Int
ern
at
i
onal
Journ
al of
P
ower E
le
ctr
on
i
cs a
n
d
Drive
S
ystem
(I
J
PE
D
S
)
Vo
l.
11
,
No.
3
,
Septem
be
r 2020
, pp.
1519
~
1526
I
SS
N:
20
88
-
8694
,
DOI: 10
.11
591/
ij
peds
.
v11.i
3
.
pp
1519
-
1526
1519
Journ
al h
om
e
page
:
http:
//
ij
pe
ds
.i
aescore.c
om
Impact o
f field r
ough
ness and p
ower losse
s, tur
bu
l
ence i
nte
nsity
on ele
ctricity pr
oducti
on for a
n on
shore wi
nd farm
Be
dri Dr
agus
ha
,
Bukuri
je
Ho
xh
a
Facul
ty
of
Me
ch
ani
c
al
Engi
n
ee
ri
ng,
Dep
art
m
ent
of
The
r
moe
n
erg
et
i
cs
and
Rene
w
abl
e
Energy,
Uni
ver
sity
of
Prishti
na,
Kos
ovo
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Dec
1
8
, 201
9
Re
vised
M
a
r
14
, 2
0
20
Accepte
d
Apr
1
3
, 20
20
Whe
n
design
ing
a
powe
r
gene
r
a
ti
on
project
fro
m
a
diff
ere
nt
so
urc
e,
and
in
our
c
ase
study
,
wind,
when
calc
ul
at
ing
th
e
an
nual
ene
rgy
pro
duce
d,
i
t
is
nec
essary
to
def
i
ne
and
ca
l
culate
the
losses
in
cur
r
ed
in
the
sys
te
m.
The
m
ai
n
ca
use
o
f
losses
in
a
win
d
p
ar
k
is
du
e
to
th
e
oscillations
c
a
used
by
the
turbul
en
ce
of
th
e
a
ir
aro
und
the
turbi
ne
be
ca
use
of
roughne
ss
of
te
rra
in.
Th
e
pape
r
desc
r
ibe
s
t
wo
me
thods
of
e
stim
ating
turbule
nce
in
te
nsity
:
on
e
base
d
on
the
mean
and
standa
rd
deviat
i
on
(SD
)
of
wi
nd
spee
d
fro
m
th
e
n
acel
l
e
ane
mo
me
t
er,
th
e
othe
r
fro
m
me
a
n
power
output
and
i
ts
SD
.
Thes
e
an
al
yses
are
ver
y
im
port
a
nt
for
under
stan
ding
the
f
at
igu
e
and
mecha
ni
cal
stress
on
the
wind
turbi
nes.
T
hen
significance
of
the
sit
e
rugg
edne
ss
inde
x
(R
IX)
and
the
associa
t
ed
p
erf
o
rma
nc
e
ind
icato
r
(ΔRIX
)
are
co
nfirm
ed
for
te
rr
ai
n
and
th
e
conse
quences
o
f
appl
y
ing
W
As
P
outside
its
oper
ating
en
vel
ope
ar
e
quant
ified.
Ke
yw
or
d
s
:
Win
d
e
nerg
y
Win
d
sim
ulati
on
s
Energ
y pro
du
c
ti
on
Eff
ic
ie
nc
y
Wak
e
losse
s
Turb
ulence
This
is an
open
acc
ess arti
cl
e
un
der
the
CC
BY
-
SA
l
ic
ense
.
Corres
pond
in
g
Aut
h
or
:
Bukurije
Hox
ha
,
Faculty
of
M
ec
han
ic
al
E
ngine
erin
g,
Dep
a
rtme
nt of
Thermoe
ne
rg
et
ic
s and Rene
w
able E
nergy,
U
niv
e
rsity
of
Pr
i
sh
ti
na, K
osovo
Emai
l:
b
uk
ur
ij
e.hoxha
@uni
-
pr.ed
u
1.
INTROD
U
CTION
The
st
ru
ct
ur
e
,
op
e
rati
on,
a
nd
plan
ning
of
e
le
ct
ric
powe
r
netw
orks
will
unde
rgo
c
onsidera
ble
a
nd
rap
i
d
c
hanges
du
e
to
i
ncr
ease
d
global
ene
r
gy
c
on
s
umpti
on
.
Th
ere
fore,
el
ect
ric
util
it
y
com
pan
ie
s
are
s
trivin
g
to
instal
l
wind
turb
i
nes
as
e
ne
rgy
res
ource
s
to
mee
t
gro
wing
c
us
to
mer
l
oa
d
de
man
d
[1,
2]
.
The
mech
anical
energ
y
f
rom
w
ind
a
fter
getti
ng
c
onver
te
d
i
nto
el
ect
rical
ene
rgy
th
r
ough
tu
r
bin
e
a
nd
gen
e
r
at
or
,
enter
s
int
o
the
colle
ct
or
s
ys
te
m
of
the
wi
nd
far
m
[
3].
Win
d
tu
rb
i
ne
te
c
hnol
ogy
has
un
derg
on
e
a
re
vo
luti
on
duri
ng
the
la
st
centu
ry
[4,
5].
Be
cause
of
exc
essive
in
vestm
ent,
cl
imat
e
limit
at
ion
s,
a
nd
mainte
na
nce
di
ff
ic
ulti
es
of
off
sh
or
e
wind
farms
c
ompa
red
to
ons
hore
wind
farms,
onsho
re
w
ind
fa
rms
a
re
al
so
at
tract
in
g
increa
sin
g
at
te
ntion
from
wi
nd
po
wer
co
mp
a
nies
[
6].
More
f
undame
ntall
y,
un
der
sta
nd
i
ng
a
nd
reli
ably
pre
di
ct
ing
wind
dy
namics
remains
a
ce
nt
ral
pr
ob
le
m
i
n
orde
r
to
f
oreca
st
wi
nd
power
[
7].
T
urb
ulence
in
the
wind
is
ca
us
e
d
by
dissipati
on
of
the
wind’s
ki
netic
energ
y
into
the
rmal
energ
y
th
rou
gh
the
creati
o
n
an
d
destr
uction
o
f
pro
gr
essi
vely
s
mall
er
ed
dies
[
8,
9].
To
cal
cu
la
te
the
am
ount
of
ene
r
gy
obt
ai
ned
f
rom
a
t
urbine
locat
ed
at
the
aforeme
ntio
ne
d
l
ocati
on,
t
he
analyti
cal
pa
th
is
us
e
d,
rec
ognizing
the
stre
ng
t
h
of
the
wind
a
nd
the
me
chan
ic
a
l
stren
gth
ob
ta
in
ed
f
rom
it
[
10,
11].
M
ore
f
unda
mental
ly,
unde
rstan
ding
an
d
reli
ably
pre
dicti
ng
wind
dyna
mics
remains
a
ce
nt
ral
pro
blem
in
order
t
o
f
or
eca
st
wind
powe
r
[12].
T
he
r
ugge
dn
e
ss
in
dex
c
on
ce
pt
ha
s
bee
n
us
e
d
extensi
vely
ov
er
t
he
la
st
10
ye
ars
in
wind
res
ource
as
se
ssme
nt
a
nd
sit
ing
stu
dies
in
c
omplex
te
rr
ai
n
,
especial
ly
i
n
te
rr
ai
n,
w
hich
is
ou
tsi
de
t
he
ope
rati
on
al
en
vel
ope
of
li
ne
arise
d
flo
w
m
od
el
s
su
c
h
as
W
AsP
[13].
Ko
s
ovo
as
a
si
gn
at
ory
of
t
he
Energ
y
Co
mm
un
it
y
Treat
y
plans
the
co
ns
tr
uc
ti
on
a
nd
us
e
of
re
new
a
ble
e
ne
rgy
capaci
ti
es
in
the
value
of
20
%.
On
e
of
the
wa
ys
of
us
i
ng
ren
e
wa
ble
e
nerg
y
i
n
Koso
vo
is
t
he
us
e
of
wi
nd
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
11
, N
o.
3
,
Se
ptembe
r
2020
:
15
19
–
15
26
1520
powe
r
[
14].
Si
te
ru
gge
dness
ind
e
x
cal
culat
ion
s
ha
ve
no
w
been
im
pleme
nted
in
t
he
W
AsP
program
f
or
al
l
pr
e
dictor
a
nd
pr
e
dicte
d
sit
es
[15].
The
ma
jor
go
al
of
O
ptimal
Power
Flow
(OPF
)
is
to
imp
rove
a
ta
rg
et
capaci
ty
s
uc
h
as
cost
of
f
uel
by
mea
ns
of
ideal
cha
nge
of
the
co
ntr
ol
var
ia
bles
sim
ul
ta
neo
usl
y
dif
f
eren
t
equ
al
it
y an
d
i
ne
qu
al
it
y co
ns
tr
ai
nts [16,
17].
Win
d
a
nd
ot
he
r
mete
orolo
gic
al
measu
reme
nt
s
with
a
m
et
mast
on
sit
e
st
arti
ng
f
rom
Augu
st
2017
t
o
Decem
ber
2017
are
ma
de
f
or
Kitka
Win
d
Par
k
in
Ka
me
nica.
This
farm
is
locat
ed
a
t
about
7km
nort
h
of
Kame
nica ci
ty
of Kos
ovo, w
hi
ch
is at
the
eas
te
rn
par
t
of
Kosovo.
Ba
sed
on
the
c
orrelat
ion
c
oeffici
ent
a
nd
c
on
cur
ren
t
meas
ur
ement
pe
rio
d
of
dataset
s,
30
–
yea
r
ti
me
series
hav
e
b
ee
n
s
yn
t
hesized
. R
egardin
g un
c
ertai
nties we
re
ta
ken
i
nto
acco
un
t i
n
a
nalysis.
It
is
note
d
that
meas
ur
eme
nt
per
i
od
co
vers
on
l
y
3.5
m
onths
a
nd
ave
ra
ge
wind
s
pee
d
is
aff
ect
e
d
by
seaso
nal v
a
riat
ion
s
. A
s
measu
reme
nt p
eri
od
i
s co
inci
ding w
i
th low
wind, s
e
aso
n
6%
i
ncr
ea
se in av
e
rag
e
win
d
sp
ee
d
is
c
omp
at
ible
with
lo
ng
-
te
r
m
data
tre
nd
s
.
Fi
rst,
we
will
inv
est
igat
e
to
wh
at
e
xte
nt
the
sit
e
rug
gedness
ind
e
x
a
nd
orogra
ph
ic
pe
rformance
in
dic
at
or
co
nce
pts
ar
e
sti
ll
sup
ported
w
he
n
us
in
g
c
onte
m
porar
y
cal
culat
ion
pr
oc
edures
a
nd
t
opog
raphic
data
.
The
st
udy
of
the
fiel
d
r
ough
ness
in
de
x
wa
s
done
us
in
g
WA
s
P
so
ft
war
e
.
Figure
1. Lo
ng
-
te
rm mo
nt
hly
wind s
pee
d
tre
nd
s
(
ref
e
re
nce
ti
me series
)
Table
1
.
C
omp
ariso
n of wi
nd
data meas
urem
ent
Lon
g
-
term
Measu
r
em
en
t Sou
rce
Co
rr
elatio
n
Co
eff
i
cien
t R
2
(Week
ly
Average)
Period
W
in
d
Speed
Ad
ju
stment
ME
RR
A
-
2
Rean
al
y
sis
Data
0
.22
1
0
1
.01
.1995
–
2
1
.1
0
.20
1
7
-
ME
RR
A
-
2
Raw R
eanaly
sis
Data
0
.69
4
0
1
.01
.1987
–
3
0
.1
1
.20
1
7
1
0
5
.99
%
NCAR
-
C
FSR R
e
an
aly
sis
Ser
ies
0
.25
8
0
1
.01
.2011
–
3
0
.1
1
.20
1
7
-
2.
METHO
D
O
F STU
D
Y
The
sit
e
co
ns
is
ts
of
a
gr
ic
ultura
l
la
nd,
short
s
hru
bs
a
nd
fore
st
areas
.
Most
of
the
w
oode
d
areas
a
re
on
the
nort
hern
hi
ll
side
of
th
e
sit
e.
Woo
ded
ar
eas
are
s
howing
simi
la
r
c
harac
te
risti
c
of
ab
ou
t
10m
hei
gh
t
an
d
cl
os
ed
forma
ti
on. Gene
ral El
ect
ric 3
.
6MW
and G
e
ne
ral El
ect
ric 3
.
2MW
t
urbines
are
stu
died f
or
t
he
sit
e.
Turbine
s c
har
a
ct
erist
ic
s ar
e
giv
en
in belo
w
ta
ble.
Table
2.
T
ech
ni
cal
ch
aracte
ris
ti
cs o
f win
d
t
urbines
take
n
in
study
Variables
Gen
eral Elect
ric
3
.
2
MW
Gen
eral Elect
ric
3
.
6
MW
Po
wer
3
2
0
0
KW
3
6
0
0
KW
Hu
b
Heigh
t
1
1
0
m
1
1
0
m
Ro
to
r
Diam
et
er
1
3
0
m
1
3
7
m
Sweep
in
g
Ar
ea
5
3
,09
3
m
²
5
8
,96
5
m
²
Tower
Tub
u
lar
Tub
u
lar
Grid co
n
n
ectio
n
5
0
/6
0
Hz
5
0
/6
0
Hz
IE
C Clas
s
II
B (
Mediu
m
T
I
)
II
IC
(L
o
w
T
I
)
Po
wer
Cu
rve A
ir
Den
sity
1
.22
5
kg
/m
3
1
.22
5
kg
/m
3
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
P
ow Elec
& Dri S
ys
t
IS
S
N: 20
88
-
8
694
Impact
o
f f
ie
ld
ro
ug
hness
and po
we
r
losse
s,
t
ur
bule
nce in
te
ns
it
y on elec
tri
ci
ty
…
(
Bedri
Drag
usha
)
1521
Figure
2. C
e
an
d
C
t
c
urves f
or
GE win
d
t
urbines, an
d p
ow
e
r
curve
b
ase
d o
n win
d
s
pee
d
Turb
ulence
int
ensity (
T
I)
is o
ne
of
the
va
ria
bles
to
est
imat
e
fati
gue
l
oads o
n
the
tu
rb
i
ne
com
pone
nts.
It
is
def
i
ned
as
the
rati
o
of
th
e
wi
nd
spe
ed
s
ta
nd
a
rd
de
viati
on
to
the
wi
nd
sp
ee
d.
As
th
e
res
ult
of
e
xp
ect
ed
wind
s
pee
d
i
nc
rease
due
to
boun
dary
la
yer
w
in
d
s
hea
r
c
ha
racteri
sti
c
of
t
he
wind,
a
pro
portio
nal
de
cre
ase
in
tur
bu
le
nce
i
ntensity
is
al
s
o
e
xp
ect
e
d
at
up
pe
r
heig
hts,
as
in
Fi
gure
3
is
sh
ow
n.
IEC
propose
a
meth
od
t
o
def
i
ne
tu
rb
i
ne
cl
ass
acco
rd
i
ng
to
tu
r
bu
le
nc
e
intensit
y
at
15m/s
wind
sp
ee
d.
Av
e
ra
ge
am
bie
nt
tur
bu
le
nce
intensit
y
at
me
t mast
locati
on
at
h
ub
heig
ht is
obtai
ned as
fo
l
lows
.
Met
Mast
Locatio
n
at
Hu
b
Heigh
t
Am
b
ien
t T
u
rbu
len
ce I
n
ten
sity
Am
b
ien
t T
I
1
.28
S
D
110m
9%
12%
Figure
3. Cha
ngin
g of wi
nd s
peed b
y
al
ti
tu
de
in
st
udy
te
r
ra
in, Kit
ka
On
s
hore
wind
farms:
Wi
nd
tur
bin
es
sit
ti
ng
over
c
omplex
te
rr
ai
ns,
Atm
os
phe
ric
sta
bili
ty
is
rar
el
y
cl
os
e
to
ne
ar
ne
utral
(
highly
c
onvecti
ve
–
un
sta
ble
durin
g
t
he
da
y
ti
me
an
d
hi
gh
l
y
sta
bl
e
no
ct
urnal
co
ndit
ion
s
with
hi
gh
she
ar
at
night
ti
me),
m
uch
higher
a
mb
ie
nt
turbulence
l
evel
[
18,
19
].
The
s
urface
flo
w
char
act
e
risti
cs are
highly
de
pe
nd
e
nt
on the
s
lop
e
of the
h
il
l
[20, 2
1].
Wak
e
e
ff
ect
s
hav
e
bee
n
cal
culat
ed
by
N
.
O
Jense
n
Wa
ke
model.
Pa
r
k
e
ff
ic
ie
ncy
of
95.
4%
is
achieve
d.
T
urb
ines
are
no
t
obser
ve
d
to
be
al
ign
ed
i
n
pre
vaili
ng
wind
di
recti
on
a
nd
i
n
non
-
prevail
in
g
wi
nd
directi
on
horiz
on
ta
l
distances
are
c
onside
red su
it
able.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
11
, N
o.
3
,
Se
ptembe
r
2020
:
15
19
–
15
26
1522
Figure
4. Plac
e
ment
of w
i
nd turbines
Figure
5. Wi
nd f
ar
m tu
r
bin
es l
ayout a
nd K
it
ka
w
in
d r
os
e
3.
RESU
LT
S
AND DI
SCUS
S
ION
S
Ver
ti
cal
e
xtra
po
la
ti
on
uncer
ta
inty
is
re
su
l
te
d
from
tw
o
di
ff
e
ren
t
s
ources.
Di
ff
e
rence
betwee
n
measu
reme
nt
he
igh
t
a
nd
hub
heig
ht
is
on
e
of
the
m
a
nd
c
onsidere
d
as
0.7
%
pe
r
10m
dif
fer
e
nce
i
n
acc
orda
nce
with
th
e
c
omplexit
y
of
the
sit
e.
T
he
oth
e
r
s
ource
of
uncerta
inty
is
t
he
dif
fe
ren
ce
bet
ween
el
evati
on
of
t
he
met
mast
an
d
the
t
urbines
,
w
hich
is
con
si
der
e
d
0.7%
per
10m
el
evati
on
dif
f
eren
ce
for
the
sit
e.
Re
su
lt
ing
wind
sp
ee
d
an
d
a
nnual
ene
rgy
producti
on
unce
rtai
nties
for
e
ach
tu
rb
i
ne
is
pr
ese
nted
bel
ow.
D
ue
to
th
e
high
el
evati
on
di
ff
e
ren
ces
betwee
n
met
mast
a
nd
the
tu
rb
i
ne
T1
G
hi
gh
e
r
AEP
unce
rtai
nt
y
is
e
xpect
ed
.
CF
D
analysis
of
t
he
sit
e
will
su
bs
ta
ntial
ly
reduce
t
hese
un
ce
rtai
nt
i
es.
The
se
ns
it
ivit
y
of
the
flo
w
fiel
d
t
o
the
t
err
ai
n
descr
i
ption
has
di
rect
c
on
se
quence
s
for
nu
merical
m
odel
li
ng
,
wh
ic
h
is
t
he
prefe
rr
e
d
t
ool
for
est
imat
i
ng
how
the terrai
n
a
ff
e
ct
s w
in
d res
our
ces at
w
i
nd
-
tu
r
bin
e
sit
es [22,
23].
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
P
ow Elec
& Dri S
ys
t
IS
S
N: 20
88
-
8
694
Impact
o
f f
ie
ld
ro
ug
hness
and po
we
r
losse
s,
t
ur
bule
nce in
te
ns
it
y on elec
tri
ci
ty
…
(
Bedri
Drag
usha
)
1523
Table
3
.
Win
d t
urbines del
ta
e
le
vation an
d de
lt
a h
ei
gh
t
w
it
h
annual e
nerg
y pro
du
ct
io
n
W
TG
Delta ele
v
atio
n
Delta heig
h
t
Res
u
lt,
win
d
sp
eed m/s
Res
u
lt,
AEP
%
T1
-
9
3
.4m
26m
6
.8
1
2
.4
T2
-
6
0
.0m
26m
4
.6
8
.5
T3
-
2
0
.2m
26m
2
.3
4
.1
T4
-
4
0
.1m
26m
3
.3
5
.8
T5
-
3
0
.0m
26m
2
.8
4
.9
T6
-
0
.0m
26m
1
.8
3
.2
T7
-
2
8
.2m
26m
2
.7
4
.9
T8
1
0
.0m
26m
1
.9
3
.5
T9
20m
26m
2
.3
4
.0
T10
-
20m
26m
2
.3
4
.2
Total
5
.6
Figure
6. A
nnua
l ener
gy
producti
on from
wind tu
rb
i
nes
in
Kitka,
base
d o
n delt
a ele
vatio
n of t
heir pla
ce
ment
Figure
7.
Com
par
is
on b
et
wee
n delt
a ele
vatio
n
a
nd d
el
ta
hei
gh
t
of
wind tu
r
bin
es
and
wind
sp
ee
d o
n
the
m
The
cha
racteri
sti
cs
of
t
he
t
urbi
nes
obta
ine
d
i
n
t
he
stu
dy,
to
gethe
r
with
the
te
st
c
ondi
ti
on
s
un
der
wh
ic
h
the
test
char
act
e
risti
cs f
or
the
res
pecti
ve powe
r
re
su
l
t, are
giv
e
n bel
ow.
Table
4: Tec
hnic
al
ch
aracte
ris
ti
cs o
f
t
wo w
i
nd tu
r
bin
e
mode
ls
Turb
in
e typ
e
7
x GE
3
.6
-
137
3
x GE
3
.2
-
130
Hu
b
heig
h
t
1
1
0
m
Rated
po
wer
3
6
0
0
k
W
& 3
2
0
0
k
W
Nu
m
b
er
o
f
turb
in
es
10
Ins
talled
capacity
3
4
.8 MW
Table
5
.
A
nnua
l ener
gy
producti
on from
wind f
a
rm
Exp
ected lifeti
m
e
o
f
W
TG
2
0
years
Gros
s an
n
u
al ener
g
y
p
rod
u
ctio
n
1
2
2
,883 MWh
W
ak
e ef
fe
cts
9
5
.4%
Av
ailab
ility
9
6
.3%
Turb
in
e Per
form
an
ce
9
9
.7%
Electr
i
cal los
ses
9
8
.0%
Env
iron
m
en
tal los
ses
9
8
.5%
Cu
rtailm
en
t los
ses
1
0
0
.0%
Total Los
ses
8
8
.4%
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
11
, N
o.
3
,
Se
ptembe
r
2020
:
15
19
–
15
26
1524
Tables
6
an
d
7,
sho
ws
t
he
resu
lt
s
of
cal
c
ulati
on
s
of
th
e
mea
n
value
of
t
urb
ulence
intensit
y
a
t
diff
e
re
nt
meas
ur
i
ng
heig
hts
durin
g
the
on
e
-
yea
r
meas
ur
ement
per
i
od.
On
l
y
te
n
-
mi
nu
te
i
nterval
s
wer
e
consi
der
e
d w
he
re th
e
w
i
nd spee
d was
higher
than 3
m/s.
Table
6
.
A
ver
a
ge value
of
wind tu
rbulence
intensit
y i
n K
oz
nica wi
nd p
a
r
k, f
or d
i
ff
e
ren
t
measu
rin
g heig
hts,
durin
g
t
he
mea
su
re
ment
pe
rio
d durin
g o
ne y
ear (w
≥
3
m
/ s
)
Heig
h
t of m
e
asu
re
m
en
t
84m
60m
40m
TI
(
w=3
m
/s)
(
%
)
1
0
.5
1
0
.9
1
1
.4
Table
7
.
A
ver
a
ge value
of
wind tu
rbulence
intensit
y i
n K
oz
nica wi
nd p
a
r
k, f
or d
i
ff
e
ren
t
measu
rin
g heig
hts,
durin
g
t
he
mea
su
re
ment
pe
rio
d durin
g o
ne y
ear (w
≥
15 m
/
s)
Heig
h
t of m
e
asu
re
m
en
t
84m
60m
40m
TI
(
w=1
5
m
/s)
(%)
1
3
.79
1
4
.3
1
4
.34
It
is
e
xtremel
y
val
uab
le
–
an
d
s
om
et
imes
r
equ
i
red
f
or
ba
nk
a
ble
est
imat
es
–
to
instal
l
two
or
more
masts
at
the
w
ind
fa
rm
sit
e;
cro
ss
-
pr
e
dicti
on
bet
ween
s
uc
h
masts
will
pro
vid
e
asse
ssm
ents
of
the
acc
ur
ac
y
and
uncertai
nt
y
of
the
flo
w
modell
ing
over
the
sit
e.
Tw
o
or
more
masts
are
al
so
requir
ed
in
c
omplex
te
rr
ai
n,
wh
e
re
rug
gedness
i
nd
e
x
(R
IX)
a
nd
∆R
I
X
a
nalyses
ar
e
necessa
ry,
a
s
in
fi
gures
8
an
d
9,
wh
e
r
e
RIX
char
act
e
risti
c is a
very im
port
ant point
[
24, 2
5].
Figure
8. Re
fere
nce site
a
nd
WT
R
IX, a
nd
diff
e
r
e
nce
between t
hem
Figure
9. Com
par
is
on of
ref
e
ren
ce
sit
e r
oughne
ss
base
d on wind t
urbine
num
ber an
d bas
ed on t
heir
elevat
ion
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
P
ow Elec
& Dri S
ys
t
IS
S
N: 20
88
-
8
694
Impact
o
f f
ie
ld
ro
ug
hness
and po
we
r
losse
s,
t
ur
bule
nce in
te
ns
it
y on elec
tri
ci
ty
…
(
Bedri
Drag
usha
)
1525
As
ex
pected
i
n
t
his
ca
se,
f
r
om
the
fi
gures
prese
nted
in
Figure
9,
t
he
rou
ghness
of
t
he
te
rr
ai
n
is
un
c
ha
ng
e
d
a
nd
le
ss
t
han
t
hat
of
the
wind
t
urbines
,
w
hile
th
e
dR
IX
in
t
he
wind
t
urbines
i
ncr
ease
s
t
his
va
lue
because
of
a
dd
it
ion
al
turb
ulen
ce creat
ed
in
t
he
m.
4.
CONCL
US
I
O
NS
Kitka
has
a
m
od
e
rate
wind
r
eso
ur
ce
f
or
wind
power
dev
e
lop
me
nt.
Wi
nd
be
ha
vior
at
th
e
met
t
ow
e
r
sit
es
dem
onstr
at
es
low
ext
re
me
wi
nd
pr
ob
a
bili
ty
but
m
od
eratel
y
high
-
to
-
hi
gh
tur
bule
nc
e
due
t
o
heavy
brus
h
to
the
north
.
Fr
om
the
ab
ov
e
analysis
it
can
be
cl
early
se
en
how
t
he
rel
at
ion
s
hip
betw
een
RI
X
of
sit
e
an
d
RIX
of
wind
tur
bin
es
.
The
r
oughne
ss
i
nd
e
x
for
the
case
of
the
gr
ound
su
r
face
w
her
e
the
wind
t
urbi
nes
a
r
e
locat
ed
is
co
nst
ant
no
matt
er
how
the
al
ti
tude
of
the
te
rr
ai
n
changes
wh
e
r
e
they
a
re
loca
te
d
within
t
he
w
in
d
park.
As
ΔR
IX
values
betwee
n
met
mast
an
d
tur
bin
e
l
ocati
on
s
a
re
in
the
range
of
2.2
%
and
7.8%,
simi
la
rit
y
pr
i
nciple
ca
n
be
sai
d
to
be
m
os
tl
y
achie
ved.
Since
t
he
de
gree
of
rou
ghne
ss
of
the
tu
r
bin
es
de
pe
nd
i
ng
on
the
te
rr
ai
n
wh
e
re
they
a
re
t
o
be
l
ocated
i
s
withi
n
t
he
acce
pte
d
ra
nge
t
he
n
it
f
ollows
t
hat
si
mil
ar
tu
r
bin
es
can
be
instal
le
d
in
Kitke
without
pr
ej
ud
ic
e t
o
th
e
qual
it
y
of ene
r
gy
that ca
n be
produ
ce
d.
REFERE
NCE
S
[1]
Ahmed
Al
Amer
i,
Aouch
enni
Ou
nissa,
Cristi
an
N
ic
hita
Aouz
el
l
ag
Djam
a
l
“Power
Loss
Analy
sis
for
Wi
nd
Pow
er
Grid
Inte
gr
at
i
on
Based
on
Weibu
ll
Distribu
ti
on
”,
Ene
rgies
,
vol
.
10
,
no
.
4
,
pp
.
463
-
479,
2017
.
[2]
M.
Pus
hpava
ll
i
,
N.
M
Jothi
Sw
ar
oopan,
“Per
form
anc
e
analysis
of
hybrid
photovolt
ai
c
/wind
en
erg
y
sys
te
m
using
KY
boost
conve
rte
r”
,
Int
ernati
o
nal
Journal
of
P
ower
El
e
ct
ronic
s
and
Dr
iv
e
Syst
em
(IJ
PE
DS)
,
v
ol.
10,
no
.
1.
pp
.
433
-
443,
201
9
.
[3]
Akhile
sh
Praka
s
h
Gupta,
Stude
nt
Memb
er,
A
.
Mohapa
tr
a,
an
d
S.
N.
Singh
,
“Appa
ren
t
Po
wer
Loss
Base
d
Equi
valent
Mod
el
of
Wi
nd
Far
m
Co
ll
e
ct
or
Sys
te
m
”,
Proc
ee
d
in
gs
of
th
e
Na
ti
o
nal
Powe
r
System
s
Confe
renc
e
(NPSC)
,
NI
T
T
ir
uchi
rap
p
al
l
i, Ind
ia
,
Dec
em
be
r
14
-
16,
2018
.
[4]
Z.
Chen
,
“Issue
s
of
Connecting
W
ind
Farms
in
to
Pow
er
Sys
tems”,
IE
EE
/
PE
S
Tr
ansm
ission
and
Distribut
ion
Confe
renc
e
&
E
xhi
bition:
Asia
a
nd
Pacific
Dal
ia
n
,
Chin
a
,
2005.
[5]
Benoua
z
Idr
iss
Yass
ine
,
Al
la
ou
a
Bou
me
di
ene,
“Re
newa
b
le
en
e
rgie
s
eva
lu
ation
and
l
inki
ng
to
smar
t
g
rid”,
Inte
rna
ti
ona
l
Jou
rna
l
of
Pow
er
E
l
ec
tron
ic
s
and
Dr
ive
Sys
tems (IJP
EDS),
vol
.
11
,
n
o.
1
,
pp
.
107
-
11
8,
2020
.
[6]
Xiawe
i
Wu,
Weiha
o
Hu
,
Qi
Hu
ang,
Co
ng
Ch
en
,
Zhe
Chen
,
Fre
de
Bl
aa
b
je
rg
,
“
Optim
ized
Place
me
nt
of
Ons
hor
e
Wi
nd
Far
ms Co
nsideri
ng
Topog
rap
hy”, E
n
erg
i
es,
Vol
.
12
,
pp
.
29
44
,
2019
.
[7]
Patri
ck
Mil
an,
Matt
hia
s
Wa
¨ch
t
er,
and
Joac
hi
m
Peinke
,
“T
urbu
l
ent
Char
ac
t
er
of
Wi
nd
En
erg
y”
,
Phy
sica
l
Revi
ew
Lett
ers
,
vo
l
.
110,
2013.
[8]
Kristine
Mikk
els
en,
“E
ffe
ct
of
f
ree
str
eam
tu
rbu
le
nc
e
on
wind
t
urbine
p
erf
orm
a
nce
”
,
Norw
egia
n
Univer
sity
of
Scie
nc
e and
T
echnology,
2013
.
[9]
Bukurij
e
Hoxha
,
Bedri
Dr
agusha
,
“
Wi
nd
Energy
Predic
t
ion
in
Kos
ovo
by
WAsP
11
softw
are
”
,
Przegląd
El
e
kt
r
otechnic
zn
y
,
vo
l. 1, no. 3, p
p.
127
-
130
,
202
0.
[10]
Sabrij
e
OS
MA
N
AJ
,
Bukurie
HO
XH
A,
Rexhe
p
S
EL
IMA
J,
“An
e
xper
imental
study
of
W
ind
D
at
a
of
a
Wi
nd
Farm
in
Kos
ovo”,
Przegląd El
e
kt
rot
echnic
zny
Vol
.
1
,
No,
7,
pp.
23
-
27,
2013.
[11]
Mohame
d
Nado
ur,
Ahmed
Essadki,
Tamou
N
a
ss
er,
Moham
m
ed
Fdaili,
“Rob
ust
coor
dinated
cont
rol
using
bac
kstepp
ing
of
flywhe
e
l
en
er
gy
storage
sys
te
m
and
DF
IG
for
power
smo
othi
ng
in
wind
power
pl
ant
s”
,
Inte
rnational
Jo
urnal
of Powe
r
El
e
ct
ronics
and
Dr
iv
e
Syst
em
(I
J
PE
DS)
,
vo
l. 10,
no.
2
.
pp
.
1110
-
1122,
2019
.
[12]
I.
Van
d
er
Hov
en,
“Power
Spe
ct
rum
of
Horiz
o
nta
l
Wi
nd
Spe
e
d
in
Th
e
Frequ
enc
yra
ng
e
from
0.
0007
To
900
Cycl
es
Per
Hour
”
Journal
of
Me
t
eorology
,
vol
.
1
4
160
-
164,
1957
[13]
Firas
A.
Hadi
,
Sama
h
Shyaa
Oudah,
Raf
a
A
.
Al
-
Bal
dawi
,
“Pr
e
-
fea
sib
il
i
ty
Stu
dy
of
Hypotheti
ca
l
W
ind
Ene
rg
y
Projec
t
Us
ing
S
im
ulated
and
M
ea
sured
Dat
a”
,
2nd
Int
ernati
on
al
Conf
ere
nc
e
f
or
Engi
ne
ering,
Technol
og
y
an
d
Sci
en
ce of
Al
-
Ki
t
ab
Unive
rs
ity
,
2
018
.
[14]
Bukurij
e
Hoxh
a,
Rexh
ep
Sel
i
ma
j
,
Sabrije
O
smana
j
,
“An
E
xper
im
ent
a
l
Stu
dy
of
We
ibu
ll
and
Rayleigh
Distribut
ion
Functi
ons
of
W
ind
S
pee
ds
in
Kos
ovo”,
TEL
KOMNIK
A
(Tel
ec
omm
uni
cat
ion
,
Computi
ng,
E
lectronic
s
and
Control),
vo
l.
16,
no.
5
,
pp
.
24
51
-
2457,
2018
.
[15]
Rat
hma
nn
,
O.,
N.G.
Mortense
n
,
L.
La
ndb
erg
a
nd
A.
Bow
en,
“
As
sess
ing
the
a
cc
ura
cy
of
WAsP
in
non
-
si
mpl
e
te
rra
in”,
Procee
dings
of
th
e
18th
Bri
ti
sh Wi
nd
En
ergy
Associat
ion
Conf
ere
nc
e,
Wi
nd
En
ergy
Conv
ersion
,
pp
.
413
-
418,
1996
.
[16]
P.
Naga
le
shmi
,
“Solut
ion
for
op
ti
mal
powe
r
flo
w
proble
m
in
w
ind
en
erg
y
sys
tem
using
hybrid
mul
ti
obj
ec
t
iv
e
art
if
ic
i
al
physi
c
al
opt
im
i
za
t
ion
al
gori
thm
”
,
In
te
rnational
Jou
rnal
of
Pow
er
El
e
ct
roni
cs
and
Dr
iv
e
Syst
em
(IJ
PE
DS
)
,
vol
.
1
0,
no
.
1
,
pp
.
486
-
503,
2019
.
[17]
We
i
Ti
an
,
Ah
m
et
Ozba
y
,
W
ei
Yuan,
Parth
a
Sa
rak
ar,
Hui
Hu,
“
An
Expe
rimental
Study
on
the
P
erf
orma
n
ce
s
o
f
Wi
nd
Turbi
n
es
over
Comp
le
x
Te
rra
in”,
51st
A
IAA
Ae
ros
pace
Sci
en
ce
s
M
eetin
g
including
the
New
Hor
izons
Forum and
A
erospace
E
xposit
ion
,
2013
.
[18]
Ahmet
Ozba
y
,
We
i
T
ia
n
,
Hui
Hu,
“
An
Exp
eri
m
ent
a
l
Study
on
th
e
Eff
e
ct
s
of
W
ake
Int
er
fer
ence
on
the
Perform
ance
of
Wi
nd
Turbi
n
es
over
Fla
t
and
C
ompl
ex
T
err
ai
ns
”,
Iowa
St
at
e
U
nive
rsity
,
Educ
a
ti
onal
Resour
ce
s
on
Wi
nd
En
erg
y
,
2013
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
11
, N
o.
3
,
Se
ptembe
r
2020
:
15
19
–
15
26
1526
[19]
We
i
T
ia
na
,
Ahm
et
O
zba
y,
Hui
H
ub,
“T
er
rai
n
ef
fe
ct
s
on
ch
a
ra
ct
er
i
stic
s
of
surfac
e
wind
and
wind
t
urbine
wake
s”
,
Proce
dia
Engi
n
e
ering,
Vol
.
126
,
pp.
542
-
548
,
20
15.
[20]
Jihane
Kart
it
e
,
Mohame
d
Cher
kaoui
,
“Im
prov
e
d
ba
ckt
r
ac
king
sea
rch
opt
im
i
za
t
ion
al
gori
thm
f
or
PV
/W
ind
/FC
sys
te
m”
,
TEL
KOMNIKA
(Telec
omm
unic
ati
on,
Computin
g,
El
e
c
tronic
s
and
Con
t
rol),
Vol
.
18
,
No
.
1
,
pp
.
456
-
464
,
2020.
[21]
Julia
La
ng
e,
Ja
kob
Mann,
Jac
ob
Berg
,
Dan
Parvu,
Ry
an
Ki
lpa
trick,
Adria
n
Costac
h
e,
Juba
yer
Chowdhury,
Kamr
an
Siddiqu
i,
Horia
Hang
an
“
For
wind
turb
in
es
in
com
pl
ex
t
e
rra
in,
the
d
evi
l
i
s
in
th
e
de
t
ail”
,
Env
ironmenta
l
Re
search
Le
tters
,
Vol
.
12
,
2017
.
[22]
Le
andr
o
Jos
e
Leme
s
Stiv
al,
Al
ex
andr
e
Kolodynskie
Guetter,
Fer
nando
Ol
ive
ir
a
d
e
Andrad
e,
“T
he
im
p
act
of
wind
shea
r
and
turbu
l
enc
e
intensity
on
wind
turb
ine po
wer
per
for
ma
nc
e
”,
Espaço Ene
rg
ia
,
2017
.
[23]
Niel
s
G.
Mort
en
sen,
“
Plann
ing
a
nd
Deve
lopment
of
W
ind
Farms:
W
ind
Resourc
e
As
sess
me
nt
and
Siti
ng
”,
DTU
Wi
nd
Ene
rgy
,
2
013
.
[24]
Hima
ni
,
Navn
e
et
Sh
arm
a
,
“H
ard
ware
-
in
-
the
-
l
oop
simul
at
or
of
wind
turb
ine
em
u
la
tor
using
la
bv
ie
w”
,
Inte
rnational
Jo
urnal
of Powe
r
El
e
ct
ronics
and
Dr
iv
e
Syst
em
(I
J
PE
DS),
vo
l. 10,
no.
2
.
pp
.
971
-
9
86,
2019
.
[25]
Wa
ri
t
W
era
pun
,
Yuttha
na
Ti
r
a
wanic
haku
l,
Jo
mpob
Waewsak,
“Win
d
Shear
Coeff
icient
s
and
the
ir
Eff
e
ct
on
Ene
rgy
Produc
tion”,
Ene
rgy
Pro
ce
dia
,
Vol
.
138
,
pp.
1061
-
1066
,
2017.
Evaluation Warning : The document was created with Spire.PDF for Python.