See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/320310804

Recent trends in the design of hydropower components subjected to cycling and

fatigue; towards improved technical design specifications

Conference Paper · June 2017

CITATIONS READS

2 1,452

1 author:

Alexandre Trudel
AECOM Consultants
5 PUBLICATIONS   65 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Experimental study on the propagation of fatigue cracks in the heat affected zone of hydraulic turbine runner welds View project

Fatigue in hydropower View project

All content following this page was uploaded by Alexandre Trudel on 10 October 2017.

The user has requested enhancement of the downloaded file.

Recent trends in the design of hydropower components subjected to
cycling and fatigue; towards improved technical design
specifications.

By Alexandre Trudel, Eng. M.A.Sc. AECOM, Power, Canada East

Mario Turgeon, Eng. AECOM, Power, Canada East
Isabelle Lanctôt, Eng. AECOM, Power, Canada East

Abstract
Hydropower is regarded as being a highly flexible energy source, a coveted
characteristic that is well earned by the fact that it can provide: peaking generation
capabilities, balance to grids with significant intermittent renewable energy integration,
ancillary services, highly dispatchable power, stable grid-scale energy storage at high
overall efficiency and so on.

Depending on the degree of operation flexibility of a given hydro plant, a

correspondingly varying extent of cyclic loads can act on critical generation components
subjected to fatigue, such as turbine runners, rotor spiders, poles, etc. Technical
specifications detailing the design requirements for components subjected to fatigue are
consequently highly variable and site dependant, in addition to often differing between
regions and owners.

Furthermore, high competitiveness between turbine-generator unit manufacturers

results in designs highly optimized for efficiency and/or cost. The counterpart of such
quest for high efficiency can be aggressive mechanical designs that, when coupled with
a higher extent of cycling induced by increased operation flexibility, must be
accompanied by thorough, consistent and safe design requirements in technical
specifications, thus ensuring that critical components subjected to fatigue operate
without major failure throughout their life expectancy.

Through a review of the most up-to-date literature, as well as recent project case
studies, this paper aims to provide guidance and contribute towards the standardization
of technical design specifications for reliable long-term operation of critical hydropower
components subjected to fatigue.
1. Introduction
Though the concept of fatigue in hydropower is not new, it certainly has gained a
significant amount of attention in the last few decades. Increased operation flexibility of
hydropower units is no doubt a driver for such growing interest in fatigue aspects. This
quest for high flexibility is motivated by various well known reasons; deregulation of
energy markets encouraging more dynamic operation of units, balancing of grids with
increasing integration of intermittent renewable sources and, to a lower but potentially
rising extent, climate change, which enhances the variability of water availability. A
counterpart of increased operation flexibility is a greater toll on hydropower components
subjected fatigue, where frequent operating condition changes and sustained operation
in off-rated conditions result in such components being subjected to a greater number of
cyclic and dynamic loads. Furthermore, high competitiveness between turbine and
generator manufacturers often result in designs highly optimized for efficiency, which
can lead to shrinking mechanical safety margins. Such potentially aggressive
mechanical designs, when coupled with flexible operation, must be governed by
thorough, consistent and safe fatigue requirements in technical design specifications,
thus ensuring that components subjected to fatigue operate without major failure
throughout their life expectancy.

The primary goal of this paper is to provide guidance and contribute towards the
improvement and standardization of technical design specifications related to the
reliable long-term operation of hydropower components subjected to fatigue. To achieve
this goal, an extensive review of recent studies addressing various fatigue-related
aspects, as well as typical industry practices are presented in order to establish the
state of the art in design against fatigue in hydropower. Next, the different modes of
operation and grid services that a hydropower plant and its units can provide are
reviewed and their impact on the service lifetime of components subjected to fatigue is
described. Lastly, guidelines for defining adequate technical design specifications
governing fatigue aspects are proposed.

2. Fatigue assessments in hydropower; a review

Table 1 presents a summary of studies that address various aspects related to the
fatigue of hydropower components. It can be seen that turbine runners, especially
Francis runners, have received the most attention, as they represent the hydropower
component that is by far the most solicited by cyclic loads. In addition, typical runner
materials, namely martensitic stainless steels such as ASTM A743 grade CA6NM, have
also been extensively studied in the last decade or so as their metallurgy, fatigue and
fracture mechanics properties were previously not well known compared to conventional
carbon and structural steel grades used to manufacture structural components such as
rotor spiders, headcovers, brackets, etc. The dynamic loads to which turbine
 components are subjected and which are induced by complex hydraulic phenomena
occurring at specific head and flow conditions have also received a great deal of
attention in recent years. This illustrates a strong and growing interest from turbine and
generator manufacturers, as well as Owners to enhance knowledge regarding the
mechanical reliability of such components over wide operating ranges. The reader is
highly encouraged to carefully study the references in Table 1 in order to get a sense of
the actual state of the art in design against fatigue of critical hydropower components.
Table 1 Summary of recent studies addressing fatigue-related aspects in hydro turbines and generators
Component Subject References
Turbine runner
Fatigue and fracture
mechanics
Material properties [1]–[7]
Reviews [8], [9]
Failure case studies [10]–[17]
Calculation methods [18], [19]
Residual stresses [20]–[25]
Reliability [26]–[35]
Variable amplitude loading [36]
Dynamic Loads
General [37]–[50]
No-load operation [51]
Part load operation [52], [53]
Rotor-stator interaction [54]–[59]
Transients [60]–[63]
Von Kármán vortex shedding [64], [65]
Metallurgy [6], [20], [66]–[70]
Advanced manufacturing [23], [25], [71]–[79]
Non-destructive testing [80]
Stay vanes
Review [81]
Fatigue failure case studies [82]–[84]
Guide vanes Von Kármán [85]
Draft tube Fatigue failure case studies [86]–[88]
Shaft Fatigue failure case studies [89], [90]
Generator
Review [91]
Poles fatigue [92]–[94]
Windings thermal fatigue [95]
2.1. Fatigue theory and applications in hydropower
Fatigue is the process by which materials cumulate damage when subjected to
repeating loads, even if the stresses induced by such loads are below the material’s
tensile strength. The fatigue life of a mechanical component from the moment it is put in
service up to final failure consists of two phases; the fatigue-initiation and fatigue-
propagation phases. The following sections provide a brief theoretical introduction of
these two fatigue phases and how they apply to hydropower. For further reading, the
reader is referred to the books of Suresh [96] and Anderson [97], among many others,
for comprehensive reviews of the theories of fatigue and fracture mechanics.

2.1.1. Fatigue-initiation phase; high-cycle and low-cycle fatigue

High-cycle fatigue is by far the fatigue regime that has received the most attention and
which concerns the most practical applications. It involves components subjected to
repeating loads that induce cyclic stresses well below the yield strength. In the case of
low-cycle fatigue, the yield strength is exceeded during each cycle. The exact transition
in terms of number of cycles to failure between low-cycle and high-cycle fatigue differs
between materials, but a rule of thumb calls for high-cycle fatigue being concerned with
a number of cycles to failure from 104 and higher. The resistance of a material against
high-cycle fatigue is characterized using a stress-cycle (S-N) curve, which represents
the amplitude of stress vs the number of cycles to failure. In the case of low-cycle
fatigue, the fatigue resistance is characterized using strain-cycle (ε-N) curves. A
common misconception is that the number of applied load cycles to which a component
is subjected during its operation determines if it is subjected to high- or low-cycle
fatigue, whereas it is the number of cycles to failure, i.e. the number of cycles that a
component can sustain at a given stress amplitude, that determines which applies.

Figure 1 shows a schematic representation of an S-N curve. This curve allows one to
predict the number of cycles a given component can sustain when loaded at a given
stress amplitude, or on the opposite, the stress amplitude to which the component can
be subjected for a predetermined number of design cycles. S-N curves for several
materials typically show a plateau for a high number of cycles (>107-108). This plateau
occurs at a stress amplitude commonly referred to as the fatigue endurance limit, where
components subjected to stress amplitudes below the endurance limit will exhibit infinite
life. Furthermore, one can notice three regions of interest when looking at Figure 1; (1)
the infinite-life region, where stress amplitudes lower than the endurance limit result in
an infinite fatigue life, (2) the safe-life region, where stress amplitudes are higher than
the endurance limit but the number of applied or design cycles is lower than the
associated number of cycles to failure and (3) the region where the number of applied
or design cycles for a given stress amplitude is higher than the associated number of
cycles to failure, thus resulting in premature initiation of a fatigue crack which can be
assessed using fracture mechanics. This is described in further details later.
 Figure 1 Schematic representation of an S-N curve

A noteworthy aspect related to design against high cycle fatigue is when a component is
subjected to variable amplitude loading, i.e. several loads of different amplitudes. In this
case, Miner’s rule, which is the most widely used cumulative fatigue damage model, can
be applied to determine if the total fatigue damage is acceptable. The main steps are (1)
to segregate all different loads by stress amplitude level using an appropriate method
such as the rainflow cycle counting algorithm [98], (2) to determine the number of cycles
to failure corresponding to each different load using an appropriate S-N curve, and (3)
to sum the ratios of applied cycles to number of cycles to failure for each different stress
amplitude and make sure that the sum does not exceed a specified criteria. The
mathematical expression of Miner’s rule for design fatigue analyses is ∑𝑘𝑖=1 𝑛𝑖 ⁄𝑁𝑖 ≤ 𝐶 ,
where ni is the number of applied or design cycle at stress amplitude Si, Ni is the
number of cycles to failure at stress amplitude Si and C is a design criteria that must not
be exceeded (often equal to 1). Miner’s rule can be thought of as a sum of the fractions
of total fatigue life consumed by several loads of different stress amplitudes. It is
important to understand that while fatigue loads might not lead to fatigue failure when
treated individually, the cumulative fatigue damage induced by all loads can.

In hydropower, all rotating components, as well as non-rotating components whose

mechanical stresses are dependent on power output, flow or head are subjected to
high-cycle fatigue. Low-cycle fatigue can also occur for extreme conditions, but seldom
result in significant damage over typical design lifetimes because of their very low
number of occurrences. Standard North American hydropower industry practices reveal
that components subjected to high-cycle fatigue are often designed by following the
guidelines and fatigue curves of ASME BPVC Section VIII Division 2 [99]. The
guidelines of welding codes, such as AWS D1.1 Structural Welding Code – Steel and
CSA W59 - Welded steel construction (metal arc welding) can also be followed for the
sizing of welds in structural welded components such as rotor spiders, brackets and
headcovers. Proprietary fatigue design curves can also be used by turbine and
generator manufacturers if they are supported by extensive laboratory testing in
representative conditions and environments. For instance, an S-N curve for runners
made of welded martensitic stainless steel castings should be produced using cast
coupons welded with parameters similar to the ones used for the prototype runner and
tested in water. Coupons cut out from real runner castings, such as blades, crowns or
bands are ideal since their metallurgy and internal casting defects are fully
representative of the actual component. Technical specifications for turbine and
generator equipment should specify a default S-N curve to use for the design of
components subjected to fatigue, but should also allow the use of proprietary curves
upon demonstration of their validity and acceptance by the Owner and/or his Engineer.
Such proprietary curves can represent a competitive element between manufacturers
and should be encouraged.

2.1.2. Fatigue-propagation phase; fatigue meets fracture mechanics

When a component subjected to fatigue has accumulated a number of cycles such that
it has depleted its resistance to fatigue-initiation, a crack is initiated and the fatigue-
propagation phase begins. The transition between fatigue-initiation, where there is no
crack, and fatigue-propagation, where there is a crack of appreciable length, is however
not clearly defined. The study of small cracks, which is a field on its own, aims at
characterizing the propagation of cracks in this transition region and is usually
concerned with cracks smaller than the size of characteristic microstructural elements
(e.g. grain size) [100]. A detailed explanation of the mechanisms of small crack
propagation is beyond the scope of this paper. It is however noteworthy to mention that
small fatigue crack theory concepts are implemented in the reliability model used by
Hydro-Quebec for the assessment of turbine runner reliability, as indicated by the works
of Gagnon et al. [26]–[33]. Once a crack has reached an appreciable length, i.e. when it
can be detected using typical NDT techniques, its propagation behavior, and
consequently the remaining life of its containing component, can be predicted using
fracture mechanics. The most common tool to determine the fatigue crack propagation
rate is the da/dN-ΔK curve proposed by Paris et al. in 1961 [101] and which is still widely
used in most major industries where such analyses are required. This curve allows one
to predict the propagation rate (da/dN) of a crack as a function of stress intensity factor
range (ΔK). The stress intensity factor is a fracture mechanics parameter that depends
on crack length, stress and geometry. A detailed account of the mathematics behind the
stress intensity factor is beyond the scope of this paper, but it can be stated that the
stress intensity factor is to a cracked part what stress is to an un-cracked part. The
fatigue crack propagation curve can typically be divided in three regimes (see Figure 2
below). First, the near-threshold regime is characterized by a threshold stress intensity
factor range (ΔKth) below which cracks do not propagate. This threshold stress intensity
factor range is analogous to the fatigue endurance limit in the case of high-cycle fatigue.
Secondly, the stable propagation regime is characterized by the Paris law, which is of
the form 𝑑𝑎⁄𝑑𝑁 = 𝐶𝛥𝐾 𝑚 , where C and m are material-specific constants determined
experimentally. The crack propagation curve is linear in a log-log plot in this regime.
Thirdly, the fast propagation and failure regime is marked by acceleration of the crack
propagation rate until the material’s fracture toughness (KIC) is exceeded and sudden
failure occurs.

Figure 2 Schematic representation of a fatigue crack propagation da/dN-ΔK curve.

An important aspect that differentiates high-cycle fatigue and fatigue crack propagation
is the concept of variable amplitude loading. As previously explained, the order in which
load cycles of different amplitudes occur does not matter for high-cycle fatigue. In the
case of fatigue-propagation however, it has long been recognized that a succession of
loads of different amplitudes influences the crack propagation rate such that overloads
can result in a temporary retardation of the crack, while underloads can accelerate
cracks [102]. Consequently, a simple linear summation of the contribution to total crack
propagation of loads of different amplitudes can in some cases be non-conservative. If
deemed significant for a particular material, such phenomena should be considered for
hydropower components being subjected to loads of different amplitudes [36].

In hydropower, design stage fatigue crack propagation analyses are typically only
performed for turbine runners, if at all. This approach is appropriate for runners
manufactured from welded castings since they inherently contain casting defects such
as shrinkage porosities. Major defects are usually detected during quality control and
repaired, but undetectable defects persist. Fatigue crack propagation studies are
therefore useful to make sure that undetectable defects do not lead to failure before the
end of a specified inspection interval. Another use of fracture mechanics in the design of
turbine runners is for the sizing of partial penetration T-welds joining blades to their
support [2]. This manufacturing method can allow significant cost reductions while also
potentially leading to lower weld-induced residual stresses, which have been shown to
be detrimental to the fatigue resistance of turbine runner welds [7], [22], [24]. However,
the resulting un-welded portion of partial penetration joints can be regarded as a crack
and must be sized using fatigue-propagation analyses. In order to promote
competitiveness at the tender stage, technical design specifications should allow turbine
manufacturers to use partial penetration welding in Francis and propeller runners
provided that the design calculation method and the considered material properties are
suitable and validated by the Owner and/or his Engineer. Ideally, an extensive list of
references of runners containing partial penetration welds and that have been in
operation for several years should also be provided in order to demonstrate the
reliability of the design and manufacturing methods used.

As for other hydropower components, fatigue crack propagation analyses are typically
only performed during condition and residual life assessments. When cracks are found
during inspections, fitness-for service assessments based on fatigue and fracture
mechanics can be performed in order to determine if the cracks must be repaired or can
be tolerated and left as-is until the end of the expected service lifetime of the
component. When no cracks can be tolerated for a given component, such studies can
be performed upon detection of a crack in order to ensure that the unit can operate
safely while a new component is being manufactured and supplied, this to minimize
unnecessary production revenue losses.
In all cases, fatigue and fracture mechanics assessments of hydropower components
are typically based on internationally recognized standards, most often BS7910 Guide
to methods for assessing the acceptability of flaws in metallic structures from the British
Standard Institute and FKM’s Fracture Mechanics Proof of Strength for Engineering
Components. These standards describe methodologies for the assessment of existing
cracks, but can also serve as a basis for the elaboration of design methods.

2.2. Cyclic loads in hydro turbines and generators

Various cyclic loads act on critical hydropower components and can cause significant
fatigue. The design of such components must consider these cyclic loads in order to
achieve the specified service life. Typical cyclic loads in hydro turbines and generators
can be classified as either high-amplitude/low-frequency (HA/LF) cycles or low-
amplitude/high frequency (LA/HF) cycles. The former are induced by steady-state stress
variations from changes in operating conditions and are directly linked to the mode of
operation and grid services of a given power plant as will be further explained below.
The latter mainly relate to complex hydraulic phenomena that occur during steady state
and transient operations and can be extremely damaging if their amplitude are such that
they significantly contribute to the fatigue of a given component.

Criteria found in technical specifications regarding LA/HF stress cycles can sometimes
be inappropriate, especially when safe operation over wide operating ranges is
required. As an example, the maximum amplitude of dynamic stresses in runners are
sometimes expressed as a percentage of the static stress, which can be suitable for
rated operating conditions, though not for operating conditions at low discharges. This
can be demonstrated with the following. Figure 3a shows static and dynamic (peak-
peak) stress values as measured by a strain gage mounted on a prototype Francis
runner of the Chute à la Savanne power plant in Quebec, Canada. The values and
figures have been reproduced from [103]. It can be seen that while static stresses (filled
markers) monotonously increase with power output in an almost linear fashion, the
dynamic stress (empty markers) is highly non-linear, with a significant peak around 40%
of rated power output. Furthermore, as can be seen in Figure 3b, the ratio of dynamic to
static stresses is high at low power outputs and stabilizes at a few percents near rated
conditions. This demonstrates that criteria for LA/HF stress fluctuations based on a
percentage of the static stress in runners are inadequate at low power outputs. The
most obvious example supporting this claim can be seen in Figure 3a, where at approx.
20% of rated power, the static stress is null, while the dynamic stress is equal to approx.
9 MPa. Alternatively, technical specifications should safely require that LA/HF stress
fluctuations be well below the endurance limit in the case of high-cycle fatigue, and
below the threshold stress intensity factor range in the case of fatigue crack
propagation, so that such dynamic and often random loads do not cause fatigue
damage.
 a) ○ Dynamic stress
● Static stress

b)

Figure 3a) Static and dynamic (peak-peak) stresses measured in a prototype runner as a function of power
output and b) ratio of dynamic and static stress as a function of power output.

Table 2 synthesizes the operating conditions and associated phenomena that can
induce significant fatigue cycles in hydro turbines and generators. Typical orders of
magnitudes of associated frequencies and of total number of cycles that can be induced
over the course of a turbine-generator unit lifetime are also shown. As a concrete
example, Figure 4 reproduces the S-N curve of Figure 1 along with colored areas
delimiting typical ranges of stress amplitude and number of cycles for different types of
fatigue loads in a Francis runner. Note that in this particular example, the x-axis of the
figure also represents the number of cycles actually seen by the runner in operation.
 Table 2 Classification of operating conditions and phenomena that induce fatigue cycles in hydropower components.
Operating conditions and Description of induced fatigue cyclic Frequency Total number of cycles over 50 years
phenomena stresses
4
Start-stops Static stress variation between rest and full load 5/day >10
4
Load variations Static stress variation between two loads 5/day >10
HA/LF

Static stress variation between rest and 2

Load rejections 5/year >10
maximum full load rejection overspeed
Static stress variation between rest and 1
Runaway 1/year >10
maximum runaway overspeed
Steady state - No-load (1) 8 (1)
Wideband >10 , rainflow
Stochastic phenomena
Steady state - Part load
Inter-blade vortices Stress fluctuations induced by pressure Wideband
(1) 8
>10 , rainflow
(1)

pulsations 8
Helical vortex rope ~0.3f0 in static frame >10
Steady state - Rated load Structural forced response and amplifications of (2) 8
stress fluctuations can occur for some NGV·f0 >10
Rotor-stator interaction
phenomena (e.g. rotor-stator interaction)
Steady state - High load
(2) 10
Rotor-stator interaction NGV·f0 >10
LA/HF

8
Axial vortex rope ~0.1·f0 - 0.4·f0 >10
Steady state – Others
10
Von Kármán vortex Stress fluctuations induced by mechanical 𝑆𝑡∙𝑣 (3) >10 , depends on excited natural
𝑓𝑉𝐾 =
shedding response to exciting pressure pulsations 𝑡 frequency
Stress fluctuations induced rotating bending
10
In-operation cycles loads for horizontal machines, mechanical/ Variable >10 , depends on phenomena frequency
magnetic imbalance, structural resonance, …
8
Transients – start-ups Stress fluctuations during speed increase Variable >10 , depends on start-up sequence
8
Transients – shutdown Stress fluctuations during speed decrease Variable >10 , depends on shutdown sequence
6
Transients – load rejections Stress fluctuations during transient Variable >10 , depends on load rejection sequence
(1) Some operating conditions and phenomena induce random loads over a wide range of frequencies. The best tool to assess such loads and determine the number of induced
fatigue cycles over a given period of time is the rainflow cycle counting algorithm [98].
(2) Represents the guide vane passing frequency as seen by the runner, where NGV is the number of guide vanes, and fo is the synchronous rotational frequency. Harmonics of this
frequency can also cause resonance of runner.
(3) fVK is the frequency of Von Kármán vortex shedding, St is the Strouhal number, v is the flow velocity in the wake of vane or blade and t is the wake thickness.
 Figure 4 Schematized S-N curve along with areas delimiting the stress amplitudes and number of cycles of
typical loads seen by a Francis runner over its service lifetime.

The fatigue cycles shown in Table 2 affect different hydropower components in different
manners. Table 3 relates these operating conditions and phenomena to the main
hydropower components subjected to fatigue. The mechanical design of these
components is typically optimized to manage the fatigue damage induced by high-
amplitude/low-frequency cycles (safe-life design approach). On the other hand, the
amplitude of stress fluctuations induced by hydraulic and dynamic phenomena should
never be above the fatigue endurance limit or crack propagation threshold, depending
on the type of analysis performed (infinite-life design approach). Otherwise, fast and
premature failure is most likely to occur because of the high frequency of these stress
fluctuations.
 Table 3 Main hydropower components affected by fatigue
Operating conditions and phenomena that induce fatigue cycles
HA/LF LA/HF
Steady state Transients

Load rejections and

In-operation cycles

Load rejections
Load variations

Von Kármán

Shutdowns
Rated load
Start-stops

High load
Part-load

Start-ups
runaway

No-load
Runner X X X X X X X X X X X
Stay vanes X X X X
Guide vanes X X X X
Main components subjected to fatigue

Spiral case X X
Head cover X X
Draft tube steel
X X X
liner
Shaft X X X X
Rotor spider X X X X
Rotor rim X X X X
Rotor poles X X X X
Guide bearings X X X X
Thrust bearing X X X X
Upper/lower
X X X X
bracket
Stator frame X X X X
Stator windings X X X X
3. Impact of plant operation and characteristics on fatigue
3.1. Modes of operation and their impact on fatigue
The mode of operation intended for a given hydropower plant has a direct influence on
the extent to which fatigue will impact the lifetime of its turbine and generator
components. McManamay et al. [104] have proposed a classification of the different
modes of operation for hydropower plants which has been adopted by the
U.S. Department of Energy in their 2016 Hydropower Vision report [105]. These are, in
descending order of total capacity in the U.S.:

ROR
 Peaking (PK) 29% INT-PK ROR-PK
 Run-of-river (ROR) 19% 5%

 Intermediate peaking (INT-PK)

PK ROR-UP
 Run-of-river/peaking (ROR-PK) 41% 3%

 Run-of-river/upstream Peaking (ROR-UP) CC

2%
 Canal/Conduit (CC) RERG
1%
 Reregulating (RERG)
Figure 5 Operation modes and their relative
proportion by capacity in the U.S.

The impact of these modes of operation on the lifetime of components subjected to

fatigue is directly linked to the operation flexibility that they provide. To each of the
previously described modes of operation is associated a certain extent of operation
flexibility, where units designed for peaking provide the most flexible operation, while
units in run-of-river plants without significant pondage are dependent on seasonal
natural flows and discharge from upstream plants and provide the least operation
flexibility. The counterpart of high operation flexibility is an accentuation of fatigue
damage mechanisms that occur for components that “see” the cyclic loads induced by
flexible operation. As an example, units operating in highly flexible peaking plants must
be designed to sustain high-amplitude/low-frequency stress cycles induced by
start-stops and load variations. Turbine components in the water passage, especially
the runner, must also be designed to sustain low-amplitude/high-frequency stress
cycles induced by various hydraulic phenomena occurring at specific operating
conditions. In order to ensure reliable design of hydropower units subjected to fatigue,
the intended mode of operation and its associated number of load cycles should be
clearly defined in the technical design specifications.

The difference in operation flexibility of the peaking and run-of-river modes of operation,
which represent the two extremes of the operation flexibility spectrum, can be
appreciated from Figure 6. This figure shows in a) the plant power output through year
2016 for three power plants operated by the U.S. Army Corps of Engineers
(data obtained from the U.S.A.C.E. Northwestern Division database [106]), which can
be classified as follows in accordance with [104]: peaking (Detroit Dam), run-of-river
without pondage (Lower Monumental Dam) and run-of-river with pondage (Dworshak
Dam). Though both Lower Monumental Dam and Dworshak Dam are classified as run-
of-river plants as per [104], the power output trend is significantly different between both
plants. In the case of Lower Monumental Dam, the power output is seen to fluctuate
significantly, which can be explained by the fact that this power plant is the third
downstream of a four run-of-river power plants complex along the Snake River with no
significant pondage. In the case of Dworshak Dam, significant pondage allows the plant
to operate for extended periods of time without frequent and significant power output
variations, thus acting as a base load plant. As for the peaking plant,
Detroit Dam, frequent start-stops can be seen to occur daily throughout the year.

The extent to which the modes of operation of these plants induce fatigue cycles can be
further visualised by applying the rainflow cycle counting algorithm [98] on the
plants’ power output time series (Figure 6b). This algorithm is normally used to reduce a
spectrum of randomly varying stresses into a set of discrete stress ranges and
corresponding number of cycles, thus allowing performing fatigue damage calculations
using a cumulative damage model such as Miner’s rule. It is thus also well suited to
reduce hydropower plants power output time series, which can also be highly variable
and random, in a set of discrete power output variation ranges and corresponding
number of cycles over a given time period. As can be seen in Figure 6b, the shape of
the resulting curves can be directly linked to the mode of operation of their
corresponding plant, where a curve initially remaining flat and extending towards the
upper-right portion of the graph indicates a high number of cycles of large power output
variations, whereas a curve immediately falling down indicates a low number of large
power output variations. Such results can be used to properly characterize the mode of
operation of a given plant based on past operation and also estimate the fatigue
damage of turbine and generator components induced by start-stops and load
variations, which are often the most critical fatigue loads to consider.
a)

b)

Figure 6a) Hourly power output of three hydropower plants operated by the U.S.A.C.E. b) Cumulative number of
power output variations obtained with the rainflow cycle counting algorithm [98].
3.2. Grid services and their impact on fatigue
In addition to the mode of operation, specific grid services can be provided by
hydropower plants. McGowin and Steeley from EPRI have proposed a classification of
these grid services in the context of a study on grid-scale wind integration [107]. As
stated in [105], Key has shown that hydropower has the flexibility to provide all these
grid services [108], though at a varying extent depending on, among others, turbine and
generator technologies. From the perspective of critical hydropower components
subjected to fatigue, the most relevant grid services that hydro can provide are:

 Load-following (LF)
 Primary frequency control (PFC)
 Operating reserve (OR), which include spinning and non-spinning reserve
 Base load (BL)

3.2.1. Load-following
Load-following ensures a balance between load and generation through smooth ramps
over relatively long time frames (e.g. hours). From a fatigue point of view, load-following
results in additional cycling for components whose mechanical response is dependent
on load. It can also result in sustained operation at off-rated operation conditions, where
hydraulic phenomena can induce high-frequency fatigue cycles to the main hydraulic
components. While it is understood that the total number and amplitude of load
variations cannot be predicted in advance, technical specifications should nevertheless
provide conservative estimates of the number and amplitude of load variations to
consider for the mechanical design of related components. These can be established
from past operation of similar reference power plants.

3.2.2. Primary frequency control

Primary frequency control, or regulation, is associated with continuous adjustments of
the power output in order to account for fast fluctuations in the load and maintain the
grid frequency [109]. Hydropower units are well suited to provide this service when
equipped with automatic generation control (AGC), which allows high frequency
adjustments of the turbine flow, and hence power output. For double regulated turbines,
such as Kaplan turbines, this can result in a significant increase in blades adjustments
in order to maintain the cam relationship. Some components of the blade regulation
mechanisms can therefore be subjected to additional load cycles and increased fatigue
[16], [17]. When technical specifications cover requirements for such regulation
operation by providing design values for the number and amplitudes of blade
adjustments, mechanical components can be designed accordingly. However, when
existing units are newly equipped with AGC, the additional fatigue cycles for which the
unit was not originally designed for can be an issue. In this case, a field assessment
and fatigue analysis of the unit can be required in order to assess the impact of the
additional fatigue cycles induced by the AGC mode.

3.2.3. Operating reserve

When providing operating reserve services, hydropower components can be subjected
to additional fatigue cycles specific to this grid service. The nature and extent of those
additional fatigue cycles depends on the manner in which a given unit is operated while
waiting to respond to a reserve demand. Whether acting as spinning or non-spinning
reserve, the response of a hydropower unit to reserve demands will induce additional
start-stops and/or load variation cycles. Technical specifications should therefore
specify a conservative estimate of the number of time the units are expected to be
called for reserve demand purposes throughout their intended service lifetime. In
addition, units providing spinning reserve, whether while operating offline at speed-no-
load or online at part loads will sustain fatigue cycles induced by high frequency
hydraulic phenomena as detailed in Table 2. In this regard, technical design
specifications should clearly specify the yearly number of hours and the load at which
the unit is to be operated in spinning operating reserve.

3.2.4. Base load

Base load units operate near rated conditions most of the time, and are therefore not
subjected to extensive cycling and dynamic loads. Such units are expected to start and
stop very rarely, typically only for maintenance purposes. Since the capacity factor, and
hence total generation, is to be maximized for base load units, they are typically
designed for the highest efficiency possible. However, hydraulic components must be
designed in order to reach an optimum between hydraulic performances and
mechanical integrity. For instance, the thickness of runner blades and vanes are kept as
low as possible to maximize capacity and minimize hydraulic losses, but not too low that
there is an unacceptable risk of failure before the intended service lifetime. Since fatigue
cyclic stresses are infrequent and of low amplitudes for base load units, one could
argue that typical mechanical safety factors could be alleviated in order to enhance
hydraulic design optimization and maximize efficiency. As an example, peak static
stresses in turbine runners during normal operation are often specified to be lower than
183 MPa for martensitic stainless steel cast grade CA6NM, which represents a safety
factor of 3 when considering the minimum yield strength specified by ASTM A743 [110].
Such a high safety factor has historically been adopted in order to provide an adequate
safety margin against previously not well known dynamic loads induced by part-load
operation and rotor-stator interaction. However, with the actual extensive knowledge
acquired in recent years regarding, among others, dynamic loads and materials
science, hydraulic components of base load units could be more aggressively designed
in order to maximize efficiency without compromising reliability.
 Table 4 shows the operating conditions and associated fatigue cycles of Table 2 for
which the extent, whether in number or amplitude, is significantly influenced by the
previously described modes of operation and grid services. Some elements of Table 2
were omitted since they are not affected by a specific mode of operation or grid service.
Table 4 Fatigue cycles induced by operating conditions and phenomena involved in the different modes of
operations and grid services
Modes of operation Grid services
ROR-PK
Operating conditions PK INT-PK ROR period PK period ROR-UP ROR LF PFR OR BL
Start-stops X X X X
Load variations X X X X X X X
Steady state - No-load X X X X
Steady state - Part load X X X X X X X X X
Steady state - Rated load X X X X X X X
Steady state - High load X X X X X X X
Transients – start-ups X X X X
Transients – shutdown X X X X
4. Towards improved and standardized technical specifications
Technical specifications governing fatigue design requirements on which manufacturers
base their analyses are often highly variable. Though this is understandable because of
the unique features of every power plant, a certain level of standardization should be
promoted. This section provides a compilation of guidelines derived from experience,
industry practices and literature in order to contribute towards such standardization of
fatigue-related technical specifications.

4.1. Specifications governing fatigue loads

Table 5 provides guidelines on the requirements that should be included in turbine and
generator technical specifications in order to properly establish the number of cycles to
be considered for design fatigue analyses. Applicable elements from Table 2 have been
reproduced and a description of the requirements to include is shown.
Table 5 Technical design specifications for fatigue loads induced by various operating conditions
and phenomena
Operating conditions and
phenomena Technical design specification requirements
Start-stops Daily/monthly number of start-stops
Load variations Daily/monthly number of power output variations from
HA/LF

full load to a specified minimum load(1)

Load rejections Monthly/yearly number of load rejections
Runaway Monthly/Yearly number of runaways
Steady state - No-load Yearly number of hours of operation at speed-no-load.
Steady state - Part load Yearly number of hours of operation at various power
outputs in accordance with weighting
LA/HF

Steady state - Rated load

Yearly number of hours of operation at specific power
Steady state - High load outputs outside specified operating range
Transients – Start-ups Daily/monthly number of start-stops(2)
Transients – Shutdowns Daily/monthly number of start-stops(2)
Transients – Load rejections Monthly/yearly number of load rejections(2)
(1) The minimum load to be considered for load variation cycles can be established based on the weighting of the weighted average
efficiency formula. A simplified and more conservative approach is to consider that every load variation cycle is from full load to the
minimum load of the specified operating range.
(2) The manufacturer shall determine a suitable duration over which significant transient stress fluctuations occur during the
transient event. The total number of cycles can then be determined considering the specified number of start-stops and load
rejections.
4.2. Specifications governing fatigue stress determination methods
Another important aspect that should be addressed in technical specifications is the
method by which the stresses considered for design fatigue analyses are to be
determined. A simple categorization of these methods can be based on the types of
fatigue loads as defined in section 2.2, namely high-amplitude/low-frequency or low-
amplitude/high-frequency. Further classification of methods for low-amplitude/high-
frequency loads can also be established based on the type of response of the structure.
Some phenomena, such as the part-load vortex rope, induce dynamic stresses through
a quasi-static response of the structure, i.e. inertial and damping forces can be
neglected. Other phenomena, such as rotor-stator interaction and Von Kármán vortices
can cause a forced response (resonance) of the excited structure which leads to
amplified dynamic stresses. Table 6 shows the proposed classification along with
descriptions of the methods that can be specified to determine the stresses required as
inputs for fatigue analyses. Design criteria are also proposed for the different types of
fatigue loads. Technical design specifications can specify a specific method from which
the stresses are to be determined for a given component and type of fatigue load. The
complexity, and thus accuracy of the specified method should be coherent with the
expected severity and extent of the fatigue load considered. For instance, a turbine
runner expected to operate at part load for significant periods of time, where dynamic
stresses can be quite high, should specify a method to be used that will allow an
accurate determination of the expected dynamic stresses. Minimally, specifications
should dictate that suitable methods shall be proposed by the turbine-generator
manufacturer for validation and acceptance by the Owner and/or his Engineer.
 Table 6 Methods for fatigue stress determination
Type of fatigue load Stress determination methods description Fatigue design criteria
High-amplitude/low-frequency
Minimum Analytical calculation at considered operating conditions
Better(1) Static FEA at considered operating conditions Total fatigue damage lower
than a specified criteria
Best(1) Static FEA at considered operating conditions with numerical (Miner’s rule)
model calibration from reference prototype measurements
Low-amplitude/high-frequency
Quasi-static response Minimum Transposition from reference prototype dynamic strains
measurements High-cycle fatigue
Better Quasi-static FEA with fluctuating pressure field calculated by Dynamic stresses lower than
unsteady CFD(2) fatigue endurance limit
Better Quasi-static FEA with fluctuating pressure field measured on
homologous model turbine Fatigue crack propagation
Stress intensity factor range
Better Scale-up from homologous model dynamic strain measurements lower than threshold
Best Quasi-static FEA calibrated with dynamic strains measured on (no propagation allowed)
homologous model
Forced response Minimum FEA modal analysis to determine natural frequencies in relevant Natural frequencies different
(resonance)(3) environment and qualitative assessment of risk of forced than exciting frequency by a
response against exciting frequencies specified margin (e.g. 20%)
Better Transposition of reference prototype dynamic strains High-cycle fatigue
measurements Dynamic stresses lower than
fatigue endurance limit
Better Dynamic FEA with fluctuating pressure field calculated by
unsteady CFD
Fatigue crack propagation
Best Dynamic FEA with fluctuating pressure field calculated by Stress intensity factor range
unsteady CFD and validated with model turbine measurements lower than threshold
(no propagation allowed)
(1) For many structural components subjected to fatigue, such as the rotor spider, FEA calculations of stresses are not required. Analytical calculations provide adequate results.
(2) For some complex hydraulic phenomena, such as inter-blade vortices, it is difficult or impractical with actual computational tools to obtain accurate numerical results in a
reasonable amount of time. In this case, the use of experimental measurements is recommended.
(3) The stresses induced by a forced response against exciting phenomena can typically not be transposed from model measurements because of the great difficulty of achieving
proper homology of the phenomena exciting frequencies vs the structure’s natural frequencies.
4.3. Guidelines on specifications governing fatigue material properties
As previously stated, technical design specifications should specify fatigue curves to be
used for the mechanical design of critical components subjected to fatigue. Minimally,
specifications should state that the design of such components shall be done according
to the guidelines of an internationally recognized code or standard, which implies the
use of its fatigue curves. The codes and standards typically used for fatigue
assessments are (items in bold are the ones often found in hydro):

 High-cycle fatigue
▫ BPVC Section VIII Division 2 by ASME
▫ BS7608 Guide to fatigue design and assessment of steel products by BSI
▫ D1.1 Structural Welding Code – Steel by AWS
▫ W59 - Welded steel construction (metal arc welding) by CSA
 Fatigue-crack propagation
▫ BS7910 Guide to methods for assessing the acceptability of flaws in
metallic structures by BSI
▫ Fracture Mechanics Proof of Strength for Engineering Components
by FKM
▫ R6 Assessment of the integrity of structures containing defects by the
British Energy Generation Report
▫ SINTAP Structural integrity assessment procedures for European industry
by the European Union Brite-Euram Programme
▫ API 579-1/ASME FFS-1 Fitness-For-Service Assessment Standard by API
and ASME

On the other hand, turbine and generator manufacturers, who often perform advanced
R&D activities on material science and fatigue-related aspects, should also be
encouraged to propose their own design fatigue curves, provided that they are suitable
for the designed component. The use of fatigue curves specifically derived for a given
component and material is highly beneficial in terms of design optimization and can help
avoid under- or over-design.

5. Conclusion
There is no doubt that fatigue is a growing element of interest for the reliable design of
hydropower generation equipment. This is illustrated by the impressive quantity of
research studies addressing various aspects of fatigue in hydropower that have recently
been conducted. Furthermore, hydro turbines and generators are nowadays operated in
a more dynamic fashion in order to provide the operation flexibility required to perform in
deregulated markets, provide grid services, balance grids with high penetration of
intermittent renewable sources and counteract the increasingly variable water
availability induced by climate change. Higher risk of fatigue failure is a major trade-off
against increased operation flexibility and must be thoroughly considered during design
and operation of hydro generation equipment. This begins at the stage of technical
specifications writing, where requirements governing fatigue aspects play a major role in
how fatigue is adequately considered during design. This paper has provided a
comprehensive review of various fatigue-related aspects applied to hydropower, an
overview of the impact of modes of operation and grid services on fatigue, as well as
guidelines on how to properly define technical specifications governing fatigue aspects.
By doing so, it hoped to have contributed to the improvement and standardization of
fatigue-related technical specifications, and ultimately to the end of fatigue failures in
hydropower.
References
[1] W. J. P. Casas, S. L. Henke, and N. Novicki, “Fracture toughness of CA6NM alloy,
quenched and tempered, and of its welded joint without PWHT,” Weld. Int., vol. 23,
no. 3, pp. 166–172, Mar. 2009.
[2] M. Sabourin, D. Thibault, D. A. Bouffard, and M. Lévesque, “New parameters
influencing hydraulic runner lifetime,” IOP Conf. Ser. Earth Environ. Sci., vol. 12, no.
1, p. 012050, 2010.
[3] Y. Iwabuchi and I. Kobayashi, “A Study of Toughness Degradation in CA6NM
Stainless Steel,” Mater. Sci. Forum, vol. 654–656, pp. 2515–2518, 2010.
[4] B. L. da Silva, J. L. de A. Ferreira, and J. A. Araújo, “High-cycle notch sensitivity of
alloy steel ASTM A743 CA6NM used in hydrogenator turbine components,” Fract.
Struct. Integr., vol. 0, no. 14, pp. 36–44, 2010.
[5] B. L. da Silva, J. L. de A. Ferreira, F. Oliveira, and J. A. Araújo, “Influence of mean
stress on the fatigue strength of ASTM A743 CA6NM alloy steel,” Fract. Struct.
Integr., vol. 0, no. 14, pp. 17–26, 2010.
[6] D. Thibault, P. Bocher, M. Thomas, J. Lanteigne, P. Hovington, and P. Robichaud,
“Reformed austenite transformation during fatigue crack propagation of 13%Cr–
4%Ni stainless steel,” Mater. Sci. Eng. A, vol. 528, no. 21, pp. 6519–6526, Aug.
2011.
[7] A. Trudel, M. Sabourin, M. Lévesque, and M. Brochu, “Fatigue crack growth in the
heat affected zone of a hydraulic turbine runner weld,” Int. J. Fatigue, vol. 66, pp.
39–46, Sep. 2014.
[8] A. Trudel and M. Sabourin, “Metallurgical and fatigue assessments of welds in cast
welded hydraulic turbine runners,” IOP Conf. Ser. Earth Environ. Sci., vol. 22, no. 1,
p. 012015, 2014.
[9] X. Liu, Y. Luo, and Z. Wang, “A review on fatigue damage mechanism in hydro
turbines,” Renew. Sustain. Energy Rev., vol. 54, pp. 1–14, Feb. 2016.
[10] Y. Luo, Z. Wang, J. Zeng, and J. Lin, “Fatigue of piston rod caused by unsteady,
unbalanced, unsynchronized blade torques in a Kaplan turbine,” Eng. Fail. Anal.,
vol. 17, no. 1, pp. 192–199, Jan. 2010.
[11] D. Frunzǎverde et al., “Failure analysis of a Francis turbine runner,” IOP Conf.
Ser. Earth Environ. Sci., vol. 12, no. 1, p. 012115, 2010.
[12] E. Egusquiza, C. Valero, X. Huang, E. Jou, A. Guardo, and C. Rodriguez,
“Failure investigation of a large pump-turbine runner,” Eng. Fail. Anal., vol. 23, pp.
27–34, Jul. 2012.
[13] G. Urquiza et al., “Failure analysis of a hydraulic Kaplan turbine shaft,” Eng. Fail.
Anal., vol. 41, pp. 108–117, Jun. 2014.
[14] X. Liu, Y. Y. Luo, and Z. W. Wang, “Fatigue Analysis of the Piston Rod in a
Kaplan Turbine Based on Crack Propagation under Unsteady Hydraulic Loads,” IOP
Conf. Ser. Earth Environ. Sci., vol. 22, no. 1, p. 012017, 2014.
[15] A. Luna-Ramírez, A. Campos-Amezcua, O. Dorantes-Gómez, Z. Mazur-
Czerwiec, and R. Muñoz-Quezada, “Failure analysis of runner blades in a Francis
hydraulic turbine — Case study,” Eng. Fail. Anal., vol. 59, pp. 314–325, Jan. 2016.
[16] A. Tapper, “A fatigue investigation in a Kaplan hydropower station operated in
frequency regulating mode,” 2016.
[17] E. Wurm, “Consequences of Primary Control to the Residual Service Life of
Kaplan Runners,” presented at the Russia Power 2013 & Hydrovision Russia 2013,
Moscow, Russia, 2013.
[18] S. Vantadori, A. Carpinteri, and D. Scorza, “Simplified analysis of fracture
behaviour of a Francis hydraulic turbine runner blade,” Fatigue Fract. Eng. Mater.
Struct., vol. 36, no. 7, pp. 679–688, Jul. 2013.
[19] A. Carpinteri, C. Ronchei, D. Scorza, and S. Vantadori, “Fracture mechanics
based approach to fatigue analysis of welded joints,” Eng. Fail. Anal., vol. 49, pp.
67–78, Mar. 2015.
[20] D. Thibault, P. Bocher, and M. Thomas, “Residual stress and microstructure in
welds of 13%Cr–4%Ni martensitic stainless steel,” J. Mater. Process. Technol., vol.
209, no. 4, pp. 2195–2202, Feb. 2009.
[21] D. Thibault, P. Bocher, M. Thomas, M. Gharghouri, and M. Côté, “Residual stress
characterization in low transformation temperature 13%Cr–4%Ni stainless steel
weld by neutron diffraction and the contour method,” Mater. Sci. Eng. A, vol. 527,
no. 23, pp. 6205–6210, Sep. 2010.
[22] A. Trudel, M. Brochu, and M. Lévesque, “Residual stress effects on the
propagation of fatigue cracks in the weld of a CA6NM stainless steel,” presented at
the 13th International Conference on Fracture, Beijing, China, 2013.
[23] S. Godin, D. Thibault, and J. B. Lévesque, “An Experimental Comparison of
Weld-Induced Residual Stresses Using Different Stainless Steel Filler Metals
Commonly Used for Hydraulic Turbines Manufacturing and Repair,” Mater. Sci.
Forum, vol. 768–769, pp. 628–635, 2014.
[24] A. Trudel, M. Lévesque, and M. Brochu, “Microstructural effects on the fatigue
crack growth resistance of a stainless steel CA6NM weld,” Eng. Fract. Mech., vol.
115, pp. 60–72, Jan. 2014.
[25] S. Sarafan, J.-B. Lévesque, P. Wanjara, J. Gholipour, and H. Champliaud,
“Distortion and residual stresses in electron beam-welded hydroelectric turbine
materials,” Sci. Technol. Weld. Join., vol. 21, no. 6, pp. 473–478, Aug. 2016.
[26] M. Gagnon, S. A. Tahan, P. Bocher, and D. Thibault, “Impact of startup scheme
on Francis runner life expectancy,” IOP Conf. Ser. Earth Environ. Sci., vol. 12, no. 1,
p. 012107, 2010.
[27] M. Gagnon, S. A. Tahan, P. Bocher, and D. Thibault, “The role of high cycle
fatigue (HCF) onset in Francis runner reliability,” IOP Conf. Ser. Earth Environ. Sci.,
vol. 15, no. 2, p. 022005, 2012.
[28] M. Gagnon, A. Tahan, P. Bocher, and D. Thibault, “On the stochastic simulation
of hydroelectric turbine blades transient response,” Mech. Syst. Signal Process., vol.
32, pp. 178–187, Oct. 2012.
[29] M. Gagnon, A. Tahan, P. Bocher, and D. Thibault, “On the Fatigue Reliability of
Hydroelectric Francis Runners,” Procedia Eng., vol. 66, pp. 565–574, 2013.
[30] M. Gagnon, A. Tahan, P. Bocher, and D. Thibault, “A probabilistic model for the
onset of High Cycle Fatigue (HCF) crack propagation: Application to hydroelectric
turbine runner,” Int. J. Fatigue, vol. 47, pp. 300–307, Feb. 2013.
[31] M. Gagnon, A. Tahan, P. Bocher, and D. Thibault, “Influence of load spectrum
assumptions on the expected reliability of hydroelectric turbines: A case study,”
Struct. Saf., vol. 50, pp. 1–8, Sep. 2014.
[32] M. Gagnon and D. Thibault, “Response Spectra and Expected Fatigue Reliability:
A Look at Hydroelectric Turbines Behavior,” Procedia Eng., vol. 133, pp. 613–621,
2015.
[33] D. Thibault, M. Gagnon, and S. Godin, “The effect of materials properties on the
reliability of hydraulic turbine runners,” Int. J. Fluid Mach. Syst., vol. 8, no. 4, pp.
254–263, 2015.
[34] N. Abolfathi Nobari, “Bayesian updating of hydroelectric turbine fatigue reliability,”
masters, École de technologie supérieure, Montréal, 2016.
[35] M. Berdai, A. Tahan, and M. Gagnon, “Imprecise Probabilities in Fatigue
Reliability Assessment of Hydraulic Turbines,” ASCE-ASME J. Risk Uncertain. Eng.
Syst. Part B Mech. Eng., vol. 3, no. 1, pp. 011006-011006-8, Nov. 2016.
[36] M. Hassanipour, Y. Verreman, J. Lanteigne, and J. Q. Chen, “Effect of periodic
underloads on fatigue crack growth in three steels used in hydraulic turbine
runners,” Int. J. Fatigue, vol. 85, pp. 40–48, Apr. 2016.
[37] P. Y. Lowys, F. Paquet, M. Couston, M. Farhat, S. Natal, and F. Avellan,
“Onboard Measurements of Pressure and Strain Fluctuations in a Model of low
Head Francis Turbine. Part 2 : Measurements and Preliminary Analysis Results,”
Proc. 21st IAHR Symp. Hydraul. Mach. Syst., pp. 873–880, 2002.
[38] H. G. Poll, nther, J. Zanutto, Carlos, and W. Ponge-Ferreira, “Hydraulic Power
Plant Machine Dynamic Diagnosis,” Shock Vib., vol. 13, no. 4–5, pp. 409–427,
2006.
[39] L. Zhou, Z. Wang, R. Xiao, and Y. Luo, “Analysis of dynamic stresses in Kaplan
turbine blades,” Eng. Comput., vol. 24, no. 8, pp. 753–762, Nov. 2007.
[40] Z. W. Wang, Y. Y. Luo, L. J. Zhou, R. F. Xiao, and G. J. Peng, “Computation of
dynamic stresses in piston rods caused by unsteady hydraulic loads,” Eng. Fail.
Anal., vol. 15, no. 1–2, pp. 28–37, Jan. 2008.
[41] R. Xiao, Z. Wang, and Y. Luo, “Dynamic Stresses in a Francis Turbine Runner
Based on Fluid-Structure Interaction Analysis,” Tsinghua Sci. Technol., vol. 13, no.
5, pp. 587–592, Oct. 2008.
[42] Z. Wang et al., “Hydroturbine operating region partitioning based on analyses of
unsteady flow field and dynamic response,” Sci. China Technol. Sci., vol. 53, no. 2,
pp. 519–528, Feb. 2010.
[43] Y. Luo, Z. Wang, J. Zhang, J. Zeng, J. Lin, and G. Wang, “Vibration and fatigue
caused by pressure pulsations originating in the vaneless space for a Kaplan turbine
with high head,” Eng. Comput., vol. 30, no. 3, pp. 448–463, 2013.
[44] X. Huang, J. Chamberland-Lauzon, C. Oram, A. Klopfer, and N. Ruchonnet,
“Fatigue analyses of the prototype Francis runners based on site measurements
and simulations,” IOP Conf. Ser. Earth Environ. Sci., vol. 22, no. 1, p. 012014, 2014.
[45] U. Seidel, C. Mende, B. Hübner, W. Weber, and A. Otto, “Dynamic loads in
Francis runners and their impact on fatigue life,” IOP Conf. Ser. Earth Environ. Sci.,
vol. 22, no. 3, p. 032054, 2014.
[46] F. Bouloc, J. Guillozet, F. Duparchy, P. Y. Lowys, and A. Duparchy, “Mechanical
risks prediction on Francis runner by Spatial Harmonic Decomposition,” IOP Conf.
Ser. Earth Environ. Sci., vol. 49, no. 7, p. 072015, 2016.
[47] A. Lyutov et al., “Modelling of a Francis Turbine Runner Fatigue Failure Process
Caused by Fluid-Structure Interaction,” IOP Conf. Ser. Earth Environ. Sci., vol. 49,
no. 7, p. 072012, 2016.
[48] C. Monette, H. Marmont, J. Chamberland-Lauzon, A. Skagerstrand, A. Coutu,
and Jens Carlevi, “Cost of enlarged operating zone for an existing Francis runner,”
IOP Conf. Ser. Earth Environ. Sci., vol. 49, no. 7, p. 072018, 2016.
[49] C. Trivedi and M. J. Cervantes, “Fluid-structure interactions in Francis turbines: A
perspective review,” Renew. Sustain. Energy Rev., vol. 68, Part 1, pp. 87–101, Feb.
2017.
[50] A. Coutu et al., “Specific speed effect on francis runner reliability under various
operating conditions,” in XXXVII Iberian Latin-American Congress on Computational
Methods in Engineering, Brazil, 2016.
[51] B. Nennemann et al., “Challenges in Dynamic Pressure and Stress Predictions at
No-Load Operation in Hydraulic Turbines,” IOP Conf. Ser. Earth Environ. Sci., vol.
22, no. 3, p. 032055, 2014.
[52] F. Duparchy et al., “Analysis of the part load helical vortex rope of a Francis
turbine using on-board sensors,” J. Phys. Conf. Ser., vol. 656, no. 1, p. 012061,
2015.
[53] S. Bouajila, T. D. Colombel, P.-Y. Lowys, and T. Maitre, “Hydraulic Phenomena
Frequency Signature of Francis Turbines Operating in Part Load Conditions,” IOP
Conf. Ser. Earth Environ. Sci., vol. 49, no. 8, p. 082001, 2016.
[54] F. Dompierre and M. Sabourin, “Determination of turbine runner dynamic
behaviour under operating condition by a two-way staggered fluid-
structureinteraction method,” IOP Conf. Ser. Earth Environ. Sci., vol. 12, no. 1, p.
012085, 2010.
[55] H. Tanaka, “Vibration Behavior and Dynamic Stress of Runners of Very High
Head Reversible Pump-turbines,” Int. J. Fluid Mach. Syst., vol. 4, no. 2, pp. 289–
306, 2011.
[56] R. Guillaume, J. L. Deniau, D. Scolaro, and C. Colombet, “Influence of the rotor-
stator interaction on the dynamic stresses of Francis runners,” IOP Conf. Ser. Earth
Environ. Sci., vol. 15, no. 5, p. 052011, 2012.
[57] U. Seidel, B. Hübner, J. Löfflad, and P. Faigle, “Evaluation of RSI-induced
stresses in Francis runners,” IOP Conf. Ser. Earth Environ. Sci., vol. 15, no. 5, p.
052010, 2012.
[58] A. Coutu, M. D. Roy, C. Monette, and B. Nennemann, “Experience with rotor-
stator interactions in high head Francis runner.”
[59] X. Huang, C. Oram, and M. Sick, “Static and dynamic stress analyses of the
prototype high head Francis runner based on site measurement,” IOP Conf. Ser.
Earth Environ. Sci., vol. 22, no. 3, p. 032052, 2014.
[60] C. Trivedi, B. Gandhi, and C. J. Michel, “Effect of transients on Francis turbine
runner life: a review,” J. Hydraul. Res., vol. 51, no. 2, pp. 121–132, Apr. 2013.
[61] F. Léonard and M. Gagnon, “Transient response and life assessment: Case
studies on the load rejection of two hydroelectric turbines,” presented at the
International Conference Surveillance 7, France, 2013.
[62] É. Moisan, D.-B. Giacobbi, M. Gagnon, and F. Léonard, “Self-excitation in
Francis runner during load rejection,” IOP Conf. Ser. Earth Environ. Sci., vol. 22, no.
3, p. 032025, 2014.
[63] M. Gagnon, J. Nicolle, J.-F. Morissette, and M. Lawrence, “A look at Francis
runner blades response during transients,” IOP Conf. Ser. Earth Environ. Sci., vol.
49, no. 5, p. 052005, 2016.
[64] R. K. Fisher, U. Seidel, G. Grosse, W. Gfeller, and R. Klinger, “A case study in
resonant hydroelastic vibration: The causes of runner cracks and the solutions
implemented for the Xiaolangdi hydroelectric project,” presented at the Proceedings
of the Hydraulic Machinery and Systems 21st IAHR Symposium, Lausanne,
Switzerland, 2002.
[65] Q. Shi, “Abnormal noise and runner cracks caused by von Kàrmàn vortex
shedding: A case study in Dachaoshan hydroelectric project,” presented at the
Proceedings of the 22nd IAHR symposium on hydraulic machinery and systems,
Stockholm, Sweden, 2004.
[66] Y. Y. Song, D. H. Ping, F. X. Yin, X. Y. Li, and Y. Y. Li, “Microstructural evolution
and low temperature impact toughness of a Fe–13%Cr–4%Ni–Mo martensitic
stainless steel,” Mater. Sci. Eng. A, vol. 527, no. 3, pp. 614–618, Jan. 2010.
[67] D. Thibault, M. Gagnon, and S. Godin, “Bridging the gap between metallurgy and
fatigue reliability of hydraulic turbine runners,” IOP Conf. Ser. Earth Environ. Sci.,
vol. 22, no. 1, p. 012019, 2014.
[68] M. M. Amrei, Y. Verreman, F. Bridier, D. Thibault, and P. Bocher, “Microstructure
Characterization of Single and Multipass 13Cr4Ni Steel Welded Joints,” Metallogr.
Microstruct. Anal., vol. 4, no. 3, pp. 207–218, Jun. 2015.
[69] M. M. Amrei, H. Monajati, D. Thibault, Y. Verreman, L. Germain, and P. Bocher,
“Microstructure characterization and hardness distribution of 13Cr4Ni multipass weld
metal,” Mater. Charact., vol. 111, pp. 128–136, Jan. 2016.
[70] M. M. Amrei, H. Monajati, D. Thibault, Y. Verreman, and P. Bocher, “Effects of
Various Post-Weld Heat Treatments on Austenite and Carbide Formation in a
13Cr4Ni Steel Multipass Weld,” Metallogr. Microstruct. Anal., vol. 5, no. 1, pp. 50–
61, Feb. 2016.
[71] R. Simoneau, D. Thibault, and J.-L. Fihey, “A Comparison of Residual Stress in
Hammer-Peened, Multi-Pass Steel Welds–A514 (S690Q) and S41500,” Weld.
World, vol. 53, no. 5–6, pp. R124–R134, May 2009.
[72] L. B. Winck, J. L. A. Ferreira, J. A. Araujo, M. D. Manfrinato, and C. R. M. da
Silva, “Surface nitriding influence on the fatigue life behavior of ASTM A743 steel
type CA6NM,” Surf. Coat. Technol., vol. 232, pp. 844–850, Oct. 2013.
[73] E. Boudreault, B. Hazel, J. Côté, and S. Godin, “In situ post-weld heat treatment
on martensitic stainless steel turbine runners using a robotic induction heating
process to control temperature distribution,” IOP Conf. Ser. Earth Environ. Sci., vol.
22, no. 1, p. 012016, 2014.
[74] J. Arakawa et al., “Fatigue strength of USP treated ASTM CA6NM for hydraulic
turbine runner,” Surf. Eng., vol. 30, no. 9, pp. 662–669, Sep. 2014.
[75] S. Sarafan, P. Wanjara, H. Champliaud, D. Thibault, and L. Mathieu, “Evaluation
of Electron Beam Welded AISI 415 Stainless Steel,” p. V02BT02A013, Nov. 2014.
[76] S. Sarafan, P. Wanjara, H. Champliaud, and D. Thibault, “Characteristics of an
autogenous single pass electron beam weld in thick gage CA6NM steel,” Int. J. Adv.
Manuf. Technol., vol. 78, no. 9–12, pp. 1523–1535, Jun. 2015.
[77] F. Mirakhorli, X. Cao, X.-T. Pham, P. Wanjara, and J.-L. Fihey, “Hybrid Laser-Arc
Welding of 10-mm-Thick Cast Martensitic Stainless Steel CA6NM: As-Welded
Microstructure and Mechanical Properties,” Metall. Mater. Trans. A, vol. 47, no. 7,
pp. 3545–3563, Jul. 2016.
[78] F. Mirakhorli, X. Cao, X.-T. Pham, P. Wanjara, and J.-L. Fihey, “Post-weld
Tempered Microstructure and Mechanical Properties of Hybrid Laser-Arc Welded
Cast Martensitic Stainless Steel CA6NM,” Metall. Mater. Trans. B, vol. 47, no. 6, pp.
3245–3256, Dec. 2016.
[79] F. Mirakhorli, X. Cao, X.-T. Pham, P. Wanjara, and J. L. Fihey, “Phase structures
and morphologies of tempered CA6NM stainless steel welded by hybrid laser-arc
process,” Mater. Charact., vol. 123, pp. 264–274, Jan. 2017.
[80] H. H. Boukani, M. Viens, S. A. Tahan, and M. Gagnon, “On the performance of
nondestructive testing methods in the hydroelectric turbine industry,” IOP Conf. Ser.
Earth Environ. Sci., vol. 22, no. 1, p. 012018, 2014.
[81] J. H. Gummer and P. C. Hensman, “A review of stayvane cracking in hydraulic
turbines,” Int. Water Power Dam Constr., vol. 44, no. 8, pp. 32–42, 1992.
[82] E. Goldwag and D. G. Berry, “Von Karman Hydraulic Vortexes Cause Stay Vane
Cracking on Propeller Turbines at the Little Long Generating Station of Ontario
Hydro,” J. Eng. Power, vol. 90, no. 3, pp. 213–217, Jul. 1968.
[83] A. Y. Aronson, V. M. Zabelkin, and I. M. Pylev, “Causes of cracking in stay vanes
of Francis turbines,” Hydrotech. Constr., vol. 20, no. 4, pp. 241–247, Apr. 1986.
[84] A. D. Neto, D. H. Gissoni, D. M. Gonçalves, R. Cardoso, D. A. Jung, and P. J.
Meneghini, “Engineering diagnostics for vortex-induced stay vanes cracks in a
Francis turbine,” IOP Conf. Ser. Earth Environ. Sci., vol. 49, no. 7, p. 072017, 2016.
[85] B. Papillon, J. Brooks, J.-L. Deniau, and M. Sabourin, “Solving the guide vanes
vibration problem at Shasta,” presented at the Hydrovision International, United
States, 2006.
[86] F. Casanova, “Failure analysis of the draft tube connecting bolts of a Francis-type
hydroelectric power plant,” Eng. Fail. Anal., vol. 16, no. 7, Jan. 2009.
[87] F. C. García and C. A. M. Viveros, “Experimental analysis of the vibration on the
draft tube of a Francis hydraulic turbine during operation at different power levels,”
Rev. Fac. Ing., vol. 0, no. 55, pp. 90–98, Mar. 2013.
[88] “Standardizing Parameters for Managing Rough Load Zones and No-Run
Zones,” 2016. [Online]. Available:
http://www.hydroworld.com/articles/hr/print/volume-35/issue-
10/articles/standardizing-parameters-for-managing-rough-load-zones-and-no-run-
zones.html. [Accessed: 10-Feb-2017].
[89] T. Maricic, D. Haber, and S. Pejovic, “Standardization as Prevention of Fatigue
Cracking of Hydraulic Turbine-Generator Shaft,” in 2007 IEEE Canada Electrical
Power Conference, 2007, pp. 103–110.
[90] D. Momčilović, Z. Odanović, R. Mitrović, I. Atanasovska, and T. Vuherer, “Failure
analysis of hydraulic turbine shaft,” Eng. Fail. Anal., vol. 20, pp. 54–66, Mar. 2012.
[91] O. Morin and B. Cournoyer, “Assessing the Impact of Cyclic Loading on
Generators Life: An Emerging Concern,” presented at the Hydrovision International,
United States, 2016.
[92] M. Himmelreich, H. Schifferl, M. Gerhold, and F. Neumayer, “Fatigue Strength
Analysis of Highly Loaded Pole Endplates,” presented at the HYDRO 2013, Austria,
2013.
[93] T. Hildinger, D. Ludwig, H. Henning, and M. Hagmeyer, “Fatigue Assessment in
Hydro Generator Pole Fixation,” presented at the Hydrovision International 2015,
Portland, OR, USA, 2015.
[94] D. Ludwig and T. Hildinger, “Bring new life to Hydro Generator,” presented at the
Hydrovision International, United States, 2016.
[95] W. Ladstatter, M. Gerhold, and G. Lemesch, “Do Grid Stability Measures Reduce
the Residual Life Time of Hydro Generators?,” presented at the 10th ICOLD
European Club International Symposium : Dams and HEPPs for Sustainable
Development, Turkey, 2016.
[96] S. Suresh, Fatigue of Materials:, 2nd ed. Cambridge: Cambridge University
Press, 1998.
[97] T. L. Anderson, Fracture Mechanics: Fundamentals and Applications, Second
Edition. CRC Press, 1994.
[98] “ASTM E1049-85(2011)e1, Standard Practices for Cycle Counting in Fatigue
Analysis,” ASTM International, West Conshohocken, PA, 2011.
[99] “Boiler Pressure Vessel Code Section VIII-Rules for Construction of Pressure
Vessels Division 2-Alternative Rules,” ASME, 2015.
[100] K. S. Ravichandran, Y. Murakami, and R. O. Ritchie, Small Fatigue Cracks:
Mechanics, Mechanisms and Applications. Elsevier, 1999.
[101] P. Paris, M. Gomez, and W. Anderson, “A rational analytic theory of fatigue,”
Trend Eng., vol. 13, pp. 9–14, 1961.
[102] “Fatigue Crack Growth under Variable-Amplitude Loading,” in Fatigue of
Structures and Materials, Springer Netherlands, 2004, pp. 274–309.
[103] M. Sabourin, J.-L. Gagné, S. Giroux, A. St-Hilaire, and J. de la Bruyère-Terreault,
“Mechanical Loads and Fatigue analysis of a Francis Runner,” presented at the
Hydrovision, Canada, 2004.
[104] R. A. McManamay, C. O. Oigbokie, S.-C. Kao, and M. S. Bevelhimer,
“Classification of US Hydropower Dams by their Modes of Operation,” River Res.
Appl., vol. 32, no. 7, pp. 1450–1468, Sep. 2016.
[105] “Hydropower Vision: A New Chapter for America’s 1st Renewable Electricity
Source,” U.S. Department of Energy’s Wind and Water Power Technologies Office,
2016.
[106] “U.S. Army Corps of Engineers Northwestern Division, dataquery.” [Online].
Available: http://www.nwd-wc.usace.army.mil/dd/common/dataquery/www/.
[Accessed: 03-Mar-2017].
[107] C. McGowin and W. Steeley, “Wind Power Integration Technology Assessment
and Case Studies,” EPRI, 1004806, 2004.
[108] T. Key, “Quantifying the Value of Hydropower in the Electric Grid: Final Report,”
EPRI, 2013.
 [109] B. J. Kirby, “Frequency Regulation Basics and Trends,” Oak Ridge Natiobnal
Laboratory, ORNL/TM-2004/291, 2004.
[110] “ASTM A743 / A743M-13ae1, Standard Specification for Castings, Iron-
Chromium, Iron-Chromium-Nickel, Corrosion Resistant, for General Application,”
ASTM International, West Conshohocken, PA, 2013.

View publication stats