Professional Documents
Culture Documents
Maintenance Engineering and Maintainability: An Introduction
Maintenance Engineering and Maintainability: An Introduction
net/publication/227046791
CITATIONS READS
35 31,733
1 author:
Krishna B. Misra
RAMS Consultants
132 PUBLICATIONS 3,465 CITATIONS
SEE PROFILE
All content following this page was uploaded by Krishna B. Misra on 07 May 2016.
Krishna B. Misra
RAMS Consultants, Jaipur, India
Abstract: Maintenance is another important aspect of system performance after reliability. There are
several facets of the the maintenace management and in this introductory chapter, we would like to have
these surveyed. Broadly speaking, maintenance is the process of maintaining equipment in its operational
state either by preventing its transition to a failed state or by restoring it to an operational state following a
failure. This leads to various types of maintenace activities that can be planned to realize the objective of
maitenance, such as preventive, predictive or corrective. Recent developments in the maintenace
engineering and management are also discussed in this chapter.
Maintenance Task: Maintenance tasks [12,13,15] parameter to verify if the maintenance was up to
are generally of two types, viz., planned and the standard is to determine availability of the
unplanned. The planned maintenance can be system since the maintenance is result dependent
preventive or corrective (including deferred and the test lies in effectiveness than efficiency.
maintenance) whereas the unplanned maintenance
is mainly corrective, which includes any
emergency maintenance. Again the preventive 46.1.2 Maintenance Philosophy
maintenance can be scheduled maintenance and the
other can be based on condition of the equipment- Based on the timing and the work contents
known as condition-based maintenance. Therefore, involved in the maintenance task, different
the form of the maintenance can be: maintenance philosophies [13, 15] can be put in the
1. Pre-planned maintenance: This includes following categories. viz.,
early maintenance tasks such as cleaning, 1. Timing known, Content known: Preplanned
greasing, lubricating, zero-setting and Maintenance (PPM), Planned Shut downs,
recording key measurements. This is often Routine Inspections, Schedules Change-outs
conducted by non-maintenance staff and fall in this category.
observed equipment deterioration that cannot 2. Timing known, Content unknown: Statutory
be corrected will be reported to regular Surveys, Third Party Inspections, Condition-
maintenance staff. This is also called First- based maintenance
line maintenance. 3. Timing unknown, Content known:
2. Planned maintenance: This is also known as Anticipated Maintenance work, Contingency
scheduled maintenance and its timing and work awaiting shutdown, Run to destruction
scope are both known in advance. 4. Timing unknown, Content unknown:
3. Shutdown maintenance: It is a planned Breakdown maintenance, immediate Repairs
maintenance but is carried out when arising from inspection, Run to failure.
production or plant is shut down. From the point of view of work management,
4. Breakdown maintenance: This is carried out activities falling in category 1 are most welcome
when equipment fails to meet its desired and in category 4 are least welcome. It is always
function. This may involve repairs, advantageous to shift work to a more manageable
replacements or adjustments as considered category. By using effective maintenance regime,
necessary. the intention should be to keep category 4 as empty
5. Emergency maintenance: This is carried out as possible. it is worthwhile to observe that in “run
only when either inspection or breakdown to failure”, in which equipment is run till it fails
maintenance has identified its necessity. and repairs are carried out but work content is
Inspection: All plant machines, equipment and unknown. However, in “run to destruct” category,
structures are subjected to regular inspection in which equipment is run till it fails and is
scheduled to detect performance or safety problems discarded, a replacement is required thus work
and to ensure that all items receive necessary content is known. Also by careful evaluation of
maintenance. maintenance histories and shutdown programmes
the timing uncertainty in category 3 can be
Verification: Verification is a process, not a minimized. Such work followed by full job
measurement, and it has two primary objectives: preparation can provide a fast response when
• To check that the maintenance work is being situation finally requires it. Also the requirements
done in category 2 can be anticipated with care once
• To confirm that maintenance standards have operations are routinely established.
not been compromised. Out of all maintenance work preplanned
Verification is usually done or supplemented by (in maintenance (category 1) is most widely used but it
addition to in-house verification) a third party (to is often argued that this method requires work to be
be completely impartial or fair). One of the key
Maintenance Engineering and Maintainability: An Introduction 749
done that is not necessary since it is done on the This is resulting in change in the concept of
basis of calendar irrespective of machine condition. maintenance organization and responsibilities.
Run to destruct (category 3) the equipment is There is a tremendous increase in the number, size
usually used until it fails and is discarded or and variety of physical assets (plants, equipment
replaced. The works requires switching off the and buildings), which have be maintained.
machinery or isolate it, remove or replace the old There is a rapidly growing awareness of product
unit and connect the new unit. Work management quality affecting maintenance activities and of the
is straight forward as materials technical content is extent to which equipment failure can affect safety
all known. and the environment, coupled with the requirement
Breakdown maintenance (category 4) is simple of achieving high plant availability at reduced
and can be applied quickly with limited resources costs. With the increase in complexity and
and information but work arising from unexpected sophistication, coupled with organizational changes
breakdown is difficult to manage and may involve have resulted in changes the maintenance work is
high costs. Breakdown maintenance is not safe and carried out today. These changes are testing
may sometimes involve danger to life. attitudes and skills in all sectors of industry to their
limits. Management and engineers [7,11] are
46.1.3 Maintenance Scope Changed with Time beginning to adopt completely new ways of
thinking and strategies towards maintenance, so
Over the past several decades starting from the pre- that they can evaluate them sensibly and apply
World War II period, the concept and scope of those likely to be useful to them and their
maintenance has changed considerably, more than companies. In fact, the changes that have taken
any other management discipline. These changes place during this period are well documented by
can basically be attributed to the advent of more Moubray [14], who split this period into three
complex designs of systems requiring new distinct periods (which Moubray [14] calls as
maintenance techniques. generations I, II and III) and are summarized in
Table 46.1.
• A major shift in organizational thinking efficient and cost-effective operation over the life
towards participation, team-working and of a product.
flexibility. Maintenance has been categorized based on the
A major challenge facing maintenance people nature and purpose of the maintenance work and its
nowadays is not only to learn what these frequency. Generally, there are four types of
techniques are, but to decide which are worthwhile maintenances in use, viz., preventative, corrective,
and which are not in their own organizations. If we predictive and fault-finding.
make the right choices, it is possible to improve Maintenance can also be classified according to
asset performance and at the same time contain and the degree the maintenance work is carried out to
even reduce the cost of maintenance. If we make restore the equipment in relation to its original
the wrong choices, new problems are created while state. This leads to the following categorization:
existing problems only get worse. Perfect Maintenance: is one which restores the
In fact RCM provides a framework which enables equipment to as good as new condition.
users to respond to these challenges, quickly and Minimal Maintenance: results in equipment having
easily. Since every physical asset is put into service the same failure rate as it had before the
because someone wants it to do something, they maintenance action was initiated. This is called- as
expect it to fulfill a specific function or functions. bad as old state.
When an asset is maintained, obviously the state Imperfect Maintenance: is one in which the
we wish to preserve is the one in which it continues equipment is not restored to as good as new but
to do whatever its users want it to do. relatively younger (a state in between as good as
Of recent, emphasis is being placed on core new and as bad as old).
business by several major companies of the world Worse Maintenance: This type of maintenance
and under this thinking; companies are transferring results (unintentionally) in an increase of
previously in-house functions to external specialist equipment’s failure rate or actual age but does not
companies. Use of experts or specialists and result in break down.
reasons for choosing in-house teams on-site Worst Maintenance: This type of maintenance
accommodated teams or manned-up technicians results (unintentionally) results in equipment’s
brought from outside for the completion of a job, is breakdown.
becoming routinely common these days. In the foregoing classification maintenance can be
Subcontracting for maintenance job is also preventive (PM) or corrective (CM) and
becoming quite common in the globalizing world accordingly the PM or CM would be belong to one
of trade these days and the number and size of of the above categories. The have been discussed in
companies engaged in maintenance work is nicely in several texts [20, 22, 24].
growing very fast. In this handbook, we have included Chapter 47
and Chapter 48 on some aspects of maintenance
modelling and optimization as well as on the trends
46.2 Approaches to Maintenance on maintenance technology and management.
equipment on a regular basis based on the expected It is important to make it explicitly clear that if a
life of the equipment and the frequency of the component has a constant failure rate (i.e. defined
maintenance is generally constant. Preventive by an exponential distribution), then preventive
maintenance is a schedule of planned maintenance maintenance of the component will have no effect
actions aimed at the prevention of breakdowns and on the component's failure occurrences.
failures since the primary objective of preventive In fact, the objective of a good preventive
maintenance is to prevent the failure of equipment maintenance program is to either minimize the
before it actually occurs. For example, lubrication overall costs (or downtime, etc.) or meet a
of mechanical systems is done after a certain reliability/ availability goals. In order to achieve
number of operating hours or replacement of this, an appropriate interval of time must be
lightning arresters in jet engines after a certain determined for the scheduled maintenance. One
number of lightning strikes. It is designed to way to do that is to use the optimum age
enhance the equipment reliability by replacing replacement model, where the model satisfies the
worn components before they actually fail. conditions mentioned previously, viz.,
Preventive maintenance activities include • The unit is exhibiting behavior associated
equipment checks, partial or complete overhauls at with a wear-out mode. That is, the failure rate
specified periods, oil changes, lubrication and so of the unit is increasing with time.
on. In addition, workers can record equipment • The cost for planned replacements is
deterioration so they know which worn out parts to significantly less than the cost for unplanned
replace or repair before they can cause system replacements.
failure. The technological advances in tools for Long-term benefits of preventive maintenance
inspection and diagnosis have enabled even more include:
accurate and effective equipment maintenance. The • Improved system reliability.
ideal preventive maintenance program would
• Decreased cost of replacement.
prevent all equipment failure before it occurs.
• Decreased system downtime.
Preventive maintenance [17] is a logical alternative
of choice if the following two conditions are • Better spares inventory management.
satisfied: Thus long-term effects and cost comparisons
usually favor preventive maintenance over
• The equipment in question has an increasing
performing maintenance actions only when the
hazard rate. In other words, the hazard rate of
system fails.
the equipment increases with time, implying
a wear-out situation Preventive maintenance 46.2.1.1 Maintenance Policies
of a component that is assumed to have an
The literature is full of numerous maintenance
exponential distribution (which implies a
models [22] that have been presented by
constant failure rate) does not make sense!
researchers using various assumptions and cost
• The overall cost of the preventive
structures but all of them can be broadly
maintenance action must be less than the
categorized under certain maintenance policies.
overall cost of a corrective action. The
These policies, however, are with regard to single
overall cost for a corrective action includes
equipment.
ancillary tangible and/or intangible costs,
Age-dependent PM Policy: Under this policy, PM
such as downtime costs, loss of production
is carried out at some predetermined age or
costs, lawsuits over the failure of a safety-
repaired (CM) upon failure. PM or CM can be
critical item, loss of goodwill, etc.)
perfect, minimal or imperfect.
As stated earlier, if the unit has an increasing
Periodic PM Policy: Under this policy,
failure rate, then a carefully designed preventive
equipment’s PM is carried out at fixed time
maintenance program may improve system
intervals regardless its failure history.
availability. Otherwise, the costs of preventive
Failure Limit Policy: Under this policy the PM is
maintenance might actually outweigh the benefits.
performed only when the failure rate (or some
Maintenance Engineering and Maintainability: An Introduction 753
index of performance) reaches a predetermined inspection, all failed equipment are replaced with
level and the intervening failures are repaired when new ones and all functioning equipment are
they occur. serviced so that they become as good as new.
Sequential PM Policy: Under this policy, Opportunistic Maintenance Policies: Under this
equipment’s PM is carried out at unequal time policy in multi-component system, it is possible to
intervals, which go on becoming shorter with the do PM of non-failed equipment at a reduced
age. additional cost while failed equipment is being
Repair Limit Policy: Under this policy, two cases repaired. There are many variations the in strategy
arise, viz., repair cost limit policy and the other is under this category.
repair time limit policy. Under the former, the Warranty Models: Maintenance needs to be
repair cost is assessed when the equipment fails incorporated in the warranty models so that if
and repair is performed only if the repair cost is warranted equipment fails, the failed components
less than a predetermined limit, otherwise the that caused the equipment failure will be replaced
equipment is replaced, whereas under the repair and in addition PM can be carried out to reduce the
time limit policy, the limit is set on the repair time chances of failure in future. Therefore, warranty
instead of cost. policies [23] with integrated PM should be
Repair Number Counting Policy: Under this preferable.
policy, the equipment is replaced at kth failure and It can easily be shown that the corrective
the first (k-1) failures are removed by minimal replacement costs increase as the replacement
repair. Upon replacement the process repeats. interval increases. In other words, the less often we
As said before, the foregoing maintenance perform PM, the higher will be the corrective costs.
policies were applicable to single equipment with Obviously, the longer we let a equipment operate,
increasing failure rate. Recently, there has been its failure rate increases to a point where it is more
increasing interest in multi-equipment maintenance likely to fail, thus requiring corrective actions.
models. In fact maintenance of a multi-equipment However, just the opposite is true for the
system differs from that of single equipment in that preventive replacement costs. The longer we wait
there exists economic or failure dependence. Due to perform PM, less shall be the costs. On the
to former, the PM non-failed subsystems can be other hand, if we do PM too frequently, the costs
performed at reduced cost while the failed would go up. If we combine both costs, it is easy to
subsystems are being repaired. In case of failure realize that there is an optimum point that
dependence or correlated failures, failure of one of minimizes the costs. In other words, one must
the subsystem may affect the functioning of other strike a balance between the risk (costs) associated
subsystems. In case of a system with number of with a failure while maximizing the PM interval
subsystems or equipment, the following analysis.
maintenance policies are applicable: Generally, preventive maintenance (PM) is
Group Maintenance Policy: Under this policy considered beneficial but it must be mentioned that
there are three different cases: there are risks of equipment failure and human
T-Age group Replacement Policy: The units or errors committed while performing PM, just as in
equipment are replaced when the system is of age any other maintenance operation. This logic
m-Failure group policy: this calls for a system dictates that it would cost more for regularly
inspection after m- failures have occurred scheduled downtime and maintenance than it
m-Failure and T-Age Policy: This policy combines would normally cost to operate equipment until
the advantages of m-failure and T-age policy and repair is absolutely necessary. This may be true for
calls for a group replacement when the system is of some components; however, one should compare
age T or when m-failures have occurred, whichever not only the costs but the long-term benefits and
comes first. This policy requires inspection at savings associated with preventive maintenance.
either the fixed age T or the time when m-failures Without preventive maintenance, for example,
have occurred, whichever comes first. At costs for lost production time from unscheduled
754 Krishna B. Misra
equipment breakdown will be incurred. Also, inspections are performed while equipment is in
preventive maintenance results in savings due to an service, thereby minimizing disruption of normal
increase of effective system service life. system operations. This type of maintenance is
There are several excellent references available generally carried out on mechanical systems where
for preventive maintenance modelling and analysis historical data is available for validating the
but the books by Nakagawa [20], Wang and Pham performance and maintenance models for the
[22] and Jardine and Tsang [24] stand out systems and the failure modes are known.
distinctly. The book by Jardine and Tsang [24] Predictive maintenance (PdM) helps determine the
presents different optimal replacement policies and condition of in-service equipment in order to
models, spare parts provisioning besides discussing predict when maintenance should be performed.
optimal inspection policies under various The predictive component of the term predictive
conditions. Besides capital equipment replacement maintenance comes from the goal of predicting the
decisions and maintenance resource requirements future trend of the equipment's condition. This
are also presented. Nakagawa and Wang and Pham approach uses principles of statistical process
[20,22] describes perfect and imperfect preventive control to determine at what point in the future
maintenance models and several optimum maintenance activities will be appropriate.
preventive maintenance policies. Inspection models In fact, for condition monitoring or predictive
with different types of units such as standby, technique any relevant means is acceptable to
storage and different types of failures such as determine equipment condition, and to predict
extended failures (catastrophic, partial or degraded) potential failure, which may even include the use
and intermittent failures. Faults revealed or of the human senses (appearance, sound, feel,
unrevealed are also considered. In fact both these smell etc.), machine performance monitoring, and
books present an up-to-date status of various statistical process control techniques. But the other
models that have been available so far. sophisticated technologies used in monitoring
In this handbook, we have included Chapter 49 include:
by Nakagawa on Replacement and Preventive • Vibration Measurement and Analysis
Maintenance models. • Acoustic Emission (Ultrasonics)
• Oil Analysis
46.2.2 Predictive Maintenance • Infrared Thermography
• Motor Current Analysis
Predictive maintenance (PdM) or condition based These will be explained in some details in the
maintenance (CBM) is carried out only after following paragraphs.
collecting and evaluating enough physical data on
performance or condition of equipment such as 46.2.2.1 Vibration Measurement and Analysis
temperature, vibration or particulate matter in oil This technology owes its popularity to Fast Fourier
etc. by performing periodic or continuous (on-line) Transform (FFT) technique developed in 1964,
equipment monitoring. Analysis is then performed when spectrum analyzers became available that
on the collected data to prepare an appropriate could be used with special transducers to measure
maintenance plan. PdM technologies used to machine vibrations. When portable FFT-based data
collect information of equipment condition can collectors became available around 1980, the use
include infrared, acoustic (partial discharge and vibration as a tool for machinery fault diagnosis
airborne ultrasonic), corona detection, vibration went a sea change and found its applications in the
analysis, sound level measurements, oil analysis petrochemical, electrical power, paper and other
and other specific online tests. process industries. Today, On-line Vibration
The basic aim in PdM [18] is to perform Analysis System (rather than portable analyzers) is
maintenance at a scheduled point in time when the widely used for monitoring vibrations with
maintenance activity is most cost effective but piezoelectric accelerometers as vibration probes
before the equipment fails in service. Most PdM and computer-based specialized data acquisition
Maintenance Engineering and Maintainability: An Introduction 755
systems for collecting, storing and archiving FFT Lubricating oil contains a good deal of
data. The on-line dynamic vibration monitoring information about the envelope in which it
system enables mechanical fault detection at the circulates. Wear of metallic parts, for example,
earliest detectable time. But the system installation produces a lot of minute particles, which are
costs are usually high because of the capital cost of carried by the lubricant. Wear means the loss of
the hardware as well as the installation labour. solid material due to the effects of friction of
However, vibration analysis is the most appropriate contacting surfaces. These small metal particles
solution on high speed rotating machines and can can give information about the machine parts that
be the most expensive part of a PdM program. It are wearing, and can be detected by various
still remains the most widely used method for methods, such as atomic emission spectrometry.
detecting rubbing in rotating machines such as Determination of larger particles can be done using
electrical power generating turbines through optical or electronic microscopy, or ferrography.
changes in amplitude and phase of the 1X (rotating The acidity of oil can indicate whether the oil has
frequency) vibration component. been oxidized as a result of operation at high
temperature, if there is a high percentage of
46.2.2.2 Acoustic Emission
moisture, or due to oil having been in service for
Acoustic emission (AE), particularly, for bearing long.
diagnosis is becoming quite indispensable. The viscosity of the oil is also an important
Sometimes acoustical analysis may also be done at parameter and must be in conformity with the
sonic or ultrasonic levels. Sonic technology is requirements of the machine. The alkalinity or the
useful for mechanical equipment whereas loss of alkalinity of the oil proves that the oil is in
ultrasonic equipment is useful for detecting contact with some inorganic acids such as
electrical problems. Ultrasonic testing is a sulphuric or nitric acid. Oil undergoes destructive
relatively new predictive technology. It is capable changes in property when it is subjected to oxygen,
of detecting sounds that lie outside the human combustion gasses and high temperatures.
hearing range, and are indicators of failing Viscosity change, as well as additive depletion and
mechanical conditions. As a result, ultrasonic oxidation occur to degrade the oil.
testing has become an essential part of predictive Several methods are used to analyze oil
maintenance program. Ultrasonic inspections can condition and contamination. These analyses may
help detect leaking gases or fluids in heat include spectrometry, viscosity analysis, dilution
exchangers, compressors, and valves. This analysis, water detection, acid number assessment,
technique not only locates leaks in steam traps and base number assessment, particle counting, and
valves, but can also be helpful in detecting certain microscopy. Oil and wear particle analysis is a
electrical faults, such as arching and corona. combination of spectrometric, ferrographic, and
filter analysis. Oil and wear particle analysis can
46.2.2.3 Oil Analysis
detect abnormal wear modes particularly in
Oil analysis in relevant cases can be more reliable aviation systems, long before the wear can cause
technique in predictive maintenance. The early use any serious damage.
of oil analysis dates back to the early 1940s by the The oil data modelling and analysis can be of
railway companies in the Western United States, great help in fault detection procedure and we have
where the technicians used simple spectrographic included Chapter 50 in this handbook on this
equipment and physical tests to monitor aspect.
locomotive engines. When diesel locomotives
46.2.2.4 Infrared (IR) Thermography
came in use, oil analysis practices by railways
became very frequent. By the 1980s oil analysis Infrared (IR) thermography is used to detect
formed the basis of condition-based maintenance temperature changes in bearings and shafts. The IR
(CBM) in most railways in North America. thermographic analysis detects abnormal
temperatures that may signify corrosion, damaged
756 Krishna B. Misra
wiring, loose connections, and/or insulation a continuous basis, such as fluids processes in
breakdown and hot spots. Infrared monitoring and chemical industries, motor generator systems,
analysis help reduce unexpected electrical and serial trunk conveyor systems, and general line
mechanical equipment (from high to low speed production in manufacturing industries.
equipment) failures. This vital information can Motor Current Signature Analysis (MCSA) is a
alert in advance of the catastrophic failures and is technique used to determine the operating
considered cost-effective technique of monitoring. condition of AC induction motors without
interrupting production. MCSA techniques can be
46.2.2.5 Motor Current Analysis
used in conjunction with vibration and thermal
Motor Diagnostic technologies have become more analysis to confirm key machinery diagnostic
prevalent through the 1990’s in recent time. The decisions.
technologies include: Motor Circuit Analysis MCSA works on the principle that induction
(MCA) and Motor Current Signature Analysis motor circuits can actually be viewed as a
(MCSA) applied to both energized and de- transducer. By clamping a Hall Effect current
energized electric motor systems. Motor current sensor on either the primary or the secondary
analysis techniques are non-intrusive methods of circuit, fluctuations in motor current can be
detecting mechanical and electrical problems in observed. It has been observed that if a high
motor driven rotating equipment. Motor Current resistance exists, (for example due to broken rotor
Monitor (MCM) uses the electric motor of the bars) harmonic fluxes are produced in the air gap.
equipment as a sensor and the information about These fluxes induce current components in the
the equipment is extracted from the line current of stator winding that can cause modulation of the
the motor. The MCM first learns the motor-based supply current at ± the number of motor poles
system for a certain period of time and acquires times slip. Available signal processing techniques
and processes the motor data. These data are stored can help extract the modulating frequency and
in the internal database and a reference model, represent the amplitude relationship of modulating
which consists of parameters and their means and frequency to line frequency. This relationship
standard deviations, is established. Now, during allows one to estimate the presence and severity of
actual monitoring the data being acquired is the defect.
compared with the results stored in the internal
46.2.2.6 Problems in Condition Monitoring
database. If the acquired data is significantly
different from the reference model, it indicates a Although many organizations use sophisticated
fault level that is determined by the magnitude and techniques for condition monitoring these days,
the time duration of the difference. This approach there are still many problems that remain
is used for diagnostics of electrical equipment, unresolved. For example, it is known that the main
especially with rotating components. The main determinant of frequency of condition monitoring
advantages lie in its being a noninvasive is the lead time to failure (or PF interval), which is
methodology for diagnosing health and operations the time at which an incipient failure can first be
of motor-actuated valves, generators, electric detected, until functional failure occurs. For
motors, and other types of electric equipment. example, in case of a bearing, PF interval is the
With new, inexpensive RF components and time interval from when overall bearing vibration
integration techniques, along with advances in levels reach an alarming limit, until the bearing
microprocessor technology, it is possible to seizes completely. In order to be sure that the
provide an economic solution for wirelessly failure is detected prior to the functional failure;
monitoring motor operating parameters - such as the bearing must be monitored at a frequency less
temperature, vibration, current, etc. - for all classes than the PF Interval. But PF interval can hardly be
of motors. The generic technology will find determined accurately. For instance, in case of a
applications in areas where motors, pumps, bearing, the PF Interval may vary depending on the
gearboxes or drive chains need to be monitored on type of bearing installed, the severity of its
Maintenance Engineering and Maintainability: An Introduction 757
inherent design or built-in reliability is and mean • Failure and repair data: Failure modes, failure
downtime reflects how good the maintainability is? and repair distributions.
If the steady state availability (SSA) is to be kept • Repair strategy: number of repair crews,
high, one should try to design for high value of independent repair facilities inspection or
mean uptime (MUT) and mean down time (MDT) overhaul schedules etc.
to be as low as possible since SSA is the ratio of Once the above information is available, one can
mean uptime to mean cycle time which is the sum proceed using any of the following techniques:
of mean uptime and mean downtime. There can be • State Space Approach or Markov Method
several combinations of MUT and MDT which can • Block Diagram Approach
offer the same value of availability. Thus • Conditional Probability Approach
availability is dependent on reliability and • Monte Carlo Approach
maintainability and by clever manipulation one can The detailed information on these approaches can
get the desired value of availability. But the life be found in references quoted above or in [4].
cycle costs must not be lost sight of while However Markov method appears to be more
designing maintained equipment or system. popular technique to analyze maintained systems
There are other measures of performance of the although it suffers the disadvantage of the
maintained equipment such as point availability dimensionality of the formulation even for
and interval availability. moderately sized system. But with the fast
Point availability is defined as the probability computing facilities having large memories
that the equipment or system is available at a given available, this should not be considered as a
point of time and the interval availability is defined disadvantage any more. In any case, one can
as the expected fraction of an interval of specified always decompose a large problem into
length that the equipment or system is in an up manageable sized problems and also use a
state. Naturally if the interval becomes very large, combination of solution techniques rather than
the interval availability would approach steady using just one technique.
state availability. There are other measures such as
frequency of failures and mean duration of failure 46.2.4.4 Availability Trade-offs
of interest in case of maintained systems. The
frequency of failures can also be further classified In designing repairable systems, it is generally
as interval frequency and steady state frequency. desired to optimize system availability subject to
The interval frequency of failure is defined as the some cost constraints. Alternatively, a designer
expected number of times a failure state may be specified a value of availability to achieve
encountered in a specified interval; whereas the and determine the optimal pair of mean time
steady state frequency of failure or simply the between failures (MTBF) and the mean time to
frequency of failure is defined as the expected repair (MTTR) while minimizing the cost of the
number of times a failure state is encountered over system. These formulations are discussed in
a long period of time. Chapter 32 of this handbook and the several
models are discussed in texts like [1, 8]. However,
46.2.4.3 Assessment Techniques it is necessary to have some relationship between
For quantitative analysis of performance measures the variables MTBF and cost and MTTR versus
like reliability, availability or any other measures cost before any tradeoff can take place. Also the
mentioned in earlier section, it is necessary to have lower and upper limits on MTBF and MTTR
the following information about the equipment or should be established from practical considerations
system: as well as the state-of-the-art of the available
• System configuration: such as number of technology. This will help establish the feasibility
units, identical or non-identical, series, pairs for MTBF and MTTR.
parallel, standby, nature of redundancy such
as warm, cold or active etc.
760 Krishna B. Misra
demonstrated to outweigh the involved costs. nor production, but involve only the direct
However, the financial benefits and costs cost of repair.
associated with RCM are difficult to assess due to
the fact that areas of savings are vague as there are
no clear cause and effect relationships in the 46.4 Total Productive Maintenance
evaluation process. Costs can easily be identified
than benefits. Costs include initial outlays It is a well-known fact that in many factories, the
primarily for training, and ongoing costs, including operating time is less than 50% of the gross
maintenance and support personnel, and available hours per year; it is obviously shows that
expenditures associated with the maintenance the assets are not being used to the fullest extent.
introduced as a result of RCM findings. Benefits This is partly due to scheduled downtime, which
can also be identified through a series of steps. includes holidays, no production planned due to
Firstly, one can start by identifying the current limited load, spare capacity to cope with volume
problems that can be resolved through RCM. flexibility etc. The other part is caused by the fact
Secondly, by estimating how much improvement that the production is not wholly efficient. The
would result through the adoption of RCM for the reasons for this can be categorized into losses,
identified problems. Lastly, one should quantify which can be influenced during development and
each of the improvements in the larger sense of production phase. The total productive
company’s performance (profits, plant availability, maintenance is a proactive equipment maintenance
personnel cost, etc.). When that quantification is strategy designed to improve overall equipment
done, then the economic benefits of RCM can be effectiveness .It actually breaks the barrier between
evaluated to see if its adoption is justified or not. maintenance department and production
The value of RCM lies in the fact that it department of a company.
recognizes that the consequences of failures are far Total Productive Maintenance: It is an approach to
more important than their technical characteristics. optimize the effectiveness of production means in a
In fact, it recognizes that the only reason for doing structured manner.
any kind of proactive maintenance is not to avoid TPM focuses on improving the planned loading
failures per se, but to avoid or at least minimize the time. The gap (losses) between 100% and actual
consequences of failures. In fact, the RCM process efficiency can be categorized into three categories:
classifies these consequences into the following Availability, Performance and Yield (Quality
four groups: Rate).
• Hidden failure consequences: Hidden failures Availability losses: These include breakdowns and
have no direct impact, but they may lead to changeovers situations when the line is not running
multiple failures with often catastrophic while it should be.
consequences. (Often these failures are Performance losses: These are basically due to
associated with protective devices which are speed losses and small stops/idling/empty positions
not fail-safe.) when the line is running, but is not providing the
• Safety and environmental consequences: A quantity it should.
failure has safety consequences if it has Yield losses: These occur when the line producing
potential to injure or kill someone. It has products, but there are losses due to rejects and
environmental consequences if it breaches start-up quality losses.
any environmental standard. These losses lead to the Overall Equipment
• Operational consequences: A failure has Effectiveness (OEE) indicator, which tells you how
operational consequences if it affects efficiently the planned production process is. TPM
production, product quality, customer service helps to improve the OEE by providing a structure
or operating costs or cost of repairs. to quantify these losses, and by subsequently
giving priority to the most important ones. TPM
• Non-operational consequences: Failures
provides concepts and tools to achieve both short
belonging to this category nether affect safety
and long term improvements.
762 Krishna B. Misra
Total Productive Maintenance is not the same as a is to maximize plant and equipment effectiveness
maintenance department that repairs breakdowns to achieve the optimum life cycle cost of
(breakdown maintenance). TPM is a critical production equipment.
adjunct to lean manufacturing. If machine uptime Nippondenso of Japan was first to implement
is not predictable and if process capability is not TPM. It had already had quality circles which
sustained, we cannot produce at the velocity of involved the employees in changes. Therefore, all
sales. One way to think of TPM is deterioration employees took part in implementing Productive
prevention and maintenance reduction, not fixing maintenance. Based on these developments
machines. For this reason many people refer to Nippondenso was awarded the distinguished plant
TPM as Total Productive Manufacturing or Total prize for developing and implementing TPM, by
Process Management. TPM is a proactive approach the Japanese Institute of Plant Engineers (JIPE).
that essentially aims to prevent any kind of slack Thus Nippondenso of Toyota group became the
before occurrence. Its motto is "zero error, zero first company to obtain the TPM certification.
work-related accident, and zero loss." TPM has TPM identifies 16 types of wastes (Muda) and
five goals: then works systematically to eliminate them by
1. Maximize equipment effectiveness. making improvements (Kaizen). TPM has 8 pillars
2. Develop a system of productive maintenance of activity, each being set to achieve a “zero”
for the life of the equipment, target. These pillars are:
3. Involve all departments that plan, design, use, 1. Focused improvement (Kobetsu-Kaizen): for
or maintain equipment in implementing eliminating waste.
TPM. 2. Autonomous maintenance (Jishu-Hozen): in
4. Actively involve all employees. autonomous maintenance, the operator is the
5. Promote TPM through motivational key player. It involves daily maintenance
management. activities carried out by the operators
For this concept to function properly, the machines themselves that prevent the deterioration of
must be ready when we need them and they must the equipment.
be shut down in such a fashion as to be ready the 3. Planned maintenance: for achieving zero
next time. Key measures include efficiency while breakdowns.
running and quality. Overall Equipment 4. Education and training: for increasing
Effectiveness (or OEE) tells us how TPM is productivity.
working, not just the typical measures of uptime 5. Early equipment/product management: to
and throughput. OEE is actually the product of reduce waste occurring during the
availability, performance efficiency and the quality implementation of a new machine or the
rate. Operators know what maintenance tasks are production of a new product.
theirs; they also know what tasks are appropriate 6. Quality maintenance (Hinshitsu-Hozen): This
for the skilled trades’ maintenance crew. TPM is a is actually “maintenance for quality”. It
philosophy that helps create ownership of the includes the most effective quality tool of
manufacturing process among all employees. TPM: “poka-yoke”, which aims to achieve
Teamwork is vital to the long-term success of zero loss by taking necessary measures to
TPM. The maintenance group performs equipment prevent loss.
modification that would improve its reliability. 7. Safety, hygiene, and environment: for
These modifications are then incorporated into new achieving zero work-related accidents and for
equipment. The work of the maintenance group is protecting the environment.
then to make changes that will lead to maintenance 8. Office TPM: for involvement of all parties to
prevention. Thus preventive maintenance along TPM since office processes can be improved
with Maintenance prevention and Maintainability in a similar manner as well.
Improvement are grouped as Productive
Maintenance. The aim of productive maintenance
Maintenance Engineering and Maintainability: An Introduction 763
having a poor maintenance process - adding [13] Knezevic Jezdimir. Systems maintainability:
CMMS will not make it better. Analysis, engineering and management. Kluwer
Academic Publ., 1997.
[14] Moubray John. Reliability centered maintenance
References (2nd edition). Butterworth-Heinemann, Oxford,
1997.
[15] Stoneham Derek. Maintenance management and
[1] Rau John G. Optimization and probability in
technology handbook. Elsevier, U.K., 1998.
systems engineering. Van Nostrand Reinhold
[16] Kumar UD, Crocker J, El-Haram M. Reliability,
Company, 1970.
maintenance and logistic support- A life cycle
[2] Moss Marvin A. Designing for minimal
approach. Kluwer Academic Publisher, Boston,
maintenance expense: The practical application of
2000.
reliability and maintainability. Marcel Dekker,
[17] Gertsbakh IB. Reliability theory with applications
New York, 1985.
to preventive maintenance. Springer, Berlin, 2000.
[3] Anderson Ronald T, Neri Lewis. Reliability-
[18] Mobley Keith. An introduction to predictive
centered maintenance: Management and
maintenance. Butterworth-Heinemann, U.K.,
engineering methods. Elsevier Applied Science,
2002.
London, 1990.
[19] Narayana V. Effective maintenance management:
[4] Misra KB. Reliability Analysis and Prediction: A
Risk and reliability strategies for optimizing
Methodology Oriented Treatment. Elsevier
performance. Industrial Press, Inc., New York,
Science Publishers, Amsterdam, 1992.
2004.
[5] Smith AM. Reliability-centered maintenance.
[20] Nakagawa Toshio. Maintenance theory of
McGraw-Hill, New York, 1993
reliability. Springer-Verlag, London, 2005. .
[6] Knezevic Jezdimir. Reliability, maintainability and
[21] Bloom Neil B. Reliability centered maintenance:
supportability. McGraw-Hill, New York, 1993
Implementation made simple. McGraw-Hill, New
[7] Blanchard Benjamin S, Verma Dinesh, Peterson
York, 2005.
Elmer L. Maintainability: A key to effective
[22] Wang Hongzhou, Pham Hoang. Reliability and
serviceability and maintenance management. John
Optimal Maintenance. Springer Verlag, U.K.,
Wiley, New York, 1995.
2006.
[8] Ebeling Charles E. An introduction to reliability
[23] Murthy DNP, Blischke WR. Warranty
and maintainability engineering. McGraw-Hill
Management and Product Manufacture. Springer
Company, New York, 1996.
Verlag, U.K., 2006.
[9] Elsayed AE. Reliability engineering. Addison
[24] Jardine Andrew KS, Tsang Albert HC.
Wesley Longman, Inc. New York, 1996.
Maintenance, replacement, and reliability: Theory
[10] Lewis EE. Introduction to reliability engineering.
and applications. Taylor & Francis, Boca Raton,
John Wiley & Sons, Inc., New York, 1996.
2006.
[11] Kelly Anthony. Maintenance strategy - Business-
[25] Rao BKN. Toward the universal theory of failure.
centered maintenance. Butterworth Heinemann,
In Proceedings of 19th International Congress
Oxford, 1997.
COMADEM 2006. Kumar, Parida and Rao (Eds.)
[12] Kelly Anthony. Maintenance organization and
Luleå, Sweden, 2006.
systems - Business-centered maintenance.,
Butterworth Heinemann, Oxford, 1997.