical energy consumption. However, the

comminution process remains inherently

inefficient. It is understood that 85% of the energy

used comminution is dissipated as heat, 12% is

attributed to mechanical losses and only 1% of the

total energy input is used in size reduction of feed

material (Alvarado et al. 1998).

Another aspect of comminution which contributes

significantly to the energy intensity of the process

is the consumption of media and mill liners. The

quantity of energy necessary for the fabrication of

the media and/or liners from extraction, to

transport, to manufacture and assembly is

considered part of the total energy utilisation of

the comminution process. This is also quite an

energy intensive process and opportunities to

minimise steel consumption in comminution

circuits should be realised.

The third factor which contributes significantly to

the energy intensity of comminution circuit is the

chosen target product size, or grind size. As the

target product size decreases the energy required

to achieve the product size increases significantly.

As the particles become smaller their strength and

therefore their resistance to breakage increases.

For small particles as internal flaws are depleted,

the observed strength approaches the high

intrinsic strength of the solid.

This is the prospect facing all comminution design

and operating engineers. Hence, there is an

increasing need for more efficient grinding

systems and innovative comminution technologies

that not only increase the efficiency of the

comminution process but also act to minimise the

consumption of steel media and liners. This

together with a new approach to choosing a target

product size can significantly increase the energy

efficiency of a comminution circuit.


The two parts of this investigation were

undertaken according to the following:

A modified grinding strategy

All of the modifications made to the configuration

of an existing comminution circuit were

undertaken using the modelling and simulation

package JKSimMet. It is a valuable tool in

predicting the performance of a comminution

circuit, including the equipment power draw,

product size distributions, classification

efficiency, etc.

The aim of the eco-efficient modifications was to:

• Reduce the direct energy usage. That is,

decrease the grinding power used in the size

reduction process.

• Reduce the indirect energy usage. That is,

decrease the consumption of comminution

consumables such as steel media and mill

liners.

The strategy used to achieve these aims involved

better utilisation of the comminution energy. This

can be achieved by purposely targeting the

comminution energy at specific problematic size

fractions such as the critical size material in the

primary mill and the very fine fraction in the

secondary mill. Through the use of more efficient

comminution devices such as HPGR (high

pressure grinding rolls) a product size distribution

that is narrower in size range can be generated.

This size distribution contains less material in the

coarse fraction to aid overall mineral liberation

and less material in the fine fraction to minimise

the effects of slimes.

The strategy also exploited opportunities to

minimise the load in grinding mills and in some

cases employ autogenous grinding techniques to

decrease indirect energy consumption.

Selection of a target product size

This concept was illustrated using a sample from

of rod mill discharge from Mount Isa. Rod mill

discharge can be equated to SAG mill discharge in

terms of product size and product size distribution

in SABC circuit (SAG mill with recycle crushing

and a single stage ball mill).

The sample underwent staged grinding, using a

laboratory scale rod mill and ball mill to produce

samples of product sizes ranging from

P80=0.015 m to P80=1.500mm.

Samples from each of the product sizes were

analysed using MLA techniques to assess the

extent of liberation of individual mineral

components.

In order to maximise the potential target grind size

and therefore minimise required energy input for

the circuit, the following concepts were applied:

• minerals were grouped to form larger entities

and therefore deemed liberated at coarser

sizes

• particles were deemed recoverable in a

flotation process, as opposed to liberated,

when the particle contains greater than 15%

mineral of interest

Together the above two points act not only to

increase the target product size but it also enable

the separation and rejection of material (typically

liberated gangue) from the circuit. The rejection

of gangue from the feed stream can limit the

amount of over-grinding but more importantly it

will reduce the mass of material proceeding to the

next stage of size reduction and/or separation.

This allows for more efficient separation and

significantly lower energy requirements for

downstream size reduction.


A modified grinding strategy

Eco-efficient modifications were made to an

existing SABC according to the following:

• A pre-crush stage using a HPGR was included

to minimise the generation of critical size

material in the primary mill. The aim was to

reduce a portion of the feed material below

critical size and generate more fines to

decrease the load in the primary mill. Not all

of the feed material was pre-crushed to allow

for sufficient abrasion breakage in the primary

mill.

• The SAG mill was converted to an AG mill

and the ball mill was converted to a pebble

mill to decrease liner and media consumption

• Existing mill sizes were reduced

When the performance of this new circuit,

denoted ABC-HPGR-PM, was compared to the

original SABC circuit, the following (based on

JKSimMet and mass balancing predictions) was

observed.

• The ABC-HPGR-PM circuit produced a both

a finer and steeper product size distribution

which promoted better flotation performance,

approximately a 2% increase gold recovery

• Direct energy comparisons show that the

ABC-HPGR-PM circuit had an operating

work index of 15.9kWh/t compared to

24.3kWh/t for the base case SABC circuit.

This is an energy saving greater than 8kWh/t

for the whole circuit. Significant to say the

least.

• Comparison of indirect energy consumption

for the two circuits is showed that again the

ABC-HPGR-PM out performed the standard

SABC circuit with a 79% saving in indirect

energy based on media and liner wear alone.

The SABC circuit consumed 3535kW while

the ABC-HPGR-PM circuit consumed

753kW. These figures are based on 6kWh/kg

required to manufacture the grinding

consumables.

Selection of a target product size

Analysis of the MLA data for the various product

sizes to include liberated and recoverable

particles, showed that:

• 70-80% of galena and sphalerite is liberated at

P80=76 m

• 70-80% of galena and sphalerite is

recoverable at P80=150 m

• When galena and sphalerite are combined and

analysed as one mineral, 85% of this mineral

is recoverable at P80=150 m

• >95% of the sulphides (combined galena,

sphalerite, pyrite and pyrrhotite) are

recoverable at P80=1607 m

• 45% of non-sulphide gangue is liberated at

P80=1607 m

These points can be used in the design a

comminution circuit where product size is finer

than 1.5mm.

The aim is to exploit the recoverability of valuable

minerals and the rejection of liberated gangue in a

separation step at the coarsest product size.

Figure 1: Design of separation circuit post primary

grinding.

The valuable sulphides are then separated in the last

Table 1: Energy saved by rejecting the liberated non-

sulphide gangue at the head of the separation circuit.

size reduction, the rejected material is primarily coarse


circuit to include:

problematic fractions such as critical size material

mineral liberation characteristics of coarse stream

stream and therefore the comminution circuit thereby

the liberation and recoverability characteristics of

Acknowledgements

on liberation data analysis and coarse particle

References

2. Atasoy, Y., W. Valery, et al. (2001). Primary verses

secondary crushing at St. Ives (WMC) SAG mill circuit. SAG 2001, Vancouver.

3. Daniel, M, 2007. “Energy efficient liberation using

HPGR technology”. PhD thesis. University of Queensland, Brisbane. (unpublished)

4. Grano, S. (2006). Ian Wark Research Institute. Annual Report 2006. U. o. S. Australia.

5. Huls, B. J. and G. S. Hill (2006). "The coarse particle recovery process." CIM Bulletin 99(1093): 68-74.

6. Johnson, N. W. and P. D. Munro (2002). Overview of

flotation technology and plant practice for complex

sulphide ores. Mineral processing plant design, practice and control, Vancouver, SME.

7. Laslett, G. M., D. N. Sutherland, P.Gottlieb, and N.R.

Allen, (1990). "Graphical assessment of a random

breakage model for mineral liberation." Powder Technology 60(2): 83-97.

8. Manlapig, E. V., D. J. Drinkwater, P.D.Munro, N.W.

Johnson, and R.M.S Watsford, (1985). Optimisation of

grinding circuits at the lead/zinc concentrator, Mount Isa

Mines Ltd. Symposium on Automation for Mineral Resource Development, Brisbane, Australia.

9. Munro, P. D. (1993). Lead-zinc-silver ore concentration

practice at the lead-zinc concentrator of Mount Isa

Mines Limited, Mount Isa, QLD. Australasian Mining

and Metallurgy - The Sir Maurice Mawby Memorial

Volume. J. T. Woodcook and J. K. Hamilton. Melbourne, AusIMM. 1: 498-503.

10. Napier-Munn, T. J., Morrell, S., Morrison, R.D. and

Kojovic, T., (1996). Mineral Comminution Circuits:

Their Operation and Optimisation, pp. 49-92, (Julius

Mineral Kruttschnitt Mineral Research Centre: Brisbane)

11. Pokrajcic, Z, and Morrison, R, 2008, A simulation

methodology for the design of eco-efficient comminution circuits. IMPC 2008, Beijing. IN PRINT.

Improving Grinding Efficiency with the IsaMill™

M. Larson1, R. Morrison

1, F. Shi

1 and M.F. Young

2

1University of Queensland, JKMRC, QLD 2Xstrata Technology, QLD


As mining companies have been called on to

process ever finer grained deposits they have had

to turn to new technology to economically treat

these ores. One example of this is the IsaMillTM

.

The IsaMillTM

was originally developed to grind

fine grained Mount Isa and McArthur River

lead/zinc ores to sizes finer than 7 micron.

The IsaMillTM

is a horizontal stirred mill that takes

advantage of high rotational speeds and fine inert

media to efficiently grind to liberation sizes that

were previously uneconomical, at best, and

sometimes were simply not achievable.


A laboratory investigation was conducted to

determine the effects of the IsaMillTM

operating

variables on power draw and power efficiency,

using a 4 litre M4 IsaMillTM

at the JKMRC. The

results of the test work were used to construct a

preliminary IsaMill model.

The laboratory standard signature plot is used to

characterize IsaMill performance and grinding

energy requirements (Weller and Gao, 1999). This

has been shown to reliably translate laboratory

scale results to mine site operations. Laboratory

testing was undertaken with the M4 IsaMill to

determine which operating variables affect the

grinding energy efficiency and to what extent they

impact the grinding energy efficiency. A

campaign of test work was conducted using

copper concentrate from the Mount Isa Mine that

had a F80 =64 microns. Variables tested included

the mill speed, feed pump volumetric flow rate,

grinding media type, grinding media size, media

filling and feed pulp density. In addition viscosity

was investigated as it relates to the feed pulp

density and grinding energy efficiency

Mill Speed vs Power

y = 0.0017x - 1.866

R2 = 0.986

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

1400 1600 1800 2000

Mill RPM

Ne

t P

ow

er D

ra

w (k

W)

15 micron

power draw

Figure 1: Mill speed vs. net mill power draw (at constant

media load).


The feed pump volumetric flow rate did not have

an effect on grinding energy efficiency.

Regardless of the pump speed the same signature

plot will still be created. The mill speed did not

have an effect on grinding energy efficiency and it

was found to be a reliable indicator of net mill

power draw between 1400 and 2000 RPM, as

shown in Figure 1. Less residence time was

required to grind to a given target size as the mill

speed was increased.

Similarly, a linear relationship was a good fit for

the net power draw when the media load was

varied (Figure 2). The two relationships should

provide a means in the future to better predict the

outcomes of varying operating conditions while

performing signature plot tests

Media Loading vs Net Power Draw

y = 0.0161x + 0.0156

R2 = 0.9679

0.4

0.6

0.8

1

1.2

1.4

1.6

20 40 60 80 100

Media Load (%Full)

Ne

t P

ow

er D

ra

w (k

W)

15 micron

power draw

Figure 2: Media loading vs. net mill power draw (at

constant mill speed).

The overall grinding efficiency tends to increase

as the media volume increases in the mill, with the

most dramatic effect occurring from the low end

of filling to the beginning of the generally

accepted operating range of 60-80% full, when

grinding to a target of P80=10 microns, as shown

in Figure 4. The media volume is the bulk volume

of the media at rest divided by the net volume of

the mill (volume of shaft and grinding discs

subtracted from total mill volume). The

inefficiency was most clearly seen when the mill

had very low media filling. When the mill had low

media filling the initial discharge density was

lower than the feed density indicating there was a

build up of solids inside the mill. The solids that

build up in the volume that would normally be

filled with grinding media could cause increased

power draw and be reducing the power efficiency.

M4 IsaMill Model

Two simple equations were developed relating the

overall efficiency of varying media loads and feed

pulp densities, based on the previous grinding

energy requirements graphs. The 100% efficiency

was set at a media filling of 80% and solids feed

pulp density of 45%, then the equations were

developed to relate the energy efficiency loss to

the changing variables. These equations were

applied to a set of six model validation tests.

Table 1: Model Validation Results using Mount Isa Cu

Concentrate.

Test Media

Volume

(L)

Density

(%weight)

kWh/t

model

kWh/t

actual

%

error

1 2.8 39 68.5 69.5 1.44

2 2.3 54.4 86 84.8 1.42

3 1.9 35 90.5 83.7 8.12

4 2.6 52 76.6 75.6 1.32

5 2.5 45 75.5 77 1.95

6 2.7 57.1 80.1 82.2 2.55

Fines Production Model

It was decided to investigate the form of the

relationship between new surface area generated

and energy input, based on work by McIvor

(2007) and later Musa and Morrison (2008).

McIvor and Musa both demonstrate that new

production of particles finer than a certain target

size is approximately linear with energy for rod

and ball mills. A similar simple relation was

attempted to allow grinding energy predictions

with changing the feed size.

The selected micron size for the modelling was

chosen as the 50% passing size of the final pass

from the Isamill test, and then the percent passing

the selected micron size for each pass in the

Isamill test was plotted against cumulative energy

for the Isamill test to that point.

This did not fit a linear relationship or any

variation of a power plot. It was determined

however that by squaring the percent passing

value and plotting it versus cumulative energy on

a normal plot a straight line connecting each

IsaMill pass with the feed at the 0 kWh/t could be

formed. This would have the advantage over the

standard signature plot that the feed material

could be included in the line created.

The main disadvantage of the signature plot is that

each percent passing line cannot be extrapolated

back to the feed size. This imperfection has to be

taken into account when estimating the effect of

changing feed sizes. With the new method, using

the percent passing a particular size, the line

should be able to be shifted left or right depending

on the change in feed size and using the same

slope to estimate the new power requirement.

This should be valid for any case where the same

media is capable of grinding the new feed size.


The work done thus far shows that through an

improved understanding of basic mill operating

variables, showing grinding energy efficiency of

the Isamill can be predicted and improved. The

ability to predict the mill performance with

coarser feed sizes using the fines squared function

will allow the design of grinding circuits

incorporating the IsaMill. This along with the

optimization of media size will assist to fully

optimize the entire circuit grinding efficiency

Acknowledgements

The authors wish to thank Xstrata Technology for

funding this project. This work was carried out

under the auspice of the Centre for Sustainable

Resource Processing, which is established and

supported under the Australian Government’s

Cooperative Research Centres Program.

References

1. Weller K. and Gao M., 1999, “Ultra-fine Grinding”,

AJM Crushing and Grinding Conference, Kalgoorlie, 27-28th April 1999

2. McIvor R. and Finch J., “The Finch-McIvor Functional Performance Based Grinding Circuit Modelling System”

3. Musa F. and Morrison R., 2008. “A More Sustainable

Approach to Assessing Comminution Efficiency”, Comminution 08 Falmouth

Multiple-pass High Pressure Grinding Rolls Circuits

M. Hilden and M. Powell



High Pressure Grinding Rolls (HPGR) are widely

considered to be more energy efficient devices for

comminuting rock than tumbling mills such as

SAG mills. Recently, HPGR have been installed

and commissioned at Cerro Verde (Peru),

Mogolokwane (South Africa), and Freeport

(Indonesia) with reported energy savings of

around 19% over more conventional grinding

routes [1].

To date, HPGR have been installed as single unit,

either in open circuit or closed circuit with a

screen. A single HPGR unit is typically limited to

around 3 kWh/t, compared with SAG mills which

may have energy inputs exceeding 10 kWh/t, and

therefore despite being less efficient can generate

a finer product.

This project aims to determine whether a series of

HPGR units could be used to perform most of the

grinding duty in a process plant, increasing the

amount of breakage performed in these highly

efficient devices. Such a circuit could

substantially reduce the energy and grinding

media used in milling circuits and therefore

reduce the carbon footprint of new processing

plants.


Three ore samples have been tested to date. The

samples were a hard copper-gold ore, a softer

lead-zinc sulphide ore, and the third was a

porphyry copper ore. The samples were crushed to

-32 mm top-size and processed in a three-pass

HPGR circuit (Figure 1).

HPGR

HPGR

HPGR

Figure 1: Three HPGR in series.

Test work was carried out on two large pilot-scale

Köppern HPGRs at AMMTEC, Perth. The first

two samples were tested using 0.22 m (W) x 0.75

m (D) rolls in July 2007. The third sample was

tested using a newer 0.25 m (W) x 1.0 m (D)

HPGR (Figure 2) in March 2008.

Figure 2: Köppern’s Andrew Gardula and JKMRC’s

Malcolm Powell discussing the pilot-scale HPGR at

AMMTEC Perth.

The third sample was also tested in a wide range

of circuit configurations that included screening

between passes. Test work was also carried out

using one or two passes with the Köppern HPRG

with subsequent passes carried out at CSIRO

Pinjarra Hills using a laboratory-scale HPGR 0.10

m (w) x 0.25 m (D). In each case, up to three

passes of the HPGR were used.

The energy consumption and product size was

measured for each stage in the flow sheets to

allow energy efficiency of each circuit to be

determined.


The feed and product sizes from the samples

tested to data show that the three-HPGR circuit

produces a broad product size distribution, but

also that the magnitude of size reduction reduces

with subsequent passes. Two passes were found to

be effective; however the third pass continues to

generate fines without substantially reducing the

top size. The result is a flatter or broader product

size distribution as shown in Figure 3.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.01 0.1 1 10 100

Size, mm

% P

assin

g

Feed

Pass 1

Pass 2

Pass 3

Figure 3: Product sizing for the three-stage circuit.

A number of three-stage circuit HPGR

configurations were compared with the above

three-pass HPGR circuit. Two approaches were

found to be effective in generating a steeper

product size distribution:

1. Removing excess fines from the feed to the

HPGR (Figure 4a), and

2. Using a smaller diameter rolls in the second

and/or third pass (Figure 4b). A B

HPGR

HPGR

HPGR

2.0 mm Screen

48%

52%

1.0 m HPGR

1.0 m HPGR

0.25 m HPGR

HPGR

8 mm Screen

HPGR

HPGR

1.0 m HPGR

22%

78%

t

0.25 m HPGR

0.25 m HPGR

Figure 4: Two of the alternative flow sheets tested.

As shown in Table 1, these circuits produce a

similar-sized product using less energy compared

with three passes in the 1.0 m rolls. In the case of

flow sheet 4b, just two HPGR passes are needed

to grind to an equivalent product fineness

(measured using the 80% passing size, P80, and

the quantity of -150 microns fines generated), and

uses 33% less energy.

Table 1: Comparison of various three-pass HPGR circuits.

Pass 1 Pass 2 Pass 3

P80, um 9391 3788 2414

% -150um 24% 38% 47%

Energy, kWh/t 2.77 5.23 7.53

P80, um 8804 2824 1425

% -150um 22% 35% 45%

Energy, kWh/t 2.83 4.13 5.48

P80, um 4710 1764 1195

% -150um 27% 46% 53%

Energy, kWh/t 3.23 5.03 5.83

Triple-pass circuit Fig 1.

Modified Circuit Fig 4a.

Modified Circuit Fig 4b.


• Test work confirms that the HPGR can be

used in series to efficiently generate a very

fine product. After three passes, around half

of the product can be made into final flotation

feed size.

• Two flow sheet modifications further improve

the performance of the three-stage HPGR

circuit and use less energy, albeit using a

potentially more complex flow sheet.


This work has shown that a multi-pass HPGR

circuit can generate a high proportion of fines, but

only two passes appear to be beneficial without

either a classification stage or a reduction in roll

diameter. One or two additional samples will be

tested in coming months.

Acknowledgements





Centres Program. The large-scale test work was

carried out with the assistance of Köppern

Machinery Australia and the small-scale HPGR

test work was carried out by Steve Suthers of

CSIRO, Pinjarra Hills.

References

1. C. M. Rule, I. Smit, A. J. Cope, and G. A. Humphries,

"Commissioning of the Polycom 2.2/1.6 5.6MW HPGR

at Anglo Platinum's new Mogalakwena North

Concentrator," presented at Comminution 08, Falmouth, 2008.

Update on the JKRBT (JKMRC Rotary Breakage Tester)

T. Kojovic and F. Shi



The long history of comminution studies at the

JKMRC has included the development of

laboratory testing procedures, to provide material

characterisation data that are used in conjunction

with machine specific data in modelling and

simulation (Napier-Munn et al, 1996). The key

developments include:

• Twin Pendulum

• Drop Weight tester (DWT)

• Automated Linear Impact Comminution

Evaluator (ALICE)

• Short Impact Load Cell (SILC)

The DWT is an established and globally accepted

impact characterisation device, and has become

one of the standard impact tests for AG/SAG

milling ore characterisation. The DWT apparatus

and its associated data reduction technique were

developed so that the relationship between

specific energy input and resultant product size

could be determined (see Equation 1):

( )EcsbeAt .

10 1= (1)

where t10 is a size distribution ‘fineness’ index

defined as the progeny percent passing one tenth

of the initial mean particle size, Ecs is the specific

comminution energy (kWh/t), and A and b are the

ore impact breakage parameters determined from

DWT results. The index A*b has become well

known in the mining industry as a reliable

indicator of impact ore hardness, and essentially

describes the rate at which fines are produced

(t10) for a set amount of specific energy (Ecs).

The t10-Ecs relationship and A*b parameters are

used in size-reduction-modelling for crushers and

mills in the JKSimMet mineral processing

simulator (Wiseman and Richardson, 1991).

Since all tests using the DWT, SILC (Bourgeois

and Banini, 2002), or the Hopkinson bar are

conducted on single rock specimens positioned

manually on the anvil or flat surface, they are both

time consuming and expensive. For the test to

remain practical the rock samples being tested are

limited to 10 – 30 pieces for each size fraction,

which inevitably throws into question the

statistical validity of the derived ore

characteristics. Recent developments in DEM of

milling have revealed that small energy impacts

occur much more frequently than high energy

impacts (Djordjevic et al, 2004). As such it

appears a new requirement in breakage testing

will be the characterization of incremental

breakage at small impact energies. However,

testing of repetitive impacts at small energies

using the DW tester is very time-consuming and

hence impractical. Research into finding a rapid

breakage characterization device was clearly

warranted in an effort to overcome these

limitations.

The concept of using kinetic energy to crush rocks

seemed like a viable alternative for rapid breakage

characterization, since it no longer requires the

positioning of rock specimens manually on the

anvil. Industrial applications of this concept are

found in Vertical Shaft Impact (VSI) crushers and

laboratory pulverizers, which employ a rotor-

stator impacting system. However, both of these

devices are employed merely for particle size

reduction, not for breakage characterization, as the

exact amount of energy applied in the process is

not well controlled or measured. However, the

JKMRC found that VSI crushers can be modelled

from first principles, allowing useful simulations

(e.g. effect of rotor speed) to be carried out with

only the basic design data on the machine and

rock characteristics required (Kojovic, 1996;

Djordjevic et al, 2003). This suggests that the rock

characteristics may be inferred from the product

of the VSI, if the amount of energy applied in the

process can be precisely controlled and measured.

In response to the body of evidence supporting the

use of kinetic energy, in 2005 the JKMRC

comminution research team led by Shi and

Kojovic decided to thoroughly investigate the

feasibility of using the kinetic energy concept for

particle impact breakage characterization. A

prototype Rotary Breakage Tester (JKRBT), with

a rotor diameter of 360 mm, was designed and

manufactured by the JKMRC pilot plant

workshop team (Figure 1). The operating system

consists of a vibrating feeder, a rotor-stator

impacting device with its drive system, and an

operation control unit. Like the DWT, the JKRBT

also requires the ore particles be pre-sized into

narrow fractions. Particles of the selected size are

fed into the rotor-stator impacting system via a

vibrating feeder.

Figure 1: Photograph of the prototype JKRBT device, with

rotor-stator showing through the inspection window.


The JKRBT uses a rotor-stator impacting system,

in which particles gain kinetic energy while they

are spun in the rotor. They are then ejected and

impacted against the stator, causing particle

breakage. The specific energy of each impact in

the JKRBT, Ecs, is defined as the kinetic energy

Ek per particle mass m:

22

5.05.0

iik V

m

Vm

m

EEcs === (2)

Since the particle mass does not affect the specific

energy in this type of impact breakage device, the

Ecs becomes solely dependent on the impact

velocity Vi. The research team has shown that the

specific energy of impact can be accurately and

precisely controlled in the JKRBT. They also

confirmed that the actual impact velocity in

practice is less than theoretical, requiring the

JKRBT to be calibrated first to achieve precise

breakage energy levels. The calibration is

machine specific and can be programmed into the

JKRBT control system to allow the user to enter

the required specific energy directly, or it can be

tabulated in the JKRBT Operating Manual as pre-

set speeds which can be entered manually to

achieve a desired specific energy level.


The JKRBT overcomes some of the limitations of

existing impact tests such as a lack of precision of

the energy input, the amount of time required to

run individual tests and the difficulties in testing

the smallest particles. To date, comparative

breakage tests using the prototype JKRBT device

and the traditional JKMRC Drop Weight Tester

showed that the two devices generate the identical

breakage–energy relationship for the same ore of

the same size, as shown in Figure 2.

0

5

10

15

20

25

30

35

40

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

Ecs (kWh/t)

t 10 (

%)

DWT

RBT

RBT repeat

Model Fit

Figure 2: Comparison of t10 vs. Ecs relation determined by

the JKRBT and DWT.

Statistical analysis indicates that the two testing

methods can generate statistically similar

breakage parameter A*b values, as illustrated in

Figure 3 which compares the A*b values of 16 ore

types determined by industrial JKRBT and DWT

respectively. This analysis accounts for the

difference in contact points and strain rates

between the DWT and JKRBT.

0

20

40

60

80

100

120

140

160

180

200

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Data set

Axb

valu

es

Figure 3: Comparison of breakage parameters A*b

determined by the JKRBT and DWT (10% error bars on

DWT result).

The very positive feedback received from mining

companies to the JKRBT has prompted the

JKMRC to further validate the device through a

formalized experimental test program on

professionally designed and manufactured JKRBT

machines. The validation project started in March

2007 with the first industrialised unit installed and

commissioned at Anglo Research in

Johannesburg, South Africa. Barrick, BHP

Billiton, Rio Tinto and Teck Cominco have also

signed up to participate in the project. The current

status and site deployment is as follows:

• Anglo Research (Crown Mines, South Africa)

x 2 machines

• Barrick Gold (JKtech, Brisbane) - 12 months

agreement

• BHP-Billiton (Newcastle Technology Centre,

Australia)

• Rio Tinto (Kennecott, USA)

• Teck Cominco (Trail, Canada)

Figure 4 shows a diagram of the industrialised

JKBRT, designed and manufactured by Russell

Mineral Equipment (RME).

Figure 4: The first industrialised JKRBT installed at Anglo

Research in South Africa.

Unlike the prototype device, the industrialised

JKRBT can be opened for easy access to clean the

breakage chamber. The lid is operated by an

electronically driven linear actuator. The counter-

weight at the back of the lid is designed to prevent

the lid from falling down by gravity should the

electronic system fail while the lid is opening or

closing. The industrialised JKRBT employs a

rotor 450 mm in diameter, and can treat particles

from 1 to 45 mm, at specific energy levels from

0.001 to 3.8 kWh/t. The use a rotary feeder has

provided a simple feeding system that also offers

an effective noise suppression mechanism,

reducing the noise to below 85dB at the highest

operating speed.

It is worth noting that the new breakage model

developed by the JKMRC (Shi and Kojovic,

2007) enables the user to accurately match the

DWT derived breakage parameters, A and b, from

JKRBT data on four (13.2 to 45 mm) instead of

the standard five DWT size fractions (13.2 to 63

mm). The model and application of JKRBT

concept to ore testing has been patented by The

University of Queensland.


Feedback from Anglo Research has indicated

higher productivity and better repeatability by the

first commercial JKRBT when compared with the

Drop Weight Tester. The results to date look very

promising and suggest the JKRBT will provide a

powerful and easy to use tool for rapid ore

breakage characterisation to the mining industry.

Tests have confirmed the device offers a rapid

method for determining the hardness of drill core

samples within the context of the AMIRA P843

geometallurgical project. For comparative testing,

Walters and Kojovic (2006) found the JKRBT

appears to be the best choice, followed by SMC

and PLT.

Some of the key features of the JKRBT are:

• Treats many more particles – hence can

generate statistically more valid results

• Wide particle size range (1.4 – 45 mm), wide

energy range (0.001 – 3.9 kWh/t)

• Rapid characterisation (1/8 - 1/10 of DWT

time confirmed by Anglo Research)

• Accurately and precisely control energy

• Excellent reproducibility


A new rapid breakage characterization testing

device, the JKRBT, has been developed by the

JKMRC. A detailed study using a prototype

machine confirmed the concept of using

controlled kinetic energy to characterize ore

particle breakage, prompting a comprehensive

validation using an industrialized JKRBT,

professionally designed and fabricated by Russell

Mineral Equipment (RME). The validation project

will be completed in 2008, with commercial

release expected in early 2009.

Application of the prototype JKRBT in

geometallurgical testing with the AMIRA P843

GeM project has confirmed the JKRBT device

offers a rapid and consistent method for

determining the hardness of drill core samples.

The JKRBT has the potential to revolutionize ore

testing, with applications in laboratory impact

breakage characterisation of rock samples, drill

cores, coal and other materials. Automation with

lab image sizing technology like the Camsizer

would further enhance its productivity. It also

stands poised to significantly advance the process

control of comminution circuits since it is readily

adaptable to on-line ore hardness measurement.

This concept will be explored in the current P9O

project.

Acknowledgements

The research and development of the JKRBT has

been funded by The University of Queensland,

JKMRC, JKTech, AMIRA/P9N Project, and

Centre for Sustainable Resource Processing

(CSRP). In-kind support from Anglo Research

staff during the commissioning of the first

industrial JKRBT unit is greatly appreciated.

References

1. Bourgeois, FS and Banini, GA, 2002. A Portable load

cell for in-situ ore impact breakage testing. Int. Journal of Mineral Processing. 65, 31–54.

2. Djordjevic, N, Shi, F and Morrison, RD, 2003. Applying

discrete element modelling to vertical and horizontal

shaft impact crushers. Minerals Engineering 16: 983-991.

3. Djordjevic, N., Shi, F. and Morrison, R., 2004.

Determination of lifter design, speed and filling effects

in AG mills by 3D DEM. Minerals Engineering, 17, 1135-1142.

4. Kojovic, T., 1996. Vertical shaft impactors: predicting

performance, Quarry Aust Jrnl, Vol 4, No 6, pp35-39

5. Napier-Munn, T.J., Morrell, S., Morrison, R.D., and

Kojovic, T., 1996. Mineral comminution circuits: their

operation and optimisation. ISBN 0 646 28861. Julius

Kruttschnitt Mineral Research Centre.

6. Shi, F. and Kojovic, T. 2007. Validation of a model for

impact breakage incorporating particle size effect. Int. Journal of Mineral Processing, 82, 156-163.

7. Van Latum, LA, 1985. The evaluation,

conceptualization and development of single particle

breakage testing apparatus. Masters Thesis

(unpublished), Technical University of Delft, Netherland.

8. Walters, S., and Kojovic, T., 2006. Geometallurgical

Mapping and Mine Modelling (GEMIII) – the way of

the future. Proc SAG 2006, Vancouver, Vol IV, pp411-425.

9. Wiseman D.M. and Richardson J.M., 1991. JKSimMet

- the mineral processing simulator. Proceedings 2nd

Can Conf on Comp Applications in the Min Ind, (Eds

Paulin, Pakalnis and Mular), Univ B.C. and CIM, Vol II,

pp427-438.

Improvement of Energy Efficiency of Rock Comminution through Reduction of

Thermal Losses

N. Djordjevic


Introduction

Theoretical calculations show that only some 0.1

to 2 % of the energy supplied to rock during

comminution is effectively utilised for fracturing.

This project investigates the accumulation in

thermal energy of rock, using an advanced

thermal imaging camera, to quantify how much

energy is lost into increased thermal energy of

rock, and to explore potential relationships

between this heat accumulation, the physical

and/or mineralogical characteristics of the rock,

the produced fragment size distribution and

operational parameters of the comminution

process. Ultimately it is hoped that such better

understanding of heat losses, will generate options

for modification of operational and equipment

parameters of comminution equipment to reduce

heat losses and improve efficiency of rock

crushing.

Obtained results demonstrated that substantial

fraction of net energy delivered to rock ends up as

the heat energy radiating from the rock after

crushing. Obviously this indicates more energy is

delivered to the rock that what is required to crush

the rock. From the practical point of view,

question is what can be done with existing

equipment to reduce such thermal losses. We were

focused on the potential improvement of the

efficiency of crushing that can be archived

through modification of the feed size distribution.

Testing was performed during HPGR crushing of

Rio Tinto’s Bingham Canyon copper ore.

Thermal Radiation

Any object will emit energy due to its

temperature, as long as temperature of the object

is above absolute zero. From the measured

radiation energy and emissivity constant of the

surface, it is possible to calculate surface

temperature of the materials. Assuming that

surface temperature represents average

temperature of the materials, it is then possible to

calculate thermal energy stored within materials,

from equation.

Experimental Results

We performed a range of rock crushing

experiments which was monitored using sensitive

infrared (IR) camera. Measurements of thermal

losses that occurred during High Pressure

Grinding Roll (HPGR) crushing of local basalt

were performed. HPGR crushing was performed

on the six samples of equal mass (18.86kg). Initial

pressure of the HPGR were set to 20, 30 (2

samples), 40, 50 and 60bar. Infrared imaging was

performed during crushing, Figure 1. Crushing

products were sized and compared against feed

size distribution and introduced net comminution

energy.

Figure 1: Setting used during HPGR testing.

Increase in the specific comminution energy

(applied pressure), results in finer fragmentation.

However, there is tendency that after certain

amount of energy is introduced, further increase in

consumed specific energy, produce only minor or

insignificant improvements in fragmentation. This

is observed in the diagrams of fragment size

distributions which show asymptotic behaviour of

the fragment size distribution as applied pressure

increase, or amount of net energy consumed

increases, Figure 2.

HPGR Crushing of Mt. Morrow Basalt

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

0.1 1 10

size (mm)

cu

mu

lati

ve (

%)

passin

g

Feed

20bar

30bar

40bar

50bar

60bar

Figure 2: Feed size distribution and fragment size

distribution after HPGR crushing under different pressure.

Results of IR imaging of the crushed rock, shows

that large fraction of net energy supplied to rock is

transformed in to heat, rising temperature of the

rock, Figures 3.

Figure 3: Infrared image for set pressure of 30bar, copper

ore (temperature scale min. 30degC, max. 60degC).

From the temperature and measured net energy

supplied to rock, efficiency of energy utilization

was calculated as function of the HPGR pressure

setting, Figure 4.

Efficiency of Utilization of Net Energy Consumed

0

10

20

30

40

50

60

20 30 40 50 60

Pressure (bar)

%

Figure 4: Efficiency of energy utilization as function of

HPGR pressure.

Conclusions

Presented results and interpretation are restricted

on comminution aspects of energy transfer and

thermal losses. We were able to perform

reproducible measurements of temperature

increase that occur during highly transient events

such as dynamic rock breakage. Obtained results

show that with increase of energy introduced,

there is increase in the temperature along the

fractured surface as well as increases in the

overall amount of thermal energy.

Results obtained during HPGR testing clearly

indicate that there is optimum intensity of

pressure to which given rock needs to be exposed.

Any further increase in pressure results in only

marginal increase in fragmentation and significant

increase in unproductive heating of rock.

Feed characterised with relatively narrow size

distribution (i.e., without large amount of very

fine material) is prone to more efficient crushing

at significantly lower pressure than feed with

same top size, but including relatively large

amount of fines. Experimental results indicate

that up to 40% of energy can be saved through

optimization of the applied pressure and

modification of feed size distribution.

Acknowledgements





Centres Program.

Towards a Virtual Comminution Machine

R.D. Morrison1 and P.W. Cleary

2

1University of Queensland, JKMRC, QLD 2CSIRO, Mathematical and Information Sciences, VIC


Towards the end of the 1990’s readily available

personal computers became sufficiently powerful

– when combined with an efficient numerical

code – to use Discrete Element Modelling (DEM)

in two dimensions for models involving a few

hundred to a few thousand particles in

commercially available packages. Some

proprietary codes reported up to 200,000 particles

(Herbst and Nordell, 2001 and Cleary, 2001)

In early 2000, JKMRC and CSIRO-MIS agreed to

an informal collaboration with the objective of

testing various DEM approaches against detailed

process measurements. The initial collaboration

demonstrated that 3D-DEM using spheres was

sufficiently realistic for flow patterns and power

estimation within tumbling mills. The results were

reported in papers which were presented at SAG

2001 and in the technical literature. (Morrison et

al, 2001; Cleary et al, 2003; Cleary et al, 2008)

The VCM will allow a comminution machine

design which exists as a suitably detailed design

in a 3D Computer aided design file (CAD) to

simulate processing an ore (which has been

characterised by suitable test work) to predict

progeny, power consumption, wear and even

machine component loadings.

Methodology

A “new” comminution device typically requires

15-25 years for development and to gain a

sufficient degree of industrial acceptance to have

any chance of commercial survival.

Unsurprisingly, many promising devices suffer

from interruptions to their development schedule

or fail altogether for lack of development finance

as much as any technical shortcomings.

There are also broader reasons for seeking more

efficient ways to carry out comminution. It has

been estimated that comminution processes world

wide consume about 3% of all electricity

generated (La Nauze et al, 2002). This compares

with 1.3% by ore milling in North America (DOE

report, 1981). Consumption of grinding media

and wear resistant liners consumes about the same

order of energy in terms of green house gas

production (Musa and Morrison, 2007). Hence

the total energy equivalent is more like 6%. In

Australian terms, 10 and 20% are reasonable

estimates. The current substantial pressure to

reduce greenhouse gas production therefore

provides a strong incentive to devise more energy

efficient comminution devices, and to reduce the

calendar time required for their development.

The most serious obstacle in the traditional

approach to developing a new comminution

device is the time and cost of building successive

prototypes. Each prototype represents a

substantial investment in development time and

capital cost. Mechanical shortcomings such as

component rigidity may cause rejection of a

design with potentially good process performance.

If at least some of the required prototypes can be

constructed and assessed by simulation, then a

large reduction in development time and cost

should be possible. In many ways worse still are

machines which work well at lab and pilot scale

but cannot be scaled up to commercially viable

sizes.

Each prototype cycle can easily cost 6-12 months

of development time and tens to hundreds of

thousands of development dollars. Hence, a

system which can allow a good deal of the

development work to reside in a simulation model

has much to recommend it.

Experimental Technique

To be credible as a VCM, both the simulation

program and the ore characterisation process must

be thoroughly verified. Hence the VCM should

also be useful for modelling and further

development of existing comminution devices.

A VCM has the further advantage that variations

on a theme can more easily be considered

allowing estimation of the total energy “footprint”

in terms of electrical energy and wear whilst

treating “identical” simulated feed materials.

The Virtual Comminution Machine (VCM) must

be able to relate each collision type and energy

within a DEM simulation to the probable degree

of breakage of that particle – based on

characterisation tests of a particular ore. As

several modes of comminution occur within

tumbling mills, relating collision spectra to

comminution represents a particular challenge.

The greater part of comminution energy is

expended in a slurry environment in “wet”

autogenous and semi-autogenous (AG and SAG)

mills and ball mills. The proportion of fine (minus

37 micron) material in the slurry has a strong

effect on its viscous properties. A reasonably

viscous slurry which can retain particles on the

surface of the grinding media helps to ensure

selection of those particles for breakage. Hence

the fluid and particle interactions are also

important for particle breakage and transport.

Models based on computational fluid dynamics

(CFD) are not well suited to the high shear, high

viscosity environment with complex free surface

behaviour found in wet tumbling mills. DEM can

be adapted to fluid modelling using Smoothed

Particle Hydrodynamics (SPH), see Cleary (1998)

and Cleary et al. (2007). The shear dependant

viscosity model developed by Shi and Napier-

Munn (2002) has been implemented in the CMIS

code using SPH as a way to model mineral

slurries.

In a DEM simulation, the forces exerted by each

particle on the equipment surfaces are estimated.

Therefore the relative wear rates can also be

estimated using an appropriate wear model.

Similarly, the total energy requirements for the

machine can be estimated by summing the

energies consumed by particle movement.


In short, the VCM objective is to model each

process within any comminution machine to an

acceptable level of accuracy so that it can be used

as virtual development environment for existing

and new comminution machines.

The VCM will always require powerful

computational capabilities. A parallel

devolvement is the UCM (Universal

Comminution Model) which will embed VCM

type relationships into a system suitable for

desktop PCs (Powell et al, 2008)

Acknowledgements





Centres Program.

References

1. Cleary, P. W., (1998), Modelling confined multi-

material heat and mass flows using SPH, Applied Mathematical Modelling, Vol. 22, pp. 981-993.

2. Cleary, P. W., (2001), Charge behaviour and power

consumption in ball mills: Sensitivity to mill operating

conditions, liner geometry and charge composition, Int. J. Min. Processing, Vol. 63, pp. 79-114.

3. Cleary, P. W., Powell M.S and Morrison R.D.(2008),

Applying Computationally Intensive Techniques to Modelling of Comminution Devices, IMPC Beijing

4. Cleary, P. W., Morrison, R., and Morrell, S., (2003).

Comparison of DEM and experiment for a scale model SAG mill, Int. J. Min. Processing, 68, pp. 129-165.

5. Cleary, P. W., Prakash, M., Ha, J., Stokes, N., and Scott,

C., (2007), Smooth Particle Hydrodynamics; Status and

future potential, Progress in Computational Fluid Dynamics, 7, pp. 70-90.

6. DOE, 1981. Committee on Comminution and Energy Consumption, NMAB-364

7. Herbst, J. A., and Nordell, L., (2001), Optimization of

the design of sag mill internals using high fidelity

simulation, Proceedings of the SAG conference, Eds.

Barratt, Allan and Mular,, (University of British Columbia), Vol. IV, pp. 150 -164.

8. La Nauze, R.D and Temos, J., (2002), Technologies for

sustainable operation. CMMI Congress 2002, Cairns Queensland, May AusIMM, pp. 27-34.

9. Morrison, R.D., Cleary, P.W. and Valery, W., (2001).

Comparing power and performance trends from DEM

and JK modelling. SAG 2001, Univ of British Columbia, Vancouver, pp. 284-300.

10. Musa F, and Morrison, R. D., (2007). Assessing

comminution efficiency.XXII ENTMME / VII MSHMT – Ouro Preto-MG, Brazil

11. Powell M.S., Govender I. and McBride A.T. (2008)

Challenges in Applying the Unified Comminution

Model, IMPC Beijing

12. Shi, F. N. and Napier-Munn, T. J., (2002). "Effects of

slurry rheology on industrial grinding performance", IntJMP, 65(3-4), pp. 125-40.

The New Energy Logging from the Discrete Element Method

N. Weerasekara and M. Powell



A discrete element modelling (DEM) simulation

for a typical mill section with several thousands of

particles generates massive amounts of output

data, running to several gigabits. Even with good

sampling techniques1 this generates files of close

to a gigabyte.

Although some DEM software provides a GUI

based data analysis environment, their techniques

either require a considerable amount of processing

power or are not capable of delivering the

required inputs for the unified comminution

model (UCM) kind of modelling2.

Therefore a new data logging system was

developed to fulfill the following objective:

• Provide flexibility in handling huge data sets

generated by typical DEM code.

• Extract particle collision information from

DEM output database.

• Summarise data to provide more useful

information for further comminution

modelling process.

Methodology

A C/C++ based framework was developed in

extracting and analysing the massive data sets

generated though the DEM simulation. The

following structure shown in Figure 1 and 2 were

adapted in implementation.

Figure 1: Phase I data extraction.

As show in Figure 1, Phase I of this system

interacts directly with the DEM output and

generates data sets which can be easily managed

and handled by the Phase II. The method adapted

in the Phase I will vary slightly depending on the

DEM output data structure of the DEM code used.

But the basic framework is similar to the

flowchart in Figure 1.

The basic framework for the Phase II is shown in

Figure 2. In this phase the extracted and grouped

data is further summarised depending on the

requirements for further modelling work.

Figure 2: Phase II data summarisation.

A series of DEM simulations were carried out

using EDEM software (Figure 3) and CSIRO

DEM code, for a 1.6m diameter pilot SAG mill

with roughly 30,000 particles. This new data

logging framework was tested and used in

analysing the outputs generated though both these

codes. These results are being used as inputs to

the UCM modelling.

Figure 3: EDEM output graphical view.

Group collision energy data based on

particle size classes and mill liner

Particle Class data

Particle ID’s, No of collisions at each time step

Particle ID’s

Collision energy at each collision

Analysis and summaries collision energy

data for required particle size classes and mill liner

Summarized output

(e.g. Cumulative collision

energy spectrum)

e.g. Modules

AbsCollision

AbsNormalEnergy

AbsTangEnergy

RelCollision

BinEnergyLamda

DEM Output (Raw Data Base collision info, energy etc.)

Extract particle and surface specific information

Extract collision information (e.g. normal/tangential energy)

Particle Class data Particle ID’s, No of collision at each time step

Group data base on particle classes and liner surface

Log energy data at each collision

Particle ID’s Collision energy at each collision

Key Results

From this energy logging system, normal and

tangential collision energy distribution for

different particle classes were observed (Figure 4)

across the full energy spectrum. The effect of

particle size and the number of particles present in

the mill on the cumulative collision energy per

particle is clearly shown. This provides an

excellent input for estimating the modes and

degree of breakage.

Figure 4: Cumulative collision energy spectrum for 12

particle classes (labelled C1 to C12), using CSIRO DEM

simulation.

It was also interesting to observe the number of

collisions occurring within a given particle size

class (Figure 5).

Figure 5: Cumulative number of collision across the normal

energy spectrum for 12 particle classes (labelled C1 to C12),

using EDEM simulation.

Number of collisions per particle (DEM

) is another

important parameter for UCM modelling2, the

developed framework generated that quite well,

for example Figure 6 shows DEM

variation across

the normal energy spectrum. The data logging

system also generates DEM

variation for all the

particle classes simulated in the DEM code.

Figure 6: Number of collisions per particle (DEM

) across

the normal energy spectrum, from 5 different DEM test cases

using EDEM.

Highlights and Benefits

• Robust, efficient and faster data extraction,

analysis and summarisation process.

• Demands less computational power and

memory.

• Has the flexibility and potential to extract and

summarise other relevant information needed

out of the DEM simulation.


The C/C++ based framework was tested and used

in logging and summarising energy from DEM

outputs from two different DEM codes. This was

tested and used in Windows and Linux platforms.

It also proved to be more efficient and fast.

The future direction is to extend this framework to

extract and analysis multiple sets of DEM outputs

in parallel.

Acknowledgements

This work was carried out under the auspice and

funding of the CSRP, and with additional funding

from the JKMRC

References

1. P. W. Cleary, Recent advances in DEM modelling of

tumbling mills, Minerals Engineering, Volume 14, Issue 10, , October 2001, Pages 1295-1319.

2. M.S. Powell, I. Govender, A.T. McBride, Applying

DEM outputs to the unified comminution model,

Minerals Engineering, Volume 21, Issue 11, Pages 744-750.

3. EDEM user manual, Copyright © 2002-2008 DEM

Solutions.

Australian Life Cycle Initiative (AusLCI) and CSRP Database: Australian Data

D. Giurco1, P. Schmidt

2 and B. McClellan

2

1University of Technology Sydney, Institute for Sustainable Futures, NSW 2University of Queensland, Sustainable Minerals Institute, Centre for Social Responsibility in

Mining, QLD


The Australian Life Cycle Initiative (AusLCI) is

being developed by CSIRO and the Australian Life

Cycle Assessment Society (ALCAS). Its aim is to:

"provide a national, publicly-accessible database

with easy access to authoritative, comprehensive

and transparent environmental information on a

wide range of Australian products and services over

their entire life cycle"[1].

In parallel, CSRP researchers are working to

develop a database of readily accessible information

for exchange between CSRP participants to assist

with decision making for sustainability.

The aim of this extended abstract is to:

1. provide an overview of activities being

undertaken by AusLCI and, in particular, the

"metals" working group within AusLCI

2. describe the progress of a CSRP database and

how this relates to AusLCI

3. outline benefits of the CSRP database and

AusLCI initiative and future research needs.

AusLCI Initiative Overview and Approach

Data on which Life Cycle Assessment studies are

based is largely drawn from outside Australia (e.g.

Europe) and can be inappropriate to our local

context. AusLCI is the hub for gathering accurate

and reliable Australian data for use in LCA studies.

The initiative has several technical committees

considering issues of allocation and data quality and

guidelines covering data format, quality, and critical

review aimed at ensuring transparent data is used.

Several sector-based groups have formed to set the

research agenda for data collection within sectors

and engage industry participation (e.g. agriculture,

plastics, water, construction, metals). See [2] for

further information.

The metals group has identified:

• need to further engage industry (e.g. through

roundtable workshop)

• need to address data confidentiality when

seeking involvement

• limited Australian data is publicly available.

AusLCI Future Directions

The AusLCI initiative will continue gathering data

across sectors for incorporation into its database

over the coming years. The priorities for the metals

sector include consulting with industry to identify

data gaps and to find funding to commission new

research for accurate data collection.

CSRP Database Overview

Independently of the AusLCI process the CSRP SD

program involves the task of producing a database

to support CSRP researchers. This commenced

independently of the AusLCI project due to the

longer time-frames required to deliver AusLCI data.

The aim of this is to provide a database of

predominantly quantitative relevant information that

would assist CSRP researchers in evaluating the

overall benefits or impacts of project outcomes.

This database is also intended to feed into the

developing SUSOP® mechanism options which will

incorporate Sustainable Development principles

into the design and operation of industrial

processing plants.

Methodology: CSRP Database Development

Development of the CSRP database over the recent

months builds on initial work during the

development of SD CAT (the SD Contributions

Assessment Tool). The database has been initially

structured to fit with the perceived needs of SD

assessment in the CSRP, as per Figure (overleaf).

To date, the database covers data categories of

electricity, transport, grinding media and reagents.

The data has been collated from public domain

sources such as ABARE, the Department of Climate

Change and industry reports. Minerals processing

data is contributed through other projects and

studies (for example see [3]).

CSRP Database: Interim Findings

Table 1 gives a brief overview of the current

coverage of data found from public domain

literature to date. Whilst transport and electricity

data are increasingly available ahead of emissions

trading, proprietary products such as reagents and

grinding media will require consultation with

producers, or estimated ranges of values.

Table 1: Data overview.

Data

category

Example Data Data

source

Coverage

Transport

(per net

tonne

kilometre)

Articulated

trucks

0.0705 kg CO2

0.0784 kg CO2

(FFC)*

AGO,

NPI,

ARA

Australian

average figures

for key modes of

freight transport;

Electricity

(per MWh)

Qld Grid

903 kg CO2

and

1046 kg CO2 /

MWh (FFC)

AGO,

NPI,

ESAA,

ABARE

ABS

Australian, fuel-

specific and state-

by-state data on

production and

emissions;

Grinding

Media

~ ~ Data collection

continuing

Reagents ~ ~ Data collection

continuing

*Full fuel cycle

The rapid expansion of data quantity and quality

for transport and electricity has been supported by

increasing requirements of carbon reporting. Data

availability for grinding media and reagents has

been less supported by these processes as shown

by limited public domain material.

A ready reference of useful data for CSRP

projects and from CSRP projects for sharing with

other CSRP participants will facilitate assessing

the sustainability benefits of new projects

CSRP Database Benefits

• A comprehensive database of current data for

Australia will allow quicker, reliable

assessment of sustainable development

benefits of CSRP research outcomes and other

projects.

• Provide a ready reference for assessing the

carbon intensity of ongoing operations and

areas for reduced impacts.


The next phase for database development will

involve direct application of energy data to more

specific minerals processing data, especially

around grinding media and reagents. Further

verification of data from multiple sources will

clarify data ranges and potential error margins.

Acknowledgements


the Centre for Sustainable Resource Processing,


Australian Government’s CRC Program.

References

1. AusLCI website. http://www.auslci.com.au

2. Aus LCI Wiki site http://alcas.asn.au/auslci/pmwiki/pmwiki.php/AusLCI/AusLCI

3. Norgate, T. E. and S. Jahanshahi (2006). Project 1A2:

Energy Issues Paper - Energy Use in Metallurgical

Processes and Related Greenhouse Gas Emissions.

Perth, Co-operative Research Centre for Sustainable Resource Processing.

Figure 1: Initial structure of database (dark orange blocks in the figure have been completed to date, lightly shaded blocks are

partially complete, white blocks need data).

Quantitative Phase Analysis of Fly Ashes for Use as Geopolymer Source Materials

R.P. Williams, N. Chen-Tan and A. van Riessen

Curtin University of Technology, Centre for Materials Research, WA


How would it be if we could use cheap industrial

waste products to make concrete with superior

properties that will last for thousands of years,

similar to the concretes used in Ancient Rome?

That is exactly what is achievable by making

geopolymers with clays, fly ash or almost any

aluminosilicate material. The possible

applications of geopolymers are only limited by

scientists’ imaginations. Geopolymers are

currently in use as a concrete (cement

replacement) and have also shown promise in

specialised applications such as fire proof barriers,

ceramic precursors and refractory adhesives to

name just a few. Geopolymers have only short-

range ordering, hence are considered to be x-ray

amorphous.

Fly ash is the fine particle residue from burning

coal in thermal power stations. Fly ashes have

complex variable composition due to the

composition of the coal from which they are

derived and the operating conditions of the power

station. Fly ash consists of a mineral fraction,

which is inert, and an amorphous fraction, which

takes part in the geopolymerisation reaction. This

means the concentration and composition of the

amorphous fraction needs to be determined, in

order to manufacture geopolymers from different

fly ashes with reproducible properties.

The bulk elemental composition of the fly ash can

be obtained by XRF and the mineralogical (phase)

composition can be obtained using a quantitative

XRD technique such as Rietveld refinement,

allowing the amorphous composition to be

determined by difference [1, 2]. Further

understanding of the fly ash internal structure and

of significant factors in the chemistry of the

geopolymer formation, such as the role of

crystalline and amorphous phases and particle

size, can be obtained by from x-ray absorption

spectroscopy (XAS), scanning electron

microscopy (SEM) and transmission electron

microscopy (TEM). Compositional analysis of the

amorphous component of the precursors provides

a better understanding of the geopolymerisation

reaction.


For powder diffraction, each fly ash was

micronized for 5 minutes with a known mass of

internal standard (fluorite - CaF2), the resulting

slurry was dried at 105°C for 24 hrs, then loaded

into 0.5 mm borosilicate capillaries (GLAS,

Germany). The remaining powder was stored in a

sealed vial for laboratory powder diffraction.

Powder diffraction (PD) data was collected at the

Australian Synchrotron (Melbourne, Australia). A

wavelength of 1.000625 angstrom was utilised,

the diffraction data was collected on the

MYTHEN position sensitive detector system. This

detector is comprised of 16 MYTHEN modules

each covering ~4.8° 2 , such that the system

covers ~80° 2 with 0.2° gaps every 4.8° 2 . Two

dataset were collected per sample, each for 300

seconds, the repeat collected after moving the

detector bank 0.5°, the two datasets were

subsequently joined using a conversion tool

described elsewhere [3].

Figure 1: Photo of the Powder Diffractometer at the

Australian Synchrotron.

The crystalline phases present in each sample

were determined using search match routines in

both Bruker-AXS EVA 10.0 and MDI Jade 6.0

software packages to search the powder

diffraction file (PDF) database.

Rietveld Quantitative Phase analysis was

performed using the Bruker-AXS TOPAS 4.0

software package. The crystal structures used

were selected from the Inorganic Crystal Structure

Database (ICSD) (version 2008/1). The line

profile function was refined using relevant

standards utilising the Thompson-Cox-Hastings

model. The background was modelled using an

order 2 Chebychev polynomial and two broad

split pseudo-Voigt peaks. The following

parameters were refined for each crystalline

phase: scale, lattice parameters, Lorentzian

crystallite size (i.e. sample contribution is a

Lorentzian function convolved such that the

FWHM varies as a function of cos ), atomic

displacement parameter for phases > 5 wt%. The

data resulting from this process is the absolute

crystalline phase quantification and the

amorphous quantification.


Preliminary results from Rietveld quantitative

phase analysis of laboratory based x-ray

diffraction data showed that the method is viable,

however many of the minor phases are near the

detection limit providing poor accuracy and

precision in the quantification. Powder diffraction

data has been collected at the Australian

Synchrotron, the improved signal to noise and 2-

theta resolution lowers the detection limit of these

minor phases, improving accuracy and precision

of the quantification, these results will be

presented.

Table 1: A comparison with other studies of the

Quantitative Phase analysis of NIST SRM 1633b Fly

ash.

Th

is s

tud

y

Win

bu

rn e

t a

l

20

00

[1

]

Wa

rd a

nd

Fren

ch

20

06

[2

]

Phase Name wt% ± (2 esd) wt% ± (2 sd) wt%

Mullite 20.9 0.2 21.8 1 22.2

Quartz 6.53 0.08 6.1 0.4 7.1

Maghemite 3.3 0.2 2

Magnetite 2.6 0.1 3.9 0.4 2.2

Hematite 1.5 0.1 2.2 0.4 1.1

Gypsum 1 0.2

Calcium

aluminate 1.1

Amorphous 65.1 0.3 65 2.2 64.3

To gauge the accuracy of the method employed in

this study a NIST standard reference material fly

ash was also characterised, the results are

compared to other studies in Table 1. In this study

the concentration of the iron oxides phases

(maghemite, magnetite and hematite) were

determined to have slightly higher total

concentrations than in the other studies (7.4, 6.1

and 5.3 wt% for this study, Winburn et al. [2] and

Ward and French [1] respectively), this is to be

expected, as the other studies employed Cu Kalpha

radiation for which there is significant

microabsorption effects with iron, resulting in a

negative bias on phase quantification of iron

containing phases [4]. The other phases are

comparable, however no gypsum nor calcium

aluminate were detected in the sample.


Rietveld quantitative phase analysis (RQPA) was

undertaken on several local fly ashes, the

amorphous concentration has been determined.

Future work involves using this data to calculate

the amorphous composition of each fly ash and

produce viable geopolymers with each.

Acknowledgements





Centres Program.

This research was undertaken on the Powder

Diffraction beamline at the Australian

Synchrotron, Victoria, Australia. The views

expressed herein are those of the authors and are

not necessarily those of the owner or operator of

the Australian Synchrotron

Thank you to Hans Fairhurst of Cockburn Cement

and Craig Heidrich of Ash Development

association for providing the fly ashes.

References

1. Ward, C.R. and D. French, Determination of glass

content and estimation of glass composition in fly ash

using quantitative X-ray diffractometry. Fuel, 2006. 85(16): p. 2268-2277.

2. Winburn, R.S., et al., Rietveld quantitative X-ray

diffraction analysis of NIST fly ash standard reference materials. Powder Diffraction, 2000. 15(3): p. 163-172

3. Williams, RP and van Riessen A, Data processing of

diffraction data from the MYTHEN detector system on

the Powder Diffraction beamline at the Australian

Synchrotron. Centre for Materials Research (CMR)

Technical Note, 2008, Curtin University of Technology.

Note# 001

4. Madsen I.C., Scarlett N.V.Y., Cranswick L.M.D. and

Lwin T. Outcomes of the IUCR CPD Round Robin on

Quantitative Phase Analysis: Samples 1a to 1h, J.Appl.

Cryst, 2001 34, 409-426.

Synthesis Paths and Performance Evaluation of Geopolymer Binder Systems

Derived from Major Mineral Processing and Mining Waste Feedstock Materials

K. Sagoe-Crentsil and T. Brown

CSIRO, Materials Science and Engineering, VIC

Introduction

The base chemical and mineralogical composition

of several large volume industrial, mining and

mineral processing byproducts fall within the

compositional limits of feedstock materials

deemed suitable for Geopolymer binder synthesis.

Such materials include, fly ash from coal-fired

electric power plants, red mud residues from

alumina processing as well as blast furnace slag

from iron ore processing. Geopolymer research

has thus attracted considerable research interest

[1-2] in recent years, partly due to its potential to

utilize diverse industrial wastes in large scale

mining, geotechnical and civil construction

applications. In essence, the underpinning

chemical reactions governing aluminosilicate

solids and alkali silicate solutions [3-4] to form

Geopolymer binders can, in principle, be

amenable to a range of industrial waste forms.

This study therefore attempts to examine the

relationships between chemical formulation,

micro structure and mechanical properties of

selected fly-ash based Geopolymer systems to

explore compositional limits relevant to mining

waste streams eg red-mud mineralogy.

Specifically, this paper draws on experimental

studies to examine the relationships between

chemical formulation, microstructure and

mechanical properties of high silica Geopolymer

systems with SiO2/Al2O3 > 15 as an alternative

binder to conventional geopolymers having

SiO2/Al2O3 = 3-4.

Methodology

Several Geopolymer formulations with

conventional Si contents, termed low-Si

Geopolymer, were prepared with different

proportions of ingredients selected to allow the

effect of silica content to be assessed. Samples

with compositional ratios i.e. SiO2/Al2O3 = 2.7 to

3.9 differing in the proportion of silica added in

solution compared to that obtained from the solid

fly ash component, as detailed in Table 1. The

high-silica samples required additional solid silica

to achieve required Si dosages. The formulations

were cured at 85°C for 2 hours. Compressive

strengths were measured on 25.4 mm cubes and

SEM was performed on fracture faces.

The investigation focussed on the effects of molar

concentrations of SiO2 through a combination of

electron microscopy and strength testing.

Table 1: Nominal chemical composition of Geopolymer

systems.

Nominal mix composition

Silicate series

SiO2/Al2O3

Ratio

Na2O 2.7SiO2 Al2O3 10H2O 2.7

Na2O 3.0SiO2 Al2O3 10H2O 3.0

Na2O 3.5SiO2 Al2O3 10H2O 3.5

Na2O 3.9SiO2 Al2O3 10H2O 3.9

Key Results

Low-silica content Geopolymer mixtures:

The microstructures of Geopolymer samples

appeared similar to images previously published

for these materials [5, 6]. The condensed gel

phase in the samples consisted of rounded

growths with bridging between them (Fig. 1).

Variations exist for microstructures of

geopolymer systems with SiO2/Al2O3 = 2.7 and

3.9. The fineness of the texture and the density of

the geopolymer increases greatly from SiO2/Al2O3

= 2.7 to 3.9. The images reveal higher silica

samples contain more unreacted particles, and

also that crystals have grown on a number of the

unreacted fly ash particles, particularly in the

higher silica mixture (Si/Al 35) as shown in the

microstructure of Figure 1.

The compressive strengths mirrored observed

trends in the microstructure in the expected way,

in that the densest, finest grained samples were

also the strongest. In particular, the change in

strength with silica content was quite dramatic,

given that a molar ratio increase of 11% Si from

SiO2/Al2O3 = 3.5 to 3.9 yielded a corresponding

62% strength increase.

Figure 1: SEM image of a low-silica Geopolymer

SiO2/Al2O3 =3.9. Scale bar represents 1 m.

High-silica Geopolymer mixtures:

The high silica geopolymer systems exhibited

extremely dense microstructures; much denser

than conventional geopolymers. The dense

microstructures suggest that high silica

geopolymers could possess good mechanical

properties and bonding characteristics. Figure 2

shows the strength development characteristics of

Si-rich geopolymer systems with varying

SiO2/Al2O3 content ranging from SiO2/Al2O3 =

15-56. The compressive strength was a maximum

at around SiO2/Al2O3 = 32. It must also be noted

that the pH dropped with increasing silica content

to about SiO2/Al2O3 = 38 and remained constant

beyond that. The high-silica content formulations

were also characterised by excess water demand.

0

10

20

30

40

10 30 50 70

SiO2/Al2O3

Co

mp

re

ss

ive

Str

en

gth

(MP

a)

Figure 2: Effect of varying silica content on compressive

strength of high- Si Geopolymer systems.

Discussion

It was observed that the investigated fly ash-based

formulations had characteristic complex

microstructures and structural variety. In

particular, samples with the highest strength and

the densest microstructure tended also to contain

crystalline components.

Compressive strengths in general can be related to

both composition and microstructure in a logical

way. These general observations and their

implications on feedstock material selection may

be best understood by examining the fundamental

dissolution and condensation reactions occurring

during synthesis. Given that Si solubility, is

generally much less than aluminium under

alkaline conditions, the condensation reaction in

geopolymers will likely occur between

monomeric [Al(OH)4]– and a variety of silicate

species, including monomers and oligomers. The

silicate solutions used in the production of

geopolymers, typically have high SiO2

concentrations above 5M, and M2O/SiO2 ratios of

0.66~0.83 by the addition of NaOH which is

further likely to generate other polymeric silicate

species, as demonstrated by Barbosa et al[7]. In

relation to the potential use of industrial waste

streams e.g. red mud residues from the Bayer

process, the key reaction determinant will likely

occur during the condensation reaction between

monomeric [Al(OH)4]– and a variety of silicate

species, including monomers and oligomers.

Conclusions

Observations described in this paper broadly

follow expected theoretical correlations between

geopolymer mix composition, microstructure and

strength. It is observed that the high silica

geopolymer formulations investigated achieve

near optimal high strengths at SiO2/Al2O3 35

and show characteristic low porosity and dense,

fine grained microstructures. The observed

synthesis parameters further suggest potential

beneficial and novel applications of red-mud as

raw feedstock material in a variety of building

product and civil construction applications.

References

1. Davidovits, J, Davidovits, M, Davidovits, N, US Patent

5342595, 1994.

2. Fernández-Jiménez, A.; Palomo, A.; Criado, M.,Cem. Concr. Res.,35, Issue: 6, June, 2005, pp. 1204-1209.

3. Van Jaarsveld, J.G.S, van Deventer, J.S.J, and

Lorenzen, L, Met. Mater. Trans. B 29B (1998) 283– 291.

4. Steveson, M, and Sagoe–Crentsil, K, J. Mater. Sci.,40 (2005) 4247-4259.

5. Phair, J.W, and Van Deventer, J.S.J, Miner. Eng. 14

(2001) 289–304.

6. Van Jaarsveld, J.G.S, Van Deventer, J.S.J, and Schwartzman, A, Miner. Eng. 12(1) (1999) 75–91

7. Barbosa, V, MacKenzie, K, Thaumaturgo, C, Int. J. Inorg. Mater. 2 (2000) 309.

Managing Coal-Fired Power Station Solid By-Products

J.T. Gourley

Geopolymer Alliance, WA


Coal-fired power stations are focused on producing

Energy as their primary product. As a general rule

they are forced to treat their secondary products

(flyash, bottom ash and heat) as waste products

since most see the effort to process and market

these bi-products as a distraction from the core

business. From the view of sustainable resource

processing this is unfortunate. With minimal effort

these materials can be converted to useful products

which can bring with them considerable greenhouse

gas credits when used to supplement or replace

existing products.

If addressed as a separate business all solid wastes

can be converted to useful products. Future market

demand will be such that all currently produced

wastes will be saleable as products and previous

wastes that are currently stored in tailings dams or

dry storage sites will be required to be recovered

and processed.


It is theoretically possible to convert all solid wastes

(flyash and bottom ash) from coal fired power

stations, into useful products for the construction

industry. Potential by-products include:

• Activated carbon or char, as a filtering medium

or in the steel industry.

• Coarse flyash as an extender for Ordinary

Portland Cement (OPC).

• Cenospheres.

• Conventional cement-sized flyash (Classified

flyash) as a supplementary cementitious

(SCM’s) material for inclusion in OPC blends.

• Coarsely milled Run-Of-Station (ROS) flyash

as a supplementary cementitious material for

inclusion in OPC blends.

• Finely milled ROS ash as an OPC substitute.

• Very finely milled classified flyash as a

substitute for silica fume.

• Bottom ash as sand for use in geopolymer

concrete or in high volume flyash OPC

concretes.

If a low temperature drying stage is included, flyash

currently stockpiled in tailings dams can be brought

on stream to produce milled ash.

Why does the construction industry need such products?

Figure 1 illustrates the predictions of the cement

industry for cement usage out to 2020, and the

consequent production rates of carbon dioxide. If

the industry is to attempt to meet community

expectations of significant reductions in greenhouse

gas emissions over this period then large volumes of

supplementary cementitious materials (SCMs) are

going to needed very quickly. The environmental

challenge facing the cement industry over the next

40 years means that ordinary Portland cement will

have to be largely replaced SCMs or fully, by the

use of geopolymer cements.

0

5 0 0

1 , 0 0 0

1 , 5 0 0

2 , 0 0 0

2 , 5 0 0

3 , 0 0 0

3 , 5 0 0

4 , 0 0 0

1 9 9 0 2 0 0 5 2020 (1) 2020 (2) 2020 (3)

Glo

bal O

utp

ut

(millio

n t

on

nes / y

ear)

SCM

Cemenn t

CO2

Figure 1: Predicted cement production.

(Ref:http://www.concretesdc.org/Sustainability/Sustainability_h

ome.htm).

1990 and 2005 values are the actual global

quantities used (cement and SCMs) and the CO2

masses generated from the production of these

binder masses.

2020 (1) is the prediction for “business-as-usual”

but with some attempt to increase SCM proportions

to 20% by mass of binder (the current hopes of the

cement and concrete industry; Ref

http://www.concretesdc.org/Sustainability/Sustaina

bility_home.htm).

2020 (2) is the prediction if the CO2 levels are to be

kept at 2005 levels.

2020 (3) is the prediction if they are to be reduced

to 1990 levels.

The CO2 contributions are based on the following

assumptions:

• the cement clinker contribution reduces from

1.0 t/t in 1990 to 0.8t/t in 2020 as the fuels used

to fire the clinker kilns change;

• the SCM contribution increases to 0.15 t/t in

2020 as milling of coarse flyash becomes

necessary to provide sufficient feedstock at

these high replacement levels.

Scenario (1), business-as-usual, is not an option if

the construction industry is take up its share of the

burden of addressing climate change.

Scenario (2) is just possible with OPC concrete as

usually OPC (the primary binder component) must

be in excess of flyash content for usable concrete

strengths.

Scenario (3) is not possible with OPC concrete

and normal classified flyash and is only

achievable with OPC-Engineered flyash/slag

blends, or with geopolymer concrete where the

SCM (flyash, slag, etc) becomes the (dominant)

binder fraction.

This latter scenario requires nearly 3 billion

tonnes of SCM per annum, worldwide.

The most likely SCMs to be used will be processed

fly ashs (EFAs) and/or ground granulated blast

furnace slags (GGBFSs).

The cost to grind flyash is significantly lower than

that of slags as the vast majority of medium to

coarse flyash particles are hollow spheres that crush

like eggshells. In addition such a process tends to

produce a single-mode particle size distribution

whereas grinding hard, solid slag particle tends to

produce bi-modal distributions; a population of

small chips and a population of large chipped parent

particles.

Just as particle size distribution (PSD) is important

in designing a low cost concrete aggregate skeleton,

so it is in binder design. Optimum particle packing

in binders allows the use of minimum quantities of

fluids to achieve the required workability or flow-

ability of the concrete. As OPC concrete strength

and durability characteristics are inversely

proportional to fluids content (air and water in the

case of OPC concrete) a perfectly packed binder

fraction allows the production of a superior

concrete. As Australian cements do not in general

get close to perfect packing with their PSDs, the

incorporation of EFAs and GGBFSs will produce

superior concretes and at the same time convert

wastes to valuable by-products and reduce

greenhouse gas emissions.

Geopolymer binders just use SCMs as the binder

and as a result of the kinetics of the

geopolymerisation process, allow the use of sintered

or pelletised aggregate particle made from flyash or

slag. Thus geopolymer concretes are well suited to

the use of aggregates which would otherwise be

shunned in conventional concrete mix designs.

In fact, it is possible to make a very commercial

geopolymer concrete entirely from ash and slag

wastes. Flyash can be pelletised using a

geopolymer process to form coarse aggregate.

Larger particles of granulated slag can be used as

coarse aggregate. Bottom ash or fine granulated

slag can be used as the fine aggregate (sand).

Coarse flyash can be used as the binder coarse

fraction. Classified flyash or lightly crushed Run-

Of-Station flyash can be used as the cement-sized

binder faction and fully crushed flyash (EFA) can

be used as the fine binder fraction.

Design your concrete mix for optimum particle

packing, add a small quantity of concentrated

alkaline solution and you have a low-cost, high

performance concrete which enhances the

sustainability of both the power generation and

construction industries.


The Geopolymer Alliance is working to encourage

the power industry to review these sustainability

opportunities and join with research providers

within CSRP to investigate possible process and

product options.

Thermal Character of Geopolymers Synthesised from Class F Fly Ash Containing

High Concentrations of Iron and -Quartz

W.D.A. Rickard and A. van Riessen


Background

Geopolymers are a synthetic inorganic polymer

produced by the alkali activation of

aluminosilicate raw materials. In recent years fly

ash has been proposed as a suitable

aluminosilicate source for use in geopolymeric

materials. The use of fly ash allows for a less

expensive alternative to pure sources as well as

utilising an industrial waste. Due to their

inorganic framework, geopolymers have excellent

thermal stability, far in excess of traditional

cements [1]. These characteristics allow

geopolymers and geopolymer composites to be

considered for use in high temperature

applications such as furnace linings, thermal

insulation and wall panels.

Previous investigations into to the thermal

properties of geopolymers have been conducted

using pure aluminosilicate sources or have not

considered the effect of the impurities introduced

through the fly ash on the bulk thermal properties.

This paper reports the thermal characteristics of

geopolymers prepared with a class F fly ash

containing 15 wt% iron oxide and 20 wt% -

Quartz, upon heating to 900 °C. Characterisation

techniques to be presented include thermal

expansion, TGA, DTA, XRD and SEM.


Geopolymer samples in this study were

synthesised using a Western Australian fly ash as

the aluminosilicate source material. Samples were

prepared to achieve compositional ratios of Si:Al

= 2.3, Na:Al = 0.85 and H2O:SiO2 = 2.0 by

calculating concentrations based on the

amorphous components of the fly ash only. A DI-

24 Adamel Lhomargy dilatometer was used for

the thermal expansion measurements (Figure 1).

All measurements were conducted in accordance

with ASTM E831. Moulded dilatometry samples

were cast into a 1 mL syringe and cut to length to

achieve a cylinder of 15 mm length and 5 mm

diameter. Other dilatometry samples were cut

from larger samples using a diamond saw to

achieve an 8 x 8 x 15 mm rectangular specimen.

Larger, ‘bulk’, samples were moulded into 15 mm

diameter cylinders and cut to a 20 mm length.

Thermo-Gravimetric Analysis (TGA)/Differential

Temperature Analysis (DTA) measurements were

performed using a SDT Q600 V3.4 Texas

Instrument in both air and nitrogen atmospheres.

Figure 1: DI-24 Dilatometer.


Figure 2 shows the dilatometer curves for fly ash

geopolymers of various sample dimensions.

Included on the graph is a regional segregation of

the various expansion / contraction stages

observed during heating to 900 °C. The regional

model expands on what was initially proposed by

Duxson et al. [2]. The geopolymers in this study

exhibited fundamentally different thermal

expansion at high temperatures to what has been

observed previously in the literature. It is believed

by the authors that this is due to the presence of

impurities introduced via the fly ash precursor and

the effect of sample dimensions.

Figure 3 compares the simultaneous TGA / DTA

thermograms of the geopolymers used in this

study exposed in both air and nitrogen. An

endothermic peak associated with the dehydration

of adsorbed water was observed at approximately

80 °C for both atmospheres. A broad exothermic

peak at 400 °C was observed for the sample in air

which was not observed for the sample in

nitrogen, suggesting the presence of an oxygen

dependant reaction.

Figure 2: Thermal expansion of fly ash geopolymers. Si :Al

= 2.3, H2O:SiO2 = 2.0.

Three major mass change events were observed to

occur in the geopolymer samples upon heating to

1000 °C in air. Dehydration of free water

accounted for 80 % of the total weight loss,

whereas dehydroxylation was approximately 10

%. Interestingly the sample gained weight above

600 °C as the iron species oxidised (figure 3, air

TGA).

Figure 3: TGA / DTA results for a Collie fly ash

geopolymer.

The authors of this study suggest the following

two stage phase transition of ferrihydrite to

hematite describes the oxidation reaction observed

in figure 3;

OHFeOFeOOH + Eq. 1

3222 OFeOFeO + Eq. 2

Optical investigations revealed the exposure of

geopolymers to 900 °C in air changed the colour

of the specimens from grey (similar to that of the

original fly ash) to dark red (similar to that of

hematite). The colour change was not uniform in

the interior of the larger samples. Large radial

cracking was observed in exposed samples.

Observed differences in the microstructure after

high temperature exposure included an increase in

porosity and a reduction in the amount of intact

fly ash spheres, as most had melted during

exposure.

Analysis of XRD patterns indicated an increased

concentration of hematite after heating to 900 °C

(Figure 4). Other crystalline phases did not vary

significantly.

Figure 4: XRD patterns of the major hematite peaks.

Continuous line: Unexposed geopolymer, Dashed line:

Thermally exposed geopolymer.


Geopolymers synthesised from fly ash containing

high concentrations of iron oxides and crystalline

silica exhibited variable thermal properties at

temperatures greater than 500 °C. This was due to

interactions of the secondary phases, most notably

the iron oxides, with the atmosphere and sample

degradation through cracking. The results of this

study suggest that quartz rich fly ashes do not

significantly reduce the thermal performance of

the resulting geopolymer as key characteristics

such as thermal expansion and phase stability

were largely unaffected during thermal exposure.

Geopolymers synthesised from iron rich fly ashes

were observed to reduce the thermal performance

of the sample due to phase transitions and

oxidisation reactions inducing variable thermal

expansion at high temperatures.

Acknowledgements





Centres Program.

References

1. Duxson, P., Provis, J.L., Lukey, G.C., Mallicoat, S.W.,

Kriven, W.M., van Deventer, J.S.J., Understanding the

relationship between geopolymer composition,

microstructure and mechanical properties. Colloids and

Surfaces A: Physicochemical and Engineering Aspects

2005. 269(1-3): p. 47-58.

2. Duxson, P., G.C. Lukey, and J.S.J. van Deventer,

Physical evolution of Na-geopolymer derived from

metakaolin up to 1000 °C. Journal of Materials Science,

2007. 42(6): p. 3044-3054.

Rational Utilisation of Fly Ash for Geopolymer Processing

J. Temuujin, R. Williams and A. van Riessen



As demand for concrete as a construction material

increases, the production of Portland cement will

also increase. However, production of Portland

cement liberates a considerable amount of

greenhouse gas as a result of de-carbonation of

limestone in the kiln during manufacturing of

cement and the combustion of fossil fuels.

Furthermore, Portland cement is also among the

most energy-intensive construction materials,

after aluminium and steel [1]. Therefore, the

search for an alternative binder to replace Portland

cement binder is important endeavour.

Importantly, such so-called “geopolymer”

materials not only have comparable or superior

properties to traditional cementitious binders, but

also have low greenhouse emissions. The source

materials for the production of geopolymer are

usually fly ash, metakaolin and slag. Fly ash is

thought to be good candidate for source material

because it is the residue from coal burnt in a

thermal power plant and regarded as a waste

product. However, fly ash also contains some

unburnt carbon with the following effects [2]:

• Increases the electrical conductivity of

concrete;

• Changes colour of mortar and concrete (may

appear black);

• Increases demand for water and additives.

Therefore, carbon free fly ash based geopolymers

are expected to have better mechanical properties

than carbon containing fly ash based

geopolymers.

For production of geopolymers low calcium class

F fly ash is preferred as the source material rather

than high-calcium class C fly ash as the presence

of high amounts of calcium may interfere with the

polymerisation process and alter the

microstructure. However, addition of a small

amount CaO to fly ash/cement system accelerates

the hardening process and increases strength [3].

In fact, one of the most important aspects of

adding calcium to a geopolymer system is its

ability to harden at ambient temperature.

In an attempt to improve the performance of fly

ash-based geopolymers and facilitate curing at

room temperature, fly ash was calcined and

calcium compounds were included in the

formulation, respectively.


Collie fly ash was used as the source material for

production of the geopolymer. For de-

carbonation, the fly ash was calcined at 500 and

800oC. Sodium silicate D-51 (PQ Australia Pty

Ltd.) and sodium hydroxide solution with 14 M

concentration were used as the activating

solutions.

Analytical grade calcium oxide (CaO) and

calcium hydroxide Ca(OH)2 were used as calcium

sources. One, 2 and 3 wt. % calcium oxide was

substituted for fly ash in the geopolymer mixture.

The geopolymer samples were made with a Si:Al

ratio of 2.3 and a Na:Al ratio of 0.88. Mechanical

properties were evaluated by measuring

compressive strength of cylindrical samples.


Calcination of fly ash

Preliminary calcination of fly ash caused

approximately a 3% decrease of the amorphous

component of the fly ash while at same time

increasing the crystalline phases; mullite and

hematite. Therefore, 500 and 800oC calcined fly

ash can be considered as carbon free, but with

slightly different amorphous and crystalline

compositions. However, carbon free fly ashes

show weaker compressive strength than raw fly

ash. Figure 1 shows influence of fly ash

calcination temperature on compressive strength

of the geopolymers.

Figure 1: Influence of fly ash calcination temperature on

compressive strength of the geopolymers.

0

20

40

60

80

0 500 1000

Temperature, oC

Co

mp

res

siv

e s

tre

ng

th,

MP

a

Calcination of the fly ash resulted in a decrease of

the compressive strength of the samples from

55.7(9.7) MPa in non treated fly ash based

geopolymer to 44.3(5.4) MPa for 800oC calcined

fly ash based samples. Geopolymer prepared

using raw fly ash shows 0.54 (0.1) % volume

shrinkage while for the 500 and 800oC calcined

fly ash samples the shrinkage was 0.83 (0.25) and

5.45 (0.40)%, respectively. Also efflorescence

was observed on the surface of the calcined fly

ash based geopolymers. The results indicate that

the geopolymerisation reaction in the calcined fly

ash sample was less intensive and consequently

depends strongly on the surface conditions of the

fly ash. The difference between the raw and

800oC calcined fly ash in terms of the amorphous

and crystalline components is only 3 wt%. But

there are very large differences in terms of

geopolymerisation. The heat treatment or

calcination causes crystallisation of the

aluminosilicates and hematite on the surface of

the spherical particles. Such crystallised

aluminosilicate (mullite) and hematite prevent

dissolution of the aluminate and silicate species

by the activating alkali, consequently there is

excess alkali which is able to migrate to the

surface of the geopolymer during curing which

then crystallises as sodium hydrate carbonate.

Addition of Ca

Figure 2 shows compressive strength changes of

the calcium added geopolymer samples cured at

ambient and 70oC.

0

10

20

30

40

50

60

70

80

90

0 1 2 3 4

Added Calcium Compound, %

Co

mp

ress

ive

stre

ng

th,

MP

a

CaO added and

20C cured

Ca(OH)2 added

and 70C cured

Ca(OH)2 added

20C cured

CaO added

70C cured

Figure 2: Compressive strength versus wt% calcium

addition.

Adding calcium compounds improved the

mechanical strength of samples cured at room

temperature while reducing the strength of

samples cured at 70oC suggesting the possibility

of different mechanisms of hardening. The data

also indicates that the addition of Ca(OH)2 is more

beneficial than CaO in terms of compressive

strength. Calcium compound addition is likely to

result in precipitation of calcium silicate hydrate

or calcium silicate aluminate hydrate phases and

at the same time improve the dissolution of the fly

ash in the alkaline medium and subsequently the

geopolymerisation reaction.


• Based on the fly ash calcination experiments

presented here, it is suggested that removal of

unburnt carbon from fly ash by calcination is

counter productive. The calcination may

remove the carbon but the concomitant

changes to the surface of the fly ash limit the

degree of polymerisation.

• Addition of the calcium compounds CaO and

Ca(OH)2 improves mechanical properties of

the fly ash based geopolymers cured at room

temperature. Calcium compound addition

reduces mechanical properties of geopolymer

cured at elevated temperatures. Ambient

temperature cured samples exhibit

efflorescence because of the presence of the

excess alkaline solution caused by incomplete

dissolution of the fly ash spheres and low

dissolution rate. Efflorescence formed on the

Collie fly ash based geopolymer is sodium

phosphate hydrate.


As a means of removing unburnt, carbon

calcination of the fly ash is not considered to be

beneficial for both economic reasons and the

reduced physico mechanical properties of the

geopolymers. Addition of the calcium

compounds into geopolymer mixtures cured at

ambient temperature improves mechanical

properties. The reason for the improved

mechanical properties is believed to be due to an

increase in the geopolymerisation reaction and

possible formation of Ca-containing components.

Acknowledgements





Centres Program.

References

1. D.Hardjito, S.E.Wallah, D.M.J.Sumajouw, and

V.Rangan, “Fly ash-based geopolymer concrete”, Aust. J. Struct. Eng. 2005, 6: 1-9

2. T.H.Ha, S.Muralidharan, J.H.Bae, Y.C.Ha, H.G.Lee,

K.W.Park, D.K.Kim, “Effect of unburnt carbon on the

corrosion performance of fly ash cement mortar”. Construction and Build.Mater. 2005, 19: 509-515

3. S.Antiohos and S.Tsimas, “Activation of fly ash

cementitious systems in the presence of quicklime Part

1. Compressive strength and pozzolanic reaction rate”, Cement and Concrete Res. 2004, 34: 769-77

Sustainability Roadmap for the Kwinana Industrial Area / Extended Abstract

K. Schianetz1, M. John

1, D. van Beers

2, C. Oughton

3 and D. Cooling

3

1Curtin University of Technology, Centre of Excellence in Cleaner Production, WA 2GHD, WA 3Kwinana Industries Council, WA


The Kwinana Industrial Area (KIA) is by far the

largest and most diverse industrial processing

region (with supporting industries) in Western

Australia. As with elsewhere in Australia (and

globally), the KIA is facing sustainability

challenges on various fronts, such as increasing

water and energy scarcity, climate change, an

aging workforce, and growing community

expectations.

Through the Kwinana Industries Council (KIC)

the industries within Kwinana are working closely

on common issues in the KIA, including the

sustainable development of the region. Over the

past decade, significant progress has been made

by the industries towards the improvement of the

sustainability performance at the company and

regional level. Given the ever increasing

sustainability pressures, there is strong desire to

address sustainability challenges in the KIA,

through a pro-active and collaborative industry

approach. Since April 2004 the CSRP research

project on ‘Capturing Regional Synergies in the

KIA’ is assisting the KIC and its members on

their journey towards sustainability. The overall

objective of the project is to provide practical

support to the companies in the KIA to develop,

evaluate and implement regional synergy

opportunities, thus creating greater eco-

efficiency1.

Within the framework of this project a

sustainability roadmap is now being developed to

assist the KIC and its industry members with

strategic decision making on the long-term

sustainability of the KIA. The broad aims of the

Sustainability Roadmap are to:

• Guide KIC strategic decision-making on the

long-term sustainability of the KIA;

• Provide the basis for stakeholder discussion

on sustainable development;

• Scope sustainable development initiatives;

• Document progress towards sustainability

goals and targets; and

• Benchmark the sustainability performance of

the KIA with other industrial regions and

industry sectors in Australia and

internationally.

The paper will report on the development process

and the preliminary results of the sustainability

roadmap for the KIA, including a historical

perspective on milestones achieved so far and

future aspirations.

Development of a Framework of the KIA

Sustainability Roadmap

The Sustainability Roadmap for the KIA is work

in progress. In a first step a draft framework of the

KIA Sustainability Roadmap was developed

(Figure 1), consisting of a KIA Case-Studies

Matrix and KIC Sustainability Management

Matrix.

Past Present Future

Documentation of

KIA case-studies

KIA SUSTAINABILITY ROADMAP

“Where are we now, where do we want to be, and how do we get there?”

INDIC

ATORS

Current

performance

Aspirational

performance

targets

Time

KIA

su

sta

ina

bil

ity

pe

rfo

rm

an

ce

Low

High

KIC action plans for

selected priority areas

Past

performance

KIC SUSTAINABILITY

MANAGEMENT MATRIXINDIC

ATORS

KIA CASE-STUDY MATRIX

Figure 1: KIA Sustainability Roadmap – Framework1

The figure illustrates how sustainability indicators

are used to measure past and current performance

of the KIA and to set aspirational targets. The

KIA Case-Study Matrix could be used to

document significant past events and initiatives

that have led to the KIA sustainability

performance of the present. The KIC

Sustainability Management Matrix will assist the

KIC with formulating action plans for selected

priority areas to achieve the aspirational targets.

The Choice of Sustainability Indicators

The roadmap is based on sustainability indicators,

which relate economic, social and environmental

factors, and therefore provide integrated

information as to how the KIA is progressing

toward Triple Bottom Line sustainability.

The choice and use of indicators for sustainable

development is a critical determinant to sound

decision-making processes2. In particular, it is

argued that tailor made indicator sets, addressing

specific stakeholder concerns and supporting the

organisations’ strategies, are more likely to assist

stakeholders in the achievement of their goals

than approaches which prioritise reporting against

generic ‘off the shelf’ indicators3.

The search for sustainability indicators is

evolutionary2. The necessary process is one of

collective learning. This learning process at the

KIC started out with a literature review of

regional sustainability indicator systems being

used internationally, and indicators used by

companies in the KIA. The Sustainability

Reporting Framework of the Global Reporting

Initiative (GRI) was used as a basis for the

development of the initial set of KIA

sustainability indicators4. The GRI framework is

the world’s most widely used framework for

reporting on an organisation’s economic,

environmental, and social performance.

The initial list of possible indicators has been

discussed with different stakeholder groups and

generated the framework of the KIC Sustainability

Management Matrix. At the heart of the matrix is

a set of key performance indicators for each of

seven priority themes, namely greenhouse gas

emissions and energy, water, process residues,

economic performance, air quality and noise,

community, and workforce. Through the KIC

Eco-Efficiency Committee, the matrix is being

populated with relevant primary and secondary

performance indicators, aspirational targets, and

subsequent strategies and action plans to achieve

the targets.

The selection of relevant sustainability indicators

is subject to the following criteria:

• Measurable (relatively easily obtainable and

build upon existing data);

• Understandable for internal and external

communications;

• Influenceable by KIC and its industry

members;

• Relatively small number of key indicators;

• Provide holistic overview of the sustainability

performance of the KIA; and

• Provide the basis for benchmarking KIA with

other (inter)national industrial areas.

As part of the extension of the CSRP Kwinana

Synergies Project, the KIC Sustainability

Management Matrix will be further developed

with KIC internal and external stakeholders. This

will include collection of historical industry data

to quantify selected sustainability indicators

(historical trend analysis), documentation of

Kwinana case-studies, and engagement with KIC

industry members, government and community.


The KIA Sustainability Roadmap is work in

progress. The work so far focused on the

development of the roadmap framework, the

review international sustainability indicators, and

the development of the KIC Sustainability

Management Matrix, including the selection of

indicators relevant to the KIC. The next steps

include setting of five year aspirational targets and

formulating subsequent strategies and action plans

to meet these targets. The focus will be on

activities where the KIC has an influence and can

make a positive impact. Engagement and

discussions with KIA stakeholders will be an

important part of this process. In its final form, the

KIA Sustainability Roadmap will enable ongoing

constructive stakeholder discussions (e.g.

government and community) on emerging local

issues and possible future directions for the KIA.

Acknowledgements





Centres Program. The authors wish to

acknowledge the commitment and support from

the Kwinana Industries Council and its members

to the research presented here.

References

1. Van Beers, D. 2008 “Capturing Regional Synergies in

the Kwinana Industrial Area. 2008 Status Report”, CSRP, Perth, Western Australia.

2. Meadows, D. 1998 “Indicators and Information Systems

for Sustainable Development. A Report to the Balaton

Group”. The Sustainability Institute, Hartland Four Corners, VT.

3. Warhurst, A. 2002 “Sustainability Indicators and

Sustainability Performance Management”. Report

commissioned by the Mining, Minerals and Sustainable

Development (MMSD) project of International Institute for Environment and Development (IIED).

4. Global Reporting Initiative (GRI) 2000 “Sustainability

Reporting Guidelines”. Global Reporting Initiative, Amsterdam.

Production and Application of Red SandTM

E. Jamieson, A. Jones, L. Guilfoyle, D. Cooling and S. Attiwell

Alcoa World Alumina, WA


The extraction of alumina from bauxite produces

a high volume by-product which is stored in

secure impoundments close to the refining

operations.

In Western Australia, typically half of this by-

product has a particle size in excess of 90 m due

to the high quartz content of bauxite from the

Darling Range. This coarse fraction can be further

processed to produce a general purpose sand

termed Red Sand™.

Alcoa World Alumina Australia (Alcoa) operates

three alumina refineries in Western Australia,

producing a potential resource of up to 20,000

tonnes of Red Sand™ per day.

A variety of uses for bauxite residues has

previously been outlined by Jamieson et al (2005).

Primarily Red SandTM

has been deemed to have

physical properties suitable for use as a general

fill material within the construction industry and

road construction.

Attiwell (2007) has shown that Red Sand™ can

be produced to a standard suitable for branding

with the Centre for Sustainable Resource

Processing (CSRP) ReSand® trade mark. The

CSRP brand ReSand® represents by-product sand

from the mining industry independent of source

having met application and sustainability criteria.

Alcoa has now embarked upon production of

suitable quantities of Red Sand™ to provide to the

CSRP for demonstration projects as a part of the

CSRP’s broader evaluation of ReSand®.


To demonstrate the use of Red Sand™ as

ReSand® a pilot plant has been constructed to

provide approximately 5,000 tonnes of material

for use in these trials.

The pilot plant was commissioned on 13/06/2008.

Production capacity has reached design

specification of 10 t/hr, with the plant being run

typically over an 8 hour period.

Figure 1: Delivery of the flat bottom classifier to

construction site.

Figure 2: Installation of dewater sand screen.

Key operational goals include consistent sand

washing and neutralisation via carbon dioxide

dosing within the flat bottom classifier.


Standard operation has been able to produce clean

sand washed free of fine residue and neutralised to

a pH of less than 10.5. This value can be lowered

depending upon product use requirements.

Figure 3: Sand stockpile in production.

Demonstration 1

The First application being demonstrated is the

use of sand for top dressing of turf.

Assessment of the Red SandTM

has ensured all

factors have been assessed as acceptable

including, leaching, composition, particle size,

dusting potential, phosphate retention and pH.

Preliminary trials have been successfully

completed on a cored golf course fairway,

resulting in improved drainage and greater

moisture penetration.

A larger demonstration is to use 1000 cubic

meters of sand at Fairbridge Village for top

dressing of the sports oval. Trials of Red SandTM

application to the oval (1 m2) have shown typical

top dressing performance as expected.

Figure 4: Grass growth through the sand top dressing,

before and after 25 days.

Alcoa has provided the sand including cartage to

site as part of its CSRP collaborative research

project.

Demonstration 2

The second major trial will be the use of Red

Sand™ in road construction. Assessment of

particle size, composition, leach analysis and

strength parameters have indicated trial materials

are well suited for this application.

Sand for this trial is currently being produced and

samples are being assessed to confirm results. A

freeway onramp has been earmarked by Main

Roads for construction using some 2000 m3 of this

material in the first quarter of 2009.


• Pilot plant operation has demonstrated

technical and operational capability.

• Planning for demonstration projects utilising

the Red SandTM

are well underway

• Collaboration under the guidance of the CSRP

leading to demonstrated viability of by-

product utilisation.


With the successful demonstrated uses of Red

SandTM

, Alcoa will be looking toward larger scale

production of Red Sand™ and commercialised

uses for the product.

Acknowledgements


the CSRP, which is established and supported

under the Australian Government’s Cooperative

Research Centres Program. Special thanks to the

Wagerup Refinery Residue Area Group of Alcoa

World Alumina, Australia, for continued

assistance to pilot plant operations. Thanks to

Linatex Pty Ltd for delivery of the pilot plant and

associated assistance. Thanks are extended to the

Main Roads Department WA and also Chemistry

Centre WA for assistance with demonstration

trials.

Finally thanks to Iggy Castle of Fairbridge Village

for her enthusiastic support of the trial.

References

1. Jamieson, E.J., Cooling, D.J. & Fu, J. (2005) High

volume resource from bauxite residue. Proceedings of

the 7th International Alumina Quality Workshop. Perth October 2005. pp 210- 213.

2. Attiwell, S. (2007). “Classification of ReSand® and

Leach Testing of Alkaloam® and Red Lime™”. Centre

for Sustainable Resource Processing (CSRP) Inaugural Conference. Melbourne 21st November, pp 73-74.

Using Life Cycle Assessment to Compare the Environmental Performance of

Bauxite Residue, Lime and Bauxsol™ in the Treatment of Acid Mine Drainage at

Mount Morgan Mine

D. Tuazon1, G.D. Corder

1, L.Q. Harris

2

1University of Queensland, Sustainable Minerals Institute, Centre for Social Responsibility in Mining, QLD 2Hatch Engineering, QLD

Introduction to Seawater Neutralised Red

Mud and Bauxsol™ LCA

In December 2005, a supplementary project to the

CSRP’s Project 3C1 (Developing Local Synergies

in the Gladstone Industrial Area) was initiated to

investigate the holistic environmental merits of

utilising bauxite residue (or red mud) for acid

mine drainage (AMD) remediation at Mount

Morgan, a historical gold and copper mine in

regional Queensland approximately 40 km south-

west of Rockhampton.

The findings of the initial study [1] found that red

mud was better than lime from an environmental

perspective, with the former having a much lower

carbon dioxide inventory. The study and the

results gave further impetus to pursue an

extension LCA by considering the life cycle of

another neutralant – Bauxsol™. Bauxsol™ is a

proprietary, red mud derived product from Virotec

International which is designed to treat AMD.

Like red mud, Bauxsol™ was considered as a

candidate for remediation and the Department of

Mines and Energy commissioned a study to

investigate the feasibility of using different

neutralants at Mount Morgan. The study by EWL

Sciences Pty Ltd [2] concluded that lime was the

most favourable option based on economics,

practicality and technological considerations. The

result of this study was the planning and

subsequent installation of a lime dosing plant at

Mount Morgan [3].

Note that Bauxsol™ actually refers to many

different blends of related products which are

tailored to suit different remediation scenarios. In

the case of Mount Morgan, the blend Acid B

Extra™ was suggested. Nevertheless, in this

paper, the product analysed will be referred to as

Bauxsol™.

Goal and Scope Definition

The life cycles of each of lime, seawater

neutralised red mud and Bauxsol™ involved

tracking the inventories of fuel usage, electrical

consumption and carbon dioxide emissions. The

carbon dioxide emissions inventory was decided

as the primary yardstick of comparison between

all neutralants. The functional unit and basis was

1000 cubic metres of untreated, acid mine

drainage water.

Life Cycle Determination

The life cycle of each neutralant involved the

extraction of raw materials and manufacture at

Gladstone (Queensland) before being transported

to Mount Morgan for remediation application and

disposal. In the cases of the lime and red mud

scenarios, the remediation application was using a

dosing plant which used grid electricity, whereas

with the Bauxsol™ scenario, a portable, on-site

processing facility was used for the remediation

process. The Bauxsol™ processing facility is

powered by fuel-powered generators. In all cases,

the grave of each life cycle involves disposing the

spent neutralant into the Mount Morgan pit.

Bauxsol™ presented another interesting life cycle

aspect, which was the possibility of additive lime

in the Bauxsol™.

The cradles for each of the life cycles were the

locations where the raw materials for each of the

neutralants were extracted; these were:

• Lime Scenario: East End Mine (Gladstone).

• Seawater Neutralised Red Mud Scenario: Red

Mud Dam (Gladstone)

• Bauxsol™ Scenario: Red Mud Dam

(Gladstone)

Life Cycle Inventory and Impact Analysis

Table 1 summarises the inventory data for the

base case. In the base case, it is seen that using red

mud produces 81% less carbon dioxide than using

lime and Bauxsol™ up to 92% less. However the

incremental benefit achieved by using Bauxsol™

instead of red mud is lessened or even reversed if

lime is required to be added to the Bauxsol™ as

could be the case in at least some applications.

Table 1: Key Life Cycle Assessment Results (Basis: 1000 m3

of AMD water).

Scenario Fuel

(L)

Electricity

(kWh)

CO2

(kg)

Lime 16 377 4378

Seawater

neutralised redmud

190 164 853

Bauxsol™ 70 0 335

In addition to benefits in terms of carbon dioxide

inventory, the use of Bauxsol™ requires less

electricity than using either lime or red mud

(although it should be noted that, in this study, the

Bauxsol™ production and application facilities

are powered by fuel-powered generators) and less

fuel than using red mud. The reason for less fuel

is that Bauxsol™ is more potent as a neutralant

than red mud per unit mass, thereby reducing the

amount of fuel required to transport neutralant

materials to Mount Morgan (this reduction in fuel

usage is reflected in a drop in the carbon dioxide

inventory).

Sensitivity analyses revealed that the carbon

dioxide inventory of the Bauxsol™ scenario is a

strong function of the amount of lime that is

added into the treatment media. The reason for

this is that manufacturing lime produces a

significant amount of carbon dioxide and the

environmental burdens associated with adding

lime to Bauxsol™ quickly compromise the carbon

dioxide benefits of the neutralant compared to red

mud. Analysis showed that even with the worst

base neutralisation power of Bauxsol™ of 5 mol

H+/kg, the carbon dioxide inventory of Bauxsol™

was still lower than that of both the red mud and

lime scenarios. When additive lime was

introduced into the Bauxsol™ treatment media,

the carbon dioxide inventory of Bauxsol™ was

still much lower than using lime but could easily

be higher than that of red mud depending on the

base neutralisation power of Bauxsol™. The

combined effects of neutralisation power of

Bauxsol™ and additive lime on the Bauxsol™

scenario carbon dioxide inventory can be

summarised graphically in Figure 1.

Conclusion

The study illustrates that Bauxsol™, over its life

cycle, generates less carbon dioxide and uses less

fuel than both red mud and lime, although the

benefits of Bauxsol™ compared to red mud are

less than those of red mud compared to lime. The

results also show that the improved environmental

benefits of Bauxsol™ compared to red mud may

be marginal (or even negative) depending on the

amount of lime added.

Acknowledgements

This project is carried out under the auspice and

with the financial support of the Centre for

Sustainable Resource Processing, which is

established and supported under the Australian

Government’s Cooperative Research Centres

Program.

The authors would further like to acknowledge the

contributions and support from Mansour Edraki

(The University of Queensland, Sustainable

Minerals Institute, Centre for Mined Land

Rehabilitation), Philip Bangerter (Hatch

Engineering) and Daniel Blair (Virotec

International).

Bauxsol™ and Acid B Extra™ are trademarks of

Virotec International.

Disclaimer

This study was intended to provide an objective

desktop investigation of the environmental aspects

of using red mud, lime or Bauxsol™ for

remediating acid mine drainage. It was not

intended to endorse (or otherwise) the use of any

particular neutralant, nor attest to their efficacies

as AMD neutralants.

References

1. Tuazon, D. and Corder, G.D. (2006) Life Cycle

Assessment of Seawater Neutralised Red Mud from

Gladstone for Treatment of Acid Mine Drainage at

Mount Morgan Mine. Centre for Social Responsibility in Mining: Brisbane.

2. Jones, D. (2002) Technical Input to Contingency

Planning Process and Review of Water Treatment

Options for Water in the Open Cut Pit at Mount Morgan. EWL Sciences: Darwin.

3. Unger, C., et al. (2003) Rehabilitation Plan for the

Mount Morgan Minesite Central Queensland.

Department of Natural Resources and Mines, Queensland Government: Brisbane.

Figure 1: Bauxsol™ CO2 LCI vs. Neutralisation Power of Bauxsol™ and Additive Lime Content.

Date post:	15-May-2017
Category:	Documents
Upload:	tapante
View:	222 times
Download:	1 times

CSRP08 Proceedings Lowres.pdf#Page=49

Documents