TELKOM
NIKA
, Vol.10, No
.1, March 2
0
1
2
, pp. 33~4
2
ISSN: 1693-6
930
accredited by DGHE (DIKTI
), Decree No: 51/Dikti/Kep/2010
¢
33
Re
cei
v
ed Juli
18
th
, 2011; Revi
sed Septe
m
ber 15
th
, 2011; Accepted
Jan
uary 5
th
, 2012
Demand Shifting Bidding in a Hybrid System with
Volatile Wind Power Generation
D.K.
A
g
ra
w
a
l, N.P.
Patidar, R.K.
Nema
Dep
a
rtment of Electrical E
ngi
neer
ing,
MANIT, Bhopal, India
e-mail: dk
agra
w
a
l
_
2
0
0
@
y
a
h
oo.co.in ,n
ppati
dar@
y
a
h
o
o
.co
m
, rk_nema@
yaho
o.com
Abs
t
rak
Mekan
i
sme “
p
rice res
p
o
n
sive
de
ma
nd
shifti
ng b
i
d
d
in
g
”
d
i
bah
as se
ba
gai
solus
i
a
l
tern
atif unt
u
k
me
na
nga
ni i
n
termitansi
dal
a
m
pe
mba
ngkit
tenaga
ang
in
. Makalah i
n
i
me
ng
usulk
an
sebu
ah for
m
u
l
asi
persa
maan
p
e
ngur
ang
an
h
a
rga
dan
p
e
m
bat
asan
e
m
is
i ek
o
n
o
m
i
de
ng
an
a
ksentuas
i p
a
d
a
inte
grasi
ten
a
g
a
angin. Analisis
ini didas
arkan dat
a pembangkitan sistem
uji bus I
EEE 30 pada pembangk
it konv
ens
iona
l
dan ten
a
g
a
angi
n sela
ma peri
ode 2
4
ja
m. Hasi
l pen
elitia
n men
unj
ukkan b
ahw
a
pend
ekata
n
ya
n
g
dius
ulka
n da
pa
t mere
duksi h
a
r
ga da
n men
a
n
gan
i inter
m
ita
n
s
i dal
a
m
pe
mb
angk
it tenag
a ang
in.
Ka
ta
k
unc
i
: pe
mb
an
gkit tena
ga an
gi
n, pr
ice
respons
ive d
e
m
a
nd sh
ifting b
i
ddi
ng, siste
m
uji b
u
s IEEE 30
A
b
st
r
a
ct
Price resp
onsi
v
e de
ma
nd sh
i
fting bid
d
i
ng
mecha
n
is
m is d
i
scussed as
an
alternativ
e sol
u
tion t
o
dea
l w
i
th intermittency i
n
w
i
nd gen
eratio
n. T
h
is pap
er
pro
poses a for
m
u
l
ation of
soci
al
w
e
lfare equ
ati
o
n
w
i
th price res
pons
ive d
e
m
a
nd shifti
ng b
i
ddi
ng a
nd ec
ono
mic e
m
issi
on dis
patch
w
i
th emp
has
is
o
n
integr
ation
of wind power.
The analys
is is based on
the IEEE 30 bus test system
generation data, wit
h
conve
n
tio
nal a
nd w
i
nd g
e
n
e
r
a
tion p
l
a
n
t over a per
i
od of
24 ho
urs. It
has be
en d
e
m
onstrated th
at th
e
prop
osed
appr
oach l
e
a
d
s to reducti
on i
n
e
m
i
ssion as w
e
ll d
eal w
i
th inter
m
i
ttency in w
i
nd gen
eratio
n.
Ke
y
w
ords
: IEEE 30 bus test system, price r
e
spo
n
sive
de
mand sh
ifting b
i
d
d
in
g, w
i
nd pow
er gen
erati
o
n
1. Introducti
on
Due to enviro
n
mental an
d energy
se
cu
ri
ty benefits there is a p
o
siti
ve shift towards the
prod
uctio
n
of
ele
c
tri
c
al
en
ergy from
re
newable
source
s of
en
erg
y
esp
e
ci
ally from
win
d
whi
c
h
are cl
ean a
n
d
abun
dantly
available in
nature.
On
regul
atory si
de in India
and many ot
her
cou
n
trie
s, there a
r
e ne
cessities to g
enerate a
certain am
ou
nt of electri
c
al en
ergy from
rene
wa
ble
so
urces.
China
is the
co
untry with th
e la
rgest i
n
stall
e
d
win
d
power
cap
a
city in
th
e
worl
d at the end of year 2
010 wh
erea
s India’s total
installe
d wind
powe
r
ca
pa
city is fifth in the
worl
d. It is reported by the
Global
Wind
Energy C
oun
cil (G
WEC) that global inst
alled win
d
po
wer
cap
a
city in
creased
by 24
.1% duri
ng t
he yea
r
a
n
d
stan
ds at 1
97.0 G
W
i
n
2010
[1]. Large
cap
a
city win
d
powe
r
gen
erators
are
con
necte
d to
t
r
a
n
smi
ssi
on o
r
sub
-
t
r
a
n
smi
s
sion
sy
st
em
s.
At
the end of 2
010, India h
ad 13.1 G
W
of insta
lled
wind
cap
a
cit
y
, with 40%
operating in
th
e
south
e
rn
state of Tamil Na
du and
wind
power pote
n
ti
al estimated
by the Centre
for Wind Ene
r
gy
Tech
nolo
g
y (C-WET) i
s
49
.13 GW [2].
The ge
neration of ele
c
tri
c
po
we
r by conv
e
n
tional
sou
r
ces p
r
odu
ce
s main
ly sulfur
dioxide, ca
rb
on, NOx, an
d mercury e
m
issi
on
s ca
u
s
ing a
c
id rai
n
, urban
smo
g
, and event
ually
global
climat
e ch
ange i
n
addition to
posi
ng si
gn
if
icant he
alth risks.
Ren
e
wabl
e ele
c
tri
c
ity
gene
ration
s
mainly from
Wind
farm
s
h
e
lp to
prev
en
t relea
s
e
of
e
m
issi
on
s into
the atm
o
sph
e
re
preventin
g e
n
vironm
ent damage. On t
he other h
a
n
d
, unpre
d
icta
ble, intermittent and vola
tile
nature of
win
d
energy ma
y threat
en po
wer
system
chara
c
te
risti
c
s such as volt
age
s, freque
ncy
and ge
neratio
n adeq
ua
cy whi
c
h can pot
entially enl
arge the we
akn
e
ss of
power
system
s.
Dema
nd
sid
e
ma
nage
m
ent (DSM) i
n
co
rpo
r
ate
s
energy effici
ency
(EE),
Energy
Con
s
e
r
vation
(EC)
and
De
mand
Respo
n
se
(DR). In
the m
o
st
ele
c
tricity ma
rkets; the
con
s
um
ers
play a
mu
ch
more limite
d
rol
e
tha
n
p
r
odu
cers. It i
s
wid
e
ly a
c
kn
owle
dged
tha
t
a mo
re
a
c
tive
partici
pation i
n
the ma
rket
by the dem
an
d sid
e
could
have si
gnifica
nt benefits [3]. A good d
eal
of
resea
r
ch ha
s been
re
port
ed on
mea
s
urem
ent of l
oad el
asti
city, pre
s
umin
g
that indu
stria
l
,
resi
dential a
n
d
comme
rcial
consume
r
s
will re
spo
nd to price
sign
als [3]-[8]. DR at end user’s
Evaluation Warning : The document was created with Spire.PDF for Python.
¢
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 10, No. 1, March 2
012 : 33 – 42
34
premi
s
e
s
ca
n
re
du
ce
gen
eration,
tran
smissi
on
and
distrib
u
tion
capa
city of uti
lity at relativ
e
ly
fraction
al cost as compa
r
e
d
to investm
ent re
q
u
ire
d
to provide
ne
w capa
city. The ap
pro
p
riat
e
deman
d man
ageme
n
t in
mode
rn inte
rcon
ne
cted po
wer
syste
m
with dispe
r
se
d gene
ratio
n
may
also
result in
redu
ce
d a
s
so
ciated
CO2
e
m
issi
on
s in
d
a
y-ahe
ad
ele
c
tri
c
ity marke
t
s throug
h p
r
i
c
e
respon
sive
d
e
mand
shiftin
g
bid
d
ing
(P
RDS
). Su
an
d Kirschen
[9
] pro
p
o
s
ed
th
e PRDS
bidd
ing
for ma
rket
cl
earin
g m
e
ch
anism
of
da
y-ahea
d m
a
rkets. P
R
DS
biddin
g
q
uan
tify the dem
and
respon
se i
n
day-ah
ead m
a
rket, and
so
me re
spo
n
siv
e
cu
stome
r
s
are a
b
le to shift the dema
n
d
from p
e
rio
d
s
of high l
o
cational m
a
rgi
nal
pri
c
e
(LMP
)
to the p
e
rio
d
s
of l
o
w
LMP
s
. However, t
h
e
market cle
a
ri
ng me
chani
sms devel
ope
d in [9] do
not take into
accou
n
t the operational a
n
d
se
curit
y
c
o
n
s
t
r
aint
s of
t
r
a
n
smi
ssi
on n
e
t
wor
ks.
Kan
w
ardee
p Sin
gh et al. [10] discu
ssed t
he
influen
ce of PRDS bid
d
ing
on co
nge
stio
n and LMP in
Pool-Based
Electri
c
ity Market
s. Impact
s
of
availability based tariff on
wind p
o
wer trading o
p
tion
were analy
z
e
d
in [11].
This pa
per in
vestigate
s
th
ese
un
match
ed
chall
eng
e
s
cau
s
e
d
by
wind
po
we
r
plants to
the optimization problem.
PRDS bid,
e
m
issi
on
s con
s
traint
s, and f
uel co
sts
are
con
s
id
ere
d
in
the
reali
z
ation
of most favora
ble gen
eration mix
for a system
with
wind p
o
wer
gene
ration. F
uel
co
sts, enviro
n
mental cost
s and emi
ssi
ons a
r
e
co
n
s
ide
r
ed in th
e implement
ation of optimal
gene
ration m
i
x for a system with win
d
gene
ration
along with
PRDS bid to
maximize social
welfare. The
rest of the p
aper
i
s
orga
nize
d as foll
ows: Next
se
ction 2 de
scribes
wind p
o
w
er
scena
rio redu
ction an
d PRDS bidi
ng m
e
cha
n
ism. Se
ction 3 de
scri
b
e
s research
method. Results
are p
r
e
s
ente
d
and di
scussed in se
ction
4.
Finally, section 5 con
c
lu
des the p
ape
r.
2. The Propo
sed Me
thod
The
win
d
p
o
w
er p
a
rtici
pat
ion into
total
prod
uctio
n
of
electri
c
al
en
e
r
gy d
epen
ds
upon
the
forecast
of
wi
nd mo
mentu
m
. A pri
n
ci
pa
l difficulty wit
h
mod
e
ling
wind p
o
wer
produ
ction i
s
th
at
the relation
sh
ip of
wind
sp
eed to
wi
nd
p
o
we
r p
r
o
d
u
c
tion i
s
extreme
l
y nonlin
ear.
The
wind
po
wer
gene
rato
rs require
no
fo
ssil so
urce
s hen
ce,
the
o
peratio
nal
co
st of win
d
u
n
its ha
s b
e
e
n
assume
d to be ze
ro. Diff
erent forecasting approa
ches avail
able
can be
stud
ied in [12]-[1
3
].
Re
sult of win
d
power
unp
redi
ctability can be
st
udie
d
by applyin
g
different
scenari
o
s i
n
to the
model.
M
ont
e
Carlo
si
m
u
lation wa
s popul
ari
z
ed
by
scienti
s
t in
the 195
0
s
.
Mo
nte Carlo
simulatio
n
is
a method tha
t
can mod
e
l thou
san
d
s
of
scena
rio
s
an
d help
s
to mo
del un
certai
nties
of win
d
p
o
we
r o
u
tput. It provides a
ra
n
ge of
po
ssibl
e
out
com
e
s
togethe
r with there
proba
bi
lity.
Modelin
g all
main an
d po
ssible
scen
ario
s dete
r
min
ed
by the un
ce
rtain varia
b
le
s
can
be d
one
by
Monte Carlo
simulatio
n
. These scen
ari
o
s a
r
e def
in
e
d
by the pro
bability distri
bution
s
and t
heir
simulation parameters. M
a
ny types
of probability distri
butions
are used in
dif
f
erent situations
su
ch a
s
norm
a
l, uniform an
d triangul
ar di
stributio
ns.
There are m
any sam
p
ling
techni
que
s
su
ch
a
s
Imp
o
rtan
ce
sam
p
ling, Sobol
numbe
rs
sampli
ng, Mi
dpoint
sam
p
li
ng, Latin
hyp
e
rcube
sampl
i
ng (LHS
), an
d L
H
S Mo
nte
Ca
rlo
sa
mpli
ng
to eliminat
e
scenarios with very
low
probability. T
hese techniques are en
gaged to
reduce t
h
e
comp
utationa
l req
u
irem
ent
to simul
a
te
larg
e
num
b
e
r of
scena
ri
os. L
H
S ha
s the be
nefit of
gene
rating a
set of strati
fied sampl
e
s that more pre
c
isely refl
ect the sh
ap
e of a samp
led
distrib
u
tion a
nd red
u
ces t
he numb
e
r o
f
runs. The
g
eneral effect is that the mean of a set of
simulatio
n
re
sults mo
re
q
u
ickly ap
pro
a
c
he
s th
e
‘tru
e’ value,
particularly fo
r
model
s that
are
simply ad
din
g
or
subtract
ing a nu
mbe
r
of
variabl
e
s
. The trade
off betwee
n
the numb
e
r
of
redu
ce
d scen
ario
s and the
simulation p
r
eci
s
ion i
s
po
ssible by choo
sing the n
u
m
ber of re
du
ce
d
scena
rio
s
so
that the obje
c
tive function
woul
d not
ch
ange m
u
ch o
r
the rel
a
tive d
i
stan
ce b
e
twe
e
n
origin
al
scen
ario
s
and
re
d
u
ce
d
scena
ri
os i
s
within
a
n
a
c
ceptabl
e
level [14], [15
]. In this
mod
e
l
the numbe
r o
f
reduced sce
nario
s is cho
s
en to be ten
since the valu
e of objective
function at this
numbe
r
doe
s not
cha
nge
much.
The
scenari
o
s taken
we
re
having
the hi
ghe
r p
r
obabilitie
s. T
he
deviation
s in
wind
po
wer
h
a
ve bee
n taken in to a
c
co
unt by co
nsi
d
ering
differe
n
t
sce
na
rio
s
. The
forecaste
d
wi
nd po
wer g
e
n
e
rated a
nd re
duced sce
narios data a
r
e take
n from [16
].
Not all consumers have the facility or the in
centive to adjust thei
r
demand when pri
c
es
cha
nge. La
rg
e part of the power requi
rement will th
er
efo
r
e totally inelasti
c. In price takin
g
b
i
ds,
the dem
and
aggregato
r
i
s
rea
d
y to a
c
cept
a spe
c
ified am
ount o
f
powe
r
at prevailing m
a
rket
price, an
d its power
co
nsumption
rem
a
ins co
nsta
nt
irre
sp
ective of
variation
s
in
market pri
c
e.
This
kind
of bid is
req
u
i
r
ed to m
eet
necessa
ry
d
a
ily servi
c
e
s
to indu
strial
, resid
ential
and
dome
s
tic loa
d
s. In
pri
c
e
re
spo
n
sive
bid
s
, the p
r
ice to
be p
a
id
by a
bidde
r d
e
cre
a
s
e
s
con
s
i
s
ten
t
ly
decrea
s
in
g
with re
sp
ect to
incre
a
se in
power
u
s
e. Details of
PRDS
biddi
ng scheme
ha
s be
en
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
¢
Dem
and Shifting Bidding in
a Hyb
r
id System
with Volatile Wind Power …. (D.K. Agra
wal
)
35
formed in [9]
.
In PRDS bids, an ag
gregator
on
be
half of con
s
u
m
ers is abl
e
to incre
a
se or
decrea
s
e
its
certai
n p
e
rce
n
tage of
dem
and in
re
sp
o
n
se to
ma
rke
t
price. Key factor in P
R
DS
biddin
g
sche
me is p
r
ice resp
on
sive co
nsum
er
wh
o can tran
sfer i
t
s dema
nd from pea
k d
e
m
an
d
perio
ds of hig
h
market pri
c
e to off peak deman
d pe
rio
d
s in which market pri
c
e is
some
wh
at low.
The sco
pe o
f
present pa
per i
s
to stu
d
y the influe
nce of P
R
DS bidding in
a hybrid
system
with v
o
latile Wi
nd
Powe
r Ge
neration in
d
a
y-a
head m
a
rket
s.A usual P
R
DS bid is
sh
own
in [10]. Main
dissimila
rity betwee
n
PRDS bidding
sch
e
me an
d p
r
ice re
spo
n
sive
deman
d bid
d
i
ng
is that i
n
P
R
DS bid
d
ing
d
u
ring
a
pa
rticular
th
t
pe
riod,
maximum d
e
mand li
mit ca
n be
greate
r
than earli
er, to inclu
de the
fall of load occurre
d
duri
ng pea
k pe
ri
ods d
ue to soarin
g ele
c
tri
c
ity
rate, whe
r
ea
s in
pri
c
e
re
spo
n
sive
bid
d
ing
schem
e
loss
of loa
d
duri
ng
pea
k hou
rs can
not
recovered in
off peak
hou
rs. In PRDS
b
i
dding
agg
reg
a
tor’s
on b
e
h
a
lf of respon
sive con
s
um
ers
specifies for
particular
th
t
period, its
maxi
mum, an
d mi
nimum
pri
c
e
bids an
d m
a
ximum p
o
wer
deman
d which in sim
p
le
st form can b
e
sum of e
n
tire ene
rgy ne
e
d
of re
spo
n
si
ve part. Du
e
to
negative sl
op
e of PRDS bi
dding
re
spon
sive pa
rt
of demand
ca
n b
e
less than it
s maximum v
a
lue
becau
se
ag
gregator wo
uld accept
o
n
ly that part of d
e
m
and fo
r whi
c
h its
willing
price is l
e
ss t
han
or e
qual
to m
a
rket de
cla
r
e
d
pri
c
e. In
PRDS bid
d
ing
a
ggre
gato
r
’s p
r
ice
respon
si
ve part
of en
e
r
gy
of sch
eduli
n
g
period
can b
e
con
s
um
ed i
n
few su
b-in
t
e
rvals. In sim
p
lest form it can be con
s
u
m
ed
in a singl
e pe
riod. Mathem
atically, it can
be rep
r
e
s
ent
ed as:
,m
a
x
,
,
0
kt
kt
RS
RS
DD
t
T
≤≤
∀
∈
(1)
,
,
kt
RS
R
S
k
tT
Dt
E
∈
⋅Δ
≤
∑
(2)
ma
x
,
,
,
kt
RS
RS
k
DE
t
t
T
=Δ
∀
∈
(3)
whe
r
e:
,
kt
R
S
D
Consumption of
th
k
demand shifting consumer at peri
od
t
;
max
,
,
kt
R
S
D
Maximum consu
m
ption of
th
k
deman
d shifting consu
m
er at pe
riod
t
;
,
Tt
∀
Set of scheduling sub-intervals and
duration of o
n
e
sub-interval;
,
R
Sk
E
Maximum limit on energ
y
consum
ed under
deman
d shifting bid of
th
k
DistCo during en
tire
scheduling period;
Price
taki
ng
consume
r
s ha
ve infinite ma
rginal
value due to vertic
al
c
u
rve with res
p
ec
t to
power
so it
can
not be i
n
clu
ded in
consume
r
gro
ss
su
rplu
s. In optimizatio
n pro
b
lem g
r
oss
surplu
s of
thi
s
type
co
nsu
m
ers a
r
e
not
inclu
ded
an
d assume
d con
s
tant.
Th
us
e
quation
s
(4
) and
(5) sho
w
s
ho
w the
consu
m
er
gro
s
s
su
rplu
s i
s
cal
c
u
l
ated b
a
se
d
on the
acce
p
t
ed dem
and
-side
bids a
nd the
margi
nal valu
e that con
s
u
m
ers attach t
o
these bi
ds:
,,
,
,
,
1
RS
kj
t
k
j
t
RS
J
kt
j
GS
MB
D
=
⋅
=
∑
(4)
,,
,
1
RS
RS
kt
k
j
t
J
j
DD
=
=
∑
(5)
whe
r
e:
,,
kj
t
R
S
MB
Marginal benefit
of
th
j
segment of
th
k
demand shifting bidder du
ring
th
t
period;
J
Number of seg
m
ent of the bid of
bidder;
,,
RS
kj
t
D
Demand of
th
k
bidder during
th
t
hour on
th
j
segment of its bid;
,
kt
GS
th
k
demand shifting consumer gross
surplus at period
t
;
In this model
certain pe
rcentage of po
wer for e
a
ch
hour whi
c
h i
s
pri
c
e takin
g
can be
pro
c
u
r
ed by con
s
um
ers regardle
s
s
of
market-c
le
aring p
r
ices. F
o
r a
co
mplet
e
and
comp
act
formulatio
n of the PRDS bidding m
e
ch
ani
sm th
e rea
der i
s
referred to
[9], [10].
In the
Evaluation Warning : The document was created with Spire.PDF for Python.
¢
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 10, No. 1, March 2
012 : 33 – 42
36
optimizatio
n prog
ram use
d
for
ma
rket-cle
ari
ng,
th
e bid spe
c
ification
s
a
r
e
transl
a
ted int
o
con
s
trai
nts:
(i) on the
d
e
mand
duri
n
g each pe
rio
d
and (ii
)
on
the total demand ove
r
the
optimizatio
n hori
z
on. Thi
s
last sp
ecification
i
s
impl
emented
a
s
an in
equality
rathe
r
th
an
an
equality con
s
traint be
ca
use a d
e
man
d
-side
bid b
e
lo
w the l
o
west
price at
whi
c
h
gen
erators are
willing to produce
woul
d otherwi
se prev
ent the
mark
et from
cleari
ng. The price responsive bids
are conve
r
ted
into a form suitable for mi
xed-integ
e
r li
near p
r
o
g
ra
m
m
ing su
ch a
s
in [9].
3. Resear
ch
Method
The obj
ective
is to maximi
ze the
so
cial
we
lfare, i.e., the differen
c
e between th
e value
that con
s
um
e
r
s atta
ch to the ele
c
trical
energy t
hat they buy and the co
st
of ge
neratio
n that has
been formulat
ed ba
sed on
cla
ssi
cal EL
D with emissi
o
n
. Equation (6) as p
r
op
osed in [9] is used
to co
nsi
d
e
r
e
c
on
omic loa
d
dispatch
with emi
s
si
on. T
he
p
r
o
p
o
s
ed
so
cial welfa
r
e
eq
uation
af
ter
mo
d
i
fic
a
tion
is
r
e
pr
es
en
te
d a
s
:
()
,,
,
,
11
1
,
g
N
TK
k
t
it
i
t
it
up
tk
i
it
G
M
a
x
G
S
u
FC
S
t
EEC
φ
⋅
==
=
⎛⎞
=−
+
+
⎜⎟
⎜⎟
⎝⎠
∑∑
∑
(6)
whe
r
e:
φ
Optimal social welfare;
,
it
F
C
Fuel cost of
th
i
gene
rator at time
t
;
,
it
G
u
Status of
th
i
generat
or at time
t
(1:on,
0:off);
,
it
EE
C
Blended emissio
n
cost of
th
i
generat
or at time
t
;
,
it
up
St
Start up cost of
th
i
generato
r
at time
t
;
g
N
Total numbe
r of
generato
r
s in the
net
w
o
rk;
K
Total numbe
r of
demand side bidders;
Fuel co
st of
th
i
g
enerator in te
rms of real po
wer o
u
tput
,
g
it
P
ca
n be expre
ssed as:
,
2
,,
$/
hr
it
i
g
it
i
g
it
i
FC
a
P
b
P
c
=⋅
+
⋅
+
(7)
whe
r
e:
,
gi
t
P
Real po
w
e
r
outp
u
t of an
th
i
generato
r
at time
t
;
, ,
ii
i
ab
c
F
uel cost cur
v
e coefficients;
The
th
i
gene
rato
r sta
r
tup
co
st is de
scrib
ed
in te
rm of the
numbe
r of h
ours the g
e
n
e
rato
r
has b
een d
o
w
n:
,
,
1e
x
p
it
off
it
i
i
i
up
H
St
κρ
τ
⎡⎤
⎛⎞
=+
−
−
⎢⎥
⎜⎟
⎝⎠
⎣⎦
(8)
whe
r
e:
i
κ
The
th
i
unit fixed start-up cost part
in $;
i
ρ
Start-up cost of
th
i
unit at
th
t
hour fr
om cold condition in $;
,
it
off
H
The numbe
r of h
our’s
th
i
unit do
w
n
a
t
th
t
hour;
i
τ
Rate of cooling o
f
th
i
unit
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
¢
Dem
and Shifting Bidding in
a Hyb
r
id System
with Volatile Wind Power …. (D.K. Agra
wal
)
37
In this
optimiz
a
tion
problem, for the
sak
e
of
sim
p
li
cit
y
,
t
he st
art
-
up co
sts are
take
n
con
s
tant, whi
c
h can be giv
en as:
(
)
,,
,
1
it
i
i
t
i
t
up
G
G
St
u
u
κ
−
=⋅
−
(9)
,
0
it
up
St
≥
(10
)
Emissi
on of
th
i
u
n
it can be ex
pre
s
sed a
s
:
,
2
,,
lb/
h
r
it
i
g
it
i
g
it
i
EC
P
P
αβ
γ
=⋅
+
⋅
+
(11
)
whe
r
e:
, ,
ii
i
α
βγ
are emission coefficients;
The fuel co
st curve
s
an
d emissi
on curve
s
of
the powe
r
plants a
r
e modele
d
as step-wi
se
linear fun
c
tio
n
, in o
r
de
r to
approximate t
he typica
l
qu
adrati
c
sha
p
e
d
cost
cu
rve
of a po
we
r pl
ant.
The d
ual-obj
ective combi
ned e
c
o
nomi
c
emi
s
sion
d
i
spat
ch p
r
obl
em is
co
nve
r
ted into
sin
g
le
optimizatio
n
probl
em by i
n
trodu
cin
g
p
r
ice
penalty fa
ctor
h
which
convert e
m
issi
on o
u
tput int
o
equivalent e
m
issi
on cost
as follo
ws:
,,
$/hr
it
it
EEC
h
E
C
=⋅
(12
)
The pri
c
e
pe
nalty factor
h merg
e the
emis
sion
wi
th fuel co
st and sum of
merg
ed
emis
sion
c
o
st
,
f
uel c
o
st
an
d st
a
r
t
u
p
co
st
giv
e
s t
o
tal
prod
uctio
n
co
st in
$/hr [17]
. Modified
pri
c
e
penalty fa
cto
r
ap
proa
ch
wa
s p
r
op
ose
d
in [1
8] to
give exa
c
t to
tal ope
rating
co
st. The
p
r
ice
penalty facto
r
i
h
is the ratio between
maximum fu
el co
st an
d
maximum e
m
issi
on of
th
i
gene
rato
r:
,m
a
x
,m
a
x
$
/
l
b
1
,
2
,
3
.
....
...
ii
ig
hF
C
E
C
i
N
==
(13
)
whe
r
e:
,m
a
x
i
FC
Fuel cost of an
th
i
unit at maximum p
o
wer out
put;
,m
a
x
i
EC
Emission of an
th
i
u
n
it at maximum p
o
wer out
put;
To find out the modified
price pen
alty factor for a particula
r loa
d
deman
ds f
o
llowin
g
step
s are p
r
o
posed in [18].
•
The pri
c
e p
e
n
a
lty factor
i
h
is cal
c
ulate
d
.
•
Price p
enalty factors
i
h
are arrang
ed in a
s
cendin
g
ord
e
r.
•
Add the maximum ca
pa
city of each un
it
,m
a
x
gi
P
one at a time, starting
from the sm
allest
value
i
h
until
,m
a
x
g
id
PP
≥
∑
.
•
Then the mo
dified pri
c
e p
enalty factor
m
h
is com
puted
by interpolati
ng the value
s
of
i
h
for
the last two u
n
its by satisf
y
i
ng the co
rrespondi
ng load
deman
d.
In this pa
per
shifted p
r
ice
penalty facto
r
appr
oa
ch i
s
use
d
to a
cco
mmodate l
o
a
d
shifting
behavio
r. The
inequality co
nstrai
nt on re
al power ge
n
e
ration
,
g
it
P
of
th
i
generato
r
is:
,m
i
n
,
,
m
a
x
gi
gi
t
g
i
PP
P
≤≤
(14
)
Evaluation Warning : The document was created with Spire.PDF for Python.
¢
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 10, No. 1, March 2
012 : 33 – 42
38
whe
r
e:
,m
i
n
gi
P
Minimum val
ue of real po
wer g
ene
ratio
n
of
th
i
generato
r
;
,m
a
x
gi
P
Maximum value of real po
wer g
ene
ratio
n
of
th
i
generato
r
;
If a unit m
u
st
be
“on
”
fo
r a
ce
rtain
num
ber
of
ho
urs before
it
can be shut do
wn
,
then a
minimum up-time (
i
UP
T
) is im
posed. On t
he contra
ry, the minimum
down-time
(
i
D
N
T
) is the
numbe
r of hour(s) a u
n
it must sta
y
off-line before it can
be brou
ght
on-line a
g
a
in.
Mathemati
c
al
ly, the minimum up/do
wn time con
s
train
t
s for
th
i
unit ca
n be expre
ssed as:
()
(
)
,1
,
,
1
0
ii
t
i
t
i
t
UP
in
G
G
TH
u
u
−−
−⋅
−
≥
(15
)
()
(
)
,1
,1
,
.0
ii
t
i
t
i
t
DN
of
f
G
G
TH
u
u
−−
−−
≥
(16
)
whe
r
e:
,
it
in
H
Amount of time
th
i
unit has b
een ru
nnin
g
;
Equation
s
(1
3), (14
)
are n
online
a
r. It c
an be line
a
ri
zed usi
ng the
method p
r
e
s
ented by
Cha
ng et al. [20]. The UC sche
dule
should p
r
ov
id
e the exact
amount of p
o
we
r to meet
the
con
s
um
er’
s
d
e
mand. Th
erefore:
,,
11
1
1
,,
g
RS
T
N
NW
K
M
kt
m
t
in
w
k
m
gi
t
n
w
t
P
PD
D
==
=
=
+=
+
∑∑
∑∑
(17
)
whe
r
e:
,
T
mt
D
Con
s
um
ption
of
th
m
pri
c
e ta
kin
g
bidde
r at pe
riod
t
;
M
Maximum value of real po
wer g
ene
ratio
n
of
th
i
generato
r
;
,
P
nw
t
Fore
ca
sted g
eneration of
th
nw
wind p
o
wer u
n
it at time
t
;
NW
Numb
er of wi
nd po
wer u
n
its;
4. Results a
nd Analy
s
is
Mathemati
c
al
ly so
cial
welf
are
eq
uation
is a
de
cisi
o
n
p
r
oble
m
wi
th an
obje
c
ti
ve to b
e
maximize
d with re
spe
c
t to a seri
es
of prevailin
g
equality an
d ineq
uality con
s
trai
nts.
The
equatio
n is a
mixed-integ
e
r no
n linea
r probl
em an
d inclu
d
e
s
a
large
numb
e
r of integers
and
contin
uou
s variabl
es.
No
n linea
r pa
rt is co
nver
te
d into pie
c
e
w
ise linea
r functio
n
usi
n
g the
techni
que
given in [9]. The
market-clea
r
i
ng alg
o
rithm
descri
bed i
n
Section
-
3 h
a
s been
applie
d
to
several sce
n
a
rio
s
to asse
ss
eco
nomi
c
viability
of demand
shifting
and evalu
a
te its impa
ct on
emission
di
spatch
and
o
n
win
d
scen
ario
s. The
pl
atform u
s
ed
for the im
ple
m
entation
of this
prop
osed
ap
proa
ch
is on
INTEL[R], P
entium [R
]
4
CPU 3.06
GHz, 51
2 M
B
of RAM.
Many
c
o
mmerc
i
al pack
ages
s
u
c
h
as
CPLEX, LINDO,
OSL and XPRESS-MP exis
t in the mark
et place
have been suc
c
ess
f
ully applied to
UC
problems
.
In this
paper,
we
us
e XP
RESS-MP to
s
o
lve t
h
e
probl
em [19].
The tes
t
s
y
stem used in the
s
t
udies c
o
ns
is
ts of
IEEE 30 bus s
y
s
t
em with a total
capacity of 435 MW. The IEEE 30 bus
system ha
s
six generating units. The
characteristi
c
s o
f
gene
rato
rs,
u
n
it co
nstraint
s a
nd the
e
m
issi
on
co
efficient
s a
r
e
gi
ven in T
able
I. The maxim
u
m
and mi
nimu
m load
s
are
396.7
6
M
W
and
183.4
MW, respe
c
tively, while t
he total
syst
em
forecaste
d
en
ergy de
mand
is 693
4.76 M
W
h. T
he
stud
y period i
s
2
4
-
hou
rs. The 2
4
-ho
u
r
syste
m
load an
d fore
ca
sted wi
nd
power a
r
e p
r
ese
n
ted in
T
able 2. The
prop
ortio
n
of the dema
nd that
respon
ds to
prices affect
s the shap
e o
f
the
dem
and
cu
rve. Loa
d
parti
cipatio
n
factor (LPF
) is
defined a
s
th
e ratio of the price re
sp
on
sive demand t
o
the total possi
ble dem
an
d [9].
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
¢
Dem
and Shifting Bidding in
a Hyb
r
id System
with Volatile Wind Power …. (D.K. Agra
wal
)
39
The
simul
a
te
d wi
nd
po
wer scen
ario
s
are a
s
sumed
to
follow a
normal di
strib
u
tion
with a
standard
devi
a
tion (volatilit
y) of 10%
of
expected val
ues
whi
c
h i
s
t
he forecasted value and in
the
followin
g
stu
d
ies, th
e n
u
m
ber of redu
ced
sce
nari
o
s i
s
cho
s
e
n
t
o
be
ten, ta
ken fro
m
[16].
The
followin
g
thre
e ca
se
s are d
i
scusse
d in this pap
er:
Ca
se 1: Ca
se 1 is the b
a
se
ca
se wit
hout
PRDS biddin
g
whi
c
h applie
s 11
commitme
n
t and
disp
atch
for
the forecaste
d
wi
nd
po
wer
and
the other
ten si
mulated win
d
po
we
r
scena
rio
s
, without co
nsi
dering the
co
rrel
ation betwee
n
scena
rio
s
.
Ca
se 2: In this ca
se, we
obse
r
ve the im
pact of PRDS on sy
stem
operation
and comp
are
gene
ration di
spat
ch an
d total system op
erating
co
st with and
with
out PRDS.
Ca
se 3: De
monst
r
ate th
e relation
shi
p
betwe
en
L
P
F and emi
s
sion at different value of load
partici
pation f
a
ctor from 0 to 0.1.
4.1. Social
Welfare
w
i
th
out PRDS bidding
With
fo
re
ca
sted wind po
wer given
in Table
2, we solve
the so
cial welfare equatio
n
without in
clu
s
ion of dem
an
d shifting
and
determi
ne
th
e dispatch of
non-win
d
unit
s
given i
n
Ta
ble
3(a
)
. The
che
ape
st Unit 1
and 2 a
r
e
al
ways
com
m
itted. The mo
re expen
sive
Unit 3 an
d 4
are
committed
be
tween
Ho
urs
1 and
22.
Un
it 5 is
com
m
i
tted between
Hou
r
s 1
6
an
d 21.
Unit 6 i
s
committed
b
e
twee
n Hours 2
-
7 a
nd 1
5
-21. T
he
sy
stem g
ene
rat
i
on cost i
s
$
1608
0.296.
To
observe the
impact of
wi
nd po
we
r scenari
o
s,
we
solve ten
so
cial welfare
equatio
ns. T
h
e
operation
co
sts a
r
e
sho
w
n in Figu
re
1(a
)
wh
ich range from $
1618
5.804 fo
r Sce
nari
o
5
to
$158
37.47
5 for Scena
rio
9. Table
4(a). sho
w
s the
social
welfa
r
e
equatio
n solu
tions fo
r the
10
scena
rio
s
. Each value in
the table sho
w
s the num
b
e
r of times certain unit
s
are ON in the
10
scena
rio
s
. Here, Unit 3 and 4 are ON
mainly bet
we
en Ho
urs 1–
22, while
Uni
t
5 is ON ma
inly
betwe
en Hou
r
s 16
–21, whi
l
e Unit 6 is O
N
mainly bet
wee
n
Ho
urs 3-7 an
d Ho
urs 15–
21.
4.2. Social Welfare
w
i
th PRDS bidding
To ob
serve t
he impa
ct of deman
d shifti
ng we
ta
ke L
P
F 0.1. The load an
d win
d
profile
are
same
as Ca
se 1. Ta
ble 3(b). pre
s
ent
s the
di
spatch of n
o
n
-
win
d
unit
s
for 24
hou
rs.
The
system
gen
eration
co
st red
u
ce
s
to
$1
56
57.407.
In
all 10 scena
rio
s
operation co
sts
a
r
e sh
own
in
Figure 1
(
b
)
which
ra
nge
from $1
577
3.6
16 for Scena
rio 5 to
$15
33
0.454 fo
r S
c
e
nario
10. In
this
ca
se, for fore
ca
sted a
nd
a
ll ten sce
nari
o
s the
expe
n
s
ive unit
5 is not di
spat
ch
ed a
s
sho
w
n
in
Table 4(b). T
hese re
sults
sho
w
the lower co
st of
usi
ng PRDS bid
d
ing for su
pp
lying the load in
the system.
Gen N
o
1234
56
Max
(MW
)
200
80
50
35
30
40
Mi
n
(MW
)
50
20
15
10
10
12
γ
0
.
013
0.
02
0.
027
0.
0
2
9
0
.
0
2
9
0.
027
β
-
0
.9
-0
.1
-
0
.
0
1
-
0
.
0
1
-0
-0
.0
1
1
232.
4
4
4
13
238
84
α
2
2
.
9
8
25.
31
25.
51
24.
9
24.
7
2
5.
3
2
274.
4
70.
2
14
259
80
c
0000
00
3
320.
6
7
6
15
291.
2
7
8
b
21
.
7
1
3
.
2
5
3
3
4
373.
8
8
2
16
324.
8
3
2
a
0
.
004
0.
018
0.
063
0.
0
0
8
0
.
0
2
5
0.
025
5
396.
8
8
4
17
344.
4
4
Mi
n Up
T
i
m
e
(Hrs
)
1211
21
6
380.
8
8
4
18
337.
4
8
Mi
n Do
w
n
T
i
m
e
(Hrs
)
1212
11
7
344.
4
100
19
330.
4
1
0
S
hut
Dow
n
c
o
s
t
50
60
30
85
52
30
8
298.
2
100
20
315
5
Col
d
s
t
art
(Hrs
)
2111
11
9
268.
8
7
8
21
285.
6
6
I
n
i
t
i
a
l
uni
t
s
t
at
us
-1
-3
2
3
-
2
2
10
225.
4
6
4
22
254.
8
5
6
Hot
S
t
art
up c
o
s
t
70
74
50
1
1
0
7
2
4
0
11
205.
8
100
23
225.
4
8
2
C
o
l
d
S
t
art
up c
o
s
t
176
187
113
2
6
7
180
113
12
224
92
24
183.
4
5
2
T
able 1. Data f
o
r
IE
EE-30 bus
sy
stem
Load
(MW
)
W
i
nd
(MW
)
T
abl
e 2. Hourly
lo
ad and f
o
rec
a
st
w
i
n
d
power
Pe
r
i
o
d
Load
(MW
)
W
i
nd
(MW
)
Pe
r
i
o
d
Evaluation Warning : The document was created with Spire.PDF for Python.
¢
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 10, No. 1, March 2
012 : 33 – 42
40
S
y
s
t
em
oper
at
i
on c
o
s
t
i
n
t
e
n
s
c
enar
i
o
s
w
i
t
h
o
u
t
P
R
D
S
bi
ddi
ng
15600
15700
15800
15900
16000
16100
16200
16300
13579
S
c
enar
i
o
s
(A
)
Co
s
t
(
$
)
S
y
s
t
em
oper
at
i
o
n
c
o
s
t
i
n
t
en
s
c
enar
i
o
s
w
i
t
h
P
R
D
S
bi
ddi
ng
15000
15200
15400
15600
15800
16000
1
3
579
S
c
en
ar
i
o
s
(B
)
Co
s
t
(
$
)
T
o
t
a
l
em
i
s
s
i
on w
i
t
h
and w
i
t
h
o
u
t
w
i
nd pow
er
6200
7200
8200
9200
10200
0
0
.
02
0
.0
4
0
.
06
0
.0
8
LP
F
(C
)
T
o
t
a
l
em
i
s
s
i
on
(l
b)
W
i
th
o
u
t w
i
n
d
pow
er
W
i
t
h
w
i
nd pow
er
Figure 1. System operation cost
Table
3.
G
e
n
e
ration Dis
p
a
t
ch
(M
W) with
Fore
ca
sted Wind
Po
we
r
Ho
ur
U1
U2
U
3
U4
U5
U
6
Hou
r
U
1
U2
U3
U
4
U5
U6
1
100
4
0
26.67
21.73
0
0
1
100
40
2
6
.67
0
0
0
2
100
4
0
24.53
18.33
0
21.33
2
100
50.09
2
6
.67
0
0
0
3
100
6
0
26.67
27.27
0
30.67
3
100
40
2
6
.67
26.67
0
2
1.33
4
139.5
6
0
26.67
35
0
30.67
4
102.1
6
0
2
6.67
35
0
3
0.67
5
150
6
0
37.09
35
0
30.67
5
120.8
6
0
2
6.67
35
0
3
0.67
6
144.5
6
0
26.67
35
0
30.67
6
106.4
6
0
2
6.67
35
0
3
0.67
7
100
6
0
26.67
27.07
0
30.67
7
100
40
2
6
.67
26.67
0
2
1.33
8
100
44.8
7
26.67
26.67
0
0
8
100
40
2
6
.67
26.67
0
2
1.33
9
100
4
0
26.67
24.13
0
0
9
100
40
2
6
.67
26.67
0
2
1.33
10
100
36.
4
1
5
1
0
0
0
1
0
100
40
2
6
.67
26.67
0
2
1.33
11
60.8
2
0
1
5
1
0
0
0
1
1
100
40
2
6
.67
26.67
0
2
1.33
12
87
2
0
15
10
0
0
12
100
40
2
6
.67
26.67
0
2
1.33
13
100
2
9
15
10
0
0
13
100
40
2
6
.67
26.67
0
2
1.33
14
100
4
0
20.67
18.33
0
0
14
100
40
2
6
.67
26.67
0
2
1.33
15
100
4
0
26.67
25.2
0
21.33
15
100
40
2
6
.67
26.67
0
2
1.33
16
110.5
6
0
26.67
35
3
0
30.67
16
108
60
2
6
.67
3
5
0
3
0
.67
17
150
6
0
34.73
35
3
0
30.67
17
150
60
3
0
.29
3
5
0
3
0
.67
18
147.1
6
0
26.67
35
3
0
30.67
18
143.3
6
0
2
6.67
35
0
3
0.67
19
138.1
6
0
26.67
35
3
0
30.67
19
135
60
2
6
.67
3
5
0
3
0
.67
20
127.7
6
0
26.67
35
3
0
30.67
20
126.2
6
0
2
6.67
35
0
3
0.67
21
112.3
6
0
26.67
26.67
23.
33
30.67
21
107
60
2
6
.67
26.67
0
3
0.67
22
100
45.4
7
26.67
26.67
0
0
22
100
46.65
2
6
.67
0
0
0
23
100
43.
4
0
0
0
0
23
100
40
2
6
.67
0
0
0
24
100
31.
4
0
0
0
0
24
100
41.75
2
6
.67
0
0
0
(
a
)
w
i
tho
u
t
PRD
S
bidd
in
g
(
b
)
w
i
th
PRD
S
bidd
in
g
4.3. Social Welfare
w
i
th L
P
F
v
a
r
y
ing fr
om 0 to 0.1
In this sectio
n, we study the
sen
s
it
iv
it
y
of
social w
e
lf
are re
sult
s t
o
t
he size of
P
RDS
biddin
g
with
and without wind po
we
r.
For
th
e
same
system
we u
s
e
different P
RDS
with
sev
e
ral
LPF option
s
as
sho
w
n i
n
fig. 1(C). It ca
n be
see
n
th
at with wi
nd
power total e
m
issi
on
red
u
c
e
s
from 9
275.7
4
8
lb
to 6
628.
245 l
b
whe
n
we
d
o
n
o
t
consi
der PRDS. By increa
sing
the
si
ze
of
PRDS, the total emission fu
rther
redu
ce
s to 6449.473 l
b
.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
¢
Dem
and Shifting Bidding in
a Hyb
r
id System
with Volatile Wind Power …. (D.K. Agra
wal
)
41
Table 4. Statistics of Unit
Commitments
in Sc
enarios
5. Conclusi
on
A MIP base
d
soci
al
welf
are
problem
i
n
clu
d
in
g
win
d
and
PRDS b
i
dding
is de
scrib
ed i
n
this
paper. The case s
t
udies
based on
an IEEE 30
bus
s
y
s
t
em
generator data indicate that the
appli
c
ation o
f
PRDS can
impact the
pea
k load
re
ductio
n
, syst
em ope
rating
co
st, emissi
on
reduction, commitment
and dispatch
of the units
. Much of the benefits
listed here will depend on
the loa
d
p
a
rti
c
ipatio
n fa
cto
r
of P
R
DS bi
dding.
The
e
x
ample
on IE
EE 30
bu
s
system g
ene
rator
data sho
w
ed
the effectiven
ess of
the
propo
sed m
ode
l. The propo
sed alg
o
rithm
can
be u
s
e
d
f
o
r
the ope
ration
plannin
g
in the day-ahe
a
d
as
well
a
s
the long term
plannin
g
of wind u
n
its in
a
con
s
trai
ned t
herm
a
l po
wer system.
Referen
ces
[1]
Global Wind E
nergy
Counc
il. [Online
]. Available: http://
w
w
w.gw
ec.net/.
[2]
CWET, Govt
. o
f
India. [Online]
. Available h
ttp://
w
w
w
.
c
w
et.tn.
n
ic.in/html/departments_
w
r
a.html
[3]
DS
Kirsch
en. Dema
nd-si
de vie
w
of
e
l
ectri
c
it
y
m
a
rkets.
IEEE Trans. P
o
wer Syst.
20
03; 1
8
(2): 5
2
0
–
527.
[4]
S Borenstein,
J Bushnell. An em
p
i
rica
l a
n
a
l
y
sis
of the
potenti
a
l for
market po
w
e
r
in
Cal
i
forni
a
’s
electricit
y i
ndu
str
y
.
T
he Journ
a
l of Industria
l Econo
mics
, 19
99; 47(3): 2
85–
333.
[5]
S Stoft. Pow
e
r Sy
stem Economics:
Designing Ma
rkets for Electricit
y
.
New
York: Wiley
-
Interscience.
200
2.
[6]
Natio
nal Institu
t
e of Econom
ic
s and In
dustr
y
Rese
arch
. T
he price e
l
asticit
y
of dema
nd for
electricit
y i
n
NEM regi
ons. T
e
ch. rep.,
Nation
al Electric
ity Market Manag
ement Co
mpa
n
y
, Victoria, 20
02.
[7]
Ho
w
a
nd W
h
y
Customers R
e
spon
d to Electr
icit
y
Pr
ice
Var
i
abil
i
t
y
. C
h
a
p
ter
5: Implicit Pric
e Elasticiti
es
of Dem
and
for El
ectricit
y
an
d P
e
rformanc
e
Results.
200
5. [Onlin
e]. Avai
la
ble
:
http://certs.lbl.gov/pdf/ny
i
so
[8]
RH Patrick, F
A
W
o
lak. Esti
mating th
e cus
t
omer-l
ev
el d
e
m
and for
el
ectricit
y
un
der re
al-time mark
e
t
prices. 20
01. [Onlin
e]. Availa
ble: http://
w
w
w
.
nber.org/p
ap
e
r
s/
w
8
21
3.pdf
[9]
CL Su,
D Kirs
chen. Qu
antif
yi
ng th
e effect
o
f
deman
d res
p
onse
on
el
ectri
c
it
y
m
a
rkets.
IEEE Trans.
Power Syst.
2009; 24(3): 1
199
–12
07.
[10]
K Sing
h, NP P
adh
y, J S
har
ma. Influe
nce
of
price r
e
sp
on
sive d
e
man
d
s
h
ifting
bi
ddi
ng
on co
ng
esti
o
n
and LMP i
n
po
ol-b
ased d
a
y
-a
hea
d el
ectricit
y markets.
IEEE
Trans. Power
Syst
. 2011; 26(
2).
[11] DK Agra
w
a
l, RK Nema, NP Patidar.
W
i
nd p
o
w
e
r trading o
p
tions i
n
India
n
electricity ma
rket: inclusio
n
of avail
abi
lity b
a
sed tariff an
d day ah
ea
d trad
ing
. IEEE ICPS. 2009: 1-6.
[12]
Sánch
e
z I. Short-term predicti
on of
w
i
n
d
en
e
r
g
y
pr
od
uction.
Int. J
.
Forecast.
2006; 22: 43
–56.
[13]
Li S, W
unsch DC, O’Hair
EA, Giesselman
n MG. Using neur
al net
w
o
rk
s to estimate w
i
nd tur
b
in
e
po
w
e
r gen
erati
on,”
IEEE Trans. Energy Conversion
. 200
1; 16: 276
–2
82.
[14]
J Dup
a
cová,
N Grö
w
e-K
u
s
k
a, W
Römis
c
h.
Scen
ario
reducti
on in
s
t
ochastic prog
ramming: A
n
appr
oach
usin
g prob
abi
lit
y
m
e
trics.
Math. Program
. Series A
. 2003: 3: 49
3–5
11.
[15]
N Grö
w
e-Kusk
a, H Heitsch,
W Römisch.
Scenar
io re
ducti
on a
nd sce
nari
o
tree construc
tion for pow
e
r
ma
na
ge
me
nt p
r
obl
ems.
Proc.
IEEE Po
w
e
r T
e
ch Co
nf. Italy
. 2003. 3: 23–
2
6
.
[16]
Jian
hui
W
ang,
Moh
a
mmad
Shah
ide
h
p
our,
Z
u
yi
L
i
. Sec
u
rit
y
-
C
onstr
ain
ed U
n
it
Com
m
itment
w
i
t
h
Volatile Wind Po
w
e
r Generation.
IEEE Trans. Power Syst.
2008; 23(
3): 131
9–1
32
7.
[17]
P Venkates
h, R Gnana
dass,
NP
Padh
y. C
o
mparis
on a
n
d
Applic
at
ion of
Evolutio
nar
y Programmi
n
g
T
e
chniques to
Combi
n
e
d
Eco
nomic Emiss
i
o
n
Disp
a
tch
w
i
t
h
Lin
e
F
l
o
w
C
onstrai
nts.
IEEE Trans. On
Power Syst.
2003; 18(2): 6
88-
697.
[18]
I Jacob
Rag
l
e
nd, Nar
a
ya
na
Prasad P
a
d
h
y
.
Soluti
ons
to
Practical Unit Commitment
P
r
obl
ems
w
i
t
h
Operatio
nal, P
o
w
e
r F
l
o
w
and
Enviro
nmenta
l
Constra
i
nts.
IEEE
, 1-424
4-0
493-
2/06. 20
06
.
[19]
Mode
lin
g
w
i
th
Xpr
e
ss-MP, D. Associates, Ed
., 2005. Avail
a
ble: http://
w
w
w
.
dasho
ptimizati
on.com
Evaluation Warning : The document was created with Spire.PDF for Python.
¢
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 10, No. 1, March 2
012 : 33 – 42
42
[20]
Cha
ng GW
, Agan
agic M, W
a
ig
ht JG, Medina J, Bu
rton T
,
Reeves S, C
h
ristofori
d
is M. Exp
e
rie
n
ce
s
w
i
t
h
Mi
xe
d Int
eger
Lin
ear Pr
ogrammi
ng B
a
sed Ap
proac
h
e
s on S
hort-T
e
rm H
y
dro Sc
hed
uli
ng.
IEEE
Transactions on Power Syst..
200
1; 16(4): 74
3-74
9.
Evaluation Warning : The document was created with Spire.PDF for Python.