130
CHAPTER 8
FAILURE MODES AND EFFECTS ANALYSIS OF WT
8.1 INTRODUCTION
Wind energy stands out to be one of the most promising new and
renewable sources of generating electrical power for any country. Wind
electric generator converts kinetic energy available in wind to electrical
energy by using rotor, gear box and generator. The wind energy is widely
used because of its environmental friendliness and many countries have good
wind potential to harness energy.
Failure Modes and Effects Analysis (FMEA) is a step-by-step
approach for identifying all potential problems in a design, a manufacturing or
assembly process, or a product or service. “Failure Modes” means the ways,
or modes, in which something might fail. Failures are any errors or defects,
especially one that affects the end user, and can be potential or actual.
“Effects Analysis” refers to studying the consequences of those failures. Its
most visible result is the documentation of collective knowledge of cross
functional teams. FMEA can be adopted during design stage to prevent
failures afterwards it used for control the failures during operation. The
different types of FMEA are System FMEA, Design FMEA, Process FMEA,
Service FMEA and Software FMEA. FMEA helps us to identify the different
prospective failures and used to develop the requirements that minimize the
likelihood of those failures.
131
FMEA technique is designed for the following reasons.
1. To identify potential failure modes for a product or process
2. To assess the risk associated with those failure modes
3. To endow with numerical rank for issues in terms of
importance
4. To prioritize according to their serious consequences
5. To identify and carry out corrective actions for the most
serious issues.
8.2 REQUIREMENTS OF FMEA
In general, FMEA requires the identification of the following basic
information:
1. Item(s)
2. Function(s)
3. Failure(s)
4. Effect(s) of Failure
5. Cause(s) of Failure
6. Current Control(s)
7. Recommended Action(s)
8.3 NECESSITY OF FMEA
1. When a process, product or service is being designed or
redesigned
132
2. When an existing process, product or service is being applied
in a new way
3. Before developing control plans for a new or modified
process
4. When enhancement goals are planned for an existing process,
product or service
5. When analyzing failures of an existing process, product or
service
6. Round the clock life of the process, product or service
8.4 FMEA PROCEDURE
1. Assemble the cross functional team with diverse knowledge
about the process, product or service and define the scope
2. Establish the ground rules
3. Collect data and review all relevant information
4. Identify the parts, components, system or processes to be
analyzed
5. Identify the potential failures, effects and causes
6. Determine the severity rating, or S
7. Determine the occurrence rating, or O, used to estimate the
probability of failure
8. Determine the detection rating, or D, used to estimate the
controls systems used to detect the cause or its failure mode
133
9. Calculate the Risk Priority Number, or RPN, which equals
S × O × D. and also calculate the Criticality by multiplying
Severity by Occurrence, S × O
10. Evaluate the risk associated with the issues identified by the
analysis
11. Prioritize and assign corrective actions
12. Perform corrective actions and re-evaluate risk
13. Distribute, review and update the analysis as appropriate
8.5 GROUND RULES AND ASSUMPTIONS OF FMEA
Before detailed analysis takes place, ground rules and assumptions
are usually defined and it must include,
1. Standardized mission profile with specific fixed duration
mission phases
2. Sources for failure rate and failure mode data
3. Fault detection coverage that system built-in test will realize
4. Whether the analysis will be functional or piece part
5. Criteria to be considered (mission abort, safety, maintenance,
etc.)
6. System for uniquely identifying parts or functions
7. Severity category definitions
134
8.6 WT HIERARCHY FOR FMEA
The WTS can be classified into two main categories according to
pitch control, constant and variable pitch. Similarly, according to speed, it is
classified into constant and variable speed turbines. The WT consists of rotor
system, gear system, brake system, generator system, hydraulics system,
electrical system etc as shown in Figure 1.3
A system FMEA is made up of various systems. A FMEA sub
system is a subset of a system FMEA. A component FMEA is a sub set of sub
system. In this research, the WT sub assemblies are considered for evaluation.
The main parts of the rotor considered for the analysis are hub, main bearing,
blade and rotor control system.
Figure 8.1 WT Hierarchy for FMEA
The failure of the WTS is defined through three stages as shown in
Figure 8.1. The WT is placed in the first stage (stage I); whereas, WT sub
135
assemblies and parts are in second (stage II) and third (stage III) levels
respectively.
In rotor system the blades, hub, main bearing and nose cone were
considered and in gear box system the low speed shaft, high speed shaft, input
shaft, input pinion, bearings, gears, cooling system, sensors and temperature
control were considered In brake system the important parts are brake pad,
spring, brake shoe and mechanical control. In generator system the important
components are the contactors, shaft, bearing, flexible coupling, generator
cooling system, top terminal box and generator control at control panel.
The key yaw system elements taken into account of analysis are
yaw motor, yaw gear, yaw planetary, yaw bearing, yaw brake, drives,
controllers and wind vane. The important rotor hydraulic system assembly
constituents are accumulator, hydraulic cylinder, hydraulic pump, oil
reservoir, pump and hydraulic transmission system. The hydraulic brake
system with hydraulic pump, oil reservoir and hydraulic brake control, brake
solenoid, accumulator and hydraulic cylinder were considered.
The Table 8.1 shows the severity, occurrence and detection rating
scale for WT in high uncertain wind area of Muppandal site.
136
Tabl
e 8.
1 Se
veri
ty, O
ccur
renc
e an
d D
etec
tion
ratin
g sc
ale
for
WT
Scal
eN
o.
Seve
rity
Occ
urre
nce
Det
ectio
n
Ver
y H
igh
Ver
y H
igh
likel
ihoo
d th
at th
ecu
rren
t con
trols
will
det
ect t
heFa
ilure
Mod
e
Hig
h
Hig
hlik
elih
ood
that
the
curr
ent
cont
rols
will
det
ect.
Mod
erat
ely
Hig
h
Mod
erat
ely
high
like
lihoo
d th
at th
ecu
rren
t con
trols
will
det
ect t
hefa
ilure
mod
e.
Mod
erat
e
Mod
erat
e lik
elih
ood
that
the
curr
ent c
ontro
ls w
ill d
etec
t the
failu
re m
ode.
137
Tab
le 8
.1 (C
ontin
ued)
Scal
eN
o.
Seve
rity
Occ
urre
nce
Det
ectio
n
Low
Low
like
lihoo
d th
at th
e cu
rren
tco
ntro
ls w
ill d
etec
t the
Fai
lure
Mod
e
Rea
sona
bly
Prob
able
Ver
y Lo
wV
ery
Low
like
lihoo
d th
at th
e cu
rren
tco
ntro
ls w
ill d
etec
t the
Fai
lure
Mod
e
8R
emot
e
Rem
ote
likel
ihoo
d th
at th
e cu
rren
tco
ntro
ls w
ill d
etec
t the
Fai
lure
Mod
e
9V
ery
Rem
ote
Ver
y R
emot
e lik
elih
ood
that
the
curr
ent c
ontro
l will
det
ect t
he fa
ilure
mod
e
10N
o kn
own
cont
rol i
s ava
ilabl
e to
dete
ct th
e fa
ilure
138
8.7 FMEA
A failure mode is defined as the behavior in which a system, sub
assembly, part etc. could potentially fail to meet the design intent. FMEA is
an important reliability tool to explore the ways or modes in which each
system element can fail and assesses the consequences of each of these
failures. Failures are prioritized according to how serious their consequences
are, how frequently they occur and how easily they can be detected.
The purpose of the FMEA of wind turbines is to take actions to
eliminate or reduce failures, starting with the highest-priority ones. Risk
Priority Number (RPN) is used to evaluate the risk associated with the
potential problems and it helps to identify the critical failure modes associated
with the design or process. The RPN can be calculated by using
Equation (8.1).
RPN = Severity (S) X Occurrence (O) X Detection (D) (8.1)
The severity, occurrence and detection are rated for 10 point scales.
Therefore, RPN varies from 1 to1000, that is 1 represents absolute best and
1000 represents the worst.
8.7.1 Severity
The Severity (S) is used to estimate the most serious effect of
failures. This is the severity rating, or S. Severity is rated on a scale from 1 to
10 in this research, based on the seriousness required for the WT components,
where 1 means ‘No Effect’ and 10 means ‘Failure’ affecting the system
operation and safety without any warning.
139
8.7.2 Occurrence
The Occurrence (O) is used to estimate the likelihood that the cause,
if it occurs, will produce the failure mode and its particular effect and it is also
called as likelihood. This rating estimates the probability of failure occurring
for that reason during the lifetime of the system. Occurrence is usually rated
on a scale from 1 to 10, where 1 is extremely unlikely and 10 is inevitable. A
failure cause is used to identify the design weakness of the system that may
result in a failure.
8.7.3 Detection
The Detection (D) is termed as the effectiveness. It is used to
estimate the effectiveness of the controls to prevent or detect the cause or
failure mode. The assumption is that the cause has occurred. This rating
estimates how well the controls can detect either the cause or its failure mode
after they have occurred, but, before the customer is affected. Detection is
usually rated on a scale from 1 to 10, where 1 means the control is absolutely
certain to detect the problem and 10 means the control is certain not to detect
the problem values. The inspection is carried out by visual, auditive and
olfactive. The maintenance method followed in wind farms are time based
preventive maintenance.
8.8 FMEA FOR WT
In this research, the twenty numbers of 250 kW WT and its seven
major sub assemblies such as rotor, gear box, brake system, generator, yaw
system, rotor hydraulic control system and brake hydraulic system were
considered. The sub assemblies considered for FMEA analysis include parts
and its control system.
140
8.8.1 Failure Modes Considered For WT
The twenty failure modes considered for 250 kW WT and its sub
assemblies.
Mechanical
1. Misalignment of shaft
2. Fracture failure
3. Bed bolt shear
4. Friction
5. Wear out
6. Over heat
7. Brittle seals
8. Body crack
9. Material
10. Lightening
11. Flutter
12. Impact load
Electrical
1. Voltage fluctuation
2. Sensor fault
3. Control panel fault
4. Irregular power supply and output
141
5. Cable fault
6. Control system fault
8.8.2 Failure Mechanism of WT
1. Prolonged excessive vibration of individual components
2. Manufacturing defect in gear box casing
3. Fatigue and dynamic load
4. Excessive wear
5. Contamination
6. High voltage
7. Prolonged high temperature
8.8.3 Failure Effect of WT
1. Loss of power generation
2. Failure of WT components
3. Shutdown
4. Generator burn out
5. Control panel failure
6. Oil pitting
7. Tip open when running
8. Yaw motor rapture
142
8.9 RANKING OF EFFECTS OF WT BY INITIAL SEVERITY
The average severity of twenty numbers of 250 kW WT is from 5 to
8 as shown in Figure 8.2.
Figure 8.2 Ranking of Effects of WT by Initial Severity
The maximum severity of 8 is allotted to rotor. The severity for
rotor is high because if there is any mild crack or damage caused in the blade
or tiny fault of the main bearing and rotor shaft, which will reduce the
generation significantly. Next to rotor, the generator and gear box have a
severity of 7. The present design of the gear box of the WT has a short life
due to improper material, improper lubrication system and inefficient cooling
system. The electrical fault in the generator is somewhat predictable than
mechanical fault in the rotor. But the generators are the second key element
for power generation. The severity allotted for yaw system, brake system and
rotor hydraulic system is 6 based on their individual component seriousness to
the system. The low severity is given for hydraulic control of the brake
system as 5, because, once hydraulic control fails, the mechanical and
electrical actuated brake will take care.
143
8.10 RANKING OF CAUSES OF WT BY INITIAL OCCURRENCE
The occurrence of individual component of the WT is assigned
according to the failure rates as specified range in the Table 8.1. The high
frequency of failures of yaw system, hydraulic system of rotor and generator
had brought the occurrence rate to 6 as shown in Figure 8.3. The reason for
the variation of occurrence from 4 to 6 is that the single failure mode
probability of failure occurs in between 0.0001 and 0.0004. In the constant
pitch and constant speed WT, the average failure rates of the yaw, hydraulic
control of rotor and generator lie between 0.00038 failures per hour,
0.000354and 0.000313. The failure rate per hour of gear box system is in
between 0.0000103 and 0.000467. The average failure rate for 20 wind
turbines per hour of gear box system is 0.00286. Therefore, the occurrence
allocated as 5. The failure rate per hour of hydraulic control of brake system
is in between 0.000122 and 0.000398 and its average failure rate is 0.000259.
The lowest failure rate per hour in the rotor is obtained as 0.0001833 and
brake is 0.000191. The occurrence value is assigned as 4.
Figure 8.3 Ranking of causes of WT by Initial Occurrence
144
8.11 RANKING OF CAUSES OF WT BY INITIAL DETECTION
The detection is rated by the scales as rated on the Table 8.1. The
detection refers to the likelihood of detecting a root cause before a failure can
occur. It enforces requirement of any measurement device. The brake system
of the WT has poor level of design control of detection as indicated in Figure
8.4.since it is difficult to identify the wear of brake pad. Therefore, the highest
detection value assigned for brake system as 9. In the same way, the fault
detection is unfeasible for the yaw system and all hydraulic rotor and
hydraulic system of brake controls. The detection value assigned to them is 7.
The brake system, yaw system and hydraulic control of brake and rotor
system require urgent action to detect the failure mode before it occurs. The
gear system has a detection device such as temperature sensor, vibration
sensor and gear oil level indicator. The rotor and generator have a good
detection device.
Figure 8.4 Ranking of causes of WT by Initial Detection
145
8.12 CAUSES RANKED BY INITIAL RPN
The causes ranked by initial RPN is shown in Figure 8.5 and
Appendix 2. It shows the relative risk of the WT components which have high
priority. RPN is calculated by multiplying the Severity by the Occurrence by
the Detection of the risk. In this study, the highest value of RPN obtained is
252 for yaw system and the lowest is 160 for rotor. It is noted that the most of
the failures are contributed by yaw, hydraulic control of rotor, brake and gear.
The RPN of yaw system and rotor hydraulic control are high and this is due to
the high severity, absence of detective mechanisms and increased frequency
of failures over a period of time. The brake system and gear system have a
RPN of 216 and 210 respectively. Although the rotor has high severity, the
RPN is less (168) due to low detection and occurrence. But, rotor as a whole
system of hydraulic and mechanical system, the risks are extremely large. The
generator has high severity and occurrence, but it has low RPN (168) because
of diminutive detection value.
Figure 8.5 Ranking of causes of WT by Initial RPN
146
8.13 MODIFICATION AND ANALYSIS OF WT COMPONENTS
The modification and redesign is implemented on trail basis in a
specific location number 10, where the failures of individual components are
high. The failure of yaw system during high wind causes reduction in power
generation and imparts high thrust on the blade which causes major failure.
To reduce the yaw failure an active redundant soft yaw system is fixed. The
soft yaw system provides smooth start and stop without noise and yaw brakes
are totally eliminated. Initially, the numbers of failures were 6 to 10 per year
over a span of fifteen years. After implementation, the failures come down by
20 to 33 % when compared with conventional yaw system. The redundant
yaw system not only affords smoothness to the system but also eliminates the
risks of the rotor system.
The gear oil test facility is incorporated in the site to test kinematic
viscosity, density, flash point and metal content. The gear oil analysis is
performed on site laboratories so that the information about the future damage
and the correct time for oil change are easily revealed. The vibration meter is
fixed on various parts of WT components like shaft, gear, yaw system,
generator and various positions of the tower. Based on the outcome of the
vibration meter, the bolts are adequately torque to make the structure rigid for
increasing the span of MTTF. The pressure gauges have been set in rotor and
brake hydraulic system to gauge the system pressure to ensure smooth
operations of the turbine. To measure the rate and extend of wear of the brake
pads, infrared position sensor was installed at appropriate place. This helps
for planning predictive maintenance instead of breakdown maintenance.
8.14 RANKING OF CAUSES BY INITIAL AND REVISED RPN
The modification of WT components are shown in Appendix 3 and
Figure 8.6. After modification, the RPN varies from 96 to 150 indicating
147
reasonable improvement in reducing the relative risks considerably. The
reduction of occurrence by diminishing the failure rate and the reduction of
detection by augmenting new measurement systems can trim down the RPN
to some extent. In two other important components, the rotor and generator,
with a reduction of failure rate and high detection rate, the net RPN is
relatively low though they have high severity indices.
Figure 8.6 Ranking of causes of WT by Initial and Revised RPN
The gear system is ranked third in the RPN causes. It is proved that
the gear system is highly critical and it requires best control system. The
hydraulic control of brake has the least RPN of 175.
8.15 CAUSES RANKED BY RPN % REDUCTION
The improvements in the components RPN in term of percentage
are represented in the graph shown in Figure 8.7. It can be observed that it is
maximum in yaw system as 52% and lowest of 40 %in the rotor. The highest
reduction of RPN in yaw is due to the provision of active standby redundant
yaw system. The installation of the infrared position sensor in brake system
resulted in 50% reduction in RPN. Simple periodic maintenance and
148
predictive maintenance of shaft, sleeve ring, bearing and cooling system of
generator brings the RPN reduction. The new design, pressure regulation
system brought the improvements in hydraulic controls of brake and rotor.
Figure 8.7 Causes Ranked by RPN % Reduction
8.16 CRITICALITY ANALYSIS USING SEVERITY AND
OCCURRENCE OF WT
The Criticality analysis is a procedure by which each potential
failure mode is ranked according to the combined influence of severity and
probability of occurrence. The criticality value is obtained by the product of
severity and occurrence. The criticality of failure is independent of detection
because detection is failure prevention through design.
The criticality index is the quantitative measure of criticality of the
failure mode that combines the probability of the failure mode’s occurrence
with its severity ranking. For each severity categorization, the criticality
index is calculated for each of the corresponding failure. The result is a rank
149
ordering of failure modes within each severity classifications. The index is
defined as
k k p k pC t (8.2)
Where
Ck = Critical index of failure mode ‘k’
kp = Fraction of the component p’s failure having failure
mode k given component ‘p’ has failed
k = Conditional probability that failure modes ‘k’ will
result in the identified failure effect
p = Failure Rate of component ‘p’
t = Duration of time in analysis
The generator is considered as a highly critical component and it has
a criticality value of 42 as shown in Figure 8.8.
Figure 8.8 Criticality Ranking of WT
150
8.16.2 Initial Occurrence and Severity matrix
Subsequently, yaw is placed in the second level of criticality,
because of the high frequency of failure and moderate severity. After that, the
gear system and rotor system have high criticality ranks due to the high
severity and very moderate occurrence. The brake has low criticality because
of low severity. The criticality is trivial in hydraulic controls of brake and
rotor.
Figure 8.9 Initial Occurrence and Severity matrix
8.15.1 Revised Occurrence and Severity matrix
The Figure 8.8 compares the failure modes by way of a Criticality
ranking. The occurrence / severity matrix shown in Figure 8.9 and 8.10
exposes that all the seven components of wind turbines are having very high
priority causes because, they are above the high priority line. It implies that
the immediate rectification action is required to solve the crisis. The yaw
(coordinates 6, 6) and the generator (co ordinates 7, 6) are fall on very high
priority causes. Both rotor and brake hydraulic systems require action slightly
151
lower than the other priority causes. After implementation, the occurrence is
reduced with less investment and priority is brought nearer to the high priority
line as shown in Figure 8.10. But, the graph exposes that it requires some
serious design modifications.
Figure 8.10 Revised Occurrence and Severity matrix
Figure 8.11 Yearly generation at redesign implemented location
152
After implementation, the generation obtained in the year 2012 is
5,99,874kW.hr, which is very high as shown in Figure 8.11, when compared
with the past generations over 15 years.
The FMEA study carried out successfully to twenty numbers of
constant speed and constant pitch 250 kW wind turbines over a span of 15
years by using Xfmea software. This study has a prospective of improving the
reliability and availability of wind turbines by modifying and redesigning the
components, redundancy system, effective monitoring mechanism and proper
preventive maintenance. From the FMEA, it is evident that immediate actions
are required in generator, yaw system, rotor system and gear system. The
generator is constructed overhanging and this leads to repeated misalignment
of shaft due to vibrations. The high uncertainty wind causes voltage
fluctuations and the inadequate performance of sleeve ring, carbon brush and
bearing diminish the reliability and availability of the generator. The failures
in the generator are early detected and restricted with help of the continuous
monitoring of vibration meter. It is inevitable that the yaw system requires
redundant system to increase its performance. The improvement in lubrication
system and efficient cooling system makes 47% improvement in RPN of the
gear box.
The rotor is overstressed due to the frequent failure of yaw
mechanism. It is found that the smooth and efficient yaw mechanism with
active redundancy reduces 52 % of RPN. The crack on the blade edge and
tip, improper blade material and fatigue load brought down the rotor
efficiency. It is necessary to set up design control for detecting the brake
failure. The frequent inspection of the brake pad with a help of infrared
position sensor, reduces failure to 50%. It is suggested that the yaw system
and brake system should be inspected in weekly schedule and preventive
maintenance be carried out effectively. The FMEA has the potential to
153
improve the reliability of WTSs at high uncertain wind environment, where
reliability plays an important role. Furthermore, it is believed that in time, it
will play a key function in the monitoring and maintenance to make wind
power generation a more efficient.
8.17 SUMMARY
The step-by-step Failure modes and effects analysis is carried out in
this chapter for identifying all potential problems of WT and its sub assembly.
The RPN is computed for existing WT based on severity, occurrence and
detection. After modification, the revised RPN is obtained. The reduction of
occurrence by diminishing the failure rate and the reduction of detection by
augmenting new measurement systems can trim down the RPN to some
extent. The revised RPN indicates 40 to 52 percent improvement in reducing
the relative risks considerably. The criticality index derived from FMEA is
calculated for each consequent failure.