Risk-Based Preventive Maintenance Planning using Failure Mode and Effect

Analysis (FMEA) for Marine Engine Systems

Kadir CICEK1, Hasan H.TURAN2, Y. Ilker TOPCU3, M. Nahit SEARSLAN4
1
Marine Engineering Department, Istanbul Technical University, Tuzla 34940, Istanbul, Turkey
([email protected])
2
Department of Industrial Engineering, University of Yalova, Merkez 77100, Yalova, Turkey
([email protected])
3
Department of Industrial Engineering, Istanbul Technical University, Macka 34367, Istanbul, Turkey
([email protected])
4
Department of Industrial Engineering, Istanbul Technical University, Macka 34367, Istanbul, Turkey
([email protected])

ABSTRACT
Equipments and systems on board ship, although well designed, will not remain safe or reliable if they are not maintained.
The general objective of the maintenance process is to make use of the knowledge of failures and accidents to achieve the
maximum possible safety with the lowest possible cost. In this insight, this study proposes a risk-assessment methodology
for preventive maintenance planning evaluation based upon a reliability model for marine engine systems, which allows
the use of flexible intervals between maintenance interventions. Failure Mode and Effect Analysis (FMEA) approach is
chosen as a risk assessment methodology in this paper to synthesize the potential failure modes in order to identify early
response and smoothly promote a preventive maintenance planning. As a case application fuel oil system is comparatively
assessed. In the paper, it is seen that the proposed risk-assessment methodology provides invaluable supports to construct
risk-based preventive maintenance planning for fuel oil system onboard ship.
Keywords: risk-based preventive maintenance planning, marine systems, reliability, FMEA

1. INTRODUCTION failure modes in order to identify early response and to

take appropriate actions into account. FMEA is a
Maintenance, repair, and overhaul of complex industrial proactive analysis tool, allowing engineers to define,
and marine systems have received considerable attention identify, and eliminate known and/or potential failures,
in the last decades, due to the high amounts of capital problems, errors, and so on from the system, design,
invested and the high availability rates requested. In process, and/or service [2-7].
particular, attention has been given to the determination
of closeness to the optimal maintenance policies, With in this direction, as a case application crucial
balancing the costs of preventive maintenance actions troubles in fuel oil system onboard ship are investigated
against the costs of malfunctioning (sub) systems. Once deeply to adopt an effective preventive maintenance
these policies have been determined, it is often assumed strategy for fuel oil system that allows upgrading system
that sufficient system-knowledge and capacity is reliability.
available to carry out any action needed to increase
system performance to the desired level [1]. Especially 2. LITERATURE REVIEW
to prevent the risky situations and to increase systems’
reliability onboard ships, the prestigious marine engine Although in the literature it is possible to find great
manufacturers and ship operators have continuously numbers of study related with risk-based preventive
analyzed evidences gathered from the past experiences. A maintenance planning and FMEA analysis, a few of the
great level of effort has been expanded to constitute a current works deals with the application of these tools to
preventive maintenance planning. marine engine systems more specifically fuel oil systems.
This study deviates from existing research by its unique
Risk assessment integrates reliability and therefore can case application which has not been investigated before.
be used as a decision tool in order to construct a In this section extensive literature review is presented
preventive maintenance planning. Maintenance planning about current FMEA methodologies and their application
based on risk assessment minimizes the probability of areas.
system failures and its consequences. In this paper,
failure mode and effect analysis (FMEA) is used as a risk Souze et all [8] evaluates the application impact of the
assessment technique which synthesizes the potential RCM (Reliability-Centered Maintenance) methodology
on a power generating system. Case study is also failure’s impacts so the impacts can be reduced [18]. The
conducted on hydraulic turbines via help of FMEA and process of FMEAs has three main focuses:
Fault Tree Analysis (FTA) tools. Garcia et all [9] presents The recognition and evaluation of potential failures
a data envelopment analysis approach to determine and their effects [18];
ranking indices among failure modes in which the typical The identification and prioritization of actions that
FMEA parameters are modeled as fuzzy sets. Franceshini could eliminate the potential failures and reduce
and Galetto [2] study a method for carrying out calculus their chances of occurring or reduce their risks [18];
of the risk priority of failures in FMEA. They also argue The documentation of these identification,
that their method works well without need of any evaluation and corrective activities so that system
arbitrary and artificial numerical conversion. Almannai et reliability improves over time [18].
all [10] discusses usage of both the quality function
deployment (QFD) technique and FMEA technique on The first formal FMEAs were conducted in the aerospace
manufacturing automation decision making as a decision industry in the mid-1960s and were specifically focused
support tool. Keskin and Ozkan [11] use so called on safety issues [19]. The goal with safety FMEA was,
Adaptive Resonance Theory (ART) which is a learning and remains today, to prevent safety accidents and
algorithm without consulting to evaluate risk priority incidents from occurring [19]. While engineers have
numbers in FMEA. Papodopoulos and Petersen [12] always analyzed processes and products for potential
describe a method that combines system analysis in the failures, the FMEA process standardizes the approach
design process with safety analysis, drawing on and establishes a common language [19]. FMEA is
techniques such as hazard and operability studies and mainly applied in industrial production of machinery,
FTA. Pinna et all [13] conducts very detailed FMEA motor of cars, mechanical and electronic components
study about application of remote handling transfer [20]. The introduction of FMEA onboard ship operations
systems at International Thermonuclear Experimental can be considered as a step in a new direction.
Reactor. Chang et all [14] applies grey theory to solve
encountered problems when conventional subjective An application of a FMEA follows a series of successive
linguistic assessment scores are used. They believe that steps: analysis of the process, product or system in every
this procedure is far more effective to enhance product single part, listing of identified potential failures,
reliability and process stability. Chin et all [15] presents evaluation of their frequency, severity (in terms of effects
that traditional risk priority number and fuzzy rule-based of the failure to the process and to the surroundings) and
approximate reasoning methodologies are not sufficient detection technique, global evaluation of the problem and
to corporate with cross-functional and multidisciplinary identification of the corrective actions and control plans
nature of FMEA. Due to the indicated reasons, they that could eliminate or reduce the chance of the potential
propose evidential reasoning (ER) approach and illustrate failures [3,20].
newly proposed method on a fishing vessel. Chiozza and
Ponzetti [16] apply FMEA to blood cross- The most important aspect of FMEA is the evaluation of
matchingclinical chemistry analyses, as well as to point- the risk level of potential failures that are identified for
of-care testing. Farquharson et all [17] describes the every sub-system or component. The global value of the
result of studies that are carried out by American Bureau damages caused on the function or on the surroundings
of Shipping to provide risk-based guidelines for by every failure is indicated with the risk priority number
performing FMEA on high-speed crafts and propulsion (RPN) [20]. A FMEA uses RPN to assess risk in three
remote control systems. Xu et all [7] conducts research categories: Occurrence (O), is the assessment of how
about overcome highly encountered issue during FMEA frequently the specific failure cause is projected to occur,
analysis for engine systems. Mentioned problem arouses Severity (S), is an assessment of the seriousness of the
from when there are interdependencies among various effect of the potential failure to the system, and Detection
failure modes. Authors suggest fuzzy-logic-based method (D), such as the assessment of the probability that the
to address this issue and test proposed method on diesel operating parameters monitoring system will detect a
engine’s turbocharger. cause/mode of failure before the component/system is
damaged and stopped. The rating is done on a scale from
3. OVERVIEW ON FAILURE MODE AND EFFECT 1 to 10 for each category. The occurrence is related to the
ANALYSIS (FMEA) probability of the failure mode and cause. The severity
index measures the seriousness of the effects of a failure
The FMEAs methodology is one of the risk analysis mode. Thus, a severity index is assigned to the end effect
techniques recommended by international standards [18]. of a failure. The detection index is generated on the basis
It is a systematic process to identify potential failures to of the likelihood of detection relevant design reviews,
fulfil the intended function, to identify possible failure testing and quality control measures. The RPN is
causes so the causes can be eliminated, and to locate the generated by taking the product of these three indices
(occurrence, severity and detection). The RPN represents 9 Very Very remote change the controls will
the risk associated with each failure mode [21]. remote detect a potential cause and subsequent
failure mode
RPN = O x S x D. (1) 8 Remote Remote change the controls will detect a
potential cause and subsequent failure
mode
The FMEA document shows a list of items that identify; 7 Very low Very low change the controls will detect
(i) the system or component part, (ii) the potential failure a potential cause and subsequent failure
mode, (iii) the potential effect of failure, (iv) the severity mode
index (S), (v) the potential cause of failure, (vi) the 6 Low Low change the controls will detect a
frequency of occurrence index (O), (vii) the design potential cause and subsequent failure
verification action, (viii) the detectability index (D), (ix) mode
the Risk Priority Number (RPN). The characteristic 5 Moderate Moderate change the controls will detect
failure mode indexes are expressed on ordinal qualitative a potential cause and subsequent failure
mode
scales [22, 23] identifying the various levels of 4 Moderately Moderately high change the controls
‘dangerous’ situations. Tables 1, 2 and 3 show the high will detect a potential cause and
qualitative scales commonly used for the occurrence, subsequent failure mode
severity, and detectability indexes. 3 High High change the controls will detect a
potential cause and subsequent failure
Table 1. Ratings for occurrence of a failure mode
Rating Possible 2 Very high Very high change the controls will
Probability of occurrence detect a potential cause and subsequent
failure rate
10 Very high: failure is almost inevitable failure mode
9 1/3 1 Almost Controls will almost certainly detect a
8 High: repeated failures 1/8 certain potential cause and subsequent failure
7 1/20 mode
6 Moderate: occasional failures 1/80
5 1/400 4. APPLICATION OF CASE STUDY
4 1/2000
3 Low: relatively few failures 1/15000 4.1 Introduction to Fuel Oil System
2 1/150000
1 Remote: failure is unlikely The fuel oil system is extremely important system on a
Table 2. Ratings for severity of a failure ship which is designed to supply clean fuel oil to main
engine, diesel generators and emergency diesel generator.
Rating Effect Severity of Effect The reliability and the performance of the system are
10 Hazardous System failure resulting in hazardous
directly responsible for accommodation effective voyage
without effects almost certain.
warning
of the ship. On the other hand, it is not easy to keep the
9 Hazardous System failure resulting in hazardous system performance always stable, high and reliable
with effects highly probable. because the fuel oil system is a complex system and
warning tendency to malfunction.
8 Very high System inoperable but safe.
7 High System performance severely affected. The fuel oil system is comprised of storage tanks, service
6 Moderate System operable and safe but tanks, transfer pump, heaters, purifiers, manifolds,
performance degraded. valves, strainers, filters, flow meters and piping which
5 Low Reduced performance with gradual
are illustrated in figure 1.The fuel oil storage and service
performance degradation.
4 Very Low Minor effect on system performance. tanks are equipped with level indicators and level
3 Minor Slight effect on system performance. transmitters. The heavy oil service tank uses a suction
Non-vital faults will be noticed most of heater to adjust oil viscosity while storage tank uses a
the time. whole tank heating system to maintain the oil above the
2 Very Negligible effect on system pour point temperature. The pumps are used to transfer
Minor performance. fuel oil throughout the whole system. The fuel oil
1 None No effect. purifiers are utilized for purification (separation of
Table 3. Ratings for detectability liquids) and clarification (removal of solids). Strainers
are responsible for removing coarse particulate matter
Rating Detection Criteria that may damage rotating equipment. Fuel oil is filtered
10 Absolutely Controls will not and/or cannot detect a to extract particulate matter and to remove water. Filters
impossible potential cause and subsequent failure which are designed to remove water from fuel oil are
mode
known as coalescing filters. The fuel oil pipeline flow
meter is used to measure fuel oil quantities for depending upon conducted literature review and expert
accounting purposes. opinions.

It is obvious that most crucial failure mode is dripping of

fuel valve based on calculations of FMEA analysis which
is illustrated in appendix 1. Therefore, to overcome this
frequently occurred incident, it is wise to decrease inter
arrival times of fuel valves checking and to increase
number of inspections for maximum combustion pressure
measurement. The second highest RPN values belong
two major failure modes namely: early opening of fuel
valve that is aroused from not correct adjustment of an
injector valve opening pressure and seizing of three-way
valve.. First two failures are related to fuel valve can be
taken care by same type of corrective action such as
controlling the opening and closing pressure of a fuel
injection orifice and Pmax control. On the other hand,
remaining failure mode requires different control scheme.
Figure 1. Fuel system diagram [24] The suggested control procedure ,which must be
performed regularly, is measuring the clearance of pump
4.2 Methodological Procedure of FMEA gears, checking shaft displacement and inspection of
bearings.. Moreover, there are some remaining failure
Conducting an initial survey to gather modes (plugged strainer, abnormal temperature in heat
feedbacks/evidences from different fuel system exchangers, inadequate flow range to separator,
casualties, a required level of knowledge to apply FMEA etc..),which are not as severe as above mentioned
was enabled. The main goal of the methodological failures, require different preventive maintenance plans.
approach is to constitute a preventive maintenance plan As a maintenance action, these operative parts of fuel oil
on fuel oil system to identify all aspects of failure and to systems have to be inspected by respect to their operation
suggest precautions on operational duties. Complying parameters such as temperature, pressure and viscosity
with the FMEA application principles, the investigation and overhauled continually by expert operators.
team which constitutes a FMEA practitioner and a couple
of professional marine engineers were guided the 5. CONCLUSION AND FUTURE STUDY
following six main steps:
Step 1- Brainstorm about potential failure modes In this proposed research, firstly failures -which may lead
Step 2 - List potential effects of each failure mode to severely dangerous situations- encountered in fuel oil
Step 3 - Assign a severity ranking for each effect systems on board ships, are analyzed via FMEA method.
Step 4 - Assign an occurrence ranking for each It is observed that the general structure of FMEA
failure mode procedure is appropriate and beneficial for constituting a
Step 5 - Assign a detection ranking for each failure risk based preventive maintenance plan for marine
mode/effect systems. FMEA successfully enables evaluation of
Step 6 - Calculate the risk priority number for each failures that are important in real sense in fuel oil system
failure mode by its allowing structure to feedback, brainstorming and
expert judgements. Preventive maintenance plan which is
Based on the application steps, the results obtained for an output of FMEA analysis increases reliability by
fuel oil systems to constitute a risk-based preventive assigning priority on corrective actions for dripping and
maintenance plan are illustrated in appendix 1. With early opening of fuel valve, abnormal sound in
respect to appendix 1, the analysis and discussion are transfer/supply/booster pumps, low supply pressure on
conducted in next section. high pressure fuel pump and plugged strainer. In the
same time it reduces number of occurrence of failures on
4.3 Analysis and Discussion fuel oil system.

The quantitative results which are illustrated on appendix Although FMEA is regularly used during operational
1 derived from FMEA application on fuel oil system. The stage, it can be also more advantageous if it is employed
necessary precautions will be emphasized in this section. during manufacturing and design stage of marine systems
First of all, RPN value “150” is recognized as the and products. Towards this direction, combined operation
threshold value whether precautions are required or not ,
and designing process in FMEA shed lights to marine Analysis”, IMDC03, the 8th Int. Marine Design
systems producers in whole production stages. Conference, Athens, May 2003.
[13] T. Pinna, R. Caporali, A. Tesinic, “Failure Mode
and Effect Analysis for remote handling transfer
REFERENCES systems of ITER”, Fusion Engineering and Design,
Vol.83, pp.1710-1714, 2008.
[1] Boer, R., Schutten, J.M.J., and Zijm, W.H. “A [14] Chang C. L., Liu P. H, Wei C. C., “Failure Mode
Decision Support System for Ship Maintenance and Effect Analysis Using Grey Theory”, Integrated
Capacity Planning”, CIRP Annals, Vol. 46, No.1, pp. Manufacturing Systems, Vol.12, pp. 211-216, 2001.
391-396, 1997. [15] Chin K. S., Wang Y. M., Poon G. K.K., Yang J.
[2] Fiorenzo, F., and Maurizio, G. “A New Approach for B., “Failure Mode and Effects Analysis Using a
Evaluation of Risk Priorities of Failure Modes in Group-Based Evidential Reasoning Approach”,
FMEA”, International Journal of Production Computers & Operations Research, Vol. 36 , pp.
Research, Vol. 39, No. 13, pp. 2991-3002, 2001. 1768-1779, 2009.
[3] Ireson, G., Coombs, W., Clyde, F., and Richard, Y.M., [16] Chiozza M. L., Ponzetti C. “FMEA: A model for
Handbook of Reliability Engineering and Reducing Medical Errors”, Clinica Chimica Acta 404,
Management - 2nd ed., McGraw-Hill Professional, pp. 75–78, 2009.
New York, 1995. [17] Farquharson J., McDuffee J., Seah A. K.,
[4] Omdahl, T.P., Reliability, availability and Matsumoto T., “FMEA of Marine Systems: Moving
maintainability dictionary, American Society for from Prescriptive to Risk-based Design and
Quality, Quality Press, Milwaukee - United States of Classification”, Annual RELIABILITY and
America, 1988. MAINTAINABILITY, Symposium, 2002.
[5] Sankar, N.R., and Prabhu, B.S. “Modified approach [18] Guidelines for Failure Mode and Effects
for prioritization of failures in a system failure mode Analysis for Automotive, Aerospace and General
and effects analysis”, International Journal of Manufacturing Industries, Dyadem Press, Ontario –
Quality & Reliability Management, Vol. 18, No. 3, pp. Canada, 2003.
324–35, 2001. [19] McDermott, R.E., Mikulak, R.J., Beauregard,
[6] Stamatis, D. H., Failure Mode and Effect Analysis: M.R., Basics of FMEA, Productivity Press - Taylor &
FMEA from Theory to Execution, Second Edition, Francis Group, New York, 2009.
American Society for Quality, Quality Press, [20] Scipioni, A., Saccarola, G., Centazzo, A., and
Milwaukee - United States of America, 2003. Arena, F., “FMEA methodology design,
[7] Xu, K.,Tang, L.C., Xie, M., Ho, S.L., and Zhu, M.L., implementation and integration with HACCP system
“Fuzzy assessment of FMEA for engine systems”, in a food company” Food Control, Vol. 13, pp. 495-
Reliability Engineering & System Safety, Vol. 75, pp. 501, 2002.
17–29, 2002. [21] Seung, J.R., and Kosuke, I., “Using Cost Based
[8] Souza R. Q., Alvares A. J., “FMEA and FTA Analysis FMEA to Enhance Reliability and Serviceability”,
Journal of Advanced Engineering Informatics, Vol.
for Application of Reliability-Centered Maintenance
17, pp. 179-188, 2003.
Methodology: Case Study on Hydraulic Turbines ”,
[22] Fraser, N.M., “Ordinal Preference
ABCM Symposium Series in Mechatronics- Vol.3, pp.
Representations”, Journal of Theory and Decision,
803-812, 2008. Vol. 36, No. 1, pp. 45-67, 1994.
[9] P. A. A. Garcia, R. Schirru & P. F. Frutuoso E. Melo [23] Franceschini, F., and Rossetto, S., “Design for
“A Fuzzy Data Envelopment Analysis Approach for Quality: Selecting Product’s Technical Features”,
FMEA”, Progress in Nuclear Energy, Vol. 46, No. 3- Journal of Quality Engineering, Vol. 9, No. 4, pp.
4, pp. 359-373 2005. 681-688, 1997.
[10] B. Almannai, R. Greenough, J. Kay, “A [24] Demirel, K., Er, I.D., Gemi Yardimci Makineleri
Decision Support Tool Based on QFD and FMEA II, Birsen Yayinevi, Istanbul, 2008.
for the Selection of Manufacturing Automation
Technologies”, Robotics and Computer-Integrated
Manufacturing 24 (2008) 501–507.
[11] Keskin G.A., and Ozkan C. “An Alternative
Evaluation of FMEA: Fuzzy ART Algorithm”, Qual.
Reliab. Engng. Int. DOI: 10.1002/qre (2008).
[12] Papadopoulos Y., Petersen U. “Combining Ship
Machinery System Design and First Principle Safety
 Appendix 1. FMEA Analysis Worksheet

Component Failure mode Failure causes Failure effect Occurrence Severity Delectability RPN

Transfer/Supply/Booster Abnormal sound Defective bearing / Shaft displacement Overloading of electric motor 6 7 4 168
Pump Running without oil Gear wear-out Low transfer/supply pressure 5 8 5 200
High pressure fuel pump Low supply pressure Opening suction valve early or late Decreasing in the engine power output 5 8 5 200
Poor atomisation and combustion, timing
Early opening of fuel
Service pressure too light problems, power balance and temperature 9 7 5 315
valve
Fuel valve variations
Sticking of piston rings in their grooves,
Dripping Oversized injection mechanisms 7 8 6 336
fire blow
Strainer Plugged strainer High viscous oil Pressure drop in line 9 5 3 135
Abnormal sound Defective bearing Overloading of electric motor 6 8 5 240
Separator Inadequate pumping pressure Gear wear-out 5 6 6 180
Inadequate flow range
High viscous oil Decreasing efficiency of separator pump 7 5 4 140
Operational capability of the heat
Heat Exchanger Abnormal Temperature Fouling in heat exchanger 8 5 4 160
exchanger is correspondingly reduced.
Three-way valve Inoperable Seizing of valve Threaten safe ship manoeuvring 5 9 7 315