POLISH MARITIME RESEARCH 4 (116) 2022 Vol. 29; pp.

97-104
10.2478/pomr-2022-0047

OPTIMISATION OF RELIABILITY AND MAINTENANCE

PLAN OF THE HIGH-PRESSURE FUEL PUMP SYSTEM
ON MARINE ENGINE

Vlatko Knežević 2*
Ladislav Stazić 1
Josip Orović 3
Zoran Pavin 3
1
University of Split, Faculty of Maritime Studies, Croatia
2
University of Zadar, Department of Maritime Studies, Croatia
3
University of Zadar, Maritime Department, Croatia

* Corresponding author: [email protected] (V. Knežević)

Abstract

This paper presents a method of adjusting and designing the maintenance scheme for the high-pressure fuel pumps of
a slow-speed two-stroke marine engine, MAN 6S70MC-C. The maintenance database for the marine fuel system was
obtained from the planned maintenance software, and covered a period of 11 years. During this period, 29 failures occurred
that required corrective actions. Our methodology includes failure mode analysis, risk analysis, reliability calculation and
maintenance interval adjustment. Each failure is described using a failure mode analysis, based on a combination of the
mode and cause of failure. The objective of this study is to recommend a new preventive maintenance interval based on the
exponential reliability results and the analysed maintenance data. The initial maintenance plan for each fuel pump was
set to 8,000 running hours, whereas in the modified plan, it is recommended to set this to 4,000 hours. Our results show
an increase in the system reliability from 60% to 95% when the new modified maintenance plan is applied. In addition,
the results and the recommended initial maintenance schedule are validated based on three similar types of engine with
the same fuel pump system. The new maintenance approach can reduce the risk of component failure, which will lead to
increased reliability of the fuel pump system and the optimisation of maintenance costs.

Keywords: reliability, maintenance modification, high-pressure fuel pump, risk analysis

INTRODUCTION activities, which is needed to plan efficient overhauls and optimise

maintenance costs. The CMMS database can also be used to
The functionality and safety of ship operations depend on adjust the initial maintenance plan according to experience [2]
a high level of reliability of the engine room systems and associated and maintenance records, with the aim of increasing reliability
equipment. The most important aspect of high reliability is an and reducing the consumption of spare parts.
optimised and appropriate maintenance schedule. On modern The most efficient method of improving the maintenance
ships, every engine room maintenance activity (planned plan for any engine component is to continuously monitor its
or unplanned) is regularly documented via a computerised technical condition. The importance of early fault detection and
maintenance management system (CMMS), and according to effects of incorrect operation of the main engine are presented
the ISM code [1], this system is mandatory for every ship and in the article [3]. This methodology is called condition-based
shipping company. The main reason for implementing a CMMS maintenance (CBM), and its main advantage is the early
in the engine department is to provide a database of maintenance detection of faults. This can be achieved by recording several

POLISH MARITIME RESEARCH, No 4/2022 97

variables such as vibration, temperature, acoustics, and oil accomplish, due to the behaviour and customs in the shipping
residues for analysis; however, these methods use tools and industry, such as unlawful requests and corrupt demands. The
technologies that are typically expensive for ship owners. maritime sector is vulnerable to bribes, due to a lack of transparency,
A complete overview of CBMs is provided in the article in which can result in the interruption of normal operations,
[4], which also explains the ways in which a CBM can be used navigational risks and illegally signed documents. Clearing agents
to optimise maintenance strategies. In [5], an artificial neural and marine surveyors are frequently affected by bribery, and this
network is used to analyse the database of a container ship in has an impact on the availability of correct information about the
order to develop a CBM strategy. technical condition and port calls of a ship. According to DNV-GL
Improvements to a failure database were achieved in [6] CEO Henrik O. Madsen, the shipping industry is “too conservative
through a combination of reliability-centred maintenance (RCM), and too passive” in this regard [16]. This opinion is also shared by
failure simulation and expert knowledge. This methodology is some other authors [17, 18].
useful for optimising the failure database and obtaining accurate
predictions of failures. Failure mode and effect analysis (FMEA)
was used in [7] to develop an optimised preventive maintenance CASE STUDY: HIGH-PRESSURE
schedule for a system. Applying this method could ensure higher FUEL PUMPS
reliability of the system components and lower maintenance
costs. The same method was applied in [8] to ship components For this study, the HP fuel pumps were selected as one of the
such as a fin stabiliser, high elastic couplings and diesel generator crucial components of a complex fuel system. These components,
sets, in order to develop a new repair strategy. An improvement along with the fuel injectors and exhaust valves, have a high failure
on the FMEA method was presented by Nguyen [9], who used rate (λ) in the fuel system [19], defined as the ratio between the
failures in a tanker ship system as a case study. The improvement number of failures and the total number of running hours. We
was based on risk estimation and judgements by experienced consider six HP fuel pumps (Fig. 1) in a slow-speed S70 MC-C
marine engineers. The reliability, availability and maintainability two-stroke marine engine, and the maintenance database was
(RAM) method was used in [10] based on failure records for four obtained from a real shipping company, which allowed access
similar ships. The results highlighted the importance of regular to its database on condition of confidentiality. The ship was
periodic maintenance, especially for cylinders, as these are the a crude oil tanker with a maximum output power of 16,780 kW
components with the highest unavailability value. The Real-time (82 RPM main engine) and three auxiliary engines with a power
Anomaly Detection Intelligent System (RADIS) framework [11] of 910 kW each.
has been applied to a diesel generator of a tanker ship to establish
a smart maintenance scheme, with particular reference to the
maritime industry, where this type of maintenance is still under
development. With the aim of optimising the maintenance costs
and availability of ships and naval vessels, risk-based maintenance
was investigated in [12]. All the authors mentioned above agree
that existing maintenance schedules are limited and require
further development in order to provide improvements in the
maintenance and reliability of these systems.
In this paper, the maintenance and reliability of a high-pressure
(HP) fuel pump is improved through a method called Maintenance
Concept Adjustment and Design (MA-CAD) [13]. The objective
of this approach is to reduce the probability of a failure occurring
by analysing the maintenance database and adjusting the initial
maintenance interval. In general, components with the most frequent
operational failures are selected for this analysis. For example, the
study in [14] examined maintenance adjustments for exhaust valves,
while the work in [15] focused on air compressors, and it was found Fig. 1. Cross section of the HP fuel pump [20]
that the frequency of failure and the maintenance costs could be To ensure the functionality and availability of each HP fuel
reduced by changing the initial maintenance schedule. To achieve pump, all components must be maintained at scheduled intervals.
a relevant and meaningful change in the maintenance plan, it is These maintenance intervals are proposed by the manufacturer,
necessary to obtain a reliable database (maintenance history) of the Classification Society [21] and the shipping company, and
the component under study. This task is usually very difficult to together they form the initial maintenance plan (Fig. 2).

Fig. 2. Initial maintenance plan for the HP fuel pump in the CMMS [22]

98 POLISH MARITIME RESEARCH, No 4/2022

 In the CMMS, the initial maintenance plan is divided into • Maintenance task (explanation of the maintenance task
three different maintenance tasks, labelled with codes C0500, performed),
S0001 and 00147. Each code represents a different preventive • MTTR (average time needed to repair a failed component),
maintenance task to be performed by the engine crew on • TF (running hours for each component at the time of
board (with the exception of task S0001). It is recommended failure).
to perform maintenance jobs with code C0500 every 8,000
running hours, which includes inspection of the fuel pump Tab. 2. Summary of the exploitation data for the six HP fuel pumps
suction and puncture valve, and adjustment of the pilot Com- Failure Cause Maintenance MTTR TF
valve, pump cam and shock absorber. Overhaul of the fuel ponent number task (h) (h)
pump (top cover, barrel, and roller guide) at intervals of 1
Leakage and
improper operation
Actuator
disassembled and 1 1,050
16,000 running hours is required for job code 00147. The of VIT actuator cleaned
last maintenance job (S0001) is carried out every 60 months 2 Leakage of shock Exchanged with
2 8,680
absorber spare part
by the Classification Society or as part of the Chief Engineer’s
Leakage of shock Exchanged with
overhaul inspection while ship is in dry dock, which is then 3 absorber spare part 1 16,357
confirmed by a surveyor. HP
fuel Leakage of shock Exchanged with
pump 4 absorber spare part 0.5 31,005
MAINTENANCE DATABASE ANALYSIS 1
Leakage of shock Exchanged with
5 absorber spare part 0.5 46,335

The maintenance database for the six HP fuel pumps was Puncture valve Valve cleaned and
6 fouled all seals replaced 2 51,445
analysed over a period of 11 years. According to this database Overhauled and
Wear- out of
and the CMMS, the first maintenance operation took place in 7 a suction valve replaced with 2 54,805
spare part
2007, and the last updated entry was made in 2018. During this
Leakage of Exchanged with
period, the total running hours of the six fuel pumps amounted 8 puncture valve spare part 2 5,886
to 59,098 h. When this value is divided by the number of years Leakage of shock Exchanged with
HP 9 1 8,826
of operation, the average operating rate is 5,372 h/year. Table fuel absorber spare part
1 shows the maintenance work carried out during this period pump Leakage of shock Exchanged with
2 10 absorber spare part 0.5 40,820
for each HP pump; the last column contains the number of
Leakage of Exchanged with
corrective (unexpected) maintenance operations carried out 11 puncture valve spare part 2 48,193
after failure of the component. Overhauled and
Wear-out, damaged
12 puncture valve replaced with 1 6,473
Tab. 1. Maintenance jobs in the initial maintenance plan spare part
HP
Corrective fuel Leakage of shock Exchanged with
Component S0001 C0500 00147 pump 13 absorber spare part 0.5 46,513
jobs
3
Fuel pump top
Fuel pump 1 2 8 5 7 Leakage of
14 puncture valve cover parts 2 48,193
overhauled
Fuel pump 2 2 8 4 4
Leakage of shock Exchanged with
15 absorber spare part 1 13,193
Fuel pump 3 2 8 4 3
Leakage of shock Exchanged with
16 absorber spare part 1 45,463
Fuel pump 4 2 8 4 4 HP
fuel Leakage of Exchanged with
Fuel pump 5 2 8 4 6 pump 17 puncture valve spare part 2 48193
4
Fuel pump 6 2 8 4 5 Components
Wear-out of overhauled and
18 a suction valve replaced with 2 54,805
Σ29 spare part
Overhauled and
FAILURE MODE ANALYSIS 19 Leakage of shock
absorber replaced with new 2 10,253
gaskets
Leakage of Exchanged with
The purpose of failure mode analysis (FMA) is to show 20 puncture valve spare part 1 27,893
the relationship of events between failure cause and failure Leakage of shock Exchanged with
HP 21 1 28,314
effect [23]. In the MA-CAD method, FMA is based on fuel absorber spare part
a combination of failure mode and failure cause (FMCC). pump Leakage of shock Exchanged with
5 22 absorber spare part 1 41,915
Three crucial properties of FMCC are taken into consideration:
Parts overhauled
predictability, reaction time (MTTR) and evidence/cause 23 Plunger blocked in and replaced with 4 45,275
barrel
(possibility of observing component failure). Based on the new seal rings
operating data for the six fuel pumps, the following failure 24 Leakage of Exchanged with
2 46,093
puncture valve spare part
characteristics were highlighted (Table 2):
• Component (identification of the engine component),
• Failure number (number of failures that occurred),
• Cause (reason for each failure),

POLISH MARITIME RESEARCH, No 4/2022 99

 years, depending on the size and type of ship [24]. In the case
Com- Failure Maintenance MTTR TF
Cause of the HP fuel pumps, 29 failures occurred within 11 years, and
ponent number task (h) (h)
Leakage of Exchanged with the life of the component was set to 20 years. ELFF is calculated
25 1 5,970
puncture valve spare part using Eq. (1):
Leakage of shock Exchanged with
26 absorber spare part 2 11,679
HP ELFF = F(T ) = 52.7272  (1)
fuel Leakage of Exchanged with
27 puncture valve spare part 1 38,955
pump
6 Parts overhauled where:
Leakage, fault of a
28 plunger and replaced with 4 42,735 F(T) – is the probability of failure within interval T (29
new seal rings
failures);
Leakage of Exchanged with
29 puncture valve spare part 2 48,193 L – is the component life span (20 years);
T
∫0 R(t)dt – is the reliability weighted cycle for interval T
Most malfunctions were due to the leakage of fuel oil in certain (11 years).
parts of the fuel pump. The corrective maintenance task for this
type of malfunction was to replace faulty units with spare parts. In order to determine the magnitude of risk associated
These malfunctions varied in terms of the mean time to repair with the occurrence of a failure, it is necessary to calculate
(MTTR) for a failed component, which was recorded in the the risk index. There are two categories of risk indices, which
CMMS by the engineer responsible for this maintenance task. also represent safety and operational effects. Each category
The component with the most failures was the shock absorber, represents the criteria used to rank the importance of the failure
with an MTTR of 1 h, and the failure with the longest repair time effect. The magnitude of the failure effect is expressed using
was the blockage of the plunger, with a downtime of 4 h. The the significance index (SI).. For the case of failure of the HP
MTTR depends on the experience of the person performing the fuel system, the safety significance index (SIs) is set to 0.0001,
maintenance task and the availability of spare parts. representing an event with no injury and no damage to the ship
The basic effects of failure in the case of leaks or wear-out or environment. The operation significance index (SIo) is set
faults (shock absorber, puncture valve) are vibrations and to 0.001, which represents partially available operation during
defects, leading to insufficient fuel supply to the injector. In the failure, meaning that the failure effect is that the ship is
the case of a variable injection timing (VIT) actuator, the main unavailable for operation for some hours [13].
function of which is to reduce fuel consumption and achieve The safety risk index is calculated as:
optimum pressure even at low loads, failure results in uneven
load distribution in the engine, and can lead to an incorrect RI(S) = SI(S) · ELFF = 0.00527   (2)
fuel-air ratio or even an increase in fuel consumption. For all
of the fuel pump failures that occurred, the cause was internal. and the operational risk index as:

RI(o) = SI(o) · ELFF = 0.0527    (3)

RISK ANALYSIS AND RELIABILITY
For the HP fuel pumps, the safety risk index is 0.00527, since
The effects of a fault on the ship, the engine room system and a failure affects only one cylinder, which could be out of service
the environment are described by risk analysis, where risk is until the failure is corrected. The operational risk index is slightly
defined as the expected loss associated with the occurrence of an higher, as the vessel could be without function for several hours
adverse event [23]. The magnitude of the expected loss is indicated (depending on the MTTR). There are two ways of reducing the
by the significance index, while the probability of occurrence is risk index: through preventive maintenance, or by modifying the
expressed by the expected life failure frequency (ELFF). component. Although the importance of a failure cannot be reduced
The significance index is divided into two classes, representing by maintenance, the frequency of failure can be reduced, and this
safety and operational effects, and is measured on a scale of will also reduce the ELFF. The results of a risk index analysis for
zero to one, where one represents a catastrophic event (loss of the HP fuel pump are shown in Table 3.
the ship, environmental catastrophe), and zero represents an
effect of no significance. In the case of HP fuel pumps, which Tab. 3. HP fuel pump risk index analysis
are crucial for the operation of the main engine, the safety and FMCC Component SI(s) SI(o) ELFF RI(s) RI(o)
operational indices are determined as follows: 1 HP fuel pump 0.0001 0.001 52.72 0.00527 0.0527
• Safety significance index SI (s) > 0.001 (marginal - possible
injury, possible damage to the vessel) EXPONENTIAL RELIABILITY OF THE SYSTEM
• Operational significance index SI (o) > 0.0001 (reduced
performance - the ship operates with reduced performance) In reliability engineering, failures due to completely random
events are represented using an exponential failure distribution.
The ELFF is determined as the number of failures that occur In this distribution, the reliability function R(t) is expressed
during the lifetime of a component. It has been estimated that mathematically as [25]:
the lifetime of a ship and its equipment is between 20 and 30 R(t) = e–tλ       (4)

100 POLISH MARITIME RESEARCH, No 4/2022

 In general, the reliability of the entire system is not easy
to determine, since many components within the system are
interconnected in a serial or parallel configuration. Failure of
a serial system occurs if any of its components fails; however,
this is not the case for the HP fuel pump system, as one pump
is designated for each cylinder, and in the case of failure of one
pump, the engine can still run with a lower load.
The reliability of the HP fuel system can be expressed as [23]:
n
Rs = 1 – ∏[1 – Ri(t)]     (5)
i=1

For the case of six equal HP fuel pumps, the reliability is

calculated as follows: Fig. 3. Exponential reliability analysis for six HP fuel pumps

n
Rs = 1 – ∏(1 – Ri)n     (6)
i=1
MA-CAD MAINTENANCE MODELLING
To determine the exponential reliability of the system and
each component, it is necessary to calculate the failure rate (λ). The MA-CAD method for modification of the maintenance
The failure rate is the frequency with which a component fails plan uses the three-parameter Weibull distribution. The
within a specified period of time, and it is calculated by dividing following equation was used to estimate the reliability function
the number of failures by the total number of running hours. R(t) with the Weibull method [26]:
Table 4 shows the failure rate for the HP fuel pump system,
and Table 5 shows the failure rate for each HP fuel pump. The R(t) = e–( )β
, t ≥ γ, β > 0, η > 0    (7)
mean time between failures (MTBF) is useful in predicting how where:
likely a component is to fail in a certain time period, or how t – is the time of failure,
often a certain type of failure may occur. η – is a scale parameter (representing the mean life of
the component),
Tab. 4. Parameters of the HP fuel pump system γ – is a threshold parameter,
Total running hours t = 59,098 [h] β – is a shape parameter.
Number of failures n = 29
Each parameter must be determined or estimated (if the
Failure rate n =0.4907 ∙ 10−3
λ =— maintenance history is not available) to properly calculate the
t
reliability function of the system. The time (t) of each failure
n =2038 [ℎ]
Mean time between failures MTBF = —
λ was recorded in the CMMS, and based on this information the
scale parameter (η) was calculated as 8,443 h. This parameter is
Tab. 5. Failure rate for each HP fuel pump considered to be the mean life (time to failure) of a component
or system under normal operating conditions.
HP fuel pump/cylinder number Failure rate (λ)
The shape parameter (β) has a positive value, and indicates
Pump 1 1.1844 ∙ 10 –4 whether the distribution has a decreasing, increasing or constant
Pump 2 6.7684 ∙ 10 –5 rate. This parameter is related to the predictability of failure,
which depends on the shape of the failure rate. Table 6 shows
Pump 3 5.0763 ∙ 10 –5
the relationship between the failure rate, predictability and the
Pump 4 6.7684 ∙ 10 –5 shape parameter. The threshold parameter (γ) is also known as
Pump 5 1.0152 ∙ 10 –4
the location parameter, as it defines the lowest possible value
in this distribution. All other values must be greater than the
Pump 6 8.4605 ∙ 10 –5 threshold parameter; in this case, the value is zero, which means
failures are possible as soon as operation starts.
Fig. 3 shows the reliability for each HP fuel pump, which depends
on the number of failures (i.e. the failure rate). It can be seen that Tab. 6. Relation between predictability and the shape parameter [6]
pump 3 has the highest reliability, due to the lowest number of Shape
Description Predictability
failures, and pump 1 has the lowest reliability, as it has the most parameter (β)
failures. Pumps 2 and 4 are shown on the same curve, as they FMCC with a decreasing or none (p=0) 0<β≤1
have the same number of failures. The first maintenance interval, constant failure rate
according to the CMMS, is set to 8,000 running hours, and the FMCC with an increasing low (0 < p < 0.5) 1<β<2
failure rate
reliability of pump 1 at this time is 38.78%. Pump 3 has the highest
reliability of 66.62%, followed by pumps 2 and 4, with 58.19%. The FMCC with a significantly high (0.5 ≤ p ≤ 1) 2≤β
increasing failure rate
reliability of each fuel pump is calculated using Eq. (4).

POLISH MARITIME RESEARCH, No 4/2022 101

 The predictability is typically used to assess whether the
periodic preventive maintenance in the maintenance plan is
effective. If the failure rate decreases or remains constant, it
is assumed that preventive maintenance can be replaced by
corrective action. However, this cannot be applied to the HP
fuel pumps, where there are significant numbers of failures, and
for this reason, preventive maintenance measures are crucial.
With a significantly increasing failure rate, the shape parameter
(β) is set to two, with a predictability of between 0.5 and one.
Fig. 4 shows the reliability of the HP fuel pump system with
an exponential Weibull distribution. Fig. 5. Reliability comparison of the initial and modified maintenance intervals

The results of the modified maintenance intervals are

compared in Fig. 5. The reliability for both the initial and
modified maintenance plans was calculated using Eq. (6). The
initial maintenance period was reduced by 50% and the system
reliability was increased from 60% to 95% in the modified plan,
with a period of 4,000 operating hours. When the number of
preventive maintenance activities is increased, the maintenance
cost typically also increases, and it is important to plan the
supply of spare parts. Using the MA-CAD method, the required
number of spare parts can be calculated by dividing the average
usage rate (λu) by the MTBF. Most of the spare parts used in
the overhaul of the HP fuel pump were new gaskets, which are
Fig. 4. Exponential reliability of the HP fuel system used to prevent fuel oil leakage. These parts are relatively cheap,
and having more than necessary in stock does not significantly
increase the maintenance costs.
RESULTS AND DISCUSSION It is important to emphasise that this maintenance
recommendation applies to HP fuel pumps that are driven
When modifying the initial maintenance plan, the objective by camshafts. For ME engines with hydraulic and electronic
is to minimise the occurrence of failures without increasing the components, the occurrence of failures is much lower, and the
maintenance costs. Our reliability analysis of this system with first maintenance overhaul could be done at 32,000 hours of
an initial maintenance interval of 8,000 running hours was operation [27]. In addition, this type of engine has an automotive
compared with other HP fuel pump systems from the database fuel system, so if more than one fuel oil booster pump fails, the
[22]. The examples used for comparison were drawn from operating mode will be automatically changed to the slower mode
similar types of engine (six cylinders) and HP fuel pumps, and the engine will run at a lower load. It should also be noted
with the first maintenance interval set to 4,000 running hours. that for some types of engine, the first maintenance interval is
Table 7 shows the number of failures and the failure rates for already set to 4,000 h, such as for the MAN S/L80MC. However,
the three examples, with an initial maintenance interval of this includes only the suction and puncture valve (checks on
4,000 running hours. The benefit of a 50% shorter maintenance condition), while the overhaul of the shock absorber is still done
interval (a reduction from 8,000 to 4,000 hours) is verified by at 8,000 h. The modified maintenance plan can increase the
a reduction in unexpected failures. reliability of the system by about 50% compared to the initial
In the first example, there were eight HP fuel pump failures, maintenance period of 8,000 h of operation. The reliability of
and in the second there were 11 failures over a total operating the HP fuel system without optimisation of the maintenance
time of 48,247 hours. In the third example, there were only five plan was 0.11% over 8,000 running hours, whereas with the
failures over 90,174 operating hours. These three examples had modification, it was 0.65% for the same running hours. According
significantly lower numbers of failures compared to the engine to the maintenance plan in the CMMS (Fig. 2), the first major
in our case study, with a total of 29 failures, and can be used as overhaul is set to 16,000 h; this could also be reduced to 8,000 h, in
validation of the modified initial maintenance plan. view of the importance of maintenance tasks such as the overhaul
of the complete top cover of the fuel pump, and the overhaul of
Tab. 7. Parameters with a modified maintenance plan the fuel pump barrel and roller guide. Also, the authors have
Example 1 Example 2 Example 3
considered the problem of increased workload with the new
recommended maintenance plan. However, the nature of the first
Total running hours 48,247 48,247 90,174 maintenance plan is to check on condition and not the complete
Number of failures 8 11 5 overhaul. This action includes adjustments of the pilot valve, fuel
pump cam, inspection of the shock absorber and the adjustments
Failure rate (λ) 1.658 ∙ 10 –5
2.279 ∙ 10 –5
5.544 ∙ 10 –5
are based only on observations, so if there are no failures the

102 POLISH MARITIME RESEARCH, No 4/2022

workload won’t be increased. The proposed modification of the REFERENCES
maintenance plan is validated based on three examples in which
the first maintenance action is carried out after 4,000 operating 1. Revised ISM code, Effective from 1 January 2015. [Online].
hours. Our results show that the MA-CAD methodology with Available: https://www.classnk.or.jp/hp/pdf/activities/
an adequate maintenance history database can achieve higher statutory/ism/ISM_Cd/ISM-Code-e.pdf. [Accessed: Sept. 2,
reliability for the system or a given engine component. 2022].

2. V. Knežević, L. Stazić, J. Orović, and L. Mihanović, “Use of

CONCLUSIONS PMS continuous improvement scheme for maintenance
adjustments in shipping industry,” Book of Proceedings -
Choosing the most efficient maintenance plan is important in 8th International Maritime Science Conference,12.04.2019.,
order to increase the reliability and safety of shipping through pp. 415-420, Budva, Montenegro.
reducing the frequency of failure of a component or system. To
achieve this goal and to obtain accurate results, it is necessary 3. C. Karatuğ and Y.Arslanoğlu, “Importance of early fault
to analyse the maintenance database of a component from diagnosis for marine diesel engines: a case study on efficiency
a CMMS or maintenance record books. In this paper, the failures management and environment”, Ships and Offshore Structures,
of six HP fuel pumps are analysed using a comprehensive vol. 17, pp. 472-480, 2022, doi: 10.1080/17445302.2020.1835077.
MA-CAD method, and the following conclusions can be drawn:
• Of a total of 29 failures, 26 were fuel oil leaks or wear-out 4. A. Prajapati, J. Bechtel, and S. Ganesan, “Condition based
of a shock absorber, puncture or suction valve. Given the maintenance: a survey”, Journal of Quality in Maintenance
high level of predictability of these failures, preventive Engineering, vol. 18, no. 4, pp. 384-400, 2012, doi.org/10.1108/
periodic maintenance and condition-based (inspection) 13552511211281552.
maintenance are strongly recommended.
• The results of the exponential reliability function showed 5. Ç. Karatuğ and Y. Arslanoğlu, “Development of condition-
that the initial maintenance interval should be modified in based maintenance strategy for fault diagnosis for ship engine
order to reduce the failure rate and consequently increase systems,” Ocean Engineering, vol. 256, 2022, doi.org/10.1016/j.
the reliability of the system. According to the initial oceaneng.2022.111515.
maintenance schedule proposed by the manufacturer, the
initial interval is set to 8,000 running hours. In the modified 6. F. Vera-García, J. A. Pagán Rubio, J. Hernández Grau, D.
plan, it is reduced by 50% (4,000 h), which will ensure Albaladejo Hernández, “Improvements of a failure database
higher reliability of the system. Our results are confirmed for marine diesel engines using the RCM and simulations,”
based on three examples with the same maintenance plan. Energies, vol. 13, p. 104, 2020, https://doi.org/10.3390/
• The most important function of the CMMS is to enable en13010104.
updating and control of the spare parts inventory. In
a scenario where the modified maintenance plan is 7. T. O. Alamri and J. P. T. Mo, “Optimisation of preventive
implemented and the number of failures is reduced, maintenance regime based on failure mode system modelling
savings in maintenance costs could be achieved, while considering reliability,” Arab. J. Sci. Eng., vol. 49, 2022, https://
the maintenance costs for increased preventive actions in doi.org/10.1007/s13369-022-07174-w.
the case of the HP fuel pumps are negligible.
In order to choose the most efficient maintenance plan, it 8. T. Song, T. Tan, and G. Han, “Research on preventive
is necessary to continuously monitor and record the technical maintenance strategies and systems for in-service ship
condition and maintenance actions during the period of equipment,” Polish Maritime Research, vol. 29, no. 1, pp. 85-96,
operation. The changes to the maintenance plan should also be 2022, https://doi.org/10.2478/pomr-2022-0009.
implemented in new maintenance software programs that allow
the maintenance plan to be adjusted based on the condition 9. H. Nguyen, “Fuzzy methods in risk estimation of the ship
of each engine component. This methodology could also be system failures based on the expert judgments,” Journal of
applied to other major engine components to achieve a longer KONBiN, vol. 43, pp. 393-403, 2017, doi:10.1515/jok-2017
and more reliable system life and safer navigation. -0058.

ACKNOWLEDGEMENTS 10. M. Bayraktar and M. Nuran, “Reliability, availability, and

maintainability analysis of the propulsion system of a fleet,”
The authors are grateful to the shipping company that gave Scientific Journals of the Maritime University of Szczecin,
us access to their maintenance software program to obtain the Zeszyty Naukowe Akademii Morskiej w Szczecinie, vol. 70,
database. This company allowed us access under conditions of no. 142, pp. 63-70, 2022.
anonymity, so further details about this shipping company and
the vessels have not been published in this paper. 11. C. Velasco-Gallego and I. Lazakis, “RADIS: A real-time
anomaly detection intelligent system for fault diagnosis of

POLISH MARITIME RESEARCH, No 4/2022 103

 marine machinery,” Expert Systems with Applications, vol. 204, 24. O. Dinu and A. M. Ilie, ”Maritime vessel obsolescence, life
p. 117634, 2022. cycle cost and design service life,” IOP Conference Series:
Materials Science and Engineering, vol. 95, 2015.
12. J.Cullum et.al., “Risk-based maintenance scheduling with
application to naval vessels and ships,” Ocean Engineering, 25. D. Matika and L. Mihanović, “Reliability of a light high speed
vol. 148, 2018, pp. 476-485, https://doi.org/10.1016/j. marine diesel engine,” Shipbuilding: An International Journal
oceaneng.2017.11.044. of Research and Development, vol. 62, pp. 28-36, 2011.

13. B. Vučinić, “MA-CAD, Maintenance Concept Adjustment 26. Life Data Analysis: Weibull Analysis. [Online]. Available:
and Design,” University of Delft, Faculty of Mechanical https://www.weibull.com/basics/lifedata.htm. [Accessed:
Engineering and Marine Technology, 1994. Sept. 12, 2022].

14. A. Bukša, I. Šegulja, and V. Tomas, “The adjustment of the 27. 1st Classification Society Event 2017: ME Fundamentals
maintenance approach for improved operability and safety of and Critical Principles. [Online]. Available: https://pbm.hr/
ship navigation,” Promet—Traffic & Transportation, vol. 22, media/1100/man-me-hcu.pdf. [Accessed: Sept. 21, 2022].
no. 2, pp. 95-103, 2010, doi.org/10.7307/ ptt.v22i2.168.

15. L. Stazić, et al., “Maintenance interval adjustment based on

the experience: Case study of marine air compressor system,”
Naše More: International Journal of Maritime Science and
Technology, vol. 67, no. 2, pp. 146-152, 2020, doi:10.17818/
NM/2020/2.7.

16. L. Vogdrup-Schmidt, “The shipping industry is too CONTACT WITH THE AUTHORS
conservative and passive.” [Online] Available: https://
shippingwatch.com/Services/article7817989.ece. [Accessed: Vlatko Knežević
Sept. 13, 2022]. e-mail: [email protected]

17. M. E. Manuel, “Maritime risk and organizational learning,” University of Zadar

ebook (1st ed.). CRC Press, 2012, doi: 10.1201/9781315593937. Department of Maritime Studies
Croatia
18. T. Teijl, “Innovating in the maritime cluster,” Master’s thesis,
Delft University of Technology, The Netherlands, 2014.
Ladislav Stazić
19. A. Bukša, D. Martinović, and M. Vidaković, “Failure analysis
of the diesel engine ship propulsion system,” Pomorstvo: University of Split
Scientific Journal of Maritime Research, vol. 23, no. 2, pp. 391- Faculty of Maritime Studies
404, 2009. Croatia

20. MAN B&W: Volume II Maintenance: S70MC-C, E[Online]

Available: https://engine.od.ua/man-70m [Accessed: Dec.
02, 2022]. Josip Orović

21. Lloyd’s Register, “Rules and regulations for the classification of Zoran Pavin
ships, Part 5: Main and auxiliary machinery,” London, United
Kingdom, 2013. [Online]. Available: http://docshare04. University of Zadar
docshare.tips/files/30667/306670934.pdf. [Accessed: Sept. 12, Maritime Department
2022]. M. Pavlinovića 1
23000 Zadar
22. Amos Business Suite version 9.0 – Maintenance software Croatia
program, [Online] Available: https://amos-business-suite.
software.informer.com/ [Accessed: Dec. 2, 2022].

23. C. E. Ebeling, An Introduction to Reliability and Maintainability

Engineering, 3rd ed. Waveland Press: Long Groove, IL, USA,
2019. ISBN 978-1-4786-3933-6.

104 POLISH MARITIME RESEARCH, No 4/2022