TTHHEE UUNNIIVVEERRSSIITTYY OOFF AALLBBEERRTTAA
RROOLLEE OOFF MMAAIINNTTEENNAANNCCEE IINN CCOOMMPPOONNEENNTT
AANNDD SSYYSSTTEEMM RREELLIIAABBIILLIITTYY:: A Diesel Power Generation Case Study
By
SYED JEHANZEB ADEEL HAIDER ID: 1135539
A Project Report
Presented to,
FACULTY OF GRADUATE STUDIES AND RESEARCH IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
MASTERS OF ENGINEERING
DEPARTMENT OF MECHANICAL ENGINEERING UNIVERSITY OF ALBERTA
EDMONTON, ALBERTA T6G2V4
WWiinntteerr--22000077
AACCKKNNOOWWLLEEDDGGEEMMEENNTT
First of all the author wishes to thank his mother for all her prayers and support because of whom he is at this
stage of completing his M.ENG Program, and to his family for the love and support they have always provided.
Dr. John Doucette provided continuous encouragement, knowledge, advice, optimism, and uplifting
supervision throughout the duration of the project. The author would like to sincerely thank and express
appreciation for the supervisor role in this effort.
Many individuals have contributed to this report by giving their time, knowledge, expertise, and experience.
Their contributions are mainly in the form of ideas, concepts, approaches, guidelines, recommendations, and
reviews. The authors extend their thanks and appreciation to the contributors
As well, thanks to Dr. David Checkel and Mr. Zhigang Tian, whose instructions in the courses, “Combustion
Engines” and “Reliability for Design”, played a vital role in this report. Thanks to the faculty staff, librarians, and
other colleagues for providing assistance, technical support, comments, and social atmosphere throughout the
project report.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
1
CCOONNTTEENNTTSS EEXXEECCUUTTIIVVEE SSUUMMMMAARRYY..............................................................................................................................................................................................................................................................................................22
11.. IINNTTRROODDUUCCTTIIOONN ..............................................................................................................................................................................................................................................................................................................33
22.. MMAAIINNTTEENNAANNCCEE SSTTRRAATTEEGGIIEESS ................................................................................................................................................................................................................................................................44
2.1. CORRECTIVE MAINTENANCE (CM) ...............................................................................................................4 2.2. PREVENTIVE MAINTENANCE (PM) ................................................................................................................5
2.2.1. SCHEDULED MAINTENANCE............................................................................................................................5 2.2.2. PREDICTIVE MAINTENANCE (PdM) .................................................................................................................5 2.2.3. RELIABILITY CENTERED MAINTENANCE ...........................................................................................................6
2.3. COMPARISON OF MAINTENANCE METHODS ..............................................................................................8 2.4. CONDITION MONITORING SYSTEMS (CMS) .................................................................................................9
2.4.2. WHAT IS POSSIBLE TO BE MEASURED? ............................................................................................................9 2.5. DESIGN IMPROVEMENT TO MINIMIZE THE MAINTENANCE .....................................................................10
33.. RREELLIIAABBIILLIITTYY TTHHEEOORRYY......................................................................................................................................................................................................................................................................................1111
3.1. RELIABILITY....................................................................................................................................................11 3.1.1. SYSTEM RELIABILITY ...................................................................................................................................12
3.1.1.1. System Reliability Analysis ...............................................................................................................12 (a) Reliability Block Diagrams....................................................................................................................12 (b) Fault Tree Analysis ..............................................................................................................................13 (c) Failure Mode and Effect Analysis .........................................................................................................13
3.1.1.2. Binary Reliability Analysis .................................................................................................................13 3.1.1.3. Typical System Reliability Models.....................................................................................................14
(a) Series Systems ....................................................................................................................................14 (b) Parallel Systems ..................................................................................................................................14 (c) Series-Parallel Systems .......................................................................................................................15 (d) Parallel-Series Systems .......................................................................................................................15 (e) Mixed Series-Parallel Systems ............................................................................................................15 (f) k-out-of-n Configuration ........................................................................................................................16 (g) Complex Systems / Bridge Structures .................................................................................................16
3.2. AVAILABILITY..................................................................................................................................................17 3.2.1. UNAVAILABILITY ..........................................................................................................................................18
3.3. MAINTAINABILITY...........................................................................................................................................18 3.4. BATHTUB CURVE...........................................................................................................................................19 3.5. RELIABILITY, AVAILABILITY, AND MAINTAINABILITY .................................................................................19
44.. EEQQUUIIPPMMEENNTT LLIIFFEE AANNDD IIMMPPAACCTT OOFF MMAAIINNTTEENNAANNCCEE ..................................................................................................................................................................................2211
4.1. FAILURE MODELS..........................................................................................................................................21 4.2. THE IMPACT OF MAINTENANCE ..................................................................................................................21
4.2.1. AN EXAMPLE OF IMPACT OF MAINTENANCE ON EQUIPMENT LIFE .....................................................................22
55.. AANN OOVVEERRVVIIEEWW OOFF DDIIEESSEELL EENNGGIINNEESS ..................................................................................................................................................................................................................................2244
5.1. HOW DIESEL ENGINES WORK .....................................................................................................................24 5.2. WHY DIESEL? .................................................................................................................................................27 5.3. ADVANTAGES OF A DIESEL ENGINE...........................................................................................................27 5.4. DIESEL POWER GENERATORS....................................................................................................................28 5.5. TODAY'S DIESEL GENERATION ...................................................................................................................28
66.. CCAASSEE SSTTUUDDYY:: PPOOWWEERR GGEENNEERRAATTIIOONN TTHHRROOUUGGHH DDIIEESSEELL GGEENNEERRAATTOORRSS ........................................................................................................2299
6.1. SITE OVERVIEW.............................................................................................................................................29 6.2. COMPOSITION OF THE POWER GENERATION SYSTEM ..........................................................................29 6.3. RELIABILITY ANALYSIS OF THE SYSTEM ...................................................................................................30 6.6. PROBLEMS FACED ........................................................................................................................................34
77.. CCOONNCCLLUUSSIIOONNSS ............................................................................................................................................................................................................................................................................................................4400
RREEFFEERREENNCCEESS ........................................................................................................................................................................................................................................................................................................................4422
Appendix A: Specifications of Cummins NT855 Set Used for Case Study
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
2
EEXXEECCUUTTIIVVEE SSUUMMMMAARRYY This report explains different types of maintenance strategies which are frequently observed in mechanical
engineering industry, how they are carried out, their advantages and disadvantages, how they differ from each
other, how they are related to one another, and above all these, how they can affect reliability of components
as well as of the system. As we all know that reliability of a new component is higher than a used one,
similarly, proper and timely maintenance of a component results in improved reliability of that component and
of the system of which this component is a part. This report also throws light on this aspect of maintenance. In
relation to the project title, this report will also present a case study which includes personal experience and
actual data analysis of Power Generation System through Diesel Generator Sets. This report explains what a
diesel engine is and how it works. Following this introduction, the report explains actual problems faced by our
company, and remedial steps regarding maintenance, reliability, and cost. We also discuss some of the steps
which were actually taken by management, their impact, positive and negative, and if negative, then what were
the causes. We conclude with some suggestions which could have helped resolve the problem regarding
system reliability and maintenance.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
3
11.. IINNTTRROODDUUCCTTIIOONN What is maintenance and why is it performed? Past and current maintenance practices would imply that
maintenance is the set of actions associated with equipment repair. The Merriam-Webster’s Collegiate
Dictionary defines maintenance as follows: “the work of keeping something in proper condition; upkeep.” [23].
This means that maintenance is the action or a series of actions that are taken to prevent a component, a
device, or a system from failure or to repair regular degradation that a component or a system experience with
the normal operation to keep it functioning properly. It is unfortunate that the data available in many recent
studies shows that proper measures of equipment maintenance are not taken to keep it in good running
condition. Rather, they wait for component and system failures and then take necessary actions to repair or
replace the component or the system as needed. We know that every equipment and system has a predefined
operational life and there is no such system or component that lasts forever. For instance, a system may be
designed to work at full design load for 3,000 hours and to bear 10,000 start and stop cycles.
Maintenance is required to reach the design life and to make the most of any particular system or equipment.
For example, belts need adjustment, alignment needs to be maintained, rotating equipment requires proper
lubrication, etc. Sometimes, replacement of certain components is required, e.g., a motor vehicle’s wheel
bearing, to ensure the main system, in this case a car, lasts for its design life. Anytime we fail to perform
equipment’s or system’s maintenance as recommended by the designer/manufacturer, we cut down the life of
the system or equipment. Over the last 30 years, different approaches of maintenance activities have been
developed to ensure equipment or system completes or goes beyond its design life. We can make the most of
any equipment or system by using preventive maintenance, predictive maintenance, or reliability centered
maintenance in addition to the corrective maintenance which is simply to wait for a system to fail [1].
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
4
22.. MMAAIINNTTEENNAANNCCEE SSTTRRAATTEEGGIIEESS
Two basic types of maintenance philosophies or strategies can be identified, namely corrective and preventive.
Preventive maintenance can be sub-divided into scheduled, predictive, and reliability centered maintenance.
Predictive maintenance is the most recent development. In practice, all these types are used in maintaining
engineering systems. The challenge is to optimize the balance between these types for maximum profitability.
In general, corrective maintenance is the least cost effective option when maintenance requirements are high.
As Rao stated in [2], “An ideal maintenance strategy meets the requirements of machine availability and
operational safety, at minimum cost.”
2.1. CORRECTIVE MAINTENANCE (CM)
Corrective maintenance is carried out when a defect in an item is recognized and is anticipated to put that item
into a state in which it can carry out its intended function. This type of maintenance is often called repair and is
carried out after the failure of a component. The purpose of the corrective maintenance is to bring the
component immediately back in to a running state, either by repairing or replacing the failed component. To
use only corrective maintenance is seldom a good solution. This means that you will run your system until a
breakdown occurs and in some literature this is referred to as a breakdown strategy. By choosing breakdown
strategy, the preventive maintenance can be reduced to a lowest and the system can function unless a major
breakdown of a component resulting in a system shutdown. This strategy is uncertain, as failures of relatively
small and dispensable components can lead to severe consequential damage. Another aspect of this strategy
is that most breakdowns are likely to occur during high load conditions when the system is at peak
performance and needed the most. One more drawback of this strategy is that when repair or replacement of
any component is needed the downtime can be extensive as logistics gets more complicated if things are not
Maintenance
Corrective Maintenance
Preventive Maintenance
Scheduled Maintenance (Time-based)
Predictive Maintenance
(Condition-based)
Reliability Centered
Maintenance
Figure 2.1: Types of Maintenance
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
5
planned in advance. Comparing with other strategies, corrective maintenance has some benefits too, such as
low maintenance cost during operation of the system and that the components of the system will be used for
maximum lifetime. The lifetime of the component cannot be predicted and true assessment of the cost and
lifetime can only be made once the component has failed [2].
2.2. PREVENTIVE MAINTENANCE (PM)
Preventive maintenance is carried out at fixed intervals or in accord with approved norms and intended to
reduce the likelihood of failure or the degradation of functioning of an item. Preventive maintenance is
performed regularly to postpone failures or to prevent failures from occurring. There are three different types of
preventive maintenance; scheduled maintenance or time-based maintenance, predictive or condition based
maintenance, and reliability centered maintenance. What differs between these three are the way of deciding
when to perform the preventive maintenance [2].
2.2.1. SCHEDULED MAINTENANCE
Scheduled or Time-Based maintenance is a type of preventive maintenance that is carried out in accordance
with an established time schedule or established number of units of usage. The schedule for the preventive
maintenance can be either clock-based or age-based maintenance. Clock-based maintenance means that the
preventive maintenance is carried out at specified calendar times and age-based maintenance means that the
maintenance is carried out when a component reach a certain age. The age does not need to be calendar
time, but measured for example, in revolutions or operational time, etc.
Scheduled maintenance should be designed to minimize the probability of failures. The cycle times of the
maintenance program should be harmonized with the system requirements. The system should be examined
and maintained periodically, see Figure 2.2. This maintenance approach implies that components showing
wear will be replaced on a regular basis even if they are not at the end of their lifetime. Scheduled
maintenance requires regular contact with the system and the supply for maintenance tools and workforce
shares a considerable cost for the maintenance. The advantage of preventive maintenance is that it can be
scheduled ahead of time and the management of logistics can be made simple [2], [3].
2.2.2. PREDICTIVE MAINTENANCE (PdM)
Predictive maintenance refers to an approach that deals with the actual condition of a component or a system.
This type of maintenance is not carried out according to periodic schedules but rather when a change is
observed in characteristics. Predictive maintenance is a type of preventive maintenance that is based on
equipment’s performance and/or condition monitoring and the consequent actions. In this type of preventive
maintenance, a designated monitoring device collects data of the system or component. The condition
monitoring could be scheduled, on request, or continuous. The collected data can forecast required
maintenance prior to predicted failure. Maintenance is carried out when a condition variable reaches or
exceeds a threshold value. The system components are operated to a predefined condition of wear and
fatigue. When this condition is reached, the component has to be maintained or replaced. Examples of such
condition variables are vibrations in a machine, temperature, measure of quality of the lube oil, etc.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
6
Condition monitoring of components helps to plan the maintenance before failure and reduces downtime and
repair costs. The components are therefore used closer to their lifetimes and the coordination of spare parts is
easy. Another advantage of putting condition based system into practice is that trends and statistical data such
as mean time to failure can be made available. The statistical data from condition monitoring system provides
reliable data for remaining lifetime of components in the system [2], [3]. Figure 2.2 shows an example of
condition based maintenance along with corrective and scheduled maintenance.
Figure 2.2: Condition based / Predictive maintenance compared to scheduled and corrective maintenance [3]
2.2.3. RELIABILITY CENTERED MAINTENANCE
Reliability centered maintenance (RCM) is in fact establishing or improving a maintenance program in the most
cost-effective and technically feasible manner. This is a systematic, planned approach which is based on the
consequences of failure. It actually represents a shift away from scheduled or time-based maintenance and
puts emphasis on the functional significance of system components and their records of failures and
maintenance. RCM involves a methodical approach to the following three main matters.
1. Understanding failure
2. Consequences of failure - if consequence of failure is zero, one would choose “run to failure”.
3. Impact of preventive maintenance
The application of RCM results in a scheduled maintenance program that includes all scheduled tasks that are
required for safe and economical operations. There are two activity branches for RCM:
1. Operational
2. Analytical
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
7
Operational RCM is concerned with designing and implementing effective scheduled maintenance tasks and
procedures for performing them. Analytical RCM evaluates and applies inherent reliability characteristics to
optimize the scheduled maintenance program [13]. RCM measures consider the following key objectives of a
maintenance program:
• Minimise Costs
• Meet Safety and Environmental Goals
• Meet Operational Goals
The RCM process starts with a failure mode and effects analysis which spots the critical failure types in a
methodical and structured manner. The process then requires the assessment of each critical failure mode to
determine the best maintenance strategy to reduce the severity of every failure. Maintenance strategy that is
chosen, should consider cost, safety, environmental, and operational consequences. The results of
redundancy, spares costs, maintenance team costs, equipment aging and repair times must be taken into
account along with all other factors. Once most favourable maintenance procedures have been recorded the
RCM practice gives us predictions and costs of system performance. It also provides expected requirements of
spare parts and manpower requirements for maintenance.
RCM is in fact a method which relates reliability to maintenance. It is a qualitative approach to systematize
maintenance. It was first put into practice in 1960s by the aircraft industry when Boeing 747 series was
introduced, and the objective was to lower Preventive Maintenance (PM) costs in achieving a predefined
reliability level. This methodology was developed further after getting successful outcome. The key feature of
RCM is its emphasis on preserving system function where critical components for system reliability are
prioritized for PM measures. However, the method is generally not capable of showing the benefits of
maintenance for system reliability and costs [8].
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
8
2.3. COMPARISON OF MAINTENANCE METHODS In the light of above explanations of different maintenance approaches, Table 2.1 throws light on some merits
and demerits of these methodologies.
Table 2.1: Comparison of different maintenance methods
Method Advantages Disadvantages
Corrective Maintenance • Low maintenance costs during operation.
• Components will be used for a maximum lifetime.
• High risk in consequential damages resulting in extensive downtimes.
• No maintenance scheduling is possible.
• Spare parts logistics is complicated.
Preventive Maintenance (Scheduled/Time based)
• Expected downtime is low. • Maintenance can be scheduled. • Spare logistics is easy.
• Components will not be used for maximum lifetime.
• Maintenance costs are higher compared to corrective maintenance.
Predictive Maintenance (Condition based)
• Components will be used up to almost their full lifetimes.
• Expected downtime is low. • Maintenance activities can be
scheduled. • Spare part logistics is easy
because a failure can be detected early in time.
• Reliable information about the remaining lifetime of the components is required.
• High effort for condition monitoring hardware and software is required.
• Cost of another layer in the system. • Identification of appropriate condition
threshold-values is difficult.
Reliability Centered Maintenance
• Can be the most efficient maintenance program.
• Lower costs by eliminating unnecessary maintenance or overhauls.
• Minimize frequency of overhauls. • Reduced probability of sudden
equipment failures. • Able to focus maintenance
activities on critical components. • Increased component reliability. • Incorporates root cause analysis.
• Can have significant start-up cost, training, equipment, etc.
• Savings potential not readily seen by management.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
9
2.4. CONDITION MONITORING SYSTEMS (CMS)
Is condition monitoring system (CMS) is a fine method of reducing the maintenance for any component or a
system? Actually, what is required is more acquaintance of how to monitor the system and information that
whether condition monitoring is applicable to the components of this system or not [4].
2.4.1. BENEFITS OF A CONDITION MONITORING SYSTEM
Methods to measure the condition of the components in any system are essential to put condition based
maintenance into practice, and this is achieved by using a condition monitoring system which continually
observes the system. The implementation of a CMS leads to some new advantages because of its
characteristics. Table 2.2 indicates some of the major characteristics, advantages, and benefits of CMS.
Table 2.2: Characteristics of Condition Monitoring Systems
Characteristics Advantages Benefits
Early warning • Avoid breakdowns. • Better planning of maintenance.
• Avoid repair costs. • Minimize downtime.
Identification of problem
• Right service at the right time. • Minimizing unnecessary
replacements.
• Prolonged lifetime. • Lowered maintenance costs.
Continuous monitoring • Constant information that the system is working.
• Security. Less stress.
The main characteristic of a CMS is its recognition of problems in the system in advance. An early warning can
provide the time to plan the repair work and to order required parts for replacement if necessary. The ability to
spot the trouble at an early stage is helpful to employ the right maintenance activity at the right time and the
ability to forecast where the problems rising from can assist in pointing out which parts need replacement and
also possible reasons and causes of the failure. As Davies stated in [4], “A condition monitoring programme
can minimize unscheduled breakdowns of all mechanical equipment in the plant, and ensure that repaired
equipment is in an acceptable mechanical condition. The programme can also identify machine train problems
before they become serious.”
2.4.2. WHAT IS POSSIBLE TO BE MEASURED?
When applying a condition monitoring system on a system mainly comprised of mechanical components such
as bearings, gearboxes, etc., there are different parameters to examine. These parameters may include
vibration, noise, debris and water in oil, alignment, pressure, temperature, stress, strain, erosion, corrosion,
etc.
Applying a condition monitoring system also includes applying new methods for maintenance planning. Davies
stated, “The output of a condition monitoring programme is data. Until action is taken to resolve the deviations
or problems revealed by the programme, plants performance cannot be improved. Therefore, a management
philosophy committed to plant improvement must exist before any meaningful benefit can be derived” [4].
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
10
2.5. DESIGN IMPROVEMENT TO MINIMIZE THE MAINTENANCE
Engineered modifications and design improvements in any system or component could result in minimizing the
requirement of its maintenance and even in elimination of several maintenance tasks. These changes and
improvements utilize the recorded data of previously employed procedures to design the need for
maintenance. We can use analogy of an automobile as an example. If we compare the modern automobile to
a 1970s era automobile, a significant drop in the maintenance requirements are quite obvious. One of the main
areas is engine tune ups. Automobiles of 1970s required tune ups every 30,000-to-40,000 miles. Today’s
automobiles require tune ups at 100,000 miles, with no degradation in performance. These improvements
were studied, re-engineered, and implemented. Same is the case for plant and facility equipment today.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
11
33.. RREELLIIAABBIILLIITTYY TTHHEEOORRYY
“The main objective of a reliability study should always be to provide information as a basis for decision.” -
Rausand and Hoyland, 2004
The results provided by a reliability study will not tell us exactly what to do, but in what direction to look. For
example, a reliability study can be useful in areas of risk analysis, optimization of operations and maintenance.
The risk analysis is a way of identifying causes and consequences of failure events, and the optimization is a
way of telling how failures can be prevented and how to improve the availability of a system. One can see
reliability theory as a tool for analysing and improving the availability of the system [5].
3.1. RELIABILITY
The word reliability is vast in meaning. In common, reliability measures the capacity of a system to carry out its
assigned task successfully and takes into account the past experiences and behaviour to assess the upcoming
performance of the system. Another description that shows a different aspect of the concept is that reliability is
the probability that a component or a system is performing its role satisfactorily, for the intended period of time,
under the intended operating conditions. Reliability can be measured through the mathematical model of
probability by identifying the probability of successful performance with the degree of reliability. Normally, a
component or system is said to perform satisfactorily if it does not fail during the time of service. On the other
hand, many components are expected to experience failures, be repaired and then returned to working state
during their entire useful life. In this case a more suitable measure of reliability is the availability of the
component.
We can measure reliability in many ways depending on particular circumstances. The mean time to failure
(MTTF) is defined as the mean time between first operation of a component or a system and the first event of
a breakdown or malfunction. When a breakdown is observed, the system is repaired and put back into
operation and the system is then considered as fully functioning. The mean down time (MDT) is defined as the
average time that the system is not functioning when a component is being repaired, and is basically the time it
takes to repair a failure that is why it is also called mean time to repair (MTTR). The mean time between
failures (MTBF) takes into account the mean time to failure and the mean down time/mean time to repair. The
down time is usually much shorter than the time of operations, and if so, then the two measurements can be
viewed as MTTF ≈ MTBF [3].
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
12
Figure 3.1: Measurements of reliability [3]
3.1.1. SYSTEM RELIABILITY
A system is a collection of components arranged according to a certain structure. A large system may be
comprised of a number of smaller subsystems. Reliability modeling is actually analyzing components and their
relationships and relating components’ states (working or failed) to system state.
3.1.1.1. System Reliability Analysis
System reliability analysis involves evaluation of the system reliability based on the reliabilities of its
components. A number of methods can be applied to the system reliability analysis, such as:
(a) Reliability Block Diagrams
Designing a reliable system starts with the development of a model. A reliability block diagram (RBD) is a
logical model that gives a graphical presentation to evaluate the relation ship between different parts of a
system. Basic RBDs are built of series, parallel, or combinations of series and parallel components. Every
block symbolizes an event or purpose of a system component. If all components are required to operate
successfully for the system to operate successfully then these blocks are arranged in series. In contrast, if only
one component needs to operate successfully for the system to be functional then these blocks are arranged
in parallel. The main purpose of producing reliability assessments is to enable the user to:
(i) compare one product against another.
(ii) compare the virtues of different distribution systems.
(iii) compare the variation in the scale of the reliability evaluation between the different systems under
assessment.
The graphic representation of RBDs makes them easily understandable and hence readily checked. The
overall system reliability can be very easily calculated from the reliability of the blocks using the laws of
probability. Block diagrams can also be used for the evaluation of system availability
Some limitations of this method are that it cannot model systems in which the sequence of failures affects the
outcome and this method cannot model maintenance strategies also.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
13
(b) Fault Tree Analysis
Fault tree analysis (FTA) is a conclusive, failure-based approach which helps us derive a conclusion based on
logical deduction. FTA begins with an unwanted event, such as a failure, and then deduces its causes using a
methodical logical process in backward direction. In order to determine the cause of failure, a fault tree (FT) is
constructed as a logical diagram of the events and their relationships that are necessary and sufficient to
cause an undesired event, which is the top event.
In fact FT does not consider all possible failures of any system or all the causes of the failures. It takes into
account any one particular failure which is the top event and include only those faults that contribute to this top
event. Further, only those faults are covered in FTA that are assessed and considered to be realistic. It is also
important to know that a fault tree is a qualitative model that can be analyzed quantitatively but it is not a
quantitative model in itself.
As stated in the NASA handbook [25]: "FTA provides critical information that can be used to prioritize the
importance of the contributors to the undesired event. The contributor importance provided by FTA vividly
shows the causes that are dominant and that should be the focus of any safety or reliability activity”.
(c) Failure Mode and Effect Analysis
A failure mode and effects analysis (FMEA) is a logical process of recognizing and avoiding any failures
associated with product, service, or the process before they actually occur. The center of attention of FMEA is
prevention (i.e., “fix it before it breaks”). The purpose of FMEA is to observe all of the means of failure of a
product or a process, analyze associated risks, and take action where necessary. General applications of
FMEA include prevention of defects, improvement of processes, identification of probable safety issues, and
increasing satisfaction of the client. FMEAs should be performed by organizations from initial product or
process design through the entire life of the product. Design FMEAs or process FMEAs are chosen by
organizations accordingly. Organizations can also get advantage by performing FMEAs on existing products or
processes for improvements [26].
As explained in [26], implementation of an FMEA can help
1. identify possible failures.
2. identify chances of cost reductions and quality improvement.
3. identify issues in advance that can reduce the cost of making changes.
4. exhibit an organization’s dedication to an inclusive quality system.
5. provide an outstanding preventive device for ISO 9001.
3.1.1.2. Binary Reliability Analysis
Let’s consider a system with n components. A component may only be in one of two possible states, working
or failed. Xi is used to indicate the state of component i.
0, if component i is failed by time t Xi =
1, if component i is working up to time t
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
14
Similarly, in a binary reliability analysis, a system can be in only one of two possible states, working or failed;
Φ(X) is used to denote the state of the system, i.e., structure function. The state of the system is completely
determined by the states of the components [12]. The component and system reliability can be stated as:
Ri = Pr(Xi = 1): reliability of component i
Fi = Pr(Xi = 0) = 1-Ri: failure probability or unreliability of component i
RS = Pr(Φ = 1): reliability of the system
FS = Pr(Φ = 0): 1-RS: failure probability or unreliability of the system
3.1.1.3. Typical System Reliability Models
(a) Series Systems: In a series system, if one component fails, the entire system fails. For instance, a car is
comprised of four basic subsystems: the engine, the transmission, the battery, and the wheels. These
subsystems are in a series combination and a failure of any of these four will result in a system failure. In other
words, all of the units in a series system must succeed for the system to succeed. Analytically, it is
represented as follows:
nxxxxxX .....)( 4321=Φ
nS RRRRR .....321=
)()......().().()( 321 tRtRtRtRtR nS =
(b) Parallel Systems: A parallel system is working if one component is working. In a simple parallel system, at
least one of the units must succeed for the system to succeed. Units in parallel are also referred to as
redundant units. Adding redundancy is one of a number of methods through which system reliability can be
improved. It is widely used in the aerospace industry and generally used in exceptionally critical systems. So in
a parallel system, all n units must fail for the system to fail.
)1).....(1).(1(1)( 21 nxxxX −−−−=Φ
nS FFFFF .....321=
∏=
=n
iiS tFtF
1
)()(
Here, SS FR −= 1
… 1 2 n
…
1
2
n
Figure 3.2: Series System
Figure 3.3: Parallel System
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
15
(c) Series-Parallel Systems: In a series-parallel system, the system requires at least one of the modules from
each column to be functioning in order to remain operational. It is represented as follows:
Figure 3.4: Series-Parallel System (d) Parallel-Series Systems: In a parallel-series system, the system requires that all the modules of at least
one row to be functioning in order to remain operational. It is represented as follows:
Figure 3.5: Parallel-Series System (e) Mixed Series-Parallel Systems: Most of the simple and smaller systems can be correctly presented by a
series or parallel arrangement, but there are many larger complicated systems that involve both series and
parallel composition of components in the entire system. Reliability of such systems can be evaluated by
calculating the reliabilities of the individual series and parallel segments and then combining them properly. An
example of mixed series-parallel systems is a computer-projector system which is quite common in our
classrooms, meeting rooms, and conference rooms in the offices:
Figure 3.6: An example of Mixed Series-Parallel System [12]
… …
1
2
n1
…
1
2
n2 …
1
2
nN
… 1 2 n1
… 1 2 n2
… 1 2 nN
Laptop Video cable Socket
USB
Internet Desktop PC
Power Cable Projector Screen
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
16
(f) k-out-of-n Configuration: The k-out-of-n configuration can be considered as a special type of parallel
configuration of components. This type of configuration requires minimum k components to work successfully
out of the total n parallel components for the system to be functional. For instance, consider an airplane that
has four engines. The plane is designed in a manner that at least two engines are required to operate for the
plane to remain in the air. This means that the engines are in a k-out-of-n configuration, where k = 2 and n = 4.
This example is elaborated in Figure 3.7:
Figure 3.7: An example of k out of n configuration
In this type of configuration, when the number of components required to keep the system running approaches
the total number of components in the system, the system's nature inclines to that of a series system. And
when k = n, the system is a series system. Mathematically,
)1,()1,1(),( −+−−= nkRFnkRRnkR nn
Boundary conditions; ,1),0( =nR ,0),( =nkR for kn <<0
(g) Complex Systems / Bridge Structures: In many cases, it is not easy to recognize which components are
in series and which are in parallel in a complex system. The network shown in Figure 3.8 is a good example of
such a complex system. Several methods exist for obtaining the reliability of a complex system including
decomposition method, event space method, and path-tracing method. We will here briefly discuss the
decomposition method.
Decomposition Method: The decomposition method is an application of the law of total probability. It involves
choosing a "key" component and then calculating the reliability of the system twice; once as if the key
component failed (R5 = 0) and once as if the key component succeeded (R5 = 1). These two probabilities are
then combined to obtain the reliability of the system, since at any given time the key component will be failed
or operating. Using probability theory, the equation is:
( ))(1)0|()()1|()( 5555 tRxtRtRxtRtR −⋅=+⋅==
The key component is 5 in the following complex system which is simplified by decomposing method [10], [12].
(a) System Working (b) System Failed
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
17
Figure 3.8: An example of decomposition of a Complex/Bridge Structure
3.2. AVAILABILITY
Availability is the probability that a component or a system is successfully operating at any point of time. For
instance, if a computer can be used for 18 hours in a day, the availability of the computer is said to be 18/24.
The most simple representation for availability is as a ratio of the expected value of the uptime of a system to
the sum of the expected values of up and down time, or
][][][
DowntimeEUptimeEUptimeEA+
=
If we define the status function X(t) as
1, sys functions at time t X(t)=
0, otherwise
therefore, the availability is represented by
A(t) = Pr[X(t) = 1]
A(t) = Pr[X(t) = 1] = E[X(t)], t > 0
The availability performance is defined as “The ability of an item to be in a state to perform a required function
under given conditions at a given instant of time or during a given time interval, assuming that the required
external resources are provided” – Maintenance terminology, SIS 2001 [2], [5].
A commonly used measurement of availability is the amount of operational time divided by the nominal time,
the nominal time or uptime is usually a period of one year and then the availability is presented as percentage
of operational time per year [12].
1 3
2 4
1 3
2 4
(a) When R5 = 1 (b) When R5 = 0
1 3
2 4
5
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
18
A common equation for measurement of availability is:
MTTRMTBFMTBFtyAvailabili
+=
+=
λµµ
when t → ∞
Here, µ = repair rate (following exponential distribution) and λ = failure rate (following exponential distribution)
3.2.1. UNAVAILABILITY
The term unavailability is the likelihood that a system is not available at a point of time when it is required.
Unavailability may be expressed mathematically,
Unavailability = 1 – Availability, or as the ratio,
MTTRMTTFMTTRU+
=+
=λµ
λ when t →∞ ,
where, µ = repair rate (following exponential distribution), λ = failure rate (following exponential distribution),
MTTR = mean time to repair, and MTTF = mean time to failure 3.3. MAINTAINABILITY
Maintainability is defined as the probability that the system is successfully restored after failure within a
specified time. It is in fact a measure that how conveniently and rapidly a system can be repaired and restored
to working state after a failure. It is an attribute of design and mechanism, workforce availability in the required
skill levels, sufficiency of maintenance methods and tools. Maintainability is generally expressed as the
probability that a system will be restored to running condition within a given time, when the maintenance is
carried out according to recommended procedures. e.g. 90% maintainability in an hour [20].
Maintainability Function, M(t), for a system with the repair times distributed exponentially is given by:
Where, µ = Repair rate and MTTR = mean time to repair
Maintainability represents the ease with which maintenance can be done to repair a system after a failure.
Maintainability addresses all scheduled and unscheduled events, which are performed to:
a. repair or replace a component that shows undesirable physical condition or performance degradation,
repair or replace a component that has failed, and verify the restoration of a component or system to
working state,
b. determine the running status of the system or a component, and
c. adjust, align, or service components.
MTTR
etM t
=
−= ⋅
µ
µ
11)(
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
19
3.4. BATHTUB CURVE
“Normal mechanical failure modes degrade at a speed directly proportional to their severity. Thus, if the
problem is detected early, major repairs can be prevented in most instances.” - Davies, 1998 [4].
Generally the chances of a component to fail are high in the early period of its lifetime. The component might
have hidden flaws which could lead it to instant failure at the very beginning of its life span. If the component
survived this early phases after its birth, the chances of failure reduces and it enters a stable phase where it
remains for a definite period unless it enters the final stages of its life where the probability of its failure starts
to raise again because of wear and tear. If this failure is plotted on a graph, the shape of the curve is similar to
that of a bathtub, for this reason it is known as bathtub-curve. Figure 4.9 shows the bathtub curve with the
three typical phases.
The initial phase is called burn-in period; the stable phase is called useful life period and the end phase is
called wear out period. In a Burn-in process, a component is checked before being used as a part of the
system and the motive is to detect those components that would fail as a result of infant mortality. The system
can be made more reliable and trustworthy especially for reduced early failures if the burn-in period is made
sufficiently long.
Figure 3.9: Typical Bathtub curve [3]
Figure 3.9 exemplify a possible shape of the failure function. There are some other shapes of failure function
curves as well, but the bathtub curve generally represents the failure rates for mechanical components. Since
majority of the mechanical components goes through wear and tear, the failure function curve shows a little
escalating trend during the useful life period, consequently their actual curves may vary to some extent from
the one shown in Figure 3.9 [3].
3.5. RELIABILITY, AVAILABILITY, AND MAINTAINABILITY
We’ve already discussed that, reliability of a component or a system is the probability that it will not fail and will
carry out its assigned function successfully. Maintainability is the probability the system is successfully
restored after failure. Availability is the probability that the system is successfully operating at any point of time,
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
20
and it depends on both reliability and the maintainability. For this reason, one must realize how vital the
probability theory is to understand the reliability, maintainability, and availability of the system [9]. The relation
between reliability, availability, and maintainability can be understood better from Figure 3.10.
Reliability Maintainability Availability
Constant Decreases Decreases
Constant Increases Increases
Increases Constant Increases
Decreases Constant Decreases
Figure 3.10: Relation between Availability, Reliability, and Maintainability [12]
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
21
44.. EEQQUUIIPPMMEENNTT LLIIFFEE AANNDD IIMMPPAACCTT OOFF MMAAIINNTTEENNAANNCCEE Here we will discuss some important aspects of maintenance, equipment lifetime, how failure can be avoided,
types of failures, and impact of maintenance on equipment life.
4.1. FAILURE MODELS
Failure is the termination of the ability of a device to perform a required function. In many applications,
component failures can be divided into two categories i.e. random failures and deterioration. In random
failures, the rate of failure is constant. In other words, the chance of a failure occurring in any future time-
interval is the same as it is now. This is a failure whose rate of occurrence (intensity) is constant, and
independent of the device’s condition. In deterioration, failures arise as a consequence of the component
aging or wearing out. This is a process by which the rate of failure increases due to loss of strength, the effects
of usage, environmental exposure or passage of time. Simple failure-repair processes in the two cases are
shown in Figure 4.1. The various state designations are explained in the legend of the figure. The deterioration
process is represented by a sequence of stages of increasing wear, finally leading to equipment failure [6].
Figure 4.1: State diagrams for (a) random failure and (b) deterioration failure. W: working state, F: failure state, D1, D2, …Dk: stages of deterioration [6]
4.2. THE IMPACT OF MAINTENANCE
According to the concepts available in [6], the prime intention of maintenance is to improve the mean time to
failure. One approach of including maintenance to the system models presented in Figure 4.1 is shown in
Figure 4.2. In Figure 4.2(b), it is supposed that maintenance will result in an improvement to the situation in the
previous stage of wear and tear (with least maintenance). This differs with many approaches explained in this
study, where maintenance involves replacement (i.e., a return to the “new” conditions). In the case of random
failures, Figure 4.2(a), the hypothesis of constant failure-rate explains that maintenance cannot make any
improvement, since the probability of a breakdown during any upcoming time span are the same with or
without maintenance. The situation is quite different for deterioration processes, where the times from the new
condition to failure are not exponentially distributed even if the times between subsequent stages of
deterioration are. In such a process the hazard function denoted by h(t), is increasing, and maintenance will
bring about improvement independently of the types of distributions between stages. Hazard function can be
defined as instantaneous failure rate at any point in time. For this reason, conditions cannot be improved by
maintenance for random failures, but maintenance has a major role to play when failures are the consequence
of aging.
W F D1 D2 Dk F
(a) (b)
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
22
In Fig. 4.2(b), the dotted-line transitions to and from state M1 indicate that maintenance out of D1 should really
not be performed because, as noted, it would be meaningless. State M1 could be indeed omitted if the
maintainer knew that at the time of next maintenance, the deterioration process would still be in its first stage
and, therefore, no maintenance would be needed. Otherwise, maintenance must be carried out regularly from
the beginning, and state M1 has to be part of the diagram. Note that deterioration processes are a subset of
the failure models where the hazard function h(t), is increasing with time. Maintenance is advantageous in all
such cases. If the failure is random, the hazard function is constant, and maintenance is of no use.
Figure 4.2: State diagrams including maintenance states for (a) random failure, (b) failure following a three-stage deterioration process [6].
4.2.1. AN EXAMPLE OF IMPACT OF MAINTENANCE ON EQUIPMENT LIFE
In order to understand the impact of maintenance on equipment life better, let us consider an example a
system assuming that without maintenance, the system would fail after exactly 10 years, the scheduled
maintenance time is 3 years and the effect of this maintenance is a 1-year improvement in life. This system is
shown in Figure 4.3(a) and the horizontal line represents as a scale of deterioration. It can be seen that the
time to failure is now extended to 14 years as a result of the four maintenances carried out in the interval.
Figure 4.3: Maintenance every 3 years, resulting in (a) 1-year improvement, (b) 3-year improvement if total wear is 6 years or more, otherwise as in (a).
M—maintenance MM—overhaul F—failure [6].
W F D1 D2 D3 F
(a) (b)
M1 M2 M3 M
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
23
It is not hard to understand that improvement in the system as a result of maintenance is less than the wear
and tear of the system between two consecutive maintenances, particularly in the early life of the system when
only small maintenances are carried out. After this initial phase, the effect of maintenance should be equivalent
or more than the wear and tear going on between maintenances. This can be done by planning overhauls
(major maintenances) before a given state of deterioration. Now considering the example of Figure 4.3(a), if
the overhaul is required after 6 years, and if the effect of this overhaul is a 3-year improvement in system life,
Figure 4.3(b) will be the correct represent of this scenario. An important point to notice here is that now the
expected time to failure is infinity; means that failure will never take place. This is a false conclusion and is
completely because of the hypothesis that all quantities involved have fixed values. If variables are used
instead of fixed values, the failure will occur sooner or later [6].
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
24
55.. AANN OOVVEERRVVIIEEWW OOFF DDIIEESSEELL EENNGGIINNEESS 5.1. HOW DIESEL ENGINES WORK
The Diesel cycle is a combustion process of a reciprocating internal combustion engine, in which fuel is ignited
by heat generated when air is compressed in the combustion chamber, into which fuel is injected. This is as
opposed to igniting it with a spark plug as in the Otto cycle (four-stroke/petrol) engine. Diesel engines (heat
engines using the Diesel cycle) are used in automobiles, power generation, diesel-electric locomotives, and
submarines. Invented by Rudolph Diesel in 1897, it was initially planned to work on Peanut oil. A Diesel engine
generates power (while continuously injecting fuel) to maintain the cylinder at constant pressure during its
power stroke.
We all know that if we compress any gas, its temperature will increase, and through this process, fuel is ignited
in diesel engines. The process starts when air is drawn into the cylinders and then compressed by the pistons
at compression ratios of about 25:1, which is quite higher than the compression ratios of 7:1 to 10:1 in the
spark-ignition Otto cycle engines. When the compression stroke is at or near to the end, diesel fuel is injected
into the combustion chamber through an injector. As the fuel comes into contact with air, it ignites. This is
because the air has been heated to a temperature of about 700–900 °C (1300–1650 °F) due to compression.
The combustion resulting from the fuel ignition causes a sudden rise in heat. This results in expansion in the
cylinder and moves the piston downward due to increased pressure. A connecting rod transmits this motion to
a crankshaft to convert linear motion to rotary motion. This rotary motion is used as power source in many
applications such as a diesel engine vehicle. Mechanical valves, present in the cylinder heads, generally
control the air intake. Most of today’s advance diesel engines are supplied with a turbocharger for increased
power output, and in some types, a supercharger is used to increase the amount of intake air [21]. A simple
piston-cylinder arrangement of a four stroke diesel engine is shown in Figure 5.1 [23].
Figure 5.1: Simple piston-cylinder arrangement of a four stroke diesel engine [23]
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
25
A four stroke Diesel Cycle Engine is discussed onwards in this study. The four-stroke operating principal
applied to a diesel engine is illustrated in Figure 5.2 [19]:
Figure 5.2: Four stroke operating principle applied to diesel engine [19]
In thermodynamic terms, in a diesel engine, air is compressed adiabatically in the cylinder. An adiabatic
process is one in which no heat is gained or lost by the system. As already mentioned above, this compression
of air increases the temperature and causes the fuel mixture to ignite. The ideal air-standard cycle is
considered as a reversible adiabatic compression followed by a constant pressure combustion process, then
an adiabatic expansion as a power stroke and a constant volume exhaust. At the end of exhaust, a fresh
amount of air is taken in as indicated in Figure 5.3. As the compression and power strokes of this idealized
cycle are adiabatic, the efficiency can be calculated from the constant pressure and constant volume
processes. The input and output energies and the efficiency can be calculated from the temperatures and
specific heats.
)(1 bcp TTCQ −= and )(2 dav TTCQ −=
1
21
QQQ
Efficiency+
==η
Here, Q1 = heat input, Q2 = heat Output, η = efficiency, Cp = Specific heat at constant pressure,
Cv = specific heat at constant volume, Ta = temperature at point a in diesel cycle,
Tb = temperature at point b, Tc = temperature at point c, and Td = temperature at point d
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
26
Figure 5.3: Air standard diesel engine cycle [23] According to the concepts in [20] and [23], it is a little hard to start diesel engines in cold weather as the low
metal temperature of the cylinder and head consume part of the heat generated in the cylinder during the
compression stroke, making it difficult for the fuel to ignite. This is why some of the diesel engines use little
electric heaters, called glow plugs, inside the cylinder to help ignition when starting.
An essential part of every diesel engine is a mechanical or electronic governor. The purpose of this governor is
to manage the engine speed by controlling the rate of fuel deliverance.
Figure 5.4: Cutaway view of an in-line overhead cam four-cylinder diesel engine [20]
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
27
For two reasons, diesel engines are heavier than Otto cycle or gasoline engines of the equal power. The first is
that diesel engines require a larger displacement of cylinders in each power stroke than a gasoline engine to
produce the same amount of power. The second reason for the greater weight of a diesel engine is it has to be
tough to endure the high combustion pressures required for ignition of the fuel, and also to bear the shock from
the detonation of the fuel mixture after ignition. Consequently, the weight of the piston and the connecting rod
are considerably heavier, and the cylinder and rest of the engine are also heavier in order to accelerate and to
decelerate these masses [21], [17], [18].
A diesel generator is basically a diesel engine that drives an alternator which converts the mechanical energy
into electrical energy. An alternator works by creating a magnetic field that rotates in such a manner that the
field passes through a set of coils. As the field passes through the coils, an electric current is induced on the
coils. This electric current is then regulated to become the electricity and is made available on the generator's
output. For very small generators, the magnetic field may be produced by permanent magnets, whereas for
larger generators, the field is generated by a set of coils.
5.2. WHY DIESEL? On account of growing demand and low supply, the prices of fuel are rising on a regular basis and therefore
we need a cost effective fuel to fulfil our requirements. The diesel engine has proved to be extremely efficient
and economical. The energy density of diesel is greater than gasoline, means that diesel fuel can provide
more energy as compared with the same amount of gasoline. For this reason, diesel engines can offer better
mileage in automobiles, and consequently the first choice for heavy-duty transportation and equipment. Diesel
engines draw larger interest today mainly because of higher efficiency and cost effectiveness [21], [22].
5.3. ADVANTAGES OF A DIESEL ENGINE
If studying in detail, there is a big list of advantages and distinct features of diesel engines over gasoline/petrol
engines. Briefly stating here, the diesel engine is more efficient and preferable as compared with gasoline
engine due to the following reasons [22]:
• Modern diesel engines have lower noise as well as lower maintenance costs. They need smaller
amount maintenance as compared with older models and gas engines of similar size.
• They are more rough and reliable.
• In a diesel engine, there is no sparking because of auto-ignition of fuel. Since there are no spark plugs
or spark wires, the maintenance cost is less.
• The cost of fuel consumed to produce one kilo-Watt power is 30% to 50% lower than that of gasoline
engines.
• A water-cooled 1800 rpm petrol engine normally runs 6000-10,000 hours before it needs maintenance
but an 1800 rpm water-cooled diesel engine runs 12,000 to 30,000 hours prior to any major servicing
is required.
• Petrol/Otto engines burn hotter than diesel engine, and therefore they have a considerably shorter life
compared with diesel engines.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
28
5.4. DIESEL POWER GENERATORS
Diesel engines are generally used as mechanical engines, power generators and in mobile drives. They are
extensively utilized in automobiles, locomotives, construction equipment, and numerous other industrial
applications. Industrial diesel engines and diesel powered generators are widely used in construction, marine,
mining, hospital, forestry, telecommunications, underground, and agricultural applications. Power generation
for prime or standby backup power is a major application of today's diesel engines.
Diesel powered electrical generators are used in many industrial and commercial organizations. These
generators can be used in homes for small loads, as well as in hospitals, industrial plants, and commercial
buildings for big loads. They can be used both as a source of prime power or standby power back-up. They
are available in various specifications and sizes. Diesel generator sets rating 5-30KW are normally used in
simple home and personal applications. Industrial applications cover a wider range of power ratings (from 30
kW to 6 Megawatts) and are used in different industries all over the world. For home use, single-phase power
generators are sufficient. Three-phase power generators are mainly used for industrial purposes [22].
5.5. TODAY'S DIESEL GENERATION
Modern diesel generators are extremely powerful and advanced. They are designed to last longer and to give
reliable and durable source of power. Commercial businesses can depend on them when a storm arrives or a
crisis crop up with the local power network. A good, solid diesel generator can give any commercial business
a safe, secure backup power source that can save a possible loss of production. All important industries such
as construction, marine, manufacturing, forestry, and mining, to name a few, needs a reliable backup power in
cases of blackouts. With a diesel generator, commercial industries can continue functioning normally with little
to no downtime [22].
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
29
66.. CCAASSEE SSTTUUDDYY:: PPOOWWEERR GGEENNEERRAATTIIOONN TTHHRROOUUGGHH DDIIEESSEELL GGEENNEERRAATTOORRSS
-- Power Generation by a Reputable Multi-national Engineering Firm in Karachi, Pakistan
There are many aspects to consider in power generation system reliability. Here, we try to give a brief
overview of personal experiences at an unnamed industrial site in Karachi, Pakistan to show the impact of
effective maintenance on the performance and reliability of such a system.
6.1. SITE OVERVIEW
The objective was to supply continuous electric power to a commercial building through diesel generator sets
which acted as a source of prime power for the building. This is why the generators operated at prime working
cycle. The reason for using diesel power as the prime source is the unreliable local power supply in Karachi,
which include voltage problems, total blackouts, and long downtimes. Most of the firms use diesel power as a
backup source but some, like this site, use diesel power as their only source which is much more reliable than
local power.
This power generation site actually started up in June 2000, when our company installed the entire setup and
the client took over operation and maintenance (O&M) on its own. Due to the inability of the client to manage
the system prudently and not observing the desired results, O&M was also subsequently awarded to our
company. The Power Generation Department of our company took over the site in July 2002. The author was
employed by this firm as a mechanical engineer in the maintenance department from 2002 to 2004.
6.2. COMPOSITION OF THE POWER GENERATION SYSTEM
The power generation site discussed in this study had three diesel generator sets listed in Table 6.1:
Table 6.1: List of Diesel generators at site
DG No. kVA Rating Safe Load (kilo-Watt) Make/Model Working Cycle
1 300 kVA 180 kW Cummins NT 855 Prime
2 300 kVA 180 kW Cummins NT 855 Prime
3 70 kVA 42 kW Perkins UK 1004 TG Prime
The ratio of the actual power transmitted (“real power”) to the apparent power that could have been transmitted
if the same current were in phase and undistorted is known as the power factor (PF). It is always less than or
equal to 1 [28].
)()(.
kVAPowerApparentkWPowerealRPF =
DG-1 and DG-2 had apparent power of 300 kVA with power factor of 0.8 which gives 240 kilo-Watt of real
power transmission for each set. Safe working load of each set is 75% of the full load which was 180 kW for
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
30
DG-1 and DG-2. These two sets were synchronized (i.e., can share the load while working simultaneously)
and are used in the daytime. DG-3 was a 70 kVA set which gave 56 kW of real power and 42 kW of safe
working load, and worked nights and all day on Saturdays and Sundays (when the site required minimum
load). DG3 was not synchronized with DG1 and 2.
Synchronizing is the term given to electrically connecting two or more generators together. Paralleling means
electrically connecting a generator or set of generators to the electrical supply network. Both synchronizing
and paralleling add additional cost to a project but have significant advantages. Synchronizing can allow a
stepped approach to power supply using several generators that automatically switch in or out depending on
the required load. This means the generators are used in the most fuel efficient manner. It is sometimes
better to run three 500 kVA generators in a load stepped configuration than it is to run a 1500 kVA generator.
Running a generator lightly loaded for an extended period is inefficient and can cause a diesel engine to glaze
its bores. Having a synchronizing panel gives you the ability to add additional generating capability as required
without losing power to the building [14].
For detailed specifications and configuration of the Cummins C300-D5, Engine NT855 Diesel Generator Set,
refer to Appendix-A [24].
6.3. RELIABILITY ANALYSIS OF THE SYSTEM
The entire power generation setup of the site can be referred as a series system for around 10 hours of
daytime in weekdays. This was mainly due to the absence of a standby generator at site. This situation is
elaborated in the load distribution graph in Figure 6.3. These circumstances were quite alarming and required
client’s immediate attention. There was a 100 % chance of a complete blackout even if a single generator has
failed during the daytime, (i.e., entire system would have failed if one of its components has failed). The client
was well aware of the situation and was planning to install a standby generator to avoid any possible blackouts
as well as to improve reliability of the entire system. Our company was working on the feasibility of this project
to get the client’s approval until I left in May 2004. As indicated in the load curve in Figure 6.3, the maximum
load required by the site never exceeded 250 kW. We were dealing with this problem in such a manner that
whenever we faced problems with any of the generators during peak load hours, we requested the client to
turn off the air-conditioning system of the building. We had full support of the client in case of such a difficulty.
This trick reduced the building load to a level of less than 180 kW, which was well bearable for the only
available 300 kVA generator at that particular time. In the meantime, the broken generator set was attended
and fixed, and the system became functional again. As we already studied in Section 2.1, this is an example of
corrective maintenance.
Now we will briefly look into the reliability analysis of this entire power generation system. During the peak load
hours in the daytime, the system block diagram can be shown as in Figure 6.1.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
31
Figure 6.1: RBD of the Power Generation System during daytime in weekdays
DG-3 is not included in the above diagram as it was not synchronized with DG-1 and DG-2 as mentioned
before. DG-1 & DG-2, both were 300 kVA Cummins NT855 models of 1998 and 2001 respectively. We do not
have the exact reliability figures of these generators, but here we try to make some logical assumptions for
their reliabilities for the study purpose, taking the generators’ condition, history, and year of manufacture into
account. By assuming R1 = 0.9 for DG-1 and R2 = 0.95 for DG-2, we can calculate the reliability of this system
during peak hours of daytime as explained in Section 3.1.1.3 (a):
855.0)95.0).(9.0(. 21 === RRRS
During the weekends, nights, and public holidays, the system reliability block diagram can be shown as in
Figure 6.2.
Figure 6.2: RBD of the Power Generation System during weekends and nights
Since during the minimum load hours, all the three generators of the site were available. DG-3 was 70 kVA
Perkins model of 1998, so we assume its reliability as R3 = 0.9. Now we can calculate the reliability of the
power generation system as it worked on weekends and nights as following:
9995.0)9.01).(95.01).(9.01(1)1).(1).(1(1 321 =−−−−=−−−−= RRRRS
We can see that the reliability of this parallel configuration at minimum load is 0.9995 which is considerably
higher than the reliability of series configuration at peak load hours in daytime, which is 0.855.
DG-1 180 kW
DG-2 180 kW
DG-3 42 kW
DG-2 180 kW
DG-1 180 kW
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
32
The load distribution shown in Table 6.2 is one of the daily observations recorded at site. These observations
were taken on August 01, 2002, a month after our company took over the site. This table gives a very clear
picture of site load conditions. It tells us the total load of the building, the power generated by the entire
system, and the load each generator was bearing at that time. The adverse affects of this type of loading,
when there is a big difference in generated load and consumed load, are discussed in upcoming section. The
load distribution curve in Figure 6.3 based on these observations gives a better understanding of site load
conditions.
Table 6.2: Load Distribution at Site observed on August 01, 2002
Time DG1 Load (kW)
DG2 Load (kW)
DG3 Load (kW)
Building Load (kW)
Available Load (kW)
7:00 90 0 0 90 180
8:00 130 0 0 130 180
9:00 100 100 0 200 360
10:00 110 110 0 220 360
11:00 110 110 0 220 360
12:00 120 120 0 240 360
13:00 120 120 0 240 360
14:00 120 120 0 240 360
15:00 120 120 0 240 360
16:00 120 120 0 240 360
17:00 120 120 0 240 360
18:00 120 120 0 240 360
19:00 90 90 0 180 360
20:00 0 110 0 110 180
21:00 0 100 0 100 180
22:00 0 90 0 90 180
23:00 0 90 0 90 180
0:00 0 0 40 40 42
1:00 0 0 30 30 42
2:00 0 0 30 30 42
3:00 0 0 30 30 42
4:00 0 0 25 25 42
5:00 0 0 25 25 42
6:00 0 0 25 25 42
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
33
Figure 6.3: Load Distribution Curve at Site observed on August 01, 2002
LOAD
CURV
E
90
130
200
220
220
240
240
240
240
240
240
240
180
110
100
9090
180
180
360
360
360
360
360
360
360
360
360
360
360
180
180
180
180
4242
4242
4242
42 2525
2530
3030
40
050100
150
200
250
300
350
400
7:00
8:00
9:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
0:00
1:00
2:00
3:00
4:00
5:00
6:00
TIME
L O A D -K W
Build
ing Lo
ad (k
W)Av
ailab
le Lo
ad (k
W)
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
34
6.6. PROBLEMS FACED
The load curve based on one of the daily observation charts dated August 01, 2002, is shown in Figure 6.3
and the data is available in Table 6.2. It shows that the total load of the building varied between 90 kW to 240
kW during daytime and between 25 kW to 45 kW on the weekends and nights. The major problems that we
faced when the site was handed over to us in July 2002 were:
1. Absence of standby generator: Absence of standby generator for peak hours in daytime as both 300
kVA generators (i.e., DG-1 and DG-2) were running simultaneously for around 10 ~ 11 hours without
any back up.
2. High Fuel Consumption: At this kind of load distribution, where both 300 kVA generators which can
bear 180 kW of load each, runs at 100 or 90 kW of load as indicated in the load curve in Figure 6.3,
the specific fuel consumption (SFC), which is the amount of fuel consumed to produce 1 kWh of
power, is at maximum. Another significant reason for higher fuel consumption was poor maintenance
of generators.
3. Lack of Maintenance: The preventive maintenance was very much neglected by the previous
management and all the generators needed immediate attention for maintenance. Due to this reason,
the site had a history of large downtimes and failures when it was handed over.
In response to the first problem, as we already mentioned, the client was planning to install a standby
generator and our company was working on the feasibility of this project to get the client’s approval.
For the remaining two problems, we will throw some light on the actions that were taken to solve these
problems and what was their outcome.
As specified in manufacturer’s maintenance procedures and recommendations, preventive maintenance
should be followed according to the schedule provided in Table 6.3.
Table 6.3: Schedule for preventive maintenance of generators
1. Changing Lube oil, Oil filter, Fuel filter and Air filter Every 250 running hours
2. Fuel injectors/Tappet Calibration Every 1500 running hours
3. Top Overhauling (Minor OH) Every 6000 running hours
4. Major Overhauling Every 10,000 running hours
When the site was taken over by our company in July 2002, it was revealed that the fuel injectors’ calibrations
of DG-1 and DG-2 have been delayed for more than 3500 running hours. As mentioned in Table 6.3, this
activity was supposed to be carried out at every 1500 running hours. We took a set of observation on August
01, 2002 before carrying out the injectors’ calibrations of both generators to examine the fuel consumption of
both generators without calibration. These observations are shown in Table 6.4. The capacity of the main
diesel tank at was 8,000 liters through which the fuel is supplied to all the three generators at site. Each
generator has its own fuel sump at the base of skid which is programmed to be automatically refilled when the
diesel level is reduced to half of its actual capacity. The fuel sump capacity of the 300 kVA Cummins generator
sets (i.e., DG-1 and DG-2) is 800 liter.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
35
Table 6.4: Observation of fuel consumption of DG-1 and DG-2 before tappet calibration on August 01, 2002
Diesel Consumed
(Liters) Time Synchronizing Panel Meter Reading (kWh)
DG1 DG2
DG Set Fuel Sump Level
Laod DG1 (kW)
Laod DG2 (kW)
7:00 1019488 Full – 800 liters 100 100
8:00 - 100 100
9:00 - 105 110
10:00 - 90 95
11:00 - 95 100
12:00 1020160 400 400 Half – 400 liters 100 100
13:00 Re-filled – 800 liters 100 100
14:00 - 100 110
15:00 - 110 110
16:00 - 100 100
17:00 - 90 90
17:50 1021472 400 400 Half – 400 liters 90 100
For the above set of observation we can see in the ‘fuel sump level’ column that the amount of diesel
consumed by each generator during the observed time is 800 liters. Synchronizing panel has a meter installed
on it which tells us the cumulative power generated by the entire system. The difference in the synchronizing
panel’s meter reading at the start and end of observation is in fact the total power generated in kWh within the
observed time.
Synchronizing panel’s meter reading difference = 1021472 - 1019488 = 1984 kWh
SFC, as already defined, is the amount of diesel consumed to produce 1 kWh of power. Therefore, SFC of the
power generation system before injectors’ calibration is:
SFC = 8000 / 1984 = 0.403 liters / kWh
After conducting the fuel injectors’ calibrations of both generators, DG-1 and DG-2, on August 10, 2002, we
observed the fuel consumption of these generators again on August 18, 2002. These observations are shown
in Table 6.5.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
36
Table 6.5: Observation of fuel consumption of DG-1 and DG-2 after tappet calibration on August 18, 2002
Diesel Consumed
(Liters) Time Synchronizing Panel Meter Reading (kWh)
DG1 DG2
DG Set Fuel Sump Level
Laod DG1 (kW)
Laod DG2 (kW)
7:00 1096384 Full – 800 liters 100 100
8:00 - 100 100
9:00 - 100 110
10:00 - 100 95
11:00 - 95 100
12:00 1097600 400 400 Half – 400 liters 100 95
13:00 Re-filled – 800 liters 100 100
14:00 - 100 110
15:00 - 110 100
16:00 - 100 110
17:00 - 95 90
17:50 1098848 400 400 Half – 400 liters 90 100
Here, we can see again that the fuel consumed by each generator is 800 liters for the observed time.
Synchronizing panel’s meter reading difference, which is the total power generated by the system in observed
time, is:
Synchronizing panel’s meter reading difference = 1098848 - 1096384 = 2464 kWh
Now the SFC after injectors’ calibration of both generators, DG-1 and DG-2, is:
SFC = 8000 / 2464 = 0.325 liters / kWh
This set of observation clearly shows a considerable improvement in SFC from 0.403 to 0.325 after we carried
out fuel injectors’ calibrations of DG-1 and DG-2. Although the ideal SFC for these generators as claimed by
the manufacturer is 0.25 liters/kWh, at this kind of minimum load of the building, as indicated in the load
distribution curve in Figure 6.3, each generator was running at extremely light load for an extended period of
time and therefore the SFC was exceedingly high. For this reason, there was still much room for improvement
of SFC at this site which could have only been achieved if the building load would have remained around the
maximum load these generators can bear, i.e., between 300 to 360 kWh. It was also observed that there was
a tremendous improvement in the engines’ sounds of DG-1 and DG-2 after the tappet calibrations were done.
In Table 6.6, we are listing the record of failures and downtimes at the site from June 2001 to June 2002,
before it was handed over to our company.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
37
Table 6.6: Record of failures and downtimes from June 2001 to June 2002
Number of failures S.No Parts DG1 DG2 DG3
Downtime caused by
failures (hrs) A Engine 1 Flexible coupling and flywheel 0 0 0 0 2 Dry type air filter with clogged condition indicator 0 0 0 0 3 Cooling radiator 2 1 0 3 4 Fuel pump 0 1 1 4 5 Gasket 1 0 1 4 6 Electronic governor 0 1 0 0 7 Dual fuel filter with on line filter changing provision. 0 0 0 0 8 Lube oil pump, oil cooler and filter 1 0 1 0 9 Turbo charger 1 1 0 2 10 24V DC starter & battery charging alternator 0 0 0 0
11 Engine mounted microprocessor based control panel that displays engine and electrical parameters
a Lube oil pressure indicator and temperature gauge 1 0 0 0 b Tacho meter for speed indication with hour meter 0 0 0 0 c Battery charging Ammeter 0 0 0 0 d Starting switch with key 0 0 0 0 e Over speed stop switch with contacts 0 0 0 0 f Low lube oil pressure switch 0 0 0 0 g High water temperature alarm & trip 0 0 0 0
12 Stainless steel flexible for engine exhaust 0 0 0 0
13 Control cables from engine to AMF (Automatic Mains Fail) panel 0 0 0 0
14 Batteries 0 0 0 0 B Alternator 1 Continuous damper winding 0 0 0 0 2 RTDs (Resistance Temperature Detectors) 0 0 0 0 3 Anti condensation heaters 0 2 1 3 4 Pilot exciter 0 0 0 0 5 3-Phase sensing AVR (Auto Voltage Regulator) Card 2 0 1 4 TOTAL 8 6 5 20 hrs
Here, we can see that the total number of failures at site from June 2001 to June 2002, which is the sum of
failures occurred in each generator, (i.e., DG-1, DG-2, and DG-3), at site, is: Total number of failures at site from June 2001 to June 2002 = 8 + 6 + 5 = 19
The downtimes caused by these failures are also recorded in Table 6.6 which sums up to 20 hours for the
observed year. We can now calculate the total operating hours for the observed year:
Total operating hours from June 2001 to June 2002 = (365 x 24) - 20 = 8740 hours
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
38
Now, for the observed year, we can calculate the Estimated Failure Rate, which can be defined as the number
of failures in a particular span of time.
Estimated Failure Rate from June 2001 to June 2002 = Number of failures / Operating hours
= 19 / 8740 = 0.002174 failures / hour
Now we will list the record of failures and downtimes at the site from June 2002 to June 2003 in Table 6.7.
Taking you back, our company took over the site in July 2002 and these records are of one of the year when
our company managed the site.
Table 6.7: Record of failures and downtimes from June 2002 to June 2003
Number of failures S.No. Parts DG1 DG2 DG3
Downtime caused by
failures (hrs) A Engine 1 Flexible coupling and flywheel 0 0 0 0 2 Dry type air filter with clogged condition indicator 0 0 0 0 3 Cooling radiator 0 1 0 1 4 Fuel pump 0 0 1 1 5 Gasket 0 1 0 1 6 Electronic governor 0 0 0 0 7 Dual fuel filter with on line filter changing provision 0 0 0 0 8 Lube oil pump, oil cooler and filter 1 0 0 0 9 Turbo charger 0 1 0 0 10 24V DC starter & battery charging alternator 0 0 0 0
11 Engine mounted microprocessor based control panel that displays engine and electrical parameters
a Lube oil pressure indicator and temperature gauge 0 0 0 0 b Tacho meter for speed indication with hour meter 0 0 0 0 c Battery charging Ammeter 0 0 0 0 d Starting switch with key 0 0 0 0 e Over speed stop switch with contacts 0 0 0 0 f Low lube oil pressure switch 0 0 0 0 g High water temperature alarm & trip 0 0 0 0
12 Stainless steel flexible for engine exhaust 0 0 0 0
13 Control cables from engine to AMF (Automatic Mains Fail) panel 0 0 0 0
14 Batteries 0 0 0 0 B Alternator 1 Continuous damper winding 0 0 0 0 2 RTDs (Resistance Temperature Detectors) 0 0 0 0 3 Anti condensation heaters 0 0 1 0 4 Pilot exciter 0 0 0 0 5 3 Phase sensing AVR (Auto Voltage Regulator) Card 1 1 0 1 TOTAL 2 4 2 4 hrs
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
39
Here, we can sum the total number of failures at site from June 2002 to June 2003 as follows: Total number of failures at site from June 2002 to June 2003 = 2 + 4 + 2 = 8
The downtimes caused by these failures are recorded in Table 6.7 and these failures sums up to 4 hours for
this observed year, which is considerably less than the previous year’s downtime. Now calculating the total
operating hours for the observed year:
Total operating hours from June 2002 to June 2003 = (365 x 24) - 4 = 8756 hours
Estimated Failure Rate from June 2002 to June 2003 can be calculated as follows:
Estimated Failure Rate from June 2001 to June 2002 = Number of failures / Operating hours
= 4 / 8756 = 0.000914 failures / hour
It is can be clearly noticed by comparing Table 6.6 and Table 6.7 that the number of failures reduced 58%,
(i.e., from 19 to 8), Estimated Failure Rate also reduced 58%, (i.e., from .002174 to .000914 failures per hour,
as shown in Figure 6.4(a)), and the downtime was reduced to 80%, (i.e., from 20 hours to 4 hours, as shown in
Figure 6.4(b)), which is a huge difference. This difference was mainly because of negligence of preventive
maintenance by the previous management as already discussed, and because our company very strictly
followed the preventive and scheduled maintenance procedures after taking over the site as recommended by
the manufacturer.
0.002174
0.000914
0
0.0005
0.0010.0015
0.002
0.0025
Failure Rate
2001-2002 2002-2003
Year
Figure 6.4 (a): Estimated Failure Rate Comparison
20
4
0
5
10
15
20
Hours
2001-2002 2002-2003
Year
Figure 6.4 (b): Downtime Comparison
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
40
77.. CCOONNCCLLUUSSIIOONNSS Maintenance processes are meant to increase the mean time to failure of a component or system. As defined
earlier in this study, maintenance is a set of actions carried out to avoid a component or system failure or to
repair regular degradation that a system or component goes through with the passage of time to keep it in
proper working order.
The concept of corrective maintenance allows any system to work until it fails completely, only providing the
option of repair or replacement of defected component when evident problems arise. Recorded data and
studies show that operating cost in this style is maximum of all types of preventive maintenance. The benefits
of this method are that it works well if downtimes do not have an effect on production and if labour and material
costs do not matter.
The idea of preventive (scheduled/time-based) maintenance deals with the scheduling of maintenance actions
at planned time intervals, where broken component is fixed or replaced before any noticeable problem arises.
Studies show that the costs of operating a system in this manner are very less than corrective maintenance.
The concept of predictive maintenance involves planning of maintenance actions only if and when some
specific operational situations are observed such as increased vibration, temperature and/or lubrication
degradation, or any other detrimental trend through timely monitoring of the system. When any predefined
specific characteristic or parameter reaches a predetermined undesirable level, the system is shut off for
repairing or replacing broken components to prevent a more pricey failure from taking place in future. Studies
show that operating cost for this approach is even lesser than scheduled maintenance. Also, since
maintenance is only acted upon when it is needed, the production capacity is likely to improve.
Reliability centered maintenance approach involves all of the above mentioned methodologies of predictive
and preventive maintenance, in accordance with root cause failure analysis. This not only identifies exact
problems that occur, but also guarantees that highly developed repair and replacement procedures are
adopted, including possible modification of the system, and therefore assisting to stay away from problems or
keep them from taking place. Studies show that in this fashion, the operating cost is the least. This approach
works exceptionally well if the maintenance workforce if well aware, skilled, and knows the time to carry out all
of the required actions. In RCM approach, system maintenance works can be scheduled in an orderly manner
as with the predictive based method. Moreover, extra improvement efforts also can be taken on to minimize or
eliminate probable problems from happening again and again. It also gives opportunity to purchase materials
for necessary repairs ahead of time, hence reducing the requirement of huge inventory. Since maintenance
work is carried out only when needed, and additional efforts are put forward to comprehensively examine the
reason of the breakdown and determine the techniques to enhance system reliability, there can be a
considerable improvement in production capacity [1].
We know that while there have been dramatic technological advances in energy efficiencies that have resulted
in significant energy savings, (an example of which is discussed in Section 5.5 “Today’s Diesel Power
generation”) there is still room for improvement. In any power generation system (for instance the one
mentioned in the case study in Section 7), the most immediate opportunity to improve the reliability of the
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
41
system and save energy is taking the simple and relatively easy steps of preventive maintenance in a timely
manner to ensure that the system is maintained at peak efficiency and hence reducing failure rate and
downtimes.
According to a survey conducted in Karachi, Pakistan in 2004, over 90% of Commercial Diesel Generators
were under-performing due to one problem or another [29]. In many cases, the problem was as simple as a
dirty filter. In the commercial arena, up to 50% more energy would be saved through proper installation, sizing
and maintenance of Diesel Generators and related equipment [29]. Improving system efficiency by 10% to
20% is a conservative estimate of the impact of proper maintenance [29]. For systems that are seldom or
never serviced, the savings could reach 100% [29]. To achieve this efficiency, we would recommend to strictly
following the requirements for system maintenance timely and efficiently. Just like the system mentioned in the
case study, the results that have been compared for both scenarios (i.e. system conditions with maintenance
and without maintenance), clearly shows how much impact can be made on the system reliability and
availability by following proper preventive maintenance procedures.
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
42
RREEFFEERREENNCCEESS [1] Pro-Active Maintenance for Pumps, Archives, Piotrowski, J. April 2, 2001. [2] Underhall Terminology - Maintenance terminology, Svensk Standard SS-EN 13306, Stockholm: SIS Förlag AB 2001 [3] System Reliability Theory, M. Rausand, A. Hoyland, Hoboken: John Wiley & Sons 2004, ISBN 0-471-47133-X [4] Handbook of condition Monitoring, A. Davies, London: Chapman & Hall 1998, ISBN 0-412-61320-4 [5] Reliability Theory with Application - Preventive Maintenance, I. Gertsbakh, Berlin: Springer-Verlag 2000, ISBN 3450-67275-3 [6] S. H. Sim and J. Endrenyi, “Optimal preventive maintenance with repair,” IEEE Trans. Reliability, vol. 37, no. 1, pp. 92–96, Apr. 1988. [7] J. Endrenyi and S. H. Sim, “Availability optimization for continuously operating equipment with maintenance and repair,” in Proceedings of the Second PMAPS Symposium, September 1988, Nov. 1989, EPRI Publication EL-6555. [8] L. Bertling, “Reliability centred maintenance for electric power distribution systems,” Ph.D. dissertation, Dept. Elect. Power Engineering, KTH, Stockholm, Sweden, 2002. [9] Reliability, Availability, and Maintainability, “RAM”, website: www.answers.com/topic/reliability-availability-and-maintainability (accessed in March 2007) [10] System Reliability, Maintainability, and Avalability Analysis, “Overview”, website: http://www.weibull.com/basics/system_reliability.htm (accessed in March 2007) [11] Corrosion Maintenance, “The maintenance revolution”, website: http://www.corrosion-doctors.org/Inspection/Maintenance.htm (accessed in February 2007) [12] Lecture Notes for MecE 514 “Reliability for Design” Fall- 2006 by Zhigang Tian (University of Alberta) [13] What Analytical RCM is?, “Analytical RCM”, website: http://www.ca-advisors.com/quals/epri/a_rcm/ (accessed in April 2007) [14] Power Hire – about generators, “Synchronizing”, website: http://www.powerhire.co.nz/tools/faqs.html [15] The value of reliability in power systems - pricing operating reserves - José Fernando Prada, MIT-EL99-005 WP July 1999 [16] Perkins Engines Company – Industrial, “Engine Genetics”, website: http://www.perkins.com/cda/ (accessed in March 2007) [17] How Diesel Engine Works?, “ Diesel Fuel”, website: http://auto.howstuffworks.com/diesel.htm (accessed in February 2007) [18] Modern High-Speed Oil Engines Volume II by C. W. Chapman published by The Caxton Publishing Co. Ltd. reprinted in July 1949 [19] DP Chip Power, “Generat Diesel Information”, website: http://www.dpchip.com/pumpinfo.html (accessed in June 2007)
M.Eng. Project Report, Winter 2007
Role of Maintenance in Component and System Reliability Syed J A Haider
43
[20] Engine-AutoZone, “How engine works” website: http://www.procarcare.com/icarumba/resourcecenter/encyclopedia/icar_resourcecenter_encyclopedia_engine1.asp (accessed in March 2007) [21] Internal Combustion Engines Fundamentals – John B. Heywood [22] Industrial Diesel Generators and Engines, “Types and Applications”, website: http://www.dieselserviceandsupply.com/ (accessed in April 2007) [23] The Diesel Engine, “Diesel Engine Cycle”, website: http://hyperphysics.phy-astr.gsu.edu/hbase/thermo/diesel.html (accessed in March 2007) [24] Cummins N Series Generator Sets (200kVA ~ 450kVA), “Detailed Lead Description”, website: http://honny1992.en.alibaba.com/offerdetail/52720984/Sell_Cummins_N_Series_Generator_Set.html (accessed in April 2007) [25] PBMA-KMS, “What is FTA?”, website: http://pbma.nasa.gov/framework_content_cid_356 (accessed in June 2007) [26] CIRAS-FMEA, “What is FMEA?” website: http://www.ciras.iastate.edu/quality/fmea.asp (accessed in June 2007) [27] Understanding High Availability of IP and MPLS Networks, “Defining key terms – Unavailability” website: http://www.ciscopress.com/articles/article.asp?p=361409&seqNum=2&rl=1 (accessed in June 2007) [28] BC Hydro. Guides to Energy Management (GEM) Series: Power Factor. Vancouver, B.C. 1999. [29] Prospects of energy in Pakistan by Aamir Kabir, March 03, 2003, The Daily Dawn
AAPPPPEENNDDIIXX –– AA
Specifications of Cummins C-300 D5 Generator with NT- 855 Engine
Diesel PoweredGenerating SetsC300 D5
Standard Genset Features Generator Set Performance Generator Set OptionsCummins water cooled Diesel engine, Voltage Regulation Mechanical OptionsOil and fuel filter fitted, water separator,, Maintains voltage output to within ±1.0%. Compliance - CE Certification (Guarding)Lube-oil drain valve fitted At any power factor between 0.8 and 1.0Electric starter & Charge alternator 24 v D.C. At any variations from No load to Full load. Fuel sytem optionsElectronic governor At any variations from Cold to Hot. Low Fuel Level ShutdownNormal duty air filter At speed droop variations up to 4.5%.Single bearing alternator, class H/H , IP23 Frequency RegulationStandard voltage 400/230 volts 50 Hz Isochronous under varying loads from no Exhaust OptionsExciter/Voltage reg - Torque Match as std load to 100% full load when electronic Exhaust Silencer - Industrial (9 dB), In-Line PCC2100 without Bargraph governor is ?tted. Exhaust Silencer - Residential (25 dB), In-Line 3 pole MCCB Random Frequency Variation Flexible and fixing kit for Industrial silencerWelded steel base frame with A/V mounting, Random Frequency Variation Flexible and fixing kit for Residential silencerAnti Vibration Mounts Will not exceed ±0.25% of its mean value forSingle skin metal fuel tank constant loads – no load to full load.Tank capacity of min 8 hours operation at Waveform Warranty70% standby load Total harmonic distortion open circuit voltage Warranty - 5 Year Extended Standby ApplnLoose 9 dB(A) silencer waveform in the order of 1.8%. Three-phase Warranty - 2 Year Extended Prime ApplnSet mounted starting battery balanced load in the order of 5.0%.Engine Tractor Blue & Alternator Munsell Jade Green Telephone In?uence Factor (TIF) Voltage ConnectionsRadiator and Guarding black TIF better than 50. 277/480V, 3 PhasePacking under shrunk plastic film THF to BS 4999 Part 40 better than 2%. 254/440V, 3 PhaseOperation & Maintenance manual Alternator Temperature Rise 240/416V, 3 PhaseStandard set of labels Class H insulation. 230/400V, 3 Phase
220/380V, 3 PhaseRadio Interference 127/220V, 3 PhaseIn compliance with BS 800 and VDE levels 115/200V, 3 PhaseG and N. 110/190V, 3 Phase
Miscellaneous OptionsEngine Specification Alternator Specification Coolant heater -240/120VCummins NT855G6 Type Battery Charger 110-277V, 3AIn-line direct injection Brushless single bearing, revolving ?eld, PCC2100 with bargraph6-cylinder diesel engine. pole, drip proof, screen protected. Automatic Transfer SwitchesType Class H Insulation, IP23 Protection Packing - Export BoxWater cooled, four cycle IC 01 cooling system. Compliance StandardsTurbocharged Fully interconnected damper winding. To BS4999/5000 pt 99,Construction AC exciter and rotating recti?er unit. VDE 0530, UTE5100,Two valves per cylinder, forged steel Epoxy coated stator winding. NEMA MG1-22, CEMA,crankshaft and connecting rods, cast iron Rotor and exciter impregnated with tropical IEC 34, CSA A22.2,block. grade insulating oil and acid resisting AS1359, BSS 5514,Starting polyester resin. Dynamically balanced rotor ISO 3046 and ISO 852824 volt negative earth. Battery charging BS 5625 grade 2.5.alternator 35 amp on engine. Cranking Sealed for life bearings.current 640 amps at 0°C. Layer wound mechanically wedged rotor.Fuel System24 volt fail safe actuator. Spin-on paper Exciterelement fuel ?lters with Bosch fuel Triple dipped in moisture, oil and acidpump injection system with integral resisting polyester varnish and coated withElectronic governor. Dual ?exible fuel lines anti-tracking varnish.and connectors. Standard fuel waterseparator.Filters Output windings with 2/3 pitch for improvedAir cleaner with dry element and restriction harmonics and paralleling ability.indicator. Spin-on full ?ow lube oil ?lter. Close coupled engine/alternator for perfectCooling alignment.50°C radiator as std Oil cooler. Drain Tap
Model name
ESP PRP ESP PRP
C300 D5 300 275 240 220
Specifications may change without noticeC300 D5
kVA kWe
Technical Data
PRIME POWER (PRP)Prime power is available for an unlimited number of annual operating hours in variable load applications, in accordance with ISO8528-1.A 10% overload capability is available for a period of 1 hour within a 12-hour period of operation, in accordance with ISO 3046-1.STANDBY POWER RATING (ESP)
In installations served by unreliable utility sources (where outages last longer or occur more frequently), where operation is likely to exceed 200 hours per year, the prime power rating should be applied. The Standby Power rating is only applicable for emergency and standby applications where the generator set serves as the back up to the normal utility source. All ratings are based on the following reference conditions:- Ambient temperature – 27oC
Dimensions and Weights
Model Engine Length (mm) Width (mm)Height (mm) (50Deg Rad)
Set weight wet (Kg)
Set weight dry (Kg)Enclosed Weight
Wet (Kg)
C300 D5 NT855G6 3549 1100 2078 3360 3170 5005
06/2004 C300 D5
Model
Set output
C300 D5
380-440 V 50 Hz
Speed
Alternator voltage regulation
Prime Rating
Standby Rating
Engine Make
Engine Model
Starting/Min °C
Cylinders
Engine build
Standard Governor/Class
Aspiration and cooling
100 A/hrBattery capacity
Gross Engine output – Prime
Cummins
NT855G6
Six
In-line
Bore and stroke
Compression Ratio
Cubic capacity Air intake – engine (Prime)
220 kWe 275 kVA
240 kWe 300 kVA
280 kWm
Electronic/Class G2
Turbocharged
140 mm x152 mm
14:1
14 Litres
Unaided / 4°C
Alternator insulation class
Fuel consumption (Prime)
Exhaust gas flow – prime
Exhaust gas back pressure max
Pusher fan head (duct allowance)*
Heat radiated by eng (Prime)
Fuel consumption (Standby)
Lubrication system oil capacity
Base fuel tank capacity – open set
Coolant capacity
Exhaust temp – prime
Minimum air opening to room
Minimum discharge opening
Air flow – radiator*
1500 rpm
±1.0%
H
56 l/hr
62 l/hr
38.6 Litres
750 or 900 Litres
63.9 Litres
574°C
1071 l/s
76 mm Hg
4.92m3/s
361 Litre/s
2.10 sq m
1.39 sq m
13 mm Wg
- Altitude above sea level – 150 metres - Relative humidity – 60%
The Standby Power Rating is applicable for supplying emergency power for the duration of a utility power interruption. No overload, utility parallel or negotiated outage operation capability is available at this rating.
50 kWm310 kWmGross Engine output – Standby
Diesel Power Generator – Cummins NT855 [24]