Content uploaded by P. Knights
Author content
All content in this area was uploaded by P. Knights on Mar 18, 2015
Content may be subject to copyright.
Maintenance/
Engineering
Best-in-class maintenance benchmarks
in Chilean open-pit mines
P. F. Knights*,The University of Queensland, and CRCMining, Brisbane, Australia, and
P. Oyanader, IGT, Santiago, Chile
KEYWORDS: Maintenance, Benchmarking,
Performance, Efficiency.
Paper reviewed and approved for publication
by the Maintenance/Engineering Division of
CIM.
ABSTRACT
In 2001, the Catholic University of Chile
undertook a maintenance benchmarking study
of six open pit copper mines having mill capac-
ities varying between 18,000 t/d and 156,000
t/d, and collectively responsible for 58% of
Chilean copper production. This paper
describes the methodology used to conduct the
study, as well as the overall results. Key per-
formance indices were selected to measure the
effectiveness, efficiency, and development of
the maintenance process. Using a balanced
scorecard philosophy, these indicators were
divided into client satisfaction, financial, inter-
nal process, and learning and growth indices.
Maintenance was found, on average, to
be responsible for 44% of mine production
costs. Percentage planned maintenance of
equipment fleets was found to be low by world
standards, averaging 35%, 56%, and 44%,
respectively, for blasthole drill, shovel, and haul
truck fleets. Fleet availabilities were found to be
significantly influenced by the percentage of
planned maintenance achieved, while mainte-
nance cost per equipment was found to
decrease non-linearly with increases in percent-
age planned maintenance. Investment in tech-
nical training (including planned maintenance
practice) was found to be low by global stan-
dards. The paper concludes that fleet availabil-
ity and maintenance costs in Chilean open pit
operations could be considerably improved by
improving and/or developing maintenance
planning standards, and investing in training
personnel in planned maintenance techniques.
Introduction
In today’s capital-intensive mining indus-
try, maintaining high equipment availability,
utilization, production, and quality (or yield)
rates is vital to the financial performance of
mining companies. Equipment maintenance
and repair (referred to hereafter as mainte-
nance) play a vital role in assuring productive
capacity and equipment capability.
Due to a lack of publicly available bench-
marks, in order to identify opportunities for
improving current equipment management
strategies, mining companies should partici-
pate in maintenance benchmarking studies.
These studies involve comparing the mainte-
nance performance of mines operated by the
same owner (internal benchmarking); compet-
ing companies with a mutual interest in sharing
data (competitive benchmarking); or compa-
nies operating in different industry sectors gen-
erally acknowledged to be industry leaders in
maintenance (functional benchmarking). Since
equipment operating parameters can vary
markedly from one mine to another, an essen-
tial element of benchmarking studies is to find
ways for eliminating bias in the data to enable
accurate comparison, the so called comparison
of “apples with apples.”
In 2000, the Mining Council of Chile
(Consejo Minero a.g.) agreed to assist the Min-
ing Centre of the Catholic University of Chile in
undertaking a competitive benchmarking study
of key maintenance performance indices in
Chilean mines and concentrator plants. The
chief executives of the 17 companies that then
formed the membership of the Mining Council
were approached with regard to participating
in the benchmarking study; eight mines
decided to participate. All eight mines pro-
duced copper as the principal product. Since
one mine exclusively used underground extrac-
tion methods and another mine submitted data
in an aggregate manner that made it impossi-
ble to separate mine and mill performance, the
maintenance performance of the six remaining
mines employing open pit mining methods
were benchmarked. In order to maintain the
commercial confidentiality of the participants,
the names of these mines cannot be published.
In the study, they are referred to by letters as
mines A to F inclusively. The milling capacities
of these mines varied between 18,000 t/d and
156,000 t/d, and in 2000, the six mines were
collectively responsible for 58% of Chilean
copper production.
Study Objectives
The objectives of the benchmarking study
were fourfold:
•Ascertain the relative importance of main-
tenance as a percentage of mine production
costs;
•Identify the leaders in mine maintenance
performance and determine their associ-
ated best-in-class performance indices;
•Identify global improvement opportunities
for the Chilean mining industry; and
•Identify specific improvement opportunities
for participating companies.
This paper outlines the methodology used
and principal findings associated with the
benchmarking study.
Benchmarking Methodology
The first task in any benchmarking study
(beyond identifying the potential participants)
is to determine which performance indices to
compare (Watson, 1993;Camp, 1994). Consid-
erable differences exist in the number and type
of indices measured by companies. Some com-
panies measure only the basic indices, whereas
others have determined additional perform-
ance indices to be useful. It is therefore neces-
Paper 30 June/July 2005 CIM Bulletin TECHNICAL PAPER
Peter F. Knights
holds a Ph.D. in mining engineering
from McGill University, a M.Eng.in
systems engineering from the Royal
Melbourne Institute of Technology, and
a B.Eng. in mechanical engineering
from the University of Melbourne. He
was recently appointed Principal
Research Fellow with the Division of
Mining and Mineral Processing
Engineering at the University of
Queensland and program leader,smart
mining systems, for CRCMining in
Brisbane,Australia.
Patricio Oyanader
holds a M.Sc. degree in mining
engineering and a B.Sc. in industrial
engineering from the Catholic
University of Chile. He is a principal
consultant with the management
consulting firm IGT,in Santiago, Chile.
His current work focuses on process
and product innovation in a wide range
of industries, including mining,
manufacturing, energy, and
transportation.
June/July 2005 1
*This study was undertaken while Dr. Knights was employed at
the Pontificia Universidad Católica de Chile.
sary to determine which performance indices
are commonly used by all benchmarking partic-
ipants; and, to narrow the focus of the study,
which of these indices should be solicited for
inclusion in the benchmarking study.
Furthermore, companies have adopted
widely different definitions when it comes to
calculating performance indices. For example,
equipment availability may be calculated on
the basis of calendar hours or programmable
hours (calendar hours minus unavoidable
losses due to energy cuts and natural phenom-
ena) in a given time period. In any benchmark-
ing study it is essential to understand and
locate these differences, otherwise, apples will
be compared with oranges or bananas!
Maintenance Performance Indices
Maintenance is essentially a support
process, as distinct from a core production
process. As Figure 1 shows, the maintenance
system responds to mine operations, which in
turn must respond to the business environment.
Maintenance can be modelled as a transforma-
tion process that transforms resources (labour,
materials, spares and repairables, support
equipment and tools, knowledge, finance, and
external services) to a set of measurable out-
comes designed to maintain productive capacity
and capability assurance. These outcomes are:
equipment availability, reliability, maintainabil-
ity, product throughput, product quality (or
yield), compliance with safety and environmen-
tal legislation, and most importantly, company
profitability. If we accept that this model accu-
rately portrays the objectives of a maintenance
organization, three classes of performance
indices can be identified. These are: effective-
ness indices (how the achieved outcomes com-
pare with target outcomes); efficiency indices
(how efficiently the resources are utilized in
producing the outcomes); and development
indices (how much is being invested to improve
maintenance service levels and efficiency).
Effectiveness Indices
Much has been published concerning
maintenance performance indices (e.g. Camp-
bell, 1995; De Groote, 1995; Wireman, 1998;
Arts et al., 1998; Dwight, 1999; Duffuaa et al.,
1999). The key maintenance effectiveness
indices used in the mining industry are equip-
ment availability,reliability, and maintainability.
Reliability is often measured by mean time
between failures (MTBF), and maintainability
as the mean time to repair or restore service
(MTTR). At the time of undertaking the bench-
marking study (2001), only two of the six par-
ticipating mines were routinely measuring
these parameters. For this reason it was
decided that the study should focus on com-
paring fleet availabilities and forego the collec-
tion of fleet reliability and maintainability
data1.Fleet availabilities were solicited in the
form of available hours/(calendar hours – lost
hours), where lost hours include energy shut-
downs and stoppages due to storms and other
natural phenomena.
Safety statistics, namely accident rate and
severity (expressed as the number of lost-time
incidents and lost-time hours per million man
hours, respectively) were solicited for mainte-
nance personnel. This included both company
employees and contractors working on non-
capital projects. Production data, in terms of
tons of ore and waste mined per year, were
solicited from the participating mines. As well
as ensuring fleet availability and reliability,
maintenance contributes to a mine’s profitabil-
ity via its cost efficiency. Total annual mainte-
nance costs for the mines were solicited which,
combined with the production data, enabled
the unit costs of maintenance (US$/ton) to be
calculated. These costs include the direct and
indirect costs of maintaining the mobile equip-
ment and work shops. In one case (mine B), it
was not possible to separate the primary
crusher maintenance costs from aggregate
data submitted. This is an acknowledged
source of bias in the financial results.
Efficiency Indices
Efficiency indicators can effectively be
divided into two groups: financial and internal
process efficiency indicators. The following
financial indicators were solicited from the
companies participating in the study:
•maintenance costs as a percentage of min-
ing costs;
•annual cost of mine maintenance, including
the indirect costs associated with supervi-
sion and planning;
•labour costs as a percentage of annual
mine maintenance costs;
•spares and repairables cost as a percentage
of annual mine maintenance costs;
•contractor costs as a percentage of annual
mine maintenance costs; and
•overhead costs as a percentage of annual
mine maintenance costs.
Internal process efficiency indicators were
solicited for:
•the percentage of planned maintenance
(preventive, predictive, and programmed
major component replacement) carried out
on each equipment fleet (blasthole drills,
hydraulic and cable shovels, wheel loaders,
haul trucks, and auxiliary equipment);
•scheduled maintenance compliance (actual
planned hours vs scheduled planned hours);
•organizational efficiency in terms of the
ratio of supervisors to maintenance techni-
cians (the inverse of this ratio is known as
the span of control);
•investment in maintenance planning as evi-
dent from the ratio of planners to mainte-
nance technicians (planning clerks and
statisticians were also considered to be
maintenance planners); and
•efficiency in inventory practices as evident
from the stores turnover ratio (value of
spares and repairables consumed vs aver-
P. F. KNIGHTS and P. OYANADER CIM Bulletin June/July 2005
2CIM Bulletin ■Vol. 98, N° 1088
1It is interesting to note that, in the three years that have
passed since undertaking this study, most Chilean mines now
routinely report these variables. Considerable difficulty still exists
in benchmarking MTBF and MTTR values between mines, since
no industry standard has been developed for the measurement
of these variables. For instance, some mines include all stop-
pages, including PMs, in their calculation, while others incorpo-
rate only those events which they classify as failures.
Fig. 1. The maintenance resource transformation process.
Labour
External
services
Finance
Knowledge
Tools
Spares &
Repairables
Materials
Availability
Reliability
Maintainability
Throughput
Product
Quality
Safety
Environmental
Norms
Profitability
age annual on-hand inventory value) and
service level (percentage of requisitions that
the store is able to meet without delay).
An efficiency index commonly used to
compare the efficiency of maintenance across
different industry sectors is the annual cost of
maintenance as a percentage of replacement
asset value. Due to difficulties in determining
equipment replacement values, it was decided
not to include this index in the study.
Development Indices
The development indices solicited con-
cerned:
•the level of training commitment to mainte-
nance employees (both contractor and com-
pany workers), in terms of annual training
hours and investment per employee; and
•a breakdown of the educational status of
maintenance employees, in terms of the
highest level of educational certification
achieved.
The Balanced Scorecard Approach
The management of any large enterprise
involves the allocation of resources where they
will best create value. Different entities within
the same company may have competing objec-
tives requiring compromises to be sought. The
balanced scorecard, first developed by Kaplan
and Norton (1992), professors at the Harvard
Business School, is a means of balancing
opposing goals that also effectively links per-
formance indices with company strategy.
Although it has been successfully implemented
by many Fortune 500 companies, the balanced
scorecard has to date had limited application in
the management of mine equipment mainte-
nance (Tsang, 1998).
The balanced scorecard identifies four
classes of performance indices: client satisfac-
tion; financial performance; internal process;
and learning and growth (innovation and
development) indices. The first class of indices
equates to process effectiveness. The second
and third classes (financial and internal process
indices) are efficiency indices, and the last class
corresponds to the development indices. It was
therefore determined to structure and present
the results of the benchmarking study using the
balanced scorecard format.
The Structured Questionnaire
Benchmarking studies should ideally be
conducted by an independent team that visits
each of the participating mines. However, due
to financial restrictions, it was decided to con-
duct the benchmarking study via the use of a
structured questionnaire. A benchmarking
questionnaire was devised to capture the nec-
essary performance data. An annex to the
questionnaire was also prepared which speci-
fied the form in which the performance indica-
tors should be measured. A copy of the
questionnaire and annex was sent by email to
the maintenance managers of all participating
companies. In addition to maintenance per-
formance data, information was solicited con-
cerning mine operating parameters (altitude
above sea level, average haulage distances,
rock hardness, and abrasivity) and fleet charac-
teristics (fleet size, composition, and capacity).
Data Collection and Quality Assurance
Data collection and revision took place
over a period of six months. Five of the mines
participating in the study submitted data via
the questionnaire, and a visit was made to one
mine in order to directly collect data (mine C).
Quality assurance of the data was undertaken
by a three-step process. The completed ques-
tionnaires were scrutinized for missing, incom-
plete, or inexact data before being re-sent to
each participating company with a list of
queries. When the revised questionnaires were
re-submitted, a preliminary analysis of the per-
formance data was undertaken. This involved
constructing histograms of each of the per-
formance parameters, and in some cases, x-y
plots of key performance indicators vs mine
operating parameters. This process enabled
outlier data to be identified, resulting in addi-
tional queries being sent to the corresponding
companies. Following the receipt of explana-
tions and/or revisions for this data, a prelimi-
nary report was prepared and submitted to
each mine. This in turn generated queries from
some of the participating mines, resulting in
further revision of the data.
Results
The aggregate results of the benchmark-
ing study are shown in Table 1. All data are for
maintenance performance registered in 2000.
Discussion
Fleet Availabilities
Equipment availabilities are a complex
function of many factors. For example, haul
truck availability depends on planned mainte-
nance practices (preventive, predictive, and
programmed); mine altitude above sea level;
haul route distance and profile; haul route
maintenance practices; equipment type
(diesel-electric or mechanical) capacity and
age (cumulative operating hours); climatic
conditions and truck/shovel operating prac-
tices. In order to determine the best-in-class
availability for a specific mining fleet, it is
important to quantify the influence of each of
these variables.
Figure 2 shows the relationship between
truck fleet availability and percentage planned
maintenance for five of the six participating
mines. Fleet availabilities increase on average
by 1% for every 4% increase in planned main-
tenance over the range examined2.The
increase will be slightly higher or lower for spe-
cific mines and truck fleets when the other fac-
tors listed previously are taken into account. It
should be stressed that the availability values
recorded are for overall fleet availability. The six
June/July 2005 CIM Bulletin P. F. KNIGHTS and P. OYANADER
June/July 2005 3
2Recent non-published work by two mining companies operat-
ing in Chile (2004) suggests that, when planned maintenance
levels exceed 60%, there exists on optimum planned mainte-
nance level beyond which declining fleet availabilities can be
expected. The best fit curve in this case is not linear; a quadratic
fit may be more appropriate.
Fig. 2. Haul truck fleet availability vs percentage planned maintenance.
mines participating in the study operate mixed
haul trucks fleets, and while a new fleet may
have an availability in excess of 90%, the
availability of an older truck fleet will tend to
reduce overall fleet availability.
tenance represents between 40% and 50% of
mining costs (including operating and mainte-
nance costs but excluding general and admin-
istrative costs). The average value is 44%. This
highlights the importance of maintenance in
the overall cost structure of mining.
Unit Maintenance Costs
The cost of maintenance per ton extracted
is dependent on environmental variables and
operational factors that are specific to each
mine. For this reason, a comparison based on
this indicator alone cannot be made. The main
factor affecting unit cost is the size of the total
equipment fleet operated by the mine per tons
milled. An indicator that takes into account
fleet size (a function of haul route distance and
profile, equipment availability and utilization,
and operational factors such as blending) and
that can be better used for comparison is the
maintenance cost per equipment. However, this
indicator does not take into account cost dif-
ferences arising from the use of equipment of
different capacities.
Planned Maintenance
The low levels of planned maintenance
(preventive, predictive and programmed com-
ponent replacement) for the equipment fleets
indicate that company resources are being
used for reactive maintenance, resulting in
higher costs, lower fleet availabilities, and mak-
ing it more difficult to schedule proactive main-
tenance. Analyzing the data per company, it
can be seen that as planned maintenance
increases, the overall maintenance costs are
reduced as a result of a more efficient use of
resources (Fig. 3).
Figure 3 provides a useful empirical
means of estimating potential maintenance
cost savings through improvement in planned
maintenance practices.
Scheduled Maintenance Compliance
A graph prepared for fleet availability as a
function of scheduled maintenance compliance
showed strong positive correlations for shovel
and haul truck fleets. This indicates that not
only is the percentage planned maintenance
important, but it is important to meet targets
set by weekly maintenance plans.
Span of Control
Supervision ratios vary between 1/6 and
1/16 and were shown to be strongly correlated
with the use of contractors. Overall supervision
ratios are higher for companies making more
use of contractors, since not only must a con-
P. F. KNIGHTS and P. OYANADER CIM Bulletin June/July 2005
4CIM Bulletin ■Vol. 98, N° 1088
Table 1. Maintenance benchmark results
Performance Index Mines Inferior Superior Mean
Client (Effectiveness) Indices
Availability per fleet (%)1
Blasthole drills 6 64% 89% 78%
Shovels 6 68% 93% 84%
Wheel loaders 6 69% 88% 78%
Haul trucks 6 76% 89% 83%
Auxiliary equipment 6 63% 85% 79%
Accident frequency25 8.1 0 4.12
Severity index25 450 0 153
Financial Indices
Maintenance cost3per ton extracted (US$/ton)45 0.46 0.29 0.36
Maintenance cost as per cent of mining cost 5 40% 50% 44%
Maintenance cost per equipment (KUS$/equipment)55 486 684 549
Labour cost as per cent of direct cost of maintenance 5 0% 28% 14%
Spares and repairables as per cent of direct cost of maintenance 5 0% 49% 32%
Contractors as per cent direct cost of maintenance 5 16% 99% 50%
Overhead as per cent direct cost of maintenance 5 1% 7% 4%
Internal Process Indices
Planned maintenance per fleet (%)6
Blasthole drills 5 11% 65% 35%
Shovels 6 45% 67% 56%
Wheel loaders 6 19% 71% 40%
Haul trucks 5 25% 58% 44%
Auxiliary equipment 5 21% 66% 42%
Scheduled maintenance compliance 3 47% 91% 67%
Maintenance man hours as per cent production man hours7537% 54% 45%
Contractor man hours as per cent production man hours8520% 98% 56%
Overtime (%)950%5.3% 2.5%
Ratio of supervisors to technicians (span of control)10 41/6 1/16 1/8
Ratio of maintenance planners to technicians10 41/10 1/24 1/14
Stores turnover11 2 1.7 1.9 1.8
Stores service level11 380% 100% 92%
Learning and Growth Indices
Training cost as per cent total maintenance cost12 4 0.06% 0.39% 0.21%
Annual training hours per maintenance employee 4 27 75 55
Employees in training as per cent total employees 4 32% 100% 77%
Training investment per maintenance employee (US$/person) 4 71 870 497
Education status of the maintenance workers:13
University (%) 4 3% 24% 13%
Technical professional qualification (%) 4 4% 42% 27%
Technical training (%) 4 24% 52% 40%
High school (%) 4 13% 17% 15%
Primary school (%) 4 0% 17% 5%
1Availabilities recorded by mines C, D, and F are on the basis of calendar hours and do not take into account lost hours. Fleet availabilities for the remaining
companies should be marginally lowered for comparison.
2Safety indicators for mines A and D correspond to contractor personnel only,responsible for the majority of man hours.
3Cost data for mine B includes the primary crusher.
4Includes ore and waste (does not include ore from stockpiles). Maintenance and production costs are annual, do not include depreciation and amortization,
and are expressed in US dollars as of December 2000. The conversion factor used considers both Major Price Index (MPI) and Consumer Price Index (CPI):
Factor = 0.68*Variation CPI + 0.32*Variation MPI.
5Number of equipment includes: drills, shovels, loaders, haul trucks,and auxiliary equipment.
6Total maintenance hours are divided into two categories: planned maintenance (under schedule) and unplanned maintenance (reactive).
7Total man hours worked includes operations and maintenance for company and contractor personnel.
8Contracted man-hours include only maintenance tasks and exclude capital improvement projects.
9Mine A and mine D values correspond to contracted personnel representing 84% and 98% of total personnel, respectively.
10 Mine A and mine D values correspond to contracted personnel.
11 Average values for six months (July to December 2000).
12 Training delivered to own personnel, with the exception of mine D, which includes company and contracted personnel.
13 Mine D value corresponds to contracted personnel only.
Maintenance Costs
For the companies participating in the
study, representing 58% of the production of
Chilean fine copper in 2000, the cost of main-
tractor employ on-site supervisors,but the mine
must employ administrators to supervise the
contractors. As Table 2 shows (see subsequent
section), supervision ratios for the mines are
generally higher than those for other global
industries.This reflects the fact that the data in
Table 2 is strongly weighted by process plant
data, and the geographic dispersion of blast-
hole drills, shovels, and associated infrastruc-
ture requires that mines employ proportionally
more maintenance supervisors than is the case
for maintaining fixed process equipment.
Planning Ratio
Planning ratios vary between 1/10 and
1/24. This positions companies between the sec-
ond and fourth quartile results of the global
industry data shown in Table 2. Since mainte-
nance planning has been shown to be a key vari-
able in influencing equipment availability and
maintenance cost per equipment, an opportunity
exists for some of the participating companies to
increase their respective planning ratios.
Training Indicators
Training and development indicators for
maintenance staff show a low investment by
the participating companies. This is further
emphasized by the global industry comparison
in Table 2. Given that the majority of training
hours are dedicated to safety training, it can be
concluded that, on average in 2000, Chilean
mining companies were investing little in main-
tenance technical training. Considerable bene-
fits could be attained by investing more
resources in improving and/or developing
maintenance planning standards, and training
maintenance personnel in the use of planned
maintenance techniques.
Educational Level
The private sector mines participating in
the study employed a higher percentage of uni-
versity and technically qualified maintenance
professionals than did the state-run mine that
participated in the study. In addition, a trend
towards hiring better qualified maintenance
personnel was shown by the two recently com-
missioned mines participating in the study.
Comparison with Other Industries
In 1998, James Humphries, vice president,
manufacturing, for the engineering company
Fluor Daniel published a paper entitled “Best-
in-class maintenance benchmarks” in which
maintenance benchmarks were gathered for
148 global companies considered to be in the
top quartile of their industry according to earn-
ings and/or market share (Humphries, 1998).
The industries represented in this study
included mining and metals, refining and petro-
June/July 2005 CIM Bulletin P. F. KNIGHTS and P. OYANADER
June/July 2005 5
3For each company, percentage planned maintenance is calcu-
lated as the average of the percentage planned maintenance for
each equipment fleet weighted by fleet size. Mine B does not
consider the blasthole drill and auxiliary equipment fleets. Mine
C does not include the haul truck fleet.
Fig. 3. Annual maintenance cost per equipment vs percentage planned maintenance. 3
Table 2. Comparison of results with global industry benchmark results (Humphries, 1998)
Performance Index Quadrant
Fourth Third Second First
Client (Effectiveness) Indices
Availability per fleet (%)
Blasthole drills <78% 78 to 84 (78%) 85 to 91 >91%
Shovels <78% 78 to 84 (84% 85 to 91 >91%
Wheel loaders <78% 78 to 84 (78%) 85 to 91 >91%
Haul trucks <78% 78 to 84 (83%) 85 to 91 >91%
Auxiliary equipment <78% 78 to 84 (79%) 85 to 91 >91%
Accident frequency
Severity index
Financial Indices
Maintenance cost per ton extracted (US$/ton)
Maintenance cost as per cent of mining cost
Maintenance cost per equipment (KUS$/equipment)
Labour cost as per cent of direct cost of maintenance
Spares and repairables as per cent of direct cost of maintenance
Contractors as per cent direct cost of maintenance <8 8 to 19 20 to 40 >40 (50%)
Overhead as per cent direct cost of maintenance
Internal Process Indices
Planned maintenance per fleet (%)
Blasthole drills <65% (35%) 66 to 78 79 to 94 95
Shovels <65% (56%) 66 to 78 79 to 94 95
Wheel loaders <65% (40%) 66 to 78 79 to 94 95
Haul trucks <65% (44%) 66 to 78 79 to 94 95
Auxiliary equipment <65% (42%) 66 to 78 79 to 94 95
Scheduled maintenance compliance <15 15 to 35 36 to 70 (67%) >70
Maintenance man hours as per cent production man hours
Contractor man hours as per cent production man hours
Overtime (%)
Ratio of supervisors to technicians (span of control) <1/9 (1/8) 1/9 to 1/17 1/18 to 1/40 >1/40
Ratio of maintenance planners to technicians <1/25 (1/14) 1/25 to 1/59 1/60 to 1/80 >1/80
Stores turnover <0.5 0.5 to 0.7 0.7 to 1.2 >1.2 (1.8)
Stores service level <93 (92%) 93 to 96 97 to 99 >99
Learning and Growth Indices
Training cost as per cent total maintenance cost
Annual training hours per maintenance employee <40 40 to 69 (55) 70 to 80 >80
Employees in training as per cent total employees
Training investment per maintenance employee (US$/person)
chemical, paper, automotive, textiles, food pro-
cessing, rubber products, power, electronics,
consumer products, and chemicals and phar-
maceuticals. Of these companies, 75% were
North American, 16% Asian, and 9% Euro-
pean. Table 2 presents a comparison of
selected mine maintenance benchmark results
(the figures in parenthesis are the average
benchmark results) with the results published
by Humphries.
It can be seen that, in 2000, fleet avail-
abilities for Chilean open pit operations were
low by world standards (third quartile)4.In
addition, planned maintenance percentages
and annual training hours per maintenance
employee are in the lowest quartile. This rein-
forces the point made in the previous discus-
sion; in Chile,fleet availability and maintenance
costs could be considerably improved by
improving and/or developing maintenance
planning standards, and training personnel in
planned maintenance practices.
Conclusions and Recommendations
For the six open pit mines that partici-
pated in the study and collectively represented
58% of the copper production of Chile in 2000,
maintenance costs averaged 44% of mining
costs. This highlights the importance of the
direct cost of maintenance to the financial per-
formance of mines.
Percentage planned maintenance of
equipment fleets are low by world standards,
averaging 35%, 56%, and 44%, respectively,
for blasthole drill, shovel, and haul truck fleets.
Fleet availabilities were found to be sig-
nificantly influenced by the percentage of
planned maintenance achieved (preventive,
predictive, and programmed component
replacement). For example, haul truck fleet
availabilities were found to increase on average
by 1% for every 4% increase in planned main-
tenance over the range examined.
Maintenance cost per equipment was
found to decrease non-linearly with increases in
percentage planned maintenance. An empirical
relationship was determined that can be used to
estimate potential cost savings as a result of
improved planned maintenance practices.
The ratio of maintenance planning staff to
maintenance technicians varied between 1/10
and 1/24, indicating improvement opportuni-
ties for some of the participating companies.
Investment in technical training (including
planned maintenance practice) was found to
be low by global industry standards.
It is concluded that fleet availability and
maintenance costs in Chilean open pit opera-
tions could be considerably improved by
improving and/or developing maintenance
planning standards, and investing in training
personnel in planned maintenance techniques.
Acknowledgments
The authors would like to acknowledge
the assistance of the Mining Council of Chile
for supporting this study, as well as the main-
tenance and business improvement personnel
of the eight participating companies who gave
so generously of their time and knowledge.
References
ARTS, R.H.P.M, KNAPP, G., and MANN, L., 1998.
Some aspects of measuring maintenance per-
formance in the process industry. Journal of
Quality in Maintenance Engineering, 4, p. 6-11.
CAMP, R., 1994. Business Process Benchmarking.
American Society for Quality Control
(ASQC)/Quality Press, Milwaukee, 464 p.
CAMPBELL, J., 1995. Uptime: Strategies for Excel-
lence in Maintenance Management. Productivity
Press, Portland, Oregon, 192 p.
DE GROOTE, P., 1995. Maintenance performance
analysis: A practical approach. Journal of Quality
in Maintenance Engineering, 1, p. 4-24.
DUFFUAA, S., RAOUF, A., and CAMPBELL, J., 1999.
Planning and Control of Maintenance Systems:
Modeling and Analysis. John Wiley & Sons Inc.,
New York, 371 p.
DWIGHT, R., 1999. Searching for real maintenance
performance measures. Journal of Quality in
Maintenance Engineering, 5, p. 258-275.
HUMPHRIES, J., 1998. Best-in-class maintenance
benchmarks. AISE Steel Technology Magazine,
75.
KAPLAN, R. and NORTON, D., 1992. The balance
scorecard measures that drive performance. Har-
vard Business Review,January-February,p. 71-79.
TSANG, A., 1998. A strategic approach to managing
maintenance performance. Journal of Quality in
Maintenance Engineering, 4, p. 87-94.
WATSON, G., 1993. Strategic Benchmarking: How to
Rate Your Company’s Performance against the
World Best. John Wiley & Sons Inc., New York,
288 p.
WIREMAN, T., 1998. Developing Performance Indica-
tors for Managing Maintenance. Industrial Press
Inc., New York, 195 p.
P. F. KNIGHTS and P. OYANADER CIM Bulletin June/July 2005
6CIM Bulletin ■Vol. 98, N° 1088
4The availability figures in Table 2 correspond to those listed for
discrete processes by Humphries (1998).
