Efficiency and Productivity Improvements at a

Platinum Concentrator
Development of a Management Tool to Measure and Monitor
OEE and Process Pain
Master of Science Thesis in Quality and Operations Management, Mechanical Engineering

ANTON KULLH
JOSEFINE LMEGRAN
Department of Product and Production Development
CHALMERS UNIVERS ITY OF TECHNOLOGY
Gothenburg, Sweden, 2013

Efficiency and Productivity Improvements at a

Platinum Concentrator
Development of a Management Tool to Measure and Monitor
OEE and Process Pain

ANTON KULLH
JOSEFINE LMEGRAN

Department of Product and Production Development

CHALMERS UNIVERSITY OF TECHNOLOGY
Gothenburg, Sweden, 2013

Efficiency and Productivity Improvements at a Platinum Concentrator

Development of a Management Tool to Measure and Monitor OEE and Process Pain
ANTON KULLH (B. Sc. Mechanical Engineering)
JOSEFINE LMEGRAN (B. Sc. Mechanical Engineering)
ANTON KULLH, JOSEFINE LMEGRAN, 2013.
Published and distributed by
Chalmers University of Technology
Department of Product and Production Development
SE-412 96 Gothenburg
Sweden
Telephone + 46 (0)31-772 1000
In collaboration with
University of Cape Town
Anglo American Platinum
Printed by
Reproservice
Gothenburg, Sweden 2013
Cover photo:
Parts of the comminution process at Mogalakwena North Concentrator, Limpopo,
South Africa (Josefine lmegran, 2012)
Efficiency and Productivity Improvements at a Platinum Concentrator
Development of a Management Tool to Measure and Monitor OEE and Process Pain

II

ANTON KULLH
JOSEFINE LMEGRAN

Abstract
The current low platinum price has put
high pressure on the industry and forced
companies to introduce cost cutting efforts
as well as productivity increasing actions.
Increasing the productivity can be done
either by increasing the output or
decreasing the amount of consumed
resources. This project has focused on the
latter. There are several productivity
increasing methods, such as Total
Productivity Maintenance (TPM) and Lean,
to utilise. In the mining industry these
methods have not been used to the same
extent as in, for instance, the automobile
industry to improve productivity. Existing
research in mining mostly deals with the
technical aspects of the process, such as
optimising single units.

Concentrator (MNC), Limpopo, South

Africa. MNC is ranked as the biggest
single-stream platinum concentrator in
South Africa and one of the largest
facilities of its type in the world (Mining
Weekly, 2008).
The Masters thesis writers have developed
a method for calculating Overall
Equipment Effectiveness (OEE) in a
comminution
process.
The
method
incorporates a method to calculate quality,
which is a parameter that has previously
not been defined for a comminution
process.
A method called Pain analysis has been
developed by the Masters thesis writers to
display duration and frequency of the
reasons that cause the stops in the process.
This new way of displaying stop data has
been appreciated by its users and has
received positive response from the
organisation.

This project has three distinct phases and

will use the incorporated tools of TPM and
Lean to, firstly define a calculation model
of equipment performance metrics for a
single stream comminution process.
Secondly, a tool to perform real time
calculations of defined metrics will be
developed. Thirdly, a method for using the
tool output in the organisation in a value
creating way, with primary focus on finding
root-causes to productivity limiting issues,
will be designed.

The developed Overall Productivity Tool

(OPT) is at this stage a fully functional
software used by MNC in daily work. The
methods and day-to-day tools developed in
this Masters thesis project will be
incorporated in new software developed by
Anglo American Platinum. The software is
to be implemented throughout the
organisation.

The project is a collaboration between

Chalmers University of Technology,
Gothenburg, Sweden, and the University of
Cape Town, South Africa. The project
sponsor is Anglo American Platinum and
the plant where the project has been
conducted
is
Mogalakwena
North

Answers to the research questions are

provided at the end of the report as well as
recommendations for the operations at
Mogalakwena North Concentrator.
efficiency, quality, availability, performance,
overall utilization, process pain, platinum,
concentrator, Overall Productivity Tool
(OPT)

Keywords: Overall Equipment

Effectiveness (OEE), productivity,

III

IV

ACKNOWLEDGEMENTS
A number of people have been highly
valuable to us in this Masters thesis project.
We would like to mention them here and
send them our deepest gratitude for
assisting us throughout the project.

every day. We would also like to thank the

metallurgists on site, Albert Blom, Sithi
Mazibuko, Felix Mokoele, Herman Kemp
and Howard Saffy, who always have been
patient with our questions and given their
input to our work.

Great thanks to our examiner Prof. Magnus

Evertsson
(Chalmers
University
of
Technology) as well as to our supervisors
Dr. Erik Hulthn (Chalmers University of
Technology) and Dr. Aubrey Mainza
(University of Cape Town), who together
initiated this project. Without their early
vision of creating a Masters thesis project
focused on increasing the productivity at
Mogalakwena North Concentrator (MNC),
this project would not have been born.

We also want to send our gratitude to the

brilliant PhD students and at Chalmers
Rock Processing Systems Johannes Quist
and Gauti Asbjrnsson for their invaluable
input and encouragement in all sorts of
matters throughout the entire project.
We extend a huge thanks to Barbara
Andersen at the University of Cape Town
who arranged everything we could possibly
need during our entire stay in South Africa.
Without her help we would have been
forced to put more effort into booking
flights than typing code.

We would also like to thank Neville Plint

and Gary Humphries at Anglo American
Platinum, who always have been just an
email away and provided guidance and
feedback throughout the project.

Finally, we would like to thank all

employees
at
Mogalakwena
North
Concentrator for supporting our work,
patiently allowing us to ask questions and
providing useful answers.

Thanks to Senior Concentrator Manager

Barry Davis for his positive attitude to this
project from day one, for authorising us to
get access to the plant as well as his office
and providing us feedback throughout the
project. We give many thanks also to Plant
Manager Ellie Moshoane and Technical
Manager Dane Gavin who have acted as
supervisors on site and supported our work
and provided reflections and feedback

Last but not least, we would like to thank

our families and friends for all
encouragement and support during this
exciting journey from the 5th floor of the
Mechanical Engineering building at
Chalmers to one of the worlds largest
platinum concentrators.

TABLE OF CONTENTS
.

Chapter One - Introduction ....................................................................................................................... 1

1.1 Project Introduction.............................................................................................................................. 2
1.2 Company Introduction ......................................................................................................................... 2
1.3 Background............................................................................................................................................ 2
1.4 Earlier Efforts In This Area ................................................................................................................ 3
1.5 Project Scope ......................................................................................................................................... 3
1.6 Research Questions .............................................................................................................................. 4
1.7 Delimitations ......................................................................................................................................... 4
Chapter Two Theoretical Framework ................................................................................................... 5
2.1 Productivity............................................................................................................................................ 6
2.2 Total Productive Maintenance (TPM) ............................................................................................... 6
2.3 Overall Equipment Effectiveness (OEE) .......................................................................................... 6
2.4 Continuous improvement - Kaizen () ....................................................................................... 11
2.5 Performance Measurements .............................................................................................................. 12
2.6 Value Stream Mapping (VSM) ......................................................................................................... 13
2.7 RACI Matrix ....................................................................................................................................... 14
2.8 Platinum ............................................................................................................................................... 15
2.9 Extraction of Platinum-Group Metals (PGMs) ............................................................................. 16
Chapter Three - Methodology ................................................................................................................. 17
3.1 Research Strategy ............................................................................................................................... 18
3.2 Research Approach ............................................................................................................................ 18
3.3 Theory .................................................................................................................................................. 18
3.4 Reliability ............................................................................................................................................. 20
3.5 Validity ................................................................................................................................................. 20
3.6 Ethical aspects ..................................................................................................................................... 21
Chapter Four - Data ................................................................................................................................. 23
4.1 Quantitative Data ............................................................................................................................... 24
4.2 Qualitative Data .................................................................................................................................. 25
4.3 Empirical Data .................................................................................................................................... 25
4.4 Current Value Stream Maps .............................................................................................................. 29
VI

Chapter Five - Results .............................................................................................................................. 31

5.1 Calculation Model............................................................................................................................... 32
Final OEE Calculation ......................................................................................................................... 33
Availability ............................................................................................................................................ 35
Utilised Uptime ..................................................................................................................................... 36
MTBF & MTTR.................................................................................................................................... 36
Pain Analysis ......................................................................................................................................... 36
5.2 The Overall Productivity Tool (OPT) .............................................................................................. 39
5.3 OPT Method ........................................................................................................................................ 43
User Expertise ....................................................................................................................................... 43
Five WHYs ............................................................................................................................................ 43
OPT Guidelines..................................................................................................................................... 43
OPT Meeting Procedure ...................................................................................................................... 44
OPT Action List .................................................................................................................................... 44
5.4 OEE For a General Single Stream Process ..................................................................................... 45
5.5 Crusher and Mill Stops Reporting Procedure ................................................................................. 45
Chapter Six Discussion & Conclusions ............................................................................................... 47
6.1 Calculation Model............................................................................................................................... 48
6.2 The Overall Productivity Tool (OPT) .............................................................................................. 57
6.3 OPT Method ........................................................................................................................................ 58
6.4 OEE for a General Single Stream Process ...................................................................................... 59
6.5 Crusher and Mill Stop Reporting Procedure ................................................................................... 60
6.6 Research Questions and Answers ..................................................................................................... 60
6.7 Observations ........................................................................................................................................ 62
6.8 Recommendations .............................................................................................................................. 64
6.9 Future Research .................................................................................................................................. 64
References ................................................................................................................................................. 66

VII

APPENDIX
APPENDIX I
Process Maps
APPENDIX II
Lynxx Cameras Map
APPENDIX III
Job Card Inspections and Maintenance Map
APPENDIX IV
OPT Print Screens
APPENDIX V
OPT Manual
OPT Meeting Procedure
OPT Action List
Organisational Chart

VIII

CHAPTER ONE

INTRODUCTION
This chapter will introduce the Masters Thesis project by presenting a background of the
problem, a company introduction, earlier efforts in the area, the project aim and finally the
research questions

CHAPTER 1 - Introduction
refined platinum produced by their own
mines amounted to about 44 tons in 2011.

1.1 PROJECT INTRODUCTION

The project is a collaboration between
Chalmers University of Technology,
Gothenburg, Sweden, and the University of
Cape Town, South Africa. The sponsor of
the project is Anglo American Platinum
and the plant where the project has been
conducted
is
Mogalakwena
North
Concentrator (MNC), Limpopo, South
Africa. MNC is ranked as the largest singlestream platinum concentrator in South
Africa and one of the largest facilities of its
type in the world (Mining Weekly, 2008).

To operate more effectively and efficiently

Anglo Platinum recently accomplished a
thorough reconstruction and they now
operate nine individual mines around
South Africa. One of them is Mogalakwena
Mine, which is situated 30 kilometres
northwest of Mokopane in the Limpopo
province and operates under a mining right
covering a total area of 137 square
kilometres (Anglo American, 2012).
Mogalakwena Mine provides ore to
Mogalakwena South Concentrator (MSC)
and Mogalakwena North Concentrator
(MNC). MNC is ranked as the largest
single-stream platinum concentrator in
South Africa and one of the largest
facilities of its type in the world and is the
plant where this Masters thesis project was
conducted (Mining Weekly, 2008).

The examiner of this thesis is Prof. Magnus

Evertsson
(Chalmers
University
of
Technology). Dr. Erik Hulthn (Chalmers
University of Technology) and Dr. Aubrey
Mainza (University of Cape Town) have
acted as supervisors. The writers of this
Masters thesis are B.Sc. Anton Kullh
(Chalmers University of Technology) and
B.Sc. Josefine lmegran (Chalmers
University of Technology).

1.3 BACKGROUND
South Africa accounts for nearly 80% of
the global platinum production, which
makes the platinum price highly influenced
by the economy of the country.

This project was initiated by the supervisors

who wanted to investigate how to increase
productivity in a comminution process. The
project scope has thereafter evolved
throughout the project as the writers have
gained more knowledge in the area of
research. It is important to have a valid
measure of productivity in order to be able
to increase it and to be able to judge if your
efforts are contributing to productivity
improvements. The final scope is set as
three distinct phases and can be viewed
under 1.5 Project Scope.

During the last five years there has been a

large decrease in the platinum spot price. In
March 2008 the price peaked at 2273
USD/oz t (Kitco, 2012), which can be
compared to the spot price at the start of
this project (mid-August 2012); 1485
USD/oz t (Kitco, 2012).
The price drop has several explanations.
Firstly, the price spiked in 2008 due to a
prospected supply shortage. Secondly, the
sector has experienced wage inflation in
excess of the general inflation. Thirdly, the
strengthening of the rand has decreased the
gap between dollar-denominated sales and
rand-based costs. As a result of this, some
high cost mines have had trouble running

1.2 COMPANY INTRODUCTION

Anglo American Platinum Limited is a
South African company which holds about
40% of the worlds newly mined platinum,
making them the world leading primary
producer of platinum. The equivalent of

CHAPTER 1 - Introduction
profitable operations during recent years.
(Mail & Guardian, 2011)

can be monitored. Further, it will allow for

the results of performed productivity
increasing actions to be analysed and
evaluated. This is in agreement with the
author of the book TPM Vgen till stndiga
frbttringar rjan Ljungberg:

This significant drop in price has put high

pressure on the industry and forced
companies to introduce cost cutting efforts
as well as productivity increasing actions.
This project will deal with the aspects of
productivity.

What you do not measure, you cannot control

and what you cannot control, you cannot

An increase in productivity can gain several

stakeholders and be profitable not only for
the company itself, but also for the nearby
communities, the region and the country.

improve.

(Ljungberg, 2000 p.37)

The methods TPM and Lean and their

incorporated tools will found a basis for
this Masters thesis project, which is divided
into three distinct phases. Firstly, a
calculation
model
of
equipment
performance metrics for a single stream
comminution process will be defined. This
calculation model will be aligned with the
Anglo American Equipment Performance
Metrics Standard and fully functional for a
single-stream
comminution
process.
Secondly, a tool to perform real-time
calculations of the defined metrics will be
developed. To achieve this, user friendly
software which automatically computes the
defined equipment performance metrics for
the equipment included in the project will
be developed. Thirdly, a method will be
designed for using the tool output in the
organisation in a value creating way. The
primary focus will be on designing a
method for finding root-causes to
productivity limiting issues.

1.4 EARLIER EFFORTS IN

THIS AREA
In the mining industry TPM and Lean have
not been used to the same extent to
improve productivity as in, for instance, the
automobile industry. There is some
research
literature
on
productivity
increasing efforts in comminution processes
but it mostly deals with the technical
aspects of the process, such as optimising a
single unit, i.e. a ball mill. This approach
can result in an unintended suboptimisation instead of an optimisation of
the entire process chain. The research has a
gap concerning the usage of the above
mentioned methodologies to improve the
total productivity of the comminution
process. Therefore, this thesis aims to help
fill the gap and explain how to work with
productivity increasing efforts throughout
the entire process, instead of only in single
units.

1.5 PROJECT SCOPE

To increase productivity, the process needs
to be completely comprehended and
controlled. It is highly important to have a
valid and accurate method of calculating
equipment performance metrics, so that it

CHAPTER 1 - Introduction

Define a calculation
model for OEE &
other equipment
performance metrics
in a single stream
process

Develop a tool that

calculates OEE &
other equipment
performance metrics
in real time

Design a method
describing how to use
the tool output in the
organisation with
primary focus on
finding root causes

Figure 1 The three phases of this Masters thesis project

1.6 RESEARCH QUESTIONS

1.7 DELIMITATIONS

The following research questions have been

formulated for this Masters thesis. The
research questions cover the areas of
Overall Equipment Effectiveness (OEE),
Performance indicators, productivity and
SHE (Safety, Hygiene, Environment).

To fulfil the purpose of this Masters thesis,

the project concentrated on the initial part
of the comminution circuit of a minerals
processing plant, which at MNC includes
the primary gyratory crushing, the
secondary
crushing,
HPGR-crushing,
classifiers, feeders, and conveyors.

1. How can a method be developed to

define and rank process units
critical to productivity in a
comminution process?

The research was limited to this area for

two reasons; it is a good idea to start the
implementation in a small scale and it
would have been too time consuming for
the Masters thesis to include a larger
section of the plant.

2. How should OEE numbers be

calculated in a single-stream
comminution process?

The decision to limit the project to a subset of the plant is supported by Idhammar
(2010) who argues that implementing OEE
in a part of the plant will facilitate an
implementation throughout the plant at a
later stage. Idhammar also states that an
early pilot can eliminate issues and provide
useful training and experience for staff
members.

3. Which factors in the process chain

are more critical to productivity
according to the OEE philosophy?
4. How can OEE be used as a
performance measure of equipment
and process performance?
5. How can OEE help to improve
SHE (Safety, Hygiene,
Environment)?

The developed tool will be fully functional,

however not completely integrated into the
Scada and PI-system at the plant.

6. How can measuring OEE help to

improve productivity?

CHAPTER TWO

THEORETICAL FRAMEWORK
This chapter will present the theoretical framework used for the Masters thesis project.
The theoretical framework shall act to facilitate understanding for readers with no or little
former experience of the industry and the methodologies used in the project.
However, the targeted readers of the report are assumed to already possess a basic knowledge
in the area. Some basic definitions are therefore left out.

CHAPTER 2 Theoretical Framework

maintenance
team
do
everything
(Nakajima, 1989). Ahuja (2011) recognises
that for a long time companies have seen
maintenance as a static support activity
instead of as a key component for revenue
generation, which he claims it is. TPM
demands
collaboration
between
all
functions in an organisation even though
the most extensive is found between
production and maintenance where the
greatest synergies can be created. TPM is
aimed at changing peoples mind-set
towards maintenance rather than providing
a perfect tool that will solve all problems
(Ahuja, 2011). According to Smith and
Hawkins (2004) the primary focus of TPM
should be to reduce and eliminate the
effectiveness losses, since this is where the
biggest gains can be achieved.

2.1 PRODUCTIVITY
A general definition is that productivity is
the amount of output per unit of consumed
resources or total costs incurred. Thus,
productivity is defined as how efficiently
the resources are being utilised in the
production of goods or services. Increasing
the productivity can hence be done by
increasing the output, in terms of volume
and quality, or decreasing the amount of
consumed resources with a constant or
increased output. (Prokopenko, 1992)

Productivity =

Output
Input

Equation 1

The essence of productivity improvement is

working more intelligently, not working
harder. There are several methods to help
increase productivity. Two of the most
well-known
are
Total
Productivity
Maintenance (TPM) and Lean. These
methods, and some of their incorporated
tools, will be described in more detail in the
following sections.

2.3 OVERALL EQUIPMENT

EFFECTIVENESS (OEE)
There are many ways of measuring how
well-functioning a certain unit is. One of
the most common methods in the industry
is to use availability as a metric. Other
common methods are MWT (Mean
Waiting Time), MTTR (Mean Time To
Repair) and MTBF (Mean Time Between
Failure). However, these parameters, and
many other common parameters, do not
give a comprehensive view of units and
equipment if displayed alone. Included in
TPM, there is a metric called OEE (Overall
Equipment Effectiveness), which gives a
more inclusive view of the value added by
the unit, compared to other metrics.
(Ljungberg, 2000)

2.2 TOTAL PRODUCTIVE

MAINTENANCE (TPM)
Total Productive Maintenance (TPM) was
first coined in Japan in the beginning of the
nineteen seventies as a way of increasing
the availability of machines and equipment
by better utilising the maintenance
resources. Simply put, it is about keeping
machines and equipment in good condition
without interfering with daily production.
The methodology was created in order to
support the Japanese effort to implement
Lean in most of their industries. The
fundamental idea of TPM is to involve the
operators in the maintenance and make
sure that they conduct most of the day-today activities instead of having the

It is important to identify the factors that

limit the process from receiving higher

CHAPTER 2 Theoretical Framework

OEE Availability x Performance x Quality

Equation 2

Gross Operating Time

Planned Production Time

Planned Production Time Unplanned Downtime
=
Planned Production Time

Equation 3

Availability

Performance

Quality

Net Operating Time

Ideal Cycle Time

Operating Time
Gross Operating Time
Total Pieces

Valuable Operating Time

Good Pieces

Net Operating Time

Total Pieces

effectiveness and to understand how the

process is performing. OEE is a method
that can help to do this by giving a better
understanding of how well a process is
performing and by identifying what the
limiting factors are. (Hansen, 2001)
An OEE number of 100% corresponds to a
unit which is performing at its maximum
capacity always running, always at the
optimal speed and producing perfect
quality.
It is essential to note that OEE is more
than just one number; it is, in total, four,
which are all individually useful. The OEE
measurement combines the availability of
the machine, the performance rate and the
quality rate in one equation (M. Maran, et
al., 2012)

Equation 4

Equation 5

CHAPTER 2 Theoretical Framework

2.3.1 GENERAL OEE DEFINITION

There is no established standard for how to
calculate OEE for a single stream
comminution process. A version frequently
used in the manufacturing industry, derived
from (Nakajima, 1989), is presented in
Equation 2, and explained in more detail in
Equation 3, 4, 5 and Figure 2.

Quality is the ratio between Valuable

Operating Time and Quality Losses, where
Valuable Operating Time is Net Operating
Time minus Quality Losses, or the ratio
between Good Pieces and Total Pieces.
The OEE calculation parameters are
described in Figure 2.

Availability is the ratio between Gross

Operating Time and Planned Production
Time, where Gross Operating Time is
Planned
Production
Time
minus
Unplanned Downtime. For definitions of
parameters, see Figure 2.
Performance is the ratio between Net
Operating Time and Gross Operating Time,
where Net Operating Time is Gross
Operating Time minus Speed Losses, or the
ratio between the actual speed and nominal,
budgeted, or target cycle time.

Figure: Freely after Method and a system for improving the operability of a production plant

Figure 2 Definitions of the general OEE parameters

CHAPTER 2 Theoretical Framework

2.3.2 ANGLO AMERICAN

EQUIPMENT PERFORMANCE
METRICS OEE DEFINITION

represents the ratio between Primary

production (P200) time and Total time
(T000), where Primary production is
defined as Time equipment is utilised for
production and Total time is defined as
The total possible hours available.
(Anglo American Equipment Performance
Metrics,
2012)
For definitions
of
parameters, see Figure 3.

The
Anglo
American
Equipment
Performance Metrics (a company internal
standard) offers another way of calculating
the OEE, which will be presented in this
section.
The metric corresponding to the general
OEE calculations availability is titled
Overall Utilisation (see Equation 6). This

Figure: Freely after Anglo American Equipment Performance Metrics Time Model

Figure 3 Time definitions by Anglo American Equipment Performance Metrics

CHAPTER 2 Theoretical Framework

The Performance metric (see Equation 7) is
stated as The portion of the OEE Metric
which represents the production rate at
which the operation runs as a percentage of
its targeted rate. It is calculated as the
ratio between Actual Production Rate and
Target Production Rate, where Actual
Production Rate is the ratio between
Actual Production Achieved and Primary
Production (P200), whereas the Target
Production Rate is defined as an input.
(Anglo American Equipment Performance
Metrics, 2012)

defined the world-class level at 85 % and it

is composed of an Availability of 90%,
Performance of 95% and Quality of 99%,
which creates Equation 9.

The Quality is stated as The portion of the

OEE Metric which represents the Quality
achieved at an operation as a percentage of
its targeted Quality and is calculated as
the ratio between Actual Quality and
Target Quality (see Equation 8). Both the
numerator and the denominator are stated
as inputs and no calculation method for
them is provided in the Anglo American
Equipment Performance Metrics. (Anglo
American Equipment Performance Metrics,
2012)

2.3.3 OEE BENCHMARK

It is often hard to measure the financial

benefits of proposed improvement projects
and it is easy to oversee important projects
and instead prioritise average projects.
Bottlenecks are what prevent a process

According to Hansen (2001) very few

companies calculate OEE or use it to
maintain and set new priorities. He has
defined different levels of OEE for
companies to aim for, which can be
described as follows:

According to Hansen (2001) a batch type

process should have a world-class OEE
of >85 %, for continuous discrete processes
the OEE value should be >90 % and for
continuous on stream processes the OEE
value should be 95 % or better.

2.3.4 OEE ECONOMICS

According to M. Lesshammar (1999) most

equipments OEE ranges from 40-60 %,
whereas the world-class level is said to be
85 %. Smith and Hawkins (2004) have

Overall Utilisation

< 65 % Unacceptable. Money is

constantly lost. Take action!
65-75 % OK, only if improving trends
can be shown over a quarterly basis.
75-85 % Pretty good. But keep working
towards the world-class level.

Primary Production P200

Total time T 000

Actual Production Rate

Target Production Rate

Actual Production Achieved / Primary Production
=
Target Production Rate

Equation 6

Performance

Quality

Actual Quality
Target Quality

Equation 7

Equation 8

90 % Equipment Availability x 95 % Performance Efficiency

x 99 % Rate of Quality = 84.6 % OEE

10

Equation 9

CHAPTER 2 Theoretical Framework

throughput and limits a plant from
becoming effective; therefore, bottlenecks
should be the first place where OEE is
applied. In order to prioritise the OEE
improvement projects relative to the
average ones, it is important to be able to
show the financial gains. (Hansen, 2001)

Continuous improvement is the process of

making incremental improvements, no
matter how small, to achieve the lean goal
of eliminating all waste that does not add
any
value
but
only
adds
cost.
Kaizen teaches employees skills to work
more effectively in smaller groups, solving
problems, documenting and improving
processes, collecting and analysing data,
and also to self-manage within the peer
group. (Liker and Convis, 2012)

Hansen (2001) has shown that there is a

link between OEE and critical financial
ratios and that a company that understands
and applies OEE improvement projects
will harvest dividends year after year, since
OEE improvement projects work to
eliminate the root causes of problems.

The concepts of Kaizen started in the early

days of Toyota and included the now
famous concepts of just-in-time (JIT),
process flow and quality improvements.

According to Hansen (2001) a 10 %

increase of OEE from 60 % to 66 % will
give:

Kaizen can be divided into six main steps

which became the basis for the Toyota
Kaizen course developed by the company
in the 1970s. (Kato and Smalley, 2011)

21 % increase of Return on assets

(ROA)
10 % increase of capacity
21 % improvement of the operating
income

1. Discover Improvement Potential

2. Analyse the Current Methods
3. Generate Original Ideas

He also states that starting on a low OEE,

rather than on a high OEE, makes it easier
to find opportunities to improve.

4. Develop an Implementation Plan

5. Implement the Plan

Ahlmann (2002) discusses the financial

implications of an increased OEE from 60 %
to 80 % in Swedish industry and argues that
it shows a 20 % economic improvement.

6. Evaluate the New Method

It has been understood that, in realtity,
continuous improvements are impossible
since some parts of the process sometimes
need to be operated in the same way as the
day before. Everything cannot be changed
to the better every day. Continuous
improvement is a vision, a dream, which no
company can totally master. (Liker and
Franz, 2011)

2.4 CONTINUOUS IMPROVEMENT

KAIZEN ()
The term kaizen, in Japanese, literally
means change (kai) for the better (zen).
Kaizen is defined by Oxford Dictionaries as:

a Japanese business philosophy of continuous

improvement of working practices, personal
efficiency, etc..

11

CHAPTER 2 Theoretical Framework

2.4.1 THE FIVE WHYS 5 WHYS

2.5 PERFORMANCE
MEASUREMENTS

One root-cause finding technique included

in the kaizen methodology is the 5 WHYs.
It implies to ask why a problem exists five
times, going to a deeper level with each
why until the root cause of the problem is
found. The user of the technique should
take countermeasures at the deepest level
feasible of cause and at the level that will
prevent reoccurrence of the problem.
(Liker and Convis, 2012)

Using performance measures is a

procedure aimed at collecting and
reporting information regarding the
performance of an operation or individual
parts thereof. This procedure can help the
organisation to define and measure the
goals it is aiming to achieve.
In the industry, performance measures are
most often denoted as KPIs (Key
Performance Indicators). Widely used KPI
metrics are, for instance, cycle time, Mean
Time Between Failure (MTBF) and
utilisation. (Taylor Fitz-Gibbon, 1990)

To visualize the root causes, which may be

multiple, an Ishikawa diagram (also known
as a fishbone diagram) can be used in order
to create a clear picture of the current
situation and to map out the possible root
causes (see Figure 4). (Perrin, 2008)

Halachmi (2005) elaborates on the logic of

reasons in support of introducing
performance measurement as a promising
way to improve performance. This
strengthens the motives for measuring the
performance of the operations.

Picture: Real World Project Management

Figure 4 - Ishikawa diagram

12

CHAPTER 2 Theoretical Framework

One factor that will influence MTTR is the
severity of the breakdown; another factor is
the quality of the maintenance itself. A
high MTTR should be approached with
good troubleshooting methods, to quickly
identify the root cause, and the
maintenance actions should be reviewed
regularly
to
identify
improvement
opportunities. (Mahadevan, 2009)

If you do not measure results, you cannot tell

success from failure
If you cannot recognize failure, you will repeat
old mistakes and keep wasting resources.
If you cannot relate results to consumed
resources, you do not know what is the real
cost...

(Halachmi, 2005, p.504)

2.6 VALUE STREAM

MAPPING (VSM)

2.5.1 MEAN TIME BETWEEN

FAILURES (MTBF)

Value Stream Mapping (VSM) is a lean

manufacturing technique used to analyse
the flow of materials and information in a
system.

Mean Time Between Failures (MTBF) is a

measure of asset reliability. It is the average
time between one failure and another
failure for repairable assets (see Equation
10). An increasing MTBF indicates
improved asset reliability. MTBF is best
when used on asset or component level and
should be performed on critical assets and
trended over time. Low MTBF numbers
should be approached with analysis (i.e.,
root-cause failure analysis (RCFA) or
failure mode and effect analysis (FMEA))
in order to identify how the asset reliability
can be improved. (Gulati, 2009)

Value Stream Mapping involves all process

steps, both value added and non-value
added ones. In that way Value Stream
Mapping can be used as a visual tool to
help identify the hidden waste and sources
of waste. Preferably, a current state map
should be drawn to document how things
actually proceed in the process. Thereafter,
a future state map should be developed to
shape a lean process which has eliminated
root causes of waste.
Rich et al. (2006) defined the seven Value
Stream Mapping tools as:

2.5.2 MEAN TIME TO REPAIR

(MTTR)

Mean Time To Repair (MTTR) is a

measure of the average time required to
restore an assets back to working condition
(see Equation 11). In the context of
maintenance, MTTR is comprised of two
parts; the first is the identification of the
problem and the required repairs; the
second is the actual repair of the equipment.

Process Activity Mapping

Supply Chain Responsiveness Matrix
Product Variety Funnel
Quality Filter Mapping
Forrester Effect Mapping
Decision Point Analysis
Overall Structure Maps

MTBF Uptime / Number of stops

Equation 10

Equipment Downtime Time

Number of stops

Equation 11

MTTR

13

CHAPTER 2 Theoretical Framework

2.7 RACI MATRIX

In
a
large
organisation
where
responsibilities are divided between several
parties and decisions impact many core
functions,
it
is
important
that
responsibilities are clear and involve
different parties across the firm, especially
in the decision-making process. The RACI
Matrix is a method to manage decision
allocation processes (see Table 1). RACI is
an
acronym
for
Responsibility,
Accountability,
Consulting
and
Information and stands for different roles
in the decision process. (Dressler, 2004)

Responsibility (R) The role

responsible for decisions that fall under
their area of responsibility within the
organisation. This is an active and
important role in the decision making
process.

Accountability (A) This role is the

person in charge of the individual
taking on the Responsibility role and
carries the accountability for the
decision made.

Consulting (C) This role is not

accountable or responsible for the
consequences of the decision made but
shall be consulted in the decision
making process.

Information (I) This group includes

other persons in the organisations that
will be impacted by the decision and its
outcome shall be kept in the
information loop.

According to Dressler (2004), the RACI

Matrix is used by many effective
organisations to clarify ambiguous decision
areas and solve decision conflicts upfront.
This to give people a clear understanding
about their roles in regards to contributing
to an efficient decision making process.

Dressler (2004) defines the building blocks

of the RACI Matrix as follows:

Table 1 An example of how a RACI Matrix can be created. R=Responsible,

A=Accountable, C=Consulted, I=Informed.

Person A

Person B

Decision A

Decision B

Decision C

14

Person C

Person D

Person E

CHAPTER 2 Theoretical Framework

must perform reliably for long periods of
time at high temperatures under oxidising
conditions. Platinum is also used as a
catalyst in cracking petroleum products.
Currently there is a high interest in the use
of platinum as a catalyst in fuel cells and in
antipollution devices for automobiles. (The
PGM Database, 2012)

2.8 PLATINUM
Platinum was discovered in 1735 in South
America by Ulloa and can be found
occurring naturally, accompanied by small
quantities of iridium, osmium, palladium,
ruthenium and rhodium, all of which
belong to the same group of metals, the
Platinum Group Metals (PGM) (The PGM
Database, 2012). Platinum is one of the
rarest elements in the Earth's crust and has
an
average
abundance
of
approximately
platinum
5g/kg.
Other
precious metals like
gold, copper and
nickel
denote
concentration in ores
in percentages, but
platinum
denotes
this in parts per
million. Based on a Figure 5 Platinum in
typical
conversion the periodic system
rate of 25%, 14 tons
of ore are required to produce 10 grams of
platinum. (Probert, 2012)

The price of platinum has varied widely; in

the 1890s it was cheap enough to be used
to adulterate gold. But in 1920, platinum
was nearly eight times as valuable as gold.
The spot price on 2012-08-15 was
approximately 1395 USD/oz t (1 oz t =
31.103 g), which can be compared to the
gold price for the same day; 1594 USD/oz t
(Kitco, 2012).

78

Pt

A comparison between platinum and its

more well-known periodic table neighbour
gold can be viewed in Table 2.

195.08

Table 2 - A comparison between platinum and

gold.
Freely from The PGM Database (2012)

Platinum Gold

Platinum, iridium and osmium are the

densest known metals. Platinum is 11%
denser than gold and about twice the
weight of the same volume of silver or lead.
Platinum is soft, ductile and resistant to
oxidation and high temperature corrosion.
It has widespread catalytic uses. (Platinum
Today, 2012)

Chemical Symbol
Atomic number

Pt
78

Au
79

Atomic weight

195.084

196.967

Density (g/cm )
Melting point (C)

21.45
1769

19.30
1064

Vickers hardness
(MPa)
Electrical resistivity
(nohm-cm at 20C)

549

216

105

22.14

Tensile strength
(MPa)

125-240

120

In 2009, approximately 45% of the world's

platinum was used in automotive catalytic
converters, which reduce noxious emissions
from vehicles. Jewellery accounted for 39%
of demand and industrial uses accounted
for the rest. (Anglo Platinum, 2011)
Examples of its industrial uses are hightemperature electric furnaces, coating
missile nose cones, jet engine fuel nozzles
and gas-turbine blades. These components

15

CHAPTER 2 Theoretical Framework

matte that is richer than the
concentrate in platinum-group metals.

2.9 EXTRACTION OF
PLATINUM-GROUP METALS
(PGMS)

Step four is to produce a, either

through magnetic concentration or by
leaching
separate
platinum-group
elements in the converter matte, very
rich platinum-group metal concentrate
containing about 60% platinum-group
elements.

The last step is to refine this

concentrate to individual platinumgroup metals with purities in excess of
99.9%.

To get pure platinum a long process has to

be followed. The extraction of platinumgroup metals is described by Crundwell, et
al. (2011) in the following five steps:

Step one is to mine ore with a high

concentrate of platinum-group metals
while leaving rock lean in platinumgroup metals behind.
Step two is to comminute the mined ore
into powder and isolate the platinumgroup elements in the ore by creating a
flotation concentrate consisting of
nickel-copper-iron sulfides that has a
high
content
of
platinum-group
elements.

In general the concentrating and

smelting/converting are done in or near the
mining region while the refining is done in
the region or in distant refineries.
(Crundwell et al., 2011)

Step three is to smelt and convert this

concentrate to a nickel-copper sulphide

16

CHAPTER THREE

METHODOLOGY
In this chapter the methodology of the thesis will be presented and analysed in order for the
reader to better understand the approach leading up to answering the research questions and
fulfil the project scope. First, the research strategy and approach are presented. Next, details
are given regarding the data collection methods used in the thesis project. Finally, reliability,
validity and ethical aspects of the study are discussed.

17

CHAPTER 3 Methodology

3.1 RESEARCH STRATEGY

LONG-TERM SUCCESS

The research has been conducted in

collaboration with the stakeholders at
Mogalakwena North Concentrator (MNC),
in order to jointly solve the problem. This
is, according to Bryman and Bell (2011), a
typical case where action research should
be chosen as a research strategy, since
answering the research questions demands
an iterative process. Action research also
allows the researchers to keep an open
mind toward the problem at hand and to
go back and forth between theory and
practice in order to compare the results
from data collection with theory, and to
generate a thorough analysis which can be
revised.

Action research is also an appropriate

strategy when looking at the long term
success of the project, long after the
completion of the thesis. In order for the
new activities to be deeply rooted within
the organisation it is very beneficial that
key power groups within the organisation
have been a part of the improvement
process from the beginning (Nadler &
Tushman, 1997). This strategy also makes it
easier for the researchers to truly
understand the system since it demands a
lot of interaction between the different
stakeholders (Bryman & Bell, 2011).

The assumption in action research is that

the natural surroundings, in which the
problem occurs, is the best place to study
the problem. The data collection associated
with action research includes both
qualitative and quantitative methods.
Qualitative research is a research in which
words are more emphasised than quantities
during data collection, whereas quantitative
research puts more emphasis on quantified
data.

3.2 RESEARCH APPROACH

The chosen approach for this thesis is an
abductive one. When using an abductive
approach theory, empirical data and
analysis are developed simultaneously in an
iterative process (Bryman & Bell, 2011).
This research approach is suitable when
using action research since it involves the
mind-set of involving theory and empirical
information at the same time in order to
develop an analysis that will answer the
research questions. As the understanding
of the situation increases, the need for new
theory will most likely occur and this suits
the description of an abductive approach
very well.

GROUNDED THEORY
A proper methodology when conducting
action research is to use Grounded theory,
which is defined as theory that is derived
from data, systematically gathered and
analysed through the research process.
(Strauss & Corbin, 1998, pp.12). This
implies that data collection, analysis and
eventual theory are closely related in an
iterative process.
Strauss and Corbin
(1998) claim that grounded theories, since
they are drawn from data, are likely to
offer insight, enhance understanding and
act as a meaningful guide.

3.3 THEORY
Before travelling down to South Africa the
pre-study was initiated in order for the
researchers to gain the basic knowledge
needed to understand the problem, become
familiar with the available process data and
the available theories in the area. Since
Anglo American Platinum already used

18

CHAPTER 3 Methodology
OEE as a measure, the study of OEE and
therefore also TPM became a natural part
of the studies. TPM also led the path to
studies of other maintenance fields, such as
Lean maintenance, which is closely linked
to improving productivity. The studies of
OEE were broadened by studying the
economic factors associated with improving
OEE numbers, which also provided
arguments for working with OEE
improvements.

well as the daily work and methods used.

The observations also acted as indications
of what theoretical fields were interesting
for further study and in that way led to a
more focused literature study.

INTERVIEWS
A major part of the qualitative data was
collected through numerous interviews with
persons involved in the production process
as well as management. The character of
the interview was dependent on the
position of the interviewee and the type of
information
sought
(qualitative
or
quantitative). The basic approach was an
unstructured interview in order for the
interviewee to further elaborate on the
questions asked. In
an
unstructured
interview, the interviewers do not follow
a strict structure of questions, but instead
might have only one or a few questions to
answer. The interview therefore resembles
a conversation. (Bryman & Bell, 2011)

After core studies regarding the process in

general and TPM methodologies, methods
for identifying bottlenecks such as Value
stream mapping, were studied more closely
since they offered a very clear path to
evaluate the production chain and direct
the improvements to the right sections.
The remaining part of the literature study
was conducted during the empirical study
on site in cases where the empirical findings
resulted in new theoretical aspects to study.
This is also in line with the chosen
abductive approach (Bryman & Bell, 2011).

SECONDARY DATA
The major part of the data is secondary in
the sense that it stems from information
from metallurgists and specialists, and that
no long-term observations beside the
machines have been conducted. Still, the
data is in most cases measured over a long
period of time, which in a sense increases
its accuracy. This gathering of data also
limited the cost of the project, though the
information might be difficult to
understand and interpret and there is
always a risk that some important
information may be left out of the material
handed to the researchers (Bryman and
Bell, 2011). The process data used for
calculation can also be seen as secondary
data. The process data comes from the PI
database were more than hundreds of
thousands of different measurement points
are logged.

3.3.1 DATA COLLECTION

METHODS
The data needed for this thesis has mainly
been
collected
through
interviews,
observations and by using secondary data.
Most of the quantitative data was collected
through gathering data from the PI
database whereas the qualitative data was
collected
through
semi-structured
interviews, which is also the usual first
step of engagement in the action
research approach (Scheinberg, 2009).

OBSERVATIONS
In order for the researchers to acquire their
own understanding of the situation,
observations were conducted on site. This
helped with the understanding of the
specific steps in the production process as

19

CHAPTER 3 Methodology

3.4 RELIABILITY

relevant to this thesis since one research

question aims to explore how certain
changes affect the productivity of the
process. The cause and effect relations are
closely linked to the internal validity and
these
have
been
tested
through
triangulation and pattern matching. The
validity has also been ensured by using well
known methods and tools such as TPM,
OEE and Lean.

The reliability concerns the results of the

project and whether they are repeatable or
not (Bryman & Bell, 2011). Achieving high
reliability when using action research is a
difficult task since the purpose is to change
peoples mind-set and the environment in
which they act. By taking field notes
throughout the entire project and also by
keeping a diary, the researchers intention
has been to write down all important
aspects of the thesis. The research is based
on a combination of quantitative and
qualitative data compared with existing
methodologies within the area of increasing
productivity. This approach ensures that
the results are best practice from a
theoretical point of view applied in the
specific context.

3.5.3 EXTERNAL VALIDITY

The validity of the project deals with the

issue of whether the right aspects were
studied in order to answer the research
questions. Bryman and Bell (2011) propose
to measure four different aspects in order
to determine the validity of the thesis;
construct validity, internal validity, external
validity and ecological validity.

At first, the external validity can be

regarded as rather low since the study aims
at improving the specific site in question.
However, one of the research questions
(number 1, see section 1.6) deals with how
to develop a method, applicable at a
general plant, to define and rank process
units, which gives the thesis an increased
external validity. To create an external
valid thesis, proven methods have been
used and general equations and definitions
have been presented. This will help others
to interpret the content of the thesis into
other contexts and therefore increase the
external validity. The aim for the
researchers has been to develop an
aggregated method which is generic and
can be successfully implemented at similar
plants.

3.5.1 CONSTRUCT VALIDITY

3.5.4 ECOLOGICAL VALIDITY

The construct validity is regarded as high

since data triangulation was used in cases
where previous measurements were
compared with data collection on site.
Since the researchers have spent extensive
time on site, the possibility of measuring
critical aspects on several occasions was
good.

The ecological validity concerns whether

the findings are applicable to everyday life.
In this case, the findings are highly
applicable to day-to-day operations in the
process. Since the data has been collected
from
the
daily
operations
and
improvements have been done in the actual
production equipment, even though in a
small scale at first, the ecological validity
has to be considered as high. The pitfall
might be that of the Hawthorne studies,
which implies that people perform better

3.5 VALIDITY

3.5.2 INTERNAL VALIDITY

The internal validity deals with the issue of
causality. The internal validity is highly

20

CHAPTER 3 Methodology
just because they are being studied. Due to
the long time period and the size of the
production system this has to be considered
as a low risk, but should still be kept in
mind in projects of this kind.

3.6 ETHICAL ASPECTS

The researchers believe that a wellfunctioning mining industry in the region
benefits all stakeholders and that this thesis
should be a part of that development. The
aim of this thesis is ultimately to increase
the competitiveness of the company and
the region, which has been a guideline
throughout the project. Interviewees and
other participants in this research have
been informed about the purpose and have
had the option to decline. This is to ensure
that the participants do not feel harmed
due to lack of informed consent, invasion of
privacy or deception (Scheinberg, 2012).

21

22

CHAPTER FOUR

DATA
This chapter will first present how the quantitative and qualitative data has been gathered.
Next, the empirical data, including a process map, presentation of the units included in the
project as well as some remarkable facts about the comminuting units, are presented. The
quantitative and qualitative data has, together with the empirical data, served as input to the
obtained results of this Masters thesis project.

23

CHAPTER 4 - Data
To provide a picture of how the developed
tool (called the Overall Productivity Tool
(OPT)) uses tags from the PI-database to
calculate the required metrics, see Table 3.
The table displays the measures used for
process area 406 to perform the required
calculations.

4.1 QUANTITATIVE DATA

Most of the quantitative data has been
extracted from a highly technological
process data system which collects and
stores process data continuously in a
database. This is known as the PI-system
and PI-database. The PI-database consists
of hundreds of thousands of measurement
points, called tags, which can be
manipulated and combined in almost
limitless combinations.

Quantitative data has also been gathered

through interviews and observations. When
applicable, technical specifications have
been gathered from original plant drawings.

Table 3 The data extracted from the PI-database in order to calculate the OEE measures for area 406
UNIT

MEASURE

UNIT

MEASURE

HPGR Crusher
406-CR-001

Power [kW]

Conveyor
406-CV-003

Running [ON/OFF]

HPGR Screen 1
406-SC-001

Running [ON/OFF]

Conveyor
406-CV-004

Running [ON/OFF]

HPGR Screen 2
406-SC-002

Running [ON/OFF]

Conveyor
406-CV-005

Running [ON/OFF]

HPGR Silo Feeder 1

406-FE-001
HPGR Silo Feeder 2
406-FE-002
HPGR Feed Bin
Feeder 1
406-FE-003
HPGR Feed Bin
Feeder 2 406-FE-004
HPGR Screen Feeder 1
406-FE-005
HPGR Screen Feeder 2
406-FE-006
Conveyor
406-CV-001
Conveyor
406-CV-002

Conveyor
406-CV-006
Conveyor
406-CV-007

Running time [min]

Speed [%]

Belt scale
406-WT-010B

Speed [%]

Belt scale
406-WT-402
Belt scale
406-WT-416
Belt scale
406-WT-433
Lynxx camera
(placed at 406-CV-007)

Speed [%]
Speed [%]
Speed [%]
Running [ON/OFF]
Running [ON/OFF]

24

Running [ON/OFF]
Running [ON/OFF]
Mass [t]
Mass [t]
Mass [t]
Mass [t]
Size [mm]

CHAPTER 4 - Data

4.2 QUALITATIVE DATA

Classification and Flotation

The qualitative data has been gathered

through unstructured interviews. In an
unstructured interview, the interviewers do
not follow a strict structure of questions,
but instead might have only one or a few
questions for the interviewee (Bryman &
Bell, 2011). The interview therefore
resembles a conversation. The unstructured
personal interviews have been conducted
both in Sweden and in South Africa.

Prof. Magnus Evertsson, Chalmers

Dr. Erik Hulthn, Chalmers Rock
Processing Systems

Johannes Quist, PhD student, Chalmers

Ludvig Lindlf, PhD student at the

Division of Operations Management

During the time spent at the University of

Cape Town in South Africa, unstructured
interviews were held with Dr. Aubrey
Mainza, Department of Minerals Research,
from September 12th until December 20th.

Plant Manager

Technical Manager

Operators

Planners

Head of R&D

Head of Control and Instrumentation

Head of Engineering

Leader Process Control Engineer

Control Engineer

4.3.1 ORGANISATIONAL CHART

To understand the organisational structure
at Mogalakwena North Concentrator, an
organisational chart (Appendix I) over the
divisions involved in areas critical for this
project, has been developed. This chart
formed the basis for an understanding of
the divisional and inter-divisional processes.

Unstructured interviews were held on site

with persons with the following positions
from September until mid-December 2012:

Production Leaders

The empirical study serves to give the

reader an understanding of the current
state in interesting areas of the operations,
both in production technical and in
organisational terms.
Along with the
theoretical framework it will serve as a
basis for the analysis. The data presented
has been captured during interviews and
observations on site.

Torbjrn Ylip, Senior Lecturer in

Producton Systems

4.3 EMPIRICAL DATA

Rock Processing Systems

Engineering Specialists

Observations have been made on site from

September to December 2012.

Gauti Asbjrnsson, PhD student,

Chalmers Rock Processing Systems

The following persons at Anglo American

Platinum in Johannesburg have been
consulted during the project (September mid-December 2012):

Rock Processing Systems

Metallurgist Graduates for Milling &

Classification and Flotation

During
the
pre-study
in
Sweden
unstructured interviews were held with the
following persons at Chalmers University
of Technology from August 15th until
September 10th:

Metallurgist for Dry Section, Milling &

25

CHAPTER 4 - Data

4.3.2 PROCESS MAP

COMMINUTION

The comminution circuit is divided into

seven sub-areas (102, 401, 405, 406, 407, 408
and 440, see Figure 6). These areas are
mapped in more detail in Appendix I.
Except for the information gathered from
interviews and observations; this map
contains information from the original
plant drawings.

Comminution is a part of the concentrator

process and is defined as:

The action of reducing a material,

especially a mineral ore, to minute particles
or fragments.
(Oxford Dictionaries, 2012)

Figure 6 - Process map displaying area 102-408 at MNC (For more detailed maps, see
Appendix I)

4.3.3 UNITS IN A CONCENTRATOR

PLANT

Units dedicated to comminution in a

concentrator are, for instance, the primary
crusher, secondary crusher, High Pressure
Grinding Roll crusher (HPGR) and the ball
mill.

In the following section, the different types

of process units included in the project and
their dedicated tasks in the plant will be
described.

CONVEYOR
A conveyor is a unit which has the purpose
of moving or transporting bulk material or

26

CHAPTER 4 - Data
objects in a path predetermined by the
design of the conveyor. The conveyor can
be horizontal, inclined or vertical in its
design. At MNC, all bulk transports are
performed by conveyors and the total
length of all conveyors is approximately
9000 metres.

FEEDER
A feeder is a unit that puts material in
motion. Its purpose is to regulate the
amount of material that, for example, is fed
into a crusher or from a storage silo onto a
conveyor.

CLASSIFIER

4.3.4 AREA AFFILIATIONS OF

UNITS

A classifier is a unit which classifies the

material physically by separating it based
on its particle size. At a concentrator this
can be performed by, for instance, a grizzly,
screen or cyclone, all which have the same
purpose but perform the separation of
material in different ways.

The following section will present the name

and type of the units included in the project
based on their area affiliation. The
presentation order of the following tables is
equal to the material flow order in the
process.

AREA 102
Table 4 The name and number of units in process area 102

TYPE OF UNIT
Comminution unit
Feeder
Conveyor
Classifier

NO. OF
UNITS
1
2
2
0

NAME
102-CR-001 Primary crusher
102-FE-001 & 102-FE-002
102-CV-001 & 102-CV-002
-

AREA 401
Table 5 - The name and number of units in process area 401

TYPE OF UNIT
Comminution unit
Feeder
Conveyor
Classifier

NO. OF
UNITS
0
6
1
1

NAME
401-FE-001 - 401-FE-006
401-CV-001
401-GY-001 Grizzly

AREA 405
Table 6 - The name and number of units in process area 405

TYPE OF UNIT
Comminution unit

NO. OF
UNITS
3

Feeder
Conveyor
Classifier

5
6
2

NAME
405-CR-001 405-CR-003 Secondary Crushers
1,2,3
405-FE-001 405-FE-005
405-CV-001 405-CV-006
405-SC-001 & 405-SC-002 Secondary Screen 1 & 2

27

CHAPTER 4 - Data

AREA 406
Table 7 - The name and number of units in process area 406

TYPE OF UNIT
Comminution unit
Feeder
Conveyor
Classifier

NO. OF
UNITS
1
6
7
2

NAME
406-CR-001 HPGR Crusher
406-FE-001 406-FE-006
406-CV-001 406-CV-007
406-SC-001 & 406-SC-002 HPGR Screen 1 & 2

AREA 407
Table 8 - The name and number of units in process area 407

TYPE OF UNIT
Comminution unit
Feeder
Conveyor
Classifier

NO. OF
UNITS
0
2
2
0

NAME
407-FE-001 & 407-FE-002
407-CV-001 & 407-CV-003
-

IN TOTAL
Table 9 - The total number of units in process included in this project

TYPE OF UNIT
Comminution unit
Feeder
Conveyor
Classifier

NO. OF
UNITS
5
21
15
5

28

CHAPTER 4 - Data

4.3.4 COMMINUTION UNIT FACTS

Since Mogalakwena North Concentrator is
designed to be the worlds largest single
stream platinum concentrator (Mining
Weekly, 2008), it has several remarkable
units.

Anglo Platinum. The drives are powered by

a 17.5 MW motor, five times as big as a
similarly-sized throughput mill. When the
plant was commissioned in 2009 these
GMDs, with a diameter of 26 feet (7,9 m),
were the largest installed in the world..
(Probert, 2012)

CRUSHERS

LYNXX CAMERAS

The circuit includes in total three crushing

systems one primary crusher, three
secondary crushers and one tertiary crusher
(HPGR).

The plant has five Lynxx cameras, which

optically determine the particle size of the
material which is being transported on a
conveyor. The map displaying the installed
positions can be viewed in Appendix III.

The primary crusher, which is the largest

primary gyratory crusher in the world,
weighs 480 tonnes, is 60-feet (18.3 m) in
diameter and has a 1 MW motor. The
crusher can handle ore pieces up to one
square sectional metre and has a maximum
capacity of 3000 tonnes per hour.

4.4 VALUE STREAM MAPS

The Value Stream Mapping process is
based on the interview results and
observations and consists of individual
maps of the current state of processes and
the operations found to be of interest for
this project. The maps were developed to
give an enhanced understanding of the
current state of the processes and are
presented here to give the reader an
enhanced understanding of the state.

The secondary crusher system consists of

three identical hydrocone crushers with
different settings due to the different sized
materials with which they are fed. This
design makes the plant setting quite rare
and requires customised equipment
performance calculations.
The tertiary crusher is a HPGR (High
Pressure Grinding Roll) crusher, which
utilises two 100 tonne rolls, one fixed in
position and the other moving horizontally
to adjust the gap between them, to crush
the ore. Mogalakwena North Concentrator
is the first platinum plant in the world to
utilise a HPGR crusher for this purpose.

CRUSHER AND MILL STOPS

REPORTING PROCEDURE MAP
An analysis of the production process
called Pain analysis has been developed by
the Masters thesis writers. To be able to
perform this analysis, data from crusher
and mill stops is required. To understand
the internal reporting process for these
reports the entire process was mapped
accordingly (Figure 4).

MILLS
The primary and secondary mills are the
two first gearless mill drives (GMD) at

29

CHAPTER 4 - Data

Figure 4 Current Crusher and Mill Stops Reporting Procedure Map

The Crusher and Mill Stops Reporting
Procedure Map displays the procedure for
reporting the stops. It can be seen that the
input to the report is a daily activity,
whereas the submitting of the report is
done on a monthly basis. This can be
regarded as a poorly synchronised
procedure and will eventually result in
missing or delayed data regarding the stops.
The analysis of crusher and mill stoppages
will, due to this process, only be possible
once a month, which is insufficient for the
Pain analysis.

JOB CARD INSPECTIONS MAP

A plants performance is highly affected by
the quality of the performed maintenance
work. Due to this, the current process for
daily maintenance work at the plant was
mapped to acquire an understanding of
how it works. The map of the job card
inspections can be viewed in Appendix IV.

30

CHAPTER FIVE

RESULTS
This chapter will present the results that have been obtained in the project. The results are
presented in three distinct sections Define, Develop and Design. The same sections are used
in the discussion and conclusions in order for the reader to more easily follow the red thread.

31

CHAPTER 5 - Results

CHAPTER 5 INTRODUCTION TO RESULTS

Define a calculation
model for OEE &
other equipment
performance metrics
in a single stream
process

Develop a tool that

calculates OEE &
other equipment
performance metrics
in real time

Design a method
describing how to use
the tool output in the
organisation with
primary focus on
finding root causes

Figure 7 Project phases

The results are presented according to the
three distinct project phases (see Figure 7).
The first phase was to define a calculation
model for OEE and other equipment
performance metrics. The second phase
was to develop a tool (OPT) which uses the
calculation model to perform real time
calculations of OEE and other equipment
performance metrics. The third phase was
to design a methodology to use in the
organisation to find root causes to
productivity limiting issues. Moreover, this
chapter includes a presentation of how to
calculate OEE in a General Single Stream
Process and results from the new procedure
for Crusher and Mill Stops Reporting that
was introduced at MNC by the two
Masters thesis writers.

The calculation model covers the five

metrics; OEE, Availability, Utilised
Uptime, Mean Time Between Failure
(MTBF) & Mean Time To Repair (MTTR)
and Pain (see Figure 9). The final
calculation for these metrics will be
presented in the following sections. The
calculations are based on the Anglo
American Equipment Performance Metrics
definitions and have taken inspiration from
a customised calculation model developed
by the Masters thesis writers (see section
5.4) to suit the process of MNC, since the
Anglo American definition is not
comprehensive enough to be used in this
process. However, since it is a company
standard, the aim has been to use it as
extensively as possible and it has been a
wish from the company not to deviate from
the standard whenever possible.

5.1 CALCULATION MODEL

In the following
section, the results
of the first phase
(see Figure 8) of
this project will be
presented.
This
part contains the Figure 8 Project
calculation model phase 1 - Define
for OEE and other
equipment performance metrics in a 24/7
single stream comminution process.

Define a
model

32

CHAPTER 5 - Results

OEE
Utilised
Uptime

Availability

Overall
Utilisation

Performance

MTBF
&
MTTR

Process
Pain

Quality

Figure 9 - Equipment metrics included in the calculation model

5.1.1 FINAL OEE CALCULATION

The final OEE equations used in all
calculations will be presented in this section.
The metrics consist of three parts - Overall
Utilisation, Performance and Quality (see
Equation 12). The three components of
OEE can also be used as individual metrics.
The components of the OEE equation are
presented in the following sections.

To clarify, Overall Utilisation is the metric

corresponding to the general OEE
calculations measure known as Availability,
displaying the equipment usage, however
not calculated in the same way or with the
same result.

OVERALL UTILISATION

Equipment Performance is calculated

according to the Anglo American
Equipment Performance Metrics as the
ratio between Actual Production Rate and
Target Production Rate. The Performance
essentially indicates how efficiently the unit
has been working, i.e. to what degree the
unit has been doing things in the correct
way.

PERFORMANCE

To determine the utilised time of a unit, the

ratio between Primary production time and
Total time is used. Primary production time
is defined as Time equipment is utilised
for production and Total time is defined
as The total possible hours available.
These definitions are taken from the Anglo
American Equipment Performance Metrics.
For explanations of the equation
components, see Figure 10.

OEE Overall Utilisation x Performance x Quality

Overall Utilisation

Primary Production
Total time

Equation 12
Equation 13

Actual Production Rate

Target Production Rate

Actual Production Achieved / Primary Production
=
Target Production Rate
Performance

33

Equation 14

CHAPTER 5 - Results

Figure: Freely after Anglo American Equipment Performance Metrics Time Model

Figure 10 - Time definitions by Anglo American Equipment Performance Metrics

The Actual Production Rate is the ratio
between Actual Production Achieved and
Primary Production, which both can be
calculated with data drawn from the PIdatabase. The Target Production Rate,
however, is an input measure which has to
be defined for every single unit.

equation compares the Actual Particle Size

at a certain point in the process to the
Target Particle Size for that point (see
Equation 15).
The Quality is defined as the mean
deviation above Target Size as a percentage
of the Target Size (see Figure 11). This
implies that all particles below target size
result in zero deviation, hence a 100%
Quality. To achieve the metric Quality
suitable for the OEE calculation and not
the deviation, the ratio is subtracted from 1.
For instance, if all particles are below
Target Size, the Quality will be 100%. If
some particles are above Target Size, the
equation will compute their individual
deviation from the Target Size. Together
with the particle sizes below the Target
Size, which all are regarded as having no

QUALITY
The method to calculate the product
Quality was developed by the Masters
thesis writers. The Quality looks at the
particle size and shows to what extent the
particle size is below the targeted size. The
target and actual particle size concerns the
P80 value, which is a commonly used value
in the comminution industry. P80 is defined
as the size where 80 percent of the material
passes a certain upper size limit. The

Quality 1

Mean deviation from Target Size

Target Size
n

Deviation from target size

i 1

n
Target Size
34

Equation 15

CHAPTER 5 - Results
deviation, an average deviation is
computed. This average deviation is then
divided by the Target Size and subtracted
from 1 to get a percentage between 0 and
100.

Oversized
particles

Approved particle sizes

[mm]
Figure 11 - Example size distribution of P80 with target size marked

5.1.2 AVAILABILITY
OEE and its components is the first metric
in the calculation model. The second metric
is Availability.
In some cases, the Overall Utilisation
equation is not sufficient and/or applicable,
for instance when there is a lack of data. In
that case, Availability (see Equation 16) is
used. The Availability shows the
percentage uptime of the unit and is
defined as the ratio between Uptime and
Total Time.

35

CHAPTER 5 - Results

Availability

Uptime
Total Time

Utilised Uptime

Equation 16

Overall Utilisation
Availability

Equation 17

MTBF

Uptime
Number of stops

Equation 18

MTTR

Equipment Downtime Time

Number of stops

Equation 19

5.1.3 UTILISED UPTIME

5.1.5 PAIN ANALYSIS

The Masters thesis writers have developed

a new metric to display the ratio between
Overall Utilisation and Availability, called
Utilised Uptime (see Equation 17). This is
the third metric in the calculation model.
The Utilised Uptime shows the percentage
of the available time that is used for
Primary Production. This metric is possible
to compute for the units where Overall
Utilisation can be calculated.

The fifth metric in the calculation model is

Pain. In order to help focus the efforts in
the daily work at the plant, this analysis
tool, which includes a new way of
displaying failures, has been developed by
the Masters thesis writers. The analysis
provides a metric called Pain and looks at
how much pain a certain stop has caused a
unit. Instead of the user being forced to
look at and compare both the frequency of
a stop and its downtime in order to find the
most painful problem, Pain can be used to
give an aggregated view.

5.1.4 MTBF & MTTR

The Pain is calculated as the product of

frequency of the stop and total downtime
caused by the stop (see Equation 20). This
gives a total view of the pain the problem
has caused. The unit of Pain is time
(minutes) but can be regarded with minor
importance since it does not display an
actual downtime but the sum of all
downtimes multiplied by the number of
stoppages. Therefore, the Pain is used as a
unit-less metric and displayed unit-less in
the tool.

The fourth metric in the calculation model

is the combination of Mean Time Between
Failures (MTBF) and Mean Time To
Repair (MTTR). These two metrics are
calculated to give an indication of asset
reliability and the quality of maintenance
work. For explanations of the equation
components, see Figure 10.

Pain failure, type _ x Frequency of error Total downtime caused by error =

n

n failure, type _ x n failure, type _ x, n

Equation 20

i 1

n failure, type _ x t failure, type _ x, 1 t failure, type _ x, 2 ... t failure, type _ x, n

36

CHAPTER 5 - Results
Along with the Pain equation, the Masters
thesis writers have developed the Pain
diagram which displays the errors with the
top 6 highest Pains, sorted in ascending
order (see Figure 12). This diagram gives a
clear view of the currently most alarming
errors in the process.
To facilitate the drilling down in stop
reasons, the frequency and downtime can
be displayed and analysed in a pyramid
diagram (see Figure 13).

Figure 12 The Pain chart as it is displayed in OPT

Figure 13 Pyramid diagram of Pain components

5.1.6 SETTING FILTERS & TARGET

VALUES

large amount of data analysis. These

determined levels shall be reviewed
periodically due to possible process
changes.

In the calculation of Overall Utilisation,

Availability, Performance and Quality
target values are used. These target values
are critical components of the calculations
since they each have a large impact on the
results. This section will present the
methods used when determining the
individual target values.

OVERALL UTILISATION &

AVAILABILITY
The Overall Utilisation and Availability
calculations use filters to determine in what
type of activity the unit is engaging. For
instance, when a comminuting unit is
running above a certain power level, it is
assumed to be performing primary
production activities. In this project these
levels have been determined based on a

37

CHAPTER 5 - Results
should therefore be taken into account
when setting target production rates.

PERFORMANCE
When calculating Performance, a Target
Production Rate is used for comparison
with the Actual Production Rate. The
Anglo American Equipment Performance
Metrics describes three ways of setting the
Target Production Rate, these ways are as
follows:

QUALITY
When calculating Quality, a Target Particle
Size is used for comparison with the Actual
Particle Size. The Target Particle Size at
different points in the process has been set
with advice from production experts at
MNC. The target is based on the particle
size demands of the downstream
comminuting unit.

1. Best demonstrated production rate,

which is defined as the best demonstrated
performance determined by calculating the
average of the five best monthly production
rates.

5.1.7 CLASSIFICATION

2. Equipment nameplate/design capacity

rate.

The equipment has been divided into four

different classification groups; A, B, C and
D. These have been made in order to
divide equipment
based
on their
complexity and need of monitoring. The
available measuring points have also
affected what group in which the units were
placed.
Table
10
presents
the
measurements
for
the
different
classification groups.

3. Budgeted production rate.

The Masters thesis writers propose a
fourth way of setting target production rate.
The fourth way should be based on the
design capacity of the unit and take
changes on parts and settings that have
affected
installed
capacity
into
consideration. For a crusher, critical
changes could, for instance, be changing
chamber or closed side setting (CSS). Such
changes will result in a capacity change and
Table 10 Classification of units

Classification

Equipment

Circuits

Comminution
units

Supporting
equipment
(with available
measures)

Details

OEE
Overall
Utilisation
Performance
Quality
Availability
Utilised Uptime

OEE
Overall
Utilisation
Performance
Availability
Utilised Uptime
MTBF
MTTR
Pain

Overall
Utilisation
Availability
Utilised
Uptime

Supporting
equipment
(without
available
measures)
Availability

38

CHAPTER 5 - Results
The program extracts real time data from
the PI-database and computes the defined
metrics from the calculation model defined
in the first phase of the project.

5.2 THE OVERALL

PRODUCTIVITY TOOL (OPT)
In the following
section,
the
results of the
second phase of
this project will
be presented.

For detailed screen shots of the OPT

software, see Appendix V.

Develop a
tool

The result of the Figure 14 Project

second phase of phase 2 - Develop
this Masters thesis project is a
management tool which combines the
parameters identified during the first phase
and calculates them in real time. The tool
displays charts and tables of equipment
metrics and the metric Pain. The developed
tool is called the Overall Productivity Tool
(OPT). The abbreviation suits the purpose
of the tool well since the word opt is
defined as (Merriam-Webster, 2012):

Figure 15 - Structure of OPT and its

included metrics

5.2.2 OEE MODULE

The OEE module displays charts with the
OEE for the circuit and the crushing unit in
the monitored area. It also displays charts
with the Overall Utilisation, Performance,
Quality, Availability, and Utilised Uptime
for the units included in the monitored area.
The charts display data for the period of
the previous 7 days as well as for the
previous 30 days, which gives the user a
presentation of the metrics in both short
and medium term.

To make a choice; especially: to decide in

favour of something

5.2.1 OPT STRUCTURE

The metrics in the tool are organised in a
manner that provides a dynamic overview
of the equipment in the process. OPT is
built up of four modules in an excel
document OEE, OEE Table, Pain
analysis and Stop Table. The modules will
be presented in the following four sections.
OPT includes the five metrics OEE,
Availability, Utilised Uptime, Mean Time
Between Failure (MTBF) & Mean Time To
Repair (MTTR) and Pain Analysis. OPT is
covering the process areas 102, 401, 405,
406 and 407 at MNC. Figure 15 displays the
structure of OPT.

The charts displaying Overall Utilisation,

Availability and Utilised Uptime are sorted
in ascending order to visualise the units
with lowest current values.
The charts displaying Quality and
Performance show those values for the
crushing unit in the monitored circuit as
well as for the entire circuit (see Figure 21
& 17). In the OEE module, explanations of
the charts and their metrics are provided to
give a clear view of what is being presented.
At the top of the sheet, the monitored time
intervals are displayed.

OPT is developed and programmed in

Excel and Visual Basic Editor and has
direct communication with the PI-database.
39

CHAPTER 5 - Results

Figure 17 Quality graph from OPT

Figure 16 OEE graph from OPT

Figure 18 Availability graph from OPT

Figure 19 Overall Utilisation graph from OPT

Figure 21 Utilised Uptime graph from OPT

Figure 20 Performance graph from OPT

40

CHAPTER 5 - Results

5.2.3 OEE TABLE MODULE

5.2.4 PAIN MODULE

The OEE Table module displays the

components of the OEE calculations
categorised by unit type (crusher, classifiers,
feeders and conveyors). The OEE Table
module provides the calculations with a
transparency and can be used to acquire a
more thorough understanding of the charts
displayed in the OEE module. At the top
of the module, the monitored time intervals
are displayed. Additional metrics displayed
in the module are Average Size and
Average Deviation.

Pain is a way of visualising the combination

of stop frequency and downtime for a unit
in the monitored area. The Pain Module in
OPT displays charts with the top 6 Pains in
the monitored process area sorted in
descending order (see Figure 23).
At the top of the sheet, the monitored time
intervals are displayed.
The charts are sorted in descending order
to visualise which stop reasons currently
are causing the greatest damage to the
process. The stop reasons are labelled
below each bar in the chart. The unit for
the y-axis is thousand minutes, however,
Pain is displayed as a unit-less metric.

The definition of any displayed component

can be viewed by hovering over the
component name (see Figure 22, where the
pointer is hovered over Availability).
The displayed rates included in OEE have
a colour code which visualises the current
status. Green is for Satisfactory, yellow for
Poor and red for Alarming.
The colour limits for the different units can
be seen at the bottom of the sheet. The
limits shall be set based on the business
targets and can only be changed by the
administrator of OPT.

Figure 22 - Information appearing when hovering over Availability

41

CHAPTER 5 - Results
At the top of the sheet, above the charts,
Mean Time Between Failures (MTBF) and
Mean Time To Repair (MTTR) are
displayed for the crushing unit in the area.

Figure 23 Pain chart as it is presented in

OPT

Stop time: The time the unit stopped

Start-up time: The time the unit started
up after the stop
Duration: The duration of the stop
Stop reason: The reason for the stop
Manually entered comment: Possible
manually entered comment by stand-by
official
Downtime categories
- Downtime sub-category code
- Downtime sub-category name
- Downtime category code
- Downtime category name
Scheduled/Unscheduled: Indicates if
the stop was scheduled or not

The downtime categories are used to

facilitate the allocation of stops in
alignment with the Anglo American
Equipment Performance Metrics Time
Model.

5.2.5 STOP TABLE MODULE

The Stop Table module presents the
information upon which the Pain Analysis
is based in a more detailed way (see Figure
24). The stop information is drawn from
stop reporting through the PI-database.
The following information is presented to
the user:

At the top of the sheet the total number of

stops and the total downtime in the chosen
time interval are displayed. At the extreme
top of the sheet, the monitored time
intervals are displayed.

Figure 24 The stop table as it is presented in OPT

42

CHAPTER 5 - Results
successful
combination.
The
group
members different backgrounds help to
create a broader view of the OPT output. If
a cross functional group is used, it also
helps to increase the collaboration between
departments and limits dual work. The
OPT Methods incorporated action list
facilitates tracking of issued actions, which
has been seen necessary in large
organisations.
Another positive effect is
that cross-functionality unites the users
among a common systematic problem
solving technique.

5.3 OPT METHOD

In the following
section,
the
results of the
third phase of
this project will
be presented.

Design a
method
Figure 25 Project
phase three - Design

The result of the

third phase of this Masters thesis project is
a methodology describing how to use the
output of OPT in a productive way with
primary focus on finding root causes to
productivity limiting issues and follow up
on action taken. The three supportive areas

5.3.2 FIVE WHYS

Figure 26 A model describing the intended usage of OPT

to the right of OPT (in Figure 26) are User
expertise, the 5 WHYs and OPT
Guidelines. These areas will be presented
in more detail here.

The root cause finding technique, the 5

WHYs, is proposed as a suitable method to
use to find root causes to issues
encountered when analysing the OPT
outcome. For more information on the 5
WHYs, see section 2.2.1 Five Whys.

5.3.1 USER EXPERTISE

In order to achieve a valid result when
analysing the OPT output, a certain user
expertise is required. The user shall possess
good knowledge of the process and have
previous experience from working with the
process. It is also important that the user
possesses a systematic problem solving
technique.

5.3.3 OPT GUIDELINES

The Masters thesis writers have developed
user guidelines for how to use OPT to
facilitate usage of the tool. The guidelines
will be presented here and consist of the
following documents:

The success of the OPT method is also

determined by its users and their expertise.
It has been seen that a cross functional user
group, i.e. a group consisting of personnel
from different functional groups of the
organisation, including for instance both
technical and engineering staff, is the most

OPT Manual

OPT Meeting Procedure

OPT Action List

OPT MANUAL
The OPT Manual is a complete guide on
how to use the tool. It provides a step-by43

CHAPTER 5 - Results
step guide on how to run the OPT software,
explanations of the sheets in OPT (OEE,
OEE Table and Pain and Stop Table),
definitions of the metrics displayed and
examples of how to interpret the output
values. The OPT Manual shall be used to
facilitate the usage of OPT for both
experienced users and beginners.

d. Seek support in OPT Manual if

needed
4. Fill in OPT Action List
a. Add actions based on encountered
issues to OPT Action List

OPT ACTION LIST

The OPT Manual can be viewed in full in

Appendix VI.

The OPT Action List is a document used to

capture and keep track of actions formed
by issues found when analysing the OPT
output. A proposed OPT Action List
design has been developed and can be
viewed in full in Appendix VIII. The action
list consists of the following categories to
be filled in:

OPT MEETING PROCEDURE

The document called OPT Meeting
Procedure proposes a structured method
for holding a meeting focused on OPT. It is
suggested that a meeting be held once a
week to review and analyse the OPT
outcome as well as to create actions to
solve encountered issues and follow up on
issued actions. A full OPT Meeting
Procedure document has been developed
and can be viewed in full in Appendix VII.

ACTION

Action: Description of action to be

taken

RACI responsibility roles

The steps of the OPT meeting are as

follows:

Responsible: The person(s) working

with achieving the task

1. Meeting Preparations

Accountable: The person answerable

for the correct completion of the task

a. Update OPT
b. Review Action List from previous
meeting

Consulted: The person(s) to ask for

advice in regards to the task

2. Start meeting with follow-ups of

outstanding actions

Informed: The person(s) to keep

informed

DATE

a. Go through issued actions

i. If completed, fill in OPT Action
List
ii. If outstanding, refine action
and/or update due date

Due date: When to be completed

Completion date: Actual completion

OUTCOME

3. Running OPT

Result: Result of action taken

Conclusions: Conclusions that can be

drawn from the action and its result.

a. Review outcome
b. Analyse outcome
c. Use the 5 WHYs to find root
causes of encountered issues

May result in a new action to be added

to the OPT Action List

44

CHAPTER 5 - Results

5.4 OEE FOR A GENERAL

SINGLE STREAM PROCESS

The general quality calculation was

customised to be valid for a single stream
comminution process. Instead of using
good pieces as a measure, the Masters
thesis writers have developed a new
method to calculate quality based on
particle size. The Quality calculation is
described in more detail in section 5.1.1
Final OEE Calculation.

Based on the general method of

calculating OEE and the time definitions
from the Anglo American Equipment
Performance Metrics Time Model, a
customised version has been developed in
this Masters thesis project to better suit a
general single stream process. However,
this is not the method used in the
developed tool, OPT. To view the OEE
calculation model used in OPT, see 5.1.1
Final OEE Calculation.

Uptime
Total Time

Availability

Performance

Quality 1

Equation 21

Actual Production / Target Rate

Uptime

Equation 22

Mean deviation from Target Size

Target Size

Deviation from target size

i 1

Equation 23
i

n
Target Size

The general calculation of Availability was

customised in order to better fit a single
stream process. Instead of using Planned
Production Time as the denominator in the
general OEE definition, Total Time is used,
which is the total hours available. The
Availability is therefore determined by
dividing the Uptime by the Total Time.

5.5 CRUSHER AND MILL STOPS

REPORTING PROCEDURE
To enable the Pain Analysis to be
performed, there is a need for daily reports
of stops and their causes. There is a daily
record of crusher and mill stops which is
entered manually every morning. However,
when this project started, the stop data was
extracted into a report on a monthly basis,
and its lead-time was longer than required.

The general performance calculation had to

be customised to be valid for a single
stream process. Performance for a general
single stream process can be calculated as
stated in Equation 22. This gives in the
ratio between the targeted time to produce
the actual tonnes produced and the actual
time consumed (uptime).

This means that the required data to

perform the Pain Analysis existed but a
system to extract it daily was not in place
and the data required a great deal of

45

CHAPTER 5 - Results
preparation to get it to a stage where it
could be analysed. Therefore, a new system
that allows all the crusher and mill stops
data to be recorded and reported daily in
electronic format has been developed and
implemented at MNC by the Masters
thesis writers. This procedure has
eliminated the manual work involved when
creating the crusher and mill stop reports
and also reduced the risk for manual data
entry errors. A map of the new procedure

is presented in Figure 27 and can be

compared to the former procedure found in
section 4.4, Figure 4.
Beyond saving manual work, the procedure
automatically allocates the lost time events
to the time definitions presented in the
Anglo American Equipment Performance
Metrics, which facilitates the tracking of
stoppages.

Figure 27 - New Crusher and Mill Stops Reporting Procedure Map

46

CHAPTER SIX

DISCUSSION & CONCLUSIONS

This chapter will discuss the findings of the project, drawn conclusions,
answers to the research questions, observations and recommendations for the organisation are
also presented and finally areas for future research are identified. The discussion and
conclusions are presented in three distinct sections Define, Develop and Design.

47

CHAPTER 6 Discussion & Conclusion

CHAPTER 6 - INTRODUCTION
TO DISCUSSION

6.1 CALCULATION
MODEL

The mining industry has lagged behind

manufacturing industry when it comes to
process control and process optimisation.
For instance, the mining industry does not
use modern methods when it comes to
measuring and calculating equipment
performance metrics in an accurate way
and using this information to monitor and
improve processes. One of the goals in this
project has been to use some of the
knowledge from the manufacturing
industry and apply it in the mining industry.

The baseline for

developing the
calculation
model for OEE
in
a
single
stream
Figure 29 Project
comminution
process was to phase one - Define
make it as generic
as possible and avoid making it site specific.
During the pre-study in Sweden, the
Masters thesis writers developed a
functional method to calculate OEE in a
single stream comminution process which
was tested and validated by using historical
production data from MNC. This was a
good learning point which facilitated the
understanding of the characteristics of the
model and how certain parameters affect
output.

Define a
model

The discussion and conclusions are

presented according to the three distinct
project phases (see Figure 28). The initial
challenge, as well as the first phase of this
project, was to define an equipment
performance calculation model for a single
stream comminution process. The second
phase of this project was to develop a tool
(OPT) that uses the calculation model to
perform real time calculations of OEE and
other equipment performance metrics. The
third phase was to develop a methodology
describing how to use the tool output in the
organisation in a value creating way, such
as finding root causes to productivity
limiting issues.

Later, the Masters thesis writers were

introduced to the Anglo American
Equipment Performance Metrics, which is
an internal company standard describing
how to calculate equipment performance
metrics including, for instance, OEE. The
model developed by the Masters thesis
writers was found to be well aligned with
the company standard which is very good.
The
standard,
however,
was
not
comprehensive enough regarding quality
definitions and calculations. Therefore, the
quality definitions from the thesis writers
OEE model were adopted into the
companys standard OEE definitions. The
OEE model was continually under

In the following sections a discussion on

findings from the different phases of the
project will be presented as well as
conclusions drawn, answers to the research
questions, observations, recommendations
for the organisation and finally future
research proposals.

Define a calculation
model for OEE &
other equipment
performance metrics
in a single stream
process

Develop a tool that

calculates OEE &
other equipment
performance metrics
in real time

Figure 28 - Project phases

48

Design a method
describing how to use
the tool output in the
organisation with
primary focus on
finding root causes

CHAPTER 6 Discussion & Conclusion

development during the project and as
knowledge in the area grew, the model was
refined and additional parameters were
added.

weight, hence all are equally important

when it comes to OEE. The OEE provides
a good measure of the status of a unit;
however, it cannot tell what causes the
OEE number or how the OEE can be
changed. By looking into the three included
parameters, a slightly better view of the
current unit status will be provided. Still,
answers to possible issues will be hard to
determine. To give a more inclusive picture
of the unit status, two additional
parameters are presented in OPT Availability and Utilised Uptime.

OEE is a good performance measure, but it

is important to not read it as one parameter;
it is actually four. The individual
parameters give a broader understanding of
the equipments performance and provide
different approaches as to how the
equipments performance can be improved.
The downside of OEE is that it cannot
provide the user with the reason for an
eventual increase or decrease of the
measure. It would of course be a great
feature if the OEE could tell exactly what
happened in the process, but that is not the
character of the measure. This gap can be
partially filled by using a systematic
analysis method developed by the Masters
thesis writers, discussed under section 6.3
OPT Method.

OVERALL UTILISATION
The metric Overall Utilisation shows the
units time distribution as the percentage of
time the unit is used for primary production.
This is the metric which represents the time
usage in OEE. However, it will not show
the percentage of time the unit has been
available for production, merely the time it
has been utilised. By looking at the Overall
Utilisation one cannot tell whether the unit
has been utilised all the available time or if
there is more available time to utilise. That
is, one cannot tell if the available time has
to be increased in order to increase the
utilisation or if the utilisation can be
increased without increasing the availability
of the unit. To be able to determine this,
OPT presents both Availability and Overall
Utilisation (see section 6.1.2) for all
possible units. Displaying of both the
Overall Utilisation and the Availability
facilitates the understanding of the
distribution of the equipments total time.
To clarify the relation between Overall
Utilisation and Availability, the Masters
thesis writers defined a metric referred to
as Utilised uptime (see section 6.1.3), which
is the ratio between Overall Utilisation and
Availability.

The fact that OEE already existed as an

internal standard has only been beneficial
for the project since it has created a
smoother introduction of the OPT and its
parameters. However, the OEE methods
have not yet been fully implemented in the
organisation and the OPT can act as a
facilitator in the full implementation of
OEE. In this way, OPT and the
organisational OEE implementation can
interact to create an OEE proficient
organisation.

6.1.1 OEE CALCULATION

A units OEE can be determined by
multiplying
its
Overall
Utilisation,
Performance
and
Quality
(see
Equation 24). The OEE is based on these
three metrics, each one carrying the same

OEE Availability x Performance x Quality

49

Equation 24

CHAPTER 6 Discussion & Conclusion

Due to lack of required data, it is not
possible to determine Overall Utilisation
for all units in the process. In those cases,
Availability is being calculated instead. The
user has to be aware that these two metrics
differ and shall not be compared. The
Availability can, however, be compared
between similar units since it is being
calculated for all units included in the
project.

assurance. The third metric in OEE

provides a view of the quality of the
performed work.
Due to lack of data, the Performance
metric cannot be determined for all units.
This is the case for all the classifiers,
conveyors and feeders at MNC. From the
available data one cannot tell the rate at
which the units have been performing;
hence, the Performance cannot be
determined. For these units, the OEE will
consist of Overall Utilisation and/or
Availability. This is acceptable since the
concerned
units
are
not
primary
contributors to the main task of the
production process to comminute ore.
Their main function can be regarded as
supportive, therefore their main concern is
to be available to perform their dedicated
task.

PERFORMANCE
The Performance of a unit has two
components, Target Production Rate and
Actual Production Rate, and is computed
as the ratio between the two. This means
that the Performance is as affected by the
Actual rate as the Target rate. The Actual
rate is determined by data extracted from
the PI-database and is only dependant on
the performance of the process unit. The
Target rate is set by the organisation
following certain guidelines (see section
5.1.6 Setting Filters & Target Values). This
in turn means that a rate set by the
organisation has a huge part in deciding the
Performance rate of a unit. Therefore, the
setting of the Target Production Rate needs
to be done very carefully, otherwise
Performance can turn out to be a
misleading metric.

QUALITY
The new definition of Quality (see
Equation 25) combined with the method of
how to, in practice, determine quality in a
single stream comminution process has not
been seen before. The development of a
quality metric makes the OEE calculation
complete and provides a more accurate
OEE value than previously when the
quality most often was assumed to be 100%
in a process such as this one. It is a wellworking method, however it could be
refined. It does not take into account the
magnitude of the deviations below target
size but assumes all sizes below target to
have a quality of 100%. The method could
be refined to take those variations into
account, which would result in a more
accurate quality measure. At MNC, there
was no need for lower particle size limits
since all particles smaller than target size
were accepted.

It should be noted that the Performance

can result in a ratio greater than 100%.
This will occur when the Actual Rate
exceeds the Target Rate, which obviously
happens when the Target Rate is defined at
a too low value. In such a case the Target
Rate shall be reviewed and possibly
adjusted.
The Performance metric shows how
efficiently the unit is working but will not
show if the right things have been done,
which is defined as effectiveness. The
effectiveness has to be ensured by other
organisational processes, such as quality

The method could also allow a certain span

of sizes around the target size, if the unit
50

CHAPTER 6 Discussion & Conclusion

would allow it. In this project, such defined
parameters have not been found. The
development of such target values should
be the next step in improving the quality
calculation.

have a large impact on the output making it

crucial to measure quality continuously in
the process. However, since quality is
monitored continuously in the process, the
quality calculation assumes that the input is
within the given target limits.

According to the performed literature

reviews, no way of determining quality in a
comminution process has been previously
defined. Normally the quality is assumed to
be 100%. The Anglo American Equipment
Performance Metrics quality measure
provides a definition of quality (Equation 8)
but no applicable definitions of the input
values, i.e. Actual Size and Target Size.
Therefore, the method developed by the
Masters thesis writers to determine
Quality (Equation 25) can be regarded as a
major finding in this project. The
developed quality measure gives the
possibility to perform a more complete
OEE for a comminution process, since
there now exists a way to measure quality.

The Actual Particle Size input to the

Quality calculation is extracted from the
PI-database which stems from Particle Size
Distribution curves. The curves are created
with data from Lynxx cameras that
optically determine the particle size
distribution (PSD) of bulk material on a
conveyor belt. The method provides a
complete size distribution curve and the
measure used in this calculation is P80. The
cameras are installed at certain points in
the process (see section 4.3.4 Lynxx
Cameras and Appendix III). The usage of
cameras to determine particle size is
beneficial since it does not interfere with
the process. The calculation uses the P80
size provided by the system and compares
it with the target P80 size for that point in
the process. That is, the Quality metric is
fully dependant on correct data from the
Lynxx cameras. If the data provided by the
cameras is regarded to be accurate, the
Quality metric also can be regarded as
accurate. The accuracy of the camera data
is ensured by regular calibration of the
equipment and the Quality calculations can
therefore be trusted.

The Quality metric provides OEE an

aspect of effectiveness to what degree the
right activities have been performed. If the
particles are within accepted size, the right
activities can be assumed to be performed.
The other OEE parameters will then tell
the amount of time the process needed to
produce the output (Overall Utilisation)
and at what rate the output was produced
(Performance). If the Quality is not
satisfactory, the unit is not performing the
intended activities and the case has to be
investigated. The Quality can be
unsatisfactory due to, for example,
improper
equipment
settings
or
unsatisfactory input to the unit. In a single
stream process an unsatisfactory input can

Quality 1

The new definition is valid given that the

required inputs are available. The input
does not have to stem from Lynxx cameras;
it can as well be drawn from any other
reliable size measuring system. This widens
the application of the model. However, a

Mean deviation from Target Size

Target Size
Equation 25

Deviation from target size

i 1

n
Target Size

51

CHAPTER 6 Discussion & Conclusion

target size will always be required. In this
case some of the target sizes for the process
were already determined, others were not.
At those points in the process where actual
size is possible to determine, but the target
size is undefined, a method should to be
developed to determine the target size.

the quality measured at 406-CV-007 is the

quality of the product performed by the
whole 406 circuit, not for any single piece
of equipment. The same applies for the
other process areas.

As previously mentioned, the Quality

measure will not take the magnitude of the
deviation below the target size into
consideration. It will only take into account
the percentage deviation above the target.
This presentation of the number is chosen
because all deviations below are regarded
as positive. However, it is understood that
the actual deviation is an important factor
to consider. Therefore, in addition to
displaying the quality, the tool displays the
actual size deviation in millimetres
including the sign of the value, plus or
minus. Based on those two values the user
of the tool can conclude how severe the
deviation is. For instance, a negative
deviation might be acceptable up to a
certain limit, whilst all positive deviations
might be unacceptable. This is a decision
point in the tool where the users expertise
has to be utilised (see section 5.3.4 OPT
Users).

The Overall Utilisation is used as the main

measure of time usage for a unit in the
calculation model. However, if the Overall
Utilisation equation for some reason is not
applicable (e.g. lack of data), Availability is
used. For units which lack the sufficient
data one cannot tell whether the unit is
performing primary production or not.
Therefore, the metric Availability is being
used in OPT for some classifiers, conveyors
and feeders instead of Overall Utilisation.

6.1.2 AVAILABILITY

Availability shows the ratio of time that the

unit is available for production, i.e. the time
it is not standing still and therefore has the
possibility to contribute to the production
process. However, the metric cannot
provide information on the productivity of
the equipment, which is presented through
the metric Performance.
The Availability is also used as an
additional parameter for the crushing unit
in OPT. Displaying both the Overall
Utilisation and the Availability facilitates
the understanding of the distribution of the
equipments total time, since the two
metrics use different parameters in their
respective equations. What has to be
considered is that in most instances the
metric Availability will give a higher value
than (or equal to) Overall Utilisation since
the Uptime, which Availability is based on,
is higher than (or equal to) the Primary
Production, which Overall Utilisation is
based on. This is because Uptime is more
inclusive (also including for example idling
time) than Primary Production (see Figure
10). This is important to regard when
comparing the two metrics. It is always

Further, this method calculates the quality

at given points in the process. For instance,
for area 406, the quality is measured at the
conveyor named 406-CV-007, which is the
conveyor belt between the secondary
screens and the mill feed silo. The actual
size at that point is the result of the entire
406 circuit working together. Obviously,
the HPGR crusher has alone reduced the
size of the particles, which is the main task
of the circuit. Still, all other equipment has
to be in place and functional in order to
bring the material through the circuit. The
screens have to split the material accurately,
the feeders have to feed, the conveyors
have to transport and so forth. Therefore,

52

CHAPTER 6 Discussion & Conclusion

highly recommended to analyse the
parameters included in a metric before
comparing between metrics.

condition, which often leads to high capital

investments. During the Masters thesis
writers time on site, the Utilised Uptime
was frequently calculated and analysed and
one important conclusion that could be
drawn was that Availability is not a critical
problem at MNC. The Utilised Uptime is
most often low and primary focus should
therefore be put on increasing the amount
of primary production, hence the Overall
Utilisation.

Availability is often used as an indication

of how well maintenance work is carried
out on an asset, i.e. how much of the total
time the asset is available for production.
However, efficiency of maintenance work is
not the only factor affecting the available
equipment time. At MNC, and in almost
every single stream process, interlocks are
used to control the process units relative
behaviour. This can result in, for instance,
an upstream unit standing still due to a
breakdown downstream. For this reason
the analysis of available time has to take
into account the reason for the downtime,
which can sometimes be out of the control
area of that certain unit.

The Utilised Uptime metric can be used to

read out various information about a unit.
A low Utilised Uptime indicates a low
utilisation of the time the unit actually has
been available for production. This shows a
possibility to increase the production of the
unit by only increasing the utilised time,
without increasing the available time of the
unit or reducing the unit downtime. In fact,
if the availability increases and the
production time is constant, the Utilised
Uptime will decrease. A Utilised Uptime of
100% indicates that all available time has
been utilised for production. This means
that both the available time and the utilised
time have to be increased in order to
increase production.

6.1.3 UTILISED UPTIME

To clarify the relation between Overall
Utilisation and Availability, the Masters
thesis writers came up with a metric
referred to as Utilised Uptime, which is the
ratio between Overall Utilisation and
Availability (see Equation 26). This ratio
shows the percentage of the uptime that is
used for primary production. The metric
Utilised Uptime will highlight the
difference between available time and
utilised time, which is an unutilised time
share and therefore an area of possible
improvement.

6.1.4 MTBF & MTTR

The two metrics Mean Time Between
Failures (MTBF) and Mean Time To
Repair (MTTR) are two useful measures
for indicating asset reliability and the
quality of the maintenance work. Since
they both represent a mean time, the
metrics will give the average time between
the failures and the average time to get the
asset back in working condition. But none
of the metrics will give the distribution of
the failures or the time consumed to repair
the unit. This is critical information for the

Plants within the mining industry are often

battling with trying to increase their
equipment availability by improving the
asset reliability and the maintenance
quality. This is often done through the
updating and changing of parts more
frequently than required by the equipment

Utilised Uptime

Overall Utilisation
Availability

Equation 26

53

CHAPTER 6 Discussion & Conclusion

site maintenance team in order to be able
to improve their asset reliability as well as
their routines. That is why MTBF and
MTTR should be used as indicators
trended over time together with a
systematic analysis of the metrics, as
discussed in section 6.3. Continuous logging
of downtime, stop location, cause of
downtime, etc. is important, not only to
make it possible to calculate MTBF and
MTTR but to also facilitate the analysis.
This type of logging has been automated by
the Masters thesis writers at MNC. Further
details of the stop reporting procedure can
be found in section 5.5 and 6.5.

Figure 30 The relationship between

frequency and total downtime for
unplanned stops in January 2012 at MNC
The Pain concept has been very well
accepted, both on site and at the head
office. The users highly appreciate the
possibility to view frequency of error and
downtime in one metric and one single
diagram. The Pain concept as well as other
parts of OPT will be implemented in newly
developed software to be introduced
company-wide. This can therefore be
regarded as one of the major achievements
in the project.

The over-time trending of the metrics

should be used when comparing the current
status with previous results to understand if
actions taken are improving the asset
reliability and the quality of maintenance.
The longer the time span reviewed, the
more accurate the metrics will be. It is
therefore preferable to analyse a time span
of 30 days rather than 7 days.

6.1.6 PAIN ANALYSIS

When introducing the new concept, Pain, it

is important to clarify how the metric is
computed so that no confusion arises. Most
importantly, the Pain does not represent
the total downtime in any sense but the
product of frequency (n) and sum of
downtimes, which makes Pain n times
greater than the total downtime. To not
confuse the user, Pain is presented as a
unit-less metric.

The Pain analysis has been developed by

the Masters thesis students. It provides the
user with an understanding of the
downtime situation and its distribution
between total downtime and frequency or
error. The usage of the Pain analysis saves
the user the often complex task of
combining frequency and downtime from
two separate graphs to find the most critical
error in the process.

The input to the Pain analysis is extracted

from downtime reports which are
performed only on crushing units; hence
the Pain analysis is limited to those units. If
a downtime reporting procedure would be
in place for any other unit, a Pain analysis
would be possible to perform for that unit.

It was discovered that frequency and total

downtime of an error often do not
correspond (see Figure 30). Sometimes the
two parameters, rather, are inverted, i.e.
when downtime is high, frequency is low
and vice versa.

Since the input to the analysis is drawn

from downtime reports created by

54

CHAPTER 6 Discussion & Conclusion

employees on site, the reporting has to be
done properly. It is crucial that the
reporting employee knows the process and
what to report, i.e. the root-cause to the
downtime and not the consequence of it.
The human involvement will create a
possibility of human errors in this otherwise
highly automatic system. It has to be
considered that errors can occur. To
minimise errors in the reporting, the
reporting employees shall be well educated.
To ensure this, a workshop was held with
the concerned parts on site.

be done if one of the components is found

to be of more importance.

6.1.7 SETTING TARGET VALUES

To enable some of the metrics to be
calculated, target values need to be
determined. This is the case when
calculating Overall Utilisation. It has to be
defined when the unit is performing
primary production. For instance, a crusher
might be defined to perform primary
production when its power exceeds 140 kW.
For other units, the limit can be defined as
a speed or a weight etc. This concludes that
the defined limit to a high degree decides
the calculated equipment performance.
Therefore it is highly important that
accurate limits are defined. If so, the result
will be truthful.

The internal document, Anglo American

Equipment Performance Metrics, not only
includes OEE definitions but also
downtime categorisation used to allocate
downtimes and facilitate tracking and
comparison between company sites. This
categorisation model is included in OPT
and in that way completely aligned with the
company standard. The former reporting
system was not aligned with the company
standard and its downtime categories. The
tracking was therefore not possible and has
been made possible through the Pain
analysis and OPT.

It is complex to set target values, especially

as parameters are ever changing. According
to the Masters thesis writers proposal in
section 5.1.6, it would be a good idea to set
targets based on equipment changes. It is
understood that small process equipment
changes are being performed frequently
and that targets cannot be changed as
frequently, therefore, the suggestion is to
review targets periodically. The user needs
to find an appropriate interval to review
the different targets, since it might not be
suitable to review all targets simultaneously.

When comparing Pain values between units,

one should be cautious and not compare
different time spans since the values most
often are higher for a longer time span.
Also, caution has to be taken when
determining an acceptable level for an
error. For instance, the downtime named
Shift change might always have the same,
relatively high, level due to a predefined
time dedicated for shift change and might
therefore not need as much attention. Of
course, the aim shall always be to decrease
downtimes, but one should be aware that
certain downtimes are more critical than
others.

When targets are modified, the results will

also change. For instance, if a unit is
observed to always have high Performance,
it performs close to its targeted rate, and a
setting change is done to make it perform
even better, the Performance value will
most probably change. The user needs to
be aware of this when changing targets and
analysing data. This is particularly
important when data prior to and after a
target change is compared, since the results
can change drastically when a target is
changed.

The Pain concept can be refined and

developed by giving either component a
factor to put more emphasis on it. This can
55

CHAPTER 6 Discussion & Conclusion

In the end, it is highly recommended not to
compare OEE values and the values of its
included components between units and
sites. If such comparisons are not being
made, the exact numbers are not as
important as the relative numbers for a
single unit, which are of much more
importance and interest. The handicap of a
golfer can be used as an analogy when it
comes to only competing and comparing
results individually. Still, the ambition
should always be to set as accurate target
values as possible.

parameters for all units, but keeping the

measures to a minimum reduces the risk for
information overflow as well as makes it
easier for the user to read the output from
OPT. Full measures for all types of
equipment would demand an investment as
well since all equipment at MNC is not
prepared with measuring equipment. The
Masters thesis writers suggest that a proper
evaluation should be performed to find out
if there are any missing measure points
before any investments are carried out.

6.1.8 EQUIPMENT CLASSIFICATION

There are four different classifications of
the equipment in the calculation model.
The
reasons
for having
different
classifications of the equipment are two.
Firstly, all equipment is not equally
complex and does not require the same
detailed
monitoring.
Secondly,
all
equipment does not have the same
technical set-up and possibilities to
measure all parameters. However, there
will always be a demand for measuring all
Table 11 Classification of equipment

Classification

Equipment

Circuits

Comminution
units

Supporting
equipment
(with available
measures)

Supporting
equipment
(without
available
measures)

Details

OEE
Overall
Utilisation
Performance
Quality
Availability
Utilised Uptime

OEE
Overall
Utilisation
Performance
Availability
Utilised Uptime
MTBF
MTTR
Pain

Overall
Utilisation
Availability
Utilised
Uptime

Availability

56

CHAPTER 6 Discussion & Conclusion

users so that they could start using OPT
immediately. The tool has received very
good feedback from the users at MNC as
well as the senior team at the Head Office
in Johannesburg. The plan is for MNC to
use the OPT prototype and provide
feedback to the process control team in
Johannesburg, which is currently working
on a new software platform that will use
some parts of OPT. The fact that Anglo
Platinum will use parts of the project
proves that the outcome is practically
useful. Hopefully, this project will fill an
existing
gap
in
the
productivity
improvement work within the organisation.

6.2 THE OVERALL

PRODUCTIVITY TOOL
(OPT)
During the early
parts of the
project,
the
Masters thesis
writers
developed
a
small scale OPT Figure 31 Project
prototype to test phase two - develop
the
calculation
model with production data from MNC.
The idea of the OPT prototype was to learn
as much as possible about the
characteristics of the process and test the
calculation model as well as the coding of
the software. It was beneficial to run the
calculation model at an early stage in the
project since it gave the possibility to refine
it and get feedback from the process reality.
The learning curve was steep for the
process knowledge but even steeper for the
art of coding. Since the two Masters thesis
writers
are Mechanical engineering
students and not Software engineering
students. Throughout the project, the OPT
prototype
was
constantly
under
development, where module after module
was tested and added to the code. This gave
a thorough understanding of the dynamics
of the code. The coding could definitely
have been done differently if it had been
done by professionals from the beginning.

Develop a
tool

The development of a suitable way to

present the OPT output has been a long
iteration process. It was a balancing act to
keep it simple and clean while still
providing the user with enough, and the
right, information to enable the user to
perform an analysis and make accurate and
valuable conclusions. There is an infinite
amount of information that could be
presented in OPT, but the Masters thesis
writers have been very selective in the
decision on what to present and what to
leave out. The information in OPT is
presented on two different ways - overview
graphs and detailed data. The overview
graphs are to be used to get a quick
overview over the current status, while the
detailed data can be used for more detailed
systematic analysis. This applies to all the
metrics in OPT, i.e. OEE, Availability,
Utilised Uptime, MTBF & MTTR as well
as Pain analysis. The presentation of data in
OPT is consistent, which is important since
it speeds up the user learning curve as well
as facilitates the analysis of large amounts
of data.

The overall concept is well aligned with the

company standard of metric definitions,
which helps to lower the learning curve for
the user of the tool. OPT was developed
with a product development approach and
hence customised for the end users and
their requirements.

OPT is built in Visual Basic Editor and the

user interface is Microsoft Excel. There are
several benefits from this. OPT extracts
data from the process database PI and

At the end of the project the final OPT

prototype, as well as the OPT guidelines
and manual, were handed over to the end
57

CHAPTER 6 Discussion & Conclusion

performs calculations according to the
calculation model and presents the results
in Microsoft Excel automatically. Microsoft
Excel is a very common software, which
means that most of the users already are
familiar with the interface and are capable
of using OPT. It will be easy for the more
advanced users to make changes and
amendments to the code, but this has been
restricted to only certain users to avoid
mistakes and corruption of data. For MNC,
the use of OPT will not incorporate any
investments since Microsoft Excel is
already a part of the company software
package.

The Five Whys is an internationally

recognized systematic problem solving
methodology that has a proven record of
finding root causes. The Five Whys is
already implemented in the organization as
the main problem solving methodology. It
has therefore been incorporated in the
OPT method.
To further facilitate ease of use for the
users of OPT, the Masters thesis writers
have developed structured guidelines to
follow when working with OPT. The
guidelines consist of three parts; the OPT
Manual, the OPT Meeting procedure and
the OPT Action list. The OPT Manual is a
complete guide on how to use the tool with
examples of how to interpret various results.
The OPT Meeting procedure proposes a
structured way of holding a meeting
focused on OPT and its outcome. The OPT
Action list is a document to capture and
keep track of actions that have evolved
from analysing the OPT output.

The dry section at MNC consists of five

production areas, 102, 401, 405, 406 and 407.
All these areas are covered by OPT to get a
comprehensive view of the dry sections
productivity. OPT can be extended to cover
all process areas at MNC to get an
aggregate view.

6.3 OPT METHOD

The OPT Manual will most likely be used

during the introduction period of OPT. The
manual is a good guide for someone who
has not previously worked with OPT and
therefore is not familiar with all the metrics.
The manual should also be used whenever
a new problem is detected since it
addresses different ways to interpret OPT
output. However, OPT is developed to be
so user-friendly and intuitive that no
manual is necessary, so the intention is that
the manual should not be needed
constantly.

The OPT method

is based on three
parts,
User
expertise, the 5
Whys and the
OPT guidelines.
They found the Figure 32 Project
basis for how OPT phase three - design
should be utilised to gain as much valuable
output from it as possible.

Design a
method

The OPT users background knowledge of

the process is the key to understanding the
information presented in OPT. It has been
assured that the intended users of OPT at
MNC have the required knowledge. If this
requirement is not met by the user, the
result will most certainly not be as
satisfying as is could be.

The meeting procedure document was

developed to create a focused meeting with
the aim to analyse and find root causes to
problems as well as follow up on issued
actions. The predefined procedure will
hopefully guide the meeting participants
through the meeting and help to keep the
meeting productive and not too time
consuming.
58

CHAPTER 6 Discussion & Conclusion

The main reason for using an action list as a
supportive technique when working with
the tool is that the actions shall be
documented and it should be stated who is
responsible for what. The emphasis should
be put on analysis of outcome and followup of actions taken, since those two areas
tend to sometimes be neglected at MNC.

OPT belong to the engineering and

technical teams. This is considered to be a
successful combination of users since skills
from different departments are important
for getting everyone focused on the most
critical problems. The cross functional
collaboration around the tool will hopefully
help to increase the general cross functional
collaboration in the organisation. It has
been observed by the Masters thesis
writers that an increased cross functional
collaboration is possible and is therefore
advisable. An improved cross functional
collaboration will create a common focus in
the organisation and help the employees to
reach their goals and at the same time
reduce the risk for dual work.

When OPT and the OPT Guidelines were

handed over to the organisation, the
manual incorporated in the guidelines was
highly appreciated by the organisation,
since the Masters thesis writers were
leaving the site upon project finalisation.
The users now have the possibility to
further train themselves in using OPT as
well as to train new users. It will also work
as a support if something with OPT is not
working properly.

What has not been done is a proper test

and evaluation period, similar to what was
carried out for the calculation model and
tool. This is currently carried out by the
organisation itself and the end users of
OPT. A proposal for an evaluation project
by the Masters thesis writers is under
development.

There will always be a need to analyse the

OPT outcome since merely reading the
numbers cannot provide any complete
answer. The goal with the OPT analysis is
to identify and eliminate root-causes to
encountered problems that affect the
productivity. The analysis of metrics
displayed in OPT are suggested to be
carried out in two major ways and can be
summarised as follow:

6.4 OEE FOR A GENERAL

SINGLE STREAM PROCESS
During the pre-study, the Masters thesis
writers developed a general model for
calculating OEE in a single stream process.
The aim was to keep the model separated
from a specific site or company. The
process to develop the model gave good
knowledge in the subject, which helped
later during the development of the final
calculation model customised for MNC.
The general model has not been used in
OPT because the organisation standards
had to be considered. However, the Quality
calculation developed by the Masters
thesis writers is used both in OPT and in
the general calculation model since there
was a gap in the definitions created by the

1. From metrics to process. If there is

noticeable change in metric values,
find reasons in the process.
2. From process to metrics. If certain
changes are being performed in the
process, investigate if the metric
values are changing.
The end users were identified during the
time at MNC and the reason for using them
is their good process knowledge as well as
their cross functional positions where they
can
exchange
valuable
information
between their respective departments. The
employees chosen to be the main users of

59

CHAPTER 6 Discussion & Conclusion

organisation, which prevented the use of
the proposed Quality definition.

6.6 RESEARCH QUESTIONS

AND ANSWERS

The model suitable for a general single

stream process was tested on historical data
and was found to be working very well. It
would be interesting to test it in another
single stream process, for example in a
different industry such as the paper and
pulp industry.

This section will present the answers to the

Research Questions.
1. How can a method be developed to
define and rank process units critical to
productivity in a comminution process?
Firstly, the process needs to be
completely understood by the person
developing the method. Both interprocess relations and individual unit
functions have to be mapped and
comprehended. This should be done
in order to identify critical parts of
the process.

6.5 CRUSHER AND MILL STOP

REPORTING PROCEDURE
The new automatic stop reporting
procedure developed by the Masters thesis
writers has created a way for the
downtimes to be allocated and categorised
according to the Anglo American
Equipment Performance Metrics downtime
categories. This facilitates tracking and
comparison
of
downtimes
between
company sites, which is important in large
businesses.

Secondly, there is a need for a

thorough understanding of the
organisation running the operations.
Thirdly, a measure of productivity
has to be defined in order to be able
to evaluate the productivity of the
process
units.
OEE
(Overall
Equipment Effectiveness) is such a
measure. It gives an inclusive view of
the value added by the unit since it
includes three measures (availability,
performance and quality).

Beyond the company-wide standardisation

benefits, it also facilitates the analysis of
downtimes and errors on site since the
reporting is being performed daily, instead
of once a month as before. This shortening
of lead-time has resulted in a process where
downtimes can be investigated very soon
after their occurrence which helps minimise
their negative impact on the process.

Fourthly, which units to include in

the ranking need to be defined. The
selection of units can be done based
on the knowledge assimilated in the
previous steps.

Previously, the downtime table was created

by
manually
entering
downtime
information and manually categorising the
downtimes. In the new procedure, a script
draws data from the PI-database and
organises it into a downtime table. The new
downtime reporting procedure can be
argued to be more robust since it has
eliminated several manual steps.

Fifthly, based on the understanding

of the operations, critical process
parameters have to be determined.
Every
single
unit
within
a
comminution process has certain
parameters to address when looking
at productivity. Among those
parameters, some are more critical to
productivity than others. These have

60

CHAPTER 6 Discussion & Conclusion

to be identified and will be used
further on in the ranking.

The Performance will be maximised

when the unit is running better than,
or as close to its target rate, as
possible. To achieve this the unit has
to receive a satisfactory and
continuous feed, run with optimal
settings and be in good condition.

Sixthly, a rating based on the critical

parameters should be developed. The
rating has to take in to account the
different criticality of the parameters.
For instance, safety shall have the
highest
criticality
among
the
parameters.

The Quality will be maximised when

the unit is producing the right particle
size, i.e. minimising the deviation
from target size. The actual particle
size will be dependent upon the
quality of the feed, the settings of the
unit and the condition of the unit.

To keep the method aligned with the

current operations the parameters
within it need to be periodically
reevaluated.
2. How should OEE numbers be calculated
in a comminution process?

4. How can OEE be used as a performance

measure of equipment and process
performance?

The traditional OEE calculation was

developed for the manufacturing
industry and is therefore not suitable
for a comminution process. Several
changes have to be made to suit a
comminution
process.
The
calculation model developed to suit
this particular process is presented in
section 5.4 OEE for a General Single
Stream Process. Given that the
required data is available, this
method should be suitable in a
general case. The major difference
from the general OEE calculation is
the new way to define Quality, which
is customised for a comminution
process.

Since OEE includes three measures,

i.e. availability, performance and
quality, it is a comprehensive
performance measure in comparison
to single-parameter measures.
It is highly important that the target
KPI of each unit is established based
on the conditions of that particular
unit and that the targets are being
reviewed on a regular basis. It is
important to note that the OEE of a
unit shall not be compared to OEEs
of other units, but only to itself. This
is crucial since the conditions and
target definitions between units may
differ.

3. Which factors in the process chain are

more critical to productivity according
to the OEE method?

5. How can a high OEE help to improve

SHE (Safety, Hygiene, Environment)?

To achieve a high OEE, the included

metrics must all be high. Overall
Utilisation will be maximised when
the unit has a high running ratio, i.e.
few stops. This will be facilitated by
the good condition of the unit, which
can be assured through high quality
maintenance.

High OEE measures imply a well

running plant. This facilitates the
planning of scheduled stops and most
definitely
results
in
fewer
breakdowns. A process with few
unscheduled stops, i.e. a large
proportion of scheduled stops, is a
safer process than a process with a
61

CHAPTER 6 Discussion & Conclusion

large amount of unscheduled stops.
This is the case since scheduled
maintenance gives the opportunity to
plan the maintenance actions and
creates better conditions for the
performance of safe operations.
Hence, a high OEE creates
opportunities to improve SHE.

the procedure followed in problem solving

processes cannot be refined or improved
when evaluated because there is no proper
documentation. In addition to this, it is
difficult to train new people because there
is no database with information on the
problems encountered in the plant and how
they were resolved in the process. For
instance, the 5 WHYs method is frequently
mentioned as the correct method to follow,
however, no documentation has been
presented on how it had been used to
resolve problems on the plant. From the
outsider- it does not seem to be used to the
same extent as planned.

6. How can measuring OEE help to

improve productivity?
The measuring will not improve
productivity directly but measuring
individual OEEs of the units in the
process will help to identify where in
the process bottlenecks exist and will
therefore highlight possibilities for
improvement.
A
successful
elimination
of
the
identified
bottlenecks will result in an
improvement in OEE and can
consequentially give a productivity
improvement.

Clear problem solving procedures should

be developed and communicated to the
employees who are intended to master and
apply the methods. In cases where
education is required to use the methods, a
concerted effort should be applied to
provide it. The existing problem solving
method (the 5 WHYs) is a suitable method
which can help to eliminate root causes. A
proper follow-up of the usage of the
communicated method should be done.

6.7 OBSERVATIONS
One of the reasons for spending a
considerable period of time on site was for
the researchers to observe the day-to-day
activities and gain a greater understanding
of the operations. Various observations
have been made during the time spent on
site. Only those that were deemed
important for plant productivity are
reported here.

FOLLOW-UP
The plant has done well in following up on
most of the issues that have arisen. It is
commendable for a plant with such a large
capacity to carry out most of the follow-up
tasks as they do. However, the
documentation and formal report back on
the actual effects of performed process
changes targeted for follow-up on tasks that
need a review is not stringent. This leads to
failure to attend to some of the cases
targeted for follow-up. If the records and
report backs do not capture the follow-up
information and the effects of the change,
the plant would take it for granted that
follow-ups are done continuously even
though some of the key matters are not
getting any attention. This can be the case

PROBLEM SOLVING
Observations have been made regarding
the problem solving procedures in the
operations. Although many tasks in the
operations involve problem solving, there
seem to be no defined structure and
documentation of procedures used in
recurring tasks. Granted, while most of the
employees have many years of experience,

62

CHAPTER 6 Discussion & Conclusion

for some major process changes such as
changing the liner in a crusher or adjusting
the crusher gap settings.

of efforts because the divisions will have a

better insight into the activities taking place
in other divisions at the same time. Further,
it will work as a learning opportunity for
the persons not directly involved but well
informed. In that way they can gain a
greater understanding of the work of other
divisions.

The responsibility for follow up should be

shared throughout the entire organisation.
When a particular recommendation for
process changes is made, a person should
be assigned to implement proper
evaluation and follow-up. This will make it
easy for all involved to understand the
effects of process change

Although a lot has been done to promote

communication and cooperation there is
still a divide between divisions. There is still
a divide in reporting structures and
development of tasks which leads to
duplicated efforts. The communication and
cooperation between divisions should be
enhanced and the duplicated efforts should
be eliminated. This can be done by, for
instance, holding common meetings
involving only the people relevant for the
discussion. It is recommended to keep the
meetings action oriented and focused on
the dedicated subject in order to optimise
the numbers of persons involved and to
minimise time spent on meetings.

Follow-ups are not only important after

major process changes, but also for regular
tasks assigned to people. These can be
listed as action items for weekly meetings
and the tick box approach can be taken at
such meetings. This should be done to
capture a record which may be very useful
in providing insight on jobs that take a long
time and the reasons for such delays which
can feed directly into planning meetings.
Such a record can also provide information
on problematic areas of the plant which
may require more resources with time. This
can result in a better understanding of
recurring tasks as well as a learning
opportunity for the other meeting
participants.

JOB CARDS
The current system for maintenance relies
on the SAP to generate job cards which in
most cases works well. However, some
areas in the process do not have access to
job cards because some tasks lack a
dedicated functional code in the system.
The problem that arises from this is that
some jobs are performed without job cards
and therefore cannot be easily tracked.
Further, all existing job cards should be
continuously reviewed to keep them
aligned with the continuously changing
process.

INTER-DIVISIONAL
COMMUNICATION & COOPERATION
Throughout the organisation there is a
common drive to produce concentrate as
effectively as possible and to maintain the
plant in a good operational state. This is
clearly visible even for external observers
like the researchers who prepared this
report. However, the plant is fairly large
and it takes a long time for information to
reach all the relevant people in various
sections of the operations. There is an
opportunity to implement information
structures that can help visualise
information between divisions of the
operation. This will also reduce duplication

GOVERNANCE
MNC has a defined structure in terms of
the
sections
of
engineering,
technical/metallurgy and production which
is commendable. It is also evident that
63

CHAPTER 6 Discussion & Conclusion

when repair work is required, teams from
all sections are involved, which reinforces
the team spirit that is present at the
concentrator. However, it has been
observed that due to the integration
between those teams, some areas of
responsibility for certain categories of
employees seem to be undefined. In a case
when a task does not require handling in a
routine manner, it can easily fall under no
ones area of responsibility and this can
create a problem. A good example of this
would be equipment failure due to an
unidentified problem. In this case it is
better for the maintenance division to focus
on the required repair work, while
production teams continue with production
tasks. This will minimise the impact of the
repair work on production and will also
ensure that the responsibilities for various
tasks are streamlined. It will also assist in
eliminating duplication of efforts on the
same task. This mode of operation can only
be achieved if all the sections have full staff
complements and all the teams are skilled
in specific tasks. It also requires a common
decision making platform and approach.

The outputs from the tool can help create

standards which can be formulated and
implemented in the organisation. Followups on matters arising from using the tool
can be structured and implemented with
buy-in from all divisions.

6.8 RECOMMENDATIONS
Recommendations for the outcome of this
project mainly concern the usage and
future development of the Overall
Productivity Tool (OPT). The development
of OPT should continue before the new
platform is completed, it is highly
recommended that the current users of
OPT continue to provide feedback on how
the tool is used and how it can be improved
by suggestions for improvements can be
incorporated in the follow up version. The
users are encouraged to thoroughly test the
different methods suggested by the
Masters thesis writers since these have not
been fully evaluated on a production plant.
Another important aspect of the methods is
that they should be tested by several
different users and not only the main users
to provide information on how user
friendly the tool is. It is crucial to do this in
order to receive feedback from experienced,
as well as new, users before further
development options proposed.

UTILISING OPT IN THE

ORGANISATION
In addition to the suggested general
improvements, a new weekly meeting
should be initiated; this should involve key
people from all divisions. The crossdivisional meeting participants should use
the tool developed in this Masters thesis to
create a continuous improvement forum.
This will allow plant personnel more
opportunities to communicate and resolve
plant communication problems seamlessly.
Having cross-divisional participants in the
meetings when applying the tool will
provide a good platform for tracking and
learning from actions taken which will lead
to an increase in productivity.

6.9 FUTURE RESEARCH

The Masters thesis writers have found
many interesting areas of research and
would like to propose a few subjects for
future research.

64

Evaluate and develop the methods

suggested in phase three of this project
since this has not been done in the
project due to the limited available time.

CHAPTER 6 Discussion & Conclusion

The setup for OPT in its current form

focuses on the comminution areas of
MNC. It is suggested that OPT should
be extended to all process areas at
MNC.

The calculation model could be tested

in a different industry such as the paper
and pulp industry or in a single stream
manufacturing plant.

The financial benefits of OEE and

other
performance
metric
improvements is an area where detailed
research is proposed. More detailed
background information on this would
be beneficial for an organisation when
deciding what investments to make in
order to improve process productivity.

The authors of this Masters thesis have

seen that it would be possible and
beneficial to integrate OPT with a
condition measurement system. This
can increase the maintenance quality in
the process which is beneficial for
improving productivity.

65

REFERENCES

REFERENCES
Ahlmann, H., (2002). From traditional practice to the new understanding: the significance of the
Life Cycle Profit concept in the management of industrial enterprises. In: IFRIM (International
Foundation for Research in Maintenance), Maintenance Management & Modelling conference.
Vxj 2002, Vxj, Sweden: International Foundation for Research in Maintenance
Anglo Platinium webpage (updated 2011), About Anglo Platinum (Electronic). Available at:
http://www.angloplatinum.com/about/about_sub/comp_profile.asp [Accessed August 15, 2012]
Bryman, A., and Bell, E., (2011). Business Research Methods (3rd ed.). Oxford: Oxford
University Press.
Crundwell, F., Moats, M., Ramachandran, V., Robinson, T., Davenport, W.G., (2011),
Extractive Metallurgy of Nickel, Cobalt and Platinum Group Metals, Elsevier, ISBN 978-008096-809-4
Dressler, S., (2004), Strategy, Organization and Performance Management. From Basics to Best
Practices, Universal-Publishers, ISBN 978-1-58112-532-0
Gulati, R., (2009), Maintenance and Reliability Best Practices, Industrial Press, Inc., ISBN: 9780-83112-211-5
Halachmi, A., "IntroductionPerformance Measurement is Only One Way of Managing
Performance" International Journal of Productivity and Performance Management:
Performance Measurement and Performance Management, Volume 54, Number 7,
2005. Emerald Group Publishing 2005.
http://common.books24x7.com.proxy.lib.chalmers.se/toc.aspx?bookid=13077
[Accessed November 2, 2012]
Hansen, R C., (2001). Overall Equipment Effectiveness A Powerful Production / Maintenance
Tool for Increased Profits. New York, Industrial Press, Inc. ISBN 0-8311-3138-1
Idhammar Systems Ltd., (2010), Implementing OEE Systems. [online] Bristol: Idhammar
Systems Ltd., Available at:
http://www.idhammarsystems.com/resource%20library/overview/whitepapers/ [Accessed
September 13, 2012]
Kato, I. and Smalley, A., (2011), Toyota Kaizen Methods: Six Steps to Improvement. Taylor and
Francis Group, Productivity Press. ISBN: 978-1-4398-3854-9
Kitco webpage (updated 2012), Live Charts (Electronic), Available at:
http://www.kitco.com/charts/livegold.html [Accessed October 25, 2012]
Lesshammar, M., (1999), Evaluation and improvement of manufacturing performance
measurement systems the role of OEE, International Journal of Operations and Production
Management, vol. 19, no. 1, pp. 55-78
Liker, J. K. and Convis, G. L., (2012). The Toyota Way to Lean Leadership: Achieving and
Sustaining Excellence through Leadership Development. McGraw-Hill. ISBN: 9780071780780

66

REFERENCES
Liker, J. K., and Franz, J. K., (2011), The Toyota Way to Continuous Improvement: Linking
Strategy and Operational Excellence to Achieve Superior Performance. McGraw-Hill USA.
ISBN: 978-0-07-176215-1
Ljungberg, ., (2000), TPM Vgen till stndiga frbttringar. Lund: Studentlitteratur. ISBN:
978-91-44-00837-0
Luotojarvi, M., 2009-04-09, Method and a system for improving the operability of a production
plant, US 2009/00940880 A1, United States of America.
Mahadevan, B., (2009), Operation Management: Theory and Practice, Prentice Hall, ISBN:
978-8-17758-564-3
Mail & Guardian webpage (updated 2011), Platinum's pluses and minuses (Electronic),
Available at: http://mg.co.za/article/2011-11-25-platinums-pluses-and-minuses [Accessed
October 25, 2012]
Maran M., Manikandan G., and Thiagarajan, K., (2012). Overall Equipment Effectiveness
Measurement By Weighted Approach Method. In: IMECS (International MultiConference of
Engineers and Computer Scientists), International MultiConference of Engineers and
Computer Scientists 2012. Hong Kong, 14-16 March 2012.
Merriam-Webster, Opt Definition (Electronic) Available at:
http://www.merriam-webster.com/dictionary/opt [Accessed November 30, 2012]
Mining Weekly webpage (updated March 14, 2008), Biggest platinum concentrator achieves
safety milestone at PPRust, Available at: http://www.miningweekly.com/article/biggestplatinum-concentrator-achieves-safety-milestone-at-pprust-2008-03-14 [Accessed November
30, 2012]
Nadler, D, Tushman, M., (1997) Implementing New Designs: Managing Organisational Change.
Managing Strategic Innovation and Change: A collection of readings, First edition, pp 596-606.
Oxford University Press, New York.
Nakajima, S., (1989). Introduction to TPM. Cambridge, Mass.: Productivity Press
Oxford Dictionaries, Definition of comminution (Electronic) Available at:
http://oxforddictionaries.com/definition/english/comminution [Accessed November 30, 2012]
Oxford Dictionaries, Definition of kaizen (Electronic) Available at:
http://oxforddictionaries.com/definition/english/kaizen [Accessed December 11, 2012]
Perrin, R., (2008), Real World Project Management: Beyond Conventional Wisdom, Best
Practices and Project Methodologies, John Wiley & Sons, ISBN: 9780470170793
The PGM Database, Metals Data (Electronic) Available at:
http://www.pgmdatabase.com/jmpgm/index.jsp [Accessed August 15, 2012]
Probert, T., (2011). Rocking around the clock in SA. African Review of Business and
Technology, February 2012, pp.56-57.

67

REFERENCES
Platinum Today webpage (updated 2012), Properties of platinum group metals (Electronic),
Available at: http://www.platinum.matthey.com/applications/properties-of-pgm/ (2012-08-15)
Prokopenko, J., (1987). Productivity management: A practical handbook. International Labour
Office, ISBN: 92-2-105901-4
Rich, N., Esain, A., Bateman, N., Massey, L., Samuel, D., (2006). Lean Evolution: Lessons
from the Workplace. Cambridge University Press.
Scheinberg, S. (2009). From Inner to Outer Integration of Refugees and Immigrants in
the Swedish Society: A Gestalt Intervention in Joseph Melnick and Edvin C. Nevis eds.
Mending the World: Social Healing Interventions by Gestalt Practitioners Worldwide, A
Gestalt International Study Center Publication, p. 206-240
Smith, R. and Hawkins, B., (2004). Lean Maintenance, Butterworth Heinemann ISBN: 978-07506-7779-0
Strauss, A., and Corbin J.M., (1998). Basics of Qualitative Research: Techniques and
Procedures for Developing Grounded Theory. Thousand Oaks, California: Sage.
Taylor Fitz-Gibbon, C., (1990). "Performance indicators", BERA Dialogues (2), ISBN 978-185359-092-4

68

ORGANISATIONAL MAP MNC 2012

Created by: Anton Kullh & Josefine lmgeran

APPENDIX I - ORGANISATIONAL CHART

PROCESS MAPS MNC 2012

Created by: Anton Kullh & Josefine lmgeran

APPENDIX II - PROCESS MAPS

II

PROCESS MAPS MNC 2012

Created by: Anton Kullh & Josefine lmgeran

III

PROCESS MAPS MNC 2012

Created by: Anton Kullh & Josefine lmgeran

IV

PROCESS MAPS MNC 2012

Created by: Anton Kullh & Josefine lmgeran

PROCESS MAPS MNC 2012

Created by: Anton Kullh & Josefine lmgeran

VI

PROCESS MAPS MNC 2012

Created by: Anton Kullh & Josefine lmgeran

Lynxx Cameras Positioning Map MNC 2012

Created by: Anton Kullh & Josefine lmgeran

APPENDIX III - LYNXX CAMERAS POSITIONING MAP

II

Job Card Inspection anf Maintenance Map MNC 2012

Created by: Anton Kullh & Josefine lmegran

APPENDIX IV - JOB CARD INSPECTIONS AND MAINTENANCE MAP

II

Print Screens from the Overall Productivity Tool (OPT)

Displayed area: Area 102

APPENDIX V - OPT PRINT SCREEN

II

Print Screens from the Overall Productivity Tool (OPT)

Displayed area: Area 102

III

OPT Manual
Developed by: Anton Kullh & Josefine lmegran, 2012

APPENDIX VI - OPT MANUAL

OPT MANUAL
GUIDELINES TO THE OVERALL PRODUCTIVITY TOOL (OPT)

HOW TO UPDATE CALCULATIONS PAGE 2

EXPLANATIONS OF OEE CALCULATIONS PAGE 3
EXPLANATIONS OF PAIN ANALYSIS PAGE 5
DEFINITIONS OF METRICS PAGE 6
EXAMPLES OF HOW TO INTERPRET VALUES PAGE 9

IV

OPT Manual
Developed by: Anton Kullh & Josefine lmegran, 2012

HOW TO UPDATE CALCULATIONS

1. Open the excel-file named OPT_areacode.
If a Security Warning appears, click on Options, in the bar just above the sheet (see Figure 1),
and choose Enable this content in the dialogue box. Click OK (see Figure 2).

Figure 1 Security Warning that might appear when launching OPT

Figure 2 How to enable macros

2. To update the OEE calculations for the previous seven days, go to the sheet OEE, click the
grey button named 7 days and wait until a dialogue box opens and confirms the update.

OPT Manual
Developed by: Anton Kullh & Josefine lmegran, 2012
3. To update the OEE calculations for the previous month, click the grey button named Previous
Month and wait until a dialogue box opens and confirms the update.
4. To update the Pain analysis for the previous month, go to the sheet named Pain and click the
grey button named Previous Month and wait until a dialogue box opens and confirms the
update.
Note that the Pain analysis for the previous month has to be updated before updating the
previous 7 days in order to display the stop table for the previous 7 days under OEE Table sheet.
5. To update the Pain analysis for the previous seven days, go to the sheet named Pain and
click the grey button named 7 days and wait until a dialogue box opens and confirms the
update.
6. OPT is now updated according to the dates displayed on top of each sheet under Start and
End.
Note that the document shall be saved before closing down.

OEE CALCULATIONS
Two sheets in the Overall Productivity Tool (OPT) concern OEE calculations; those are named
OEE and OEE Table.

OEE SHEET
The OEE sheet displays charts with the OEE for the circuit and the crushing unit in the
monitored area. It also displays charts with the Overall Utilisation, Performance, Quality,
Availability, and Utilised Uptime for all units included in the monitored area.
The charts are sorted in ascending order to visualise what units that currently have the lowest
Overall Utilisation, Availability and Utilised Uptime. The abbreviations of the units are explained
in the OEE Table sheet.
To the right of the charts boxes with explanations of the charts and their metrics are provided.
At top of the sheet, the monitored time intervals are displayed. For information on how to
update these and the tables to the current end time, see section How to Update Calculations.

OEE TABLE SHEET

The OEE Table sheet provides the calculations a transparency and can be used to get a more
thorough understanding of the charts displayed in the OEE sheet.
The OEE Table sheet displays the components of the OEE calculations categorised by unit type
(Crusher, Classifiers, Feeders, and Conveyors).

VI

OPT Manual
Developed by: Anton Kullh & Josefine lmegran, 2012
The definition of any displayed component can be viewed by hovering over the component name
(see Figure 3, where the pointer is hovered over Availability).

Figure 3 Information appearing when hovering over Availability

All displayed rates (except for Utilised Uptime) have a colour code which visualises the current
status. Green is for Satisfactory, yellow for Poor and red for Alarming.
Satisfactory

Poor

Alarming

The colour limits for the different units can be seen at the bottom of the sheet. The limits shall be
set based on the business targets of those values and should only be changed by the
administrator of OPT.
The cells with a grey background colour are target values which shall be changed if process
changes resulting in target changes are performed. The target values should be changed only by
the administrator of OPT.
All set targets shall be reviewed if the process has been changed in such way that the current
target parameters are invalid.
Parameters to be reviewed:

Target rates (tph)

Target particle size

Running definition limits for units

Primary production limits for units

VII

OPT Manual
Developed by: Anton Kullh & Josefine lmegran, 2012
At top of the sheet, the monitored time intervals are displayed. For information on how to
update these and the tables to the current end time, see section How to Update Calculations.

PAIN ANALYSIS
Two sheets in the Overall Productivity Tool (OPT)concern the Pain analysis; those are named
Pain and Stop Table.

PAIN SHEET
Pain is a way of visualising the combination of frequency and downtime of stops occurred in the
monitored area. The Pain sheet displays charts with the top 6 Pains in the monitored area.
The charts are sorted in descending order to visualise what stop reasons that currently are
causing the largest Pain. The stop reasons are labelled below each bar in the chart. The unit for
the y-axis is thousand minutes, however, pain is displayed as a unitless metric.
At the top of the sheet, above the charts, Mean Time Between Failures (MTBF) and Mean Time
To Repair (MTTR) are displayed for the units.
At the top of the sheet, the monitored time intervals are displayed. For information on how to
update these and the tables to the current end time, see section How to Update Calculations.
More detailed stop information, such as stop time, start-up time, duration of stop, comment and
downtime codes, can be found in the Stop Table sheet.

STOP TABLE SHEET

The Stop Table sheet presents detailed stop information drawn from stop reporting though the
PI database. The information that can be viewed is as follows:

Stop time: The time the unit stopped

Start-up time: The time the unit started up after the stop
Duration: The duration of the stop
Stop reason: The reason of the stop
Manually entered comment: Possible manually entered comment by stand-by official
Downtime categories
o Downtime sub-category code
o Downtime sub-category name
o Downtime category code
o Downtime category name
Scheduled/Unscheduled: Indicates if the stop was scheduled or not
The downtime categories are used in order to facilitate the allocation of stops in alignment with
Anglo American Equipment Performance Metrics Time Model.
At the top of the sheet the total numbers of stops and the total downtime in the chosen time
interval are displayed. At the extreme top of the sheet, the monitored time intervals are
displayed. For information on how to update these and the tables to the current end time, see
section How to Update Calculations.
VIII

OPT Manual
Developed by: Anton Kullh & Josefine lmegran, 2012

DEFINITIONS OF METRICS
The following metrics are used in OPT and has to be understood in order to utilise OPT as
effectively and correctly as possible.

OEE - OVERALL EQUIPMENT EFFECTIVENESS

OEE is a metric that displays how effectively a unit or operation is utilised. OEE is calculated as
the product of Overall Utilisation, Performance and Quality.
OEE = Overall Utilisation x Performance x Quality

OVERALL UTILISATION
The Overall Utilisation is the percentage of the total time that the unit is utilised for primary
production. It is the ultimate performance indicator of how total calendar time is utilised.
Overall Utilisation = Direct Operating Time / Total time
Direct Operating Time (T300): Time the unit is performing primary production activities
Total time (T000): Total time in chosen time interval (24/7)

PERFORMANCE
The Performance is the production rate at which the operation runs as a percentage of its
targeted rate.
Performance = Actual Production Rate / Target Production Rate
Actual Production Rate = Actual Production Achieved / Primary Production
Actual Production Achieved: Actual tonnes produced during chosen time interval
Primary Production (P200): Time equipment is utilised for production.
For time definitions, see Figure 4.

QUALITY
The Quality looks at the P80 particle size and shows to what extent the particles size is below the
targeted size. It compares the Actual Particle Size at a certain point in the process to the Target
Particle Size. The Quality is defined as the mean deviation above Target Size as a percentage of
the Target Size. This implies that all particles below target size results in zero in deviation. To
get the Quality and not the deviation, the ratio is subtracted from 1.
n

Quality = 1

Mean deviation from Target Size

= 1
Target Size

For time definitions, see Figure 4.

AVAILABILITY
IX

i 1

Deviation from target size

n
Target Size

OPT Manual
Developed by: Anton Kullh & Josefine lmegran, 2012
The Availability is the percentage of the total time that the unit is available for production
activities.
Availability = Uptime / Total time
Uptime (T200): Time the unit is available for production activities
Total time (T000): Total time in chosen time interval (24/7)
For time definitions, see Figure 4.

UTILISED UPTIME
The Utilised Uptime is the percentage of the available time that the unit is being utilised for
primary production.
Utilised Uptime = Direct Operating Time / Uptime

Figure 4 Anglo American Equipment Metrics Time Model

PAIN
Pain is calculated as the product of frequency of the error and total stop time caused by the
error.
Pain = Frequency of error x Total stop time caused by error

MEAN TIME BETWEEN FAILURES (MTBF)

The MTBF is the average elapsed time between failures of the unit.
MTBF = Uptime / Number of stops
Uptime (T200): Amount of time the unit is available for production activities
Number of stops (D000events): The number of downtime events occurred during the period of
time viewed
X

OPT Manual
Developed by: Anton Kullh & Josefine lmegran, 2012

MEAN TIME TO REPAIR (MTTR)

The MTTR is the average time required to repair the failed unit.
MTTR = Equipment Downtime Time / Number of stops
Equipment Downtime Time (D000): Downtime that renders the equipment inoperable
Number of stops (D000events): The number of downtime events occurred during the period of
time viewed

EXAMPLES OF HOW TO INTERPRET VALUES

The outcome of the Overall Productivity Tool can be analysed and interpreted in several
different ways. This section will explain some fundamentals when analysing the metrics. These
examples might not always be valid but can provide user with an idea of what information that
can be drawn from OPT.

5 WHYS

A general recommendation when analysing the outcome of OPT is to use

the method 5 WHYs. The goal is to find the root cause of the problem.
When the root-cause is found, a conclusion of what to do should be
drawn and an action to resolve the problem should be taken. If the
encountered problem is complex, a Ishikawa diagram can be used to find
multiple root causes (see Figure 5).

Figure 5 Ishikawa (or fishbone) diagram to help find multiple root

causes

OEE

OEE is not just a number; it can be up to four numbers the OEE, Overall
Utilisation, Performance and Quality. It is important to look into all the
factors when analysing an OEE number. If an OEE number found in the
OEE sheet is found to be of interest, it can be viewed in more detail in the
OEE Table sheet. There, all components of the OEE can be seen and
analysed individually.
XI

OPT Manual
Developed by: Anton Kullh & Josefine lmegran, 2012

LOW
OEE
HIGH
OEE

OVERALL
UTILISATION
LOW
OVERALL
UTILISATION

HIGH
OVERALL
UTILISATION

A low OEE indicates a non-effectively utilised unit or circuit. To help

increase the OEE, the factors included has to be known. The included
factors can be seen in OEE Table sheet. The components of a factor can be
seen when hovering over it.
A high OEE indicates an effectively utilised unit or circuit. Even though a
unit or circuit has a high OEE, it should not be neglected. A well
performing unit or circuit can provide information about how to run a
unit or circuit effectively. The user should learn from this and apply it on
other units and circuits.
The Overall Utilisation shows to what extend the unit has been utilised
for production. It is the ultimate performance indicator of how total
calendar time is utilized.
A low Overall Utilisation indicates a small proportion of Direct Operating
Time, which is when the unit is performing production activities. If the
unit has not been utilised, it could be due to internal issues or factors
outside of its boundaries, such as low feed. The Overall Utilisation can be
increased by extending the Direct Operating Time, i.e. the time when the
unit is actually producing.
A high Overall Utilisation indicates a large Direct Operating Time, which
is when the unit is performing production activities. A unit with high
Overall Utilisation can provide useful information on how this can be
achieved. The user should learn from this and apply it on other units.

The Performance shows the production rate at which the unit or circuit
runs as a percentage of its targeted rate. This means that the targeted
PERFORMANCE rate has a large influence on the achieved Performance; therefore, it is
highly important that the Target rate is carefully determined. Otherwise,
the Performance measure will be misleading.
A low Performance indicates a Production Rate far below the targeted
rate during the primary production time. This can be due to either a very
LOW
long production time or low production achieved. To increase the
PERFORMANCE
Performance, a higher achieved production has to be reached during the
production time or the same amount of production has to be reached in
shorter time.
A high Performance indicates that the production rate is close to the

HIGH
targeted Production Rate. A unit or circuit with high Performance can
PERFORMANCE provide useful information on how this can be achieved. The user should
learn from this and apply it on other units and circuits.
The Performance can exceed 100%. This will occur when the Production

PERFORMANCE Rate is greater than the Target Rate. This indicates that the target rate
>100%
has to be reviewed. If process changes have been made in such way that

XII

OPT Manual
Developed by: Anton Kullh & Josefine lmegran, 2012
the current target parameters are invalid, the Target Rate shall be
adjusted accordingly.

QUALITY

The Quality looks at the P80 particle size and shows to what extent the
particles size is below the targeted size. It looks at the mean particle size
deviation above the targeted size. This implies that all particles below
target size results in zero in deviation, hence 100% in Quality.

LOW
QUALITY

A low quality indicates that the mean particle size is far above the Target
Size. The actual particle size and mean deviation are displayed in the
sheet OEE Table. If the Quality is low, the downstream process might be
affected and it should be beneficial to look at the performance of the
downstream units.

HIGH
QUALITY

AVAILABILITY

A high Quality indicates a mean particle size below the Target Size. If all
particle sizes are below the targeted size, the Quality will be 100%. The
actual particle size and mean deviation are displayed in the sheet OEE
Table.
The Availability shows to what extend the unit has been available for
production. It does not have to be used during that time; however, it has
to be available. Availability has a strong connection to Overall Utilisation.
Those two metrics are complementary since they present the unit
running time in two different aspects. The Availability is in most cases a
larger number since it is including a broader span of time, i.e. all the time
the unit has been switched on, whereas the Overall Utilisation only
includes the time the unit has been performing production.
A low Availability indicates a large proportion of non-running time. The

LOW
Availability can be increased by extending the unit Uptime, which implies
AVAILABILITY reducing the unit downtime. For the crushing units, the downtimes can
be seen in the Stop Table sheet.
A high Availability indicates a large proportion of running time. This

HIGH
implies that the downtime and non-controllable time both are low. A unit
AVAILABILITY with high Availability can provide useful information on how this can be
achieved. The user should learn from this and apply it on other units.

UTILISED
UPTIME
LOW
UTILISED
UPTIME

To show the ratio between Availability and Utilised Uptime, a rate called
Utilised Uptime is displayed in OPT. The Utilised Uptime is the proportion
of the Available time that has been utilised for production.
A low Utilised Uptime indicates a low utilisation of the time the unit
actually has been available for production. This shows a possibility to
increase the production of the unit by only increasing the utilised time,
without increasing the available time of the unit or reducing the unit stop
time. In fact, if the availability increases and the production time is
constant, the Utilised Uptime will decrease.

XIII

OPT Manual
Developed by: Anton Kullh & Josefine lmegran, 2012

HIGH
UTILISED
UPTIME

PAIN

A high Utilised Uptime indicates a high utilisation of the time the unit has
been available for production. An Utilised Uptime of 100% indicates that
all available time has been utilised for production. This means that both
the available time and the utilised time have to be increased in order to
increase production. A unit with high Utilised Uptime can provide useful
information on how this can be achieved. The user should learn from this
and apply it on other units.
The Pain charts show the top 6 highest Pains for the crushing unit.
A high Pain of a failure indicates one of the following:

HIGH
PAIN

MTBF

High frequency of failure

Large downtime caused by failure
Both high frequency and large downtime caused by failure
High Pains points out what areas cause most problems and should be
investigated and resolved.
Mean Time Between Failures (MTBF) shows the average elapsed time
between failures of the unit.

LOW
MTBF

A low MTBF indicates that the unit fails frequently. The aim is to
maximise the MTBF. Actions should be taken to investigate how to solve
the problem.

HIGH
MTBF

A high MTBF indicates that the unit does not fail frequently. A unit with
high MTBF can provide useful information on how this can be achieved.
The user should learn from this and apply it on other units.

MTTR

Mean Time To Repair (MTTR) shows the average time required to repair
the failed unit.

HIGH
MTTR
LOW
MTTR

A high MTTR indicates that the downtime per failure is long. The aim is to
minimise the MTTR. Actions should be taken to investigate how to solve
the problem.
A low MTTR indicates that the downtime per failure is short. A unit with
low MTTR can provide useful information on how this can be achieved.
The user should learn from this and apply it on other units.

XIV

OPT Meeting Procedure

Developed by: Anton Kullh & Josefine lmegran

APPENDIX VII - OPT MEETING PROCEDURE

XV

OPT Action List

Developed by: Anton Kullh & Josefine lmegran, 2012

APPENDIX VIII - OPT ACTION LIST

Action List for OPT
Action list to be filled when using the Overall Productivity Tool (OPT)

Number of outstanding tasks

3
RESPONSIBILITY ROLES

ACTION
Completed? Action

Responsible

Accountable

Consulted

DATE
Informed

Due date

OUTCOME

Completion date

Result

Conclusions

Natalie
Albert

Natalie
Dane

Dane
-

Ellie, Natalie

2012-12-01
2012-12-10

2012-11-30

John
Phillemon

Philip
Natalie

Dane, John

Dane

2012-12-01
2012-12-30

2012-12-02

Sensor damaged, sensor Point check sensors on

changed
conveyors

No

Check "Maintenance" on CR002

Create failure reason (see above)
Check possible measure error on
CV002
Improve rutines for maintenance
Develop condition monitoring for
Grizzly clogging

"Maintenance" referred Create a new more

to liner change
specified failure reason

Dane

Natalie

John, Philip

2013-01-14

Yes

Check spigot on cyclones

Felix

Dane

Herman

2012-11-25

2012-11-23

Different spigot size on

installed cylones
Change to similar sizes

Yes
No
Yes
No

XVI