TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol. 14, No. 1, April 2015, pp. 42 ~ 48
DOI: 10.115
9
1
/telkomni
ka.
v
14i1.771
3
42
Re
cei
v
ed
Jan
uary 6, 2015;
Re
vised Ma
rc
h 6, 2015; Accepte
d
March
24, 2015
Forecasting of Utility Cost in a Deregulated Electricity
Market by Using Locational Marginal Pricing
T. Mohanapr
iy
a*
1
, T.R. M
a
nikand
a
n
2
, T
.
Ve
n
k
a
t
es
an
3
K.S. Rangas
a
m
y
Co
lle
ge of
T
e
chnolog
y
KSR Kalvi N
a
g
a
r,
T
i
ruchen
go
de, Na
mak
a
l-6
37 21
5, T
a
milnadu, Indi
a
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: mohan
atcp@
g
mail.c
o
m
*
1
, manik
and
an
26
6
@
gmai
l.com
2
,
pramoth
99@
yaho
o.co.uk
3
A
b
st
r
a
ct
In the dere
gul
a
t
ed electric
ity mark
et bid
d
in
g
cont
est is the ma
jor o
perati
o
n. Prices obta
i
ned fro
m
the result of bi
ddi
ng strategy
is essenti
a
l, since all
mark
et partici
pant
s do
not be fami
li
ar w
i
th the accurat
e
assess
me
nt of
future pric
es i
n
their
dec
isio
n-makin
g
pr
oc
ess. Locati
o
n
a
l
Margi
nal
Prici
ng (LMP)
obta
i
ns
from the O
p
ti
ma
l Pow
e
r F
l
o
w
proble
m
giv
e
s the ec
on
o
m
ic v
a
lu
e of e
l
ectrical
en
erg
y
at each
loca
tion.
Propos
ed
met
hod
is
bas
ed
o
n
l
o
ssless
DC
Optima
l P
o
w
e
r F
l
ow
. T
o
solv
e this
LMP
pro
b
le
m opti
m
i
z
a
t
i
o
n
base
d
L
i
n
ear
Progra
m
mi
ng
(LP) ap
pro
a
ch
has
be
en
i
m
ple
m
ented. I
n
this p
a
p
e
r L
M
P valu
es w
i
t
h
transmission, line outage and
generat
or outage
constraints
are studied. IEEE 14
an
d IEE
E
30
bus systems
are used as a test system
in t
h
is paper.
Ke
y
w
ords
:
locati
ona
l marg
inal
prici
ng, co
ngesti
on
ma
na
ge
me
nt, gener
ator outag
e,
lin
e outag
e, lin
ea
r
progr
a
m
min
g
.
Copy
right
©
2015 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
By tradition t
he d
e
livery
o
f
elect
r
ical e
nergy
ha
s
b
een
se
en
as a
com
m
una
l facility
provide
d
by regulate
d
utilities. In a lot
of cou
n
trie
s, a
singl
e state
-
o
w
ne
d utility was a
c
cou
n
tab
l
e
for the g
ene
ration, tran
sm
issi
on a
nd d
e
liver of p
o
we
r. We si
mpl
y
call thi
s
uti
lity as Verti
c
ally
Integrated
Utilities (VIUs). The ele
c
tri
c
al
supply i
ndu
stry all over the worl
d ha
s p
r
acti
ced a
sta
ge
of quick an
d
perma
nent
cha
nge in te
rms of stru
ctu
r
e, right
s, proce
s
s and
a
d
minist
ration.
To
improve th
e o
peratio
n effici
ency of VIUs
huge
ch
an
g
e
s
a
r
e ta
king
p
o
sition i
n
the
power in
du
stry
whe
r
eby com
petitive
markets
a
r
e
repla
c
ed by
VIUs as
to
initia
te
a co
mpetition
betwe
en p
o
w
er
prod
ucers a
nd buye
r
s [1]. These chang
es
ar
e
comm
only
referred
a
s
dere
gulatio
n
or
rest
ru
cturin
g. The m
a
in
ob
jectiv
es of re
stru
ctured m
a
rket a
r
e
se
cure and
e
c
on
omic ope
ratio
n
of
a po
wer sy
stem with
out violating a
n
y system
se
cu
rity limit. Sec
u
rity is
the
mos
t
s
i
gnificant
feature of
the
power syste
m
ope
ration and coul
d
be
facilitated by
utilizing th
e a
s
sorted
servi
c
e
s
offered to th
e
market. The
eco
nomi
c
al o
peratio
n
of th
e po
wer ma
rket would
de
cre
a
se the
cost
of elec
tric
ity [1].
Major types
of market
structure
are
p
ool,
bilate
ral and hybrid
model
s
[2]. Pool
is
a
centralized m
a
rket for buyers a
nd
sell
e
r
s where ele
c
tri
c
power se
llers and bu
yers submit bids
and pri
c
e
s
in
to the pool for the amo
u
n
t that t
hey
are willi
ng to
sell or buy. The Indep
en
dent
System Operator (ISO) i
s
a centralized
authority to
sellers and b
u
y
ers. The mai
n
obje
c
tive of the
ISO is mainta
ining reli
able
and secure o
peratio
n
of po
wer in
du
stry. ISO will forecast the dema
nd
for the day and re
ceive bi
ds that will satisfy t
he demand at the lowe
st co
st a
nd pri
c
e
s
for
the
electri
c
ity on
the b
a
si
s
of the mo
st ex
pen
sive
generator in operation. Bilateral models al
so
referred to
as Dire
ct Acce
ss Mod
e
l. As the nam
e
impl
ies, custom
ers are fr
ee to
contract di
re
ctly
with po
we
r g
enerating
co
mpanie
s
. By establi
s
hi
ng a
n
app
rop
r
iate
acce
ss
and
prici
ng
stand
ard
s
,
cu
stome
r
s transfe
r pu
rch
a
se
d po
we
r as restri
cted
to the power transmi
ssion
and di
stribut
ion
over utility wires. Bilate
ral
market do
e
s
not
requi
re ISO as in
the ca
se
of pool ma
rket
s.
Gene
rato
r unit commitm
ent and eco
nomic di
sp
atch de
cisi
on
s are depe
nd
ent on individual
market p
a
rtici
pants.
Du
e to
this
self
co
m
m
itment
ne
ed
of ind
epe
nde
nt ope
rato
r i
s
minimi
zed
[2
].
Still there is a
need of an ISO for the Indepe
ndent op
eration of tra
n
smi
ssi
on sy
stem. This
ki
nd of
ISO is norm
a
l
l
y referre
d to as Minim
a
l ISO due it minimal cent
ral o
peratio
n.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Fore
ca
sting o
f
Utility Cost i
n
a Deregul
ated El
ectri
c
ity
Market by
Usi
ng… (T. Mo
h
anap
riya)
43
Tran
smi
ssi
on
Pricing is a
major issu
e in t
he dereg
ul
ated elect
r
icit
y market [3].
Fede
ral
Energy
Reg
u
l
atory Commi
ssi
on
(FERC) re
cog
n
ized t
hat tran
smi
s
sion g
r
id i
s
th
e key i
s
sue t
o
comp
etition,
and i
s
sue
d
g
u
ideline
s
to p
r
ice
the
tran
smissi
on
syste
m
. Even thou
gh transmissi
o
n
co
sts a
r
e
sm
all as
co
mpa
r
ed to po
we
r
prod
uctio
n
ex
pen
se
s an
d repre
s
e
n
ts a
small
perce
nt of
major investor owned utiliti
e
s
ope
rating
expenses, a t
r
ansmi
ssi
on system
is
the most
import
an
t
key to com
p
etition becau
se it can
cre
a
te efficien
ci
es in the p
o
w
er
gene
rati
on market. T
h
e
con
d
ition whe
r
e overl
oad
s i
n
tran
smissio
n
lines
o
r
tran
sform
e
rs o
c
cur is
calle
d conge
stion [4-6].
Con
g
e
s
tion could preve
n
t system
op
era
t
ors
from
di
spatchi
ng
addi
tional p
o
wer
from a
spe
c
ific
gene
rato
r. Conge
stion co
uld be ca
use
d
for variou
s rea
s
on
s, su
ch as tran
smi
s
sion line outa
ges,
gene
rato
r out
age
s, ch
ang
e
in ene
rgy d
e
m
and
and
un
coo
r
din
a
ted t
r
an
sa
ction
s
.
Con
g
e
s
tion
ma
y
result in p
r
ev
enting n
e
w
contra
cts, infe
asibility
in exi
s
ting a
nd n
e
w
contract
s, addition
al out
age
s,
and mon
opol
y of price
s
in some
regi
on
s of powe
r
systems and
da
mage to sy
stem com
pon
e
n
ts
[4-6]. Co
nge
stion m
a
y b
e
prevented
to som
e
ex
t
ent by me
an
s of
re
servat
ions,
right
s
and
con
g
e
s
tion
p
r
icin
g. Th
ere
are
two types of p
r
ici
n
g metho
d
s
are
availabl
e
in p
r
a
c
tice
for
con
g
e
s
tion m
anag
ement [
7
]. They are
uniform
and
non-
unifo
rm prici
ng stru
ct
ure.
In
this p
aper
con
g
e
s
tion is manag
ed by
mean
s of Lo
cation
al Ma
rg
inal Prici
ng (LMP) i.e. non
-unifo
rm pri
c
i
ng
stru
cture. In this p
ape
r d
a
y-ahe
ad m
a
rket and Ex
-An
t
i is con
s
ide
r
ed [7]. The
L
M
P at a lo
cat
i
on
is defined a
s
the marginal
cost to su
pp
ly an addi
tion
al MW incre
m
ent of power at the loca
tion
without violat
ing any sy
stem se
cu
rity limits [8]. This ch
arge reflects n
o
t sim
p
ly the margi
nal
cha
r
ge of po
wer p
r
o
d
u
c
tio
n
, other than
that deliv
ery cha
r
ge al
so
consi
dered. If the lowest pri
c
ed
electri
c
ity is
allocated for all Lo
cation
LM
P value
s
at all node
s will be
sam
e
. If conge
sti
on
pre
s
ent
in th
e sy
stem l
o
west
co
st en
ergy ca
nnot
re
ach
all l
o
cation, mo
re
exp
ensive
ge
nerators
will allo
cate
d
to rea
c
h
ou
t the dema
n
d
. In this
situation LMP
values
will b
e
differ fro
m
one
locatio
n
to a
nother l
o
catio
n
. Mathemati
c
ally, LMP at a bus i
s
L
a
g
r
ang
e multipli
er in
corpo
r
at
ed
with the equ
a
lity constrai
nt [8]. LMP at a
bus i
s
de
com
posed into three com
pon
en
ts.
LMP = Margi
nal Gen
e
ratin
g
unit price +
Con
g
e
s
tion p
r
ice
+
Ma
rgin
a
l
loss Pri
c
e
LMP is
obtai
ned from the
re
sult of O
p
timal
Power
F
l
ow
(OPF
). Either AC-OPF
or
DC-
OPF is u
s
ed
to determin
e
the LMP. To redu
ce t
he
compl
e
xity in the cal
c
ulati
on in this p
a
per
DC-OPF i
s
u
s
ed
[9]. In DC-OPF only
real
po
wer
flow i
s
con
s
id
ered
[10]. Dif
f
erent type
s
of
optimizatio
n model
s are use
d
for LM
P calcul
at
ion
s
like LP a
n
d
Lagrangi
an
relaxation u
s
ing
karush–
k
uh
n-Tucke
r
con
d
itions [1
1]. Evolutionary
a
l
gorithm li
ke
geneti
c
al
g
o
rithm [1
2] and
con
s
trai
ned
b
a
t algo
rithm [
12] is
also u
s
ed. Among
th
ese i
n
thi
s
pa
per
LP is
use
d
to solve the
optimizatio
n probl
em.
The
pape
r i
s
stru
ctured
as follows: Se
ct
ion 2
p
r
ovide
s
the
existin
g
tran
smi
ssi
on
pri
c
ing
method. Sect
ion 3 provid
e
s
the pro
b
le
m format
ion.
Section 4 prese
n
ts the lo
ssl
ess DC-O
PF
probl
em form
ations. Se
ctio
n 5 provide
s
the li
nea
r pro
g
rammi
ng m
e
thod. Sectio
n 6 provide
s
the
results an
d a
nalysi
s
. Section 7 de
scribe
s co
ncl
u
si
on.
2. Existing Transmissio
n
Pricing Method
Tran
smi
ssi
on
prici
ng offe
r global
acce
ss for
all pa
rticipa
n
ts in th
e market. To
re
cover
the co
sts
of transmissio
n network a
nd en
co
ur
a
g
e
market in
vestment in
transmissio
n
an
unde
rsta
nda
b
l
e pri
c
e st
ru
cture is n
e
cessary. In this
se
ction vari
o
u
s p
r
ici
ng m
e
thod
s and t
heir
cal
c
ulatio
ns a
r
e discu
s
sed.
2.1. Postage
-
Stamp
Rate Metho
d
Postage-stamp rate
scheme is
conventionally used by electri
c
utilities to
allot the
perm
ane
nt
transmi
ssion p
r
ice between the
use
r
s
of
firm tran
smi
ssi
on se
rvice. This meth
od d
oes
not ne
ed p
o
w
er flow calculation
s
an
d i
s
in
dep
end
e
n
t
of the tra
n
smissi
on
dista
n
ce
an
d
syst
em
arrang
ement.
This tra
n
smi
ssi
on p
r
ici
n
g
method allo
cate
s tran
smi
ssi
on cha
r
ge
s ba
sed
on the
amount
of th
e tran
sa
cted
power. F
o
r
each tra
n
saction the m
a
g
n
itude of
po
wer tran
sfe
r
is
cal
c
ulate
d
at the time of system pea
k.
2.2. Contr
a
c
t
Path Meth
o
d
Contract pat
h method al
so doe
s not required po
wer flow calcu
l
ation. In this method
contract
pat
h is a
co
rp
oreal
tra
n
sm
issi
on
pat
h
w
ay amon
g t
w
o t
r
an
smi
s
sion
u
s
ers t
hat
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 14, No. 1, April 2015 : 42 – 48
44
disrega
rd
s th
e fact that e
l
ectro
n
s foll
o
w
corp
oreal
paths th
at may differ dra
m
atically fro
m
contract paths. Following t
o
the
specifi
c
ation of contract paths,
transmi
ssi
on pri
c
es will then
be
assign
ed usi
ng a posta
ge
-stam
p
rate, whi
c
h is det
e
r
mine
d either individually for ea
ch of the
transmissio
n system
s or o
n
the averag
e
for the entire
grid.
2.3. MW-Mile Method
The MW-Mil
e Method is
also called a
s
line-by
-line
method sin
c
e it consid
ers, in its
cal
c
ulatio
ns,
cha
nge
s in
MW tra
n
smi
s
sion flo
w
s a
nd tran
smi
s
si
on line le
ngt
hs in mil
e
s.
The
method cal
c
ulates ch
arg
e
s
a
s
so
ciate
d
with
ea
ch wh
eelin
g
tran
sa
ction
based
on
th
e
transmissio
n
cap
a
city u
s
e
as a fu
nctio
n
of the
mag
n
itude of tran
sa
cted p
o
wer, t
he path foll
o
w
ed
by transacte
d
powe
r
, and t
he dista
n
ce traveled
by transacte
d po
wer. Th
e M
W-mil
e meth
od is
also
u
s
ed i
n
identifying tra
n
smi
ssi
on
pa
ths for a
po
wer tra
n
sacti
on. Thi
s
m
e
thod
req
u
ire
s
dc
power flo
w
calcul
ation
s
. The MW-mile
method i
s
the first pri
c
in
g
strategy p
r
o
posed for th
e
recovery of fixed transmission
co
sts ba
sed
o
n
the actual use of tra
n
smi
ssi
on net
work.
Total tran
smi
ssi
on capa
cit
y
cost is calculated a
s
follows:
T
tK
k
k
t
k
k
K
k
k
t
k
k
t
MW
L
c
MW
L
c
TC
TC
,
,
(1)
t
TC
-
c
o
s
t
allocated to trans
ac
tion t
TC
-
total cost of all lines in $
k
L
-
length of line
k in mile
k
c
-
co
st per M
W
per unit len
g
th of line k
k
t
MW
,
-
flow in line k,
due to tran
sa
ction t
T
-
set of tran
sa
ctions
K
-
set of lines
3. Problem Formation
The mai
n
ob
jective of thi
s
p
r
obl
em is minimization
of total co
st
subj
ecte
d to
ene
rgy
balan
ce
con
s
traint and tra
n
smi
ssi
on
co
nstrai
nt [
13-1
6
]. Power flo
w
is o
b
tained
by lossl
ess
DC-
OPF mod
e
l. In this OPF
re
active po
we
r
is ign
o
re
d an
d the voltage
magnitud
e
s
are a
s
sum
e
d
to
be unity [9].
Obje
ctive function
is give
n by the Equation (2
).
∑
(
2
)
Subject to de
mand con
s
tra
i
nt as sh
own in the Equatio
n (3).
∑
∑
(
3
)
Gene
ration li
mit con
s
traint
is given by the Equation (4
).
(
4
)
Tran
smi
ssi
on
line limit is
given by the Equation (5
).
(
5
)
Whe
r
e,
-
Gene
rato
r
ind
e
x
.
-
Numb
er of ge
nerato
r
s.
J
-
Line ind
e
x.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Fore
ca
sting o
f
Utility Cost i
n
a Deregul
ated El
ectri
c
ity
Market by
Usi
ng… (T. Mo
h
anap
riya)
45
4. Lossless
DC-OPF Pro
b
lem Formation
In AC netwo
rk
real
and
reactive p
o
we
r tran
smitted
from the g
e
neratin
g unit
to load
centre. Di
re
ct Curre
n
t Opti
mal Power Fl
ow give
s
ac
tive
Po
w
e
r
F
l
ow
in
AC n
e
t
w
o
rk
. T
h
is
DC
-
OPF is do
es
not have
con
v
ergen
ce
p
r
o
b
lem i.e. n
on
iterative. Fro
m
the a
c
cu
ra
cy level A
C
-OPF
is better than
DC-OPF. In DC-OPF
som
e
assumptio
n
s
are m
ade a
s
[9-10].
Powe
r inje
ction at a node
and voltage a
ngle
s
are the
importa
nt variable
s
for DC-OPF.
Active powe
r
injectio
n at a bus
is given by the Equation (6
).
∑
(
6
)
– Rea
c
tan
c
e
betwee
n
bu
s i and bus j.
Powe
r flow o
n
the tran
smi
ssi
on line i
s
g
i
ven by the Equation (7).
(
7
)
- Rea
c
tan
c
e
of line i.
DC-OPF
equ
ations an
d p
o
we
r flo
w
in
the b
r
an
ch
relation
shi
p
i
s
rep
r
e
s
ente
d
by the
Equation (8)
& (9).
(
8
)
(
9
)
Whe
r
e,
P -
N
1 vec
t
or of bus
ac
tive power injec
t
ion for bus
e
s
1,....
, N.
B -
N
N admittance matrix with R=0.
-
N
1 vec
t
or of bus
voltage angle for bus
e
s
1,.....,N.
P
L
-
M
1 vector of
bran
ch flo
w
s.
M
-
Numb
er of branche
s.
b -
M
M vector di
agon
al su
sce
p
tance matrix
.
A -
M
N bu
s
– branch in
cide
n
c
e m
a
trix. Starting a
nd e
n
d
ing b
u
s
ele
m
ents
are 1 an
d -1 resp
ectively. Otherwise 0.
5. Linear Programming
Linea
r pro
g
ra
mming is a
mathemati
c
al
model
to accompli
sh the
finest outco
me [12].
This i
s
on
e of the optimi
z
ation te
chni
que
s. It c
onsists of line
a
r objective fu
nction,
subje
c
t to
equality a
nd i
nequ
ality co
n
d
itions. In
the
lossle
ss
DCOPF optimi
z
ation p
r
obl
e
m
is form
ed
as
a
linear prog
ra
mming
pro
b
l
e
m. In this
pape
r, opt
im
ization
probl
em is solve
d
by line
a
ri
zed
approa
ch. Fi
gure
1
expl
ains the S
o
l
v
ing proced
ure
for
opti
m
al po
we
r f
l
ow
with
Lin
ear
Programmin
g
appro
a
ch usi
ng LP solve
r
.
6. Results a
nd Analy
s
is
The propo
se
d LP method
simulation
were develo
ped usi
ng M
A
TLAB 7.10 softwa
r
e
packa
ge a
n
d
the
system
config
uratio
n
is Intel
Co
re i
3
-23
28M P
r
o
c
e
s
sor
with
2
.
20 G
H
z spe
ed
and 2 GB
RAM. For sim
u
lation work two test
systems IEEE 14 and IEEE 30 bus
system
s are
-
Numb
er of bu
se
s.
-
C
o
s
t
of i
th
generato
r
unit.
- gene
ration
of
i
th
generato
r
unit
-
Minimum limit
of generatin
g
unit.
-
Maximum limi
t
of generatin
g unit.
-
Dema
nd of i
th
unit.
-
Minimum limit
of line flow.
Maximum limi
t
of line flow.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 14, No. 1, April 2015 : 42 – 48
46
con
s
id
ere
d
. The co
mputa
t
ional re
sults obtained fro
m
the test systems a
r
e a
nalyze
d
for line
outage, ge
ne
ration outa
g
e
,
transmi
ssi
o
n
and ge
nerat
or conge
sti
on. Gene
rato
r offer pri
c
e
is
cal
c
ulate
d
by the linear bi
d function. Lin
e
a
r
bid fun
c
tio
n
is given by the Equation
(10).
$
⁄
(
1
0
)
Figure 1. Flow ch
art for L
M
P calculatio
n usin
g linea
r prog
rammi
n
g
6.1. Case s
t
u
d
y
– IEEE 14 Bus Sy
stem
IEEE 14 bus system consi
s
ts of 20 lines
and 14 buses. Line and
generator data’s are
use
d
for the
simulatio
n
work. Lin
e
d
a
ta co
nsi
s
t
of sen
d
ing
and receivin
g end
bus,
line
resi
stan
ce, Li
ne re
acta
nce
,
half susce
p
t
ance a
nd
trans
former tap ratio. Two
tes
t
c
a
s
e
s
LMP
values
und
er
norm
a
l sy
ste
m
co
ndition
a
nd LMP valu
es u
nde
r tra
n
s
missio
n
con
gestio
n
condi
tion
are an
alyze
d
in this model.
Table 1.
LMP values u
nder n
o
rm
al
Condition in IEEE 14 bus sy
stem
Table 2. LMP values un
der tra
n
smi
ssion
congestion in IEEE 14
bus
system
Bus. No
L
M
P
($/MW
h
r)
Bus.
No
LMP
($/MW
h
r)
1 107.4
5
8
107.4
5
2 107.4
5
9
107.4
5
3 107.4
5
10
107.4
5
4 107.4
5
11
107.4
5
5 107.4
5
12
107.4
5
6 107.4
5
13
107.4
5
7 107.4
5
14
107.4
5
Bus. No
L
M
P
($/MW
h
r)
Bus.
No
LMP
($/MW
h
r)
1 107.455
8 272.99
2 47.615
9 279.859
3 161.532
10 281.827
4 259.948
11 293.707
5 330.756
12
303.943
6 306.010
13
302.329
7 272.99
14
289.683
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Fore
ca
sting o
f
Utility Cost i
n
a Deregul
ated El
ectri
c
ity
Market by
Usi
ng… (T. Mo
h
anap
riya)
47
LMP und
er
n
o
rmal
syste
m
con
d
ition i
s
cal
c
ul
ate
d
u
s
i
ng Lo
ssle
ss
DCOPF. In th
is case
,
simulatio
n
s
a
r
e
carrie
d out
unde
r n
o
rm
al test
syste
m
data’
s an
d
the obtain
e
d
LMP value
s
are
given in the Table 1. Con
gestio
n
is cre
a
ted in the 5
th
transmi
ssi
o
n
line by red
u
cin
g
the po
wer
flow up
per li
mit from 45
MW to 0.7
72
MW.LMP
val
ues und
er co
nge
stion con
d
ition
is pre
s
ented
in the Table 2
.
From the T
a
ble 1 & 2 it can b
e
inferred t
hat LMP values un
de
r normal con
d
ition is
same
a
s
that
of all bu
se
s.
But in case o
f
con
g
e
s
tion
occurre
d
LM
P values vary
from lo
catio
n
to
loc
a
tion.
6.2. Case Stud
y
– IEEE 30 Bus Sy
stem
IEEE 30 bus system consi
s
ts of 41 lines
and 30 bus system. It ha
s 9 generating unit. In
this test sy
stem four test
ca
se
s like L
M
P values
u
nder n
o
rm
al system
con
d
i
t
ion, Tran
smi
ssi
on
con
g
e
s
tion, g
enerator a
nd
line
outag
e condition
s are studie
d
.
Table 3. LMP
values un
der normal
c
o
ndition in IEEE 30 bus
sys
tem
Table
4.
LMP value
s
u
nde
r t
r
an
smi
ssi
on
congestion condition in IEEE 30 bus sy
stem
Bus.
No
LMP
($/MW
h
r)
Bus.
No
LMP
($/MW
h
r)
1 236.71
16
236.71
2 236.71
17
236.71
3 236.71
18
236.71
4 236.71
19
236.71
5 236.71
20
236.71
6 236.71
21
236.71
7 236.71
22
236.71
8 236.71
23
236.71
9 236.71
24
236.71
10 236.71
25
236.71
11 236.71
26
236.71
12 236.71
27
236.71
13 236.71
28
236.71
14 236.71
29
236.71
15 236.71
30
236.71
Bus.
No
LMP
($/MW
h
r)
Bus.
No
LMP
($/MW
h
r)
1
236.71
16 236.22
2
236.44
17 235.82
3
237.59
18 236.12
4
237.77
19 235.97
5
235.67
20 235.89
6
234.91
21 235.67
7
235.23
22 235.68
8
234.92
23 236.11
9
235.40
24 235.76
10
235.65
25 235.48
11
235.40
26 235.48
12
236.62
27 235.30
13
236.63
28 234.96
14
236.49
29 235.30
15
236.38
30 235.30
Table 5. LMP
values un
der Generator
outage
condition in IEEE 3
0
bus
system
Table 6. LMP
values un
der line outage
c
o
ndition in IEEE 30 bus
sys
tem
Bus.
No
LMP
($/MW
h
r)
Bus.
No
LMP
($/MW
h
r)
1 248.31
16
248.31
2 248.31
17
248.31
3 248.31
18
248.31
4 248.31
19
248.31
5 248.31
20
248.31
6 248.31
21
248.31
7 248.31
22
248.31
8 248.31
23
248.31
9 248.31
24
248.31
10 248.31
25
248.31
11 248.31
26
248.31
12 248.31
27
248.31
13 248.31
28
248.31
14 248.31
29
248.31
15 248.31
30
248.31
Bus.
No
LMP
($/MW
h
r)
Bus.
No
LMP
($/MW
h
r)
1 236.72
16
233.95
2 236.75
17
233.7
3 236.61
18
233.46
4 236.59
19
233.50
5 236.83
20
233.52
6 236.92
21
232.80
7 236.88
22
232.55
8 94.6
23
231.69
9 234.74
24
229.43
10 233.59
25
220.93
11 234.74
26
220.93
12 234.2
27
215.55
13 234.20
28
205.32
14 233.74
29
215.55
15 233.39
30
215.55
Simulation is
c
a
rried out on IEEE 30 bus
s
y
s
t
em under normal
s
y
s
t
em c
o
ndition and the
test re
sults a
r
e tabul
ated i
n
the Tabl
e 3
.
Conge
stion
is create
d
in the 7
th
transmissi
on line t
hat
con
n
e
c
ting
b
u
s
2 a
n
d
6
b
y
red
u
ci
ng th
e up
pe
r limit
of tran
smi
ssi
on lin
e from
30 M
W
to
0.2
MW
and the re
sult
s are given in
the Table 4. In addi
tion to transmissio
n con
s
trai
nt ge
nerato
r
outag
e
and line o
u
ta
ge co
nst
r
aint
s also analy
z
ed. For ge
ne
rator outa
ge condition 9
th
g
enerator pl
aced
in 30
th
bu
s is taken a
s
a
outage g
ene
rator an
d test
results a
r
e t
abulate
d
in the Tabl
e 5. For
transmissio
n line
out
age condition
10
th
line that
con
nect
s
the
bu
s 6
an
d 8 i
s
con
s
id
ere
d
a
s
a
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 14, No. 1, April 2015 : 42 – 48
48
outage lin
e a
nd simul
a
tion
result
s are gi
ven in the Table 6.
5. Conclusio
n
In a lot of restru
ctu
r
ed e
nergy ma
rkets, t
he Locational Ma
rgin
a
l
Pricing a
c
ts as an
importa
nt po
sition in
re
cent times. L
M
P is loo
ks set to b
e
the mo
st po
pular
co
nge
stion
manag
eme
n
t techniq
ue a
dopted by el
ectri
c
ity
markets aroun
d the wo
rld. To
unde
rsta
nd
the
determi
nation
of LMP L
o
ssle
ss DC O
p
timal po
wer
Flow i
s
ca
ref
u
lly analysed
whi
c
h i
s
th
e
prop
osed te
chniqu
e in this pap
er. Con
s
traint
s like tran
smi
ssi
on, gene
ration a
nd tran
smi
ssi
on
line outa
g
e
s
are
used to a
nalyze th
e m
a
rket pa
rtic
ip
ants a
bout th
e location val
ue of ele
c
tri
c
i
t
y.
LMP al
so
use
d
to mai
n
tain
the sta
b
le
op
eration
of
tra
n
smi
ssi
on sy
stem wi
thout affect
the buy
ers
and
selle
rs i
n
the m
a
rket
. LMP act
as a tru
e
p
r
ice
sign
als for
addin
g
tra
n
smissi
on
ca
pa
city,
gene
ration
capa
city and
future l
oad
s. I
t
achi
ev
es its uniq
ue a
m
bi
tion of b
e
tter
effectivene
ss of
power sy
ste
m
operation
s
in the sho
r
t-term ope
ratio
nal time fram
es by ope
nly addre
s
sing t
h
e
effects relate
d
with po
wer transmissio
n above
the
int
e
rconn
ecte
d grid. We ca
n extend
ou
r work
with hig
her b
u
s
system
an
d addi
ng mo
re co
nst
r
aint
s to our
probl
em. Instea
d
of DC-OPF,
AC
-
OPF can b
e
use
d
to solve
the powe
r
flow pro
b
lem.
Referen
ces
[1]
Shah
ide
h
p
our
M, Yamin H,
Li
Z
.
Market ope
rations
in
elect
r
ic po
w
e
r s
y
ste
m
s. Ne
w
York:
W
ile
y
.
20
02
:
1-14.
[2]
Kirsche
n
DS, Strbac G. F
undamentals of po
wer s
y
stem ec
on
omics. W
ile
y
.
2
004: 33-
39.
[3]
Sing
h H, Hao
S, Papal
e
x
op
o
u
los A. T
r
ansmission c
o
n
g
e
s
tion man
a
g
e
m
ent in com
p
e
t
itive electric
it
y
markets.
IEEE
Trans Power System
s
. 19
98; 13: 672-
80.
[4]
Sand
ip C
h
a
n
d
a
, Abhi
n an
da
n
De
.
Co
ng
esti
on
Re
li
ef
of
Contin
ge
nt
Po
w
e
r
Net
w
ork
w
i
t
h
Evol
ution
a
r
y
Optimization Algorithm.
T
E
LK
OMNIKA Indon
esia
n Journ
a
l o
f
Electrical Eng
i
ne
erin
g.
201
2; 10(1).
[5] Qun
Z
hou
,
L
e
i
gh T
e
sfatsion
,
Che
n
-Ch
i
ng
Liu.
Sh
ort-T
e
rm
Cong
estio
n
F
o
recastin
g
i
n
W
hol
esal
e
Po
w
e
r Markets.
IEEE
Trans Power Systems
. 2011; 26(
4): 218
5-21
96.
[6]
Hermag
a
sa
nto
s
Zein, Yusr
a
Sabri, Al
i Mas
h
e
r.
Implem
en
tation of E
l
ectr
icit
y
Busi
ness
Comp
etitio
n
F
r
ame
w
ork
w
i
t
h
Econ
omic Di
spatch Dir
ect Method.
T
E
LK
OMNIKA Indonesi
an Jo
urna
l
of Electric
a
l
Engi
neer
in
g.
2012; 10(
4).
[7]
Kan
w
a
r
de
ep
Sing
h, Nara
ya
na Prasa
d
Padh
y
,
Ja
yd
ev
Sharma. Influenc
e of Price Resp
onsiv
e
Dema
nd Shifti
ng Bid
d
in
g on
Con
gestio
n
an
d LMP in Pool-
B
ased D
a
y
-
Ah
ead El
ectricit
y
Markets.
IEEE
T
r
ans Pow
e
r Systems
. 20
11; 26(2): 88
6-8
9
6
/
[8]
T
Orfanogia
n
n
i
, G Gross. A Genera
l
F
o
rm
ulati
on for
LM
P.
IEEE Trans. On Power System
s
. 20
07
;
22(3): 11
63-
11
73.
[9]
AJ W
ood, BF
W
o
llen
ber
g. Secon
d
ed
ition.
Ne
w
York: Joh
n
W
ile
y
S
ons.
199
6: 93-1
08.
[10]
AJ Con
e
jo, JA
Agua
do. Multi
-
Area C
oor
d
i
n
a
ted D
e
centr
a
li
zed DC O
p
timal Po
w
e
r F
l
o
w
.
IEEE Trans
.
On Power System
s
. 1
998; 1
3
(
4
): 503-5
09.
[11]
Mohamm
ad Eb
rahim H
a
ji
ab
ad
i
a
, Ha
bib
Raj
a
bi Mas
hha
di
b.
LMP dec
ompo
sition: A n
o
ve
l
appr
oach f
o
r
structural mark
et po
w
e
r monit
o
ring,
Electric Pow
e
r Systems Researc
h
.
20
13; 99: 30-
37.
[12]
M Murali, M S
a
ilaja K
u
mari,
M S
y
d
u
l
u
. Optimal sp
ot pric
i
ng i
n
el
ectricit
y mark
et
w
i
th
inel
astic l
o
a
d
usin
g constrai
n
ed bat al
gor
ith
m
.
Electrical Pow
e
r and En
er
gy Systems
.
2
014; 62: 8
97-9
11.
[13]
Ricard
o
F
e
rná
ndez-B
lanc
o, José M Arro
yo, Nata
li
a Alg
uac
i
l
. Net
w
o
r
k-C
o
n
s
traine
d Da
y-A
hea
d Auctio
n
for Consum
er Pa
y
m
ent Min
i
mizatio
n
.
IEEE transactions on power system
s
. 20
14; 29(
2
)
: 526-53
6.
[14]
Luo
na
n C
h
e
n
, Hi
deki
Suz
u
ki
, T
s
unehisa
W
a
chi, Y
u
kih
i
r S
h
imura.
Com
p
one
nts of
No
d
a
l Pr
ices fo
r
Electric Po
w
e
r S
y
stems.
IEEE Trans. on Power System
s
. 20
02; 17(1): 4
1
-4
9.
[15] T
ongxi
n
Z
h
e
n
g
,
Euge
ne L
i
tvin
ov
.
On Ex Pos
t
Pricing
in th
e
Rea
l
-T
ime Ele
c
tricit
y
Mark
et.
IEEE Trans
Power Systems
. 2011; 26(
1): 153-
164
[16] Hon
g
y
an
Li
,
Lei
gh T
e
sfatsion
.
ISO Net Surpl
u
s Col
l
ec
tion an
d All
o
c
a
tion i
n
W
hol
esal
e Po
w
e
r
Markets und
er LMP.
IEEE Tra
n
s Power System
s
. 2
011; 2
6
(
2
): 627-6
41.
Evaluation Warning : The document was created with Spire.PDF for Python.