TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.12, No.7, July 201
4, pp
. 5052 ~ 50
7
0
DOI: 10.115
9
1
/telkomni
ka.
v
12i7.573
1
5052
Re
cei
v
ed Fe
brua
ry 2, 201
3; Revi
se
d Ma
rch 13, 201
4
;
Accepte
d
March 29, 201
4
Multi-Area Automatic Generation Control Scheme
including Renewable Energy Sources
Sandeep Bh
ongade
*
1
, Ba
rjeev
T
y
agi
2
, H.O. Gupta
3
1
Electrical En
gi
neer
ing D
e
p
a
rtment, G.S.
In
stitute of T
e
chnolog
y a
nd Sci
e
n
c
e,
Indore (M.P)-4
520
03 Ind
i
a, T
e
l: +
91 982
66
8
972
7, F
a
x: +
9
1
7312
43
254
0
2
Electrical En
gi
neer
ing D
e
p
a
rtment,
Indian In
stitute of
T
e
chnol
og
y,
Roork
ee (Uttar
a
kha
nd)-2
47
66
7 India
3
Information T
e
chno
log
y
D
e
p
a
r
tment, Ja
y
p
ee
Institute of Informatio
n
T
e
chnol
og
y,
Noid
a (U.P) -
2
013
07.Ind
i
a, T
e
le
pho
ne: +
91-
120-
240
09
73-
9
76, 240
09
87, F
a
x: +
91-
120-
24
009
86
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: bhon
ga
desa
nde
ep@
gmai
l.com
A
b
st
r
a
ct
In this
pa
per,
a
multi-
area
Aut
o
matic Ge
ner
a
t
ion
Contr
o
l (A
GC) sche
m
e w
i
th R
enew
a
b
le
Energ
y
Sources (RES)
suitable in
a restru
ctured int
e
rconnected power system
has been pr
oposed.
Photo-v
o
lt
aic
and w
i
n
d
turb
i
ne g
e
n
e
ratin
g
system h
a
s b
e
en i
n
tegr
at
ed
w
i
th the grid.
The dev
elo
p
e
d
sche
m
e
has b
een
investi
gate
d
fo
r frequ
ency c
o
ntrol w
i
th
and
w
i
thout R
ES
u
n
its. A PID co
ntroll
er h
a
s b
e
en
used
to co
ntrol
th
e
re
a
l
po
wer o
u
t
p
u
t
o
f
fo
ssi
l
s
fu
e
l
ge
ne
ra
to
rs. Th
e
p
a
r
ame
t
e
r
s o
f
PID
co
n
t
ro
l
l
e
r h
a
s
b
e
e
n
tuned
accord
ing to G
enetic A
l
gor
ith
m
(GA) bas
ed
perfor
m
a
n
ce
i
n
dices. T
he fu
n
c
tioni
ng of pr
o
pose
d
sche
m
e
has
bee
n tested
on
a 39-
bus N
e
w
Engl
and syst
e
m
a
nd
on a
75
-bus Ind
i
an
po
w
e
r system n
e
tw
ork. Frequen
cy
regulation m
a
r
k
et
scenario has
be
en cons
idered in both the system
s.
The res
u
lts of proposed A
G
C
sche
m
e w
i
th a
nd w
i
thout RE
S units have
b
een co
mpar
ed.
Ke
y
w
ords
:
re
new
abl
e e
nerg
y
sources (RE
S
), PV system,
WTG sys
tem, flyw
heel e
ner
gy storag
e sys
t
e
m
(FESS)
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
Aroun
d the worl
d, governments h
a
ve
been pay
in
g more atte
ntion on poli
c
ie
s that
prom
oting th
e cl
ean
o
r
Ren
e
wable
Energy
S
ource
s
(RES) t
o
serve
customer
dem
a
nd.
Ren
e
wable
energy curre
n
tly provides more
than
14% of th
e worl
d’s e
n
e
rgy su
pply
[1].
Acco
rdi
ng to the some p
a
rt of the world’
s we
at
her
co
ndition
s, it does not allo
w the co
nstructi
on
of photovoltaic (PV) plant
only and/or wind gene
rato
r (WG
) only stand-alon
e pla
n
ts, despite the
availability of these technologies
for energy
supply
in off grid
sy
stem
s. Thus,
the hybri
d
pl
ants
are gu
arante
ed the co
ntin
uity of the supply mixi
ng the different rene
wable
en
ergy re
so
urces –
like PV, WG
, even micro
-
hydro – limiting the
diesel gene
rator
set (DGS) u
s
e for ba
ck-up
purp
o
se o
n
ly. In ad
diction
a hydroge
n f
uel
cell
(FC)
can
be
add
e
d
to the
de
si
gned
hybri
d
plant
in orde
r to
re
alize
a
syste
m
witho
u
t DGS [2].
In co
nne
ction to
this,
RES technolo
g
y such
as
Photo-voltai
c (PV) sy
ste
m
and
Wind
Turbi
ne G
e
nerato
r
(WT
G
) a
r
e the t
w
o mo
st attractive
te
c
h
no
lo
g
i
es
.
The ph
otovoltaic (PV
)
plant
s, on the
co
st
poi
nt of view, have som
e
disa
dvantag
e
s
over
other
co
nven
tional en
ergy re
sou
r
ces.
In the
restru
cture
d
p
o
wer system
sce
nario,
witho
u
t
spe
c
ific pu
bli
c
in
ce
ntives
whi
c
h
cal
c
ul
a
t
e the
so
cial
advantag
es
o
ffered by
PV techn
o
logy,
t
h
e
photovoltai
c
i
s
n
o
t yet co
mpetitive wit
h
othe
r
re
so
urces.
The
p
r
ice
pe
r
watt
for a
PV m
odule
decrea
s
e
s
in
the recent years,
spe
c
ifi
c
ally,
the pri
c
e/Watt peak in Europe
a
n
cou
n
trie
s
has
decli
ned
fro
m
€5.5 i
n
2
001 to
€2.1
in 20
11 [3].
The the
r
mal
conve
n
tional
techn
o
logie
s
are
actually m
o
re
expen
sive in
term
of soci
al cost
s,
but
cu
stome
r
s no
dire
ctly
pay
t
h
is so
cial co
st
that is in ch
a
r
ge o
n
the society [4]. The in
fluen
ce o
f
PV system on po
wer
sy
stem freq
uen
cy
control is di
scussed in [5].
Integrating
WTG syste
m
wi
th energy sto
r
age
units in
a multi-a
r
ea
Automatic G
e
neratio
n
Control (AG
C
)
scheme,
the gene
ra
ted elect
r
ic
energy can
be effectivel
y controlle
d
the
freque
ncy
de
viation and
meet the d
e
m
and
of co
ntro
l a
r
ea. T
h
e
desi
r
e
d
characteri
stic of
WTG
system
s (win
d farm) is to
get the maximum
output
power for a
cert
ain wi
nd speed. When
the
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Multi-Are
a
Autom
a
tic Gene
ration Control
Sc
hem
e incl
uding
Rene
wable… (S
and
eep Bhon
gad
e)
5053
WTG sy
stem
s
connected
to a grid, the wind farm
will have no reserve to
supply power under
emergen
cy i.
e. wh
en
the
grid
freq
uen
cy is l
o
w.
In
o
r
de
r to
pa
rtici
pate
RES u
n
i
ts pa
rticularl
y
,
WTG
syste
m
s in f
r
eq
uen
cy regul
ation, t
he wi
nd fa
rm
s
shoul
d o
p
e
r
ate
with rese
rves.
Whe
n
t
h
e
system frequ
ency i
s
high
e
r
or l
o
we
r tha
n
the nom
i
n
a
l
value, the p
i
tch controll
er of wind fa
rm
s
help
s
to
incre
a
se
o
r
d
e
cre
a
se
the
captu
r
ed
wi
nd
po
wer
and
this fe
ature
hel
ps th
e wi
nd
farm
s
to
partici
pate in
power sha
r
in
g [6].
The gri
d
fre
quen
cy can
be controll
ed
by
cont
rolli
ng the
real
power o
u
tput
of the
conve
n
tional
gene
rating
pl
ants.
RES ca
n be
u
s
ed
fo
r frequ
en
cy re
gulation
se
rvi
c
e
s
. Advanta
g
e
of usin
g
RES is that e
n
e
r
g
y
can
be
stored in
ene
rgy
stora
ge
syste
m
wh
en it i
s
in abu
nda
nce
and late
r on,
it can used
to bring the
system
frequ
ency at nomi
nal value. No
wad
a
ys, ene
rgy
storage
devices like Flywheel
Energy
Storage Sy
stem (FESS) [
6
-8], battery
storage
energy
storage
(BESS) [9-10], advanced
capacitor [11-12], supercondu
cting magneti
c
energy storage
(SMES) [9, 13] are u
s
ed fo
r frequ
en
cy regulatio
n app
lication.
Dynami
c
p
r
o
pertie
s
of
win
d
turbi
n
have
been
di
scussed
in [14]. T
he p
r
op
osed
dynami
c
model of wi
n
d
turbin
con
s
idere
s
rotatio
nal e
ffect
s of blade in m
a
thematical eq
uation form
a
nd
then sove it using finite el
e
m
ent metho
d
. Due to
the
n
on-lin
ea
r ch
aractersti
c of fuel cell m
odel,
a
large
chan
ge
in the
output
voltage
of fu
el cell ta
ke pl
ace,
wh
en lo
ad
cha
nge
s, t
herefo
r
e
for t
he
appli
c
ation of
feul cell in
di
stribute
d
ge
n
e
ration
sy
ste
m
, a con
s
tan
t
output voltage of feul cell
is
requi
re
d [15].
This p
ape
r
pre
s
ent
s a
multi-area A
G
C
sc
hem
e inclu
d
ing RE
S
system
s suitable
in
comp
etitive electri
c
ity market. The
dev
elope
d sc
h
e
m
e an
alyze
s
the effect of
RES syste
m
s on
freque
ncy re
gulation.
In m
odelin
g
the RES
system
s,
dire
ct conve
r
sion
of the
sunlight a
nd
wind
spe
ed into
el
ectri
c
ity ha
s
been
utilize
d
in case of
PV system
s
as
well a
s
i
n
ca
se
of WTG
system
s also
. A FESS has also been i
n
tegrate
d
wi
t
h
PV and WTG system
s. In this work i
t
is
assume
d that
the
RES unit
s
a
r
e
deliveri
ng its ma
xim
u
m real p
o
we
r outp
u
t at a
given time
while
prop
ortio
nal,
integral and derivative
(PI
D
) cont
rolle
r has b
een
used to ch
ang
e
the real p
o
wer
output of
con
v
entional g
e
n
e
rato
rs. Pa
ra
meters of
the
PID co
ntroll
e
r
are tun
ed u
s
ing th
e G
e
n
e
tic
Algorithm. Int
egral
of the
squ
a
re
of th
e area
co
ntrol erro
r
(ISACE)
have
be
en utili
zed
a
s
the
fitness fun
c
tio
n
for geneti
c
algorith
m
.
The p
r
opo
se
d AGC
sche
me ha
s bee
n test
ed o
n
39-Ne
w Engl
and sy
stem
whi
c
h is
divided into t
w
o
control area and
on a
75-b
u
s In
di
a
n
power
syst
em divided i
n
to four
cont
rol
area
s. A
de
regulate
d
el
ectricity ma
rket scen
ario
h
a
s
b
een
a
s
su
med in
b
o
th t
he
system
s.
The
PV generator is in
clude
d i
n
are
a
-1 and
WTG
syst
e
m
in area-2 i
n
ca
se
of
39-bus system a
nd
simila
rly, one PV system in inclu
ded in a
r
ea-2 a
nd
on
e
WTG sy
stem
in area
-4, in ca
se of 75-bu
s
system. The Flywheel E
nergy Storage System
(F
ESS) has been incl
uded
i
n
the respective
area
s of b
o
th the sy
ste
m
, whe
r
e P
V
gene
rators and
WTG
system
s a
r
e
con
s
id
ere
d
. The
perfo
rman
ce
studie
s
have
been carried
out by us
ing
the MATLAB SIMULINK for tran
sa
ctio
ns
within an
d across the
control area b
oun
d
a
rie
s
.
2. Sy
stem
Modelling
Duri
ng the
past o
ne a
n
d
half de
ca
de, many el
ectri
c
utilities and p
o
wer netwo
rk
comp
anie
s
,
world
-
wi
de, h
a
v
e bee
n force
d
to
cha
nge
their
way
of d
o
ing
bu
sine
ss, from ve
rtical
ly
integrate
d
m
onop
oly to a
n
op
en m
a
rket enviro
n
me
nt. Electri
c
ity refo
rms a
r
e
being
broug
h
t
to
introdu
ce
co
mmercial in
centives in g
eneration,
transmi
ssion,
distrib
u
tion a
nd retailin
g o
f
electri
c
ity, wi
th re
sultant
efficien
cy gai
n, in
many
ca
se
s. The i
n
trodu
ction
o
f
comp
etition
in
electri
c
ity m
a
rket may
cause em
erg
ence of
sev
e
ral
ne
w en
tities. Such
as, G
ene
rati
ng
Comp
anie
s
(Gen
co
s),
T
r
a
n
smi
ssi
on Co
mpany (Tran
s
co), Di
stribu
tion
Compa
n
i
e
s (Disco
s), and
an Indepe
nd
ent System Operator
(IS
O
): The sy
stem operator
is
an entity entruste
d
with
the
respon
sibility
of ensu
r
ing t
he relia
bility and securi
ty of the power
system. It is an inde
pen
d
ent
entity and does n
o
t parti
cipate in the
electri
c
it
y trading. It usu
a
lly does not
own gen
era
t
ing
resou
r
ces, ex
cept fo
r som
e
re
se
rve ca
pacity in
ce
rtain cases. In
ord
e
r to m
a
i
n
tain the
syst
em
se
curity an
d reliability, the SO pro
c
u
r
e
s
variou
s
services, such as
supply of eme
r
gen
cy re
serv
es
or rea
c
tive p
o
we
r fro
m
th
e othe
r entiti
e
s in
th
e
system. The
s
e
se
rvice
s
a
r
e kn
own a
s
the
‘ancilla
ry se
rv
ice
s
’. One of su
ch se
rvice is the freq
uen
cy regul
ation.
Freq
uen
cy Reg
u
lation
Service
s
:
F
r
eque
ncy re
gulation
is the
min
u
te-t
o-min
u
te
adaptatio
n of
the ge
nerato
r
outp
u
t to m
eet t
he imb
a
l
ance bet
wee
n
total supply
and
dema
n
d
in
the system.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 7, July 201
4: 5052 – 50
70
5054
In order to
maintain the system
security
and reli
ability, the S
y
stem Operator (S
O)
pro
c
u
r
e
s
vari
ous services,
su
ch
a
s
su
pply of
em
ergen
cy re
se
rv
es, freque
ncy regul
ation
and
rea
c
tive p
o
wer from th
e ot
her entitie
s in
the
sy
stem.
These se
rvices are kno
w
n
as
the
‘an
c
ill
ary
servi
c
e
s
’. On
e of su
ch
service i
s
the
freque
ncy
regulatio
n. Freque
ncy reg
u
lation h
e
lps to
maintain i
n
tercon
ne
ction freque
ncy, min
i
mize
differe
n
c
e
s
b
e
twe
en
actual
an
d
schedul
ed
po
wer
flows bet
wee
n
control
a
r
e
a
s, a
nd m
a
tch the
gen
erat
ion to th
e loa
d
withi
n
the
control a
r
e
a
.
In
freque
ncy re
gulation
servi
c
e ma
ny co
mmercial tra
n
sa
ction
s
ca
n take pla
c
e
such a
s
Poolco,
bilateral, an
d a combi
natio
n of these two.
In Poolco ba
sed tran
sa
ctio
n [16], the Discos
an
d Ge
nco
s
of the same are
a
participate in
the frequ
en
cy regulatio
n throu
gh
syste
m
operator
.
SO accept
s
bids
(volume
and pri
c
e
)
from
power
pro
d
u
c
ers
(G
en
co
s) who
are
will
ing to q
u
ic
kly
(withi
n a
bou
t 10-1
5
min
u
tes) in
cre
a
se
or
decrea
s
e
the
i
r level of
p
r
odu
ction.
Consume
r
s
(Discos)
al
so can su
bmit bids
to
SO for
increa
sing
o
r
decrea
s
in
g th
eir level
of
co
nsum
pti
on. In
ea
ch
hou
r
of ope
ration,
th
e SO
activate
s
the most favo
rable bi
d.
In bilateral transactio
n
, Gencos an
d Discos
n
egotia
te bilateral contra
cts am
o
ng ea
ch
other a
nd
sub
m
it their contract
ual
agree
ments to
a SO. The pl
ay
ers a
r
e respon
sible for
havin
g a
comm
uni
cati
on path to excha
nge
contract data a
s
well as mea
s
u
r
ements to d
o
load follo
wing
in
real
-time. In such a
n
arran
gement, a Di
sco se
nd
s a pulse to Gen
c
o to follow t
he pre
d
icte
d load
as lo
ng a
s
it
doe
s not
exceed the
contracted val
ue.
The respon
si
bility of the Di
sco is to mo
n
i
tor
its load
conti
nuou
sly and
ensure th
e l
oad
s followi
n
g
req
u
ire
m
en
ts are m
e
t a
c
cordi
ng to t
h
e
contractu
a
l a
g
ree
m
ent. A detailed di
scu
ssi
on on bil
a
teral tra
n
sacti
ons i
s
given i
n
[17-18].
2.1.
Calcula
t
ion of Are
a
Co
ntrol Error (ACE)
In a
pra
c
tical multi
are
a
po
we
r
sy
stem, a
con
t
rol a
r
ea
i
s
interco
nne
ct
ed to
its
neigh
bori
ng
area
s
with tie lines, all forming pa
rt of the overall po
wer
pool. If
ij
P
is the tie line
real
po
we
r flo
w
fro
m
a
n
a
r
ea-i to
an
oth
e
r a
r
e
a
- j
an
d
m is the tota
l numb
e
r of a
r
ea
s, the
net
tie
line power flow from area-i will be:
m
i
j
j
ij
i
tie
P
P
1
(
1
)
In a conve
n
tional
AGC formulation,
ij
P
is gene
rally mai
n
tained at a fi
xed value. , in a
dere
gulate
d
e
l
ectri
c
ity ma
rket, a Di
sco
m
a
y have
co
ntracts with
the
Gen
c
o
s
in
th
e same
area
as
well
as with
the G
e
n
c
o
s
in
othe
r a
r
ea
s,
too.
Hen
c
e, the
sch
edul
e
d
tie-lin
e power of
any are
a
may chan
ge
as the dem
an
d of the Disco
chan
ge
s.
Thus,
the
net
ch
ang
e in
th
e sch
edul
ed
steady-state
power flo
w
o
n
the tie
line
from an
area
- i ca
n be
expresse
d a
s
:
m
i
j
j
ji
m
i
j
j
ij
i
tie
new
tie
D
D
P
P
1
1
(
2
)
Whe
r
e,
i
tie
P
is t
he chan
ge i
n
the
sched
uled tie
-
line
power
due t
o
ch
ang
e in
the
deman
d,
D
ij
is the dema
n
d
of Discos i
n
area
-j from
Gen
c
o
s
in area-i , and
D
ji
is the dema
n
d
o
f
D
i
sc
os
in
a
r
ea
-
i fr
o
m
G
enc
os
in
a
r
ea
-
j
.
Gene
rally,
0
i
tie
P
(Convention
a
l
AGC).
Du
rin
g
the tra
n
si
e
n
t peri
od, at
any given
time, the tie-line po
wer e
r
ror is given a
s
:
new
i
tie
actual
i
tie
error
i
tie
P
P
P
(
3
)
This e
rro
r si
g
nal ca
n be u
s
ed to gene
rat
e
the Area Control Error
(ACE) sig
nal a
s
:
error
i
tie
i
i
i
P
f
B
ACE
(
4
)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Multi-Are
a
Autom
a
tic Gene
ration Control
Sc
hem
e incl
uding
Rene
wable… (S
and
eep Bhon
gad
e)
5055
Whe
r
e, B
i
is the frequ
en
cy bias fa
ctor an
d
∆
f
i
is the frequen
cy devia
tion in area
-i.
2.2.
Modeling of
PV Genera
to
r
In Photovoltaic technolo
g
y, there is a dire
ct co
nversi
on of sunli
ght into electri
c
ity
throug
h the u
s
e of ph
otovoltaic a
rray.
The in
comi
n
g
sola
r ra
diation or
sunli
g
h
t
is measure
d
in
units
Watts
/meter
2
.
The a
s
sum
p
tions mad
e
in
the math
em
atical m
odeli
ng of PV g
e
nerato
r
s a
r
e:
All the
energy lo
sse
s
in
a PV g
e
nerato
r
, in
clu
d
ing
con
n
e
c
tion lo
sse
s
, wi
ring l
o
sse
s
a
nd oth
e
r l
o
sses
are a
s
sume
d
to be ze
ro.
Secon
d
, is th
e PV gene
rat
o
r ha
s a m
a
ximum po
we
r
point tra
c
ker
i.e
.
1
, where,
is the conve
r
si
on
efficien
cy of PV generato
r
.
The output p
o
we
r of the PV system ca
n
be expre
s
se
d as follo
ws [
19]:
(
1
0.005
(
25
)
)
PVG
PVG
a
ES
T
(
5
)
Whe
r
e,
is the solar i
rra
diati
on (W/m
2
),
S
is the surfa
c
e a
r
ea of the PV
module
s
in m
2
,
a
T
is the ambi
e
n
t temperature and
PVG
is the
conve
r
si
on ef
ficien
cy of PV gene
rator. F
r
om (5
)
,
it is clea
r that
the output po
wer
of PV syst
em mainly d
epen
ds o
n
a
m
bient tempe
r
ature (
a
T
), and
sola
r radiatio
n (
) becau
se
conve
r
si
on ef
ficien
cy of PV array
PVG
and
S
surface a
r
ea o
f
PV
array are cons
tant. In this
work, it is assumed that
PVG
E
is linearly vari
ed
with
only.
T
h
e
tr
an
s
f
er
fu
n
c
tion
o
f
PV is
r
e
pr
es
en
ted
b
y
a simpl
e
first ord
e
r
system an
d de
scrib
ed
in [19]:
1
1
PVG
PV
P
VG
E
G
sT
(
6
)
Whe
r
e,
PVG
T
is called t
i
me con
s
t
ant
of
P
V
syst
em.
In Figure 1, converte
r is
bi
dire
ctional i.e
., it
not only can supply a
c
power to the l
oad, but
also
can
charge the FESS by rect
ifying the surplus po
wer
when the total supply power exceeds
the load po
wer.
DC-
A
C
C
onv
e
r
ter
FESS
S
o
lar
i
r
r
adiat
ion
Photov
olt
a
ic
sy
stem
P
FES
S
T
o
pow
er
sy
st
em
RE
S
P
E
pv
g
AC
-AC
Co
nv
e
r
ter
()
P
VG
P
Figure 1. Grid
-co
nne
cted P
V
System
2.3.
Modeling of
Wind Turbin
e
genera
tor
(WTG
)
The ge
nerate
d
power of the win
d
tu
rbi
ne gen
erato
r
depen
ds o
n
the wind
sp
eed
W
V
.
The me
cha
n
i
c
al po
we
r out
put of the win
d
turbine i
s
e
x
presse
d as [
20]:
3
W
1
P
2
rp
w
A
CV
(
7
)
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 7, July 201
4: 5052 – 50
70
5056
Whe
r
e,
is the air den
sity in kg/m
3
,
r
A
is the swept a
r
e
a
of blade
s in m
2
,
p
C
is the
power
co
effici
ent (di
m
en
sio
n
less) which i
s
a
fun
c
tion
o
f
tip sp
eed
rat
i
o (
λ
) an
d bl
a
de pit
c
h
angl
e
(
β
) in de
gree
s,
W
V
is the win
d
spe
ed in m/sec.
The tran
sfe
r
functio
n
of WTGs i
s
given
by simple lin
ear first ord
e
r lag by negle
c
ting all
the non-li
nea
rity [19],
1
1
WT
G
WT
G
WW
T
G
E
G
Ps
T
(
8
)
Whe
r
e,
WTG
T
is cal
l
ed time con
s
tant of wind g
enerator, an
d
taken a
s
1.5 se
c.
A wind farm is a g
r
ou
p of wind tu
rbine
s
in the
sam
e
locatio
n
u
s
ed for p
r
od
u
c
tion of
electri
c
p
o
we
r. A larg
e wi
n
d
farm m
a
y consi
s
t of several h
und
re
d
individual
win
d
turbin
es,
a
n
d
cover an
extende
d area
of hund
red
s
of squ
a
re
mi
les. Nowada
ys, onsho
re
wind fa
rm
s
are
cap
able
of n
o
t only ge
nerating po
we
r
but also p
r
ov
iding a
n
cill
ary servi
c
e
s
[2
2]. Onshore
wind
farms
can a
c
t
ually be co
nsi
dere
d
as
WT
Gs a
s
they can be op
erat
ed as
conve
n
t
ional gen
erat
ors
[23].
In orde
r to have improved frequ
en
cy respon
se th
e wind farm
shoul
d ope
rate with
reserve
s
. Th
e most im
portant feat
ure o
f
the wind fa
rm to parti
cipa
te in po
wer
sharin
g when t
h
e
system frequ
ency deviate
s from the
n
o
minal valu
e
is that the pitch
controll
ers
of wind f
a
rm
increa
se
s or
decrea
s
e
s
th
e captu
r
ed
wi
nd po
wer [24]
.
In ca
se
whe
n
the WTG
system
s a
r
e
con
n
e
c
ted to
grid, vari
abl
e-spee
d wi
n
d
turbin
e
use
d
, the rot
a
tional spee
d
is de
cou
p
le
d from g
r
id frequ
en
cy by power
conve
r
ter. The i
nertia
con
s
tant for
wind p
o
wer i
s
time de
pen
dant. The typ
i
cal ine
r
tia
co
nstant for th
e
wind tu
rbin
e
s
is
about 2
-
6
se
c [25]. De
pe
nding o
n
the
type of gene
ra
tor unit
s
, typical in
ertia
consta
nts for t
h
e
grid p
o
wer g
e
nerato
r
s are i
n
the ra
nge o
f
2-9 sec
[2
6]. A complete
configuration o
f
WTG sy
ste
m
into AGC for
area
-i is
sho
w
n in Figu
re
2.
~
Wi
n
d
T
u
r
b
i
n
e
F
ESS
R
ES
P
Wi
n
d
s
p
e
e
d
WT
G
E
F
ES
S
P
AC
-AC
Conv
ert
e
r
DC
-A
C
In
ve
rt
e
r
WT
G
P
To
P
o
w
e
r
sy
st
e
m
Figure 2. Con
f
iguration of
WTG System
in the Propo
sed AG
C Sch
e
me [5]
2.4.
Modeling of
Fl
y
w
heel Energ
y
Storage S
y
stem (FESS)
Integrating an Energy Storage System
(ESS) into the PV and WTG sy
stems can
sup
p
re
ss the
fluctuation
s
in
sola
r ra
diatio
ns an
d wi
nd velocity (spe
e
d
). Flywh
eel
Energy Stora
ge
System (FES
S) stores
en
ergy in the fo
rm of
the ki
n
e
tic ene
rgy stored in the
rotating flywh
eel
and can be retrieved later as an electri
c
al out
put. There are
some advantages of FESS over
Battery Energy Storage
System (BESS), and they ar
e higher power densi
t
y, insensitivity to
environ
menta
l
co
ndition
s,
no h
a
za
rd
ou
s ch
emicals
et
c. The
ki
netic ene
rgy
store
d
in the
rotating
flywheel is gi
ven by:
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Multi-Are
a
Autom
a
tic Gene
ration Control
Sc
hem
e incl
uding
Rene
wable… (S
and
eep Bhon
gad
e)
5057
2
1
2
E
J
(
9
)
Whe
r
e,
E
is the
Energy sto
r
ed in the flywhe
el in N-m,
J
is the flywhe
el mome
nt of
inertia in N-m
-
se
c
2
, and
is the rotational
velocity in rad
/
sec.
A FESS inte
grates the function of a m
o
tor,
flywheel
rotor and generator into
a singl
e
integrate
d
sy
stem. The m
o
tor (while "chargi
ng" t
he flywheel
), whi
c
h u
s
e
s
ele
c
tric current fro
m
the utility grid to provide energy to ro
tat
e
the flywhee
l, spins con
s
tantly
to maintain a ready
sou
r
ce of
kin
e
tic e
nergy (
E
). The g
ene
ra
tor (whil
e
"di
s
-cha
rgi
ng" t
he flywh
eel) then
co
nverts
the kineti
c
en
ergy of the flywhe
el into ele
c
tri
c
ity (
FEES
P
).
In the present study, it is assumed that
FESS has enough capacity to store surplus
energy ge
ne
rated by th
e
gene
rating
u
n
its. When
the d
e
man
d
i
n
control
a
r
e
a
in
cre
a
ses,
the
FESS can release enough energy to
the connected load
within a
very short
time. Theref
ore,
whe
n
PV
system is i
n
cl
ud
ed in
the AG
C bl
ock fo
r a
r
ea
-i, then
th
e net
po
wer
gene
rated
in
the
system
can b
e
expre
s
sed
as:
R
E
S
P
VG
FEES
PP
P
(
1
0
)
Similarly, wh
en WT
G sy
st
em is
con
s
id
ered
i
n
the A
G
C bl
ock for area
-i, then
the net
power ge
ne
ra
ted in the system can be e
x
presse
d as:
R
E
S
W
T
G
FEE
S
PP
P
(
1
1
)
The trans
f
er func
tion of the s
t
orage s
y
s
t
em
s
FESS c
a
n be tak
e
n as firs
t order lag [19],
1
FE
ES
F
EE
S
F
EE
S
K
f
Ps
T
(
1
2
)
Whe
r
e,
K
FEES
is the gain consta
nt and
T
FEES
is
the tim
e
c
o
ns
tant.
C
ont
rol
l
e
r
Conventional
G
enc
os
PO
W
E
R
SYST
EM
Ti
e-
Li
ne
Ti
e-
L
i
ne Er
r
o
r
D
I
S
C
O
Bi
+
|
+
+
Bi
later
a
l
tr
ans
ac
t
i
on s
i
gnals
to Genc
o of
s
a
m
e
area
Bi
later
a
l
tr
ans
ac
t
i
on s
i
gnals
to Genc
o of
other
ar
ea
D
P
i
tie
P
f
r
equenc
y
deviation
s
i
gnals
f
r
om
other
ar
eas
T
i
e-li
ne devi
a
t
i
on
s
i
gnals
to o
t
her areas
B
i
l
a
t
e
ral t
r
ans
ac
tion s
i
gnal
s
f
r
om
Di
s
c
os
of
ot
her
areas
Dem
and
of
Di
s
c
os
in ot
her
areas
f
r
om
G
enc
os
in area-
i
Dem
and
of
D
i
s
c
o
s
i
n
area-
i
f
r
om
Genc
os
in
ot
her areas
1/R
i
f
i
ACE
+
+
+
-
-
-
+
j
ti
e
P
i
tie
P
G
P
+
PVG
P
+
F
ESS
P
+
R
ES
P
WT
G
sy
st
e
m
FE
E
S
PV s
y
s
t
e
m
+
WTG
P
Figure 3. AGC Block diag
ram for area-i.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 7, July 201
4: 5052 – 50
70
5058
In the pre
s
e
n
t study, Beaco
n's Sm
art
Energy 25
flywheel ha
s been
con
s
i
dere
d
to
extract
and/o
r
d
e
liver
po
wer th
at is seal
ed in
a
vacu
u
m
chamb
e
r a
nd
spin
s
bet
wee
n
8,0
00
a
n
d
16,000 rpm. At 16,000 rp
m the flywheel can
store a
nd deliver 2
5
kWh of extractable ene
rgy [27].
The ove
r
all
block
diag
ra
m of AG
C
sc
heme
in
cl
ude
s
RES (PV system
and
WT
G
sy
st
em
s) f
o
r
an i
th
area
of m-a
r
ea
power
syste
m
is
sho
w
n
in Figu
re
3
.
The ge
ne
ration
sub
s
ystem
s
compri
se conv
entional ge
ne
rators, WT
Gs and PV. The energy sto
r
a
ge sub
s
yste
m
inc
l
udes
an
FESS that is c
o
nnec
ted t
o
the lo
ad
s
i
de. Ass
u
me
only the
s
t
udied WTG; PV
and
FESS require suitable power conver
ters for exchangi
ng
energy wit
h
the studied ac sy
stem. The
FESS is assumed to have
enough
capacity to stor
e surplus energy
generated by the generati
ng
s
u
bs
ys
tems
. When the power demand inc
r
eas
e
s
,
the FESS c
an releas
e enough energy to the
c
o
nnec
ted load within a very s
h
ort time.
The fo
rm
of a
PID
cont
rolle
r
can
be
exp
r
esse
d
a
s
th
e
sum
of th
ree
terms,
propo
rtional,
integral, and
derivative co
ntrol.
Th
e tra
n
sfer fun
c
tion
of such
a
PID
cont
rolle
r
can b
e
exp
r
e
s
sed
as:
s
K
s
K
K
s
G
d
i
p
C
)
(
(
1
3
)
Whe
r
e,
d
i
p
K
K
K
,
,
are the propo
rtion
a
l, integral a
nd de
rivative gain co
nsta
nt of the
controller. Optimal values of
d
i
p
K
K
K
,
,
ca
n b
e
determi
ned
b
y
many
ways, one
of the
m
, is
sug
g
e
s
ted by
the Dond
e et al [15].
Initially, parameters (
d
i
p
K
K
K
,
,
) of PID controlle
r are
a
sel
e
ct
ed usi
ng Le
a
s
t Square
Minimization method, whi
c
h
give
s stabl
e
re
sult
s.
ACE is mi
nimize
d u
s
ing th
e
GA optimi
z
ati
o
n
toolbox GAO
T
in MATLAB prop
osed by
C. R. Hou
c
k
[28] to obtain
the optimal
PID para
m
et
ers.
The fitne
s
s fu
nction t
a
ke
n i
n
the p
r
e
s
ent
wo
rk is
i
n
teg
r
al of th
e squ
a
re
of the Area Control Error
(ISACE).
The problem
to determin
e
d
i
p
K
K
K
,
,
is formul
a
t
ed as follo
ws: Minimize (Integral of
squ
a
re of the
Area Control
Erro
r).
2
1
()
m
i
i
IS
A
C
E
A
C
E
(
1
4
)
Whe
r
e, m is the numb
e
r of
area in the system.
Subjecte
d to:
mi
n
m
a
x
,,
,
mi
n
m
a
x
,,
,
mi
n
m
a
x
,,
,
pi
p
i
pi
ii
ii
ii
d
i
di
di
KK
K
KK
K
KK
K
Whe
r
e,
i
d
i
i
i
p
K
K
K
,
,
,
,
,
are
the p
r
op
ortio
n
a
l, integral a
nd d
e
rivative
gain
s
of the
PID
controlle
r of
i
th
area.
min
,
min
,
min
,
,
,
i
d
i
i
i
p
K
K
K
and
max
,
max
,
max
,
,
,
i
d
i
i
i
p
K
K
K
are th
e lo
wer boun
ds
and
uppe
r
boun
ds of the
PID controll
e
r
.
3. Test Sy
stem
The p
r
op
ose
d
AGC
sche
me for
a mul
t
i-are
a
po
we
r system, d
e
scrib
ed i
n
the
previou
s
se
ction, ha
s
been te
sted o
n
a 39
-bu
s
New Engl
and
system [29] an
d on a 7
5
-b
u
s
India
n
syst
em
[30]. The 39-bus
system
has b
een div
i
ded into
two
control a
r
e
a
s and the 7
5
-
bu
s syste
m
into
four cont
rol area
s.
F
o
r 3
9
-bu
s
syste
m
s,
th
ree
Di
scos an
d at
lea
s
t on
e
Gen
c
o
have
bee
n
con
s
id
ere
d
in
ea
ch
are
a
.
The n
u
mb
er
of Gen
c
o
s
a
nd
Disco
s
in
the 39
-bu
s
system an
d in
the
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Multi-Are
a
Autom
a
tic Gene
ration Control
Sc
hem
e incl
uding
Rene
wable… (S
and
eep Bhon
gad
e)
5059
75-b
u
s sy
ste
m
is give
n in
Table
s
1
an
d 2, re
sp
ecti
vely. A general pu
rpo
s
e
G
o
verno
r
- Tu
rbine
model ha
s be
en used, whi
c
h is take
n fro
m
[31].
The comme
rcial multime
g
a
watt varia
b
l
e
sp
eed
wind
turbine
of rat
i
ng 1.5M
W d
e
velope
d
by Gene
ral E
l
ectri
c
(GE) h
a
s b
een
co
nsidere
d
in thi
s
study. A win
d
farm
con
s
ist
s
of
WTG
uni
t
s
of rating
1.5
M
W e
a
ch ha
s be
en
co
nsi
dere
d
in
are
a
-
2 of 3
9
-bu
s
system
and i
n
area-4 of 7
5
-bu
s
system. In
3
9
-bu
s
syste
m
, one PV
system of ratin
g
2MW an
d 4
M
W of
rating
, one PV
system
has b
een
con
s
ide
r
ed in 7
5
-bus
system.
Table 1. Co
ntrol Area
s in 3
9
-Bu
s
Powe
r
System
Control Are
a
Area Rating(
MW) Market
Participants
AREA-1
400
Genco
1,2,3,
4,5,
PV Genco-1,
Discos 1,2,3
AREA-2
500
Genco
6,7,8,
9,10
,
WTG-1,Discos 5,6,7
Table 2. Co
ntrol Area
s in 7
5
-Bu
s
Powe
r
System
Control Area
Area Rating(
MW) Market
Participants
AREA-1
460
Genco
1,2,3
AREA-2
994
Genco
4,5,6,
7,8,
P
V
GE
NC
O-2
AREA-3
400
Genco
9,10,
AREA-4
4470
Genco 11,1
2,13,
14,15, WT
G-2
The typical p
r
ofile of sola
r radiation a
n
d
wind
spee
d
,
for both the system
s, wa
s take
n
from the Nati
onal Re
ne
wa
ble Energy L
aboratory we
bsite [32], recorde
d
of May, 2010, as sho
w
n
in Figu
re
4(a
)
& 5
(
a) Power o
u
tput a
s
given in
Eq
uation
(4) ha
s b
een fin
d
o
u
t usi
ng the
PV
system
and
WTG
sy
stem
model
given
in
se
cti
on
(2
.2). The
re
al
power
output
of PV an
d
WTG
system i
s
ta
ken in
per unit
depe
ndin
g
u
pon the
ar
ea
rating. Th
e typical
re
al po
wer outp
u
t of PV
and WTG sy
stem in p.u. is sho
w
n in Fig
u
re 4
(
b) & 5(b).
Figure 4(a
)
. Real Irradi
an
ce Data used
for the Propo
sed AG
C Mo
del
0
200
400
600
800
1000
1200
1
96
191
286
381
476
571
666
761
856
951
1046
1141
1236
1331
1426
1521
1616
1711
Insolation
(W/m
2
)
Time(sec.)
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 7, July 201
4: 5052 – 50
70
5060
Figure 4(b
)
. Powe
r Output
of PV System of
area-1 of 39-b
us Syste
m
for 30 min
Figure 5(a
)
. Real
Wind Sp
eed Data use
d
for the Prop
ose
d
AGC Scheme
0
20
0
40
0
60
0
80
0
1000
1200
1
400
16
0
0
18
0
0
0
0.
2
0.
4
0.
6
0.
8
1
1.
2
1.
4
1.
6
1.
8
2
P
o
w
e
r
gen
era
t
e
d
i
n
P
V
s
y
s
t
em
o
f
ar
ea-
1
Ti
m
e
(
s
e
c
)
po
w
e
r
p
r
od
uc
ed
b
y
P
V
s
y
s
t
e
m
(
M
W
)
0
2
4
6
8
10
12
14
16
18
123456789
1
0
1
1
1
2
1
3
1
4
1
5
1
6
1
7
1
8
1
9
2
0
2
1
2
2
2
3
2
4
wind
speed
(m/s)
Time(Hour
of
day)
Re
a
l
wind
speed
da
t
a
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Multi-Are
a
Autom
a
tic Gene
ration Control
Sc
hem
e incl
uding
Rene
wable… (S
and
eep Bhon
gad
e)
5061
Figure 5(b
)
. Powe
r Gen
e
rated by WTG S
y
stem
of area
-2 of 39-bu
s System for 3
0
min
4. Simulation
Results
To an
alyze t
he effect
of
RES units o
n
fr
eq
uen
cy regul
ation, si
mulation stud
ies
h
a
ve
been
carried
out on
the
p
r
opo
sed
AG
C mod
e
l fo
r t
w
o
ca
se
s. F
o
r first
ca
se,
AGC mo
del
ha
s
been con
s
id
e
r
ed without RES
units
and for
se
co
nd ca
se
with RES
units. It is assumed that
RES
units i
n
cl
uded in the
network
will
suppl
y
the maxi
m
u
m output av
ailable at that parti
cular ti
me.
Data sou
r
ce
s
for sola
r
rad
i
ations and
wind sp
eed
has bee
n
ta
ken
from
National Ren
e
wa
ble
Energy La
bo
ratory we
bsite
[33]. If
the solar radiat
io
n
s
for 3
0
minut
e are of the f
o
rm a
s
sho
w
n in
Figure 4(a), then real po
wer outp
u
t
of PV system in
MW for a
r
ea
-1 of 39
-bu
s
system
will b
e
of
the form a
s
shown in Figu
re 4(b
)
. Simila
rly, if
wind sp
eed
s for 30
min are
of the form a
s
sh
own
in Figure 5(a
)
, then real po
wer o
u
tput of WTG sy
ste
m
s in MW fo
r area
-2 of 39
-bu
s
sy
stem will
be of the form as shown in Figure 5(b
)
.
In this wo
rk it is assu
med t
hat the RES
uni
ts a
r
e o
w
n
ed by the sy
stem ope
rator.
PV and
WTG
sy
stem
s give it
s full
output
whe
n
e
ver it i
s
req
u
ired,
whil
e t
he
conventio
nal g
ene
rators
cha
nge th
eir power
as
per th
eir p
a
rticipation fa
ctors
(pfs). T
he pf for
ea
ch g
ene
rato
r is
determi
ned by utilizing their bids submi
tted to
sy
stem operator [
16]. Th
e results for different
system
s are descri
bed b
e
l
o
w:
4.1. 39-Buse
Sy
stem
To sim
u
late t
he 39
-bu
s
system, it is a
s
sume
d that t
he ge
nerators an
d the lo
ads
are
partici
pating
in the fre
que
n
c
y re
gulatio
n
marke
t. Th
e Gen
c
o
s
an
d Di
scos bid
s
for
area-1
and
area
-2
were assume
d as
given in Tabl
e 3 and 4.
Table 3. Gen
c
o
s
and
Discos Bids in Area-1 of 39
-Bu
s
System
Gencos/Discos Price(Rs./KWh)
Capacit
y
(
MW)
Genco1
4.9
20.0
Genco2
5.1
15.0
Genco3
6.1
15.0
Genco4
5.4
25.0
Genco5
5.2
10.0
Disco1 5.2
4.0
Disco2 5.7
4.0
Disco3 6.1
4.0
0
20
0
40
0
60
0
80
0
10
00
12
00
14
00
16
00
18
00
0
5
10
15
20
25
P
o
w
e
r
g
ene
r
a
t
e
d i
n
W
T
G s
y
s
t
em
of
a
r
e
a
-
2
Ti
m
e
(
s
e
c
)
po
w
e
r
pr
oduc
ed
b
y
W
T
G
s
y
s
t
em
(
M
W
)
Evaluation Warning : The document was created with Spire.PDF for Python.