energies

Article
A New Study on the Effect of the Partial Wake Generated in a
Wind Farm
Said Zergane, Chouki Farsi, Salah Amroune *, Souad Benkherbache and Noureddine Menasri

Mechanical Engineering Department, University of M’sila, M’sila 28000, Algeria;

[email protected] (S.Z.); [email protected] (C.F.); [email protected] (S.B.);
[email protected] (N.M.)
* Correspondence: [email protected]; Tel.: +213-657271118

Abstract: In this article, we present an investigative study on the often-overlooked partial wake
phenomenon in previous studies concerning wind farm configurations. A partial wake occurs when
a portion of the actuator disk of a downstream wind turbine is affected by the wake of another
upstream turbine. This phenomenon occurs in addition to the full wake, where the entire upstream
turbine is affected by the wake of the frontal turbine, also leading to a decrease in wind speed and
consequently a reduction in power production. The proposed study is based on measuring the power
generated by the area swept by the wake of an array of turbines in a wind farm. To accomplish this,
we integrate the linear wake model of Jensen, the specifications of the ENERCON E2 wind turbine,
and the wind farm data into Matlab-developed software (version 18) to perform the calculations. In
a concrete application, this proposed method is validated by reproducing the previous works that
neglected the partial wake in wind farm configurations. The simulation results obtained are analyzed,
compared, and discussed under similar operational conditions.

Keywords: wind farm; wake model; renewable energy; partial wake; Jensen model

Citation: Zergane, S.; Farsi, C.;

1. Introduction
Amroune, S.; Benkherbache, S.;
Menasri, N. A New Study on the Wind energy is a renewable and sustainable energy source that relies on the inex-
Effect of the Partial Wake Generated haustible power of the wind. As wind forces act on the blades, they induce rotor rotation,
in a Wind Farm. Energies 2024, 17, which turns the generator, converting the wind’s energy into electricity. This generated
1498. https://doi.org/10.3390/ electric current can be either integrated into existing electrical grids or used directly in
en17061498 isolated places. Due to these advantages, the portion of wind energy to global electricity
production continues to increase, facilitating the installation of large-scale wind turbines in
Academic Editors: Davide Di Battista,
both offshore and onshore wind farms across the world [1]. However, the efficiency of wind
Fabio Fatigati and Marco Di
Bartolomeo
farms heavily relies on how the wind turbines are arranged, particularly in downstream
positions, where they are affected by the wake of the frontal wind turbines. The wake
Received: 11 January 2024 phenomenon, associated with dense groupings of wind turbines in wind farms, leads to a
Revised: 15 March 2024 speed deficit, affecting the performance of the turbines and reducing energy production [2].
Accepted: 19 March 2024 As a wind turbine can only transform a portion of the available wind energy from
Published: 21 March 2024
wind into clean electricity, another wind turbine totally or partially immersed in the
wake receives less energy. This phenomenon is called the wake effect [3,4]. In wind
farm configurations, the wake effect has been the subject of several studies carried out by
Copyright: © 2024 by the authors.
researchers employing different models and analytical methods to describe the evolution
Licensee MDPI, Basel, Switzerland. of wind speed and model the wakes generated within wind farms. The first works in
This article is an open access article this field were presented by Betz [5,6] and Lanchester [7] in the early 1920s. Their work
distributed under the terms and aimed at optimizing wind turbine performance and understanding downstream wind
conditions of the Creative Commons speed, forming the basis for future research on the characteristics of the wake zone. With
Attribution (CC BY) license (https:// advancements in computer technology, it has been observed that a number of wake models
creativecommons.org/licenses/by/ have been developed, including empirical-analytical models and computational fluid
4.0/). dynamics (CFD) models [8,9]. Empirical-analytical models, notable for their simplicity and

Energies 2024, 17, 1498. https://doi.org/10.3390/en17061498 https://www.mdpi.com/journal/energies

Energies 2024, 17, 1498 2 of 12

computational speed, are often referred to as explicit models [10–13], as highlighted by

Barthelmie [14] and Ishihara [15]. Other CFD models classified as implicit models [16,17],
such as those presented by Crasto et al. [18] and Crespo [19], generally involve more
complex mathematical tools and consume significant computational resources due to the
lack of flexibility in application to different problems [20]. Consequently, the majority of
commercial software for optimizing the positions of wind turbines in wind farms uses
simpler empirical-analytical models.
Although other analytical models exist that take into account three-dimensional wakes
and vertical wind shear, such as the one presented by Zhao [21], the linear model by
Jensen [22], which focuses on two-dimensional wakes, is widely adopted in optimization
methods. This empirical-analytical model is also widely used to estimate the wind speed
within the wake [23]. Due to its speed and ease of programming, it is particularly suitable
for wind farm layout systems, both onshore and offshore, predicting a linear expansion of
the wake [24,25]. Furthermore, it can be used to assess axial velocity profiles within the
wakes of underwater turbines.
In studies focusing on methods for optimizing the placement of wind turbines in
a wind farm, it is important to note that satisfactory agreement has been reported in
the results obtained from various methodologies employing the concept of a full wake
based on the Jensen model [22]. These methods encompass genetic algorithms [26,27],
the viral bases method [28], Monte Carlo simulation [29], and pseudo-random number
generation [30]. Throughout these previous studies, the phenomenon of partial wakes has
not been particularly pronounced. However, in other studies, this phenomenon has been
treated differently. Lio et al. [31] conducted a study aimed at estimating the lateral and
vertical position of the wake’s center using fixed four-beam LiDAR, considering that a
partial wake results from the interaction between wakes. Their conclusion indicates that,
in this context, the recovery of the wake center of upstream turbines is faster than that of
downstream turbines. By integrating both full and partial wakes, He et al. [32] presented
an analysis of the fatigue of wind turbines under shear flow. In various environmental
conditions, they indicated that yaw effects can induce significant load variation, which is
detrimental to structural stability. In another aspect of the research, Vad et al. [33] proposed
a new model of a non-symmetric Gaussian wake, where the wake affecting the actuator
disk of a leading turbine does not exhibit uniformity, thus leading to partial wake overlap.
This model uses a non-symmetric Gaussian function to represent the velocity deficit in the
wake, resulting in different forms of wake expansion and, consequently, the formation of
a partial wake. In another separate publication which focused on quantifying the wake
dynamics resulting from the turbine positioning, Scott et al. [34] conducted wind tunnel
experiments using reduced turbine models placed in both symmetric and asymmetric
configurations. Their results indicated that wake interactions are characterized by the mean
velocity over time and the turbulence intensity. The latter contributes to the full wake by
merging, thereby influencing wind turbine power production.
In conclusion, there is still no clear consensus on how to model partial wakes. To
address this gap, this article delves into the study and clarification of wakes and their
effects on energy production at a wind farm.
In this article, we present a new study that explores an additional phenomenon of the
full wake—the partial wake—as well as its influence on the power output of wind turbines
within a wind farm. This study is based on the use of the linear wake model of Jensen [22],
incorporating the characteristics of the ENERCON E2 wind turbine and the dimensions of
a typical wind park into a calculation program developed using Matlab while considering
a uniform and unidirectional wind speed on the actuator disk. The proposed methodology
is applied to a wind farm consisting of 45 turbines. The simulation results obtained are
discussed and compared to those from previous works that neglected the partial wake
effect. The rest of the article is structured as follows. In Section 2, we present the modeling
of the partial wake according to the analytical model of Jensen. Section 3 details the method
for calculating the power generated when the wake is partial. Section 4 delves into the
Energies 2024, 17, x FOR PEER REVIEW 3 of 12

glected the partial wake effect. The rest of the article is structured as follows. In Section 2,
Energies 2024, 17, 1498 3 of 12
we present the modeling of the partial wake according to the analytical model of Jensen.
Section 3 details the method for calculating the power generated when the wake is par-
tial. Section 4 delves into the analysis and discussion of the results, including a compar-
analysis
ison withand discussion
previous of In
works. theconclusion,
results, including a comparison
we summarize withofprevious
the impact works.
the present study.In
conclusion, we summarize the impact of the present study.
2. Jensen Wake Model
2. Jensen Wake Model
Jensen [22] developed a simple model to describe the variation in wind speed in the
Jensen [22] developed a simple model to describe the variation in wind speed in the
wake of a front turbine. Due to the simplicity of programming and ease of implementa-
wake of a front turbine. Due to the simplicity of programming and ease of implementation
tion offered by the Jensen model, many methods incorporate it to achieve optimal mi-
offered by the Jensen model, many methods incorporate it to achieve optimal micro-location
cro-location
of of wind
wind turbines in a turbines in aAs
wind farm. wind farm.
shown inAs shown
Figure inthis
1, in Figure 1, inthe
model, thiswind
model, the
speed
wind speed presented a “hat” profile in
presented a “hat” profile in the wake [35]. the wake [35].

Figure1.1.Jensen
Figure Jensenwake
wakemodel.
model.

Thewind
The windspeed
speedin inthe
thewake
wakeU Uww is
is described
described as
as follows
follows [36,37]:
[36,37]:
√
2𝑅 1 − 1 − 𝐶

!
𝑈 = 𝑈 1R − 1 − 1 − CT (1)
Uw = U0 1 − 𝑟 (1)
rx 2
where CT is the thrust coefficient, R is the blade radius of the wind turbine and rx is the
radiusCofT the
where is the thrust
wake at acoefficient, R is the
distance x from theblade
wind radius
turbineofthatthegenerated
wind turbine and rx is the
the wake.
radiusJensen
of the[22]
wake at a distance x from the wind turbine that generated the wake.
assumed that the increase in the wake radius rx was proportional to the
Jensen
position [22]itassumed
x, and depends onthatthe
the increase of
coefficient inwake
the wake radius
expansion rx the
α in wasfollowing
proportional to
relation
the position
[38,39]: x, and it depends on the coefficient of wake expansion α in the following
relation [38,39]:
r x 𝑟= =
R𝑅++αx𝛼𝑥 (2)
(2)
Also, the
Also, theentrainment
entrainment coefficient
coefficient ααdepends
dependsononthe
theroughness
roughnessof
ofthe
theground
groundZZ00and
and
theheight
the heightofofthe
theturbine
turbineZ:
Z:
1
α=  1 (3)
𝛼=
2ln ZZ0 𝑍 (3)
2𝑙𝑛
𝑍
2.1. Partial Wake Modeling within the Jansen Model
2.1. Partial Wake
In a wind Modeling
farm, within
we refer thepartial
to the Jansen wake
Modeland its associated effects, where a portion
of theInactuator
a wind farm, we refer to the partial turbine
disk of a downstream wind wake andis influenced by the
its associated wake
effects, of another
where a por-
wind turbine in the ∆r
tion of the actuator disk zone, as illustrated in Figure 2.
x of a downstream wind turbine is influenced by the wake of an-
other wind turbine in the Δrx zone, as illustrated in Figure 2.
Energies 2024, 17, x FOR PEER REVIEW 4 of 12
Energies 2024, 17, 1498
x FOR PEER REVIEW 4 4of
of 12
12

Figure 2. Wind turbine downstream affected by the partial wake in the Δrx zone.
Figure
Figure 2.
2. Wind
Wind turbine
turbine downstream
downstream affected
affected by
by the partial wake
the partial wake in
in the
the Δr
∆rxx zone.
zone.
In this figure, it is evident that the actuator disk of the downstream wind turbine is
scanned
In at two
In this
this different
figure, speeds:that
it is evident one the
without the wake
actuator disk ofeffect, denoted as Uwind
the downstream 0, and anotheris
turbine
with the at
scanned wake
two effect,
differentdenoted
speeds:asone
Ux. without
From this, the we
wakecan deduce
effect, the relationship
denoted as U00, and of the
and another
another
partial
with thewake
the wake
wake zone as follows:
effect,
effect, denoted
denoted as as
Ux .UFrom
x. From
this,this, we deduce
we can can deduce the relationship
the relationship of the
of the partial
wake zone
partial wake aszone
follows:
as follows:
Δ𝑟 = 𝑅 + 𝑟 − 𝑦 (4)
∆r x = R + r x − y (4)
Δ𝑟 = 𝑅 + 𝑟 − 𝑦 (4)
2.2.Method
2.2. Methodfor forDetermining
Determiningthe theRegion
RegionImpacted
Impactedby byaaPartial
PartialWake
Wakein inthe
theJansen
JansenModel
Model
2.2. Method
Inthe
In for Determining
thepartial
partial wake,
wake, it isitthe
isRegion Impacted
distinguished
distinguished by
thea actuator
thatthat Partial Wake
diskinof
the actuator the
theJansen
disk of theModel
downstream
downstream wind
wind
turbine turbine
operates
In the operates
partialinwake,in two
two different different regions:
regions: one
it is distinguished one characterized
thatcharacterized by
by aofwind
the actuator disk a wind speed deficit
speed deficit
the downstream re-
wind
resulting
sulting
turbine in in a partial
a partial
operates reduction
in reduction
two different inindeveloped
developed
regions: power
power
one andanother
and another
characterized where
by awhere
wind the thewind
speedwind speed
speed
deficit re-
remains
remains constant
constant (Figure
(Figure 3).
3).
sulting in a partial reduction in developed power and another where the wind speed
remains constant (Figure 3).

(a)
Figure 3. Cont.
(a)
EnergiesEnergies
2024, 17,2024,
x FOR17,PEER
1498 REVIEW 5 of 12
5 of 12

(b)
Figure Figure
3. (a) The upstream
3. (a) wind turbine
The upstream (right) (right)
wind turbine generates a wakea that
generates wakepartially influences
that partially the
influences the
downstream wind turbine (left). (b) Detailed partial wake area.
downstream wind turbine (left). (b) Detailed partial wake area.

At a distance x, the wake

At a distance x, the surface Aw is Aw is
wake surface
𝐴 =A𝜋𝑟= πr 2 (5) (5)
w x
In addition, the surface AR of the actuator disc of the wind turbine is
In addition, the surface AR of the actuator disc of the wind turbine is
𝐴 = 𝜋𝑅 (6)
A R = πR2 (6)
Therefore, the partial wake acting on the part δA (Figure 3b) is
ℎ wake acting on the part δAℎ(Figure 3b) is
Therefore, the partial
𝛿𝐴 = 𝑎𝑟𝑐𝑠𝑖𝑛 𝑟 − ℎ 𝑟 − ℎ + 𝑎𝑟𝑐𝑠𝑖𝑛 𝑅 −ℎ 𝑅 −ℎ (7)
𝑟  𝑅  
h p h p
with δA = arcsin r x 2 − h r x 2 − h2 + arcsin R2 − h R2 − h2 (7)
rx R

with s 𝑟 −𝑅 +𝑦
ℎ= 𝑟 − 2 2
(8)
r2𝑦
 2 2
x −R +y

2
h = rx − (8)
2y

3. Power Generated
3. Power in the in
Generated Case
theof a Partial
Case Wake Wake
of a Partial
The total
Thepower Ptot in the
total power Ptotcase of the
in the casepartial
of thewake canwake
partial be calculated as the sum
can be calculated asofthe
thesum of
products of the partial powers and the ratios of the corresponding regions,
the products of the partial powers and the ratios of the corresponding regions, which which are as are
follows:
as follows:
A R − δA δA
Ptot = P(U 𝐴 0 )− 𝛿𝐴 + P(𝛿𝐴UW ) (9)
𝑃 =𝑃 𝑈 A+R 𝑃 𝑈 A R (9)
𝐴 𝐴
δA A R −δA
where P(U0 ), P(UW ), A , and AR represent the power without the wake effect, the
𝑃 𝑈 ,with
wherepower 𝑃 𝑈 the, wake, and R
effect, therepresent
area ratio theaffected
power bywithout the wake
the wake, effect,
and the area the
ratio not
poweraffected
with the bywake effect,respectively.
the wake, the area ratio affected by the wake, and the area ratio not
affected by the wake, respectively.
Approach to Estimating the Power Produced by a Wind Farm
Approach toIn Estimating
Figure 4, the
wePower
present Produced by a Wind
the proposed Farm for determining the power developed
flowchart
Inwhen the4,wake
Figure affects a the
we present portion of the actuator
proposed flowchart disc
forofdetermining
a downstream theturbine.
power The program
devel-
begins by identifying the zone where the partial wake operates and
oped when the wake affects a portion of the actuator disc of a downstream turbine. The introducing the required
parameters, such as the turbine’s thrust coefficient
program begins by identifying the zone where the partial wake CT , the ground roughness Z
operates and introducing0 and the
,
upstream wind speed
the required parameters, such as U . If the condition of existence of the partial wake
0 the turbine’s thrust coefficient CT, the ground roughness is confirmed,
Z0, andthen
the the wind speed
upstream wind at eachUzone
speed of the actuator disc, whether affected by the wake or
0. If the condition of existence of the partial wake is
 Energies 2024, 17, x FOR PEER REVIEW 6 of 12

Energies 2024, 17, 1498 6 of 12

confirmed, then the wind speed at each zone of the actuator disc, whether affected by the
wake or not, is determined according to Equation (1). Otherwise, the wind speed remains
constant. Subsequently, the developed power for each zone can be calculated using the
not, is determined
power curve of the according
ENERCONto E2Equation
turbine. (1). Otherwise, the wind speed remains constant.
Subsequently,
Thus, thethe developed
total power Ptot power for turbine
of a single each zone
can can be calculated
be computed using the
by summing thepower
par- curve
oftial
thepowers.
ENERCON E2 turbine.

Figure4.4.Proposed
Figure Proposed flowchart
flowchartfor
forthe numerical
the simulation.
numerical simulation.

4. Results and total

Thus, the Discussion
power P
tot of a single turbine can be computed by summing the partial
To perform the numerical simulations, it was important to represent the corre-
powers.
sponding power curve of the ENERCON E2 wind turbine (see Figure 5) [40] with a pol-
4.ynomial
Resultsequation as follows:
and Discussion
To perform the numerical simulations, it was important to represent the corresponding
𝑃 𝑈 = 𝛼𝑈 (10)
power curve of the ENERCON E2 wind turbine (see Figure 5) [40] with a polynomial
equation as follows:
I
Energies 2024, 17, x FOR PEER REVIEW P (U ) = ∑ i =1 α i U i 7 of 12 (10)

Figure 5.5.Power
Figure Powercharacteristic curvecurve
characteristic of the E2
of ENERCON turbine. turbine.
the E2 ENERCON

Equation (10) and the turbine data, with a diameter D = 82 m and a hub height Z = 85
m, were introduced in the program as indicated in the flowchart (Figure 4).

4.1. Interval of the Partial Wake at a Wind Farm

In a given wind farm, we considered that the typical longitudinal and transverse
distances between the rows of turbines were to the order of 5 D (where D is the diameter
Energies 2024, 17, 1498 7 of 12

Figure 5. Power characteristic curve of the E2 ENERCON turbine.

Equation (10) and the turbine data, with a diameter D = 82 m and a hub height
Z = 85 m, were(10)
Equation introduced in the program
and the turbine as aindicated
data, with diameterinDthe
= 82flowchart (Figure
m and a hub 4). Z = 85
height
m, were introduced in the program as indicated in the flowchart (Figure 4).
4.1. Interval of the Partial Wake at a Wind Farm
In a given
4.1. Interval of thewind
Partialfarm,
Wakeweat aconsidered
Wind Farm that the typical longitudinal and transverse
distances between the rows of turbines
In a given wind farm, we considered werethat
to the
theorder D (where D and
of 5longitudinal
typical is thetransverse
diameter
of the turbine)
distances between(Figure 6). According
the rows of turbinesto were
Jensen’s model
to the [22],
order of 5we could estimate
D (where D is thethe extent
diameter
of the partial wake’s influence on neighboring rows of wind turbines using Equation
of the turbine) (Figure 6). According to Jensen’s model [22], we could estimate the extent (2),
regardless of the wind speed. The calculated results indicate that the partial wake must
of the partial wake’s influence on neighboring rows of wind turbines using Equation (2),
verify the following criteria:
regardless of the wind speed. The calculated results indicate that the partial wake must
verify the following criteria:
 x ≤ 45.16D No wake effect
𝑥 45.16𝐷
45.16D No wake effect
< x < 56.49D Partialwake (11)
45.16𝐷 ≤ x𝑥 56.49𝐷 Partialwake
56.49D Total wake (11)

56.49𝐷 𝑥 Total wake

Figure 6. Wake expansion.

Figure 6. Wake expansion.

The
The investigation
investigation into
into the
the partial
partial wake’s
wake’simpact
impacton onthe
the developed
developedpower
powerwas wascon-
con-
ducted
ducted for
for the
the ENERCON
ENERCON E2 E2 wind
wind turbine operating at
turbine operating at aa wind
wind speed
speed UU == 66 m/s
m/s and
and
ground
ground roughness
roughness Z Z00 ==0.3
0.3 m.
m. Calculations
Calculationswere werecarried
carriedout
outbyby varying
varying only
only the
the region
region
affected
affected by the
thepartial
partialwake
wakewithwiththethe variation
variation of the
of the downstream
downstream distance
distance x, ranging
x, ranging from
from 45 D to 60 D. The results obtained are presented
45 D to 60 D. The results obtained are presented in Table 1. in Table 1.

Table
Table 1.
1. Characteristics
Characteristics and
and data
data of
of the
the ENERCON
ENERCON E2
E2 turbine.
turbine.
xx(m)
(m) UU(m/s)
(m/s) rxr(m)
x (m)
PP(kW)
(kW)
45 D 6 367.743 316.027
45 D 6 367.743 316.027
50
50 D
D 5.98
5.98 367.74
367.74 314.715
314.715
55
55 D
D 5.99
5.99 440.35
440.35 314.217
314.217

For different values of the position x, the wind speed in the wake gradually returned to
6 m/s far downstream. Furthermore, the radius of the wake expanded, gradually reaching
the downstream wind turbines when starting from x = 50 D. As a result, the power output
of a single wind turbine initially decreased, eventually stabilizing at a nearly constant value.
This phenomenon can be attributed to the wind stabilizing at a specific level.

4.2. The Power Generated by the Wind Farm

As an application of this method, we propose a wind farm in given dimensions
(60 D × 60 D), comprising 45 ENERCON E2 turbines distributed across 144 identical cells
aligned with a predominant wind direction as shown in Figure 7. The symbol X denotes
the position of each turbine in the wind farm.
 4.2. The Power Generated by the Wind Farm
As an application of this method, we propose a wind farm in given dimensions (60
D × 60 D), comprising 45 ENERCON E2 turbines distributed across 144 identical cells
Energies 2024, 17, 1498 aligned with a predominant wind direction as shown in Figure 7. The symbol
8 ofX
12 denotes
the position of each turbine in the wind farm.

Figure 7. Distribution
Figure 7. Distributionofofwind
wind turbines
turbines in the
in the proposed
proposed wind
wind parkpark according
according to theto the predominant
predominant
wind direction,indicated
wind direction, indicated
by by
thethe arrows.
arrows.

With a uniform wind speed U = 6 m/s and a ground roughness Z0 = 0.3 m, the simu-
With a uniform wind speed U = 6 m/s and a ground roughness Z0 = 0.3 m, the sim-
lation results obtained, accounting for the partial wake effect, indicate that the calculated
ulation results obtained, accounting for the partial wake effect, indicate that the calcu-
power output of the wind farm was 13,137 MW.
lated power output of the wind farm was 13,137 MW.
4.3. Validity Study
4.3. Validity
In orderStudy
to study the validity of the proposed method, we replicated and integrated
the previous
In order works by Mosetti
to study et al. [27],
the validity of Zergane et al. [38],
the proposed and Ituarte-Villarreal
method, we replicatedetand al. [28],
integrated
illustrated in Figure 8a–c, respectively, into our MATLAB program along with their
the previous works by Mosetti et al. [27], Zergane et al. [38], and Ituarte-Villarreal et al.operat-
ing conditions.
[28], illustrated in Figure 8a–c, respectively, into our MATLAB program along with their
The results obtained from reproducing the previous works using the developed pro-
operating conditions.
gram are presented in Table 2.

Table 2. Comparison of the results obtained in the previous studies.

Previous Works P (MW) without Partial Wake Effect P (MW) with Partial Wake Effect U (m/s) %∆P/P
Mosetti et al. [27] 12.352 12.112 12 1.98
Zergane et al. [38] 5.058 4.862 6 4.03
Ituarte-Villarreal [28] 14.814 14.511 12 2.08

As indicated in Table 2, the greatest percentage of power difference was recorded in

the study by Zergane et al. [38] (4.03%), while the lowest was that of Mosetti [27] (1.98%).
This can be explained as follows:
− In Mosetti’s study [27], the 26 wind turbines arranged in the optimal configuration were
more widely spaced in a park of 100 cells, resulting in less partial wake interference.
− Conversely, in the study by Zergane et al. [38], the 55 wind turbines in the park were
distributed across a dense arrangement of 144 cells, leading to increased interference
among the wind turbines and consequently a significant partial wake effect.
It is important to note that in the study by Zergane et al. [38], the total power of the
45 wind turbines was lower than those calculated by Ituarté et al. [28] and Mosetti [27],
who used 30 and 26 turbines, respectively. This difference can be attributed to the varying
types of turbines used in each wind farm, as well as consideration of different wind speeds
during the calculations.
Energies
Energies2024,
2024,17,
17,x 1498
FOR PEER REVIEW 9 of
9 of1212

(a)

(b)

(c)
Figure
Figure 8. Optimal arrangement:
8. Optimal arrangement: (a)
(a)Moseti
Moseti et [27],
et al. al. (b)
[27],Zergane
(b) Zergane et and
et al. [38], al. (c)—Ituarte-
[38], and
(c)—Ituarte-Villarreal
Villarreal et al. [28]. et al. [28].

5. Conclusions
The results obtained from reproducing the previous works using the developed
program are presented
In this study, we in Table 2. the impact of a partial wake on the power generated
investigated
by a wind farm. Previous studies have often neglected the partial wake, concentrating
Table 2. on
solely Comparison of the results
the total wake when itsobtained in theradius
expanding previous studies. the center of the downstream
intersects
Previous Works P (MW)wind turbine’s
without rotor.
Partial WakeHowever,
Effect inPreality,
(MW) both
withthe total Wake
Partial wake Effect
and partial wake consistently
U (m/s) %ΔP/P
contribute to wind turbine interference. Using Jensen’s wake model, along with the
Mosetti et al. [27] 12.352 12.112 12 1.98
characteristics of the ENERCON E2 wind turbine and wind farm data, we developed a
Zergane et al. [38] 5.058 4.862 6 4.03
Matlab program to calculate the total power. The results obtained revealed the following:
Ituarte-Villarreal [28] 14.814 14.511 12 2.08
Energies 2024, 17, 1498 10 of 12

− This method enables the determination of the range within which the partial wake
operates. This range extended between 45.16 D and 56.49 D, thereby facilitating a
more precise determination of the power developed by a wind farm at a given speed.
− By comparing the results obtained in this study with those of previous works, it
becomes apparent that the application of this method is more decisive in denser parks,
where encounters with partial wakes are more frequent.
− In a broader context, this method could be used as a reference tool for accurately
evaluating the power produced by a wind farm.
Finally, as a perspective and continuation to this work, a more detailed study that
considers the influence of wind turbine components, specifically the nacelle and the tower,
could be conducted in the future. This would enable more precise prediction of the wind
turbine power within a partial wake zone.

Author Contributions: Conceptualization, S.Z.; Software, N.M.; Validation, C.F.; Formal analysis,
S.Z. and S.B.; Writing—original draft, S.Z.; Writing—review & editing, S.A. All authors have read
and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Data Availability Statement: The data presented in this study are available on request from the
corresponding author.
Conflicts of Interest: The authors declare no conflict of interest.

List of Symbols

AR Actuator disk area (m2 )

Aw Wake area (m2 )
CT Thrust coefficient
D Rotor diameter (m)
R Rotor radius (m)
P(UW ) Power with wake effect (W)
P(U0 ) Power without wake effect (W)
Ptot Total power (W)
rx Wake radius at x position (m)
∆rx Area affected by the partial wake
x Wake downstream position (m)
U Wind speed (m/s)
Ux Wind speed at x wake downstream position (m/s)
U0 Wind speed without wake effect (3/s)
UW Wind speed in the wake (m/s)
y Distance between the centers of two neighboring wind turbines (m)
Z Hub height (m)
Z0 Ground roughness (m)
α Entrainment coefficient
δA Surface swept by the partial wake (m2 )

References
1. Korolev, V.G. Development prospects of wind energy in the Russian energy complex. Electr. J. 2022, 35, 107094. [CrossRef]
2. Veena, R.; Manuel, S.M.; Mathew, S.; Petra, M.I. Wake induced Power Losses in Wind Farms. Int. J. Eng. Adv. Technol. 2020, 9,
2175–2180.
3. Lin, J.; Wei, Z.; Wei, J.; Shen, W.Z. New engineering wake model for wind farm applications. Renew. Energy 2022, 198, 1354–1363.
[CrossRef]
4. Hu, W.; Huang, Q.; Wu, X.; Li, J.; Zhang, Z. Wind farm control and optimization. In Control of Power Electronic Converters and
Systems; Academic Press: Cambridge, MA, USA, 2021; pp. 609–644.
5. Coelho, P. The Betz limit and the corresponding thermodynamic limit. Wind Eng. 2023, 47, 491–496. [CrossRef]
Energies 2024, 17, 1498 11 of 12

6. Gumilar, L.; Sholeh, M.; Triharto, R.; Rumokoy, S.N.; Monika, D.; Aji, A.F. Influence of Wind Turbine Pitch Angle on DFIG Output
Stability under Load Changes. In Proceedings of the 2021 4th International Seminar on Research of Information Technology and
Intelligent Systems (ISRITI), Yogyakarta, Indonisia, 16 December 2021; IEEE: Piscataway, NJ, USA, 2021; pp. 435–439.
7. Lanchester, F.W. A contribution to the theory of propulsion and the screw propeller. J. Am. Soc. Nav. Eng. 1915, 27, 509–510.
[CrossRef]
8. Uchida, T. Effects of inflow shear on wake characteristics of wind-turbines over flat terrain. Energies 2020, 13, 3745. [CrossRef]
9. Hertwig, D.; Gough, H.L.; Grimmond, S.; Barlow, J.F.; Kent, C.W.; Lin, W.E.; Robins, A.G.; Hayden, P. Wake characteristics of tall
buildings in a realistic urban canopy. Bound. Layer Meteorol. 2019, 172, 239–270. [CrossRef]
10. Peña, A.; Réthoré, P.E.; Van Der Laan, M.P. On the application of the Jensen wake model using a turbulence-dependent wake
decay coefficient: The Sexbierum case. Wind Energy 2016, 19, 763–776. [CrossRef]
11. Duc, T.; Coupiac, O.; Girard, N.; Giebel, G.; Göçmen, T. Local turbulence parameterization improves the Jensen wake model and
its implementation for power optimization of an operating wind farm. Wind Energy Sci. 2019, 4, 287–302. [CrossRef]
12. Thøgersen, E.; Tranberg, B.; Herp, J.; Greiner, M. Statistical meandering wake model and its application to yaw-angle optimisation
of wind farms. J. Phys. Conf. Ser. 2017, 854, 012017. [CrossRef]
13. Lopes, A.M.; Vicente, A.H.; Sanchez, O.H. Operation assessment of analytical wind turbine wake models. J. Wind Eng. Ind.
Aerodyn. 2022, 220, 104840. [CrossRef]
14. Barthelmie, R.; Larsen, G.; Frandsen, S.; Folkerts, L.; Rados, K.; Pryor, S.; Lange, B.; Schepers, G. Comparison of wake model
simulations with offshore wind turbine wake profiles measured by sodar. J. Atmos. Ocean. Technol. 2006, 23, 888–901. [CrossRef]
15. Ishihara, T.; Yamaguchi, A.; Fujino, Y. Development of a new wake model based on a wind tunnel experiment. Glob. Wind Power
2004, 105, 33–45.
16. Yang, J.; Fang, L.; Song, D.; Su, M.; Yang, X.; Huang, L.; Joo, Y.H. Review of control strategy of large horizontal axis wind turbines
yaw system. Wind Energy 2021, 24, 97–115. [CrossRef]
17. Hansen, M.H. Using an Unsteady Panel Method with Fluid-Structure Interaction Problems. In IUTAM/IFToMM Symposium on
Synthesis of Nonlinear Dynamical Systems: Proceedings of the IUTAM/IFToMM Symposium, Riga, Latvia, 24–28 August 1998; Springer:
Dordrecht, The Netherlands, 2000; pp. 139–148.
18. Crasto, G.; Gravdahl, A.R. CFD wake modeling using a porous disc. In Proceedings of the European Wind Energy Conference &
Exhibition 2008, Brussels, Belgium, 31 March–3 April 2008; Volume 15.
19. Crespo, A.; Hernandez, J.; Frandsen, S. Survey of modelling methods for wind turbine wakes and wind farms. Int. J. Prog. Appl.
Wind Power Convers. Technol. 1999, 2, 1–24. [CrossRef]
20. Ti, Z.; Deng, X.W.; Yang, H. Wake modeling of wind turbines using machine learning. Appl. Energy 2020, 257, 114025. [CrossRef]
21. Huang, H.S. Distributed genetic algorithm for optimization of wind farm annual profits. In Proceedings of the 2007 International
Conference on Intelligent Systems Applications to Power Systems, Kaohsiung, Taiwan, 5–8 November 2007; IEEE: Piscataway, NJ,
USA, 2007; pp. 1–6.
22. Jensen, N.O. A Note on Wind Generator Interaction; Technical Report; Riso-M-2411; Riso National Laboratory: Roskilde, Den-
mark, 1983.
23. Zhang, S.; Gao, X.; Ma, W.; Lu, H.; Lv, T.; Xu, S.; Wang, Y. Derivation and verification of three-dimensional wake model of
multiple wind turbines based on super-Gaussian function. Renew. Energy 2023, 215, 118968. [CrossRef]
24. Feng, D.; Li, L.K.; Gupta, V.; Wan, M. Component wise influence of upstream turbulence on the far-wake dynamics of wind
turbines. Renew. Energy 2022, 200, 1081–1091. [CrossRef]
25. Hewitt, S.; Margetts, L.; Revell, A. Building a digital wind farm. Arch. Comput. Methods Eng. 2018, 25, 879–899. [CrossRef]
26. Grady, S.A.; Hussaini, M.Y.; Abdullah, M.M. Placement of wind turbines using genetic algorithms. Renew. Energy 2005, 30,
259–270. [CrossRef]
27. Mosetti, G.; Poloni, C.; Diviacco, B. Optimization of wind turbine positioning in large windfarms by means of a genetic algorithm.
J. Wind Eng. Ind. Aerodyn. 1994, 51, 105–116. [CrossRef]
28. Ituarte-Villarreal, C.M.; Espiritu, J.F. Optimization of wind turbine placement using a viral based optimization algorithm. Procedia
Comput. Sci. 2011, 6, 469–474. [CrossRef]
29. Marmidis, G.; Lazarou, S.; Pyrgioti, E. Optimal placement of wind turbines in a wind park using Monte Carlo simulation. Renew.
Energy 2008, 33, 1455–1460. [CrossRef]
30. Zergane, S.; Smaili, A.; Masson, C. Optimization of wind turbine placement in a wind farm using a new pseudo-random number
generation method. Renew. Energy 2018, 125, 166–171. [CrossRef]
31. Lio, W.H.; Larsen, G.C.; Thorsen, G.R. Dynamic wake tracking using a cost-effective LiDAR and Kalman filtering: Design,
simulation and full-scale validation. Renew. Energy 2021, 172, 1073–1086. [CrossRef]
32. He, R.; Yang, H.; Lu, L. Optimal yaw strategy and fatigue analysis of wind turbines under the combined effects of wake and yaw
control. Appl. Energy 2023, 337, 120878. [CrossRef]
33. Vad, A.; Tamaro, S.; Bottasso, C.L. A non-symmetric Gaussian wake model for lateral wake-to-wake interactions. J. Phys. Conf.
Ser. 2023, 2505, 012046. [CrossRef]
34. Scott, R.; Viggiano, B.; Dib, T.; Ali, N.; Hölling, M.; Peinke, J.; Cal, R.B. Wind turbine partial wake merging description and
quantification. Wind Energy 2020, 23, 1610–1618. [CrossRef]
Energies 2024, 17, 1498 12 of 12

35. Katic, I.; Hojstrup, J.; Jensen, N.O. A Simple Model for Cluster Efficiency. In Proceedings of the European Wind Energy Association
Conference and Exhibition, Rome, Italy, 7–9 October 1986; Volume 1, pp. 407–410.
36. Seim, F.; GravdahL, A.R.; Adaramola, M.S. Validating of kinematic wind turbine wake models in complex terrain using actual
wind farm production data. Energy 2017, 123, 742–753. [CrossRef]
37. Sun, H.; Gao, X.; Yang, H. Validations of three-dimensional wake models with the wind field measurements in complex terrain.
Energy 2019, 189, 116213. [CrossRef]
38. Zergane, S.; Amroune, S.; Rokbi, M.; Guesmia, S. New study on the extension of a current wind farm, case of Kaberten park in
Algeria. Turk. J. Computer Math. Educ. 2022, 13, 428–434.
39. Tong, W.; Chowdhury, S.; Zhang, J.; Messac, A. Impact of different wake models on the estimation of wind farm power generation.
In Proceedings of the 12th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference and 14th AIAA/ISSMO,
Multidisciplinary Analysis and Optimization Conference, Indianapolis, IN, USA, 17–19 September 2012; p. 5430.
40. Available online: https://fr.wind-turbine-models.com/turbines/550-enercon-e-82-e2-2.300/ (accessed on 1 January 2023).

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.