1. Processing Gold Ores Good Morning, ladies and gentlemen.
I'm here to present the paper "Process design for gold ores: a diagnostic approach"
The authors are: myself;
Professor Arthur Chaves, from University of São Paulo, Brazil,
and Professor John Meech, from University of British Columbia, Canada

2. Agenda Evolution of the gold industry
Gold ore types and characteristics
Processing gold ores
processing stages
unit operations
Process development for gold ores
An expert system for gold process design
A gold mine in Amazon
Summary

3. Evolution of the gold industry

4. Evolution of gold industry
Pre-cyanidation
Ancient Egypt (3050 a.c)
Turkey - gold coins (700 a.c.)
Gold refining (560 a.c.)
Roman Empire
Gold rushes
Spanish and Portuguese Colonies
South America
California, New Zealand, Canada, Australia

5. Evolution of gold industry
“100 years of cyanidation”
Crown Mine - New Zealand
Exploitation of Witswatersrand deposits
South Africa as main gold producer
Processing route evolution
application of basic mineral processing technology
Low incentive from low gold prices

6. Gold price over the last 50 years

7. Gold price since 1344

8. Evolution of the gold industry
“Accelerated development years ( )
Driving force: gold price
Development of gold processing
CIP process
Heap Leaching
SAG Milling
Refractory ore treatment

9. Evolution of the gold industry
Evolution of gold industry
Gold industry in the 1990’s
declining gold prices
need for cost reduction
low cash cost properties
Trends for next decade
new “rump-up” of gold prices
development of projects now “on hold”
increasing control and automation

10. Gold ore types and characteristics

11. Gold ores Gold: Process and Mineralogy
Most noble metal, native occurrence
Also associated with silver, tellurium, bismuth and PGM’s
typical ore grades: 0.5 to 20 g/t
Primary gold source: ores
Secondary gold sources
gravity concentrates
flotation concentrates
plant tailings
refinery tailings
recycled gold
* PGM means platinum group metals

12. Types of gold deposits
Placer ores
Oxidized ores
Primary ores

13. Types of gold mining methods
Placer ores => Placer mining
Oxidized ores => Open pit
Primary ores => Underground mining

14. Gold Ore Types Main ore types (after Marsden) placers oxidized
free milling
silver rich
iron sulphide bearing
arsenic sulphide bearing
carbonaceous
copper bearing
antimony bearing
gold telluride bearing
easy
processing
refractory

15. Gold ore characterization and testing
Mineralogical characterization
Technological tests
exploratory testwork
diagnostic leaching
optimization testwork
pilot plant

16. Gold Liberation “Free gold” definition and discussion
Classic definition: Gaudin
Physical liberation of the valuable mineral from gangue
Limitation: Hydrometallurgical and Pyrometallurgical processes: access of gas or solution to mineral is the issue
Definition adopted: free gold = accessible gold
direct correlation with processes
used by hydrometallurgists

17. Refractory gold
Refractory gold: cannot be made accessible using only comminution
Ore needs chemical pretreatment prior to gold extraction
Main causes of gold refractoriness
gold encapsulated in quartz/silicates
gold encapsulated in sulphides
ore containing active carbonaceous matter

18. Industrial gold ore processing

19. Gold ore processing Gold Ore Gold Gold Purification Production
Preparation
Gold liberation
Particle size adjustment to next processes
Physical concentration and oxidation pretreatment
Gold
Extraction
Extraction of gold from ore into solution and/or concentrate
Gold
Purification
Purification of gold bearing solution
Gold
Production
Recovery of gold from solution or concentrate to produce a bulliion bar

20. Unit operations for each step
Intelli Gold
Ore
Preparation
Gold
Extraction
Gold
Purification
Gold
Production

21. Process development for gold ores

22. Process development

23. Process design Process routes Geology Project context Mineralogy
Industrial processes
Process routes
Project
context
Mineralogy
Technological
tests

24. Choice of oxidation processes
Gold ore
Free gold
Refractory gold
Gravity concentration
and conventional
Cyanidation
Oxidation
followed by
leaching
Direct
Roasting
Biooxidation
Chemical
oxidation
CIL
Alternative
reagents
under
pressure
Acid
Pressure
Alkaline
Flotation
concentration

25. Feed Primary Crushing Secondary SAG Milling Crushing Rod Milling
Ball Milling
(or pebble mill)
Gravity
Agglomeration
Thickening
To smelting
Concentration
Heap
Flotation t
Oxidation c
Heap
Pressure
Roasting
Bio-oxidation
Neutralization
Oxidation
Heap
Pre-aeration
Pre-aeration
Cyanidation
Carbon-in-
Carbon-in-
Cyanidation
Cyanidation
leach
leach
Carbon-in-
Carbon-in-
Carbon-in-
CCD
CCD
column
pulp
pulp
Elution
Elution
Elution
Zinc
Zinc
Electrowinning
Electrowinning
Electrowinning
Precipitation
Precipitation
Smelting
GOLD

26. Primary Crushing Secondary SAG Milling Rod Milling Ball Milling
(or pebble mill)
Thickening
Gravity
Concentration
Agglomeration
Flotation
Heap
Oxidation
Pressure
Bio-oxidation
Roasting
Neutralization
Pre-aeration
Cyanidation
Carbon-in-
leach
pulp
CCD
column
Elution
Electrowinning
Zinc
Precipitation
Smelting
Feed
To smelting t c
GOLD
Primary
Crushing
Secondary
SAG Milling
Ball Milling
Pre-aeration
Carbon-in-
leach
Elution
Electrowinning
Smelting
Feed
GOLD

27. Fazenda Brasileiro ore
Main ore characteristics
deposit type: “greenstone belt”
mineralogy: fine gold (d95=37 microns),
free or in borders/fractures, associations with sulphides and quartz
laboratory tests: high recoveries for both direct cyanidation and flotation, good recovery in gravity concentration

28. Fazenda Brasileiro ore
Conventional process route

29. Refractory sulphide ore
Main ore characteristics
deposit type: “greenstone belt”
mineralogy: very fine gold(d95=3 microns), encapsulated and in borders/fractures, association with sulphides
laboratory tests: low recoveries in direct cyanidation and gravity concentration tests, high recoveries in flotation tests, high recoveries by cyanidation after bio-oxidation or pressure oxidation

30. Refractory sulphide ore
Processing route with pretreatment

31. Placer ore Main ore characteristics deposit type: “placer”
mineralogy: medium size gold(d95=100 microns), some physically liberated,
some associated with sulphides

32. Placer ore Physical only processing route Feed Primary Crushing
Secondary
Crushing
Ball Milling
Gravity
Thickening To
smelting
Concentration
Physical only processing route
Flotation
Smelting
GOLD

33. Expert system for gold process design

34. Intelligold - expert system for process design
Objective:
Integration of gold process design knowledge into artificial intelligence software
Principles
Use of logical variables and fuzzy sets to represent information
Organization of information into a “knowledge building” stepladder

35. Intelli Gold Project start DECISION MAKING Decision rules, fuzzy sets
Ore/deposit
information
Intelli Gold
Process routes
indication
Continue investigation
. Do research
. Gather more data
Flowsheet
alternatives
Hypertext with
literature review and case studies
DECISION
MAKING
Abandon or
hold
Cost and
revenues
calculation
Mining method
user input
INTELLIGOLD aims to be an expert system for gold project design.
It works in this way: the user inputs ore and deposit information in the system,
the application of decision rules and fuzzy sets leads to possible process routes indication, which are presented in a hypertext using flowsheet alternatives together with information on these alternatives;
after that, costs, revenues, return and risk are evaluated for each alternative and presented to the user, that has to make the decision of either implement, abandon or study it more;
if the decision is to study more, new information can be now added to the system, and refined answers are produced;
this cycle continues until sufficient basis for the “go-no go” decision exists.
Implement
project
Ranking of alternatives
Alternative NPV/IRR Risk factors

36. Knowledge building Process routes Process options Behavior
combined
W2= f ( sampling,
demonstration testwork)
W1=1
inferred
measured
knowledge
Process options
combined
W1=1
W2= f ( sampling,
preliminary testwork)
inferred
measured
knowledge
Contradictions
management and
feedback
Behavior
This slide illustrates how the system "builds knowledge", to accomplish the task intended.
The system is designed like a step ladder. Its inferencing goes, step by step, verificating, in each step, the compatibility about the inferred and measured variables.
Or, in other words, knowledge and data.
Each combined variable is therefore the weighted combination of measured and inferred sub-variables, that, in theory, have to be compatible, or have an approximate meaning. Usually, the weight of measured attributes are higher than the inferred.
The system goes therefore inferencing and combining values until the suggestion and automatic drawing of feasible process routes are possible.
Sometimes, when the system finds an incompatibility, like if mineralogy suggests gold is cyanidable and testwork finds it is not. The system then feedbacks the information, asking the user to verify the data in each level. If still the contradiction persists, it goes to the lower level, in a verification of causes and consequences. Ultimately, if the ore behavior is particular, the user has the ability of changing weights, telling the system which information it has to trust more.
To have one idea of the size of the step ladder, it can be said that the process design module itself has around 600 rules and manipulates among 1300 variables and sub-variables.
combined
W1=1
W2= f ( sampling,
equipment, precision)
1300 variables
600 rules
inferred
measured
knowledge
Mineralogy
Geology

37. models and adjustment parameters
Economic evaluation
Flowsheets
and process
information
Project site
Mining method
Economic
parameters
Environmental
issues
Unit operation
models
Generic models
Infrastructure
models and adjustment parameters
Costs
Revenues
Cash flow
Economic indicators
Key operations
Key variables
Risk factors
After the metallurgical evaluation is finished, the next step is the economic evaluation.
. it uses generated flowsheets as input
. estimate costs for each unit operation, using
actual costs and adjustment factors
and applying inflation indexes
. calculate revenues based on gold recoveries and prices,
and finally,
. calculates the economic parameters
. of course, the system assigns a degree of certainty to the economic parameters and indicates the main risk factors

38. A gold mine in the Amazon

39. A gold mine in the Amazon
Igarapé Bahia Mine
Largest Brazilian gold mine
10-12 tons of gold/year
Very low cash cost
Located in the Carajas Reserve
Forest reserve area owned by a mining company - CVRD
Only 0.4% of area occupied by mine activities
Preservation of area and native cultures

40. Aerial view of mine site
Mill
Infrastructure facilities

41. Mine

42. Processing installations

43. Grinding mills

44. Gold smelting furnace

45. Waste disposal

46. Summary Gold is an important resource to economy
Gold ores are found in different types
Processing is highly dependent on mineralogical characteristics
Process design is a challenging activity
Intelligent systems can be used to design and operate gold plants
Advanced use of information technology in mineral industry is a key issue in the 21st century