Risk Based Maintenance
Risk Based Maintenance
Risk Based Maintenance
1. OBJECTIVES..................................................................................................................3
2. SYSTEM DESCRIPTION................................................................................................3
3. RISK-BASED MAINTENANCE.......................................................................................4
4. CALCULATION OF RISK...............................................................................................6
5. CRITICAL ANALYSIS.....................................................................................................7
5.1 HIERARCHICAL ANALYSIS OF PROCESSES.......................................................8
5.2 STRUCTURING THE HIERARCHY.........................................................................9
5.3 WEIGHTS OBTAINED IN THE HIERARCHICAL ANALYSIS OF PROCESSES....11
6. EQUIPMENT ANALYSED..............................................................................................12
7. OBTAINING AND PROCESSING DATA........................................................................13
8. CALCULATION OF THE CONSEQUENCES................................................................13
9. CALCULATION OF FAILURE RATES...........................................................................13
10. CALCULATION OF THE RISK OF THE EQUIPMENT EVALUATED............................16
11. CONCLUSIONS..............................................................................................................22
12. SUGGESTIONS FOR FUTURE STUDIES.....................................................................22
13. REFERENCES................................................................................................................24
1. OBJETIVES
I. To present the basis for the implementation of the methodology called Risk-Based
Maintenance (Risk Based Maintenance).
2. SYSTEM DESCRIPTION
The concepts to be described will be applied to the hypothetical pumping system of an oil
pipeline, whose equipment is installed in the storage terminal of petroleum products, consisting
of five main pumps and 4 auxiliary pumps.
Table 1 shows the arrangements of pumps used to ship each of the products.
Table 2 presents a summary of the monthly sales of the Terminal with the operation of the
pipeline.
$3.000.000
$2.500.000
$2.000.000
$1.500.000
$1.000.000
$500.000
$-
jul/18 set/18 out/18 dez/18 fev/19 mar/19mai/19 jul/19 ago/19
3. RISK-BASED MAINTENANCE
Risk assessment integrates reliability with safety, facility integrity, company financial issues,
and environmental issues and can therefore be used as a decision tool for preventive
maintenance planning.
Maintenance planning based on risk analysis minimizes the probability of failure of the
system and its consequences (related to safety, economic and environmental). Helps
management in making correct decisions about maintenance investment or related areas. This
in turn will result in better use of assets and capital.
Risk analysis is a technique to identify, characterize, quantify, and evaluate loss from an
unwanted event. The risk analysis approach integrates probability and consequence analysis at
various stages of the analysis and attempts to answer the following questions:
The risk assessment may be quantitative or qualitative. The output of a quantitative risk
assessment will normally be a number, such as cost impact ($) per unit of time. The number can
be used to prioritize a series of items that have been assessed for risk.
Quantitative risk assessment requires a large amount of data for both probability assessment
and consequence assessment. The fault tree or decision tree is often used to determine the
probability that a certain sequence of events will result in a certain consequence.
Qualitative risk assessment is less rigorous and results are often presented in the form of a
simple risk matrix, where one axis of the matrix represents the probability and the other
represents the consequences. If a value is given to each of the probabilities and consequences,
a relative value for the risk can be calculated.
It is important to recognize that qualitative risk value is a relative number that has little
meaning outside the matrix structure. Within the matrix structure, it provides a natural
prioritization of the items evaluated using the matrix. However, as these risk values are
subjective, prioritizations based on these values are always debatable.
The proposed risk-based maintenance (RBM) strategy aims to reduce the overall risk of
failure of operational facilities. In high- and medium-risk areas, a focused maintenance effort is
required, while in low-risk areas, effort can be minimized to reduce the overall scope of work and
the cost of the maintenance program in a structured and justifiable manner.
Quantitative risk value is used to prioritize inspection and maintenance activities. RBM
suggests a set of recommendations on how many preventive tasks (including type, means and
time) should be performed. The implementation of RBM will reduce the probability of an
unexpected failure.
Phase 2: Classification of teams according to the importance of their function. Once the
objectives of maintenance are defined, the next step is to know the critical areas for achieving
these objectives. The implementation of Phase 1 will establish the criteria that allow the hierarchy
of assets and areas of the company. The result of this phase will be a ranking of equipment or
areas that will allow establishing the best strategy to achieve the previously defined objectives.
Phase 3: Analysis of weak points in high impact equipment. Once the equipment is listed,
the first task will be to attack assets whose failures may have worse consequences for the
process.
Phase 4: Definition of maintenance plans and necessary resources. The maintenance policy
will be established taking into account the criticality of the equipment and the resources needed
for this policy will be defined.
Phase 5: Maintenance programming and resource allocation optimization. Once the
maintenance plans are prepared, the next step is to program them over time and minimize
maintenance costs.
Phase 6: Evaluation and control of maintenance execution. At this point, once the
maintenance policy has been defined, it is necessary to assess its effectiveness and to establish
the possible improvements to be implemented.
Phase 7: Life cycle analysis and possible equipment renewal. Once established the
maintenance strategy and policy, it is necessary to analyze its impact on equipment, as well as
its possible renovation, taking into account economic criteria and impact on reliability.
4. RISK CALCULATION
Risk is a probabilistic term, defined as "likely expenses or losses due to the probable
occurrence of an unwanted event or failure". In this simple but powerful concept coexists the
possibility that an event or statement becomes reality or is satisfied, with the respective
consequences, mathematically, the risk associated with a decision or event is given by the
universal expression:
Where:
R(t): Risk
f(t): Frequency
C$(t): Consequences
The failure frequency will be evaluated by means of the failure rate of the Weibull distribution,
whose parameters will be calculated from the statistical analyses of the data obtained in the SAP
system, considering only the events that made the equipment unavailable; corrective, preventive
or periodic inspections.
Regarding the consequences, several criteria considered critical for the billing of the
installation will be considered, these criteria were selected considering several bibliographic
references on the subject.
Where;
The weights will be determined by the technique known as Hierarchical Process Analysis
(AHP), using the software Super Decisions.
This preliminary risk analysis is known as screening (screening), the final result of the
analysis will be a risk map of the analyzed equipment classified by its risk.
Thus, the equipment classified with higher risk indices deserve a more detailed analysis,
through an FMEA (Analysis of Modes and Effects of Failure), identifying the main failure modes
in order to reduce the failure frequencies, or even make a decision to purchase new equipment.
5. CRITICALITY ANALYSIS
In Criticality Analysis (CA), relative bands are established to represent the probabilities
and/or frequencies of occurrence of events and their consequences. Criticality analysis is based
on risk theory. When evaluating a given event, it is necessary to quantify the frequency of
occurrence and the consequences of each of the scenarios that lead to the event under study.
The risk behaves as a balance that allows to weigh the influence of several alternatives in terms
of impact and probability, guiding the analyst in the decision making.
In this decision-making process, risk is used as a tool for optimizing asset service plans,
directing greater resources and efforts to high-risk assets and reducing effort and resources to
those at low risk. risk, which generally allows a justified expenditure of resources directed at
maintenance items.
According to ISO 31000 Risk Management System (SGR), the risk management process
should be an integral part of physical asset management and should meet the requirements set
out in the standard for its identification, risk analysis, assessment and treatment in its operational
context.
The purpose of this standard is to describe an efficient and rational work process that results
in an optimized maintenance program based on a risk analysis as well as a cost-risk-benefit
analysis.
Maintenance management is illustrated as a work process where products are produced with
low SMS risks and high production performance. The basic model proposed as best industry
practice is shown in figure 4.
The existing methods of criticality analysis are grouped by several authors in different ways,
and according to their type can be considered qualitative, semi-quantitative and quantitative,
these methods are based on risk assessment and aim to identify critical equipment in a
production system.
The AHP is a method based on the evaluation of different criteria that allow to organize a
specific structure and whose final objective is to optimize managerial decision-making, reducing
complex decisions to a series of comparisons that allow the classification of different aspects
(criteria) assessed.
The use of AHP (Hierarchical Process Analysis) has been growing day by day in the most
diverse areas, and in the management of asset maintenance helping in decision-making at the
time of directing financial, human and technological resources, to develop more efficient
maintenance plans, for which the level of criticality of the different systems/equipment
participating in the production process needs to be taken into account.
This situation is not easy to resolve as there is a wide variety of factors involved that
generate great uncertainty in the hierarchical process of the systems, so the application of the
AHP technique can help to specifically identify the level of criticality, allowing the optimization of
the effective distribution of maintenance resources according to the degree of importance of each
system in the production process.
The AHP helps analysts organize critical aspects of a problem into a hierarchical structure
similar to a family tree, by reducing complex decisions to a series of simple comparisons and
classifications and synthesising the results in a logical and clear way for the choices made.
When designing the hierarchical tree of the AHP, the goal is to develop a general structure
that meets the needs of analysts to solve the problem of selecting the best maintenance policy.
The AHP starts by dividing a complex problem and with multiple criteria into a hierarchy in which
each level comprises some manageable elements that are then divided into another set of
elements. The hierarchical analysis process proposed here is based on the methodology
developed by Saaty, perform the following steps:
I. Define the decision criteria in the form of hierarchical objectives. The classification is
structured on different levels: starting at the top with the definition of the main
objective of the classification process, then intermediate levels are defined (criteria
and sub-criteria to be evaluated) and, finally, at the lower level, the alternatives to be
compared are described.
II. Evaluate the different criteria, sub-criteria and alternatives according to their
corresponding importance at each level. Qualitative and quantitative criteria can be
compared using informal judgments to obtain weights and priorities. For qualitative
criteria, the AHP technique uses simple comparisons (par to par) to determine
weights and evaluate them. In this way, the analyst can focus on only two criteria at
the same time. In fact, the AHP technique is based on the assumption that the analyst
(decision maker) can more easily choose a comparison value than an absolute value.
Verbal judgments are transferred to a scoring scale (see Table 8). Subsequently, in
a judgment matrix, a priority vector is calculated and used to weigh (compare) the
elements of the matrix. Saaty (1980, 1990) demonstrated mathematically that the
normalized eigenvector calculated from the matrix M of the criteria is the best
approximation for evaluation of the analyzed criteria.
𝐶𝐼 = (4)
Where; λmax is the maximum eigenvalue of the normalized matrix. Finally, the relation of
consistency relation between the consistency index (IC) and the random consistency index (AI),
equation (5) is calculated. The random consistency index is defined as the average random
consistency index obtained by simulating 100000 randomly generated reciprocal matrices using
the Saaty scale, table 9.
𝑅𝐶 = (5)
IV. As a final result of the analysis, what we have is a classification of the weight of the
criteria adopted. For each criteria, the classification level is calculated on a scale
between 0.0000 - 1,000.
It is possible to perform all the analysis using an EXCEL spreadsheet, as the selection of the
importance of the criteria adopted is qualitative, used a free software called Superdecisions which
has several tools, mainly the one that performs a sensitivity analysis among the various criteria,
thus allowing a much more consistent analysis.
Figure 5 presents the model applied for the Hierarchical Process Analysis using the Super
Decisions software, constructed as follows:
Figure 5 - Model built in the Super Decisions program to obtain the weights.
For the present case, considering the analyzed equipment, the failure data obtained and
treated, the costs of maintenance, safety, environment and production, the weights obtained were
as follows:
In this way, replacing the weights in equation (3) that assesses the consequences for each
equipment, we have that:
𝐹
𝐶$ = [0,2328. 𝐹 + 0,5173. 𝐹 .0,01]. 0,065 + 0,0360. 𝐹 + 0,0671. 𝐹 + 0,1468.
𝐴𝑣𝑎𝑖𝑙
Where the values of F1, F2, F3, F4, F5 depend on each equipment.
The calculation of the availability of the "system" depends on the availability of each
equipment and the combination with the other equipment that constitutes the so-called "system".
As previously mentioned, the PIPELINE transfer system has 5 main pumps and 2 auxiliary
pumps, and it operates with 4 main pumps in series with 1 on standby and with 1 auxiliary pump
with the other on standby.
Thus, the system availability calculation must consider the following possible combinations
of the main pumps, table 11 shows the calculated availability value for each combination.
6. EQUIPMENT ANALYZED
The data on the Terminal's auxiliary and main pumps are shown in Table 12.
Table 13 indicates the pump arrangements used to ship each product from the Terminal.
The data of the interventions carried out by the TERMINAL maintenance management in the
auxiliary and main pumps of the PIPELINE were obtained from the records in SAP.
It was assumed that the intervention times of the analyzed equipment follow a Weibull
distribution, and the parameters were obtained according to a program written in Python,
automatically and without the intervention of the analyst.
The costs of maintaining the equipment (corrective, preventive and predictive) were obtained
with the precious help of the TERMINAL maintenance management.
The data referring to the reliability metrics (MTTR and MTTF) for each type of maintenance
performed were obtained in SAP, being calculated in an EXCEL spreadsheet.
8. CALCULATION OF CONSEQUENCES
Some values were assumed for the calculation of the consequences, described below:
• Estimated purchase value of a new main pump with motor US $ 178.253,12
• Purchase value of a new auxiliary pump with motor for light US $ 115.864,53
• Purchase value of a new auxiliary pump with motor for LPG US $ 133.689,84
The reduced cost of cleaning and disposal of leaks is due to the small volume leaked, as
soon as discovered, the management provides immediate correction of the problem. The
estimated value of a human life was adopted due to the various studies of RBI and Quantitative
Risk Analysis in which the value of 1 million dollars was assumed.
The value of 6.5% for the probability of immediate ignition and 1% for the probability of fatality
in the event of a puddle fire were adopted according to several references and studies already
carried out.
As mentioned, the distribution parameters of each pump were obtained with the aid of the
Python script and are shown in table 14, as well as the values calculated for the MTTF given by
equation (6) and failure rates equation (7).
𝜆(𝑡) = . (7)
𝐹 (𝑡) = 1 − 𝑒 (8)
𝑅 (𝑡) = 𝑒 (9)
Where.
The failure rates shown in figure 8 represent very well the behavior of the pumps with respect
to the number of failures, which impacts the final value of the risk (figure 9).
The risk calculation was obtained by the product of the consequence of the failure by its
frequency, here replaced by the failure rate, which is still a frequency (in the case of failures /
time unit).
The results of the consequences, failure rates and calculated risk are shown in table 15.
Figure 8 presents the result of the evaluation comparing the risk of the selected equipment,
as expected, the equipment with the highest failure rates has the highest risk values (B-3403A
and B-3403E).
Figure 9 represents the final risk assessment for the main light pump B-3403A, locating the
risk in a matrix similar to the matrix used by the Risk Based Inspection methodology according
to API RBI 581.
In addition to the risk criticality classification, figure 9 presents another result of the
application of the methodology in relation to the acceptability criteria and deadline for taking some
action that takes the risk value to a lower level, as well as suggests a plan of action. action to
achieve this purpose, this assessment was carried out using an EXCEL spreadsheet.
Figure 17 - Risk criticality rating of the B-4001B Booster Pump for LPG
11. CONCLUSIONS
The application of the multicriteria tool allowed a uniform analysis of the equipment between
the criticality classes in a manner consistent with their performance according to the assumed
criteria.
Criteria such as safety risk, impact on the environment, production, costs of maintenance
and acquisition of equipment and unavailability were fundamental in the application of the
proposed methodology. In this sense, the multicriteria model allowed a joint assessment of all
the criteria and the relationships between them.
Thus, it is considered that the objectives of the work were achieved, with the development
of a structured model for analysis of criticality using the multicriteria methodology. With the
support of the work developed, it is possible to apply the model in other systems in which it is
necessary to assess the criticality of their equipment, with the possibility of adapting the steps
and methodology used due to the nature of the equipment or process which is to be evaluated.
The methodology proved to be totally applicable for the classification of equipment according
to its criticality using the AHP method to obtain the weights of the criteria chosen to calculate the
consequence of a failure or event that results in the unavailability of the equipment.
In addition, the criteria used for the criticality analysis were considered adequate and
sufficient to classify the 9 pumps of the PIPELINE system installed in the Terminal terminal, and
the results of the equipment classification can be used as basic guidelines to decide the strategy
maintenance.
It should be noted that multicriteria methods do not present an optimal solution to the
problem, but a coherent solution through the balance of the decision-making process and the
existing relationships between the alternatives, criteria and preferences of the decision-makers.
The proposed application of the methodology presented here can be improved as it begins
to be used. One of the points to be included is a more refined study of the criteria and which ones
have greater representativeness in the criticality of industrial equipment, which should be carried
out in conjunction with all the actors involved, operation, maintenance, SMS, logistics and
programming.
In order to systematize and optimize the analysis process, software can be developed that
integrates all stages of the model, from the construction of the evaluation matrix, modeling
preferences and application of multicriteria methods, since at work these steps were performed
individually and with different systems (EXCEL and Super Decisions).
This program would be a script developed in Python and that integrates all these concepts,
in addition to searching for data in SAP, and performing all statistical treatment to obtain the
various parameters and perform all the calculations presented in this work since the language
has all necessary modules.
The model can also be integrated with maintenance methodologies such as the RCM, during
equipment classification stages and failure analysis, and the Total Productive Maintenance
(MPT) philosophy associated with the planned maintenance pillar, in order to optimize the
management of actions maintenance and improve the technical vision of the team on all assets
of a facility.
The work can be extended to explore other possibilities, integrating the methodology based
on the AHP with the fuzzy logic based on rules or method of neural networks.
Using the Machine Learning algorithms, it will be possible to extract the structure of the AHP
method and the respective weights for each of the criteria chosen in a fully automatic way, without
using subjectivity.
As seen, AHP is widely used as a decision model, due to the way in which it deals with
multiple criteria, due to its easy understanding and operation with qualitative and quantitative
data, however, regardless of the advantages of this method, it reflects the form of human thinking,
and the model of human evaluation is uncertain and imprecise.
For example, when assessing the relative importance of different criteria, experts can show
insecurity in their assessments, as the data and information they have about the variables is
vague and ambiguous.
Despite the fact that Saaty's discrete scale has the advantages of simplicity and ease of use,
it does not take into account the uncertainty associated with the human perception model.
This aspect is exactly what makes it interesting to associate AHP with the theory of Fuzzy
logic known in the literature as the Fuzzy Analytical Hierarchy Process (FAHP) or Extended
Hierarchical Analytical Process with Fuzzy Logic.
In this way, it is possible to give a greater mathematical formalism to the approach of the
ambiguity and uncertainty of the variables used in the AHP method, since the design of the FAHP
allows the use of fuzzy numbers on the fundamental scale to form expressions closer to natural
language, that will be used in the arrangement of the judgment matrix and peer comparison, in
order to reduce the inaccuracy inherent in the experts' judgments.
Therefore, in FAHP, the linguistic labels used are conveniently chosen to assess the relative
importance of attributes, factors, conditions and / or criteria compared to others at the same level
of hierarchy. In other words, it is enough for the specialist to make an assessment of the
importance of the element verbalized in qualitative terms, and then the decision maker can use
a scale, previously defined, to obtain the fuzzy numerical values that correspond to his
assessment.
Thus, the fuzzy variant is that if an expert considered, for example, that the variable V1 is
more important than the variable V2 in a ratio of 3 to 1 (3/1), now this judgment smoothes its
value to that of a linguistic expression like "about 3 to 1", or between 2/1, or 4/1, and so on,
expressions that will be represented by fuzzy triangular numbers.
From then on, the process methodology is similar to the traditional method, except for the
mathematical tools used, which must take into account that we will now be acting on triangular
fuzzy numbers. Thus, once the valuation matrix composed of triangular fuzzy numbers has been
created, there is a method of geometric mean that weighs the fuzzy values for each option, with
the hierarchical connection established and, finally, a pertinence function for each option
develops the ranking of priorities (or smoothed degrees of importance).
13. REFERENCES
• Al-Najjar, B., & Alsyouf, I. (2003). Selecting the Most Efficient Maintenance Approach Using
Fuzzy Multiple Criteria Decision Making. International Journal of Production Economics,
84(1), 85-100.
• Arunraj, N. S., & Maiti, J. (2010). Risk-Based Maintenance Policy Selection Using AHP and
Goal Programming. Safety science, 48(2), 238-247.
• Bertolini, M., Bevilacqua, M., 2006. A Combined Goal Programming – AHP Approach to
Maintenance Selection Problem. Reliability Engineering and System Safety 91, 839–848.
• Bevilacqua, M., Braglia, M., 2000. The Analytical Hierarchy Process Applied to
Maintenance Strategy Selection. Relaibility Engineering and System Safety 70, 71–83.
• Chemweno, P., Pintelon, L., Horenbeek, A., Muchiri, P. (2015) "Development of a Risk
Assessment Selection Methodology for Asset Maintenance Decision Making: An Analytic
Network Process (ANP) Approach". Production Economics. journal homepage:
www.elsevier.com/locate/ijpe.
• Clemen, R. T., & Winkler, R. L. (1999). Combining Probability Distributions from Experts in
Risk Analysis. Risk analysis, 19(2), 187-203.
• Dey, P. K. (2010). Managing Project Risk Using Combined Analytic Hierarchy Process and
Risk Map. Applied Soft Computing, 10(4), 990-1000.
• Dey, P. K., Ogunlana, S. O., & Naksuksakul, S. (2004). Risk-Based Maintenance Model for
Offshore Oil and Gas Pipelines: a Case Study. Journal of Quality in Maintenance
Engineering, 10(3), 169-183.
• F.I. Khan, R. Sadiq, M. Haddara, Risk-based Inspection and Maintenance (RBIM) Multi-
Attribute Decision Making with Aggregate Risk Analysis, Trans. IChemE Process Saf.
Environ. Prot. 82 (B6) (2004) 398– 411.
• F.I. Khan, S.A. Abbasi, A Criterion for Developing Credible Accident Scenarios for Risk
Assessment, J. Loss Prevent. Process Ind. 15 (2002) 467–475.
• F.I. Khan, S.A. Abbasi, Accident Hazard Index: a Multi Attribute Method for Process Industry
Hazard Rating, Trans. IChemE Process Saf. Environ. Prot. 75 (4) (1997) 217–224.
• F.I. Khan, S.A. Abbasi, Risk Analysis of a Typical Chemical Industry Using QRA procedure,
J. Loss Prevent. Process Ind. 14 (2001) 43–59.
• Harnly, A.J., 1998. Risk Based Prioritization of Maintenance Repair Work. Process Safety
Progress 17, 32–38.
• HSR. (2002). Offshore Hydrocarbon Release Statistics and Analysis: . Retrieved from UK:
• ISO19902. (2008). Petroleum and Natural Gas Industries - Fied Steel Offshore Structures.
In.
• J.-R. Chen, Y.-T. Yang, A Predictive Risk Index for Safety Performance in Process
Industries, J. Loss Prevent. Process Ind. 17 (2004) 233– 242.
• Khan, F.I., Haddara, M., 2003. Risk-Based Maintenance (RBM): a Quantitative Approach
for Maintenance/Inspection Scheduling and Planning. Journal of Loss Prevention in the
Process Industries 16, 561–573.
• Khan, F.I., Haddara, M., 2004. Risk-Based Maintenance of Ethylene Oxide Production
Facilities. Journal of Hazardous Materials A108, 147–159.
• Labib, A.W., O’Connor, R.F., Williams, G.B., 1998. An Effective Maintenance System Using
the Analytic Hierarchy Process. Integrated Manufacturing Systems 9, 87– 98.
• Moore, W.J., Starr, A.G., 2006. An Intelligent Maintenance System for Continuous Cost-
Based Prioritization of Maintenance Activities. Computers in Industry 57, 595–606.
• Moubray, J., 1991. Reliability-Centered Maintenance. Industrial Press Inc., New York.
• N. S. Arunraj and J. Maiti, 2010, "Risk-Based Maintenance Policy Selection Using AHP and
Goal Programming," Saf. Sci., vol. 48, pp. 238247, 2.
• NASA, 2002. Probabilistic Risk Assessment Procedures Guide for NASA Managers and
Practitioners. Office of Safety and Mission Assurance, Washington.
• P. Guymer, G.D. Kaiser, T.C. McKelvey, Probabilistic Risk Assessment in the CPI, Chem.
Eng. Prog. (1987) 37–45.
• P.K. Dey, A Risk-Based Maintenance Model for Inspection and Maintenance of Cross-
Country Petroleum Pipeline, J. Qual. Maint. Eng. 7 (1) (2001) 25–41.
• P.K. Dey, S.O. Ogunlana, S. Naksuksakul, Risk-Based Maintenance Model for Offshore Oil
and Gas Pipelines: a Case Study, J. Qual. Maint. Eng. 10 (3) (2004) 169–183.
• Saaty, T. L. (2008). Decision Making with the Analytic Hierarchy Process. International
journal of services sciences, 1(1), 83-98.
• Saltelli, A. (2002). Sensitivity Analysis for iImportance Assessment. Risk Analysis, 22(3),
579-590.
• T. Saaty L., 2006, Decision Making with Analytic Network Process: Economic, Political,
Social and Technological Applications with Benefits, Opportunities, Costs and Risks. USA:
Springer Science+Business Media, LLC.
• Triantaphyllou, E., Kovalerchuk, B., Mann Jr., L., Knapp, G.M., 1997. Determining Most
Important Criteria in Maintenance Decision Making. Journal of Quality in Maintenance
Engineering 3, 16–28.
• Wang, L., Chu, J., & Wu, J. (2007). Selection of Optimum Maintenance Strategies Based
on a Fuzzy Analytic Hierarchy Process. International Journal of Production Economics,
107(1), 151-163.
• Y. Masataka, T. Jun, B. Hidenari, K. Toshiharu, F. Akio, Application of Risk-Rased
Maintenance on Materials Handling Systems, IHI Eng. Rev. 37 (2) (2004) 52–58.