Reliability Centered Maintenance Book

1. Buttemlorth-lfe~ncmann
Lin;lcre House, Jordan Hrll, Oxford OX2 8DP
225 Wlldwvc~od
Avenue, Woburn, MA 01801-2041
A cttvtslon of Reed Educattonal and Professronal Publlrhtng Ltd
-
@
A mernbrr o f the Reed Elsevier plc group
OXFORD BOSTON JOIIANNESRURG
MtSLBOlJRNE NEW DELf-lI SINGAPORE
First published 1991
Reprnlted 1992, 1993, 1994 (twrce) 1995, 1996
Second edrtron 1997
Reprrntcd 1997
All rights rercived No past of thrs publrcat~on
in,l> be reproctucetl tn ,my niatelral foml (mcludlng
photocopytug or stortng In any medlum by electronic
Incan:, ntld whcthel or not transtently or tnctdentally
to ome other ue ot thtr publication) w~tho~tt
the
wlrtten permlsslon of the copyrrght holder except
In dccoiddn~e
wttli the provrhton of the Copyrtght,
Derrgns anct Patents Act 1988 or under the term5 of a
licence issued by theCopyright Lrcenslng Agency Ltd,
9
1
3 Ioteenfiam Court Road, London, England WI P 9HI:
Appltcattons for the copyr~ght
holder's wrltten permlssrotl
to reproduce any part of thls publtcation should be addressed
to the publishers
British Library Cataloguing in Publication Data
A catalogue record for this book is available
fiom the British Library
ISBN 0 7506 3358 1
For Edith
Typeset by the a~itho~
Printed ancl bot~nciIn (ircat Brrtatri by
Btddles Ltd, Gttrlctford and K~ng's
I,ynn

2. Contents
Preface
Acknowledgements
1 Introduction to Reliability-centredMaintenance
1.1 The changing world of maintenance
1.2 Maintenance and RCM
1.3 RCM: The seven basic questions
1.4 Applying the RCM process
1.5 What RCM achieves
2 Functions
2.1 Describing functioris
2.2 Performance standards
2.3 The operating context
2.4 Different types of functioris
2.5 How functions should be listed
3 Functional Failures
3.1 Failure
3.2 Functional failures
Failure Modes and Effects Analysis
What is a failure mode?
Why analyse failure modes?
Categories of failure modes
How much detail?
Failure effects
Sources of information about modes and effects
Levels of analysis and the information workrheet
5 Failure Consequences
5.1 Technically feasible and worth doing
5.2 Hidden and evident functions
5.3 Safety and environmental consequences

3. 5.4 Operationiil consequences
5.5 Non-operational consequences
5.6 Hidden failure consequences
5.7 Conclusion
6 Proactive Maintenance 1: Preventive Tasks
6.1 'Technical feasibility and proactive tasks
6.2 Age and deterioration
6.3 Age-related failures and preventive Inalntenance
6.4 Scheduled restoration tasks
6.5 Scheduled discard tasks
6.6 Failures which are not age-related
Proactive Maintenance 2: Predictive Tasks
Potcnttal fa~lures
and on-condit~on
rnalntenance
The P-F tnterval
The tcchn~cal
fea51b111tyof on-conclrtton tasks
Categortes of on co~ldit~on
techrilques
On-condttton ta&5 iome of the p~tfalls
Linear and non-llnear P-F curves
How to determine the P-F Interval
When on-cctndltron tasks are worth dotng
Selecting proactive tasks
8 Default Actions 1: Failure-finding
8.1 Default actions
8.2 Failure-finding
8.3 Failure-finding task intervals
8.5 The technically feasibility of failure-finding
9 Other Default Actions
9.1 No scheduled maintenance
9.2 Redesign
0.3 Walk-around checks
10 The RCM Decision Diagram
10.1 Integrating consequences and tasks
10.2 The RCM decision process
10.3 Completing the deciiion worksheet
10.4 Computers and RCM
11 Implementing RCM Recommendations
11.1 Implementation - the key steps
11.2 The RCM audit
11.3 Task descriptions
1 1.4 Implementing once-off changes
11.5 Work packages
11.6 Ma~ntenance
planning and control systerns
11.7 Reporting defects
12 Actuarial Analysis and Failure Data
12.1 The six failure patterns
12.2 Technical history data
13 Applying the RCM Process
13.1 Who knows?
13.2 RCM review groups
13.3 Facilitators
13.4 Implementation strategies
13.5 RCM in perpetuity
13.6 How RCM should not be applied
13.7 Building skills in RCM
14 What RCM Achieves
14.1 Measuring maintenance performance
14.2 Maintenance effectiveness
14.3 Maintenance efficiency .
14.4 What RCM achieves
15 i
i Brief History of RCM
15.1 The experience of the airlines
15.2 RCM in other sectors
15.3 Why RCM 2?
Appendix 1:Asset hiernrchres rind functiorz~l
hlock hugmtn.t
Appendix 2: Human error
Appendix 3: A rontin~~u~n
oj rzsk
Appendix 4. Condition monitoring
Bibliography
Index

4. Preface
Humanity continuesto depend to an ever-increasing extent on the wealth
generatedbyhighlymechanisedandautomatedbusinesses.We alsodepend
more and more on services such as the uninterrupted slipplyof electricity
or trains which nln on time. More than ever, these depend in turn on the
continued integrity of physical assets.
Yet when these assets fail, not only is this wealth eroded anct not only
are these services interrupted, but our very survival is threatened. Equip-
ment failurehas played apart in someof the worst accidents anct environ-
mental incidents in industrial history -incidents which have become by-
words, such as Arnoco Cadiz,Chernobyl, Bhopal and Piper Alpha. As a
result, theprocesses by which these failures occur and what must be done
to manage them are rapidly becoming very high priorities indeed, cspeci-
ally as itbecomes steadily more apparentjust Plow many of these failures
are caused by the very activities which are s~lpposed
to prevent them.
The first industry to confront these issues was the international civil
aviation industry. On the basis of research which challenges rnruiy of our
most firmly and widely-held beliefs about nraintenance, this inctustry
evolved acompletely new strategic frameworkforcnsilring that any asset
continues to perform as its users want it to perform. This framework is
known within the aviation industry as MSG3, and o~~tside
it as Reliabil-
ity-centred Maintenance, or RCM.
Reliability-centred Maintenance was developed over aperiod of thirty
years. One of the principal milestones in its development was a report
commissioned by the United StatesDepartment of Defense from United
Airlines and prepared by Stanley Nowlan and the late Howarcl Heap in
1978.Thc report provided a comprehensive description ctf the develop-
ment and application of RCM by the civil aviation industry. It forms the
basis of both editions of this book and of much of the work done in this
field outside the airline industry in the last fifteen years.
Since the early 19801s,the author and his associates have helped corn-
panies to applyRCM in hundreds of industrial locations around the world
work which led to the development of RCM2 for ind~~stries
other than
aviation in 1990.

5. xii Ke1iabilit)l-rentreti Mairztennnce P ~ ~ ~ ; E c P ...
X l l l
The firstcdition of thisbook (publishedin the UK in 1991andtheUSA
in 1992)provided a comprehensive introduction to RCM2.
Sincethen, the RCM philosophy has continued to evolve, tothe extent
that it became necessary to revise the first edition to incorporate the new
developments. Several new chaptershave been added, while others have
been revised and extended. Forernost among the changes are:
a more comprehensive review of the role of functional analysis andthe
definition of failed states in Chapters 2 and 3
a much broader and deeper look at failure rnodes and effects analysis
in the context of RCM, with special emphasis on the question of levels
of analysis ancl the degree of detail required in Chapter 4
new material on how to establish acceptablelevels of risk in Chapter 5
and Appendix 3
the acldition of more rigorous approachesto the determinationof failuire-
finding fask intervals in Chapter 8
more about the ilnplemcntation of RCM recommendations in Chapter
1I, with extra enlphasis on the RCM auditing process
more inforination on how RCM should - and should not - be applied
in Chapter 13,incl~icling
a inore con~prehensive
look at the role of the
RCM facilitator
new materia1 on the measurement of the overall performance of the
maintenance fitnction in Chapter 14
* a brief review of asset hierarchies in Appendix 1,together with a surn-
mary of the (oftenoverstated) roleplayed by functional hierarchies and
functional block diagrams in the application of RCM
areview of different types of human error in Appendix 2, together with
21 look at the part they play in the failure of physical assets
the addition of no fewer than 50 new techniques to the appendix on
condition morlitoring (now Appendix 4).
In the second impression of the second edition, the word 'tolerable' has
been substituted for 'acceptable' in discussions about risk in Chapters 5
and 8 anit 111Appendix 3, in order to align this book more with standard
tern~inology
used in the world of risk. It also includes further material on
the practicality of failure-finding task intervals in Chapter 8,and slightly
revised rnaterial on RCM irnplernentation strategies in Chapter 13.
The book is intended for maintenance, production and operation,
managers who wish to learn what RCM is, what it achieves and how it 13
applied. It will also provlde student5on buslness or management srud~el
courses with a comprehensive introctuct~on
to the fo~rnul;ltrotl
of strate-
gies for the management of phyl~cal
(as oppoc;ecl to f~nanci~ll)
d,ets
Finally, the book w11l be i~~valuable
for any student$ of ,my branch ot
engineering who seek a thorough undelstanding of the stdte-of-the-,irtIn
modem maintenance. It is designed to be read at three level3
Chapter 1is written forthose who only wish to review the key elements
of Reliability-centred Maintenance.
Chapters 2to 10describe themain elements of the technology of IICM,
and wiIl be of most value to those who seek no rnore than a reasonable
technical grasp of the subject.
Theremainingchaptersareforthose who wish toexplore RCM in rnore
detail.Chapter 11provide abrief s~~rnmary
of the key steps which must
be taken to implement the recorninendations arising fro111
RCM analy-
ses. Chapter 12 takes an in-depth look at the sometitnes contentious
subject of the relationshipbetween age arid failure. Chapter 13consid-
ershow RCM shouldbe applied,with emphasisontherole of the people
involved. After reviewing ways in which maintenance erfcctiveness
and efficiency should be measured, Chapter 14 describes what RCM
achieves. Chapter 15provides a brief history of RCM.
JOHN MOUBRAY
Lutterworth
Leicestershire
Serptember 1997

6. Acknowledgements
It has only been possible to write both editions of this book with the help
of a great ITlally people around the world. In particular, I would like to
record my continuing gratitude to every one of the hundreds of people
with whom I have been privileged to work over last ten years, each of
whom has contributed something to the material in these pages.
In addition, Iwould liketo pay specialtribute to anumber of people who
played alnajorrole inhelpingto clevelopandrefine the RCM philosophyto
the point discussedin this edition of this book.
Firstly, special thanks are due to the late Stan Nowlan for laying the
foundations forboth editionsof thisbook sothoroughly,both through his
own writings and in person, and to all his colleagues in the civil aviation
industry for their pioneering work in this field.
Special thanks are alsodue to Dr Mark Horton, for hishelp in develop-
ingmany oftheconceptsembodied inChapters 5 and8,andtoPeterStock
for researching and helping to co-author Appendix 4.
1am also indebted to all members of the Aladon network for their help
in applying the concepts and for their cot~tinuous
feedback about what
works and what doesn't work, much of which is also reflected in these
pages. Foremost among these aremy colleaguesJoel Black, ChrisJames,
Hugh Colnnan and Pan Hipkin, and my associates Alan Katchmar, Sandy
Dunn,TonyGeraghty,Frat Amarra, Phil Clarke,Michael Hawdon, Brian
Oxenham, Ray Peden, Simon Deakin and Theuns Koekemoer.
Anlong the many clients who have proved and are continuing to prove
that RCM is a viable force in industry, 1 am especially indebted to the
following:
Gino Palarchio and lion Thomas of Dofasco Steel
Mike IJ[opcraft,Terry Belton and Barry Camina of Ford of Europe
Joe Campbell of the British Steel Corporation
Vincent Ryan and Frank O'Connor of the Irish Electricity Supply Board
Francis Cheng of Hong Kong Electric
Bill Seeland of the New Venture Gear Cornpany
Denis Udy, Roger Crouch, Kevin Weedon and Malcolm Regler of the
Royal Navy
Don Turner and Trevor Ferrer of China Light & Power
John Pearce of the Mars Corporation
Dick Pettigrew of Rohn~
& Haas
Pat McRory of BP Exploration
Al Weber and Jerry Haggerty of Eastman Kodak
Derek Burley of Opal Engineering.
The roles played by Don Humphrey, Richard Hall, Brian Davies, Tom
Edwards,David Willson and the late Joe Versteeg in helping to develop
or to propagate the concepts discussed in this book are also acknowl-
edged with gratitude.
Finally, a specialword of thanks to rny farn~ly
firrcreating an environ-
ment in which it was possible to write both editions of this book, and to
Aladon Ltd forpermission to reproduce the RCM Information and Ueci-
sion Worksheets and the KCM 2 Decision Diagram.

7. 1 Introductionto
Reliability-centredMaintenance
1.1 The Changing World of Maintenance
Over the past twenty years, maintenance has changed, perhaps more so
than any other management discipline. 'The cl~anges
are cine to a huge
increase in the number and variety or physical arets (plant, e~liupmcnt
and buildings) which must be maintained throughout the world, rtluch
more coinplex destgns, new malntenancc techniyucs ,~nd
changing vrew
on rnamtenance organlsatlon anit respons~bllltte
Maintenance1salso reqpondrng to changmgexpect'ition These include
a rapldly growmg awareness of the extent to whtch ecjuipnlent iailitre
affects safety and the envtronment, a growing awaietles of the connec-
tion between mamtendnce and pioduct q~l;~lity,
'ind I I I L I C C I ~ I ~ ~
pieure
to achieve htgh plant availahilrty and to contarn cost5
The changes are testing attttucles 'mcl skills rn br;lnche 01 inciustiy
to the lim~tMamtenance people are having to .dopt ~orupletelynew
way of thtnking and acting, as engineels and as iitanciger At the 5 ~ m e
tlme the limttations of rnatntenance sy5tems itre becoming tncre&mgly
apparent, no matter how much they are computcrised
In the face of thts avalanche of change, managers everywheie arc
looking for a new approach to maintenance They want to avoid the false
starts and dead end5 which always accompany major upheaval Irl stecid
they seek a ,trategzc Jrunzen~ork
which ~yntlzecl~c~~
the new deveEo[>r7zents
latou coherentpattr~rn,
so dzut the] can evaluate them senrlbly n?tcfci[~p1y
those IlkeLy to he of nzost vnlric ro rhenz and rlzerr L owzpanzes
This book descrtbes '
1 philosophy whrch p1ovrtle5jubt such a ftaine-
work It is called Rel~ability-centred
Maintena~lce,
or RCM
If~tIS applied correctly,RCMtrandoj111sthe rcIattonhrp between the
undertaktngs wh~ch
use it, therr extsttng physical asset and the people
who operatedndmalntarn those assets It also enablesnew uett tobe put
Into effective service with great peed, confidence ,ind precrion
Thts chapter provtde abr~ef
introduction to RC'M, tarttng wrth a look
at how maintenance ha evolved over the pat fifty year

8. 2 Reliubiliv-centred Maintenance
Since the 1930's, the evolution of maintenance can be traced through
three generations. RCM is rapidly becoming a cornerstone of the Third
Generation, but this generation can only be viewed in perspective in the
light of the First and Second Generations.
The First Generation
The First Generation covers the period irpto World War TI.In those days
industry was not very highly mechanised, so downtime did not matter
much.This meant that the prevention of equipment failurewas not avery
high priority in the minds of most managers. At the same time, most
equipment was simple and much of it was over-designed. This made it
reliable and easy to repair. As a result, there was no need for systematic
maintenance of any sort beyond simple cleaning, servicing and lubrica-
tion routines. 'The need for skills was also lower than it is today.
The Second Generation
'Things changed ctr~t~natically
during World War TI. Wartime pressures
increased the demand for goods of all kinds while the supply of industrial
manpower dropped sharply.This led to increased mechanisation. By the
1950's ~nachines
of all types were more numerous and more complex.
Industry was beginning to depend on them.
As this dependence grew, downtime came into sharper focus. This led
to the idea that equipment failures could and should be prevented, which
led in turn to the concept of preventive maintenance. In the 1960's, this
consisted mainly of equipment overhauls done at fixed intervals.
The cost of maintenance also started to rise sharply relative to other
operatingcosts.This ledto the growthof mintenarzceplanning and confrol
systems. These have helped greatly to bring maintenance under control,
and are now an established part of the practice of maintenance.
Finally, thc amount of capital tied up in fixed assets together with a
sharp increase in the cost of that capital led people to start seeking ways
in which they could rnaximise the life of the assets.
The Third Generation
Sirlce the mid-seventies, the process of change in industry has gathered
even greater momentum. The changes can be classified under the head-
ings of new expectations, new re~earch
and new techniques.
New expectations
Figure 1 .I shows how expectations of maintenance have evolved.
Downtime has always affected the prod~tctivecapabiliry of physical
assetsby reducing output, increasing operating costs anciinterfering with
customer service. By the 1960's and 1970ts,this was already a major
concern in the mining, manufacturing and transport sectors.In manufac-
turing, the effects of downtime are being aggravated by the worldwide
move towards just-in-time systems, where reduced stocks of work-in-
progress mean that quite smallbreakdowns are now ril~rch
trlore likely to
stop a whole plant. In recent times, the growth of mcclianisation and
automation has meant that reli~tbilityand avnilahilitjl have now also
become key issues in sectors as diverse as health care, data processing,
teleconimunications and building management.
Greater automation also means that more and rnore failures affect our
ability to sustain satisfactory quality starz~lcrrds.
This applies as much to
standards of serviceas itdoes toproduct quality.For instance, equipmelit
failures can affect climate control in buildings ancl the punctuality of
transport networks as much as they can interfere with the consistent
achievement of specified tolerances it1 manufhcturing.
More and more failures have serio~rs
sgfity or en~~iromnentill
conse-
quences, at a timc when standarcis in these areas are rising rapidly. In
some parts of the world, the point is approaching where orgallisations
either conform to society's safety ancl envin)nmental expectations, or
they cease to operate.This adds an order of magnitude to our dependence
on the integrity of our physical assets - one which goes beyond cost and
which becomes a simple matter of orgatlisational survival.

9. At the saiile time as our dependence on physical assets is growing, so
too is their cost - to opemtc and to own. To secure the maximum return
on the investment which they represent,they itlustbe kept working effici-
ently for as long as we want them to.
Finally, the cost of rrraintenarzce itself is still rising. in absolute terms
and as a prop or ti or^ of total expei~diture.
In some industries, it is now the
second highest oreventhe highest elementof operating costs. As aresult,
in only thirty years it has moved from almost nowhere to the top of the
league as a cost control priority.
New research
Quite apart from greater expectations, new research is changing many of
our most basic beliefs about age and failure. In particular, it is apparent
that there is less and less connection between the operating age of most
assets and how likely they are to fail.
However, Third Geileration research has revealed that not one or two but
six failure patterrls actually occur in practice. This is discussed in detail
later, but it too is having a profouncl effect on maintenance.
New tec,hniquc~s
There has been explosive growth in new maintenance concepts and tech-
niques. I-fundreds have been developed over the past fifteen years, and
rnore are emerging every week.
1940 1950 1960 1970 1980 1990 2000
Figure 1.3: Changing ma~ntenance
techn~ques
The new developments include:
decisiotz siipport tools,such ashazard studies, fail~lre
inodes andeffects
analyses and expert systenis
new maintenance teclzniques, such as condition tnonitoring
destgrzzng equipment w~th
a much greater e~nph,tit
on reli,ib~lityand
rnarntarnabillty
n mqor ~hzfi
in orgnnzscztlonczl th~rzkirzg
towarcli partiotpcttlon,team-
working and flexiblhty.
A major challenge fac~ng
malntet~ance
people nowadays I not only to
learn what these techniques are, but to decide which are worthwhlle and
which are not in their own organisations. If we inake the nght choices, it
ispossible to improveassetperformance andnt thesurne tune contain ancl
even reduce the cost of maintenance. If wc inake the wrong choicei, new
problems are created while existing problerns only get worie
The chnllerzgesfucing nzaintennrzcc
In a nutshell, the key challenges facing niodern malnteilance mati'~gers
can be rummzlrired ar follows.
to select the mo5t appropriate technlcluei
* to deal wtth each type of iallure process
inorder to f~~lfil
alltheexpectatlons of the owners of the ,iets, the uiers
of the aisets and of soclety as ii whole
in the mort cost-effectwe and errcluring i,kih~on
with the actlve upport and co-operallon of all the people tnvolved

10. 6 Kc~li~zlbility-cerztrd
Mairttenance I
RCM provides aframework which enablesusers torespond tothese chal-
lenges, quickly and simply. It does so because it never loses sight of the
fact that mainterranceisaboutphysical assets. Ifthese assetsdidnotexist,
the maintenance functio~l
itself would not exist. So RCM starts with a
comprehensive, zero-based review of the maintenance requirements of
each asset in its operating context.
All too often, these requirements are taken for granted. This results in
the development of organisation structures, the deployment of resources
andthe implementation of'systemson the basis ofincomplete orincorrect
assumptions about the real needs of the assets. Onthe other hand,if these
I
requirements are defined correctly in the light of modern thinking, it is
possible to achieve quite remarkable step changes in maintenance effi-
ciency and effectiveness.
'The rest of this chapter introduces RCM in more detail. It begins by
exploring the meaning of 'maintenance' itself. It goes on to define RCM
and to describe the seven key steps involved in applying th~s
process.
1.2 Maintenance and RCM
From the engineering viewpoint, there are two elements to the manage-
ment of any physical asset. It must be maintained and from time to time
it may also need to be modified.
The major dictionaries define maintain as cause to continue (Oxford)
or keep in art existing sfate (Webster). This suggests that maintenance
means preserving something. Onthe other hand,they agreethat tomod&
something means to change it in some way. This distinction between
maintain and ~nodify
has profound implications which are discussed at
length in later chapters.However, we focuson maintenance at this point.
When we set out to maintain something, what isit that wewishtoc n u s ~
to continue? What is the existing state that we wish to preserve?
The answer to these questions can be found in the factthat every phys-
ical asset is pitt illto service because someone wants it to do something.
In other words, they expect it to fulfil a specific function or functions.So
it follows that when we maintain an asset, the state we wish to preserve
must be one in which it continues to do whatever its users want it to do.
Maintenance: Ensuring that physical assets
continue to do what their users want them to do
Introduction to Reliubilzty-cerrtredh4aintc.nunr.r 7
What the users want will depend on exactly where and how the asset is
being used (theoperating context).This leads to the following formal de-
finition of Reliability-centred Maintenance:
Reliability-centred Maintetzarzce: a process
used to determine the rnairzterzarzcereyuiremer2t.s
of any physical asset in its operating context
In the light of the earlier definition of ~naintenance,
a f~tller
definition of
RCM could be 'a process used to cleterminevvhcat rrz~tst
he done to eizsurr-
that anyphysical rrssel continues tock, k t * h ~ t ~ ~ ~ l
its usenc want it to do in
its present operating context'.
1.3 RCM: The seven basic questions
The KCM process entails asking seven questions ahout the asset or syc-
tem under review, as follows:
what are thefunctions and associatedperfornrnrzce starzdards of the
asset in its present operating corttext?
in what ways does itfail tofulfil itsfuizctions?
what caclses eachfz~nctional
failt~re?
what happens when eachfailure occurs?
in what way does eachfailure matter?
what can be done topredict orprevent eachfailure?
what shozild be done if a suitableproactive t ~ s k
canrzot befound?
These questions are introduced brielly in the following paragraphs, and
then consiclered in detail in Chapters 2 to 10.
Functionsand PerformanceStandards
Before rt 13 posslble to apply a plocess ured to deterrri~ne
what must be
done to ensure that any phyical aet contmues to (lowhatever rts useri
want it to do in ~ t c
present operating context, we need to cfo two things
deterrnlne what ~ t s
usel, want rt to do
ensure that it rs capable of dorng what 1t4user5 want trt tart w~th

11. This is why tile first step in the RCM process is to define the functions of
each asset in its operating context, together with the associated desired
standards of performance. What users expect assets to be able to do can
be split into two categories:
primayfitnctions, which summarise why the asset was acquired in the
first place. This category of functions covers issues such as speed,out-
put, carryingor storage capacity, product quality andcustomer service.
seco~zclar)!fErrzctiolzs,
which recognisethat every asset is expected to do
rnore than sirnply fulfil its primary functions. Users also have expecta-
tions in areas such as safety, control, containment, comfort, structural
integrity,econorny,protection,efficiency of operation,compliance with
environmcntal regulations and even the appearance of the asset,
The users of the assets are usually in the best position by far to know
exactly what contribution each asset makes to the physical and financial
well-be~ng
of the organisatiorl as a whole, so it is essent~al
that they are
involved in the RCM process from the outset.
Done properly, this stepalone usually takes up about athird of the time
lnvolved in an entire RCM analysis. It alsousually causes the team doing
the analyis to learn a rerllarkable amount often a frightening amount
- about how the equipment actually works.
Functions are explored in Inore detail in Chapter 2.
Functiotial Failures
'The objectives of maintenance are defined by the functions and associ-
atedperformance expectations of the asset under consideration. Buthow
does maintenance achieve these objectives?
The only occrrrrencewhich is likely tostopany asset performing to the
standard required by its users is some kind of failure. This suggests that
tnaintenance achieves its objectives by adopting a suitable approach to
the inanager~lent
of failure.However,before we c a ~
apply a suitableblend
of failuremanagement tools, we needto identify what failures can occur.
The RCM process does this at two levels:
* firstly, by identifying what circumstances amount to a failed state
then by asking what eventscan cause the asset to get into a failed state.
In the world of RCM, failed states are known asfunctionalfaibres be-
cause they occur when an asset is urlrlble tofuEfiE afunction tou standard
c$l7elCfor-munr.ewhich is ac.ceptahle to the user.
In addition to the total inability to function, this ciefinitiorlencompas-
ses partial failures, where the asset still f~~nctions
but at an unacceptable
level of performance (including sittlationswhere the asset cannot sustain
acceptable levels of quality or accuracy). Clearly these can only be
identified after the f~lnctions
and performance stanciardsof'the asset have
been defined.
Functional failures are discussed at greater length in Chapter 3.
Failure Modes
As mentioned in the previous paragraph, onceeach functional failure has
been identified, the next step is to try to identify all the evcrzts +vhic%
are
reusonabf.ylikely to cause enchfuiled state. These events are known as
failure modes. 'Reasonably likely' failure modes include those which
have occurred on the same or similar equipment operating in the same
context, failures which are currently being prevented by existing main-
tenance regimes, and failures which have not happened yet hut which are
considered to be real possibilities in the context in question.
Most traditional lists of fiiilurc modes incorporate faililres caused by
deterioration or normal wear and tear. However, the list should include
failures caused by human errors (onthepart of operators and tnaintai~~ers)
and design flaws sothat allreasonably likely causes of ecluiprnentfailure
can be identified and dealt with appropriately. It is also important to
identify the cause of each failure in enough cletail to ensure that time and
effort are not wasted trying to treat sytnptorns instead of causes. 0
1
1
the
otherhand, it is equally important to enswe that time is not wasted on the
analysis itself by going into too much cletail.
Failure Effects
The fourth step in the RCM proces entails listingfailz.~re
~flects,
whlch
describe what happens when each failure mode occur. These descrip-
tions should include all the lnforrnation needed to support the evaluation
of the consequences of the failure, such as.
what evidence (if any) that the failure has occurred
in what ways (if any) it poses a threat to safety or the environn~ent
* in what ways (if any) tt affects production or operations
what physical damage (ISany) is caused by the fatlure
what must be done to rcpair the failure

12. Failure modes anci effects are discussed at greater length in Chapter 4.
Theprocess of ide~ttif~~irzgfunctio~1s,
functiortulfail~tres,
failure mocies
un(ifi~i1ure
effectsyielcls surprising cmd often very exciting opportunities
for irnprovittgperj3rrnanc.e nncl scgety, and ulsofor clinzirratirzg waste
Failure Consequences
A detailed analysis of an average industrial undertaking is likely to yield
between three and ten thousand possible failure modes. Each of these
failures affectsthe organisation in some way,but in each case, the effects
are different. They may affect operations. They may also affect product
quality, customer service, safety or the environment. They will all take
time and cost money to repair.
It is these consequences which most strongly influence the extent to
which we try to prevent each failure. In other words, if a failure has seri-
ous consequences, we are likely to go to great lengths to try to avoid it.
On the other hand, if it has little or no effect, then we may decide to do
no routine maintenance beyond basic cleaning and lubrication.
A great strength of RCM is that it recognises that the consequences of
failures are farInoreimportant thantheir technical characteristics. In fact,
it recngnises that the only reason for doing any kind of proactive main-
tenance is not to avoid failuresper se,but to avoid or at least toreduce the
consequerzccsof failure. The RCMprocess classifies theseconsequences
into four groups, as follows:
Hiddenfailure consequences: Hidden failures have no direct impact,
but they expose the organisation to multiple failures with serious, often
catastrophic, consequences.(Most of these failures are associated with
protective devices which are not fail-safe.)
Safety and environmental consequences: A failure has safety conse-
quences if it could hurt or kill someone. It has environmental conse-
quences if it co~tld
lead to a breach of any corporate, regional, national
or international environmental standard.
Operational consequences: A failure has operational consequences if
it affects production (output, product quality, customer service or oper-
ating costs in addition to the direct cost of repair)
Non-operatiortal consequences: Evident failures which fall into this
category affect neither safety nor production, so they involve only the
direct cost of repair.
We will see later how the RCM process uses these categories as the basis
of a strategic frameworkfor maintenance decision-miiking. By forcing a
structured review of the consequencesof each failure mode in terms of the
above categories, it integrates the operational,environmental and saf'ety
objectives of the maintenance function.This helps to bring safety and the
environment into the rnainstrearn of maintenance managc~tiezit.
The consequence evaluation process also shifts emphasis away from
theideathatallfailuresarebad andmust beprevented.Insodoing,itfocuses
attention on the maintenance activities which have most effect on thepcr-
formance of the organisation, and diverts energy away from those which
have little or no effect. It also encourages us to think inore broadly about
different waysofmanagingfailure,ratherthan toconcentrateonlyonfailure
prevention.Failurenianagerncnttechniquesaredividedintotwocategories:
* proactive tusks: these are tasks undertaken before a falluse occurs, In
order to prevent the item from getting into a failed stare.They e~librace
what istraditionallyknown as 'predictive' and 'preventive' nnalntentmce,
although we will see later that KCM uses the terms sche~llll~d
rcltom-
tion, ~c/zeduEed
discard and our-conditionr~zcrintenarzcc.
* defizultacttons: these deal with the failed state,and are choen when it
is notpossible to identify aneffectiveproactive task.Defit~~lt
acllotlsin-
cludefurlitre-finding, redesign and run-to-firilur-c.
The consequence evaluation process is discussed again bncfly lutes 111
this chapter,and inmuch more detail inChapter5.The nextsectzonof thts
chapter looks at proactive tasks in rhore detail
ProactiveTasks
Many people still believe that the best way to optiniise plant availability
is to do some kind of proactive maintenance on a ro~rtille
basls. Second
Generation wisdom suggested that this should consist of overhauls or
component replacements atfixed intervals. Figure 1.4~llustrates
the fixed
interval view of failure.
Figure 1.4:
The traditional
view of failure Age

13. Irztroduction to Ke1iabili~-certtrc>cl
hlriirzt~rzcrrzc*e 13
Figure 1.4is based on the assumption that most itemsoperate reliably for
aperiod 'X', andthen wear out. Classicalthinking suggeststhatextensive
records abouthil~lre
will enableus to determinethis lifeand so make plans
to take preventive action shortly before the item is due to fail in future.
This model is true for certain types of simple equipment, and for some
complex items withdominant fail~lre
modes.Inparticular, wear-out char-
acteristics are often found where equipmentcomes into direct contact with
the product. Age-related failures are also often associated with fatigue,
corrosion, abrasion and evaporation.
However, equipment in general is farmore complex than it was twenty
years ago. This has led to startling changes in the patterns of failure, as
shown in Figure 1.5.The graphs show conditional probability of failure
against operating age for a variety of electrical and mechanical items.
Pattern A is the well-known bathtub curve. It begins with a high
incidenceof failure (known as infant mortality)followed by aconstant or
gradually increasingconttitional probability of failure, then by awear-out
Tone. Pattern B shows constant or slowly increasing conditional prob-
ability of failure, ending in a wear-out Lone (the same as Figure 1.4).
Figure 1.5:
Six patterns F
offailure
Pattern C shows slowly increasing conditional probability of failure,
butthere isno identifiable wear-out age. Pattern D showslow conclitional
probability of failure when the itern is new orjust out ofthe shop, then a
rapid increase to a constant level, while pattern E sl~ows
a constant con-
ditional probability of failure at all ages (random failure).Pattern Fstarts
with high infant mortality, which drops eventually to a constant or very
slowly increasing conditional probability of failure.
Studies done on civil aircraft showed that 4% of the items confoimed
to pattern A, 2% to B, 5% to C,7% to D, 14%~
to E and no fewer than 68%
topattern F. (The number of tirnes these patterns occur in aircraft is not
necessarily the same as in industry. But there is no doubt that as asscts
become more complex, we see more and rnore of patterns E anci F.)
These findings contradict the belief that there is always a connection
between reliability and operating age. This belief led to the idea that the
more often an item is overhaulecl, the less likely it is to fail. Nowadays,
this is seldom true. Unless there is a dominant age-related failure mode,
age limits do littleor nothing to improve the rcliability of conlplex items.
I11fact scheduled overhauls can actually incrt.a.rc.overall failure ratcs by
introducing infant mortality into otherwise stable systems.
An awareness of these factshas led some organisations to abanctonthc
idea of proactive maintenance altogether. In fact, this can be the right
thing to do for failures with lninor consequences. Rut when the failure
consequences are significant, sontethiizg niust be tione to prevent or pre-
dict the failures, or at least to reduce the consequences.
This brings us back to the question of proactive tasks. As mentioned
earlier, RCM divides proactive tasks into three categories, as follows:
scheduled restoration tasks
scheduled discard tasks
scheduled on-condition tasks.
Scheduled restoration and schecluled discard tci~kr
Scheduled restoration entailsrenianufactur-ingacomponent or overhaul-
ing an assembly at or before a specified age linxt, regiurdletc of 1t9oon-
dition at the time. Similarly, scheduled discard entailsdiscardn~g
an Iten1
at or before a spec~fied
life limit, regardless of ltt conditlotl at the time.
Collectively, these two types of tasks are now generally known aspre-
ventive maintenance. They used to be by far the sx~ost
w~ciely
used form
ofproactive maintenance. However forthe reasons cfstcussedttbuve,they
are much less widely used than they were twenty years ago.

14. On-condition tush
The continuing need to prevent certain types of failure, and the growing
inability of classical techniqites to do so, are behind the growth of new
types offailure management. Themajority ofthese techniquesrely on the
fact that rnost faili~res
give some warning of the fact that they are about
tooccur.These warningsareknown ztspoterttialfailures, and aredefined
as idc.ntiJiable physical conditions which indicate that nfunctionalfail-
ure is about to occur or is irz the proces.~
of occurring.
The new techniques are used to detect potential failures so that action
can be taken to avoid the consecjuenceswhich could occur if they degen-
erate into functional failures. They are called on-conditiontasks because
itemsareleft in serviceon the condition that theycontinue tomeet desired
performance standards. (On-condition maintenance includes predictive
muinterzance, condition-bused maintenance and condztion monitoring.)
Used appropriately,on-condtion tasks are a very good way of managing
failures, but they can also be an expensive waste of time. RCM enables
decisions in this area to be made with particular confidence.
Default Actions
RCM recognises three major categories of default actions, as foIIows:
* j2liklre-fi17ding.1Failure-finding tasks entail checking hidden functions
periodically to determinewhether they have failed (whereas condition-
basecf tasks entail checking if something is failing).
redesign: redesign entails making any one-off change to the built-in
capability of a system.Thisincluclesmodifications to the hardware and
also covers once-off changes to procedures.
no scheduled rnuintenarzce:asthe name implies,this defaultentailsmak-
ingnoeffort to anticipate orprevent failuremodes towhich itis applied,
andsothosefailuresaresimply allowed tooccur andthenrepaired.'This
default is also called run-tofailure,
The RCM Task Selection Process
A great strength of RCM is the way it provides simple,precise and easily
understood criteria for deciding which (if any) of the proactive tasks is
tt~chrzically
feasible in any context, and if so for deciding how often they
should be done and who should do them. These criteria are discussed in
more detail in Chapters 6 and 7.
Whether or not a proactive task is technically feasible is govenied by
the technicalclzaructeristics ofthetask andof the failurewhich it ismeant
toprevent. Whether itis worthdooitzg 1sgoverned by how wellit deals with
the consequences of the failure. If aproactive task cannot he fourldwhich
is both technically feasibIe and worth doing, then suitable default action
must be taken. The essence of the task selectio~i
process is as follows:
for hiddenfailures, aproactive task isworth doing if it reduces the risk
of the multiple failure associated with that futlction to an acceptably
lowlevel.If sucha taskcannotbe foundthen ascheduledfailzlre-Jirtdirtg
taskmustbe performed.If asuitablefailitre-filldingtaskcannotbe found.
then the secondary defa~~lt
decis~on
is that the item may have to be re-
designed (depending on the consequences of the rnultiple failure).
for failureswith safety or envirorlrnerztalconequertcex, a prcnctive task
is only worth doing if it reduces the risk of that failure on its own to
a very low levelindeed,if itdoes not eliminateit altogether.If a task can-
not be found which reduces the risk ofthe failureto at1acceptably low
level, the item must he redesigned or theprocess nzzcst be cftanged.
* if the failure has operatio~zulconsequences, a proactive task is o111y
worthdoing if the total cost of doing it ovrrctperiotloj'tinze is less than
the cost of the operational consequences and the cost of repair over the
same period. In other words, the task rnust be ju.c.r$crl or?econornic
grounds. If itis notjustified, the initial defaultdecision isr2oscfteduled
mainterzance.(Ifthis occurs andthe operational consequencesare still
unacceptable then the secondctry defa~tlt
decision is again redesign).
* if a failurehas non-operc~tionnl
consequences a proactive task is only
worth doing if the cost of the task over a period of time is less than the
cost of repairover the sameperiod. Sothesetasks rnust alsobejustiJied
on economic grourzds. If it is not justified, the initial default decision
is againnoschertllledmainterzance,andif therepair costsaretoohigh,
the secondary default decision is once again redesign.
This approach means that proactive tasks are only specified for failures
which really need them, which in turn leads to substantial reductions in
routine workloads.Lessroutine work alsomeansthat the remaining tasks
are more likely to be done properly. This together with the elrmination of
counterproductive tasks leads to more effective rllaintenance

15. Compare this with the traditional approach to the development of
maintenancepolic~es.
Traditionally,the maintenancerequirements of each
asset are assessed in term of itsreal or assumed technical charactenstics,
without considering the consequences of failure. The resulting schedules
are used for all si~nilar
assets, again without considering that d~fferent
consequences apply in different operating contexts. This results in large
numbers of schedules which are wastetl, not because they arc 'wrong' in
the technical sense, but because they achieve nothing.
Note also that the RCM process considers the maintenance require-
nlents ofeach assetbefore asking whether itis necessary to reconsider the
design. This is simply because the maintenance engineer who is on duty
today has to maintain the equipment as it existstudciy,not what should be
there or what might be there at some stage in the future.
1.4 Applying the RCM process
Before setting out to analyse the maintenance requirements of the assets
in any organisation,we need to know what these assets are and to decide
which of them aretobe subjected to theRCM review process.This means
that aplant register must be prepared if one does not exist already. In fact,
thevast lnajorityof industrialorganisationsnowadaysalreadypossessplant
registers which are adequate for this purpose, so this book only touches
on the most desirable attributes of such registers in Appendix 1.
Planning
If it is correctlyapplied,RCM leadsto remarkable improvementsin main-
tenance effectiveness, and often does so surprisingly quickly. However,
the successful application of RCM depends on meticulous planning and
preparation. The key elements of the planning process are as follows:
= decide which assets are most likely to benefit from the RCM process,
and if so, exactly how they will benefit
assess the resources required to apply the process to the selected assets
* in cases where the likelybenefitsjustify theinvestment, decide indetail
who is to perform and who is to audit each analysis, when and where,
ancl arrange for them to receive appropriate training
ensure that the operating context of the asset is clearly understood.
Review groupT
We have seen how the RCM process ernbodies seven basrc quest~onIn
practice, malntenancepeople s~mply
cannot answer allthese questions on
their own This 1sbecause Inany (11not 1no5t)ot the answers can only be
supplred by production or operations pcople This applies especidlly to
questions coucernrng functions, deslred pertormance, tailure ctfects dnci
fa~lure
consequences
For this reason, areview of the rna~ntendnce
requiielnents ot my asset
shouldbe done by srnallteams whlc11~nclude
at leurf oneperson fromthe
maintenance function and one from the operations funct~otlThe enlor-
~ t y
of the group members is less Important than the tact tlldt they 41ould
have a thorough knowledge of the aset under revrew Each group mern-
her shoulddlsohave been trained In RCM 'Themake-upof a typlcdl RCM
revrew group is shown rn Figitre I 6
The use of these groups not
only enables management
to gain acces5 to the
knowledge and Operat~ons
expertise of each Superv~sor
member of the group
on a systematic basis,
but the membe~
s Operator
themselves gain a
greatly enhanced under-
Facilitator
Engineering
Supervisor
Craftsman
(M and/or E)
standlng of the asset in External Specialist (ifneeded)
its operating context. (Technicalor Process)
Figure 1.6: A typical RCM rev~ew
group
Fucrlztators
RCM review groups work unde~
the gudance of highly tra~ncd
spec~~tl-
ists In RCM, known as fac~htatorsThe facilitator are tlie most irnpor-
tant people in the KCM revlew process The11role 1s to elnure that
the RCM analysis IS caxr~ed
out at the itght level, that s y ~ e m
bouncla-
rtes are clearly defined, thatno lnlportant Item ;Ire overlooked dnd th'it
the results of the analysls are properly tecorded
RCM is correctly u~iderstood
and applied by the group inenlbers
the groupreaches conensu In a brisk andorderly tashion,wluleretatn-
ing the enthus~asm
and corrlrnitrnent of n~dlvid~lal
member

16. 18 Reliability-rerzlred Maintenance
the analysis progresses reasonably quickly and finishes on trme
Facililators also work with RCM project rnariagersor sponsors to ensure
that each analysis is properly planned and receives appropriate manage-
rial and logistic support.
Facilitators and RCM review groups are discussed in more detail in
Chapter 13.
The outcomes ofan RCM unrxlysis
If it is applied in the rnanner suggested above, an RCM analysis results
in three tangible outcomes, as follows:
maintenance schedules to be done by the maintenance department
revised operating procedures for the operators of the asset
a list of areas where one-off changes must be rnade to the design of the
asset or the way in which it is operated to deal with situations where the
asset canriot deliver the desiredperformance in its current configuration.
'Two lesr tangibleoutcomes are that partictpantsm the process learn a great
deal about how the asset works, and also tend to function better as team.
Allcliting crti~irirnplerrzerztcitiorz
Irnn~ediatcly
;rSter the review has been completed for each asset, senior
managers with overall responsibility for the equipment must satisfy thern-
selves that decisions rnadc by the group are sensible and defensible.
After each review is approved, the recommendations are implemented
by incorporating maintenance scl~cdules
into nlaintenarlce planning and
control systems, by incorporating operating procedure changes into the
stanclard operating proccd~lres
for the asset, and by handing recommen-
dationsfor design changes to the appropriatedesign authority.Key aspects
of auditing and i~~lplernentation
are cliscussed in Chapter 11.
1.5 What RCM Achieves
Desirable ar they are, the outcomes listed above should only be seen as
a meancto anend. Specifically,they shouldenablethe maintenance func-
tion to f ~rlf~l
it11 the expectations listed In Figure 1.1at the beg~nning
of
this chapter. How they do so is sclmmarisedin the following paragr~phs,
and discussed again in Inore detail in Chapter 14.
Greatersafety andenvironmental integrity: RCM conrtders the e'lfety
and env~ronmental
~mplicatlons
ol evely f,iilure niode before conridel-
ing rts effect on operattons Thrs me;lns that steps are t'iken to rntnlinlsc
all identifiable equipment-related safety dncl enrrronnlentL~l
hdrarcfs, if
notel~m~nate
themaltogether By integratrngsafetyintothem,llnstrearn of
matntenance decisron-maktng, RCM alo improvec c~ttituitec
to s'lfety
Improvedoperatingperformance (output,prodact quality anrlcusto-
rnerservice): RCM recognlses thatalltypes ot maintenAncehave snrne
value, and prov~des
rules tor decld~ng
wlirch 15 moct siiitdble in every
s~tuatlonBy dorng so,~thelpcellsure th'tt only the tno5t eftect~ve
for rns
of maintenance are chosen for each aet, and that ttrtLtble'kction 1%
taken III cases where m'unlen'mcc cdriiiothelp Thrs much more tightly
focused marntenance effort leadstoquantumjunlpr In the pcrlorrn,lnce
of existzng asset5 where these are sought
RCM was developed to help arrli~les
dr'tw np 111:ilntenance progr~~rnc
for new typcs of arrcratt before they enter ccrvrct. As a ierult, rt ie .in
deal may to developsuchprograms forrzevcar wtr, e~pectally
~ornplev
equrpment tor whlch no hlrt011cdlrntorr~l~ltton
15 dvalldble Thr r'lves
much ot the tr~al
and error wh~ch
re so ottcn pLutot the ctevelopment of
new matntenance progrdrnr -- trlal which ri, tlrlie consurnrng ,inti frus-
trating, and error whrch can be very costly
Greater maintenance cost-effectiveness: RCM corttinu,illy focuses
attention on the malntenance actrvltie whlch h,lve rnoct effect on the
performance of the plant T h ~ s
helps to enulc th'it ecerythlng eltent on
malntenance IS spent where it wtll do the ~not
good
In addrtlon, dIlCM I correctly applled to exrsting maintenance sy-
tems, it reduces the amount of routlne work (111other words, rnarnte-
nance tasks ro be undeltaken on a LVCIZ( bar) ~wued
tn each penod,
usually by 40%to 70% On the other hand, 1
1RCM 1sused to develop
a new malntenance program, the result~ng
scheduled work1o;ld r much
lower than ~f the program IS developed by tract~t~onal
methods
Lorzger useful lqe of expeizsive iterns, due to a carefully focuseci em-
phasis on the use of on-condition maintenance techniclues.
A coinprehensive database: An RCM revrew ends wlth a conqrehe11-
sive and fully docunlented record of the maintenance r-eqt~ircment
of
allthe significantassets used by the organlsatroil 'I'h15make it poci~ble

17. 20 Ne1ictbilit)i-centredMaintenance
toadapttockaizgirlg~.zrc1~rnstances
(SLIC~
aschangingshiftpatternsornew
technology) without having to reconsider all maintenance policies from
scratch.It also enablesequipcnent usersto demonstratethat their mainte-
nance programsare built on rational foundations(the audit trail required
by Inore and rnore regulators). Finally, the infornlation stored on RCM
worksheets r.ect~4r.e~
the effects of ~tajfturnover
with its attendant loss
of experience and expertise.
An RCM review of the maintenance requirements of each asset also
provides a ~nuch
clearer view of the skills required to maintain each
asset, and for deciding what spares should be lzeld in stock. A valuable
by-product is also imjwove~l
dmtvings and rnanu~~ls.
Greaternzotivatiolz of ilzdividuals, especially people who are involved
in the review process. 'This leads to greatly improved general under-
standing of the equip~nent
in its operating context, together with wider
'ownerhip9of nlairltenanceproblems andtheir solutions. It alsomeans
that solutions are more likely to endure.
Better teamwork: RC'M provides acommon, easily understood techni-
cal language for everyone who has anything to do with ~naintenance.
'This gives maintenance ancl operations people a better uriderstanding
of what maintenance can (and cannot) achieve and what rnlrst be done
to achieve it.
All of these issues are part of the mainstream of maintenance manage-
ment, and many are alreadythe target of ~mprovement
programs. Amajor
featureof RCM isthat itprovides aneffective step-by-step framework for
tackling ~ ~ 1 1
of them at once, and for involving everyone who has any-
thing to do with the equipment in the process.
RCM yields results very quickly. In fact, if they are correctly focused
and correctly applied, RCM reviews can pay for themselves in a matter
of ruonthsand sornetirneseven amatterof weeks, asdiscussed in Chapter
14.The reviews transform both the perceived maintenance requirements
of the physical assets used by the organisation and the way in which the
maintenance function as a whole is perceived. Thc result 1s more cost-
effective, more harmonious and much more sirccessf~~l
maintenance.
Functions
Most people become engineers because they fee1 at least some affinity
for things, be they mechanical,electrical or structural.This affinity leads
themtoderivepleasure fromassets in good condition,hiitto feel oMencled
by assets in poor condition.
These reflexes have always been at the heart of the concept of preven-
tive maintenance. They have given rise to concepts such as 'asset care',
which as the name implies, seeksto care for assetsperse.They have also
led some maintenance strategists to believe that maintenance is all itbout
preserving the inherent reliability or built-in capability of any asset.
In fact, this is not so.
As we gain a deeper understanding of'the role of assets in business, we
begin to appreciate the significance of the fact that any physical asset is
put into service because someone wants it to do something. So it follows
that when we maintain an asset, the state which wc tc2i.shtol)re.rrrt3r
rnList
be one itz cvhich it contitz~~es
to LLU whatever its zrser-.v want it to (to.Later
in this chapter, we will seethat this state -. what the users want -is liirida-
mentally different from the built-in capability of the asset.
This emphasis on what the asset dues rather than what it is provides a
whole new way of defining the objectives of maintenance for any asset
-one which focuses on what the user wants. 'This is the most important
single feature of the RCM process, and is why many peoplc regard RCM
as 'TQM applied to physical assets'.
Clearly,inorder todefine theobjectivesof maintenance in tenrls of user
requirements, we must gain a crystal clear tulderstanding of the functions
of each asset together with the associated perfortnance standards.This is
why the RCM process starts by asking:
whatarethefunctionsandassociatedperformancestandardsofthe
asset in its present operating context?
Th~s
chapter cons~derith~s
qucstlon 1
1
1 more detd It clescnbes how
fr~nctron~
should be det~ned,
explore the two rnarr~
type, of ],crfom~ance
atandards.revrews tlrfferentcategorle4of functlon5itnd how how func-
tions should be l~rted

18. 2.1 Describingfunctions
It is a well cstablisheclprinciple of value engineering that afunction statc-
nient should consist of a verb and an object.It is also helpful to start such
statementswith the word 'to' ('to pump water', 'to transport people', etc).
However, as explained at length in the next part of this chapter, users
not only expect an asset to fulfil a function. They also expect it to do so
to an acceptable level of performance. So a f~inction
definition - and by
implication the clefinition of the objectives of maintenance for the asset
- is not complete unless it specifies as precisely as possible the level of
performance desired by the user (as opposed to the built-in capability).
For instance, the primary function of the pump in Figure 2.1 would be listed as:
To pump water from Tank X to Tank Y at not less than 800 litres per minute.
This exanlpleshowsthat acompletefunction statement consists of averb,
an object and the standard of performance desired by the user.
A ficrrctiorz statement should consist of a verb,
an object and a desired standard of performance
2.2 Performance standards
The objectiveof maintenance is to ensure that assets continue to do what
their users want them to do.'The extent to which any user wants any asset
to do anything can be clefined by a minimum standard of performance. If
we could build an asset which co~rld
deliver that minimum performance
without deteriorating in anyway, then that wouldbe the endof the matter.
The machine would run continuously with no need for maintenance.
However, in the real world, things are not that simple.
'The laws of physics tell us that any organised system which is exposed
to the real world will deteriorate.The end result of this deterioration is total
diso~ganisation(also known as 'chaos' or 'entropy'), unless steps are
taken to arrest whatever process is causing the system to deteriorate.
For instance,the pump in Figure2.1 is pumpingwater into a tank from which the
water is drawn at a rate of 800 litresiminute. One processthat causesthe pump
to deteriorate (failuremode) isimpellerwear. This happensregardlessof whether
itispumpingacidor lubricatingoil,andregardlessof whetherthe impellerismade
of titanium or mild steel.The only question is how fast it will wear to the pointthat
it can no longer deliver 800 litres/minute.
Figure 1:
Initialcapability vs
desiredperformance
Offtake from tank
800 l~tresim~nute
So~fdeteno~ation
1sinev~t~tble,
itmust
be allowecl tor Thls rnedrts that &hen
any asset 1s put into erLrlce,it mut be
able todeliver more than the lnrnlrnttrn
standxcl of perfolnnance cleired by the
L I ~
What the ,iset 14 'ible to deltver15
known asI ~ F ,
lrzrtzczlc U ~ L I I I Z I E ~ ~
(orinher-
ent reliabilrty) Etgure2 2 rllustr,~testhe
nght relatlonsh~p
between this capabl-
11tyand desrred per tormmce
Figure 2.2: Allowfng for deter1oraBon
For instance,in order to ensure that the pump shown in Figure2.1 does what its
users want and to allow for deterioration, the systern designers must specify a
pump which has an initial built-in capability of somethirig greater than 800 litresi
minute. In the exampte shown, this initial capability is 1 000 litres per minute.
This means that performance can be defined in zvvo ways, as litllows:
desired performance hut the user kviirrts tlip asset to c/o)
* built-in capability (what it crin c / o )
Later chapters look at how maintenance helps ensure that assets continue
to f~11fil
their intended functions, either by ensuring that thcir capability
remains above the minimum standard desired by the user or by restoring
something approaching the initial capability if it drops below this point.
When considering the question of restoration, bear in mind that:
* the initial capability of any asset is established by its design anclby how
it is made
maintenance can onIy restore the asset to this initial level of capability
- it cannot go beyond it.
In practice, most assetsare ;ldequatelyde4rgnetlant1hurlt, 50 11ts usually
possible to develop marntenance programs which entlre that such ctets
continue to do what their user4 wmt

19. Figure 2.3: A maintainable asset
In short, such assets are maintainable, as shown in Figure 2.3.
On the other hand, if the desired performance exceeds the initial cap-
ability, no amount of r~~airltena~lce
can deliver the desired performance.
In other words, s~tch
asset) are not maintainable, as shown in Figure 2.4.
For instance, if the pump shown in Fig- 11
the tankfull Sincethe maintenancepro- 1 I -
gramdoes notexistwhichmakespumps
bigger, maintenance cannot deliver the
desiredperformance inthis context. Sim-
ilarly,if we makea habitof tryingto draw
15 kW (desiredperformance) from a 10
kW electric motor (initial capability), the
motorwillkeeptrippingoutandwilleven-
tually burn out prematurely. No amount
of maintenance will make this motor big
enouqh. It may be perfectly adequately
desiined and built Inds own right- ltjust Figure 2.4:
cannot deliver the desired performance A non-ma,nta,nable s,tuat,on
In the context In which ~tIS being used
Two conclurons whrch can be drawn from the above examples are that
for any aacet to hc m'tintainable, the destred performance of the asset
must fall with~n
the envelope of ~ t s
in~t~al
capability
In order to detetmrne whether this 1sso, we not only need to know the
in~tlal
capab~l~ty
of the asset, but we also need to know exactly what
mrrtlmum pertormance the uer is prepared to accept 111the context m
wh~ch
the as4et 15 betng used
This underlines the importance of identifying prec~selywlznt thr urrry
want when starting to develop a mainterlance program The following
paragraphs explore key aspects ofperfonilance 4tandardstn more detanl.
Multiple perjorrncmce stcrrzda'uruls
Many function statements incorporate more than one and somctirnes
several perforrnance standards.
For example, one function of a chemical reactor in a batch-type chemical plant
might be listed as:
T o heat up to 500 kg of product X from ambient temperature to boiling point
(I25°C) in one hour.
Inthis case,the weightof product,the temperature rangeandthe time all present
differentperformanceexpectations.Similarly,the primaryfunction of a motorcar
might be defined as:
To transport up to 5 people along made roads at speeds of up to 140 km/h
Here the performanceexpectations relate to speed and number of passengers.
C)uantltnfzve peiformr~ncesturzdclrds
Performdnce standards sho~tld
he quantrfiecl where posible, becdwe
quantitative standards are ~nherently
milch more precise thcznclual~t;ttie
standards Specralcxe should be taketi to avord yuaht~ttr~e
statements like
To produce as many wrdgets as requ~red
by production', or 'to go '1s f ~ t
as poss~ble'Functlorr ctatements of thrs type 'ire meanrngIes, 11- orlly
becauce they make it ~mpossible
to deftne exactly when the rtem 14 t'itled
In reality, it can be extraord~narlly
dtlfrcult to dct111eprectsely vhdt is
requ~red,
but just because it 1s difficult does riot mean th'tt it cannot oi
should not be done One major user of RCM surrimed up this pornt by
saymg 'If the users of an asset cannot specrfyprecrscly whatperformance
they want from an asset, they cannot hold the malnt~rner
s dccountable for
sustarnlng that performance.'
QuaEitative st~znclnrds
In spite of the need to be precise, it is sometimes impossible to specify
quantitative performance standards so we have to live with qualitative
statements.
For instance,the primary function of a painting is usually 'to look acceptable' (if
not 'attractive'). What is meant by 'acceptable' varies hugely from person to
personandisimpossibleto quantify.As a result,userandmaintainerneedto take
care to ensure that they share a common understanding of what is meant by
words like 'acceptable' before setting up a system intended to preserve that
acceptability.

20. Absolute peIJi)rmarzcestc-rnd~lrcls
A f~~nction
statementwhich containsno performance standardat all usually
inlplies an absolute.
For instance, the concept of containment is associated with nearly all enclosed
systems. Functionstatementscoveringcontainment areoftenwrittenas follows:
To contain liquid X
The absence of a performancestandard suggests that the system must contain
aNthe liquid,andthat any leakageat all amountsto a failed state. Incaseswhere
an enclosed systemcan tolerate some leakage, the amountwhich can be toler-
atedshouldbeincorporatedasa performancestandardinthefunctionstatement.
V~rriable
pc$ormance sr~indards
Perfor~nance
expectations (or applied stress) sornetirnes vary infinitely
between two extremes.
Considerforexampleatruck usedtodeliver
loads of assortedgoods to urban retailers.
Assumethattheactualloadsvary between
(say)0 (empty)and5tons,withanaverage
of 2.5 tons, and the distributionof loads is
asshowninFigure2.5.Toallowfordeterio- t4
ration,the initialcapabilityof thetruck must -
z 3
bemorethan the 'worstcase' load,which in 2
this example is 5 tons. The maintenance a
program in turn must ensure that the cap-
abilitv does not drop below this level, in _i o
which case it would automatically satisfy
thefullrangeofperformanceexpectations. Figure 2.5:
Variableperformance standards
Upper and lower lirnits
In contrast to variable performance expectations, some systems exhibit
variable capability. These are systems which simply cannot be set up to
function to exactly the same standard every time they operate.
Forexample,a grindingmachineusedtofinish grindacrankshaft will notproduce
exactly the same finished diameter on every journal. The diameterswill vary, if
only by a few microns. Similarly,a filling machine in a food factorywill not fill two
successivecontainerswithexactlythesameweightof food.Theweightswillvary,
if only by a few milligrams.
Figure 2.6 incticatesthat capability variations of this nature usually vary
about a mean. In order to accommodate this variability, the associated
desired standards of performance incorporate an upper and lower limit.
For instance, the primary function of a sweet-packing machine might be:
To pack2501-1 gmof sweetsintobagsat a minimumrateof 75 bagsper minute.
The primaryfunction of the grinding ma-
chine miaht be: I
.
, ~~
To finish grind main bearingjournals in
a cycle time of 3.00 k0.03 minutesto a
diameter of 75 20.1 mm with a surface
finish of Ra0.2.
(Inpractice, thiskindof variability is
t
usually unwelcome for a number of
a a
reasons.Ideally,processes should be $j 2
so stable that there is no variation at $ ?
all and hence no need for two l~mits
QHI Figure 2.6:
In pursuli of this ideal, many indua- 2k. Upper dnd lower Iirn~ts
tnesare spendinga great deal ol time
and energy on des~gn~ng
procecses that vary as little as possible How-
ever, th~s
aspect of design and development I beyoncl the ,cope ol this
book R~ght
now we are concerned p~rrely
w~th
vaii,rb~lity
from the wcw-
point of inaintenance )
How much varlab~llty
can be toleiated In the peciti~~ition
of my prod-
uct IS usually governed by external factors
For instance, the lower limit which can be tolerated on the crankshaft journal
diameter is governed by factors such as noise, vibration and harshness,and the
upper limit by the clearances neededto provide adequate lubrication.The lower
limitof theweightof the bagof sweets(relativeto the advertisedweight) is usually
governedbytrading standardslegislation,whilethe upperlimit isgovernedbythe
amount of product which the company can afford to glve away.
Incaseslike these. the desireclperfoor-manceliixits are known as thc upper
and lower speciftcation limits. The limits of capability (usually clefinecl
as being three standard deviations either side of the mean) are known as
the upper and lower control limits. Quality management theory suggests
that in a well managed process, the difference between the control limits
should ideally be half the differencebetween the specificationlimits.This
multiple should allow ainorethan adequate margin fordeterioration from
a maintenance viewpoint.
Upper and lower limits not only apply to product cluality. Tlicy also
apply to other functiollal specifications such as the accuracy of gauges
and the settings of control systems and protective clevices. 'This issue is
discussed further in Chapter 3.

21. 28 Ueliubi1i~-centreif
Maintenarzce
2.3 The Operating Context
In Chapter 1,RCM was defined as 'a process used to determine the main-
tenance requirements of any physical asset in its operating context'. This
context pervaclcs the entire mairlte~lancestrategy formulation process,
starting with the definition of functions.
For example,consider a situation where a maintenanceprogramis being devel-
opedfor atruck usedto transport materialfrom Startsvilleto Endburg.Beforethe
functions and associated performance standards of this vehicle can be defined,
the people developing the program need to ensure that they thoroughly under-
stand the operating context.
For instance, howfar is Startsvillefrom Endburg?Overwhat sort of roadsand
what sort of terrain? What are the 'typical worst case' weather and traffic condi-
tions on this route?What load isthe truck carrying (fragile?corrosive?abrasive?
explosive?)Whatspeedlimitsandother regulatoryconstraintsapplytothe route?
What fuel facilities exist along the way?
The answers to these questions might lead us to define the primary function
of this vehicle as follows: 'To transport up to 40 tonnes of steel slabs at speeds
of upto 60 mph(average45 mph)from Startsvilleto Endburgon onetank of fuel'.
The operating context also profoundly influences the requirements for
seconciaryf~tnctions.
In the case of the tr~~ck,
the climate may demand air
conditioning, regulations ]naydemandspeciallighting, theremoteness of
Endburg Inay demand that special spares be carried on board, and so on.
Not only does the context drastically affect functions andperformance
expectations, but it also affects the nature of the failure modes which
could occur, their effects and consequences, how often they happen and
what must be done to manage them.
For instance, consider again the pump shown in Figure 2.1. If it were moved to
alocationwhereitpumpsmildlyabrasiveslurryintoaTankBfromwhichthe slurry
is being drawn at a rate of 900 litres per minute, the primary function would be:
To pump slurry into Tank B at not less than 900 litres per minute.
This is a higher performancestandardthan inthe previouslocation,sothe stand-
ardto which it hasto bemaintainedrisesaccordingly.Becauseit is now pumping
slurry instead of water, the nature, frequency and severity of the failure modes
also change. As a result,although the pump itself is unchanged,it is likelyto end
up with a completely different maintenance program in the new context.
All this means that anyone setting out to apply RCM to any asset or
process must ensitre that thcy have a crystal clear understanding of the
operating context before thcy start. Some of the most important factors
which need to be consiclered are ciiscussed in the following paragraphs.
Batclz and flow procerseJ
In manufacturlng plants, the most Iinport'lnt featureof the opercitrngcon-
text 15 the type of process This ranges from flow proces5 opelatlons where
nearly all the equipment 1s interconnected, to jobbing operation, where
most of the machlne5 are Independent
In flow processes, the failure of a single asset can elther top the entire
plant or srgnificantly red~rce
output, unless surge capac~ty
or st'tnd-by
plant 1savailable. On the otherhand, tn batch or jobbtng plLmt,mot fail-
ures only curtall the output of a s~ngle
machine or I~ne
The consecjuences
of suchfasluresare determ~ned
masnly by the durLtt1on
of the toppage and
the amountof work-~n-process
que~rlllg
m tront of srtb4equentoperations
These differences mean that the m~iintenance
strategy applied to an
asset which is part of a flow process coultlbe radically dsfferenttron~
the
5trategy appl~ed
to an ident~cal
a s w In a batch enviloilnient
Redurulzclnrzcy
The presence of redundancy -or alternative means of production - is a
feature of the operating context which must be considered in detail when
defining the functions of any asset.
The importanceof redundancyis illustrated by the three identical pumps shown
in Figure 2.7. Pump B has a stand-by, while pump A does not.
Figure 2.7:
Different
operating
contexts
Stand Alone j( Duty Stand-by
This meansthat the primaryfunctionof pumpA isto transfer liquidfromone point
to another on its own, and that of pump B to do it infhe presence of a stand-by.
This differencemeansthat the mainterlancerequirementsof these pumpswill be
different (just how different we see later), even though the pumps are identical.
Quality .starzdnrds
Quality standards andstandards of customerservice aretwo trroreaqects
of the operating context which can lead to differences between the de-
scriptions of the functions of otherwise identical machines.
For example, identical milling stations on two transfer machines might have the
same basicfunction-to mill a workpiece. However,depth of cut, cycletime, flat-
nesstolerance and surface finish specificationsmight all be different.This could
lead to quite different conclusions about their maintenance requirements.

22. Erzviror~n~ental
stcznckzrds
An increasingly important aspect of the operating context of any asset is
the impact which it has (or coulci have) on the environment.
Growing worldwide interest in environmental issues means that when
we maintain any asset, we actually have to satisfy two sets of 'users'. The
first is the people who operate the asset itself. The second is society as a
whole, which wants both the asset and the process of which it formspart
not to cause ilndue harm to the environment.
What society wants is expressed in the form of increasingly stringent
environn~ental
standardsand regulations.These are international,national,
regional, rl~iirlicipal
orevencorporatestandards. They coveran extraordi-
narily wide range of issues, from the biodegradability of detergents to the
content of exhaust gases. In the case of processes, they tend to concentrate
on unwanted liquid, solid and gaseous by-products.
Mostind~~strics
arerespondingtosociety'senvironmentalexpectationsby
ensuring that eqttipnient is designed to comply with the associated stand-
ards. However, it is not enough simply to ensure that a plant or process
is environnlentally sounct at the liloment it is commissioned. Steps also
have tobe taken toensurethat it re~rtains
incompliancethroughoutitslife.
'Taking theright steps isbecoming amatter of urgency, because allover
the worlcl, more and more incidents which seriously affect the environ-
nlent areoccilrringbeca~ise
some physical asset clid not behave as it should
-in other worcts, because sornethi~lg
failed. The associated penalties are
becoming very harsh indeed, solong-term environ~nental
integrityis now
a particularly important issue for maintenance people.
Safety hazarcir
An increasing number of organisations haveeitherdeveloped themselves
or siibscribe to formal standards concerning acceptable levels of risk. In
somecases,these apply atcorporate level,in others toindividual sitesand
in yet others to individ~ialprocesses or assets. Clearly, wherever such
standards exist, they are an important part of the operating context.
Shift frarmngrmenfL~
Shift arrangements profoundly affect the operating context. Someplants
operate for eight hours per day five days a week (ancl even less in bad
times).Others operate continuously for seven days a week, and yet others
somewhere in between.
In a singleshiftplant, production lost dueto failurescan usually be made
up by workingovertinle.'This overtime leadsto increased production costs,
so maintenance strategies are evaluated in the light of these costs.
On the otherhand, if an assetis working 24hoursper (lay,sevendaysper
week, it is seldompossible to make tip for lost time, so ciowntimccauses
lost sales. This costs a great deal more than extra overtime, so it is worth
tryingmuchhardertoprevent failures iinderthesecirc~tmstances.
However,
it is also more difficult to make equipment available for maintenance in
afully-loadedplant,somaintenance strategies need tobe for~xiulated
nith
special care.
As products move through their lifecycles or as eco~tomic
conclitions
change, organisations can rnove from one end of this spectrum to the
other surprisinglyquickly.For this reason,it is wise to review nlnintenance
policies every time this aspect of the opcrating context changes.
Work-in-process
Work-in-process refers to any material which has not yet been tliroughall
the steps of the manufacturing process. It may be stored in tanks, in bins,
in hoppers, on pallets, on conveyors or in special stores.The consecluen-
ces of the failure of any machine are greatly influencecf by the itmo~~nt
of
this work-in-process between it and the next machines in the process.
Consideran examplewherethe volume of work inthe queue is s~ffficient
to keep
the next operationworking for six hours and it only takes four hoursto repairthe
failure mode under consideration. In this case, the failure would be unlikely to
affect overall output. Conversely, if it tobk eight hours to repair, it could affect
overall output because the next operation would come to a halt. The severity of
these consequences in turn depends on
the amount of work-in-process between that operation and the next and so on
down the line, and
the extent to which any of the operations affected is a bottleneck operation (in
other words an operation which governs the output of the whole line).
Although plant stoppagescost money, it also costs n~oney
to hold stctcks
of work-in-process. Nowadays stock-holding costs of any kine1 are so
high that reducing them to an absolute minimurn is u top priority. 'I'his is
a major objective of 'just-in-time' systems and thcir derivatives.
These systems reduce work-in-process stocks, so the cushion that the
stocks provided against failure is rapidly disappearing. 'This is a vicious
circle,because the pressure on maintenance departmentsto recfuce failures
in order to make it possible to do without the cushion is also increasing.

23. Sofrom the tnaintenance viewpoint, abalance hastobe struckbetween
the ecoi~omic
inlplications of operatiorial failures, and:
the cost of holding work-in-process stocks in orderto mitigate the effects
of those failures, or
the cost of cloing proactive maintenance tasks with a view to anticipat-
ing or preventing the failures.
To strikethis balance successfully,this aspectof the operatingcontextmust
be particularly clearly understood in manufacturing operations.
Repair titne
Repa~r
titnes are influenced by the,,pc)edof resporzseto the failure, which
isafunctionof f:trlure reporting systelns andstaffing levels, andthespeed 1
ojrepz~ir
itself. which is a fitnctionof the availab~lity
of sparesand appro-
priate tools ancl of the capability of the person doing the repairs.
These factors heavily influence the effects and the consequences of
failur-es,and they vary widely fromoneorganisationto another As aresult,
this aspect of the operating context also needs to be clearly understood.
i
SpL1ri!5
It 1spos~bIe
to use a derivative of the RCM process to optimise spare5 1
tack ,lncf the ,toc~ated
ftulurcmanagement pol~clesThis der~vatlve
is
bawl on the fdct that the only reason for keeping a stock of spare parts Is
'11 ure.
to dvoid or reduce the ctmsequenccs of f' I
The relatlonshlp between spares and failure consequences hinges on I
the tlme ~ttakes to procure spates from supplrers If ~tcould be done
tnstantly there would be no need to stock any spares at all. But in the real
world procurrng pares takes time Thls 1sknown as the lead trme, and ~t
ranges from a matter of rn~nutes
to several months or years If the spare
1snot a stock ~teril,
the lead tirne often dictates how long ~ttakes to repalr
the falluse, and hence the severrtyof tts consequences On the other hand,
holcllng sparesrn tack also costs money, ro a balance needs tobe struck,
on a case-by-casebas~s,
between the cost of holding a spare rn stock and
the total cost ot not L~olcltng
~t In some cases, the welght andor d~men-
sons of the spates also need to be taken into account because of load and
space restnctlons, espectally rn factl~tles
like oil platforms and shlps
Th~s
spate5 optimization process ts beyond the scopeof this book How-
ever,whenapplymgRCMtoanexrsttng Lacrhty,onehastostartsomewhere
In most caws, the best way to deal w~th
spares is as follows
use RCM to develop a maintenance strategy based on existing spares
holding policies,
* reviewthefailuremodes associatedwithkey spareson anexceptionbzsis,
by establishingwhat impact (if any)a change in the present stockholding
policy wouldhave onthe initial maintenance strategy,and then picking
the most cost-effective maintenance strategy/spares holding policy.
If this approach is adopted, then the existing spares holctingpolicy can be
seen as part of the (initial) operating context.
Mczrket errn nand
Theoperating context sometimesfeatures cyclic variations indemand for
the products or services provided by the organisatton.
Forexample, soft drink companiesexperiencegreaterdemandfor their products
in summer than inwinter, while urbantransport companiesexperience peak de-
mand during rush hours.
In caseshkethese, the operatlonai consequences of falluseare much nlore
senous 'it the tlmes of peak dernand, so111this type of ~nd~~rtry,
tht aqect
of the opelating context needs to be especially clearly undertood when
definlng functions and assesrng fallure consecjuences
Haw mat~>rial
szrpply
Sometimes the operating context is influenced by cyclic fl~tctuations
in
the supply of raw materials. Food manufacturers often experience peri-
ods of intense activity during harvest times anct periods of little or no
activity at other times. This appliesespecially to fruit processors and sugar
mills.Duringpeakperiods, operational failures not only affect output, but
can lead to the loss of large quantities of raw materials if these cannot be
processcd before they deteriorate.
Documenting the opercttirzg content
For all the abovereawns, it 1
sessential to ensure that everyone ~nvolved
in
the development of a maintenanceprogram for any asset f~~lly
uncle~stanci
the operating context of that asset. The best way to do so 1sto document
the operating context,if necessary up to and ~ncluding
the cwerall miion
statement of the entire organlsatlon, as part of the RCM proccs,
Figure 2.8 overleaf shows a hypothetical operating context statement for the
grinding machine mentioned earlier. The crankshaft is used in a type of engine
used in motor car model X.

24. Makecar ModelXdivisionemploys4000peopletoproduce220000carsthisyear. Sales
model X forecastsindicatethatthiscouldriseto320000peryearwithin3years.Weare
(Corresponding now number 18innationalcustomer satisfactionrankings,andintendto reach
asset: Mod.elX 15thplacenext year and 10thplacethe followingyear. Thetarget for losttime
Car Division) injuriesthroughout the divisionis one per 500 000 paid hours.The probability
of a fatality occurringanywhere in the division should be less than one in 50
years.The divisionplansto conformto all known environmentalstandards.
Make The Motown EnginePlantproducesallthe enginesfor modelX cars. 140000
engines Type 1and80000Type 2 engines areproducedper year. Inorderto achieve
(Corresponding thecustomersatisfactiontargetsfortheentirevehicle,warrantyclaimsforengines
asset: Motown mustdropfromthepresentlevelof20per1000to5per 1000.Theplantsuffered
EnginePiant) three reportableenvironmental excursions last year - our target is not more
thanone inthe nextthreeyears. The plant shuts downfor two weeks peryear
to allow productionworkers to take their main annualvacations.
MakeType 2 TheType 2 enginelinepresentlyworks 110hoursperweek (2x 10hr shifts5
engines days per week and one 10hour shift on Saturdays). The assembly line could
(Corresponding produce 140000enginesper year inthese hoursif it rancontinuously with no
asset:Jype 2 defects, but overalloutput of engines is limitedby the speedof the crankshaft
Engine Line) manufacturingline.Thecompanywouldlikeas muchmaintenanceaspossible
to be done duringnormalhourswithout interferingwith production.
Machine Thecrankshaft lineconsistsof 25operations,andisnominallyableto produce
crankshafts 20crankshaftsperhour(2200perweek, 110000per50weekyear).Itcurrently
(Corresponding sometimesfails to producethe requirement of 1600 per week in normaltime.
asset: Crank- Whenthis happens,the line hasto work overtime at anadditionalcost of £800
shaftmachini~~g
per hour. (Sincemost of the forecast growth will befor Type 2 engines, stop-
line2) pagesonthislinecouldeventuallyleadto lostsalesof modelX carsunlessthe
performanceis improved,)Thereshouldbenocrankshaftsstoredbetweenthe
endof the crankshaft lineandthe engineassemblyline,but operations infact
keepa palletof about60crankshaftsto providesome'insurance' againststop-
pages.Thisenablesthecrankshaftlinetostopfor upto3hourswithoutstopping
assembly.Crankshaftmachiningdefectshavenotcausedanywarrantyclaims,
but the scrap rate on this lineis 4%. The initialtarget is 1.5%.
Finishgrind Thefinishgrindingmachinegrinds5mainand4bigendjournals. itisthebottle-
crankshaft neck operationon the crankshaft line, and the cycletime is 3.0 minutes.The
mainand big finisheddiameter of the mainjournals is 75mmf O.lmm, and of the bigends
endjournals 53mm t0.1 mm. Bothjournals have a surface finish of Ra0.2. The grinding
(Corresponding wheelsaredressedeverycycle,aprocesswhichtakes0.3minutesoutof each
asset: Ajax 3 minute cycle.The wheels needto be replacedafter 3 500 crankshafts, and
Mark5 grinding replacement takes 1.8hours. There are usually about ten crankshaftson the
machine) conveyor betweenthis machine and the next operation, so a stoppage of 25
minutescanbetoleratedwithoutinterferingwiththe nextoperation.Totalbuffer
stocksontheconveyorsbetweenthismachineandtheendofthelinemeanthat
this machine can stop for about 45 minutes beforethe line as a whole stops.
Finishgrindingcontributes0.4% to the present overallscrap rate.
Figure 2.8: A n operating context statement
The hierarchy starts with the division of the corporation which produces this
model, but itcouldhavegone up one levelfurtherto includethe entire corporation.
Notealsothat acontext statement atany level should apply to allthe assets below
it in the hierarchy, not just the asset under review.
The context statement4 at the hrghe~levels 111 this hierarchy are 41mply
broad function statements Perfornlancc stand:uds at the highest levels
quantlfyexpectationsfrom the vrcwpornt of the overall b~islnessAt lower
levels, performance 4tandards become 5tead1ly more specrf ic uiltil one
reaches the asset under review. The prlmary and secondary functlons of
the asset at this level are defined as dcscrrbed 111 the lest oSthis chapter
2.4 Different Types of Functions
Every phyr~cal
asset has rnore than one -often eve1al -functlons. It the
objective of maintenance is to ensure that the asset cdn continue to fulfll
thesef~~nct~ons,
then they mustall be identifiedtogether withthert current
desired standards of performance At t~rst
glCtitce,
tllrs may seem to be a
Safalrly straightforward exercle However In pr;lcttce ~t nearly alwdys
turns out to be the s~ngle
most chnllengrngand tlme-consu~~~tng
'Ispect ot
the maintenance strategy tornlulation piocess,
'Th~s
isespecially t~ue of older facilities Procluctsch'inge, plantcoltfig-
iiratlons change, people change, techrlology changea arid performance
expectattonschange-but dill we find assetsm service that have been there
sincethe plant was bullt Detinrng precisely what they are itpposed to be
dorng rzovv requlres very close cooperatlon between irlamtatneis ancl user.
It 1salso usually a profound learning expenence for everyone lnvolved
Functions are divided Intotwo main categories (primary dnd second-
aryfunctlons)andthen furtherdivlded into various sub-categorles These
ale revlewed on the followtng pages. starting with prlrnitry functrons
Primary functions
Organlsatlons acqurrephysical assets forone,posstbly two, seldom more
than three maln reacons These 'reason%'are def~ned
by s~iit~ihly
worded
functron ctatements Because they are the 'maln' reasonr why the asset is
acclurred,they areknown a%yrirnaryf~~rzctior2s
They atethe reasons why
the asset exrsts at all, so care shoulcl be taken to dehne them a prectsely
as posslble

25. Primary functionsare usually fairly easy to recognise.In fact,the names
of most industl-ialassetsare based on their primary functions.
For instance the primary function of a packing machine is to pack things, of a
crusher to crush something and so on.
As mentioned earlier, the real challenge lies in defining the current per-
formanceexpectatio~rs
associated with these f~inctions.
For most types of
equipment, the performance standards associated with primary functions
concernspeeds,volitmesandstoragecapacities.Productqualityal$ousually
need to be consiclered at this stage.
Chapter 1mentioned that our abilityto achieve and su~tain
satisfactory
quality stalldardadepends increasingly onthe capability andcondition of
the assets which produce the goods. These standards are usually associ-
ated with primary fi~nctions.
As a result, take care to incorporate product
quality criteria into primary function statementa where relevant. These
includedimerzsio~ls
for machining, fonning or assembly operations,purity
st~~r~cl~ir~i~s
for food, chemicals andpharmaceuticals, hur~lness
in the case
of heat treatment, filling levels or weights for packaging, and so on.
E'unctionlz( block di~igr~lms
If an asset is very complex or ifthe interaction between different systems
is poorly tincterstood,it is sometimeshelpful to clarify the operating con-
text by drawing up functional block diagrams.Theseare simplydiagrams
showi~lg
alltheprimary functions of anenterpriseatanygiven level.They
are discussedin more detail in Appendix 1.
Mttltijjle irzciej~enrlt~nt
primary fLinctions
An asset can have more than oneprimary function. For instance, the very
nameof amilitaryfighter/bornbersuggeststhatithastwoprimaryfunctions.
In such cases, both should be listed in the functional specification.
A similarsituation is oftenfound in manufacturing,wherethe sameasset maybe
usedto performdifferent functions at different times. For instance,a single reactor
vessel in a chemical plant might be used at differenttimes to reflux (boilcontinu-
ously)threedifferentproductsunderthreedifferentsets of conditions,asfollows:
Product 1 2 3
Pressure 2 bar 70bar 6bar
Temperature 180°C 120°C 140°C
Batch size 500 litres 600litres 750litres
(Itcouldbesaidthatthisvessel isnotperformingthreedifferentfunctions, butthat
it isperformingthe samefunctionto different standardsof performance.Infact, the
distinctiondoes not matterbecausewe arriveatthe sameconclusioneitherway.)
In cases like this, onecould list aseparate function statementforeachpro-
duct. This would logically lead to three separate maintenance programs
forthe sameasset.Threeprograms may be feasible-perhaps evendesirable
- if each product runs continuouslyfor very long periods.
However, if the interval between long-tenn maintenancetasks is longer
than the change-over intervals, then it is impractical to change the tasks
every time the machine is changed over to a different product.
One way around this problem is to combine the 'worst case' stanclards
associated with each product into one function statement.
Intheaboveexample,acombinedfunctionstatementcouldbe'to refluxupto 750
litres of product at temperatures up to t80°Cand pressures up to 10 bar.'
This will leadtoamaintenance program which rnight crnbody sotneover-
maintenance some of the time, but which will ensure that the suet can
handle the worst stresses to which it will he exposed.
Serial or dependent pnnzary functions
One often encounters asset wh~ch
must perform two or more priInary
functions in series. These are known as serial functions.
For instance,the primary functions of a machine in a food factory may be 'to fill
300cans with food per minute' and then 'to seal 300 cans per minute'.
The distinction between multiple primary fitnction and serial primary
functions is that in the former case, each function can be performed inde-
pendently of the other,while in the latter, onefunction must be perforined
before theother.Inotherwords, forthecanning machine to work properly
it must Sill the cans before it seals them.
Secondary Functions
Most assetsareexpectedtofulfil oneormorefiinctions in aciditiontotheir
primary functions. These are known as secorzdaryfurzcfiorzs.
For example,the primary function of a motor car might be described as follows:
to transport up to 5people at speeds of up to 140 kmih along made roads
If this was the only function of the vehicle, then the only objective of the mainte-
nance programfor this car would beto preserve its ability to carry up to 5 people
atspeedsof up 140kmihalongmaderoads.However,thisisonlypartof thestory,
because most car owners expect far more from their vehicles, ranging from the
ability to carry luggage to the ability to indicate how much fuel is in the fuel tank.
To help ensure that none of these fi~nctions
are overlookect,they are divi-
ded into seven categories as follows:

26. environmental integrity
safety/structural integrity
contro~containment/comfort
appearance
protection
econornylefficiency
supedluous functions.
The first letters of each line in this list form the word ESCAPES.
Although secondary f~~rictions
are usually less obvious than primary
filnctions, the loss of a secondary function can still have serious conse-
quences- sometimes more serious than the loss of aprirnary function. As
3 result, secondary f~mctions
often need asmuch if not inore maintenance
thanprimary functions, sothey toomustbeclearly identified.Thefollow-
ing pages explore the main categories of these functions in more detail.
Envimnnzentcil integrity
Part 2of this chapterexplained how society'senvironmental expectations
have become a critical feature of the operat~ng
context of many assets.
RCM begins the process of compliance with the associated standards by
incorporating them in appropriately worded function statements.
For instance, one function of a car exhaust or a factory smoke stack might be 'to
containno morethan X microgramsof a specifiedchemical percubic meter'.The
car exhaust system might also be the subject of environmental restrictionsdeal-
ing with noise, and the associated functional specification might be 'to emit no
more than X dB measured at a distance of Y metres behind the exhaust outlet'
SafeV
Most users want to be reasonably surethat their assets will not hurt or kill
them. In practice, most safety hazards emerge later in the RCM process
asfailuremodes. However, in somecases it is necessary to writefunction
statements which deal with specific threats to safety.
For instance,two safety-relatedfunctions of a toaster are 'to preventusersfrom
touching electrically live components' and 'not to burn the users'.
Many processes and components are unable to fulfil the safety expecta-
tions of users on their own. 'This has given rise to additional f~~nctions
in
the formof protective devices.Thesedevicespose someofthe most diffi-
cult and coinplex challenges facing the maintainers of modern indttstrial
plant. As a result, they are dealt with separately below.
A Yurthe1 subsel of safety- elated finrctronr are thoie whlch deal wlth
product contammatlonand hygiene Tliese are most oftenfound 111the food
andpharmaceutical ~ndustries.
The associatetiperformance tandLirdrare
usually tightly spectfied, and lead to r~gorot~s
and c(3rnprehenslveinain-
tenance routines (cleaning and testlng/val~dat~oii)
Structural integrity
Many assets have a structural secondary function. This usually involves
slipporting some other asset, sub-system or component.
For example, the primary function of the wall of a building might be lo protect
peopleandequipmentfrom the weather, but it mightalso be expected to support
the roof (and bear the weight of shelves and pictures).
Large, coniplex ctructure with multiple loact bc:trtng path dntl high
levels of redundancy need to be anttlyscd using a pecialised erlon of
RCM Typicalexamples of such tnlctures are airfr~trrte,
the 1-1~111of tilp
and thc ~tructural
element, of offshore or1platfortrli
Structures of this type are rare In rndustry 111 geiieral, o the relevant
analytical techniques are not covered In thrc,book FTowever,ti'l~ghtfot-
ward, single-celled structural elements can be analysed In the ame w~iy
as any other functlon ciesciibed in th~s
chapter
Control
In inany cases, users not onIy want assets to fulfil functtons to a glven
standard of performance, but they alsowant to be able to regulate the per-
fonnance. This expectationis summarised in separatefunction statement.
For instance,the primary functionof a car as suggested earlier was 'to transport
upto 5peopleat speeds of upto 140kmihalongmade roads'. Onecontrolfunction
associated with this function could be 'to enable driver to regulate speed at will
between -15 km/h (reverse) and +I40km/hl.
Ind~cation
or feedback forms an important subset of the control category
of functions. T h ~ s
includes tunctrons whlch provlde operators with real-
time intormatron about the proce7s (gauge, ~ndtc~ltors,
telltales, VDU's
and control panels), or which record such tnfor~rl,irionfor later analysis
(d~grtal
or 'tnalog recording devices, cockpit voice recottiers in iurcrdft,
etc) Peiforrnance standards assoc~ated
w~th
thew functtoni,not only le-
late to the ease with which it should be possible to read and dsstrntlateoi
to playback the ~nformation,
but alo cover it ~zccuracy
For instance,the function of the speedometerof a car might be described as 'to
indicate the road speed to the driver to within -1-5
-0% of the actual speed'.