Abbasetal

Marine Structures 71 (2020) 102718
Contents lists available at ScienceDirect
Marine Structures
journal homepage: http://www.elsevier.com/locate/marstruc
Review Article
An overview of maintenance management strategies for corroded

steel structures in extreme marine environments
Muntazir Abbas a, Mahmood Shafiee b, *
a
Department of Energy and Power, Cranfield University, Bedfordshire MK43 0AL, UK
b
School of Engineering and Digital Arts, University of Kent, Canterbury CT2 7NT, UK
A R T I C L E I N F O A B S T R A C T
Keywords: Maintenance is playing an important role in integrity management of marine assets such as ship
Maintenance management structures, offshore renewable energy platforms and subsea oil and gas facilities. The service life
Steel structures of marine assets is heavily influenced by the involvement of numerous material degradation
Degradation
processes (such as fatigue cracking, corrosion and pitting) as well as environmental stresses that
Marine corrosion
vary with geographic locations and climatic factors. The composition of seawater constituents (e.
Extreme climatic conditions
g. dissolved oxygen, salinity, temperature content, etc.) is one of the major influencing factors in
degradation of marine assets. Improving the efficiency and effectiveness of maintenance man
agement strategies can have a significant impact on operational availability and reliability of
marine assets. Many research studies have been conducted over the past few decades to predict
the degradation behaviour of marine structures operating under different environmental condi
tions. The utilisation of structural degradation data – particularly on marine corrosion – can be
very useful in developing a reliable, risk-free and cost-effective maintenance strategy. This paper
presents an overview of the state-of-the-art and future trends in asset maintenance management
strategies applied to corroded steel structures in extreme marine environments. The corrosion
prediction models as well as industry best practices on maintenance of marine steel structures are
extensively reviewed and analysed. Furthermore, some applications of advanced technologies
such as computerized maintenance management system (CMMS), artificial intelligence (AI) and
Bayesian network (BN) are discussed. Our review reveals that there are significant variations in
corrosion behaviour of marine steel structures and their industrial maintenance practices from
one climatic condition to another. This has been found to be largely attributed to variation in
seawater composition/characteristics and their complex mutual relationships.
1. Introduction
The maritime sector plays an extremely important role in the economic growth of many countries around the world [1]. However,
the costs associated with maintenance and repair of the assets operating in this sector (such as ship structures, offshore renewable
energy platforms and subsea oil and gas facilities) are extremely high and continue to rise. Currently, the maintenance costs in the
maritime sector account for between 20 and 40% of the total operating expenses (OPEX) [2–5]. The high cost of maintenance for
marine assets is mainly attributed to the involvement of various degradation/deterioration processes in aquatic environments which
* Corresponding author. School of Engineering and Digital Arts, University of Kent, Canterbury CT2 7NT, UK.
E-mail address: m.shafiee@kent.ac.uk (M. Shafiee).
https://doi.org/10.1016/j.marstruc.2020.102718
Received 17 March 2019; Received in revised form 20 December 2019; Accepted 17 January 2020
0951-8339/© 2020 Elsevier Ltd. All rights reserved.
M. Abbas and M. Shafiee Marine Structures 71 (2020) 102718
pose detrimental effects on technical integrity, safety and reliability of the assets.
The degradation in marine assets usually occurs due to a variety of mechanisms such as fatigue cracking, corrosion, pitting, scour,
etc. Many recent research studies have identified the corrosion as the most prominent degradation mechanism in the maritime sector,
which can result in catastrophic failures [6,7]. The marine assets are either totally or partially immersed in corrosive seawaters. Most
of the fixed and floating marine structures located near the harbors are exposed to domestic or industrially polluted seawaters, which
further accelerate the pace of structural degradation, in particular corrosion. Additionally, non-submerged structures in the vicinity of
coastal areas are vulnerable to corrosion damage due to the accumulation of salt and other corrosive compounds in marine
environments.
To control the rate of degradation, increase the operational uptime, reduce the life-cycle costs, and extend the service lifetime in
marine assets, a number of maintenance practices including preventive maintenance (PM), condition-based maintenance (CBM), risk-
based maintenance (RBM), and structural health monitoring (SHM) have been deployed by marine industry professionals [8]. In
today’s world, the improvement of operational and environmental safety has been the prime objective of maintenance operations in
the maritime sector. Historically, maintenance was seen as more of an economic liability than an effective tool to improve productivity
in organisations. However, after experiencing some serious incidents and environmental disasters such as Macondo oil spill in the Gulf
of Mexico in 2010, it has become incumbent for the industry to comply with requirements set out by regulatory bodies such as the
International Maritime Organisation (IMO), European Maritime Safety Agency (EMSA), United States Coast Guard (USGC), etc.
[9–11]. These regulations have created enormous hurdles for maritime operators, necessitating continuous improvement in their
existing monitoring systems and maintenance regimes. One such example is the IMO law on sulphur content control in fuel, which is
going to be implemented from 2020 onwards. It dictates that the sulphur content which is currently 3.5% m/m (mass/mass)
(equivalent to 35,000 ppm) must be capped at 0.50% m/m (equivalent to 5000 ppm) [12]. Such legislations will increase the pressure
on maritime industries to develop more reliable, risk-free and cost-effective maintenance strategies for their critical assets.
The commercial maritime industry is currently more reliant on conventional time-based maintenance procedures, which in general
are inefficient and labour intensive. In order to increase safety, operational uptime, effectiveness and reliability while reducing
maintenance costs [13,14], the maritime industry must adopt risk-based and reliability-centred asset maintenance practices from some
other industries such as the aerospace, nuclear, and chemical.
The maintenance management of marine assets is a complex task because of the uncertainties involved in long-term prediction of
the corrosion damage under different environmental conditions. It is a proven fact, supported by the scientific literature, that the
selection of a maintenance strategy for marine assets is highly influenced by climatic conditions such as temperature, relative hu
midity, wind speed and direction, etc. Therefore, it is logical to apprehend that the implementation of the same maintenance regime for
systems operating in different environmental conditions and with different degradation modes will not result in an optimal outcome.
To optimise maintenance practices and achieve greater reliability, throughput, cost-effectiveness and safety in the marine sector,
several advanced data-driven models integrated with condition monitoring (CM) and non-destructive testing (NDT) technologies as
well as risk assessment tools have been proposed over the past few decades.
This paper presents an overview of the state-of-the-art and future trends in asset maintenance management strategies applied to
corroded steel structures in extreme marine environments. The corrosion prediction models as well as industry best practices on
maintenance of marine steel structures are extensively reviewed and analysed. In this regard, we identify several deterministic and
probabilistic models that have been developed to predict the corrosion rate of marine steel structures as a function of the exposure
Fig. 1. Evolution of maintenance strategies in the marine sector.
2
period, environmental conditions and material properties. It is shown that the existing models involve considerable uncertainties in
data collection and analysis for accurate modelling of the combined effects of environmental factors on overall corrosion loss in marine
structures. To overcome this drawback, some applications of advanced technologies such as computerized maintenance management
system (CMMS), Bayesian network (BN), artificial intelligence (AI), and multi-criteria decision analysis (MCDA) to maintenance
optimization of corroded steel marine structures will be discussed. Our review reveals that there are significant variations in corrosion
behaviour of marine steel structures and their industrial maintenance practices from one climatic condition to another.
The rest of the paper is organized as follows. Section 2 describes various contemporary maintenance strategies and their signifi
cance in the marine industry. Section 3 presents the impacts of surrounding climatic conditions on the structural degradation (in
particular corrosion) of marine assets. Section 4 reports the results of a literature review on various maintenance strategies applied to
marine steel structures in corrosive environments. Section 5 provides an overview of some advanced techniques that can be used for
corrosion prediction and maintenance planning of marine steel structures. Section 6 discusses the results of the critical analysis of the
identified literature. Finally, Section 7 concludes the study with a brief summary and future directions.
2. Maintenance strategies
A maintenance strategy delineates an organisation’s vision on how to preserve the health and safety of assets throughout their life-
cycle. Generally, it is comprised of procedures for survey/inspection, repair, upkeep and renewal of the systems, subsystems, and
components [15]. In the search for greater efficiency and lower cost, a number of maintenance strategies have been conceived by the
researchers over the years [16]. Fig. 1 shows the evolution of key maintenance strategies in the marine sector. These strategies are
briefly introduced in the following subsections:
2.1. Corrective maintenance
In corrective maintenance or run-to-failure (RTF) strategy, a correction action is taken to bring the equipment back to a functional
state after it unexpectedly stops working. This action includes either repair or replacement of failed component and it can be carried
out as and when required. Therefore, this maintenance strategy is preferred only on those equipment whose failure consequences are
considered minimal. The investment required for the execution of this maintenance strategy is much less than any other maintenance
strategy, however, it may incur additional repair costs and increase downtime when applied to critical equipment [17].
2.2. Preventive maintenance (PM)
The PM is an interval-based maintenance procedure which is implemented on an operational equipment so as to avoid any potential
failure or severe degradation that may impact system reliability in near future [18]. The frequency of PM tasks is often chosen based on
the experience of technicians or recommendations from original equipment manufacturers (OEMs). PM strategy has been able to offer
higher system availability, reduced failure rates, longer equipment lifespan, and lower cost compared to the corrective maintenance.
PM is currently practiced in many marine industries as the most preferred maintenance strategy [19].
Despite several intrinsic benefits, PM does not guarantee elimination of all unexpected failures as it does not take into account the
present health state of components. For this reason, PM sometimes results in unnecessary machinery downtime, excessive repair costs
and maintenance-induced failures [20,21]. Some researchers reported that conducting time-based PM actions may lead to mis
judgement about the equipment’s health condition as the rate of usage may not be constant over time [22].
2.3. Condition-based maintenance (CBM)
The CBM includes use of modern CM methods to precisely diagnose faults and predict the future working condition of the system
[23]. According to this strategy, a maintenance is performed when one or more indicators show that equipment performance is
degrading or that the equipment is about to fail [24]. In other words, the CBM decision is made based on a set of indicators associated
with system’s physical condition or performance [25]. Many researchers have shown that the CBM strategy is more effective than the
time-based PM strategy. It is reported in the literature that the use of CM techniques may extend maintenance overhaul cycles by up to
50% and save between 25% and 45% of maintenance costs [26]. This maintenance strategy is based on the output data collected either
online and off-line from different CM technologies, such as vibration analysis, ultrasound analysis, infrared thermography, oil analysis,
wear particle analysis, acoustic emission testing, etc. The collected data is analysed to extract meaningful patterns and predict the time
for future maintenance. Some researchers have considered diagnostic and prognostic as main features of a CBM system (for example,
see Ref. [27]).
Even though the CBM is able to deliver substantial savings in maintenance cost and reduction in failure risks, the current surveys
reveal that only 10% of industries in the marine sector use CBM as their preferred maintenance strategy [15]. One of the reasons for
slow adoption of CBM is the limited access to highly skilled personnel for execution and further interpretation of the results [28].
Nevertheless, with the advancement in prognostic and health monitoring (PHM) methods, the reliance on CM and active/passive SHM
techniques in equipment maintenance has grown exponentially in recent years [29].
3
2.4. Reliability-centred maintenance (RCM)
The RCM concept was conceived for the first time in American Aviation industry in mid-1980s. It is a planned maintenance
program that retains the essential functions of a system while improving its reliability, maintainability and availability (RAM) [30,31].
It is reported in the literature that if the RCM is employed correctly with a thorough understanding of its essence, it can be helpful to
reduce maintenance effort by 40%–70% compared to other maintenance strategies such as the corrective maintenance, scheduled
overhaul, scheduled replacement, and scheduled on-condition tasks [32]. The maintenance decisions in RCM are made based on
qualitative risk information mainly derived from the knowledge/skills of the operators. RCM necessitates the default maintenance
actions to counter failure situations which arise due to unavailability of effective proactive maintenance procedures. These default
actions include failure finding, run to fail, and redesign [33].
Within the marine sector, RCM is commonly used for the maintenance of ships and their associated equipment. The RCM strategy
uses techniques such as Failure Mode Effect Analysis (FMEA), Failure Mode Effect and Critical Analysis (FMECA) and Fault Tree
Analysis (FTA) to identify possible causes of each failure, as well as some statistical techniques to estimate mean time between failure
(MTBF), mean time to repair (MTTR), etc. [34,35]. The FMEA is the essence of RCM, as it provides a procedure to identify and
recognize function(s), failure modes, failure causes, and effects and consequences of a failure on the operability of a particular
equipment, system or process [36]. For comprehensive literature on RCM and its implementation, readers are referred to Refs.
[37–41].
2.5. Risk-based inspection (RBI)
The concept of RBI was originated from the nuclear industry in the 1970’s. Over the years, it has been adopted by other sectors such
as petrochemical, electrical systems, offshore energy sector and, to a lesser extent, shipping. The RBI emphasizes on the factor of risk in
the overall maintenance of equipment. The risk is evaluated based on the likelihood (probability) of occurrence of a hazard and its
consequent effects on the operation of the equipment [42,43]. RBI is an optimised maintenance strategy which offers greater safety and
provides an overall risk mitigation plan to minimise the frequency of undesirable events [44,45]. In this strategy, the risk analysis
outcomes are utilized for maintenance scheduling and decision-making. Risk analysis permits flexibility for the use of qual
itative/quantitative methods or a combination of both. Some researchers have proposed a risk-based approach for the inspection and
maintenance of ship vessels, where the RCM is recommended for mechanical systems and RBI for hull and structures. For further
information on RBI, the readers can refer to Refs. [46–49].
2.6. Other contemporary maintenance strategies
In recent years, several other developments in maintenance management systems have been evolved and implemented in the
marine sector. Reliability database (RDB) is one of novel concepts proposed for the management of reliability datasets. It is a prime
source of information for design, development and initial deployment of advanced, cost-effective and optimised maintenance systems.
RDB records all significant maintenance activities with a core focus on equipment failures. This is considered to be the prime enabler
for maintenance strategies such as Total Life Cycle Systems Management (TLCSM) and CBM plus (CBMþ). The TLCSM deals with the
Fig. 2. Classification of RUL prediction techniques [55].
4
total system performance (including hardware, software, and human), its operational effectiveness, and suitability, survivability,
safety, and affordability [50]. The CBMþ is a novel maintenance concept which was developed on the basis of CBM strategy but by
including various advanced tools and procedures, acquired from real-time health monitoring and sensor technologies [51]. The CBM
þ facilitates the shift from conventional maintenance regimes (e.g. time-based PM) to proactive/predictive methods governed by CM
programmes.
The term remaining useful life (RUL) in maintenance implies the remaining time of a system/subsystem to perform its function
prior to failure or end of useful life [52]. The RUL estimation models are either deterministic or probabilistic, but they generally
incorporate degradation factors, material properties, and environmental conditions [53]. The RUL estimation models can be classified
into four categories: analytical (physics-based), model-based (data-driven), knowledge-based, and hybrid (fusion) (see Fig. 2) [54–57].
The RUL methods can be further refined into a more accurate maintenance model by the use of statistical and Artificial Intelligence (AI)
techniques [58]. Sometimes, manufacturers use accelerated life testing (ALT) data to predict the lifetime of equipment under different
operating and environmental conditions. Typically, the OEM’s maintenance recommendations are based on ALT results [53]. These
tests are simulated on an accelerated time scale and then the reliability of the equipment is estimated based on specified failure data
settings, operating conditions and design stresses [59].
3. Impact of environmental conditions on corrosion of marine structures
Marine structures are often exposed to severe and corrosive environmental conditions such as high or low temperatures, high
salinity, high or low pH values, etc. These conditions accelerate the material degradation rate and thus shorten the time to failure of
structures. The reliability of steel structures in marine environments is highly influenced by the variations in ambient climate, loading
conditions, applied protective measures, and the adopted maintenance strategies [60]. Corrosion and fatigue are the most prominent
degradation mechanisms in marine steel structures which adversely affect their reliability by inducing strength losses, brittle fracture,
thickness reduction, cracking, etc. Static and shock loads, erosion, and turbulent seawater velocity are additional factors contributing
to the failure of marine structures. Some research studies have reported that over 90% of ships’ structural failures are caused by
corrosion [61].
The key factors affecting the corrosion process in marine steel structures have been categorised into different types of physical
properties, chemical properties, and biological contents in seawater [62–66]. The most influential factors in physical, chemical and
biological properties have been studied widely in the literature. Table 1 lists the most influential factors in the marine corrosion
process, including seawater temperature and velocity, pressure, pH level, dissolved oxygen (DO), salinity, pollutants, etc. Some re
searchers have shown that the biological factors such as sulphur reducing bacteria (SRBs) are the most contributing element in the
anoxic seawater conditions.
The characteristics of marine conditions and seawater specifications are found to be immensely variable across the globe. For
example, the temperature of surface seawaters varies from –2 � C along poles to 35 � C along the equator. Consequently, the corrosion
rate as well as the health state of marine structures will be different from a region to another [67,68]. Furthermore, the corrosion rate
may be altered with inspection, maintenance and repair actions, which makes the RUL prediction process for marine structures more
complicated [69].
Although majority of the marine steel structures are protected with coatings to inhibit corrosion and stress corrosion cracking
(SCC), an inspection is needed to assure that the corrosion protection system is working [70,71]. The prominent corrosion resistant
methods used on external surfaces of the ships or offshore structures include: sacrificial anodes, impressed current cathodic protection
(ICCP), various types of anti-fouling, anti-corrosive paints, and ultrasonic guided wave methods [72]. However, some marine systems
such as heat exchanger tubes are not yet provided with a surface coating, causing them to be more vulnerable to corrosion damage. The
maintenance practices in such cases are even more dependent on operating ambient conditions and proportion of detrimental
corrosion factors in the seawater composition.
4. Findings of the literature review
This section reports the results of the literature review performed on various maintenance practices and asset management stra
tegies adopted for marine structures in extreme corrosive environments. Our review covers journal papers, conference proceedings,
books, academic dissertations, industry reports and government guidelines. The identified studies are categorised with respect to some
criteria such as the type of marine structure under consideration (e.g. offshore platforms, ships, oil rigs, subsea pipelines and offshore
wind turbines), degradation mechanisms and maintenance planning methodologies, and some key findings will be reported.
Table 1
A list of factors affecting the marine corrosion in steel structures.
Type Factors
Physical Seawater temperature, Water velocity, Surface wetting, suspended solids, pressure
Chemical Dissolved oxygen (DO), CO2, Salinity, pH, pollutants, carbonate solubility
Biological Bacteria, Biomass, Pollutants, Marine growth
5
4.1. Findings on maintenance management strategies for marine steel structures
Marine structures are designed for operation in hostile environments subject to corrosive sea-water, hot and cold temperature
extremes, and static/dynamic loading conditions. Although the corrosion affects the performance of an assets throughout the life cycle,
the extent of the damage varies depending on many factors such as the designed allowance for corrosion loss, the effectiveness of
preservation methods, and severity of dynamic environmental conditions [73]. The main function of a reliable maintenance scheme is
to identify the critical components, functions, failure modes, causes, effects and consequences, and then recommend a cost-effective
repair policy to attain optimal operational availability. Nowadays, different aspects of maintenance are considered during the design
phase of marine structures; however, some modifications, additions/alterations in the existing engineering design may be necessary in
later stages during the operation. Marine structures usually deteriorate more rapidly under extreme conditions than under normal
conditions. This causes the gap between designed capability and current performance to become greater and greater over time. The
deterioration of the design performance in marine structures over time is illustrated in Fig. 3.
In commercial ships, the operators/owners either rely on OEM’s recommendations or seek the expertise of engineers to determine
maintenance support requirements [56]. OEM’s maintenance procedures are often based on the age of the ships, not real-time
degradation data. The tendency of shifting from time-based PM to CBM, online monitoring, and predictive maintenance is
emerging over the years in the marine sector. CBM is considered to be an efficient approach to improve the reliability and reduce the
operating costs of marine systems, especially for those assets involving high safety risks. Emovon [4] conducted a comparative study on
the application of different maintenance strategies to ships, and finally, offline-CBM was found to be the most effective method for
maintenance of seawater pumps in a marine diesel engine. More recently, Michala et al. [74] presented a novel concept of CBM on
ships using wireless systems, where the CM data about ship machinery components is transmitted to the onshore maintenance experts
through a decision support system (DSS).
Lazakis and Olçer
€ [5] presented a Reliability and Criticality Based Maintenance (RCBM) strategy using fuzzy multi-attributive
group decision-making (FMAGDM) technique to identify an optimised maintenance strategy for maritime assets. The study
concluded that the time-based PM was the best maintenance strategy, followed by the predictive maintenance. Cicek and Celik [75]
used risk priority number (RPN) in FMEA methodology to enhance the reliability and operational safety while decreasing the failure
probability of marine diesel engines. Similarly in a comparative study of onshore and offshore wind turbines, Shafiee and Dinmo
hammadi [76] proposed a FMEA-based risk evaluation methodology integrating qualitative (expert-driven) and quantitative (data-
driven) information to formulate a maintenance strategy for wind turbines. Tang et al. [77] proposed a novel model based on Analytic
Hierarchy Process (AHP) and Fuzzy Borda Count (FBC) for identification of the most risky items in offshore oil and gas equipment.
Some researchers opined that the use of RCM in the marine sector is culturally different than that in the aviation sector. Therefore, it is
more sensible to consider RCM as a philosophy rather than a methodology [78]. From the commercial ship owners’ point of view, the
RCM is considered to be exhaustive, time-consuming and complex [3]. Wabakken [79] reported that the RCM is a long-term strategy
which requires time and resource-intensive effort. Therefore, RCM has been hesitantly adopted by maritime companies.
RBI is becoming a popular maintenance strategy in the marine sector, in particular for ship’s hull and structures. Cullum et al. [22]
proposed an RBI scheduling framework for naval vessels and ships and concluded that shifting from RCM strategy to RBI is more
convenient than shafting from PM or CBM to RBI. Dong and Frangopol [80,81] proposed quantitative risk assessment (QRA) models for
ship structures subject to corrosion and fatigue. The genetic algorithm (GA) and Bayesian networks were used to provide an optimal
inspection/repair plan and reliability/risk updating for overall mitigation of lifecycle risk. Similarly, Turan et al. [82] proposed an RBI
model to estimate the overall reliability of ships and diving support vessels and prioritize the maintenance tasks.
A comprehensive distribution of journal articles and conference papers based on maintenance strategies is shown in Table 2.
Numerous NDT and SHM methods are used for detection, quantification, and prognostics of surface and subsurface defects (such as
Fig. 3. Deterioration of design performance over time [15].
6
Table 2
Distribution of journal articles and conference papers based on maintenance strategies.
Ref Type of Structure Maintenance strategy
RCM CBM PM RBM/RBI HYBRID
[2] Ship ✓
[5] Ship ✓
[8] Ship ✓
[15] Ship ✓ ✓
[18] Ship ✓ ✓ ✓ ✓
[19] Ship ✓ ✓
[20] Subsea Pipeline ✓
[21] Ship ✓
[22] Ship ✓ ✓ ✓
[28] Ship ✓
[38] Ship ✓
[42] Subsea Pipeline ✓ ✓
[43] Subsea Pipeline ✓
[50] Defence equipment ✓
[69] Ship ✓
[73] Ship ✓
[74] Ship ✓
[75] Ship ✓
[77] Oil and gas ✓ ✓
[78] Oil and gas ✓ ✓
[79] Ship ✓
[80] Ship ✓
[81] Ship ✓
[82] Ship ✓
[83] Oil and Gas ✓
[84] Wind turbine ✓
[85] Wind turbine ✓
[86] Ship ✓
[87] Ship ✓
[88] Ship ✓
[89] Ship ✓
cracks, gouges, pits, and erosion/corrosion loss) due to various structural degradation mechanisms like corrosion, pitting, or fatigue
cracking [90]. Photographic imaging has been used to measure and bifurcate appearance of marine pitting corrosion [91]. Recently,
visual imaging and high-resolution photography were used with integration on remotely operated vehicles (ROV) or autonomous
underwater vehicles (AUV) for enhanced safety and efficiency and reduced cost of underwater repair activities. These vehicles can
utilize videos that are able to diagnose structural corrosion and anode wear, and thus facilitate the inspection of difficult-to-access
areas [92–94]. From literature review, the following CM or SHM techniques were identified for use by marine operators [95,96]:
Visual and optical testing; radiographic testing; ultrasonic testing (conventional phased array, and guided waves); metallographic
examination; electrochemical and electromagnetic testing; liquid penetrant testing; magnetic particle testing; acoustic emission
testing; infrared and thermal testing; mass loss; X-ray; eddy current.
Some newly introduced technologies in metal and composite structures are acoustic emission and guided waves ultrasonic testing
(GWUT). These technologies utilize active/passive transducers and contact/non-contact techniques to detect structural cracks,
corrosion under insulation (CUI), pits and corroded portions in metallic and composite structures [97–99]. The guided waves have
been considered as a useful defect detection technique for large structural assets and as antifouling, ice detection and de-icing missions
on marine and aircraft structures. Recently, marine inspection robotic assistant (MIRA) system and micro-aerial vehicle-based have
been used for structural fault identification of ship structures [100].
Table 3
Main types of corrosion in ships and offshore structures.
Corrosion material Corrosion Features Problem areas Effects
General mild, high tensile steels roughly uniform over extended Plating reduction in plate thickness,
areas structural capacity
Pitting limited for mild steels, mainly high highly localized penetration, Plating, local detail local reduction in thickness and
tensile and stainless steels often with clusters of pit stress intensification
Crevice mainly stainless and some high highly localized elongated at connections, washers, etc. localized stress intensification
tensile steels
Galvanic dissimilar metals in contact penetration localized different metals in contact, local failure, and severe localized
variation in metal composition material loss
7
4.2. Findings on factors affecting structural corrosion in marine environments
Structural degradation in marine environments is a time-dependent process, primarily occurring due to corrosion and fatigue
[101]. These processes also encourage several other degradation processes such as strength reduction, brittle fracture, buckling, etc.
[102,103]. In a study about reliability-based maintenance of ship’s hull under corrosion effects, the replacement of affected plate was
recommended to be carried out when the thickness reduces below 75% of its designed thickness [104].
Corrosion is an electrochemical process occurring in marine environments because of reaction between various ingredients of the
metallic surface and sea water. It occurs due to the availability of water along with an electron acceptor element, like oxygen [105].
The marine steel structures generally experience several forms of corrosion under immersion state. The most common types of
corrosion in ships and offshore structures and their effects on material degradation are presented in Table 3.
The general corrosion and pitting corrosion are more common in marine applications than other forms of corrosion such as
galvanic, crevice, SCC, groove, and edge corrosion [106–108]. In general corrosion, the thinning phenomenon occurs uniformly on the
surface of a metallic surface. On the other hand, pitting is a highly localized type of corrosion that occurs randomly in various stages
over certain areas; hence it results in perforation and thickness reduction in specific regions of the metal surface [109–111].
When the corrosion attack on metal structures is non-uniform (i.e. pitting or crevice corrosion), the collection of corrosion rate data
via conventional methods can be misleading [112]. This phenomenon is more common where the coatings or the base metal itself are
deteriorated [113]. Some research studies reveal that the reduction of tensile strength in presence of pit corrosion is 2.5 times more
than that in presence of general corrosion [114]. The number of pits increases with the deterioration of coatings which in turn leads to
corrosion growth independently [115]. During the pit formation, cathodic oxygen reduction occurs on the adjacent surfaces of pits to
reduce the corrosion process. Engelhardt and Macdonald [116] categorised the pitting process into three phases, including: nucleation
(pit initiates), propagation (pit grows), re-passivation (pit growth stops).
In general, the marine corrosion can be categorised into short-term and long-term corrosions, depending on duration of the
exposure in seawater. The duration of short-term corrosion typically ranges between 6 and 24 months of initial exposure, when the
corrosion process is led by activation, concentration, and diffusion phases. Then the long-term corrosion takes place, which is led by
biological activities and nutrients in seawater [117]. Some researchers have shown that the duration of short-term and long-term
corrosions depends on the constituents of seawater (biological and chemical) and its physical properties, temperature in particular
[118].
Marine steel structures exhibit significant variance in corrosion rates with changing zones with respect to the sea surface. These
zones include tidal, atmospheric, splash, and submerged zones. Some research literature reported that highest rates of corrosion are
observed in splash zones followed by low tidal zones. In the natural seawater conditions, Melchers [62–66] showed that the highest
corrosion losses occur in splash zone and immersed zone, followed by half-tide and coastal atmosphere. Fig. 4 presents the corrosion
rates of marine steel structures in different exposure zones.
In natural seawaters, the corrosion rate of low carbon steel structures is estimated to be between 0.1 mm and 0.3 mm/y; however it
can rise up to 2–4 mm/y in the severe marine climatic conditions [72]. It is widely believed that the rate of corrosion decreases with the
exposure period, possibly due to the hindrance offered by corrosion deposits for free exchange of ions. However, some studies have
reported that the rate of corrosion may increase with the exposure period in cases where the structure is subject to dynamic loads,
higher velocity, and pollutant factors in seawater [119,120]. In moderate marine climates, the corrosion content accumulated on steel
structures is primarily comprised of lepidocrocite (ɣ-FeO(OH)), goethite (α-FeO(OH)), maghemite (ɣ-Fe2O3) and magnetite (Fe3O4)
[121]. In case of atmospheric marine corrosion, iron samples are initially corroded into lepidocrocite (γ-FeO(OH)) – an unstable rust
form. However, because of continuous interaction between oxygen and water in surrounding environment, it is converted into goethite
– a more stable form of rust [122,123]. The corrosion in marine steel structures is oxidation of ferrous iron ions which yields a reddish
brown ferrous oxy-hydroxide (FeO(OH)) compound, i.e. rust. As an initial step of oxidation, the rust layer starts building up on to the
Fig. 4. Corrosion rates of marine steel in different exposure zones [62–66].
8
surface of the metal structure due to the presence of free oxygen in sea water and its continuous access on to the metallic surface. With
the increase in exposure period, the surface deposits on metal skin barricade the interaction between free oxygen and metal skin.
Subsequently, the rate of metal loss will vary non-linearly [124–126]. Faraday’s law is generally used to estimate the initial corrosion
rate in which the effects of bacterial actions, corrosion deposition and biofilms are assumed to be negligible [127].
The Pourbaix Potential-pH diagram graphically demonstrates the electrochemical aspects of corrosion process, and it has appli
cations in the corrosion of metals subjected to an aqueous electrolyte, batteries or fuel cells. It is used to establish the types of reaction
and stable phases of reaction products in an equilibrium state of a chemical process. This diagram gives a very effective and deep
understanding of the possible reactions and yielding products, including passivity regions in a corrosion process. It however cannot
predict the rate of corrosion and chemical processes at a given temperature in electrolytic solution [111,128].
4.2.1. Environmental factors

The environmental conditions play a significant role in degradation of offshore metallic structures and, thus, in selection of the
inspection method and its frequency. The metal alloys selected for use in marine structures generally have good corrosion resistance.
Melcher [129] reported that the composition of metal structures significantly influences the initial corrosion phase (kinetically
controlled oxidation) and the long-term anaerobic corrosion phase, whereas the diffusion phase of oxidation was found to be inde
pendent of the alloying constituents. The weather conditions and seawater constituents vary enormously across various oceans around
the globe. The seawater reservoirs in colder regions have low seawater temperature and salinity level; thus the DO concentration is
higher. However, in the hot countryside of tropical/sub-tropical regions, the sea surface temperature and hence the salinity levels are
higher and DO content ranges up to 3.5–4 mg/L. Typically, the salinity of major natural seawaters ranges between 32,000 and 45,000
ppm [130].
In summer, the sea surface temperature in some tropical regions may reach up to 33–35� . The seawater temperature of the Gulf of
Mexico is reported to be in the range of 20–32 � C. Nergis et al. [131] studied the range of various prominent seawater parameters in the
Arabian ocean and reported the seawater surface temperature range to be from 28 � C to 41 � C in summer seasons. Higher corrosion
rates have been reported in warm seawater regions. The significance of seawater temperature can be estimated from the fact that
corrosion rate of marine steel structures at a temperature of 25 � C was twice of that at 10 � C [125]. During a research study on at
mospheric corrosion in coastal regions of the Arabian sea, Jamil et al. [132] found out that the corrosion rates were in the category C5,
which is placed in ‘very high’ corrosive environment according to BS EN ISO 9223 [133,134]. More recently, Jilani [135] discussed
various levels of pollutants in coastal waters of the Arabian sea and their impacts on the transformation of intrinsic open seawater
properties (like pH, DO, total solid content, etc.). The results of her analysis are presented in Table 4.
4.2.2. Physical and chemical factors

The physical and chemical factors are closely interrelated in contributing to marine structural corrosion. The temperature, DO,
salinity and wetting duration are the dominant factors in marine immersion corrosion [136]. The seawater may contain different
concentration levels of chlorine ions, carbon dioxide (CO2), hydrogen sulphide (H2S) and ammonia [137]. A higher quantity of these
compounds adversely affects the surface of the metallic structure by accelerating the rate of corrosion. Higher content of H2S, CO2, and
seawater temperature have harmful effects on metallic marine structures [138]. The combined action of the chloride as well as the salt
deposit provide a conducive environment for flourishing microbial activities, which often results in crack formation in metals [136].
With the initial exposure of metal structures in seawater, a passive layer is formed on its surface that resists further corrosion.
Dissociated chloride ions (Cl ) in seawater may penetrate this protective film and initiate crevice/pitting corrosion. The hydroxide
ions (OH ) in aqueous electrolytic solutions assist in passive layer formation, whereas, Cl ions damage the layer and facilitate further
corrosion and pit development [91].
The corrosion factor is more detrimental in hot and moist seawater conditions of ship ballast tanks, as it contains a high con
centration of entrapped oxygen even at higher temperatures [139]. Some researchers have reported the significant effect of tem
perature, pH, calcium carbonate solubility, and exposure time on corrosion of structural steels in seawater, brackish and freshwater
(see Refs. [117,140]). Melchers [141,142] showed that the corrosion rate in moderate- and low-temperature seawaters is doubled for
each 10 � C temperature rise when controlled by kinetic process; whereas during the diffusion process it is doubled after every 30 � C rise
in temperature, given the DO concentration is constant [143]. A series of seawater corrosion tests at various temperatures were carried
out by Chandler [144]. It was found out that the corrosion loss of carbon steel at 25 � C was nearly twice larger than that at 10 � C. In the
open seawaters, DO is able to discharge freely with increasing the temperature. Corrosion rate tends to increase with temperature up to
80 � C, then onwards it declines sharply due to the rapid decrease in solubility of oxygen.
The standard fluctuation of pH in seawaters lies between 7.8 and 8.2 and it has been reported by several researchers that this
variation does not have significant impact on corrosion rate. However, it can indirectly influence accumulation of calcium carbonate
Table 4
The monthly average level of pollutants in coastal waters of the Arabian sea.
Parameters Sample location General coastal water compositions
Ocean water sandpit Fish Harbour Netty jetty MS jetty West Wharf China Creek
pH 8.1 7.1 7.23 7.07 7.14 6.9 7.5–8.3

DO (mg/l) 2 0.84 1.18 1.1 1.06 0.66 >3.5
EC (mS/m) 58 56 54 48 54 65 50
9
on cathode protected structures [118,144]. The pH variations may exert an active influence on pitting and crevice corrosion of
active–passive metals [145].
In ferrous alloys, the effect of CO2 on corrosion loss is far less than DO at same concentration level. At a CO2 concentration of 20
mL/L and same temperature conditions, oxygen is found to be ten times more corrosive than CO2 [145]. Presence of O2 and CO2 in
seawater can reduce its pH value from slightly alkaline to acidic, which in turn can enhance the corrosion of steel. Several research
studies have highlighted that in case of short-term corrosion, the nominal pH level does not affect the corrosion process; whereas, CO2
can upset the pH value in long-term exposure. The overall pattern of corrosion in marine steel structures is nonlinear. Although the
short-term corrosion may initially exhibit a linear pattern, it has been asserted by various researchers that the short-term corrosion
pattern can be highly erroneous if it is used for prediction of long-term corrosion. The influencing factors which control the long-term
corrosion are identified as exposure time, temperature, salinity, microbiologically influenced corrosion (MIC), SRBs, water velocity
and alloy effects [146–148].
Increasing chloride content in seawater can aggravate the pit corrosion in submerged metal [149]. Moreover, the combined effect
of DO and chloride concentration highly accelerates the corrosion rate. The corrosivity of structural steel specimens along the coastline
of a heavily industrialized region in Baltic Sea was investigated by Zakowski et al. [150]. It was concluded that the corrosion rate in
low-salinity seawaters is significantly lower (0.0585 mm/year) than that in nominal ocean conditions. Table 5 gives the salinity levels
across various sea regions throughout the world. As can be seen, the salinity level is highest in seawaters of hot sea regions (e.g.
Mediterranean Sea and Indian Ocean) and lowest in the cold countryside (e.g. Baltic and Caspian seas) [151].
The DO concentration in seawater is a function of the following factors: temperature, water velocity, salinity and biological ac
tivities. The oxygen solubility decreases at a higher temperature. Under standard atmospheric pressure at sea level, the DO concen
tration is found to be 8.26 mg/L at 25 � C and 12.77 mg/L at 5 � C [152]. Oxygen is the main electron acceptor for the corrosion process;
hence its quantity decreases at elevated temperatures which may reduce the overall rate of corrosion. However, this decreasing
corrosion factor is compensated by the increasing temperature and salinity level; therefore, the corrosion rate typically increases with
the rise in temperature. In addition to temperature effect on DO, its percentage in seawater reduces with an increase in the chemical
and biological content, particularly in polluted seawaters. In a study about corrosion on austenitic steel, Malik et al. [149] reported
that the content of DO decreases with the increase in water temperature. Corrosion rate was found to increase within the temperature
range of 25–65 � C; however then onward, the critical pitting potential (Epit) was found to remain constant. The DO concentration in
seawater as a function of salinity and temperature is presented in Table 6.
The influential parameters of seawater vary with the sea depth and this variability is also dependent on the geographical location
and season. As the water depth increases, the temperature reduces but the hydrostatic pressure increases; however, the latter does not
pose any significant effect on corrosion rate [153]. Due to higher nutrients and higher seawater temperature, the corrosion rate in the
shallow sea environment is found to be higher than that in the deep sea environment. In a research study, Venkatesan et al. [145,154]
showed that the short-term corrosion rate of mild steel in surface water of Indian ocean is four times more than that in deep water.
Melchers [155] showed that the effect of water depth on corrosion rate is subject to the variation in temperature, DO and nutrient
levels.
The photosynthesis process in marine ecology system causes a significant increase in DO concentration. Similarly, the air bubbling
produced by the wave propagation in open sea serves for seawater oxygenation. Various researchers have reported that the DO
concentration in seawater changes with regional surface seawater temperature from about 8.0 mL/L in the Arctic seas to 4.5 mL/L or
even less in the tropical seawater. In certain harbour conditions, it further reduces due to the presence of nutrient-rich waters, pol
lutants and industrial wastages [156–158]. Fig. 5 illustrates the relationships between various seawater parameters.
Effects of seawater velocity on the corrosion rate of immersed or semi-submerged metallic structures have been investigated by
several field/laboratory experiments and the results are presented in Refs. [159,160]. Corrosion rate was observed to increase non
linearly with the water velocity (0–1 m/s). This effect was found more prominent in the early phase but slowed down gradually with
the rise in growth of biofouling, marine growth, and corrosion products on steel coupons [105,161]. The material loss in some metals
(such as iron, copper alloys, and steel) tends to be higher beyond a critical velocity [162], however minor velocity changes can be
ignored during corrosion studies on structural steel [163]. The effect of velocity on marine structural corrosion can be more damaging
when the accumulated corrosion growth is removed mechanically or naturally by the wave action [164,165].
Table 5
Salinity levels across various sea regions throughout the
world.
Sea Salinity in ppm
Baltic Sea 8000

Caspian Sea 13000
Black Sea 22000
Irish Sea 32500
Atlantic Ocean 37,000
Red sea >36000
Mediterranean Sea 41000
Indian Ocean 40000–45000
10
Table 6
Oxygen concentration in seawater as a function of salinity and temperature.
Temperature (� C) Oxygen concentration (ppm) level at various salinity levels (in parts per thousand (ppt))
0 8 16 24 51 36
0 14.6 13.9 13.1 12.4 11.8 11.4

5 12.5 11.9 11.3 10.7 10.2 9.9
10 10.9 10.4 9.8 9.4 8.9 8.7
15 9.5 9.1 8.7 8.2 7.9 7.7
20 8.5 8.1 7.7 7.3 7.0 6.8
25 7.6 7.2 6.9 6.6 6.3 6.2
30 6.8 6.5 6.2 6.0 5.7 5.6
Fig. 5. The relationships between (a) conductivity and salinity (b) dissolved oxygen and salinity [151].
4.2.3. Microbiological factors

The exposure of steel structures into the marine conditions rapidly initiates a complex chain of electrochemical reactions which
include colonization of marine growth, biofilms and various forms of bacteria. The structural degradation under the influence of
biological activities (microorganism, bacteria, biofilms, etc.) is commonly known as microbiological induced corrosion (MIC) [166].
The microbiological ingredients play a vital role to nourish oxygen depleted regions/anaerobic conditions during the latter phases of
corrosion and promote localized corrosion, particularly during long-term corrosion phase. These effects can be visible in all exposure
zones; the tidal, splash and coastal atmospheres [167]. The microbial biofilms and biofoulings are undesirable
micro-organism/bacterial cells which deposit on metal surface and encourage a conducive environment for anaerobic corrosion [144].
The marine biofouling may be comprised of flora and fauna in the form of micro and macro bio-organisms. SRBs encrusting algae,
Fig. 6. (a) A model indicating the effect of nutrient level on corrosion (b) A model indicating the effect of temperature and DO on corrosion [134].
11
fungi, seaweeds, molluscs, barnacles, zebra mussels, worms, sea squirts, barnacles, hydroids are few common types of biofoulings in
marine environment [72,127,168]. The pollutant addition in seawater aggravates the concentration of hydrogen sulphides (H2S) and
nutrient content in the form of dissolved inorganic nitrogen (DIN) which significantly elevate the corrosion rate of low carbon steel
[134,169,170]. The models presented in Fig. 6 highlight the above phenomena as well as the effects of temperature and DO variations
on corrosion process.
The heated oil or lubricants inside the ship tanks encourage the growth of microbiological contaminations (MBCs), which further
leads to a higher corrosion rate. The seawater temperature in the range of 20–50 � C is found to be ideal for the growth of SRBs. Presence
of crucial compounds in seawater such as hydrogen sulphide (H2S), other sulphides and sulphates form an unstable/corrosive passive
film on metal surfaces, which permeates the interaction of these detrimental compounds with the metal surface and aggravates the
corrosion process [171]. In addition to the growth of general and localized corrosions, the micro-biofoulings can induce various other
types of corrosion such as hydrogen embrittlement and SCC [127,172]. Seawaters in coastal regions are also engulfed with numerous
metal and non-metal ingredients with the addition of industrial and domestic effluents. These factors enhance corrosion of steel
structures through galvanic reaction, acidic hydrolysis and cathodic reaction [135,173].
4.3. Findings on corrosion prediction models for marine structures
Over the years, several corrosion prediction models for marine steel alloys have been proposed in the literature. These models can
be divided into different types of empirical, phenomenological (qualitative analysis based on experimental data), probabilistic, and
physical models. The empirical models are based on historical data or measurement of corrosion loss, whereas the physical models are
based on actual corrosion process [174,175]. Accurate prediction of corrosion is a challenging task because the available data
sometimes is highly scattered due to the involvement of extremely dynamic environmental conditions. Earlier research studies pro
posed deterministic and linear models for corrosion prediction. However, in recent research, several nonlinear and probabilistic
methods have been proposed. In some cases, the uncertainty of prominent corrosion factors are also included in the form of random
variables (such as coating life, corrosion rate, and thickness margin) so as to develop more accurate and precise corrosion models
[176]. Southwell was the first researcher who proposed two linear and bilinear corrosion models for steel structures [166]. These
models are given by:
- Southwell’s linear model:

dðtÞ ¼ 0:076 þ 0:038t ; (1)
where t is the time period or exposure time and d(t) represents the corrosion thickness.
- Southwell’s bi-linear model:

�
0:09t ; 0 � t < 1:46y
dðtÞ ¼ ; (2)
0:76 þ 0:038t ; 1:46 � t < 16y
The Southwell models were later improved by Melchers for corrosion prediction of marine structures [177]. These corrosion
prediction models are given below.
- Melchers-Southwell’s non-linear model:
dðtÞ ¼ 0:84t0:823 ; (3)
- Melchers’ tri-linear model:

8
< 0:170t ; 0�t<y
dðtÞ ¼ 0:152 þ 0:0186t ; 1 � t < 8y ; (4)
:
0:364 þ 0:083t ; 8 � t < 16y
- Melchers’ power law model:
dðtÞ ¼ 0:1207t0:6257 ; (5)

The ship hull and offshore structures are often applied with various protective measures (metallic and non-metallic paints including
antifouling paints) as corrosion shields. Therefore, some researchers have divided the corrosion process on marine structures into three
phases: (i) no corrosion phase or coating (T0 or Tc), (ii) transition between no corrosion and corrosion initiation (Tt), and (iii) the
progress of corrosion (T). In the first phase, it is assumed that no corrosion occurs due to the protective coating being applied on
structures; while in the second phase there are slight changes in different models. Some of the most popular nonlinear corrosion models
are presented in below:
- Yamamoto-Ikegami’s non-linear corrosion model [115].
12
dðtÞ ¼ C1 ðt T0 Tt ÞC2 ; (6)
where C1 and C2 are corrosion constants, T0 is no corrosion zone during which the durability of protective coating is assumed to remain
intact, and Tt is the transition period between coating durability and corrosion initiation.
- Paik’s nonlinear model [175]:
dðtÞ ¼ c1 ðt Tcl Þc2 ; (7)
where c1 and c2 are fixed coefficients and Tcl is the life of coating. The coefficient c2 is usually assumed to be 1/3 or 1, while the
coefficient c1 is the symbolic corrosion rate per year. Paik et al. [178] proposed three types of curves for general and localized cor
rosions on ship structures. These include a convex, a concave and a linear model as shown in Fig. 7. The convex curve shows that the
corrosion rate rises initially but it tends to slow down with the increase in exposure time, because of the deposition of corrosion content
on metal surface. This curve is typically applied to marine structures under statically loaded conditions. Alternatively, in the concave
model, the corrosion rate is accelerated with aging. It is considered to be a more suitable trend in structures with dynamic loading
conditions because the corrosion trend generally decreases with the exposure period. Paik’s model is the only model which shows
increment in the corrosion rate with the exposure period, specific to the loaded structures. In some cases, the effects of nutrients and
biological content can also increase the corrosion trend; however, the same has not been considered by the Paik’s model, purely based
on statistical observations.
The implications in Paik’s model have been addressed by the nonlinear model of Soares and Garbatov [106] as given below:
2 � �3
t Tc
6 7
(8)
Tt
dðtÞ ¼ d∞ 41 e 5;
where d∞ is the thickness loss during the long-term corrosion, Tc represents the coating life of metallic structure, Tt is the transition
time (i.e., the period during which the corrosion process initiates). As shown in Fig. 8, the corrosion process in this model is divided
into three phases. During the first phase, no corrosion occurs because of the corrosion protection system. The corrosion begins during
the transition period (Tt) and increases to a certain depth in plate thickness, until it stops at a depth of d∞. The model has been adopted
by numerous studies such as [126,139,179–181].
Qin and Cui [61] proposed a prediction model using Weibull distribution, showing an increase in corrosion rate in the second phase
and a decrease in the third phase. This model is illustrated in Fig. 9. As can be seen, it describes the corrosion process in three stages:
� [0, Tst]: There is no corrosion as the corrosion protection system is completely active,
� [Tst, TA]: Corrosion process begins and the corrosion rate increases linearly,
� [TA, TL]: It associates with general corrosion,
where Tst is the time when corrosion begins, TA is the corrosion accelerating life, TL is the life of corrosion protection system where
general corrosion starts. Generally TL ranges between 2 and 10 years, depending on the quality of the protection layer and severity of
climatic conditions [182–184].
All the aforementioned corrasion models are purely based on statistical principles and theoretical or field experimental data.
Melchers [142,185] is the first to propose a five-phase phenomenological corrosion model (encompassing both short- and long-term
corrosions) for marine steels, as shown in Fig. 10. This model does not consider the surface protection and its age variability factors,
which itself is a complete and complex science with a different scope. The research revealed that the long-term corrosion in marine
Fig. 7. Paik’s corrosion prediction model [178].
13
Fig. 8. Soares and Garbatov’s corrosion prediction model [106].
Fig. 9. Qin and Cui’s corrosion model [61].
Fig. 10. Melchers’ general corrosion model for steel structures [60].
structures is as a result of a complex collaboration between electrochemical process and bacterial colonization in natural (oxygenated)
and anoxic seawaters. A nonlinear corrosion equation was formulated for almost every phase of the corrosion model. The DO con
centration, seawater temperature, and water velocity have been considered to be the main influencing factors, which may exhibit
certain interrelation with the depth of sea [186].
As can be seen in Fig. 10, the first three phases (phases 0, 1 and 2) of the Melchers’ model illustrates the short-term corrosion
pattern and these phases are almost similar to the post Tst or T0 phase in previous models. The uniqueness of Melchers’ model is the
explanation of long-term corrosion mechanism with the demonstration and justification of rapid rise in corrosion rate (phases 3 and 4)
after a stagnated period. This sharp rise is attributed to the involvement of massive biological activities and nutrients in anaerobic
conditions. The stagnation phase of corrosion is generally attributed to the accumulation of corrosion and fouling deposits on the metal
surface, which splits its connection with external and stimulates anaerobic conditions. Furthermore, Melchers later extended his model
for corrosion prediction of other alloys (aluminium and copper alloys) in marine conditions, fresh water as well as in the coastal or
14
atmospheric conditions. For further reading about corrosion prediction models, the readers are referred to Refs. [187–193].
5. Advanced maintenance management techniques
The conventional maintenance management practices in the marine sector are rapidly shifting towards advanced solutions such as
e-maintenance, computerized maintenance management systems (CMMS), and remote SHM. The e-maintenance and CMMS can
provide refined data at the right time to facilitate decision-making for maintenance. From the literature review, several intelligent
techniques, statistical and stochastic analysis tools and MCDA methods were identified that can be used for improved maintenance
management of corroded steel marine structures. Some of these modern asset maintenance techniques include BN, genetic algorithms
(GA), artificial neural network (ANN), deep learning and fuzzy inference systems [194–196].
Numerous mathematical models have also been proposed to predict the complex nonlinear relationships between corrosion rate
and varying environmental conditions [181]. The evolution of advanced modelling and simulation techniques has enabled more
sophistication in corrosion prediction of aging marine structures. Various artificial intelligence (AI) and machine learning (ML) tools
(such as ANN, support vector machine (SVM)) as well as probabilistic techniques (such as BNs, Markov chain, Monte-Carlo simulation)
have been proposed by researchers for corrosion modelling and risk/reliability-based inspection planning of marine structures
[197–199]. The results acquired with the use of these methods have been found to be very promising and more accurate (see Refs. [68,
107,200]). The most prominent methodologies used for marine maintenance and corrosion prediction modelling are briefly discussed
in followings:
5.1. Fault tree analysis (FTA)
FTA is one of the most important analytical methods used for fault identification and reliability assessment of systems/components
[201]. It is a graphic tool comprised of sequential combinations of faults, which can subsequently result in the occurrence of unde
sirable events [202]. A typical fault tree consists of a top event and a set of basic events organized with the logic gates (AND, OR, etc.)
[203]. FTA has been used for both qualitative as well as quantitative reliability analyses in many industries. Various researchers have
used this method individually or in combination with other techniques (such as event tree analysis (ETA), Markov chain Monte-Carlo
(MCMC) and BN) for failure analysis of marine structures/equipment [24,204–206]. Laskowski [207] performed a structural reliability
analysis on the marine diesel engine of a ship and its components using qualitative FTA. Lazakis et al. [19] developed a hybrid
FTA-FMEA strategy for identification of critical systems/subsystems in a marine engine.
5.2. Bayesian network (BN)
BN is a probabilistic graphical method which uses Bayes’ theorem for updating the prior occurrence probability of failures. It
indicates a set of random variables and associated conditional dependencies in form of a directed acyclic graph (DAG), containing a set
of nodes to represent variables and edges to denote probabilistic causal dependence [208]. It involves independent and dependent
variables known as causes and consequences respectively, which are connected via direct arrows pointing from the causes to the
consequences [209,210]. BN signifies the joint probability distribution and it is flexible to perform predictive (forward) as well as
diagnostic (backward) analysis [148]. In recent years, BNs have been extensively used for modelling of corrosion in marine structures
as well as optimising the RBI plans [148,191,211–214]. For an inclusive understanding of BNs, the readers are referred to Refs. [210,
215,216].
5.3. Statistical and stochastic models
Numerous statistical and stochastic techniques have been employed for degradation modelling and maintenance planning of
marine assets. These methods have been instrumental to develop the relations between various dependent/independent process
variables and estimate the likelihood of occurrence of events. The statistical/stochastic techniques that are commonly adopted by
researchers include: multivariate analysis, regression models, Copulas, Markov process, Poisson process, Monte-Carlo simulation,
Cox’s approximation, and Weibull analysis [35,56,118,217]. Detailed deliberation on the maintenance procedures, their planning,
inspection and prediction trends using various statistical models and methodologies are explained in details in Refs. [33,66,218–220].
5.4. Multi-criteria decision analysis (MCDA)
The MCDA techniques have gained a huge momentum in decision making for the selection of an efficient and effective inspection/
maintenance strategy. This approach comprises a finite set of alternatives (i.e. maintenance strategies) amongst which the decision-
makers have to select, evaluate or rank, in accordance with the weights of a finite set of criteria (attributes). Each substitute is
given an evaluation rating using a suitable measure followed by the aggregation process to acquire the prioritized alternatives from the
best to the worst [221]. The simple additive weighting (SAW), AHP, analytic network process (ANP), TOPSIS, PROMETHEE and the
elimination and choice translating reality (ELECTRE), etc. are some MCDA methods used in maintenance management [222]. Several
research studies on maintenance strategy selection using MCDA techniques were reviewed in Ref. [223].
15
5.5. Artificial intelligence (AI) and machine learning (ML)
AI models and ML techniques have been used as a revolutionary tool in the corrosion and fatigue modelling as well as the opti
mization of risk/reliability-based maintenance [19,181]. They require certain input parameters which are processed through single or
multiple layers to generate outputs. These methods are sometimes also known as Soft Computing Techniques [55]. Some commonly
adopted AI techniques are ANN, fuzzy logic, SVM and GA. Recently, Shirazi and Mohammadi [187] formulated a hybrid intelligent
model to predict the corrosion rate of 3C steel using ANN and swarm particle optimization (PSO).
A detailed distribution of the journal papers by methodologies used to model marine corrosion and maintenance strategy is shown
in Table 7.
6. Discussion and analysis
Over the past few decades, numerous maintenance procedures have evolved for an optimal management of physical equipment and
effective planning of inspections to reduce cost and/or risk of failure. Literature reveals that the marine asset maintenance practices
started from conventional RTF concept and then shifted towards time-based PM in early 1960s. The PM concept is still the most widely
used maintenance strategy in the commercial maritime industry. However, in recent years, some advanced strategies such as CBM,
RCM/RBI and CBM plus have been adopted as alternative strategies to achieve maximum system/subsystem availability/reliability
with minimal cost, failure risks, manpower and material resources. The advancements in failure sensing equipment and data analytics
approaches have provided superior platforms to inculcate improved online and offline health monitoring techniques. The generic
concept of reliability-based maintenance has become more effective and optimised by digitalised revolution in marine maintenance
industry and its integration with some other sophisticated tools such as NDT and SHM.
This review study has primarily focused on the corrosion aspects of submerged/partially submerged marine and ship structures.
The environmental conditions considered in this study were mainly the seawater composition (chemistry), physical factors (such as
temperature) and amalgamated pollutants from various domestic, agricultural and industrial sources into the seawater, which tend to
affect the ratio of intrinsic seawater constituencies, especially in the coastal seawaters. The reviewed literature revealed that the
degradation of marine structures due to uniform and localized corrosions is far more than all other type of corrosion. Since the marine
corrosion is known as a highly nonlinear process during the long exposures due to the involvement of numerous dependent/inde
pendent variables, a multidisciplinary knowledge of material science, structural mechanics, electrochemistry, topography, and hy
drodynamic is required.
The corrosion prediction models developed up to date are subject to several limitations because of the complexities involved in
understanding of the relationship between environmental factors and corrosion rate. Many researchers have highlighted the variation
Table 7
Distribution of papers by methodologies for corrosion modelling and marine maintenance.
Ref Type of structure Methodology for assessment
BN AI/ANN FTA MCDM FEA FMEA HYBRID
[5] Ship ✓
[6] Ship ✓
[12] Ship ✓ ✓
[18] Ship ✓
[19] Ship ✓ ✓ ✓
[23] Offshore wind turbine ✓ ✓ ✓ ✓
[28] Ship ✓
[42] Maintenance ✓
[46] Offshore wind turbine ✓ ✓
[56] Oil and gas ✓
[67] Ship structure ✓
[75] Ship machinery ✓
[76] Offshore wind turbine ✓
[77] Marine structures ✓ ✓
[114] Marine structures ✓
[175] Ship ✓
[187] Offshore structures ✓ ✓
[190] Underground pipelines ✓
[193] Marine ✓
16
in corrosion behaviour in the various zones above and below the seawater surface; however, it has been agreed that higher corrosion
rates are generally found in the splash zone, mean lower tidal region, and just below the low-tide level, respectively. The corrosion
phenomenon in the latter region is known as accelerated low water corrosion (ALWC) which is more common in the pollutant near-
coast seawaters and generally is attributed to the high presence of bacterial activities, and high DIN content [225]. Some researchers
have also attributed the high corrosion to the formation of local galvanic cells due to the difference in corrosion potential in high and
low aerated zones, just below the water line [226].
The water temperature, DO, salinity, water velocity, pH and biological activity are found to be the most influencing factors in
corrosion of marine steel structures. Both laboratory and experimental based research studies have concluded that corrosion initiates
rapidly within hours of immersion in seawater. However, there is a continuous variation in corrosion rate with the rise in exposure
duration and the rate of corrosion stagnates during the diffusion phase of Melchers’ modal, prior initiation of biological activity led by
anaerobic conditions. The long-term corrosion mostly comes into play during anaerobic conditions with subsequent involvement of
nutrients, SRB activity, MIC, biofoulings. Due to highly nonlinearity in marine corrosion process, the prediction of long-term corrosion
based on the short-term corrosion data is not recommended. Moreover, the field experiment results in comparatively larger corrosion
losses than the simulated laboratory-based experiments using artificial seawaters, probably due to absence of biological corrosion
factors in the controlled laboratory environments and higher variability in the influential corrosion parameters during the field
experiments.
It has been deduced that certain interdependent relationships exist between some prominent environmental contributors, which
further complicate corrosion mechanism in marine conditions. The DO in water generally tends to accelerate corrosion rate by rapid
oxidation, however its concentration declines with the rise in temperature. Similarly, salinity goes high in warm seawaters and DO
decreases in these conditions. A significant rate of corrosion has been reported with the increase in seawater velocity but it slows down
with extended exposure durations because of the adhesion of marine growth and corrosion deposits on metallic skin. The DO and pH
values of coastal seawater decrease and become more acidic with the influx of effluents and nutrients. Moreover, the corrosion rate in
cold seawaters is found to be far less than the seawaters of hot countryside, because of the direct relationship between the corrosion
rate and seawater temperature.
Although the changing climatic conditions across the globe are found to be highly effective in dictating corrosion rate of marine
assets, an amalgamation of pollutants, various industrial/agricultural wastes, heavy metals and effluents near coast regions further
complicates the understanding and modelling of corrosion process. The final product formed after incorporation of these run-offs in
seawater becomes highly detrimental towards structural deterioration. Therefore, it has been recorded from the literature that the
severity of ambient conditions in harbours and coastal regions flooded with wastewater addition is more detrimental towards
corrosion than the open seawater environments. It also implies that the installed marine assets (such as wind energy, oil rigs, and
harbour infrastructures) and vessels stationed for long durations in pollutant mixed harbour or coastal areas may experience more
rapid deterioration than seagoing vessels or fixed platforms, away from coastal/harbour areas.
The marine structures are protected from corrosion using various organic/nonorganic coatings as well as other protection methods.
The life expectancy of the protective coating has been reported to range between 3 and 5 years, depending on the severity of climatic
conditions, seawater chemistry, nature of pollutant contamination, etc. Some corrosion prediction models have been developed based
on the assumption that no corrosion takes place as long as the protective coating is intact. Corrosion process is believed to kick on with
the fracture initiation in the protective coating. The paint-fracturing phenomenon may also result in highly localized corrosion as the
exposed bare metal acts as an anode, while the remaining protected areas act as a cathode. Similarly, several corrosion models have
been developed based on historical data from ship structures applied with the protective surface coatings as well as other corrosion
protection measures such as sacrificial zinc anodes and ICCP system. Therefore, it implies that actual corrosion rates are much higher
in the bare surface metal. Hence, these models may underestimate the actual corrosion losses in the absence of any of the protective
measures. Secondly, majority of the corrosion models in the literature have purely been developed statistically based on experimental
data, which do not have any link with the theoretical knowledge of electrochemistry. Therefore, these empirical and mathematical
models may have several limitations, particularly in the seawater with higher pollutant content where corrosion rates are mainly led by
the biological activities and nutrients and sulphide content. The basic phenomenological corrosion models of Melchers [142,185] have
the ability to correlate the various phases of corrosion and variability in the environmental factors with the scientific knowledge on
corrosion and electrochemistry.
It can be deduced from the detailed literature review that despite the enormous research studies on how to model the corrosion in
steel structures, substantial uncertainties still exist because of the involvement of various potential contributors and their complex
relation with the rate of corrosion. Therefore, there is still room for further improvement in corrosion modelling accuracy. Recently,
various AI and ML algorithms (such as BN, ANN and SVM) have been successfully used to model the corrosion process with involving
all influential environmental factors such as temperature, DO, pH level, salinity level, SRBs, etc. Inculcation of the sophisticated and
rational digital technologies, such as big data, Internet of Things (IOT), AI, and digital twins can significantly improve the accuracy of
corrosion process modelling.
From literature review, it has been revealed that the maritime sector is still more reliant on time-based PM concept. The CBM has
also been used in recent years as part of the PM strategy. In the shipping industry, the RCM and CBM plus approaches have been widely
adopted for the maintenance of naval ships. Advanced technology driven fault diagnosis and prognosis, SHM, remote maintenance and
e-maintenance technologies have provided great opportunities for the marine industry to adopt more efficient and optimised main
tenance procedures. Using MCDAs, integrated maintenance methodologies, advanced sensing technologies, realistic prediction
modelling for structural degradation mechanisms can be instrumental to develop data-driven, risk-based maintenance plans for the
marine assets operating in extreme environmental conditions.
17
7. Conclusion and future works
The aim of this review paper was to analyse the effect of marine environmental conditions on corrosion-based degradation of steel
structures. It also highlighted the prognostic models on marine corrosion phenomenon and its impact on the reliability, health
assessment, inspection intervals and overall maintenance strategy selection of assets. Due to significant variability of environmental
factors, the corrosion in marine steel structures shows a great variation in different immersion zones. Hence, it is necessary to update
corrosion models or their parameters according to the metal loss in different immersion zones, phases of corrosion, compositions of
seawater, geographical regions, etc. Subsequently, it warrants a dynamic approach for inspection/maintenance planning of marine
assets, capable of updating its interval according to the severity of climatic conditions by dynamic degradation prediction models
integrated with the online/offline SHM tools.
In this review paper, the following conclusions have been deduced regarding the impact of environmental factors on corrosion
mechanism, its complexities in marine steel structures, and the challenges associated with their maintenance:
� In natural seawater conditions, the sensitivity of environmental factors towards structural corrosion is fairly complex. This is partly
because of the complicated relationships between the coexisting factors (e.g. temperature, DO concentration, salinity level, pH
level, velocity, etc.) and corrosion rate of marine steel structures.
� For the same duration of exposure in natural seawater conditions, sensitivity of sea water temperature towards corrosion is found to
be the most influencing factor. Therefore, the majority of the corrosion models are based on either seawater temperature or
exposure period.
� Variability of temperature in natural seawater around the globe is enormous ( 2 � C–35 � C). Therefore, the corrosion rate tends to
be tremendously higher in hot seawater conditions by the combined effect of high temperature and subsequently higher salinity
level. Although DO concentration tends to reduce with higher seawater temperature, the influence of temperature surpasses the
effect of DO. Subsequently, the rate of structural degradation as well as frequency of inspection and maintenance actions are also
dependent on the extremeness of environmental conditions.
� Seawater salinity and pH level can influence the rate of corrosion; however, under normal sea conditions their sensitivity for
corrosion loss is merely insignificant. Nevertheless, a great variation of seawater salinity has been found in specific hot regions,
where corrosion rates have been reported on the higher side than other regions. In the highly polluted seawaters, pH level tends to
be more acidic (5.5–6.6); therefore, susceptible for accelerated uniform and localized corrosions.
� The influx of effluents causes significant chemical variation in seawater chemistry that can substantially increase the corrosivity
factor by transforming the intrinsic specifications of seawater, such as DO content, salinity, bacterial content, total dissolved solids
(TDS), heavy metal ion concentration, turbidity, and pH level. In addition, presence of nutrients in the form of DIN (compounds of
nitrates, nitrides, ammonia) rapidly increases the rate of biological activity and promotes MIC.
� Influence of pollutants in seawater in the form of DIN, sulphides, etc. on corrosion rate can be far higher than the effect of physical
factors such as seawater temperature.
� The inspection/maintenance of structures in polluted seawater conditions (rich in DIN and sulphides) needs to be more frequent
than that in the nearest natural seawater condition because of the susceptibility for higher corrosion losses in polluted waters. A PM
schedule in the absence of online/offline CM, SHM or prior knowledge of the corrosion rates in the specific climatic conditions will
be likely to fail or ineffective to predict the PF curve for the exposed structure.
� Very few studies have been reported on integration of the outcome of degradation process models (such as corrosion and fatigue
prognostic models) as an input for scheduling/optimising the inspection and maintenance management system (see Ref. [227]).
The degradation of ships and other marine structures is a highly complicated phenomenon, mainly because of their inherent
extreme operating conditions, corrosive environment, and extended operations away from maintenance facilities. The technology-
driven advancement in marine equipment have its own risks and the maintenance demands have been amplified subsequently.
Over the years, various maintenance practices and inspection methods have been adopted in the marine industry in order to attain
higher reliability, safety and maintenance efficiency. It has been observed that the use of prognostic and health management (PHM)
techniques, degradation prediction models, integrated risk- and reliability-based analysis and decision-making techniques have
enhanced the overall maintenance paradigm of marine assets. However, with this need of highly skilled workmanship, the budget
requirement for acquisition of advanced health monitoring technologies has also enhanced accordingly.
Developments in remote sensing and diagnostics, prognostics, SHM, and wireless data transferring methods play a significant role
in the modern day maintenance of marine asset. To some extent, the novel maintenance strategies of RCM and CBM þ have been
adopted by the naval shipping sector. The PM scheme currently holds the highest market share in commercial ship maintenance.
Although a paradigm shift towards more advanced concepts such as CBM, RCM, and RBI has been noticed in recent years, the pace of
this transformation may take several years. It is probably due to the certain precincts of the ship maintenance industry, including the
high initial cost of implementing new strategies and training of operators/maintenance teams, prevailing hired maintenance concept
in offshore energy and shipping sectors.
References
[1] Eruguz AS, Tan T, van Houtum GJ. A survey of maintenance and service logistics management: classification and research agenda from a maritime sector
perspective. Comput Oper Res 2017;85:184–205.
18
[2] Turan O, Olcer AI, Lazakis I, Rigo P, Caprace JD. Maintenance/repair and production-oriented life cycle cost/earning model for ship structural optimisation
during conceptual design stage. Ships Offshore Struct 2009;4:107–25.
[3] Shafiee M, Brennan F, Armada Espinosa I. A parametric whole life cost model for offshore wind farms. Int J Life Cycle Assess 2016;21(7):961–75.
[4] Emovon I. Ship system maintenance strategy selection based on DELPHI-AHP-TOPSIS methodology. World J Eng Technol 2016;4:252–60.
[5] Lazakis I, Olçer
€ A. Selection of the best maintenance approach in the maritime industry under fuzzy multiple attributive group decision-making environment.
J Eng. Marit Environ 2016;230:297–309.
[6] Juri�si�c P, Parunov J, Garbatov Y. Aging effects on ship structural integrity. Brodogr Shipbuild 2017;68:15–28.
[7] Guo J, Wang G, Ivanov L, Perakis AN. Time-varying ultimate strength of aging tanker deck plate considering corrosion effect. Mar Struct 2008;21:402–19.
[8] Lazakis I, Turan O, Aksu S. Increasing ship operational reliability through the implementation of a holistic maintenance management strategy. Ships Offshore
Struct 2010;5:337–57.
[9] International Maritime Organization (IMO). International convention for the prevention of pollution from ships. MARPOL; 1973. Available Online: http://
www.imo.org/en/About/Conventions/ListOfConventions/Pages/International-Convention-for-the-Prevention-of-Pollution-from-Ships-(MARPOL).aspx.
[Accessed 21 November 2019].
[10] International Maritime Organization (IMO). International convention for the safety of life at sea. SOLAS; 1974. Available Online: http://www.imo.org/en/
About/Conventions/ListOfConventions/Pages/International-Convention-for-the-Safety-of-Life-at-Sea-(SOLAS),-1974.aspx. [Accessed 21 November 2019].
[11] United Nations. United nations convention on the law of the sea. 1982. Available Online: http://www.un.org/Depts/los/convention_agreements/texts/unclos/
unclos_e.pdf. [Accessed 21 November 2019].
[12] Royal Institution of Naval Architects (RINA). IMO confirms 2020 date for 0.5% sulphur limit fuels. Shiprepair eNews; 2016. Available Online: https://www.
rina.org.uk/IMO_2020_sulphur_limit_fuels.html. [Accessed 21 November 2019].
[13] Conachey R, Serratella CM, Wang G. Risk-based strategies for the next generation of maintenance and inspection programs. WMU J. Marit. Aff. 2008;7:151–73.
[14] American Bureau of Shipping (ABS). Guide for surveys based on machinery reliability and maintenance techniques. 2016. Available Online: https://ww2.
eagle.org/content/dam/eagle/rules-and-guides/current/survey_and_inspection/121_machineryreliabilitymaintenancetechniques/MRM_Guide_e.pdf.
[Accessed 21 November 2019].
[15] Tomlinson NA. What is the ideal maintenance strategy ? A look at both MoD and commercial shipping best practice. In: Proceedings of the 13th international
naval engineering conference and exhibition; 26-28 April 2016 [Bristol, UK].
[16] Shafiee M. Maintenance strategy selection problem: an MCDM overview. J Qual Mainten Eng 2015;21(4):378–402.
[17] Houshyar A. Reliability and maintainability of machinery and equipment, Part 2: benchmarking, life-cycle cost, and predictive maintenance. Int J Model
Simulat 2005;25(1):1–11.
[18] Goossens AJM, Basten RJI. Exploring maintenance policy selection using the Analytic Hierarchy Process; an application for naval ships. Reliab Eng Syst Saf
2015;142:31–41.
[19] Lazakis I, Raptodimos Y, Varelas T. Predicting ship machinery system condition through analytical reliability tools and artificial neural networks. Ocean Eng
2017;152:404–15.
[20] Selvik JT, Scarf P, Aven T. An extended methodology for risk based inspection planning. Electron J Reliab Risk Anal Theory Appl 2011;2:115–26.
[21] Anantharaman M. Using reliability block diagrams and fault tree circuits, to develop a condition based maintenance model for a vessel’s main propulsion
system and related subsystems. TransNav: Int J Mar Navig Saf. Sea Transp 2013;7:409–13.
[22] Cullum J, Binns J, Lonsdale M, Abbassi R, Garaniya V. Risk-based maintenance scheduling with application to naval vessels and ships. Ocean Eng 2018;148:
476–85.
[23] Shafiee M, Sørensen JD. Maintenance optimization and inspection planning of wind energy assets: models, methods and strategies. Reliab Eng Syst Saf 2019;
192:105993.
[24] International Organization for Standardization (ISO). ISO 13372: condition monitoring and diagnostics of machines — vocabulary; Geneva, Switzerland. 2012.
p. 15. Available Online: https://www.iso.org/standard/52256.html. [Accessed 21 November 2019].
[25] Rausand M, Hoyland A. System reliability theory: models, statistical methods, and applications. second ed. New Jersey, USA: John Wiley & Sons Inc.; 2003,
ISBN 978-0-471-47133-2.
[26] Ahmad R, Kamaruddin S. An overview of time-based and condition-based maintenance in industrial application. Comput Ind Eng 2012;63:135–49.
[27] Jardine AKS, Lin D, Banjevic D. A review on machinery diagnostics and prognostics implementing condition-based maintenance. Mech Syst Signal Process
2006;20:1483–510.
[28] Lazakis I, Dikis K, Michala AL, Theotokatos G. Advanced ship systems condition monitoring for enhanced inspection, maintenance and decision making in ship
operations. Transport Res Procedia 2016;14:1679–88.
[29] Giurgiutiu V. Structural health monitoring with Piezoelectric wafer active sensors. second ed. Sandiago, USA: Academic Press; 2014. p. 1024.
[30] Nowlan FS, Howard HF. Reliability centered maintenance. Springfield, Virginia, USA: U.S. Department of Commerce; 1978.
[31] Cheng Z, Jia X, Gao P, Wu S, Wang J. A framework for intelligent reliability centered maintenance analysis. Reliab Eng Syst Saf 2008;93:806–14.
[32] Johnston DC. Measuring RCM implementation. In: Proceedings of the annual reliability and maintainability symposium. 28-31; January 2002. p. 511–5.
Seattle, WA, USA.
[33] National Aeronautics and Space Administration (NASA). Reliability centered maintenace guide for facilities and collateral equipment; Washington, D.C, USA.
2008. Available Online: https://fred.hq.nasa.gov/Assets/Docs/2015/NASA_RCMGuide.pdf.
[34] Ministry of Defence (MoD). UK. Reliability and maintainability assurance guide Part 3: R & M case. 2016. p. 38. https://standards.globalspec.com/std/
10146255/def-stan-00-42-part-3. Available at:.
[35] Naval Surface Warfare Center (NSWC). Handbook of reliability prediction procedures for mechanical equipment. West Bethesda, Maryland, USA: Logestic
Technology Support Group; 2010. Available at: https://kscddms.ksc.nasa.gov/Reliability/Documents/HandbookofMechanicalReliability.pdf.
[36] Ebrahimi A. Effect analysis of reliability, availability, maintainability and safety (RAMS ) parameters in design and operation of dynamic positioning (DP)
systems in floating offshore structures. MSc Thesis. Stockholm, Sweden: Royal Institute of Technology (KTH); 2010.
[37] Mobley RK. Maintenance engineering handbook. eighth ed. McGraw-Hill Education; 2014. p. 704.
[38] Conachey RM, Montgomery RL. Application of reliability-centered maintenance techniques to the marine industry. ABS Tech. Pap; 2002. p. 34. Available at:
http://citeseerx.ist.psu.edu/viewdoc/download?doi¼10.1.1.447.7222&rep¼rep1&type¼pdf.
[39] Smith AM. Reliability-centered maintenance (RCM). McGraw-Hill; 1993. p. 216.
[40] Department of Defense (DOD). DOD 4151.22-M reliability centered maintenance (RCM); Washington D.C., USA. 2011. Available at: https://www.wbdg.org/
FFC/DOD/DODMAN/415122-M.pdf.
[41] Smith AM, Hinchcliffe GR. RCM–Gateway to world class maintenance. second ed. Oxford: Butterworth-Heinemann; 2004. p. 337.
[42] Selvik JT, Aven T. A framework for reliability and risk centered maintenance. Reliab Eng Syst Saf 2011;96:324–31.
[43] Dawotola A. Risk based maintenance of petrolium pipelines. MSc Thesis. Netherlands: Delft University of Technology; 2012.
[44] Serratella C, Wang G, Tikka K. Risk-based inspection and maintenance of aged structures. In: Paik JK, Melchers RE, editors. Condition assessment of aged
structures. Woodhead Publishing; 2008. p. 487–518.
[45] Dinmohammadi F, Alkali B, Shafiee M, B�erenguer C, Labib A. Risk evaluation of railway rolling stock failures using FMECA technique: a case study of
passenger door system. Urban Rail Transit 2016;2:128–45.
[46] Shafiee M. A fuzzy analytic network process model to mitigate the risks associated with offshore wind farms. Expert Syst Appl 2015;42:2143–52.
[47] DNVGL. DNVGL-RP-G101 Risk based inspection of shore topsides static mechanical equipment. 2002. Høvik, Norway.
[48] API. Risk-based inspection technology. 2008. Washington D.C., USA.
[49] API. Risk-based inspection. 2009. Washington D.C., USA.
19
[50] Millar RC. The role of reliability data bases in deploying CBMþ, RCM and PHM with TLCSM. In: Proceedings of IEEE aerospace conference; 1-8 March 2008
[Big Sky, Montana, USA].
[51] Deaprtment of Defense. Condition based maintenance plus - DoD guidebook. Washington DC, USA. 2008. available at: https://www.dau.edu/guidebooks/
Shared%20Documents%20HTML/Condition%20Based%20Maintenance%20Plus%20(CBMþ)%20Guidebook.aspx.
[52] Sikorska JZ, Hodkiewicz M, Ma L. Prognostic modelling options for remaining useful life estimation by industry. Mech Syst Signal Process 2011;25:1803–36.
[53] Vaidya P, Rausand M. Remaining useful life, technical health, and life extension. Proc Inst Mech Eng Part O J Risk Reliab 2011;225:219–31.
[54] Medjaher K, Tobon-Mejia DA, Zerhouni N. Remaining useful life estimation of critical components with application to bearings. IEEE Trans Reliab 2012;61:
292–302.
[55] Okoh C, Roy R, Mehnen J, Redding L. Overview of remaining useful life prediction techniques in through-life engineering services. In: Proceedings of the 6th
CIRP conference on industrial product-service systems, vol. 16. Ontario, Canada: Windsor; 2014. p. 158–63.
[56] Animah I, Shafiee M. Condition assessment, remaining useful life prediction and life extension decision making for offshore oil and gas assets. J Loss Prev
Process Ind 2018;53:17–28.
[57] Shafiee M, Animah I, Simms N. Development of a techno-economic framework for life extension decision making of safety critical installations. J Loss Prev
Process Ind 2016;44:299–310.
[58] Shafiee M, Animah I. Life extension decision making of safety critical systems: an overview. J Loss Prev Process Ind 2017;47:174–88.
[59] Elsayed EA. Reliability prediction and accelerated testing. In: Kobbacy KA, Murthy DNP, editors. Complex system maintenance handbook, vol. 7. Springer;
2008, ISBN 978-1-84800-010-0. p. 155–78.
[60] Soares CG, Garbatov Y, Zayed A. Effect of environmental factors on steel plate corrosion under marine immersion conditions. Corrosion Eng Sci Technol 2011;
46:524–41.
[61] Qin S, Cui W. Effect of corrosion models on the time-dependent reliability of steel plated elements. Mar Struct 2003;16:15–34.
[62] Melchers RE. Probabilistic models for corrosion in structural reliability assessment—Part 1: empirical models. J Offshore Mech Arctic Eng 2003;125:264.
[63] Melchers RE. Probabilistic models for corrosion in structural reliability assessment—Part 2: models based on mechanics. J Offshore Mech Arctic Eng 2003;125:
272.
[64] Melchers RE. Probabilistic model for marine corrosion of steel for structural reliability assessment. J Struct Eng 2003;129:1484–93.
[65] Melchers RE. Principles of marine corrosion. In: Dhanak R, M, Xiros NI, editors. Ocean engineering. London: Springer; 2016, ISBN 9783319166490. p. 111–23.
[66] Melchers RE. Effect on marine immersion corrosion of carbon content of low alloy steels. Corrosion Sci 2003;45:2609–25.
[67] Soares CG, Garbatov Y, Zayed A, Wang G. Non-linear corrosion model for immersed steel plates accounting for environmental factors. In: Marine technology
conference & expo; New Jersey, USA; 2005. p. 193–211.
[68] Paik JK, Kim DK. Advanced method for the development of an empirical model to predict time-dependent corrosion wastage. Corrosion Sci 2012;63:51–8.
[69] Ventikos NP, Sotiralis P, Drakakis M. A dynamic model for the hull inspection of ships: the analysis and results. Ocean Eng 2018;151:355–65.
[70] Soares CG, Garbatov Y. Reliability assessment of maintained ship hulls with correlated corroded elements. Mar Struct 1997;10:629–53.
[71] Melchers RE. Development of new applied models for steel corrosion in marine applications including shipping. Ships Offshore Struct 2008;3:135–44.
[72] Valdez B, Ramirez J, Eliezer A, Schorr M, Ramos R, Salinas R. Corrosion assessment of infrastructure assets in coastal seas. J Mar Eng Technol 2016;15:124–34.
[73] Garbatov Y, Soares CG. Reliability based maintenance of marine structures. Mar Technol Eng 2011;2:1101–20.
[74] Michala AL, Lazakis I, Theotokatos G, Varelas T. Wireless condition monitoring for ship applications. In: RINA, royal institution of naval architects - smart ship
technology; london, UK; 2016. p. 51–8.
[75] Cicek K, Celik M. Application of failure modes and effects analysis to main engine crankcase explosion failure on-board ship. Saf Sci 2013;51:6–10.
[76] Shafiee M, Dinmohammadi F. An FMEA-based risk assessment approach for wind turbine systems: a comparative study of onshore and offshore. Energies 2014;
7:619–42.
[77] Tang Y, Liu Q, Jing J, Yang Y, Zou Z. A framework for identification of maintenance significant items in reliability centered maintenance. Energy 2017;118:
1295–303.
[78] Mokashi AJ, Wang J, Vermar AK. A study of reliability-centred maintenance in maritime operations. Mar Pol 2002;26:325–35.
[79] Wabakken I. Application of RCM to construct a maintenance program for a maritime vessel. MSc Thesis. Trondheim, Norway: Norwegian University of Science
and Technology; 2015.
[80] Dong Y, Frangopol DM. Risk-informed life-cycle optimum inspection and maintenance of ship structures considering corrosion and fatigue. Ocean Eng 2015;
101:161–71.
[81] Dong Y, Frangopol DM. Incorporation of risk and updating in inspection of fatigue-sensitive details of ship structures. Int J Fatig 2016;82:676–88.
[82] Turan O, Lazakis I, Judah S, Incecik A. Investigating the reliability and criticality of the maintenance characteristics of a diving support vessel. Qual Reliab Eng
Int 2011;27:931–46.
[83] Animah I, Shafiee M, Simms N, Erkoyuncu JA, Maiti J. Selection of the most suitable life extension strategy for ageing offshore assets using a life-cycle cost-
benefit analysis approach. J Qual Mainten Eng 2018;24(3):311–30.
[84] Yeter B, Garbatov Y, Soares CG. Risk-based multi-objective optimisation of a monopile offshore wind turbine support structure. In: Proceedings of the ASME
36th international conference on ocean, offshore and arctic engineering, trondheim, Norway, june 25–30; 2017. p. 1–10.
[85] Nielsen JJ, Sørensen JD. On risk-based operation and maintenance of offshore wind turbine components. Reliab Eng Syst Saf 2011;96:218–29.
[86] Hecht M, An X. A stochastic model for determining inspection intervals for large marine vessels. In: Annual symposium reliability and maintainability, 26-29
jan; 2004. p. 559–64. Los Angeles, CA, USA.
[87] Akpan UO, Koko TS, Ayyub B, Dunbar TE. Risk assessment of aging ship hull structures in the presence of corrosion and fatigue. Mar Struct 2002;15:211–31.
[88] Hamada K, Fujimoto Y, Shintaku E. Ship inspection support system using a product model. J Mar Sci Technol 2002;6:205–15.
[89] Soares CG, Garbatov Y. Reliability of maintained hull girders of two bulk carrier designs subjected to fatigue and corrosion. J Sh Ocean Technol 1999;3(1):
27–41.
[90] Abbas M, Shafiee M. Structural health monitoring (SHM) and determination of surface defects in large metallic structures using ultrasonic guided waves.
Sensors 2018;18(11):26.
[91] Caines S, Khan F, Shirokoff J. Analysis of pitting corrosion on steel under insulation in marine environments. J Loss Prev Process Ind 2013;26:1466–83.
[92] Kros H. Performing detailed level 1 pipeline inspection in deep water with a remotely operated vehicle (ROV). In: Offshore technology conference; houston,
Texas, USA, 2-5 may; 2011. p. 1–11.
[93] Terribile A, Schiavon R, Rossi G, Zampato M, Indrigo D. A remotely operated tanker inspection system (ROTIS). In: Offshore mediterranean conference and
exhibition, 28-30 march; 2007. p. 1–9. Ravenna, Italy.
[94] Ortiz A, Bonnin-Pascual F, Garcia-Fidalgo E, Company JP. Visual inspection of vessels by means of a micro-aerial vehicle: an artificial neural network approach
for corrosion detection. Adv Intell Syst Comput 2016;418:223–34.
[95] Bonnin-Pascual F, Ortiz A. Corrosion detection for automated visual inspection. In: Developments in corrosion protection, Chapter 25. London, UK:
IntechOpen; 2014, ISBN 978-953-51-1223-5. p. 619–32.
[96] Giurgiutiu V, Roman C, Lin B, Frankforter E. Omnidirectional piezo-optical ring sensor for enhanced guided wave structural health monitoring. Smart Mater
Struct 2015;24(1). https://doi.org/10.1088/0964-1726/24/1/015008.
[97] Moheimani SOR, Fleming AJ. Piezoelectric transducers for vibration control and damping. In: Grimble MJ, Ferrara A, editors. Advances in industrial control.
London, UK: Springer; 2006, ISBN 9781846283314.
[98] Carellan I G De, Moustakidis S, Legg M, Dave R, Selcuk C, Jost P, Krause HJ, Seton J, Gan T, Hrissagis K. Characterization of ultrasonic wave propagation in the
application of prevention of fouling on a ship’s hull. In: International conference on maritime technology; 7-9 july; 2014 [Glasgow Scotland].
[99] Moustakidis S, Kappatos V, Karlsson P, Selcuk C, Gan TH, Hrissagis K. An intelligent methodology for railways monitoring using ultrasonic guided waves.
J Nondestr Eval 2014;33:694–710.
20
[100] Ahmed M, Eich M, Bernhard F. Design and control of MIRA: a lightweight climbing robot for ship inspection. In: World symposium on mechatronics
engineering & applied physics; 18-20 june; 2014. p. 58–62. Sousse, Tunisia.
[101] Soares CG, Garbatov Y. Reliability of maintained ship hulls subjected to corrosion and fatigue under combined loading. J Constr Steel Res 1999;52(1):93–115.
[102] Soares CG, Garbatov Y, Zayed A, Wang G. Influence of environmental factors on corrosion of ship structures in marine atmosphere. Corrosion Sci 2009;51:
2014–26.
[103] Hussein AW, Soares CG. Reliability and residual strength of double hull tankers designed according to the new IACS common structural rules. Ocean Eng 2009;
36:1446–59.
[104] Soares CG, Garbatov Y. Reliability of maintained ship hulls subjected to corrosion. J Ship Res 1996;40(3):235–43.
[105] Melchers RE. Modeling and prediction of long-term corrosion of steel in marine environments. Int J Offshore Polar Eng 2012;22(4):7.
[106] Soares G, Garbatov Y. Reliability of maintained, corrosion protected plates subjected to non-linear corrosion and compressive loads. Mar Struct 1999;12:
425–45.
[107] Khedmati MR, Nouri ZHME, Roshanali MM. A comparative computational investigation on the effects of randomly distributed general corrosion on the post-
buckling behaviour of uniaxially loaded plates. J Mech Sci Technol 2012;26:767–83.
[108] Bhandari J, Khan F, Abbassi R, Garaniya V, Ojeda R. Modelling of pitting corrosion in marine and offshore steel structures - a technical review. J Loss Prev
Process Ind 2015;37:39–62.
[109] Melchers RE. Transient early and longer term influence of bacteria on marine corrosion of steel. Corrosion Eng Sci Technol 2010;45:257–61.
[110] Wang Y, Wharton JA, Shenoi RA. Influence of localised pit distribution and bench-shape pits on the ultimate compressive strength of steel plating for shipping.
Corrosion 2014;70(9):915–27.
[111] Fontana MG. Corrosion engineering. third ed. New York, USA: McGraw Hill Education; 2005, ISBN 0070214638.
[112] British Standards Institution (BSI). BS EN ISO 11306: corrosion of metals and alloys - guidelines for exposing and evaluating metals and alloys in surface sea
water. 1998. London, UK.
[113] Hifi N. Decision support system for risk-based inspection and maintenance planning for ship hull structures. PhD thesis. University of Strathclyde; 2013.
[114] Rahmdel S, Kim K, Kim S, Park S. A novel stepwise method to predict ultimate strength reduction in offshore structures with pitting corrosion. Adv Mech Eng
2015;7:1–10.
[115] Yamamoto N, Ikegami K. A study on the degradation of coating and corrosion of ship’ s hull based on the probabilistic approach. J Offshore Mech Arctic Eng
1998;120:121–8.
[116] Engelhardt G, Macdonald DD. Unification of the deterministic and statistical approaches for predicting localized corrosion damage. I. Theoretical foundation.
Corrosion Sci 2004;46(11):2755–80.
[117] Melchers RE. The marine corrosion of structural steels in brackish and fresh waters. Struct Infrastruct Eng 2006;2:53–61.
[118] Bhandari J, Khan F, Abbassi R, Garaniya V, Ojeda R. Pitting degradation modeling of ocean steel structures using Bayesian network. J Offshore Mech Arctic
Eng 2017;139(5):11.
[119] Paik JK, Kim SK, Lee SK. Probabilistic corrosion rate estimation model for longitudinal strength members of bulk carriers. Ocean Eng 1998;25:837–60.
[120] Melchers RE. Probabilistic models for corrosion in structural reliability assessment—Part 1: empirical models. J Offshore Mech Arctic Eng 2003;125(4):
264–71.
[121] Morcillo M, Chico B, de la Fuente D, Almeida E, Joseph G, Rivero S, Rosales B. Atmospheric corrosion of reference metals in Antarctic sites. Cold Reg Sci
Technol 2004;40:165–78.
[122] Zise W, Chunchun X, Xia C, Ben X. The morphology, phase composition and effect of corrosion product on simulated archaeological iron. Chin J Chem Eng
2007;15(3):433–8.
[123] Khan MI, Bano H, Khan HTS, Mahmood A, Kazmi SA. Atmospheric corrosion kinetics and dynamics of Karachi onshore areas. J Chem Soc Pakistan 2015;37(1):
179–89.
[124] Melchers RE. The effect of corrosion on the structural reliability of steel offshore structures. Corrosion Sci 2005;47:2391–410.
[125] Melchers RE. Statistical characterization of pitting corrosion - Part 2: probabilistic modeling for maximum pit depth. Corrosion 2005;61:766–77.
[126] Zayed A, Garbatov Y, Guedes Soares C. Corrosion degradation of ship hull steel plates accounting for local environmental conditions. Ocean Eng 2018;163:
299–306.
[127] Gu J-D, Ford TE, Mitchell R. Microbial degradation of materials: general processes. In: Revie RW, editor. Uhlig’s corrosion handbook the electrochemical
society series. John Wiley and Sons Inc.; 2011, ISBN 9780470080320. p. 1–20. Pennington, NJ, USA.
[128] Jones DA. Principles and prevention of corrosion. second ed. London, UK: Pearson Education; 2001, ISBN 0133599930. p. 592.
[129] Melchers RE. Effect of small compositional changes on marine immersion corrosion of low alloy steels. Corrosion Sci 2004;46:1669–91.
[130] Kalogirou SA. Seawater desalination using renewable energy sources. Prog Energy Combust Sci 2005;31:242–81.
[131] Nergis Y, Sharif M, Choudhry AF, Hussain A, Butt JA. Impact of industrial and sewage effluents on Karachi coastal water and sediment quality. Middle-East.
J Sci Res 2012;11:1443–54.
[132] Jamil I, Bano H, Castano JG, Mahmood A. Characterization of atmospheric corrosion near the coastal areas of Arabian Sea. Mater Corros 2018;69(7):898–907.
[133] British Standards Institution (BSI). BS EN ISO 9223: corrosion of metals and alloys — corrosivity of atmospheres — classification, determination and
estimation; London. 2012. Available Online: https://shop.bsigroup.com/ProductDetail/?pid¼000000000030209288.
[134] Peng L, Stewart MG, Melchers RE. Corrosion and capacity prediction of marine steel infrastructure under a changing environment. Struct Infrastruct Eng 2017;
13:988–1001.
[135] Jilani S. Present pollution profile of Karachi coastal waters. J Coast Conserv 2018;22:325–32.
[136] Wiener MS, Salas BV, Quintero-Nú~ nez M, Zlatev R. Effect of H2S on corrosion in polluted waters: a review. Corrosion Eng Sci Technol 2006;41:221–7.
[137] Al-Thubaiti MA, Hodgkiess T, Ho SYK. Environmental influences on the vapourside corrosion of copper-nickel alloys. Desalination 2005;183:195–202.
[138] Zayed A, Garbatov Y, Soares CG, Wang G. Environmental factors affecting the time dependent corrosion wastage of marine structures. Marit Transp 2005;1:
589–98.
[139] Soares CG, Garbatov Y, Zayed A, Wang G. Corrosion wastage model for ship crude oil tanks. Corrosion Sci 2008;50:3095–106.
[140] Melchers RE. Examples of mathematical modelling of long term general corrosion of structural steels in sea water. Corrosion Eng Sci Technol 2006;41:38–44.
[141] Melchers RE. Effect of temperature on the marine immersion corrosion of carbon steels. Corrosion Sci 2002;58:768–82.
[142] Melchers RE. Modeling of marine immersion corrosion for mild and low-alloy steels — Part 1 : phenomenological model. Corrosion Sci 2003;59:319–34.
[143] Ijsseling FP. General guidelines for corrosion testing of materials for marine applications: literature review on sea water as test environment. Br Corrosion J
1989;24:53–78.
[144] Chandler KA. Marine and offshore corrosion. London, UK: Butterworth-Heinemann; 1985, ISBN 0408011750.
[145] Venkatesan R, Venkatasamy MA, Bhaskaran TA, Dwarakadasa ES, Ravindran M. Corrosion of ferrous alloys in deep sea environments. Br Corrosion J 2002;37:
257–66.
[146] Melchers RE. Microbiological and abiotic processes in modelling longer-term marine corrosion of steel. Bioelectrochemistry 2014;97:89–96.
[147] Melchers RE, Jeffrey RJ. Long-term corrosion of mild steel in natural and UV-treated coastal seawater. Corrosion 2014;70:804–18.
[148] Bhandari, J., Khan, F., Abbassi, R., Garaniya, V., Ojeda, R. Reliability assessment of offshore asset under pitting corrosion using Bayesian Network. In: NACE
corrosion conference; 6-10 march 2016, vancouver, British columbia, Canada, pp. 1–15.
[149] Malik AU, Ahmad S, Andijani I. Corrosion behavior of steels in gulf sea water environment. Desalination 1999;123:205–13.
[150] Zakowski K, Narozny M, Szocinski M, Darowicki K. Influence of water salinity on corrosion risk - the case of the southern Baltic Sea coast. Environ Monit Assess
2014;186:4871–9.
[151] Aromaa J, Fors�en O. Factors affecting corrosion in Gulf of Finland brackish water. Int J Electrochem 2016:9. Article ID 3720280.
[152] Mcneill LS. The importance of temperature in assessing iron pipe corrosion in water distribution systems. Environ Monit Assess 2002;77:229–42.
21
[153] Traverso P, Canepa E. A review of studies on corrosion of metals and alloys in deep-sea environment. Ocean Eng 2014;87:10–5.
[154] Venkatesan R, Dwarakadasa ES, Ravindran M. Biofilm formation on structural materials in deep sea environments. Indian J Eng Mater Sci 2003;10:486–91.
[155] Melchers RE, Jeffrey R. Corrosion of long vertical steel strips in the marine tidal zone and implications for ALWC. Corrosion Sci 2012;65:26–36.
[156] Taleb-Berrouane M, Khan F, Hawboldt K, Eckert R, Skovhus TL. Model for microbiologically influenced corrosion potential assessment for the oil and gas
industry. Corrosion Eng Sci Technol 2018;53:378–92.
[157] Melchers RE. Influence of dissolved inorganic nitrogen on accelerated low water corrosion of marine steel piling. Corrosion 2013;69:95–103.
[158] Wang X, Melchers RE. Corrosion of carbon steel in presence of mixed deposits under stagnant seawater conditions. J Loss Prev Process Ind 2017;45:29–42.
[159] Melchers RE, Jeffrey R. Influence of water velocity on marine immersion corrosion of mild steel. Corrosion 2004;60(1):11.
[160] Melchers RE. Mathematical modeling of the effect of water velocity on the marine immersion corrosion of mild steel coupons. Corrosion 2004;60(5):8.
[161] Melchers RE. Effect of nutrient-based water pollution on the corrosion of mild steel in marine immersion conditions. Corrosion 2005;61:237–45.
[162] Jingjun L, Yuzhen L, Xiaoyu L. Numerical simulation for carbon steel flow-induced corrosion in high-velocity flow seawater. Anticorros Methods Mater 2008;
55:66–72.
[163] Melchers RE, Jeffrey R. Early corrosion of mild steel in seawater. Corrosion Sci 2005;47(7):1678–93.
[164] Li SX, Akid R. Corrosion fatigue life prediction of a steel shaft material in seawater. Eng Fail Anal 2013;34:324–34.
[165] Hansom JD, Barltrop NDP, Hall AM. Modelling the processes of cliff-top erosion and deposition under extreme storm waves. Mar Geol 2008;253(1–2):36–50.
[166] Schumacher M. Seawater corrosion handbook. Noyes Data Corp.; 1979, ISBN 0815507364. p. 494.
[167] Melchers RE, Jeffrey R. The critical involvement of anaerobic bacterial activity in modelling the corrosion behaviour of mild steel in marine environments.
Electrochim Acta 2008;54(1):80–5.
[168] Vhanmane S, Bhattacharya B. Ultimate strength analysis of ship hull girder under random material and geometric properties. J Offshore Mech Arctic Eng 2011;
133(3):031602 (8 pages).
[169] Melchers RE. The effects of water pollution on the immersion corrosion of mild and low alloy steels. Corrosion Sci 2007;49:3149–67.
[170] Melchers RE. Long-term immersion corrosion of steels in seawaters with elevated nutrient concentration. Corrosion Sci 2014;81:110–6.
[171] Habib K, Fakhral-Deen A. Risk assessment and evaluation of materials commonly used in desalination plants subjected to pollution impact of the oil spill and
oil fires in marine environment. Desalination 2001;139(1–3):249–53.
[172] Pedersen A, Hernandez-Duque G, Thierry D, Hermansson M. Effects of biofilms on metal corrosion. In: Sequeira CAC, Tiller AK, editors. Microbial corrosion,
proceedings of the international EFC workshop on microbial corrosion. London, UK: The Institute of Materials; 1992, ISBN 0901716081.
[173] Mashiatullah A, Qureshi RM, Ahmad N, Khalid F, Javed T. Physico–chemical and biological water quality of Karachi coastal water. Nucleus 2009;46(1–2):
53–9.
[174] Shafiee M, Ayudiani PS. Development of a risk-based integrity model for offshore energy infrastructures - application to oil and gas pipelines. Int J Process Syst
Eng 2016;3(4):211–31.
[175] Paik JK, Thayamballi AK, Park Y, Hwang JS. A time-dependent corrosion wastage model for seawater ballast tank structures of ships. Corrosion Sci 2004;46:
471–86.
[176] Luque, J.; Hamann, R.; Straub, D. Spatial model for corrosion in ships and FPSOs. In Proceedings of the ASME 33rd international conference on ocean, offshore
and arctic engineering; june 8–13, 2014, San Francisco, California, USA, 11 pages.
[177] Melchers RE. Corrosion uncertainty modelling for steel structures. J Constr Steel Res 1999;52(1):3–19.
[178] Paik JK, Jae L, Joon H, Young P. A time-dependent corrosion wastage model for the structures of single and double hull tankers and FSOs and FPSOs. Mar
Technol 2003;40(3):201–17.
[179] Silva JE, Garbatov Y, Soares CG. Reliability assessment of a steel plate subjected to distributed and localized corrosion wastage. Eng Struct 2014;59:13–20.
[180] Zayed AA, Garbatov YY, Soares CG. Reliability of ship hulls subjected to corrosion and maintenance. Struct Saf 2013;43:1–11.
[181] Wang Y, Wharton JA, Shenoi RA. Ultimate strength analysis of aged steel-plated structures exposed to marine corrosion damage: a review. Corrosion Sci 2014;
86:42–60.
[182] Qin S, Cui W. A discussion of the ultimate strength of ageing ships, with particular reference to the corrosion model. Proc Inst Mech Eng, Part M: J Eng Marit
Environ 2002;216(2):155–60.
[183] Qin S, Cui W. A new corrosion model for the deterioration of steel structures in marine environments. In: 1st int. ASRANet colloq; 8-10 July 2002. p. 9.
Glasgow, UK.
[184] Qin S, Cui W. A discussion of the ultimate strength of ageing ships, with particular reference to the corrosion model. J Eng Marit Environ 2015;216:155–60.
[185] Melchers RE. Modeling of marine corrosion of steel specimens. In: W, Kain R, editors. Corrosion Testing in natural waters: second volume; young. West
Conshohocken, Pennsylvania, USA: ASTM International; 1997. p. 20–33.
[186] Melchers RE, Jeffrey R. Surface “roughness” effect on marine immersion corrosion of mild steel. Corrosion 2004;60(7):697–703.
[187] Shirazi AZ, Mohammadi Z. A hybrid intelligent model combining ANN and imperialist competitive algorithm for prediction of corrosion rate in 3C steel under
seawater environment. Neural Comput Appl 2017;28(11):3455–64.
[188] Alc�antara J, Chico B, Díaz I, de la Fuente D, Morcillo M. Airborne chloride deposit and its effect on marine atmospheric corrosion of mild steel. Corrosion Sci
2015;97:74–88.
[189] Sun B, Ye T, Feng Q, Yao J, Wei M. Accelerated degradation test and predictive failure analysis of B10 Copper-Nickel alloy under marine environmental
conditions. Materials 2015;8(9):6029–42.
[190] Wang H, Yajima A, Liang RY, Castaneda H. Bayesian modeling of external corrosion in underground pipelines based on the integration of Markov chain Monte
Carlo techniques and clustered inspection data. Comput Civ Infrastruct Eng 2015;30:300–16.
[191] de Farias BV, Netto TA. FPSO hull structural integrity evaluation via Bayesian updating of inspection data. Ocean Eng 2012;56:10–9.
[192] Cui W, Wang F, Huang X. A unified fatigue life prediction method for marine structures. Mar Struct 2011;24:153–81.
[193] Ling, W., Dong-Mei, F. A novel approach using SVR ensembles for minor prototypes prediction of seawater corrosion rate. In: Second international workshop
on computer science and engineering, 28–30 oct. 2009, qingdao, China.
[194] Cui J, Wang D, Ma N. Case studies on the probabilistic characteristics of ultimate strength of stiffened panels with uniform and non-uniform localized corrosion
subjected to uniaxial and biaxial thrust. Int J Nav Archit Ocean Eng 2019;11(1):97–118.
[195] Shabarchin O, Tesfamariam S. Internal corrosion hazard assessment of oil & gas pipelines using Bayesian belief network model. J Loss Prev Process Ind 2016;
40:479–95.
[196] Garbatov Y, Soares CG. Bayesian updating in the reliability assessment of maintained floating structures. J Offshore Mech Arctic Eng 2002;124(3):139–45.
[197] Valor A, Caleyo F, Alfonso L, Vel� azquez JC, Hallen JM. Markov chain models for the stochastic modeling of pitting corrosion. Mathemaical Probl. Eng. 2013;
13.
[198] Caleyo F, Vel� azquez JC, Valor A, Hallen JM. Markov chain modelling of pitting corrosion in underground pipelines. Corrosion Sci 2009;51:2197–207.
[199] Zhang Y, Kim C-W, Tee KF. Maintenance management of offshore structures using Markov process model with random transition probabilities. Struct
Infrastruct Eng 2017;13:1068–80.
[200] Baz�an FAV, Beck AT. Stochastic process corrosion growth models for pipeline reliability. Corrosion Sci 2013;74:50–8.
[201] Shafiee M, Enjema E, Kolios A. An integrated FTA-FMEA model for risk analysis of engineering systems: a case study of subsea blowout preventers. Appl Sci
2019;9(6). Article No. 1192.
[202] Shafiee M, Animah I, Alkali B, Baglee D. Decision support methods and applications in the upstream oil and gas sector. J Petrol Sci Eng 2019;173:1173–86.
[203] Vesely WE, Goldberg FF, Roberts NH, Haasl DF. fault tree handbook. Washington, D.C., USA: U.S. Nuclear Regulatory Commission; 1981.
[204] Atehnjia DN, Zaili Y, Wang J. Application of fault tree-Bayesian network for graving dock gate failure analysis. Int J Adv Sci Res Eng 2018;4(1):27–37.
[205] Khakzad N, Khan F, Amyotte P. Safety analysis in process facilities: comparison of fault tree and Bayesian network approaches. Reliab Eng Syst Saf 2011;96:
925–32.
22
[206] Choi I-H, Chang D. Reliability and availability assessment of seabed storage tanks using fault tree analysis. Ocean Eng 2016;120:1–14.
[207] Laskowski R. Fault tree analysis as a tool for modelling the marine main engine reliability structure. Sci J Marit Univ Szczecin 2015;41:71–7.
[208] Li KX, Yin J, Bang HS, Yang Z, Wang J. Bayesian network with quantitative input for maritime risk analysis. Transp A Transp Sci 2014;10:89–118.
[209] Gelman A, Carlin JB, Stern HS, Dunson DB, Vehtari A, Rubin DB. Bayesian data analysis. third ed. Boca Raton, Florida, USA: Chapman and Hall/CRC; 2013,
ISBN 978-1439840955.
[210] Nielsen TD, Jensen FV. Bayesian networks and decision graphs. New York, USA: Springer-Verlag; 2007, ISBN 978-1-4419-2394-3.
[211] Caleyo F, Valor A, Alfonso L, Vidal J, Perez-Baruch E, Hallen JM. Bayesian analysis of external corrosion data of non-piggable underground pipelines.
Corrosion Sci 2015;90:33–45.
[212] Pui G, Bhandari J, Arzaghi E, Abbassi R, Garaniya V. Risk-based maintenance of offshore managed pressure drilling (MPD) operation. J Petrol Sci Eng 2017;
159:513–21.
[213] Abbassi R, Bhandari J, Khan F, Garaniya V, Chai S. Developing a quantitative risk-based methodology for maintenance scheduling using Bayesian Network.
Chem Eng Trans 2016;48:235–40.
[214] Enjema E, Shafiee M, Kolios A. A study on the reliability of oil and gas Blowout Preventer (BOP) technologies under deep-water erratic conditions. In: Safety
and reliability – theory and applications. CRC Press, Taylor & Francis Group; 2017. https://doi.org/10.1201/9781315210469-302. 346-346.
[215] Kjærulff UB, Madsen AL. In: Jordan M, Nowak R, SchOlkopf B, editors. Bayesian networks and influence diagrams: a guide to construction and analysis. New
York: Springer-Verlag; 2008. p. 318.
[216] Xu Y, Choi J, Dass S, Maiti T, Başar T, Bicchi A. In: Krstic M, editor. Bayesian prediction and adaptive sampling algorithms for mobile sensor networks.
Heidelberg, Germany: Springer International Publishing; 2016.
[217] Si XS, Wang W, Hu CH, Zhou DH. Remaining useful life estimation - a review on the statistical data driven approaches. Eur J Oper Res 2011;213(1):1–14.
[218] Dentcheva D. Optimization models with probabilistic constraints. In: Calafiore G, Dabbene F, editors. Probablistic and randomized methods for design and
uncertanity. London, UK: Springer; 2006.
[219] Kvam P, Lu J-C. Statistical reliability with applications. In: Pham H, editor. Engineering statistics. London, UK: Springer; 2006. p. 49–60.
[220] US Department of Defence. MIL-HDBK-189C: handbook reliability growth management. 2011. Available at: http://www.barringer1.com/mil_files/MIL-HDBK-
189C.pdf.
[221] Shafiee M. Maintenance logistics organization for offshore wind energy: current progress and future perspectives. Renew Energy 2015;77:182–93.
[222] Emovon I, Norman RA, Murphy AJ. Hybrid MCDM based methodology for selecting the optimum maintenance strategy for ship machinery systems. J Intell
Manuf 2018;29(3):519–31.
[223] Shafiee M. Maintenance strategy selection problem: an MCDM overview. J Qual Mainten Eng 2015;21:378–402.
[224] Emovon I, Norman RA, Murphy AJ. The development of a model for determining scheduled replacement intervals for marine machinery systems. Proc Inst
Mech Eng, Part M: J Eng Marit Environ 2017;231:723–39.
[225] Gubner RJ. Biofilms and accelerated low-water corrosion of carbon steel piling in tidal. PhD Thesis. Univrsity of Portsmouth; 1998.
[226] Jeffrey R, Melchers RE. Corrosion of vertical mild steel strips in seawater. Corrosion Sci 2009;51:2291–7.
[227] Yamamoto N. Prediction of corrosion condition considering effect of maintenance. In: Proceedings of the 33rd international conference on ocean, offshore and
arctic engineering (OMAE), 8–13 june 2014, san francisco, California, USA; 2014. p. 7. https://doi.org/10.1115/OMAE2014-23851”.
23

Abbasetal

Uploaded by

Copyright:

Available Formats

Abbasetal

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Abbasetal

Uploaded by

Copyright:

Available Formats

Marine Structures 71 (2020) 102718

Contents lists available at ScienceDirect

An overview of maintenance management strategies for corroded

Fig. 1. Evolution of maintenance strategies in the marine sector.

2.1. Corrective maintenance

2.2. Preventive maintenance (PM)

2.3. Condition-based maintenance (CBM)

2.4. Reliability-centred maintenance (RCM)

2.5. Risk-based inspection (RBI)

2.6. Other contemporary maintenance strategies

Fig. 2. Classification of RUL prediction techniques [55].

3. Impact of environmental conditions on corrosion of marine structures

4. Findings of the literature review

4.1. Findings on maintenance management strategies for marine steel structures

Fig. 3. Deterioration of design performance over time [15].

RCM CBM PM RBM/RBI HYBRID

4.2. Findings on factors affecting structural corrosion in marine environments

Fig. 4. Corrosion rates of marine steel in different exposure zones [62–66].

4.2.1. Environmental factors

4.2.2. Physical and chemical factors

pH 8.1 7.1 7.23 7.07 7.14 6.9 7.5–8.3

Baltic Sea 8000

0 14.6 13.9 13.1 12.4 11.8 11.4

4.2.3. Microbiological factors

4.3. Findings on corrosion prediction models for marine structures

- Southwell’s linear model:

- Southwell’s bi-linear model:

- Melchers-Southwell’s non-linear model:

dðtÞ ¼ 0:84t0:823 ; (3)

- Melchers’ tri-linear model:

- Melchers’ power law model:

dðtÞ ¼ 0:1207t0:6257 ; (5)

- Yamamoto-Ikegami’s non-linear corrosion model [115].

dðtÞ ¼ C1 ðt T0 Tt ÞC2 ; (6)

- Paik’s nonlinear model [175]:

dðtÞ ¼ c1 ðt Tcl Þc2 ; (7)

Fig. 7. Paik’s corrosion prediction model [178].

Fig. 8. Soares and Garbatov’s corrosion prediction model [106].

Fig. 9. Qin and Cui’s corrosion model [61].

5. Advanced maintenance management techniques

5.1. Fault tree analysis (FTA)

5.2. Bayesian network (BN)

5.3. Statistical and stochastic models

5.4. Multi-criteria decision analysis (MCDA)

5.5. Artificial intelligence (AI) and machine learning (ML)

6. Discussion and analysis

BN AI/ANN FTA MCDM FEA FMEA HYBRID

7. Conclusion and future works

You might also like