CORE Metadata, citation and similar papers at core.ac.

uk
Provided by Scientific Journal of PPI - UKM

Scientific Journal of PPI-UKM Sciences and Engineering

ISSN No. 2356 - 2536

Availability Improvement Methodology in Thermal Power Plant

H. Pariamana*, I. Garniwaa, I. Surjandarib, B. Sugiartob
a
Department of Electrical Engineering, Faculty of Engineering - University of Indonesia, Depok City, West Java16424, Indonesia
b
Department of Industrial Engineering, Faculty of Engineering - University of Indonesia, Depok City, West Java 16424, Indonesia

Abstract

Availability of a complex system of thermal power plant is strongly influenced by maintenance program and component reliability.
Various maintenance techniques, likes RCM (reliability-centred maintenance), RBM (risk based maintenance) and CBM (condition-based
maintenance), have been applied to improve the availability. Implementation of RCM, RBM, CBM alone or combined RCM and RBM or
RCM and CBM is a maintenance technique used in thermal power plants. This study develops an new maintenance methodology
integrating RCM, RBM and CBM to increase the availability of thermal plants. The method generates MPI (Priority Maintenance Index)
and FDT (Failure Defense Task). MPI is used to determine the priority of components in maintenance program. FDT consists of the tasks
of monitoring and assessment of conditions other than maintenance tasks. Both MPI and FDT obtained from development of functional
tree, failure mode effects analysis, fault-tree analysis, and risk analysis (risk assessment and risk evaluation) were then used to develop and
implement a plan and schedule maintenance, monitoring and assessment of the condition and ultimately perform availability analysis. The
results of this study indicate that the reliability, risks and conditions-based maintenance methods, in an integrated manner can increase the
availability of thermal power plant.

Keywords: Integrated Maintenance Techniques, RCM, RBM, CBM, Availability, MPI

1. Introduction decreases. For keeping the system availability, a proper

maintenance program is needed.
A complex system with high production loss value, such In a complex system, there are several maintenance
as a power plant, keeping availability and reducing costs problems i.e. how to identify critical components, how to
related to maintenance are at the top of management priority and how to maintain them. Maintenance techniques
concerns in a large enterprise. The system availability is developed in the literature propose various rules to category
determined the component reliability and the maintenance the system components into critical component. Then, they
program implemented. That program influences the repair determine the priority maintenance and finally plan a
time and the reliability of component and system. maintenance program which consists of preventive
Normally, a system’s reliability will deteriorate and this maintenance and condition monitoring tasks.
increases the probability of failure. Further, the availability Various maintenance techniques have been proposed to
of the system will decrease. In case of the power plant, the increase an availability of any systems [3]. Generally, the
unavailability occurred results a high cost of production established maintenance techniques have been developed in
loss depend on the time duration of shutdown condition. the literature i.e. Reliability Centred Maintenance (RCM),
Availability measures are related to how long time the Risk Based Maintenance (RBM) and Condition Based
unit can operate in the certainty period time. Most power Maintenance (CBM).
plants use the index proposed by IEEE std.762tm (2006) to The development and applying of RCM, RBM and
define availability [1]. That index represents the percentage CBM can be found in [4]-[10], [11]-[15] and [16]-[19]
of a given period of time, expressed in hours that the unit is respectively. Several researchers have been integrated two
in service (including reserve shutdown state). Reference [2] of those maintenance techniques in one maintenance
defined expressed by the ratio of the mean time to failure to program. Integration between RCM and CBM, RCM and
the sum of the mean time to failure plus the mean time to RBM has been done in [20] and [21], respectively. This
repair. The index, usually evaluated monthly, is reported in integration increases the availability higher than that each
a Generating Availability Data System (GADS) and can be the techniques is implemented separately. In the present,
used for comparison between different generating systems. the study of integration of those techniques has been not
A reduction in availability is caused by planned developed. Integration all techniques are expected to yield
maintenance and unplanned maintenance actions. Improper the increasing of availability higher than that resulted of
maintenance can result a repair time longer than that based integration between RCM – RBM or RCM – CBM.
on manufacturer’s recommendation or the system reliability

*
Corresponding author. Tel.: N/A; fax: N/A.
E-mail address: [email protected].

43
 44

Nomenclature This paper is structured in the following way: Section 1

t = time period [h] provides a background of the methodological maintenance
R(t) = reliability at time t
M(t) = reliability at time t development. The steps of the methodological maintenance
β = Weibull distribution shape parameter development are explained in Section 2. Selected example
η = Weibull distribution characteristic life [h] of application to thermal power plant is discussed in
μ = Mean in the logarithmic domain, lognormal distribution Section 3. Conclusions are then discussed in the last
σ = Standard deviation in the logarithmic domain lognormal section.
distribution
AF = Availability Factor
EAF = Equivalent Availability Factor
MTTF = Mean Time to Failure 2. Development Method
MTTR = Mean Time to Repair
RPN = Risk priority number The first step is the elaboration of a thermal power plant
S = Severity
O = Occurrence
functional tree that describes the functional relationship
D = Detection between the subsystems of equipment to the relationship
MC = Maintenance Cost between – components. The structure of the system is
RI = Risk Index described in the function tree from top to bottom. The top
Cf = Fixed cost of failure (cost of spare parts)
level describes the main function of system analysed. The
DT = Down Time
Cv = Variable cost per hour of down time bottom level describes the components of the system.
PLC = Production Loss Cost Between of them there are several equipment. All the
PL = Production loss in Mega Watt hour components forming the function of the level above and
SP = Selling price of generated electricity subsequently formed the main functions of the system.
FDT = Failure Defense Task
MPI = Maintenance Prioritization Index
Event of components will affect the above system-level
POH = Sum of all hours experienced during Planned Outage events and subsequent events on top system. This
(PO) + Planed outage Extension (PE) of any Planned functional tree will be the basis for further process analysis
Outage (PO) on the method developed.
MOH = Sum of all hours experienced during Maintenance
Outage (MO) + Maintenance Outage Extension of any
The next stage is to perform FMEA analysis for each
Maintenance Outage (MO) component mentioned in the function tree. Each component
SH = Service hours, Sum of all Unit Service Hours analysed is based on modes of damage and the effects of
PH = Periods hours, Number of hours in the period being the system. FMEA can be seen from the analysis of the
reported that the unit was in the active state. most critical components. This analysis is to identify the
AH = Availability hours, Sum of all Service Hours (SH) +
Reserve Shutdown Hours (RSH) + Pumping Hours + level of criticality of each component and then determined
Synchronous Condensing Hours. the sequence of critical levels of all components. The
EFDH = Each individual Forced Derating (D1, D2, D3) is critical level is determined severity, occurrence, and
transformed into equivalent full outage hour(s). This is detection. The order of criticality based FMEA analysis
calculated by multiplying the actual duration of the
derating (hours) by the size of the reduction (MW) and denoted with ranking priority number (RPN) which is
dividing by the Net Maximum Capacity (NMC), these multiplication of severity (S), occurrence (O) and detection
equivalent hour(s)are then summed (D).
ESDH = Equivalent Scheduled Derated Hours, Each individual The level of criticality of each index is denoted by
Planned derating (PD,DP) and Maintenance Derating
(D4, DM) is transformed into equivalent full outage
numerical code of the value of 1 to 10. A value of 1
hour(s). This calculated by multiplying the actual indicates the lowest critical level and the value of 10
duration of the derating (hours) by the size of reduction indicates the highest level of criticality. Range 1 to 10
(MW) and dividing by the Net Maximum Capacity illustrates the qualitative scale defined through expert panel
(NMC). These equivalent hour(s) are then summed.
and [22] on this study. Expert panel consists of experienced
technicians, engineers and planners in maintenance
activities the thermal power plant.
Based on the severity, component failure causing
This paper presents a new methodological maintenance
unavailability of the system is categorized as critical
development to yield an integrated maintenance program.
components. This unavailability caused by the activities of
In this method, a new approach to identify the most critical
the component repair failed. The longer time to repair
components in a thermal power plant by combining the
means that the higher the level of criticality of the
concepts of RCM, RBM and CBM. The criticality is
associated with the component reliability and the risk of its component. The severity values that categorized critical
failure. The higher the criticality of the component, the components begin grades 6. The description of the severity
more technical and financial resources should be expended of the value of 6 to 10 is given in Table 1.
in the maintenance activities to keep the thermal power The occurrence includes criteria for determining the
plant availability for operation. The method results a FDT criticality of components. Components with the emergence
aiming at the overall thermal power plants availability. of failure in a short time mean lower component reliability.
FDT is defined as maintenance activities including
preventive maintenance and condition monitoring tasks.
 45

Table 1 maintenance performed. The description occurrence index

Critical severity index of the thermal power plant
of the value of 6 to 10 is given in Table 2.
Critical
Description
Index Table 2
The potential component failure causes the failure of Critical occurrence index of the thermal power plant
equipment , but does not cause damage to other Critical
Description
components but the equipment is still available, the index
failure resulted trip equipment, potential damage to 6 Failure occurred under 4 years
the environment, failure potentially not meet 7 Failure occurred under 2 years
(6) government regulations regarding environmental, 8 Failure occurred under 1 years
Major failure lead to or replacement of components which 9 Failure occurred under 6 months
fail, failure resulting plan trips , and can be recovered 10 Failure occurred under 3 months
in less than 2 days, ability ramp rate down to 45 %,
impact decrease in efficiency increase 15 %, failure
of the parent system functions in one ( 8 hours ) or In general, the criticality of components defined the
100 % redundancy possibility of failure occurring during the system operation.
Potential failure hardware components cause but no
damage other components, potential severe damage For mechanical-electrical system component damage is not
to the environment, failure to comply with the sudden. Failure was preceded by a decrease in reliability
government resulted in no, the failure resulted in the and this can be monitored. Components that failure cannot
(7) repair or replacement of components failed ,plant /
Severe equipment not be operated hearts Short Time ( 2-7 be monitored with existing technology categorized critical
days ),ability street level down- up to 55 %, impact components. Level of the monitoring capabilities the failure
efficiency 18 % reduction of component denoted by detection index. Components are
failure parent system function hearts 4 hours
redundancy or 50 %
most easily detectable failure given index value 1. Index
The potential component failure causes the failure of value 10 for components failure cannot be detected with
equipment , but does no cause damage to other existing technology. In the case of thermal power plant, the
components, the failure resulted trip equipment,
component failures can be detected. This means there is no
inflicting severe damage to the environment, failure
result does not meet government regulations, the critical component categories based on detection. All the
(8)
Very Severe
failure resulted in the repair or replacement of components can be controlled to avoid a failure during the
components which fail, plant / equipment cannot be system operation.
operated in a long time (7 days -1 month), ability
ramp rate down to 65 %, impact decrease in FMEA analysis produces a sequence component
efficiency of 21%, failure of the parent system criticality. Then the next step is to develop a failure
functions in 1 hour or redundancy of 50 % diagnostic procedure that allows rapid recovery actions in
Failure cause severe damage to the components or case of failure occurrence. That procedure determines the
other equipment, equipment failures result trip, cause
harm to the environment , hazardous material leaks root cause of the component failure aiming at directing
into the environment, failure result does not meet technicians to investigate the cause of the detected failure.
(9) government regulations, the failure resulted in the We use Fault Tree Analysis (FTA) technique for
Hazardous repair or replacement of many components, plant / diagnosing a failure. This technique is suitable for tracing
equipment cannot be operated in a long time ( 1-3
months ), ability ramp rate down to 75 %, impact the cause of failure to a complex system with many
reduction efficiency of 25 %,system malfunction components [23]. Search the cause of the failure is done by
within 30 minutes or no back up decomposing the groove cause the failure from system to
Failure cause severe damage to other equipment, the the subsystems to component level. Thus, using FTA in the
failure resulted trip a system equipment, cause harm
to the environment , hazardous material leaks into the developed method can improve maintainability by speed-up
environment, failure result does not meet government failure diagnosis.
(10) regulations, the failure resulted in the replacement of FTA generates a series of maintenance actions that
Catastrophic almost all components, plant / equipment cannot be include preventive maintenance task and condition &
operated in a long time ( > 3 months ), ability ramp
rate of up to > 75 %, system failures have a major assessment task. A set of maintenance actions aimed at
impact on reducing the efficiency , failure of the avoiding the unexpected failures and is called Failure
parent function immediately or no back-up Defence Task (FDT) in this study. This FDT also FDT term
commonly is used in the practice of maintenance of the
Unavailability is often caused by low reliability of case study. FDT can be either repair or replacement
components and should be avoided. The higher rate of actions. The repair action restores the component like a
occurrence means that the lower the reliability of new condition as replacement actions.
components and it also means that the higher the degree of Determination of the critical level components in the
criticality. Criticality components based occurrence index FMEA analysis has not considered the aspect of risk
determined by the maintenance cycle. In the case of this associated with the costs of the failure. Risk will need to be
study, the major inspection & overhaul (MI) is done every considered because the cost is an important factor in
four years. Component failure that occurred prior to the MI managing a thermal power plant. Risk of component
is categorized critical component. This means that failure damage caused substantial costs should be prioritized in
has occurred prior to the preventive maintenance. Category maintenance. This study considers the risk in determining
critical component starts from the value 6 in which the the criticality of components through risk analysis.
failure can occur before four years or the preventive
 46

There are two steps in the analysis of risk i.e. the risk periodic preventive maintenance actions, determination the
assessment and evaluation. At this step of the risk type of condition monitoring (on-line or off–line) and
assessment were calculated probability of occurrence of maintenance intervals. Determining the type of
failure and its consequences. That probability is obtained maintenance actions are based on the FDT of each
through the use of software Weibull++ (Reliasoft, 2003). component that has been formulated at the FTA analysis.
The consequence is calculated from the cost of repair The integrated maintenance program is formed to
required and the cost of lost production. improve the availability of power plant. Three aspects are
The calculation of the cost of repair using Eq. (1) in important in the maintenance of the reliability, risks and
[14], conditions have been adopted in this maintenance program.
Each of these aspects has been used in the RCM, RBM and
MC  C f  DT .C v (1) CBM. The development of this program has been to
integrate two maintenance bases, namely time and
And the calculation of the cost of lost production using Eq. condition-based maintenance.
(2) in [14], The final step of the methodology is the analysis of
reliability and availability. This analysis is done to see the
PLC  DT .PL .SP (2) effect of the implementation of an integrated maintenance
program to improve the availability of the system. The
The second step is to calculate the risk evaluation. This reliability analysis is based on the time to failure data
evaluation result the risk index which is obtained from analysis. Meanwhile, the availability analysis is based on
dividing the value of risk assessment to the level of the analysis of the repair time. The effectiveness of the
acceptable risk in [23]. The component criticality is maintenance methodology can be measured from the extent
determined by the value of the risk index. In this study, the to which the increased availability after application of the
critical component is a component that has a risk index integrated maintenance program.
above 0.8. Conversely, components that have a risk index The main steps of the methodology of maintenance
of less than 0.8 are not considered critical components. proposed are shown as Fig. 1.
Thus the critical level is not only determined by the
reliability aspects but involves aspects of risk.
Two important aspects of maintenance have been
considered to determine the criticality of components,
namely reliability and risk. The next step is to set
maintenance priority based on those aspects. Weighted
RPN is extended the definition of RPN by multiplying it
with a weight parameter, which characterize the importance
of the failure causes within the system [24]. In the same
manner, we use to calculated Maintenance Priority Index
(MPI) which to determine the maintenance priority of
components. The higher value of MPI means higher
priority component in the maintenance.
The maintenance priority is indicated by the value of
MPI that obtained by using Eq. (3).

MPI  RPN x RI (3)

MPI value becomes an important element in preparing

the maintenance program. Component with the highest MPI
value is a top priority in the maintenance. The MPI value
also used to determine condition monitoring & assessment Fig.1. The main steps of the methodological of maintenance developed
plan. Condition monitoring is most required for groups of
components that have a high value of MPI. Maintenance In the methodology, the integrated maintenance
action based on condition monitoring is necessary to program is formed of four input factors, namely the data
maintain the component with a high level of criticality is history of operation and maintenance, FDT, MPI and
not fail during the system operation. Such components condition monitoring. The mechanism of the formation of
failure reduces the availability significantly and produces the integrated maintenance program from four input factors
great risk costs. is given as Fig. 2.
The combination of MPI, FDT and condition
monitoring shows three important aspects of the
maintenance has been integrated in setting up a
maintenance program that is called the integrated
maintenance program. This program includes the type of
 47

functions. A failure in a component at the bottom of the

tree affects all subsystems above it, causing a possible
degradation in the thermal power plant operation,
represented by any reduction in the nominal power output
or even environmental degradation. The tree was according
developed according to the operation manual furnished by
the manufacturer.

3.2. Failure mode and effect analysis

The FMEA analysis was performed for each component

listed in the end of a given branch of the functional tree.
The failure modes for the components were developed
according to manufacturer’s information’s and other failure
analysis from FMEA team. The FMEA team identifies,
evaluates, and prioritizes potential failures [25].
The analysis pointed out that the most critical
components for the thermal power plant are:
Fig. 2. Flow chart of the integrated maintenance program 1. Cooling water system: Shaft of main circulating water
pump.
2. Feed water system: thrust bearing of BFP variable speed
hydraulic coupling, journal bearing of scoop tube assy,
3. Application Mechanical seal of BFP#A, shaft of scoop tube assy,
primary shaft of variable speed hydraulic coupling.
The method has been developed is applied to the 3. Boiler system: Platen super heater inlet header,
thermal power plant Unit 4 with an output capacity of 200 Finishing superheater inlet header, Primary superheater
MW, located in North-Jakarta. They have 1664 kinds of tube, Primary superheater inlet header, Finishing
mechanical and electrical components. The reliability and superheater element, Primary reheater element,
availability of the thermal power plant are simulated based Finishing reheater element, Primary reheater inlet
on a three-year failure database. header, Boiler pressurized second superheater.
4. Turbine system: Thrust pad passive side, Thrust pad
active side, Turbine journal bearing, Journal bearing
3.1. Functional tree
tilting pad, Low Pressure rotor blade, Main stop valve
stem, High pressure rotor blades.
The functional tree the power plant is presented in Fig.
5. Generator system: Rotor copper conductor, Winding
3 and was divided into thirteen subsystems. For detail of
copper, Stator winding insulation.
the functional tree of the system which is divided into
subsystems until each component, each one performing a
specific function in connection with the subsystem main

Fig. 3. The functional tree of Thermal power plant Unit 4

the subsystem system level, the fault tree of thermal power

3.3. Fault tree analysis plant is showed in Fig. 4. This fault tree can be elaborated
including lower subsystem until their components at the
The functional tree is fundamental for the understanding bottom level. The events at component level are named a
of the functional relation between system components. At basic event. These events are analyzed to plan a
 48

maintenance program. Based on the functional tree, we use

the fault-tree technique to identify the component failure of
Thermal power plant
combination of component failures that cause the failure of Failure

system or subsystem. For those components a maintenance

program can be formulated to avoid their failures. The
maintenance program consists of preventive maintenance
and condition-monitoring tasks. In the present study, those
tasks is named failure defense task (FDT). FUEL SYSTEM
Failure Equipment Failure

3.4. Failure defense task

The maintenance program of thermal power plant is

based on four or five year cycles depending on system
condition. Some annually based basic preventive tasks are GAS FUEL FUEL OIL
SYSTEM Failure SYSTEM Failure
performed. In the middle of the cycle a more complex 03 04 TURBINE ailure
MONITORING
SYSTEM Failure
09 12
inspection is performed. After that the basic tasks are
performed annually and at the end cycle major inspection COOLING WATER CONDENSATE AIR AND FLUE TRANSFORMATOR
SYSTEM failure WATER SYSTEM GAS SYSTEM
and overhaul maintenance is performed. 02 Failure failure
failure
11
05 07
Based on the results of the FMEA and Risk analysis, the
PLANT
RCM and RBM concepts can be used to recommend FDT CONTROL
SYSTEM Fails FEED WATER
GENERATOR
PLANT Failure
ELECTRICAL
INSTALLATION
01 BOILER Failurel
SYSTEM Failure 10 Failure
to those components that have a criticality index greater 06
08
13

than 6 (severity) and/or 6 (occurrence) and/or 0.8 (risk

index). The failure of those components can cause the Fig. 4. The fault tree of the thermal power plant
unavailability of the thermal power plant. In the present
study the detectable index do not used as critical index 3.6. Maintenance Priority Index
because all the component failures can be detected by
technology available. MPI value calculation applied to all components of the
Generally, the power plant has a complex monitoring bottom of the tree function. This value is obtained from the
system based on temperature, pressure, vibration, chemical calculation by using Eq. (3). Once identified based on the
concentrates, acoustic, gas, current measurements and criticality index, from 1664 components in the system there
analysers. That system is used to monitor and assess the are 858 types of critical components. This criticality is
real-time condition of the critical components of the based on one of the index value of the index the severity,
thermal power plant allowing the use of FDT to improve occurrence and risk. In this paper all critical components
the system availability. Those data can be used to define are not shown. The highest value is on the blade of the
the trend and pattern in the equipment condition and a limit turbine system components. This component is certainly
value must be selected as a potential failure indicator. not suffered failure during system operation. Therefore, the
That value allows identification of the alert level, components which have high MPI values should get
providing to schedule FDT before failure occurs. The priority maintenance. Maintenance priorities presented
analysis is used for the implementation of the FDT. within the scope of activities of the FDT.
Moreover, most of the critical components of the thermal
power plant can be assisted with FDT. 3.7. Reliability and Availability Analysis

3.5. Risk analysis Reliability can be defined as the probability that a

system the probability that a system will perform properly
Risk analysis is done to identify the level of risk arising for a specified period of time under a given set of operating
from a component failure to the system. There are two conditions. For the thermal power plant the failure is any
consequences of the costs incurred as a result of failure of component failure that causes incapacity of generating the
the components i.e. the maintenance cost and production nominal power output.
loss cost. Those costs are calculated respectively using Eq. The reliability analysis is performed on the power plant.
(1) and (2). In the case of thermal power plant Unit 4, PLC It is based on the time to failure data analysis. Probably the
calculation per hour is around US$ 11,840. Results single most used parameter to characterize reliability is the
calculated from both equations for each component are MTTF. It is just the expected or mean value of the failure
added. time, expressed as in Eq. (4):
Then, the total of those costs is divided by the value of
acceptable risk to obtain risk index. The acceptable risk 

was determined based on yearly maintenance expenditure MTTF   R(t )dt (4)
0

Unit 4 (US$ 2,080,000).

 49

Random failures constitute the most widely used model for maintainability modelling are   7 .9652 and
for describing reliability phenomena. They are defined by   0.9333 .
the assumption that the rate of failure of a system is The thermal power plant is an electrical-mechanical
independent of its age and other characteristics of its complex system. It means there are two categories of the
operating history. The Weibull probability distribution is component, i.e. electrical and mechanical component. In
one of the most widely used distributions in reliability mechanical component, the causes of failure are likely to be
calculations involving time related failures. Through the quite obvious. The primary time entailed in the repair is
appropriate choice of parameters a variety of failure rate then determined by how much time is required to extract
behaviours can be modelled, including constant failure rate, the damage parts and install the new component. In
in addition to failure rates modelling both wear-in and contrast, If an electronic device fails, maintenance
wear-out phenomena. personnel may spend most of the repair procedure time in
The thermal power plant is modelled as one block. For diagnosing the problem, for it may take considerable effort
that block the reliability and maintainability distribution are to understand the nature of the failure well enough to locate
estimated based on failure data recorded. The two- the part that is the cause. Conversely, it may be a rather
parameter Weibull distribution, typically used to model straightforward procedure to replace the faulty component
wear-out that failure rate increases. This distribution once it has been located.
represented in Eq. (5): Once the reliability and maintainability parameters are

t  calculated the system availability can be estimated. The
  

R (t )  e   (5) availability is controlled by two parameters. Firstly, MTTF
which is a measure of how long, on average, the thermal
The distribution parameters are estimated through the power plant will perform as specified before an unplanned
use of parametric estimation methods that fit the failure will occur, being associated with equipment
distribution to the ‘time to failure’ data. There are reliability. Secondly, MTTR which is a measure of how
procedures for estimating the Weibull distribution long, on average, it will take to bring the equipment back to
parameters from data, using what is known as the normal serviceability when it does fail. Although reliability
maximum likelihood estimation method. For the thermal can be at least estimated during the thermal power plant
power plant reliability analysis the software Weibull ++ design stages, its availability is strongly influenced by the
version 6 (Reliasoft,2003 ) was used for parameter uncertainties in the repair time. Those uncertainties are
estimation. influenced by many factors such as the ability to diagnose
The results of parameter values estimation for the the cause of failure or the availability of equipment and
thermal power plant are   5,3533 and   15211 . The skilled personnel to carry out the repair procedures. In the
case of a thermal power plant, the same equipment model
thermal power plant presented 39 failures that caused
can present different availability in different sites due to
equipment unavailability in the analysis period. Several of
difference in the skill of personnel responsible for
those 22 failures occurred in the first two operational years.
maintenance.
Most of them were related to leakage in line drain gland
Considering the thermal power plant operating one year
turbine and economizer. In the last three years there were 1
or 8760 hours and using parameters of the reliability and
leakage in line drain gland and 2 leakages in glad seal
maintainability probability distribution is found the
steam. The failure root-cause was improperly monitoring
availability for the thermal power plant given in Table 4.
feed water, pressure and temperature.
Availability is an index dependent on reliability and
The failures that may affect thermal power plant
maintainability. The availability will increase if the
availability were associated with components listed at the
reliability increases and/or the maintainability increase.
bottom of functional tree branches presented in Fig. 2 and
In Table 4, the average availability of thermal power
were considered as critical components in the FMEA
plant within 2008 to 2012 is 76.95 %. The thermal power
analysis.
plant analysed in the present study has lower availability
The other aspect from reliability related to increasing
than the value presented in the NERC. The availability can
availability is maintainability that is defined the probability
be increased through FDT, These maintenance tasks will
of an equipment will be repaired in a given period of time.
reduce the probability of failure during the thermal power
Typically, the describing of the probability distribution was
plant operating.
used lognormal distribution to model the time to repair
distribution of complex systems. The maintainability can be
expressed according to Eq. (6) in [21]:
3.8. Availability Improvement
 ln t   
M (t )     (6) The maintenance policy performed in the present study is
   called FDT that consists of the preventive and predictive
maintenance tasks. FDT is the results of the availability
Based on the time to repair database for the thermal analysis listed in Table 4. Management can implemented
power plant using the software Weibull ++ Version 6 our recommendation of maintenance improvement to
(Reliasoft, 2003). The lognormal distribution parameters increasing the thermal power plant availability.
 50

Implementation of the FDT improves two aspects of the Weibull distribution parameter values after the
maintenance performance i.e. reliability and improvement are   7 .3307 and   17721 . Fig. 7 shows
maintainability. The chance of those performances can be the graphic of the one’s maintainability in the same period.
seen in Fig. 6 and 7. In this graphic, the maintainability after FDT performed
higher than that before FDT performed. Those graphics
Fig. 6 shows the graphic of the thermal power plant show FDT has increased the reliability and maintainability
reliability in two periods i.e. 2011 to 2012 and 2013 to of the thermal power plant and hence the availability of the
2014. The first period describes the condition which thermal power plant increases which is shown in Table 5.
maintenance tasks are not improved yet. The second cycle The lognormal distribution parameter values for
describes the condition which maintenance tasks have maintainability after improvement are   5 .6239 and
implemented the FDT. The comparison of those provided
  2.1182 .
the reliability improvement after FDT performed. The
Table 4
The availability before FDT performed
SH POH MOH FOH RSH AH PH EFDH ESDH
Year EAF
(HOUR)
2008 7056.53 0 1261.62 37.15 428.70 7485.23 8784 397.88 635.04 73.46%
2009 6403.42 1413.80 459.88 455.73 27.17 6430.59 8760 152.99 421.58 66.85%
2010 501.47 1066.00 1155.62 198.62 638.30 6339.77 8760 135.28 184.73 68.72%
2011 1105.03 1416.00 45.50 0 6193.47 7298.50 8760 4.11 0 83.27%
2012 6649.84 0 325.15 335.38 1473.62 8123.46 8784 2.49 0 92.45%

Average 76.95%

Table 5 shows that the average level of availability

achieved 94.59 %. This value is higher than the average
level of availability is achieved before applied FDT. This
increasing can be seen clearly from the FOH which shows
the frequency of occurrence of unexpected failure. The
condition prior to the FDT, the total time of repair due to
damage to the components is 705.53 hours and this value
decreased after FDT be 33.82 hours. This means FDT
applied can reduce unplanned maintenance activities and
further increase the availability of the system.

4. Conclusion

Fig.6. The reliability comparison between before and after FDT In this paper we proposed a new methodological
maintenance development for a complex system with high
production loss such as thermal power plant with high
capacity output. Such a system, availability will become the
most important factor that will be concerned by top
management. The method developed is used to increase the
system overall availability. This method has considered
maintenance based of reliability, risk and condition in
which these three maintenance basis have not been
considered in an integrated manner. The integration of
these three maintenance basis is expected to increase the
availability of better than previous methods.
The proposed method has a new rule of the determining
critical components based on FMEA, FTA and risk
analysis. Analysis performed yields MPI for action and
maintenance planning, is called FDT. MPI also is used to
prepare the procedure of condition monitoring for critical
components. Integration between FDT and condition
Fig. 7. The availability comparison between before and after
 51

monitoring yields an integrated maintenance program that the proposed maintenance program found that the average
increases the availability of system. availability of the system within a period of two years
The method developed has been applied to the case of a increased from 76.95% to 94.59%. Thus, this method can
thermal power plant with an output capacity of 200MW. improve the system performance especially the availability
Availability analysis is done to determine the effectiveness of the system.
of an integrated maintenance program. Application of this
method follows the maintenance cycle that has been run on
the system in the present. From the comparison of
performance between before and after implementation of
Table 5
The availability after FDT performed
SH POH MOH FOH RSH AH PH EFDH ESDH
Year EAF
(HOUR)
2013 7323.94 912 24.00 0 500.07 7824.00 8760 0 0 89.32%
2014 6320.52 0 0 9.82 2429.67 8750.18 8760 1.72 0 99.87%
Average 94.59%
Source: The operation and maintenance data of the thermal power plant

[10] N.S. Bhangu, R. Singh, G.L. Pahuja, Reliability centered

maintenance in a thermal power plant: a case study, Int. J.
Productivity and Quality Management, Vol. 7, No. 2, 2011.
Acknowledgment [11] Dag Eirik NORDGÅRD, Eivind SOLVANG, Geir SOLUM,
Developing and implementing a risk base maintenance strategy for
The authors would like to thanks the BOD of PT PJB for distribution companies, 18th International Conference on Electricity
Distribution, CIRED, Turin, 2005.
support this research. We also like to give our appreciation [12] Faisal I. Khan, Mahmoud R. Haddara, Risk-based maintenance of
to Engineering Dept. of PT PJB UPHB for their assistance. ethylene oxide production facilities, Journal of Hazardous Materials,
108(3): 147-59, 2004.
[13] Faisal Khan, Mahmoud Haddara, Mohamed Khalifa, Risk-Based
Inspection and Maintenance (RBIM) of Power Plants, Springer
References Series in Reliability Engineering, Springer-Verlag London Limited,
2012.
[1] American National Standards Institute, 2006, IEEE [14] Loganathan Krishnasamy, Faisal Khan, Mahmoud Haddara,
Std.762tmDefinitions for Use in Reporting Electric Generating Unit Development of a risk-based maintenance (RBM) strategy for a
Reliability, Availability, and Productivity, Availability and power-generating plant, Journal of Loss Prevention in the Process
Productivity, IEEE, New York. Industries 18, p. 69–81, 2005.
[2] Faisal I. Khan, Mahmoud M. Haddara, Risk-based maintenance [15] Yatomi Masataka, Takahashi Jun, Baba Hidenari, Ohinata
(RBM): a quantitative approach for maintenance/inspection Toshiharu, Fuji Akio, Application of Risk-Based Maintenance on
scheduling and planning, Journal of Loss Prevention in the Process Materials Handling Systems, IHI Engineering Review, Vol. 37, No.
Industries 16, 561–573, 2003. 2, June 2004.
[3] Amik Garg and S.G. Deshmukh, Application and case studies [16] Christos Emmanouilidis, Luca Fumagalli, ErkkiJantunen,
maintenance management: literature review and directions, Journal PetrosPistofidis, Marco Macchi, Marco Garetti, Condition
of Quality in Maintenance Engineering, Vol. 12 No. 3, pp.205-238, monitoring based on incremental learning and domain ontology for
2006. condition-based maintenance, Proceedings of APMS 2010
[4] Carazas F.G.J, Souza G.F.M, Availability Analysis of Gas Turbines International Conference on Advances in Production Management
Used in Power Plants, Int. J. of Thermodynamics, ISSN 1301-9724, Systems, Cernobbio, Como, Italy, 2010.
Vol. 12 (No. 1), pp. 28-37, March 2009. [17] Rosmaini Ahmad, Shahrul Kamaruddin, An overview of time-based
[5] Carazas F.G.J, Souza G.F.M, Availability analysis of heat recovery and condition-based maintenance in industrial application, Elseiver,
steam generators used in thermal power plants, Energy, 36, 3855- computer & industrial engineering, 2012.
3870, 2011. [18] Stephan, Laird, Condition Based Maintenance on turbogenerator :
[6] Dacheng Li, Jinji Gao, Study and application of Reliability-centered What Makes It Real?, ALSTOM Power, 2003.
Maintenance considering Radical Maintenance, Elsevier, Journal of [19] Y.G. Li , P. Nilkitsaranont, Gas turbine performance prognostic for
Loss Prevention in the Process Industries, 2010. condition-based maintenance, Elseiver, computer & industrial
[7] François Besnard, Katharina Fischer, LinaBertling, Reliability engineering, 2009.
Centred Asset Maintenance-A step towards enhanced reliability, [20] Gang Niu, Bo-Suk Yang, Michael Pecht, Development of an
availability and profitability of wind power plants, Division of optimized condition-based maintenance system by data fusion and
Electric Power Engineering, Department of Energy and reliability-centered maintenance, Elsevier, reliability engineering and
Environment, at Chalmers University of Technology in Gothenburg, system safety, 2010.
Sweden , 2010. [21] J.T. Selvik, T. Aven, A framework for Reliability and Risk Centered
[8] Fredrik Backlund, Vattenfall, Managing the introduction of RCM Maintenance, Elsevier, Pecht, Development of an optimized
Experiences from a Swedish Hydropower Company, 2003. condition-based maintenance system by data fusion and reliability-
[9] Katharina Fischer, Francois Besnard, LinaBertling, Reliability- centered maintenance, Elsevier, Reliability engineering and system
Centered Maintenance for Wind Turbines Based on Statistical safety, 2010.
Analysis and Practical Experience, IEEE Transactions on Energy [22] Franceschini F, Galetto M. A new approach for evaluation of risk
Conversion, Vol. 27, No. 1, 2012. priorities of failure modes in FMEA. Int. J. Prod. Res Vol 39, No.
32. 2001.
 52

[23] Carazas F.G.J, Souza G.F.M, Risk-based decision making method

for maintenance policy selection of thermal power plant equipment,
Energy 35, pp.964-975, 2010.
[24] Ningcong Xiao, Hong-Zhong Huang, Yanfeng Li, Liping He,
Tongdan Jin, Multiple failure modes analysis and weighted risk
priority number evaluation in FMEA, Engineering Failure Analysis
18, 1162–1170, 2011.
[25] Kmenta S. and Ishii K., Scenario-based FMEA: A life cycle cost
perspective, ASME design engineering technical conference, 2000