This article was published in an Elsevier journal.

The attached copy

is furnished to the author for non-commercial research and
education use, including for instruction at the author’s institution,
sharing with colleagues and providing to institution administration.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:

http://www.elsevier.com/copyright
 Author's personal copy

Engineering Failure Analysis 15 (2008) 129–141

www.elsevier.com/locate/engfailanal

Steam turbine blade failure analysis

Zdzislaw Mazur *, Rafael Garcia-Illescas, Jorge Aguirre-Romano,
Norberto Perez-Rodriguez
Instituto de Investigaciones Eléctricas Turbomaquinaria, Reforma 113, Col. Palmira 62490, Cuernavaca, Morelos, Mexico

Received 3 August 2006; accepted 20 November 2006

Available online 22 January 2007

Abstract

A last stage turbine blades failure was experienced in two units of 660 MW. These units have one high-pressure turbine
and two tandem-compound low-pressure turbines with 44-in. last-stage blades. The blades that failed were in a low pres-
sure (LP) turbine connected to the high pressure turbine (LP1) and in LP turbine connected to the generator (LP2). The
failed blades had cracks in their roots initiating at the trailing edge, concave side of the steeple outermost fillet radius. Lab-
oratory evaluation of the cracking indicates the failure mechanism to be high cycle fatigue (HCF).
The last-stage blades failure evaluation was carried out. The investigation included a metallographic analysis of the
cracked blades, natural frequency test and analysis, blade stress analysis, unit’s operation parameters and history of events
analysis, fracture mechanics and crack propagation analysis. This paper provides an overview of this failure investigation,
which led to the identification of the blades torsional vibrations near 120 Hz and some operation periods with low load low
vacuum as the primary contribution to the observed failure.
Ó 2006 Elsevier Ltd. All rights reserved.

Keywords: Failure analysis; High cycle fatigue; Last-stage failure; Metallurgical examination; Steam turbine failures

1. Introduction

Energy produced by the boiler is converted to rotational motion of the shaft through the turbines which
drive the generator rotor. The rotor windings with excitation from the exciter create current flow through
the three phase generator stator windings. The stator is connected to the grid through transformers and such.
The generator mass and the grid load coupled through the stator to the rotor create a torsional resistance to
the turbine driven shaft’s rotational motion. Ideally under steady state conditions (constant rotational speed
and constant electrical load) the torque remains essentially constant. Two types of changes in the turbo gen-
erator (t/g) system can cause torsional variation: (a) turbine perturbations and (b) electrical grid perturba-
tions. The electrical perturbations are due to sudden changes on the grid such as large motor starts and arc
furnace operation. In general, there are always relatively small perturbations that create a sort of broadband
background noise, but with large equipment, such as, arc furnaces the perturbations may be in the order of

*
Corresponding author. Tel.: + 52 777 3623811; fax: + 52 777 3623834.
E-mail address: mazur@iie.org.mx (Z. Mazur).

1350-6307/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.engfailanal.2006.11.018
 Author's personal copy

130 Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141

Nomenclature

da/dN velocity of crack propagation

da crack size increment during one cycle of fatigue
dN number of fatigue cycles
Dt time of the fracture propagation

20–100 MWs with similar Var changes. These transients may cause phase imbalance and negative sequence
currents [1].
All the sources of perturbations have time, frequency and amplitude characteristics that will uniquely affect
the shaft rotation and manifest as torsional vibration. Fast transients will excite higher vibration modes, neg-
ative sequence currents will cause excitation forces near 120 Hz, and inter-harmonic currents may cause exci-
tation at any frequency, not just the harmonics [1]. The coincidence of any of these transient forces with a
torsional resonance will increase vibration amplitudes significantly. The torsional forcing functions can have
both a pseudosteady state and transient characteristics where the of power amplitude changes over many sec-
onds (large motor starts, grid frequency) to relatively instantaneous changes (several milliseconds) furnace arc-
ing, line faults. The effect of the speed of power change (rise time) or the frequency composition of the
perturbation will determine the amplitude and frequency of the response.
Historically, failures of turbine and generator (t/g) rotor components due to torsionally induced high cycle
vibration fatigue (retaining rings, shaft cracks, blade root cracks) have been catastrophic and with little warn-
ing. Torsional vibration of rotors, the cause of the fatigue, is generally a sporadic, transient phenomenon pro-
voked by sudden load changes on the grid and/or inter-harmonic loading which lasts from seconds to minutes
[1]. Most of the time these transient events do not overly excite the t/g shaft torsional resonances, but occa-
sionally there is a coincidence of the transient’s wave form characteristics and torsional resonances resulting in
several cycles of high stress. The accumulation of these cycles may lead to crack initiation and fatigue failure.
In addition, modification to the t/g rotor; i.e., exciter or turbine replacement, shaft modification may change
the torsional resonances into a range of susceptibility.
Several power plants have had fatigue cracks on the rotor shaft, in retaining rings and at the turbine blade
roots [2]. The failures ranged from small cracks found during inspection to complete failures. These failures
can cause sever damage to the turbine and generator and are potential human safety problem. The expense of
downtime and repair may be in order the millions of dollars. The cause of these failures is vibrational fatigue
initiated and driven by the t/g rotor torsional vibration.
According to [3–5], the steam turbine operation with low load/low vacuum induces last-stage (L-0) blade
excitation (vibration) by unstable flow developing high vibratory stresses. Due to reduced mass flow, the steam
conditions are variable along the steam path; there are zones of different pressure, radial flows, counter flows,
flow recirculation (flow instabilities). These operation conditions generate blade excitation forces (X mode tor-
sional vibrations), which can lead to blade failures. Unit operation with reduced mass flow is also causing a
reduction in the flow velocity in the same degree. This results in changes of the blade entry flow incidence angle
(change of stage velocity triangle); the flow is entering into the L-0 blades with negative incidence angle strik-
ing the suction surface of the blades airfoil and exciting them. The pressure fluctuation, flow recirculation and
counterflows, in conjunction with the negative incidence angle flow striking on the blades, can develop exces-
sive vibratory stresses causing fatigue fracture of the blades.
This paper provides an overview of the last-stage blades failure investigation, which led to the identification
of the combined effect of torsional vibrations near 120 Hz and operation periods with low load/low vacuum as
the primary contributors to the observed failure.

2. Methodology

The blade under evaluation was the 44-in./1800 rpm last-stage blade (L-0) of a 660 MW turbine (two units)
which consists of one high-pressure turbine and two tandem-compound low-pressure turbines. Each L-0 row
 Author's personal copy

Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141 131

consists of 180 blades in 45 groups of 4, with the 4-blades in each group connected by three tiewires. During
the unit’s last overhauls, the turbines visual examination [6,7] revealed that in the unit 1, LP1 rotor, 8 L-0
blades on the governor side (4.5%) and 23 blades on the generator side (13%) were cracked, and also in
LP2 rotor, 37 blades on the governor side (21%) and 18 blades on the generator side (10%) were cracked
too. Similarly in the unit 2, LP1 rotor, 12 L-0 blades on the generator side (7%) were cracked with no-cracked
blades on the governor side, and also in LP2 rotor, 21 blades on the governor side (12%) and 8 blades on the
generator side (4.5%) were cracked too. The blade is made of 17–4 PH precipitation hardening stainless steel.
The failed blades had cracks in their roots initiating at the trailing edge, concave side of the steeple outermost
fillet radius as is shown in Fig. 1 [6].
All the cracks initiated from blade root serrations, at the corner under the trailing edge where the concave
side of the root meets the exhaust side endface (Fig. 2).
The detail of the fracture surface is shown in Fig. 2b with various features identified by metallographic
examination. As it can be seen, the fracture surface is divided by a beach marks (BM) in the zones of different
roughness where the presence of striations–fracture sliding planes (S), which are characteristic for fatigue
mechanism of fracture propagation, are noticeable. The distances between beach marks corresponds to differ-
ent events of fatigue and the number of beach marks commonly indicates the number of events which partic-
ipate in fatigue propagation. Each zone between beach marks contain thousands of striations (fine marks)
which corresponds to the stable fracture propagation velocity da/dN and the fatigue striation distances on
the fracture surface correspond to the crack size increment, da during one cycle of fatigue. The number of
beach marks on the failed blades fracture surface was variable. The smallest number of beach marks found
in unit 2 was 3 and the highest was 13 (see Fig. 2a). The smallest distance between the beach marks was

Fig. 1. Fracture in the L-0 blade root (unit 2).

 Author's personal copy

132 Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141

Fig. 2. Detail of the L-0 blade fracture surface (unit 2). BM – beach marks, S – striations.

0.26 mm and the biggest was 12.2 mm. The longitude of the fracture in all failed blades was varied within the
range of 11.6–19.02 mm.

3. Metallurgical investigation of the L-0 blade

The metallurgical investigation of the failed L-0 blade of unit 2 was carried out, and included metallogra-
phy, scanning electronic microscopy (SEM) fractography and chemical analysis. The microstructure of the
blade root zone is shown in Fig. 3. The microstructure consists of tempered homogenous martensite obtained
due to precipitating hardening heat treatment, free of failures, typical for forged stainless steel according to
specification 17-4 PH (AISI 630).
The chemical composition and hardness tests of the failed blades were carried out which confirmed concor-
dance of the blade material used to the design specification. The average hardness of blade material was 26.5
HRC which fall within the design limits.
Fractography evaluation was carried out on the exposed crack surface of blade root using SEM to deter-
mine the origin of the fracture. Fig. 4 shows the zone of the crack initiation where the transgranular crack
initiation and propagation was found which is typical of a fatigue failure mechanism.

4. Blades natural frequency analysis

To asses whether the dynamic response characteristics of the L-0 bladed-disk row could have contributed in
any way to the observed failure, the zero speed natural frequency tests were carried out. These included a row-
ing impact-hammer modal test of all 4-blade groups in all L-0 rows of unit 2. The results of these 0 rpm test
were transformed to 1800 rpm rotational speed and are presented in Fig. 5 (LP2 rotor, high-pressure side) as a
representative for the all L-0 stages. The vibration response (amplitude) of the groups of blades is expressed in
g’s RMS normalized dimensionless.
 Author's personal copy

Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141 133

Fig. 3. Microstructure of 17-4PH blade alloy.

Fig. 4. Detail of the zone of the fracture initiation (SEM).

As it can be seen from Fig. 5 there is some frequency scatter between groups of blades particularly for the
second mode which can be explained by variation in erosion rate between blade to blade (Fig. 6), variation of
the blades geometrical tolerances of manufacturing and blade group’s installation tolerances variation. The
reported natural frequencies of the groups of the blades confirm that no mechanical resonances of the blade
structure exist.
 Author's personal copy

134 Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141

LP2- high pressure side

1

0.8

Amplitude
0.6

0.4

0.2

0
30 60 90 120 150
Frequency-Hz

Fig. 5. The 1800 rpm dynamic response of the first three modal forms of vibration of the L-0 blades of the LP2 rotor, high-pressure side,
unit 2.

Fig. 6. Erosion of the leading edge of the L-0 blade.

5. Blade failure analysis

Considering the results of the L-0 blades natural frequency analysis and blades real operation period (nine
years), it may be concluded that the mechanical resonance of the blades does not contribute to blade failure.
This conclusion also indicates that fatigue failure of the blades was not originated during continuous opera-
tion under vibration stresses, but during transition events. If the fatigue initiation and propagation were dur-
ing continuous operation under resonance vibratory stresses, the blade failure would occur practically
immediately (after a few hours of operation).
Analyzing the unit 1 and unit 2 operation history since the date in which the L-0 blades were installed, there
were found some transient events which may be related to the blades failure. In the case of unit 1, one period
of operation with low load/low vacuum was registered (50 min) and 13 events in total of operation with poor
vacuum and reduced load (vacuum below the design value but within the operation range limits) and sudden
load changing. In the case of unit 2 there were registered in total 30 events of sudden load changing and oper-
ation periods with poor vacuum/reduced load. Some part of these events may be related to fracture initiation
(incubation) period and other part to fracture propagation period. During unit’s last overhaul, it was carried
out non destructive examination of all L-0 blades and there was not found any cracks indications in the blades
roots. Since this time, during current inspection which revealed blades failure, it was found maximum 13 beach
marks on one blade fracture surface of unit 2 (Fig. 2a), which probably corresponds to all fatigue propagation
events registered in this period. On the other hand, at the same period since last overhaul, there were registered
14 events in unit 2, which may be related to blade fracture propagation period (3 events of sudden load
 Author's personal copy

Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141 135

changing and 11 events of operation with poor vacuum/reduced load). Comparing these data, it is observed a
very good agreement between number of beach marks on the blade fracture surface and number of hypothetic
fatigue fracture propagation events.
In Fig. 7 it is shown an example of one transient event of sudden load changing in electrical grid due to
failure of neighborhood electric line (load drop of 74 MW) related to unit 2 and in Fig. 8 vibration increase
of this unit due to this event.
From Ref. [1] it is known that the analyzed train of rotors of the 660 MW units is susceptible to sporadic/
transient torsional vibration. As it was mentioned before, two types of changes in the turbo generator (t/g)
system can cause torsional vibration: turbine perturbations and electrical grid perturbations. The electrical
perturbations are mainly due to sudden changes on the grid such as motor starts, failures of the neighborhood
electric lines of the other power stations, relative isolation of the t/g station from other generating stations on

Load vs time
700

650
Load vs time
600

550
MWe

500

450
SLC
400

350

300
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Time (April 2002)

Fig. 7. Transient event of sudden load changes in electrical grid (SLC).

Vibrations
SLC
0.12 BEARING 9 HORI
BEARING 9 VERT

0.1
Displacement p-p (mm)

0.08

0.06

0.04

0.02

0
03-Apr-02 00:00

04-Apr-02 21:00
05-Apr-02 12:00
06-Apr-02 03:00

08-Apr-02 00:00

09-Apr-02 21:00

11-Apr-02 03:00

12-Apr-02 09:00

14-Apr-02 06:00
14-Apr-02 21:00

16-Apr-02 03:00

18-Apr-02 15:00
19-Apr-02 06:00

20-Apr-02 12:00

22-Apr-02 09:00
23-Apr-02 00:00
03-Apr-02 15:00

10-Apr-02 12:00

11-Apr-02 18:00

13-Apr-02 00:00

18-Apr-02 00:00

21-Apr-02 03:00
06-Apr-02 18:00

08-Apr-02 15:00

16-Apr-02 18:00

19-Apr-02 21:00
13-Apr-02 15:00

21-Apr-02 18:00
04-Apr-02 06:00

07-Apr-02 09:00

09-Apr-02 06:00

15-Apr-02 12:00

17-Apr-02 09:00

23-Apr-02 15:00

time

Fig. 8. Unit vibration increase due to sudden load changes in the grid (SLC).
 Author's personal copy

136 Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141

the grid resulting in very large and rapid power changes if any perturbation occur, arc furnace operation and
other electric grid variations. These transition events can excite torsional resonance of the rotor resulting in
exciting of 1st mode of vibration (tangential) grouped blades which results in high quantity of cycles of high
stresses specifically in L-0 blades due to their major length and weight. Accumulation of these cycles can lead
to the cracks initiation and propagation due to fatigue. It is also facilitate because of significant length of the
turbo generator rotor (50 m approximately) which represent high torsional inertia.
Fig. 9a provides the frequency spectrum of the power time history of similar units with ‘‘low excitation’’
while Fig. 9b provides the ‘‘high excitation’’ [1]. Note the 114.75 Hz addition to the signal with the 60 and
120 Hz background noise.
Fig. 10 provides a time history of the torsional responses of the rotor. Notice the change in torsional ampli-
tude which varies with power.
Fig. 11 provides a frequency spectrum of the time history. Note the coincidence at Hz; the perturbation in
power at 114.5 Hz is driving the vibration response at 114.26 Hz.
According to references [4,5,8] turbine operation with low vacuum/low load can excite a torsional vibration
mode of the L-0 blades (X-shaped mode) by unstable flow, developing high vibratory stresses. Fig. 12 shows
steam flow stream lines distribution at the L-0 stage during low load/low vacuum operation [5].
Due to reduced mass flow, the steam conditions are variable along the steam path; there are zones of different
pressure, radial flows, counter flows and flow recirculation (flow instabilities). These operation conditions gen-
erate blade excitation forces (X mode torsional vibrations), which can lead to blade failures. Unit operation with
reduced mass flow is also causing a reduction of the flow velocity in the same degree. This results in changes of the
4.8
4.0
RMS
3.2
MWATTS
2.4
1.6
0.8
0.0

0 40 80 120 160 200

FREQUENCY IN HZ
AMP: 0.01168 FREQ: 114.75
4.8
4.0
3.2
RMS
MWATTS
2.4
1.6
0.8
0.0

0 40 80 120 160 200

FREQUENCY IN HZ
AMP: 3.054 FREQ: 114.75

Fig. 9. Frequency spectra of generator power: (a) low excitation and (b) high excitation [1].
 Author's personal copy

Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141 137

Low response High response

2.00
RAD/SEC

0.00

-2.00
2.00 Low excitation High excitation
RAD/SEC

0.00

-2.00
0.000 10.432 38.912 59.392 79.872 90.112 102.400 112.640
TIME IN SECONDS

Fig. 10. Time history of rotor torsional vibration [1].

0.048

0.040

0.032
RAD/SEC RMS

0.024

0.016

0.008

0.000
0 40 80 120 160 200
FREQUENCY IN HZ
AMP: 0.006174 FREQ: 114.26

0.048

0.040
RAD/SEC RMS

0.032

0.024

0.016

0.008

0.000
0 40 80 120 160 200
FREQUENCY IN HZ
AMP: 0.3117 FREQ: 114.26

Fig. 11. Frequency spectrum of torsional vibration: (a) low excitation and (b) high excitation.

blade entry flow incidence angle (change of stage velocity triangle); the flow is entering into the L-0 blades with
negative incidence angle striking the suction surface of the blade airfoil and exciting the blades.
The typical vibration spectrum of the grouped blades in X-shaped mode, which corresponds to the funda-
mental vibration mode, is shown in Fig. 13 [8]. The pressure fluctuation, flow recirculation and counterflows,
in conjunction with the negative incidence angle flow striking on the blades, can develop excessive vibratory
stresses causing fatigue fracture.
 Author's personal copy

138 Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141

Fig. 12. Steam flow stream lines distribution at the L-0 stage during low load/low vacuum operation [5].

Vibration stress
1

0.8

0.6
Normalized vibration stress

0.4

0.2

-0.2

-0.4

-0.6

-0.8

-1
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
time (sec)

Fig. 13. Vibration spectrum of the grouped blades in X-shaped mode (3rd mode) [8].

The characterization of stresses existing at the blades during operation for 1st tangential mode and 3rd
X-shaped mode of vibration in group of 4 blades is shown in Fig. 14 [8].
As it can be seen from Fig. 14, for 3rd mode of vibration (X-shaped) the highest stresses are developed at
the 1st blade in the group, next at the 4th and 2nd, and finely at 3rd blade. For the 1st vibration mode (tan-
gential) the highest stresses are at the 4th blade in the group (last blade of the group), next at the 3rd, 2nd and
1st blade. For the 3rd mode of vibration it can be observed large differences in stress levels between first and
 Author's personal copy

Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141 139

100
Normalized stresses in the zone of failure 90

80

70

60

50

30
3rd. Mode
20 1st. Mode
10

0
1 2 3 4
Blade position in the group

Fig. 14. Effect of vibration for each blade for 1st and 3rd modes.

the rest of the blades in the group (especially third blade). According to this stress distribution at the blades the
frequency of the blades failure at each position in the group should have similar distribution. Failure of the
‘‘end’’ blades of a group could be consistent with an ‘‘X-shaped’’ mode of vibration.
In the case of 1st mode of vibration, the stress distribution at the blades is more uniform and the frequency
of the blades failure at each position in the group should have more random distribution, i.e. anywhere within
the group.
The characterization of stresses existing at the blades during operation for mixed mode (1st tangential mode
and 3rd X-shaped mode together) in 4 blades of the group is shown in Fig. 15. As it can be seen, there are
small differences in stress levels between the blades; and stress distribution is similar as for 3rd mode of

100

90
Normalized stresses in the zone of failure

80

70

60

50
1st and 3rd
40 mixed mode

30

20

10

0
1 2 3 4
Blade position in the group

Fig. 15. Effect of blade vibration for each blade for 1st and 3rd mixed mode.
 Author's personal copy

140 Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141

20
LP2-HP
18
LP2-GEN
16 LP1-HP
Number of failed blades

LP1-GEN
14

12

10

0
1 2 3 4
Blade position in the group

Fig. 16. Distribution of the failed blades in the L-0 stages of the unit 2.

vibration. In this case, the frequency of the blades failure at each position in the group should have more ran-
dom distribution than for the 3rd X-shaped mode.
Analyzing the distribution of the failed blades in the L-0 stages of unit 2 (see Fig. 16) it can be observed that
the distribution of the cracked blades is consistent with the 3rd X-shaped mode and also with the 1st tangential
mode, depending on stages (rotors) analyzed. This distribution is influenced by the number and duration of
fatigue events (operation periods with low load low/poor vacuum and events of sudden load changes) and
which of these two groups of events predominated.

6. Conclusions

On the basis of the analysis of the results of the last stage blade metallographic examination, units opera-
tional parameters, blade/rotor natural frequencies, blade stresses and fracture mechanics, it may be concluded
that the L-0 blades failure initiation and propagation was driven by a high cycle fatigue mechanism. This was
probably due to the combined effect of the units operating in low load low/poor vacuum conditions and events
of sudden load changes in the grid, resulting in transient excitation of the blades (3rd X-shaped mode and 1st
tangential mode).
Considering the real L-0 blades operation period (nine years), it may be concluded that the dynamic char-
acteristics of the blades (mechanical resonances) did not contribute to blade failure. This means that of blade
root fatigue fracture was probably originated during transient events and not during continuous (stable) oper-
ation under vibratory stresses. It is consistent with other references [9].
Analyzing the units operational history and the blade root fracture surface, a good congruence between
number of high cycle fatigue transition events and number of beach marks on the fracture surface, was found.
The turbine operation with low load and low vacuum typically results in steam flow instabilities-pressure
fluctuation and flow recirculation. Moreover, due to reduced steam flow/steam velocity, the flow is entering
into the stage with negative incidence angle striking the suction surface of the blade airfoil. As a result, high
vibratory stresses were developed in the blades structure, which participate in the blades fatigue failure.
It was found that the analyzed train of rotors of the 660 MW units is susceptible to sporadic/transient tor-
sional vibration due to turbine perturbations or electrical grid perturbations. The electrical perturbations are
due to sudden changes on the grid such as motor starts, failures of the neighborhood electric lines of other
power plants, relative isolation of the t/g station from other generating stations on the grid resulting in very
large and rapid power changes, arc furnace operation and other electric grid variations. These transition
 Author's personal copy

Z. Mazur et al. / Engineering Failure Analysis 15 (2008) 129–141 141

events can excite torsional resonance of the rotor at frequency near 120 Hz resulting in excitation of the 1st
mode of vibration (tangential) grouped blades which results in high quantity of cycles of high stresses specif-
ically in L-0 blades due to their major length and weight. Accumulation of these cycles can lead to cracks ini-
tiation and propagation due to fatigue. It is also facilitated because of significant length of the turbo generator
rotor (50 m approximately) which represent high torsional inertia.

References

[1] Dorfman LS.,Trubelja M. Torsional monitoring of turbine-generators for incipient failure detection. In: Proceedings of the sixth EPRI
steam turbine/generator workshop, St. Luis, Missouri; August 17–20, 1999. p. 1–6.
[2] Rosario DA. L-0 Failure investigation at south Texas project. In: Proceedings of the 8th EPRI turbine-generator conference, Nashville,
Tennesee; August 25–27, 2003. p. 1–8.
[3] Mazur Z, Hernández A, Garcı´a R. L-0 Blades failure investigation of a 110 MW geothermal turbine. In: Proceedings of the ASME
power conference, Paper PWR2007-88024, Atlanta, (GA, USA); May 2–4, 2006. p. 1–9.
[4] Suzuki T, et al., Recent upgrading and life extension technologies for existing steam turbines, ASME power, Chicago; April 5–7, 2005.
p. 577–82.
[5] Troyanowskij BM. Filipow GA, Bulkin AE. Parovyje i gazovyje turbiny atomnych elektrostancyj, Moscow, Energoatomizdat; 1985 [in
Russian].
[6] Robles E. et al., Análisis de Causa de la Falla de los Álabes del Disco L-0 de la Turbina de Baja Presión de la Unidad 2, internal report
IF.FIN.SCT/TM.04/CLV2.2004, REV.0., Mexico; October 2004 [in Spanish].
[7] Gachuz ME, et al., Análisis de Causa Raı´z de la Falla de los Álabes de Rueda L-0 de la Turbina de Baja Presión U-1, internal report
IF.FIN.SCT/TM.04/CLV1.2004, Mexico; May 2004.
[8] Report of L-0 rotating blades failure, T8-04109, Mitsubishi Heavy Industries, Ltd., May 2004.
[9] Brose WR, Marshall LS, Sarber RL. Simplified methods of steam turbine blade root life analysis. In: Proceedings of the joint power
generation conference, ASME PWR, vol. 26, Phoenix (AZ, USA); 1994. p. 213–21.