Ocean Engineering 234 (2021) 109261

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

A failure analysis of floating offshore wind turbines using

AHP-FMEA methodology
He Li a, b, H. Díaz a, C. Guedes Soares a, *
a
Centre for Marine Technology and Ocean Engineering (CENTEC), Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal
b
Center for System Reliability and Safety, University of Electronic Science and Technology of China, Chengdu, Sichuan, 611731, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: An AHP-FMEA methodology is proposed to analyse the floating offshore wind turbines failure causes. The
Analytical hierarchy process method introduces the expert opinions to generate a risk index through the Analytical Hierarchy Process criteria
Failure analysis weighting technique. Main components, critical systems, failure modes, and failure causes are analysed. Fifteen
Floating offshore wind turbine
failure scenarios are identified and corresponding preventive and corrective actions are suggested to cut off paths
Failure mode and effect analysis
Risk analysis
of failure propagations and to reduce the failure impacts. This method allows to minimize the catastrophic
failures on floating offshore wind turbines. Moreover, a comparison study is conducted to remove the uncertainty
when implementing the presented method, which validates the proposed methodology’s creativeness and
feasibility.

1. Introduction (Kang et al., 2017; Bento and Fontes, 2019; Li and Guedes Soares, 2019;
Li et al., 2020b). Through risk assessment, effective maintenance can be
Floating offshore wind technology has experienced extensive and guaranteed, which are predictions of weak links in the system in the
worldwide growth during the past several years due to the surge of early design stage (Sinha and Steel, 2015; Kang et al., 2019a; Scheu
renewable energy demand (Turner, 1999; Dincer, 2000; Díaz and et al., 2019; Wu et al., 2019; Li et al., 2020b, 2021).
Guedes Soares, 2020b). The floating wind represents the next step of the Floating offshore wind power generation has apparent advantages
offshore wind market. This new concept permits accessing depths of seas over fixed-bottom foundation structures. Nevertheless, there exists the
up to 1000 meters, where the wind resource is typically higher (Col­ possibility to appear a large volume of damages of blades, gearboxes,
menar-Santos et al., 2016; Musial et al., 2016; Salvacao and Guedes mooring systems, and other parts of floating wind turbines as their
Soares, 2018). operation time increases due to the severe working conditions at sea
Alongside the wind potential access to infeasible locations so far may (Santos et al., 2015; Kang et al., 2019b; GWEC, 2020). The failure rate
provide increased opportunity for spatial diversification (Moore et al., and economic losses can be effectively reduced through fault detection
2018; Díaz and Guedes Soares, 2020a). Floating wind turbines could and risk or failure analysis of floating wind turbine systems. Therefore,
lead to lower wind energy costs due to the benefits already mentioned, to analyse the failure modes and their effects, the method of Failure
but are currently slightly more expensive than fixed structures (Atha­ Mode and Effect Analysis (FMEA) is taken to conduct failure assessment
nasia and Genachte, 2013; Castro-Santos et al., 2016; Uzunoglu and in this paper.
Guedes Soares, 2020). The main objective of FMEA is to identify the deficiencies at the
Due to these new concepts’ immaturity, energy planners and stake­ product design stage and, based on that, analyse failures of each element
holders’ decisions are highly dependents on costs and vulnerabilities and their impact on each system or even at a holistic level (Spreafico
(Leimeister and Kolios, 2018; Castro-Santos et al., 2020). The floating et al., 2017; Liu et al., 2019; Li et al., 2020a). FMEA incorporates sub­
technology requires frequent maintenance, and such a procedure’s dif­ jective assets like severity, occurrence, and detection to assess a sub­
ficulty involves a considerable amount of associated costs (Sinha and jective and dimensionless index, namely Risk Priority Number (RPN),
Steel, 2015; Li et al., 2020b). Due to the distance to shore and high repair which is a representation of each potential failure (Stamatis, 2003).
consuming time, all capital and operating costs are assumed to escalate However, limitations of FMEA have been criticized (Stamatis, 2003;

* Corresponding author.
E-mail address: [email protected] (C. Guedes Soares).

https://doi.org/10.1016/j.oceaneng.2021.109261
Received 28 February 2021; Received in revised form 26 April 2021; Accepted 30 May 2021
Available online 19 June 2021
0029-8018/© 2021 Elsevier Ltd. All rights reserved.
H. Li et al. Ocean Engineering 234 (2021) 109261

Xiao et al., 2011; Dinmohammadi and Shafiee, 2013; Shafiee and Din­ FMEA process. The analysis outcomes provide fundamental guidance for
mohammadi, 2015; Sinha and Steel, 2015; Du et al., 2017; Tazi et al., condition-based maintenance implementation and benefit optimal oper­
2017; Cevasco et al., 2018; Scheu et al., 2019; Li et al., 2020b, 2021): (i) ating and maintenance (O&M) strategy planning.
Three risk factors (severity, occurrence, and detection) are not differ­ On the other hand, new FMEA models have been created to release
entiated in the RPN due to the relative importance of them considered practical limitations, especially neglecting risk factors’ weights, when
equal. (ii) Various combinations of risk factors may result in the same applied to the offshore wind energy sector. Xiao et al. (2011) put for­
RPN value hindering risk level determination. (iii) Discrete and fixed ward that severity, occurrence, and detection should be graded by
risk factor levels generate RPNs most likely at the bottom of the scale weights according to maintenance features. Accordingly, Li et al. (2021)
and several specific values resulting in uneven RPN distributions. considered the relative importance of severity, occurrence, and detec­
Therefore, conventional FMEA may not be adequate to determine the tion of FMEA to identify risky failures of floating offshore wind turbines.
risk level of failures. As a result of this, a weighted-FMEA was developed and implemented to
By improving the traditional method of FMEA so that it can be free identify critical failure items of floating offshore wind turbine support
from the practical restrictions already mentioned above and adequate to structures. Rather than implementing the weighted-FMEA method at the
failure analysis of floating offshore wind turbines, a weighted FMEA system level, Dinmohammadi and Shafiee (2013) finalized a risk anal­
technique is proposed in this paper. The novel contributions of this study ysis of a holistic wind turbine using the method mentioned above.
as follows: However, the mentioned studies assign directly subjective weights of
risk factors of FMEAs, and such a rough modelling manner makes the
(1) To apply the Analytic Hierarchy Process (AHP) methodology to failure analysis results discussable. Recently, ascertaining mentioned
extract weights of risk factors of FMEA. weights by Analytic Hierarchy Process (AHP) method tends to be a hot
(2) To present a normalization algorithm to convert the risk factors’ topic for today’s research in the risk analysis sector.
values into a comparable scale as their weights to avoid results’ The AHP structure contains multiple layers and weights designed to
bias. connect a specific layer and the upper one. On this basis, the information
(3) To propose an AHP-FMEA technique to complete a convincing propagation among the AHP structure includes information trans­
and close-to-practice failure analysis of floating offshore wind formation in the same layer and between layers, computationally say
turbines, which are relatively new and very few be investigated. comparison calculation in layer and weighting calculation between
layers (Stamatis, 2003; Huang et al., 2020). The AHP method extracts
This paper is organized as follows: The state-of-the-art is provided in information from the given data by computing relative differences of
Section 2. The methodologies presented are illustrated in Section 3. factors and then transfers the comparison results to the upper layers by
Modules classification of the floating offshore wind turbine and data using a weighting algorithm. The AHP is structurally flexible and con­
collection is demonstrated in Section 4. The results and recommenda­ forms to the weight-acquire demand of FMEAs. Accordingly, Hu et al.
tions are in Section 5. Comparisons and discussions are conducted in (2009) evaluated green components’ failures by the weighted-FMEA, in
Section 6. Conclusions are drawn at the end in Section 7. which AHP methodology has been applied to infer the weights of risk
factors. The results recommended a more feasible quality checking
2. State-of-the-art procedure. Wang et al. (2020) assigned weights of severity, occurrence,
and detection by the AHP methodology and prioritized 21 failure modes
The efforts on conducting failure analysis of wind turbines (onshore of structure, propulsion, electrical and auxiliary systems of an ocean
and offshore, fixed-bottom and floating structures) using FMEA or its fishing vessel. The study suggested that the hydraulic failure of the
upgraded version (Failure Mode, Effects and Criticality Analysis, auxiliary system is the most severe failure of the ocean fishing vessel. To
FMECA) are implemented into two aspects: analyse actual wind turbines overcome the situation that various combinations of risk factors could
by using conventional methods and develop new models to release obtain the same RPN hiding failure properties of systems, Fattahi and
practical limitations. Khalilzadeh (2018) put forward a hybrid method based on FMEA and
As for analysing actual wind turbines by using conventional FMEAs/ AHP methods to prioritize eight accidents, as substitutions of failure
FMECAs. Analyses aim to drive more comprehensive and instructive re­ modes, of the Kerman Steel Industries Factory in Iran. Two dangerous
sults by an elaborate system grading and exhaustive failure collection in accidents, plates and rollers falling on the feet, were found to be critical
operating wind turbines and then to suggest recommendations to optimal than others. Carpitella et al. (2018) proposed a developed FMECA model
design and update preventive activities of daily operations. For instance, to prioritize 42 failure modes of a cleaning vehicle, and the AHP tech­
Arabian-Hoseynabadi et al. (2010) decomposed a 2 MW indirect drive and nique had been implemented to infer weights of risk factors. This study
variable speed wind turbine into 11 components. They collected 16 failure ascertained 12 critical failure modes and accordingly optimized the
modes with 27 root causes to drive a failure analysis by applying the FMEA maintenance activities of the vehicle.
method. The study identified the wind turbine’s decisive failure modes: However, the mentioned studies neglected the individuality of fail­
materials failure, fracture, rupture, and electrical device failures. Bhar­ ures knowing that they assign the same weights of risk factors to all
atbha (2015) analysed 16 components of a 5 MW wind turbine with 33 failures. For floating offshore wind turbines, failure mechanisms are
failure modes by the FMEA method. The analysis applied probability different from failures to failures. For instance, failures of support
ranking and consequence ranking to generate the criticality index of each structures are significantly different from gearboxes as the former
failure mode identified that gear teeth slip (blades) and distorting the mainly resulted from harsh sea conditions. Still, gearboxes’ failures are
shape of the shaft (main shaft) are the most critical failures. Sinha and mostly the consequences of degradations of gears and bearings. To
Steel (2015) analysed 36 failure modes of a gearbox of offshore wind design weights of risk factors to each component or failure mode or
turbines by FMEA technique, accordingly ascertained the dangerous fail­ failure cause, considering related failure properties is still an open
ures by the developed FMEA software. The study was proved to be more problem of the risk analysis domain and the floating offshore wind
close-to-practice and tends to improve the accuracy of RPN numbers. turbine sector.
Dinmohammadi and Shafiee (2013) broke offshore wind turbines down to
16 assemblies for conducting the FMEA schedule. The analysis identified 3. Methodology
that the tower has the highest priority for failure prevention of offshore
wind turbines. To reach more comprehensive and constructive results of The AHP-FMEA technique is constructed in this section, aiming at
failure analysis for offshore wind turbines, Scheu et al. (2019) collected addressing shortcomings of the conventional FMEA method that is already
337 failure modes and employed more than 40 specialists to drive the mentioned in Section 1. The presented AHP-FMEA technique is a

2
H. Li et al. Ocean Engineering 234 (2021) 109261

combination of a proposed normalization algorithm and the AHP meth­ is applicable for other risk/failure analysis cases if the pairwise evidence
odology, which allows one to generate risk index (RPN) of a failure by two between risk factors are available.
data sets: relative importance of severity, occurrence, and detection (by
the proposed normalization algorithm) as well as their weights (by the 3.1. Normalization algorithm
AHP methodology). The procedure of driving the proposed AHP-FMEA
methodology is demonstrated as follows, also as shown in Fig. 1. Denote βSi , βOi , and βDi the average values of severity, occurrence,
and detection of failure cause i, which can be arranged by:
(1) System identification. Determine the floating offshore wind tur­ ⎡ ⎤
bine that to be analysed. βS 1 βS 2 ⋯ βS i ⋯ βS n
⎣ βO1 βO2 ⋯ βOi ⋯ βOn ⎦ (1)
(2) System decomposition. Break the floating offshore wind turbine
βD1 βD2 ⋯ βDi ⋯ βDn
down to components and record the hierarchical structure.
(3) Failures collection. Identify failure modes of each component and Denote
analyse their root causes. Record the hierarchical relationship
βK i
between components and failure modes until to root causes, ξK ij = (2)
βK j
obtain values of risk factors (severity, occurrence, and detection)
of failure causes from specialists.
where, K reflects risk factors including severity, occurrence, and
(4) Relative importance computation. Calculate the relative impor­
detection.
tance of each failure cause by the proposed normalized algo­
Accordingly, the comparison matrix ξKij of risk factors is represented
rithm, see Section 3.1.
(5) Weights calculation. Collect pairwise indices (severity over as:
occurrence, severity over detection, and occurrence over detec­ ⎡ ⎤
ξS 11 ξS12 ⋯ ξS1n
tion) from specialists and infer weights of risk factors of each ξKij = ⎣ ξO11 ξO12 ⋯ ξO1n ⎦ (3)
failure item by AHP technique, see Section 3.2. ξD11 ξD22 ⋯ ξD1n
(6) RPNs computation. Calculate RPNs of failure causes (use Eq.
The normalized vector ΦKij is defined as:
(10)), failure modes (use Eq. (11)), and components (use Eq.
(12)) of the floating offshore wind turbine. ΦK = [ ϕS ϕO ϕK ] (4)
(7) Recommendations. Identify weak links of the floating offshore
wind turbine and suggest corrective actions and preventive O&M where
measures to avoid their occurrence. ∑
ξKij
(5)
j
Overall, several concerns should be highlighted for a better under­ ϕK = ∑∑
ξKij
standing of the proposed AHP-FMEA methodology: The method employs i j

both risk factors and weights, indicating that it requires more extensive It is obvious that ϕKij ∈ [0, 1] and
∑
ϕK = 1.
data collection than conventional methods; The scales of risk factors’ K=S,O,D

values are normalized to be the same as that of weights to avoid that either
indices or their weights affect significantly the RPN generation than 3.2. AHP based weights computation
another; The weights of risk factors are changeable for different failure
items as a consequence of extensive data collection, detailed modelling, The importance matrix of risk factors of failure mode i is constructed
and much more complicated computations; The AHP-FMEA methodology as:

Fig. 1. The procedure of implementing the proposed AHP-FMEA/FOWT: Floating Offshore Wind Turbine; S: Severity; O: Occurrence; D: Detection.

3
H. Li et al. Ocean Engineering 234 (2021) 109261

⎡ ⎡ ⎤ ⎤
αi αis/o αis/d 1 αis/o αis/d
⎢ s/s ⎥ ⎢ / i ⎥
⎢ ⎥ ⎢
αio/d ⎥ = ⎢ 1 αs/o 1 αio/d ⎥
Ai = ⎢ αio/s αio/o ⎥
⎣ ⎦ ⎣ / / ⎦
αid/s αid/o αid/d 1 αis/d 1 αio/d 1
⎡ ⎤
αS1 αS2 αS3
= ⎣ αO1 αO2 αO3 ⎦ (6)
αD1 αD2 αD3

where, αis/o , αis/d , and αio/d are pairwise weights of severity over occur­
rence, severity over detection, and occurrence over detection of failure
mode i, respectively.
Define a matrix Δ as:
⎡ ⎤
δS1 δS2 δS3
ΔKj = δO1 δO2 δO3 ⎦
⎣ (7)
δD1 δD2 δD3

where
α
(8)
Kj
δKj = ∑
αKj
K=S,O,D
Fig. 2. Floating offshore wind turbine/(a) Wind turbine; (b) Tower and tran­
Accordingly, the weights of indices are calculated by:
sition piece; (c) Floating foundation; (d) Mooring system.
⎡ ⎤T
∑
3
∑
3 ∑
3
ωj = ⎣ δSj /
3 δOj /
3 δDj 3 ⎦
/
(9) maintaining an accurate position of the tower. The supporting role,
j=1 j=1 j=1
tower, and transition piece are also expected to resist force and moment
produced by the variable wind, wave, and current loads (Lee et al.,
Hence, the RPN of failure cause i (RPNW
FC
j ) is computed as: 2016). The failures of support structures occur mainly at the flange bolt
i

( ) and weld area, caused by fatigue loads. Meanwhile, failure modes like
W
RPNFC j = f ΦKij , ωj = ΦKij × ωj (10) hit by blades, strong wind or wave, lightning strike, ice storm, storm,
i

and material fatigue also exist under the effect of extreme loads. Severe
where ΦKij denotes the relative importance of failure cause i of jth consequences, such as tower collapses, will result in the wind turbine’s
failure mode. ωj reflects weights of failure mode j. sinking and severe losses (Chou and Tu, 2011).
Hence, the RPN of failure mode j which contains m failure causes can Floating foundation: The main types of floating offshore wind
be calculated as: turbines under development are the Semi-Submersible, Spar, and Ten­
∑
m sion Leg Platform (TLP) (Uzunoglu et al., 2016). The main floating
W
RPNFM i
= W
RPNFC j (11) foundation failures are pillar damage, corrosion of pipe joint, weld effect
of the pipe joint, fatigue joints, insufficient detection, typhoon (not exist
i
i=1

Subsequently, the RPN of component h (RPNComp

W
) can be computed for all seas), a crash of planes, biological collision, and capsize.
Mooring system: The mooring systems for floating offshore wind
h

as:
turbines have benefited from the offshore oil and gas experience (Ara­
p
∑ pogianni et al., 2013). The main mooring configurations are the Cate­
W
RPNComp = W
RPNFM (12)
nary system and the Taut leg system (Xu et al., 2019). The mooring
j
h
i
j=1
system failures result in the unbalance of upper structures. The possible
failures are the sea’s limited state, corrosion off-air lead, fatigue off-air
4. Modules classification of the floating offshore wind turbine
lead, mooring lines brake, unsatisfied sea state for operation, insuffi­
and data collection
cient emergency measures for abrupt scenes and analysis, and calcula­
tion fault. When the mooring module fails, the system will be at risk.
4.1. Modules classification

The offshore floating wind turbine comprises various modules, in­

Table 1
dependent or auxiliary to each other. The floating offshore wind turbine
Main components of floating offshore wind turbines.
system is functionally divided into four parts: wind turbine, tower and
transition piece, floating foundation, and mooring system (see Fig. 2). System Component
The main components of floating offshore wind turbines are listed in Code System
Table 1. WT Wind Turbine Blade (BL), Hub (HB), Main Bearing (MB), Main
Wind turbine: The wind turbine is mainly composed of a rotor Shift (MSH), Generator (GE), Gearbox (GB),
system, nacelle, and yaw system, among which the internal components Converter (CV), Transformer (TR), Pitch Subsystem
of nacelle are crucial and complicated. The blades/pitching control (PS), Yaw Subsystem (YS), Controller and Electrical
Facilities (CE)
system, nacelle, and gear’s failure rates are the three highest. Other wind
TT Tower and Tower (TO), Transition Piece (TP)
turbine failures can be described as rotor system failure, bearing or shaft Transition Piece
failure, gearbox failure, braking system failure, generator failure, sensor FF Floating Pipe (PP), Pillar (PL), Lights (LT), Helicopter
failure, yawing failure, drivetrain control system failure, hydraulics Foundation Assistance Equipment (HAE), Handrails (HD),
Ladders (LD), Vents (VE), Sensors (SS), Manholes
failure, structural failure, and electrical failure.
(MH)
Tower and transition piece: The tower is used to support the wind MS Mooring System Mooring Lines (ML), Fairlead (FL), Anchor (AC),
turbine components. The transition piece performs the function of Rests (RE)

4
H. Li et al. Ocean Engineering 234 (2021) 109261

4.2. Data collection Table 3

The rating guidance of risk factors.
Five specialists with diverse backgrounds and with an overall 28 Rating Severity Occurrence Detection
years’ experience in floating offshore wind turbines or offshore wind
Probability Description
farms were employed, see Table 2. Overall, 42 failure modes with 104 (P)
failure causes were identified by the specialists. It should be noted that 5
1 The effect is not P < 10− Extremely Certain
the failure modes and causes identified partly reviewed from publica­ noticed less
tions, and the remains either have been observed in the operating 2 Very slight effect P = 10− 5
Remote Very high
offshore wind farms or ascertained by specialists under the assumption noticed
5
of those failures are potentially occurred according to their experience 3 Slight effect causing P = 10− Very slight High
annoyance
and knowledge. − 5
4 Slight effect causing 10 < P < Slight Moderate
Failure items of the floating offshore wind turbine are listed in Ap­ return of product 4*10− 4
pendix A. The standardization and consistency of specialists’ evalua­ 5 Moderate effect 4*10− 4 < P < Occasional Medium
tions are guaranteed by unified rating guidance, see Table 3. Moreover, causing return of 2*10− 3
the normalized values of risk factors and computed weights are listed in product
6 Significant effect 2*10− 3 < P < Moderate Low chance
Appendix B. Besides, pairwise comparisons of risk factors representing 1*10− 2
the relative importance among risk factors are performed using Saaty’s 7 Major effect 10− 2 < P < Frequent Slight
method that is designed to evaluate the degree to which a factor is 4*10− 2
important than comparing one, see Table 4. 8 Extreme effect, 4*10− 3 < P < High Remote
system inoperable, 0.2
safety issue
5. Results and recommendations 9 Critical effect, 0.2 < P < 0.33 Very high Very remote
system shutdown,
5.1. Systems and components safety risk
10 Hazardous, without P > 0.33 Extremely No chance, no
warning, life- high inspection
Finding out critical systems and components of floating offshore threatening
wind turbines initiates the understanding of failure natures and benefit
for arranging O&M activities e.g. failure warning and diagnosis, ele­
ments backup and transportation, inspections and preventive mainte­
Table 4
nance scheduling.
Pairwise comparisons of Saaty’s method.
The criticality rank of the floating offshore wind turbine systems is
demonstrated in Fig. 3(a). The wind turbine is recognized to be the most Intensity of Description Intensity of Description
Importance Importance
critical system of the floating offshore wind turbine, with an RPN of
0.44, followed by the mooring system (0.18), the floating foundation 1 Equal 7 Very Strong
Importance Importance
(0.17), as well as the tower and transition piece system (0.16). Fig. 3 also
3 Moderate 9 Extreme
concluded that: the wind turbine is critical as they are structurally Importance Importance
complicated and contains more failure items; Mooring systems tend to 5 Strong 2,4,6,8 Intermediate
be more vulnerable in dealing with complex and harsh sea conditions Importance Values
than floating foundations; The tower and transition piece system is the
most reliable system that likely free from the failures due mainly to its
floating platforms at a selected sea space, are recognized as weak links of
simple structure and limited functions.
the floating offshore wind turbine as which hold the highest RPNs, see
Overall, 18 components of the floating offshore wind turbine were
Fig. 3(c). According to Fig. 3(b) and (c): the impacts of fairlead and
analysed. Mooring lines, jointed with fairleads and anchors to locate
anchor malfunctions on the holistic floating offshore wind turbine are
minor; The tower is riskier than transition pieces. In general, the tower’s
failures are unlikely to be detected, giving rise to disasters to floating
Table 2
offshore wind turbines. Failures of towers will directly result in upper
Specialists involved/FOWTs: Floating Offshore Wind Turbines.
wind turbines shaking (reduce energy generation efficiency) or sink into
Code Employer Duty Working Country Reason of
the water. Fortunately, the likelihoods of tower failures are relatively
Period Employment
low as displayed in Appendix B. Such an idea is also supported by spe­
<1> Wind System 4 Years China Specialized in cialists who distributed the highest weight of severity to the tower
Energy Design systems design
(73%). Still, the weights of occurrence (18%) and detection (9%) are the
Company
<2> Wind Components 3 Years China Specialized in lowest, see Fig. 3(b).
Energy Design components For the components within the nacelle, the generator (RPN = 0.099)
Company design and gearbox (RPN = 0.098) are remarkably critical than other devices
Wind Quality 4 Years China Experienced in
<3>
(RPNs<0.04). The results are also obtained from Du et al. (2017),
Energy Engineer quality issues
Company
Bhardwaj et al. (2019), Scheu et al. (2019), and Li et al. (2020b). The
<4> University Researcher 6 Years China Experienced in gearbox and generator are structurally complex with a selection of ele­
failure analysis ments and accomplishing pivotal and primary functions. The failure
of FOWTs, with consequences of both components are severe and impact the electricity
several
generation of floating offshore wind turbines immediately. The afore­
publications
related mentioned results are in line with the practice that engineers had fol­
<5> Floating Technical 11 years Spain Experienced in lowed in gearboxes and generators by implementing sensors to
Wind Officer Chief the operation of monitoring their health conditions as a basis of that warning and predict
Company FOWTs
their malfunctions ahead of unwanted events happen. As typical elec­
especially in the
maintenance
tromechanical equipment, generators fail more frequently than me­
sector chanical devices (installed within nacelles) such as gearboxes. These

5
H. Li et al. Ocean Engineering 234 (2021) 109261

Fig. 3. Failure analysis of the floating offshore wind turbine at systems and components levels/CE: Controller and electronic elements.

Fig. 4. Failure analysis of the floating offshore wind turbine at failure modes level.

6
H. Li et al. Ocean Engineering 234 (2021) 109261

results can explain that specialists assigned a higher weight to genera­ 5.3. Failure causes
tors’ occurrence, as in Fig. 3(b).
Failure causes are the root reasons for failures, which randomly give
rise to failure modes of floating offshore wind turbines and cannot be
5.2. Failure modes eliminated and mostly cannot be observed visually such as material
deterioration and human errors. Aggregately, 104 failure causes were
Failure modes are representations of floating offshore wind turbines identified, and RPNs of which were computed by the AHP-FMEA, see
working in unexpected states. In this study, 42 failure modes are ana­ Fig. 5 and Appendix B. The mean and median of failure causes’ RPN are
lysed, see Fig. 4. A total of 15 failure modes are recognized to be critical 0.0097 and 0.0096, respectively. The critical failure causes wear, fa­
as their RPNs exceed the average value (0.024), that is, bearing defor­ tigue, lubrication fails, corrosion, environmental factors, and human
mation, generator overheat, winding failure of the generator, gearbox errors, see Fig. 6.
overheat, open circuit of converter and transformer, tower collapse, From the single failure cause point of view, destructive environ­
tower crack, transition piece crack, floating foundation hitting, water­ mental conditions are top failure factors e.g. lightning strikes and ice
tight and additional structures failure of floating foundations, abnormal storms that introduce fragility to towers and strong wind/wave bring
and broken mooring lines, as well as abnormal functions of mooring about the additional risk of failures to both towers and floating foun­
lines. dations. The results are supported by Kang et al. (2017) and Kang et al.
The generator and gearbox are the weak links of the floating offshore (2019b), in which harsh sea conditions were ascertained to be risk
wind turbine (see Section 5.1). To be specific, bearing deformation (with factors to floating offshore wind turbines from risk assessment and
the RPN of 0.0362 and RPN share of 3.6%), overheat (0.03, 3%), and reliability analysis (failure probability) perspectives. The results also
winding failure (0.0335, 3.4%) are risky failure modes of the generator. explain the leading differential of failure features between floating
Overheat (0.0268, 2.87%) is the most critical failure mode of the offshore wind turbines and fixed-bottom structures or between onshore
gearbox. Differently, relatively high failure frequency and remarkable wind turbines and offshore ones. According to results, floating founda­
weight of detection are distinguished properties of the gearbox failures, tions and towers are influenced significantly by harsh environmental
indicating the gearbox failures are frequent while hard to be detected situations as such components are huge and exposed to adverse weather
and predicted. The transformer failures are severe as they directly affect conditions. In contrast, smaller size components installed within na­
the electricity yield of the floating offshore wind turbine. Failures of celles are likely free from sea conditions’ decisive impacts.
tower and transition pieces are few but risky. Tower collapse is critical as Besides, material-related factors like fatigue, wear, and corrosion are
such a failure is unlikely to be detected in advance of its occurrence. also critical failure causes. The aforementioned results are confirmed by
Moreover, the transition piece is an essential component, and the pri­ Arabian-Hoseynabadi et al. (2010), Kahrobaee and Asgarpoor (2011),
mary failure of which is crack. Bharatbhai (2015), Sinha and Steel (2015), Kang et al. (2017), Li and
Devices failures such as watertight, vent, pump, sensor, and manhole Guedes Soares (2019), Scheu et al. (2019), and Li et al. (2020b), which
are important among other components of floating foundations. These determined that materials and structural degradation are dangerous
failures are common but hard to be detected. Note that high failure factors to failures of floating offshore wind turbines. Unlike harsh sea
frequencies are the unique nature of the floating foundations. Mooring conditions that make failures of floating offshore wind turbines different
lines failures, including abnormal and broken mooring lines, call for from onshore and fixed-bottom structures, the results also display that
engineers’ particular attention, especially designers. Broken mooring material-related factors may promote failures of several types of wind
lines occupying the top place of all failure modes by contributing 10% turbines in a similar way.
RPN to the total. Abnormal mooring lines initiate broken mooring lines,
make 5% RPN of the holistic equipment, which can be an indicator to
predict and avoid disasters mooring lines failures.

Fig. 5. Failure analysis of failure causes of floating offshore wind turbines.

7
H. Li et al. Ocean Engineering 234 (2021) 109261

Fig. 6. Critical failure causes of floating offshore wind turbines, see Appendix A.

5.4. Recommendations release unexpected wear. (ii) Pay attention to the quality and states of
the anchor pickup device and mooring winch. (iii) Improve the training
Failure behaviour is a path of the initiation, propagation, and to operators and maintenance members to avoid human failures.
occurrence of a failure. To be specific, failure behaviour is the series Note that common cause failures are identified according to the
form that a failure cause gives rise to a failure mode(s) of a component(s) failure behaviours ascertained, which makes failures of those compo­
and subsequently results in failure of a system(s) until the malfunction of nents correlated and will introduce additional failure probability to the
floating offshore wind turbines. It accomplishes the following failure correlated components and ultimately decrease the reliability of floating
chain: failure cause-failure mode of component-system failure-mal­ offshore wind turbines. However, correlated failures will not be dis­
function of floating offshore wind turbines. Failure analysis is to cussed in this paper as they are under reliability estimation, not the topic
recommend preventive and corrective actions to cut off the failure of this paper that is failure analysis.
propagation chain already mentioned or avoid (at least release or
decrease probabilities) the occurrence of critical failure causes. Ac­ 6. Discussion on expert opinions
cording to Sections 5.1 to 5.3, 15 essential behaviours of failure were
identified, as displayed in Fig. 7. To this end, 15 recommendations, both 6.1. Comparison between experts about risk priority number
preventive and corrective, are suggested as follows:
Wind turbines: (i) Strengthen surfaces of bearing tracks (inner ring, Table 5 gives the RPN values obtained from the proposed AHP-FMEA
outer ring, and rolling elements) of bearings and improve electric module (Section 3) for the fifteen sub-assemblies of floating wind tur­
corrosion preventions of generators. (ii) Upgrade or backup cooling bine systems considered in this study. Transition piece, main bearing,
systems to enhance the heat dissipation of generators and converters. tower, and mooring system account for almost two-fifths of all critical
(iii) Anti-corrosion treatments of windings. (iv) Introduce wear and fa­ failure modes identified during this paper. This value is unprecedented
tigue preventive actions to gears in gearboxes. (v) Guarantee high- in the context of the literature available in this subject area. Moreover,
quality lubrications by better design of the hydraulic systems and these categories contribute more than 10% of the critical failure modes,
implement additional leak-proof measures. (vi) Monitor performance of whereas the contribution is rather significant from the expert’s back­
transforms, backup additional devices, and open circuit function design ground of floating offshore wind turbines. Experts (1) and (2) deal with
to avoid overvoltage. systems design. They considered higher risk of the transition piece,
Tower and Transition pieces: (i) Design and implement reliable blades, hub, main bearing, main shaft, and generator; experts (3) and (4)
and less vulnerable lightning prevention systems. (ii) Reinforce tower focus on quality and research deal with offshore, identified tower, pitch
and transition piece welding quality. (iii) Consider the impacts of strong system, main bearing, main shaft, and hub as risky components in this
wind/waves in the full life cycle of floating offshore wind turbines. study. Nevertheless, Expert (5) focus on floating structures identified
Floating foundations: (i) Additional design to release the impacts of with higher RPN the transition piece, mooring system, and floating
strong wind/waves. (ii) Enhance pipe joints design and welding quality foundation. Therefore, it can be concluded that the background and
of floating foundations to avoid watertight fault. (iii) Periodical in­ work experience in an area is key for an FMEA analysis. The different
spections to bilge piping and pumps. experts’ backgrounds cover the systems involved in a floating wind
Mooring systems: (i) Enhance the strength of mooring lines, buoys turbine, guarantee equity in the results.
friction chain, and transitional chain to eliminate abnormal stress and

8
H. Li et al. Ocean Engineering 234 (2021) 109261

Fig. 7. Critical failure behaviours of floating offshore wind turbines.

6.2. Experts consensus (2013) and Grošelj et al. (2015).

The consensus is calculated as the weighted arithmetic mean of all
The proposed AHP-FMEA methodology implemented, and the di­ hierarchy nodes (categories). The consensus indicator of the consoli­
versity of experts involved generated a general result from the different dated alternative evaluation matrix weighted with the global priorities
opinions’ aggregation. It is necessary to measure the (severity, occur­ for alternatives was used. The consensus indicator ranges from 0% (no
rence, and detection) consensus for the aggregated group result consensus) to 100% (full consensus). The results are shown in Table 7.
(Table 6). Shannon entropy and the three independent components A particular observation is made for the consensus of results. With
(alpha, beta, and gamma) to derive the consensus indicator were used. less than 36%, the percentage of agreement between the different ex­
The priority distribution of indices of severity, occurrence, and detection perts is low. This low consensus is expected due to the various stake­
among different experts was analysed using Refs’ relations Goepel holders’ different expertise regarding the 15 components analysed. The

9
H. Li et al. Ocean Engineering 234 (2021) 109261

Table 5

Expert 5
Individual evaluation of the components as a function of the RPN.
Components Expert Expert Expert Expert Expert

2
1
5
1
2
3
1
1
1
1
1
3
4
4
4
1 2 3 4 5

Expert 4
Blades 5.36% 10.82% 5.27% 7.14% 2.40%
Hub 10.68% 5.56% 2.31% 11.68% 0.19%
Main bearing 8.08% 8.31% 10.58% 7.45% 14.59%

5
3
6
3
3
4
3
3
3
3
4
4
3
3
3
Main shaft 10.34% 9.35% 2.43% 11.83% 0.19%

Expert 3
Generator 11.73% 7.10% 6.08% 7.59% 4.17%
Gearbox 9.07% 5.06% 9.61% 6.93% 6.40%
Converter 2.73% 3.69% 4.38% 3.30% 0.19%

1
2
1
1
2
1
1
1
2
1
1
2
2
2
2
Transformer 0.64% 3.20% 4.92% 6.31% 0.19%
Pitch system 9.75% 4.51% 11.11% 6.73% 0.19%

Expert 2
Yaw system 6.61% 6.06% 4.14% 4.87% 0.65%
Controller and electrical 0.63% 8.31% 4.87% 4.87% 0.19%

6
3
3
3
6
5
4
2
2
4
5
3
6
4
5
facilities

Detection
Tower 3.25% 6.10% 13.81% 11.39% 13.93%

Expert 1
Transition piece 11.47% 10.09% 4.74% 3.41% 20.71%
Floating foundation 4.51% 3.83% 6.03% 2.69% 16.01%

4
6
7
5
6
5
3
1
6
5
1
2
5
3
4
Mooring system 5.15% 7.99% 9.71% 3.83% 19.97%

Expert 5
conclusions also indicate the background and knowledge of the

2
1
5
1
2
3
1
1
1
1
1
3
4
4
5
employed specialists are diverse, which helps prevent bias results and
validated that specialists’ selection is reasonably feasible.

Expert 4

10
5

5
9
7
6
3
6
5
7
5
7
5
4
4
6.3. Comparison between experts about components criticality

Expert 3
The criticality number is one of the most important outcomes of an

3
1
6
1
3
4
3
3
5
2
4
3
4
2
5
FMEA application. Criticality Number is very similar to RPN, a tradi­
tional FMEA metric. The difference is that criticality only includes

Expert 2
severity and occurrence. The criticality value for the floating wind tur­
bine components is estimated in Table 8.

6
8
7
6
6
5
5
6
8
7
4
4
6
5
5
Occurrence
Table 8 shows that the transition piece’s criticality index, main Expert 1
bearing, main shaft, and hub is significantly higher in floating wind
turbine components. Moreover, other components as the tower and
5
5
5
5
7
5
2
2
5
6
2
3
8
6
4
mooring system carry a high criticality factor. The main bearing, main
shaft, and hub are the most critical components in floating wind turbines
Expert 5

for experts (2), (3), and (4), while components as transition pieces and
mooring systems are the critical ones for experts (1) and (5). On the
2
1
5
1
2
3
1
1
1
1
1
7
7
6
5
other hand, converter, controller, and electrical facilities show the
Expert 4

lowest criticality values. The dominant failure cause for floating turbine
design are the tower, main shaft, main bearing, and hub.
7
8
7
9
6
6
7
7
8
5
5
9
5
4
7
Expert 3

6.4. Comparison of the AHP-FMEA and the conventional method

10

10
5
6
8

6
7
6
5
5
9
5

3
5
5
In this section, a quantitative comparison is made between the
Severity, occurrence, and detection based on individual considerations.

Expert 2

ranking orders of the traditional FMEA and the proposed AHP-FMEA. In

Table 9, the results obtained for floating offshore wind turbines from the
5
4
7
9
3
4
3
5
4
3
7
8
5
3
4

traditional FMEA using the RPN method are compared with the results
obtained from the proposed AHP-FMEA using the normalization method
Expert 1
Severity

(Section 3) and RPN method.

As can be seen, the main problem in the traditional FMEA method­
5
9
6
9
6
7
8
8
7
6
7
9
6
7
8

ology is that it puts two critical assemblies, the yaw system, and the
Components/CE: Controller and electrical facilities

converter, as having the same priority in Shafiee and Dinmohammadi

(2015).
A useful outcome of the proposed methodology could compare the
ranking orders obtained from both methods. From Table 9, it is observed
that both approaches lead to different ranking orders for seven indi­
Controller and electrical facilities

vidual assemblies (i.e., generator, gearbox, and tower). Nevertheless,

there is a noticeable similitude between the results obtained using two
approaches for some major components. The hub, controller and elec­
Floating foundation

trical facilities, main shaft, pitch system, transition piece, mooring sys­
Transition piece

tem, and the floating foundation occupy the same position in both
Mooring system
Main bearing

Pitch system
Transformer

rankings. Besides, as the assemblies’ failure rates in both ranks are

Yaw system
Main shaft

Conventer
Generator

almost equal, they are prioritized from the RPN perspective.

Gearbox
Table 6

Blades

Tower

For a better understanding of the proposed AHP-FMEA methodology,

Hub

the following aspects should be highlighted to remove the confusing

10
H. Li et al. Ocean Engineering 234 (2021) 109261

Table 7 support consensus in the field. Luckily, this study unprecedent­

Consensus analysis. edly provides such evidence by comprehensively collect judg­
S Consensus 35.22% β 1.0673 α 94.2453 ɣ 100.5908 ments from field engineers.
O Consensus 9.31% β 1.0969 α 92.7169 ɣ 101.6982 (2) Instead of applying absolute values of severity, occurrence, and
D Consensus 15.68% β 1.0895 α 91.9624 ɣ 100.1892 detection as been done by other published studies, this paper
converts such parameters into a comparable scale of their weight
to avoid biased outcomes derived from the situation that a group
Table 8 of parameters would impact more on the results than others.
Criticality index of the components. (3) This study contributes to the failure analysis of engineering cases
Components Criticality (in practice) and the development of FMEA methodology as well
as its extended versions (in academia). Accordingly, the method
Expert Expert Expert Expert Expert
1 2 3 4 5
applies to all failure analysis scenarios and is not limited to
floating offshore wind turbines or the wind energy sector.
Blades 25.33 28.67 15.00 29.33 4.33
Hub 37.50 29.50 5.50 80.00 1.00
Main bearing 24.00 44.50 43.50 27.00 25.00 7. Conclusions
Main shaft 43.00 46.50 10.00 81.00 1.00
Generator 38.09 19.27 13.91 45.91 6.00
This paper proposes an AHP-FMEA methodology to complete a
Gearbox 33.40 18.10 27.50 36.10 9.30
Converter 20.00 15.40 18.00 22.60 1.00 comprehensive failure analysis of floating offshore wind turbines. The
Transformer 13.40 29.40 16.60 43.20 1.00 presented method infers the importance of severity, occurrence, and
Pitch system 35.33 28.33 27.00 40.00 1.00 detection of the floating offshore wind turbine’s failures by the AHP
Yaw system 36.00 20.33 17,00 33.33 2.00 technique, and accordingly, weights of such risk factors can be extracted
Controller and electrical 13.00 26.00 20.00 25.00 1.00
facilities
to drive the implementation of the proposed AHP-FMEA methodology.
Tower 22.88 25.75 27.63 62.13 18.50 The wind turbine (with an RPN of 0.44) is the most critical system of the
Transition piece 48.67 28.67 11.00 23.33 27.17 floating offshore wind turbine followed by the mooring system (0.18),
Floating foundation 37.53 14.68 11.84 17.89 24.42 floating foundation (0.17), as well as tower and transition pieces (0.16).
Mooring system 27.61 21.43 27.00 26.00 25.13
Subsequently, broken mooring lines and the other 14 failure modes are
identified to be risky failures. This study subsequently distinguished the
critical failure causes including wear, fatigue, lubrication fails, corro­
Table 9 sion, environmental factors, human errors, and so on, and accordingly
Risk priority comparison between conventional and proposed methods. 15 preventive and corrective activities such as enhance the strength of
Rank Conventional method Proposed method mooring lines are suggested to prevent the floating offshore wind tur­
Components Value Components Value bine from catastrophic failures. Moreover, a comparison study is con­
ducted to clear some aspects when implementing the presented method.
1 Mooring system 25.79% Mooring system 18.98%
2 Floating foundation 14.84% Floating foundation 17.68%
The comparison results indicate that the RPN derived by different spe­
3 Generator 11.91% Tower 10.97% cialists is discrepant, meaning that FMEAs are subjective methods, and
4 Gearbox 11.03% Generator 10.48% personal judgments impact results; the selection of specialists is plau­
5 Tower 9.36% Gearbox 10.25% sible as their backgrounds are diverse; the correctness of the AHP-FMEA
6 Transition piece 7.60% Transition piece 6.22%
results is confirmed by the results of conventional methods.
7 Main bearing 3.16% Transformer 4.42%
8 Blades 2.79% Converter 4.18%
9 Pitch system 2.51% Pitch system 2.84% CRediT authorship contribution statement
10 Yaw system 2.18% Blades 2.78%
11 Converter 2.18% Yaw system 2.73%
12 Main shaft 2.06% Main shaft 2.49% He Li: Methodology, Formal analysis, Writing – original draft. H.
13 Transformer 1.84% Main bearing 2.33% Díaz: Methodology, Formal analysis, Visualization, Writing – original
14 Hub 1.84% Hub 2.03% draft. C. Guedes Soares: Writing – review & editing, Supervision.
15 Controller and electrical 0.91% Controller and electrical 1.65%
facilities facilities

Declaration of competing interest

situation that has long been existing in the failure analysis of the floating
offshore wind energy sector, which are: The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
(1) This study makes it clear to practitioners how to distribute the work reported in this paper.
weights to indices of FMEAs. One conviction of practitioners of
failure analysis is that weights of indices of FMEAs have to be Acknowledgements
considered for a convincing failure analysis result. However, such
parameters were distributed personally without clear evidence. This study was completed within the project ARCWIND - Adaptation
For instance, Dinmohammadi and Shafiee (2013) consider the and implementation of floating wind energy conversion technology for
weights of severity, occurrence, and detection to be 0.21, 0.26, the Atlantic region, which is co-financed by the European Regional
and 0.53, respectively, considering if failure can be detected in Development Fund through the Interreg Atlantic Area Programme under
advance, the frequency and consequence of which would be no contract EAPA 344/2016. The first author has been supported by the
longer important. It can be compared with the idea of Li et al. scholarship from China Scholarship Council (CSC) under Grant No.
(2021) that give the weights mentioned above by 0.4, 0.35, and 201806070048. This work contributes to the Strategic Research Plan of
0.25, respectively, assuming that potential failure of a system is the Centre for Marine Technology and Ocean Engineering (CENTEC),
an intrinsic and natural characteristic, detection measures cannot which is financed by the Portuguese Foundation for Science and Tech­
reduce physical risks. It is evident that the two assumptions have nology (Fundação para a Ciência e Tecnologia - FCT) under contract
inherent differences and reveal no substantial evidence to UIDB/UIDP/00134/2020.

11
H. Li et al. Ocean Engineering 234 (2021) 109261

Appendix A. Failure modes and causes of the floating offshore wind turbine

Code Failure Mode Level Code Failure Cause Level

Failure Modes End Effects Failure Causes

WT-BL-FM1 Blades cracks Wind turbine stop working #1 Manufacturing error

WT-BL-FM2 Delamination Wind turbine stop working #2 Insufficient lighting protection
WT-BL-FM3 Gear teeth slip Blades fail to attack wind properly #3 Wear, fatigue, etc.
WT-HB-FM4 Fracture in the shell Rotor break #4 Manufacturing error
WT-HB-FM5 Error in positioning Blades break away from the hub #5 Manufacturing error and/or fitting error
WT-MB-FM6 Bearing damage Wind turbine stop working #6 Wear, fatigue, etc.
WT-MB-FM7 Bearing vibration Abnormal working condition #7 Substandard lubrication
WT-MS-FM8 Cracks Collapse of wind turbine #8 Welding defects
WT-MS-FM9 Fracture Collapse of wind turbine #9 Fatigue
WT-GE-FM10 Bearing deformation No, abnormal or unbalanced electricity generation #10 Improper grease
#11 Over tighten
#12 Electric corrosion of rollaway nest
#13 Shaft wear deformation
WT-GE-FM11 Overheat (GE) Offshore wind turbine shutdown #14 Turbine overload
#15 Cooling system failure
#16 Partial short circuit on stator winding
WT-GE-FM12 Winding failure No, abnormal or unbalanced electricity generation #17 Cable insulation failure
#18 Connecting plug fall off
#19 Interturn short circuit
#20 Winding corrosion
WT-GB-FM13 Wear gears Exceeded vibration or unstable electricity output #21 Wear, fatigue
#22 Dirty or lacking lubrication
WT-GB-FM14 Seized gears No electricity output #23 Sudden shock exceed limitation
WT-GB-FM15 Fractured gear teeth Exceeded vibration or unstable electricity output #24 Sudden shock exceed limitation
#25 Fatigue
WT-GB-FM16 Wear bearing Exceeded vibration #26 Fatigue
WT-GB-FM17 Overheat (GB) Offshore wind turbine shutdown #27 Wear
#28 Lubrication dried out
#29 Leaking
WT-GB-FM18 Shift crack Offshore wind turbine shutdown #30 Fatigue
WT-CV-FM19 Short circuit Converter shutdown #31 Over heat
WT-CV-FM20 Open circuit Disconnect to grid #32 Load mutation
#33 Invert power input fault
#34 Overload
#35 Cooling system fault
WT-TR-FM21 Short circuit Transformer shutdown #36 Over heat
WT-TR-FM22 Open circuit Disconnect to grid #37 Constant overload
#38 Iron core corrosion
#39 Overcurrent
#40 Overvoltage
WT-PS-FM23 Wrong pitch angle Decrease of electricity output #41 Poor calibration
WT-PS-FM24 Pitting Gears Vibration increase #42 Wear, fatigue
WT-PS-FM25 Misalignment bearings Decrease of electricity output #43 Wear, excessive vibration
WT-YS-FM26 Seizure bearings Over heat #44 Poor lubrication
WT-YS-FM27 Corrosions Pitting of raceways #45 Presence of corrosive substances
WT-YS-FM28 Hydraulic leakage Rotor fails to stop #46 Wear or degradation on hydraulic lines
WT-CE-FM29 Short circuit Offshore wind turbine shutdown #47 Moisture penetration
WT-CE-FM30 Open circuit Offshore wind turbine shutdown #48 Lightning strike
TT-TO-FM31 Tower collapse Failure of whole facility and vast economic loses #49 Strong wind/wave
#50 Lightning Strike
#51 Hit by blades
#52 Ice storm
#53 Braking system failed
TT-TO-FM32 Abnormal vibration Potential collapse #54 Resonance
TT-TO-FM33 Crack Potential collapse #55 Faulty welding of Tower
#56 Material fatigue
TT-TP-FM34 Transition piece crack Potential collapse #57 Material fatigue
#58 Corrosion
#59 Plastic deformation
#60 Cyclic degradation
#61 Strong wind/wave
#62 Faulty welding
FF-FF-FM35 Hit by dropped objects Damage to the facility, vast economic loses #63 Planes crash
#64 Biological collision
#65 Strong wind/wave
FF-FF-FM36 Watertight fault Potential failure #66 Inefficient detection
#67 Pipe joint corrosion
#68 Pipe joint weld defect
#69 Pipe joint fatigue
#70 Pillar damage
#71 Excessive fouling of platform
(continued on next page)

12
H. Li et al. Ocean Engineering 234 (2021) 109261

(continued )
Code Failure Mode Level Code Failure Cause Level

Failure Modes End Effects Failure Causes

FF-FF-FM37 Additional structures fail Potential failure #72 Navigation and work lights fail
#73 Helicopter assistance equipment fail
#74 Handrails corrosion
#75 Ladders corrosion
#76 Dynamic umbilical connection fail
#77 Towing brackets/bollards fail
#78 Vents fail
#79 Bilge piping/pumps fail
#80 Sensors for platform monitoring fail
#81 Manholes fail
MS-ML-FM38 Abnormal mooring lines Mooring line strength decrease or broken #82 Mooring lines wear
#83 Mooring lines fatigue
#84 Mooring lines corrosion
#85 Abnormal stress
#86 Not effective maintenance
MS-ML-FM39 Mooring lines broken Malfunction of the whole system, the facility cannot locate in water #87 Transitional chain wear
#88 Friction chain wear
#89 Mooring winch failure
#90 Buoys friction chain wear
#91 Anchor pickup device damage
#92 Hydraulic motor failure
#93 Accumulator failure
#94 Over pressure
#95 Connectors failure
#96 Mooring interface structure failure
MS-FL-FM40 Fairlead failure The anchor cannot be dropped and lift #97 Fairlead corrosion
#98 Fairlead fatigue
MS-AC-FM41 Anchor failure Anchor failure #99 Abnormal working conditions
#100 Cyclic degradation
MS-RE-FM42 Abnormal functions Anchoring accuracy decrease #101 Poor operation environment
#102 Insufficient emergency measurement
#103 Human Error
#104 Analysis and calculation fault

Appendix B. Weighted failure indices, Indices’ Weights, RPNs and their ranking of failure causes

Code Weighted Indices (S Indices’ Weights (S RPN (* Rank Code Weighted Indices (S Indices’ Weights (S RPN (* Rank
O D) O D) 10− 2) O D) O D) 10− 2)

WT-BL-FM1- (0.0094 0.0088 (0.729 0.163 0.108) 0.96 52 WT-GE-FM12- (0.0073 0.0102 (0.674 0.226 0.1) 0.81 85
#1 0.0120) #19 0.0090)
WT-BL-FM2- (0.0094 0.0097 (0.731 0.188 0.081) 0.96 54 WT-GE-FM12- (0.0087 0.0084 (0.674 0.226 0.1) 0.89 70
#2 0.0108) #20 0.0114)
WT-BL-FM3- (0.0059 0.0093 (0.687 0.186 0.127) 0.73 98 WT-GB-FM13- (0.0084 0.0110 (0.661 0.208 0.131) 0.99 47
#3 0.0114) #21 0.0156)
WT-HB-FM4- (0.0080 0.0102 (0.614 0.268 0.118) 0.86 78 WT-GB-FM13- (0.0094 0.0102 (0.661 0.208 0.131) 1.02 39
#4 0.0078) #22 0.0138)
WT-HB-FM5- (0.0108 0.0110 (0.763 0.176 0.061) 1.08 23 WT-GB-FM14- (0.0105 0.0097 (0.674 0.226 0.1) 1.02 38
#5 0.0090) #23 0.0090)
WT-MB-FM6- (0.0122 0.0093 (0.627 0.28 0.093) 1.16 13 WT-GB-FM15- (0.0119 0.0049 (0.709 0.179 0.112) 1.00 44
#6 0.0144) #24 0.0060)
WT-MB-FM7- (0.0098 0.0141 (0.731 0.188 0.081) 1.06 25 WT-GB-FM15- (0.0119 0.0093 (0.709 0.179 0.112) 1.10 16
#7 0.0096) #25 0.0084)
WT-MS-FM8- (0.0126 0.0106 (0.717 0.195 0.088) 1.18 12 WT-GB-FM16- (0.0084 0.0110 (0.687 0.186 0.127) 0.92 62
#8 0.0078) #26 0.0108)
WT-MS-FM9- (0.0136 0.0084 (0.691 0.218 0.091) 1.20 10 WT-GB-FM17- (0.0087 0.0097 (0.731 0.188 0.081) 0.90 65
#9 0.0078) #27 0.0102)
WT-GE-FM10- (0.0077 0.0115 (0.674 0.226 0.1) 0.92 63 WT-GB-FM17- (0.0070 0.0115 (0.731 0.188 0.081) 0.80 87
#10 0.0138) #28 0.0096)
WT-GE-FM10- (0.0077 0.0110 (0.674 0.226 0.1) 0.90 68 WT-GB-FM17- (0.0091 0.0106 (0.731 0.188 0.081) 0.98 49
#11 0.0132) #29 0.0138)
WT-GE-FM10- (0.0077 0.0119 (0.674 0.226 0.1) 0.92 61 WT-GB-FM18- (0.0101 0.0102 (0.729 0.163 0.108) 1.05 30
#12 0.0132) #30 0.0138)
WT-GE-FM10- (0.0077 0.0102 (0.674 0.226 0.1) 0.88 73 WT-CV-FM19- (0.0094 0.0084 (0.614 0.268 0.118) 0.91 64
#13 0.0132) #31 0.0090)
WT-GE-FM11- (0.0084 0.0106 (0.558 0.32 0.122) 0.95 55 WT-CV-FM20- (0.0080 0.0062 (0.614 0.268 0.119) 0.75 95
#14 0.0120) #32 0.0072)
WT-GE-FM11- (0.0108 0.0124 (0.558 0.32 0.122) 1.14 14 WT-CV-FM20- (0.0084 0.0049 (0.614 0.268 0.119) 0.72 99
#15 0.0114) #33 0.0066)
WT-GE-FM11- (0.0087 0.0102 (0.558 0.32 0.122) 0.94 57 WT-CV-FM20- (0.0091 0.0053 (0.614 0.268 0.119) 0.77 93
#16 0.0102) #34 0.0060)
(0.674 0.226 0.1) 0.86 80 (0.614 0.268 0.119) 0.84 81
(continued on next page)

13
H. Li et al. Ocean Engineering 234 (2021) 109261

(continued )
Code Weighted Indices (S Indices’ Weights (S RPN (* Rank Code Weighted Indices (S Indices’ Weights (S RPN (* Rank
O D) O D) 10− 2) O D) O D) 10− 2)

WT-GE-FM12- (0.0070 0.0128 WT-CV-FM20- (0.0094 0.0066

#17 0.0096) #35 0.0066)
WT-GE-FM12- (0.0066 0.0102 (0.674 0.226 0.1) 0.79 91 WT-TR-FM21- (0.0098 0.0088 (0.661 0.208 0.131) 0.89 71
#18 0.0108) #36 0.0042)
WT-TR-FM22- (0.0091 0.0066 (0.674 0.226 0.1) 0.80 88 TT-TO-FM33- (0.0126 0.0132 (0.717 0.195 0.088) 1.26 7
#37 0.0042) #55 0.0108)
WT-TR-FM22- (0.0094 0.0084 (0.674 0.226 0.1) 0.88 74 TT-TO-FM33- (0.0129 0.0137 (0.717 0.195 0.088) 1.32 6
#38 0.0054) #56 0.0144)
WT-TR-FM22- (0.0087 0.0079 (0.674 0.226 0.1) 0.82 84 TT-TP-FM34- (0.0080 0.0110 (0.634 0.174 0.192) 0.99 46
#39 0.0048) #57 0.0150)
WT-TR-FM22- (0.0087 0.0084 (0.674 0.226 0.1) 0.83 83 TT-TP-FM34- (0.0087 0.0128 (0.634 0.174 0.192) 0.96 53
#40 0.0048) #58 0.0096)
WT-PS-FM23- (0.0080 0.0097 (0.665 0.245 0.09) 0.83 82 TT-TP-FM34- (0.0091 0.0097 (0.634 0.174 0.192) 0.95 56
#41 0.0066) #59 0.0108)
WT-PS-FM24- (0.0094 0.0115 (0.696 0.229 0.075) 0.99 45 TT-TP-FM34- (0.0084 0.0110 (0.634 0.174 0.192) 1.01 41
#42 0.0096) #60 0.0150)
WT-PS-FM25- (0.0080 0.0110 (0.674 0.226 0.1) 0.88 72 TT-TP-FM34- (0.0091 0.0124 (0.634 0.174 0.192) 0.96 51
#43 0.0090) #61 0.0090)
WT-YS-FM26- (0.0084 0.0084 (0.729 0.163 0.108) 0.87 77 TT-TP-FM34- (0.0098 0.0137 (0.634 0.174 0.192) 1.05 29
#44 0.0108) #62 0.0102)
WT-YS-FM27- (0.0080 0.0106 (0.0080 0.0106 0.88 75 FF-FF-FM35- (0.0143 0.0079 (0.618 0.297 0.085) 1.20 9
#45 0.0102) 0.0102) #63 0.0090)
WT-YS-FM28- (0.0084 0.0119 (0.661 0.208 0.131) 0.86 79 FF-FF-FM35- (0.0108 0.0088 (0.618 0.297 0.085) 1.04 34
#46 0.0042) #64 0.0126)
WT-CE-FM29- (0.0084 0.0071 (0.614 0.268 0.118) 0.80 89 FF-FF-FM35- (0.0150 0.0115 (0.618 0.297 0.085) 1.36 4
#47 0.0084) #65 0.0102)
WT-CE-FM30- (0.0084 0.0071 (0.594 0.249 0.157) 0.77 94 FF-FF-FM36- (0.0112 0.0093 (0.614 0.268 0.118) 1.05 32
#48 0.0060) #66 0.0096)
TT-TO-FM31- (0.0168 0.0053 (0.731 0.188 0.081) 1.38 1 FF-FF-FM36- (0.0105 0.0093 (0.614 0.268 0.118) 1.02 37
#49 0.0072) #67 0.0108)
TT-TO-FM31- (0.0161 0.0071 (0.731 0.188 0.081) 1.37 2 FF-FF-FM36- (0.0098 0.0106 (0.614 0.268 0.118) 1.00 43
#50 0.0078) #68 0.0096)
TT-TO-FM31- (0.0164 0.0062 (0.731 0.188 0.081) 1.36 3 FF-FF-FM36- (0.0105 0.0102 (0.614 0.268 0.118) 1.09 18
#51 0.0054) #69 0.0150)
TT-TO-FM31- (0.0150 0.0044 (0.731 0.188 0.081) 1.25 8 FF-FF-FM36- (0.0108 0.0097 (0.614 0.268 0.118) 1.00 42
#52 0.0084) #70 0.0066)
TT-TO-FM31- (0.0154 0.0084 (0.731 0.188 0.081) 1.33 5 FF-FF-FM36- (0.0084 0.0106 (0.614 0.268 0.118) 0.87 76
#53 0.0060) #71 0.0060)
TT-TO-FM32- (0.0129 0.0102 (0.729 0.163 0.108) 1.19 11 FF-FF-FM37- (0.0056 0.0093 (0.493 0.311 0.196) 0.68 101
#54 0.0078) #72 0.0060)
FF-FF-FM37- (0.0066 0.0071 (0.493 0.311 0.196) 0.70 100 MS-ML-FM39- (0.0119 0.0088 (0.661 0.208 0.131) 1.08 22
#73 0.0078) #89 0.0084)
FF-FF-FM37- (0.0042 0.0097 (0.493 0.311 0.196) 0.66 103 MS-ML-FM39- (0.0108 0.0119 (0.661 0.208 0.131) 1.08 20
#74 0.0078) #90 0.0090)
FF-FF-FM37- (0.0042 0.0097 (0.493 0.311 0.196) 0.66 103 MS-ML-FM39- (0.0108 0.0115 (0.661 0.208 0.131) 1.09 19
#75 0.0078) #91 0.0102)
FF-FF-FM37- (0.0056 0.0079 (0.493 0.311 0.196) 0.68 102 MS-ML-FM39- (0.0105 0.0102 (0.661 0.208 0.131) 1.02 36
#76 0.0078) #92 0.0090)
FF-FF-FM37- (0.0059 0.0084 (0.493 0.311 0.196) 0.74 96 MS-ML-FM39- (0.0108 0.0102 (0.661 0.208 0.131) 1.05 31
#77 0.0096) #93 0.0096)
FF-FF-FM37- (0.0066 0.0093 (0.493 0.311 0.196) 0.73 97 MS-ML-FM39- (0.0091 0.0097 (0.661 0.208 0.131) 0.92 59
#78 0.0060) #94 0.0090)
FF-FF-FM37- (0.0073 0.0097 (0.493 0.311 0.196) 0.81 86 MS-ML-FM39- (0.0112 0.0084 (0.661 0.208 0.131) 1.06 28
#79 0.0072) #95 0.0108)
FF-FF-FM37- (0.0066 0.0106 (0.493 0.311 0.196) 0.80 90 MS-ML-FM39- (0.0119 0.0066 (0.661 0.208 0.131) 1.06 27
#80 0.0072) #96 0.0102)
FF-FF-FM37- (0.0066 0.0084 (0.493 0.311 0.196) 0.78 92 MS-FL-FM40- (0.0094 0.0097 (0.709 0.179 0.112) 0.98 48
#81 0.0096) #97 0.0120)
MS-ML-FM38- (0.0105 0.0115 (0.614 0.268 0.118) 1.07 24 MS-FL-FM40- (0.0105 0.0106 (0.709 0.179 0.112) 1.09 17
#82 0.0102) #98 0.0144)
MS-ML-FM38- (0.0108 0.0106 (0.614 0.268 0.118) 1.11 15 MS-AC-FM41- (0.0091 0.0088 (0.674 0.226 0.1) 0.90 67
#83 0.0138) #99 0.0090)
MS-ML-FM38- (0.0101 0.0124 (0.614 0.268 0.118) 1.05 33 MS-AC-FM41- (0.0087 0.0088 (0.674 0.226 0.1) 0.92 60
#84 0.0078) #100 0.0132)
MS-ML-FM38- (0.0101 0.0102 (0.614 0.268 0.118) 1.06 26 MS-RE-FM42- (0.0087 0.0093 (0.652 0.235 0.112) 0.90 69
#85 0.0138) #101 0.0096)
MS-ML-FM38- (0.0101 0.0088 (0.614 0.268 0.118) 0.97 50 MS-RE-FM42- (0.0091 0.0097 (0.652 0.235 0.112) 0.93 58
#86 0.0096) #102 0.0096)
MS-ML-FM39- (0.0115 0.0093 (0.661 0.208 0.131) 1.08 21 MS-RE-FM42- (0.0091 0.0119 (0.652 0.235 0.112) 1.01 40
#87 0.0096) #103 0.0126)
MS-ML-FM39- (0.0108 0.0093 (0.661 0.208 0.131) 1.03 35 MS-RE-FM42- (0.0091 0.0075 (0.652 0.235 0.112) 0.90 66
#88 0.0090) #104 0.0120)

14
H. Li et al. Ocean Engineering 234 (2021) 109261

References Kang, J., Sun, L., Sun, H., Wu, C., 2017. Risk assessment of floating offshore wind turbine
based on correlation-FMEA. Ocean Eng. 129, 382–388.
Kang, J., Sobral, J., Guedes Soares, C., 2019a. Review of condition-based maintenance
Arabian-Hoseynabadi, H., Oraee, H., Tavner, P.J., 2010. Failure modes and effects
strategies for offshore wind energy. J. Mar. Sci. Appl. 18 (1), 1–16.
analysis (FMEA) for wind turbines. Int. J. Electr. Power Energy Syst. 32 (7),
Kang, J., Sun, L., Guedes Soares, C., 2019b. Fault Tree Analysis of floating offshore wind
817–824.
turbines. Renew. Energy 133, 1455–1467.
Arapogianni, A., Genachte, A.B., Ochagavia, R.M., Vergara, J.P., Castell, D.,
Lee, Y.S., González, J.A., Lee, J.H., Kim, Y.I., Park, K.C., Han, S., 2016. Structural
Tsouroukdissian, A.R., Korbijn, J., Bolleman, N.C.F., Huera-Huarte, F.J., Schuon, F.,
topology optimization of the transition piece for an offshore wind turbine with jacket
Ugarte, A., Sandberg, J., Laleu, V.D., Maciel, J., Tunbjer, A., Roth, R., Gueriviere, P.
foundation. Renew. Energy 85, 1214–1225.
D.L., Coulombeau, P., Jedrec, S., Philippe, C., Voutsinas, S., Weinstein, A., Vita, L.,
Leimeister, M., Kolios, A., 2018. A review of reliability-based methods for risk analysis
Byklum, E., Hurley, W.L., Grubel, H., 2013. Deep Water—The Next Step for Offshore
and their application in the offshore wind industry. Renew. Sustain. Energy Rev. 91,
Wind Energy. European Wind Energy Association (EWEA). Available at: https://teth
1065–1076.
ys.pnnl.gov/sites/default/files/publications/EWEA-2013.pdf. (Accessed 25
Li, H., Guedes Soares, C., 2019. Reliability analysis of floating offshore wind turbines
February 2021).
support structure using hierarchical Bayesian network. In: Beer, M., Zio, E. (Eds.),
Athanasia, A., Genachte, A.B., 2013. Deep offshore and new foundation concepts. Energy
Proceedings of the 29th European Safety and Reliability Conference. Research
Procedia 35, 198–209.
Publishing Services, Singapore, pp. 2489–2495.
Bento, N., Fontes, M., 2019. Emergence of floating offshore wind energy: technology and
Li, H., Guedes Soares, C., Huang, H.Z., 2020a. Reliability analysis of floating offshore
industry. Renew. Sustain. Energy Rev. 99, 66–82.
wind turbine using Bayesian Networks. Ocean Eng. 217, 107827.
Bharatbhai, M.G., 2015. Failure mode and effect analysis of repower 5M wind turbine.
Li, H., Teixeira, A.P., Guedes Soares, C., 2020b. A two-stage Failure Mode and Effect
Int. J. Adv. Res. Eng. 2 (5), 2394–2444.
Analysis of offshore wind turbines. Renew. Energy 162, 1438–1461.
Bhardwaj, U., Teixeira, A.P., Guedes Soares, C., 2019. Reliability prediction of an
Li, H., Diaz, H., Guedes Soares, C., 2021. A developed failure mode and effect analysis for
offshore wind turbine gearbox. Renew. Energy 141, 693–706.
floating offshore wind turbine support structures. Renew. Energy 164, 133–145.
Carpitella, S., Certa, A., Izquierdo, J., La Fata, C.M., 2018. A combined multi-criteria
Liu, H.C., Chen, X.Q., Duan, C.Y., Wang, Y.M., 2019. Failure mode and effect analysis
approach to support FMECA analyses: a real-world case. Reliab. Eng. Syst. Saf. 169,
using multi-criteria decision making methods: a systematic literature review.
394–402.
Comput. Ind. Eng. 135, 881–897.
Castro-Santos, L., Martins, E., Guedes Soares, C., 2016. Cost assessment methodology for
Moore, A., Price, J., Zeyringer, M., 2018. The role of floating offshore wind in a
combined wind and wave floating offshore renewable energy systems. Renew.
renewable focused electricity system for Great Britain in 2050. Energy Strat. Rev. 22,
Energy 97, 866–880.
270–278.
Castro-Santos, L., Silva, D., Bento, A.R., Salvacao, N., Guedes Soares, C., 2020. Economic
Musial, W., Heimiller, D., Beiter, P., Scott, G., Draxl, C., 2016. 2016 Offshore Wind
feasibility of floating offshore wind farms in Portugal. Ocean Eng. 207, 107393.
Energy Resource Assessment for the United States (No. NREL/TP-5000-66599).
Cevasco, D., Collu, M., Lin, Z., 2018. O&M cost-Based FMECA: identification and ranking
National Renewable Energy Lab. (NREL) Golden, CO, United States.
of the most critical components for 2-4 MW geared offshore wind turbines. J. Phys.
Salvacao, N., Guedes Soares, C., 2018. Wind resource assessment offshore the Atlantic
Conf. 1102 (1), 012039.
Iberian coast with the WRF model. Energy 145, 276–287.
Chou, J.S., Tu, W.T., 2011. Failure analysis and risk management of a collapsed large
Santos, F.P., Teixeira, A.P., Guedes Soares, C., 2015. Review of wind turbine accident
wind turbine tower. Eng. Fail. Anal. 18, 295–313.
and failure data. In: Guedes Soares, C. (Ed.), Renewable Energies Offshore. Taylor &
Colmenar-Santos, A., Perera-Perez, J., Borge-Diez, D., dePalacio-Rodríguez, C., 2016.
Francis Group, London, UK, pp. 953–959.
Offshore wind energy: a review of the current status, challenges and future
Scheu, M.N., Tremps, L., Smolka, U., Kolios, A., Brennan, F., 2019. A systematic Failure
development in Spain. Renew. Sustain. Energy Rev. 64, 1–18.
Mode Effects and Criticality Analysis for offshore wind turbine systems towards
Díaz, H., Guedes Soares, C., 2020a. An integrated GIS approach for site selection of
integrated condition based maintenance strategies. Ocean Eng. 176, 118–133.
floating offshore wind farms in the Atlantic continental European coastline. Renew.
Shafiee, M., Dinmohammadi, F., 2015. An FMEA-based risk assessment approach for
Sustain. Energy Rev. 134, 110328.
wind turbine systems: a comparative study of onshore and offshore. Energies 7,
Díaz, H., Guedes Soares, C., 2020b. Review of the current status, technology and future
619–642.
trends of offshore wind farms. Ocean Eng. 209, 107381.
Sinha, Y., Steel, J.A., 2015. A progressive study into offshore wind farm maintenance
Dincer, I., 2000. Renewable energy and sustainable development: a crucial review.
optimisation using risk based failure analysis. Renew. Sustain. Energy Rev. 42,
Renew. Sustain. Energy Rev. 4 (2), 157–175.
735–742.
Dinmohammadi, F., Shafiee, M., 2013. A fuzzy-FMEA risk assessment approach for
Spreafico, C., Russo, D., Rizzi, C., 2017. A state-of-the-art review of FMEA/FMECA
offshore wind turbines. Int. J. Prognostics Health Manag. 4, 59–68.
including patents. Comput. Sci. Rev. 25, 19–28.
Du, M., Yi, J., Guo, J., Cheng, L., Ma, S., He, Q., 2017. An improved FMECA method for
Stamatis, D.H., 2003. Failure Mode and Effect Analysis: FMEA from Theory to Execution.
wind turbines health management. Energy Power Eng. 9, 36–45.
Quality Press.
Fattahi, R., Khalilzade, M., 2018. Risk evaluation using a novel hybrid method based on
Tazi, N., Châtelet, E., Bouzidi, Y., 2017. Using a hybrid cost-FMEA analysis for wind
FMEA, extended MULTIMOORA, and AHP methods under fuzzy environment. Saf.
turbine reliability analysis. Energies 10 (3), 276.
Sci. 102, 290–300.
Turner, J.A., 1999. A realizable renewable energy future. Science 285, 687–689.
Global Wind Energy Council (GWEC), 2020. Global wind report 2019. Available at: http
Uzunoglu, E., Guedes Soares, C., 2020. Hydrodynamic design of a free-float capable
s://gwec.net/global-wind-report-2019/. (Accessed 29 July 2020).
tension leg platform for a 10 MW wind turbine. Ocean Eng. 197, 106888.
Goepel, K.D., 2013. Implementing the analytic hierarchy process as a standard method
Uzunoglu, E., Karmakar, D., Guedes Soares, C., 2016. Floating offshore wind platforms.
for multi-criteria decision making in corporate enterprises – a new AHP excel
In: Castro-Santos, L., Diaz-Casas, V. (Eds.), Floating Offshore Wind Farms. Springer
template with multiple inputs. In: Proceedings of the International Symposium on
International Publishing Switzerland, pp. 53–76.
the Analytic Hierarchy Process, pp. 1–10.
Wang, Z., Ran, Y., Chen, Y., Yu, H., Zhang, G., 2020. Failure mode and effects analysis
Grošelj, P., Stirn, L.Z., Ayrilmis, N., Kuzman, M.K., 2015. Comparison of some
using extended matter-element model and AHP. Comput. Ind. Eng. 140, 106233.
aggregation techniques using group analytic hierarchy process. Expert Syst. Appl. 42
Wu, X., Hu, Y., Li, Y., Yang, J., Duan, L., Wang, T., Adcock, T., Jiang, Z., Gao, Z., Lin, Z.,
(4), 2198–2204.
Borthwick, A., Liao, S., 2019. Foundations of offshore wind turbines: a review.
Hu, A.H., Hsu, C.W., Kuo, T.C., Wu, W.C., 2009. Risk evaluation of green components to
Renew. Sustain. Energy Rev. 104, 379–393.
hazardous substance using FMEA and FAHP. Expert Syst. Appl. 36 (3), 7142–7147.
Xiao, N., Huang, H.Z., Li, Y., He, L., Jin, T., 2011. Multiple failure modes analysis and
Huang, J., You, J.X., Liu, H.C., Song, M.S., 2020. Failure mode and effect analysis
weighted risk priority number evaluation in FMEA. Eng. Fail. Anal. 18 (4),
improvement: a systematic literature review and future research agenda. Reliab.
1162–1170.
Eng. Syst. Saf. 199, 106885.
Xu, S., Wang, S., Guedes Soares, C., 2019. Review of mooring design for floating wave
Kahrobaee, S., Asgarpoor, S., 2011. Risk-based failure mode and effect analysis for wind
energy converters. Renew. Sustain. Energy Rev. 111, 595–621.
turbines (RB-FMEA). In: 2011 North American Power Symposium, pp. 1–7.

15