TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.12, No.5, May 2014, pp
. 3625 ~ 36
3
3
DOI: http://dx.doi.org/10.11591/telkomni
ka.v12i5.5096
3625
Re
cei
v
ed
No
vem
ber 1
0
, 2013; Re
vi
sed
De
cem
ber 1
3
,
2013; Accep
t
ed Jan
uary 2
,
2014
Economic Evaluation for Peak Shaving of Wind Power
Integrated Syst
em
Tingting Ho
u*, Suhua Lo
u, Yao
w
u
Wu, Zhilei Wang, Lin Yi
S
tate Ke
y
Lab
orator
y of Adva
nced El
ectrom
agn
etic En
g
i
ne
erin
g and T
e
ch
nol
og
y, Col
l
e
g
e
of Electrical
and
Electron
ic Engi
neer
ing, Hu
azh
ong U
n
iver
s
i
t
y
of Science a
n
d
T
e
chnol
og
y,
W
uhan 4
3
0
074
, Hubei Prov
inc
e
, Chin
a
*Corres
p
o
ndi
n
g
author, e-Ma
i
l
: holtting
@hus
t.edu.cn
A
b
st
r
a
ct
T
he variab
le a
nd no
n-dis
patc
hab
le outp
u
t of w
i
nd
farms bri
ngs gre
a
t difficulty to the qua
ntitative
analysis
of the econom
ic
operation f
o
r peak
shaving
of power system
int
egrated
with high penetration of
w
i
nd pow
er. F
o
r the r
and
o
m
nature
of w
i
nd p
o
w
e
r,
the
pap
er esta
bl
i
s
hes a
pe
ak
shavi
ng c
apac
ity
requ
ire
m
e
n
t mode
l for w
i
nd
pow
er. Based
on the p
eak
shavi
ng ca
pac
ity requir
e
me
n
t
mod
e
l for w
i
n
d
power, a
m
o
del for ec
onom
y
evaluation of
p
eak shaving
of power system wi
th a high
penetration of wind
pow
er is pro
p
o
sed. T
he
mo
del
mak
e
s it p
o
ssibl
e to q
u
a
n
titatively
ana
l
y
z
e
th
e i
n
flue
n
c
es of w
i
nd p
o
w
er
integr
ation on the
pe
ak
shav
ing
ca
pac
ity requ
ire
m
e
n
t an
d oper
atio
n ec
ono
my. T
he c
a
se studi
es w
e
re
carried
out for
a system, a
nd
the resu
lts veri
fied the
effectiveness and ac
curacy
of
the prese
n
ted mo
d
e
l.
Some c
onclusions
are s
u
mmari
z
ed about the
econom
y evaluation m
e
thod of
peak s
h
aving of system
s
integr
ated w
i
th signific
ant w
i
nd
.
Ke
y
w
ords
:
econ
o
m
y eval
u
a
tion of pe
ak shavi
ng, w
i
nd fa
rm, peak sh
avi
ng cap
a
city re
quir
e
ment sce
nari
o
,
backw
ard scen
a
rio re
ductio
n
Copy
right
©
2014 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
The diffe
ren
c
e bet
ween
pe
ak l
oad
and
valley load
of power grid
is gr
o
w
ing
year
by year
with the
ch
an
ge of the
st
ru
cture
of el
ectricity
dema
nd,
and th
e p
e
a
k
shaving
of
power
syste
m
is
increa
singly
d
i
fficult. In pa
rticula
r
, with
th
e qui
ck d
e
vel
opment
of re
newable
ele
c
tricity ge
ne
rati
on
su
ch as wind
power
and
so
lar po
wer
[1], esp
e
ci
a
lly the
progress of
wind
po
we
r
a
r
oun
d th
e
world
is
con
s
i
s
tentl
y
impre
s
sive
be
cau
s
e
of
its l
o
w cost
, mature
te
chnolo
g
y, ri
ch
sto
r
e
and
f
r
ee
pollution, the
large
-
scal
e integratio
n of
wind p
o
we
r beco
m
e
s
a
global tre
n
d
of wind po
wer
developm
ent
[2-3]. As
a
re
sult, the l
a
rg
e-scal
e in
te
gration of th
e
stocha
stic an
d
unp
redi
ctabl
e
power
will create
a much greater
chall
e
nge for peak
shaving
of
power system. Furthermore, the
eco
nomi
c
op
eration fo
r p
eak
shavin
g
of pow
e
r
system occu
p
i
es an i
m
po
rtant position
in
electri
c
ity p
r
o
ductio
n
, an
d i
t
s imp
a
ct
on
the e
c
on
omic ben
efits of
the
whol
e
system is far mo
re
than 1/3, th
o
ugh th
e p
e
a
k
sh
aving
peri
od a
c
cou
n
ts
f
o
r ju
st 1/3
of
the total o
peration p
e
rio
d
,
and
plays a de
ci
sive role i
n
the ope
ration
econ
om
y of powe
r
sy
ste
m
[4]. So it’s impo
rtant and
necessa
ry to
study th
e
chara
c
te
risti
c
of
pea
k sh
aving cap
a
city
requi
rem
ent of
po
wer
sy
stem
integrate
d
wi
th large
-
scal
e wind
po
we
r and
pr
e
s
e
n
t
the econ
o
m
y evaluatio
n model
of pea
k
shavin
g.
In Chi
na, a
ccordin
g to th
e
national
win
d
po
wer pla
nni
ng, seven
wi
nd p
o
wer
ba
ses
with
an installed
wind power
c
apacity up to ten million
KW respecti
vely will be built com
p
letel
y
in
2020, an
d they are respectively located at Ha
m
i
in Xinjiang
province, Ji
uqua
n in Ga
nsu
provin
ce, He
bei province, Jian
gsu
province
, e
a
st
ern in
ner
M
ongoli
a
, and
we
stern i
n
ner
Mongoli
a
. The total installed wind p
o
wer ca
pa
city
of the seven
wind ba
se
s will be up to
5808
×1
0
4
KW in 2015, and
9017×10
4
KW in 2020. T
h
is large-scal
e and ce
ntrall
y explored wi
nd
power will g
r
eatly incre
a
se the difficulty of
peak sh
aving of power syste
m
wi
th wind po
wer
integrate
d
.
I
n
re
cent
y
e
a
r
s,
con
s
ide
r
a
b
le re
se
ar
ch
has
bee
n co
ndu
cted o
n
the integ
r
atio
n of win
d
power [5-9], and there is a gro
w
ing
co
nce
r
n for
the
pressin
g
issues fa
cing p
eak
shaving
of
power
syste
m
integrate
d
with
larg
e-scale
wind p
o
w
er [1
0-1
5
]. Referen
c
e [1
1] analyze
d
the
cha
r
a
c
teri
stic of neg
ative pea
k
shaving
of co
nventio
nal ge
ne
rators an
d p
r
op
osed a m
odel
t
o
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3625 – 36
33
3626
cal
c
ulate the
limit of capability of negative
pea
k shavin
g ba
sed on the d
e
termini
s
tic
unit
commitme
n
t and win
d
po
wer o
u
tput, ignori
ng the stocha
stic nat
ure of wind p
o
we
r. Refere
nce
[12] presente
d
a peak sha
v
ing sch
eme
of therma
l po
wer for the in
tegrated
Jiuq
uan win
d
po
wer
base, com
b
in
g its wi
nd p
o
w
er
ch
aracte
ristic.
Re
fe
re
nce [1
3] pre
s
ented a m
e
th
od to calcula
t
e
pea
k shaving
ability
of
Northeast Chi
na power grid
int
egrate
d
with l
a
rge
-
scale
wi
nd farms ba
sed
on the feature of Northea
st China p
o
wer grid,
an
d prop
osed a prin
ciple for
pea
k shavin
g
by
coo
r
din
a
ting hydro
p
o
wer and
the
r
mal power
for wi
n
d
farm
s. Refe
ren
c
e [14] an
alyzed the
pe
ak
shaving ability of Beijing-Tianjin-T
angsan power
gri
d
accordi
ng t
o
its l
oad charac
teri
stic and
power so
urce comp
ositio
n, and
p
r
ovi
ded th
e
a
p
p
r
oximate
ca
p
a
city range
of win
d
p
o
wer
integrate
d
th
at the power grid
can
ab
sorb. Re
fe
re
nce
s
[12
-
14]
all analyzed t
he pe
ak
sha
v
ing
ability
the system
can supply
based on a
specifi
c
sy
st
em. Ref
e
rence [15] eval
uat
ed the effect
of
wind p
o
wer
on the load
pea
k-to
-valle
y difference of
powe
r
sy
stem integrate
d
with large-scale
wind
po
wer a
c
cordi
ng to th
e variatio
n of
pea
k-to
-v
alle
y of net load,
based o
n
chronolo
g
ical
lo
ad
time
serie
s
a
nd chronol
ogi
cal
time
seri
e
s
of wi
n
d
p
o
w
er
outp
u
t si
mulated by Weib
ull
di
stri
bution
function
of wi
nd spee
d. In fact, the pe
ak
s
havin
g a
b
ility and op
e
r
ation
econo
my of the po
we
r
system
integ
r
ated
with
win
d
po
we
r
are
clo
s
ely relate
d to the
loa
d
ch
aracte
risti
c
, po
we
r
so
u
r
ce
comp
ositio
n and win
d
po
wer ch
aracte
ristic, only
analyzing wi
nd po
wer
cha
r
a
c
teristic or the p
e
a
k
shaving abilit
y of the power sy
stem
int
egrated with
wind power i
s
incompl
e
te. Most of current
literature an
a
l
yzed the
ch
alleng
e brou
ght by integ
r
ated wi
nd
p
o
we
r fa
cing
pea
k shaving
of
power
syste
m
macroly a
nd roug
hly, and the
r
e i
s
a serio
u
s l
a
ck of a
ge
neral
metho
d
of
quantitative a
nalysi
s
and
modelin
g of operation e
c
onomy of pe
ak shaving.
So there i
s
u
r
gent
need for in
-d
epth study o
n
effect
s of stocha
stic na
ture of
wind
power on p
e
a
k shaving a
nd
eco
nomy eva
l
uation of pe
ak shaving o
f
power
syst
em integrate
d
with high penetration
wind
power.
The pea
k
sh
aving ca
pa
city requiremen
t
is
the difference betwee
n
the maximum loa
d
and
minimum
load
du
rin
g
all the
ope
rati
on p
e
ri
od, a
s
a
re
sult,
stu
d
y on
pe
ak shaving
of po
we
r
system i
n
teg
r
ated
with la
rge-sc
ale
win
d
po
we
r n
e
e
d
s
co
ordi
nati
ng the
load
chara
c
te
risti
c
and
wind po
we
r output cha
r
a
c
teri
stic du
rin
g
the w
hole
operation pe
riod. At the s
a
me time, the
stocha
stic an
d unpredi
cta
b
le
nature of wind po
wer output ma
kes the pea
k
shavin
g cap
a
c
ity
stocha
stic
greatly, whi
c
h i
s
the
key differen
c
e
of p
e
a
k
sh
aving i
s
sue
bet
ween
traditional
po
wer
system a
nd o
ne integrated
wi
th larg
e-scale win
d
power.
For the sto
c
h
a
stic n
a
ture
of wind po
we
r, th
is pap
er
pre
s
ent
s a p
eak
shavin
g cap
a
city
requi
rem
ent
model
of po
wer sy
stem i
n
tegrate
d
wi
th large-scale
wind
po
we
r, com
b
ing th
e
load
cha
r
a
c
teri
stic, and ch
oo
se
s typical p
e
a
k
shavi
ng
ca
pacity re
quirement sce
n
a
r
ios to m
odel
its
stocha
stic n
a
t
ure. Based o
n
the pea
k
shaving capa
city requi
re
me
nt model, the
pape
r propo
se
s
a eco
nomy e
v
aluation mo
del of pea
k shaving of po
wer
system i
n
tegrate
d
hig
h
penet
ration
wind
power. Ta
kin
g
into a
c
cou
n
t
the sto
c
ha
st
ic natu
r
e
of wind po
we
r, th
e load
shed
ding fee
be
cau
s
e
of lack of po
sitive pea
k
shaving a
b
ility and p
enal
ty
fees fo
r win
d
spill
age b
e
cau
s
e
of lack of
negative p
e
a
k
shaving
abi
lity also be
co
me a p
a
rt of
t
he obj
ective f
unctio
n
of the
model. Fin
a
ll
y,
the ca
se stu
d
ies
were ca
rrie
d
out for a system
, an
d the results
verified the e
ffectiveness
and
accuracy
of
the presente
d
mod
e
l. So
me co
n
c
lu
sio
n
s a
r
e
su
m
m
ari
z
ed
abo
ut the e
c
on
o
m
y
evaluation m
e
thod of pea
k shaving of sy
st
em
s integra
t
ed with sig
n
ificant wi
nd.
This paper i
s
divided into
t
he followi
ng
sections. Section III
presents the peak
shaving
cap
a
city
req
u
irem
ent mo
del of
po
wer syste
m
inte
grated
with
high pen
etrat
i
on wind
po
wer.
Section IV formulates the
economy evaluation mo
del
of peak shaving. Se
ction V illustrates
the
methodol
ogy usin
g a test system. Sectio
n
VI draws so
me releva
nt concl
u
si
on
s.
2. Peak Shav
ing Capa
cit
y
Requireme
n
t
Model
The pea
k sha
v
ing
capa
city requi
rem
ent origin
ates in
variation
of wind p
o
wer
out
put an
d
the differe
nce bet
ween
p
eak load
an
d
valley load,
and i
s
clo
s
el
y related
to t
heir
ch
ron
o
lo
gical
correl
ation. T
o
model
the
pea
k shaving
cap
a
city
req
u
irem
ent of p
o
we
r
system
integrate
d
wit
h
large
-
scale
wind
po
we
r, the pa
pe
r
use
s
th
e p
r
i
n
cipl
e of
ch
oosi
ng typical load
curv
e for
referen
c
e, according to
statistical a
nal
ysis
on the
pea
k-to
-valle
y of net load and sce
n
a
r
io
redu
ction
p
r
oce
dure, a
n
d
obtai
ns typical
wi
nd
power outp
u
t
cu
rves an
d corre
s
po
n
d
ing
prob
ability for analy
s
is on
pea
k
shaving
of po
we
r
s
ystem integ
r
ate
d
with
high
p
enetratio
n
wi
nd
power.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Econom
ic Evaluation for P
eak Sha
v
in
g of Wi
nd Power Integrated
System
(Ting
t
ing Hou
)
3627
2.1. Classific
a
tion of Pea
k
Shav
ing Capacity
Requirement Sc
enarios
In ope
ration
simulatio
n
of
power
syste
m
int
egrated
with la
rge
scale
win
d
po
wer, t
o
absorb
re
ne
wable e
n
e
r
gy p
r
ioritly, win
d
power
i
s
con
s
ide
r
ed
a
s
a
negative lo
ad
and
obtain
s
t
he
net load. Th
e
analysi
s
o
n
pea
k shaving
cap
a
ci
ty
re
q
u
irem
ent
is condu
cted ba
sed on
the n
e
t
load of
power system i
n
teg
r
ated
wi
th la
rge-scale
win
d
po
we
r, wh
i
c
h ta
ke
s into
accou
n
t both
the
pea
k shaving
of origin
al lo
ad and va
riati
on of win
d
po
wer, b
u
t also their correlati
on. The n
e
t lo
ad
is formul
ated
as follo
ws:
ne
t
W
LL
P
(1)
Whe
r
e
W
P
and
L
are respe
c
tively wind po
wer
output a
nd load of th
e whol
e ope
ration
perio
d, and
1
,,
,
T
T
W
W
Wt
WN
PP
P
P
,
1
,,
,
T
T
tN
LL
L
L
,
P
Wt
and
L
t
a
r
e re
spe
c
tivel
y
wind po
we
r
output and lo
ad in peri
od t, and
t
=1
，
2
，
……
N
T
.
The pea
k sh
aving ca
pa
city requiremen
t
of
power system integrated with large-scal
e
wind p
o
wer i
s
the differe
n
c
e bet
wee
n
the maximum
net load and
minimum ne
t load duri
ng
all
the op
eratio
n
pe
riod, th
at
pea
k-to
-valle
y of t
he
net l
oad.
Due
to
the un
ce
rtain
natu
r
e
of wi
nd
power, the
r
e
may are
m
any scen
ario
s of wi
nd p
o
w
er output,
so the
net l
oad o
b
taine
d
by
formulatio
n (1) al
so
ha
s m
any scen
ario
s. The
pea
k
shaving
cap
a
city requireme
nt of wind
po
we
r
output scen
ario i is formula
t
ed as follo
ws:
ma
x
(
)
m
i
n
(
)
ii
i
p
vn
e
t
n
e
t
PL
L
(2)
Whe
r
e
ne
t
ii
W
LL
P
,
i
W
P
is th
e win
d
p
o
wer output
scena
rio i, a
n
d
i=1,2,……
N
s
,
i
p
v
P
a
nd
i
ne
t
L
are re
spe
c
tively peak sha
v
ing capa
city
r
equi
reme
nt and net load
of the wind p
o
we
r output
scena
rio i,
ma
x(
)
and
mi
n
(
)
are
re
spe
c
tively the max function an
d min function.
From fo
rmul
a
t
ion (2
), the
set of pea
k
sh
aving capa
cit
y
requi
rem
e
n
t
scena
rio
s
p
v
, can
be obtain
ed
from the
set
of wind p
o
w
er
output
scen
ario
s
1
,,
iN
s
WW
W
W
PP
P
, and the
probability of
every scenari
o
is
1
,,
p
ro
i
N
s
pp
p
, and
1
1
S
N
i
i
p
, which i
s
al
so t
h
e probably
of the proba
bi
lity of corre
sp
ondi
n
g
wind
power outp
u
t scena
rio.
Usi
ng well-b
e
ing an
alysi
s
[16] for refe
ren
c
e, the p
eak
shavin
g
cap
a
city re
q
u
irem
ent
scena
rio
s
ca
n be
divided i
n
to thre
e g
r
o
ups
He
althy, Margi
nal a
n
d
Risk, a
c
cordi
ng to the
effects
of wind
po
we
r integ
r
ation
on the
pea
k
shavin
g of
p
o
w
er sy
stem.
Cla
ss
He
alth
y represents
tha
t
the pe
ak sha
v
ing capa
city re
quirement
is le
ss tha
n
t
hat befo
r
e
wi
nd p
o
wer inte
gration,
L
pv
, that
is, wind p
o
wer integ
r
ation
improve
s
the pea
k sh
aving of power
system
. C
l
ass
R
i
s
k
re
pr
ese
n
t
s
that the pea
k shaving
cap
a
city req
u
ire
m
ent is g
r
eat
er than that t
he syste
m
ca
n sup
p
ly by other
gene
rato
rs,
C
pv
, that is, the system
can
not meet the
pea
k shaving
cap
a
city req
u
irem
ent of t
he
power sy
ste
m
integrate
d
with win
d
po
wer. So,
cla
s
s Marginal re
pre
s
ent
s the scena
rio
s
except
that of the
two ab
ove
cla
s
se
s,
that
i
s
, wind po
we
r integratio
n increa
se
s pea
k shavin
g pre
s
sure,
but the
C
pv
can meet the pe
ak shaving
ca
pacity req
u
ire
m
ent of the powe
r
syste
m
integrate
d
wit
h
wind po
wer.
The classifi
cation rep
r
e
s
ent
s
thre
e
different effe
ct situatio
ns of wind
po
wer
integratio
n on
peak
shavin
g, and distin
g
u
ish
e
s
three
different pea
k shaving
states cl
early.
Based
o
n
the
above
classif
i
cation
meth
o
d
, the
set
of p
eak shaving
capa
city re
quirement
scena
rio
s
p
v
ca
n be divide
d i
n
to three
su
b
s
ets,
,
H
pv
,
,
M
pv
and
,
R
pv
,
and thei
r sce
nario
s
and corre
s
po
nding p
r
ob
abi
lity are respe
c
tively formul
ated as follo
ws:
,
,
,
0,
,
1
,
2
,
,
,,
1
,
2
,
,
,,
1
,
2
,
,
hh
Hp
v
p
v
p
v
p
v
p
v
H
jj
M
pv
p
v
pv
p
v
pv
p
v
M
kk
Rp
v
p
v
p
v
p
v
p
v
R
PP
L
h
N
PP
L
C
j
N
PP
C
k
N
(3)
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3625 – 36
33
3628
1
1
1
H
M
R
N
H
h
h
N
M
j
j
N
R
k
k
pp
pp
pp
(4)
Whe
r
e
,
H
pv
,
,
M
pv
an
d
,
R
pv
rep
r
e
s
ent
the
set of
pea
k
shaving
ca
pa
city req
u
irem
ent o
f
Healthy
state
,
Margin
al sta
t
e and
Risk
state, respecti
vely;
h
p
v
P
,
j
p
v
P
and
k
p
v
P
repre
s
e
n
t sce
nario
h
,
j
an
d
k
of
,
H
pv
,
,
M
pv
an
d
,
R
pv
, respec
tively
;
N
H
,
N
M
and
N
R
are re
sp
ectiv
e
ly
the
total
numbe
r of scenari
o
s of
,
H
pv
,
,
M
pv
and
,
R
pv
.
2.2. Peak Shav
ing Capac
i
t
y
Requirem
e
nt Mod
e
l
For
comp
utational compl
e
xity, it’s impo
ssi
ble to an
a
l
yze and
eva
l
uate for eve
r
y pea
k
shavin
g ca
p
a
city requi
re
ment scena
ri
o in det
ail, so backward scena
rio red
u
ction te
chni
que
based o
n
Ka
ntorovich di
stance, KD, [1
7] is ap
p
lied
to trim do
wn
the num
ber
o
f
peak
sh
avin
g
cap
a
city re
qu
ireme
n
t scen
ario
s, obtaini
ng typi
cal scenari
o
s and correspon
ding
prob
ability
while
kee
p
ing mo
st of stochast
i
c informatio
n embed
ded
in these scenari
o
s. Backward
scena
ri
o
redu
ction
techniqu
e ba
se
d
on K
D
is an
optimal
p
r
o
c
edure, an
d el
iminates the
scena
rio
whi
c
h
has the mini
mum pro
babil
i
ty distance u
n
til the stoppi
ng crite
r
io
n h
a
s be
en met.
Based
on the
cla
ssifi
cation
method of p
eak
sh
aving
cap
a
city re
qu
ireme
n
t scen
ario a
n
d
scena
rio redu
ction techniq
ue de
sc
ribe
d above, the p
eak
shavin
g cap
a
city req
u
i
reme
nt mod
e
l is
as
follows
:
,1
,
,
,1
,
,
,1
,
,
,,
,,
,,
ll
ll
ll
lp
v
p
v
p
v
i
p
v
N
lW
lW
lW
i
l
W
N
lp
r
o
l
l
i
l
N
l
PP
P
PP
P
pp
p
p
(5)
Whe
r
e
l
=
H
,
M
,
R
,
rep
r
e
s
ents pe
ak
shaving capa
city requi
re
state He
althy, Marginal,
Risk,
respe
c
tively, and it has the same
meanin
g
in
the followin
g
s
.
lp
v
,
lW
and
lp
ro
are
respe
c
tively obje
c
tive sce
nario
sets of
pea
k shaving
cap
a
city req
u
irem
ent, win
d
po
we
r outp
u
t
curve and correspon
ding probabilities of state L.
,
l
lp
v
i
P
,
,
l
lW
i
P
and
,
l
li
p
are re
spe
c
tively peak
shavin
g ca
pa
city requi
rem
ent, wind po
wer
output
cu
rve and
corre
s
po
ndin
g
pro
bability of typ
i
ca
l
scena
rio il in obje
c
tive sce
nario
set of peak
shavin
g state l.
The pe
ak
sh
aving ca
pa
city requireme
n
t
model
takes into accou
n
t the sto
c
ha
sti
c
natu
r
e
of wind p
o
wer a
c
cording
to choo
sin
g
typical pea
k shavin
g ca
p
a
city req
u
ire
m
ent scen
ari
o
s,
based on
whi
c
h the e
c
on
o
m
y evaluati
on of peak
sha
v
ing is co
ndu
cted.
3. Economic Ev
aluation of Peak Sha
v
i
ng of Po
w
e
r Sy
stem In
tegra
t
ed
w
i
t
h
Wind
3.1. Objectiv
e Functio
n
The go
al of study on p
e
a
k
shaving
e
c
on
omy
of p
o
we
r sy
stem
integrate
d
with high
penetration wind po
wer i
s
to minimize the expecte
d cost (F
) in terms of ab
sorbi
ng win
d
po
we
r in
prio
rity and
meet corre
s
p
ondin
g
con
s
traints. Ta
kin
g
into a
c
count
the sto
c
h
a
sti
c
natu
r
e
of wind
power, the lo
ad sh
eddi
ng
fee becau
se
of lack of
p
o
sitive peak sh
aving ability and pe
nalty fees
for wind
spill
age be
cau
s
e
of lack of negative pea
k
shaving abili
ty also beco
m
e a part of the
obje
c
tive function of the model.
Base
on
the
pea
k
shavin
g
ca
pa
city re
q
u
irem
ent
m
o
del a
bove, th
e obj
ective fu
nction
of
the eco
nomi
c
evaluation of
peak
shavin
g is formul
ate
d
as follo
ws:
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Econom
ic Evaluation for P
eak Sha
v
in
g of Wi
nd Power Integrated
System
(Ting
t
ing Hou
)
3629
,
1
l
l
l
N
lN
li
MinF
p
,,
,
,
11
()
()
()
u
T
ll
l
N
N
ul
i
l
i
l
i
tu
f
tO
tC
t
(6)
Whe
r
e,
,
,
1
,
,
.
,,
,,
2
,
,
2
()
(
)
(
)
(
.
)
ll
l
l
l
uli
u
li
u
t
li
ul
i
u
li
x
f
tf
P
S
t
f
N
O
S
O
(7)
Whe
r
e
N
u
is t
he numb
e
r of
conventio
nal
units,
,,
()
l
ul
i
f
t
is fuel cost of unit u
in period t u
nder the
typical wi
nd
p
o
we
r o
u
tput
scena
rio
,
l
lW
i
P
,
including ope
rati
onal co
st
1,
,
.
,
,
()
ll
ul
i
u
tl
i
fP
, s
t
art-up cos
t
,,
()
l
ul
i
St
and emi
ssi
o
n
co
st of
SO
2
and
NO
x
2,
,
2
(.
)
l
ul
i
x
f
NO
S
O
,
.,
,
l
u
tli
P
is po
we
r outp
u
t of unit
i
in
perio
d
t
u
n
d
e
r the
typical wi
nd
power o
u
tput
scenari
o
,
l
lW
i
P
,
,
()
l
li
Ot
and
,
()
l
li
Ct
are l
o
a
d
she
ddin
g
fee
becau
se
of la
ck of
po
sitive pea
k shavin
g ability a
n
d
penalty fee
s
for
win
d
spilla
ge
because of lack
of
negative peak
shavi
ng ability in period
t under
the typica
l
wind power out
put
scena
rio
,
lW
i
P
, respec
tively.
,
()
l
li
Ot
and
,
()
l
li
Ct
are form
u
l
ated as follo
ws:
,.
,
()
()
ll
li
N
S
W
l
i
Ot
E
t
(8)
,.
,
()
()
ll
li
N
A
W
l
i
Ct
E
t
(9)
Whe
r
e
γ
is the co
st pe
r l
oad shed
din
g
,
ρ
is the cost per wi
nd
power energy spillage, and
.,
()
l
NS
W
l
i
E
t
,
.,
()
l
NA
W
l
i
E
t
are
ele
c
tri
c
ity not su
pplie
d, wind
po
we
r en
ergy
spill
age in
pe
riod
t unde
r
the typical wi
nd po
wer o
u
tput scena
rio
,
l
lW
i
P
, respec
tively.
3.2. Cons
trai
nts
For eve
r
y typical
win
d
p
o
w
er outp
u
t scen
ario
,
l
lW
i
P
,
the
f
o
llowin
g
con
s
traints mu
st be
sat
i
sf
ie
d.
a) System op
eration
con
s
traints
1) Power bal
ance co
nstrai
nt
.,
,
,
,
.
1
0,
(
1
,
2
,
,
)
ll
U
ut
l
i
l
W
i
t
t
L
t
T
u
PP
L
P
t
N
(10)
Whe
r
e
,,
lW
i
t
P
is win
d
power outp
u
t in period t of
the typical wind p
o
wer o
u
tput scena
ri
o
,
lW
i
P
,
P
L.t
is net loss in
period
t
.
2) Spinni
ng rese
rve co
nst
r
aint
.m
a
x
,
,
,
.
1
l
U
uu
t
l
W
i
t
L
t
t
t
u
Px
P
P
L
R
(11)
Whe
r
e
P
u
.max
is rate
d
capa
city of unit
i
,
R
t
is
spi
nning
re
serve i
n
pe
riod t,
x
u
,
t
is
0/
1 variabl
e whi
c
h
is equ
al to 1 if unit
i
is online in scena
rio
k
and p
e
rio
d
t
.
b) Co
nventio
nal units o
peration co
nstrai
nts
1) Generation limits
.m
i
n
.
,
,
.
m
a
x
,(
1
,
2
,
,
)
uu
t
l
i
u
T
PP
P
t
N
(12)
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3625 – 36
33
3630
Whe
r
e
P
u
.min
is minimum
capa
city of unit
i
.
2) Ramp rate
limits
..
1
uu
t
u
t
u
DP
P
U
(13)
Whe
r
e
U
u
and
D
u
are, resp
ectively, up and do
wn ra
m
p
rate limit of unit
u
.
3) Minimu
m o
n
and off time con
s
traint
s
..
.
m
i
n
..
.
m
i
n
()
()
on
u
o
n
u
of
f
u
of
f
u
tt
t
tt
t
(14)
c)
Wind farm operation con
s
traint
s
1) Generation limits
0
to
ta
l
Wt
W
PP
(15)
Whe
r
e
total
W
P
is rated ca
pa
city of wind farm.
2) Win
d
po
we
r elect
r
icity co
nstrai
nts
ex
p
WE
N
A
W
E
EE
(16)
Whe
r
e
E
W
an
d
E
ENAW
are wind energy absorbed
and
wind
energy spillage, respectively.
E
exp
is
expecte
d ele
c
tri
c
ity of wind farm.
4.
Case Stud
y
To analy
z
e th
e impa
ct of large
-
scal
e win
d
pow
er o
n
p
eak
shavin
g cap
a
city re
qu
ireme
n
t
and econo
m
y
, the model is tested o
v
er a 24-h
hori
z
on o
n
a real sy
ste
m
with its load
cha
r
a
c
teri
stics and
gen
erators
of 201
5. The dat
a
for gen
erato
r
s an
d load
a
r
e, re
sp
ectiv
e
ly,
listed as T
abl
e 1, Table 2 and Tabl
e 3. The ma
ximu
m load is 870
00MW, an
d the rated
cap
a
city
of the wind fa
rm integrated
is assign
ed a
s
174
00M
W, whi
c
h is 2
0
% of the maximum load.
Table 1. Co
n
v
entional Ge
nerato
r
’
s
Dat
a
Pmax
/MW
Pmin
/MW
Number
/
台
Coal consumption
at rated ou
put
/g/KWh
SO2 emission
/g/Kg.Tce
1000
500
9
280
1.6
600 300
90
300
1.6
360 216
29
320
3.2
300 180
123
320
3.2
200 140
48
340
3.2
135 94.5
43
360
16
100 70
12
360
16
60 42
3
380
16
25 17.5
5
450
16
Table 2. Co
n
v
entional Ge
nerato
r
’
s
Ch
a
r
acte
ri
stic of Coal
Con
s
um
ption
Unit
Power Output Ra
te
[0.5,0.6) [0.6,0.7) [0.7
,0.8) [0.8,0.9) [0.9,1)
1000
1.0556
1.037
1.0185
1.0074
1
600
1.063
1.046
1.029
1.012
1.006
360
/
1.026
1.015
1.003
1.002
300
/
1.026
1.015
1.003
1.002
200
/
/
1.024
1.013
1.006
135
/
/
1.044
1.032
1.016
100
/
/
1.044
1.032
1.016
60 /
/
1.088
1.044
1.02
25 /
/
1.088
1.044
1.02
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Econom
ic Evaluation for P
eak Sha
v
in
g of Wi
nd Power Integrated
System
(Ting
t
ing Hou
)
3631
Table 3. Loa
d
Data
t/h Load/MW
t/h Load/MW
t/h
Load/MW
1 0.758
9 0.852
17
0.946
2
0.743
10 0.903
18
0.897
3
0.734
11 1 19
0.912
4
0.727
12 0.926
20
0.986
5
0.725
13 0.928
21
0.958
6
0.741
14 0.949
22
0.899
7
0.74
15 0.963
23
0.839
8
0.779
16 0.952
24
0.787
4.1. Analy
s
is
on Peak Shav
ing Economy
Based
on typical wi
nd po
wer outp
u
t cu
rves of
the sele
cted sce
n
a
r
io
s, the com
p
a
r
ison of
eco
nomi
c
ev
aluation
re
sul
t
s of pea
k sh
aving bet
wee
n
system
s wi
th wind p
o
we
r integrated a
n
d
not is listed in
Table 4.
From Ta
ble 4
it can be co
nclu
ded that
coal
con
s
um
ption, SO2 e
m
issi
on of the system
both de
cre
a
se after wind
power integ
r
a
t
ion. Howe
ve
r, the coal co
nsum
ption pe
r thermal p
o
w
er
gene
ration
in
cre
a
ses for
that the va
ri
ation of
win
d
po
we
r o
u
tput in
cre
a
se
s p
e
a
k
shav
ing
pre
s
sure of
convention
a
l u
n
its. Thi
s
sho
w
s th
at
wi
nd
power i
n
tegra
t
ed save
s
op
eration
a
l
co
st of
conve
n
tional
units, at the same time, increa
se
s pea
k
shavin
g co
st for its variatio
n. However, the
operational
cost save
d is
more tha
n
pe
ak shav
ing
co
st increa
sed
by wind po
we
r integrated.
Table 4. Indices of Eco
nom
ic Ope
r
ation f
o
r Pea
k
sh
aving
indices
no w
i
nd
po
w
e
r
in
tegrated
w
i
th w
i
nd po
w
e
r
integrated
load shedding/MW
/
0
electricity
not su
pplied/GWh
/
0
w
i
nd e
nerg
y
spillage/GWh
/
0.04
w
i
nd p
o
w
er
abso
r
ption/%
/
99.97%
thermal po
w
e
r g
eneration/
GWh
1796
1673
coal consumptio
n/104 t
56.10
52.29
thermal coal consumption/g/kWh
312.36
312.59
sy
stem coal cons
umption/g/kWh
312.36
291.13
SO2 emission/104 t
0.0172
0.0157
thermal SO2
emission/g/kWh
0.6530
0.6460
Peak shaving depth/%
36.1%
41.4%
4.2 Anal
y
s
is
on the ability
to
absorb w
i
nd po
w
e
r
To analyze t
he ability to
absorb
wind
power
of the system, the
system
with
different
wind
power
p
enetratio
n
(proportio
n
of th
e maxi
mum l
oad) i
n
teg
r
at
ed is
simul
a
ted ba
se
d mo
dels
pre
s
ente
d
ab
ove.
Table 5. Re
sults of Econo
mic Op
eratio
n fo
r Peak Sh
aving with Dif
f
erent Wi
nd Powe
r
Penetratio
n
w
i
nd
penetration
thermal po
w
e
r
generation
/GWh
coal
consumption
/104 t
peak shaving
depth/%
thermal coal
consumption
/g/KWh
w
i
nd p
o
w
er
absorption
/%
10%
1734
54.25
38.2%
312.80
100.0%
20%
1673
52.29
41.4%
312.59
99.4%
30%
1616
50.72
42.8%
313.82
96.7%
40%
1562
49.07
44.4%
314.18
94.5%
50%
1506
47.37
45.4%
314.56
93.6%
60%
1462
46.49
46.0%
317.89
89.7%
70%
1446
46.09
46.2%
318.73
80.6%
80%
1441
46.00
46.2%
319.15
71.5%
Table 5
p
r
e
s
ents re
sults of
economi
c
operati
on fo
r pea
k
sh
avin
g with
differe
nt win
d
power pe
netration. It can be se
en that thermal
po
we
r gene
ration
and coal co
n
s
umptio
n of the
system d
e
crease as
win
d
power p
e
n
e
tration in
cr
e
a
se
s,
whi
c
h
sho
w
s t
hat
t
he sy
st
e
m
c
an
absorb m
o
re
wind po
we
r based on
wind po
wer p
e
netration 2
0
%
, and the added
wind p
o
we
r
integrate
d
sa
ves the
r
mal
power g
ene
ration but al
so
coal
co
nsum
ption,
but the
amount
of them
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3625 – 36
33
3632
repla
c
e
d
by the sam
e
add
ed wind po
wer de
cre
a
se
as win
d
powe
r
penetration
increa
se
s. At
the
same
time, the coal
co
nsumption
per
thermal
po
wer g
ene
ration
increa
se
s from wi
nd p
o
w
er
penetration 3
0
%, and the
rea
s
on i
s
tha
t
the peak
sh
aving pressu
re of thermal
units in
crea
ses
whe
n
win
d
po
wer p
enet
rati
on increa
se
s,
and their po
wer o
u
tput rat
e
decrea
s
e
s
.
4.3. Accur
a
c
y
Verification of the Mo
d
e
l
To verify th
e
accu
ra
cy of
propo
sed
m
odel
s,
all th
e
win
d
p
o
wer
output
scena
rios a
r
e
evaluated by
enume
r
ation
method, and
compa
r
i
s
on
of
result
s of the two different method
s is
listed in Tabl
e 6.
From t
he
co
mpari
s
o
n
of
result
s in
Tabl
e 6, it
can
be
se
en th
at th
e calculation
errors of
operational i
n
dice
s of th
ermal po
we
r a
r
e sm
all,
and
all are smalle
r than
1%, the large
s
t one
is
pea
k shavin
g depth
wit
h
an e
r
ror
0.48%, in
a
ddition, thermal po
we
r
gene
ration,
coal
con
s
um
ption,
co
al
con
s
um
ption p
e
r th
ermal po
we
r g
e
neratio
n,
SO
2
e
m
iss
i
on
and
SO
2
emi
s
si
on
per the
r
mal
power g
ene
ration are, re
spe
c
tive
ly, 0%, -0.19%, -0.19%, -0.34
%
and -0.34
%
.
Comp
ared with operatio
n
a
l indices of
thermal po
wer, the cal
c
ulatio
n errors of operatio
nal
indices
of wi
n
d
po
we
r a
r
e
greate
r
. Th
ou
gh the
cal
c
ul
ation e
rro
r of
wind
ene
rgy i
s
slightly gre
a
t,
it’s sm
all a
s
compa
r
ed
with
the total wi
n
d
ene
rg
y the
system
ab
sorbs, an
d the
calcul
ation e
r
ror
of the total wind ene
rgy the
system ab
so
rbs i
s
0.01%.
Table 6. Co
m
pari
s
on of Re
sults of the T
w
o Different Method
s
indices
the proposed m
o
del
enumeration met
hod
load shedding/MW
0
0
electricity
not su
pplied/GWh
0
0
w
i
nd e
nerg
y
spillage/GWh
0.04
0.02
w
i
nd p
o
w
er
abso
r
ption/%
99.97%
99.98%
thermal po
w
e
r g
eneration/
GWh
1673
1673
coal consumptio
n/104 t
52.29
52.39
thermal coal consumption/g/kWh
312.59
313.19
sy
stem coal cons
umption/g/kWh
291.13
291.69
SO2 emission/104 t
0.1081
0.1084
thermal
SO2 emission/g/kWh
0.6460
0.6482
Peak shaving depth/%
41.4%
41.2%
In a word, ba
sed o
n
the th
ree
sele
cted
wind
p
o
wer o
u
tput cu
rves
usin
g the pre
s
ente
d
model,
th
e calcul
ation error
i
s
sm
aller than
1%
fr
om
cal
c
ul
ation
e
rro
rs of
ope
rational i
ndi
ce
s of
thermal
po
wer a
nd
wind
power. If increasi
ng the
n
u
mbe
r
of typi
cal
scena
rio
s
, the cal
c
ul
ation
error will be smaller. In the practical
work, t
he num
ber of typical
sce
nari
o
s
can be decided by
the accuracy
requi
rem
ent.
5. Conclusio
n
The
sto
c
ha
stic a
nd
unp
red
i
ctable
natu
r
e
of wi
nd
po
wer b
r
in
gs gre
a
t ch
allen
ge t
o
pe
ak
shavin
g ope
ration of po
we
r syste
m
inte
grated
wi
th la
rge
-
scal
e win
d
power. Thi
s
pape
r propo
se
s
a pea
k shavi
ng capa
city requireme
nt model ta
king
into acco
unt sp
e
c
ial cha
r
a
c
teri
stics
of wind
power, u
s
ing
KD sce
nari
o
s
red
u
ction
techni
que t
o
ch
oo
se ty
pical
pea
k
shaving
cap
a
c
ity
scena
rio
s
. Base
d on th
e
pea
k shavi
ng capa
city
requi
rem
ent
model, the
p
aper present
s a
method
of e
c
onomi
c
evalu
a
tion of
pea
k sh
aving.
F
r
o
m
the
ca
se
study, som
e
concl
u
si
on
s a
r
e
summ
ari
z
ed
as
foll
ows: (1
).
Wi
nd po
we
r
integ
r
at
io
n
can
save o
p
e
ration
al
co
st of conventio
nal
units, at
the
same time,
it i
n
crea
se
s th
e
co
st of
pea
k
shavin
g fo
r th
at its
va
riatio
n st
re
sses pe
ak
shavin
g p
r
e
s
sure of th
e p
o
we
r
system.
(2
). Th
e
r
mal
power gen
eration
an
d
coa
l
co
nsumptio
n of
the system repla
c
ed by the sa
m
e
ad
ded win
d
po
wer d
e
crea
se as win
d
powe
r
pen
etra
tion
increa
se
s, an
d the coal co
nsum
ption pe
r therma
l po
wer gen
eratio
n
incre
a
ses. T
he optimal wi
nd
power capa
ci
ty integrated sho
u
ld comp
rehen
sively
consi
der o
perational cost o
f
thermal po
wer
and wi
nd p
o
w
er
ab
sorption indi
ce
s a
nd so
on. (3). The p
r
e
s
ented mo
del
for econom
ic
evaluation of
peak
shavin
g of powe
r
system
integrated high p
e
netration
win
d
power, whi
c
h
gives a hig
h
cal
c
ulatio
n accuracy, re
du
ces the co
mp
u
t
ational com
p
lexity greatly, what’s m
o
re,
it
make
s it conv
enient for po
wer
system pl
anne
r to
anal
yze the effect
of wind power integ
r
ation
on
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Econom
ic Evaluation for P
eak Sha
v
in
g of Wi
nd Power Integrated
System
(Ting
t
ing Hou
)
3633
the pea
k sha
v
ing cap
a
city
requi
rem
ent and e
c
on
om
y
of powe
r
sy
stem intuitively and cl
early o
n
the whol
e, an
d can b
e
ea
si
ly applied to pra
c
tical p
r
oj
ects.
Ackn
o
w
l
e
dg
ements
This
work was supp
orted
by Key Project of
Nation
al Natu
ral Science Foun
d
a
tion of
Chin
a (No. 5
0937
002,Ba
si
c re
se
arch o
n
theory an
d
methodol
ogy
of se
cure op
e
r
ation of mo
d
e
rn
power
syste
m
s ba
se
d on
energy st
ora
ge), an
d Fun
d
of the Nati
onal
Prio
rity Basic
Re
se
arch of
Chin
a (No. 2
012
CB215
10
0, Basi
c scie
nce
re
se
arch
on la
rge
-
scal
e win
d
p
o
we
r integration),
an
d
the National
Natural S
c
ie
nce
Fou
ndati
on of
Ch
in
a
(No. 51
207
062, O
p
timizi
ng allo
catio
n
of
multiple energy storage units in
the
po
wer
system
with hi
gh
pen
etration
of
wi
nd p
o
wer). M
r
. Ho
Simon Wa
ng
has p
r
ovide
d
tutorial as
si
stance to impro
v
e the manuscript.
Referen
ces
[1] Maeg
aard
P.
W
i
nd e
ner
gy d
e
vel
o
p
m
e
n
t a
n
d
a
ppl
icatio
n
p
r
ospects
of no
n-grid-c
on
nect
ed w
i
n
d
p
o
w
e
r
.
Procee
din
g
s
of W
o
rld
No
n-Grid-Co
n
n
e
cted
W
i
nd P
o
w
e
r
a
nd E
ner
g
y
C
o
nferenc
e, W
N
W
E
C, Nan
jin
g,
Chin
a. 20
09; 1
-
3.
[2]
Daut I, Ir
w
a
nto
M, Su
w
a
rno,
et al. P
o
tenti
a
l
of
W
i
n
d
P
o
w
e
r Gener
atio
n i
n
Perlis,
North
e
r
n Mal
a
ysi
a
.
T
E
LKOMNIKA Indon
esi
an Jou
r
al of Electrica
l
Engin
eeri
ng.
2
011; 9(3): 5
75-
582.
[3]
F
a
rajia
np
our S
,
Mohamma
di
A,
T
a
vakoli S,
et al.
Improv
ed
Bacteria
l F
o
ra
gin
g
Alg
o
rithm
for Optimu
m
Econom
ic Emi
ssion D
i
spatc
h
w
i
th W
i
n
d
Po
w
e
r.
T
E
LK
OMNIKA Indonesi
an Jo
urna
l of Electrica
l
Engi
neer
in
g
. 2012: 10(
4): 640
-648.
[4]
Li Z
h
im
in
g, Ya
ng W
e
i, Z
han
g
Jun
y
a
ng,
Z
h
e
ng M
eng
w
e
i.
E
c
onom
ic p
eak
avoi
danc
e
on
electric
po
w
e
r
supp
l
y
.
H
e
il
on
gjia
ng e
l
ectric
pow
er.
200
1; 23 (5): 315-3
18.
[5]
Driese
n J, Bel
m
ans R.
Distri
buted g
e
n
e
rati
on: chal
len
ges
and poss
i
bl
e
soluti
ons
. 2006 IEEE Pow
e
r
Engi
neer
in
g Societ
y Gener
al
Meet
in
g, Montreal, Ca
na
da. 2
006.
[6]
H Bludszu
w
e
it,
JA Domíngue
z-Navarro,
A Ll
ombart.
Stat
isti
cal
ana
l
y
sis
of
w
i
nd
p
o
w
e
r for
e
cast err
o
r.
IEEE Transactions on power system
s.
20
08; 23(3): 98
3-9
9
1
[7]
Li W
e
n
y
i, Z
h
a
ng Bao
h
u
i
, Ba Gen.
Relia
bi
li
ty imp
a
cts of large sca
le
uti
l
i
z
a
t
i
o
n
of w
i
nd
energy
on
electric power system
s.
Proc
eed
ings
of CSEE. 2008; 28 (
1
): 100-1
05.
[8]
Z
hang J
i
etan,
Che
ng H
aozh
o
ng, Hu Z
e
ch
un
, et al.
Pow
e
r system pr
ob
ab
ilistic pr
oducti
o
n
si
mul
a
tio
n
inclu
d
i
ng w
i
nd
farms.
Proce
e
d
i
ngs of the CS
EE. 2009; 29(
2
8
): 34-39.
[9]
Z
hou W
e
i, Pe
ng Yu, Su
n H
u
i, W
e
i Qin
g
h
a
i.
Dyn
a
m
ic e
c
ono
mic dis
p
a
t
ch in w
i
nd
po
w
e
r integrate
d
system
.
Proce
edi
ngs of the C
SEE. 2009; 29(
25): 13-1
8
.
[10] Z
hang L
i
yin
g
, Ye T
i
nglu, Xi
n
Yaozh
o
n
g
, et al.
Probl
e
m
s a
nd meas
ures o
f
pow
er grid ac
commod
a
ting
larg
e scale w
i
n
d
pow
er.
Proce
edi
ngs of the C
SEE. 2010; 30(
25): 1-9.
[11]
Yang H
o
n
g
, Li
u Jian
xin, Yu
a
n
Jinsh
a
.
Res
e
arch of pe
ak lo
ad reg
u
l
a
ti
on o
f
conventi
ona
l gen
erators
i
n
w
i
nd pow
er gri
d
. Proceedings
of the C
SEE.
2010; 30(16): 26-31.
[12]
Xi
ao C
hua
ng
yi
ng, W
ang
Ni
ng
bo, Din
g Kun,
et al
. System
pow
er regu
lati
on sche
m
e for Jiuqu
an w
i
nd
pow
er base.
Pr
oceedings of the C
SEE. 2010; 30(10): 1-7.
[13]
Yi Lid
o
n
g
, Z
h
u
Min
y
i, W
e
i Le
i
,
et al. A comput
ing met
hod f
o
r peak
loa
d
re
gul
ation
abi
lit
y
of North
w
est
Chin
a po
w
e
r g
r
id con
necte
d w
i
t
h
lar
ge-sca
l
e
w
i
n
d
farms.
Pow
e
r System T
e
chnol
ogy
. 201
0;
34(2):
129-
132.
[14]
Li F
u
q
i
a
ng, W
ang B
i
n, T
u
Shao
lia
ng, et
al.
Anal
ys
is o
n
p
eak l
o
a
d
reg
u
l
a
tion
perform
a
n
ce of B
e
ij
ing-
T
i
anjin-T
angsh
an Po
w
e
r Grid
w
i
th
w
i
n
d
farms connecte
d.
Pow
e
r System T
e
ch
nol
ogy
.
2010; 3
3
(18):
129-
132.
[15]
Z
hang
Ni
ng, Z
hou
T
i
anrui,
D
uan
Ch
an
gg
an
g, et a
l
. Impac
t of lar
ge-sc
ale
w
i
n
d
farm
co
n
nectin
g
w
i
t
h
po
w
e
r grid o
n
peak l
o
a
d
regu
latio
n
dema
nd.
Pow
e
r System T
e
chnol
ogy
. 2
010; 34(
1): 152
-159.
[16]
Dan
ge Hu
an
g, Billio
nt
on R. Effects of
w
i
nd
po
w
e
r on b
u
l
k
s
y
stem ade
q
uanc
y ev
alu
a
ti
on usi
ng th
e
w
e
ll-
be
ing a
n
a
l
ysis frame
w
o
r
k
.
IEEE
Transac
tions on power system
s
. 20
09;
24(3): 123
2-1
240.
[17]
Razal
i
NMM,
Hashim A
H
.
Backw
ard red
u
c
t
ion a
ppl
icati
o
n for mini
mi
z
i
ng w
i
nd
pow
e
r
scenari
o
s i
n
stochastic pro
g
ra
mmin
g
. Po
w
e
r Eng
i
ne
eri
ng an
d Optimizatio
n
Confer
ence (PEOCO
), 2010 4t
h
Internatio
na
l. I
EEE, 2010: 43
0-43
4.
[18] Rach
ev
ST
.
Proba
bil
i
ty metric
s and the st
ab
il
ity of stochastic mod
e
ls
. Ch
ic
hester, U.K.: Wile
y, 19
91.
.
Evaluation Warning : The document was created with Spire.PDF for Python.