Cut-oƨ Grades

and Optimising the

Strategic Mine Plan
Spectrum 20

Brian Hall
 Cut-off Grades
and Optimising the
Strategic Mine Plan
by Brian Hall
The Australasian Institute of Mining and Metallurgy
Spectrum Series 20 | ISBN 978 1 925100 22 8

COPYRIGHT DISCLAIMER

© The Australasian Institute of Mining and Metallurgy 2014

No part of this publication may be reproduced, stored in a retrieval system or

transmitted in any form by any means without the written consent of the publisher.

The AusIMM is not responsible as a body for the facts

and opinions advanced in any of its publications.

Front cover image:

A risk officer looks out over the Mogalakwena Platinum Mine, South Africa.
Image supplied and reproduced by the kind permission of Anglo American.

Published by:
THE AUSTRALASIAN INSTITUTE OF MINING AND METALLURGY
Ground Floor, 204 Lygon Street, Carlton Victoria 3053 Australia
Contents
CHAPTER 1 Introduction
Introductory comments 1
The evolution of cut-off theory – increasing numbers of dimensions 2
Conducting cut-off evaluations and strategy optimisations 5
General terminology and units used 6

PART 1
Developing an Understanding of Cut-Off Theory
– From Simple Break-Evens to Full Strategy
Optimisation

CHAPTER 2 Establishing Underlying Principles

Introductory comments 13
What is cut-off grade? 13
Classification of materials for cut-offs and strategy optimisation 15
Classification of costs for cut-offs and strategy optimisation 17
Applying cut-offs to distinguish ore from waste 20
Planning and operational cut-offs 21
Cut-offs and company goals – or how the junior engineer determines
the CEO’s bonus 23
Summary 23

CHAPTER 3 Break-even Analysis – Accounting for Costs and Prices

A common but limited cut-off model 25
The simple break-even formula 26
A more general definition of break-even – equal net benefits 28
Costs in ore–waste break-even calculations 33
Break-evens at various points along the production process 42
Iterative break-even convergence and divergence 43
Problems with break-even cut-offs 45
Summary 48

CHAPTER 4 Mortimer’s Definition – Accounting for Geology

Mortimer’s Definition defined 51
Calculations for Mortimer’s Definition 51
 Mortimer’s Definition illustrated 53
Mortimer’s Definition in practice 56
Summary 57

CHAPTER 5 Lane’s Methodology – Accounting for System Capacities

Demystifying Lane 59
Establishing underlying concepts 60
Applying Lane’s methodology in practice 67
Basic principles of finding the optimum cut-off 70
Limiting cut-offs 71
Balancing cut-offs 77
Completing the analysis 83
Practical processes to derive optimum cut-offs 87
The effect of price and cost changes on cut-offs 90
Complicating factors 92
Summary 97

CHAPTER 6 Mine Strategy Optimisation – Accounting for Everything

Introductory comments 101
The goal of strategy optimisation 102
The real drivers of value in a mining operation 103
Specifying the size of the mine 105
Developing the basic rationale of the strategy optimisation process 106
Strategy optimisation versus typical strategy selection 117
The strategy optimisation process 121
Optimising one project versus optimising all projects 121
Chapter summary 123

CHAPTER 7 Cut-off Derivation Methodologies and Their Place in the

Overall Mine Planning Process
Introductory comments 125
The mine planning process 126
Preproduction studies 127
A recommended planning process at operating mines 130
The importance of corporate planning policies 138
Including cut-off determination and strategy optimisation in the
planning process 139
Other issues affecting cut-off specification 142
Chapter summary 149
PART 2 Important Considerations for Strategy
Optimisation

CHAPTER 8 The Mine Strategy Optimisation Process

Introduction 155
Identifying the options and scenarios 155
Uncertain parameters and relationships 156
Geological inputs 157
Mining inputs 159
Metallurgical inputs 168
Operating cost inputs 170
Capital cost inputs 173
Social issues 175
Revenue items and economic parameters 176
Taxation 177
End-of-life issues 177
Building and using the optimisation model 178
Working with different commodities 182
Chapter summary 183

CHAPTER 9 Specifying the Grade Descriptor

An introduction to grade descriptors 189
Selecting the correct grade descriptor 190
Types of polymetallic grade descriptors 192
Generating an equivalence formula 196
Dollar values versus metal equivalents 199
Recalculating equivalent grades 203
Beyond grade or value per tonne 205
Handling value-reducing constituents 206
Costs varying with location in underground mines 207
Chapter summary 208

CHAPTER 10 Goals and Value Measures

Measuring the attainment of the corporate goals 211
Value, goals and constraints 212
Discounted cash flow measures of value 216
Accounting measures 222
Cost measures 225
 Physical and time-related measures 229
Satisfying various stakeholders 235
Chapter summary 236

CHAPTER 11 Valuation Methods

Deriving value measures for optimisation 239
Traditional valuation 240
Monte Carlo simulation 242
Orebody simulation using conditional simulation 244
Dynamic DCF – modelling future management decisions 244
Option value 246
Real options valuation 246
Some conclusions on valuation methods 251
Chapter summary 253

CHAPTER 12 Strategy Optimisation and Risk Management

Accounting for the impact of uncertain parameters 255
Identifying upside reward and downside risk 255
Rewarding management performance? 258
A case study 258
Making the strategic decisions 260
Chapter summary 260

CHAPTER 13 Optimisation Methods

Different ways of finding the optimum strategy 263
Exhaustive calculations 263
Genetic algorithms 265
Dynamic programming 266
Linear and mixed integer programming 267
Concluding comments 269
Chapter summary 270

CHAPTER 14 Answering Misconceptions and Objections

Misconceptions or valid concerns? 273
Misconceptions 273
Valid concerns 276
Chapter summary 280
CHAPTER 15 In Conclusion
A review 283
Where have we come from? 283
Where are we now? 285
Where are we going? 287
In conclusion 289

References 291

APPENDIX Case Studies

Introductory comments 293
Case study 1 293
Case study 2 294
Case study 3 296
Case study 4 298
Case study 5 299
 HOME

Chapter 1

Introduction

introductory comments
Cut-off grade is perhaps the least understood driver of value for a mining operation.
Many believe that it is determined relatively easily by knowing the mining and
extraction costs, metallurgical recovery and an estimate of the price for the metal/s under
consideration. Determining the values for these parameters may not necessarily be easy
but, having determined them, calculating the cut-off as a break-even is straightforward.
For others, the cut-off is what it is because that’s what it has always been.
Nothing could be further from the truth. Unfortunately, it would seem that many of
the key executives and decision-makers in the industry are unaware of this, as are many
of the industry analysts who commentate on the actions and decisions of company
executives. The result would appear to be that strategies and policies that maximise
returns for stakeholders and reduce risk are given the thumbs down, while those that
are guaranteed to minimise value and increase the risk of the operation are applauded.
Cut-off specification is an integral part of optimising the mining strategy. It should
be an outcome of the strategic planning and optimisation process, not an input into it.
That statement alone may raise questions in the minds of some readers. Typical practice,
particularly for an underground mine, is first to derive a cut-off that defines the size
and shape of the orebody and then to develop mining and extraction plans for the one
orebody thus defined. This commonly used model for cut-off derivation is inadequate
in many cases.
What is meant by a ‘cut-off model’? Simply, it is the mental picture we have of the
what, why and how of cut-off specification – what a cut-off is, why we have one and
how we determine it.
By way of analogy consider classical or Newtonian physics, the physics of forces
and motion described by Sir Isaac Newton in the late 17th century. It works in three-

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 1
 CHAPTER 1 | Introduction

dimensional space, and is satisfactory for most day-to-day activities. The physical
formulas relating mass, length, time, force, velocity and acceleration are useful because
they work consistently at the scales of time, distance and accuracy that most of us are
dealing with for most if not all of the time.
Early in the 20th century, Albert Einstein turned all this upside down with his special
and general theories of relativity. Suddenly we no longer existed in three-dimensional
space, but in four-dimensional space–time. Whereas we’d always thought such quantities
as mass, length and time were constant, it transpired that all of these are relative, and
the speed of light relative to any observer is what is constant. The concepts of relativity
are counter-intuitive based on our common experience, but they have been proven to be
correct. They work at extremely large scales and high velocities, while the Newtonian
model does not. Yet at the scales and velocities with which we are generally accustomed,
the Newtonian formulas are accurate approximations of the more complex theory.
At the other extreme, classical physics also breaks down. Quantum mechanics, initially
developed in the early decades of the 20th century, describes behaviour at this scale but
the concepts are perhaps even more difficult to comprehend than those of relativity.
‘Quantum weirdness’ is so far beyond our day-to-day experience that even physicists
who are experts in the field admit to having difficulty comprehending it; however, they
understand the mathematics and can draw appropriate inferences, even though it is all
beyond normal understanding.
The current ‘holy grail‘ of physical research is ‘a grand unified theory of everything’
that effectively combines relativity and quantum mechanics and explains all four of the
fundamental forces of nature in one overarching theory. Classical physics, relativity
and quantum mechanics are all seen to be approximations of the single unified theory
that work satisfactorily within different sets of limits. At the time of writing, super-
string and M-theory working in ten- or 11-dimensional space–time appears to have
the best chance of leading to this goal,1 but as in all fundamental scientific research,
nothing is assured.
How does this relate to cut-off? Just as we see an increasing number of dimensions
in our model of the cosmos and how it works, so too are there an increasing number
of dimensions in progressively more complex (but also more useful and generally
applicable) models of cut-off and strategy optimisation. The more dimensions in the
cut-off and strategy optimisation model, the more parameters that can be accounted
for and the better the potential outcome. As in physics, simpler cut-off models might
nevertheless still be useful under certain circumstances.

The Evolution of Cut-off Theory – Increasing Numbers of

Dimensions
The first part of this book deals with the principles underlying progressively more
complex models of approaches to cut-off and strategy optimisation.

1. For example, see Hawking and Mlodinow (2010). Even M-theory now seems to be not the desired single unified
theory but rather several models that work in different circumstances. These models appear to overlap and, in
the areas of overlap, give similar results, thereby generating a continuum of models that potentially cover the
whole range of physical reality.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 2
 CHAPTER 1 | Introduction

The development of these concepts is supported by a number of important principles

and definitions, which are established in Chapter 2. In particular, the discussion deals
with definitions of cut-off and the classification of costs and materials dealt with.
Many of these are counter to common industry practice and much of the confusion
surrounding cut-off derivation stems from the general misunderstanding of how the
distinctions apply.
Cut-off has always been an important concept in mining. Old texts talk of pay
limits, and pay dirt is a common term in popular culture when dealing with old-time
prospectors and miners. The term cut-off is used in a number of these texts but without
any formal definition or explanation. It appears that the writers have assumed that cut-
off is a fundamental concept that is so well understood that it needs no description. The
author is not aware of any formalised definitions but suggests that the terminology and
the way it appears to be applied implies some sort of break-even concept.
Break-even analysis is essentially what we might call a one-dimensional process.
It is based on financial parameters only. Prices, adjusted to account for metallurgical
recovery, are compared with costs. Simplistically, if the grade of mineralised material
is sufficient for the revenue obtained to pay the costs associated with extracting,
processing and selling, it is classed as ore. Break-evens are discussed in more detail in
Chapter 3.
Although widely used in the industry, break-even is a very limited cut-off model.
In particular, it does not account for the geology, especially the grade distribution of
the mineralisation, nor the capacities of the mining and processing plant. One could be
excused for asking how a cut-off model that ignores geology and production system
capacities could be useful. The answer is that the simple break-even model can be useful
so long as it is used within the constraints of the outcomes of a more comprehensive
model; however, the reality is that in practice a simple break-even model is all that is
used by many operations. As we shall see, this can lead to mine plans that are almost
guaranteed to not deliver the company’s stated goals. Indeed there is no guarantee
that an orebody delineated by a break-even cut-off – derived according to typical
industry practices – will even deliver a profit. Use of break-even cut-offs has probably
been the primary reason why mining companies have not been generating returns for
shareholders in line with increases in metal prices in recent years.2
Cut-off theory advanced somewhat with the publication of a little-known paper on
South African gold mining grade control practices in the late 1940s (Mortimer, 1950), a
section of which gives rise to what the author refers to as Mortimer’s Definition. This can
be thought of as a two-dimensional model of cut-offs. As well as the financial aspects of
break-evens, it brings in the need to consider the geology and nature of the mineralisation
as expressed by the tonnage and grade versus cut-off relationships. This ensures not
only that break-even conditions are satisfied but also that the orebody delineated will
deliver a specified profit target. Mortimer’s Definition is discussed in more detail in
Chapter 4.
Ken Lane published what has, for many years, been state-of-the-art in cut-off
optimisation, initially in a technical paper (Lane, 1964) and later in a text book

2. The author has not carried out any research to justify this comment but is recounting here complaints by
analysts at several conferences in recent years, as reported by various industry-related online news sites.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 3
 CHAPTER 1 | Introduction

(Lane, 1988, 1997). Continuing the dimensional metaphor, Lane’s methodology could
be described as a three-dimensional process. As well as the financial and geological
dimensions of the break-even and Mortimer approaches, Lane accounts for the
production system’s capacities to handle three classes of material: rock, ore and
product. The rationale for this classification of materials is fundamental for all cut-off
determinations, and is therefore described in more detail in Chapter 2.
Lane developed his theory using rigorous mathematical processes. From discussions
with mining industry professionals from a number of companies and countries, this
author perceives that the vast majority of mining engineers and geologists who attempt
to read and understand Lane give up when they get to the mathematical derivations of
Lane’s formulas. However, these formulas may be derived logically, if not completely
rigorously, without resorting to complex mathematics. This process and other important
aspects of Lane’s methodology are described in Chapter 5, which is this author’s attempt
to ‘demystify Lane’.
Lane and other workers in the field in the 1970s and 1980s introduced a number of
concepts to the development of cut-off theory, which are also described in Chapter 5.
Worth noting is the concept of a balancing cut-off, which ensures that two of the three
production system components (for rock, ore and product) are operating at their
capacities. A balancing cut-off is a function of the geology and plant capacities only and is
not at all related to costs and prices. It will often be the optimum cut-off to apply. In other
words, the optimum cut-off may actually not depend in any way on prices and costs, a
revelation for many who encounter it for the first time after believing all their working
lives that cut-off and break-even are synonymous. Also worthy of note is the concept
of opportunity cost, which has always been implicitly understood in the industry. It
is, for example, understood that one can mine down to the marginal break-even to ‘fill
the mill’ if there is a shortfall in ore supply. It is also recognised that this is only valid
so long as lower-grade material is not displacing higher-grade feed. The opportunity
cost quantifies the extent to which such displacement can occur economically. Lane’s
methodology automatically accounts for it in the cut-off optimisation process.
With the development of powerful digital computers from the 1990s on, full mine
strategy optimisation has increasingly become the goal of cutting-edge mine planners
and mining software developers.3 This may be thought of as a multidimensional analysis
taking account of, notionally, everything. The values of the strategic decision variables
can all be optimised while taking account of the uncontrollable projected economic,
financial, social and geological parameters in the future. The decision variables include
(but are not limited to) such items as cut-offs; production capacities in various parts
of the overall production process, perhaps including multiple plants; sequencing and
timing of mining from stopes, mining blocks, pits and separate mines; products and
product mixes; and stockpiling policies. All of these can vary over time or by location,
or both.
In principle, if something can be described, it can be modelled and included in analyses
and evaluations. The only limits are computing power and the evaluators’ imagination
and ingenuity in identifying important parameters, the relationships between them and

3. Complete strategy optimisation, however, is still too large a problem for current and reasonably foreseeable
computer power. Although significant advances have been made, we are still only looking at subproblems of
the whole problem, or we model the whole problem crudely or work on small, simple problems.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 4
 CHAPTER 1 | Introduction

suitably simple, accurate and efficient ways to model these within the limitations of the
evaluation methodology and, where appropriate, the software being employed.
In all the models of cut-off prior to strategy optimisation, all parameters other than
cut-off are effectively assumed to be pre-specified and the cut-off is derived for that
case. In strategy optimisation, cut-off is but one of many decisions to be made, and it
is optimised concurrently with all other decisions. Chapter 6 describes the principles
underlying a full strategy optimisation evaluation and the processes typically involved.
It can be seen that there are a number of cut-off models available to the mine planner.
Depending on the cut-off derivation methodology being used, it might at times be
necessary to make complex assumptions regarding some inputs for the methodology
to work. Or perhaps some important parameters are not adequately accounted for. This
will usually be an indication that the cut-off model used is inadequate and a higher-
dimensional model is required to deal with the real complexities of the situation.
Chapter 7 suggests when particular cut-off models may be appropriate and how
they fit into the overall mine planning process.

Conducting Cut-off Evaluations and Strategy Optimisations

The first part of this book deals with the principles underlying progressively more
complex models of approaches to cut-off and strategy optimisation. The second part
deals with specific aspects of the associated activities.
Chapter 8 describes in some detail the inputs required for a full strategy optimisation
evaluation. A number of the practicalities involved in developing and using an
evaluation and optimisation model are discussed.
An early consideration is deciding upon the appropriate grade descriptor. This
is the value given to each block of rock to describe its value relative to other
blocks – the number that tells us that ‘this bit’ is more valuable than ‘that bit’.
Actual metal grades, metal equivalents and money equivalents are commonly used;
however, the derivation of equivalents is not necessarily done as well as it could be and
other measures may be more appropriate. Chapter 9 describes how alternative grade
descriptors might be derived and how the best one is selected.
Chapter 10 discusses some of the measures of value that might be used to enable
us to say that Strategy A is better than Strategy B. If cash is indeed king, then value
measures that quantify cash generation, such as net present value (NPV) and real
options value (ROV), must be the key measures for any strategy optimisation. Cut-offs
and other components of operating strategies must be focused on maximising these.
Many companies have other value measures that interest them, and are prepared to
accept a lower NPV for improvements in other measures. Chapter 10 also discusses
some typical alternative measures. Unfortunately, it would seem that in many cases
these other measures are treated as surrogates for cash generation; however, many of
these are not correlated with cash generation, and in some cases are actually negatively
correlated. The popular ‘ounces in reserve’ measure is one such parameter, which
receives a lot of attention from senior managers and industry analysts, but can be
simply demonstrated to be negatively correlated with cash generation – in other words,
the more ounces the less cash, and vice versa. Chapter 10, as well as discussing various
measures also describes how they each correlate with cash generation.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 5
 CHAPTER 1 | Introduction

Chapter 11 describes alternative methodologies to derive some of these value

measures. Chapter 12 then considers how optimisation techniques can be applied to
deliver more complex corporate goals than simply, for example, maximising NPV or
some similar single measure of value. In particular, the trade-off between maximising
the reward from making correct assumptions about future conditions, such as price, and
minimising the downside risk of making wrong predictions is discussed.
Chapter 13 describes briefly some of the techniques available to optimise strategies.
This is not intended to replace a detailed text on operations research nor detail particular
algorithms, but rather give the flavour of several commonly used techniques, touching
on their advantages and disadvantages. All of these are useful and none can necessarily
be relied upon to give the best answer. The project evaluator might well be advised
to use a combination of techniques to allow the strengths of one to complement the
weaknesses of another.
Chapter 14 addresses some of the misconceptions and objections that apply to various
aspects of cut-off and strategy optimisation. Chapter 15 summarises the major themes
established in the preceding chapters and poses some philosophical challenges
for the industry into the future.
To conclude, the appendix describes several case studies, indicating how many
principles and practices described have been applied to identify optimum strategies and
how the value obtained from each has been increased.

General Terminology and Units Used

The strategic mine plan

Throughout this book, unless the context implies otherwise,4 mining, the mine plan and
similar terms relate to the total extraction process of valuable material from the earth to
the point of sale of product. These mining terms are assumed to include:
•• the drill, blast, load, haul and processes typically associated with the mine to extract
ore and related waste
•• the metallurgical treatment processes performed on ore to extract valuable product/s
in the plant or the mill
•• integrated smelting and refining or similar downstream processes to add value to
an already saleable product (such as the smelting of a base metal concentrate) if the
operation or the company has such facilities
•• transport of all products to the point of sale, accounting for any physical or market
limitations on the amount of product that can be handled or sold into the market.
The reason for this usage is that optimisation of the corporate business plan for an
operation or multiple operations must take account of the complete process stream,
from having the mineralisation identified and described by geologists through to final
sale and receipt of revenue. While it is possible in practice to limit an evaluation to one
part of the process, such as the mine, and assume that all other processes are not going
to change, the reality is that suboptimisation of components of a complex process cannot

4. The primary exception will be the use of the term ‘mining’ to refer to a clearly defined part of the overall
production process when discussing Lane’s cut-off methodology in Chapter 5.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 6
 CHAPTER 1 | Introduction

be guaranteed to find the best overall plan. This can only be done if the whole process is
considered together in one integrated set of evaluations.
Much of Part 1 of this book, however, will be dealing purely with selecting a cut-off
grade, and this is largely a mine-focused activity. By implication, parameters relating to
processes outside the mine will be taken as given; nevertheless, they will enter into the
cut-off derivation in a number of ways, and the need for concurrent optimisation of
the total production process will become evident.

Cut-offs
The term cut-off may be used as both an adjective and a noun. It is defined more fully in
Chapter 2.
Cut-off policy refers to a planned sequence of cut-offs for an operation over time.
The term recognises that it is rare for the best cut-off to be the same unchanged value
over the life of the mine, even if such issues as costs and product prices were to remain
unchanged.
Grade is used generically to refer to any parameter that is used to describe the relative
value of a block of rock. It could be the metal grade in a single-metal deposit, or a metal
equivalent grade or monetary value measure in a polymetallic deposit. Therefore, there
is generally no distinction in this book between such terms as cut-off grade and cut-off
value, or indeed cut-off on its own. In any discussion where the distinction between
actual metal grades and other grade descriptors or measures of value is important,
it will be explicitly explained. In practice, of course, a cut-off should be expressed in
unambiguous units so that it is clear which grade or other measure its value is to be
compared with to distinguish ore from waste.

Ore and reserves

The terms orebody, reserve and similar are used in their colloquial sense and not
according to the strict definitions specified in the JORC (Joint Ore Reserves Committee)
Code, NI 43-101 (National Instrument), SAMREC (South African Mineral Resource
Committee) and similar codes for public reporting. The cut-off derivation and strategy
optimisation processes described in this book are internal to the company performing
the evaluation, where the use of such terminology is not required by law or other
external codes of practice. This should be conducted before a reserve is publicly
reported in accordance with the code of the country where publication occurs. These
processes will typically involve identifying a number of sets of potential reserves at
different cut-offs, one of which will ultimately form the basis for the mine plan that
supports the publicly reported reserve.
As this book discusses processes that will ultimately lead to specifying a cut-off that
will potentially lead to a publicly reported reserve, ore has a specific meaning that is quite
different from the definitions in the reserves reporting codes. It is defined more fully in
Chapter 2, along with rock and product, but for now may be thought of as the mineralised
material above a particular cut-off or that which is intended to be sent through the
mine’s ore-handling system for processing through the treatment plant for the case being
considered. It should be noted, however, that specific resource categories – Measured,
Indicated and Inferred – are commonly used in many companies’ planning procedures,
and nothing here should imply that this is wrong or be discouraged. Indeed, the opposite

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 7
 CHAPTER 1 | Introduction

is the case, as the definitions for these terms convey commonly understood information
regarding the certainty of the underlying geological information and other factors.
It will always be appropriate to comply with company policies that impose restrictions
on the use of, for example, Measured and Indicated Resources for final plans; however,
it is not appropriate to restrict investigations to these categories while developing the
mine plans. All available information should be used, regardless of its resource category
or other descriptor of certainty, while obviously taking account of the different levels
of accuracy and certainty associated with each item of data.5 This, of course, allows for
the possibility that some plans might be developed using only Measured and Indicated
Resources and others using all known mineralisation. This would help to identify
whether strategies would differ if what is currently classified as lower confidence
material was subsequently found to actually exist or not, and therefore how important
it might be to conduct exploration sooner rather than later.

Units
The International System of Units (SI) and related units are generally used throughout
this book, although some non-SI units that are commonly used in the industry are also
employed. Key units and the corresponding abbreviations used in this book are:
•• Mass – tonnes (t), typically for rock, ore and waste quantities but also for product
quantities for base metals; ounces (oz), typically for product quantities for precious
metals.
•• Length – metres (m).
•• Volume – cubic metres (m³).
•• Density – tonnes per cubic metre (t/m3). Note that specific gravity (SG) is the ratio of
the density of the material to that of water, and is therefore dimensionless; however,
in SI units, density and SG are numerically the same.
•• Grade – grams per tonne (g/t), typically for precious metals; per cent (%), typically
for base metals; parts per million (ppm), typically for trace elements, contaminants
and penalty elements. When discussing general principles, the term grade unit, or
gu is used to refer to the quantity in the numerator of the grade symbol, with the
generalised grade being reported in grade units per tonne or gu/t.
•• Recovery – per cent (%).

5. As an aside, the author strongly recommends that when resource categories are recorded in geological block
models, these should represent geological confidence only. For all mine planning purposes, it is just as important
to know how certain the characteristics of low-grade material are as it is for high-grade material; however, it is
not uncommon to find that material with a grade below the resource cut-off is given a resource category code
that identifies it as waste, and only material above the resource cut-off is identified as Measured, Indicated,
Inferred, etc. This practice effectively destroys any record of the geological confidence of low-grade material
without any corresponding gain to offset that loss of information. Whether it is resource or not can be assessed
by reference to the resource cut-off grade: a separate waste flag is redundant. Extraction of resource quantities
of various categories for reporting is therefore not compromised by having the resource category code defining
only geological confidence. It is simply a matter of applying simultaneously the two essential characteristics
that define whether material falls into one of the various categories of resource: its geological confidence and
its grade relative to the resource cut-off grade. Indeed, if the resource cut-off is reduced, the recommended
procedure simplifies matters; it is only necessary to rerun the data extraction process with the new cut-off. It
is unnecessary to repeat whatever was done to specify the geological confidence unless there is additional
geological information that changes the level of confidence.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 8
 CHAPTER 1 | Introduction

Monetary amounts are generally referred to as dollars.

When expressing grades as monetary units, the term dollar value ($ value) is used, and
the unit is typically dollars per tonne ($/t).

Value drivers, options and scenarios

The term value driver is used to refer to a parameter that has a major influence on the
value of the operation. Value drivers will be described in more detail in Chapter 6 and
later, but for now may be thought of as such items as ’production rate’ and ’metal prices’.
The two examples noted here also exemplify the different classes of value drivers:
1. Those for which the company can specify a value or setting, such as specifying a
production rate or the type of metallurgical processes to be used. These are referred
to as options.
2. Those which the company cannot control or for which it does not have adequate
knowledge. Metal prices, for instance, are typically predictions of future conditions
and are intrinsically unknowable at the time an evaluation is being performed.
Certain costs or metallurgical recoveries, on the other hand, are not selectable as an
option, but may be uncertain at the time an evaluation is being undertaken. In both
cases, alternative possible values may need to be considered in the evaluation. These
are referred to as scenario parameters.
The term case is used to refer to a complete set of specifications of all the values used
for all value drivers, both options and scenarios, in one particular evaluation.
The terms cost drivers and revenue drivers will be used for such things as ore and product
quantities, which will typically be the production outcomes of the strategic decisions
made regarding various value drivers classed as options.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 9
Part 1
Developing an
Understanding of
Cut-Off Theory –
From Simple
Break-Evens to Full
Strategy Optimisation
 HOME

Chapter 2

Establishing Underlying
Principles

introductory comments
Before proceeding to discuss cut-offs and mine strategy optimisation, it is essential
first to establish some underlying principles and vocabulary. Every specialised field
has its unique set of terms with specific meaning within that context. Often these will
be common terms with a wider, colloquial meaning for laypersons, and the difference
between the informal and formal meanings often being a source of confusion; this is
compounded by specialists alternating between meanings depending on the context.
Cut-off specification is no exception. Lane (1988, 1997) generated cut-off-specific
meanings for several common mining industry terms. From discussions with technical
staff from various countries and companies over the years, the distinction is reasonably
well understood for some terms but not for others. This chapter therefore sets out a
number of terms with special meanings for cut-off and strategy optimisation as they are
used in this book. Many of the issues discussed here will be developed in more detail in
later chapters.

What is Cut-off Grade?

Much of the misunderstanding surrounding cut-offs would appear to come from
a failure to distinguish what a cut-off actually is, and why and how it is derived. In
particular, there seems to be confusion relating the ‘what’ with the ‘how’. It is therefore
important to distinguish between cut-off definition, purpose and derivation.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 13
 CHAPTER 2 | Establishing Underlying Principles

Definition
A cut-off is simply a number that indicates the point between two alternative courses of
action. Material with a grade above the cut-off is dealt with in one way, while material
with a grade below the cut-off is dealt with in another way. In particular, a cut-off is
used to distinguish between ore and waste – material with a grade above the cut-off is
ore and below the cut-off is waste.1
There are, of course, numerous cut-offs that can be specified. Other common cut-offs
are for stockpiles of ore that is to be treated at some time after it has been mined, rather
than immediately. This is often separated into a number of categories, such as high,
medium and low-grade, based on the primary metal or value measure, and/or on the
basis of grades of deleterious constituents. Other cut-offs, commonly referred to as
cut-over grades, separate the ore into components to be treated by different processes;
typically, higher-grade material will go to a higher-recovery but higher-cost process,
while lower grades will go to a lower-recovery but lower-cost process.

Purpose
The purpose of a cut-off is to achieve some economic or financial goal, such as to ensure
profitability and maximise value, etc. Depending on the location of the project, there
may be various social goals such as maximising the life of the operations or employment
in the region and taxes paid to the government. And wherever the project is located,
minimising environmental damage and greenhouse gas emissions is becoming
increasingly important and the standard when doing business in the mining industry.
In many jurisdictions, companies are required by law to operate in the best interests of
shareholders. In most cases, this will mean boosting cash returns. Achieving various
societal goals may nevertheless be important for obtaining the so-called ‘social licence
to operate’, which is an essential precursor to being able to operate at all, and hence
work in the best interests of shareholders. Other goals set by companies or analysts are
perhaps thought to be proxies for cash generation, but many in reality are not. Common
goals are discussed in more detail in Chapter 10.
Naturally, a cut-off that delivers one goal will not necessarily deliver other goals. It
is therefore imperative that the goals of the cut-off (and other strategic decisions) are
stated before the derivation commences; however, in practice this is often not the case.

Method of determination
As is already evident, there are a number of ways in which the value used as a cut-
off can be derived. Break-even calculations are common and are discussed in detail in
Chapter 3. Lane’s methodology is in use at some open pit operations, and has also been
applied at a small number of underground mines though its application is often not
easy. This is discussed in more detail in Chapter 5. Lane’s methodology can be applied
using modern spreadsheet software. It is also embedded in some commercially available
strategy-optimisation software packages.

1. The question of whether material with a grade exactly equal to the cut-off should be classed as ore or waste
sometimes arises, but this will depend on how both rock grade values and cut-offs have been derived and
expressed, and what the cut-off is actually intended to achieve. This is explored further in subsequent chapters
that deal with different methods of deriving the cut-off.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 14
 CHAPTER 2 | Establishing Underlying Principles

Overall mine strategy optimisation is perhaps the other major process by which cut-
offs are specified, and this is discussed in more detail in Chapter 6. As previously noted
in Chapter 1, the cut-off policy should be an outcome of the planning process, rather
than a cut-off being derived (perhaps as a simple break-even) as a fixed input into the
planning process.

Warning regarding terminology

In this book, cut-off is merely a number while break-even is a generic term describing a
method of calculation. This is in contrast to a number of companies, where cut-off is used
both for the number and as the generic term for the method by which it is derived, which
in this book is called a break-even calculation. In some companies, the term break-even is
used to define a specific set of costs included in the calculation of a break-even cut-off.
This usage effectively implies that break-even calculations are the only valid way to
determine the number to use as the cut-off, since the calculation method is included in
the meaning of the term as defined by these companies’ cut-off policy manuals.
This book uses the term break-even cut-off to describe a cut-off whose value has been
derived by a break-even calculation method. Certain break-even cut-offs include
marginal break-even cut-off and full-cost break-even cut-off, in which the included costs are
indicated. This is discussed in more detail in Chapter 3.

Classification of Materials for Cut-offs and Strategy

Optimisation
In order to accurately determine cut-offs, we must correctly allocate costs and identify
production system capacities. First, it is necessary to identify the key cost drivers. Lane
(1988) introduces these concepts in the development of his cut-off theory, described in
more detail in Chapter 5; but the concepts are essential for a full understanding of all
types of cut-offs, so they are introduced here.
The term cut-off grade itself points to the necessary classification. Grade in its simplest
form is the ratio of the amount of product to the amount of ore or rock in which it is
contained. The primary purpose of the cut-off is to distinguish ore from the total rock
mass—to differentiate what is ore and what is not. Therefore, cut-off theory and practice
needs to deal with three classes of material: rock, ore and product.

Rock
In this publication, rock refers to all material mined or accessed before it has been
distinguished as ore or waste, or alternatively it is the total of ore and waste.
In an open pit mine, rock is typically all the material mined in and transported from
the pit. Simplistically, once the material arrives at the rim of the pit, it is then separated
into ore and waste by specifying the destination of the truck carrying a load of rock—
waste is sent to the waste dump while ore is sent to the treatment plant or to stockpiles
for treatment, either immediately or at some time in the future.
In an underground operation, rock is typically the total mineralised material. It may be
delimited by, for example, a geological boundary such as the boundary of the lithological
unit that contains the mineralisation. Alternatively, it might be by a grade-defined
boundary, where the delimiting grade would be at most the lowest possible cut-off to be
applied to distinguish ore from waste under very optimistic economic conditions.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 15
 CHAPTER 2 | Establishing Underlying Principles

Underground, rock is separated into ore and waste by, for example, planners creating
designs for stopes where the rock within the design limits is extracted as ore and the
remainder remains in situ as waste. The common classification of development as being
in ore or waste, or as being capital or operating expenditure, is not relevant for our
understanding of cut-off theory. The important issue for development is not the grade
of the material being mined through, nor how it will be recorded in the company’s
accounts. Rather it is whether it is providing access to rock that can be then be stoped or
not, or if it is directly required for and associated with ore extraction. The classification
of development for cut-off derivation is discussed more in Chapter 3.
The author is aware of some companies that use the term rock to refer to what is in this
book referred to as waste. Clarifying the terminology is essential to avoid confusion. In
this book, for open pits:

Rock = ore (treated or stockpiled) + waste = total material moved

Underground rock, as indicated earlier, is the total contents of the mineralised

envelope, however that might be defined. What is excavated underground and usually
classified as waste or rock, depending on local terminology, is typically from major
access development, often outside the mineralised envelope. This material, however,
is not waste in the specialised sense applied in cut-off theory – it is simply an outcome
of the activities undertaken to expose the rock so that the ore it contains can be stoped.
Note that rock being classified as such ‘before distinguishing between ore and waste’
is a logical concept for the purposes of cut-off specification and strategy optimisation. In
practice, the distinction may well have been made before mining occurs so that different
mining processes can be applied if appropriate and long-term plans can be developed,
but this implies that a cut-off has already been applied. To determine the cut-off in the
first place, the total rock mass must be viewed as having a separate identity before the
cut-off is imposed.

Ore
Ore is the material intended to be processed. It will typically be the mineralised material
to be sent through the mine’s ore-handling system and for processing through the
treatment plant. It may be treated at the time of mining or stockpiled for treatment at a
later date. There may be a sequence of processes dealing with the ore.
As noted in Chapter 1, the term ore is used in this book as just described and not
according to the strict definitions specified in the JORC Code, NI 43-101, SAMREC and
similar codes for public reporting. The cut-off derivation and strategy optimisation
processes described in this book are internal to the company performing the evaluations,
where the use of such terminology is not required. This typically involves identifying
a number of sets of potential reserves at different cut-offs, one of which will ultimately
form the basis for the mine plan that supports the publicly reported reserve.

Product
Product is the valuable material separated from the ore. It will typically be the material
in the grade unit – for example, gold or copper metal for grades expressed as ounces
or grams per tonne, or as a metal percentage; however, this does not necessarily imply
that the product for cut-off derivation purposes is the actual product produced and

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 16
 CHAPTER 2 | Establishing Underlying Principles

sold by the operation. For instance, a gold mine might produce and sell gold bullion,
so the mine’s product is also the product for cut-off derivation purposes. However, for
a base metal operation, the product for cut-off purposes is the metal, but the product
produced and sold by the mine will typically be a concentrate. Some operations may
have multiple process stages producing progressively more refined forms of product
such as copper concentrate, blister, anode and cathode.
For base and precious metal operations, the product should be a relatively small
proportion of the ore but for bulk commodities, such as iron ore and coal, there may be
no separation of product from ore (the ore is the product). If there is some beneficiation,
the product may be a large proportion of the ore.
In most of the discussions in this book, the terms product and metal are used
interchangeably. Where metal is used, it does not preclude the consideration of coal or
other non-metallic products. The term metal grade will often distinguish a grade that is
expressed in terms of units of product per unit of ore – such as gold grams per tonne or
copper per cent – from any generic grade measure of value attributed to a block of rock.

Material types in open pit and underground mines

It should be noted that while the operations in an open pit mine deal solely with rock that
is subsequently separated into ore and waste, most of the operations in an underground
mine typically deal, or are associated, with ore. The exception is usually access to and
infrastructure development for a new mining block: this can be thought of as exposing a
new block of rock, which is subsequently separated into ore to be stoped and waste to be
left in situ. Major access development is therefore generally the only rock-related activity
in an underground mine for cut-off derivation purposes.
All mines separate ore from waste in some way. There may, however, be operations
where all the rock mined in an open pit is ore. This would commonly occur where the
nature of the mineralisation is such that the pit is only excavated in material above a
specified cut-off, with all lower-grade material left in situ. In this case, the ore–waste
boundary within the overall rock mass has been defined by the mine planning processes
and the operations are, for cut-off derivation and strategy optimisation purposes, more
akin to an underground operation, though mined by open pit methods.

Classification of Costs for Cut-offs and Strategy

Optimisation
As previously noted, in order to accurately determine cut-offs, it is necessary that we
classify and allocate costs. The three material types with which we are dealing – rock, ore
and product – have been identified. We’ll now identify how costs behave with respect
to each of these. Rock, ore and product are also the key physical cost drivers for cut-off
derivation and optimisation.

Cost drivers
Although it is essential to separate costs into those driven by rock, by ore and by product,
the operation’s cost accounting system may not easily facilitate this. For example, in the
treatment plant all costs will often be quoted as a cost per tonne of ore, regardless that
some are related to the amount of product generated and some incurred by the passing
of time, irrespective of quantities of both ore and product handled. It may be necessary to

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 17
 CHAPTER 2 | Establishing Underlying Principles

go back to more detailed records to reallocate costs according to the physical parameters
that drive them. More detail on applying this cost classification is provided in Chapter 3
in the description of break-evens.

Variable costs
Variable costs are those where the total dollar cost varies in direct proportion to the
physical activity or quantity. Unit costs expressed as, for example, dollars per tonne or
dollars per ounce are an accurate representation of cost relationships and behaviour.
Each additional tonne or ounce of material dealt with requires spending the specified
number of dollars to complete the action.

Fixed costs
Fixed costs are identified as costs that do not change relative to physical activity: they
are the same for any specified time at all levels of activity. Total dollars of fixed costs
over the life of the operation may also be thought of as being variable with respect
to elapsed time, rather than to some physical quantity such as tonnes. Fixed costs are
commonly expressed as dollars per year or per month. Again, the correct expression
of a fixed cost describes its behaviour: each additional month or year of operation will
require spending the stated number of dollars, regardless of how much or how little
production activity occurs.

Lump sum costs

A lump sum cost may be defined as a cost for which the quantum does not change
with either physical activity or time. Usually it is expended to provide some capability
and will be a capital item, such as the construction or expansion of a part of the plant.
It may also be the cost of development to expose and make available new rock. In an
underground mine it will be the cost to access a zone of mineralisation remote from
the current mining operation or, for example, to extend the decline to make the next
sublevel block of the resource available. Although the quantum of the lump sum may
derive from what appears to be a variable cost relationship (for example, the required
metres of development multiplied by the cost per metre), it is nevertheless a cost that is
not driven by the tonnage of rock, ore or product that may be exposed, mined or treated,
nor by the time taken. The cost of exposing a new sublevel block is not dependent on
how much rock (or mineralised lithology) is exposed, nor on how much ore is delineated
within that rock (it could be all the mineralisation, none of it, or any amount in between),
nor on whether the development is done quickly or slowly.
Certain lump sum costs will, of course, be related to rock, ore or product quantities. The
cost to construct a concentrator with a specific ore treatment rate will obviously be less
than that for a plant of double the capacity; but this is the cost of acquiring a capability
or alternative capabilities, not the cost of actually treating the material. A discussion of
capital costs follows in this chapter, and the handling of different production capabilities
is covered in Chapters 5 and 8.

Other cost behaviours

Not all costs behave in a purely linear fashion that can be separated into fixed and variable
components. There may be a continuous non-linear relationship between dollars spent

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 18
 CHAPTER 2 | Establishing Underlying Principles

and physical activity, or step changes at certain levels. These types of behaviours are
discussed in more detail in Chapter 8. At this point in our understanding, it is sufficient
to recognise that the various cut-off models we will discuss effectively require costs to be
either fixed or variable. Whatever the true relationship of cost versus activity may be, that
relationship will have a gradient that can be resolved into fixed and variable components.

Project capital and sustaining capital costs

The financial accounting distinction between capital and operating costs is largely
irrelevant for cut-off derivation purposes. These classifications have been developed by
accountants for public accounting reports for the same sorts of reasons that categories of
Mineral Resources and Ore Reserves have been defined by codes for public reporting. This
classification of costs is, however, misleading for cut-off derivation. What is important
is the way the costs behave and what they achieve. It should be noted though that if a
cut-off evaluation is being conducted on an after-tax basis, the accounting classification
of capital and operating costs must be retained to ensure that tax deductibility and tax
costs are applied and derived appropriately.
For cut-off derivation and strategy optimisation we do, however, need to identify two
types of capital expenditure: sustaining capital and project capital.
Sustaining capital maintains existing capabilities. It is a regular, ongoing cost incurred
for such things as fleet and plant replacement, and may be thought of as maintenance
costs spent irregularly. For maintaining existing capital items, there is no conceptual
difference between replacing an oil filter every 50 hours (an operating cost) and
replacing a whole truck every 50 000 hours (a capital cost). Sustaining capital for an
individual item or category of plant or fleet can therefore be thought of as ‘lumpy’ or
irregular maintenance expenditure that, by virtue of its spending pattern rather than its
behaviour, is classed by accountants as a capital rather than an operating cost.
Sustaining capital costs should normally be included in the cut-off derivation in the
same way as similarly behaving operating costs – as variable or fixed costs. Because of
the irregular nature of replacement capital expenditure for any type of asset, sustaining
capital will be used to acquire different items in successive years.Companies will
often manage their replacement programs so that the aggregate amount spent will
be relatively constant over time. This sometimes results in sustaining capital costs
being viewed as fixed costs, but in reality sustaining capital is incurred because of
the wearing out of plant due to the level of mining and processing activities. These
activities are often planned to be similar from year to year, so sustaining capital will
therefore appear to be a regular fixed cost per year. However, if the production rate
were to double or halve, the reality is that annual sustaining capital spending would
do the same since equipment would be wearing out faster or slower. Since cut-off
and strategy optimisation often considers changing production rates, it is necessary to
evaluate carefully what is driving sustaining capital costs. Often all or most of what
is quoted as a fixed cost needs to be converted to a variable cost related to the current
level of production activities, so that variations in production in each time period can
be accounted for.
Project capital buys additional capability such as mill expansion, an additional truck,
etc as well as improved product quality. It is typically a lump sum cost, though the
lump sum dollars could be determined like a variable cost as recently described. The

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 19
 CHAPTER 2 | Establishing Underlying Principles

resulting lump sum cost does not depend on the amount of ore mined but rather ‘buys’
the capability to do something different.
Under this definition, project capital is not for the ‘projects’ that are the current focus
of the existing capital plan; many of these will in fact be sustaining capital. Whereas
sustaining capital is a normal, ongoing cost of operating, spending of project capital is
discretionary, justified by the difference in values of mine plans with and without the
capital expenditure.
There are other ways to consider this issue. The argument is that replacing the truck
is an ongoing cost of continuing to produce at the planned rate. That presupposes that
the mine will continue to operate at a particular production rate, but it need not do so.
One could alternatively say that the base case is to not spend the replacement capital
but instead allow the production rate to decline as the availability and performance of
the ageing piece of equipment deteriorates, accepting that the production rate will drop
to zero at some time when there is a catastrophic failure. In that scenario, replacing a
truck is classed as a project capital item, where the benefit is in avoiding the decline or
sudden stoppage of production, justifying the expenditure. But one could also argue the
same logic for replacing the oil filter, which is normally classed as an operating cost. The
same thinking will apply: if the oil filter is not replaced the engine will deteriorate and
fail, and production will cease.
It can again be seen that there is no practical distinction between capital and operating
costs, as long as cost behaviours and other assumptions are identified and consistently
applied.

Applying Cut-offs to Distinguish Ore from Waste

There are a number of ways that cut-offs can be applied to delineate the orebody to be
mined and processed, taking account of mining practicalities. Often the grade variability
in the mineralisation is such that it is not practical to mine and treat as ore only the
material above cut-off while classifying as waste all the material below cut-off. It is
common for the practical application of the cut-off to result in excluding some ore-grade
material above cut-off from the mining inventory and including some sub-cut-off waste.
There is no correct way to do this – each case must be considered on its own merits.
It will be part of the cut-off evaluator’s task to identify the rules by which cut-offs are
to be applied in order to best achieve corporate goals. The following sections describe
the more common ways in which cut-offs can be applied, but do not purport to be
exhaustive. These descriptions are to establish underlying principles that are important
in the discussions in later chapters.

Boundary cut-offs
A boundary cut-off rule is applied to delineate the ore-waste boundary of an orebody,
stope, etc. In essence it takes account only of the location of the specified cut-off grade
in the rock mass. If the geological ore–waste boundary at the specified cut-off is regular,
the boundary cut-off rule may be trivial – a practical mining boundary between ore and
waste may be simply defined by following the geological boundary or grade contour.
If, however, the geological boundary in the rock at the specified cut-off is irregular, the
boundary cut-off rule will indicate how the practical boundary is to be placed relative to
the actual boundary. At one extreme, the rule might specify that the mining boundary

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 20
 CHAPTER 2 | Establishing Underlying Principles

is to be located entirely inside the geological boundary so that no sub-cut-off material is

included in the mining inventory. Conversely, the mining boundary could be entirely
outside the geological boundary so that no above-cut-off material is excluded. Within
these extremes could be a number of rules specifying how the trade-off between ore loss
and waste inclusion is to be handled.
When a boundary cut-off rule is applied in isolation, no cognisance is taken of grades
of the resulting stopes, orebodies, etc. Depending on the grade distribution in the rock,
the average ore grade defined could be significantly or only marginally above the
specified cut-off. Applying a boundary cut-off alone implies nothing about the grade of
the resulting ore. There is no guarantee that the ore defined will deliver any particular
corporate goal, and depending on the grade distribution and cut-off specified, this may
or may not be a problem. At low cut-offs, there will potentially be sufficient higher grade
material above cut-off to deliver an average grade higher than the cut-off, whereas at
higher cut-offs, the remaining material above cut-off may only be marginally above it. If
the boundary cut-off has been applied to the in situ grades of the rock and allowance is
subsequently made for overbreak and dilution, the final grade of the volume may even
be less than the cut-off applied.
Unless otherwise described, cut-offs in this book are assumed to be boundary cut-offs.

Volume cut-offs
Applying a volume cut-off rule addresses the potential shortcomings of the boundary
cut-off by accounting for the resulting average grade. A boundary rule would first be
applied to generate an initial mining inventory. Volumes, such as individual stopes
or subregions within the orebody defined, with average grades below the specified
volume cut-off would then be deleted from the inventory.
The volume cut-off need not be the same as the boundary cut-off; however, in the
author’s experience operations that apply a volume cut-off typically apply the same
value as the boundary cut-off.
The volume cut-off rule may allow for factors not accounted for in the boundary cut-
off. In particular, the effect of ore recovery and dilution on tonnages and grades to be
dealt with if the material is classed as ore may be applied to change the average grade of
the material or the unit costs included in the cut-off derivation.

Other rules
Depending on the nature of the mineralisation, other rules may be applied to cull material
above other cut-offs from the mining inventory or to include sub-cut-off material in the
inventory. For example, an outlier rule may have to be applied to exclude small volumes
of otherwise ore-grade material remote from the main ore zone(s) if including it in the
inventory destroys rather than adds value. Conversely, rules may need to include small
unmined waste or low-grade areas within an area that is generally mined out if, for
example, damaging stress concentrations could develop in the unmined pillars.

Planning and Operational Cut-offs

Most mines will operate with a number of cut-offs. These have different purposes and
will be applied in distinct circumstances. The cut-offs to be used must be clearly defined,

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 21
 CHAPTER 2 | Establishing Underlying Principles

and their application specified in company documentation. It should be noted that there
are few if any standard definitions used consistently throughout the industry, from
company to company and country to country. The definitions that follow are therefore
for consistency throughout this book. Companies could choose to use different terms
and definitions; however, these descriptions will be typical of those used at mines.
A planning cut-off is used in developing mine plans and designs as well as in long-term
studies, such as feasibility studies and strategy studies, to distinguish between ore and
waste. It sets the overall mine strategy.2
An operational cut-off is used on a day-to-day basis by mine operators for making
decisions as to what is ore or waste. Because of the variability in orebodies and our
inability to predict perfectly what will be in the mine, not to mention product prices, it
may be both necessary and legitimate to vary the cut-off from the long-term planning
cut-off in the short term. The variations must be within the context and constraints of
the overall plan, and that optimum long-term plan includes specifying the long-term
cut-off; however, the reasons for the variation must be understood, and part of the aim
of short- to medium-term mine planning will be to return to the planning cut-off. The
operational cut-off must be derived within the context of the planning cut-off. If it is not,
it may result in a definition of what is ore (and what is not) that may be a long way from
optimal and very difficult to change.3 This is discussed in more detail late in Chapter 3,
which covers the dangers of break-even cut-offs.
Although related, planning and operational cut-offs are distinct. One or the other may
have more prominence in different types of mines, but they must not be confused. For
example, in an open stoping underground mine, the planning cut-off will tend to have
prominence. Planning engineers will design stopes to the planning cut-off, and there
may be little or no scope to change the cut-off used by the operators on a short-term
basis: what is designed is what gets blasted and produced. Alternatively in a cut-and-
fill mine, although the initial stope sills and infrastructure development may have been
planned according to the planning cut-off—and in general the stoping operations will be
controlled to comply with this—there is much more scope to vary the cut-off applied in
the short-term; for example, to keep the mill supplied with lower grade but marginally
profitable material below the planning cut-off when there is, for whatever reason, a
shortfall of ore available above the planning cut-off.

2. The planning cut-off could be thought of as analogous to a motorway or freeway: it is the main route to get
from A to B. So long as we remain on the freeway, we will reach where we want to be quickly and efficiently.
In the mining business, it represents the path to achieving the corporate goal—whatever that is specified to
be—most effectively. We can of course go from A to B by way of side roads off the main freeway, but at the cost
of additional time, increased fuel consumption, etc. So too we may finally deplete our orebody and close our
mine, but if we have gone by way of suboptimal production rates, cut-offs and other options, we will not have
achieved the corporate goals as well as we could have. We will have taken longer, spent more or earned less
revenue than we might have. And just as a freeway will never be perfectly straight, having both its direction and
gradient change to suit the topography and environment, so too will we expect that the planning cut-off will
change over time to account for longer-term variations in, for example, geology and economic conditions.
3. Continuing the freeway metaphor, the operational cut-off may be thought of as the lane in which we are
travelling. As we see traffic ahead slowing in the lane we are in, we change lanes into an adjacent faster-moving
lane, so maintaining our efficient progress towards our destination. The freeway keeps us travelling on the right
path, but we change lanes as problems and congestion manifest themselves in one course or another. The
important point when changing lanes is that we do not lose track of the through route or find ourselves in a
position in which there is no option but to exit the freeway.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 22
 CHAPTER 2 | Establishing Underlying Principles

The danger in the latter case is that, having identified that it is easier to ‘fill the mill’ at
the lower operational cut-off, the longer-term cut-off can be changed to this lower cut-
off. The planning cut-off grade has become the grade derived for the operational cut-off,
so rather than moving back to the original planning cut-off after the short-term problem
has been resolved, the mine remains on a lower cut-off strategy. If the higher planning
cut-off delivered the optimal plan, short-term considerations will have been allowed to
subvert the previously-identified long-term plan and become the long-term plan. This
can only lead to suboptimal outcomes.

Cut-offs and Company Goals – Or how the junior engineer

determines the CEO’s bonus
Specifying the corporate and societal goals for an operation must be a primary part
of the cut-off derivation process. Without specifying these goals, the goals implicit in
the cut-off derivation process will become the de facto high-ranking corporate goals,
whether they are recognised as such or not.
In many companies, cut-offs are derived at the mines by relatively junior mining
engineers or geologists. In the absence of any other direction or by following past
practices, they determine a break-even grade that is then applied as the cut-off for the
next planning cycle. The how of cut-offs has become indistinguishable from the what –
cut-off and break-even have become synonymous by default. Using a simple break-even
grade as the cut-off merely ensures that the break-even is achieved. Break-evens will be
discussed in more detail in the next chapter, but for now, we could say that simplistically
the goal implicit in a break-even is that ‘every tonne pays for itself’.
As already noted, the cut-off that delivers one goal will not necessarily deliver another.
The author is not aware of any public company whose stated principal aim is to ensure
that every tonne of ore pays for itself. More commonly we see goals that speak of
‘maximising returns for shareholders’ or similar. There is no reason why a cut-off that
ensures that every tonne pays for itself should also maximise shareholder value; indeed,
the experience of strategy optimisation practitioners suggests that it will not! So while
senior executives are telling shareholders that they are seeking to maximise their returns,
operating staff—albeit with the best intentions—are implementing mining strategies
that are almost guaranteed not to deliver the promised returns to shareholders.
The corporate goals must therefore be integral to the cut-off and strategy evaluation
process from senior managers through to the person on-site deciding what is ore and
what is waste. In many companies the corporate goals are not transmitted down the line,
which means decision-making that is aligned with the goals is difficult, if not impossible.
It should be noted that this assumes that the company is at liberty to select a cut-off that
is aligned with and helps to deliver the corporate goals. In some jurisdictions, however,
the way the cut-off is determined is mandated by law. In such places, the cut-off clearly
needs to be whatever is specified, and any variations can only be what is legally permitted.
Achieving the corporate goal will be constrained by the allowable limits within the law.

Summary
It is important to establish key concepts and definitions from the start of the discussion
on cut-offs and strategy optimisation.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 23
 CHAPTER 2 | Establishing Underlying Principles

Essentially, we must distinguish between what a cut-off is – a number that separates

two courses of action, in particular the separation of ore and waste – and its purpose and
method of determination. Break-even calculations are but one way of deriving the value
for the cut-off. Break-even and cut-off are not the same. Much of the misunderstanding
surrounding cut-offs stems from a failure to distinguish between the what, why and how.
We need to distinguish three classes of material:
1. Rock—the total ore and waste (or total material) before the distinction between ore
and waste has been made.
2. Ore—the portion of the rock that we intend to treat and derive value from.
3. Product—the component/s of the ore that actually generate the revenue and hence
the value. There are costs and production capacities associated with or driven by
each of these.
The way in which costs behave is crucial to correctly determining the cut-off grade:
•• Variable costs are those costs for which total dollars vary in direct proportion to the
quantity of some physical parameter, such as tonnes or metres.
•• Fixed costs are usually identified as costs not varying relative to physical activity; they
are the same for any specified time at all levels of activity. Total dollars of fixed costs
are variable with respect to elapsed time, rather than to physical quantity.
•• A lump sum cost may be defined as a specified cost that does not change with either
physical activity or time. Typically it is expended to provide some capability and will
be a capital item, such as the construction or expansion of a part of the plant. It may
also be the cost of development to expose and make available new rock.
The accounting distinction between capital and operating costs is irrelevant and
potentially misleading for cut-off derivation. The critical issue is how costs behave:
•• Project capital buys the capability to do something new or different and will typically
be a lump sum cost.
•• Sustaining capital maintains existing capabilities, and is typically irregular maintenance
or replacement expenditure that is nevertheless driven by the wearing out of plant
and equipment because of its use in operations. Although often expressed as an
annual (fixed) cost in a stable, steady-state operation, sustaining capital will more
usually be incurred because of some physical activity, such as tonnage that has been
treated or handled. Sustaining capital should therefore be handled in the same way
as variable operating costs related to the relevant activity.
Cut-offs may be applied to define the boundary between ore and waste without
reference to the grade of ore thereby delineated. They may also be applied to the grade
of the volume enclosed by a boundary that has been defined. There is no one correct
method of application, and different mines may apply cut-offs differently for a number
of valid reasons.
Planning cut-offs are used to set the long-term big-picture cut-off strategy. Operational
cut-offs used on a short-term basis in operations may vary from the planning cut-off due
to, for example, natural variations and fluctuations within the deposit, conditions in the
plant, and economic conditions that vary from the long-term prediction. Planning and
operational cut-offs must not be confused, however, and the operational cut-off should
always be varying within the confines of the longer-term planning cut-off policy, which
should be directed at achieving the corporate goals.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 24
 HOME

Chapter 3

Break-even Analysis –
Accounting for Costs and Prices

A common but limited cut-off model

Many mining operations worldwide are operating with what may be called a break-even
cut-off. What is meant by the term break-even? In its simplest manifestation as it applies
to cut-off derivation, the break-even is the grade at which revenue obtained is equal to
the cost of producing that revenue.
The implication of this for cut-off purposes is clear: any grade higher than this will
generate more revenue than the costs to produce that revenue, while any grade lower
will generate less revenue than costs. This makes a break-even grade an obvious one to
use as a cut-off – any material with a higher grade is apparently worth classifying as ore
while lower-grade material is not. This, however, is not necessarily true.
It is important at this early stage of the discussion to recognise that break-even and cut-
off are not synonymous: cut-off is not break-even, and break-even is not cut-off. In many
cases, however, the terms are used interchangeably (noted in Chapter 2), and much
of the confusion surrounding cut-offs comes from this misapprehension. The previous
chapter identified that the cut-off is simply a number. A break-even calculation is one
way of deriving a number that could be used as a cut-off; however, it is not the only
valid way, as later chapters will illustrate.
Break-even analysis is essentially what might be called a one-dimensional (1D)
process. It is based on financial parameters only. Costs and prices, adjusted to account
for metallurgical recovery, are compared. Although widely used in the industry, break-
even is a very limited cut-off model. In particular, it does not explicitly account for the
geology, the nature of the mineralisation and the capacities of the mining and processing

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 25
 CHAPTER 3 | Break-even Analysis

plant. Yet the simple break-even model can be useful so long as it is used within the
constraints of the outcomes of a more comprehensive model.
The reality is that in practice, a simple break-even model is all that is used by many
operations. This can lead to mine plans that are almost guaranteed to not deliver the
company’s stated goals. Indeed, there is no guarantee that an orebody delineated by
a break-even cut-off derived according to typical industry practices will actually even
deliver a profit.

The Simple Break-even Formula

The simple break-even formula to determine the grade at which revenue obtained is
equal to the cost of producing that revenue is:

cost [$/t]
Break-even grade [gu/t] =
product price [$/gu] # recovery

Conversion factors may be required to convert the currencies in which costs and/or
prices are specified into the same currency and to convert the price into the unit in which
the grade is expressed. For example, gold price is usually quoted in US dollars per ounce
(US$/oz), but in many countries, gold grade is expressed in grams per tonne (g/t). Since
there are approximately 31.1 grams per Troy ounce, the gold price in US$/oz must be
divided by 31.1 to obtain the price per gram, so that the formula can be used to obtain
a break-even grade in grams per tonne. So, for example, if the price is US$1244/oz,
this equates to US$40/g. Including such conversions adds slightly to the calculation but
does not change the inherent simplicity of the basic break-even formula.
Although the formula is simple, many companies spend a lot of time and energy
arguing over what costs should be included and then extracting them from the
corporate cost accounting systems. The principal aim of this chapter is to establish
principles regarding what costs should go into break-even calculations, and perhaps
more importantly, why. If the why is understood, defining the what should become
relatively straightforward.
Ultimately, the costs are whatever the user chooses to specify; however, there are no
industry-wide standard definitions. Each company has its own. In this book, break-even
is used as a generic term. The author is, however, aware of at least two major mining
companies that use the term to define specific costs to be included in the calculation. It
is therefore essential, when dealing with break-even cut-offs, to identify clearly which
costs are (and which are not) included in the cost term of the formula.

A simple ore–waste break-even calculation

Consider this simple break-even calculation example:
•• product price $10/gu
•• recovery 90 per cent, and therefore
•• recoverable revenue $9.00/gu
•• variable cost $22.50/t
•• total cost $60/t

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 26
 CHAPTER 3 | Break-even Analysis

$22.50/t
Variable cost break-even =
$10/gu # 90%

= 2.5 gu/t

$60.00/t
and total cost break-even =
$10/gu # 90%

= 6.7 gu/t

The grade is expressed as grade units per tonne (gu/t); for example, per cent (or
product tonnes per ore tonne) for base metals, or grams per tonne for precious metals.
The metal price has been converted from the price per unit of metal as normally quoted
to a price per grade unit to simplify the break-even calculation.
Figure 3.1 shows these break-even relationships graphically. Variable and total unit
costs and unit revenues are plotted as straight-line functions of material grade. Costs per
tonne are assumed to be independent of grade and are therefore the same at all grades,
being $22.50 and $60.00/t, as in the data above. The revenue received increases at the
rate of $9/t/gu, as calculated above. As expected, the revenue line crosses the cost lines
at 2.5 and 6.7 gu/t, the break-even grades calculated.
Figure 3.2 also shows these break-even relationships graphically but this time as net
contributions, positive and negative. The costs already noted have simply been deducted
from the revenue. At zero grade with no revenue generated, the contribution is negative
and numerically equal to the cost per tonne. The contribution based on variable costs
increases linearly at the rate of $9/t/gu from –$22.50/t at zero grade. Similarly, the
contribution based on total costs increases at the rate of $9/t/gu from –$60/t at zero
grade. As expected, the net contribution lines cross the zero contribution axis at 2.5 and
6.7 gu/t, again the break-even grades calculated previously.

FIGURE 3.1
Graphical representation of break-evens: revenue = cost.

100

80 Total Cost Break-even

(6.7 gu/t)
60
$/tonne

40 Variable Break-even
(2.5 gu/t)

20

0
0 1 2 3 4 5 6 7 8 9 10
Grade (grade units)

Variable Cost ($22.5/t) Total Cost ($60/t)

Net Revenue ($9/t/gu)

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 27
 CHAPTER 3 | Break-even Analysis

Figure 3.2
Graphical representation of break-evens: net contribution = 0.

100
80
60
40
Variable Break-even
$/tonne

20 (2.5 gu/t)

0
–20 Total Cost Break-even
(6.7 gu/t)
–40
–60
–80
0 1 2 3 4 5 6 7 8 9 10
Grade (grade units)

Variable Contribution Total Contribution

A More General Definition of Break-even – Equal Net Benefits

The simple definition of break-even, as previously described, is the grade at which
revenue obtained is equal to the costs of producing that revenue. A more general
definition is the grade at which the benefit from one course of action equals the benefit
from an alternative course of action.
Using this definition, the concept of the ore–waste break-even becomes the grade at
which the benefit obtained from treating a block of rock as ore is the same as treating
it as waste. This leads us to what is commonly known as a cut-over grade, where the
benefit from treating ore by one process is the same as treating it by another. Typically
this is applied to two distinct metallurgical processes, one with relatively high costs and
recoveries, the other with relatively low costs and recoveries.
The equal benefits break-even occurs when:

Price × grade × recovery1 – cost1 = price × grade × recovery2 – cost2

Rearranging the terms of this formula, we obtain:

cost1 - cost2
Break-even grade [gu/t] =
product price # (recovery1 - recovery2)

An equal net benefits ore–waste break-even calculation

Let us reconsider the break-even calculation example, the key features of which are
reproduced here:
•• product price $10/gu
•• recovery 90 per cent, and therefore
•• recoverable revenue $9.00/gu
•• variable cost $22.50/t
Assume now that the variable costs comprise the following components:

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 28
 CHAPTER 3 | Break-even Analysis

•• treatment $10/t of ore

•• mining $12.5/t of ore
•• strip ratio 1.5 t of waste per tonne of ore
(therefore 2.5 t of rock mined per tonne of ore)
•• thus, mining cost $5/t of rock
Therefore, the variable cost incurred by each tonne of rock if classified as:
•• ore: $10 (treatment) + $5 (mining) = $15/t total
•• waste: $5/t (mining only)
Break-even occurs when:

Price × grade × recoveryore - costore = price × grade × recoverywaste - costwaste

costore - cost waste

Break-even grade =
product price # (recoveryore - recovery waste)

$15 - $5
=
$10 # (0.9 - 0.0)

= $10/$9

= 1.11 gu/t

Figure 3.3 shows graphically the net benefits of classifying material as ore or waste.
The net benefit of waste is negative – a cost of $5/t at all grades of material. The benefit
of ore increases linearly at the rate of $9/t/gu from –$15.00/t at zero grade. At zero
grade with no revenue generated, the net benefit is negative and equal to the cost per
tonne of material that is both mined (at $5.00/t) and treated (at $10.00/t). The benefit-
of-ore line crosses the benefit-of-waste line at a grade of 1.11, as calculated. Note that
the net contribution from ore does not become positive until the grade exceeds 1.67,

FIGURE 3.3
Graphical representation of ore–waste break-evens – equal benefits.
80

60
$ / tonne

40
Ore/Waste
20 Break-even
1.1 1 gu/t

–20
0 1 2 3 4 5 6 7 8 9 10
Grade (grade units)

Contribution (Ore) Contribution (Waste)

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 29
 CHAPTER 3 | Break-even Analysis

at which point the revenue covers the costs of both mining and treatment. However,
given the assumption that the mine is operating, the $5.00/t mining cost is a committed
or sunk cost and the net revenue from treating a tonne of rock as ore need only cover
the $10.00/t cost of treatment to make it worthwhile doing so. As noted in the preceding
chapter, there is no guarantee that the value obtained when using the appropriate cut-
off will be positive.
Our simple break-even calculation generated a value of 2.5 gu/t, compared with the
equal benefits value of 1.11. What causes this difference? The distinctions in rationale
and the arithmetic of the two calculations are self-evident, but what is perhaps more
important for our understanding of cut-off derivation is the recognition that mines’ cost
systems often report all costs as a cost per tonne of ore, which may be satisfactory for
many purposes but can be dangerously misleading when used for cut-off derivation
without critical thought. As previously indicated, it is essential that costs be allocated to
the physical activities that actually cause them to be incurred.
It can be seen from the derivation of the mining cost of $5.00/t of rock above that the
$12.50/t total mining cost for ore is made up of $5.00 for the one tonne of rock that is
classed as ore, plus $7.50 for the associated 1.5 t of rock classed as waste. The split of the
costs of mining the total 2.5 t of rock between ore and waste is a function of the cut-off
applied. If the cut-off were such that 0.5 t was ore and 2.0 t was waste, the cost per tonne
of ore would be reported as $25.00/t. Conversely, if 2.0 t was ore and 0.5 t was waste,
the cost per tonne of ore would be reported as $6.25/t. It is now evident that in the
first simple calculation above, where the reported cost per tonne of ore was used in the
break-even calculation, one of the main component costs used to determine the cut-off is
itself a function of the cut-off. The value used is based on a prior assumption about what
the cut-off is, generating the 1 t of ore/1.5 t of waste split.
This introduces a circularity into the calculation. If a set of iterative calculations were
conducted, the process might eventually converge to the right answer; however, that
would necessitate making use of the tonnage versus cut-off relationships for the block
of rock to identify the ore–waste split at the cut-off calculated, from which one could
recalculate the mining cost per tonne of ore, the break-even and so on. This iterative
process is described in more detail later in this chapter.
There is no problem with this in principle, but in the development of cut-off theory
we are at this stage considering break-even calculations that have been identified as
1D, considering only financial factors (prices and costs). The necessity to consider the
geology, as expressed by the tonnage versus cut-off relationships, means we need to go
beyond the 1D models’ limitations and bring other parameters into the analysis. At first
glance, this implies that the simple break-even cut-off model is inadequate for such a
situation, and so it is if costs are expressed only per tonne of ore.
By using the more general definition and calculation of break-evens, however, and by
correctly allocating costs to rock and ore, the circularity of having costs dependent on
cut-off in the break-even calculation has been removed. A break-even cut-off grade can
be derived by a simple 1D break-even calculation.

Cut-over calculation between two processes

Let us consider now two alternative processes. Mining costs remain at $5/t of rock and:

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 30
 CHAPTER 3 | Break-even Analysis

1. Process 1 has a 90 per cent recovery and costs $10/t of ore treated, as in the previous
example
2. Process 2 has a 98 per cent recovery but incurs costs of $15/t of ore treated.
The break-even cut-over occurs when:

Price × grade × recoveryprocess 2 - costprocess 2

= price × grade × recoveryprocess 1 - costprocess 1

costprocess 2 - costprocess 1
Break-even grade =
product price # (recoveryprocess 2 - recoveryprocess 1)

$(5 + 15) - $(5 + 10)

=
$10 # (0.98 - 0.90)

= $5.00/$0.80

= 6.25 gu/t

As determined previously, the ore–waste break-even for Process 1 is 1.11 gu/t.

Applying the same calculation for Process 2 generates an ore–waste break-even of
1.53 gu/t but this is not used as a cut-off or cut-over. Combined with the derivation of
the Process 1 - Process 2 cut-over of 6.25 gu/t, it can be seen that rock with a grade:
•• less than 1.11 should be classified as waste
•• greater than 1.11 and less than 6.25 should be classified as ore and treated by Process 1
•• greater than 6.25 should be classified as ore and treated by Process 2.
Figure 3.4 shows graphically the net benefits of classifying material as waste and as ore
treated by both Process 1 and Process 2. The first two lines are the same as in Figure 3.3.
The net benefit of waste is –$5/t at all grades of material. The benefit of ore treated by
Process 1 increases linearly at the rate of $9/t/gu from –$15.00/t at zero grade. Similarly,

FIGURE 3.4
Graphical representation of ore–waste break-evens and cut-over between two processes.
80
Ore to Ore to
Waste Process 1 Process 2
60

40
$/tonne

P1/P2 Cut-over
(6.25 gu/t)
20

Ore/Waste Break-even (P1: 1.11 gu/t P2: 1.53 gu/t)

–20
0 1 2 3 4 5 6 7 8 9 10
Grade (grade units)

Contribution (Process 1) Contribution (Process 2)

Contribution (Waste)

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 31
 CHAPTER 3 | Break-even Analysis

the benefit of ore treated by Process 2 increases linearly at the rate of $9.80/t/gu from
–$20.00/t at zero grade, being the total cost of mining and treatment by Process 2. The
Process 2 line crosses the Process 1 line at 6.25 gu/t.
Note that the ore–waste break-even for Process 2 (1.53 gu/t) does not enter into the
specifications of how rock is classified. In the absence of Process 1, rock with a grade
above 1.53 would be classed as ore and treated by Process 2, but Process 1 provides
better returns than Process 2 for all grades up to the identified cut-over of 6.25.
As mining progresses through the deposit, there is no guarantee that the distribution
of grades will be such that both processes will operate at full capacity. If one process has
spare capacity while the other is overloaded, the simple 1D break-even calculation does
not suggest what to do with the excess material which, according to the cut-over value,
should be processed by the overloaded process. It could be stockpiled until it can be
treated by that process or treated immediately through the less profitable (for that grade
of material) process with spare capacity. The simple 1D break-even cut-off model does
not do everything we might reasonably require in order to maximise the cash generated.

Rock grades equal to cut-off

The preceding description of the split of rock between destinations does not address
grades that are equal to the specified cut-off or cut-over. As with specifying costs to
include in the calculation, there is no commonly agreed position on including or
excluding material with grade equal to the cut-off.
In many cases this issue may be purely academic. For example, if grades are modelled
to more significant figures than the cut-offs, the number of blocks in the geological
model with an estimated grade the same as the cut-off is likely to be negligible and of
no practical concern. The author would, however, suggest that it is pointless to classify
material with the same grade as the ore–waste break-even grade as ore – it would then
be processed for no net benefit.
If the cut-off and modelled grades have the same number of significant figures, both
will probably have been calculated to a greater precision and then rounded. If the cut-
off has been rounded up, it might be logical to say that rock with the same grade as the
cut-off should be classified as ore. Conversely, if rounded down, you might specify the
cut-off to be classified as waste. There will still be some potential for misclassification if
the unrounded values for the rock grade and cut-off had been used. Again, this is likely
to be negligible relative to other uncertainties implicit in the grade estimation and break-
even calculation processes.
This rationale is applicable to the ore–waste cut-off, but the cut-over between the two
processes is a different issue. As noted, the 1D break-even calculation takes no account
of treatment plant capacities. Since the same net return is derived in either process,
material whose grade is exactly equal to the cut-over would be sent to the process with
spare capacity. If there is a significant amount of such material, it might be that some
is sent to one process and the rest to the other. Again, the simple 1D break-even model
does not account for all issues that need to be considered, and additional information
must be brought into the decision-making process.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 32
 CHAPTER 3 | Break-even Analysis

Costs in Ore–Waste Break-even Calculations

As stated, there are no standard definitions for names to be used for break-even grades
and the costs that are included in the calculations. Four examples are described below,
but these should not be seen as any particular recommendation for names or definitions;
rather, they illustrate the range of costs that might be included in a break-even calculation:
•• marginal break-even – costs to be covered = mining variable costs + milling variable
costs
•• mine operating break-even – costs to be covered = total mining costs + total milling costs
•• site operating break-even – costs to be covered = total mining costs + total milling costs
+ total site administration and services costs
•• full cost break-even – costs to be covered = total mining costs + total milling costs + total
site administration and services costs + head office charges + allowance for capital.
Many operations use at least two of these types of calculations for specifying cut-offs
in a given situation. Companies tend to use a cut-off similar to the first of these as an
operational cut-off for identifying marginal ore that can be treated opportunistically
when there is spare capacity in the ore handling and treatment processes. Similarly,
cut-offs similar to the third or fourth options are often used as the planning cut-off for
identifying ore and waste in the long-term mine plan; however, the costs used in the
break-even calculations are often in practice indicated with little understanding of what
is being achieved.
As previously stated, many operations use the terms break-even and cut-off
synonymously, so cut-off might be substituted for break-even in these definitions in
company documentation. The author prefers in these circumstances to use the compound
term break-even cut-off to indicate that the cut-off is to be specified by a break-even
calculation. It should be clearly understood that in this book break-even is a generic term
to describe the derivation methodology, with additional descriptors used to indicate the
costs included in the calculation.
As can be seen from the formulas and example calculations, costs to be applied in
break-even calculations ought to be the differential costs of classifying rock as ore
rather than waste. A similar rationale applies when dealing with cut-overs between
two processes. In the examples, costs that are common to both options (ore or waste,
Process 1 or Process 2) effectively cancel out, leaving only the cost differential between
the two options in the calculation.
What are the differential costs to be accounted for? If there is spare capacity in both
the ore and product streams, the extra costs of classifying a piece of rock as ore are only
the marginal variable costs of treating it as ore and dealing with the derived product.
If either the ore or product stream is at capacity, choosing to treat a piece of rock as ore
extends the operation’s life; the extra costs of classifying it as ore are the variable costs of
treating it as ore and handling the derived product, plus the fixed costs associated with
the extension of the mine life.
The following subsections discuss the variable and fixed costs that arise in different
parts of the operation and how they apply in break-even cut-off derivation.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 33
 CHAPTER 3 | Break-even Analysis

Open pit – in the mine

When a decision is made to mine a pit to a particular size, the variable costs of mining
(typically drill, blast, load and haul) are incurred for all rock removed from the pit,
irrespective of its classification as ore or waste. All rock must be transported to the point
where ore and waste streams diverge. The location of the rock in the pit – whether
from deep or shallow areas – is therefore not relevant to the ore–waste decision.1 Hence,
open pit mining variable costs should typically not go into an ore–waste break-even
calculation. This was seen in the more complex example, where the mining costs were
included in the costs of both ore and waste and cancel out in the resulting calculation.
The exception, which follows directly from this rationale, is any differential cost of
mining ore and waste. This could be the result of different:
•• grade control procedures
•• drill and blast patterns
•• mining fleets
•• haul distances.
The base mining variable cost should therefore be thought of as the cost of mining all
the rock as waste, with any incremental cost of mining ore considered an extra cost that
must be covered by the net revenue received from classing it as ore. The cost difference
may be positive or negative. For example, drilling patterns for ore may be closer-spaced
than for waste (generating a higher unit cost or positive incremental cost for ore), while
haulage distances to waste dumps may be longer than for ore (resulting in a lower unit
cost or negative incremental cost for ore).
Many readers may quite rightly raise an objection that, although the mining costs in
general should not be included in the ore–waste break-even cut-off calculation, they
nevertheless must be accounted for somehow or the operation might make a loss. In
practice, this could lead to including these costs in the break-even cut-off grade, even
though the preceding discussion indicates that they should not be. How is this resolved?
That the problem arises at all indicates that the simple 1D break-even model of cut-off
derivation is inadequate for generating a cut-off that can be used in long-term planning.
The issue will be addressed and resolved in the more complex cut-off derivation models
discussed in Chapters 4 and 6.

Underground – in the mine – excluded from break-evens

Costs to access new mining areas are typically independent of the amount of ore that is
delineated for stoping within the new exposed rock. These are lump sum costs (described
in Chapter 2) and are therefore not included in ore–waste break-evens. These excluded
costs would typically be for:
•• decline extensions downdip below existing stoping
•• access and infrastructure for new orebodies
•• footwall access development, to the extent that the strike length of the orebody does
not vary with cut-off
•• stope or production preparation that is not directly related to ore tonnage.

1. Location, and hence haulage cost, is of course very relevant to determining the best mining schedule, but that
is beyond the scope of the simple 1D process to specify a break-even grade.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 34
 CHAPTER 3 | Break-even Analysis

In the first two cases, there is no compulsion to access the mineralised material – the
remote area could be left unmined, or the downdip extension could remain unaccessed
and the mine closed instead. The decision to spend such project capital is justified by
the additional value of having the area available to be stoped. Similarly, developing the
footwall drive provides the choice of how much exposed strike length will be stoped,
without compulsion to mine any or all of it. The initial development in a stope block,
if it is simply providing the initial accesses and working faces such as sill drives in a
cut-and-fill operation or raises in ore in a narrow tabular reef, is independent of what is
subsequently stoped as ore from the rock made available.
Figure 3.5 illustrates how the strike length of the orebody and hence of access
development may be independent of cut-off. Although the strike length of the material
identified as ore may vary with cut-off, its spatial distribution is such that the accessed
strike length is the same at all cut-offs considered.

FIGURE 3.5
Strike length of development unaffected by cut-off.

Low grade
Medium grade
High grade

The break-even grade for an area that may have been accessed by expending project
capital does not change depending on whether the access has been developed or not;
however, it is not uncommon in underground mines to see two different break-even
cut-offs specified for an area, depending on whether it has been accessed or not. The
difference between the two cut-offs purportedly accounts for the cost of accessing the
area. This is fallacious reasoning. If the cut-off is going to be a particular value after
the area is accessed, that was always going to be the appropriate cut-off to apply,
whether the area was accessed or not at the time of the evaluation. The cost of the access
development will have been sunk at the time the cut-off decision for the area needs to
be finally made for mining purposes, so costs that will already be incurred at that time
should be excluded from cut-offs being derived ahead of production for such areas.
As with open pit mining, many readers will quite rightly claim that the mining costs
described here must be accounted for somehow or the operation might make a loss,
and responding to this may lead to a decision to include these costs in the break-
even cut-off grade, even though they should not be. Again, that the problem arises at
all indicates that the simple 1D break-even model of cut-off derivation is inadequate
for generating a reliable long-term planning cut-off. The issue will be addressed in
Chapters 4 and 6.

Underground – in the mine – included in break-evens

Activities generating costs that do vary with ore tonnages, and which would therefore
usually be included as ore-related costs in a break-even calculation, typically include:

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 35
 CHAPTER 3 | Break-even Analysis

•• stope development in ore

•• cross-cut stope access development in waste from strike access development
•• strike or footwall access development in waste, to the extent that the strike length of
the orebody varies with cut-off
•• stope drilling and blasting
•• ore haulage and hoisting
•• backfill
•• operating maintenance and sustaining capital costs for plant and equipment
associated with the above.
Figure 3.6 shows a deposit for which the strike length varies for all cut-offs considered.
Note that there may be a range of cut-offs over which the strike length does not change
and a range where it does. Figure 3.7 shows an example similar to Figure 3.5, where
the strike length of the mineralisation is unaffected by cut-off to the left of centre,
but becomes variable at higher cut-offs to the right of centre. The variability may be
continuous as illustrated in Figure 3.6, or have step changes as in Figure 3.7.
It is generally assumed that the listed mining costs are directly variable with ore
tonnage and the unit costs are unrelated to cut-off, but in reality they may vary with
cut-off. For example, if the width of the orebody varies with cut-off, the ore tonnes per
metre of stope development will be a function of cut-off. This introduces the same sort
of circularity into the break-even calculation as described with the total mining cost in
the pit ascribed to ore rather than rock. Unfortunately, there is no simple mechanism in
the underground situation to split this type of mining cost between ore and rock as there
was in the open pit case. In the pit, mining costs are attributed to rock rather than ore for

FIGURE 3.6
Strike length of development variable with cut-off.

Low grade
Medium grade
High grade

FIGURE 3.7
Strike length of development partly unaffected by and partly variable with cut-off.

Low grade
Medium grade
High grade

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 36
 CHAPTER 3 | Break-even Analysis

cut-off derivation, solving the problem; however, underground costs such as these are
genuinely ore-related, though not in a linear way, rather than rock-related.
Depending on how variable the affected costs are with cut-off, this may or may not
be a practical concern. If the variability is high, it may be necessary to perform iterative
calculations for the appropriate break-even grade. This necessitates evaluating the
effect of cut-off on the size and shape of the orebody, and takes the analysis beyond the
simple 1D break-even model. Each iteration of the calculation is one-dimensional, but
the process overall is not. Iterative calculations to converge on a consistent break-even
cut-off are described below.
In an open pit as noted, all material within the pit limits is to be mined and only
differential mining costs for ore would be included in break-evens. Where the rock
has come from in the pit is generally irrelevant. Apart from access development,
underground material that is mined as ore would have remained unmined had it not
been intentionally classified as ore. All mining costs downstream for those materials
classed as ore are therefore typically ore-related; therefore underground location may be
important. For example, if ore is hauled from the mine by truck, deeper zones will incur
higher costs than those closer to the surface. The break-even grade for deep ore will be
higher than for near-surface ore. But if all ore is dropped to a haulage level and hoisted
in a shaft, there may be no cost difference depending on depth. There could, however, be
additional costs associated with the lateral distance to be hauled – ore close to the shaft
may have a lower break-even than more-distant material.
Many companies address this issue by conducting a break-even analysis on individual
stopes, accounting for all the costs directly associated with each. Effectively, each stope
could have a different break-even grade, but there are some dangers with this, which
are covered in Chapter 7.

Processing – ore-related
Metallurgical activities generating costs varying with ore tonnages typically include:
•• comminution (crushing and grinding) wear protection and media
•• power (ore-related)
•• maintenance of plant – for which wear is primarily driven by the amount of ore
processed, such as
•• comminution plant
•• flotation cells/leach tanks
•• tailings thickeners, etc
•• tailings-related activities such as thickeners, pumps and dam raisings (adjusted to
account for any tailings removed from the stream for backfill)
•• sustaining capital costs for ore-related plant and equipment.
Power is often a major component of ore-related treatment costs, and the bulk of it will
typically be associated with grinding. It is often thought of as flowing from a specified
energy requirement per tonne of ore to achieve desired grind sizes and liberation of the
valuable mineral. There are two issues that have the potential to complicate this:
1. There may be ore types with different grinding energy requirements and hence
milling rates. The simple break-even cut-off model may be inadequate to handle the
impact of different blends of materials with various milling rates; however, it is at

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 37
 CHAPTER 3 | Break-even Analysis

least feasible to calculate break-even cut-offs for different ore types, but how to then
apply these is not so easy.
2. Mills will typically have a maximum-rated power draw, and may at times be operated
at full power regardless of the throughput rate. In the extreme case, power could be
thought of as a fixed rather than variable cost. It may be necessary to consider the
mill’s mode of operation and identify fixed and variable components of the power
usage and hence costs. It should also be noted that, if power draw is constant but
mill throughputs vary, in addition to the impact of different ore types noted, the
grind size, liberation, recovery and various treatment costs will become functions of
the feed rate. This implies that key inputs – costs and recoveries – are functions of
other parameters not accounted for in the break-even calculation. Break-evens can
therefore notionally be calculated for any particular set of values for these external
parameters. Yet using these break-evens to define ore and waste and subsequently
specifying different treatment rate and power use strategies will invalidate the whole
process, or alternatively require iterative calculations to converge on a break-even
cut-off that is consistent with all the other operating parameters.

Processing – product-related
Activities that generate costs varying with product quantities usually include:
•• reagents and reagent preparation
•• power related to product processing and handling
•• maintenance of plant for which wear is primarily driven by the amount of product
handled, such as concentrate pumping, filtration and thickening
•• sustaining capital costs for product-related plant and equipment.

Other product-related costs

Activities that generate costs that vary with product tonnages downstream of the
treatment plant typically include:
•• product transport
•• roasting, smelting and refining of product
•• royalties.
Royalties are usually classed as either:
•• Ad valorem – these are usually expressed as a percentage of the net price after
deduction of allowable costs, which are typically off-site realisation costs such as
treatment charges (TCs) and refining charges (RCs) – collectively referred to as
TC/RCs – and perhaps freight and transport from site to smelter. On-site operating
costs associated with product (concentrate pumping, filtration and the like) are often
not deductible for royalty purposes. Some jurisdictions charge ad valorem royalties on
gross value of product.
•• Tonnage-based – expressed as a simple dollar charge per tonne of product, or
sometimes per tonne of ore.
•• Profit-based – expressed as a percentage of net profit.
In some jurisdictions, royalty specifications may comprise a combination, or the rate
specifications to be used may be more complex than a unit cost or single percentage.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 38
 CHAPTER 3 | Break-even Analysis

Tailings costs were listed as ore-related costs. For precious metals operations, the
amount of product extracted from ore is relatively small, and for all practical purposes
the tailings tonnage is the same as the ore tonnage. In base metals operations producing
a concentrate, the product tonnage may be significant and hence the tailings tonnage is
much lower than ore tonnage. Break-even calculations are dealing only with costs for
rock, ore and product. Tailings costs will be applied to all ore tonnes, which include
concentrate tonnes. The simplest, most rigorous way to account for this in break-
even calculations may be to include a cost credit in the product-related costs for the
concentrate that is diverted from the ore stream and not incurring the tailings-related
costs applied to all ore tonnes.
Mine cost accounting systems often express product-related costs in the mill (and
frequently some or all of the downstream costs) as a unit cost per tonne of ore. Doing
this is similar to expressing rock-related costs in an open pit mine as a cost per tonne
of ore, as discussed earlier. This may be satisfactory for management control if head
grades and hence the amount of product produced per tonne of ore remain constant.
If cut-offs are under investigation, it needs to be recognised that the head grade is a
function of cut-off, and hence the product cost per tonne of ore is itself a function of cut-
off. In the same way that rock-related costs are related to rock, not ore, when dealing
with open pit mining costs, so product-related costs reported as costs per tonne of ore
must be extracted and reported separately as a cost per unit of product. This avoids the
circularity of calculating a break-even cut-off using a cost that is a function of cut-off.
Product-related costs per unit of product are effectively a reduction in the net price
received per unit of product. They should therefore be applied in the break-even
calculation to reduce the price per unit of product in the denominator of the break-even
calculation, not included as a cost per unit of ore in the numerator. The handling of
product-related costs will depend on the way that they behave. Where costs are incurred
because of the quantity of product (such as pumping, filtration and transport of product),
they should be deducted as a cost per unit from the price per unit. Where product-related
costs are driven by the value of the product (such as ad valorem royalties), they should be
applied as a proportional or absolute reduction in price. The net value used as the price
in the denominator of the break-even calculations, as shown earlier, will therefore be:

Net price = (price – TC/RCs) × (1 – royalty rate) – product-related costs

Note that this formula applies when royalties are ad valorem based on the price received
by the mine after allowing for realisation costs. Some jurisdictions may charge royalties
on the gross value of the product. If royalties are actually levied as a cost per tonne
of ore, they may be correctly included in the cost per tonne of ore in the break-even
calculation. Simple profit-based royalties may be excluded from break-even calculations,
as notionally at the break-even no profit is generated and hence no royalty is payable.2

2. Care should be taken to ensure that royalties behave in the way stated. In most cases this will not be an issue;
however, the author is aware of one large mining operation whose royalty was expressed as a dollars-per-tonne-
of-ore charge. The rate was varied by way of a complex formula which, when reduced to its simplest terms,
showed the royalty to be in fact ad valorem.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 39
 CHAPTER 3 | Break-even Analysis

Where royalties are derived by more complex mechanisms, it may be appropriate but
possibly not feasible to incorporate them in a simple break-even formula.3

Stockpiling and reclaim

Stockpiling and reclaim costs are costs that are associated only with ore. Depending on
how a company chooses to look at it, this could be included in company reports as a
mining cost (if it is an intermediate process before delivery to the mill for treatment, or
because the mining fleet is being used to handle it) or as a treatment plant cost (if mining
is deemed to finish at the point where the material is first dumped). Either way, there
will be an additional cost per tonne of ore to be included in the break-even calculation.
Several future economic scenarios may be assumed to calculate break-even cut-offs
to define the grade limits of various stockpiles for different categories of material, often
referred to by such names as high, medium and low grade and mineralised waste.

Fixed costs in break-evens

As indicated, fixed costs need to be included in break-evens when classifying a piece
of rock as ore will extend the life of the operation. This will occur if the ore or product
stream in the production process is operating at capacity. The additional ore therefore
extends the life of the operation and thereby incurs additional fixed costs for the extra
time – these costs must be paid for by the revenue obtained from the additional ore,
and hence included in the break-even calculations. A similar rationale applies when
stockpiles continue to be treated after the mining operation is complete.
A break-even planning cut-off will usually include fixed costs, as the implied purpose
is to identify the life-of-mine reserve: every tonne thus identified is essentially generating
extra mine life.
Which fixed costs should be included in such cases? The rationale suggests only fixed
costs that are incurred while the operation continues should be included. This implies
that all site administration costs and fixed costs in the mine (if still operating) and
treatment plant should be counted. In the case of a one-mine company, virtually all
corporate administration costs (except ongoing exploration) are incurred because the
mine is operating, and will cease when the mine ceases. But if the mine in question is
part of a corporate group, closing the one mine may or may not impact on the overhead
costs incurred in any one year or on the life of the group’s combined operations. Each
mine and group’s circumstances must be considered to determine whether overheads
will vary if more or less rock is classed as ore at the operation in question.
It should be noted that corporate overheads are frequently charged to the group’s
operations on the basis of ore tonnes, product quantities or product values. All of these
are arbitrary processes; they do not imply that the costs are actually driven by or related
to those parameters, for if they were they would be variable costs, not fixed. It is therefore
irrational to assume for cut-off derivation that if a mine producing part of a group’s total
ore tonnage is closed, that proportion of fixed costs will be eliminated. The proportion
by definition has to be less than that, and may even be zero. One needs to identify if, and

3. Whenever it is difficult to fit an important cost or revenue component into a break-even calculation, it indicates
that the simple break-even cut-off model is inadequate and a more complex model may be required.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 40
 CHAPTER 3 | Break-even Analysis

to what extent, corporate fixed overhead costs should be included in any one operation’s
break-even calculations.
The same reasoning applies if, for example, an open pit mine closes but treatment of
stockpiled ore continues. There is usually no simple common formula that relates site
administration fixed costs to rock tonnes, ore tonnes and product quantities in order to
identify the costs that will change if the life of the mine or the treatment facilities alters.

Capital costs in break-evens

Capital costs are significantly more expensive than operating costs if the effects of
taxation and the time value of money are taken into account. Consider a reserve of 10 Mt
being mined and treated at the rate of 1 Mt/a, an operating variable cost item of $1.00/t
and a capital cost item of $10 M, typically representing $1.00/t. Table 3.1 shows the
derivation of net cash flow and net present value (NPV)4 for each of these, assuming a
30 per cent tax rate and a ten per cent discount rate. Some values in the column for year 0
and in the bottom row have been rounded.
The columns for years 4 to 7 have been collapsed for simplicity. The operating cost of
$1.00/t generates a cash cost of $1 M per year, which in turn generates a tax credit of
$300 000 per year (assumed to be realised in the year of expenditure), and hence a net
cash cost of $700 000 per year. This can be seen to generate an after-tax NPV of some
$4.3 M.
The capital expenditure is assumed to occur at time zero, with straight line depreciation
for tax purposes over the ten-year life. This generates a depreciation charge of $1 M
per year, similar to the cash operating cost, and hence generates the same $300 000 per
year tax credit. The net cash flow for the capital item is therefore a $10 million outlay

TABLE 3.1
Effective costs of operating and capital expenditure.
Year 0 1 2 3 8 9 10
Ore tonnes 1000 kt 1000 kt 1000 kt 1000 kt 1000 kt 1000 kt
Operating costs $1.00/t $1 000 000 $1 000 000 $1 000 000 $1 000 000 $1 000 000 $1 000 000
Tax benefit 30% $300 000 $300 000 $300 000 $300 000 $300 000 $300 000
Operating costs after tax $700 000 $700 000 $700 000 $700 000 $700 000 $700 000
Net present value of 10% $4 300 000
net operating costs
Capital cost $10 000 000
Depreciation $1 000 000 $1 000 000 $1 000 000 $1 000 000 $1 000 000 $1 000 000
Tax benefit 30% $300 000 $300 000 $300 000 $300 000 $300 000 $300 000
Capital cost after tax $10 000 000 ($300 000) ($300 000) ($300 000) ($300 000) ($300 000) ($300 000)
Net present value of 10% $8 200 000
net capital cost
Equivalent annual value of net capital cost $1 327 000 $1 327 000 $1 327 000 $1 327 000 $1 327 000 $1 327 000

4. As it is beyond the scope of this book to discuss discounted cash flow techniques, readers unfamiliar with
concepts such as discount rate, net present value and equivalent annual value should refer to standard texts on
project evaluation and discounted cash flow techniques.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 41
 CHAPTER 3 | Break-even Analysis

at time zero, followed by what is effectively an annual cash inflow of $300 000. The
NPV of this cash flow pattern is $8.3 M or 1.9 times – that is, nearly double – the net
cost of the operating cost. The last row of Table 3.1 shows the after-tax NPV distributed
over the mine life as an equivalent annual value of some $1.33 M per year. The table
indicates that the net after-tax cost of what is nominally a $1.00/t operating cost is
$0.70/t; however, the net after-tax cost of a $1.00/t capital outlay is $1.33/t when the
net spending (considering the time value of money) is expressed at the same time as the
physical activities, similar to operating costs.
The ratio between capital and operating costs will depend on the tax rates, discount
rates and inflation rates. Inflation is not shown in Table 3.1 but its effect is to erode the
value of the depreciation charge over time, and hence increase further the net after-
tax cost of capital expenditure relative to operating costs. Is it therefore sufficient to
merely include capital in break-evens at the nominal cost (capital cost divided by the
tonnage to which it relates) or should a suitable factor be calculated and applied to
account for the extra cost of capital costs relative to operating costs when considering
tax and the time value of money?

BREAK-EVENS AT VARIOUS POINTS ALONG THE PRODUCTION PROCESS

In an underground operation, it is not uncommon to find that mines will have a number
of break-even cut-offs calculated for material at different points along the production
process. This is logical because there are a number of sequential ore-related activities
with underground mining, so we can reassess the ore–waste cut-off decision at each
point. It is only when there is no potential to halt the movement of rock through the
system that the irrevocable decision as to whether the rock is ore or waste is made.
At each point in the process, the costs associated with classifying the rock as ore must
be compared with classifying it as waste. Typically the cost of waste will be zero at any
point – we simply walk away at whatever stage that happens to be; however, there may
still be some costs, such as making safe or rehabilitating the partly prepared material.
Working backwards, the ore-related costs that would go into these break-evens
(depending on how advanced the production preparation is) would typically include:
•• metallurgical treatment costs versus the cost of discarding as waste material that
could be processed as ore
•• backfill costs if fill is used
•• haulage/hoisting costs and primary crushing costs in the mine, if relevant
•• stope loading
•• blasting
•• stope drilling
•• stope development
•• waste access development, to the extent that these may be included in an ore–waste
break-even calculation.
For an open pit operation, there will be one break-even cut-off-defining ore to be
treated at any time, since an open pit operation lacks the sequence of ore-related
production steps found in underground operations in which the ore–waste decision
can be re-evaluated. There might be different break-even cut-offs for freshly mined and
stockpiled material to be fed to the treatment plant, since there may be different costs

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 42
 CHAPTER 3 | Break-even Analysis

for material treated from each source. If nothing else, direct feed is notionally already
in trucks ready for tipping into the process stream, whereas stockpiled material must
incur the additional costs of reclaim. There will also typically be stockpiles of potential
ore between the mining and treatment operations, with different cut-offs applied.
These may be constructed so that declining grade stocks provide potential feed to the
plant if conditions change.

ITERATIVE BREAK-EVEN CONVERGENCE AND DIVERGENCE

It has been noted that some unit costs used in a break-even calculation are themselves
functions of cut-off. This may manifest itself as a cost structure – a set of unit costs for
activities comprising the planned underground production operations – that is different
to the cost structure used to derive the break-even cut-off to generate that plan. This
results in a break-even grade different from the initial break-even cut-off used. There is
then an understandable desire to apply this new break-even grade as a revised cut-off
and rework the reserve and mine plan. The same outcome may recur, and hence the
calculation and mine plan revision may need to be repeated. The implicit aim of this
iterative process will be to converge to the break-even that results in costs that generate
the same break-even grade as the break-even cut-off used.
Figure 3.8 shows how this process may operate, and how there is no guarantee that it
will actually converge as desired. The axes on the plot are the specified cut-off, either
arbitrary or from a previous iteration, and the break-even grade resulting from using the
specified cut-off. The curved line plots the break-evens derived from the orebody costs
defined by each of the cut-offs on the horizontal axis. The dashed straight diagonal line
refers to where the values on the axes are equal; that is, where the break-even grade for
an orebody is the same as the cut-off applied to generate that orebody.
The break-even versus cut-off curve must start and end above the reference line. At a
cut-off of zero, which implies that all the mineralised rock is classed as ore, there will be

FIGURE 3.8
Iterative break-even determinations converging or diverging.
8

6
Calculated Break-even at
Specified Cut-off

B
4

2
A

0
0 1 2 3 4 5 6 7 8
Specified Cut-off Grade

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 43
 CHAPTER 3 | Break-even Analysis

a cost per tonne for the orebody defined and the resulting break-even must be greater
than zero. At high cut-offs the orebody becomes small, but there will be costs that are not
purely variable. As the cut-off used increases, the break-even will increase rapidly and
at the theoretical limit where increasing the cut-off reduces the reserve tonnage to zero,
the break-even will become infinite.
Depending on the nature of the mineralisation (which is strictly outside the bounds
of the 1D break-even model of cut-off) the break-even versus cut-off plot may cross the
reference line. If so, it must cross in two places as shown by the points marked ‘A’ and
‘B’ in Figure 3.8. It can be seen now that there are three situations:
1. At specified cut-offs less than at Point A, the break-even versus cut-off curve is above
the reference line but converging to Point A with increasing cut-off. If an initial cut-off
is specified in this range (the black dot at approximately 0.25 on the horizontal axis)
the break-even resulting from the orebody would be approximately 1.0, indicated by
the vertical dashed arrow. If it is now reassessed using that break-even as the cut-off
(shown by the dashed horizontal arrow) and then vertically up again, a new break-
even of approximately 1.2 results. If this process continues, the cut-offs and break-
evens will increase and converge to Point A where the cut-off and break-even are the
same, as is to be expected when any cost circularities used to derive the break-even
have been resolved.
2. At cut-offs between those at Point A and Point B, the break-even versus cut-off curve
is below the reference line. If an initial cut-off is specified in this range (the black dot at
around 2.8 on the horizontal cut-off axis), the break-even resulting from the orebody
would be approximately 2.1, indicated by the vertical dashed arrow. Continuing the
iterative process with the orebody defined by using the previous break-even as the
cut-off, the successive cut-offs progressively reduce to converge to Point A.
3. At cut-offs greater than at Point B, the break-even versus cut-off curve is above the
reference line and diverging from it with increasing cut-offs. If an initial cut-off is
specified in this range, (the black dot at approximately 4.4 on the horizontal cut-off
axis), the break-even resulting from the orebody would be roughly 6.2 (the vertical
dashed arrow). Continuing the iterative process, the successive cut-offs progressively
increase but diverge instead of converging, and eventually reach a value that reduces
the tonnage of ore to zero.
In the special case where the break-even versus cut-off plot touches the reference line
tangentially at one point, Points A and B become coincident, and the second situation
described for cut-offs starting between Points A and B ceases to exist. If the break-even
versus cut-off curve does not cross the reference line, the third situation applies at all
initial cut-off values; the iterative break-even process will generate cut-offs that increase
until the ore tonnage is reduced to zero.
Hence, there is no guarantee that the iterative break-even process is able to generate a
consistent break-even cut-off. If it is feasible, whether the convergence is achieved or not
will depend on the initial cut-off used in the process.
Both Point A and Point B generate consistent break-even cut-offs, but if convergence
does occur, it will generate the lower of the two. The higher cut-off at Point B is equally
valid, but will only be found if it is, by coincidence, the cut-off used initially.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 44
 CHAPTER 3 | Break-even Analysis

PROBLEMS WITH BREAK-EVEN CUT-OFFS

Philosophically, the author’s biggest objection to the use of break-even cut-offs is that
it surrenders responsibility for specifying the corporate policy to the vagaries of the
market place. As prices fluctuate, so too do break-even cut-offs. The company does not
actively control what is ore and what is not, but passively accepts whatever the price
fluctuations deliver. There are, however, important practical ramifications.

Achieving the corporate goals

As has been indicated, there are various cost components that are not included in an
ore–waste break-even cut-off calculation; however, these costs must be covered in some
way or the operation potentially makes a loss.
Applying a break-even cut-off simply identifies the ore tonnage with a grade above
that cut-off. Yet it tells us nothing about the overall grade of that tonnage. There is no
guarantee that the grade above the break-even that is required to cover the costs excluded
from the break-even – and provide some desired profit margin – actually exists.
By way of analogy, let us consider the difference between a coral atoll and a volcanic
island. Both grow from the sea floor but by different mechanisms. The coral atoll
essentially stops growing when it reaches sea level, and sand can then accumulate to
a relatively small height above sea level. The volcanic island, however, can continue to
grow vertically and attain a great height after reaching the surface of the sea.
The sea level is analogous to the ‘every tonne pays for itself’ break-even cut-off, and
simply defines the areal extent of the dry land. The coral atoll and the volcanic island
might have the same plan area, which compares to the tonnage above cut-off. But it is
the average elevation that determines how affected the island will be if the sea level
rises – the average elevation is akin to the ore grade, which is what generates revenue.
If the sea level rises, similar to adding the costs excluded in the break-even cut-off to
achieve a minimum profit, the coral atoll may be completely flooded; although there are
tonnes that ‘pay for themselves’, these tonnes do not cover the additional costs to ensure
that a profit is made. The volcanic island, however, rises significantly above this cost
requirement, and profitability may be assured under all realistic economic conditions.
It is important to note that achieving a level of profitability is totally different from
ensuring that every tonne pays for itself. It is inappropriate to simply add these
additional costs and required profit margin to the cost used to generate a break-even
cut-off and then apply that higher cut-off to redefine what is ore and what is not. It
is the average grade of the ore that pays for all costs and generates a profit; not every
tonne mined covers those additional costs and profit margins.
This issue will be addressed in the more complex cut-off derivation models discussed
in Chapters 4 and 6.

Marginal versus full cost break-evens

In the discussion of costs to include in break-evens, only marginal costs need to be
included in the calculations if both the ore and product stream are operating below
capacity; however, fixed costs must be included if either of these streams is at capacity.
System capacities is outside the scope of the 1D break-even derivation, so the need to
consider that at all indicates that the break-even model of cut-offs has limitations.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 45
 CHAPTER 3 | Break-even Analysis

Let us accept that consideration of capacities does not ‘break’ the break-even cut-
off model, but simply sets conditions for including or excluding certain costs. It could
be argued that we can thereby generate satisfactory break-even cut-offs for the two
situations and apply the appropriate one for the state of the operation at any point in
time; however, the issue is not necessarily resolved.
If it happens to be that the tonnage available between the marginal and full-cost break-
even cut-offs (referred to as the marginal tonnage) is insufficient for filling the capacity
of either the ore or product circuits, the marginal break-even cut-off will be appropriate.
Yet this will often not be the case – there will frequently be more marginal tonnage
available than is needed to fill the capacities of the ore or product circuits. Should we
therefore treat all the available marginal tonnage if there is some spare capacity in the
ore and product circuits? The answer of course is no; if we did, the logical extension
would then be to use a marginal break-even for all ore definition. We know – both from
previous discussion and general industry-wide common sense – that using a marginal
break-even as the overall cut-off is asking for disaster.
What is required, of course, is utilising the highest grade marginal material to fill the
available spare capacity. The 1D break-even cut-off model is now definitely lacking: an
analysis will require considering both the grade distribution and production system
capacities. What this can lead to in practice, though, is a decision that if all the marginal
grade material cannot be treated, none of it should be. Both extremes are suboptimal.
Nevertheless, when the understanding of cut-off is that it is the break-even and the
marginal cost break-even does not work, then the only other alternative is the full cost
break-even, and there is no acceptable position between the two.
The author is aware of an operation where a full-cost break-even cut-off was calculated
but, when applied, the nature of the resulting orebody was such that a production rate
to fill the mill could not be maintained. The cut-off was then reduced to increase the
size of the orebody so that the mill could be operated at its capacity. In the absence of a
more comprehensive model of cut-off, company staff were uneasy: ‘The cut-off should
be X (the break-even), but we’re actually using Y. It’s wrong, but we have to do it!’ The
discussions of Lane’s methodology in Chapter 5 and strategy optimisation in Chapter 6
will indicate that they were intuitively correct, but without the understanding of more
complex cut-off models, were unable to recognise the optimum strategy as being correct.
It is also worth noting here that marginal actually means extra or differential costs. In
practice, marginal costs are often assumed to be the same as variable costs, with the terms
marginal and variable used interchangeably. This will be true when there is capacity in
the ore and product streams, and additional material classed as ore need only cover
its direct variable costs to add value; however, as noted, if either stream operates at
capacity, additional material extends the life of the operation, thereby incurring more
fixed costs. The same applies when stockpiles only are being treated after mining has
finished. In these cases, marginal costs include both the variable and fixed costs.

Rising and falling prices

When prices are rising, the break-even cut-off reduces and proportionally more of the
rock is classified as ore. In an underground mine, mineralisation that was previously
waste in areas already accessed by development becomes ore, so the pressure to
extend development into new areas is reduced. In addition, the ore tonnes per metre

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 46
 CHAPTER 3 | Break-even Analysis

of development will increase with reducing cut-off, so the pressure on development is

relaxed further. Similarly, in an open pit mine more of the exposed mineralisation will
be ore and less will be waste. The stripping ratio reduces and a lower mining rate of
total material is needed to maintain the same tonnage of ore though the mill. If prices
remain high for a period of time, a new scheduling balance of access development, stope
development and production of waste and ore mining will be established. In a desire
to minimise spending, access development and overburden removal capital will not be
spent until it is absolutely required.
The average grade of the ore produced must, however, decline. One possible outcome
of this lowering of grade is that product targets are no longer met, even with the ore
stream operating at capacity. Significant capital may be needed to increase ore treatment
rates through the metallurgical treatment plant so that product targets can again be met.
All will be well until prices start to fall, and the larger or faster the fall, the worse the
potential problem. Lower-grade material that was ore at low cut-offs will no longer be
ore at the higher break-even cut-offs with lesser prices. The now sub-cut-off material
will be removed from the production plan, resulting in a significant reduction in stocks
of developed or exposed ore. Production will be rescheduled to maintain the planned
rate from remaining sources above cut-off in the short-term, but there is potential for a
major shortfall in available ore down the track. This results from the reduction in the
rate of access development into new areas or overburden removal when prices were
higher and cut-offs were lower, which will have restricted the amount of ore available
above the new higher cut-offs in the medium term. At this stage, if previously developed
but now sub-cut-off material is still available, it may have to be brought back into the
production schedule to maintain planned ore production rates, but at lower grades. If it
is no longer accessible, production targets may not be met.
It is hoped that such reintroduced low-grade material is at least above a marginal
break-even cut-off. The author is aware of cases where the value of the low-grade
material was below its direct variable costs of production, so additional production was
actually losing money. If this were necessary to maintain the production rate above some
minimum level necessary for the mill to operate, so be it, though in those circumstances
periodic mill shutdowns and batch processing might have been a better solution. If,
however, the additional material was simply meeting an arbitrary ore tonnage target,
the potential cost of value-destroying management targets would have to be addressed.
Either way, a development or stripping rate crisis is now manifest. When prices were
rising and cut-offs falling, the ore tonnes per metre of development or per tonne or cubic
metre mined increased; however, with falling prices and rising cut-offs, the ratio will
reduce. To maintain the planned ore production rate at the higher cut-off, the steady
ongoing development or stripping rate must increase to a new level. On top of this,
there must be a campaign of development or stripping to re-establish ore stocks so that
production can be maintained. This will be happening at a time when prices are falling
and cash availability is reducing. The pressure will be to reduce these costs at a time
when the critical operational need is to increase them.
Simplistically, it is usually not possible to change the cut-off in an established operation
without changing something else, such as the development or stripping rate as indicated.
When prices are rising and break-even cut-offs are falling, the ‘new’ low-grade ore
is gradually worked into the production schedule and the overall development or

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 47
 CHAPTER 3 | Break-even Analysis

stripping rate is adjusted downwards without any perceptible problems. The necessary
changes are made but they may be unrecognised, or, if acknowledged, applauded as a
cost reduction. The increase in reserves is also viewed favourably; however, when prices
fall and the break-even cut-off is increased, potentially major changes to the schedule
need to be made almost immediately. The price falls that are the immediate cause of the
problem may also result in a reluctance or inability to spend the extra needed – in the
short-term, for re-establishing the developed or exposed ore stocks, and in the longer-
term, for maintaining an ongoing rate higher than before the cut-off increase. Production
shortfalls may therefore occur for several years until stocks are slowly rebuilt and the
appropriate balance between development or stripping and production is re-established.
Many of the reported problems in the mining industry in recent times, when prices
for many commodities have reduced significantly after a period of substantial rises, can
be attributed to this type of scenario. This is driven directly by a lack of appreciation
by industry decision-makers and market analysts that simple 1D break-even cut-offs,
when used to develop long-term mining strategies, do not result in maximised cash
generation. They can actually drive a mine into a perilous financial position when prices
turn down.5 This is discussed in more detail in Chapter 12.

SUMMARY
The implicit goal of a break-even cut-off is, in its most general terms, to ensure that every
tonne classified as ore ‘pays for itself’ at the time it is treated. There are no industry-
standard definitions of what this means or what costs go into a break-even calculation.
Most companies using break-even cut-offs specify the costs that are to be included in
perhaps several different break-even-style cut-offs.
An alternative definition of a break-even cut-off ensures that every tonne is dealt with
by the course of action that generates most benefit. As a general principle, the costs
used should be the differences associated with the two potential courses of action. If the
decision is between classifying a piece of rock as ore or waste, it is only the differential
costs of ore that need to be considered.

5. For many years, while prices were rising and break-even grades were falling, industry analysts and mining
company executives seemed happy to accept the increased reserves and lowered operating costs (though
costs per unit of metal did not necessarily fall significantly due to large tonnages of lower-grade material being
brought into reserves). In more recent times, reports of discussions at a number of mining financial conferences
suggest that analysts are now chastising company executives for not delivering more cash when prices
were high, charging them with ‘deciding to reduce head grades’. With prices falling and what was previously
considered ‘ore’ no longer being ore, analysts are complaining about production shortfalls. For those involved
in strategy optimisation and long-term mine planning, all of this was predictable. It has happened in the past
and will continue to happen again in the future as business cycles cause prices to rise and fall, unless things
change. As long as the industry as a whole – mining executives, analysts who comment upon their decisions and
technical staff who develop the plans that seek to implement corporate strategies – continues to believe that
cut-off and break-even are synonymous, the situation will stay the same.
The author suggests that company executives did not ‘decide to lower head grades’, but passively accepted
that cut-offs and hence head grades had to fall because the price was rising and cut-off is assumed to be a
break-even grade. The key issue is that break-even cut-offs are not focused on maximising cash returns. Ounces
and tonnes in reserves, the measures that managers and analysts appear to have focused on in the past, are
negatively correlated with cash generation. This critical issue with break-even cut-offs is discussed further in
Chapter 14.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 48
 CHAPTER 3 | Break-even Analysis

In the case of an open pit, all planned material is mined from the pit regardless of
whether it is ore or waste. The differential costs will therefore be the variable costs of
dealing with ore and product through the treatment circuits of the metallurgical plant
and downstream to the point of sale. Mining costs will typically not come into this
calculation, except where there are different mining costs associated with ore and waste
– perhaps as a result of different drill and blast patterns or haulage distances.
Underground, virtually all activities are associated with ore. The main exception is
major waste access development, which is exposing new mineralisation within which
ore to be produced can be delineated. This major access development typically does
not depend on the amount of mineralisation exposed. It is a function of the lateral or
vertical distance to be covered to access a remote zone of mineralisation or to expose
the next sublevel block. Because the cost of this access does not depend on the amount
of ore delineated, it does not change according to how much of the mineralisation is
classified as ore rather than waste, and is therefore not a differential cost for a break-
even calculation. Some access stope preparation development, though associated with
and maybe even mined in ore, is also not directly related to the tonnage of ore to be
extracted from a block and should similarly be excluded from a break-even calculation.
If there is spare capacity in both the ore and product circuits in the plant, material may
be included in the production plan if its grade is above the break-even cut-off derived
using downstream ore and product-related variable costs; however, not all such material
can necessarily be added to the ore stream. Only the highest grade material that fills
the capacity shortfall should be mined and treated, but the simple break-even cut-off
mechanism does not facilitate this – it is typically an all-or-nothing scenario analysis.
There is no guarantee that the average grade of an orebody defined by a break-even
cut-off will deliver a profit. Generating a profit is completely different from merely
ensuring that every tonne pays for itself. The level of profitability comes from the
average grade of material above cut-off. A simple break-even cut-off takes no account
of the grade distribution of the mineralisation, nor the various stages of the mining and
treatment process; however, many of the problems identified with break-even cut-offs
have indicated that these must be considered. Unless they are specifically brought into
the cut-off model, attempts to account for them will be arbitrary and suboptimal.
Using a break-even grade as a cut-off effectively surrenders responsibility for specifying
what is ore and what is waste to uncontrollable changes in the market. In principle, a
mining company should consciously decide what is ore and what is waste in order to
achieve its corporate goals. Break-even cut-offs are incompatible with this.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 49
 HOME

Chapter 4

Mortimer’s Definition –
Accounting for Geology

Mortimer’s Definition Defined

In a paper describing grade control practices in South African gold mines in the late
1940s, Mortimer (1950) established an important though little recognised advance in
cut-off theory. He identified that what was classified as ore had to meet two criteria:
1. the average grade of rock must provide a certain minimum profit per tonne treated
2. the lowest grade of rock must pay for itself.
In this book, these two goals comprise Mortimer’s Definition. Break-even cut-offs,
discussed in the previous chapter, have an implicit goal that ‘every tonne pays for itself’,
but as noted this does not guarantee achieving a level of profitability. Mortimer, however,
in describing what is to be classed as ore, has stated a goal that is more closely aligned
with the company’s true goals: that some specified profit target is to be achieved. Yet he
recognises that ensuring that every tonne pays for itself is also important, stating it as the
second part of his definition. What ore is – the cut-off decision – has been defined in terms of
profitability. Mortimer clearly states the goals upfront and these therefore drive the method
and calculations by which the cut-off will be derived. Mortimer’s Definition may be thought
of as a two-dimensional (2D) cut-off model, because as well as making use of costs and
prices, like break-evens, it also accounts for the geology, in particular, the grade distribution.

Calculations for Mortimer’s Definition

The grades associated with each part of Mortimer’s Definition may be determined by
break-even calculations, but since each part comprises a distinct goal, different ‘costs’
will be associated with each.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 51
 CHAPTER 4 | Mortimer’s Definition

To ensure that the average grade of rock provides a minimum profit per tonne, the
‘costs’ to be covered in the break-even calculation need to include:
•• total mining costs
•• total milling costs
•• total site administration and services costs
•• depreciation and amortisation
•• head office charges
•• interest on debt
•• the required profit margin.
Note that while all or most of the rock-related mining costs would be excluded from
an equal benefits ore–waste break-even cut-off calculation, they need to be included
in the calculation of a profit, and so they are legitimately included in this break-even
calculation. The purpose of any break-even calculation must be clearly specified to
ensure that the appropriate costs are included. The calculation for the first part of
Mortimer’s Definition is not defining a cut-off to distinguish between ore and waste; it
is identifying the average ore grade that delivers the required profit. This will be used to
derive the cut-off that distinguishes between ore and waste, but is not itself that cut-off.
To ensure that the lowest grade of rock pays for itself – the second part of Mortimer’s
Definition – the costs to be covered in the break-even calculation should include the
direct variable costs associated with classifying a piece of rock as ore. As previously
discussed, this typically includes:
•• the differential costs associated with mining ore, relative to the costs of waste
•• variable costs associated with ore and product, including sustaining capital
•• fixed costs, depending on what meaning the company applies to ‘pays for itself’.
This second calculation in Mortimer’s Definition is an ore–waste break-even cut-off,
and it effectively sets the lower bound of possible cut-offs. It cannot be any lower or the
material classified as ore will generate a loss. It is perhaps an equal benefits differential
costs break-even calculation as described in the previous chapter.
Both parts of the Definition lead to break-even calculations, and the first part – with
more cost components – will have higher costs and therefore a higher break-even grade.
Many readers may be questioning why we have both parts of the definition and not just
the first. They may also be wondering how geology comes into it to make it a 2D cut-off
model.
To answer the first question, we note that the second break-even grade of Mortimer’s
Definition (that every tonne pays for itself) sets the minimum criterion for a block of
rock on its own to be included in the inventory of material that could become ore. It
could be seen as a boundary cut-off, which is used to define the limits of material above
that grade without taking any cognisance of the grade distribution of the enclosed
material with higher grades. It suffers from all the problems identified for 1D break-
even cut-offs; however, the break-even grade delivering the first part of Mortimer’s
Definition can be thought of as a volume cut-off. It is the minimum average grade of
all the material classified as ore for a planning time frame, not just a single block or
of only one stope, and is the minimum head grade required to deliver the specified
minimum profit.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 52
 CHAPTER 4 | Mortimer’s Definition

Upon reflection we can deduce qualitatively that, if the grade distribution has a lot of
high-grade material above the every-tonne-pays-for-itself break-even grade, the average
grade of all that material could easily be above the minimum profit break-even grade.
Conversely, if there is not a lot of material higher than the minimum profit grade, the
average grade of all the material above the every-tonne-pays-for-itself break-even grade
may be below the minimum profit grade. For the average grade of what is finally classed
as ore above the minimum profit grade, we must drop some of the lower-grade material
out of the ore inventory. Although it pays for itself on a standalone basis, it drags the
average grade below what is required.
This leads to the second question regarding geology. As mentioned in Chapter 2, in
situations where both boundary and volume cut-offs are used, the two cut-off values are
typically the same; however, when applying Mortimer’s Definition, the volume cut-off
(the minimum required head grade) will usually be higher than the boundary cut-off
(ensuring that every tonne pays for itself). If we were to generate an initial ore inventory
at that boundary cut-off and then apply the higher volume cut-off to the stopes or mining
blocks in that inventory, some would have an average grade that was too low and would
be excluded. If we were to examine those, we would find that some contained higher-
grade portions above the volume cut-off that ought to be in the final inventory. If a
higher boundary cut-off were applied to eliminate the lower-grade material, the higher-
grade portions would then comprise a revised stope or mining block with a grade above
the volume cut-off.
If we similarly examined stopes or mining blocks in the initial inventory that were
above the volume cut-off, we would also find material with a grade between the initial
boundary cut-off and the higher boundary cut-off applied to the lower grade block to
get its average grade above the required volume cut-off. If we took no further action,
this lower-grade material would be included in the final inventory merely because it
happened to be adjacent to enough higher-grade material. Also, material of the same
grade has been excluded from the final inventory merely because it was not adjacent to
higher-grade material.
But this is illogical. We do not, in principle, mine low-grade material in some areas
simply because there is sufficient high-grade material nearby to raise the overall average
grade. If that were to be done, then taking it to its logical extreme we would dilute all
high-grade material down to some desired head grade, even using waste if necessary.
Rather, there should be a balance between head grade and boundary cut-off so that the
lowest-grade material mined from all areas is the same, regardless of where zones of
higher-grade material might be. For consistency, therefore, the same higher boundary
cut-off should be applied to the whole resource, not just to low head-grade blocks.
The geology must be examined to identify what boundary cut-off will deliver the
required head grade. This can be done by examining either a table or a plot of tonnes
and grades versus cut-off. The following section will illustrate this graphically.

Mortimer’s Definition Illustrated

Figure 4.1 shows a typical plot of tonnage and grade versus cut-off curves. The horizontal
axis is the boundary cut-off, expressed generically in grade units per tonne. The vertical
axis on the left shows the ore tonnage above cut-off, represented by the line reducing
from a total of 8 Mt at a cut-off of zero (the total tonnage of rock) to almost zero at a

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 53
 CHAPter 4 | Mortimer’s Definition

FIGURE 4.1
Tonnage–grade curves – every tonne pays for itself.
12 12

Grade of Material above Cut-off

Tonnage above Cut-off (Mt)
10 10

8 8
"Pay" Grade
6 6

4 4

2 2

0 0
0 1 2 3 4 5 6 7 8 9 10
Cut-off Grade

Tonnes above Cut-off Grade above Cut-off

cut-off of 10 gu/t. The grade of the material (also in generic grade units per tonne) is
reported on the right vertical axis, and increases from approximately 2.7 at zero cut-off
in the total rock tonnage to 12 in the small tonnage above a cut-off of 10. This is indicated
by the line rising from left to right.
Let us start with the second part of Mortimer’s Definition – that every tonne mined
as ore must pay for itself – and suppose that we have completed the second break-even
calculation suggested and identified that the cut-off required is 3.0 gu/t. Moving up to
the tonnage and grade lines, and then left and right to the tonnage and grade axes, we
see in Figure 4.1 that this cut-off will deliver some 2.5 Mt of ore at an average grade of
6.0 gu/t. So far we have not done anything more than a simple ore–waste break-even
cut-off calculation and identified the resulting ore tonnes and grades.
Let us assume that we have done the first break-even calculation suggested to
determine the grade that delivers the required profit margin, and that grade is 8.0 gu/t.
It is not necessary that every tonne of ore delivers this profit, just that the average grade
does. Figure 4.2 shows how this break-even is applied using the grade versus cut-off
curve to identify the cut-off grade that delivers this head grade. Starting at 8.0 on the
grade (right vertical) axis, we move left until we intersect the grade versus cut-off curve.
From here we travel down to the cut-off axis where we note that, in this example, the
required boundary cut-off is 6.1 gu/t.
The two parts of Mortimer’s Definition have now provided these possible cut-off values:
6.1 and 3.0. To ensure that we meet the conditions of both parts of Mortimer’s Definition,
we must select the larger of these two: 6.1. This ensures that we simultaneously deliver
the required profit margin while every tonne is at least paying for itself.
The process has been illustrated graphically here, but it can be done equally well
by inspecting the tabulated tonnage and grade versus cut-off data, with interpolation
between data records if appropriate.
In the scenario described, both the ‘minimum profit’ grade and the cut-off that
delivers it are higher than the every-tonne-pays-for-itself break-even cut-off grade, but
this need not be the case. Suppose that the scenario is for a relatively new mine with

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 54
 CHAPter 4 | Mortimer’s Definition

FIGURE 4.2
Tonnage–grade curves – average grade provides a specified profit.
12 12

Grade of Material above Cut-off

Tonnage above Cut-off (Mt)
10 10
"Minimum Profit" Grade
8 8
"Pay" Grade
6 6

4 4

2 2

0 0
0 1 2 3 4 5 6 7 8 9 10
Cut-off Grade

Tonnes above Cut-off Grade above Cut-off

high depreciation and amortisation charges, and being the only mine operated by the
group must bear all head office charges. Suppose also that some years later, the mine
in question is nearing the end of its life. The capital has been fully depreciated and the
company now has a number of operations to share head office costs, and does not require
this operation to contribute a large profit margin to justify its continuation. Applying the
same minimum profit break-even calculations as before, and assuming prices and site
costs have not changed, the reduction in three major cost components in the break-even
calculation results in a required average mined ore grade of 5.0 gu/t. Figure 4.3 shows
the new situation. It can be seen now that the cut-off that delivers a head grade of 5 is of
the order of 2.0, which is less than the every-tonne-pays-for-itself cut-off of 3.0. Again,
the larger of the two cut-offs – 3.0 – should be selected.

FIGURE 4.3
Tonnage–grade curves – average grade provides a specified profit (reduced costs).
12 12
Grade of Material above Cut-off
Tonnage above Cut-off (Mt)

10 10

8 8
"Pay" Grade
6 6

4 "Minimum Profit" Grade 4

2 2

0 0
0 1 2 3 4 5 6 7 8 9 10
Cut-off Grade

Tonnes above Cut-off Grade above Cut-off

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 55
 CHAPTER 4 | Mortimer’s Definition

It is interesting to note that the cut-off that delivers the required profit via the grade
versus cut-off curve will vary depending on how much high-grade material is in the
volume of rock under consideration. Suppose that the small tonnage above a cut-off
of 10 grade units has a grade of 15 rather than 12 as illustrated in the grade versus
cut-off curves in the figures in this chapter. Although the distribution of tonnages and
grades below a cut-off of 10 may be the same as before, the higher grade in the high-
grade portion will result in the average grade at all lower cut-offs being higher than
shown in the figures. The grade versus cut-off curve will be higher, and the cut-off that
will deliver the required average grade of 8.0 will therefore be lower. Conversely, if the
grade of the material above a cut-off of 10 is lower than 12, all average grades will be
lower than shown. The grade versus cut-off curve will be lower, and a higher cut-off will
be needed to deliver the same minimum profit head grade.
Depending on the distribution of high-grades in the deposit and the relativities of the
two Mortimer cut-offs, it is possible that the mineral resource might never deliver the
desired level of profitability, even though a mineable orebody is defined by the every-
tonne-pays-for-itself cut-off, as was noted when discussing the analogy of the coral atoll
and the volcanic island in the previous chapter.

Mortimer’s Definition in Practice

In practice, volume cut-offs (if used) will typically be applied to individual stopes or
relatively small contiguous regions defined by a boundary cut-off during the initial
generation of a mining reserve; however, the minimum profit grade required by
Mortimer’s Definition needs to be delivered by the blend of material from multiple
areas to be processed within a particular period of time. The minimum profit grade
and hence volume cut-off applicable to a Mortimer analysis may vary over time, even
if the every-tonne-pays-for-itself break-even cut-off does not change. This implies that
the tonnage–grade relationships developed for a Mortimer-style analysis should be for
an inventory of mining units identified by boundary cut-offs only and not prefiltered by
applying volume cut-offs to smaller mining units. Different boundary and volume cut-
offs can be applied to the combinations of mining units scheduled for concurrent mining
and processing to obtain the best cut-off policy over time. There are no absolute rules:
applying volume cut-offs to the component mining blocks might be appropriate in some
circumstances; however, the author suggests that these should be lower than the lowest
minimum profit-grade derived for Mortimer analyses over the life of the operation, so
that mining blocks that might add value are not excluded from the reserve inventory.
The author has encountered situations where staff at a mine have identified that
the break-even cut-off grade fails to deliver a profit-generating head grade, and have
completed calculations to derive a grade similar to those here for the minimum profit
grade of Mortimer’s Definition; however, this has been applied as a boundary cut-off
rather than a volume cut-off, which implies that every tonne has to be at or above this
grade, not just the average grade.1 By applying a cut-off that is now significantly higher

1. The sample cut-off calculation in the first edition of the SME Mining Engineering Handbook referred to in the
preface includes an allowance for return on investment. A cost equivalent to the initial capital investment
multiplied by the required rate of return is included in the costs to be covered in the break-even calculation, but
this is implicitly applied as a boundary cut-off (that is, every tonne has to return at least that much) rather than
a volume cut-off (the average grade has to generate that return), as suggested by Mortimer.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 56
 CHAPTER 4 | Mortimer’s Definition

than it might need to be, the orebody is effectively reduced to nothing and the process
is discarded as inappropriate. In the absence of better processes and understanding,
the cut-off is arbitrarily raised until an acceptable head grade is obtained by trial and
error. The more rigorous process of Mortimer’s Definition could have been applied
successfully if it had been correctly understood.
Having said that, it is entirely appropriate for a company to decide that every tonne
mined must generate a minimum profit if that is what it wishes to do. The cut-off then
derived would be a pure 1D break-even grade based on the simple definition of break-
evens described in the previous chapter. If that did not result in an impractical orebody,
the author suggests that there would be a lot of pressure to bring marginal material that
still ‘made money’ into the schedule, which would potentially push the effective cut-off
down towards what would have been the Mortimer-style cut-off anyway. The point
of discussing Mortimer’s Definition as defined here is to indicate that, once we move
beyond the break-even concept of every tonne covering specified costs, the introduction
of targets focused on profit leads to an extension of the cut-off derivation process to
consider the nature of the geology – the grade distribution – in addition to costs and
prices.
It should be noted, however, that there will be an unresolved circularity. Mortimer’s
Definition makes use of unit costs. For the minimum profit break-even calculation, the
total unit cost will include the cost per tonne of costs that are not directly related to
ore tonnes, which will therefore be a function of the cut-off. The iterative processes to
converge on an appropriate cut-off described in Chapter 3 may therefore be required.
While the specifications of the grades required for the two parts of Mortimer’s Definition
are related to costs, prices and profits, it is worth noting that practical constraints in
some treatment plants may require a minimum and/or maximum head grade, which
could also be accounted for using the processes described. The required head grade
would then be the larger of the practical minimum and the minimum profit grade.

Summary
Mortimer’s Definition uses break-even calculations to identify grades that ensure that:
•• the average grade of ore provides a specified minimum profit per tonne treated
•• the lowest grade of ore pays for itself.
Both of these grades are determined by break-even calculations, but with different
costs included. The required profit margin is treated as a cost for this purpose.
The corporate goal is explicitly stated to achieve a minimum level of profitability while
ensuring that no material treated generates a loss. Compare this with the implicit every-
tonne-pays-for-itself goal of a simple break-even cut-off.
The second component of Mortimer’s Definition is focused on cost and product price,
and generates a simple break-even cut-off as described in the previous chapter.
The first component of Mortimer’s Definition accounts for the nature of the
mineralisation. The break-even grade required to achieve the defined level of profitability
is not used directly as a cut-off but is the required head grade ore treated. The grade
versus cut-off relationship within the material to be evaluated during a specified time is
used to identify the cut-off that will deliver this required head grade.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 57
 CHAPTER 4 | Mortimer’s Definition

The cut-off to be used will be the larger of the two derived for the components of
Mortimer’s Definition. Although the required ‘minimum profit’ head grade will be
greater than the every-tonne-pays-for-itself break-even, the cut-off delivering that head
grade will not necessarily be higher than the every-tonne-pays-for-itself break-even.
Perhaps one of the key understandings provided by Mortimer’s Definition is the
distinction between costs that must be borne by every tonne and those that must be paid
for by the average tonne and the impact it will have on the cut-off derivation method. In
Chapter 3, this distinction was noted by identifying that a simple ore–waste break-even
should include only those differential costs that are incurred because of the decision
to classify a block of rock as ore instead of waste. Costs such as capital development
and waste stripping were specifically excluded because they do not vary with the ore
tonnage. Rather, these are lump sum amounts expended to make more rock available,
then allowing delineation of ore within that rock mass. Those costs must still be covered
somehow if the operation is to be profitable. Mortimer’s Definition indicates that such
costs must be covered by the average grade of what is classed as ore, but not necessarily
by every tonne that contributes some value.
As with break-even cut-offs, neither of the Mortimer Definition cut-offs generated
accounts for the capacity of various stages of the mining and processing plant. Yet the
geology of the deposit via the grade distribution relationships is now accounted for to
ensure that some specified minimum level of profitability is achieved by the operations,
making this what we might call a 2D process. This is a significant improvement on
simple 1D break-even cut-offs, but still not in any way guaranteeing the maximisation
of any value measure.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 58
 HOME

Chapter 5

Lane’s Methodology –
Accounting for System Capacities

demystifying lane
Ken Lane published the foundation for shareholder value-based cut-off optimisation,
first as a technical paper (Lane, 1964) and later as a text book (Lane, 1988, 1997). In
developing our understanding of cut-off models with increasing numbers of parameters
or dimensions, Lane’s methodology could be described as three-dimensional (3D).
As well as the financial and geological dimensions of break-evens and Mortimer’s
Definition, Lane accounts for a third – the production system’s capacities to handle the
three materials identified earlier: rock, ore and product.
Lane’s methodology views the cut-off purely from the mine owner’s point-of-view.
It identifies that a resource’s value is ultimately the value of the cash that was generated
from it, and taking account of the time value of money, explicitly sets out to maximise
the net present value (NPV) of the resource.1 For break-even cut-offs the goal is implicitly
ensuring that every tonne of ore ‘pays for itself’. Using Mortimer’s Definition, the goals
are explicit: every tonne has to pay for itself, and the ore’s average grade has to deliver
a minimum profit. Similarly, for Lane’s methodology there is an explicit goal, which
is to maximise NPV. It should be noted, however, that there is no guarantee that the
maximum NPV will be positive.

1. Despite mines being essentially in business to make money for their owners, focusing on NPV – probably the
best single-value surrogate for quantifying cash generation – is often criticised for leading to ‘high-grading’ of
the deposit. Whether maximising NPV is an appropriate goal is not the main issue of this chapter – corporate
goals will be discussed in Chapter 10. This chapter is simply concerned with presenting a further step in the
evolution of cut-off theory: including system capacities in developing an optimised cut-off policy.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 59
 CHAPTER 5 | Lane’s Methodology

Lane, being a mathematician, developed his theory using rigorous mathematical

processes. From discussions with mining industry professionals, it is this author’s
experience that the majority who attempt to understand Lane’s book give up when they
reach the mathematical derivations of his formulas, missing the non-formulaic part of the
theory completely. The formulas may, however, be derived logically without resorting
to complex calculations. This chapter is this author’s attempt to ‘demystify Lane’.

establishing underlying concepts

Lane and other workers in the field in the 1970s and 1980s introduced several important
concepts to cut-off theory development, which are described in this chapter. These must
be understood if a Lane-style cut-off optimisation is to be carried out correctly.

Simple Lane versus complex Lane

At the most fundamental level, Lane identifies that the value of an operation is a function
of the remaining reserve, time and strategies adopted. Maximising value will result in a
plan that specifies the optimum strategies to be adopted.
Optimum strategies will include items such as sizes of final pits and phasing of mining,
maximum mining and treatment rates, cut-offs, and sequencing and scheduling of
various deposits or parts of deposits. Lane’s writings are directed largely at optimising
cut-offs. For most processes described by Lane, strategic options other than cut-off are
in effect assumed to be fixed, and the cut-off is optimised within that framework of
other locked-in decisions. This is the way that most commentators see Lane, and for the
purpose of this book, this approach will be referred to as simple Lane.
Within this framework of predetermined mining and treatment rates and mining
sequences and the like, we will see that once all the required inputs are specified, a series
of break-even calculations and simple algorithms can be used to derive the optimum
cut-off/s. This can be done for a single time period or over the life of the operation in
much the same way as simple break-even grades and Mortimer-style cut-offs; though
certainly the implementation of the process is more complex than for the cut-off models
described in the preceding chapters. Simple Lane may therefore be thought of as being
available to a competent technical expert using, for example, standard spreadsheet
software, and not requiring access to specialist optimisation software.
Lane’s basic statement of the problem includes all strategic options available to the
company, so his theory is not limited to cut-off optimisation. Whereas Lane’s writings
present readily available algorithms for cut-off optimisation if other strategic decisions
are taken as given, the same cannot be said for optimising those other strategic factors.
The author is aware of one software package on the market that includes proprietary
algorithms that reportedly optimise other strategies. We will refer to the general
application of Lane’s theory to strategic decisions as complex Lane.
For building our understanding of cut-off theory, this chapter will deal almost
exclusively with simple Lane. The ‘Complicating Factors’ section towards the end
of this chapter will, however, take an evaluator beyond the simple towards complex
Lane. Chapter 6 will go on to describe how optimisation of strategic decisions can be
approached in principle without recourse to proprietary algorithms.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 60
 CHAPTER 5 | Lane’s Methodology

Lane’s critical concepts

In his 1964 paper (and later his book published in 1988), Lane identified the concepts of:
•• Optimum cut-off – set to achieve some optimised goal (in this case, maximising NPV).
•• Cut-off policy – a planned sequence of cut-offs over time. Using an optimum cut-off
policy will maximise the remaining NPV in every year of the mine’s life. Constant
cut-offs, typically used at operating mines, are often incompatible with this.
•• Balancing cut-off – ensures that two of the three production system components (for
rock, ore and product) are operating at capacity. This is a function of the geology and
plant capacities, and is not related to costs and prices, but will often be the optimum
cut-off to apply.
•• Opportunity cost – this has always been understood qualitatively across the industry.
For example, one can mine down to the marginal break-even to ‘fill the mill’ if there
is a shortfall in ore supply, but this is only valid so long as low-grade material is not
displacing higher-grade feed. The opportunity cost quantifies the extent to which
such displacement can economically occur, and Lane’s methodology automatically
accounts for it in the cut-off optimisation process.
As Lane’s theory and cut-off optimisation methodology are easiest to comprehend in
the context of an open pit mine, we will first look at it in that situation. With underground
mining, there are significant differences in thinking and these will be explained once the
rationale has been developed for open pit mining.

Major underlying assumptions

Lane’s methodology in its simplest form assumes a predefined mining sequence that
generates a single stream of material (ie rock) from the mine. This also presupposes
that the final pit limits have been identified. Ore will be separated from the rock, with
product subsequently extracted from the ore. The methodology therefore does not
attempt to optimise the final size of an open pit, mining sequences or schedules; these
are givens.
Maximum handling rate capacities of the material flows – rock, ore and product – are
also specified, though these may include capacity changes over time.

Definitions
In developing his theory, Lane uses common industry terms with specific meanings,
some of which differ from their normal usage. These must be understood to avoid
confusion. Lane identifies three production stages as mining, treatment and marketing.
Mining constitutes all the processes dealing with rock (undifferentiated ore + waste)
and Lane (1988) refers to this as ‘mineralised material’. This is potentially misleading,
as some readers might take this to mean only the parts of the rock mass that contain
mineralisation. The reality is that material that is totally barren waste must also be mined
in an open pit and must be accounted for in the theory. Rock is all mined material, not
just material containing mineralisation.
Treatment covers all the processes dealing with ore.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 61
 CHAPTER 5 | Lane’s Methodology

Marketing covers all the processes dealing with product, and can include smelting
and refining. Because of the way that product is accounted for in the theory, marketing
comprises all the processes that are dealing with product, not just selling.2
Most mines deal with all three material components as they distinguish ore from waste,
and separate some form of product from the ore. The exceptions will be mines where the
total material moved is classified as ore, or where the ore as mined is sold directly to a
buyer without any additional processing by the mining company.
To apply Lane’s methodology correctly, it is important to identify transitions between
rock, ore and product. Let us now consider how these transitions are identified in open
pit and underground operations.

An open pit example

Consider a low-grade disseminated base metal sulfide orebody being mined as an open
pit with no stockpiling of low-grade ore. The implication with the valuable component
being disseminated is that the same mining methods and mining fleet are used for all
the material mined in the pit: there is no distinction in what we colloquially call mining
between ore and waste. The no-stockpiling condition (which is not uncommon for mines
in mountainous regions) implies that the ore–waste decision, once made, is irrevocable.
This helps us to understand the development of the theory initially, but these conditions
will be relaxed later in the discussion.
Although in practice grade control procedures will have identified which regions
are ore and which are waste and these will be drilled and blasted separately, for the
development of cut-off theory this early knowledge is irrelevant. In one sense, the
operators do not need to know whether they are dealing with ore or waste; their
mining practices are not impacted by this distinction. For cut-off derivation purposes,
operations in the pit, which we would usually call mining, generate rock only. All in-pit
operations are therefore mining as defined by Lane. Separating ore from rock does not
occur until a truck has to branch to the treatment plant or the waste dump on surface
after it has exited the pit.3 Mining, whether the normal meaning of the word or Lane’s
definition, is the same thing.
Crushing, grinding and all other ore-related processes are treatment. Lane’s specific
usage corresponds with the normal use of the term.
Product becomes separated from ore at the flotation banks where concentrate overflows
into the launders and the ore stream becomes physically separated into concentrate and
tailings streams. For now we will assume that the tonnages of tailings and ore feed are
the same, so the tailings stream may be thought of as a continuation of the ore stream
after the product stream has been separated from it.

2. In Lane (1964) mineralised material is simply referred to as material, treatment is referred to as concentrating and
marketing is referred to as refining. Lane identifies two product stream subcomponents: concentrate moving
from concentrator to refinery, and product from refinery to market.
3. In some operations, this is the case in practice. For example, in some uranium mines classifying a truckload as
ore or waste won’t occur until the truck has passed through a detector to measure the level of radioactivity and
the truck’s destination is signalled to the driver. In most mines, however, ore and waste are identified prior to
drilling and blasting, so that they are not mixed when blasted.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 62
 CHAPTER 5 | Lane’s Methodology

Concentrate pumping, thickening, filtration, drying and transport are all processes
dealing solely with product, and are therefore classified by Lane’s terminology as
marketing. While smelting, refining and selling are identified by Lane (and correctly by
many users of the theory) as marketing, Lane’s omission in discussing the inclusion of
the concentrate (or other product) handling activities within the concentrator or the mill
has caused confusion to many of his readers.
The concentrator or mill – because of its primary purpose of separating product from
ore – must have some activities that are ore-related and classed as treatment, and other
activities that are product-related and hence classified in a Lane-style cut-off derivation
as marketing. The mill is not treatment and treatment is not the mill. It is not uncommon
in practice to find a milling constraint expressed as a tonnage of ore per unit of time that
is constant up to a specified head grade, above which the milling constraint reduces with
increasing grade. This is a practical recognition that, above a certain grade, the product-
related parts of the circuit will be overloaded if treatment proceeds at the maximum rate
at which ore could be handled at lower grades; hence, the treatment rate of ore must be
reduced so as not to overload the product circuits. Similarly, all costs in a base metals
concentrator are, in many mine cost-reporting systems, recorded as a cost per tonne
of ore. This ignores that some costs are driven by ore tonnages and others by product
quantities. The average cost per tonne of ore can therefore vary with grade, even though
the underlying cost driver relationships have not changed.
So while for a simple open pit operation the rock–ore interface is easy to identify and
the cut-off-specific meanings of the terms mining and treatment correspond with the
normal meanings, the ore–product split in the concentrator, while fully appreciated
operationally and technically, is not always well recognised as the treatment–marketing
interface when using those terms in their cut-off-specific context. Often this interface
is perceived to lie in the downstream part of the process flow, perhaps starting with
concentrate transport after the concentrate has left the physical boundaries of the
mill. The product constraint specified will then only be based on the limitations of the
downstream concentrate-handling processes. But the real constraint on concentrate
production or handling will frequently be located within the mill, and may be lower than
the downstream concentrate handling capacity, so the product or marketing constraint
will be overstated.4 Additionally, because the starting point of the product stream is not
correctly identified, costs in the concentrator will be allocated incorrectly between ore
and product. For a Lane-style analysis, we need to distinguish between ore and product-
related processes, and hence costs and capacities associated with each stream.
Note that for an open pit, rock and ore are mined (in the colloquial sense)
simultaneously. Ore is merely reclassified rock, and is separated from the rock stream
by directing trucks carrying rock to different ore and waste destinations. In exactly the
same way, product is reclassified ore that is separated from the ore stream for further
processing, if appropriate, and sale, while the remaining ore is disposed of as tailings.

4. From time to time one will see a technical paper that declares: ‘We are doing a Lane-style cut-off optimisation. We
can sell everything we produce; therefore we do not have a marketing constraint’. This will usually be incorrect,
and indicates that the team conducting the analysis has missed the point. If a mine is not one of the relative few
that does not separate a product from the ore stream, there logically has to be a limitation on the amount of
product that can be dealt with somewhere in the production process. For a base metals operation, it will typically
be one of the concentrate pumping, thickening, filtration, drying and transport activities noted in the text, or, at
a gold mine, one of the activities associated with the activated carbon into which gold is adsorbed.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 63
 CHAPTER 5 | Lane’s Methodology

More complex open pit scenarios

A relatively straightforward extension of the open pit example described to consider
different costs for material classed as ore during mining activities in the pit. For
example, ore may use a closer-spaced drilling pattern to break the ore finer, use different
mining equipment and have different haulage distances relative to waste, as described
in Chapter 3. In these cases, the differential cost between ore and waste will need to be
classified as an ore-related treatment cost. All rock mined will incur a basic mining cost,
assuming it is classed as waste, and what is classed as ore then incurs the differential
cost as an ore-related cost. The difference may be positive or negative: tighter drilling
patterns and smaller, more selective mining fleets may incur additional costs for ore
over those for waste, while shorter haul distances for ore could generate a negative cost
differential for ore.
When considering capacities, problems may arise. At the simplest level for an operation
with different-sized ore and waste mining fleets, the capacity for the mining process
(that is, mining rock and removing it from the pit) will be the sum of the capacities of the
two fleets; the capacity of the ore fleet in the pit will generate a treatment (ore-related)
constraint, and if this is less than the ore-handling constraint in the treatment plant, the
treatment-limiting constraint will be in the mine, not in the mill. This will not usually
prevent the use of Lane’s methodology.
The problem is exacerbated when the fleet capacity is a function of the cut-off used.
This can occur if there are significantly different haulage distances for ore and waste.
With different distances, the haulage rates (expressed in, for example, tonnes per
hour) will be different for ore and waste. If changing the cut-off used generates more
of the material with the lower haulage rate, the overall tonnage capacity of the fleet
will be lower than previously. Lane’s methodology in its simplest form cannot handle
capacities that are functions of the cut-off they are being used to determine; a more
complex analysis is required, but this may still fit within the conceptual limits of a Lane-
style cut-off policy optimisation and does not necessarily invalidate the approach. Often
such cut-off-dependent changes can be handled within the iterative processes that will
be described for a complete life-of-mine cut-off optimisation, so long as the relationships
required have been correctly formulated in the optimisation model being used. This is
discussed shortly.
Other complex mining requirements may extend this. For example, if the smaller,
more selective fleet is also required to be used for mining a certain amount of waste to
assist with the efficient separation of ore and waste in the pit, a portion of the smaller
fleet’s capacity must be assigned to waste and the ore capacity will be that much lower;
however, as in the previous paragraph, this quantity may be a function of cut-off. One
may also come across situations where the acid-generating potential of waste to be
dumped must be countered by the mining of additional acid-neutralising rock from
a quarry outside the pit limits. The requirement to do this will depend on the waste
tonnage to be dumped and its acid-generating potential, and this requirement will
reduce the total rock mining capacity of the fleet in the producing pit. This will be a
function of the cut-off applied, and again a more iterative analysis is required, but the
methodology will remain robust.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 64
 CHAPTER 5 | Lane’s Methodology

An underground mine example

Consider now an underground mine with trucks hauling and tipping waste from
development into stopes and hoisting ore in a shaft.
Most activities in an underground mine are dealing directly with ore. All these ore-
related activities will therefore, by Lane’s terminology, be treatment. For the example
described, production drilling, blasting and shaft hoisting operations are only conducted
for ore so these activities are therefore treatment.
It’s already been noted that in an underground operation, rock is typically the total
mineralised material. It is separated into ore and waste by planners creating designs
for stopes: the rock within the design limits is extracted as ore while the remainder
remains in situ as waste. In an underground mine, it is major access development that
exposes rock so that it can be separated into ore that is stoped and waste that remains
in situ. Major waste access development is therefore mining by Lane’s definition, and
is frequently the only mining activity in an underground mine for cut-off derivation
purposes. It will usually be the extension of a shaft or decline to access deeper ore,
or lateral development to access a new mineralised zone some distance from current
resources. The key distinguishing feature of waste access development classed as
mining is that its quantity does not depend on the amount of rock exposed or ore
defined. Following similar logic to that in Chapter 3 – identifying what costs should
and should not be included in a break-even cut-off – footwall access development that is
independent of cut-off would also be classed as mining.
Lane (1988, 1997) suggests that the mining component of development includes
‘driving headings, raises, etc and forming stopes in preparation for the extraction of
ore.’ The author considers this definition to be ambiguous or perhaps specific to certain
mining methods, and suggests that mining development is any for which quantity does
not depend on the amount of ore exposed by the development in question (typically
major access development in waste); any that depends to some extent on the quantity
of ore is treatment development. ‘(H)eadings and raises’ may be correctly classed as
mining if they form part of the overall access and infrastructure openings outside the
orebody, but will often be treatment if they are in the orebody to form stopes.5
Some waste development may, however, be ore-associated and therefore classified
as treatment. Footwall access development that varies with cut-off (for example, if the
strike length of the orebody varies with cut-off) and stope access cross-cuts in waste,
the number and hence total length of which will vary with cut-off, would be classed as
treatment-related activities.
It should be noted that the relationships between treatment development metres and
tonnes of ore may not be purely variable. For example, if an orebody expands in plan
both along strike and perpendicular to strike as the cut-off is lowered, the length of strike
development for both footwall access and strike drill drives may be directly proportional
to the length of the orebody; however, because of the expanding width, the metres per
tonne of ore may have two terms: a constant reflecting the proportionality between
development and orebody length, and a term that is perhaps inversely proportional

5. Lane’s classification of particular types of development as mining as quoted here may be valid for certain mining
methods, such as South African narrow tabular reef mining for gold and platinum group metals. Each mine and
mining method must be considered on its own merits.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 65
 CHAPTER 5 | Lane’s Methodology

to tonnes to account for the greater tonnages per metre of development as ore widths
increase with decreasing cut-off.
If these relationships cannot be accommodated by formulas, the simplistic response is
that the theory fails to adequately address all the real-world issues and cut-off models
with more dimensions may be required; however, in many cases, these more complex
interrelationships can be handled by the iterative processes involved in applying
Lane’s methodology in practice, which are described shortly, and the implementation’s
complexity does not invalidate the underlying principles.
Underground operations introduce other complexities, but careful thought will
often resolve these. For the example given, both ore and waste are trucked to tips.
To correctly classify this activity, it is useful to note that ore and rock-associated
activities are both temporally and spatially separated in an underground mine. In
a pit, waste and ore are mined together simultaneously, but underground, today’s
access development exposes tomorrow’s ore, and today’s ore is a result of yesterday’s
development. The rock-associated access development carried out during a particular
period is the result of a policy decision that production will occur in the region being
exposed in the future, and is not related to today’s ore: it is therefore not relevant in
determining what is today’s ore.
In order to identify the optimum cut-off for the available material that may potentially
be classed as ore, it is necessary to classify the mining waste development costs as
mining costs. Additionally, because of the policy to mine this waste to expose future
potential ore, the waste loading and hauling requirements must be deducted from the
total loading and hauling capacity, with the remaining balance becoming the effective
ore-handling or treatment constraint for determining what is ore today.
It may be useful to consider the treatment development costs as those that are in or
adjacent to the orebody and are directly related to making the selected mining method
work properly. Mining development costs are those that are needed to expose the
mineralisation and would still be necessary, even if nothing was stoped from the area
made available and production moved on to the next block of mineralisation exposed.
It could be argued that all mine development is required to make the method work and
should therefore be classed as treatment development. The semantics become important.
While it is necessary to have connections between operating levels so that activities
related to the mining method can occur as required, the development between levels is
merely access. The mining method does not depend on the nature of this development
(for example, whether the access is by inclined drives or by raises), nor in many instances
the quantity of this development, nor its routing through the rock mass. This is perhaps
the distinguishing feature of mining development. For treatment development, on the
other hand, the nature of the development – sizes, shapes and their placements relative
to the orebody boundary – is critical to the mining method description.
This discussion has shown how the rock–ore or mining–treatment interface might be
dealt with in an underground mine, but these will not be universally applicable. Each
mine’s circumstances must be examined individually to separate mining and treatment
operations and hence distinguish their costs and define capacities.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 66
 CHAPTER 5 | Lane’s Methodology

Applying Lane’s Methodology in Practice

Having established some important concepts, let us now move on to discussing how
Lane’s methodology can be applied in practice. As indicated, this discussion assumes
simple Lane, with all strategic options apart from cut-off predetermined. In particular, the
mining sequences result in a predefined sequence of rock coming from the pit. Capacity
limits for mining rates and treatment rates have also been decided. The optimisation
process will seek to determine cut-offs that maximise NPV within these constraints.
It was seen in previous chapters that the 1D break-even process, given a specification
of costs to use, will generate one number that could be used as the cut-off. In Mortimer’s
Definition, a 2D process, two numbers are generated and the larger of the two is selected
as the cut-off. Lane’s methodology, a 3D process incorporating financial and geological
data and mining and processing capacities, generates six potential cut-off values, one
of which will be optimal at any point in time. A simple two-stage process is applied to
reduce the six, firstly to three and then those three to the one to be used. The values of
the six will most likely change over time as both geological and economic conditions
change. The cut-off to use, and its value, will also potentially change over time.
To develop Lane’s methodology, we’ll use algebraic symbols to represent the
variables in the analysis. The notation in the following subsections is the same as
that used by Lane (1988, 1997), so a reader familiar with Lane’s work can make the
necessary correlations.

Physical parameters
Since cut-off derivation deals with three types of material (rock, ore and product),
we require variables for the associated stages in the overall production process. Lane
denotes these and other physical quantities as follows:
•• M = maximum mining rate (expressed as units of rock per time period)
•• H = maximum treatment rate (expressed as units of ore per time period)
•• K = maximum marketing rate (expressed as units of product per time period)
•• y = metallurgical recovery or yield (ie, product recovered/product contained in ore)
•• g = each of the six potential cut-off grades initially identified by the process
•• G = the three intermediate cut-offs and one final cut-off grade obtained during the
process of reducing the six, firstly to three and finally to the one cut-off to be used.
It should be remembered that, for an underground operation, the ore stream bottleneck
and hence the process defining the maximum ore tonnage that can be treated (the
treatment capacity) may be in the mine. In some circumstances this may be the case in
an open pit.

Cost and revenue parameters

We’ll also require variables to represent the costs associated with each stage in the
production process – mining, treatment and marketing. Lane denotes these and other
financial quantities as follows:
•• m = mining variable cost (per unit of rock)
•• h = treatment variable cost (per unit of ore)
•• k = marketing variable cost (per unit of product)
•• f = fixed costs (per time period)

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 67
 CHAPTER 5 | Lane’s Methodology

•• p = product sale price (per unit of product)

•• F = opportunity cost (per time period – discussed shortly in more detail).
The lower case letters ‘m’, ‘h’ and ‘k’ are also used as subscripts with cut-off variables
‘g’ and ‘G’ to identify the process stages under consideration.
Chapters 2 and 3 detail the types of costs that would be included and excluded in each
of these components. As noted there, the accounting distinction between operating and
capital costs is irrelevant for cut-off determination. Sustaining capital should therefore
be included in each cost component according to how the costs behave and which
physical parameters are driving them.

Capital development costs

It could be argued that ongoing capital development in an underground mine is
sustaining capital and should be included in the cut-off derivation as a tonnage or time-
based cost. It cannot be included as cost per tonne of ore: if it were, the ore tonnage
exposed by capital access development would depend on the cut-off, which would
make the process circular – a cut-off would have to be assumed to derive the cut-off.
Lane (1998, 1997) suggests that capital development should be included in the mining
cost, in which case it would need to be applied to the tonnes of rock exposed or as a
fixed cost, which assumes that the duration of mining has been determined. This may
also depend on assuming a cut-off to identify the time over which the capital must be
expended to make the rock available for production. The author suggests that this is
generally not the right way to deal with capital development expenditure.
It is certainly appropriate to consider this cost in the break-even calculation used to
determine the head grade in order to deliver the minimum required profit with Mortimer’s
Definition. For a normal full-cost break-even cut-off, and in using Lane’s methodology,
capital development spending enables producing from the area made available, and is
a lump sum cost not dependent on the amount of mineralisation (or rock) exposed and
ultimately classified as ore and stoped. Expenditure is justified because the difference
between the optimum mine plan’s value with the area made available and the optimum
plan without it is greater than the cost of the access development. The optimum plan,
though it may be generated before the capital is spent, considers what would be done
after the development is complete, at which stage the development cost would be a sunk
cost that is irrelevant to decisions about what constitutes the optimum mine plan. The
planner’s task is to make the most of all assets available, such as fleet and fixed plant, as
well as capital development that accesses mineralisation. As with all project capital, the
access capital development costs are not part of a break-even cut-off derivation process.

What is ‘opportunity cost’?

This is not a textbook on discounted cash flow (DCF) techniques, but to understand
opportunity cost, a brief introduction is necessary for readers unfamiliar with DCF
concepts.
Let us assume that we invest $100 today at an interest rate of ten per cent. In one year’s
time, we will have earned $10 interest, and our investment will have a balance of $110.
All things being equal, $100 now therefore has the same value for us as $110 in a year’s
time. If we have $100 in a year’s time rather than now, it is the same as having $100/1.1,
or $91 now, rounded to the nearest dollar. This is known as the present value – the value

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 68
 CHAPTER 5 | Lane’s Methodology

to us now of a cash amount at some time in the future. So, if we could have $100 now,
but for whatever reason its receipt is deferred for one year, we have effectively lost $9
of present value. This loss of value due to the delay in receiving it is the essence of the
opportunity cost.
We could think of our operation as generating a sequence of parcels of rock. Some
proportion of each of these will be classified as ore to generate product and hence
some (we hope) positive net cash flow. We can therefore also think of the results of the
operation as a sequence of net cash flows associated with each sequential parcel of rock.
Our aim is to identify the strategies to maximise the value of that sequence of cash flows.
We may identify optimum strategic plans for many years into the future; however, the
reality is that we only have to make decisions for the next parcel of rock to be dealt with
in the sequence. Future strategies that we have identified are generally just provisional.
After we have mined, treated and generated revenue and profit from the product,
we will notionally be able to make decisions based on the knowledge we then have
regarding the resource remaining.
With our current assumptions about future conditions, the resource remaining after
dealing with the next parcel of rock will have a maximum value. From both a DCF and
a cut-off optimising point-of-view, the operation can be represented simply by the next
parcel of rock that we are dealing with, plus the receipt of a lump-sum cash benefit that
represents the value of the resource remaining after we have dealt with the next parcel
of rock. Any decisions that increase the length of time it takes to deal with the next parcel
of rock will defer receiving the rest of the resource’s value. As previously indicated, this
results in a reduction in the value of the rest of the resource to us in present value terms.
This loss of value is the opportunity cost. The implication is that additional tonnage of
rock from the mining parcel being evaluated that is classified as ore will add material
to both the ore and product streams. If one of those streams is already operating at full
capacity, this will extend the duration of dealing with that parcel of rock, and therefore
defers receiving the rest of the resource’s value. To make that treatment economic, the
additional value obtained must cover not only the marginal cash costs of treating the
ore, handling the product and extending the period, but also the opportunity cost - the
time-value-of-money cost of deferring receipt of the value of the rest of the operation.
Lane (1988, 1997) presents an extensive discussion of opportunity cost. An
approximation of the opportunity cost (F) is given by the formula:

F = rV – dV/dt
where:
r = discount rate6 (expressed as a decimal fraction)
V = NPV
dV/dt = the rate of change of NPV with time
The first component, rV, is approximately the loss of value due to the deferral in
receiving the NPV of the rest of the operation. This assumes that the NPV does not

6. It is beyond the scope of this book to discuss the derivation of discount rates. For this the reader is referred to
any standard text on project evaluation or discounted cash flow techniques.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 69
 CHAPTER 5 | Lane’s Methodology

change with deferring, implying that economic conditions stay the same in the future.
The lower the discount rate, the better the approximation.7
The second term is the rate of change of NPV with time as a result of changing
economic conditions, expressed mathematically as the differential of NPV with respect
to time, dV/dt. If economic conditions deteriorate – for example, if the price is predicted
to fall – moving the mining and treatment plan back into a period of worse conditions
will result in a reduction in the NPV of that plan. In these circumstances the change
in value is negative: the value becomes lower with deferral. The inclusion of dV/dt in
the opportunity cost formula with a negative sign converts this change in value into a
positive, or extra, cost.
It is, however, possible that deferring the project for a year moves the production plan
into a period in the future with higher prices. In that case, the NPV will increase as a
result, producing a negative component that will reduce the total opportunity cost. If
dV/dt numerically is greater than rV, the overall opportunity cost may become negative,
representing an opportunity gain rather than cost.
It is also possible, even with profitable projects, for the remaining NPV to be negative at
some point. For example, this can occur towards the end of the mine life when the major
remaining cash flows are payments for mine closure, redundancy and rehabilitation.
Deferral of these closure payments will also result in a negative opportunity cost.
Finally, it should be noted that because the discount rate (r) is typically expressed as an
annual rate, both rV and dV/dt have units of dollars per year. F is therefore equivalent
to an extra annual fixed cost, expressed in dollars per year.

Basic Principles of Finding the Optimum Cut-off

Having established a vocabulary, a notation and an understanding of how costs are to
be allocated to cost parameters, we can now establish the optimum cut-off. Three of the
six potential cut-offs to be derived are known as limiting cut-offs. They are derived by
break-even formulas for each of the three production stages considered independently
to be the limiting factor on production. The other three are known as balancing cut-offs:
using these will result in two of the three production process stages operating at their
capacity limits concurrently.
The analysis addresses an increment of rock to be mined, rather than a specified time
frame; 8 however, it might be convenient to use the existing schedule to identify these rock
increments, each of which will be referred to as a mining step. It should be remembered
that applying simple Lane assumes a prespecified mining sequence. The mining steps
are therefore the increments or parcels of rock that are generated sequentially by the
mining process. Although the steps might initially represent sequential years, quarters
or months, the cut-off optimisation process may potentially change each step’s duration.

7. The loss of value is more accurately given by V.[1 – 1 / (1 + r)t]. If t = 1 year, this reduces to rV / (1 + r), so rV will
overestimate this component of the opportunity cost somewhat. Since both r and V are approximations based
on forecasts and the time of deferral will almost certainly not be 1, rV can be demonstrated to be a reasonable
approximation for the time value of money loss over a range of deferred times, at least up to one year.
8. In practice it is feasible to construct a model to conduct a Lane-style evaluation using time periods rather than
rock increments; however, for developing this methodology we will consider sequential increments of rock, as
illustrated in Lane (1988, 1997).

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 70
 CHAPTER 5 | Lane’s Methodology

What is important is that by definition steps are sequential and do not overlap; mining
steps must be mined in their entirety so that the sequence is unchanged.
Mining steps are therefore not the same as the pushback or cut-back stages normally
used in open pit mine planning. Rather, a mining step might comprise rock from two
or more pushbacks or cut-backs; for example, ore from the bottom of the pit from
the first cut-back, a mix of ore and waste from the middle levels of the pit from the
second cut-back and waste only from the upper benches of a third cut-back. For cut-
off derivation using simple Lane with mining sequences predetermined, it is irrelevant
what pushback stage or where in the pit the rock emanates from. Rather, the task of the
cut-off optimisation process is to identify the best cut-off to apply to all material that
reaches the rim of the pit simultaneously. (In full strategy optimisation or complex Lane,
optimising the sequence is also part of optimisation – but that is for a later chapter.)
The same time is the time frame for the planning purposes involved, and is by definition
the time taken to deal with a mining step. Therefore, mining step tonnages might be
relatively small for short-term planning, but become progressively larger in the future
for long-term planning where plans are less precise. The primary attribute of each mining
step is its total tonnage. The maximum mining rate will then set the minimum time to
deal with the step; however, the actual time could be greater, but will only be identified
at the completion of the cut-off optimisation process at which point the cut-off to be
applied, the actual mining, treatment and marketing rates, and hence the duration of the
mining step, will all be specified. The implication is that what is mined within a mining
step is assumed to be homogeneous. If not, redefining the steps might be appropriate to
achieve a satisfactory approximation of homogeneity within each step.
The aim of the cut-off optimisation process is to maximise the value of each mining step
while accounting for its impact on the rest of the resource’s value, which then maximises
the combined value of all steps. As noted, maximisation of NPV does not necessarily
imply that the NPV is positive: the maximum NPV may be the least negative value.

Limiting Cut-offs
Three of the six potential cut-offs to be derived are known as limiting cut-offs. These are
derived by break-even formulas considering each of the process stages to be the limiting
factor on production. When each stage is considered to be the rate-limiting bottleneck in
the overall process, the other two production stages are assumed to not be constraints:
their capacities are ignored and may be infinite. This is, of course, unreal but this lack of
realism is resolved when the six potential are reduced to the one optimum cut-off.
Lane (1988, 1997) derives these formulas by a rigorous mathematical process. It is this
feature in the early chapters of his book that causes many mining industry technical
staff to stop reading and consequently fail to grasp many of the important concepts he
has developed. To demystify Lane, therefore, the following subsections derive these
formulas by simple logical discussion rather than by mathematical manipulations.
The formulas generated below are presented using the same notation as Lane uses, so
that the reader can make appropriate correlations between this book and Lane’s if desired.

Mining-limited cut-off (gm)

The mining-limited cut-off is derived by assuming that the maximum mining rate
(for removal of rock from the pit or extending underground access development to

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 71
 CHAPTER 5 | Lane’s Methodology

previously unexposed parts of the resource) is the overall constraint on the production
process. If the mining rate is the constraint, mining will proceed at the maximum rate
rather than any slower rate, which would reduce the value.
The rock associated with the mining step must be completely mined (so that
subsequent steps can also be mined). Mining variable costs are therefore incurred
regardless of what is classed as ore. It is a principle of economic analysis of alternative
options that factors such as common costs will not influence the relativities of each
option and can therefore be ignored in the comparison. The total mining variable
costs, being the same for all cut-offs, are irrelevant to the decision and are therefore
not included in the calculations to support it. These are therefore excluded from the
mining-limited break-even calculation. We could, if we wished, perform a full-equal-
benefits-type of break-even calculation to explicitly include mining costs and see them
cancel out, as shown in Chapter 3; however, having established the principle that only
differential costs need to be included, we will simply exclude them.
The tonnage of rock in the mining step is known and the maximum mining rate is
specified. Together these will determine the time taken to deal with the step. Given that
mining is the overall rate-limiting activity, the duration is therefore fixed and does not
depend on cut-off. All fixed costs are incurred for that duration regardless of what is
classed as ore, and are therefore also irrelevant to the cut-off decision and excluded from
the mining-limited break-even calculation.
The revenue from what is classified as ore must therefore only cover the variable
treatment and marketing costs associated with that ore and the product derived from
it in order to break even. Any grade higher than the break-even will make a positive
contribution to cash flow and hence NPV.
The variable cost per tonne of ore is the treatment-variable cost (h). As described in
Chapter 3, the net revenue per unit of product in the ore is obtained by deducting the
variable cost per unit of product (k) from the price per unit of product (p). The difference
is then multiplied by the metallurgical recovery (y) to derive the recoverable net revenue
for a unit of product contained in the ore.
The mining-limited cut-off is therefore given by:9
gm = h/(p–k)y
This is similar to a traditional marginal break-even cut-off, and indicates the lowest
grade of material to be treated when the mine is unable to fill the mill. Note, however,
that product-related costs are deducted from the price in the denominator, not included
in the cost per tonne of ore in the numerator. Its rationale is fully explained in Chapter 3.
Note also that the numerator of the formula has units of ‘dollars per tonne of ore’ and the
denominator ‘dollars per unit of product’. The resulting units for the cut-off grade are
therefore ‘units of product per tonne of ore’, a grade unit, as is to be expected.

9. It should be noted that the formula as presented here is not strictly mathematically correct. As it is written,
the recovery ‘y’ is a factor in the numerator of the calculation. To place it correctly in the denominator where it
belongs – to generate the recoverable net revenue per unit of product in the ore – there should be a second set
of brackets around the product of terms in the denominator; that is: gm = h/[(p–k)y].
Similar inconsistencies exist for other limiting cut-off formulas. To retain Lane’s style of presentation, the author
has not added these additional brackets. Readers should note that, in these formulas, all terms to the left of the
‘/ ’ are in the numerator, and all terms to the right are in the denominator.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 72
 CHAPTER 5 | Lane’s Methodology

Treatment-limited cut-off (gh)

As for the mining-limited case, the rock in the step under consideration must be
completely mined. Mining variable costs are again incurred regardless of what is classed
as ore, and hence are irrelevant to the cut-off decision and do not come into the break-
even calculation. However, with treatment being the overall rate-limiting process, the
amount of the rock classed as ore and the maximum treatment rate will determine the
minimum time taken to deal with the mining step. In the extreme case where the cut-off
were so high that none of the rock were classed as ore, the duration of the step would
be zero, implying an infinite mining rate. As noted, the unreality of this will be resolved
when all the six potential cut-offs are considered to settle on the single optimum cut-off.
Given that the duration of the mining step depends on cut-off, fixed costs are incurred
for the time that results from what is classed as ore. Fixed costs, therefore, are now
marginal costs and relevant to the cut-off decision. The value of the ore should cover:
•• variable treatment and marketing costs (as for the mining-limited case)
•• fixed costs incurred during the time taken to treat the ore
•• the time-value-of-money opportunity cost of deferring the NPV of the rest of the
operation by the time taken to deal with the mining step under consideration.
The net revenue calculation in the denominator is the same as for the mining-limited
formula. Ore-related costs in the numerator include not only the treatment-variable
costs (h), but also the fixed costs since they now depend on cut-off. These consist of both
the normal cash fixed costs (f) and the time-value-of-money opportunity cost (F), which
as described is also a cost per year (or whatever the relevant unit of time is). Given a
treatment rate limit of ‘H’, every extra tonne of rock classed as ore extends the mine life
by ‘1/H’ years (or time units). The additional fixed costs associated with that extension
of life are therefore represented by ‘(f+F)/H’.
The treatment-limited cut-off is therefore given by:10

gh = {h+(f+F)/H}/(p–k)y

All costs in the numerator are expressed in dollars per tonne of ore. Since the ore
stream is the production bottleneck, fixed costs expressed in dollars per unit of time
are converted to dollars per tonne of ore by dividing by the treatment limit, H, which
is expressed in ore tonnes per unit of time. Expressing fixed costs as a cost per tonne of
ore by dividing the fixed costs per year by the production rate is common at most mines.

Marketing-limited cut-off (gk)

A similar rationale applies when the marketing operations, dealing with product,
constrain the overall production process.
As with the mining and treatment-limited cases, the rock in the step under consideration
must be completely mined. Mining variable costs are again incurred regardless of what
is ore, so are irrelevant to the cut-off decision and break-even calculation.
With marketing being the overall rate-limiting process, the amount of recoverable
product in the rock classed as ore and the maximum marketing rate (to deal with the
product recovered from the ore) will determine the minimum time taken to deal with

10. Or more correctly mathematically: gh = {h+(f+F)/H}/[(p–k)y].

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 73
 CHAPTER 5 | Lane’s Methodology

the mining step. Again, in the extreme case if the cut-off were so high that none of the
rock were classed as ore, no product would be generated and the mining step duration
would be zero, implying an infinite mining rate.
Given that the duration of the step again depends on cut-off, fixed costs are incurred for
the time that results from marketing the product recovered from the ore, and therefore
are relevant to the cut-off decision. The value of the classified ore must therefore cover:
•• variable treatment and marketing costs (as with mining and treatment-limited cases)
•• fixed costs incurred during the time taken to market the product recovered
•• the time-value-of-money opportunity cost of deferring the NPV of the rest of the
operation by the time taken to deal with the mining step under consideration.
For the treatment-limited cut-off, an extra tonne of ore treated extends the life by 1/H
years, and the term (f+F)/H is added to the numerator of the mining-limited formula
to represent the extra fixed cost incurred per tonne of ore treated. However, when the
maximum marketing rate – the rate of dealing with product – is the constraint limiting
the total production process, every extra unit of product recovered from rock classed as
ore extends the mine life by 1/K years (or time units). That extra unit of product must
therefore bear not only the associated direct variable costs, but the additional fixed costs
associated with that extension of life. The fixed cost incurred is given by ‘(f+F)/K’, which
has units of dollars per unit of product. For the marketing-limited cut-off, (f+F)/K is a
product-related cost and is deducted from the price (p) in the denominator of the break-
even formula in addition to the direct product-related variable cost (k).
The marketing-limited cut-off is therefore given by:11

gk = h/{p–k–(f+F)/K}y

Relationships between limiting cut-offs

The remaining NPV and hence the opportunity cost (F) will usually be positive, therefore
the total fixed cost (f+F) will be positive.
Relative to the mining-limited cut-off, the treatment-limited cut-off has an additional
cost term in the numerator. The treatment-limited cut-off is thus usually greater than the
mining-limited cut-off. Similarly, relative to the mining-limited cut-off, the marketing-
limited cut-off has an additional cost term reducing the denominator. The marketing-
limited cut-off is also usually greater than the mining-limited cut-off.
If towards the end of the life of the operation the remaining NPV and hence the
opportunity cost (F) becomes negative, it is possible that the total fixed cost (f+F) will
become negative. In this case, the relativities between the mining-limited and the other
two limiting cut-offs will be reversed. The practical implication is that if there are large
closure costs, it may be rational to reduce the cut-off to less than the marginal break-
even defined by the mining-limited cut-off, thereby extending the life and deferring the
incurrence of the closure costs.
The relativities of the treatment and marketing-limited cut-offs cannot be specified
simply by comparisons of the break-even formulas.

11. Or more correctly mathematically: gk = h/[{p–k–(f+F)/K}y].

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 74
 CHAPTER 5 | Lane’s Methodology

Examples of limiting cut-offs

The following inputs are used to derive mining-, treatment- and marketing-limited
cut-offs (gm, gh and gk respectively) and the limiting cut-offs are derived by simple
substitution of these values in the formulas.

Physical parameter assumptions

•• Maximum mining rate M = 5.0 Mt/a of rock
•• Maximum treatment rate H = 2.0 Mt/a of ore
•• Maximum marketing rate K = 150 000 t/a of product
•• Metallurgical recovery y = 90 per cent

Cost and price parameter assumptions

•• Mining variable cost m = $5.00/t of rock
•• Treatment variable cost h = $7.50/t of ore
•• Marketing variable cost k = $40.00/t of product
•• Fixed cost f = $16.0 M/a
•• Product price p = $800/t of product

Opportunity cost assumptions and derivation

•• Discount rate r = 10 per cent/a
•• Net present value V = $120 M
•• dV/dt = 0 (ie economic conditions are forecast not to change in the future)
•• Opportunity cost F = rV – dV/dt = $12 M/a

Mining-limited cut-off derivation

gm = h/(p–k)y
= 7.5/[(800 – 40) × 0.9]
= 0.0110
gm = 1.10 per cent

Treatment-limited cut-off derivation

gh = {h+(f+F)/H}/(p–k)y
= {7.5 + (16 M+12 M)/2 M}/[(800 – 40) × 0.9]
= 0.0314
gh = 3.14 per cent

Marketing-limited cut-off derivation

gk = h/{ p–k–(f+F)/K}y
= 7.5/[{800 – 40 – (16 M + 12 M)/150 k} × 0.9]
= 0.0145
gk = 1.45 per cent
These cut-off values will be used to derive the optimum cut-off.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 75
 CHAPTER 5 | Lane’s Methodology

The impact of opportunity cost and location of the production bottleneck

Opportunity cost is a component of both the treatment- and marketing-limited cut-offs.
The treatment-limited cut-off gh, for example, is derived by this formula:
{h+(f+F)/H}/(p–k)y
When substituting the values of the parameters, this becomes:
{7.5 + (16 M + 12 M)/2 M}/[(800 – 40) × 0.9]
The total cost in the numerator is {$7.5/t + ($16 M + $12 M)/2 Mt}, which evaluates to
$21.50/t. The opportunity cost component is $12 M/2 Mt, or $6.00/t, which equates to
approximately 28 per cent of the total cost or approximately 39 per cent of the $15.50/t of
cash costs. This is a significant component of the cut-off. In fact, for a high-value long-life
project, an opportunity cost in excess of 100 per cent of cash costs is not unrealistic in the
early years of the operation’s life.
There is often an emotive argument advanced that all material with a grade above a
break-even cut-off should be mined, since it makes money. If ore above full cost break-
even has been prepared for mining but there is adjacent material with a grade below the
full cost break-even but above the marginal break-even, the argument is that the value
of this marginal material will be lost forever by not being mined at the same time as the
above-break-even material. So what is the value of that marginal material that we might
sterilise if we don’t mine it now?
If this lower-grade material is displacing higher grade ore, the unpopular answer is:
nothing at all – it is rather a significant value destroyer. The fixed and opportunity cost
components of Lane’s treatment- and market-limited break-even cut-offs now show us
the additional costs that need to be covered by each tonne of material added to the ore
stream if it displaces higher-grade material and extends the mine life. However, if it
is not displacing higher grade ore, the common, popular answer is that it is worth
treating if its value exceeds marginal cost or, in Lane’s terminology, if its grade is
greater than the mining-limited cut-off.
Lane’s mining-limited and treatment-limited conditions, and the formulas derived to
generate cut-offs applicable in each case, are formal and rigorous representations of what
are colloquially referred to as mill-limited and mine-limited. These are situations where
additional ore is and is not displacing higher-grade material. But, what are colloquially,
organisationally and geographically known as the mine and the mill are not necessarily
mining and treatment in the strict sense required for cut-off derivation. In an open pit
operation they often will be, but in an underground operation much of what occurs in
the mine is treatment. To speak of an operation as mine-limited in the way this term is
commonly used may be misleading when it comes to assessing what break-even cut-offs
can be applied to material potentially added to the ore stream. Although the constraint
may geographically be underground in the mine, it is essential to identify whether the
output-limiting activity is in:
•• the rock (mining) stream – the rate of access development or waste stripping is
constraining the exposure of mineralisation that may be then classed as ore or waste
•• the ore (treatment) stream – limited by activities such as drilling, loading or hoisting,
or by scheduling constraints that limit available ore sources at any particular time.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 76
 CHAPTER 5 | Lane’s Methodology

Mine-limited does not necessarily mean mining-limited, and the mining-limited

cut-off formula is therefore not always the appropriate one to use just because the
constraint is in the mine. We would do better to avoid the common terminology of
mine-limited and mill-limited and instead use rock-limited and ore-limited.
Finally, if there is an ore shortfall and a marginal cut-off may be necessary to fill the
mill, there is one more unpopular question that must be asked: why do we not have
higher grade ore available now? If it is simply part of the normal geological variability
and uncertainty associated with mining, and therefore a short-term fluctuation from
average conditions that will right itself naturally in the near future, then there is no
major concern. But perhaps something has gone wrong with the planning or decision-
making in the past. For example, deferring access development or exploration to
identify further resources to be brought into the mine plan may mean that in the longer
term there will regularly not be sufficient ore above the optimum cut-off available. In
this case, we need to learn from such events, generate plans to bring ourselves back
into a position where we can implement the optimum mine plan – which includes
the optimum cut-off policy – and put procedures in place to ensure that short-term
expediency does not lead to such problems again in the future.
Once we identify that there is spare capacity, this does not necessarily mean we can
classify as ore all the material down to the marginal break-even, or in Lane terms, the
mining-limited cut-off. Lane’s methodology, however, now gives us a process to identify
what the cut-off should be in this case. It will often be what Lane identifies as a balancing
cut-off, which is the subject of the next subsection.

Balancing Cut-offs

Balancing cut-off concepts

We’ve identified that Lane’s methodology generates six potential cut-offs, three of
which are limiting cut-offs derived as break-evens by assuming that each production
stage is separately constraining the overall production rate. The other three are known
as balancing cut-offs, which occur when two production stages operate at capacity.
Consider the mining and treatment stages of the overall process. For each stage as the
throughput-limiting stage of the overall process, curves of value versus cut-off could be
generated as illustrated schematically in Figure 5.1.
The mining and treatment capacities limit value generation to the two curves,
considering each stage on its own. The values on the curves are the maximum values
at each cut-off, generated by the process stages operating at their capacities. Process
stages can operate below their capacities, in which case the full value potential will not
be achieved, indicated by the two points in Figure 5.1 illustrating one stage operating at
capacity and the other at less than capacity.
When two stages are being considered together, the two curves constrain value
generation potential to the area that is under both curves. The maximum value potential
is defined by the lower of the two curves at any cut-off, as illustrated in Figure 5.2.
As indicated in the figure, the point where the two curves cross is physically where
both stages of the process are operating at their limiting capacities, at which point they
are said to be in balance. The cut-off at this point is the balancing cut-off, which in the

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 77
 ChApter 5 | Lane’s Methodology

Figure 5.1
Value versus cut-off curves for mining and treatment process stages.

Value

Mining capacity Mining capacity

under-utilised fully utilised
Treatment capacity Treatment capacity
fully utilised under-utilised

Cut-off

Mining-limited – mining capacity limits value

potential to this line
Treatment-limited – treatment capacity limits
value potential to this line

Figure 5.2
Value versus cut-off curves for mining and treatment process stages, showing the balancing cut-off.
"Balancing Cut-off"
Both stages are operating at capacity

Together the mining

and treatment capacities
Value

limit value potential to the

lower line at each cut-off

gm gmh gh

Cut-off

Mining-limited
Treatment-limited

case illustrated is the mining–treatment balancing cut-off denoted by gmh. For limiting
cut-offs a single subscript is used to indicate the process stage that is assumed to be
the overall constraint; the double subscript identifies the two process stages that are in
balance for a balancing cut-off.
Figure 5.2 also shows that the mining and treatment-limited cut-offs generate the
maximum values of the mining and treatment-limited curves.12

12. Lane (1988, 1997) does not specifically generate these curves but does derive their formulas with rigorous
mathematics. he then applies the principles of differential calculus to obtain the first derivatives of the value
curve functions, which are the formulas that specify the gradients of the curves. these are set to zero to identify
the maximum, and the resulting equations solved to derive the formulas for the three limiting cut-offs, which,
when used, will maximise the NpV in each case. these same formulas are derived by logical argument rather than
rigorous mathematics in the preceding subsections of this chapter.

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 78
 CHAPTER 5 | Lane’s Methodology

An important feature of the balancing cut-off is that it may deliver maximum value.
As will be seen, the balancing cut-off depends solely on the capacities of the process
stages and the distribution of value, or tonnage and grade versus cut-off relationships
that exist in the material to be mined in the time frame considered. Neither of these
controlling factors is in any way related to price or cost. The balancing cut-off is derived
from physical parameters only, completely independently of financial factors.

General principles for deriving balancing cut-offs

With three process stages, there are three pairs of capacities that may be balanced:
1. mining and treatment (rock and ore), for which the balancing cut-off is gmh
2. treatment and marketing (ore and product), for the balancing cut-off of ghk
3. mining and marketing (rock and product), with a balancing cut-off of gmk.
To find the balancing cut-offs, tonnage and grade versus cut-off tables or curves are
inspected. Over a range of cut-offs, the tables or curves required will show:
•• tonnages of ore above each cut-off
•• average grade of the material above each cut-off
•• average recoverable grade of the material above each cut-off
•• ratio of tonnage of material above each cut-off to total rock
•• recoverable product in the material above each cut-off
•• ratio of recoverable product to total rock.
Some of these would be normal components of a conventional tonnage–grade data set,
whereas others may be thought unusual. The reasons for needing these will become evident
in the following explanations of the derivation of balancing cut-offs. These tonnage and
grade values should be for extracted ore, after ore loss and dilution, and accounting for
mining selectivity, not the in situ resource, as it is the tonnes and grades that are actually
mined and processed that use and fill the capacities of the various production process
stages. These data sets should be generated for the each increment of rock, or mining step,
not for the global resource. Changing tonnage–grade relationships will result in changes
of the balancing cut-offs even if production capacities do not change.
Balancing cut-offs for each pair of stages are obtained by finding the cut-off where
the ratios of each pair of material components – rock, ore and product – are in the
same proportions as the process capacities for each pair of components – mining (M),
treatment (H) and marketing (K).

Mining–treatment balancing cut-off (gmh)

The mining–treatment balancing cut-off will be the cut-off that results in both the mining
and ore processes operating at capacity. This occurs when the ratio of ore (that is, rock
tonnes above cut-off) to total rock equals the ratio of treatment capacity (H) to mining
capacity (M). Using the same assumptions for deriving limiting cut-offs:

H = 2 Mt/a and M = 5 Mt/a

H/M = 2 M / 5 M = 40 per cent

The mining–treatment balancing cut-off (gmh) will therefore be the cut-off that
demarcates the best 40 per cent of the total rock; that is, with the best grade. Using

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 79
 ChApter 5 | Lane’s Methodology

the same tonnage–grade curves as were used for developing Mortimer’s Definition in
Chapter 4, it can be seen in Figure 5.3 that, for a mining step of 8 Mt of rock, the best
40 per cent will be 3.2 Mt. From the curves, this will be generated using a cut-off of 2.3,
and the corresponding grade of the ore above this cut-off will be approximately 5.3.
At this cut-off, generating 3.2 M ore tonnes from 8 Mt of rock and applying the stated
mining and treatment capacities (5 Mt/a and 2 Mt/a respectively), both the ore and rock
tonnages will take 1.6 years to be dealt with. The same durations are required for both
process stages and they are indeed in balance.

Figure 5.3
Tonnage and grade versus cut-off curves illustrating the mining-treatment balancing cut-off.

12 12

Grade of Material above Cut-off

Tonnage above Cut-off (Mt)

10 10

8 8

6 6

4 4

2 40% 2
of 8Mt
0 0
0 1 2 3 4 5 6 7 8 9 10
Cut-off Grade

Tonnes above Cut-off Grade above Cut-off

treatment–marketing balancing cut-off (ghk)

The treatment–marketing balancing cut-off results in both the ore and product
processes operating at capacity. This occurs when the ratio of recovered product to ore
– that is the recoverable grade – equals the ratio of marketing capacity (K) to treatment
capacity (H). Using the same assumptions for deriving limiting cut-offs:

K = 150 000 t/a and H = 2 Mt/a

K/H = 0.15 M / 2 M = 7.5 per cent

The treatment–marketing balancing cut-off (ghk) will therefore deliver a recovered

grade of 7.5 per cent. The recovered grade must be used rather than the head grade or
actual average grade of the ore because it is the recovered product that will be handled
by the product circuit and restricted by its capacity, not the product contained in the
ore. Figure 5.4 shows the same tonnage–grade curves as were used for developing
Mortimer’s Definition in Chapter 4, with the assumed metallurgical recovery of
90 per cent applied to the actual grade to obtain the recoverable grade curve.
It can be seen that the required recoverable grade will be obtained using a cut-off
of approximately 6.7. The head grade will be 8.3 per cent and the corresponding ore
tonnage will be only approximately 800 000 t.

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 80
 CHAPTER 5 | Lane’s Methodology

Figure 5.4
Tonnage and grade versus cut-off curves illustrating the treatment-marketing balancing cut-off.

12 12

Grade of Material above Cut-off

Tonnage above Cut-off (Mt)
10 10

8 8

6 6

4 4

2 2

0 0
0 1 2 3 4 5 6 7 8 9 10
Cut-off Grade

Tonnes above Cut-off Grade above Cut-off

Recoverable Grade above Cut-off

Mining–marketing balancing cut-off (gmk)

The mining–marketing balancing cut-off is where the ratio of recovered product to total
rock equals the ratio of marketing capacity (K) to mining capacity (M).
The ratio of product to the total rock in which it is contained is a grade, but it
is not the grade with which we are most familiar, the grade of product in ore. As
the cut-off applied to a given tonnage of rock is increased, the amount of ore must
decrease. So, too, must the amount of product contained in the ore, but at a slower
rate so that the ratio of product in ore – what we typically call the ore grade – rises
with increasing cut-off. Although it is unusual, it is nevertheless legitimate to derive
a grade value for product contained in rock rather than ore – what we might term
the rock grade. Although the amount of product in the rock component above cut-off
will decrease with greater cut-off, the tonnage of rock remains constant, so the rock
grade will reduce with increasing cut-off. As with the treatment–marketing balancing
cut-off, it is the recovered product that will take up capacity in the product stream of
the process. To derive the mining–marketing balancing cut-off we must therefore, as
for the treatment–marketing balancing cut-off, derive a recovered rock grade using
recovered product.
Using the same assumptions for deriving limiting cut-offs:

K = 150 000 t/a and M = 5 Mt/a

K/M = 0.15 M / 5 M = 3 per cent

Figure 5.5 shows the same tonnage–grade curves used for Mortimer’s Definition
in Chapter 4, with the metallurgical recovery of 90 per cent applied to the contained
product at each cut-off to create the recovered grade in rock curve. As can be seen, this
decreases with increasing cut-off, as expected from the rationale previously described.
The required recoverable rock grade is three per cent, and it can be seen that the
recoverable rock grade is less than this at all cut-offs, including zero. The implication is

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 81
 CHAPTER 5 | Lane’s Methodology

Figure 5.5
Tonnage and grade versus cut-off curves illustrating the mining-marketing balancing cut-off.

12 12

Grade of Material above Cut-off

Tonnage above Cut-off (Mt)
10 10

8 8

6 6

4 4

2 2

0 0
0 1 2 3 4 5 6 7 8 9 10
Cut-off Grade

Tonnes above Cut-off Grade above Cut-off

Recoverable Metal above Cut-off : Total rock

that, even if all the rock in this mining step were classed as ore, mined at the maximum
mining rate and treated, there would not be sufficient product to fully utilise the capacity
of the product stream. One could hypothetically extrapolate the recovered rock grade
curve to the left, increasing with reduced cut-off, and identify a negative cut-off grade
that would deliver this balance. This is mathematically reasonable in the abstract, but it
has no physical meaning. We therefore consider the balancing cut-off to be zero.
To illustrate the generation of a positive mining–marketing cut-off, consider the
situation if K were reduced from 150 000 t/a to, say, 110 000 t/a. The recoverable rock
grade to generate the mining–marketing balance would be 2.2, and as can be seen in
Figure 5.6, gmk would be of the order of 1.3.

Figure 5.6
Tonnage and grade versus cut-off curves illustrating the mining-marketing
balancing cut-off with lower product stream capacity.
12 12
Grade of Material above Cut-off
Tonnage above Cut-off (Mt)

10 10

8 8

6 6

4 4

2 2

0 0
0 1 2 3 4 5 6 7 8 9 10
Cut-off Grade

Tonnes above Cut-off Grade above Cut-off

Recoverable Metal above Cut-off : Total rock

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 82
 CHAPTER 5 | Lane’s Methodology

Completing the Analysis

It was noted that, using Mortimer’s Definition, two numbers are generated and the
larger of the two is selected as the cut-off. Lane’s methodology, a 3D process, generates
six potential cut-off values – three limiting and three balancing – one of which will be
optimal at any point in time. Lane presents a simple two-stage process where the six are
reduced to three, then to the one optimum value to be used.

The process of finding the optimum cut-off

The first stage of reducing the six potential cut-offs to one considers each of the three
pairs of production stages. The optimum cut-off for a pair of stages is the middle-ranked
value of the two limiting cut-offs and the balancing cut-off associated with the two
stages being considered. (The reason why this is so will be illustrated graphically.) With
three pairs of stages to consider, this process results in three intermediate optimum cut-
offs, one for each pair of stages. The final optimum cut-off is then the middle-ranked
value of the three intermediate optimum cut-offs. Upper case ‘G’ is used to denote these
intermediate and final optimum cut-offs. So for:
•• mining and treatment: Gmh = middle-ranked value of gm, gh and gmh
•• treatment and marketing: Ghk = middle-ranked value of gh, gk and ghk
•• mining and marketing: Gmk = middle-ranked value of gm, gk and gmk
•• all three production stages: Gmhk = middle-ranked value of Gmh, Gmk and Ghk.

Completing the analysis – numerically

For the examples developed in the preceding subsections:
•• limiting cut-offs calculated were: gm = 1.10, gh = 3.14, gk = 1.45
•• balancing cut-offs derived were: gmh = 2.3, ghk = 6.7, gmk = 0.0.
The optimum cut-offs for pairs of stages are as follows:
•• Gmh = middle value of gm, gh and gmh = middle value of 1.10, 3.14 and 2.3 = 2.3
•• Ghk = middle value of gh, gk and ghk = middle value of 3.14, 1.45 and 6.7 = 3.14
•• Gmk = middle value of gm, gk and gmk = middle value of 1.10, 1.45 and 0.0 = 1.10.
The optimum cut-off for all stages is then given by:
•• Gmhk = middle value of Gmh, Gmk and Ghk = middle value of 2.3, 3.14 and 1.10 = 2.3.
For this set of circumstances with six potential cut-offs that range from 0.0 to 6.7, the
optimum is 2.3, which happens to be the mining–treatment balancing cut-off gmh.

Completing the analysis – graphically

It is easy to see why using the larger of the two potential cut-offs rule works for applying
Mortimer’s Definition. But why should the somewhat more complex but still simple
mechanical process described reduce the six potential cut-offs to one using Lane’s
methodology? Using the same information, we will now see how the process works
geometrically because of the generalised shapes of the value versus cut-off curves for
each of the three production stages.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 83
 ChApter 5 | Lane’s Methodology

Figure 5.7 shows the graphical derivation of the intermediate optimum cut-off for
mining and treatment process stages.
As shown in Figure 5.2, the mining and treatment-limited cut-offs define the maximum
values of the value versus cut-off curves for the two stages. The balancing cut-off identifies
the cut-off value where the two curves cross, since at this point both stages are operating
at full capacity. For these two stages, the balancing cut-off lies between the two limiting
cut-offs. The general shapes and relativities of the curves must therefore be as shown in
Figure 5.7. The maximum value obtainable at any cut-off is the lower value of the two
curves at that point; the value cannot be any higher as that would imply that a stage
is operating at greater than its capacity. With this configuration, the maximum value
has to be at the balancing cut-off, as shown in Figure 5.2. The geometrical relationships
in these circumstances require that the middle value of the three relevant cut-offs, the
balancing cut-off – gmh – must be the cut-off that delivers maximum value, and hence is
the intermediate optimum cut-off Gmh.
Figure 5.8 shows the graphical derivation of the intermediate optimum cut-off for
treatment and marketing process stages.
Again, the limiting cut-offs define the maximum values of the value versus cut-off
curves for the two process stages, and the balancing cut-off identifies the cut-off value
where the two curves cross, since at this point both stages are operating at full capacity.
For treatment and marketing stages, the balancing cut-off derived is greater than both
the limiting cut-offs. The general shapes and relativities of the curves must therefore be
as shown in Figure 5.8. Again, the maximum value obtainable at any cut-off is the lower
value of the two curves at that point. Although mathematically it might be possible for
the curves to cross in two places – that is, at two different cut-offs – the physical reality of
the balancing cut-off implies that in the special case of these value versus cut-off curves
they can only cross once. The curve symmetries imply that the maximum value of the
curve with the lower maximum value will occur at a cut-off between the balancing cut-

Figure 5.7
Graphical derivation of intermediate optimum cut-off for mining and treatment.
NPV

Gmh

gm gmh gh

0 1 2 3 4 5 6 7 8
Cut-off

Mining-limited Treatment-limited

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 84
 ChApter 5 | Lane’s Methodology

off and the limiting cut-off of the higher value curve.13 The geometrical relationships
in these circumstances require that the middle value of the three relevant cut-offs must
deliver maximum value, but with the configuration shown, the maximum value has to
be at the higher of the two limiting cut-offs – that is, the middle value of the three cut-
offs, gh – and this is therefore the intermediate optimum cut-off Ghk.
Figure 5.9 shows the graphical derivation of the intermediate optimum cut-off for
mining and marketing process stages. The two curves are the same as for these two

Figure 5.8
Graphical derivation of intermediate optimum cut-off for treatment and marketing.
NPV

Ghk

gk gh ghk

0 1 2 3 4 5 6 7 8
Cut-off

Treatment-limited Marketing-limited

Figure 5.9
Graphical derivation of intermediate optimum cut-off for mining and marketing.

gk
NPV

Gmk

gmk gm

0 1 2 3 4 5 6 7 8
Cut-off

Mining-limited Marketing-limited

13. Lane (1988, 1997) explains that it can be shown from the mathematical forms of these curves that they must
be of this geometrical form. the argument in this text is not mathematically rigorous, but qualitatively provides
an explanation of why the geometries of the curves are as they are. Users of the methodology can, however, be
assured of the mathematical rigour underlying the processes used.

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 85
 ChApter 5 | Lane’s Methodology

production stages in the two previous figures illustrating the mining–treatment and
treatment–marketing balancing cut-offs.
It has been noted, when deriving the balancing cut-off for mining and marketing, that
it is not feasible for the two stages to be in balance, given the assumptions used for the
example developed. Therefore, the two value versus cut-off curves do not intersect, at
least within the physically practical range of non-negative cut-offs. It was seen that,
even if the mining process were operating at capacity at a cut-off of zero, there would
be insufficient product in the ore feed to fill the product stream capacity. This implies
that mining is the constraint, which in turn implies that the value is higher if marketing
is the constraint than if mining is the constraint at all cut-offs. This can be seen in
Figure 5.9. It was also indicated, that, although the balancing cut-off is taken to be zero,
by projecting the trends of the tonnage and grade versus cut-off data, the balancing
cut-off could notionally be considered negative. Applying similar logic to geometrical
relationships of the treatment and marketing, with a zero or negative balancing cut-off
and two positive limiting cut-offs, the lower of the limiting cut-offs will be the cut-off
delivering the smaller of the maximum values, which as described will be the mining-
limited cut-off. This is the optimum cut-off for the combined two stages, and is also the
middle value of the three. The mining-limited cut-off gm is the intermediate optimum
cut-off Gmk.
Figure 5.10 shows the final graphical derivation of the optimum cut-off for the
processes of mining, treatment and marketing. The value versus cut-off curves are as
shown in the three preceding figures, with the intermediate optimum cut-offs Gmh, Ghk
and Gmk indicated.
Since each of the three curves plotted shows the maximum value that can be generated
when each process stage is individually the throughput-constraining process, the
maximum value of the overall operation must be the lowest of the three values at every
cut-off. For the case illustrated, this is the treatment curve for cut-offs lower than the
mining–treatment balancing cut-off gmh, and the mining curve for higher cut-offs.

Figure 5.10
Graphical derivation of optimum cut-off for mining, treatment and marketing.
NPV

Gmhk

Gmk Gmh Ghk

0 1 2 3 4 5 6 7 8
Cut-off

Mining-limited Treatment-limited
Marketing-limited

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 86
 CHAPTER 5 | Lane’s Methodology

In this case, gmh is Gmh, and this is also the cut-off that delivers the maximum value, and
therefore is the overall optimum cut-off Gmhk.
It can be seen that this is the middle cut-off of the three intermediate optimum cut-offs.
Considering the geometries involved will indicate why the middle value will be the
overall optimum cut-off delivering maximum value. Similarly, considering the shapes
of the curves and their geometrical relationships will indicate why this will be true for
all possible combinations of the six potential cut-offs generating the three intermediate
cut-offs and one overall optimum cut-off.
It is not essential to derive curves as illustrated to determine optimum cut-offs. Although
Lane (1988, 1997) derives the general mathematical form of these curves, generating the
actual curves does not form part of the process required to find the optimum cut-offs.
This is simply done using cost and price data, as well as tonnage and grade versus cut-
off data that derive actual cut-off values, as illustrated in the previous sections.

Practical Processes to Derive Optimum Cut-offs

At the time when he was developing his methodology, Lane had to work with large
corporate mainframe computers. Now the simple process can be applied by anyone
using modern spreadsheet software, especially for simple single-metal resources. There
are two separate but related processes to determine the short-term cut-off for a single
mining step, increment of rock or short-term planning period and the long-term life-of-
mine cut-off policy (that is, the planned sequence of optimum cut-offs) for all mining
steps and hence for the life of the operation.

Short-term single-period optimum cut-offs

Short-term optimum cut-off derivation for a single mining step, or, for instance, the next
month’s short-term plan, is straightforward. Current cost and price information and the
tonnage and grade versus cut-off relationships for the rock increment to be mined are
used exactly as in the previous example.
While the process and calculations are simple, obtaining the required data may not be.
It may be necessary to rework reports to reallocate costs to their correct classifications of
fixed and variable, and to the mining of rock, ore treatment and marketing of product.
Capital cost records may need to be reviewed to allocate capital to project and sustaining
categories, the latter also assigned to material types in the same way as operating costs.
Mining, treatment and marketing capacities should be known. The first two do not
usually present problems for an open pit operation; however, for an underground
operation, identifying the maximum mining rate – the rate at which mineralised
material (rock) is exposed to allow subsequent stoping in what is ultimately ore – may
not be easy to define. It should also be noted that this is not necessarily the rate at which
access development is actually scheduled to advance, but the rate at which it could be
scheduled, given the existing mining fleet, labour force and system capacities. The same
may, however, apply to the stripping rate in an open pit.
The maximum marketing rate may be poorly understood in either type of operation.
As noted earlier, it is defined by the process that limits the throughput of product
wherever it occurs in the overall treatment and selling process. Depending on the type
of product, the constraint may apply to another material containing the product such

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 87
 CHAPTER 5 | Lane’s Methodology

as a concentrate for a base metal; in this case the concentrate tonnage constraint must
be converted to a metal product constraint using the concentrate grade to effect the
conversion.
Long-term plans may need to be examined to obtain the NPV and current discount
rate used to determine the opportunity cost (F) required for the treatment and market-
limited cut-offs. For a short-term analysis, it is often satisfactory to assume that economic
conditions are not changing in the future; that is, that dV/dt is zero, so the first term of
the opportunity cost formula – the simple product of discount rate and NPV – will be a
sufficient estimate of the opportunity cost.
Tonnage and grade versus cut-off relationships must also be obtained for the next
increment of rock to be treated. As indicated earlier, Lane’s methodology as originally
formulated deals with increments of rock, not time periods (though the underlying
principles can be applied to time periods as well). If the intention is to develop the
optimum cut-off for the next short-term planning increment – say, a week or a month
– the analysis might start with the assumed rock to be mined during that period;
however, the application of Lane’s methodology may result in a different duration
when the optimum cut-off is applied using the existing mining, treatment and
marketing constraints.
If a short-term Lane-style optimum cut-off is to be calculated regularly, technical staff
should not have to reallocate costs in order to do so. Regardless of how the costs might
be reported for other purposes, they should then also be classified and reported in the
categories needed to correctly calculate Lane-style limiting cut-offs. Similarly, geological
operating procedures must be put in place to generate the required tonnage and grade
versus cut-off data for the next increment of rock, whatever size that might be, on a
regular basis.

Long-term multiperiod optimum cut-offs

For determining the long-term optimum cut-off policy, similar processes must be
conducted for each increment of rock or mining step. As indicated, Lane’s simple
methodology assumes that mining sequences have already been determined, and the
increments of rock are therefore mined successively with no overlaps. The pushback or
cut-back stages used in normal open pit mine planning are irrelevant – the rock in each
mining step may be made up of rock mined in two or more pushback stages in the same
time frame.
Although the Lane-style analysis deals with rock increments, time periods in an
existing schedule may be used to define the rock in each mining step, but as with the
short-term process, the final durations for the mining steps may be different from those
used initially to generate the steps.
Calculations for each mining step will require the same data as for the single-
increment short-term process. Costs, prices and capacities for the production stages
may vary with each mining step, but not within a step. As before, costs must be
provided according to Lane’s definitions of what is included and excluded with each
cost category.
The NPV for each step will not be provided from an externally derived mine plan as
a piece of input data. Rather, the derivation of the optimum cut-off policy, which also
determines the optimum duration of each step, is in fact determining the optimum

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 88
 CHAPTER 5 | Lane’s Methodology

long-term plan. The NPV values must therefore come from this same process. As a
general rule a set of iterative calculations will be required, working forwards mining
step by step in each iteration. An initial feasible plan must be generated by some
means.
Forecast economic conditions – prices and costs, etc – are applied to generate NPV
(V). Those conditions are then moved forward by a year to derive an estimate of
the value for V if the operation were deferred by a year. These two values of V are
used to estimate dV/dt and hence the opportunity cost (F). The optimum cut-off is
then calculated for the first iteration, potentially changing the duration of any or all
mining steps, hence changing the remaining NPV for each step. The process must
then be repeated, using the resulting NPV values to derive values of F to be used
in the treatment and marketing-limited cut-offs in the next iteration. This process is
repeated until it has converged satisfactorily to a stable set of cut-offs and mining step
durations over the life of the operation.
If prices, costs and processing capacities change over time, the situation becomes more
complex. We determine the average prices, costs and capacities to apply to each mining
step; however, as successive iterations of the cut-off optimising process change the
durations of mining steps, the average prices, costs and capacities for each step will also
change. The process described is iterative to account for the fact that changing durations
will change the remaining NPVs and hence opportunity costs for each mining step. It is
therefore possible that other altered parameters, such as various physical relationships
that are cut-off dependent, may be accounted for by this iterative process.
In the special case where prices and costs do not shift over time, changes in capacities
are aligned with mining step boundaries rather than time14 and there is no stockpiling
of lower grade material, it is theoretically possible to complete the Lane-style cut-off
optimisation in one calculation, working backwards from the last mining step. Since
these conditions do not change over time, it does not matter when each mining step
occurs: the economic conditions will be the same regardless. It is therefore reasonable
to derive the optimum cut-off for the last mining step, generate its NPV, use that to
determine the opportunity cost to use for the second-last step, optimise the second-last
step and so on, back to the first step. In the more usual practical case where there are
changes in prices and costs over time and stockpiling of low-grade material for treatment
later, the duration of each step affects the timing and hence the economic conditions for
the last step. The end point is unknown and the single-step working-backwards process
is not applicable; the iterative working-forwards process is necessary.
If future economic conditions are predicted to be stable, the iterative process will
usually converge reasonably quickly to a stable plan. If there are significant changes
predicted, it is possible that the process will not converge, but rather oscillate between
a couple of cases that flip iterations from the one to the other. The study team may then
have to manually intervene to specify an appropriate final plan.

14. If mining steps have been delineated so that each is homogeneous, aligning capacity changes with mining step
boundaries is a more rational approach than arbitrarily specifying a date that may fall within a mining step, since
production and economic conditions are assumed to be unchanging throughout a step, but may change from
step to step.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 89
 CHAPTER 5 | Lane’s Methodology

The Effect of Price and Cost Changes on Cut-offs

For many associated with the mining industry, cut-off and break-even are synonymous
terms. Break-even generally varies in direct proportion to cost changes, and in inverse
proportion to price changes. If prices double or costs halve, the break-even will halve. If
the break-even is assumed to be the cut-off, the cut-off has halved; however, if cut-offs
are derived to be Lane-style optimum cut-offs, the aim of which is to maximise NPV,
what impact might price and cost changes have?
We have seen that a Lane-style analysis generates six potential cut-offs for each mining
step evaluated, one of which will be the optimum. Three of these are limiting cut-offs
that account for prices and costs, but although these are break-even calculations, the
responses to price and cost changes may not be what is expected.
As already derived, the mining, treatment and marketing cut-off formulas are:
•• gm = h/(p – k)y
•• gh = {h + (f+F)/H}/(p – k)y
•• gk = h/{p – k – (f+F)/K}y
If there is a price rise, p increases. Since p is in the denominator of all formulas, the
three cut-offs will reduce. Because the product-related costs are deducted from prices
the change in cut-off will be more rapid than an inverse-proportional change.15 This is
the primary effect of a price change and the only effect for the mining-limited cut-off.
For treatment and marketing-limited cut-offs, there is a secondary effect. A price
rise over the life of the operation will also increase its value, so the NPV and hence
the opportunity cost (F) will increase. For the treatment-limited cut-off, F is in the
numerator, and an increase in F will increase the cut-off. For the marketing-limited cut-
off, F is in the denominator as a deduction from price. The increase in F will therefore
reduce the denominator and, as for the treatment-limited cut-off, increase the cut-off.
This secondary effect will offset the primary effect, at least to some extent, but it may
more than compensate for the primary effect, so that a price increase may result in an
increase in the treatment or marketing-limited cut-offs.
The net result of the primary and secondary effects of a price rise is therefore that the
mining-limited cut-off will reduce, but the treatment- and marketing-limited cut-offs
may reduce or increase. As a minimum, the movement in the latter two cut-offs will be
less than is normally expected for a classical break-even.
Cost reductions will have the same effect. This is, of course, to be expected: both a price
rise and a cost reduction will increase the margin, and qualitatively, that should lead to
the same general outcomes, regardless of how that margin increase was obtained.
The exact mechanism, of course, will be somewhat different. If there is a cost reduction,
any or all of the cash variable and fixed cost components h, k and f will reduce. Reduced
components in the formula numerators will reduce the corresponding cut-offs, while
reduced costs in the denominators, where they are deducted from the price, will increase
the denominators and reduce the cut-offs. Regardless of which cost components reduce,
the primary effect of the cost reduction will be a reduction in the three limiting cut-offs.

15. If cut-off were plotted as a function of price, a simple break-even cut-off would approach infinity asymptotically
as the price approached zero. With product-related costs deducted from price in the denominator of the limiting
cut-off formulas, the cut-off will approach infinity as the price approaches the product-related costs, not zero.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 90
 CHAPTER 5 | Lane’s Methodology

As for a price increase, a cost reduction over the life of the operation will also increase the
operation’s value, so the NPV and opportunity cost (F) will rise. As for a price increase,
there is the secondary effect of a cost reduction that will increase the treatment- and
marketing-limited cut-offs. Similarly, the net result of the primary and secondary effects
of a cost reduction will cause mining-limited cut-off to decrease, but the treatment- and
marketing-limited cut-offs may increase or decrease.
The other three Lane cut-offs are balancing cut-offs, which are totally independent of
prices and costs. If one of these is the optimum, increases or decreases in the limiting cut-
offs may not result in any change to the cut-off, unless the relativities of all six potential
cut-offs change so that one of the other six cut-offs becomes the optimum cut-off.
It is therefore impossible to state the effect of price or cost changes on optimum cut-
off without completing a full analysis.16 Price rises and cost reductions may result in
optimum cut-off reductions or increases, or no change at all, and the effects may vary.
The author suggests that it is likely that, in a long-life operation with a relatively
consistent grade distribution, debottlenecking of the mining and treatment processes
over time will result in a balance being established between at least two of the
production processes, and that the balancing cut-off thereby created will then remain as
the optimum cut-off regardless of price and cost changes over time. To put this another
way, if constraints do not change, prices and costs can change significantly but the
optimum (balancing) cut-off will not change. This is quite counter-intuitive for those
who have always believed that cut-off and break-even are synonymous – that if the price
goes up, the cut-off has to go down, and vice versa.
If not rigorous proof, consider as a logical explanation the curves in Figure 5.2. If we
postulate that we spend additional capital to increase the treatment capacity, the value
versus cut-off curve for treatment will rise. The intersection point will move up the
mining curve, and the balancing cut-off will therefore reduce; however, because the
mining value versus cut-off curve is flattening with reducing cut-off, the additional value
for each increment of treatment capacity will lessen. At the point where the balancing
cut-off is the same as the mining-limited cut-off – that is, where the treatment curve has
moved to the point where it passes through the maximum value of the mining curve,
and the mining-limited cut-off becomes the optimum cut-off for the two stages shown
– there is no additional value gained from investing in additional treatment capacity.
In reality, the treatment plant expansion would not be pushed to that limit – the
relationship between additional value and the cost of expansion would show that
the expansion would become uneconomic before the treatment capacity is increased
to the point where the balancing cut-off is the mining-limited cut-off. The optimum
cut-off would therefore continue to be the balancing cut-off, greater than the mining-
limited cut-off. The further implication is that, if the optimum cut-off during longer-

16. From time to time one will see an article in the technical press suggesting that, when prices are high, cut-offs
should be increased rather than reduced to increase output and cash in on the higher price. Simple examples
may be presented, sometimes almost apologetically, to show the effects of two cut-off policies, demonstrating
that an increased or elevated cut-off may generate better outcomes than a break-even cut-off.
The author suggests these writers have intuited that the best value to use as the cut-off will not necessarily be
the break-even (which of course will reduce when prices increase), but they do not have sufficient understanding
of the issues to be quantitative rather than qualitative. Lane’s formulas confirm the logic of this possibility, and
the complete process provides a quantitative methodology to determine the best strategy.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 91
 CHAPTER 5 | Lane’s Methodology

term steady-state operations is a limiting cut-off, the capacities of the various stages
are mismatched and at least one stage has been oversized.17 Unless another production
stage can be debottlenecked to add further value, excess capital will have been spent on
the oversized plant stage and the value destroyed.
If the optimum cut-off is a balancing cut-off, it is illogical to change the cut-off – for
example, because changes in costs and prices have changed the break-even grade
– unless one of the capacities in balance is also altered. This will require a conscious
decision either to operate below the capacity of one of the processes involved or spend
capital to increase a capacity. If reducing a capacity or, more accurately, operating below
capacity to accommodate a change in cut-off from a balancing optimum to a break-
even, value will by definition be destroyed. We will have moved, for example, to one
of the underutilised capacity situations illustrated in Figure 5.1. Spending capital to
increase capacity must be justified by an improvement in value resulting from using the
additional capacity, or again, value will be destroyed.

Complicating Factors
There are a number of real-world situations that may make the simple Lane-style
cut-off policy derivation more difficult. Some, with varying degrees of ingenuity, can
be handled by the iterative processes described, but others may require a change to
complex Lane or full strategy optimisation, which will be described in the next chapter.

Costs, prices and capacities varying over time

As noted, if prices and costs change over time, the situation becomes more complex.
Calculations identify the averages for the various mining steps as their durations change,
but in principle these can be handled by the iterative process described. Processing
capacities shifting over time may likewise be handled iteratively if the rates and times
are specified, a feature of simple Lane; however, this author advocates that, unless
capacity changes are linked to some external factor, such as predicted changes in prices
or costs, it is likely that they will be prompted by changes in the nature of the material
dealt with. This would suggest that the timing of such capacity changes should not be
specified as dates, but rather by reference to the boundaries between mining steps.
We can generate optimum cut-off policies for different sets of capacities – for example,
should we expand the mill from one capacity to another? Mining or treating different
quantities because of changes in capacities will change the value of the operation. Such
evaluations are common, but they are typically evaluated using the same cut-offs for
each option. But different capacities will potentially change the optimum cut-off policy
and hence impact on the value of the operation. The value gained from capital spent to
change capacities can only be identified by comparing the values of the optimum mine
plans, including their different cut-off policies, for each set of capacities. Simply changing
tonnes mined or treated and failing to account for these changes in cut-off policies may

17. This will often be a practical outcome of an operation transitioning from open pit to underground – the
treatment plant on surface will have been constructed to handle a relatively large tonnage at a low grade.
With higher costs per tonne of ore underground, and given the practical limitations on the rate at which the
underground operation can be opened up, it may be impossible to mine at a rate sufficient to fill the mill. It
may never be practical or economic to increase the mining rate (the rate of access development that exposes
the rock) and the optimum cut-off will therefore be and remain the mining-limited cut-off.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 92
 CHAPTER 5 | Lane’s Methodology

give rise to a poor investment decision if the value of an optimal plan for one set of
capacities is compared to a suboptimal plan for a different set of capacities. Capacities
must be appraised in separate Lane-style cut-off optimisation evaluations, and the set
that generates the maximum value, using its optimised cut-off policy, identified.
Simple Lane does not optimise system capacities – that is a task for complex Lane or
strategy optimisation. We may evaluate a number of options – for example, for mining
and treatment rate capacities – but unless we employ suitable optimisation methods,
there is no guarantee that by simply evaluating a few alternatives, the option with
optimum capacities has been included in the cases evaluated. So while the optimum cut-
off policies for the capacities evaluated may have been identified, the overall optimum
strategy for both capacities and cut-offs may not have been.

Capacities varying with cut-off

The capacities of the three stages of the production process are integral to a Lane-style
cut-off optimisation. They are part of the derivation of all three balancing cut-offs and
two of the three limiting cut-offs. It is possible for a capacity to be a function of the cut-
off. This will introduce complexities that might or might not be able to be dealt with
using a simple Lane analysis.
For example, in an open pit mine, the mining capacity (M) may be the truck haulage
capacity. If the haulage distances vary for ore and waste, truck productivity and the
rock-handling capacity of the fleet will depend on the proportional split of rock into ore
and waste, which is defined by and dependent on the cut-off.18 Haulage distances to
stockpiles and allocation of loading and trucking fleets to stockpile reclaim introduces
further complexities. The mining capacity, M, only comes into two of the three balancing
cut-offs: the rock–ore and rock–product balances. It is possible to derive the mining
capacity for each cut-off in the tonnage and grade versus cut-off data tables for each
mining step and include it in those tables. Instead of comparing the rock–ore and rock–
product ratios at each cut-off with the capacities’ constant ratio, the comparisons would
be with capacity ratios varying with cut-off within the data tables. Graphically, rather
than identifying the cut-off at which the material type ratio (expressed as a grade) crosses
a grade value (representing the ratio of capacities) on the grade axis, the intersection
would be with another line plotting the capacity ratio as a function of cut-off.
In a similar way, the treatment capacity (H) may be the milling capacity, which is
the product of the maximum plant operating hours and the milling rate of the ore. The
milling rate will be a function of the ore hardness, which may be a function of the head
grade that in turn depends on the cut-off. The two balancing cut-offs, including the
treatment capacity, could be accounted for similarly to the mining capacity; however,
the treatment capacity is also included in the treatment-limited break-even cut-off
formula. If the relationship between cut-off and capacity can be expressed as a formula,
this could be substituted, generating a more complex formula that potentially requires
solving numerically. If the capacity cannot be expressed as a function of the cut-off, it
would be necessary make use of the tabulations of tonnages and grades versus cut-off.

18. This effect may often be small, to the point of being insignificant for the evaluation, but the author has encountered
a situation where the acid-generating potential of the pit waste created a need to mine additional acid-neutralising
waste from elsewhere. The cut-off selected therefore had a significant impact on the rock-handling capacity of the
fleet in the main pit where a simple Lane-style analysis might have been applied.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 93
 CHAPTER 5 | Lane’s Methodology

One would derive the treatment capacity for the head grade of the material above each
cut-off in the table, and then calculate a hypothetical limiting cut-off using each capacity.
The treatment-limiting cut-off would be the cut-off where the hypothetical cut-off is the
same as the initial cut-off in the table used to calculate it.

Polymetallic orebodies
Lane (1988, 1997) presents formulas for polymetallic or multiproduct orebodies that
have (p–k)y terms for each metal component, introducing multiple grade/cut-off grade
dimensions into the analysis. Lane also indicates that search techniques may be needed
for the best solution, rather than the analytical and graphical processes described for the
single-product case, because of the increased number of dimensions. This is classified as
complex Lane or strategy optimisation.
Dollar value or metal equivalent grades use19 is often suggested as a way around
this problem. The product effectively becomes dollars or equivalent metal tonnes, and
this may be satisfactory for the mining and treatment stages of the production process
where costs and capacities are associated with rock and ore tonnages. For the marketing
stage dealing with product, it will often be difficult to express the marketing constraint
(K) when grade is a metal equivalent. The capacity must be a single value derived by
combining the capacities for each product stream in a way that represents the capacity
of the single equivalent metal process stream. There is also the problem of determining
a product-related variable cost (k) when the product is expressed as a metal equivalent
comprising two or more constituents.
Using an equivalent dollar value may remove some of these problems, depending on
how the dollar equivalent has been derived. If generated as a net smelter return (NSR) type
of value, the product represented by the grade is effectively net dollars after recoveries and
product-related costs, so the (p–k)y calculations are already included in the dollar value.
There is still, however, the problem of identifying a capacity (K) for the product expressed
as net dollars per period of time for the mining–marketing and treatment–marketing
balancing cut-off derivations as well as the marketing-limited cut-off calculation.
In some cases (for example, where precious metals are contained within a base metal
concentrate), it may be feasible to revert to the base metal component when dealing
with the three cut-offs that include product considerations. We would then have some
components dealing with dollar value grades and others dealing with metal grades.
Depending on the situation, it might be feasible to specify cut-offs appropriate to both
dollar values and metal grades, with material classified as ore satisfying both criterion.
There is no theoretical justification for this – it may be a good solution in some cases,
and not apply in others. It may also require the use of software that can identify the
optimum blend of material that satisfies both criteria, but this is well beyond the limits
of the simple Lane cut-off optimisation process.

Multiple mining areas, mines and mills

A number of mines feeding one mill can derive consolidated tonnage and grade versus
cut-off relationships to find balancing cut-offs; however, Lane’s methodology, within
the bounds of simple Lane, is based on an underlying assumption of a single rock stream

19. These are discussed in more detail in Part 2, Chapter 9.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 94
 CHAPTER 5 | Lane’s Methodology

passing through one set of rock mining, ore treating and product marketing facilities.
Multiple mines and/or mills will almost certainly have capacity constraints applied
to different components of the various rock, ore and product streams. Furthermore,
different cost structures in parallel parts of the operations may invalidate the derivation
of Lane-style limiting cut-offs for the combined material flow.
It was noted in Chapter 3 that the cost per unit of ore may vary with location
in underground mines, leading to different cut-offs throughout the mine. Lane’s
methodology does not handle this; however, it may be feasible to account for these
cost differentials in dollar value grade descriptors.20 Lane’s methodology in its simple
form applies to a prespecified stream of rock. In other words, the mines schedule has
been fixed before the cut-off optimisation begins. With multiple mines and/or mills in
the analysis, the mix of materials from different sources to various destinations might
change for the best outcome. Lane’s methodology does not handle this easily. Searching
through a range of mixes may be the only way to arrive at a solution, and complex
calculations of multiple options or specialist optimisation software may be needed.

Open pit versus underground

Lane’s methodology is relatively easy to apply in an open pit mine with a single product.
For a specified pit size, the mining sequence can be specified,21 and the optimum cut-off
will then be dependent on capacities, geology and economics. There is an assumption
that the prespecified mining sequence is independent of cut-off, and this is often the case
in reality. In addition, today’s ore is a subset that is separated out of today’s rock, as the
latter is mined.
In an underground operation, however, the cut-off defines the size and location of
the various ore lenses that will make up the mine; hence, the mining sequence may be
highly dependent on cut-off. Today’s rock (that is, access development) exposes ore
that may be mined at any time in the future, but not today, while today’s ore was made
available by yesterday’s development. Furthermore, today’s stoping may sterilise sub-
cut-off material that might otherwise have become payable at a time in the future.
There are some styles of underground mining that are readily amenable to applying
Lane’s methodology. In general, this will occur where there is a single production front
advancing in one direction, such as downdip or along strike, with access development
maintaining a lead ahead of production and no return to extract additional ore from
parts of the resource that it has already moved through. This effectively establishes a
single fixed mining sequence. Although rock mining and ore treatment are temporally
and spatially separated and ore is not merely separated out of mined rock, as for an open
pit, Lane’s assumptions apply. For example, a sublevel caving (SLC) operation has a
production front that is moving downdip, but cannot obviously advance faster than the
development rate of new sublevels. Nor generally will it be desirable for the production
front to be advancing slower than the development, as this will lead to increased costs
– either from a time value of money consideration with development being done before
it is required, or from the need to rehabilitate development that has been standing open

20. These are discussed in more detail in Part 2, Chapter 9.

21. The specified sequence may not, however, be optimal, but simple Lane does not optimise the mining sequence.
Some of the commercially available pit optimisation software applications do have modules – often optional at
extra cost – that derive an optimal or near-optimal mining sequence.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 95
 CHAPTER 5 | Lane’s Methodology

unnecessarily, often in a highly stressed environment. There is therefore a clear balance

between the rates at which mining is exposing new rock and treatment is producing ore
from the SLC.
Anecdotal evidence suggests that it is not uncommon to find in such cases that a break-
even cut-off is calculated, a suitable ore production rate is nominated for the resulting
orebody and the development rate is then specified to achieve the same vertical advance
rate, maintaining the required balance between rock mining and ore production. This
makes the calculated break-even the de facto mining–treatment balancing cut-off also. It
may be that a better solution is to develop the decline access at maximum rate, thereby
raising the balancing cut-off and working through the resource faster, but potentially
generating higher grades earlier and increasing the value.
Where there are multiple independent mining areas, or in cases where the mining
sequence becomes dependent on the size of the orebody by the cut-off grade applied,
Lane’s simple methodology may cease to be applicable and more complex optimisation
methods required.
Similarly, costs that may vary with cut-off, such as development costs per tonne of ore
in an underground mine, may make Lane’s methodology unsuitable, but in many cases
this may be adequately handled in the iterative calculations required to derive a life-of-
mine optimum cut-off policy.

Mining method constraints

Depending on the nature of the mineralisation, cut-offs may need to be applied to
mining blocks rather than to mine-wide tonnage increments. For example, the geological
conditions in some open stoping operations may mean that, once the stoping limits have
been established, they cannot be changed without causing operational problems. The
production schedule may then change to a number of mining blocks being mined in
parallel, but starting and ending at different times. This type of requirement is similar
to the stages used in open pit mine plans, but whereas the open pit plan is rearranged
into sequential steps, this may be unfeasible in an underground operation. Parallel ore
streams are not handled by the simple form of Lane’s methodology.

Stockpiles
Stockpiling of mineralisation with a grade below the current cut-off is often viable with
an open pit,22 with stockpiled material becoming additional rock that can be reclassified
as ore along with the rock that is freshly mined at a later date. There are costs associated
with building, maintaining and reclaiming a stockpile.
Difficulties in applying Lane’s methodology may arise when mining capacities need
to be considered. Drilling and blasting capacity, and perhaps that of the in-pit loading
fleet, will impose a limit on the amount of new rock mined; however, if the same trucks
are used to haul both newly mined rock from the pit to waste dumps, stockpiles and
treatment plant and reclaimed stockpile material to the treatment plant, the mining

22. Stockpiling of underground material is less common – we would not usually incur the costs of mining material
simply to put it on a stockpile; however, where the mine can produce more potential ore than the mill can handle
and there is a need to mine through lower-grade material to reach higher grades, production and stockpiling of
that lower-grade material may be appropriate if separation of material of different grades is feasible.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 96
 CHAPTER 5 | Lane’s Methodology

capacity may be a function of the cut-off. This may depend on the tonnage and grade
versus cut-off relationships in both newly mined and stockpiled material and haulage
rates associated with each. The problem is not insurmountable: there is no reason why
the tonnage and grade versus cut-off curves or tables that would be used to find the two
balancing cut-offs cannot include mining capacities that also vary with cut-off rather
than remain constant, as described earlier.
Reclaimed stockpile material is ore, so the cost of reclaiming it is an ore-related
treatment cost, even if the reclaim is being done by the rock-mining fleet and recorded
as mining costs in the company’s cost reports.
Lane (1988, 1997) presents a formula for calculating the stockpile break-even grade.
It is similar to Lane’s treatment-limited cut-off formula, but includes the cost of reclaim
as part of the cost per tonne of ore. The recovery term (y) is for the stockpiled material,
which might suffer some degradation during the years it is stockpiled. This grade is
the lower boundary of material that should be stockpiled in any year, and represents
the lowest grade that could be economically treated at the end of the operation’s life.
The upper bound will, of course, be the cut-off applied to run-of-mine material. The
construction of the stockpile is therefore easily defined.
However, the optimum stockpile withdrawal strategy cannot be determined in the
same way as newly mined material using Lane’s methodology without stockpiling, as
described earlier in this chapter. Part of the reason is that, as indicated previously, simple
Lane assumes a single stream of mined material delivered in sequential steps, from which
ore is separated. The details of the material in each mining step are known before the
cut-off optimisation process begins. Stockpiling not only adds a second potential source
of ore, but the tonnage is not fixed – it can fluctuate over the life of the mining operation
– so the tonnage and its grade distribution are functions of the cut-off policy applied
to the newly mined material. Lane describes briefly how an appropriate algorithm will
draw material for treatment from both newly mined material and stockpiles, such that
the net contribution from the lowest grade material from each is the same. The capacities
for mining, treatment and marketing – that is, for handling rock, ore and product –
must also be balanced when dealing with these components of both the newly mined
material and stockpiled material. These more complex algorithms fall within the realm
of complex Lane or strategy optimisation, though some of the issues identified may be
handled by the iterative processes required for simple Lane.

Summary
Lane’s methodology takes account of costs and prices, the grade distribution in the rock
mass and the capacities of the various stages of the production process. In principle,
Lane’s methodology can be applied to optimise any strategic decisions, including cut-
off. To develop an understanding of the parameters that influence cut-off derivation,
we refer to simple Lane, which seeks to optimise cut-offs given a specified size of mine,
mining sequence and capacities for dealing with rock, ore and product. The explicit goal
of a Lane-style cut-off optimisation is maximising NPV; however, there is no guarantee
that the maximum NPV will be a positive value.
Six potential cut-offs are derived for each sequential increment of rock (or mining
step). These increments are the constants in the analysis. The optimum cut-off policy will
determine the duration of each mining step and hence the overall life of the operation.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 97
 CHAPTER 5 | Lane’s Methodology

Changing the assumptions about costs, prices or capacities may change the duration
and overall life.
Three of the six cut-offs are limiting cut-offs, derived by assuming that the processes
dealing with rock (identified by Lane as mining), ore (identified as treatment) and
product (identified as marketing) are each independently the constraint on throughput.
Limiting cut-offs are derived by break-even formulas. Since each increment of rock must
be mined regardless of the amount of rock that is classified as ore, the mining variable
costs will be the same for all possible cut-offs. Costs common to all options will not
impact on the cut-off decision and do not need to be included in the analysis. Mining
variable costs such as drilling, blasting, loading and hauling do not enter into any of the
limiting cut-off calculations.
The mining-limited cut-off calculation includes only ore- and product-related variable
costs. As well as these costs, the treatment- and marketing-limited cut-offs also include
fixed costs and opportunity cost. The latter is the time value of money cost of deferring
receipt of the NPV of the rest of the operation by treating additional material from the
mining step under consideration as ore, as well as accounting for the impact of any
changes in economic conditions in the period within which the displaced material
would be produced. When the ore or product stream is assumed to be the capacity-
limiting process, any additional material classed as ore from the rock increment will
extend the mine life and hence incur additional fixed costs and opportunity cost, which
must be paid for by the grade of the additional material processed.
The other three cut-offs are balancing cut-offs. These are determined for each of the
three pairs of two of the production stages, and ensure that the two stages in each case
are operating at their capacity limits. These cut-offs are physical functions of the grade
distribution in the rock in each mining step and the capacities of the production stages
considered. They are in no way defined by costs and prices.
Having generated all six potential cut-offs, there is a basic two-stage process to reduce
the six, firstly to three and finally to the one optimum cut-off to be used.
For determining a short-term single-period optimum cut-off, the process derives the
necessary cost data from mine cost accounting systems, though it may require effort to
ensure that costs are attributed to the correct quantities. The opportunity cost is usually
extracted from an existing long-term plan. For determining a long-term optimum cut-off
policy – the sequence of optimum cut-offs over time – an iterative process uses NPVs
remaining at the end of each mining step in one iteration to determine the opportunity
costs for the next iteration. This process normally converges to a stable cut-off policy,
though there is no guarantee.
It is likely that in a long-life operation with a consistent grade distribution over time,
debottlenecking of the mining and treatment processes will result in a balance between at
least two of the production processes, and the balancing cut-off created will then remain
as the optimum cut-off regardless of price and cost changes over time. If constraints do
not change, prices and costs can alter significantly, but the optimum (balancing) cut-off
will not change. If the optimum cut-off is a balancing cut-off, it is not logical to change it
(for example, because changes in costs and prices have changed the break-even grade)
unless one of the capacities in balance is also changed. This must be justified by an
evaluation of the change’s impact on value.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 98
 CHAPTER 5 | Lane’s Methodology

In a single-metal open pit, Lane’s theories and methods permit analytical optimisation
of the cut-off policy for a given mining sequence. There are no simple analytical
techniques available to optimise concurrently both the schedule and the cut-off policy.23
These comments may also prove true underground, but only for some mining methods;
for example, sublevel caving.
For such issues as multiple mines, ore sources or ore types; polymetallic orebodies; and
complex stockpiling requirements, analytical techniques may be inadequate. Exhaustive
search techniques or specialist software may be required to optimise the overall plan.
This will be the key focus of the following chapter. In many cases, complex issues can
be dealt with by the iterative processes used to determine an optimum cut-off policy.
In the simplest analyses, only the opportunity cost is varied from iteration to iteration. In
reality, every component accounted for may change from one iteration to another, with
values in a later iteration being derived from the outcomes of the preceding iteration.
Lane’s methodology represents a major advance in cut-off theory. The insights it
provides help to explain many of the sometimes counter-intuitive outcomes of a full
strategy optimisation. In particular, the concept of opportunity cost – often referred to
in discussions of strategy optimisation – is rigorously accounted for to indicate to what
extent future production can be deferred to immediately treat additional material as
ore. In addition, the concept of the balancing cut-off is introduced in cut-off theory. This
shows that it is feasible for the optimum cut-off to be purely a function of the physical
features of the resource and the operation’s infrastructure – the grade distribution and
the capacities of the various stages of the mining and processing operations – and be
unaffected by price and cost changes.

23. The key words here are simple and analytical, implying analyses that could be done by a competent engineer using
spreadsheet software, for example. More complex techniques that could potentially optimise concurrently both
the schedule and the cut-off policy, such as dynamic programming and linear and mixed integer programming,
will be discussed in Part 2.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 99
 HOME

Chapter 6

Mine Strategy Optimisation –

Accounting for Everything

Introductory Comments
In the last three chapters, the focus has been on determining the cut-off grade only. We
have identified how progressively more of the assumed givens may be accounted for
in the analysis, with potentially different cut-offs resulting from both the differences
between the cut-off derivation processes and the contrasting goals in the processes. We
have considered:
•• Break-even analysis – a one-dimensional (1D) process, with the implicit goal of ensuring
that every tonne classed as ore pays for itself, whatever that might be taken to mean.
•• Application of Mortimer’s Definition – a two-dimensional (2D) process, accounting
for everything in the break-even analysis plus the nature of the mineralisation as
described by the tonnage and grade versus cut-off relationships. The explicit goals
ensure that every tonne classed as ore pays for itself and the average grade of ore
treated delivers a minimum profit per tonne.
•• (Simple) Lane’s methodology – a three-dimensional (3D) process, accounting for
everything in Mortimer’s Definition plus the capacities of the components of the
overall production process handling rock, ore and product. It has the explicit goal of
maximising NPV.
For all of these methodologies, there are two situations that are common:
1. An implicit assumption that everything other than the cut-off is fixed. The mining
plan, which includes sequences, pushback stages, mining rates and treatment rates,
has been specified or previously developed. It is then necessary to specify what

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 101
 CHAPTER 6 | Mine Strategy Optimisation

material will be classed as ore and classed as waste. The cut-off is determined to
effect this discrimination. (For this discussion, stockpiled material may be seen as ore
for which treatment has been deferred.)
2. A mining method is assumed and, given the expected costs of mining by that method,
a break-even cut-off is derived. This is applied to delineate the boundaries of the ore,
for which mine designs and schedules are then prepared, taking account of mining
and processing limits that are arbitrarily specified.
The first is more common for an open pit operation, while the second is typical of an
underground operation. Regardless of the operation, the cut-off derivation is usually
separate from other mine planning, design and scheduling processes. This stems from a
misconception that the cut-off is intrinsic to the operation or the orebody, or is a different
class of decision from which we select mining rates, treatment rates, mining sequences
and so on. This is borne out by the common practice in the mining industry worldwide
to recalculate the cut-off using latest cost information and price forecasts, typically using
break-even calculation methods. This is usually done at least annually to prepare the end-
of-year Ore Reserves report. The cut-off is perceived as having to be accepted, driven by
outside forces (in particular, product prices), over which the company has no control.
This is not saying that cut-offs should not be reviewed on a regular basis – they should
be. What is being challenged is the common belief that cut-offs are outside the company’s
control. This leads to the situation where, rather than actively selecting the best cut-offs
to weather the storms of change in economic conditions, companies allow themselves to
be blown about uncontrollably by passive acceptance of inappropriate cut-off policies,
with potentially dire results.
As noted in Chapter 2, this almost unthinking application of break-even cut-offs
confuses the definition of a cut-off with how it is derived. The cut-off is a value that
can be selected by the company. The discussion of Lane’s methodology has already
indicated that the optimum cut-off may be a balancing cut-off that is unaffected by price
or cost changes. The cut-off decision is just as important as decisions regarding mining
rates, treatment rates, mining sequences and so on. As will be seen, the best combination
of all decisions can only be made by optimising all strategic decisions simultaneously.
The cut-off, therefore, is not calculated and specified first, after which we optimise the
strategy. Rather, an optimised cut-off policy is only one of the outcomes of an overall
strategy optimisation study.
The mine strategy optimisation process can therefore be thought of as a multidimensional
analysis that is accounting for everything, or at least, anything of importance. It seeks
to identify the best combination simultaneously of cut-off grades and other design and
strategic decision parameters – the options within the set of value drivers – in order to
achieve the corporate goal.

The Goal of Strategy Optimisation

It is essential that the corporate goal is specified at the beginning of a strategy optimisation
evaluation. As noted in Chapter 2, without the goal being specified, the goal implicit
in the cut-off or strategy derivation process will become the de facto corporate goal,
whether recognised as such or not. Goal is deliberately singular here. Ultimately, there is
only one mine plan implemented, and that one implementation can only hit one target
– unless by coincidence the same strategy delivers two potential targets simultaneously.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 102
 CHAPTER 6 | Mine Strategy Optimisation

It is not this book’s intention to specify what the goal or goals of any company should
be. Yet it is worth noting that when a deposit is depleted and the gates closed on a fully
rehabilitated site, the benefit is the net cash generated. This cash may be distributed
to different stakeholders in various ways, but it is still essentially cash. If we are to
maximise the operation’s benefits to the stakeholders, maximising the cash generated
must be a primary goal. For decision-making ahead of time, we must make predictions
about the future outcomes of decisions that must be made now. Arguably the best single
number surrogate for cash generated over a period of time – accounting for the pattern
of cash flows and the time value of money – is the NPV.
Maximising NPV is a common corporate goal, but it is by no means the only one
considered in practice – many companies will specify other goals or multiple goals. This
book does not posit a preference for the relative merits of NPV or real options value
(ROV), both of which are measures of cash flow. Each is simply a value measure, the
maximisation of which could be included in the list of goals dealt with in the evaluation.
Chapter 10 discusses briefly certain aspects of a number of common goals, while
Chapter 11 describes several valuation techniques, including ROV.
What’s important to note is that the goal of the optimisation process must be clearly
specified. The goal should be related to some parameter that can be quantified as a
measure of value. The optimisation process will then focus on finding the strategy that
maximises (or, less commonly, minimises) the value of the measure of value. Again,
maximisation of cash generated over the life of the operation, measured by NPV, is a
common primary goal.

The Real Drivers of Value in a Mining Operation

There is a lot of emphasis in the mining industry on improvements in productivity,
efficiency, unit costs and the like. These are important issues for any operation, and may
be thought of as ‘doing things right’. The more fundamental issue is: are we doing the
right things? The ultimate aim, of course, should be ‘doing the right things right’, but
strategically the first concern must be ‘doing the right things’.1
Any operation deals with three big-picture parameters: the things that are given, the
things that can be changed (the strategic decisions or options) and the things that flow
out of a combination of the first two parameters.

The givens
There are perhaps two key givens with any mining operation. One is obviously the
resource with which it is endowed. The company has no control over this: it is what it is.
(What is known about the resource is a separate issue that will be discussed in the next
subsection.)
The other given might be termed the environment, which is here taken to be an all-
encompassing term that describes the social, political, economic, climatic and ecological
framework within which the company finds itself operating. Some of these may be
able to be changed as a result of corporate decisions (for better or worse), while others

1. The two parts to this distinction are often referred to as efficiency and effectiveness. An operation may be
efficient at doing what it is doing, but if it is not going in the right direction, it is not being effective in fulfilling
its purpose – we could perhaps describe it as being efficiently ineffective.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 103
 CHAPTER 6 | Mine Strategy Optimisation

may be outside the company’s control. For the sake of this discussion they are all taken
to be uncontrollable. Some may be predictable or measurable, while others may be
intrinsically unknowable, such as future product prices. Some might be considered
scenario parameters, for which the effects of different values or forecasts may need to be
considered in the evaluation.

Key strategic decision parameters

The key parameters that drive the strategic direction of any mining operation over
which the company has control – its options – are as follows:
•• Size of mine – for an open pit, this is the size of the planned ultimate pit. For an
underground operation, the mine size is defined by the cut-offs that determine the
size and shape of the orebodies or mining blocks to be mined. This is discussed
shortly.
•• Mining methods – as well as the decision as to whether a mine is to be open pit,
underground or both, there are a range of mining methods available depending on
the size and shape of the ore zones and the geotechnical conditions. Specifying access
and haulage methods and layout topologies could also be included in this category.
•• Production rates – as noted in Chapter 1, the term mining operation is used to refer
to the total production process, from delineating the geology through to selling the
final product at the point of sale. Production rates or capacities therefore apply to all
stages of the process, and may also include specifying the mix of various products.
•• Sequencing – for example, sequencing of mines within a region with multiple deposits;
sequencing of individual orebodies, mining blocks or pushback stages within a mine;
and generally specifying starting points and the direction of mining advances within
an orebody.
•• Knowledge of resource – while the resource itself is a given, our knowledge of it is
not. Decisions regarding how quickly a resource is delineated will have a major
impact on the types of decisions made. Options involving large capital expenditure
may require more certainty about the resource or more tonnage (implying more
years of mine life) than low-capital options. Financing covenants can also force an
operation to delineate more reserves than might be economically viable while project
development loans are being drawn down and repaid.
These will be discussed in more detail throughout the book. This brief description is to
set the scene for the initial discussion of the key features of overall strategy optimisation.

Consequential parameters
Consequential parameters flow out of the combination of givens and strategic decisions.
They include detailed mine designs, process flow sheets and detailed schedules of
physical activities and associated financial outcomes, such as:
•• development metres, potentially for several categories of development
•• waste tonnages or volumes
•• drilled and blasted metres
•• ore tonnages and grades
•• haulage and hoisting tonnages or tonne-kilometres
•• product quantities and qualities

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 104
 CHAPTER 6 | Mine Strategy Optimisation

•• revenue streams
•• operating costs
•• sustained capital costs
•• project capital expenditure for removing bottlenecks or enhancing products
•• taxes.
It is worth noting that these matters typically take up a lot of management time and
technical effort at operating mines. Of course, it’s important that this be done, but as
noted, these parameters are largely associated with efficiency issues. For example,
costs must be controlled but costs cannot be selected; if they could, we would naturally
select zero costs. Rather, the general level of costs is established as a result of strategic
decisions, which lead to specifying certain types of equipment with associated capital
costs, productivities and, hence, ranges of operating costs.
The key distinction, then, between strategic decision parameters (or options) and
the consequential parameters is the answer to the question: can we make a decision
about this parameter in principle? If the answer is yes, we are usually dealing with a
strategic option. If the answer is no, because it is set by other decisions we have made,
it is a consequential parameter to be controlled within the operation, but must also be
accounted for in our optimisation process.

Specifying the Size of the Mine

The initial discussion noted that a key decision involves specifying the size of the mine.
There are significant differences between underground and open pit mines in this
regard, and since it is a fundamental consideration, this will be addressed now before
moving on to a discussion of strategy optimisation in general.

Open pit
For an open pit, the size of the mine is essentially the same as the size of the planned
ultimate pit. The optimisation task is to select one of a number of alternative pits to use
as the ultimate pit limit. These alternative pits will be of increasing volume and surface
extent and/or depth, and will typically have a set of nested pit shells derived by using
commercial pit optimisation software applying well-established algorithms. Figure 6.1
illustrates the situation as a vertical section.
It should be noted that each pit in such a set will represent the pit with the maximum
value for a given volume. The value measure used by commonly available software is
undiscounted net cash flow. The term pit optimisation typically used in the open pit
mine planning context is not the same as overall strategy optimisation, this chapter’s
topic. There is no guarantee that the pit shell identified as optimum using common
industry practices is the ultimate pit that best satisfies corporate goals.2 Selecting this is
one of the primary tasks of the strategy optimisation process.

2. The pit shells derived by pit optimisation software will rarely be suitable as final pit designs. It will be necessary
to define pushback stages and locations of temporary and permanent haulage ramps in more detail than is
feasible within the pit optimisation software. The final designed pit may then be significantly different from the
pit shell from which it has been derived. The open pit planner will have to decide whether the pit optimisation
needs to be rerun incorporating the overall effects of the detailed designs into the parameters provided as
inputs to the optimisation software.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 105
 CHAPTER 6 | Mine Strategy Optimisation

Figure 6.1
Nested pits and alternative cut-off-defined orebodies.

Cut-off selection is often a secondary decision for open pit strategy optimisation,
though all the strategic decisions, including cut-off policy, should be made concurrently.
The capabilities of different types of software to carry out the required analysis will be
discussed briefly in Chapter 13, where it will be seen that computing limitations constrain
our ability to achieve real concurrent optimisation in practice. It is theoretically possible
to optimise cut-off policy, pit size and all other strategic decisions simultaneously.
In practice, some form of iterative process will be required. In this case, specifying a
number of alternative ultimate pit limits or pit shells will be necessary as a starting
point, but it should also be recognised that the cut-offs selected may drive the overall
shapes of the nested pits.

Underground
For an underground operation, the size of the mine is very much defined by the cut-offs
selected, which determine the size and shape of the orebody or orebodies to be mined.
These parameters then set bounds on such items as potential mining methods, equipment
sizes and capabilities and realistic ranges of production rates that may be achievable.
In the same way that we identify nested pits from which we select the best size of
open pit, so may we recognise several nested orebodies or stopes to form the basis of
an underground optimisation. One of these will, in essence, be selected as the orebody
that best delivers the corporate goals, and the sequence of cut-offs that defines the best
orebody over time will be the optimum cut-off policy. Figure 6.2 illustrates the type of
situation that could apply in a cross-section.
The selection of a range of cut-offs to evaluate, and generating mineable orebodies for
each of them, is typically the starting point for an underground strategy optimisation.

Developing the Basic Rationale of the Strategy Optimisation

Process
Let us consider two of the key decision parameters: production rate target and cut-off.
The following is based on an underground mine but the rationale applies to all types of

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 106
 CHAPTER 6 | Mine Strategy Optimisation

Figure 6.2
Nested cut-off-defined orebodies for an underground mine.

mines. Subsequent discussion will consider issues more specific to open pits and more
complex issues for both classes of operations. The discussion at this stage is focused on
developing principles common to all approaches, rather than the different optimisation
methods that could be employed.

Optimising one strategic decision

If we were to plot value as a function of cut-off, we would obtain a curve similar to that
shown in Figure 6.3. As the cut-off increases from a low value, the operation’s value will
increase, reach a peak and then decline. At lower cut-offs, material that destroys value is
included in the reserve, while at higher cut-offs, value-adding material is excluded from
the reserve. There is an optimum cut-off grade that maximises value. Note that the vertical
axis is simply value, and could be any measure that quantifies outcomes of different plans
so that they can be ranked, with one selected for what best achieves the company goals.
This concept should not be surprising.3 Generically this curve is the same as that
shown in Figure 5.10, representing the combination of the lowest of the three curves
generated for each of the three production stages identified in Lane’s methodology.
In that case, the maximum value occurs at Gmhk, which may be a balancing rather
than limiting or break-even cut-off. In practice, the curve may not be as smooth or
symmetrical as in this example.
The curve in Figure 6.3 is for a given mining method, capital plan, production rate,
etc. This is a 2D plot of value versus cut-off; the other parameters represent additional
dimensions. Let us consider another dimension – production rate target – but again as a
2D plot of value versus production rate, as shown in Figure 6.4.
At a coarse scale, there is a relationship of value growing then reducing with an
increasing production rate, as for cut-off, but the detail is complex. Production rate

3. This may be a new concept to some readers, but there is nothing new in it. The author has a sketch similar to
Figure 6.3 in his notes made in university mineral economics lectures in the early 1970s. Kelly and Bell (1992)
describe a study generating this type of curve for operations at Broken Hill in the mid-1980s.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 107
 chAptEr 6 | Mine Strategy Optimisation

Figure 6.3
Generic value versus cut-off curve.

Increasing Value

Optimum cut-off

Cut-off too low Cut-off too high

Unprofitable Valuable
material material
included in excluded from
the reserve the reserve

Increasing Cut-off

Figure 6.4
Generic value versus production target curve.

Production not Minor debottlenecking

utilising existing capex
capacity
Major
Increasing Value

debottlenecking
capex

Optimum Production Rate

Production not
utilising new
capacity

Increasing Production Rate

or production rate target is impacted by production limits imposed by the existing or

planned facilities. If there is no change to the facilities, value will grow as the production
rate increases up to the limits of the production facilities. If the production target were to
increase beyond this, the actual production could not because of the physical constraints,
and the curve would then plateau for higher targets at the value associated with the
production capacity. The stepped curve in the figure is assuming that, as capacities of
various parts of the production system are reached with increasing production rate
targets, appropriate debottlenecking capital expenditure is made. If capital is spent,
it will deliver a step change in capability. But if the production rate is increased only
incrementally, the value will reduce by the value of the capital spent – the new capacity
must be utilised to justify its provision.

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 108
 CHAPTER 6 | Mine Strategy Optimisation

Figure 6.4, therefore, effectively shows a resultant combined curve for three production
capacities over a continuum of production rate targets, including the effects of two
capital injections to permit the range of production targets to be achieved. Assuming
that the range of production rates on the horizontal axis is a practical, realistic range for
the orebody or mine under consideration, we see that there is an optimum production
rate. This will typically be at a rate that is fully utilising one component of the production
system, the implication being that further debottlenecking to allow increased production
will be uneconomic.
This example is combining several strategic decisions – each of which could be
represented as another dimension – into one dimension to be represented as the
production rate target. The capacities of each part of the production system are separate
decisions, as is the overall production rate target. The actual production achieved
may be yet another variable to be considered – either as an outcome of the capacities
provided or, if there are more complex interrelationships than can be modelled, as a
separate variable. For simplicity, this discussion assumes that the actual rate is the
same as the target and that the appropriate capital is expended to make the full range
of target rates feasible.

Finding and climbing the hill of value

The initial discussion has shown how each of two strategic decision parameters – cut-off
and production rate – may be optimised independently, all other decisions effectively
assumed to be fixed; however, both cut-off and production rate are independent
decisions (though, the cut-off selected may impose practical limits on the production
rate achievable from the mine). To obtain the best possible value for the project, they
must be optimised together. Figure 6.5 shows pictorially how value might respond to
simultaneous changes in the values selected for both of these decision parameters. It
can be seen that for any selected production rate target, the value grows with increasing
cut-off to a maximum, then falls, as shown in Figure 6.3. Similarly, for any given cut-off,
the value rises with an increasing production rate, then falls, then begins to rise again,
similar in principle to what is shown in Figure 6.4 but with one large capital expenditure
only.4 Figure 6.5 is taken from a real study conducted by the author, and is typical of
the outcomes of many underground mine situations where the value measure is NPV or
some similar financial or economic measure.
Because we live in 3D space, we are only able to comprehend 3D outcomes graphically
or pictorially. The hill of value (HoV) illustrated is a 3D surface showing the behaviour
of one dependent variable – value – as a function of two independent variables – in this
case, cut-off and production rate. Clearly, there are many strategic decisions to be made,
and each of these will be another independent variable to be represented on an axis in
another dimension. Similarly, there may be multiple value measures to be evaluated
and these will also be additional dimensions in the analysis. To establish the underlying

4. This and other HoV illustrations have been created using the 3D surface chart capability in Microsoft Excel™. This
facility cannot handle irregularly spaced data values on the axes; all data points are spaced equally, regardless of
their values. Hence, Figure 6.5 does not show the step changes in value seen in Figure 6.4, but the values at all
calculated points (the 42 combinations of seven cut-offs and six production rates, shown as a grid on the floor
of the chart and the surface itself ) do account for such step changes.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 109
 chAptEr 6 | Mine Strategy Optimisation

Figure 6.5
A typical underground hill of value for value versus cut-off and production target.

Increasing
Value

Increasing
Production Rate
Increasing
Cut-off

principles of the process, we will continue the discussion of a 3D case with one value
dimension and two option dimensions.
Figure 6.6 shows the typical location on the HoV surface for an operation whose
strategy has not been optimised. Typical industry planning processes do not usually
deliver a plan that is anywhere near optimal (in terms of maximising NPV).
A common proposal to improve the economics of an operation is to increase the
production rate. The idea is to spread the fixed costs of the operation over more
tonnage, thereby reducing the unit costs and increasing the average margin. This is
illustrated in Figure 6.7.
There are some possible problems with this. Firstly, it is to be hoped that the system
capacities will have been identified so that the planned production increase is to the

Figure 6.6
A typical underground hill of value showing a typical unoptimised strategy.

Typical state of
current operation
Increasing
Value

Increasing
Production Rate
Increasing
Cut-off

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 110
 chAptEr 6 | Mine Strategy Optimisation

Figure 6.7
A typical underground hill of value showing a typical improvement strategy.

Typical improvement
proposal: arbitrary
production rate
increase

Increasing
Value

Increasing
Production Rate
Increasing
Cut-off

top of the ridge of the value surface; however, there is no guarantee that this will be
done, and what may be a logical increase in some areas of the operation may require
debottlenecking capital expenditure in other parts. Furthermore, potential gains from
the arbitrary production increase proposed may be more than consumed by the capital
requirements. This can be seen in Figure 6.7 where the proposed higher production rate
produces less value than the initial rate.
It is implicitly assumed that production can be increased at the same head grade, in
which case the margin per unit of product will increase as intended. If this is not feasible
– at least in the short- to medium-term – it may be necessary to lower the cut-off to
increase the tonnage produced. And although the average unit cost per tonne of ore may
reduce, the unit cost per unit of product may not, or at least not to the extent expected.
It is also common in these circumstances for it to be proposed that, since the unit cost
will have reduced, the cut-off can also be permanently reduced, thereby increasing the
reserve. This is a natural consequence of the erroneous assumption that cut-off and
break-even are synonymous. A moment’s consideration of the HoV’s shape will show
that this is not a good plan.5 Reducing the cut-off reduces value.
Figure 6.8 indicates the optimum strategy to maximise value. It is the combination of
cut-off and production rate that puts the operation at the top of the HoV. For the case
shown, this will require both an increase in cut-off and in production rate. In this case
the underlying analysis has taken account of the production capabilities of the orebodies
at different cut-offs, so that the production strategy at each point on the value surface is
achievable in practice.
It was seen that cut-off defines the size of mine for an underground operation, whereas
in an open pit it is the size of the ultimate pit. All the preceding discussion regarding cut-

5. this conclusion assumes that the key value measure behaves as plotted. this hoV is typical of the behaviour of
NpV and long-term profit measures of value. Of course, if maximising the mine life or the reserve tonnage is the
key goal, the shape of the hoV will be completely different, falling as cut-off increases. Minimising the cut-off
then does become an appropriate strategy.

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 111
 CHAPTER 6 | Mine Strategy Optimisation

offs in an underground operation can in general be translated into the open pit situation
by replacing references to ‘increasing cut-offs’ with ‘reducing pit sizes’ and ‘reducing
cut-offs’ with ‘increasing pit sizes’. Where cut-off is shown as increasing from left to
right across the various value versus cut-off plots, the plots are equally valid for open
pits with pit size increasing from right to left on the same axis.
The strategy optimisation process is simple in concept, as has been illustrated. Still,
the practical application of the underlying principles with several option/scenario
parameters to be accounted for – each of which becomes another dimension in the
analysis – plus a time dimension (most of the option/scenario parameters may vary over
time) may not be easy. Even after the results have been successfully generated, deciding
which mine plan to implement will be an even more contentious and time-consuming
task. One of the initial problems encountered at an operation that has not been optimised
previously will be that the size of the mine as planned before the operation will be too
large; that is, for an underground operation, the existing cut-off will be too low (as shown
in Figure 6.8), or for a pit, the ultimate pit will be too big and/or the mining rate too low.
Other goals will then tend to come into play, as increasing the cut-off underground or
reducing the size of an open pit will reduce the ore tonnage and metal in the reserve,
which may not be acceptable even though it increases the cash generated. Chapter 10
discusses parameters that are often used as goals and value measures. For now, we note
simply that some are inversely correlated with cash generation. Some may be applied as
constraints rather than being optimisation goals themselves – the main value measure
should be maximised with the proviso that the outcome for another measure must be
greater than or less than some specified value.

Figure 6.8
A typical underground hill of value showing the optimum strategy.
Ideal improvement proposal:
change production rate and
cut-off to maximise value

Increasing
Value

Increasing
Production Rate
Increasing
Cut-off

Dealing with multiple goals in an underground mine

If an optimisation model has been developed to derive primary cash-related value
measures such as NPV, any other measure that might be of interest should be available
in the analysis. Considering similar plots for other value measures of interest permits

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 112
 CHAPTER 6 | Mine Strategy Optimisation

informed trade-offs to be made, so that the best strategy overall, taking account of both
quantitative and qualitative issues, can be specified.
Figure 6.9 shows what is ostensibly a 2D plot of the values of several value measures
for a range of cut-offs. This chart is taken from a real study conducted by the author for
an underground gold mine. There are five value measures shown, so in the context of
the HoV concept discussed, this could be considered to be collapsing six dimensions
– one independent option variable, cut-off, and five dependent value variables – into
a 2D representation.6 Alternatively, the discount rate could be thought of as a second
independent axis, so it is a 5D analysis with two independent axes, cut-off and discount
rate, and three dependent value variables.

Figure 6.9
Underground example – various parameters versus cut-off grade.

2 3 4 5 6
Cut-off (g/t Au)

NPV @ 0% NPV @ 5% NPV @ 10%

Avg $ / oz Avg oz / yr

The company’s major value measure was NPV at a ten per cent discount rate, but
recognising that discounting tends to favour strategies that add value in the near-term,
the company was also interested in the total undiscounted net cash flow – that is, the
NPV at zero per cent discount rate – and the NPV at an intermediate discount rate of
five per cent. Of further interest were the annual gold output, which had to be as high as
possible but not less than a specified target, and the unit cash cost of production, which
of course was to be as low as possible. Whether these are rational or appropriate goals is
not the issue here – the intention is to indicate the type of problem that companies using
multiple goals have.
The company at the time was using a cut-off of 3.0 g/t, and the calculated break-even
grade was of the order of 2.5 g/t.
It can be seen from the plot that the cut-off that maximises NPV at ten per cent discount
rate is approximately 4.5 g/t, while that which maximises net undiscounted cash flow
is approximately 4.2 g/t. The NPV at the intermediate discount rate of five per cent is
maximised by a cut-off between 4.2 and 4.5. Clearly, different goals require different

6. Other option decisions, such as production rate, are fixed here, though they could vary in a 3D perpendicular to
the values versus cut-off plane shown.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 113
 CHAPTER 6 | Mine Strategy Optimisation

strategies. Which to choose? Further inspection of the plots suggests that, near the top
of the HoVs for these cash flow value measures, the curves are relatively flat, and the
practical reality, given also the various uncertainties in the data, is that any cut-off in the
range from, say, 4.0 to 4.5 g/t will satisfy all these cash flow maximisation goals.
What about the gold production and unit cost? Interestingly, these measures are both
optimised at a cut-off of approximately 5.0 g/t, which is significantly higher than the
cut-offs maximising the cash flow measures; however, selection of a cut-off in the 4.0–
4.5 g/t range will still deliver outputs and unit costs that are close to optimum.
It becomes evident, therefore, that if there are multiple goals to be satisfied, the
optimum strategy may not optimise any individual goal but rather provide a close-
to-optimum satisfaction of a number of potentially conflicting goals. Identifying the
appropriate trade-offs may be a qualitative judgement, rather than a purely mechanical
result of the quantitative processes that generate the numbers for each value measure.
It is also interesting to note that the cut-off that maximises net cash flow (4.2 g/t) is
significantly higher than the break-even grade (2.5 g/t) stated by the company. The net
cash flow at that break-even cut-off is significantly lower than the maximum possible,
which indicates that at the break-even there is material included in the reserve that
is consuming more cash in input costs than the revenue it is generating. This in turn
implies that the break-even calculation is not adequately incorporating all costs that
should be included. This is a common outcome and typically results from excluding
sustaining capital costs7 (and sometimes some or all of the fixed costs) from break-even
calculations.
Not shown in this chart are plots of mine life and tonnes and gold ounces in the reserve.
These, of course, all fall continuously (not necessarily as straight lines) from left to right
across the chart as cut-off increases.8 When these types of measures become important,
finding a strategy that well satisfies both maximisation of cash flow or profitability and
maximisation of life or reserves measures may not be easy.
There is often a lot of discussion within companies around the appropriate value of
the discount rate to be used. It is interesting to note that the magnitude of the maximum
NPV is, not unexpectedly, strongly influenced by the discount rate. Clearly getting
the discount rate right is important if the aim of the study is to identify the desirable
purchase or sale price for the operation. If the aim is to identify the optimum operating
strategy to maximise cash returns, in this particular case, and in a number of other
studies conducted by the author, the discount rate used, perhaps unexpectedly, has little
impact. It is important to identify the purpose of the study.
In the project illustrated by the HoV in Figure 6.8, the company’s senior management
identified that the mine life – delivered by the cut-off and production rate strategy that
maximised NPV – was shorter than what was accepted as a given by both investors
and the nearby townspeople, which included most of the workers at the operation.
Management’s assessment was that the scale of the reduction in life would result in a

7. The distinction between capital and operating costs is to a large extent irrelevant for cut-off determination.
Often the distinction between project capital, usually excluded from break-even calculations, and sustaining
capital, which should usually be included, is not appreciated, and all capital is often excluded from break-even
calculations.
8. Mine life will also decrease with increasing production rate as well as increasing cut-off.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 114
 chAptEr 6 | Mine Strategy Optimisation

major reduction in the company’s share price, and cause problems with the workforce
and town. Yet after identifying a reduced but acceptable mine life – with no significant
reduction in share price and adequate time for the workforce and town to prepare for
the mine’s closure – it was found that the resulting strategy, with a somewhat lower
cut-off and production rate, was still close to the flat portion at the top of the HoV. Most
of the upside in value (from the current position on the front-left corner of the HoV)
could be achieved without the angst of a serious reduction in mine life. A number of
value versus strategy surfaces were used to make an informed decision about the best
strategy (which did not optimise any one measure of value), taking account of a number
of interrelated issues, some of which were qualitative judgements.
These case studies suggest that the ultimate aim of a strategy optimisation study is
not just to mechanically identify the optimum, if that is in fact possible, but to provide
senior decision-makers with information from which informed decisions can be made.
Decision-makers can be made aware of the benefits and dangers of strategies and,
applying their collective knowledge of bigger-picture issues not necessarily captured
by the optimisation evaluations, make the best possible decisions. Chapter 12 describes
how these techniques are applied to quantify the trade-offs between the upside of correct
assumptions, such as future metal prices, and the downside of wrong assumptions.

dealing with multiple goals in an open pit

Figure 6.10 illustrates results from an open pit case study. The HoV is shown as a
contour plot, indicating the maximum value occurring slightly above-right to the
centre of the plot. The horizontal option axis represents the run-of-mine (ROM) cut-off
– the cut-off applied to the rock mined from the pit to determine what is ore – which is
to be sent to the plant for processing. The vertical option axis is the rate of mining rock,
expressed as a percentage of the rate in the current mine plan (at the time of conducting
the study). It can be seen that there is a ridge of value running from low rate / low
cut-off to high rate / high cut-off within the range of option combinations evaluated.

Figure 6.10
Open pit example – net present value versus run-of-mine cut-off and mining rate.
200%

150%
Mining Rate
Higher Value Case as % of
Base Case Rate

100%
Base Case

Increasing ROM Cut-off

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 115
 CHAPTER 6 | Mine Strategy Optimisation

The base case – the then current mining plan – is identified by the circular marker on
the ‘100 per cent of base case mining rate’ grid line.
Open pit mining plans generally automatically optimise the ROM cut-off, given a
mining rate. There will be short-term variability, but overall the ROM cut-off will be the
Lane mining–treatment balancing cut-off, subject to its not being less than the mining-
limited break-even cut-off, and this will be the overall optimum cut-off. With short-term
grade distribution variations, there will be some variation in the ROM cut-off to suit.
This maximises value at the planned mining rates, which are set to deliver the planned
(plant capacity) ore tonnes at a planned ROM cut-off; this in turn will be calculated
as a break-even. By selecting the mining rate to balance with the ore treatment rate
at that cut-off, the ROM cut-off is optimised at those rates automatically. There is a
circular argument here. What starts as a break-even cut-off that may or may not be
optimal becomes optimal because capacities are established so that a physical balance is
achieved at that cut-off.
However, it can be seen that value can be increased, up to a point, by increasing the
mining rate and the ROM cut-off. For a given ore treatment rate, faster mining generates
more material of all grades, so the cut-off can be increased naturally to keep the mill full
of ore at a higher grade. Up to a point, the higher grade and extra revenue will cover the
additional mining costs of the faster mining rate, but eventually the distribution of grade
in the mineralisation will be such that this no longer occurs. Additionally, while mining is
in progress and fixed costs are incurred regardless of what is classed as ore, the marginal
break-even below which material cannot be economically treated is Lane’s mining-limited
cut-off; however, once mining is completed and treatment of stockpiles only is progressing,
fixed costs must be included in the marginal break-even, which therefore rises. Low-grade
material that could have added value if treated as marginal material at low mining rates
becomes unprofitable if it has been stockpiled for treatment after mining is completed
because of higher mining rates. The value therefore also drops at higher mining rates
because of the overall cut-off applied to the material within the defined ultimate pit.
This should not be seen as an argument in favour of a low mining rate and lower overall
cut-off and increased reserve. In reality the decision to mine at a slow rate has resulted
in an increased mine life, and although the fixed costs are committed and irrelevant to
the cut-off decision once the mining rate has been specified, they are, in fact, a real cost
of that marginal material. There is a direct relationship between the mining rate and the
cut-off; once we specify one, and given a pre-specified treatment rate, the other is almost
automatically determined. So a decision to reduce the mining rate and increase life is
effectively a decision to reduce the cut-off and increase the life, so the cut-off situation is
a treatment-limited situation rather than a mining-limited situation. It becomes mining-
limited by virtue of a policy decision to mine at a certain rate, rather than a genuine
physical mining rate constraint.
In the case presented here, the current ROM cut-off may be optimal given the current
mining rate, but taken together, the mining rate and the ROM cut-off are not optimal.
Both need to be increased from their current values to become optimal for the goal of
maximising NPV. However, this subsection is discussing multiple goals in the open
pit context. So what other goal/goals are evident on the NPV versus ROM cut-off and
mining rate plot? The answer is: mining costs. These will be proportional to the mining
rate. Management at this operation were being pressured to minimise annual costs, and
costs will obviously be minimised by reducing mining rates. The option selected was

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 116
 CHAPTER 6 | Mine Strategy Optimisation

therefore to minimise the mining rate. This in turn required reducing the ROM cut-
off to maintain ore feed to the mill. The net effect was to significantly reduce the NPV
achieved by moving down and to the left in the contour plot shown, rather than up and
to the right; however, cost minimisation, not NPV maximisation, was the measure by
which management was being judged.
Whether this was the corporate goal or not is another issue – it probably wasn’t. What
this case study highlights is how goals for mine staff that are thought to be aligned with
the corporate goals, but are not, can drive counter-productive decision-making. Selecting
strategies based on measures that are truly aligned with corporate goals is obvious, but
this frequently does not occur. Explicit management goals such as cost minimisation
neglect to account for situations where cost reductions reduce the revenue more than the
costs. This is what occurred in this case study. Implicit goals, such as ‘every tonne pays
for itself’ for break-even cut-offs, can also result in suboptimal strategies as has been
shown here. Optimising the mining rate and adopting the corresponding optimum cut-
off will generate a higher net cash flow, even though costs are higher and the cut-off is
not a break-even grade as commonly calculated.

Strategy Optimisation vERSUs Typical Strategy Selection

How does a strategy optimisation study differ from long-term strategy specification
practices in the mining industry?
Typical evaluation processes handle many relationships in separate, independent
processes to create a limited number of preset strategies. Technical staff produce
alternative plans and schedules for the operation being evaluated. Mine plans are
generated using industry-standard mine planning, design and scheduling software,
often to a high level of detail. By virtue of their training and experience, technical
staff (geologists, mining engineers, geomechanics specialists, metallurgists) tend to
ensure that all considerations within their fields of expertise are accounted for. The
capabilities of modern software tools also help. If anything, a false sense of accuracy
is attained – just because the scheduling package is capable of working in daily or
hourly time slices for decades into the future, doesn’t mean that this will happen. Most
technical staff would recognise this, yet the capabilities of the software almost drive us
to generate plans for much longer time frames than are warranted by the accuracy of
the underlying data, not to mention the unplanned events that will undoubtedly cause
changes to plans in the future.9
Because of time constraints and the limited availability of technical staff,10 it is
common for only a base case and a few alternative options to be generated at this level

9. One-year budgets are often obsolete by the start of the budget year to which they pertain; how much more so
for five-, ten- or 20-year plans? Yet the increasing power of modern software is leading to an expectation that
a long-term plan cannot be relied upon if it does not have this level of precision across the whole time frame
being evaluated.
10. Consider the reason for these constraints. Apart from the numbers of people available with degrees of experience
and qualifications, the topics on which these people are working do not necessarily align with the importance
of the work. It is all too common to find that the urgent displaces the important. Chapter 7 discusses a number
of mine planning process issues. It is evident that urgent issues that displace work on the important issues are
often the result of not having done the important tasks properly in the first place, if at all! It can become a vicious
circle, and it needs breaking.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 117
 chAptEr 6 | Mine Strategy Optimisation

of detail and intellectual rigour. The mining plans thus developed will then be used
to estimate the capital and labour requirements, metallurgical performance and such
items as environmental and closure plans. The schedules of various physical activities
and costs derived by each of these processes will then be consolidated in a relatively
simple model, which in reality will be little more than an arithmetic calculator to derive
revenues and costs from imported, unchangeable physical quantities, though there
may be complex calculations in some areas, such as the derivations of revenue, fleet
replacement schedules, stock movements, working capital and accounting statements.
Figure 6.11 shows schematically the study activities and flows of data and information.

Figure 6.11
The typical strategy selection process.

Production targets Debottlenecking options Closure issues

Mine design data

Prodn & Devt Project
Schedules Capex Revenue data
Schedule data

Ongoing capex
Production constraints

Simple Met. performance

Model
Operating cost drivers

Corporate
Goals

Results Decisions

The simple model will often be used to flex things that can be easily changed within
such a model. These will be such items as product prices, exchange rates, operating and
capital costs and perhaps ore grade and metallurgical recoveries, and will usually be
scenario parameters. These will be all flexed independently, up and down by specified
percentages of the base case values, regardless of whether the likely range of variations
in each of these inputs is greater or less than these arbitrary variations. As expected, on
plotting the outcomes of these sensitivity analyses (often referred to as a spider diagram),
we will find that if prices go up or costs go down, the value of the plan evaluated will
increase, and if prices go down or costs go up, the value of the plan will decrease. It is
also usual to find that value is most sensitive to variations in revenue-related parameters,
such as price, exchange rate, grade and recovery.
What we have done in this type of analysis has identified how the value of a plan varies
in response to external parameters over which we have little or no control. We have
not in any way assessed how the plan itself might actually change in response to these
external issues. The mining schedules and other underlying physicals will not change
in the model; they are fixed by the detailed work done externally, with values only
imported into the evaluation model. The absolute value of the project is not so important
when it comes to strategic decision-making. Of more importance is identifying changes
in circumstances that would cause us to change our plan. The typical strategy selection

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 118
 chAptEr 6 | Mine Strategy Optimisation

process only addresses this issue at a superficial level by evaluating a small number of
alternative operating plans, none of which is likely to be optimal.
A strategy optimisation study, on the other hand, must consider the same issues as
the strategy selection process. It must also identify the optimum operating strategy (the
best options) for the base case economic assumptions, such as prices and exchange and
inflation rates, and establish how changes in various scenarios might result in changes
to the optimum plan. The process must, theoretically, be able to evaluate all the options
to ensure the best plan can be found. For strategy selection, engineering resources are
used to develop a few detailed plans. For strategy optimisation, these resources help
to identify the relationships that experienced planners are using automatically on the
basis of their experience as the detailed plans are developed, so that the relationships
rather than specific plans are incorporated into the evaluation model. Instead of several
separate, independent processes to generate plans and schedules and then determine
their associated revenue and cost streams, the optimisation model handles all the physical
scheduling and resulting cost, revenue and value measure relationships internally. This
enables a very large number of options to be evaluated. Various inputs can be flexed
simultaneously, and the best strategy, including cut-off policy, will be identified.
Figure 6.12 shows the strategy optimisation process schematically. Note that all the
study process inputs in boxes around the outer edges of the figure are the same as in
Figure 6.11 for strategy selection. However, as indicated in the preceding discussion,
the information flowing into the model will be both data values and relationships
that enable schedules to be developed by the model, rather than predetermined
unchangeable schedules.
There are other subtle differences between the two figures for strategy selection and
strategy optimisation. Earlier chapters have identified the importance of specifying
corporate goals to drive the evaluation process, and how, if this is not done, the goal
implicit in the evaluation process will become the de facto goal of the plan developed.
Note that in Figure 6.11, the corporate goals are not linked to the evaluation processes but
are only applied to the results of those processes. In reality this is common for strategy

Figure 6.12
The typical strategy optimisation process.

Production targets Debottlenecking options Closure issues

Mine design data

Schedule data Revenue data

Robust
Model Ongoing capex
Production constraints
Met. performance

Operating cost drivers

Corporate
Goals

Results Decisions

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 119
 CHAPTER 6 | Mine Strategy Optimisation

selection. The few operating plans that are developed are not focused on achieving any
particular goal. Rather, their outcomes after they are developed are compared with each
other in light of the goals to decide which plan best delivers the goals. If none of the
plans deliver satisfactory results, attempts are made to improve the best of them; for
example, by reducing costs (often unrealistically).
In Figure 6.12, however, the corporate goals are used not only to assess outcomes, but
also feed into the evaluation process. Because the optimisation process identifies the best
plan out of a large number of options, the goals, the measures that quantify them and the
ways they are to be accounted for within the process must be key parts of the evaluation
processes. The goals are not just used to compare options after the evaluations, as is
done with typical strategy selection.
The other major difference at this stage is the difference between the sensitivity
analyses in each process. Typical sensitivity analyses flex a number of variables, such
as costs and prices, which are usually scenario parameters outside the control of the
company; then assess the impact of these changes, independently, on the value of a
specified operating plan. Strategy optimisation, on the other hand, is primarily flexing
decisions about option parameters. The HoV is in fact a sensitivity plot of value versus
decisions.
Knowing the behaviour of the value versus decisions surface is a crucial part of
strategy optimisation, indicating how much flexibility there is, as well as the optimum
strategy. Is the HoV shaped like a church steeple, so that there is a significant loss of
value if one deviates a little from the identified optimum? Or is it a broad plateau with
a range of options, indicated by little variation in value across the upper surface? In this
case, deviations from the precise optimum may result in insignificant value reductions,
as long as the limits defined by steeper sides around the flat upper surface are not
breached. Figure 6.13 illustrates the types of situations that could occur.11

Figure 6.13
The importance of the shape of the hill of value.
HoV with “flat” top HoV with “pointed” top
Precise strategy Precise strategy
specification not critical specification critical

11. In the author’s experience, most HoVs in optimisation studies are flat near the optimum but become steeper
with distance from the optimum, like the first of the two hills in Figure 6.13.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 120
 CHAPTER 6 | Mine Strategy Optimisation

While noting that three dimensions are all that we can graphically conceive, there is no
reason why we should not consider higher numbers of dimensions in the analysis and
include both options (which we can make decisions about) and scenarios (which may
vary but are outside of our control) as axes of a multidimensional HoV. To continue the
geographical analogy, consider a change in predicted prices. For strategy optimisation,
we are not so much concerned to know how much the hill rises (the value of any
operating plan options will increase), but whether the pattern of value increases is such
that the HoV also moves sideways (that is, the optimum strategy changes) as a result of
the changes in external parameters. If the axes of the HoV can include both option and
scenario parameters, the spider diagram types of sensitivity analyses can be seen to be a
subset of HoV-type outputs, depending on what parameters are selected for the axes of
various plots that can be understood graphically in 3D space.

The Strategy Optimisation Process

Preceding chapters discussing the ways to determine cut-offs describe the underlying
theory and processes to derive a number as the cut-off. Simple yet rigorous formulas
are used to obtain the desired result. Unfortunately, there is no formally accepted
recipe for strategy optimisation. This chapter has described the principles underlying
an optimisation evaluation, of which finding the optimum cut-off policy will be a
component. Although there are similarities between operations, each has its own
peculiarities, and some model customisation will be required to fit the circumstances
of each case.
There are a number of methodologies that can be applied to determine the optimum
strategy, and the details of the processes and data requirements will depend on the
methodology employed. The processes for a strategy optimisation generally involve
generating data and relationships in order to model the influence of constraints and
decisions related to geology, mining, metallurgy, operating cost, capital costs, end-
of-life and social issues and financial and economic forecasts. Data requirements and
relationships are discussed in more detail in Chapter 8. Later chapters will consider
features of various optimisation methodologies and cover some aspects of the processes
in more detail.

Optimising One Project vERSUs Optimising All Projects

Many companies own more than one operation. Each operation is often optimised
in isolation, without considering the impact on other operations. The task of senior
company executives will be to maximise the value of the company as a whole, not just
one of its operations.
For an individual operation, value will typically be maximised with a relatively large
project. If there is no limit on the amount of capital that the company can spend across
its enterprises, it may be appropriate to spend whatever is needed so that each operation
is optimised individually. However, if there are limits on what can be spent, each
operation must be considered as part of the total portfolio of projects. Each potential
capital expansion, though it may be economic and add value to that project, must be
assessed in conjunction with other potential projects at the other operations, so that
scarce capital can be allocated to projects that generate the greatest NPV for the total
capital invested.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 121
 CHAPTER 6 | Mine Strategy Optimisation

Although it is important for a team evaluating an operation to identify what strategy

maximises the value of that operation, in principle it is insufficient on its own. Results
should also indicate the maximum incremental NPVs derived from sequential capital
injections. Corporate decision-makers can then identify the optimum sizes for all
projects considered together, rather than in isolation. The overall optimum corporate
strategy may well be to have a number of smaller high-return operations instead of one
or a few larger but low-return projects. Identifying the values of expansion projects at
each operation allows the optimum allocation of scarce capital to maximise the total
value of the company.
Decision-makers should also consider the impact of sequencing various deposits
within one operation. Previous discussions have highlighted the need to flex uncertain
items to determine whether they influence current decisions. These decisions may be
significantly impacted by whether or not additional resources are mined and treated
and, if so, when. Figure 6.14 gives a simplistic representation of the type of situation that
may be found.
The horizontal axis is the cut-off to be applied to the first deposit or mine area in the
sequence of possibly many deposits/mine areas. This is the initial decision to be made:
what should the cut-off be for this resource? The lower curve shows the HoV value versus
cut-off curve for the first deposit. The optimum cut-off is clear. But if the other resources
are included in the analysis, the cut-off for the first deposit that maximises the total value
of the operation – as shown by the upper curve – is significantly higher. The reason for
this is easily explained: the loss of NPV resulting from increasing the cut-off and reducing
the reserve and mine life for the first deposit, judged on a standalone basis, is more than
compensated for by the increase in NPV resulting from the consequent bringing forward
of the cash flows of the subsequent deposits.
Decisions to mine the other deposits are not being made now but will be made at
various times in the future. There is therefore no guarantee that the other deposits will
be mined in the same way and with the same values, as indicated now, or even that they
will be mined at all.

Figure 6.14
Values of one and several sequentially mined deposits.

Optimum Cut-off for First Deposit

1. Stand-alone 2. In sequence

Reduced tonnes
Total NPV

and shorter life in

first deposit brings
forward value
of the remainder,
increasing its NPV

Increasing Cut-off, Reducing Reserve in First Deposit

All deposits First deposit only

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 122
 CHAPTER 6 | Mine Strategy Optimisation

Chapter Summary
In the previous three chapters, the main focus has been on determining the cut-off
grade, which is just as important as the decisions regarding mining rates, treatment
rates, mining sequences and so on. The best combination of decisions can only be made
by optimising all strategic decisions simultaneously. The mine strategy optimisation
process can therefore be thought of as a multidimensional analysis that is accounting
for everything. It is seeking to identify the best combination of cut-off grades and other
design and strategic decision parameters – the options within the set of value drivers –
in order to achieve the corporate goals. The corporate goal/s must therefore be clearly
specified. The strategic policies that deliver one set of goals will probably differ from the
policies that deliver other sets of goals.
Any operation deals with three types of big picture parameters: the things that are
given; the things that can be changed (the strategic decisions or options); and the things
that flow out of a combination of the first two parameters. These must be considered
in the context of external parameters, such as metal prices and inflation rates, that the
company has no control over (the scenario parameters).
There are two key givens associated with any mining operation: the resource with
which it is endowed, and the environment, which is taken to be an all-encompassing term
that describes the social, political, economic, climatic and ecological framework within
which the company finds itself operating.
The key parameters over which the company has control that drive the strategic
direction of any mining operation are typically the size of the mine, mining methods,
production rates, sequencing and knowledge of the resource. Consequential parameters
are what flows out of a combination of the givens and strategic decisions, including
detailed mine designs, process flow sheets and schedules of physical activities and their
associated financial outcomes.
Each of the strategic options can be optimised on its own, but to optimise value
the interactions between all options need to be considered. Relationships need to be
developed so that in the strategy optimisation model – in whatever software package
and with whichever technology is employed – the effects of changes made to the values
of one or more option parameters will be accurately represented throughout the model
and, of course, in its outputs.
Although powerful analytical techniques can be employed to find the optimum, it
is important to know how value responds to changes in the options settings about the
optimum.
Although formulas and defined processes exist for determining break-even, Mortimer-
style and Lane-style cut-offs, there is no such formal, well-accepted recipe for strategy
optimisation. The processes for a strategy optimisation generally involve generating data
and relationships in order to model the influence of constraints and decisions related to
geology, mining, metallurgy, operating cost, capital costs, end-of-life and social issues
and financial and economic forecasts.
There is nothing among these processes for a strategy optimisation that would not
form part of a well-conducted prefeasibility or feasibility study. The typical strategy
selection process flexes scenario items such as metal prices and costs for a limited
number of pre-specified mining and treatment plans. Strategy optimisation involves
flexing decisions. The evaluation model must therefore be able to generate the mining

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 123
 CHAPTER 6 | Mine Strategy Optimisation

and treatment plans in response to changes in strategic decisions. The key distinction
of a strategy optimisation is that multiple data sets may be required and relationships
developed so that the impact of flexing decisions is fully accounted for in the model, not
in various external design and scheduling processes. Final selection of the best option
will come from evaluating the trade-offs between relevant measures of value in light of
all the knowledge available.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 124
 HOME

Chapter 7

Cut-off Derivation Methodologies

and Their Place in the Overall
Mine Planning Process

Introductory comments
Several ways of deriving cut-offs and the optimum mine plan have been discussed to
this point. The reader may now wonder when or in what circumstances each should be
used. Determining cut-offs and strategy optimisation are part of the overall planning
process at any mine.
The author considers it a privilege to have spent most of the formative years of his
career in companies with a strong planning culture, where best practice planning is
absorbed as part of ongoing professional development. Unfortunately, many in the
mining industry – in management, operating and technical roles – do not know what
constitutes a good planning process and the benefits of it. In many cases, they often do
not know that they do not know.1 As an industry, we (quite rightly) award kudos to
those who apply their management skills to manoeuvre their operations out of difficult
problems; however, many of these operational problems could have been foreseen and

1. Possibly a result of the operations-centric culture in many companies as well as the worldwide shortage of
technical professionals, resulting in rapid staff turnover, lack of continuity and lack of mentoring. References
to staffing shortages in this book take into account the industry’s boom-and-bust cycle. During boom times
there is a shortage of qualified people available, but during downturns, while there may be plenty of qualified
people available, companies are not employing them and are running with reduced staffing levels. So, for the
work being done within companies, there is an effective shortage of staff, leaving those who are employed
overworked and without the time for thinking about long-term value-maximising strategies.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 125
 CHAPTER 7 | Cut-off Derivation Methodologies

avoided with good planning. It is easy to see the costs of the technical team in the cost
reports. The benefits are less easily quantified but will often be orders of magnitude
greater than the costs of the planning – they are the cost savings and revenue gains from
better plans and, perhaps more importantly, from the mistakes that did not get made.
This chapter discusses firstly a recommended overall mine planning process in order
to establish a framework of good process, and then indicates how the various cut-off
and strategy optimisation procedures can be included in the process. The processes
that will be described have long been well-recognised in the industry.2 While primarily
intended to give context to the cut-off and strategy optimisation techniques in this book,
the discussion may also give some insight to those who do not know (or do not know
that they do not know) what good planning processes are.

THE MINE PLANNING PROCESS

Mine plans and strategies should have a clearly defined goal (or a set of multiple goals).
As noted previously, if plans are not deliberately created to deliver specified corporate
goals, the de facto goal of the corporation will become the goal that is implicit in the
derivation of the plan (such as ensuring that every tonne pays for itself if using a break-
even cut-off), regardless of official pronouncements stating the company’s goals.
For a mining company, the primary goal is to maximise value for shareholders.
Ultimately, this is achieved by maximising the net cash generated by the operations.
Taking into account the time value of money, the net present value (NPV) is the measure
often used to represent the cash returns. It is for each company to specify its corporate
goals, and how value is to be measured. There may be a number of measures, and the best
strategy may satisfy a number of these, rather than optimising any one particular goal.
Various common value measures and goals are discussed in more detail in Chapter 10.
The discussion in the rest of this chapter will be in terms of one goal only.
Much of the traditional economic theory regarding maximising the value of a company
has been developed in the manufacturing industry where the factory is the primary
asset. In the simplest economic models, there are no constraints on inputs or outputs.
Successive time periods are essentially independent – if product is not made and sold
in any time period, the opportunity is lost forever. If the capacity of the factory is not
changed, value is optimised by maximising the value of each time period independently
by simply maximising outputs and minimising unit costs.
Mining, in general, is different. The primary asset is not the factory (the mining facilities
and metallurgical treatment plant). Rather it is the mineral resource. With a finite and
depleting resource, decisions made in any time period affect the resource remaining
at the end of it, hence affecting what is possible in all subsequent time periods. Value
cannot therefore be maximised by considering each time period independently. Plans
for all time periods must be optimised together.

2. Many authors have written on these aspects, but for simplicity, readers are referred to the Cost Estimation
Handbook published by the AusIMM, particularly Chapter 1, which deals with preproduction studies (Dewhirst,
2012). Several relevant papers are listed references to that chapter. (That so many have felt the need to write
such papers supports the author’s contention that many in the industry ‘do not know that they do not know’
what good planning procedures are.) Steffen (1997) discusses the planning processes at an operating mine.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 126
 CHAPTER 7 | Cut-off Derivation Methodologies

In a long-life operation where there are significant lag times between exploration,
planning, development or waste stripping and finally production, plans must be
developed in a hierarchy from long term to short term. Shorter-term plans should be
consistent with, and provide more detail than, the longer-term plans from which they
are derived. It is the long-term or life-of-mine plan (the LOMP) that ultimately generates
value. Once the optimum LOMP has been identified and approved, all shorter-term
plans must be developed to deliver that LOMP.
Key performance indicators (KPIs) should cascade down through the organisational
structure to ensure the ultimate delivery of the LOMP.3 Almost by definition, KPIs
for senior operating managers should focus not on short-term measures but on those
related to activities that help realise the LOMP. Focusing on short-term plans and
one-year budget KPIs is almost guaranteed to result in long-term outcomes that are at
variance with the optimum LOMP. Operations in the long term will then be suboptimal
and value-destroying. Long-term plans will be forever reactive to previous years’
short-term decisions, rather than setting the overall direction to be facilitated by the
detailed short-term plan.
Since these discussions refer to planning processes internal to each company, terms
such as ore and reserves are used in their colloquial sense and not according to the strict
definitions for public reporting in international reporting codes. Some of the studies
may result in publicly reported information, but that is beyond this book’s scope.

PREPRODUCTION STUDIES
Preproduction studies advance through a number of stages with increasing levels
of detail. These stages go by various names, one set of which has been used here.
Recommendations for incorporating cut-off determination and optimisation processes
are included in the discussion for each stage. For all preproduction study stages, the cut-
offs derived will almost certainly be planning cut-offs, as defined in Chapter 2.

Scoping or conceptual study

Typically stated to be at a high-level order-of-magnitude accuracy (±30 per cent), the
purpose of this level of study has been to demonstrate early in the resource discovery
and evaluation process that further detailed work is justified. Theoretically it is therefore
only necessary to find one case that proves this. An underlying purpose will be to avoid
relinquishing a prospect too early in the evaluation process that has the potential to add
value. So scoping studies may quite legitimately be optimistic in their assumptions.
Strategies adopted at later stages of evaluation may be different from those used in
the scoping study, without implying that the scoping study was incorrect. Anecdotally,
however, there seem to be many instances where the opposite is true, and either the
credibility of those conducting the scoping study is called into question, or subsequent
studies are constrained to the strategy used in the conceptual study; worse, they are

3. Targets for metal or ore production are often used as value measures, but they must be recognised as imperfect
surrogates for value – maximising metal output does not necessarily maximise value. Metal targets may be a
useful way to align an operation’s plans with the corporate strategy, but only after the value-optimising plan has
been identified so that optimum (not necessarily maximised) metal targets can be reliably specified.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 127
 CHAPTER 7 | Cut-off Derivation Methodologies

expected to deliver similar financial outcomes.4 In many cases, the scoping study
is becoming akin to a high-level prefeasibility study. It is common to find projects
that move straight from the scoping study to a full feasibility study without a formal
prefeasibility study, which is seen to save time and costs for getting a good project
into production. Both of these study types are discussed in more detail below, but the
implication is that, whether it is expressed in the scope of work for the scoping study
or not, there is a growing expectation that it will identify the preferred option for a
feasibility study and will be implemented. The scoping study is therefore becoming
the de facto decision point for operating strategies for a growing number of projects.
If there is a requirement – express or implied – that the outcomes of a scoping study are
as close as possible to the optimum at a later stage of investigation, it may be a necessity
to conduct a high-level strategy optimisation (which will include cut-off optimisation) at
this stage of evaluation. The quality of the data available is usually, almost by definition,
inadequate for this to be reliable, yet it is an increasingly common requirement.
If, on the other hand, it is only necessary to demonstrate that further investigations
are warranted, some form of full-cost break-even cut-off specification may be adequate.
If there is a reasonable amount of tonnage and grade data, a Mortimer-style analysis
might be a better option, particularly if the deposit is low-grade. If the case selected for
evaluation does not then demonstrate viability and a better case must be found, a high-
level optimisation style of study might be necessary, investigating a few major value
drivers such as pit size, cut-off, mining method and production rate.
If a strategy optimisation is conducted at this level, it is essential to include caveats
allowing for major changes in strategy without loss of credibility as more information
becomes available, even if the study is expected to identify the optimum plan.

Prefeasibility study
The prefeasibility study (PFS) should have a two-fold purpose:
1. demonstrate to a higher level of confidence (typically ±20–25 per cent) that the project
is technically and economically viable
2. evaluate a wide range of options to ensure that the best is selected for further detailed
study and potential implementation.
The reality is that the case selected for subsequent stages of study will be the one built
and brought into operation if it is proven to be feasible. Obviously it is desirable that the
most valuable option is the selected, investigated further and built.
In practice, many studies will evaluate a limited number of options for various design
parameters, but only considering variations from the base case for one parameter at a
time, rather than concurrently; it is deemed too costly and time-consuming to evaluate
all combinations of strategic options. This need not be a burden – most of the information
and relationships will be generated anyway – and the financial benefits of optimising
the strategy with the evaluation at the PFS stage can be far greater than the extra cost.
A full strategy optimisation study should be seen as an important and integral part of
the prefeasibility study. Some prefer to interpose an options study between the scoping
study and PFS. With this approach, the latter becomes a study of a single option or

4. As noted, it is legitimate for a scoping study to be optimistic, but that would not be appropriate for later, more
detailed stages of investigation. Somewhat poorer outcomes in later studies should not be unexpected.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 128
 CHAPTER 7 | Cut-off Derivation Methodologies

limited number of options, but less detailed and accurate than a full feasibility study.
For the purposes of this discussion, the semantics are not important. What is important
is that a comprehensive evaluation is conducted at the PFS level prior to committing to
the option to be evaluated in the feasibility study.

Feasibility study
The feasibility study (FS) is intended to demonstrate the technical and economic viability
of the project at a level of accuracy (typically ±10–15 per cent) sufficient to proceed with
the project.
So-called optimisation studies may be done after a major FS, but typically these merely
investigate ways of reducing costs or improving the efficiencies of the selected strategy,
rather than look for better strategies. Additionally, if the study results are positive, the
identified plan will be implemented quickly; delays for studies of other options are
normally unacceptable. From a public relations point-of-view, there will have been
much corporate ‘face’ invested in the published outcomes of the prefeasibility and
feasibility studies. Announcing a major change of strategy soon after the publication of
such outcomes will be politically unacceptable, destroying the company’s credibility.
If the best strategy has not been identified by the PFS and becomes the focus of the FS, it
is unlikely to be found during the FS.
The policies identified as best in the PFS, such as production rates and mining and
treatment methods, will normally carry forward into the FS, only being changed if
more detailed investigations expose previously unidentified problems. Yet in many
studies, the cut-off will have been determined as a break-even, and this may well alter
if price and cost projections have changed since the previous study. Depending on how
much the orebodies change – both because of cut-off changes and additional geological
information obtained – it may become necessary to completely rework other strategic
decisions. If cut-offs have been determined by some sort of optimisation methodology
(such as Lane’s simple methodology or a full strategy optimisation evaluation) instead
of break-even, and there are significant changes in price and cost forecasts, an early part
of the FS should be an update of the PFS optimisation. This update should use the latest
data and forecasts to assess whether the major strategic decisions going forward into the
FS should be changed. It may need to be repeated as forecasts change and information
becomes available during the course of the study; the FS may even need to be halted and
the wide-ranging PFS redone using the new information.
As noted in the discussion on scoping studies, there is a developing trend for projects
in some circumstances to go straight from a scoping study to a final FS without a PFS. In
such circumstances, a strategy optimisation study should form the initial stages of the
FS to ensure that the FS is evaluating the best option, though, as indicated, this is rarely
done. The author has seen cases where the strategy optimisation is conducted towards
the end of the FS, the strategies of which have already been publicly announced. This is
apparently done simply to tick a box to say that it has been done, with little chance that
the outcomes of the optimisation study will change the direction of the FS.
The feasibility study becomes the official LOMP until the planning processes for the
operational phase of the project start generating new LOMPs on a regular basis.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 129
 CHAPTER 7 | Cut-off Derivation Methodologies

Detailed engineering
While the feasibility study will include mine, plant and infrastructure designs to a
level of practicality and detail, its primary purpose as the name implies is to verify the
technical feasibility of the proposed designs and processes, and derive quantities to
support the predicted financial outcomes. It will usually not produce plans that can
be immediately implemented for mining and construction. Detailed engineering and
design are required to convert the principles in the FS into construction and mining
plans for practical implementation.
By this stage of the development process, mining and plant capacities should be well
defined. With improving knowledge of the mineral deposits and updated price and
cost forecasts, it may be necessary to revise cut-offs at the detailed design phase. Long-
term mining plans may require updating using optimisation techniques. Simpler cut-
off models may be appropriate for shorter-term plans developed within the context
of the longer-term plans. If the nature of the deposits and mining methods permit, a
simple Lane-style cut-off optimisation may be appropriate. If not, a Mortimer-style
cut-off specification will always be preferable to a simple break-even analysis. Lane-
style balancing cut-offs should also be identified – in the light of production capacities
being installed, these may be more appropriate if use of Mortimer-style or break-
even cut-offs were to result in only one stream of the processes operating at capacity.
A balanced operation, with two or more parts of the system at capacity, will often
generate better values. A Lane-style opportunity cost should also be considered in
break-even formulas so that, if the ore-handling and treatment processes are operating
at capacity, marginal ore adds value over the life of the operation. If prices have
increased greatly, the cut-off should not be depressed in a break-even manner to the
extent that ore grades fall too low and product targets cannot be met – a balancing cut-
off may again be more appropriate.
In practice, there is a focus on only the short-term requirements to commission the
mine so that medium- and long-term operating plans are not developed until the mine is
in its production phase. This may lead to reduced output early in the mine life after the
initial ore stocks are depleted; the feasibility study’s long-term plans will have become
inappropriate as development, production and knowledge of the orebody progress and
improve, but the long-term planning processes are not yet in place to handle this.
All aspects of the ongoing short-, medium- and long-term planning processes must
evolve during the construction and commissioning phase so that the transition to
production is seamless. From a planning point-of-view, the preproduction construction
and development phase should be treated as if the mine and plant are already operating.
The following sections recommend how planning should proceed at an existing operation
and how the cut-off derivation and strategy optimisation processes fit into this.

A RECOMMENDED PLANNING PROCESS AT OPERATING MINES

In the following discussion, plans and planning refer to all technical activities supporting
the operation and the outputs generated by these activities. They include such items
as geological data gathering, geological interpretation and models, mine designs and
schedules. The following integrated planning process has been found to deliver good
outcomes at a number of mines where similar processes have been implemented. The
key components of the recommended planning process are:

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 130
 CHAPTER 7 | Cut-off Derivation Methodologies

•• strategy options analysis (SOA)

•• life-of-mine plan (LOMP)
•• five-year plan (5YP)
•• two-year rolling plan (2YP)
•• annual budget (budget)
•• shorter-term detailed operational plans.
The process works in a hierarchy from long to short term. Long-term plans set the
overall strategic direction. Short-term plans provide progressively more detail and
accuracy at the front end of the long-term plans. Approvals processes also work from
long- to short-term and are an integral part of the process.
The importance of focusing planning activities on the corporate goal has already been
indicated. Long-term plans must focus on delivering that goal in the long run. Short-
term plans must indicate in detail how operations in the near future will contribute to
achieving long-term plans; if this is not the case, the plan will fail to deliver the corporate
goals long term, and in the worst case may preclude attaining them.
Mining schedules should allow for indirect operating functions impacting on the
schedule such as backfilling, and major planning and other non-operating activities in
the critical path such as diamond drilling, assaying, geological interpretation and plan
preparation. As well as the operations, it is also necessary to plan the planning. Planning
activities and formal approval processes are often in the critical path of establishing a
new mining area, but not considered when mining schedules are being developed, with
potentially serious impact on achieving the mining plans.5
Formally documenting the planning processes at all levels of the organisation is
essential. This is important given the rapid turnover of staff at many operations.
Documentation may seem to be unproductive, but such processes should not depend on
the knowledge or capabilities of individual people during times of high staff turnover.
Investing in an effective planning process should not be squandered because of lack of
continuity of staff in key roles. Formalising documentation standards and storage and
archival of plans is also essential so that plans and supporting information are available
to all concerned, both while they are current and in the future.
In the following discussions of different stages in the planning process, the lengths
of the plans, the time periods reported in each and the frequencies at which they are
carried out are suggested. Before the advent of computerised scheduling packages,
schedules would have been created manually at this frequency and level of granularity
in well-planned operations. Modern computerised systems permit both a much finer

5. The author has from time to time worked with planning staff at underground operations without formalised
planning processes as recommended in this chapter. To emphasise the lead times required, a skeleton schedule
of the critical path activities is developed, working from the earliest planning activities to the first blast in a new
mining area. The schedule must include planning and operating activities, and allow for two or three detailed
cycles of obtaining geological information (mapping, samples, diamond drilling, etc); obtaining survey and
assay information; updating geological databases; geological interpretation and model updates; designing and
scheduling development, development; and finally, designing the stope drilling pattern, drilling, designing the
charging pattern, charging and blasting. The lead time from the first geological data gathering to first blast,
depending on the operation and mineralisation is four to six years (hence the recommendation later in this
chapter of a 5YP). This lead time is often a revelation to improperly trained planners, who are attempting to do
all this in a fraction of the time and wondering why their plans and schedules fail to work.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 131
 CHAPTER 7 | Cut-off Derivation Methodologies

level of detail in the internal workings of the software and a full integration of short-,
medium- and long-term schedules in the one working file. As always with software
applications, users should be aware of the distinction between precision (relating to
the internal calculations) and accuracy (the overall effects of data uncertainties and the
simplification of relationships built into the calculations). The computer’s precision will
be the same at all points in time, but the accuracy of any schedule produced will be
expected to decrease with time into the future.
The author suggests that the frequency and level of granularity recommended, though
carried down from pre-computerised times, are still appropriate for official plans and
schedules distributed outside the confines of the planning office. This will avoid the
precision of the software’s calculations creating a false impression of the accuracy of
the outputs. Chiefly, it will help establish a disciplined and prepared planning process.
How the cut-off derivation and strategy optimisation processes fit into the planning
process is discussed after describing the recommended process.

Strategy options analysis

An SOA is the first step in identifying the optimum LOMP to be adopted and followed
by all detailed levels of the planning process. The SOA is the strategy optimisation
evaluation that is the subject of Chapter 6. It evaluates, at a high level, the impacts on
value, or a number of different measures of value, of all the strategic decisions that the
company can make, separately and together.
Key issues to be evaluated include, but are not limited to:
•• whether various resources are included in or excluded from the mining plan
•• mining and treatment capabilities/capacities and the timing of potential upgrades or
changes to methods and processes
•• sizes of ultimate pits (open pits) and cut-off grades (underground and open pits)
•• mining sequences6
•• the impact of changes in external or uncontrollable factors such as metal prices and the
existence of additional resources that may be discovered by exploration.
A full SOA is not normally done on an annual basis at an operating mine. It should be
reviewed annually by qualified and experienced expert planners as part of the planning
cycle, so that it remains a regular component of the process and ensures that the
approved LOMP continues to be the best long-term plan. It should not be an irregular
ad hoc process that has to be reinvented each time it is deemed necessary. It may be
useful to rerun the previous optimisation model each year with updated price and cost
forecasts (and perhaps updated tonnage and grade information for mining blocks) but
without time-consuming changes to the model logic, to check if previously identified
optimum strategies have changed. Major updates and reworking of the SOA would be
expected perhaps every three to five years, triggered by, for example:
•• significant changes in metal price forecasts
•• changes in corporate aspirations or strategic goals

6. Sequences and schedules are related terms, often used synonymously. In general in this book, sequence refers to
high-level ordering of mining activities without specific reference to time frames, and schedule refers to a more
constrained and detailed tabulation of activity quantities over time, resulting from the application of sequencing
and rate decisions. There is no clear boundary between the terms and some degree of overlap will often exist.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 132
 CHAPTER 7 | Cut-off Derivation Methodologies

•• discovering substantial new resources

•• identifying new options for processes, markets, mining strategies, etc.

Life-of-mine plan
The LOMP is the formally approved long-term plan for the mine or business. Normally
it would be selected after conducting an SOA as the plan best delivering the corporate
goal. If an SOA has not been done or is out of date, a formal LOMP is still required to
establish the framework within which all other short-term plans are developed.
The LOMP should at least be reviewed annually as part of the planning cycle. This
should be done by qualified and experienced planners who can recognise factors that
should trigger a complete reworking of the plan. This would take account of constraints
identified in short-term plans or resulting from actual events and changes identified in
the SOA. Unless a major change in strategic direction has been signalled by an update or
reworking of the SOA, changes in successive LOMPs should be minor.
Depending on the level of detail in the SOA and when it was last completed, the LOMP
may be extracted directly from the SOA, or may require some additional work to finalise
it. Either way, it should be seen as distinct from the SOA – an outcome from the SOA,
not the SOA itself.
The LOMP establishes the planning framework over time for such items as:
•• cut-off grades (underground and open pits) and sizes of ultimate pits (open pits)
•• mining rates and capacities including timing of changes
•• mining sequences and schedules, especially specifying when new mines, mining
areas, orebodies, pushback stages and the like are to commence
•• treatment rates or capacities for all plants, including timing of changes
•• sources and destinations of all products and intermediate products (for example,
if the company has its own smelting and refining facilities, the concentrates to be
smelted in-house or sold)
•• major infrastructure additions such as mine shafts and mill expansions
•• project and sustaining capital requirements to implement the plan
•• operating costs and associated resources, such as mobile equipment fleets and labour.

Five-year plan
The 5YP is an essential part of the planning process at operations with more than a
few years of life remaining. It forms a critical medium-term link between the high-
level strategies in the LOMP and the detailed short-term implementation plans. A five-
year time frame will provide sufficient lead time for all or most long-lead items to be
adequately planned, but in some cases the medium-term plan may need to be longer –
up to ten years in exceptional circumstances – or somewhat shorter. The key issue for the
time frame of this plan is establishing a reliable link between the LOMP and short-term
plans. For this discussion, five years is assumed to be typical.
The author recommends that the 5YP should be recognised as an essential component
of the planning process, with formal approval of the 5YP being part of the annual
planning cycle occurring between approvals of the LOMP from which it is derived,
and the short-term plans that emanate from it. The 5YP would be generated annually,
with quarterly periods reported, though longer periods may be suitable in the later

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 133
 CHAPTER 7 | Cut-off Derivation Methodologies

years of the plan. Monthly time periods can give a false sense of the level of accuracy
of plans in later years. Annual values, while satisfactory for high-level summaries,
may be too coarse for more detailed consideration of component activities, especially
earlier on.
The appearance of a new mining area in the last year of a new 5YP would be the flag
for designing or initiating such elements as:
•• underground
•• block infrastructure and access development with major ventilation circuits
•• stope boundaries, general layouts of stope development and stoping sequences
•• open pits
•• for new pits: environmental studies, negotiations with other stakeholders and
approvals processes
•• for new pits and new pushbacks in existing pits: overall sequencing and location
of temporary and permanent access ramps
•• underground and open pits
•• additional diamond drilling, metallurgical test work, geotechnical data gathering
and obtaining information that is required to carry out more detailed designs as
production from these areas draws nearer.
When new mining areas have advanced to the third and fourth years of the 5YP,
major design and scheduling parameters should be well established. Shorter-term
development and waste stripping schedules should be taking account of the 5YP
production schedules to include the initial access for these new areas, so that they are
available to produce when scheduled.
Production scheduled in the 5YP should be consistent with optimised schedules
identified in the SOA and formalised in the approved LOMP.

Two-year rolling plan

The 2YP, like the 5YP, is a critical link between the long-term plans and short-term
implementation plans and schedules. It provides a higher level of detail supported by
detailed engineering work at the front end of the 5YP. The author recommends that the
2YP is formally updated quarterly and reports activities on a monthly basis.
The deliberate inclusion of rolling in the name indicates that it is an ongoing part of the
short-term planning processes and is not just done once a year as part of a budgeting
cycle. Technical staff would recognise that updating short-term plans and schedules on
an annual basis is too infrequent – too much can change. Most operations would have
more frequent updates of plans and schedules when situations change, but often done
on an ad hoc basis. Formalising regular updates requires both operators and planners to
regularly look at the effects of current operating and planning issues up to two years
ahead, thereby of avoiding actions that may be expedient in the short term but create
major problems in the longer term.7

7. In some parts of the world, staff turnover rates are such that operating managers and planning staff are unlikely
to remain in the same roles for as much as two years. The organisation must therefore ensure that short-term
expediency at the expense of long-term performance is not rewarded. Having in place formalised planning
processes like those recommended here may assist with this.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 134
 CHAPTER 7 | Cut-off Derivation Methodologies

The appearance of a new stope or bench, for example, in the last quarter of a new 2YP
would be the flag for commencing the design of:
•• underground
•• stope development
•• ventilation networks for development and production
•• general layouts of drilling designs and blasting sequences
•• open pits
•• sequencing of operations on benches to maximise flexibility and the number of
working places available, and to avoid where possible operations on one bench
preventing operations on another
•• timing of the creation and removal of temporary ramps to maintain access and
flexibility
•• although not specifically related to milestone events in the pit, open pit plans
must also plan for changes or additions to surface access roads, waste dumps,
stockpiles, infrastructure construction, surface drainage and pit dewatering
•• underground and open pits
•• additional diamond drilling, mapping, geotechnical and geometallurgical data
gathering, etc if required.
When the start of these activities has advanced to being 15 to 18 months away, all
design and scheduling parameters should have been decided. Mining schedules in at
least the first half of the 2YP should allow for:
•• grade variations within stopes or benches based on planned blasting sequences
•• underground: backfilling and associated curing times required in adjacent stopes
before final stope preparation and blasting can commence
•• open pits: detailed planning and sequencing of operations on each bench
•• planning and other non-operational activities in the critical path, such as surveying
and drilling design after stope development or a bench blast is complete.
Each 2YP should be approved as realistic and achievable by management, planning
staff and operators responsible for implementing it. All potential problems with the
schedule should be identified and solutions found at least 12 to 18 months ahead of
time. It is counter-productive to approve a plan with known problems that make it
unachievable.
As with the 5YP, the duration of the 2YP could be longer or shorter to suit the situation.
The well-settled period at the start of the plan should be of the order of 12 months,
but a shorter time may be acceptable in some operations.8 There are perhaps two
considerations when defining this duration.
The first is operational, to ensure that there is adequate time for planning once a new
stope or mining level appears in the last quarter of the schedule, so that designs are
locked in and schedules well settled for a suitable time period into the future. This is to
ensure that the mine can operate with as few unpleasant surprises as possible, implying
that detailed designs and schedules need to be generated well ahead in order to identify

8. If a good planning process has not existed in the past, early efforts to establish a process will start with short time
frames with the aim of extending these over time. We must plan the planning as well as plan the operations.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 135
 CHAPTER 7 | Cut-off Derivation Methodologies

problems when they are far enough in the future to be fixable without causing undue
stress on the schedules and personnel.
Of course, not all possible problems can be foreseen and avoided. Geological and
geotechnical knowledge will always be incomplete and shortages of technical staff will
continue to ensure that plans developed are not as good as they could be. However, if the
operation is continually lurching from crisis to crisis, it is a sure sign that the planning
(in the broad sense) is inadequate and should be improved. Some do not realise that
continual crisis management does not have to be the norm in mining operations.9
The other consideration is organisational, but has similar rationale. The following
subsection discussing the budget suggests that the budget schedule should simply be
the activities in the 12-month budget period within one of the 2YPs generated shortly
before the start of the budget year. The 2YPs should ideally allow sufficient time for
designs to be completed for all stopes appearing in the budget year in the 2YP from
which the budget is extracted. This may effectively extend the well-settled schedule out
to 15 or 18 months. This may not be practical in some situations, but, while recognising
that budgets are frequently obsolete before the start of the budget year anyway, the ideal
would nevertheless be to have the best possible information regarding mining activities
in the budget. On this basis, the minimum duration of a 2YP would be 18 months, but
targeting a minimum two-year time frame is often desirable.

The annual budget

If the SOA/LOMP/5YP/2YP process is in place and plans are formally approved
sequentially from long term to short term, then plans for all time frames are continually
evolving, but are also realistic, achievable and continuously approved. The budget
plan then very simply becomes the plan for the budget year in a version of the 2YP
created three to six months before the start of the budget year. This should generate the
budget schedule with no additional work, since the rolling process of plan and schedule
generation and associated approvals ensures that it is realistic, achievable and aligned
with the corporate goal. There is no undue focus on the budget process for weeks or
months to distract planners and operators from their primary tasks of planning and
operating effectively and efficiently.
Physical quantities in the realistic and corporate-goal-oriented budget plan should
drive the budget costs via appropriate cost models. In the same way as physical plans and
schedules have increasing detail and accuracy as the time frame reduces, so too should
cost driver models. The cost models developed for the SOA will typically provide the
broad framework for what is relevant in the budget at its level of detail. The discussion
in Chapters 2 and 6 has given some indications of how costs behave, and more detailed
discussions in Chapter 8 expand on a number of issues.
Given that there is often pressure at budget time to reduce costs, it is highly desirable
that models of both physical and cost behaviours should enable identifying how
reductions can be achieved. Ideally, budget costs will be derived from production
physicals by a two-stage process. As noted in Chapter 6, focusing on and rewarding
arbitrary cost reductions without considering the impact on revenue can at times result

9. The author suggests, without proof, that, since mining tends to be a cost-focused industry and the benefits of
good planning are often unrecognised and difficult to quantify, there will often be a tendency to under-resource
the planning functions in many organisations, regardless of whether there are staff shortages or not.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 136
 CHAPTER 7 | Cut-off Derivation Methodologies

in revenue reductions larger than costs saved. Cost savings therefore can result in a
reduction in value or profit, rather than the intended improvement.
Physical quantities of inputs should be derived from activities by consumption rates.
These process inputs will include such items as labour hours, drilling consumables,
explosives, fuel, power, crushing and grinding consumables, reagents and so on. Costs
should be derived from these process inputs and the unit costs of each input. It is only
by using such a form of zero-based budgeting that requests to reduce budget costs can
be reliably dealt with. Cost reductions can only be achieved by:
•• Reducing the unit costs of inputs – this is usually a purchasing and supply issue beyond
the control of operating and planning staff; however, as with so many things, you get
what you pay for. Minimising the purchased unit cost of supplies and consumables
is counter-productive if low price results in low quality, so that the total cost of
providing a particular capability in the operation is increased because of greater
consumption rates. In the extreme case, the impact is not just on the cost of providing
a service but also leads to reduced availability of production equipment, and hence
lost production and lost revenue.
•• Reducing consumption rates – if these have been explicitly identified, there may be
some opportunity to reduce consumption rates, but in each case there will be a fully
efficient, underlying consumption rate that cannot be improved upon. Any further
attempt at reduction, for example, by artificially or arbitrarily restricting supply, can
only result in a reduction in the associated activity.
•• Reducing the physical activities scheduled – if the plans in the hierarchy from long
term to short term have been developed to achieve the corporate goal and been
previously approved, any reduction in activities must result in deviating from the
plan that delivers the corporate goal. Planning staff should have the capability to
identify the effects of reductions in physical activities in the budget year on the
long-term strategies, so that decision-makers requesting cost reductions can be
made aware of their impact on achieving the corporate goal and make trade-off
decisions.
•• Deferring sustaining capital – if the physical assets of the mine are not maintained in
the state required to continue to meet operational targets, diminished capacity in
future years (for example, by reduced availability) will result in reduced levels of
physical activities.
Without linkages of integrated models in place – both physical and financial – it is
impossible for staff to assess the impacts of short-term goals on the long-term goal,
and provide information to decision-makers. Without integrated schedules, the impact
of reducing near-term activities on later years cannot be assessed. Similarly, without
integrating the costs and revenues, the financial impacts cannot be assessed. The
intention is to avoid making short-term knee-jerk-response plans that create problems
by failing to look to the long-term health of the company.
There will be a need for some formal budget development and approval processes
specific to the budget, but if the LOMP/5YP/2YP/budget process is implemented as
recommended, the additional work required at budget time can be significantly reduced,
and more reliable plans, both physical and financial, will result as a matter of course.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 137
 CHAPTER 7 | Cut-off Derivation Methodologies

Shorter-term operational plans

Detailed operating plans, consistent with the 2YP and budget, should be developed to
ensure that all activities required to implement the approved plans are scheduled, with
all necessary inputs available as required.
These plans would typically be for three-monthly periods that are updated monthly,
with time frames of weeks and days, or even shifts. The schedules should identify
individual items of mining equipment, crews, work locations and other relevant details.

THE IMPORTANCE OF CORPORATE PLANNING POLICIES

To derive all the benefits of a good overall planning process, it is essential that policies
and procedures are supported and driven by the company’s most senior management,
and faithfully implemented at all levels. The LOMP should be specified at a high level
to deliver the corporate goal and an SOA will have defined the LOMP that best achieves
the goal. The 5YP, 2YP and budget each provide more detailed and better engineered
plans at the front end of the LOMP. Each ensures that all planned activities are focused
on achieving the goal of the next longer-term plan, and thus ultimately the LOMP.
Without that top-level commitment from the longer to the shorter term, everything
else is a waste of time. Many of the planning activities will have to take place so that
the operations can continue with a degree of future planning. However, without the
framework of the formalised process and a LOMP focused on the corporate goal to
guide short-term plans, it is unlikely that the plans will actually achieve the stated
goal. Rather than top-down, the annual planning process will more likely become
a bottom-up process. Each department within an operation or each operation in a
corporate group will typically build its own plans without any clear guidance of
what the company requires of it other than generic ‘reduce costs and increase output’
directives.
The corporate long-term plan will then be a rolled-up summation of all the individual
operations’ LOMPs. When these are aggregated, the capital requirements will be
deemed too high and arbitrary cuts or deferrals will be decreed. Unfortunately, there is
typically no time available for plans to be reworked to account for the reduced capital
spending, so the benefits of capital expenditure are built into the forecasts without the
capital having been spent to achieve them. Such plans are likely to fail, and as the old
adage says, failing to plan is planning to fail.
Similarly, without a good model linking physical mining activities, required physical
inputs and resulting operating costs, any operating cost reductions can only be
arbitrary. If budgeted unit costs are based on past history, there is no way to see how
realistic any reduction might be. If unrealistic arbitrary reductions are made to satisfy
demands from upper management, there are two extremes of after-effects during the
budget year:
1. if the primary focus is on production activities, costs will be over budget
2. if the primary focus is on costs, physical activities will be below budget, with the
dangers of
•• subverting the long-term plans (if there is a real plan to be subverted)
•• reducing the revenue generated by more than the costs saved.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 138
 CHAPTER 7 | Cut-off Derivation Methodologies

Is it any wonder that many corporate long-term plans are unrealistic and not respected or
referred to by the people who are supposed to be implementing them at the operations?10
There must be a top-level commitment to a planning process that cascades from the longer
to the shorter term, with all plans focused on achieving the corporate goal.

INCLUDING CUT-OFF DETERMINATION AND STRATEGY OPTIMISATION

IN THE PLANNING PROCESS

Strategic options analysis

A strategy optimisation evaluation (SOA) is not normally done annually as part of the
planning cycle at an operating mine. Major updates and reworking would occur every
three to five years, triggered by anything that makes the previous plan obsolete such as:
•• significant changes in metal price forecasts
•• changes in corporate aspirations or strategic goals
•• discovering substantial new resources
•• identifying new options for processes, markets, mining strategies, etc.
By definition, a full cut-off and strategy optimisation is both the key component and
purpose of conducting the SOA.

From life-of-mine plan to budget

Once the strategy optimisation is completed, the optimum mine plan will be
formalised as the LOMP. The LOMP will identify not only the immediate cut-offs
to be applied but also important inputs for ongoing cut-off derivation such as pit
limits and pushback staging in open pits, mining methods underground, mining and
treatment rates, products and the like. These will become the overall strategic decisions
implemented between successive strategy optimisations, which may be several years
apart. Nevertheless, cut-off specification or optimisation is an ongoing process in the
more detailed shorter-term planning processes.
The cut-offs in the LOMP will be planning cut-offs, as defined in Chapter 2. As noted
there, normal variations in mining would be expected to result in times when the
specified planning cut-off is temporarily inappropriate and a short-term operational
cut-off should be applied. Referring back to the freeway analogy in Chapter 2, the
LOMP cut-off policy represents the freeway to get from where we are to where we
want to be, whereas operational cut-offs in the shorter-term plans may be thought of
as changing lanes on the freeway from time to time to pass slow-moving traffic and
negotiate blockages along the way. So how should cut-offs be derived at an operating
mine within the context of an existing LOMP?
A simple break-even cut-off will rarely be optimal. A marginal break-even cut-off
may be appropriate as an operational cut-off to account for short-term fluctuations in
availability of ore to feed the treatment plant.11 If no other material is available, any

10. The author suggests that ‘challenging but achievable’ targets have a less than 50 per cent chance of being met.
11. In Lane parlance, this is the mining-limited break-even cut-off. There are, however, situations where this is the
optimum long-term planning cut-off. The marginal break-even should not become the planning cut-off through
a failure to identify what the optimum longer-term cut-off should be.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 139
 CHAPTER 7 | Cut-off Derivation Methodologies

mineralisation whose value pays for its downstream treatment and product-related
costs may be added to the ore stream and will add value. Short- and medium-term plans
should be focused on returning the operation to the point where it can operate at the
planning cut-off identified by the LOMP. The following subsections identify processes
to use at different stages for both open pit and underground mines.

Open pit shorter-term plans

For an open pit operation, the LOMP sets pit limits, pit staging and overall mining
sequences, as well as mining and treatment capacities. The 5YP, 2YP, budget and
shorter-term operational plans will refine the detail of these plans. It is likely that,
with each increasing level of detail and shorter period of time considered, there will be
variations from overall LOMP plans. The top-down process of planning time frames
and approvals will set the overall direction to be implemented by the short-term plans.
In particular, it is implicit for short-term plans that capital will not be spent to increase
mining, treatment or marketing (that is, rock, ore and product-handling) capacities.
Mining capacities could, however, be reduced by mothballing some of the mining fleet.
This does not preclude the possibility that capital expenditure could be justified on
the basis of benefits obtained within the shorter-term plan being considered, without
notionally considering the longer-term plans. In that case, however, the effect of the
capacity change on the overall SOA, and hence LOMP, may be significant enough to
affect the short-term plan under consideration. The prima facie case is that short-term
cut-off evaluations should assume the existing capacities as upper limits and not allow
for possible expansion capital expenditure.
There is no reason not to conduct simple Lane-style cut-off optimisation evaluations
for all short-term open pit plans once the pit limits, mining sequences and capacities
have been defined by the long-term plans. In practice, particularly with the budget
and shorter-term plans, the process may be simplified as more parameters become
fixed by longer-term plans, but the full simple Lane methodology should remain the
underlying reference point for any such simplifications. Any simplifications should
not be allowed to become arbitrary rules in their own right – they may become
inappropriate as conditions change in the future – but should always have firm
theoretical support. Chapter 5 indicates how Lane’s simple methodology may be
applied in short-term situations:
•• current values for each cost component of Lane’s limiting cut-offs can be derived
from the cost records
•• the opportunity cost can be estimated from the latest LOMP
•• the latest price forecasts will be available
•• capacities for each process stage are known
•• metallurgical recovery relationships are known
•• tonnage and grade versus cut-off relationships can be formed for the rock to be
mined
•• the optimum cut-off for the next increment of rock or period of time can be
determined.
In principle, the time frame is irrelevant – it may be days, weeks, months or quarters.
All schedule durations in plans with time frames shorter than the LOMP can be catered
for in a simple Lane-style analysis. Therefore, assuming that the style of mineralisation

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 140
 CHAPTER 7 | Cut-off Derivation Methodologies

is amenable,12 the prima facie argument is that a Lane-style analysis should be done for
all open pit plans shorter than the LOMP.
The optimum cut-off will often be a balancing cut-off,13 frequently the mining–
treatment (rock–ore) balancing cut-off. If the mining operations are working through
a low-grade portion of the deposit, the operation may become mining-limited. When
mining rock at the maximum rate, it is not possible to keep the treatment plant operating
at full capacity. In this case, the cut-off can be lowered to Lane’s mining-limited break-
even cut-off, but no lower.
If the mining operations are working through a high-grade portion of the deposit,
the balancing cut-off will increase and higher-grade ore can be treated. Yet at many
operations, rather than increasing the cut-off to account for the short-term variability in
the orebody, the cut-off is kept constant and the number of trucks operating is reduced,
thereby reducing the mining capacity to maintain a rock–ore balance at the specified cut-
off. This may have the apparent advantage of reducing costs. Planners should consider
carefully, however, whether the cost savings from reducing the mining rate are greater
than the revenue gain from treating higher-grade ore using the higher balancing cut-off.
Reducing costs is not the best tactic if it reduces revenue by more than the cost saving.
In the rare situations where Lane’s methodology is inapplicable, a Mortimer-style
cut-off specification is preferable to a break-even analysis. A Lane-style opportunity
cost should also be included in break-even formulas so that, if the ore-handling and
treatment processes are operating at capacity, additional ore adds value over the life
of the operation. Lane-style balancing cut-offs should be identified as accurately as
possible. If prices have increased somewhat, the cut-off should not be depressed in a
break-even manner to the extent that ore grades fall too low and product targets cannot
be met; one of the balancing cut-offs will be more appropriate.

Underground shorter-term plans

As noted in Chapter 5, applying Lane’s simple methodology to underground operations
can be difficult. The methodology assumes that the mining sequence is specified, and cut-
offs are derived for that sequence using the formulas for limiting cut-offs and the processes
for identifying balancing cut-offs. But in an underground mine, cut-offs define the size and
shape of the orebodies, and in many cases, the mining sequence will depend on the nature
of the orebodies defined; a Lane-style cut-off optimisation, which depends on sequence,
therefore depends on the cut-offs. The process has become circular and may be impractical
for a simple Lane-style analysis; however, as indicated, some cut-off-dependent inputs to
the analysis may be accounted for by the iterative process. Chapter 5 also indicates that
there are some styles of mineralisation and mining methods amenable to a Lane-style cut-
off optimisation. In particular, this occurs where there is a single production front moving
in one direction, with no returning to extract additional ore from parts of the resource that
the production front has already moved through, such as with sublevel caving.

12. Refer to Chapter 5 for a discussion of situations where a Lane-style analysis may not be appropriate.
13. If one process is limiting for an extended period of time, it may have been over-resourced, or another under-
resourced; however, this may legitimately be the case when the mine is highly constrained during early mining
stages or when running down towards closure, or if the cost to increase the capacity in one process is not justified
by the additional value generated from the new optimum mine plan.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 141
 CHAPTER 7 | Cut-off Derivation Methodologies

Where the mining method allows it, a Lane-style analysis should be applied to optimise
cut-offs for shorter-term plans where the overall development rates and ore handling
and treatment capacities have been established in the LOMP and long-term plans. The
rationale described for open pits is directly applicable for underground operations. There
are, however, some additional considerations. As noted in Chapter 5, mining in a Lane
sense relates only to major access development, while most underground operations
deal with ore and are therefore treatment activities. Particular care must be taken to
ensure that costs and capacities are attributed to the correct Lane-style process stages.
In the more common underground situations where Lane’s simple methodology is not
easily applied, a Mortimer-style cut-off specification will be preferable to a simple break-
even analysis. A Lane-style opportunity cost should also be considered in break-even
formulas so that, if the ore handling and treatment processes are operating at capacity,
additional ore adds value over the life of the operation. As with open pits, Lane-style
balancing cut-offs should be identified. If prices have increased significantly, the cut-off
should not be depressed in a break-even manner to the extent that ore grades fall too low
and product targets cannot be met – a balancing cut-off will often be more appropriate.
The discussion in Chapter 3 regarding the time required to change a cut-off in an
underground operation, particularly upwards, should also be noted. Short-term
responses to fluctuations in prices, costs or other inputs may not be practical.

OTHER ISSUES AFFECTING CUT-OFF SPECIFICATION

Planning and operational cut-offs at operating mines

As indicated in Chapter 2, there is a danger that, having identified that it is easier to
fill the mill at a lower operational cut-off, the longer-term planning cut-off is changed
to this lower cut-off. It is perfectly valid to use an operational cut-off that is different
from the longer-term planning cut-off to account for short-term variability in operating
conditions. The shorter-term planning processes must generate realistic short-term plans
and also return the operation to the optimum long-term strategy using the planning cut-
off identified in the longer-term plans.

Elevated and depressed cut-offs

Often one hears comments that a particular operation is using an elevated (or depressed)
cut-off. The implication of the terminology is that the cut-off ought to be a particular
number, but for some reason – which is explainable but not considered justifiable –
another value is being used. In such cases there is usually a rational explanation for why
the cut-off in use has been specified. It will often revolve around the fact that a break-
even calculation has been conducted but the mine plan resulting from using that break-
even as the cut-off is unsatisfactory for some reason. Typically:
•• The mill cannot be filled with ore, so the cut-off is lowered until it is. The cut-off
thereby established will be either a Lane-style mining–treatment (rock–ore) balancing
cut-off or, at the limits, a Lane-style mining-limited (rock-limited) cut-off.
•• The product targets cannot be achieved, so the cut-off is raised until they are. The
cut-off thereby established will be a Lane-style balancing cut-off, either mining–
marketing (rock–product) or treatment–marketing (ore–product).

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 142
 CHAPTER 7 | Cut-off Derivation Methodologies

•• The resulting head grade is too low to make an acceptable profit, so the cut-off is
raised until the desired outcome is achieved. The cut-off thereby established will be
a Mortimer-style minimum profit cut-off.
The underlying fallacy is that the cut-off grade ought to be a break-even grade.
Because the outcomes from such an assumption are unacceptable, there is an implicit
recognition that the break-even model for cut-off derivation is inadequate. The response
is to invoke other cut-off models discussed in this book, but without recognising that
they exist or understanding their rationale. The resulting elevated or depressed cut-
offs are perceived to be aberrations. Additionally, because the models being implicitly
applied are not understood, there is the danger that they are misapplied and the
outcomes are suboptimal. Iterative trial-and-error evaluations will often be conducted
until an acceptable outcome is obtained. Correct application of the appropriate cut-off
derivation model could have achieved a better outcome with more rigour and less work.
In one sense, if the cut-off is the number that distinguishes between ore and waste, as
described in Chapter 2, there is no such thing as an elevated or a depressed cut-off. The
cut-off being used is the cut-off. It is neither elevated nor depressed: it is what it is.
If, in light of Chapters 3 to 6, we suggest that the cut-off ought to be the value that best
delivers the corporate goal, it is proper to use adjectives such as elevated or depressed
if a different cut-off is used; however, taking this view will often result in the opposite
usage of what is encountered. Assuming that the goal of most companies is value
maximisation rather than every-tonne-pays-for-itself, cut-offs in general ought not to be
break-evens. If the cut-off being used has been changed from a break-even starting point
to achieve a mine plan that does deliver the corporate goal, a so-called elevated cut-off is
actually the right cut-off, and the original break-even is a depressed cut-off.
Using a so-called elevated cut-off is often referred to pejoratively as high-grading, with
the sense that it is somehow improper. If this is a result of achieving the corporate goal, it
should be seen as right-grading; traditional break-even cut-offs could then be described
as low-grading. To be consistent, this should be just as pejorative a term as high-grading.
In some jurisdictions, cut-off determination is mandated by law. In such places, the
cut-off ought to be what is decreed, and any variations will be elevated or depressed,
leading to high- or low-grading. Whether the latter terms are pejorative will depend on
whether the deviations are legally permitted and the extent to which they may frustrate
the underlying intent of the law.

Cut-offs and design rules

For the optimisation studies, there must be consistency in the processes used for all
options. Inconsistencies in generating mining reserves data between, for example,
cases with different cut-offs, can lead to incorrect option ranking. Design rules in the
optimisation process must then be carried forward into the ongoing planning and
design to ensure that the operational plan using the latest and most detailed information
honours the selected strategic plan identified using older, less detailed information.
Figure 7.1 illustrates the problem that may arise applying cut-offs to mineralisation
with irregular ore boundaries in an underground situation.
The figure shows an extreme hypothetical situation where the grade distribution in two
dimensions is circular, with grades highest in the centre and reducing with distance from
the centre. Three grade boundaries are shown. It is assumed for practical mining reasons

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 143
 CHAPTer 7 | Cut-off Derivation Methodologies

FIGURE 7.1
Cut-offs and design rules with irregular ore boundaries.
Optimum square stope

2 g/t

4 g/t

6 g/t

that square stopes must be mined. An optimisation evaluation (an SOA) has determined
that the square stope shown is the optimum. We wish to specify the boundary cut-off,
so that future planners will design the optimum stopes after the block model has been
updated with better geological information. What is the design cut-off to be specified?
It could be any of the following:
• 2 g/t, designing to extract the maximum tonnage of above-cut-off material but with
no sub-cut-off material included
• 4 g/t, designing to balance the tonnages of sub-cut-off material included and above-
cut-off excluded
• 6 g/t, designing to extract all the above-cut-off material, and including the minimum
quantity of sub-cut-off material necessary to achieve that.
If the outcomes of the SOA were viewed as hills of value, there would be different
hills for each design rule. Each rule that described the same optimum stope would of
necessity have the same maximum value; however, the maximum values would be
located at different cut-offs across the range from 2 g/t to 6 g/t.
What if the design rule used to generate the reserves for different cut-offs in the SOA
were not passed on to the detailed design engineers?
If the ‘no sub-cut-off material included’ rule had been used for the SOA, the optimum
cut-off would be 2 g/t. If the detailed design engineers lacked guidance with the design
rules, they might decide to use a ‘no above-cut-off material excluded’ rule. If this
were applied to the 2 g/t grade boundary – since the cut-off has now been specified
numerically as that – a significantly larger and lower-grade stope than the optimum
would be generated. Figure 7.2 illustrates this situation.
Similarly, if the ‘no above-cut-off material excluded’ rule had been used for the SOA,
the optimum cut-off would be 6 g/t. If the detailed design engineers had no guidance
regarding the design rules, they might use a ‘no sub-cut-off material included’ rule. If
this were applied to the 6 g/t grade boundary, as the cut-off has now been specified as
that, a significantly smaller and higher-grade stope than optimum would be generated.
Figure 7.3 illustrates this situation.

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 144
 CHAPTer 7 | Cut-off Derivation Methodologies

FIGURE 7.2
Cut-off applied using an inappropriate design rule 1.
Optimum square stope

2 g/t

4 g/t

6 g/t

Sub-optimal
designed stope

FIGURE 7.3
Cut-off applied using an inappropriate design rule 2.
Optimum square stope

2 g/t

4 g/t

Sub-optimal
designed stope

6 g/t

Clearly, design rules used in the optimisation evaluation must carry over into the later
ongoing mine planning and design processes so that the operational mine plan honours
the strategy selected as a result of the SOA and incorporated in the LOMP. How critical
this might be will depend on how irregular grade boundaries within the mineralisation
are and how easy or difficult it is to define stope boundaries that include all above-cut-off
material and exclude all sub-cut-off material. As soon as there is the practical necessity for
the stope design to exclude ore and include waste, the problem potentially exists.
The increasing availability of software that automatically defines stope boundaries is
making it easier to record design rules for underground operations. Rather than being
a verbal description that may be easily lost or forgotten, numerical values can now be
specified for each of the input settings in the software. These can be recorded and reused
at all stages of the planning process. If the mine is already operating and various settings
are in common use, these should be adopted in any SOA undertaken, unless there is
a good reason for doing otherwise. If the mine is not already operating, the settings

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 145
 CHAPTER 7 | Cut-off Derivation Methodologies

used in the SOA should flow on to the detailed design engineers. These may be found
to be unsuitable for the style of mineralisation encountered in practice. It will then be
necessary, initially, to compare the tonnages and grades generated by the software for
a range of cut-offs for both the SOA and detailed planning software settings. This is so
that any necessary adjustments can be made to the cut-off used in the mine to ensure that
mine plans generated using the new design rules, as reflected in the software settings,
continue to honour the strategy in the LOMP. Eventually, it will become necessary for
the SOA to be redone using the new stope design settings.
There should be consistency across all processes that generate information used
as input data for options evaluated in the SOA. Inconsistent design processes could
produce erroneous results. It should go without saying that if there are a number of
design engineers generating information, perhaps at different cut-offs, all of them
should be using the same design rules or software settings for the same orebodies to
ensure consistency of designs.
To illustrate the importance of this, consider the opposite – that each engineer chooses
their own design rules. Suppose the engineers working with 2 g/t and 6 g/t cut-offs chose
to use the rules that generated the same optimum stope illustrated in Figure 7.1. The
same tonnage and grade would be generated for both of these widely separated cut-offs.
Suppose the engineer doing the designs for 4 g/t cut-off also used one or other of these
two rules. We would now have a data set with tonnages and grades at 2 g/t and 6 g/t cut-
offs being the same, while those for a 4 g/t cut-off would be either higher or lower.
At best, the inconsistency will be identified when tonnage–grade relationships are
initially produced by combining the data from each engineer. A potentially costly and
time-consuming reworking of at least some of the input data sets would become necessary.
At worst, the inconsistencies between the data sets would remain undiscovered, leading
to unreliable outcomes of any study based on them.

Different cut-offs in different mine areas

Different cut-offs can be applied to different areas within a mine to increase value (Hall
and Stewart, 2004; Horsley, 2005). Figure 7.4 shows schematically how different cut-offs
in different areas could be optimal in an underground mine.
In the upper part of the figure, two areas with similar mineralisation, mining methods
and costs structures, and therefore the same break-even grade, used as cut-off, are mined.
Different lives for the two areas result in an unprofitable low-production-rate tail that will
be truncated by mine closure if fixed costs cannot be sufficiently reduced to allow it to be
mined profitably. During the life of the mine, production from the longer-life area will
contain material that is above the common cut-off for both areas, but which is of lower
grade than some of the material that remains unmined when the mine closes. The lower
part of the figure shows how the cut-off in this area can be increased so that, at the time the
mine does close, the best possible grades have been mined from both areas.
In this simple example, the cut-off policies to maximise value would be easily
determined. In a real operation, predecessor/successor dependencies between
mining areas and other scheduling and sequencing constraints and options may make
selecting the optimal cut-off policy for all areas over time non-intuitive and non-
trivial. Sophisticated optimisation techniques, some of which are described briefly in
Chapter 13, may be required to select the best strategy to bring into the LOMP.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 146
 CHAPTer 7 | Cut-off Derivation Methodologies

FIGURE 7.4
Value-adding by different cut-offs in similar areas.
Same cut-offs for
Areas A & B
Uprofitable Low
Production Rate Tail
Mine closed end
Production Rate of Year 6

Mining Area B
Mining Area A—Lower-grade Portion

Mining Area A—High-grade Portion

1 2 3 4 5 6 7 8 9 10
Production Year Number

Area A cut-off >

Production Rate

Area B cut-off
Mine still closed
end of Year 6
Mining Area B but all of Mining
Area A High-grade
portion extracted

Mining Area A—High-grade Portion

1 2 3 4 5 6 7 8 9 10
Production Year Number

Applying individual stope break-even cut-offs

Some underground operations, however, take this ‘different cut-offs in different areas’
principle to the extreme, effectively applying a cut-off to each stope. The process makes
use of a stope value calculation that takes account of the stope-specific, rather than
orebody-average, physical quantities and costs to determine whether an individual
stope should be included in the reserve. There is nothing wrong with this in principle
for short-term planning within the context of approved long-term plans, but the way
in which it is implemented at many operations is not aligned with the longer-term
plans and corporate goal. The process is applying a break-even volume cut-off to each
stope. If the long-term plan makes use of a break-even cut-off – that is, the corporate
goal is to ensure that every tonne of ore pays for itself – the break-even stope valuation
process will be appropriate. However, if the long-term cut-off is an optimum rather
than simple break-even,14 the individual stope valuation process has the potential to

14. It is, of course, possible that a simple break-even is the optimum cut-off ; however, this will rarely be the case, so
in this statement, optimum and break-even should be seen as different.

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 147
 CHAPTER 7 | Cut-off Derivation Methodologies

subvert the long-term plan by using short-term tactics not aligned with the long-term
strategy.
One possible way to align such a stope valuation process with the long-term strategy is
to identify the difference between the optimum cut-off and the overall break-even cut-off
and include this difference as an additional cost to be borne by each stope. Alternatively,
the Lane-style opportunity cost could be included in the costs to be accounted for in
the break-even style of analysis. These options may be appropriate if the optimum cut-
off has been identified as some form of break-even, such as a Lane-style limiting cut-
off. If the optimum cut-off is, for example, a Lane-style balancing cut-off, break-even
calculations applied to individual stopes or blocks of mineralisation may be completely
inappropriate. The only justification for such a process would be to exclude any block
that did not generate sufficient revenue to cover its direct variable costs.
The preceding discussion is in the context of, for example, open stopes that are
designed by planning and design engineers some months or years before the stope is
developed, drilled and extracted. Similar rationale applies to all other underground
mining methods. When ore and waste are identified immediately prior to ore extraction,
greater care may be required to ensure that short-term considerations do not prevent
achieving the long-term strategy. Examples include caving methods, where the grade
of the broken material in the drawpoints may be reassessed on a regular basis, and
cut-and-fill, where geologists will often mark up the ore–waste boundary in operating
stopes to direct the mining operations. In both situations, there is often the temptation to
classify as ore anything that is above its own marginal break-even grade – the rationale
being that it is profitable and, if not extracted now, will be lost forever. However, if
the treatment stream (as defined by Lane – the ore handling and treatment processes)
are operating at capacity, any additional material classed as ore must cover the costs
included in Lane’s treatment-limited break-even cut-off – ore and product-related
variable costs, fixed costs and Lane’s opportunity cost.
Assuming that these break-even processes are appropriate for short-term planning and
the costs have been correctly identified, care must be taken to ensure that costs are allocated
to the correct blocks of mineralisation. The author has, for example, seen cases where pillar
stopes with uncemented (therefore lower cost) fill are allocated lower marginal break-
even cut-offs than primary stopes with cemented fill and higher costs. The reality is that
the cemented fill has been placed in the primary stopes so that the pillar stopes, or at least
some proportion of them, can be extracted. By this rationale, the cement costs should be
allocated to the pillar stopes, not the primaries, and pillar stopes should have the higher
cut-off, all other things being equal.
The fact that both positions could be argued with some degree of logic indicates that
both are spurious. The real situation is that having specified that the mining method is
to be by a combination of primary and pillar stopes, there is a commitment to add an
overall cement content to all fill placed across primaries and pillars. The average cost of
fill for all stope types is a cost of the mining method used for the whole block. The cost of
fill in individual stopes is irrelevant. The average fill cost should therefore be attributed
to all stopes, not differential costs based on what type of stope each happens to be.
Having made the decision that a certain amount of cement, with its associated costs, is
to be included in the total fill placed in the mining block, it is irrational to then say that
higher grades should be left unmined in some areas and lower grades extracted as ore
in others simply because of the distribution of cement in the whole mining block. If two

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 148
 CHAPTER 7 | Cut-off Derivation Methodologies

stopes, a primary and a pillar, were both being extracted from the same mining block,
the lowest grade mined from each stope should be the same. The ore–waste decision
was made when the mining method was decided upon and the mining block limits
defined, before stopes were designed and allocated primary or pillar designations. The
filling decision regarding the total amount of cement has already been made.15
However, nothing in this discussion should be taken to preclude the possibility of
extracting additional low-grade stopes that will be unfilled, or filled with uncemented
fill, on the periphery of the ore boundaries defined by higher-level evaluations. This
will only be appropriate if all extra costs incurred, including the opportunity cost of
displacing higher-grade material in the ore stream, are covered.
Similar care should be taken with access and stope development. These are designed
to make the overall mining method work, and to the extent that various mine openings
provide access or other benefits for more than one stope, they should be seen as method
or block costs, distributed among all stopes, not just to the one within whose boundaries
they happen to lie.

Grade control and sampling

All the cut-off models we have considered assume the cut-off grade specified is a true
grade. The grade available for decision-making will often be a sample grade.
It is well known that samples are frequently biased. Therefore, it is necessary to know
the relationships between the sample grade and the true grade of the rock in question so
as to make an informed cut-off decision within the operating mine.
Grade control and sampling are complex specialist topics in their own right and
this book does not purport to deal with them in any way, other than to highlight their
potential importance when specifying cut-off policies.

CHAPTER SUMMARY
This chapter has discussed recommended mine planning processes and indicated how
the various cut-off and optimisation processes described in previous chapters might
be included. Mine plans and strategies must have a clearly defined goal. If plans are
not intentionally derived to deliver a specified corporate goal, the de facto goal of the
corporation will become the goal that is implicit in deriving the plan.
In a long-life operation, where there are major lag times between exploration,
planning, development or waste stripping and production, plans must be developed
in a hierarchy from long- to short-term. Short-term plans should be consistent with but
provide more detail than the long-term plans from which they are derived. The long-
term or life-of-mine plan (LOMP) is set to achieve the corporate goal and ultimately

15. Similar logic applies in an open pit. Classifying material as ore and waste is made at the pit rim. That some
material may have come from near the surface and some from the bottom of the pit is irrelevant. The cut-off
does not depend on the depth of mining. Shallower material does not have a lower cut-off than deeper material
because of different haulage costs. The haulage cost at the point where the ore–waste decision is made – the pit
rim – is a sunk cost and does not come into future decisions. In Lane parlance, it is a mining variable cost, and we
know from Chapter 5 that these do not come into Lane’s limiting cut-offs. This assumes that the mining schedule
has already been determined. Depth-related costs will, however, affect the optimum mining schedule.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 149
 CHAPTER 7 | Cut-off Derivation Methodologies

generates value. Once the optimum LOMP is approved, all short-term plans must be
developed to deliver that LOMP.
Preproduction studies for a project go through several stages with increasing levels of
detail, and potentially different ways of determining cut-offs and other strategic options.
A scoping or conceptual study is typically at an order-of-magnitude accuracy (±30 per
cent). Its major purpose is to demonstrate early in the resource discovery that further
work is justified. It is therefore only necessary to find one case that proves this. A cut-
off must be selected to derive ore tonnages and grades, but there may be insufficient
information to rigorously derive one by any of the methods discussed in previous
chapters. Some form of full-cost break-even is perhaps the most appropriate method. If
there is a reasonable amount of tonnage and grade data, a Mortimer-style analysis might
be the preferred option, particularly if the deposit is low-grade. Strategy optimisation is
generally unsuitable at this level of study but, if the cases evaluated are not viable and
a better case must be found, or if the study’s purpose is to determine what the optimum
strategy might be, a high-level optimisation study might need to be conducted.
A prefeasibility study (PFS) should demonstrate to a higher level of confidence
(±20–25 per cent) that the project is technically and economically viable. It should
evaluate a range of options to ensure that the best is selected for further study. Usually
the case selected at this stage will be the one developed if it is proven to be feasible.
A full strategy optimisation study should therefore be seen as an integral part of the
prefeasibility study.
The feasibility study (FS) is intended to demonstrate technical and economic viability
of the project at a level of accuracy sufficient to justify proceeding with the project
(typically ±10–15 per cent). The policies identified as best in the PFS, such as production
rates and mining and treatment methods, will normally carry forward into the FS,
and will only change if more investigations expose previously unidentified problems.
However, cut-offs will be recalculated if price and cost projections have changed since
the PFS was done. Depending on how much the resulting orebodies change, it may be
necessary to completely rework other major strategic decisions. The first part of the FS
may need to be an update of the PFS optimisation using the latest data and forecasts to
assess whether the major decisions going forward into the FS should be changed.
Detailed engineering is required to convert the principles in the FS into construction
and mining plans for practical implementation. With improving knowledge of the
mineralisation and updated price and cost forecasts, it may be necessary to revise cut-
offs at this stage. With many parameters such as mining and treatment capacities now
locked in, simpler cut-off models may be better. If the nature of the deposits and mining
methods permit, a Lane-style cut-off optimisation may suit. If not, a Mortimer-style cut-
off specification is preferable to a simple break-even analysis. Opportunity cost should
also be considered in break-even formulas. It would also be desirable to consider the
capacities of the now-defined production stages. Lane-style balancing cut-offs should
therefore be identified. If prices have increased significantly, the cut-off should not be
depressed in a break-even manner to the extent that ore grades fall too low and product
targets cannot be met – a balancing cut-off may be more appropriate.
At an operating mine an integrated planning process is recommended. The process
works in a hierarchy from long term to short term. Longer-term plans set the overall
strategic direction. Progressively shorter-term plans provide more detail at the front end

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 150
 CHAPTER 7 | Cut-off Derivation Methodologies

of long-term plans. Approvals processes also work from long term to short term. Long-
term plans must focus on delivering the corporate goal. Short-term plans must then
detail how operations in the near future will help achieve long-term plans.
The strategy options analysis (SOA) evaluates, at a high level, the impacts on value of
all strategic decisions that the company can make, both separately and together. A full
SOA is not normally done annually but should be reviewed annually as part of the
planning cycle to ensure that the approved LOMP continues to be the best long-term
plan. Major updates and reworking the SOA should occur approximately every three
to five years.
The LOMP is the formal long-term plan for the mine or business. It is selected after
conducting an SOA, and is the plan identified as best delivering the corporate goal. It
establishes the framework within which all other short-term plans are developed. The
LOMP should be reviewed and updated annually as part of the planning cycle, taking
account of constraints identified in shorter-term plans or resulting from actual events,
and changes in the SOA.
The five-year plan (5YP) forms a critical medium-term link between the high-level
strategies in the LOMP and the detailed short-term implementation plans. The 5YP
would usually be generated annually, with quarterly time periods reported. It provides
sufficient look-ahead time for long lead items to be identified and planned for. Formal
approval of the 5YP should be part of the annual planning cycle.
The two-year rolling plan (2YP) provides a higher level of detail and is supported by
detailed engineering work at the front end of the 5YP. It would be formally updated
quarterly and report activities on a monthly basis. It is a regular ongoing part of the short-
term planning process and is not just done once a year as part of the budgeting cycle.
Formalising these regular updates requires both operators and planners to regularly
look at the effects of current issues up to two years ahead, avoiding actions that may be
expedient in the short term but create major problems long term. The appearance of an
activity at the end of the 2YP would be the flag for detailed design work to commence.
All activities occurring in the first 12–18 months of each 2YP should have been planned
in detail, with solutions found for any problems identified.
The annual budget plan is simply the plan for the 12 months of the financial year to
be budgeted in the version of 2YP that is created three to six months before the start
of the budget year. Its preparation should not require any special attention, since the
rolling process of plan and schedule generation and associated approvals ensures that
it is realistic, achievable and aligned with the corporate goal. Physical quantities in the
realistic and corporate-goal-oriented budget plan should drive the budget costs via cost
models, whose detail will have evolved in the same way as the physical plans.
Short-term detailed operational plans, consistent with the 2YP and, when appropriate,
the budget, should be developed to ensure that all activities required to implement the
plans are scheduled, with all necessary inputs available as required. These would be
for three-monthly periods, updated monthly, with time frames of weeks, days or shifts.
Cut-off determination and strategy optimisation would be incorporated in the
planning process as follows. The SOA will identify the optimum mine plan. This would
be formalised as the LOMP, which will identify not only the immediate cut-offs but
important inputs for ongoing cut-off derivation, such as pit limits and pushback staging
in open pits, ore boundaries and mining methods underground, mining and treatment

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 151
 CHAPTER 7 | Cut-off Derivation Methodologies

rates, products and the like. These will be the strategic decisions to be implemented
between successive strategy optimisations, which may be several years apart.
For an open pit, once mining sequences and capacities of various stages of the
production process have been set by the LOMP, most of the conditions required for a
successful Lane-style cut-off optimisation have been satisfied. For an open pit, a Lane-
style analysis should be used for short-term cut-off optimisation for all time frames –
from the 5YP to short-term operational plans. If for some reason Lane’s methodology is
not easily applied, a Mortimer-style cut-off specification is preferable to a simple break-
even analysis, but a Lane-style opportunity cost should also be considered in break-even
formulas. Lane-style balancing cut-offs should be identified as a balancing cut-off will
often be more appropriate than a break-even.
Underground, cut-offs define the size and shape of the orebodies, and in many cases,
the mining sequence will depend on the nature of the orebodies defined. A Lane-style
cut-off optimisation, which in its simple form depends on sequence, depends on the cut-
offs. It may therefore be impractical, though, in some circumstances, the iterative nature
of a Lane-style optimisation will account for this and other cut-off-dependent inputs.
Lane-style cut-off optimisation is most easily applied underground when there is a
single production front moving in one direction, with no returning to extract additional
ore from parts of the resource that the production front has already moved through. In
such cases, a Lane-style analysis should be applied to optimise cut-offs for short-term
plans where the overall development rates, ore handling and treatment capacities have
been established in the LOMP and long-term plans. Care must be taken to ensure that
costs and capacities are correctly attributed to the Lane-style production stages.
In the more common situations where Lane’s simple methodology is not easily
applicable underground, a Mortimer-style cut-off will be preferred to a simple break-
even analysis for short-term plans. A Lane-style opportunity cost should also be included
in break-even formulas. Lane-style balancing cut-offs should be identified as a balancing
cut-off will often be more appropriate than a break-even. The time required to change a
cut-off in an underground operation, particularly upwards, should also be considered.
Short-term responses to fluctuations in prices, costs or other inputs may not be practical.
For both open pit and underground operations, a simple marginal break-even cut-off
will only be suitable as an operational cut-off to account for very short-term fluctuations
in availability of ore to feed the treatment plant.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 152
Part 2
Important
Considerations
for Strategy
Optimisation
 HOME

Chapter 8

The Mine Strategy

Optimisation Process

introduction
So far, we’ve discussed ways of determining cut-offs (break-evens, Mortimer’s Definition
and Lane’s (simple) methodology). We’ve looked at the underlying theory and processes
used to derive a number as the cut-off for a predefined mining plan. The concepts and
rationale underlying a strategy optimisation evaluation, of which the optimum cut-off
policy will be a part of the outcomes, has been covered, as well as the general data
requirements and procedures of performing a strategy optimisation. This chapter, the
first in Part 2, describes these requirements and procedures in more detail.
As noted, there are several methodologies that can be applied to determine the optimum
strategy, and the process details will depend to some extent on the methodology
employed. This chapter discusses common features of these. It is not possible to fully
optimise the mining strategy without considering the metallurgical treatment processes
and vice versa. But often, one or the other is attempted by teams within the operating
departments, assuming that the other process remains unchanged. A full strategy
optimisation requires an integrated evaluation of all parts of the production process
across the whole life of the operation. That this is done is an implicit assumption in the
following discussion.

IDENTIFYING THE OPTIONS AND SCENARIOS

The corporate goal must be borne in mind at all stages of the optimisation process.
What do we need to do to derive value measures associated with that goal, and what
processes and associated data do we require to achieve this? As well as the goal, it

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 155
 CHAPTER 8 | The Mine Strategy Optimisation Process

is important to identify the options that will be evaluated. Data and relationships
generated must support the full range of options, some of which will represent
different values for what may be thought of as a continuous parameter, such as the
optimum production rate within a specified range. Others may represent discrete ways
of achieving certain outcomes, such as the types of trucks/loaders; whether haulage is
to be by truck, conveyor, or shaft; and whether stoping is to be by primary/secondary
or continuous advance sequencing. If certain options are not to be investigated, the
model need not handle these. But if certain options are to be evaluated and compared,
the data must enable options to be distinguished, with the model handling all the
relationships required to do this.
As a general rule, it is impossible to evaluate something that has not been included
in the modelling. It might be feasible to use a surrogate in the model, but there is no
guarantee of a reliable outcome.1 It is therefore important at the start of the evaluation
to identify what needs to be taken into account during the studies. Items not deemed
essential at first may be easy to incorporate during the early model building, or can be
excluded from the initial stages of the project with the model structured to allow for
their easy addition later if necessary. Items that are to be added at later that have not
been anticipated in the model design may be difficult to retrofit into an existing model.
At worst, the model’s structure may make such additions impossible, necessitating a
complete rebuild. Project sponsors should ensure that the optimisation team is aware
from the outset of everything that might be of interest, and should not presuppose
what may be difficult to model; experienced teams may have developed processes that
simplify tasks that others perceive are difficult.
It is imperative that decision-makers at as high a level as possible in the company
have input into the study-scoping phase. It is not uncommon for staff at one level in
an organisation to be unaware of higher-level concerns. Similarly, matters that are
perceived to be unchangeable and not needing investigation at one level may be seen
as crucial matters by senior staff. The involvement of management in these early stages
is therefore essential, or the study may be constrained from the outset and unable to
deliver what is expected.

UNCERTAIN PARAMETERS AND RELATIONSHIPS

There will be levels of uncertainty associated with all parameters of the strategy
evaluation process. In many cases, it will be such that the study outcomes may be
considered unreliable. Sometimes there is a propensity to leave such things out of the
analysis so that only reasonably certain parameters are included. This is thought to
enhance the rigour and the reliability of the analysis.
The author suggests this is a fallacious argument. For a strategy optimisation study,
it is vital to take account of the influences of uncertain parameters. Consider that the
most uncertain parameter in any mining evaluation will usually be the product price.

1. As a simple example, if the model has not been constructed to allow for different metallurgical recoveries, one
could model the revenue effects by flexing the product price. If the cost change resulting from the different
volume of product were insignificant, using price as a surrogate for recovery might be satisfactory. However, if
the change in product-related costs resulting from a change in recovery and product volume had a significant
impact on total costs, the overall result might not be sufficiently reliable for the decisions needed.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 156
 CHAPTER 8 | The Mine Strategy Optimisation Process

We would never contemplate a study that ignored prices ‘because we don’t have good
information’. Rather, we’d estimate the most likely scenario and flex the price inputs
over a range. As has been noted, a change in the price will shift the value of any operating
plan. The purpose of an optimisation study is not so much to determine value per se, but
to identify the optimum plan and how it might change if uncertain parameters change.
So the issue is not how variable the value might be because of price uncertainty; we wish
to identify how changing price might cause the optimum plan to vary, and therefore
how it might change the decisions made.
The same rationale applies to any important but uncertain parameter. It might be
the numerical value to be used, such as a price or cost, or the tonnage and grade
of an Inferred Resource or exploration target. Or the uncertain parameter might
be a relationship, such as how metallurgical recovery varies with feed grade. It is
critical that every important (or thought to be important) but uncertain parameter is
recognised as a scenario parameter to be flexed in the analysis over the likely, though
as yet uncertain, range to identify what causes the strategic decisions to change. If the
decision-changing situation is found to be outside the likely range for an uncertain
parameter, we will be able to state that it is not vital to defining the optimum strategy.
If the decision-changing point is found to lie near the most likely value or relationship,
we know that we must either take steps to reduce the level of uncertainty or consider
the trade-offs between the benefits and costs of making decisions based on correct or
incorrect assumptions.
This risk–reward trade-off will be discussed more in Chapter 12. The issue of levels of
uncertainty is introduced here to raise awareness of its potential impact on the process.
Every data input parameter should be reviewed early in the process to identify whether
it is certain enough to be a fixed input value or relationship, or whether it should be
a scenario parameter that may influence strategic decisions. Disagreements about
whether a parameter is important or not can ultimately only be resolved by including
its effects in the analysis and investigating it quantitatively. In practice, many input
parameters that have what might be termed set and forget fixed values will also have
flexing factors applied for conducting a sensitivity analysis, as described in Chapter 6.
Such factors facilitate including those parameters in the scenario analyses, and will often
be the simplest way to investigate the impact of other uncertain parameters.

GEOLOGICAL INPUTS
A geological model is the basis of any mine planning activity. For strategy optimisation,
the model must be reliable over the range of cut-offs that may be investigated. In
practice there may be a trade-off between geological rigour and the time required
to develop such a model. It is common to find that an existing model is deemed by
the geologists to be reliable for only a portion of the range of cut-offs, leaving two
alternatives. One is to defer any further analysis until a model has been developed that
is deemed reliable across the full range of cut-offs. This is rarely an acceptable option
– the time frame for the study will already have been set and the geological staff will
have too many other tasks to attend to anyway. The other course of action – often

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 157
 CHAPTER 8 | The Mine Strategy Optimisation Process

the one adopted – is to make use of the existing model, recognising that it may have
inadequacies2 and taking account of these.
This may not be as bad as it sounds. Rigorous analyses can still be conducted with
the existing model. If the optimum strategy is found to lie within the range where the
model is reliable, the optimum will be reliable, and additional work to improve the
model would be unnecessary. If, however, the optimum strategy lies outside the reliable
range, the analysis will indicate the range of cut-offs over which the geological model
must be reliable, and whether creating such a model is worthwhile. The analysis using
the existing model will indicate the value of the optimum strategy and the value of a
strategy using the cut-off at the limit of the model’s reliability. The difference in value
will indicate the maximum cost that could be incurred for the additional reliability.
The extra cost might simply be for reprocessing the existing data. Alternatively, there
could be significant costs for such items as additional diamond drilling at a closer
spacing. If the extra cost is less than the extra value, it’s worth generating a new model.
If the cost is greater than the potential value gain, it is not worth doing – the hill of value
(HoV) will rise but, before it reaches a peak, there will be a cliff at the cut-off at the limit
of the model’s reliability, which will then be the optimum cut-off.
It is necessary to specify the level of reliability of geological information required, and
this may be better indicated by establishing what is not required. It is not necessary, for
instance, that the model locates every potential ore lens in 3D space so that final designs
for access and stope development can be generated. That is not the purpose of a strategy
optimisation study.
What is necessary is that the model represents, at a level of accuracy consistent with
the study, the tonnages and grades at various cut-offs, as well as the general size, shape
and location of orebodies defined at those cut-offs. The required modelling outcomes
for optimisation are that the shape characteristics are sufficiently accurate for mining
method options and related parameters to be reliably estimated. Underground, this may
include aspects such as sublevel interval, types of equipment and development metres.
The locations of the orebodies need only be sufficiently well established to generate, for
example, access development quantities. For an open pit, the orebody size and shape
may be important for identifying options if the changing nature of the orebody with cut-
off were to result in, for example, different mining equipment or drill and blast patterns.
There is an assumption that the nature of the geology and grade distribution is such
that, when final designs are needed in the future, it will have been possible to obtain
the additional information to generate detailed designs. This is, of course, part of the
normal planning cycle with phases of increasingly detailed geological data gathering
and generating more accurate plans, as described in Chapter 7.
The tonnage and grade versus cut-off relationships will change as the amount of
information increases. With closer-spaced drilling, grade variability will be better
identified and there will be more high-grade tonnage with greater knowledge, even if
the total resource tonnage is unchanged. This means that the tonnes and grades that exist
above certain cut-offs will be greater than predicted by early-stage geological modelling.
If there is adequate history to derive relationships, it may be appropriate to apply factors

2. In addition, it is not uncommon to find that the existing geological model is out of date and in the process of
being updated, but the new model is expected to be delivered just after the results of the optimisation study!

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 158
 CHAPTER 8 | The Mine Strategy Optimisation Process

to tonnages and grades generated at various cut-offs in preliminary studies to better

estimate the real situation. Depending on the nature of the mineralisation and the cut-
offs applied, the early models could be either underestimating or overestimating the
potential value to be obtained from the resource.
With a geological model that is sufficient for the study, it is now necessary to generate
orebodies3 at each cut-off of interest. From these, we need to identify geological
domains that allow different mining methods, scheduling constraints, metallurgical
treatment, etc. For this discussion, geology also includes geomechanical information
such as stresses, rock strengths, structures impacting designs and the like, again at a
level of accuracy commensurate with the overall study.
Exploration potential is often a critical input. This may have several categories, such
as extensions of existing orebodies, new zones within the existing operations and other
deposits in the region that could be combined with the ore feed to the existing plant. The
nature of other deposits in the region could also lead to constructing another treatment
plant or similar. These external deposits could belong to the company conducting the
evaluation or a third party, with toll treatment, ore purchase or outright purchase of
the properties as options to be evaluated. By its very nature, exploration potential is
uncertain data, though there are degrees of uncertainty. Some may be almost to the
stage of formal reporting as resources, while others may be as-yet untested targets.
Uncertainty, as has been noted, is a poor reason to exclude exploration from the
investigation. Optimum strategies could be found both with and without the exploration
potential, and scenarios including it could consider optimistic, pessimistic and mid-
range estimates. An outcome of this type of analysis could be whether the optimum
strategies with and without the exploration potential were similar or significantly
different. If similar, then, although the exploration potential is uncertain, it would not
affect operating strategy so is not a concern in that regard. If, however, strategies are
very different, particularly in the short term, decision-makers can then decide how
quickly to advance the exploration to obtain good information, or make decisions that
maximise the upside or minimise the downside of potential additional resources.4

MINING INPUTS
The preparatory geological modelling will have identified the size and shape of the
orebodies at a range of cut-offs. The first task of the mining-related processes is to
identify which of the potentially suitable methods for the orebodies at each cut-off will
be considered in the evaluation. The big-picture mining method options are open pit
and underground, and within them exist various methods depending on size, shape,
geomechanical conditions and so on.
It was identified earlier that the primary mining-related decision may be thought of as
the size of the mine, though, in reality all decisions should be optimised concurrently.
Size of mine is, however, the primary decision driving how resource data will be
generated for the underlying database for the optimisation model.

3. As previously indicated, in this book terms such as ore and orebody refer to material above a cut-off being
evaluated. They are not used in the strict sense of the codes for public reporting of resources and reserves.
4. This risk–reward trade-off is discussed in more detail in Chapter 12.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 159
 CHAPTER 8 | The Mine Strategy Optimisation Process

For an open pit mine, sizes can be defined by a number of outlines, such as nested pit
shells formed by pit optimisation software.5 For an underground mine, nested stopes or
nested orebodies need to be generated for cut-offs spanning the range within which the
optimum is likely to lie.6 Until recently, this was a labour-intensive task, and in some
cases, several designers had to be used to generate the stope shapes and reserves data
at the cut-offs required in a reasonable time frame. This had the potential for workers
to produce the information differently, so that the data sets at different cut-offs were
inconsistent. The recent development of software to generate realistic stope shapes at
specified cut-offs has reduced the time and effort required to create the underlying
database and improved the consistency of the data across the range of cut-offs considered.
Ideally, we would feed the whole geological model into the strategy optimisation
software, along with all the operating characteristics and relationships, operating and
capital costs and revenue relationships and the like, and the solution would emerge
from the process. However, the limits of computer power, now and in the foreseeable
future, requires that small geological model blocks need to be aggregated into a number
of larger mining blocks to make the optimisation problem computationally tractable.
The mining inputs of an optimisation study form the primary database of material for
mining. The geological model will be the main input, but it is the mining considerations
that result in the data that the model or software works with.
The mining blocks to be generated for the database will be different for open pit and
underground operations.

Open pit mining inventories

For open pits, the aggregated mining blocks will be bounded at the coarsest level by
pit shells corresponding to each major pushback stage and by a number of elevations
corresponding to pit bench elevations, spaced at some integer multiple of the bench
height. For small pits, a block may be the ring defined by a range of pit shells and bench
elevations around the full circumference of the pit. For larger pits, these rings may need
to be further subdivided – perhaps by specified northings or eastings, or by lithological
boundaries, or whatever is logical for the situation. Intermediate pit shells within
pushback stages and individual bench elevations may also be used to further subdivide
the mining blocks. The optimisation software will then select the blocks to be mined and
schedule them in such a way as to generate the optimum pit size and mining sequence.
This description implies that decisions regarding pushback staging of the ultimate pit
have already been made before the evaluation starts. Such sequencing and scheduling
should ideally be part of the outcome, not an input into it. So while some thought is
needed with staging at the data creation stage, it is preferable to allow as much flexibility

5. To ensure an adequate data set, these may range from the smallest pit that could realistically be mined with
pessimistic economic forecasts of prices and costs and including only mineralisation identified with a high level
of confidence, to the largest pit that might be mined with optimistic economic assumptions and including low-
confidence or hypothetical extensions of the geological resource.
6. In the author’s experience, the NPV-maximising cut-off for an underground mine is typically 1.3–1.5 times the
cut-off derived by typical break-even calculations. In the absence of other information, a range of cut-offs from
somewhat less than, to perhaps double the existing cut-off may be a good starting point for the range to be
considered. If available, a marginal break-even cut-off (Lane’s mining-limited cut-off ) with optimistic prices
could assist in defining the likely minimum cut-off, and Lane’s treatment-limited cut-off with pessimistic prices
may in many cases be a useful guide to the upper-bound of the likely range of cut-offs.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 160
 CHAPTER 8 | The Mine Strategy Optimisation Process

and as little preconception as possible. More and smaller mining blocks will assist with
this, but will also increase the computation time and make the problem potentially
less tractable. A more precise outcome may be achieved, but better decisions aren’t
necessarily produced.
The impact and timing of irrevocable decisions must also be considered. For example,
although the size of the ultimate pit using assumptions about future conditions will be a
result of an optimisation study, it may well be that the decision regarding pit size does
not need to be made for many years into the future. Notionally, the decision about the
size, mining rate and starting time of each pushback stage, including the last, which
creates the final pit walls, only has to be made just before the last stage starts mining;
however, each pushback stage must be considered in the context of the whole pit, not
independently. The number of mining blocks that can be adequately dealt with will be
a function of the capabilities of both the software and the computer hardware on which
it operates.
To facilitate cut-off optimisation, it is necessary to have tables of data for tonnages
and grades of material versus cut-off for each identified mining block, enhanced by
other available data such as geometallurgical information. Further subdivisions of the
data tables by such items as rock and ore types and different categories of deleterious
materials may also be necessary. Depending on the orebodies defined at different cut-
offs, it may be appropriate to have tonnage and grade data for two or more open pit
mining methods. For example, we might require data for selective and bulk mining
carried out by two distinct mining fleets with different productivities, ore recovery
and dilution, and capital and operating costs if this cannot be adequately estimated by
applying different factors to the one data set.
If the optimisation study includes the transition from open pit to underground mining,
it may be necessary to generate different sets of pit shells and mining blocks, accounting
for any mining of higher-grade parts of the resource from underground before the pit is
expanded into those areas. Pit slope stability may need to be re-evaluated if stopes have
been extracted adjacent to what will later be pit walls.

Underground mining inventories

For underground studies, mining blocks will best be defined after nested stopes or nested
orebodies have been produced at different cut-offs, so that the best location of boundaries
between mining blocks can take account of the shapes, potential mining methods and
similar factors. They will typically be defined by a combination of ore zones, sublevels,
northings or eastings, geological structures, changes in the size and shape of the orebodies
and boundaries defining geological domains within the rock mass or orebodies. The
mining inventory for each cut-off must be as realistic as possible, taking account of both
the ore loss and dilution required to generate realistic stope designs.
Outliers of material above any specified cut-off that would be uneconomic to extract
should be excluded. A simple economic analysis comparing costs to access and develop
an outlier with the extra revenue it generates may be adequate to exclude it. Some areas
may be economic with some price forecasts but uneconomic with others. Areas that are
economic according to such an analysis but are some distance from the nearest large zone
of mineralisation should not be automatically aggregated into those adjacent areas – it
may still be preferable to consider them separately to obtain the best plan. For example,

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 161
 CHAPTER 8 | The Mine Strategy Optimisation Process

it may be that the best allocation of scarce resources to all the tasks for a mine plan to
work is such that an apparently economic zone is never allocated mining resources.
If there is a limit on the amount of jumbo development that can be done, optimisation
software may allocate resources preferentially to sources with the highest net revenue
per metre of development. An outlying area that is notionally economic but of low net
value may well have a low return per unit of the mining resources required to develop
and produce from it. The optimiser may therefore allocate equipment and mining crews
to areas where the most money will be generated per unit of the scarce resources and
decide not to access and produce from other areas.
Underground evaluations will usually be simpler if boundaries between mining blocks
are the same at all cut-offs. If, for some reason, the natural boundary between two mining
blocks is a function of cut-off, special care may be needed. It might be essential to ensure
that, if it is feasible for the two adjoining blocks to be mined at different cut-offs, there is
no gap of unmined material in the data set, though this may be a practical outcome from
the optimisation process. It will also be important to ensure that a region that is part of
two mining blocks at different cut-offs is not mined twice by the optimisation scheduler.
If this is found to be an issue, consider whether the tonnages and grades affected are
material and must be accurately modelled, or whether they are insignificant and will not
affect the overall optimum strategy. Unnecessary modelling complexity may affect the
computational tractability of the problem.
If the optimisation study includes the transition from open pit to underground mining,
it may be necessary to generate different sets of underground mining block data for
mineralisation that is extracted within different final pit limits and to account for any
differences in methods or costs associated with proximity to the pit wall.

Other physical data

In order to model the interrelationships that may be necessary to account for all the
identified options to be investigated, other physical data will need to be generated in
association with the mining inventories described.
For open pits, it’s necessary to define haulage distances for each mining block to both
ore and waste dumping points. It is quite common to find geological block models
with a single haulage distance ascribed on the basis of a predetermined classification
as ore or waste. Strategy optimisation will make this decision. In order to facilitate
this, both distances need to be available to the optimiser. More complex information
may be required depending on dump locations, rock types and the like. Block models
may also record truck hours and fuel consumption, which may need multiple records
for combinations of different truck types and destinations if that is an issue. The
author has also encountered a situation where the acid-generating potential of the
rock necessitated mining additional waste rock from elsewhere for the sole purpose
of neutralising the acid generation in waste dumps. This in turn generated additional
loader and truck hours if a block were classified as waste rather than ore.
For underground mines, development requirements for each mining block will need
to be specified as data, though, in some cases relationships may be specified to allow
calculation within the model. Depending on the level of study conducted, development
might need to be split into categories such as major access, sublevel development
and stope development in ore and waste. How development and stoping schedule

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 162
 CHAPTER 8 | The Mine Strategy Optimisation Process

interactions are to be handled in the evaluation may also suggest what proportion of
each development category needs to be completed before the mining block can start
producing. Stope drilling and blasting parameters may need to be specified separately
at each cut-off, or, depending on the study’s level of accuracy and the materiality of
any differences, might be adequately catered for in the stoping cost per tonne of ore.
Trucking distances may vary if haulage and hoisting options are to be evaluated.
Quantities of different types of backfill may need to be specified, particularly if the
mining methods being evaluated have different proportions of the orebody to be filled
or are using different fill types.
Additional underground data requirements may also include elements such as access
methods or topologies. For example, decline layouts may differ depending on whether
the main access from the surface is by shaft or by decline, which in turn could be from
the surface or within an overlying open pit. If there are two orebodies relatively close to
each other, the decline topology options could include a single decline with accesses to
each orebody or twin declines, one adjacent to each orebody.
For open pits, the additional physical information will typically be for activities or
parameters that occur in parallel with the mining of the rock, and will be used either
to generate costs accurately or to derive activities to which constraints may be applied.
For instance, the loading fleet will constrain the tonnes mined, but the trucking fleet will
constrain the tonne-kilometres hauled or truck hours worked. Depending on variations
in the average depth of mining, the effective constraint on mining rate could vary from
one to the other over time, but both may need to be specified and appropriate data and
relationships provided.
For underground mines, some of the additional physical information will be for
activities or parameters that occur in parallel with ore production, and will be used
either to generate costs accurately or derive activities to which constraints may be
applied, such as drilling, blasting and haulage. However, some information will be
for activities that logically precede or follow the ore production activities, such as
development and backfill. If included in the analysis, these will have associated costs
and constraints on rates. Their constraints and their predecessor–successor relationships
with ore production will often have a major impact on the overall mining schedules
and therefore need to be accounted for in the scheduling component of the analysis.
Similarly, decline topology could influence access development schedules and hence,
depending on what mineralisation has been accessed, the ore production schedules.
Twin declines could make available the option of a one-way haulage loop, which could
impact on haulage productivity and on mining capacities.
There are no universal specifications for either open pit or underground, though the
preceding discussion illustrates some of the more common situations. While it is vital
to account for everything impacting on value, it is also important to minimise the
computational requirements within the optimisation component of the modelling. This is
done by combining items that logically may be combined and which, by doing so, will not
affect the accuracy of the results, such as accounting for drilling in the cost per tonne. Such
parameters may well need to be calculated and displayed separately for final outputs
presented in reports, but in the context of this discussion, that is a post-processing task for
the cases presented, and not part of the optimisation process itself.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 163
 CHAPTER 8 | The Mine Strategy Optimisation Process

Scheduling rules and capacities

Having generated data for mining blocks, it is then necessary to describe sequencing
and scheduling rules. These must result in schedules that are practical and realistic at
the level of accuracy of the study and that are able to account for all the combinations
of cut-offs, mining and treatment rates and other options that could be evaluated.
Some rules are simple and obvious, while others are less easy to describe in
accounting for all the possibilities that might potentially arise. In open pits, blocks
closer to surface are mined before underlying deeper blocks. If the pit shell boundaries
are based on pre-identified pushback stages, it is probable that minimum bench
widths in pushbacks have been accounted for. If mining blocks are smaller than this, it
may be necessary to define minimum widths of each pushback, so that mining blocks
can be aggregated by the optimiser’s scheduling processes. These thicknesses will be
functions of the size of the loaders and trucks in the mining fleet, which in turn might
vary with mining rates evaluated.
Mining rate specifications must look at what is possible practically, not just at what is
scheduled in the existing mine plan. Very often the current schedule will have been set to
deliver the required amount of ore at a pre-specified cut-off, which may not be optimal.
This will have the effect of making the current cut-off, however it has been determined, a
Lane-style balancing cut-off. Simply applying the currently scheduled rates as constraints
will result in the current cut-off appearing to be optimal. If different rates are applied as
constraints – perhaps the capacity of the existing mining fleet used to its full potential
or by the acquisition of additional equipment – different optimum cut-offs and higher
values may well result, which would more than pay for the increased mining rates. (See,
for example, Figure 6.10 and the accompanying discussion in Chapter 6.)
Although related, equipment size and mining rate are independent decisions. While
options involving high mining rates with small equipment and low mining rates with
large equipment can probably be ignored, the author suggests that the optimisation
team should avoid prejudging the mining rate at which the size of the mining fleet will
change. Discovering that should ideally be an outcome of the evaluation, not a pre-
specified input into it. To do that will require the ability to model higher mining rates
with small equipment and lower mining rates with large equipment than might be
thought to be optimal, with associated data and model logic to correctly evaluate these.
Underground, access development must precede stope development, which must
precede stope production. Stope sequencing rules may be more complex. Although
rules often specify that one stope must precede another, this is frequently done for
convenience or because of preconceived logical sequences. In reality, underground
scheduling constraints are often simply that two particular activities cannot occur
concurrently. For example, two adjacent stopes cannot be mined simultaneously: one
must be mined and filled before the other can produce, but there may be no logical
necessity for either to have priority over the other. On the other hand, it might be that
the general direction in which the production front is advancing forms part of the
description of the mining method. In that case, different predecessor and successor
relationships may be needed for each alternative. Depending on the size and shape of
the orebodies and mining methods, complex proximity rules may need to be specified
with the minimum separations between stoping activities, including filling; these will

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 164
 CHAPTER 8 | The Mine Strategy Optimisation Process

differ depending on the material between the activities, whether void or solid rock or
fill of various types.
Every scheduled activity has a notional maximum rate or capacity. In an open pit, the
maximum mining rate (or rock capacity) may be either the loading or truck haulage
capacity. As indicated, haulage distances may vary for ore and waste, and if so, the
haulage requirement and hence trucking capacity may be a function of cut-off. In hard
rock, drilling capacity may be the rock constraint. The maximum ore mining rate in an
open pit is sometimes limited by the turn-around time for grade control assay results.
Underground, we have seen that the main access development rate, which is exposing
new mineralisation that can be then classified as ore or waste, is equivalent to the open
pit rate of mining rock. Most of the processes in an underground mine are dealing
with ore. The mining data for an underground optimisation study will typically need
to identify at the least the maximum access development rate and the maximum ore
production rate within the mine. Depending on the problem’s complexity, it may be
necessary to specify capacities for a number of components of the ore production system
in the mine, such as stope development (which may be the total development capacity
less what is committed for access development for future production), and loading and
truck haulage capacities (similar to the open pit situation). The overall production rate
may be the result of complex interactions between the size of the orebody, stope layouts
and adjacency factors, and predecessor–successor relationships between development,
production and filling operations in various stopes.
The ventilation system may be the ultimate constraint on an underground operation,
but this is often not easy to bring into the modelling process. The ventilation requirement
may be a function of the depth of mining and the needs for different types of equipment
that could be used. The overall constraint may therefore be a function of the total
ventilation capacity, the average depth of scheduled mining and the equipment fleet
selected, as well as the usual constraints previously described. The overall ore capacity
will be the lowest of the maximum rates of all ore-related processes in the mine and
treatment plant.
It was noted in Chapter 6 that, for an optimisation study, it is important to ascertain
relationships between parameters so that the effects of making changes in one will be
reflected in the reported results. With regard to system capacities, linear relationships
between inputs and outcomes may be inaccurate. Figure 8.1 illustrates a common
situation with trucking, particularly in an underground mine where there may be a
limited number of places where trucks travelling in opposite directions can pass.
If there are no constraints on truck movements, no queuing, no delays for passing
and so on, capacity will increase linearly with the number of trucks, as shown by the
‘theoretical linear’ line in the figure; however, when there is interference between trucks,
such as waiting to pass a truck travelling in the opposite direction, the incremental
capacity of each extra truck is progressively less, leading to a curvilinear relationship
between capacity and number of trucks, as shown by the ‘with interference’ line in
the figure. Further complexity is added when the trucks interact with a fixed capacity
system, such as the ‘downstream’ line in the figure. The overall response will now be
a complex function of the breakdown and repair characteristics of every component,
their individual capacities and the amount of surge capacity or storage between
successive parts of the overall system. The loading capacity will impose a similar flat

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 165
 CHAPTER 8 | The Mine Strategy Optimisation Process

Figure 8.1
Production capacity versus number of trucks with various assumptions.

Production Capacity

Number of Trucks

Trucks – theoretical linear Trucks – with interference

Downstream (crush, hoist, etc) Total system capacity

line to the downstream capacity on this two-dimensional plot, but this may also be
affected by a curvilinear relationship in a third dimension, representing numbers of
loaders in the loading fleet, increasing the complexity of the interactions significantly.
A simulation of the physical system may be needed to account for these interacting
components to estimate the overall production capability. The importance of these non-
linear system responses will depend on the study’s level of accuracy. As noted when
discussing Lane’s methodology, the optimum solution will often be found to occur
when two components of the system are both operating at capacity. As can be seen in
Figure 8.1, this is often where the greatest error will occur if simple relationships, such
as the theoretical straight line and the downstream limit, are applied.
Options for the removal of constraints will usually be specified, unless the scope
of the project is clearly defined to be limited to optimising the value of the operation
subject to existing constraints. Often this will simply be the addition of more units of
development and production equipment or crews, such as drills, loaders and trucks.
Other debottlenecking options could involve, for instance, sinking a shaft or upgrading
the ventilation capacity.
It is of course possible that the schedules generated by the optimisation software will
be unconventional, but so long as they are practical and realistic, that is what counts.
Conventional wisdom is not always wise, and a logical process obeying appropriate
rules may well find a better solution to which conventional wisdom has blinded us.
The scheduling rules should therefore be as simple as possible to specify genuine
constraints, with as little imposition of a planner’s prejudgment as possible to define
a tractable problem. The more that is locked in, the less opportunity the optimiser has
to find better solutions.
It will be rare for a change in strategy to be made instantaneously. There will
typically be a period of time at the start of the evaluation that needs to have its
physical activities locked in, representing the time taken to conduct the evaluation
and make any strategy-changing decisions, plus the time taken to create a plan to
effect the transition from old to new. Depending on the corporate culture and the

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 166
 CHAPTER 8 | The Mine Strategy Optimisation Process

nature of the deposit and mining method, this unchangeable period could easily be as
much as two years; in some cases, a reduced amount of locked-in activity could also
extend into later periods as transition to the new plan is effected. The scheduling rules
will therefore need to make allowance for these locked-in activities at the beginning of
the evaluation. Depending on the accuracy of the study being conducted, limitations
may have to be placed on the transition from what is effectively the current long-term
plan – represented by the locked-in activities – and the new optimised plan generated
by the strategy optimisation processes. For example, there might be a constraint on the
rate at which production, development or waste stripping rates may be increased for
a number of years.
It is important to identify how much of the locked-in activity is a practical reality and
how much is related to corporate culture and inertia. It may be appropriate to evaluate
a range of locked-in periods, in order to identify the value generated by challenging and
overcoming organisational (as opposed to practical) impediments to change.

Concluding comments on mining

There may be practical problems generating the required information. Computer
software can produce most of the mining data required for open pits. Depending on the
level of study, it may be necessary to do some designs with human engineering input in
order to calibrate data generated automatically by software for a large number of cases.
Transforming data so that it’s suitable to input into optimisation software may present
challenges.
Automated software processes are nowhere near as advanced for underground
operations as they are for open pits. At the time of writing, stope optimisation software
that is capable of generating nested stopes is commercially available, but is not necessarily
applicable to all styles of orebodies and mining methods, though developments
continue. General purpose software for automatic generation of stope and stope access
development is limited to in-house software owned by some companies and applicable
to specific methods and styles of mineralisation. Commercially available applications
are, however, being developed. Software to optimise the placement of declines7 has been
developed but, to the best of the author’s knowledge, is not commercially available.
Generating adequate data for an underground strategy therefore requires an amount of
human-generated mine design, and availability of technical staff to do this work may be
a practical constraint on the overall strategy optimisation process.
Technical staff often produce rigorously engineered designs for every case to be
evaluated, yet this might not be necessary. If the optimisation software requires
graphical objects created by a mine planning software application as its inputs, each
case may need to be designed, and this may not be a trivial task within the constraints
of both the design and the optimisation software. If, however, the optimisation
software requires only tables of data to define the mining requirements, it may well

7. For a simple single orebody, an experienced design engineer will be able to produce a design as good as any
computer algorithm, though if multiple designs are needed (for example, to service the different orebodies
defined by different cut-offs), an automated process could be useful. For many optimisation studies, knowing
the depth of the mine and the gradient of the decline will be sufficient to derive development quantities at an
acceptable level of accuracy. For multiple orebodies that could potentially require a complex network of access
declines to link them, an access layout optimisation tool may be essential.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 167
 CHAPTER 8 | The Mine Strategy Optimisation Process

be that a small number of cases require engineering input, while data for other cases
can be interpolated. The guiding principle for input data generation is: maximum
information from minimum work. Exactly how this balance is achieved in each case
depends on the complexities of the mineralisation and mining methods and on the
engineering judgement of the technical staff. If the optimum solution is one for which
engineered designs have not been done, there should be sufficient engineering input
into other cases from which the data for the optimum case has been interpolated so
that the results can be accepted with confidence within the study’s level of accuracy.
There is no general rule as to what constitutes sufficient rigour in the data – what
is judged to be adequate by one team, or within a particular corporate culture, may
be perceived to be inadequate by another. Despite this, if any input items are deemed
to not be sufficiently accurate – perhaps because of time or availability of engineering
resources constraints – those parameters may be treated as scenario parameters
to ascertain whether changes have an impact on the decision to be made. If not, the
additional accuracy might be nice to have, but it is not needed to make a sound decision.

METALLURGICAL INPUTS
Physical metallurgical inputs to a study will generally be related to both metallurgical
performance (typically recovery and product quality) and processing constraints.
It is essential to specify recovery relationships that are reliable for the head grades
resulting from the range of cut-offs to be evaluated. Other parameters such as blends
of different rock types may also need to be accounted for. Other relationships, such
as the mathematically rigorous four-way relationship between recovery, head grade,
concentrate grade and tailings grade,8 or empirical relationships between recovery and
feed rate, may be relevant. Care must be taken to ensure that relationships are valid for
the range of head grades encountered. Often, recovery relationships have been derived
by regression of results in the current operating range, assuming a polynomial or some
form of exponential function of head grade. Plotting recoveries over the wider range of
head grades anticipated for the study may reveal unrealistic behaviours. For example,
a cubic function may produce recoveries that rise, fall and rise again with increasing
head grades, or that fall to zero or negative values with lower head grades that are
actually economic to treat.9 Some metallurgical investigations may be needed to identify
better relationships to use for the study, or scenario analysis may be necessary over any
uncertainty range in the recovery relationships to identify whether strategic decisions
are likely to change if the recovery relationships vary.

8. Recovery = [c.(h – t)] / [h.(c – t)], where h = head grade, c = concentrate grade and t = tailings grade.
This formula is particularly useful for base metals operations, where the tonnage of concentrate is a significant
proportion of the ore feed tonnage. For free-milling precious metals, where head grades are low (a few parts per
million), the tailings tonnage is the same as the ore tonnage, and the concentrate grade is very high relative to
the tailings grade. Then, as an approximation, c = (c – t) and the formula reduces to Recovery = (h – t) / h.
9. The author has also encountered relationships that, with high head grades, generated concentrates with
contained product grades in excess of 100 per cent. The point of the discussion is that quoted recovery
relationships (or any relationships) should not be blindly accepted simply because they are provided by
technical experts. The optimisation team must ensure that any relationships used will be valid across the full
range of the values that may be encountered, and the relevant technical experts should be intimately involved
in this to ensure that relationships are accurate across the range of interest.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 168
 CHAPTER 8 | The Mine Strategy Optimisation Process

Treatment plant production constraints must be identified. As a minimum, these should

be for ore feed and each final product. Depending on the complexity of the treatment
circuits, it might be necessary to specify constraints for some of the intermediate process
streams. Any points in the circuit where it may be feasible to build stockpiles of in-process
materials may potentially act as logical breaks into separate processes, each of which might
have feed and output constraints, as well as metallurgical performance characteristics.
Ore and product streams must be considered separately, taking account of the real
constraints in each independently. The ore constraint is often expressed as a maximum
tonnage rate that reduces when the feed grade increases above a specified grade.
This typically means that, at the specified grade, the ore and product streams are at
full capacity, and any increase in head grade must result in reducing the ore feed rate
to avoid overloading the product circuits. For strategy optimisation, this combined
relationship should ideally be split out and ore and product stream constraints specified
separately. The combined effects of both constraints will be taken into account in the
optimisation process, but by specifying them separately, the effects of changes in each
can be more easily handled.
Constraints may be more complex than simple ore and product tonnages. It may be
useful to consider the constraint on throughput to be the number of operating hours
available. Since different rock types in the ore feed may have different milling rates, the ore
tonnage capacity of the plant will be dependent on the ore blend. In such circumstances,
it may be preferable to apply a tonnes-per-mill-hour factor to the various components of
the ore feed, change the primary quantity measure from tonnes to required mill hours
and apply the annual operating hours as the constraint. If appropriate, this process
could be extended to account for power draw required, with the constraint becoming
the available milling power.
As for mining, options for removing constraints in the treatment plant will usually
need to be specified, unless the scope of the project is clearly defined to be limited
to optimising the operation’s value subject to existing constraints. Since a strategy
optimisation study will be evaluating a range of cut-offs and hence head grades, some
strategies investigated may be high production rate / low-grade cases, which may only
encounter ore stream constraints, while others may be low production rate / high-grade
cases, encountering only product stream constraints. Metallurgical staff will frequently
have a list of sequential upgrades for the plant. Typically these will be a mix of ore
stream and product stream upgrades, based on an assumption of increasing ore feed
tonnages at the same head grade. To properly account for the distinction between ore
and product in the analysis, it’s necessary to separate the list of plant upgrades into
separate sequential lists for each process stream.
As well as identifying upgrades to be implemented, it will be necessary to specify
any changes in metallurgical performance that result from capacity changes, such as
recovery and product quality. Operating and sustaining capital cost changes may also
need to be accounted for.
The increasing focus on geometallurgical properties of mineral resources is leading
to an increase in the amount of information that may be taken into account to optimise
operating strategies. The example of different milling rates for different rock types has
been available to evaluation teams for many years, but more sophisticated relationships
relating various metallurgical performance parameters to mineralogy and other rock

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 169
 CHAPTER 8 | The Mine Strategy Optimisation Process

properties are being developed. The availability of extra information, however, while
enabling extra opportunities to add value, potentially increases the complexity of the
analysis that must be undertaken to do so.

OPERATING COST INPUTS

Chapter 2 describes the types of cost behaviours to be accounted for in a strategy
optimisation study. As with the physical inputs to the study, the cost models should
be as simple as possible to minimise the complexity of the calculations, but sufficiently
detailed so that differences between cases brought about by the options and scenarios
selected will be described accurately to discriminate between cases.
The simplest cost model is a fully variable single cost-per-tonne model. At many
operations, cost reports only report total cost dollars and the resulting unit cost per
tonne, without any further attempt to identify what physical quantities are driving
the costs. This may be satisfactory for management control purposes, but is usually
inadequate for a strategy optimisation evaluation.
As soon as an optimisation study considers changing production rates, a fixed-and-
variable model will be necessary since not all costs increase in direct proportion to the
quantities involved. Some are related to time only, not physical activity. Similarly, if
considering changing cut-offs, ratios of various physical quantities, such as ore tonnes
per metre of development and product quantities per tonne of ore, will potentially vary.
For a given production rate and cut-off, the ratios of all the various physical parameters
will tend to be constant, and the overall cost per tonne of ore implicitly accounts for the
current ratios of physical quantities to each other.
For a more detailed optimisation study, the general principle is that variable costs
need to be specified separately for any physical parameter that does not vary in direct
proportion to ore tonnes as any of the options evaluated are varied. The reason is that
any overall average cost-per-tonne will include the costs associated with all physical
parameters, assuming that the ratios of the quantities involved are unchanging. If the
physical ratios change with changing strategies, the unit cost will as well. As discussed
in Chapter 3, this will introduce a circularity into the analysis that is resolved by relating
costs to the physical parameters that actually drive them; however, any physical
parameter for which quantity maintains the same ratio to ore tonnes across the range of
strategies need not be costed separately – its costs can be included in the variable cost
per tonne of ore.
This rationale is based on the common approach of expressing all costs as a unit cost
per tonne of ore; however, discussions in previous chapters have indicated that, as a
minimum, variable costs also need to be separated into those that are related to rock, ore
and product. If this is recognised and implemented, the rationale would be expanded
to a principle that variable costs need to be specified separately for any parameter that
does not vary in direct proportion to either rock, ore or product quantities. The costs of
those activities that do vary in direct proportion to either rock, ore or product quantities
can be included in the variable cost per tonne of rock, ore or product as appropriate.10

10. If there is a corporate requirement to report the quantities and costs of such activities, the strategy optimisation
should be as simple as possible to reduce the amount of computation required. The required outputs can then
be derived for final reporting for the final case/s presented after the optimisation has been completed.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 170
 CHAPTER 8 | The Mine Strategy Optimisation Process

If there are separate constraints for these other activities, their physical quantities at least
will need to be modelled separately so that appropriate constraints are applied, even if
costs are included in the rock, ore or product-related variable unit costs.
Fixed costs will often be best split into those that are related to the mining operation,
the treatment operations and general and administration overheads. With open pits,
treatment of stockpiles often continues beyond the closure of the mining operations, at
which point the incurring of mining-related fixed costs will cease, but plant fixed costs
continue.11 It may be appropriate to identify general and administration fixed costs with
and without mining occurring. In some organisations, where the custom of expressing
all costs as unit costs per tonne of ore is particularly entrenched, one may find that,
while variable costs are, correctly, expressed as a constant unit cost per tonne for all
levels of activity, fixed costs are expressed as a cost per tonne that varies with changes
in the production rate. If the production rate is to be varied in the strategy optimisation,
the fixed cost input has to be expressed as a function of the production rate. This is
convoluted logic that can be simplified by recognising that fixed costs are incurred
because of the elapse of time, not by the quantities of physical activities that occur in
each period. Fixed costs should be expressed as a constant cost per time period.
For open pit mines, variable costs will typically be driven by:
•• rock tonnes drilled, blasted and loaded
•• truck tonne-kilometres hauled or hours of operation.
In many cases, the haulage cost is accounted for by generating a mining cost per
tonne of rock that varies with depth, so that all costs are driven by rock quantities,
but the unit cost is a function of the average depth of mining. Either way, total mining
costs will be driven by two parameters that must be generated: rock quantities and
a distance measure, such as average haulage distance or average depth of mining.
Alternatively, the geological block model blocks, and hence the mining blocks used
in the strategy analysis, may be allocated a mining cost that accounts for differences
in haulage costs. In this case, mining costs are derived directly from parameters in
the mining blocks’ data, rather than from mining physicals and unit costs specified
as input data. In some cases, the underlying truck productivity and fuel consumption
rates, based on haulage route specifications, may be included in the block model and
hence in the mining block data.
As discussed in previous chapters, there may need to be a base case cost assuming
all rock mined is dealt with as waste, and a differential cost, which may be positive or
negative, for rock that is to be classified as ore.
For underground mines, variable costs will typically be driven by:
•• development metres, perhaps split into several categories
•• ore tonnes drilled, blasted, mucked and hoisted
•• truck tonne-kilometres hauled or hours of operation
•• backfill tonnes, perhaps split into several categories.

11. As identified in the discussion of Lane’s methodology, the mining processes are represented in an underground
operation by the major access development. It may therefore be appropriate, if the difference is material, to split
underground mining fixed costs into those related to development and production activities, as development
activities may finish some time before production.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 171
 CHAPTER 8 | The Mine Strategy Optimisation Process

In the treatment plant, variable costs will typically be driven by:

•• tonnes reclaimed from stockpiles
•• ore tonnes treated
•• product quantities generated.
In many companies, stockpile management and reclaim costs are reported as mining
costs, since the equipment used for the reclaim will often be part of the mining fleet
and the reclaim operations are managed by the mining department staff. From the
optimisation point of view, they are better seen as the first part of the treatment process,
as they will typically only be performed for ore and carried out immediately prior to the
material’s entry to the treatment plant.
Labour costs are often included in fixed costs. Depending on the level of accuracy
of the analysis, it may be appropriate to model labour numbers by fixed and variable
relationships relative to various physical activities, and then apply a cost per worker
to the number of workers thus derived. There might need to be different categories of
workers with different costs for different classes of work.
If there is a need to account for inflation rates for various cost categories, or for the
possibility of a step-change price shock – for example, for fuel or energy – it may be
necessary to model each category separately in addition to the physical cost driver splits.
Although fixed costs are, as a general rule, the same for all levels of activities across
the full range investigated, some are fixed across smaller ranges. These are sometimes
referred to as capacity or step variable costs. An example is the labour cost associated
with truck haulage. As a mine gets deeper, the truck hours needed to maintain the
tonnage targets increase. When the required hours exceed the hours available for
the existing number of trucks, an additional truck must be acquired, as well as extra
operators and maintenance personnel to maintain it. The costs of these workers will
then behave like fixed costs, incurred per time period until the haulage distance
increases further and another truck and its associated labour is acquired.
It is important to note that the split between fixed and variable costs may be dependent
on the overall time frame. Costs that are fixed in the short term will often be variable in the
longer term. Consider, for example, the labour cost of the maintenance workforce. Once
a short-term plan has been developed, such as the annual budget, the number of vehicles
in the mining fleet, their work hours and their estimated maintenance requirements will
be determined, and the maintenance workforce size will be set accordingly. This will
set the total labour cost of the maintenance workforce, which will not change if mining
and treatment activities fluctuate within the budget period. If asked what proportion
of the maintenance labour cost is fixed, the maintenance manager, whose focus is
potentially short term, will probably advise that it is 100 per cent or a value close to that.
However, if in the longer term the mining rates were to halve or double, the maintenance
requirements, and hence labour cost, would also be expected to halve or double. In the
longer term – which is what is implicitly being evaluated for strategy optimisation – this
particular cost is probably close to 100 per cent variable, not 100 per cent fixed.
Other cost behaviours are possible. Non-linear cost versus activity relationships are
sometimes used. For example, the annual cost is derived by multiplying a specified
value by the quantum of a specified activity (often ore tonnes treated), raised to a
fractional power. Site administration and general overheads are sometimes estimated
by such a relationship. In the author’s experience, the value used as the exponent is

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 172
 CHAPTER 8 | The Mine Strategy Optimisation Process

empirical and site-specific, and depends on which costs at what level of aggregation are
being modelled. Over a small range of activity, such non-linear cost behaviours may be
approximated as a simple fixed plus variable relationship, based on the formula of the
tangent to the cost versus activity curve at some base case level of activity. The resulting
cost estimates will then be less accurate the further the level of activity is from the base
case and the degree of non-linearity in the underlying relationship.
It is rare for mine cost accounts to record costs as fixed or variable (or non-linear), and
even if they did, one would have to ask whether the distinction was based on short-
or long-term considerations. Simply asking the person who is best able to provide
information regarding the split of fixed and variable cost components may not elicit
the right answer (as we just saw with our maintenance manager). Strategy optimisation
teams must ensure that they obtain information that is appropriate for their task. The
problem may perhaps be avoided by carefully framing the questions to be asked.
As indicated, asking what the fixed–variable split is may be unintentionally ambiguous
(as it does not mention time frames or activity levels). A better question might be to ask
what it is that we really want to know: how will the costs actually behave with changing
physical activities? What would happen to these costs if the quantity of the associated
activities were to halve or double? For short-term estimates, such as annual budgets, a
substantial proportion of costs will be fixed. For longer-term strategy optimisation, we
should expect that the majority of costs, especially operational costs, will be variable.
This discussion of operating costs has presented some potentially complex relationships.
As has been noted, the optimisation team must assess carefully how complex the overall
cost model needs to be to adequately discriminate between cases evaluated. The units
in which costs are expressed (per tonne, per ounce, etc) in various company reports
cannot be assumed to describe how they behave. The evaluation team must examine the
cost data provided and, when necessary, reallocate or reclassify, to ensure that the cost
model sufficiently accounts for how costs will behave over the wide range of situations
to be evaluated.
As with other aspects of the task, where there is uncertainty in the data – whether
relating to the actual cost or the appropriate split of fixed and variable components –
these should be recognised as scenario parameters to be flexed to ascertain their impact.
Variations in assumptions will change the reported values, but do they change the
strategic decisions to be made?

CAPITAL COST INPUTS

Capital costs have been discussed in some detail in Chapter 2. The accounting distinction
between capital and operating costs is in many ways irrelevant for strategy optimisation;
it is how costs behave that matters. Chapter 2 highlights the distinction between project
capital – typically a lump sum expenditure to acquire some new capability – and
sustaining capital, which may be thought of as irregular maintenance expenditure to
preserve the capabilities of the existing production facilities.

Sustaining capital
Sustaining capital will typically behave similarly to variable operating costs, and should
be treated in the same way as operating costs. The main reason to maintain the capital/
operating distinction will be to correctly handle the tax treatment of costs – they will be

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 173
 CHAPTER 8 | The Mine Strategy Optimisation Process

either expensed immediately when incurred or depreciated over a number of years. The
depreciation rules will vary from jurisdiction to jurisdiction, and discussion of these,
other than to identify an issue that the evaluation team will have to deal with, is beyond
the scope of this book.
Sustaining capital, therefore, will require input data that specifies costs incurred per
unit of activity, similar to operating costs. The unit of activity is a physical item that causes
the production facilities to wear out so that they need capital repairs or replacement. In
an open pit mine, most sustaining capital will be rock-related. An underground mine
will have a mix of rock and ore-related sustaining capital, and the treatment plant a mix
of ore and product-related capital.
Although sustaining capital may appear to be fixed costs expressed as dollars per year,
the reality is that very little will be a function of time elapsed – most will be the result of
wear and tear from handing rock, ore and product through the production facilities. The
exception is typically in the general and administrative overheads area, where buildings,
furniture and fittings and office equipment do deteriorate with time or become obsolete.
Often it is adequate to model sustaining capital by simple fixed and variable cost
relationships for each area of the production system (mine, plant, etc), but occasionally
more detail is warranted. Fleet replacement is a particular class of sustaining capital
that might require detailed calculations, depending on what information the project
sponsors want to extract from the study. This might require calculations of, for example,
operating hours for each truck per year and hence detailed rebuild and replacement
schedules. For the strategy optimisation evaluation this detailed analysis can usually
be replaced by lease-style calculations to model the sustaining capital costs,12 even if
vehicles are owned rather than leased.

Project capital
While project capital costs will usually not be included in direct calculations required for
any of the cut-off derivation processes described earlier, they will, however, be included
in an optimisation evaluation so that the value measures account for all costs. Project
capital for new capabilities may need to be specified for sequential upgrades of an item
of plant, or as mutually exclusive alternatives for acquiring various capabilities. In a
simple analysis it may be adequate to model this capital expenditure as a single lump-
sum amount, typically in the year before the capability acquired becomes available for
use in the operation. In a detailed analysis, it may be appropriate to model a spending
pattern over a number of time periods prior to the capability becoming available.
The timing of any such project capital expenditure for the provision of additional
capability will often be a strategic decision to be optimised. The data specifications and
modelling constructs must therefore be able to handle these requirements.
The analysis should also account for the impact of acquiring some new capability
on future maintenance, operating and sustaining capital costs. There will be different
effects for different types of project capital. For example, adding a process into the

12. The resulting cash flow timings may not be precise, but errors will usually be minor in the context of the overall
project cash flows, in total or in any particular year. The mathematics underlying the leasing costs ensures that
there is no impact on the accuracy of any NPV calculated. It may be appropriate to find the best strategies using
simple lease-type cost relationships, and calculate the actual replacement schedule and cash flows separately
for the small number of cases to be reported in detail.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 174
 CHAPTER 8 | The Mine Strategy Optimisation Process

existing sequence of processes in the plant without increasing the throughput would
result in additional variable maintenance, operating and sustaining capital unit costs,
and perhaps fixed costs. Duplicating an existing process to increase capacity might
increase fixed costs, but would probably not change the variable maintenance, operating
and sustaining capital unit costs. In the mine, acquiring an additional truck to deal with
the increasing depth of mining would probably not increase variable maintenance,
operating and sustaining capital unit costs per tonne-kilometre hauled or per truck
hour, but would increase the cost per tonne of rock mined.
In some situations, increases in capacity are suggested but the associated capital costs
have not been engineered. Non-linear cost versus activity relationships are often used
to derive an initial estimate. The estimated cost for a proposed capacity is derived by
multiplying the engineered cost for a base case capacity by the ratio of the proposed
and base case capacities raised to a power less than one. Six-tenths seems to be the
default consensus value for the exponent for this type of process, giving rise to what is
sometimes called the six-tenths rule13 for capital cost estimating.

Working capital
Working capital may have an impact on cash flows and value. Simplistically, working
capital can be thought of as the result of the difference in timing between expenditure
of costs and receipt of revenue. Operating costs will need to be expended for some time
before first revenues are received. Since models typically generate physical activities,
working capital is often estimated by specifying a lag (usually) for revenue cash flows
between the time of production and the receipt of revenue, and a lead or lag for cost
cash flows (depending on stock levels, consumption rates and payment terms for cost
components). First fill – the initial supply of various consumable items to make all parts
of the production process operable – may be a significant working capital item that
requires separate specification.

SOCIAL ISSUES
Depending on the operation’s location, social issues may have an impact on the costs
and timing of the start-up of operations. Items that need to be accounted for, in terms of
dollars and time, may include:
•• relocation of towns, villages and social infrastructure
•• provision of facilities for the local area, such as schools and hospitals
•• provision of job opportunities in the region, both directly in the operation and in the
supply of goods and services to the operation
•• compensation to people affected by the operation
•• requirements and expectations of the various levels of government for returns such
as taxes, duties, employment and provision of infrastructure
•• economic sustainability of the region following closure of the operation.
Many of these issues could be considered as options or scenarios for strategy
optimisation. For example, it may be useful to identify the value of the operation with or

13. The six-tenths rule may also be applied to operating cost estimates, and the exponent may vary for different components
of the cost. See Chapter 5 of the AusIMM’s Cost Estimation Handbook (Lanz, Seabrook and McCarthy, 2012).

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 175
 CHAPTER 8 | The Mine Strategy Optimisation Process

without the relocation of people or infrastructure. Similarly, the time required to reach
agreement on many issues can generate a loss of value by deferring the project. This
sort of knowledge can assist decision-makers to assess the time or money that is worth
spending on such matters. For a marginal project, these issues could make the difference
between whether it proceeds or not.
The initial capital and, if relevant, ongoing costs of these social issues are just as much a
cost of the operation as the direct mining and treatment costs. They should be accounted
for in the evaluation in the same way as similarly behaving operational costs.

REVENUE ITEMS AND ECONOMIC PARAMETERS

Revenue items include forecasts of product prices and sales terms and will be
specified by the company commissioning a study. There may be alternative sets of
price projections, often referred to as base case, optimistic or upside, and pessimistic
or downside. The strategy optimisation team should not be constrained to use only
these official forecasts. An important aspect to be derived from the study is identifying
what would cause strategic decisions to change, and price is often a decision driver.
The optimisation team should be free to evaluate prices that are well outside the range
of official forecasts, to identify how robust optimum strategies based on the official
forecasts actually are.
If the project will be a major producer in the market, it may be appropriate to have
relationships between the price and amount of product generated.
Other economic parameters affecting both revenues and costs include exchange rates
and inflation rates, which will also be specified by the company commissioning a study.
There can potentially be a number of inflation rates, such as international (perhaps
applied to product prices) and domestic (applied to cost items). There may also be
different escalation rates for prices and particular categories of costs, such as labour,
fuel and energy, in addition to the general inflation rates.
Discount rates are an important input for deriving value measures such as net present
value (NPV). As with prices, these will be specified by the company commissioning the
study. Net cash flows and discount rates can be in real or nominal terms, and before
or after tax. It is important to ensure that cash flows and the discount rate applied to
them correspond.14 There can be concern about the impact of discount rate on strategy.
As shown in Figure 6.9 and the associated discussion, the discount rate will impact
on the value of NPV, so if purchase or sale price is the issue, getting the discount rate
right may be crucial; however, the discount rate may have little impact on operating
strategies. As with other uncertain parameters, the discount rate could be treated as a
scenario variable and flexed to determine its impact on strategic decisions.

14. It is beyond the scope of this book to discuss discount rates and discounted cash flow calculations. The reader is
referred to any standard textbook on project evaluation if more information is wanted. For completeness here,
it is noted that the NPV derived from nominal cash flows and a nominal discount rate should be the same as one
derived from real cash flows and a real discount rate; however, the before-tax and after-tax NPVs can be different
and either could be the larger. It is incorrect to apply both real and nominal discount rates to the same cash flow
and call the results real NPV and nominal NPV.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 176
 CHAPTER 8 | The Mine Strategy Optimisation Process

TAXATION
Taxation generally needs to be included in a strategy optimisation analysis. Revenues
and operating costs are usually accounted for easily, while capital expenditure will
be accounted for by depreciation deductions to derive a taxable profit. Depreciation
rules will vary from jurisdiction to jurisdiction. In many jurisdictions, working capital
is not deductible for tax purposes. At the level of accuracy of most studies, it will be
satisfactory to aggregate capital expenditure into a smaller number of categories than
would be used for formal tax accounting.
Although there are many intricacies in taxation law, it is rarely necessary to take that
into account in a strategy optimisation evaluation. Indeed, one could suggest that if the
strategic decision depends on the intricacies of the tax effects, it is an entirely different
order of problem from what would normally be necessary to determine the optimum
strategy for such items as cut-offs, mining and treatment rates.
Royalties are for convenience classified here as a form of taxation. Many companies
record them as a cost, often expressed as a cost per tonne of ore. As described in
Chapter 3, they are usually classed as either:
•• ad valorem, expressed as a percentage of the net price after deduction of allowable
costs, which are typically off-site realisation costs such as treatment and refining
charges, and perhaps freight and transport from site to smelter, though some
jurisdictions charge ad valorem royalties on gross value of product
•• tonnage-based, expressed as a simple dollar charge per tonne of product, or
sometimes per tonne of ore
•• profit-based, expressed as a percentage of net profit.
In some jurisdictions, royalty specifications may be a combination of these, or the rate
specifications may be more complex than a simple unit cost or single percentage. The
optimisation model will need to generate royalties at an appropriate level of accuracy.
Expressing the royalty as a cost per tonne of ore may be satisfactory for financial
reporting, but unless the royalty is actually levied as a charge per tonne of ore, this is not
the appropriate way to model it for strategy evaluation.

END-OF-LIFE ISSUES
The major end-of-life issues that need to be accounted for are redundancy payments,
mine closure and rehabilitation costs, which include ongoing commitments, and salvage
values of both capital items and stocks of consumable items.
If the project has a long life, the impact of these on NPV could be negligible and they
may be safely ignored, although some companies’ procedures require these items
to be taken into account, regardless of the life of the operation; however, if the life is
relatively short, these cash flows may impact significantly on value. The discussion of
Lane’s limiting cut-offs in Chapter 5 has indicated how high closure costs may drive a
reduction in cut-offs to extend mine life and defer incurring closure costs.
The longer the remaining life of the operation, the more uncertain the data for these
cash flow items might be. Well-estimated costs for what is current best practice for many
of these issues may turn out to be totally inadequate years later when accepted best
practice may have altered. The impact of ongoing resource discoveries on the life of the

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 177
 CHAPTER 8 | The Mine Strategy Optimisation Process

operation will add to this uncertainty. Again, in such a situation, the values can be flexed
to ascertain whether they have an influence on current decisions.
It may be necessary to derive relationships that describe how these costs may vary if the
life of the operation changes. For example, salvage values can reduce if the age of items
disposed of increases, and redundancy payments can vary if the workforce’s average
number of years of service varies with the operation’s life. Ongoing rehabilitation during
the operating life may impact the end-of-life rehabilitation costs if the life changes.
The difference between capital expenditure cash flows and the associated accounting
depreciation charges is generally well understood. It is equally important to distinguish
between accounting provisions for mine closure costs during the life of the operation
and the actual cash flows at the end of the life. There may, however, be cash outlays
during the operating life to meet closure bond requirements, with drawdown of the
bonds helping to fund closure costs. In this case, depending on interest and discount
rates, the cash flows and accounting provisions may be similar. The way cash flows are
actually incurred should be identified and modelled appropriately.

BUILDING AND USING THE OPTIMISATION MODEL

Having identified the options and scenarios to be accounted for and all the data
inputs and relationships, they must then be combined into a model. Depending on the
optimisation techniques and software used, the model may be set to show the results of
various combinations of inputs, and/or be processed in some way so that the optimum
strategy can be identified. Chapter 13 discusses various optimisation methodologies
that might be used. For now, it is sufficient to note that some methodologies may help
produce multiple combinations of options and scenarios, so that hills of value (such
as illustrated in Figures 6.6 to 6.10) can be generated to identify the optimum strategy,
indicating both the best value and the response of value to changes in strategic decisions
(options) and external parameters (scenarios). Other methodologies may simply identify
the set of options that optimises the value measure, given the set of assumed scenario
values.

Time periods
An important consideration for the optimisation model is the lengths of the time periods
used. Existing mine planning and scheduling software applications can create schedules
with effectively continuous time scales, theoretically allowing activity durations to be
expressed to fractions of a second. Strategy optimisation models, however, typically
work with sequential time periods of specified durations. Study sponsors often like to
have a mix of durations in final reports. For example, it may be preferred to have the first
two years reported monthly, the next three years quarterly and annual periods from the
sixth year on. The question then is: is it appropriate to conduct optimisation modelling
using the final reporting time periods? The answer will frequently be no.
As a first consideration, as described for mine scheduling inputs, it will be rare for
a change in strategy to be made instantaneously. There will be a period at the start of
the evaluation time frame that needs to have all its physical activities locked in and, in
some cases, a reduced amount of locked-in activity extending into later time periods.
A significant portion of the initial time frame required to be reported in short time
periods (for example, monthly) may therefore be taken up by locked-in schedules.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 178
 CHAPTER 8 | The Mine Strategy Optimisation Process

Regardless of any initial locked-in activities, the optimisation model will need to be
capable of scheduling reasonably realistically any of the large number of alternative
cases that could potentially be evaluated. The scheduling rules will therefore have a
number of approximations. The implication is that even if very detailed schedules are
able to be generated, it is unlikely that monthly outcomes are accurate predictions of
what will happen in that month some time into the future.15 Conducting the optimisation
with monthly time frames, although giving an appearance of accuracy, is not achieving
it. One could argue that generating accurate, detailed short-term plans is not the task
of a strategy optimisation study anyway. Depending on the optimisation methodology
employed, monthly time frames may result in an unacceptably large number of periods
handled by the optimisation software.
Annual time frames are often ideal for reporting, but in many cases they will be too
coarse for optimisation modelling. For example, there will often be scheduling rules
where one activity cannot commence until another has been completed. The simplified
scheduling process in the model is often such that any conduct of the predecessor activity
in a time period precludes the commencement of the successor activity in that same time
period. Consider the situation where the predecessor activity is completed early in a
time period. If the time period is a year, then it may be almost a year before the successor
activity can commence. This may be unacceptable for realistic scheduling. Monthly time
periods can alleviate this problem but, as noted, may be impractical. The author has
found that quarterly time frames pose a good balance between schedule reality and
the number of periods modelled. A delay of up to three months or, on average, six to
seven weeks (half a quarter) between predecessor and successor activities represents a
reasonable activity duration overrun or unplanned delay between successive activities.
The accuracy of the derived schedule and the length of the modelled time periods will
also depend on the duration of activities relative to the average potential gap between
activities generated by the scheduling rules. If modelled activities take on average, say,
one month, but the average enforced delay between sequential activities is six weeks,
quarterly time frames will be unacceptable. The average duration of activities in the
schedule need to be greater than some multiple of the average duration of activities, and
that multiple will itself depend on the desired accuracy.16
If, on the other hand, the predecessor–successor relationships can be modelled in a
way that permits portions of predecessor and successor activities to be scheduled in
the same time period, some of these problems will not exist. Whatever the situation, the
optimisation team has to identify the capabilities and limitations of the software, the
complexity of predecessor–successor and other scheduling constraints and the level of
accuracy of the study, and hence determine the best durations of sequential time periods
to be used. It will often be possible to use different durations for different time periods,
typically shorter periods earlier in the operation’s life and longer later, when mining
blocks may be larger and have less accurate data.

15. Even with schedules produced by experienced engineers of detailed budgets and forecasts for one well-defined
plan, the reality is that with the normal variabilities encountered in mining, they will prove to be inaccurate
even a few weeks or months into the future, and are often obsolete by the start of the budget year. Schedules
produced by rules designed to be suitable for a range of inputs cannot be more accurate than that.
16. Simplistically, the accuracy of the schedule duration or overall production rate will be given by the ratio of the
average inter-activity delay duration to the average activity duration for activities on the critical path.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 179
 CHAPTER 8 | The Mine Strategy Optimisation Process

Shorter time periods modelled by the optimisation software (such as quarters) can
be aggregated into longer periods (such as years) for reporting purposes. Where the
required reporting periods are shorter than what is needed for optimisation, the ideal
situation is for the optimisation team to convince the project sponsors that the shorter
time periods are inappropriate and the optimisation time periods can be used for
reporting. If this cannot be done, the longer periods can be simply divided into the
required number of shorter periods, with all activities from optimisation model periods
spread evenly across the shorter reporting periods. If this is inadequate for reporting,
it may be necessary to do detailed schedules at the level of required time granularity
in order to report early years’ information for cases that are to be reported in this way.
As noted, many of the activities in these shorter time periods may in fact be locked-in
activities that will be the same for all cases reported.

Calculating cash flows

There are many ways in which cash flows can be estimated. All will depend first
on having a good model of the physical activities that are actually driving the costs.
Assuming that we have a reliable model of the physical activities, how then should
cash flows be modelled? This topic could form a book in its own right. The following
discussion, therefore, is a high-level guide to complete the major issues facing a team
conducting an optimisation study. The discussion describes what might be thought of
as the most complex and most accurate cash flow model. What is done in practice will
depend on the level of accuracy of the study. The effects of common simplifications in
cash flow models are described following the description of the ideal model logic:
1. Obtain variable unit costs, fixed costs and revenue parameters in the currencies in
which they are incurred. It may be necessary to distinguish between currencies, for
example, for expatriate and local labour costs and for consumables sourced from
various locations. Where possible, costs (and revenue items) that are quoted in one
currency but are incurred in another, such as petroleum products and product prices,
should be identified and modelled in the original rather than quoted currency.
2. Apply the appropriate cost and revenue relationships to the physical quantities
scheduled to derive periodic (monthly, quarterly, annual, etc) costs and revenues in
uninflated terms; that is, in today’s money at today’s cost levels, ignoring the future
effects of general inflation and escalation, in the currencies in which they are incurred.
3. Apply factors to account for inflation and escalation, still working in the original
currencies, to derive cash flows in nominal terms; that is, in money of the day,
the values that would be recorded at the time the costs are incurred or revenue
obtained, assuming our estimates are accurate. Typically, there is a single rate for
general inflation, but different escalation rates may be appropriate for different cash
flow items such as product prices, labour, power, fuel and consumables. Escalation
rates will be expressed relative to the general inflation rate and may be positive or
negative. General inflation and escalation may also be combined into differential
inflation rates for various cash flow items. These factors may also incorporate step
changes associated with price shocks if any are predicted. It may be necessary to
take account of the age or date of the data used for each item. Often, product price
projections will be relatively recent and expressed in today’s money; however, cost
data might be taken from the previous financial year’s cost reports and be more
than a year out of date. If relevant – for instance, if there is a high inflation rate – it

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 180
 CHAPTER 8 | The Mine Strategy Optimisation Process

may be necessary to apply inflation factors to items of data to ensure that all items
are expressed in money of the same date, typically time zero in the model.
4. If necessary, convert all cash flow items to the currency of the country taxing the
operating entity, usually the country where the operation is located. This should be
done using nominal exchange rates; that is, the exchange rates between currencies
that would be seen at the time in each period. Exchange rates are often quoted in real
terms, but differences in inflation between countries will alter the relative buying
powers of currencies in addition to any real changes that are forecast.
5. Calculate depreciation for tax purposes – and hence taxable profit and tax, and
nominal net cash flow – in the currency of the taxing country.
6. Convert this to the reporting currency in nominal terms, again using nominal
exchange rates with nominal cash flows. In some cases it may be necessary to do
further tax calculations for the country where the company is based.
7. Deflate the nominal net cash flow after tax to real cash flows using the appropriate
deflator for the reporting currency or country.
8. Discount the real after-tax cash flow using the real after-tax discount rate to obtain
the NPV.

Implications of simplifying assumptions

Simplifying assumptions are often made in cash flow analyses, but they may affect the
accuracy of the outcomes:
•• Ignoring general inflation – inflation erodes the value of depreciation. Although costs
are inflating, once capital is spent it is depreciated on the basis of the cost incurred at
the time of acquisition. Depreciation is therefore expressed in terms of the money of
the year of expenditure, but those amounts are then treated as nominal amounts when
determining tax liabilities in later years, when the purchasing power of the money
has declined due to inflation. Ignoring general inflation therefore overstates the value
of depreciation, reduces the tax payable and therefore results in an overstated value.
For comparative purposes, this might not be an issue if the options considered have
similar levels of capital intensity and spending patterns; however, if the comparison is
between a high-capital/low-operating cost option and a low-capital/high-operating
cost one, the relativities could be distorted by this simplification.
•• Ignoring differential inflation or different escalation rates – if cost inflation is greater than
revenue inflation, margins will reduce and the mine may have to close earlier than
indicated without accounting for the inflation effects. The value of the operation will
be overstated. Conversely, if cost inflation is less than revenue inflation, margins will
increase and the mine may have a longer life than indicated without accounting for
the inflation effects. The value of the operation will be understated.
•• Ignoring taxation or conducting a pretax evaluation – this may potentially create invalid
comparisons between options. It may not be a major issue if the options considered
have similar levels of capital intensity and spending patterns; however, as with
general inflation, if the comparison is between a high-capital/low-operating cost
option and a low-capital/high-operating cost one, there could be problems.
•• Using over-simplified cost driver relationships – simple cost driver models are often used
in mining economic evaluation models. Often, particularly in high-level studies,
a single unit cost per tonne will be used. This ignores the fixed and variable split

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 181
 CHAPTER 8 | The Mine Strategy Optimisation Process

of costs and will not therefore adequately account for changing activity rates over
time. A fixed cost per year plus a variable cost per tonne of ore model is also often
used, but this ignores changes in ratios of physical quantities. Chapters 2 and 3 have
discussed the various costs that may be associated with rock, ore and product, and
earlier sections of this chapter have indicated the need for more detailed physical
relationships to be identified to handle the effects of strategic decisions to be made.
Simple relationships may be adequate if there is little future variation from current
ratios of various physical quantities and activity rates over time: these are implicit
in the current unit cost per tonne. If there are variations over time, simplifying
assumptions can lead to wrong conclusions and hence suboptimal decisions.
In all these situations, there will be a trade-off between simplicity, accuracy and
reliability. There is no simple answer to this issue. In part, its resolution will depend on
the level of accuracy of the evaluations, as highlighted in Chapter 7. It may be that there
is one level of accuracy needed in the data, assumptions and relationships to achieve
a reliable result, but another more detailed level required to prove that this is so (or
perhaps to silence sceptics).

WORKING WITH DIFFERENT COMMODITIES

Base and precious metals

Virtually all the discussion to this point directly applies to base and precious metal
operations. Although there may be constraints within an operation that affect how
quickly a new strategy is implemented, as discussed, there are no constraints imposed
by the market. Products will typically be sold within the terms of existing contracts
or freely on the open market. Changes of operating strategy suggested by cut-off and
optimisation studies will be internal to the operation and have product effects that can
easily be absorbed by the market without special considerations.

Bulk commodities
In many cases, the requirement for bulk commodity operations is to maximise the
amount of material above a contracted quality specification, making the problem largely
one of scheduling, stockpiling and blending. Optimisation studies for plans of such
operations may not need to optimise cut-offs to maximise value in the same way as
base and precious metal operations, but can require some cut-off criteria to control the
properties of blends produced as the grade distribution mined varies over time.
However, if variations in resource quality translate to variations in value, Lane’s
methodology or strategy optimisation studies that do involve cut-offs may apply. With
bulk commodities such as iron ore and coal, some form of beneficiation or washing is
conducted to upgrade the ore as mined to a saleable or more profitable end product. The
only difference between bulk commodities and base or precious metals in this regard
is that with bulk commodities a large proportion of the ore (perhaps even 100 per cent)
becomes final product, whereas with the latter, a relatively small proportion of the ore
becomes product, but this does not invalidate application of the methodologies.
Bulk commodity sales contracts may provide little flexibility for product quality and
perhaps volumes, though, depending on the state of the world economy, there may be
some potential for open market sales of additional product. A wide-ranging cut-off and

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 182
 CHAPTER 8 | The Mine Strategy Optimisation Process

strategy investigation, however, may identify potentially better strategies for volumes
and product mixes that could add value in the longer term. The cut-off concept may
have to be expanded to account for the allowable grades in the final product of both the
primary commodity sold and deleterious impurities. Changing cut-offs should therefore
be thought of as changing not only the minimum grade of the main product but also the
maximum grades of the contaminants.
For example, increasing an impurity grade limit will result in a higher impurity grade
in the product; this may perhaps make it unsaleable with current contracts, but could be
saleable with new contracts, albeit with a lower price. However, the increase in tonnage
may more than outweigh the loss in the unit value of the product, resulting in increased
revenue and profit. This new product would become the focus of marketing activities
for future periods after current contracts expire. The locked-in period for activities
in a strategy optimisation may therefore be longer than for a base or precious metal
operation, but conceptually there may be little difference between the processes used.
Cut-off optimisation is not inapplicable for bulk commodities, as is sometimes postulated.
It simply requires a different outlook and time frame from base and precious metals.

Market-limited products
Some products, such as diamonds and specialty metals used in niche applications, may
be constrained by an external market limitation. In one sense, this is the case with all
products, but for many, an individual mine will often be a small contributor to the global
supply. Its profitability and operating strategies will not be affected by the physical
market limitations and influence of its own outputs on the market.
Discussions in this chapter and Chapters 5 and 6 in particular have recognised a
product-related constraint, but noted that it is usually a physical constraint within the
operations and downstream product-handling facilities. Yet the possibility of a limit on
the amount of product that can genuinely be sold into the global market was recognised.
Those discussions therefore generally cover market-limited products.
In some instances, an operation’s output may impact the price of the product.
Simplistically, the discussions assume a specified product price, but there is no reason to
exclude the assumption of a price varying in relation to the amount of product generated
by the operation. If such a relationship is derived, it can easily be accounted for in the
strategy optimisation processes described in Chapter 6 and in the iterative processes
relayed for Lane’s methodology in Chapter 5.

CHAPTER SUMMARY
This chapter has described the data requirements and procedures to perform a strategy
optimisation in detail. It is not possible to fully optimise the mining strategy without
considering the metallurgical treatment processes and vice versa. A full strategy
optimisation requires an integrated evaluation of all parts of the production process
across the whole life of the operation.
The corporate goals must be borne in mind throughout the optimisation process. They
indicate what is required to derive value measures for those goals, and the processes
and data to do that. As well as goals, it is vital to identify all the options to be evaluated;
data and relationships must support the full range of values.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 183
 CHAPTER 8 | The Mine Strategy Optimisation Process

It is necessary to take account of the influences of uncertain parameters in an

optimisation study. As with the most uncertain parameter in a mining strategy
evaluation, the product price, it’s the norm to estimate the most likely scenario and flex
the price inputs over a range. A change in price or any uncertain input will change the
value of any operating plan, but the issue is not the absolute value, but how changes
might cause the optimum plan to vary. If the decision-changing situation is found to be
outside the likely range for an uncertain parameter, it may be important for defining the
value, but not the optimum strategy. If it lies near the most likely value or relationship,
steps must be taken to reduce the uncertainty or consider the trade-offs between the
benefits and costs of making decisions based on the right or wrong assumptions.
Following is a summary of the types of inputs and processes typically required.

Geology
The major activities related to geology for an optimisation study are:
•• create a reliable model for the range of cut-offs to be investigated
•• generate potential orebodies at each cut-off
•• identify domains that require different mining methods, scheduling constraints,
metallurgical treatment, etc.
If the model is unreliable for some cut-offs, rigorous analytical processes can still be
conducted using the existing model. If the optimum strategy is then found to lie within
the range of cut-offs where the geological model is deemed to be reliable, the identified
optimum will be reliable. If, however, the optimum strategy lies outside the reliable
range, identify the best value within the reliable range, the additional value potentially
available, and whether that justifies the cost of extending the reliable range of the model.

Mining
The major activities related to mining for an optimisation study are:
•• identify suitable mining methods
•• create realistic mining shapes and aggregated mining blocks, and hence reserves for
each cut-off or pit limit and mining method
•• produce conceptual mine designs and schedules for selected representative cases (for
model calibration)
•• aim to maximise the information obtained while minimising the work to be done
•• develop suitable interpolation relationships for other cases.
For open pits, the aggregated mining blocks will be bounded by pit shells corresponding
to each major pushback stage and by a number of elevations, spaced at a multiple of the
bench height. The optimisation software will select the blocks that are to be mined and
schedule them in such a way as to generate the ultimate pit size and optimum mining
sequence. It will usually be necessary to define haulage distances for each mining block
to both ore and waste-dumping points. The optimisation will decide what is ore and
what is waste, so both distances need to be available to the optimiser. More complex
information may be required depending on dump locations, rock types and the like.
Underground, mining blocks will typically be defined by a combination of ore zones,
sublevels, northings or eastings, geological structures, changes in the size and shape
of the orebodies and boundaries defining geological domains within the rock mass or

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 184
 CHAPTER 8 | The Mine Strategy Optimisation Process

orebodies. The mining inventory for each cut-off must take account of both the ore loss
and dilution to generate realistic stope designs and the elimination of outliers of material
above cut-off that would be uneconomic to extract.
Development requirements for each underground mining block will be needed. It may
be relevant to specify what proportion of each category needs to be completed before the
mining block can start producing. Trucking distances may vary if different haulage and
hoisting options are to be evaluated. Quantities of backfill of various types may need to
be specified if mining methods being evaluated have different fill types or proportions
of the orebody to be filled. Underground data requirements may also include such items
as access methods or topologies.
Scheduling rules need to be specified for both open pit and underground optimisations.

Processing
The major activities related to processing for an optimisation study are:
•• specify recovery relationships for the range of cut-offs and hence feed characteristics
to be evaluated
•• identify constraints and how to remove them for the ore and all product streams
•• determine the impact of debottlenecking on recovery, product quality, operating and
sustaining capital costs, etc.
Because of the wider range of feed characteristics modelled in an optimisation
study, better relationships may need to be developed for the study than those used
in the operation for a narrower range. Scenario analysis may be necessary over any
uncertainty range in the relationships to identify whether strategic decisions will change
if relationships vary. The increasing focus on geometallurgical properties of mineral
resources leads to greater information being taken into account to optimise operating
strategies. This potentially increases the complexity of the analysis to account for it.

Operating costs
A suitable cost model is required, typically incorporating fixed costs (related to time
elapsed) that do not vary across the range of activity levels evaluated, and variable
costs driven by physical activities such as development metres (underground) or waste
mined (open pit), ore tonnes mined, truck tonne-kilometres or operating hours, backfill
tonnes (underground), ore tonnes milled and product quantities.
Capacity or step variable costs may be relevant. These are fixed across a portion of the
whole range, with step changes at threshold levels of activity.
The general principle is that variable costs need to be specified separately for any
physical parameter that does not vary in direct proportion to rock, ore or product
quantities. An overall average cost per tonne will include costs associated with all
parameters, assuming that the ratios of all quantities are unchanging. If the physical
ratios change with changing strategies, the unit cost will as well. This introduces a
circularity into the analysis, resolved by relating costs to the parameters that drive them.
The way in which unit costs are reported in company documents usually does not
describe how they behave. The evaluation team must examine the cost data provided
and, when necessary, reallocate or reclassify, to ensure that the cost model accounts for
how costs actually behave.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 185
 CHAPTER 8 | The Mine Strategy Optimisation Process

Capital costs
Capital costs are separated into sustaining capital, including fleet replacement and project
capital. The effect of project capital on future maintenance, operating and sustaining
capital costs must be identified. Potential discoveries from exploration expenditure, in
or near the mine and regionally, may impact on plans for known resources and often
needs to be evaluated.
The distinction between capital and operating costs is in many ways irrelevant for
strategy optimisation; what matters is how costs behave. Project capital is a lump-sum
expenditure to acquire some new capability. Sustaining capital may be thought of as
irregular maintenance expenditure for existing production facilities. Sustaining capital
behaves similarly to variable operating costs, and should be treated like operating costs.
The main reason to maintain the capital/operating distinction for an optimisation study
will be to correctly calculate tax.
Working capital may have an impact on cash flows and value. It may be thought of as
the difference in timing between expenditure of costs and receipt of revenue – operating
costs will need to be expended for some time before first revenues are received. Working
capital is often estimated by specifying a lag for revenue cash flows between the time of
production and the receipt of revenue, and a lead or lag for cost cash flows (depending
on stock levels, consumption rates and payment terms).

Social issues
For strategy optimisation, many social issues could be considered as options or scenarios.
It may be useful to identify the value of the operation with or without the expenditure or
the loss of value due to deferring the project because of time required to reach agreement
on social issues. This knowledge can assist decision-makers to assess the time or money
worth spending on these matters.

Financial and economic inputs

These inputs include such items as product prices and sales terms, taxation and discount
rates. Although corporate policies may specify values to use, an optimisation study team
should not be constrained to use only official forecasts. The study should show how
robust the recommended plans are. To this end, it might be appropriate to indicate how
far, for example, prices could vary from the forecasts without affecting the optimum
strategy.

End-of-life issues
These include issues such as redundancy payments and rehabilitation costs and
commitments. If the project has a long life, the impact of these on NPV may be negligible
and safely ignored, although some companies’ procedures require that they be taken
into account, regardless of the life of the operation. If the life is relatively short, these
cash flows may be significant. Lane’s limiting cut-offs indicate how high closure costs
may drive a reduction in cut-offs to extend mine life and defer incurring these costs.
The longer the remaining life of the operation, the more uncertain the values for these
types of cash flow items. Well-estimated costs for what is current best practice may turn
out to be inadequate many years later when accepted best practice may have changed.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 186
 CHAPTER 8 | The Mine Strategy Optimisation Process

The impact of ongoing resource discoveries on the life of the operation will also add to
this uncertainty. The values can be flexed to ascertain whether they have an influence
on current decisions.

Data adequacy
Generating adequate data for an optimisation study may require a significant amount of
human-generated mine design, and availability of technical staff to do this work may be
a practical constraint on the overall strategy optimisation process. The guiding principle
for data generation is: maximum information from minimum work. Exactly how this
balance is achieved in each case depends on the complexities of the mineralisation and
mining methods, the engineering judgement of the technical staff and the corporate
culture of the company undertaking the study. If the optimum solution found by the
software is one without engineered designs, there should be sufficient input into other
cases from which the data for the optimum case has been interpolated so that results can
be presented with confidence.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 187
 HOME

Chapter 9

Specifying the Grade Descriptor

an introduction TO Grade Descriptors

In this book, the term grade descriptor refers to any parameter that describes the value of a
block and to which we may apply a cut-off to separate material into different categories.
The purpose of the grade descriptor is to assign values to volumes of rock (such as those
in the geological block model) with larger numbers representing greater value. Grade
descriptor values should rank blocks in the same order as their true value, identifying
that this bit is more valuable for processing than that bit.
The grade descriptor does not have to be an absolute true value, and often it won’t be.
Any measure of grade is a ranking tool only. Relativity is the important issue, but only
in rank, not proportionality. As will be shown, even single metal grades will not express
value ratios accurately; for example, material with a grade of, say, 2 g/t will not be
worth twice the value of 1 g/t material. Grades expressed as dollar values – such as net
smelter return (NSR), dollars per tonne – may also not describe the actual value gained
by treating that material.
The specification of a grade descriptor is essential to a strategy optimisation study for
a polymetallic or multiproduct mine to generate a single value to quantify the relative
value of each block. For a simple single-metal deposit (for example, disseminated gold
mineralisation), the actual grade (gold grams per tonne) is usually quoted, naturally
achieving the required ranking; however, there are many grade descriptor issues that
are associated with all styles of mineralisation, including single-metal deposits. Much of
the discussion in this chapter will be relevant only for multiproduct deposits, but other
aspects (such as grade descriptors not expressed on a per tonne basis) are relevant to all
styles of mineralisation.
It is not essential that there be a linear relationship between the grade descriptor and
the true value; non-linearity is often the case. Even with a simple gold deposit, if there is

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 189
 CHAPTER 9 | Specifying the Grade Descriptor

a non-linear relationship between recovery and grade, there is a non-linear relationship

between value and grade. With a linear relationship, the relativities of the grade
descriptor value and the true value may differ. For example, if a gold treatment plant
operates with a constant 0.5 Au g/t tailings grade, an in situ grade of 2.0 g/t will have
a recoverable grade of 1.5 g/t, while an in situ grade of 1.0 g/t will have a recoverable
grade of 0.5 g/t. The ratio of in situ grades is 2:1, but the ratio of values (assumed for now
to be represented by the recoverable grades) is 3:1.
In polymetallic deposits, common practice is either to generate an NSR dollar value
for each block, or apply multipliers to the metal grades for each constituent to derive
either a dollar value (typically per tonne) or a metal equivalent (typically expressed as
the grade unit of the dominant metal). The typical value equivalence formula is:

Value = a × Metal A grade + b × Metal B grade + c × Metal C grade + …

Mathematically, this is a linear function of a number of variables, which in this specific

application are the metal grades. Geometrically, the formula describes a plane in
multidimensional space. In this formula, the value derived will often be a dollar value;
however, if a metal equivalent grade – say, based on Metal A – is required, the formula
would be:

Metal A equivalent grade = Metal A grade + b’ × Metal B grade + c’ × Metal C grade + …

where:
b’ = b/a
c’ = c/a, etc
There are several ways in which the multipliers are commonly derived, and the
calculations, advantages and disadvantages of each are described in the following
subsections. In this discussion, any value used as the grade descriptor will be termed
a grade, regardless of whether it is a single-metal grade, a metal equivalent, or a dollar
value, unless the context indicates otherwise.

SELECTING THE CORRECT GRADE DESCRIPTOR

The importance of selecting the correct grade descriptor is illustrated as follows.
Figure 9.1 shows the types of error that can occur if cut-offs are applied to a calculated
grade that does not rank blocks in the same order as their true values. The calculated
values are often derived from an equivalence formula while the true values may be NSR
dollar values.
It is assumed that the underlying metal grades are correct. The variations in Figure 9.1
are in addition to any grade estimation errors. Inappropriate multipliers can undo the
time and effort spent by resource geologists to reduce these errors.
The horizontal and vertical lines intersecting on the trend line through the data
points represent cut-offs applied to the calculated and the true (but possibly unknown)
values. Points in the upper-right and lower-left quadrants are correctly classified as
ore or waste. Points in the upper-left quadrant are truly ore, but are incorrectly classed
as waste on the basis of the calculated grade. They are excluded from the reserve and
represent a loss of ore. Points in the lower-right quadrant are waste incorrectly classed

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 190
 cHapter 9 | Specifying the Grade Descriptor

Figure 9.1
Typical plot of true and calculated grade values.

Incorrectly classed Correctly classed

as Waste—Lost as Ore

True Value

Correctly classed Incorrectly classed

as Waste as Ore—Degradation

Calculated Value

as ore. They are incorrectly included in the reserve, and represent a degrading of the
reserve with subgrade material.
This implies that at any specified cut-off, an unsuitable grade descriptor will produce a
resource or reserve with a grade lower than it could be. It’s uncertain whether the tonnage
will be higher or lower – this will depend on the nature of the data point distribution at
the cut-offs applied to the true and calculated grades; however, for any given tonnage,
the grade will be lower than it could be. Figure 9.2 illustrates this – the result of a case
study from a base metals operation where the original equivalence formula (Formula 1)
created a spread of true values for an equivalent value. A second formula (Formula 2)
was then derived to reduce that spread.
It can be seen at low cut-offs and high tonnages that there is little degradation effect;
the two grade versus tonnage lines are close or coincident. But with high cut-offs and

Figure 9.2
Tonnage–grade variations with different equivalence formulas.

Upgrade of Reserve
through better grade
description and
reduced misclassification
True Value

Calculated Value

Equivalence Formula 1 Equivalence Formula 2

Cut-off Grades and Optimising the Strategic Mine plan | Spectrum Series 20 191
 CHAPTER 9 | Specifying the Grade Descriptor

low tonnages, major improvements result from changing the equivalence formula. For
any given tonnage the grade has improved, and for any given grade the tonnage has
increased.
If there was a constant ratio in all blocks between the metal grades contributing to the
equivalent value, the relative weightings would be immaterial – all sets of weightings
would generate equivalent values showing the same rankings of blocks of mineralisation.
In the simplest case, all constituents other than one could be given zero weightings,
and the metal grade of the remaining constituents would be a good grade descriptor,
ranking all blocks of rock the same as their true values. The scatter plot in Figure 9.1
and differences between grade versus tonnage curves in Figure 9.2 would cease to exist.
Grade contours (in two dimensions) or boundary surfaces (in three dimensions) at
different cut-offs would remain nested in concentric sets regardless of the weightings
applied and the resulting numerical values of equivalent values that could be generated.
The numerical value given to a boundary or surface enclosing a volume of rock would
change with different weightings applied, but the boundary or surface itself would not.
In reality, the ratios of the constituents’ metal grades of a polymetallic deposit do
vary from block to block. Figures 9.1 and 9.2 result from combining these grade ratio
variations and different weightings to constituents of the mineralisation. Grade contours
(in two dimensions) or boundary surfaces (in three dimensions) at different cut-offs no
longer remain nested in concentric sets with different weightings applied. There is now
the potential for what was ore in one case to be waste in another, so there is also the
potential for sets of boundaries at different cut-offs using different weightings to cross
over each other rather than remain concentric. If one of the weighting sets (or calculation
methods) generates the true value, other weighting sets will generate the scatter plot
and the existence of points in Figure 9.1’s error quadrants. This effect is discussed in
more detail shortly.

TYPES OF POLYMETALLIC GRADE DESCRIPTORS

In the following discussion, dollar-value equivalent grades are assumed, but there
is generally a simple ratio between a dollar value and a metal equivalent grade. As
indicated, to obtain a metal equivalent, the dollar-value multiplier of the basis metal is
usually divided into the dollar-value multipliers of other ore constituents, so that these
are expressed in terms of their values relative to the basis metal.

In situ value
Using an in situ dollar value as the grade descriptor has the advantage of simplicity. It
merely applies the specified metal prices to the metal grades of any block, and its in situ
dollar value is immediately determined. That is, the multiplier applied to each metal
in the equivalence formula is its price. The typical in situ value equivalence formula is:

Value = Metal A price × Metal A grade + Metal B price × Metal B grade

+ Metal C price × Metal C grade + …

When metal prices are in another currency, the exchange rate has to be included in
the calculation. Where the unit of mass for the price quoted differs from the unit for
describing the metal grade, a further conversion factor is applied.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 192
 CHAPTER 9 | Specifying the Grade Descriptor

For example, gold grades are typically quoted in grams per tonne, and the gold price
is quoted in US$/oz (troy). The exchange rate from US dollars to the local currency must
be applied to convert the gold price in US$/oz to a local currency price per ounce. This
is then divided by the 31.10348 g/oz conversion factor to derive a local currency price
per gram multiplier to apply to the gold grade in grams per tonne. Similar conversions
will apply to base metals, where prices are often quoted in US cents per pound, to
derive multipliers to apply to percentage grades. Care must be exercised in these cases
to ascertain whether the percentage grades are recorded as percentage values with
magnitudes between 0 and 100, or as decimal fractions between 0 and 1. The latter might
then be displayed as percentages (which can be done with software applications such as
Microsoft Excel™).
The main disadvantage of the in situ value is that it fails to take account of the
proportion of each constituent that is recovered, and the net revenue then received for
what is recovered. There is potentially a spread of true values for any given in situ value,
as shown diagrammatically in Figure 9.1 (assuming that the horizontal axis is in situ
value), which can in turn generate the effects shown in Figure 9.2. The magnitude of the
spread will depend on the variability of the ratios of the constituent metals and of the
recoveries and payabilities of each.

Recoverable value
Using a recoverable value as the grade descriptor again has the advantage of simplicity
– the same processes are applied as for in situ value, except that metallurgical recoveries
are also applied to each metal. This is a simple process if recoveries don’t depend
on grade: the multiplier applied to each metal in the equivalence formula is its price
multiplied by its recovery. The typical recoverable value equivalence formula is:

Value = Metal A price × Metal A recovery × Metal A grade + Metal B price ×

Metal B recovery × Metal B grade + Metal C price × Metal C recovery × Metal C grade + …

Where recoveries are grade dependent, simple grade multipliers may be inadequate
and a more complex function of grade might have to be applied. The recovery terms in
the formula may need to be functions of the metal grades.
Compared with in situ values, recoverable values have the potential to reduce the
spread of true values for any given recoverable value.1 Yet recoverable values still fail
to account for differences in payments for recovered metals. Applying the specified
metal price to recovered grades implicitly assumes that the same proportion of the
price is received for each metal. This is usually not the case. Sales terms for base metal
concentrates, for instance, have different payable percentages for the various metals,
as well as treatment charges (TCs), refining charges (RCs) or combined (TC/RCs) that
represent different proportions of the various metal prices. Different grades of metal
concentrates will also result in different freight costs per tonne of metal, even if the cost
per tonne of concentrate is the same.

1. The author has, however, encountered situations where constituents with low recoveries have high payable
proportions and vice versa. In one of these cases, the in situ values were better at ranking blocks similar to the
true value than recoverable values. Every case must be judged on its own merits.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 193
 CHAPTER 9 | Specifying the Grade Descriptor

As with in situ values, the differences between recoverable and true values will depend
on the variations in the ratios of grades of the contained metals and in the proportions
of the metal prices actually received for each metal.

Payable value
Payable values take account of all the metallurgical losses, product-associated costs
and payable proportions of the recovered products. For a base metals operation selling
concentrates, this is represented by the NSR, less concentrate freight, transport and selling
costs, which may be referred to as the NSR at the mine gate. If a mine has its own smelter
and refinery, the costs that would be incurred by a concentrate seller – represented by
the payable percentage and the TC/RCs, plus any penalties for deleterious elements
– are incurred as downstream losses of recovery and direct operating costs, but the
principle is the same for both types of operation.
Simplistically, the formula for recoverable value could be enhanced by including a
payable term for each metal, but often the payability is a more complex relationship than
can be accounted for in a formula.
As has been noted, common practice on mine sites is to express all site costs as a cost
per tonne of ore, even though the ore tonnage has little or no bearing on many of them;
several costs, such as site administration and much of the labour cost across the operation,
are fixed, varying with time rather than volume of ore. Similarly, once concentrate has
been separated from the ore stream, many costs are related to concentrate or metal
quantities, not ore tonnes. Relating all costs to ore tonnes may be reasonable for certain
reporting functions; however, it can be dangerously misleading if used for operational
decision-making.
The best grade descriptor (in terms of ranking blocks of rock) will be the one that
identifies the net return from the block. It takes account not only of metal prices and
metallurgical recoveries, but also costs in the concentrator and downstream in the
smelter, refinery, etc that will vary according to the amount of metal recovered from
the block (that is, the costs dependent on metal grades), not on average costs resulting
from average feed grades. High-grade blocks will incur higher costs per tonne of ore
for product-related variable costs than low-grade blocks. This is not a concern for a
single-metal deposit; however, for a polymetallic deposit, these variations can impact
greatly on the value ranking of blocks where the ratios of constituent metal grades vary.
To correctly rank blocks, it is essential that variations in costs associated with various
products be accounted for. The only way to do this is by using a payable grade or value.
There are several ways to determine a payable grade. At many operations, complex
recovery relationships and product sales terms are built into the block model calculations
to generate individual NSR values for each block, assuming that the total mill feed
has the grade and other characteristics evaluated. Where this is possible, it will be
the best way to determine the payable value of each block; however, this may not be
an easy process to implement because of complex flows and relationships within the
metallurgical processing routes from ore to final products.
NSR-type payable grades will commonly be expressed as dollar values. However,
there is no reason why a metal equivalent grade cannot be developed for an NSR-type
grade. If an equivalence formula were used, the multiplier for the basis metal, which
is effectively the value of one grade unit of the basis metal, would be divided into the

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 194
 CHAPTER 9 | Specifying the Grade Descriptor

dollar-value multipliers for all other metals, as indicated earlier in this chapter. To derive
an NSR-based metal equivalent, the NSR value of a feed grade of one grade unit for the
basis metal and zero grade for all other constituents could be derived and this value
divided into NSR-based dollar value grades.

Inaccuracies common to recoverable and payable grades

Inaccuracies may exist even with the most accurate NSR calculations of payable grade.
Although generally unrecognised, these same inaccuracies apply to all other grade
descriptors, even when simple metal grades are used for a single-metal deposit. These
errors can arise from the recovery–grade relationships, which are often derived at an
operating mine from historical records of production performance, and are therefore
based on the average grades of feedstocks composed of ore blends from many sources.
There is no guarantee that the average recovery relationships can be applied to individual
ore sources or rock blocks. This is, however, frequently done, and in reality there may
be no alternative. As long as the potential problems are recognised they may not be
accounted for, but errors can be introduced if they are not. The increasing availability
of geometallurgical information may assist with reducing these errors, but at the time
of writing, non-linear processes for estimating geometallurgical properties of individual
blocks from sample points, and for recombining these to predict the behaviour of ore
blends, are not well established.
If the recovery is a linear function of head grade, the recovered product from a
blended feed, using the recovery calculated for the average grade of the blend, is equal
to the sum of the product quantities recovered using recoveries calculated separately
for the grades of the components of the blend. Here, applying the recovery formula
to individual block grades in isolation results in a good estimate, not only of the true
recoverable or payable grade of each block, but also of all blends when combined
using linear tonnage-weighted calculations.
If the recoverable grade versus head grade relationship is non-linear (for example,
where recovery is expressed as a polynomial function of head grade), the blend recovery
cannot be estimated by linear tonnage-weighted calculations applied to the blend
components. If an incremental tonnage with a particular grade is blended with a base
tonnage of a different grade, the grade of the blend will obviously differ from the grades
of the component tonnages. The recovery formula can be applied to the grades of the
base as well as incremental and total tonnages to determine the recovered product. It’s
usual that the sum of the recovered products from the components is not the same as
the recovered product from the blend. The additional product recovered by adding the
increment to the blend is therefore not the same as the recoverable product calculated
for the increment in isolation. The difference can be positive or negative, and can vary
depending on the recovery relationship and the relative tonnages and grades of the base
and incremental tonnages.
There is potentially an error in the estimated true value of a block of rock if it is calculated
in isolation using the recovery relationship based on average feed grade. In light of the
discussion in the preceding paragraph, this error may be negligible, but it may also be
relatively large, and either over-value or under-value the block. Furthermore, the error
is not constant, but depends on what the block is mixed with at the time it is mined.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 195
 CHAPTER 9 | Specifying the Grade Descriptor

The major implication is that if the value of a block depends on what it is blended
with, it may never be possible to specify the absolute true value of any block. In these
circumstances, the best that can be said is that the recovery calculated for a block in
isolation will result in a value that reasonably accurately ranks the block relative to
all other blocks available for mining at the same time.2 If the cut-off is derived by an
optimisation process that selects the best cut-off value to use based on the mining
inventory available at different cut-offs, such a value will still be satisfactory for ranking
material and prioritising it for treatment.
If the cut-off is based on a break-even calculation, there is no way of telling whether
– and if so, to what extent – a non-linear recovery relationship may under or overstate
the value of marginal material relative to the break-even value. Similarly, if strategy
optimisation software used requires values to be assigned to mining blocks before
the optimisation is conducted (rather than blend value being calculated within the
optimisation process), the result may be suboptimal.
It is therefore important to ascertain to which grade the metallurgical recovery
relationship is to be applied. It tends to be to average feed grade, though, due to the
practicalities alluded to before, an ‘average grade’ relationship may need to be applied
to individual blocks. The author has encountered a situation with a highly non-linear
recovery versus grade relationship, where it was assumed that small ore blocks,
equivalent in size to geological model blocks, would effectively be batch processed
sequentially rather than blended. It was therefore appropriate to apply the recovery
relationship to individual geological model blocks. The recovered metal from the ore
feed over a period of time could be reliably estimated by totalling the recovered metal
individually for each of the geological model blocks in each time period, rather than
by applying the recovery relationships to the average feed grade.

GENERATING AN EQUIVALENCE FORMULA

Despite the increasing capability of geological modelling software to generate an NSR
for each block, in some cases the metallurgical processing may be too complex for this.
Therefore it is still common to derive an equivalence formula by which equivalent
grades can be generated for all blocks of rock. For in situ and recoverable grades, the
multipliers are easily derived from forecast prices and recoveries as appropriate (though,
as indicated in complex plants, recovery may not be determined by a simple relationship
to apply to grades). When all payabilities and costs are clearly driven by individual
products only, simple extensions applying them to the recoverable calculations for each
metal separately may be satisfactory for deriving payable values. But when some costs
are common to two or more products – for example, treatment and freight charges for
base metal concentrates with two or more payable products – more complex calculations
may be necessary.

By accounting-style cost allocations

If more complex calculations are required, it is common to assume an average feed
grade and derive the NSR. Incurred costs are deducted to derive net values, and hence

2. This is a qualitative, experience-based judgement by the author and is not supported by any rigorous
mathematical or statistical analysis.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 196
 CHAPTER 9 | Specifying the Grade Descriptor

value multipliers or weighting factors, for each product component. Common costs are
distributed among the various products. This is typically done by an arbitrary process;
for example, in proportion to revenues generated, or by attributing all costs to the
major product and none to the other products. The latter is common for base metal
concentrates containing precious metals – although the precious metals may contribute
significant value, they are an insignificant component of the concentrate weight, which
is what drives transport costs. This raises questions about whether some elements are
being unduly favoured or penalised by this process.
If the metal ratios in the orebody are constant through its volume and over a range
of grades, these arguments are academic, as the constancy of the metal ratios will
automatically compensate for any cost misallocation. If the ratios are not constant,
as is more often the case, it is possible to misclassify blocks of rock to a severe extent
if the multipliers, and hence relative weightings of ore constituents, are incorrect.
Unfortunately, at operations where this process is employed, the discussion about
how to allocate common costs frequently remains focused on the resulting multipliers
themselves or on the processes used. There is little point in arguing about the relative
merits of alternative arbitrary processes – if they are arbitrary, none has any more
rational justification than any other. Such arguments are futile without an objective way
of assessing whether a set of multipliers derived using one set of assumptions about
common cost allocation is better or worse than another set with different assumptions.
The issue is not the values of the multipliers or the processes used, but the effect that they
have on the resulting orebody. Notionally, there is a rational procedure to be followed
– generate block models with block values derived using a number of multiplier sets
derived with cost allocations, and see how the orebody (or family of orebodies) varies
between cases. If variations with different multiplier sets are non-existent or small, the
method of cost allocation to derive the multipliers is not an issue, and any set may be
used if derived using reasonable processes, even if they are arbitrary.
If the differences between orebodies derived with multiplier sets are significant, it
must be determined which set is best. It would then theoretically be necessary to do
mine designs, schedules and financial evaluations for alternative orebodies defined by
the multiplier sets, to ascertain which gives the best result. This is a long and involved
process, and cannot be done every time one of the many parameters that drive the value
of an orebody changes; however, it is the logical sequence to be followed for a rigorous
answer using this accounting methodology to allocate common costs.

By multiple linear regression

There is another way. As noted, the typical value equivalence formula, which is a linear
function of the metal grades describing a value plane in multidimensional space, is:

Value = a × Metal A grade + b × Metal B grade + c × Metal C grade + …

In the same way that we require two points to define a line in two dimensions, and
three points to define a plane in three dimensions, so we require n points to define a
plane in n dimensions. If we were to investigate the relationship between two variables,
we would obtain many more than two points to perform a statistical regression and
define the straight-line relationship (the line of best fit) between variables. If we have
many more than n points in an n-dimensional problem, we can perform a multiple linear

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 197
 CHAPTER 9 | Specifying the Grade Descriptor

regression to define the plane of best fit. In a two-dimensional analysis, a simple linear
regression will generate the best fit values for the y axis intercept and the coefficient of
the independent variable x. In the same way, a multiple linear regression will calculate
the best fit values of the dependent variable axis intercept and the coefficients of the
independent variables in n dimensions. If we have a number of sets of ore grades and
their corresponding dollar values, a multiple linear regression can be performed, and
the best fit values of the multipliers for the constituent grades will be determined. There
is no need to make arbitrary allocations of costs. The maths automatically derives the
best relationship between value and metal grades.
Mines usually have a model for deriving the final net revenue from the mine head grade,
accounting for a large number of metallurgical performance and product sales terms
parameters, which have an effect on the ultimate value received by the mine for a parcel
of ore treated. If the processes are too complex to derive an NSR-type grade value for
each geological model block, a data set of grades for a representative sample of blocks
can be extracted from the block model. Assuming that each sample represents the grades
and other characteristics of the mill feed, the physical, economic and financial model can
calculate the net revenue generated in each case. Each combination of input grades and
the resulting value will be the coordinates of a point in multidimensional space. With
a number of data points greater than the number of variables, and covering the full
range of individual metal grades and ratios of grades (to avoid problems of collinearity),
the equation of the multidimensional plane can then be determined by multiple linear
regression. Software such as Microsoft Excel™ provides a simple way of doing this.
If all the points lie close to a multidimensional plane, the linear equivalence formula
model is as accurate as can be. The multipliers derived by regression will be the
theoretically correct multipliers, in that the value predicted by the simple equivalence
formula model will be the same as that resulting from the more complex economic and
financial model. The coefficient of correlation (R-squared) calculated by the regression
process will help to identify whether this is so. The closer R-squared is to 1, the better
the linear equivalence formula fits the data points. This will be the case where the net
payable grades are linear functions of head grade.
As well as generating linear multipliers to apply to each metal grade to derive an
equivalent value, the multiple linear regression process will generate a constant term. It is
unusual for equivalent grade formulas to include a constant term, but there is no reason
why they should not. There are two options regarding the constant term. Since it will be
applied to all blocks, it will not affect the ranking of blocks relative to each other. If simple
ranking of blocks is the purpose of the grade descriptor, the constant term may be ignored,
and only the multipliers for each metal applied to the metal grades. If it is desired that a
dollar value grade descriptor should represent the true value as accurately as possible, the
constant term should be included in the equivalence formula. This will need to be the case
if, for example, sets of multipliers need to be derived for different rock types and different
constants are derived from the regressions. Alternatively, the regression can be conducted
in such a way as to force the constant term to be zero. Multipliers for metal grades will
differ from those obtained with a constant term, and the correlation will not be as good.
Consideration of the resulting change in the correlation coefficient or of values derived
for actual grades, using a process similar to that illustrated in Figures 6.14 and 7.1, will
indicate whether eliminating the constant term is appropriate.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 198
 CHAPTER 9 | Specifying the Grade Descriptor

Where the net payable grades are not linear functions of head grade, or for some other
reason the value versus grades relationship is non-linear, one could potentially identify
a more complex but accurate mathematical formula relating value and the metal grades.
Alternatively, if the main purpose of the equivalent value is to define blocks above and
below cut-off, a linear formula chosen to be accurate in the region of the cut-off could be
a good solution.3

DOLLAR VALUES VERSUS METAL EQUIVALENTS

Many polymetallic mines use a dollar value as a grade descriptor. Common practice at
such operations is to calculate the cost per tonne and then use that unit cost as the cut-
off. This has the appearance of both rigour and simplicity, and is effectively applying
a break-even cut-off.4 As noted in the preceding subsection, there is no guarantee that
the dollar value assigned to a block accurately represents its true value. The application
of break-even cut-offs to dollar values of material can potentially lead to under or over-
stated cut-offs.
When price predictions change (often when new forecasts are issued annually as part
of the company’s planning cycle), typical practice is to derive and apply new multipliers
to block grades to produce new dollar values, and then re-create ore boundaries by
applying the existing cut-off value to the new block dollar values. If break-even cut-offs
are being used, this process is logical, but again there is no guarantee that the new dollar
values accurately represent the true value, nor that the errors, if any, are consistent
between the old and new values.
However, there are problems with this approach when optimum rather than break-even
cut-offs are used. Optimum cut-offs are intended to achieve some particular optimised
goal, such as maximising NPV, whereas break-even cut-offs merely ensure that every
tonne pays for itself. As indicated in previous chapters, NPV-maximising cut-offs are
typically higher than break-even cut-offs. Conventional wisdom regarding changes in
cut-offs when prices change is based on the behaviour of break-evens; changes in cut-
offs are expected to be inversely proportional to changes in prices. Yet the outcomes of
optimisation studies show that, for a single metal deposit using the metal grade as the
grade descriptor, the NPV-maximising optimum cut-off will change by a significantly
smaller proportion than the inverse proportion of the price change. Figure 9.3 illustrates
the problems that can arise when using dollar values.
The figure shows value versus cut-off curves for two price scenarios, the higher prices
being 20 per cent higher than the lower prices. There are two cut-off axes for dollar
values derived from each price set. Since all prices are assumed to increase by 20 per

3. The regression would be conducted using only the grades of sample blocks, for which the true values lie within
a small range above and below the cut-off value.
4. There is an operation with a number of polymetallic orebodies that are mined by separate mines feeding a
single concentrator. Staff in a centralised section are responsible for deriving NSR-type grade descriptors for the
blocks in the block models of all orebodies. These are derived to represent the net value of the ore at the point
of delivery to surface at each mine. Junior technical staff are tasked with determining the cost per tonne at their
mines. These costs are then applied as break-even cut-offs to the NSR-type payable value grades. By virtue of
these procedures, the company effectively allocates all responsibility for defining what is and is not its ore first to
the market, which sets prices and hence ore values; and second, to its junior technical staff, who then determine
what will be classified as ore. The company in question is not alone in doing this.

Cut-off Grades and Optimising the Strategic Mine Plan | Spectrum Series 20 199