Sunteți pe pagina 1din 44

Module 3:

Exploration Management and Targeting

Session 4:
Mineral Exploration Tactics

Jon Hronsky
21-25 March 2011

KIGAM International School for Geoscience Resources March 2011 1

Introduction

Traditionally most discussion about effective


exploration program design focuses on
application of particular search technologies

These are context sensitive

Very dependent on target type and local


environment

Objective of this training module is instead to


provide a generic set of principles that are
applicable to any context

KIGAM International School for Geoscience Resources March 2011 2

1
Prospecting vs.
Systematic Exploration

Important starting point is recognizing


fundamental difference between prospecting-
style exploration and systematic exploration

Historically, mineral exploration activities were


dominated by the prospecting approach

The systematic approach however represents a


more efficient path to discovery and is
absolutely essential under cover

KIGAM International School for Geoscience Resources March 2011 3

Prospecting-Style Exploration

Prospecting-style exploration is basically a search


f surface
for f (or
( near-surface)
f ) iindications
di ti off
mineralization
This search continues until an indication of
mineralization is found that is sufficiently interesting
to focus detailed follow-up work
The sampling methodologies employed include float-
sampling, rock
rock-chip
chip sampling and follow-up
follow up
trenching.
Prospecting-style exploration by its very nature has a
very high sampling bias when we are prospecting
we want to be biased to the relatively rare occurrence
of mineralization
KIGAM International School for Geoscience Resources March 2011 4

2
Source: Oriental Minerals Website 5

Prospecting-Style Exploration

P
Prospecting-style
ti t l exploration
l ti can beb very effective
ff ti ini
the initial stages of exploration of a particular area;
most critically it can provide the first positive
identification of potentially-economic mineralization

However, the high sampling-bias of prospecting-style


exploration makes it ineffective for the systematic
evaluation of large areas that we know to be
mineralized, and for prioritization between
mineralized target areas

KIGAM International School for Geoscience Resources March 2011 6

3
Systematic Exploration

Systematic exploration methodologies have the goal of not


j t fi
just finding
di mineralization
i li ti but
b t finding
fi di large,
l potentially-
t ti ll
economic accumulations of mineralization
Systematic project-scale exploration requires the
application of sampling methodologies (both geochemical
and geophysical) that are:
Collected on a systematic spacing
Integrate the response (e.g. Au content) from a broad
rock volume
Targets generated by a systematic exploration process are
much more likely to be associated with economic
mineralization
Examples of systematic exploration methodologies include
stream sediment sampling, grid geochemical sampling and
airborne geophysical surveys
KIGAM International School for Geoscience Resources March 2011 7

The Role of Geological Mapping in


Systematic Exploration

It is common in many modern exploration programs not to carry out


detailed geological mapping of the project area

In many areas this decision is justified by the presence of insufficient


outcrop (although even small exposures can help provide context)

However, in many other cases this decision is instead justified on the


basis that the results of the systematic sampling alone should be
sufficient to determine if follow up is required.

This is only true if the results of the program are completely


unambiguous (either negative or positive)

In many cases, geological mapping (including supporting rock chip and


trench sampling) can have a very important role to play in providing a
context for the interpretation of the results from the systematic
exploration methodologies

KIGAM International School for Geoscience Resources March 2011 8

4
Jigsaw Puzzle or Assembly Line?

KIGAM International School for Geoscience Resources March 2011 9

Good vs. Successful Exploration

It is the nature of exploration that the majority of


activities fail to yield a positive outcome, particularly at
the more advanced stages

It is important to distinguish between successful


exploration (i.e. that which produces an economically
successful result) from good exploration (i.e. that
which is an effective investment of exploration
expenditure)

The definition of good exploration is an exploration


activity that significantly changes the residual
prospectivity of the project; either up or down

KIGAM
SeniorInternational
ExplorationSchoolManagement
for Geoscience Resources
Course March
- 2011 10

5
The Exploration Stage Concept

The practice of exploration is a series of progressive


area reduction steps

In most cases, these steps are quite consistent in


general terms between different projects as they
reflect the basic set of technologies available to our
industry

The term Exploration Stage is used to describe


each of these general steps

KIGAM International School for Geoscience Resources March 2011 11

Example of an Exploration Stage Framework


(Source: WMC Resources)
12

6
The Basic Structure of
Exploration Activity

KIGAM International School for Geoscience Resources March 2011 13

Exploration Projects are like Sharks

They either keep moving forward or they die!


KIGAM International School for Geoscience Resources March 2011 14

7
Decision Point Planning

KIGAM International School for Geoscience Resources March 2011 15

Introduction

Decision Point Planning is a planning process where


resources are only
y committed to a pproject
j up
p to the point
p
where the project reaches a major go-no go decision point
Sometimes also referred to as budgeting to milestones
Best practice for any situation where:
the probability of reaching the ultimate desired
outcome of the project is low, and
a number of key intermediate milestones must be
reached on the path to the ultimate outcome
Particularly relevant to mineral exploration but also applied
generically in industrial R&D
Remember the greatest option value with a highly
uncertain early stage project is the option to stop!

KIGAM International School for Geoscience Resources March 2011 16

8
Advantages of Decision Point
Planning

Clarifies short-term goals by focusing planning


on the projects next key milestones
Drives a discipline of focusing on the most
efficient pathway to discovery
Optimizes allocation of exploration budget

KIGAM
SeniorInternational
ExplorationSchoolManagement
for Geoscience Resources
Course March
- 2011 17

Budgeting to Milestones

The probability of failure is high for any given


exploration project
Funding a project beyond the next milestone
unnecessarily constrains funding a larger portfolio of
projects
Funding a project from one milestone to the next allows
more efficient use of corporate funds because it permits
more projects to be funded at any one time
Not all projects will advance to the next milestone so
funds will be available for those projects that do
advance
Without this, contingency funding is always included,
which will either not be spent, or be spent in a sub-
optimal fashion
KIGAM International School for Geoscience Resources March 2011 18

9
Budgeting to Milestones

The greatest option value at the start of an exploration


project is the option not to proceed if discouraging
results are obtained
For this reason, it is critical to employ decision point
planning and only commit resources to a decision point

A decision point may be the completion of an


exploration stage but it may also be some intermediate
technical step

This process is also referred to as stage gating

KIGAM International School for Geoscience Resources March 2011 19

Implementing Decision-Point
Planning

Requires development of a mental model for how the


project will be explored to discovery
Technologies employed; their cost and time-frame
Sequencing of activities
Expected outcomes of each major phase of work
Two end-member approaches:
Replication of an exploration program applied
elsewhere on the assumption that it is optimum for
the project
Consideration of range of possible pathways to
discovery and conscious selection of optimum path
The latter process can be significantly assisted by the
use of Decision Trees

KIGAM International School for Geoscience Resources March 2011 20

10
Decision Trees

A decision tree is a decision support tool that


graphically depicts a model of decisions and
their possible consequences, including chance
event outcomes, resource costs, and value of
outcomes

Used to identify the strategy most likely (or the


cheapest way) to reach a goal

They can be also used to estimate the expected


value of an early-stage exploration project

KIGAM International School for Geoscience Resources March 2011 21

Decision Tree:
A Petroleum Example

Oil
.85
+$35M
Drill
-$0.5M No oil
Test favourable .15
.60 Dont drill

Oil
Seismic test .10
Test unfavourable +$35M
-$1M .40 Drill
-$0.5M No oil
.90
Dont drill

Dont run seismic


.35Oil
Decision point
+$35M
Drill
Chance event
-$0.5M No oil
$1M Cash flow .65
Dont
KIGAM International School for Geoscience Resources March 2011drill 22

11
Benefits of Decision Trees

Help
H l clarify
l if decisions
d i i
Produce better quality decisions by
encouraging us to look at how the parts
relate to the whole, rather than only looking
at the whole and attempting to make a
decision at that level
As visual representations of decisions, they
provide a good overview as the decisions
are captured on one page

KIGAM International School for Geoscience Resources March 2011 23

Benefits of Decision Trees (cont.)

Maximize consistency in decision making

The act of constructing decision trees forces us


to think ahead

If some parts of the decision-making process


change, we can simply re-work that part of the
existing tree and develop a new strategy from
that point. This also applies if we want to
experiment with different probabilities on the
branches generated by a chance node

KIGAM International School for Geoscience Resources March 2011 24

12
Expected Monetary Value
(EMV)

Ps
$ NPV
EMV ?
Pf $ Costs

EMV = {Ps * (Reward NPV-cost)} - {Pf * Failure


Costs}
Ps = probability of discovery
NPV = net present value of discovery
Pf = probability of failure = (1-Ps)
Failure costs include present exploration +/- divestment costs

KIGAM International School for Geoscience Resources March 2011 25

Drawing a Decision Tree (1)

A Decision Tree consists of 3 types of nodes:

1. Decision nodes - commonly represented by squares


1
2. Chance nodes - represented by circles
3. End nodes - represented by triangles. These include the
ultimate objective Pay-Off Node(s)

Drawn from left to right, a decision tree has only burst


nodes (splitting paths) but no sink nodes (converging
paths).

Source: Wikipedia

KIGAM International School for Geoscience Resources March 2011 26

13
Drawing a Decision Tree (2)

Step 1
Gather information that is relevant to the
decisions to be made
Step 2
Identify the decisions that need to be made,
and the chance events that may occur
Step 3
Build the information from Steps 1 & 2 into a
decision tree. The structure of the tree will
reflect the order in which decisions need to
be made and chance events occur

KIGAM International School for Geoscience Resources March 2011 27

Drawing a Decision Tree (3)

Decision nodes:
Generate branches, each of which reflects a possible
decision we could make

Chance nodes:
Generate branches, each of which relates to an
outcome
Each branch has a probability which we must estimate
These probabilities must add to 1
Calculate an EMV at each chance node; probability
weighted average of all outcomes

Terminal nodes:
End points of decision tree
Each has a monetary value (positive, zero or negative)

KIGAM International School for Geoscience Resources March 2011 28

14
Decision Tree:
A Petroleum Example

Oil
.85
+$35M
Drill
-$0.5M No oil
Test favourable .15
.60 Dont drill

Oil
Seismic test .10
Test unfavourable +$35M
-$1M .40 Drill
-$0.5M No oil
.90
Dont drill

Dont run seismic


.35Oil
Decision point
+$35M
Drill
Chance event
-$0.5M No oil
$1M Cash flow .65
KIGAM International School for Geoscience Resources March Dont
2011 drill 29

Mineral Exploration Example

Kreuzer et al (2008)
KIGAM International School for Geoscience Resources March 2011 30

15
Solving a Decision Tree

Assume branch of decision tree that has highest EMV


is optimal
p

Solve a decision tree by working backwards

Start at the terminal nodes and work from right to left.

At each decision node (only) prune the least attractive


branch

The final result is the remaining continuous branch


when the pruning has been completed back to the
initial point of the tree

This is our optimum decision-point plan

KIGAM International School for Geoscience Resources March 2011 31

Solving a Decision Tree:


A Petroleum Example
Oil
EMV= .85
(0.85 x 34.5) (0.15 x 0.5)
= $29.25M +$35M
Drill
-$0.5M No oil
Test favourable .15
.60 Dont drill

Oil
Seismic test .10
Test unfavourable +$35M
-$1M .40 Drill
EMV=
(0.6 x 28.25) + (0.4 x 2.0) -$0.5M No oil
= $16.15M
$16 15M .90
EMV=
Dont
(0.1 x 34.5) (0.9 drill
x 0.5)
= $3.0M
Dont run seismic
Oil
.35
Decision point
+$35M
Drill
Chance event EMV=
(0.35 x 34.5) (0.65 x 0.5) -$0.5M No oil
$1M Cash flow = $11.75M .65
KIGAM International School for Geoscience Resources March Dont
2011 drill 32

16
The Exploration Review Process

KIGAM International School for Geoscience Resources March 2011 33

Why do we need Exploration Reviews?

A systematic
systematic, regular review process is vital in
any exploration group

It ensures that the best expertise in the group is


used to make key investment decisions

It ensures that any decisions to proceed with


particular projects are made in the context of
the overall portfolio

Relate learning to formal review process

KIGAM International School for Geoscience Resources March 2011 34

17
Requirements for Successful
Reviews

Review process delivers better results if success criteria


and e
expected
pected timing of milestones of projects are clearl
clearly
defined at inception
Allows advanced planning of time, place and participants of
review
Not reactive, ad hoc process
Need team, peer, management, and sometimes, corporate
participation
Involvement
In ol ement of those without
itho t emotional commitment to
project essential!
May require different type of review for the different context
of each group (e.g. commercial, logistical) but only if
technical objectives are met
Beware of mission creep!
KIGAM International School for Geoscience Resources March 2011 35

Types of Reviews

Peer reviews
re ie s vs.
s Management reviews
re ie s
Actually two separate and non-competing reviews
Each has a distinct focus and intent
What is a Peer review?
Review of project results with the project team and
other g
geoscientists and internal stakeholders
Needs to be led and managed
Careful and thorough documentation of results
Needs to be timely

KIGAM International School for Geoscience Resources March 2011 36

18
Types of Reviews

What is a Management review?


Not a technical review
Determine if results are consistent with corporate
financial objectives
Determine if the project has any newly identified
commercial
i l risks
i k that
th t impact
i t its
it value
l
Consider if the project has more value to another
company

KIGAM International School for Geoscience Resources March 2011 37

Milestones

A milestone occurs when the outcome of a chance event (i.e.


phase of exploration work) is known. It is a point in time when
a decision must be made to continue with or exit the project

Oil
.85
Drill +$35M
No oil
MILESTONE

MILESTONE

-$0.5M
Test favourable .15
.60
Dont drill
Oil
Seismic test .10
Test unfavourable Drill +$35M
-$1M .40
40
-$0.5M No oil
.90
Dont drill
Dont run seismic Oil
Decision point .35
Drill +$35M
Chance event -$0.5M No oil
.65
Dont drill

KIGAM International School for Geoscience Resources March 2011 38

19
Milestone Review
Decision Tree

KIGAM International School for Geoscience Resources March 2011 39

Milestone Reviews

Successful implementation of decision point planning


d
depends
d critically
iti ll on a supporting
ti Mil
Milestone
t review
i process
A critical review of the project is essential after results of an
exploration test are received
Requires faith on the part of the project manager that funding
will be provided if the milestone is passed
Requires diligence on the part of management that funds will
b available
be il bl for
f projects
j t that
th t pass their
th i milestones
il t (i.e.
(i do d nott
fund too many projects at once)
In some cases, may be obvious whether milestones have
been achieved or not but in other cases results may be more
ambiguous

KIGAM International School for Geoscience Resources March 2011 40

20
Generic Framework for
Effective Program Design

KIGAM International School for Geoscience Resources March 2011 41

Introduction

There is no single recipe in terms of which


exploration technologies should be used

Instead, the approach chosen is to define a small


number of generic critical parameters that influence
optimum program design

The particular quantitative values of these


parameters for any project will be context sensitive:
Target
g ore type
yp
Local Geology
Logistical considerations

Generic Decision Matrices have been developed


based on qualitative values of these critical
parameters
KIGAM International School for Geoscience Resources March 2011 42

21
Systematic Sampling vs.
Contextual Data

The data collected and interpreted in any exploration


program design
d i will ill fall
f ll into
i t one off two
t categories:
t i
Systematic sampling data
Geochemistry, geophysics, drilling
Contextual data
Mapping, rock-chip sampling, geochronology

Generally
y the majority
j y of the p
program
g expenditure
p will
be incurred in systematic sampling

The key role of contextual data is to enhance the


interpretation of the significance of results derived from
systematic sampling

KIGAM International School for Geoscience Resources March 2011 43

The First Decision Point:


Stratigraphic Drilling?
In some specific situations it may be very cost-effective to
g p
drill a stratigraphic hole at the beginning
g g of an exploration
p
program

The main value created by stratigraphic drilling is the


opportunity-cost saving by evaluating a key target
parameter early (i.e. ability to stop project early)

Stratigraphic holes seek to establish either:


The presence of a favourable lithological unit
The absence of prohibitive depths of post-mineral cover

They are justified when there is significant doubt about the


presence of a favourable unit, or that cover depths are
within explorable limits
KIGAM International School for Geoscience Resources March 2011 44

22
Conditions for Stratigraphic Drilling

If testing for the presence of a favourable unit, all of


th
these th
three conditions
diti mustt be
b satisfied:
ti fi d

The validity of the target depends on a particular


critical lithological unit being present (or at the
least, the quality of the target is strongly
dependent on its presence)

The base-rate probability of this critical unit


occurring in the target area is low

The presence or absence of the critical unit can


be unambiguously verified from a single (or very
few) drill hole(s)

KIGAM International School for Geoscience Resources March 2011 45

Critical Generic Parameters

DEPENDENT ON
1. Resolution of primary targeting CONTEXTUAL
FACTORS:
2. Drill test area of Influence (AOI)
Target ore type
Local geology
3. Cost of pre-drill survey(s)
Logistical constraints

4. False positive rate of pre-drill survey(s)


y( ) Cost of commercial
risk management

(Note that pre-drill survey includes reconnaissance geochemical pattern drilling)

KIGAM International School for Geoscience Resources March 2011 46

23
Resolution of Primary Targeting

The starting point for any exploration project will be a box


on a map
map that has been defined through the application of
geological concepts to the pre-existing geoscientific data
base

The size of this box (i.e. the resolution of the primary


targeting) is an important parameter

In some cases, the resolution may be high; e.g. if a particular


bulls-eye magnetic anomaly is being targeted

However, it is more likely that the resolution will be relatively


low; e.g. targeting based on regional tectonic architecture
typically delivers a resolution of the order of 103 km2

KIGAM International School for Geoscience Resources March 2011 47

Resolution of Primary Targeting (cont.)

Targeting resolution is usually a function of:

the resolution and quality of the available data

how well the target ore-environment is known


to manifest itself in these data

KIGAM International School for Geoscience Resources March 2011 48

24
Drill Test Area of Influence (AOI)

Any drill test of a covered target that fails to intersect


mineralization,, or the recognizable
g signature
g of a major
j
deposit has, in most cases, effectively sterilized an area
of influence around the drill hole

The drill test AOI will depend on:


the mineralization footprint of the target ore-type
the area of its distinctive halo (alteration, metal
anomalism, unusual rock types etc)
the ability to employ off-hole geophysical
techniques

It can range from very low in the case of diamond


exploration to quite large in the case of porphyry
copper exploration

KIGAM International School for Geoscience Resources March 2011 49

Cost of Pre-Drill Surveys

This is a fairly obvious input parameter

It will depend strongly on the local geography of


the project area and attendant logistical issues

In some cases (e.g. Airborne EM, Airborne Gravity)


these costs can be very large (>US$ 1 M) and as
significant as drilling costs

KIGAM International School for Geoscience Resources March 2011 50

25
False Positive Rate of Pre-Drill
Surveys

Because we are dealingg with a low base-rate


environment, the false-positive rate is the
overwhelmingly most important parameter which
governs the effectiveness of a particular survey
(assuming the false-negative rate meets an
acceptable threshold)

Two different survey techniques applied to the same


area can result in a dramatically reduced false-
positive rate for coincident anomalies:
But must ensure that the two techniques are sensing
independent properties (e.g. geochemical and geophysical
responses)

KIGAM International School for Geoscience Resources March 2011 51

Bringing it all Together

Q
Qualitative
lit ti values
l off the
th four
f key
k generic
i
parameters can be integrated to develop decision
matrices to help exploration planning

In this decision model, immediate drill testing is


favoured, if it is a viable option, as it provides the
quickest path to discovery

Note that certain combinations of values for the


key parameters will mean that the project is not
explorable with currently available technology

KIGAM International School for Geoscience Resources March 2011 52

26
Resolution of Primary Targeting
LOW HIGH

HIGH (ee.g. PCDs)


Immediate
drill testing
Immediate
may be viable
drill testing
but evaluate
Drill Test AOI cost-benefit
Diamonds))
of pre-drill survey

Immediate
Pre-drill survey drill testing
may be viable
LOW (e.g. D
D

essential
but evaluate
cost-benefit
of pre-drill survey

Decision Matrix One


KIGAM International School for Geoscience Resources March 2011 53

Immediate Drill Testing

If the qualitative analysis in Decision Matrix 1


suggests that immediate drill testing may be a
viable exploration strategy, it will be necessary
to plan the possible drilling program to assess
its cost compared to other options

g
There are long-established methodologies
g (to
( be
discussed later) for planning the optimum drill
spacing given a knowledge of the drill test AOI

KIGAM International School for Geoscience Resources March 2011 54

27
Cost of Pre Drill Survey
LOW HIGH

Pre-drill survey
Pre-
Pre-drill survey
Pre-

GH
se Positive Ratte
ey
probably still

HIG
Pre Drill Surve
worth doing not viable
(unexplorable
unexplorable?)?)

Pre-drill survey
Pre-
Pre--drill survey
Pre
LOW
W
of P

probably still
Fals

very attractive worth doing

Decision Matrix Two


KIGAM International School for Geoscience Resources March 2011 55

Unexplorable Projects

It is possible that this analysis will indicate that some


projects are effectively impossible to explore in an
economic way with current technology

These will be projects that have:


Low resolution of the primary targeting
Low AOI of drill holes
Available pre
pre-drill
drill survey technology has a high
false-positive rate and a high cost

All of the above can be improved via scientific and


technological innovation

KIGAM International School for Geoscience Resources March 2011 56

28
Exploration Technology:
General Comments
Successful periods of exploration are generally
characterized by the opening up of a new search space
and the development and application of an integrated
technology package
These technology packages are a combination of
targeting concepts and technologies that are organized
in an effective way across scales and are well-suited to
the particular exploration situation
Examples include:
NiS exploration in WA 1966-72
Au exploration in Australia in the 1980s-1990s
Developing these integrated packages is far more
important than any single targeting concept or
technology (there are no silver bullets!)
KIGAM International School for Geoscience Resources March 2011 57

The 1980s-1990s
Gold Exploration Technology Package

KIGAM International School for Geoscience Resources March 2011 58

29
Summary

Understanding at least the qualitative values for four key


input parameters can provide a short-cut qualitative
assessment of the most attractive way to explore a project
(or in some cases help make the decision that the project
area cannot be explored)

The value of this structured approach to planning the


program is that all options are considered in a rigorous
framework, including going to immediate drill-testing

The application of this framework will also encourage us


to explicitly think about these key variables in the context
of our particular exploration project

KIGAM International School for Geoscience Resources March 2011 59

Optimizing Sampling Density

KIGAM International School for Geoscience Resources March 2011 60

30
Introduction
The most significant cost-input parameter at any stage of
exploration is the selected sampling density
Sampling is used here in a generic sense to include all of surface
geochemical sampling, geophysical surveys and drilling
In any systematic sampling methodology, the number of samples,
and therefore cost, increases exponentially as sampling density is
increased
Despite this, it is uncommon for sampling density to be rigorously
optimized in program design
In the absence of such optimization, the default seems to be a
tendency to over-kill, resulting in greater project expenditure
than is really required to test for the desired size of target

KIGAM International School for Geoscience Resources March 2011 61

Cultural Issues

Surprisingly, despite the obvious cost implications, it is


uncommon forf sampling
li density
d i to be
b rigorously
i l evaluated
l d
during exploration program design
The reason for this may be cultural; the cultural idea of the
strongly individualistic lone prospector has been very
strong in Western mineral exploration
In contrast, in the Soviet system (arguably as successful in
its heyday) there was a strong culture of highly-planned,
highly planned,
systematic deployment of resources in all areas of life
Not surprisingly, the first work on systematic evaluation of
sampling densities in mineral exploration came from the
Soviet system

KIGAM International School for Geoscience Resources March 2011 62

31
The key reference in the western literature is:
Igor D. Savinskii (1965): Probability Tables for Locating Elliptical
Underground Masses with a Rectangular Grid, explanatory text translated
from Russian, Consultants Bureau Enterprises Inc., New York

63

Key Parameters in Sampling


Density Optimization

Quantitative sampling density optimization requires the


adoption of values for the following key parameters:
What probability of detection is acceptable?

What is the expected response area (i.e. the 2D area of


anomalous response distinct from background noise)?

What is the expected geometry (L, W and range of angles


between strike and survey orientations) of the response area?

What degree of overlap between the target response area and


the survey line is considered a positive result? (e.g. in the case
of a gravity survey, is one anomalous station enough or would
two be required for confidence?)

KIGAM International School for Geoscience Resources March 2011 64

32
(Dentith & Galybin, 2008)
65

(Dentith & Galybin, 2008)


66

33
(Dentith & Galybin, 2008)
67

Acceptable Probability of Detection

Standard industry practice is to ensure that the spacing


between lines is less than the expected strike-length of
the minimum acceptable-sized target

Therefore a negative result is considered to have


sterilized the relevant area (although this ignores the
possibility of an unexpected orientation)

Quantitative sampling optimization shows that often a


relatively small reduction in probability of detection
(say from 1.0 to 0.7) can lead to a large cost saving

KIGAM International School for Geoscience Resources March 2011 68

34
Response Area

This is the expected 2D response area of the target in the


relevant sampling data set
Care must be taken in defining this:
It must be consistent with the size of target being
sought
The relationship between threshold values and
response area must be clearly understood
These are commonly confused in application; if we go to
lower threshold values, the size of the response area
also has to increase to be meaningful

KIGAM International School for Geoscience Resources March 2011 69

Threshold Anomaly Areas:


Porphyry Cu-Au
Log Sulphide area km2
0.16 0.4 1.0 2.5 6.3 15.8 39.8
8
nnes Cu metal

7
5mt Cu

6
Log Ton

4
Source: Donald Singer
KIGAM International School for Geoscience Resources March 2011 70

35
(Large et al, 2001)

Subtle Geochemical Haloes associated with


some Zn deposits
71

Recognizing Which Anomalies Are


Valid Drill Targets

KIGAM International School for Geoscience Resources March 2011 72

36
False Positives

In mineral exploration what we look for


(
(economici mines)
i ) are only
l rarely
l (perhaps
( h 1% off
the time) present in the target areas we test

We have a low base rate

In such a situation, the most important factor


that defines the significance of a particular
exploration technology (e.g. remote sensing) is
the FALSE POSITIVE RATE of that technology

Importantly, most exploration technologies we


use have high False Positive Rates

KIGAM International School for Geoscience Resources March 2011 73

74

37
An Example from the
War on Terror

From Savage and Wainer (2008), Chance Magazine


Let us assume:
3,000 terrorists hiding in the US population of
about 300 Million
The FBI have a magic bullet technology:
unlimited wiretapping linked to advanced voice
analysis software that can detect terrorists within
a few
f words
d off speech
h on the
th phone
h
This technology is 99% accurate: it only alerts the
FBI to non-terrorists 1% of the time
What is the probability that someone alerted to the
FBI is actually a terrorist?
KIGAM International School for Geoscience Resources March 2011 75

Savage & Wainer (2008)

Only 1 in a 1000!
76

38
Consider an Exploration Example

We have a number of conceptual targetssay magnetic anomalies


that might be diamond bearing kimberlite pipes
We know that the probability of these anomalies being a diamond
deposit is realistically about 1 in 1000 (0.001)
Now we apply an innovative new geophysical technology say
airborne gravity
We know that we get a gravity anomaly 90% of the time when a
diamond deposit is present
We also know that we get a gravity anomaly associated with a
magnetic anomaly that is not a diamond deposit 10% of the time
What is the probability that when we drill test a coincident gravity
and magnetic anomaly we find a diamond deposit?

KIGAM International School for Geoscience Resources March 2011 77

The answer is 0.89%!

P anomaly+deposit

P anomaly+ no deposit

0.89% = 0.09% / (0.09% + 9.99%)

Modified from a False Positive Calculator down-loadable from www.flawofaverages.com


KIGAM International School for Geoscience Resources March 2011 78

39
Source: Freewest website

Real Example of High False Positive Rate


79

Setting Threshold Levels


for Anomaly Significance

In any exploration survey / drill program, we need to


determine what results, short of obvious
mineralization, constitute a significant anomaly

When setting threshold levels of significance for an


exploration technology, we need to understand the
p
impact of dropping
pp g threshold levels on the false-
positive rate

Potentially small reductions in threshold might


substantially increase the false positive rate

KIGAM International School for Geoscience Resources March 2011 80

40
High Threshold

Orebody Present

True
Probability Positives
Distribution

Orebody
y Absent

Measured Value False


(eg Geochemical Sample Value) Positives

False Positive Ratio = 10%

Low Threshold

Orebody Present

True
Probability Positives
Distribution False
Positives

Orebody Absent

Measured Value
(eg Geochemical Sample Value) False Positive Ratio = 50%

81

Business Implications of the


False Positive Issue

Need to understand the large value created by


reducing the false positive rate in our exploration
programs and focus effort on doing this

Need to understand that relying on a single


exploration survey technique to define drill targets
is high-risk

Need to build this concept into the design of all new


exploration technologies up-front:
More effort on specificity rather than just a focus
on sensitivity
KIGAM International School for Geoscience Resources March 2011 82

41
How do we manage the
False Positive issue?

Prioritise targets with coincident anomalies from


multiple independent sampling strategies
Coincident surveys (e.g. geophysical and geochemical)
Assessment of all first-stage anomalies with a second
independent technology

Prioritise targets on the basis of their context within


the evolving geological picture for the project
Elimination of the obvious False Positives: Often
many false positives are readily explained
geologically remove these from consideration
Means need to study geological context of false positives,
something we are usually reluctant to do

KIGAM International School for Geoscience Resources March 2011 83

Williams (2010)
84

42
The Characteristics of a High-Quality Drill Target:
The Rule of the Six Cs

Contrast - The anomaly has a good contrast to background levels. This


y more important
is usually p than absolute values
Coherence - The anomaly hangs together; it can be easily contoured.
Coherence commonly distinguishes lithological from ore-related
anomalies
Coincidence - Two independent data sets produce an anomaly at the
same location; as a rough rule of thumb, such anomalies are about an
order of magnitude more significant than those defined on the basis of
one dataset alone
Context The anomaly has a good geological context such as a major
structural intersection
unCommon - Anomalies of that level are rare within the survey data.
Because of the low base rate, if we see a lot of similar anomalies in the
data, this alone tells us they are probably not significant
Cover - Covered anomalies are much less likely to have been previously
tested

KIGAM International School for Geoscience Resources March 2011 85

Managing Success:
What Happens after the
Discovery Hole?

KIGAM International School for Geoscience Resources March 2011 86

43
Managing Success

Get good quality, representative diamond drill sections early to


maximize geological understanding
Scope the entire mineralized system after the initial discovery
do not get sucked in too close too early (i.e. to the first
mineralized zone discovered)
Get your drilling spacing right at this scoping stage
implies
p need for knowledge
g of footprints
p and their size
Know when to quit and when to persist
clearly define threshold success criteria
walk away if these are not achieved but persist if they are

KIGAM International School for Geoscience Resources March 2011 87

44

S-ar putea să vă placă și